SCIENTIFIC POSTER SESSION ABSTRACTS

2014 AACC Annual Meeting & Clinical Lab Expo

Chicago, IL

Below are the abstract topics and posters schedule times.

TUESDAY, JULY 29, POSTER SESSIONS

9:30am 5:00pm
Cancer/Tumor Markers . . . . . . . . . . . . . . . . . A-001 A-067 . . . . . . . . . . . . . . . . . . . . . S2
Clinical Studies/Outcomes . . . . . . . . . . . . . . A-068 A-151 . . . . . . . . . . . . . . . . . . . . S20
Endocrinology/Hormones . . . . . . . . . . . . . . . A-152 A-234 . . . . . . . . . . . . . . . . . . . . S42
Factors Affecting Test Results . . . . . . . . . . . . A-235 A-279 . . . . . . . . . . . . . . . . . . . . S67
Hematology/Coagulation . . . . . . . . . . . . . . . A-280 A-321 . . . . . . . . . . . . . . . . . . . . S83
Immunology . . . . . . . . . . . . . . . . . . . . . . . . . A-323 A-374 . . . . . . . . . . . . . . . . . . . . S95
Mass Spectrometry Applications. . . . . . . . . . A-375 A-427 . . . . . . . . . . . . . . . . . . . .S111
Animal Clinical Chemistry . . . . . . . . . . . . . . A-428 A-438 . . . . . . . . . . . . . . . . . . . S127
WEDNESDAY, JULY 30, POSTER SESSIONS
9:30am 5:00pm
Automation/Computer Applications . . . . . . . B-001 B-034 . . . . . . . . . . . . . . . . . . .
Infectious Disease . . . . . . . . . . . . . . . . . . . . . B-035 B-086 . . . . . . . . . . . . . . . . . . .
Lipids/Lipoproteins . . . . . . . . . . . . . . . . . . . . B-087 B-123 . . . . . . . . . . . . . . . . . . .
Management . . . . . . . . . . . . . . . . . . . . . . . . . B-124 B-175 . . . . . . . . . . . . . . . . . . .
Electrolytes/Blood Gas/Metabolites . . . . . . . B-176 B-184 . . . . . . . . . . . . . . . . . . .
Molecular Pathology/Probes . . . . . . . . . . . . . B-185 B-220 . . . . . . . . . . . . . . . . . . .
Nutrition/Trace Metals/Vitamins . . . . . . . . . . B-221 B-254 . . . . . . . . . . . . . . . . . . .
Pediatric/Fetal Clinical Chemistry . . . . . . . . B-255 B-273 . . . . . . . . . . . . . . . . . . .
Point-of-Care Testing . . . . . . . . . . . . . . . . . . B-276 B-312 . . . . . . . . . . . . . . . . . . .
Cardiac Markers . . . . . . . . . . . . . . . . . . . . . . B-314 B-362 . . . . . . . . . . . . . . . . . . .
Proteins/Enzymes . . . . . . . . . . . . . . . . . . . . . B-364 B-387 . . . . . . . . . . . . . . . . . . .
TDM/Toxicology/DAU . . . . . . . . . . . . . . . . . B-389 B-432 . . . . . . . . . . . . . . . . . . .
Technology/Design Development . . . . . . . . . B-433 B-452 . . . . . . . . . . . . . . . . . . .

S131
S143
S158
S168
S183
S187
S196
S205
S211
S222
S237
S245
S259

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S264

Keyword Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S275
Ed. Note: These abstracts have been reproduced without editorial alteration from the materials supplied by the authors. Infelicities of
preparation, grammar, spelling, style, syntax and usage are the authors. These abstracts represent the supplement to the October issue of
Clinical Chemistry.

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

Tuesday, July 29, 2014

Poster Session: 9:30 AM - 5:00 PM
Cancer/Tumor Markers

A-001
The expression of post-translational modification of Alpha-1-antichymotrypsin
in the plasma of colorectal cancer

H. Lin1, C. Liao2, J. Lin3, K. Lee3, Y. Chen3, H. Ning4, C. Lin3. 1Chang

Gung Memorial Hospital and School of Medical Laboratory Science and
Biotechnology, Taipei Medical University, Taipei, Taiwan, 2Proteomics
Research Center, National Yang-Ming University, Taipei, Taiwan, 3School
of Medical Laboratory Science and Biotechnology, Taipei Medical
University, Taipei, Taiwan, 4Chang Gung Memorial Hospital, Tao-Yuan,
Taiwan
Colorectal cancer is the third major cause of cancer related death in the world
according to the report of cancer statistics in 2013. At present, some screening
biomarkers are applied for the detection of colorectal cancer but most of them do not
have good specificity and sensitivity. Carcinoembryonic antigen (CEA) is the most
use of tumor markers for colorectal cancer, but the specificity and sensitivity is poor.
In this study, we used proteomic approaches to investigate the expression of posttranslational modification (PTM) of alpha-1 antichymotrypsin (ACT) in the plasma of
colorectal cancer. We used immunoprecipitation, western blot and nano-LC/MS/MS
to analyze the plasma samples from normal and cancer groups. Then we compared
some types of PTM between those samples in order to find out the useful PTM sites
in ACT to be a diagnostic tool. In the result of immunoprecipitation, we identified the
accurate site of ACT in the gel electrophoresis. The following result of western blot
showed that there was no significance between the normal group and the early stage
of colorectal caner group (p=0.010). However, it was significant between the normal
and the late stage group (p<0.001). From the result of mass analysis, we identified four
types of PTM in ACT, such as Hydroxylation (Asn-323), Methylation-2 (Glu-334),
4-Hydroxy-2-nonenal (Arg-298) and N-glucuronylation (Ser-415). The expression
level of Hydroxylation was increased by 2-fold in colorectal samples when compared
with normal samples (p<0.05). The sensitivity of that was 88.89% and the specificity
was 77.78% (AUC=0.840). Our results suggest that using the expression level of
PTMs in ACT would be applicable as biomarkers for the early detection of colorectal
cancer.

Mechanism: The PdLAC enters normal and cancer cells and the mitochondrial outer
membrane by the voltage dependent anion channel, then through the inner membrane
by the Complex 1. The oxidative phosphorylation (OXPHOS) channel produces low
levels of ATP in the cancer cell. PdLAC (acting as an electrical shunt) in normal
cells would ordinarily give off electrons to the OXPHOS producing more ATP. In the
cancer cell (damaged OXPHOS) it donates electrons producing increased reactive
oxygen species (ROS). The excessive ROS builds up between the outer and inner
mitochondrial membranes. When the outer membrane ruptures, the ROS, Cytochrome
C, and the Procapases 2, 3, and 9 enter the anaerobic cytoplasm of the cancer cell and
death occurs.
Results: James Forsythe, MD, HMD conducted out-come based stage IV studies
(500+ patents 2004-2012) He found improvement in quality of life issues directly
proportional to improvement to overall response rate and found stable disease can be
tolerated and transformed into chronic livable condition.
Conclusion:This clinical and scientific documentation/data, from several public/
professional sources, provides a non-toxic adjuvant integrative nutritional therapy
option for advanced diseased cancer patients/physicians, when traditional therapy
options have been exhausted and non-traditional therapy options are under
consideration. Physician calls to other physicians with PdLAC, Coenzyme Q10, and
Vitamin D clinical experience, can determine if this therapy and monitoring of patient
progress is appropriate for a given end stage patient. It is not meant to circumvent
physician patient monitoring, good medical practice, medical ethics, and/or negatively
impact the physicians license.

A-004
Epidermal Growth Factor Receptor (EGFR) gene mutations frequency in
Brazilian lung adenocarcinoma samples by pyrosequencing.

V. D. T. Niewiadonski, P. Y. Nishimura, O. Fernandes, N. Gaburo Jr..

DASA, Sao Paulo, Brazil
Background: Lung cancer is the most prevalent life-threatening cancer worldwide
with more than 80% being non-small cell lung cancer (NSCLC). Detection of
mutations of EGFR gene is critical for predicting the response to therapy with tyrosine
kinase inhibitors (TKIs), such as gefitinib and erlotinib, in patients with NSCLC.
Patients that are EGFR mutants have constitutive TK activity and, therefore, a greater
sensitivity to anti-EGFR inhibition.
Objective: To describe the EGFR mutations frequency found in lung adenocarcinoma
samples, using pyrosequencing method.

A-002
Tumor Marker, Molecular, and Imaging Test Monitoring of a Non-Toxic
Adjuvant Integrative Nutritional Therapy Option for Stage IV Brain, Lung,
Prostate, and Breast Cancer Patients When Traditional Therapy Options Have
Been Exhausted: Palladium Lipoic Acid Complex, Coenzyme Q10, and Vitamin
D Impacting the Mitochondrial Reactive Oxygen Species (ROS) Production and
Apoptosis

E. J. Neren. Edward J. Neren, Biomedical Consultant, Suffern, NY

Background: Metal compounds (Platinum, etc.) have been used as cancer therapies
for years; however, patient toxicity usually results. Dr. Merrill Garnett synthesizing
organo-metallic compounds (1960-1990) encapsulated the metal palladium in alpha
lipoic acid (non-toxic and successful in treating mice with Ehrlich carcinoma). Cat/
dog tumors were also successfully treated. Rudy Falk, MD (1992 University of
Toronto) determined safety, improvement, and many remissions in gravely advanced
cancer cases. Since then, 200+ U.S. physicians have used the palladium/lipoic
acid complex (PdLAC), Coenzyme Q10, and Vitamin D as an adjuvant integrative
nutritional therapy for late stage cancer patients.
Objective: To provide stage IV cancer patients/physicians with a documented
nutritional therapy option that justifies physician calls to other physicians with
clinical experience, regarding their latest clinical data and determine if the therapy
monitored with tumor markers, molecular, and imaging testing is appropriate for the
given patient.
Methods: After determining patient levels (tumor markers, Coenzyme Q10, and
Vitamin D) PdLAC liquid/coenzyme Q10/Vitamin D is physician monitored and
taken as nutritionals. PDLAC is water and fat soluble and impacts the mitochondrial
ROS of both cancer and normal cells. Positive clinical response and improved

S2

quality of life results is expected within three months of intake. Eight to twelve
PdLAC teaspoons is taken in juice 4 times a day (based on patient body weight - 1
teaspoon for each 30 pounds). This therapy seeks balance between therapy, nutrition,
detoxification, and energy enhancement. Patient progress is monitored with traditional
clinical chemistries, tumor markers, Coenzyme Q10, Vitamin D testing and imaging.

Method: Thirty samples of lung adenocarcinoma were analyzed from January 2013 to
December 2013. The test was performed on formalin-fixed, paraffin-embedded tumor
specimen, after the selection of the specimen region to be analyzed by a pathologist.
The DNA was extracted using the Qiaamp FFPE Tissue kit (Qiagen, Hiden,
Germany). Concentration of DNA sample was measured spectrophotometrically
using a NanoDrop spectrophotometer (NanoDropTechnologies, Wilmington). Codons
719, 768, 790, 858, 861 and exon 19 were amplified by PCR using the EGFR Pyro
kit (Qiagen, Hiden, Germany). Successful and specific amplification of the region
of interest was verified by visualizing the PCR product on capillary electrophoresis
using Qiaxel DNA Screening Kit (Qiagen, Hiden, Germany). Preparation of singlestranded DNA was done using PyroMark Q24 vacuum workstation (Qiagen, Hiden,
Germany) according to the manufacturer instructions. The pyrosequencing reaction
was analyzed on the Pyro Mark Q24 (Qiagen, Hiden, Germany)
Results: The frequency of EGFR mutations found is presented on Table 1. All
mutations together represent only 27% of the samples.
Conclusion: The results are consistent with previous studies and reports. The singlepoint mutation L858R (CTG> CGG) on exon 21 and the frame deletions on exon 19
represents the majority mutations found in Brazilian lung adenocarcinoma samples,
although most samples showed no mutation at the target regions.
Table 1. Frequency of EGFR mutations found in lung adenocarcinoma samples.
Results
Wild type
2235del15 (exon 19)
2236del15 (exon 19)
2237_2255>T (exon 19)
2239_2248>C (exon 19)
CTG>CAG (L861Q)
CTG>CGG (L858R)

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Frequency
73%
3,3%
3,3%
3,3%
3,3%
3,3%
10%

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

A-007

A-005
Ovascreen Lateral Flow Device for Simultaneous Detection of CA125 and
WFDC2 (HE4) in Ovarian Cancer

Pca3 gene expression as biomarker to differential diagnosis of benign

hyperplasia and prostatic cancer

Y. Lebedin. Xemabio LLC, Gainesville, FL

F. F. Coelho1, O. Romano2, C. Corradi1, L. Nogueira1, W. Cabral1, E.

Mateo3, K. Borges1. 1Federal University of Minas Gerais, Belo Horizonte,
Brazil, 2Hermes Pardini Institute, Vespasiano, Brazil, 3Hermes Pardini
Institute, Belo Horizonte, Brazil

A lateral flow device named Ovascreen has been developed to simultaneously detect
two tumor markers (CA125 and HE4) in ovarian cancer. Combined determination of
serum HE4 (WFDC2 protein) and CA125 is demonstrated to increase the sensitivity
and accuracy of early diagnosis and monitoring of ovarian carcinoma. However,
a convenient and portable device for such a dual detection is not commercially
available. In this study, a panel of monoclonal antibodies of HE4 (WFDC2) was
obtained by immunization of human tumor-derived antigens at Xema Company. The
HE4 antibodies were classified into 3 epitope groups for antibody matching based on
their cross-inhibition and bindings to recombinant and natural antigens. On the other
hand, the antibodies of MUC16 (antigen CA125) were also developed and the matched
pairs (X306 and X325) were obtained by Xema. The selected pairs of both CA125
and HE4 were coated with colloidal gold particles, and combined onto a lateral flow
device. Their assay performance was evaluated based their stability and their signal
to background ratios. The best pair of CA125 and HE4 antibody-gold conjugates
for lateral flow device were determined. The cut-off values for CA125 and HE4 on
the device is set as 35 U/ml and 150 pmol/l, respectively. The resulting Ovascreen
cassette device is validated with clinical samples of whole blood, serum, and plasma.
74 samples from primarily diagnosed but untreated serious adenocarcinoma of the
ovary were evaluated. Ovascreen device showed excellent sensitivity and accuracy
during the clinical evaluations. The combined positive results on Ovascreen device
accounted for more than 50% cases in 69 patients which corresponding to 93%
clinical sensitivity. 66 out of 69 Ovascreen positive results were further confirmed
by commercial ELISA kits of CA125 (Xema Co.Ltd.) and HE4 (Fujirebio Inc.).
The Ovascreen lateral flow device is suggested for use in POC based diagnosis and
monitoring of ovarian cancer.

A-006
Characterizations of Circulating Tumor Cells Identified by Combination
of Fluorescence in situ Hybridization and Immunostaining CK, CD45 in
Pancreatic Cancer

Y. Zhang, N. Ning, Q. Chen, F. Wang, W. Cui. Peking Union Medical

College Hospital Chinese Academy of Medical Sciences, Beijing, China
Background: To improve the identification for CTCs with weak or negative CK and
diploid CTCs in pancreatic cancer, we combined the immune-staining of CK, CD45,
DAPI and fluorescence in situ hybridization with the centromere of chromosome 8
(CEP8) probe method. Methods: CTCs in 3.75 mL of blood were negatively enriched
with Epithelial Cell Adhesion Molecule independent magnetic beads coated with
anti-CD45 antibodies and identified by combining CK, CD45, DAPI and CEP8 in 61
cases including 22 pancreatic cancers, 3 borderline pancreatic solid pseudopapillary
tumors, 6 pancreatic benign tumors, and 30 healthy individuals. Results: Enriched
cells could be classified into 5 patterns (Fig. A-E): CK+CD45-DAPI+CEP8=2
(2 hybridization signals), CK+CD45-DAPI+CEP8>2 (>2 hybridization signals),
CK-CD45-DAPI+CEP8>2, CK-CD45-DAPI+CEP8=2, and CK+/-CD45+DAPI+
CEP8=2or>2. Among 22 pancreatic cancers, the patterns of CK+CD45-DAPI+
CEP8=2 and CK+CD45-DAPI+CEP8>2 were identified in 2 cases, the number of
CTCs was 6, 12 cells/3.75mL and 2, 44 cells/3.75mL, respectively. The pattern of
CK-CD45-CEP8>2 was identified in 16 cases with the range of 1-14 cells/3.75mL
and the median of 3 cells/3.75mL. The pattern of CK-CD45-CEP8=2 and CK+/CD45+CEP8=2 or >2 were detected in both pancreatic cancers and other control
cases. Dynamically monitoring CTCs and platelet count prior to and after surgery in 7
pancreatic patients revealed that they were consistent both decreased or insignificantly
decreased 3 days after surgery, whereas the count significantly increased 10 days
after surgery. Conclusion: The patterns of CK+CD45-DAPI+CEP8 =2, CK+CD45DAPI+CEP8>2 and CK-CD45-DAPI+CEP8>2 were considered as CTCs, and the
patterns CK-CD45-CEP8=2 and CK+/-CD45+DAPI+CEP8=2 or >2 were considered
as indeterminate cells. Postoperative increase in the platelet count might contribute to
CTCs dissemination, and certain correlation might exist between those two events.

Background: The Prostate Cancer (Pca) is the second most common type of cancer
in men around the world. Because of the increasing numbers of cases, it is extremely
important the development of a noninvasive test with high specificity and sensitivity
to diagnosis cancer and other prostatic alterations. Studies showed that the gene 3
of Prostate Cancer (PCA3) presents high levels of expression in tumor tissue. High
levels of PCA3 gene expression can be associated with an increased probability of
positive biopsy and has arisen as a molecular marker in the diagnosis of PCa.
Objective: We proposed to evaluate the the expression of PCA3 gene in urine from
patients with benign hyperplasia (BPH) or prostatic cancer.
Methods: The study included 33 men attended at the Clinical Hospital from Federal
University of Minas Gerais (HC-UFMG) to performer a prostatic biopsy, being 13
patients with Pca, 8 with BPH and 12 patients with no alterations (controls). It was
collected 30 mL of patients urine after prostatic massage, which was immediately
centrifuged. The pellet was added to RNA later and stored for up to 24 hours, until
the extraction of RNA. The samples were quantified in a spectrophotometer and
submitted to treatment with DNase. After, this sample was quantified in a one-step
RT-PCR for PCA3 gene and PSA/ ACTB genes for control or reaction normalization.
Results: The PCA3 gene expression was detected in 10 patients with PCa, 3 with
BPH and 2 controls. For the remaining patients was not detected any gene expression.
The test presented 77% of sensitivity for PCa screening and 38% for BPH. The
specificity was 83% for both.
Conclusion: The PCA3 screening showed median sensitivity for PCa diagnosis;
subsequently prostate biopsy is still considered the best standard procedure for detect
prostatic alterations. Some patients with PCa no presented any expression of the
PCA3 gene, which can be explained by the large number of interfering, as well as
the prostate massage, the use of drugs and the high RNA degradations rate in urine
samples. It is required the standardization of these procedures and to analyze a larger
number of samples in order to evaluate the importance of PCA3 gene expression in
differential prostatic alterations and its use in the clinical practice.

A-010
Circulating tumor markers of benign and malignant disorders of breast in
Libya.

J. R. Peela1, S. Shakila1, S. J. Dhoipode2, A. R. Said1, H. Beloch2, S. Nang2,

L. T. Peela3, A. M. Jarari4, S. O. Alsaeoti5, H. El Awamy4, F. Elshaari4,
N. M. Jarari6, M. J. Kadeer4. 1Department of Biochemistry, Faculty of
Medicine,Quest International University Perak, Ipoh, Malaysia, 2Faculty
of Medicine,Quest International University Perak, Ipoh, Malaysia, 3Great
Eastern Medical School, Srikakulam, India, 4Department of Biochemistry,
Faculty of Medicine,University of Benghazi, Benghazi, Libyan Arab
Jamahiriya, 5Department of Surgery, Faculty of Medicine,University
of Benghazi, Benghazi, Libyan Arab Jamahiriya, 6Department of
Pharmacology, Faculty of Medicine,University of Benghazi, Benghazi,
Libyan Arab Jamahiriya
BACKGROUND Breast cancer is most common malignant disorder in Libyan
females. Breast cancer is the most dreadful disease in terms of quality of life, though
heart disease is a more common cause of mortality here. The present study is a case
control study of tumour markers CA 125, CA 15-3 and Carcino Embryonic Antigen
(CEA) in serum of patients suffering from benign and malignant breast disorders in
Libya.
Materials and methods: There are 12 cases of carcinoma of breast patients with age
group ranging from 30 - 55 years of age, 10 cases of benign breast disorders i.e.,
Fibroadenosis with age group ranging from 18 to 50 years retrieved from department
of surgery, 7th October Hospital, Benghazi, Libya and there are 12 cases of age
matched controls free from both malignant and benign disorders of breast were
included in this study. Venous sampling was done to the patients, all the patients
and controls were measured serum CA125, CA15-3and CEA levels by authenticated
methods by using Cobas E 411 analyser. Statistical analysis was done by using SPSS
software by using Mann-Whitney and Wilcoxon tests.
Results: There is no significant rise of CA 125 in benign (p=0814) and malignant

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S3

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

(p= 0.676) disorders of breast when compared to controls. CA 15-3 was significantly
high in patients suffering from breast cancer (p= 0.019) when compared to controls
and also very significantly high when (p= 0.003) compared with patients with benign
breast disorders. The level of CA 15-3 was not significantly high in patients with
benign breast disorders (p=0.186) when compared to controls. The level of CEA
was significantly high in patients of breast cancer when compared to patients of
benign breast (p=0.009) disorders and (p=0.017) controls. The level of CEA is not
significantly in high patients of benign breast disorders (p=0.634) when compared
to controls.
Conclusion: The present study showing high levels CA 15-3 and CEA only in
malignant disorders of breast may be useful as diagnostic and prognostic markers. CA
125 has not shown any significance in this study proving that it is not an important
marker in malignant breast disorders.

A-011
Diagnostic value of Insulin-like growth factor-1, IGF-binding protein-3,
Chromogranin-A in differentiation between benign prostatic hyperplasia and
prostate cancer patients

S. A. K. Saleh. Umm AlQura University, Makkah, Saudi Arabia

Backgrounds: Prostate cancer (PCa) ranked the sixth most common cancer among
males in Arab world. Elevated serum Insulin-like growth factor-1 (IGF-1) level
appeared to be a possible risk factor for the development of PCa. Chromogranin A
(CgA) is the most employed serum marker to detect neuroendocrine features. Many
studies reported contradictory findings of association between IGF-binding protein-3
(IGFBP-3) and the risk of PCa. Although, the best and most sensitive screening test
available for PCa is prostate specific antigen (PSA) there is a large overlap between
PCa and benign prostatic hyperplasia (BPH) in patients with moderately increased
PSA levels. Objective: This study aimed to explore the validity of IGF-1, IGFBP-3,
CgA and thereof ratios with PSA in differentiation between BPH and PCa patients
in Saudi Arabia. Patients and Methods: The study included 62 patients with
PCa, 70 BPH patients and 56 healthy male subjects of matched age. Full history
and clinical data were recorded for all subjects. PCa patients were undergo digital
rectal examination (DRE), trans-rectal ultrasonography (TRUS) guided biopsy
of the prostate, computed tomography (CT) scan of the pelvis, bone scan and
histopathological examination, accordingly PCa stages and metastatic disease were
confirmed. PCa patients were classified into localized (n = 48) and metastatic PCa (n
= 14). Serum levels of IGF-1, IGFBP-3, CgA, PSA and free/total PSA were measured
as well as possible association between parameters were assessed. The validity
(sensitivity and specificity) were evaluated by ROC curve analysis. Results: Serum
PSA levels were significantly higher in PCa than BPH and control groups (p<0.05)
and attained sensitivity of 87% at 85% specificity with an accuracy of 86%. Although
serum IGF-1, IGFBP-3 and CgA levels did not differentiate among PCa, BPH and
control groups (p>0.05), IGF-1/PSA as well as IGFP-3/PSA ratios were found to
differentiate significantly among metastatic, localized PCa, BPH and control groups
(p<0.05). Combined use of IGF-1/PSA and IGFP-3/PSA ratios provide an overall
value of sensitivity, specificity and diagnostic accuracy (92, 84 and 88% respectively)
in the diagnosis of PCa. The addition of f/t PSA ratio to this combination seems to
improve the overall value of sensitivity, specificity and diagnostic accuracy (94, 85
and 90% respectively). Conclusion: Although there is no association of PCa risk with
serum IGF-I and IGFBP-3 levels; combination of these growth factors with PSA and
f/tPSA may be useful and can improve the overall value of sensitivity, specificity and
diagnostic accuracy of patients with PCa. Further studies are needed to elucidate the
prognostic and predictive value of these growth factors as well as their association
with PCa risk.

A-012

predictor of ovarian cancer with a sensitivity of 76% and a specificity of 95% (Moore
RM et al, 2008). The purpose of this study was to determine the stability of the serum
analyte HE4 under frozen conditions.
Methods: This is a retrospective study utilizing banked serum samples collected in an
ovarian cancer clinical trial (NCT00315692). The patients were from women greater
than or equal to 18 years of age, who selected to undergo laparotomy or laparoscopy
based on finding of a pelvic mass. The samples were collected in the US under an
IRB approved protocol. Samples were collected in red top tubes or serum separator
tubes (SST); and undergone no more than five (5) Freeze/Thaw cycles prior to the
date for stability testing. Samples were stored in a -70C freezer since the time of
collection. Sample testing was performed using the manual HE4 EIA, a sandwich
immunoassay. The initial HE4 EIA testing was carried out in July to August of 2007
and the results were retrieved from the clinical trial dataset. The stability testing with
the HE4 EIA was carried out in June of 2013. A total of 100 available samples from
this trial were tested. Among them, 84 from women with benign disease, 11 from
women with border line/low malignant potential, 4 from women with ovarian cancer,
and 1 from a woman with other gynecological cancer.
Results: The linear correlation coefficient between the two measurements was 0.986
(95% CI: 0.979 to 0.991). Weighted Deming regression gave an intercept of 2.76
(95% CI: -0.54 to 6.07); and a slope of 1.033 (95% CI: 0.970 to 1.095). The intercept
and slope are not significantly different from 0 and 1 respectively. Passing-Bablok
regression produced a similar intercept of 3.06 (95% CI: -1.26 to 6.35) and slope of
1.031 (95% CI: 0.965 to 1.108).
Conclusion: The human serum HE4 was demonstrated to be stable for at least five
(5) years for the serum samples stored at -70C and underwent no more than five
(5) Freeze/Thaw cycles.

A-013
Circulating proteolytic products of tumor-resident enzyme as potential
biomarkers for early detection of breast cancer

T. Hu. Houston Methodist Research Institute, Houston, TX

In theory, any or all of tumor-secreted proteins can serve as cancer biomarkers.
However, the reality is challenging to monitor because of the large degree of
fluctuation in abundance and localization of these tumor-secreted proteins, especially
in the early stage of tumor development and/or metastasis. As such, it seems feasible
that we might take advantage of the fact that secreted proteases/peptidases in the
tumor microenvironment generate proteolytic products, also referred to as circulating
peptides, which are detectable in bloodstream and provide ample information about
the body, coded in the patterns and quantity of these peptides. Herein we clearly link
the catalytic activity of Carboxypeptidase N (CPN) to its proteolytic products during
breast tumor progression in mouse model and clinical samples. CPN plays important
roles in regulating vasoactive peptide hormones, growth factors, and cytokines by
specifically cleaving their C-terminal basic residues. Circulating fragment profiling,
by an approach combining nanopore fractionation and mass spectrometry, revealed
the nature and extent of cleavage by CPN. These results correlated with the level
of CPN-catalyzed peptides in blood taken from the tumor-bearing mice, healthy
women and breast cancer patients. We showed that generation of C3f_R1310-L1319
specifically correlated with the CPN expression level. In both mouse and clinical
patient samples, the amount of CPN was increased in tumor tissues compared to that
seen in normal breast tissue, while its counterpart in blood remained constant. The
amount of 6 CPN-catalyzed peptides predominantly increased in sera taken from both
the mice at 2 weeks after orthotropic implantation and the patients plasma as early
as the first pathologic stage of breast cancer. In conclusion, the circulating level of
the selected 6 CPN-catalyzed proteolytic products reflects the extent of this enzymes
activity in tumors, and our results clearly indicate their strong potential as biomarkers
for non-invasive early diagnosis of breast cancer.

Stability of Serum Human Epididymis Protein 4 (HE4)

R. Radwan1, C. Fermer2, M. Lundin2, S. Raju1, S. Jones1, C. Hall2, D.

Dickson1, T. Kettlety1, Z. Li1. 1Fujirebio Diagnostics Inc., Malvern, PA,
2
Fujirebio Diagnostics AB, Gothenburg, Sweden
Background: HE4 is encoded as a 13 kDa protein and belongs to the family of whey
acidic four-disulfide core (WFDC) proteins (Israeli O et al, 2005; Bouchard D et al,
2006; Bingle L et al, 2002). HE4 was first identified in the epithelium of epididymis
(Kirchhoff C et al, 1991; 1998). Secreted HE4 has become an important biomarker for
the detection of ovarian cancer, a common cause of cancer-related death in women,
with 67% sensitivity and 96% specificity (Hellstrom I et al, 2003). Furthermore,
the combination of HE4 and CA 125 has been demonstrated to be a more accurate

S4

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

A-017
WHAT ARE THE TUMOURS MARKETS THAT BEST IDENTIFY
MALIGNANT PLEURAL EFFUSIONS?

J. D. Santotoribio, A. Garca de la Torre, C. Caavate-Solano, F. ArceMatute, S. Prez-Ramos. Puerto Real University Hospital, Cdiz, Spain
Introduction Malignant pleural effusions (MPE) are a common clinical problem in
patients with neoplastic disease. The aim of this study was to determine the accuracy
of carcinoembryonic antigen (CEA), cancer antigen 15.3 (CA 15.3), cancer antigen
19.9 (CA 19.9) and cancer antigen 125 (CA 125) measurement in pleural fluid for
diagnosis of MPE.
Materials and methods We studied pleural fluids obtained by thoracocentesis in
patients with pleural effusion (PE). We measured CEA, CA 15.3, CA 19.9 and CA
125 in pleural fluid by electrochemiluminescence immunoassay in MODULAR E-170
(ROCHE DIAGNOSTIC). Patients were classified into two groups according to the
aetiology of PE: benign PE (BPE) and MPE. PE was categorized as MPE if malignant
cells were demonstrated in pleural fluid or pleural biopsy. The accuracy for diagnosis
of MPE was determined using receiver operating characteristic (ROC) techniques by
analysing the area under the ROC curve (AUC).

A-016
Clinical Utilization of the Cancer Profile and Longevity Profile and the Role of
Thymidine Kinase (TK1) in Carcinogenesis

E. K. Schandl. American Metabolic Laboratories, Hollywood, FL

Cancer is usually detected by visualization. Clinical biochemical parameters,
e.g. Cancer Profile (CA Profile) and Longevity Profile (expanded CA Profile), are
capable to signal a developing malignancy much earlier. The methodologies used
in this study are: Chemiluminescence (HCG, CEA, TSH, DHEA-S), IRMA (HCG)
and enzyme kinetics (PHI, GGTP). HCG, the pregnancy/malignancy hormone is the
autocrine proliferative factor (APF). It is responsible for de novo DNA, RNA, protein
synthesis in pregnancy and may as well be so in malignancies. PHI is a neurokine.
Amongst its many other functions it is the autocrine motility factor (AMF). PHI
regulates anaerobic sugar metabolism by facilitating the Warburg effect. This enzyme
is responsible for cytokines and as such, may be the facilitator of micrometastesis
and circulating tumor cells by attaching to their membrane receptor (neurokine) and
jockeying cells to a distant site. Thymidine Kinase (TK1) is a dynamic measure of
tumor growth. It phophorilates deoxythymidine to deoxythymidine monophosphate, a
prerequisite for DNA replication. All three factors, HCG, PHI, TK1, must be present
for the development and sustenance of malignancies. Our data confirms the presence
of PHI and HCG in most, if not all cancers. TK1 was found in cancer biopsies, but not
in normal tissues. Reports showed the presence of the enzyme in embryonic, wound
healing, and tumor cells.
Clinical laboratory results confirmed positive CA Profile markers in approximately
90% of hundreds of pathologically established malignancies. Breast cancer yielded
92% positives, lung 97%, and colon 93%. The clinical laboratory adaptation of the
proposed profile may warn of developing, undiagnosed cases, and track progress
monitoring. The Longevity Profile is a conglomerate of laboratory tests for cancer,
coronary risk factors, sex hormones, bone health, adrenal stress, and generally
an overall examination of most organs. It is a biochemical full body scan without
radiation.

Results We studied 152 patients with ages between 1 and 89 years old (median =
61.5 years old). Fifty-one patients were MPE (22 lung cancer, 9 breast cancer, 7
mesotheliomas, 5 lymphomas, 4 kidney cancer, 2 colon cancer and 2 ovarian cancer)
and 101 were BPE (46 transudates, 42 parapneumonic, 5 tuberculous, 5 pulmonary
thromboembolism, 2 quilotrax and 1 rheumatoid arthritis). No statistically significant
differences were found between MPE and BPE patients according to CA 125 (p>0.05).
AUC values was 0.815 (p=0.0024) for CA 15.3, 0.682 (p=0,0064) for CEA and 0.639
(p=0,0359) for CA 19.9. Optimal cut off value were 16.2 U/mL (61.5% sensitivity and
86.3% specificity), 2.3 ng/mL (57.7% sensitivity and 76.6% specificity) and 2.4 U/mL
(64.3 sensitivity and 62.3% specificity) for CA 15.3, CEA and CA 19.9 respectively.
Conclusions CA 15.3 levels improved accuracy for diagnosis of MPE compared with
CEA, CA 19.9 or CA 125. CA 15.3 showed high diagnosis efficacy to predict whether
a pleural effusion is benign or malignant.

A-018
Relation of total antioxidant capacity and CA 125 in patients with epithelial
ovarian carcinoma

L. Kurti1, I. Osmani1, A. Kelmendi1, L. Zeneli2. 1University of Prishtina,

Faculty of Medicine, Institute of Biochemistry, Prishtina, Kosovo, Republic
of, 2University Clinical Centre of Kosova, Institute of Clinical Biochemistry,
Prishtina, Kosovo, Republic of
Background:We undertook the present study to investigate the possible relation
between total antioxidant capacity (TAC) and CA 125 values in serum samples of
patients with epithelial ovarian carcinoma.
Methods:Serum total antioxidant capacity was measured using trolox equivalent
antioxidant capacity (TEAC) assay in 20 serum samples with elevated values of CA
125 and 20 age- and sex-matched healthy subjects (controls) with CA 125 values
within reference ranges respectively.
Results:The measured level TAC was higher in serum samples from patients with
elevated CA 125. TAC correllated significantly with CA 125 (r: 0.516, P<0.05) when
the values of this tumor marker were pathological.
Conclusion:This study suggests that increased serum TAC of the patients with altered
levels of CA 125 may be due to the response of increased reactive oxygen species
and can be considered as a sign of oxidative stress of these patients. Therefore, the
evaluation of serum antioxidant capacity in patients with epithelial ovarian carcinoma
could contribute in diagnosis of these patients.

A-020
Experience with the use of the CellSearch system for enumeration of circulating
tumor cells (CTCs) in Asian subjects

T. C. Aw, M. Goh, S. Swe. Singapore Institute of Advanced Medicine,

Singapore, Singapore
Background: Circulating tumor cells (CTCs) are increasingly used as independent
prognostic markers as well as predictors of response to anti-tumor treatment. Solid
tumors are derived from the epithelium. Metastasis is initiated when tumor cells
invade the blood stream. Normally absent in blood, these circulating epithelial-

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S5

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

derived cells can be captured using antibodies directed against the epithelial cell
adhesion molecule (EpCAM). The CellSearch CTC system (Janssen Diagnostics) is
the only FDA approved platform for CTC detection of metastatic cancers from the
breast (since 2004), colon (2007), and prostate (2008). We acquired the CellSearch
system recently and now describe our experience with its use in providing a clinical
service for CTC analysis in the Asia-Pacific region.
Methods: The CellSearch system comprises an automated CellTracks AutoPrep
instrument to capture and label the CTCs and a semi-automated immuno-fluorescent
microscope (CellTracks Analyzer) for cell detection. Blood, collected in proprietary
tubes, are stable for up to 96 hours prior to analysis. Buffer containing ferrofluid
(nanoparticles with a magnetic core and an outer layer coated with anti-EpCAM
antibodies) is added to the sample. Following immuno-magnetic capture and
enrichment, CTCs are permeabilized before exposure to fluorescent antibodies directed
against the cytoplasm (anti-CK-8,18,19), nucleus (DAPI) and leucocytes (anti-CD45).
Thereafter the mixture is transferred into a plastic cartridge surrounded by a magnetic
sheath. The CTCs are attracted to the surface of the cartridge by the magnet. The
analyzer can accommodate up to 8 samples in each run of 2-3 hours. Each assay has
a control sample comprising SK-BR3 breast cancer cell lines tagged with 2 different
fluorescent labels one each for a population of low CTC count (approximately 40-50
cells) and another with high CTC count (approximately 1000 cells). Following a 30
min incubation, the samples are transferred singly to the Analyzer which scans the
surface of the cartridge. Fluorescent objects are imaged and displayed in a gallery for
classification. CTCs are defined as CK+,DAPI+,CD45- fluorescent objects of at least
4 microns and co-location of the cytoplasmic and nuclear images.
Results: Presumably healthy men (n=45: 27-85 years, mean age 55) had less than
3 CTCs (84% < 2 CTCs); on repeat testing CTCs declined to < 2. Healthy women
(n=56: 25-79 years, mean age 48) had less than 2 CTCs (98%). We have served 102
patients with metastatic cancer (Breast: n=43, age 34-71; Prostate: n=29,age 5079; Colon: n=30, age 28-80). We have also analyzed samples from other cancers
(brain, nose, tongue, lung, stomach, kidney, ovary, endometrium, cervix). Inter-assay
precision using the kit controls range from 10-27% for the low control and 3-13% for
the high control (n=11 control lots). External quality assurance (EQA) comprises a
gallery of 50 images sent out quarterly by the manufacturer. All users have attained
a score of > 90%. We have become aware of a third party EQA program which we
intend to subscribe.
Conclusion: CTC testing is gaining in utility. As oncologists gain more experience
with using CTC, testing will become mainstream and impact the routine clinical
laboratory.

A-023

multiple myeloma should now be investigated. The utility of Seralite in the context
of other FLC related disorders including AL amyloidosis should also be established.

A-024
Method Comparison to Quantify Free Light Chain Changes during Serial
Monitoring of Multiple Myeloma Patients

O. Berlanga, L. Adie, H. Carr-Smith, S. Harding. The Binding Site,

Birmingham, United Kingdom
Background: Serum free light chain (FLC) measurements utilising Freelite
immunoassays are an integral part of the international myeloma working group
guidelines. The assays are available on both turbidimetric and nephelometric
instruments. Here we compare responses in FLC measurements between instruments
in serial samples from multiple myeloma (MM) patients.
Methods: Sequential sera from 6 MM patients (5 IgG, 1 IgG; total sample
number=45; median (range) sample number: 8 (5-11); mean (range) follow-up:
289 (97-525) days)) treated with bortezomib were analysed retrospectively with
Freelite (The Binding Site Group Ltd, UK) on a SPAPLUS, Cobas Integra 400 and
Modular P turbidimeters and the BNII nephelometer. Percentage change in dFLC
(involved-uninvolved FLC) relative to baseline was calculated for each instrument,
and correlation and agreement assessed using Pearsons linear regression, the
coefficient of determination (R2) and Bland-Altman test. Responses were assigned
for each sequential sample following IMWG criteria. Agreement between instruments
for response assignment was assessed using weighted kappa analysis; values >0.81
indicate near perfect agreement.
Results: Median FLC, FLC and dFLC concentrations were not significantly
different between instruments (Mann-Whitney test), except for slightly higher FLC
measurements on the Modular P. Regression analysis between analysers displayed
dFLC % change slopes between 0.95 and 1.01 and intercepts between -0.06 and
0.03, with R2 0.96 for all comparisons (Table 1). Bland-Altman test revealed a
bias between 0.2 and 2.7% between instruments, with 95% limits of agreement no
greater than 16%. Weighted kappa analysis for response assignment was 0.89 for
all instrument comparisons (Table 1).
Conclusion: There is good correlation and agreement for FLC changes and response
assignment by Freelite on different analysers, indicating that routine monitoring of
the disease would not be affected by instrument selection. Small between instrument
differences in the measured FLC levels
suggest not changing analysers during
monitoring patients is preferable.

New rapid urine test for the identification and quantitation of immunoglobulin
free light chains (Bence Jones Proteins)

J. Campbell, J. Heaney, P. Patel, M. Goodall, M. Drayson. University of

Birmingham, Birmingham, United Kingdom
Background. Monoclonal and immunoglobulin free light chains (FLC) in the
urine are important biomarkers for the diagnosis and monitoring of a number of
plasma cell dyscrasias including multiple myeloma. To date, laboratory FLC tests
provide the only means of quantitating FLC and often have a slow turnaround time
that prevents early diagnosis or prompt identification of changes in disease activity.
Furthermore, the gold standards for identifying (immunofixation electrophoresis; IFE)
and quantitating (densitometry) FLC in the urine have a number of limitations. IFE
lacks analytical sensitivity (LOD >10-20 mg/L) and interpretation is often subjective.
Densitometry has high inter-test variability that contributes to an inter-lab CV% of 5095% in the UK National External Quality Assessment Service (NEQAS), and is poorly
sensitive meaning that urines need to be concentrated before measurement, sometimes
up to 150-fold. Further, clinical manifestations such as proteinuria may obscure
monoclonal FLC bands and makes identification and quantitation of monoclonal
protein bands inaccurate. Therefore, we have developed a rapid test (Seralite)
that identifies abnormal FLC levels in unconcentrated urine or blood in 10 minutes.
Seralite quantitates and FLC levels simultaneously using highly specific anti-
and anti- FLC monoclonal antibodies. Methods: Seralite validation was conducted
by retrospective analysis of urine from patients presenting with multiple myeloma
(n=100). All samples were also measured for FLC by electrophoresis immunofixation;
densitometry on concentrated urines; and a recently validated new Luminex assay
that incorporates the same mAbs as Seralite. Results: Seralite displayed excellent
clinical concordance with Luminex. Analysis of IFE results revealed that Seralite
had no false negatives, and correlated excellently with densitometry. Conclusion:
Seralite detected all FLC in urine from 100 myeloma patients at diagnosis.
Prospective use of Seralite to diagnose and monitor plasma cell dyscrasias including

S6

A-025
Differential diagnosis the property of ascites by a novel logistic regression
model

C. Lin, W. Feng, L. Leilei, S. Yuanyuan, W. Yueping, S. Jianguo. Hospital,

Nantong, China
Background: Differential diagnosis of malignant from benign ascites has a great
clinical significance for the treatment and prognosis of the disease. However,
complete discrimination between malignant ascites and nonmalignant ascites has
not yet been substantially improved in recent years. Herein, we established a logistic
regression model on the basis of multiple ascitic indices in differential diagnosis both
benign and malignant ascites. Moreover, the further assessment of its diagnostic value
is presented.
Methods: A total of 103 consecutive ascitic patients were enrolled in this study. Nine
biomarkers including telomerase, DNA ploidy, adenosine deaminase (ADA), lactate
dehydrogenase (LDH), CEA, CA125, CA19-9, Golgi Protein 73 (GP73) and serumascites albumin gradient (SAAG) were measured. The data were further analyzed

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

by using receiver operating characteristic (ROC) curve, univariate and multivariate

logistic regression to evaluate the value of differential diagnosis the property of
ascites.
Results: The median values of the ascitic telomerase, LDH, CEA, CA125, CA199, GP73 in the malignant ascites group were 0.314, 235 U/L, 20.64 ng/mL, 306 U/
mL, 45.21 U/mL and 185.1 g/L, as they were compared with those of benign group
(0.046, 109 U/L, 4.84 ng/mL, 62.13 U/mL, 19.5 U/mL, 69.8 g/L), respectively,
P<0.001. However, the concentration of SAAG in the malignant ascites group were
obviously lower than that of benign ascites group (median, 7.09 g/L vs 19.20 g/L),
P<0.001, Moreover, there was no significant difference in the concentration of
ADA between the two groups (median, 8.5 U/L vs 8.0 U/L), P>0.05. In addition,
DNA aneuploidy rate in the malignant group was significantly higher than that in
the benign group (76.0% vs 9.4%), P<0.001. By using ROC curve, univariate and
multivariate logistic regression analysis, ascitic telomerase (X1), CEA (X2), GP73
(X3), SAAG (X4) were rolled into logistic regression model:P=1/[1+e (-6.320+2.351X1+2.338X
2+4.246X3+3.459X4)
], (P: probability predictive value, e: natural logarithm). The area under
the curve of the P value of the predictive probability was 0.986, the cut-off point
was 0.469, the sensitivity was 96% and the specificity was 98.1%. When P0.469,
it was predictively diagnosed as malignant ascites; vice versa, when P<0.469, it was
predictively diagnosed as benign ascites.
Conclusion: Our study highlights that the novel logistic regression model is an
attractive strategy in differentiating property of the ascites and justifies its value in the
studies of diagnosis and therapy in malignant ascites patients.

A-026
Performance Evaluation of the Free PSA Immunoassay on the LUMIPULSETM
G1200 System

Y. Ishii, H. Murakami, H. Karasawa, C. Okamura. Fujirebio Inc, Hachiojishi, Tokyo, Japan

Background: PSA, a glycoprotein with a molecular weight of 30 kDa localized in
prostate glandular epithelial cells, is known to be released into blood when prostate
epithelial cells are damaged by malignancies. It has been demonstrated that the
measurement of PSA in blood is useful in the diagnosis of prostate cancer, followup, and evaluation of therapeutic effect as well as in screening for early detection
of prostate cancer. Percentage of free PSA has been shown to improve diagnostic
sensitivity and specificity around gray zone (4 to 10 ng/mL total PSA). Total PSA
immunoassay is already available on LUMIPULSE G1200 system1. A method has
been developed to measure free PSA and result of performance evaluation is presented.
Methods: Lumipulse G Free PSA is a chemiluminescent enzyme immunoassay
(CLEIA) that uses a two-step method for analysis. In the first step, anti-free PSA
monoclonal antibody coated magnetic particles are incubated with a patient sample.
Following a wash, the alkaline phosphatase-conjugated anti PSA mAb are added to
the mixture and incubated in the second step. Following another wash, the instrument
adds substrate solution to initiate chemiluminescence reactions. The resulting
reaction signals are proportional to the amount of free PSA in the sample and allow a
quantitative determination of free PSA in serum and plasma.
Results: The imprecision of the free PSA assay measured over 20 days using two
controls and three panels (ranging from 0.5 to 23 ng/mL) was total imprecision
of < 3.3%.The calibration range of the Lumipulse G Free PSA was 0 - 30 ng/mL
and showed a linear dose-response relationship within the calibration range. The
Lumipulse G Free PSA correlated linearly with ARCHITECT Free PSA (Slope =
0.96; r = 0.99), Access Hybritech free PSA (Slope = 1.04; r = 0.99) and ELECSYS
free PSA (Slope = 1.02; r = 0.99) within the range of 0.058 to 27.86 ng/mL via testing
120 serum samples. No hook effect was observed at 3,000 ng/mL, and no crossreactivity was observed to PAP (1000 ng/mL). No interference was observed with
unconjugated (18.3 mg/dL) or conjugated bilirubin (20.6 mg/dL), hemoglobin (487
mg/dL), triglycerides (2000 mg/dL), RF (rheumatoid factor, 1000 IU/mL) and HAMA
(1000 ng/mL). The LOB, LOD, and LOQ were 0.001, 0.002 and 0.009 ng/mL with
LUMIPULSE G1200, respectively. Correlation between serum tube and plasma tube
(EDTA-2K, Na heparin) was tested (Slope = 0.99). Serum specimens (60 prostate
cancer, 97 non cancer) whose PSA measurement values were in the gray zone (4-10
ng/mL) were tested using the Lumipulse G Free PSA and the mean percentage of
free PSA (prostate cancer; 14.8%) showed significantly lower than that of non cancer
(20.2%).

A-027
QUANTITATIVE DETERMINATION OF FREE LIGHT CHAINS ( FLC)
- KAPPA AND LAMBDA IN GROUP OF PATIENTS WITH ABNORMAL
SERUM ELECTROPHORESIS PATTERN.

Y. S. Medeiros1, R. Freitas2, M. Debiasi3. 1UFSC, Florianopolis SC, Brazil,

2
Santa Luzia Laboratrio Mdico, Florianopolis SC, Brazil, 3Santa Luzia
Laboratrio Medico, Florianopolis SC, Brazil
OBJECTIVE : The dosage of free light chains (kappa and lambda - FLC) have
been incorporated into guidelines for some hematological malignant diseases such
as multiple myeloma (MM) and other monoclonal gammopathies, in order to aid
diagnosis and monitoring. In physiological and pathological conditions such chains
are likely to be measured in the bloodstream along with the other intact molecules.
This study intends to investigate the concentration of FLC in a group of patients when
compared with the reference values.
PATIENTS AND METHODS: 56 samples of patients with ages within 19 and 88 years
old, which 22 were men and 34 were women were evaluated. Serum electrophoresis
was performed on CAPILLARYS 2 (Sebia ) and FLC dosage was performed by
nephelometry on BN - ProSpec (Siemens) using Siemens N Latex FLC kappa and
N Latex FLC lambda. From these patients, based on renal function verification by
MDRD (Modification of Diet in renal Disease Study- NKDEP) and serum protein
electrophoresis results, 12 samples were excluded due to indication of potential renal
dysfunction; whereas 17 with a record of monoclonal gammopathy and 27 with
polyclonal gammopathy result were included.
RESULTS AND CONCLUSION: Only the results of FLC in patients with normal
renal function (n = 44) were analyzed. For patients with monoclonal gammopathy
(n = 17), eight presented kappa/ lambda index with approximately six times greater
than the reference value, nine with normal values. Regarding patients with polyclonal
gammopathy (n = 27), about 50 % of FLC results were high. Based on these data, we
conclude that the inclusion of FLC assays will contribute to the clinical evaluation of
monoclonal gammopathy. Regarding to polyclonal gammopathy, overproduction of
free light chains as described in infectious processes, and autoimmune liver diseases,
among others, the clinical value of this dosage is still under study.

A-028
Best practices in incidental clinical findings associated with Multiple Myeloma
in patients attending the Emergency Service

J. L. Garca de Veas Silva1, R. Rios Tamayo2, C. Bermudo Guitarte1, P.

Navarro Alvarez2, V. Snchez Margalet1, C. Gonzlez Rodriguez1. 1Hospital
Universitario Virgen Macarena, Sevilla, Spain, 2Hospital Universitario
Virgen de las Nieves, Granada, Spain
Background: The presence of incidental clinical findings (bone pain, pathologic
fractures, anemia, hyperproteinemia, hypercalcemia, acute kidney injury) related to
Multiple Myeloma (MM) in Emergency Service and Primary Care should be studied
for screening the existence of a possible MM. A quick panel based on serum protein
electrophoresis (SPE) and quantification of serum free light chains (sFLC) enables
sensitive quantification of monoclonal component in the study of MM. The application
of this screening panel in patients with these incidental clinical finding without other
diagnosis can help us to efficiently detect a possible MM in much sorter times.
Methods: we studied 5 patients where we found incidental clinical finding
characteristic of MM (anemia, hyperproteinemia, intense bone pain). Sera of the five
patients were sent to the Immunology Lab for the screening of a monoclonal protein.
SPE were performed on CAPILLARYS 2 (Sebia) and the sFLC were measured with
Freelite (The Binding Site) turbidimetric assay. Positive results of the screening panel,
remit the patient to the Hematology Service to complete the study
Results: The results are shown in the table.
Conclusions: In the context of clinical symptoms (bone pain, pathologic fractures,
anemia, hyperproteinemia, hypercalcemia) that alerts to a possible MM case in
patients without obvious clinical diagnosis, we found the application of this protocol
(SPE+sFLC) to be efficient and advisable. The combination of SPE and sFLC yields
a fast and highly sensitivity approach in the screening of monoclonal gammopathies
which in the context of the emergency service is of particular importance.

Conclusion: The Lumipulse G Free PSA assay appears to be an accurate and precise
assay for the automated measurement of free PSA in human serum and plasma.
Not available in the US

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S7

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

Case

Sex

Age
(years)

Cause of
Emergency

Clinical Finding at
Emergency Service
Hyperproteinemia
(12 g/dl)

Female

67

Severe
abdominal
pain

Female

65

Infection and
bone pain

Hyperproteinemia,
hyperviscosity and
thrombocytopenia

Female

64

Intense lower
back pain

Intense back pain

Female

55

Lower back
pain and she
had a fall

Pathological fracture
at D12

Male

12

Lower back
pain and he
Hypercalcemia (16.6
had a fall in the mg/dl)
school

SPE

sFLC

Diagnosis

Multiple
KL= 10.47 mg/l
Large peak
Myeloma IgG
LL=99.59 mg/l
(4.18 g/dl)
Lambda Stage
Ratio=0.11
2 ISS
Multiple
KL=617 mg/l
Large peak
Myeloma IgG
LL=11.1 mg/l
Kappa Stage
(3.28 g/dl)
Ratio=55.59
2 ISS
Multiple
KL=3.15 mg/l
Large peak
Myeloma IgA
LL=102 mg/l
(3.22 g/dl)
Lambda Stage
Ratio=0.031
3 ISS
Light Chain
Two weak KL=28600 mg/l
Kappa Multiple
peaks (0.15 LL= 5.36 mg/l
Myeloma Stage
g/dl)
Ratio=5335.82
3 ISS
Multiple
Very large KL=219 mg/l
Myeloma IgA
peak (4.34 LL=1.01 mg/l
Kappa Stage
g/dl)
Ratio=216.83
3 ISS

A-029
Performance Evaluation of Tumor Marker CA15-3 on Roche Cobas e601
Immunoassay Analyzer

S. St. Romain, I. Bermudez, B. Handy, E. Wagar, Q. H. Meng. University

of Texas MD Anderson Cancer Center, Houston, TX
Backgrounds: Breast cancer is the most prevalent form of cancer diagnosed in
women and is the leading cause of cancer death in women worldwide. Cancer antigen
15-3 (CA15-3) and CA27.29 are different epitopes on the same protein antigen
product of the breast cancer-associated MUC1 gene. Substantial evidence has shown
that elevated cancer antigen 15-3 (CA15-3) levels are associated with advanced
breast cancer and metastasis. Thus, serum CA15-3 is used to monitor the therapeutic
response and recurrence of breast cancer. This study was to assess the analytical
performance of CA15-3 on Roche Cobas e601.
Methods: The Roche CA15-3 method is a sandwich electrochemiluminescence
immunoassay that employs a biotinylated monoclonal CA15-3-specific antibody and
a monoclonal CA15-3-specific antibody labeled with a ruthenium complex, forming
a sandwich complex. The evaluation was performed following CLSI guidelines.
The performance was evaluated for precision, lower limit of detection, linearity, and
accuracy. The within-run and between-run precisions were assessed by analyzing QC
material at low and high level of concentrations. The correlation between CA15-3
results on Roche Cobas e601 and CA27.29 results on Siemens Centaur was assessed.
Results: The within-run CVs for CA15-3 were 1.6% and 1.4% at the levels of 22 U/
ml and 102 U/ml, respectively. The between-run CVs at low and high levels were
2.64% and 2.59%, respectively. The measuring range was determined to be linear
between 1.00 - 300 U/ml. The lower limit of detection was 0.1 U/ml using measurable
value obtained from zero standard + 2SD (n=20). Comparison of CA15-3 on Roche
Cobas e601 with CA27.29 on Siemens Centaur showed that the slope was 1.0 (95%
CI = 0.912 to 1.087) with intercept of -12.61 and correlation coefficient r = 0.967
(Deming). The mean bias was -12.65.
Conclusion: Our data demonstrates that the CA15-3 assay on Roche Cobas e601
analyzer has an excellent precision of performance with good linearity. There is a
good correlation between serum CA15-3 and CA27.29. Serum CA15-3 can be
precisely and accurately measured on Roche Cobas e601 in monitoring response to
therapy and recurrence in breast cancer patients.

A-030
Urinary Free Light Chains in Patients With Polyclonal
Hypergammaglobulinemia And/Or Renal Impairment

V. Brunel1, B. Legallicier2, J. Wils3, M. Quillard1, M. Godin2, S. Claeyssens3.

1
Medical Biochemistry, University Hospital, Rouen, France, 2Nephrology,
University Hospital - Inserm UMR1073 - IRIB, Rouen, France, 3Medical
Biochemistry, University Hospital - Inserm UMR1073 - IRIB, Rouen,
France
Background: In plasma cell dyscrasias, monoclonal free light chains (FLC) are
involved in the pathogenesis of renal failure, a major cause of morbi-mortality.
Besides, urinary excretions of polyclonal FLC are known to increase in patients with
RI and, according to rare data, in those with hypergammaglobulinemia (H). However,
evidence for polyclonal FLC-mediated injury is limited. In this study, we assessed the
effect of H and/or RI on urinary FLC excretions.
Methods: Fresh paired samples of serum and 24h urine were analyzed in 270 patients
exempt of monoclonal gammopathy. Patients with H (n=87) had sum of serum Ig

S8

G, A and M ( Ig) concentrations 20 g/L. All patients were classified in 6 groups

according to their renal function and the presence, or not, of H; Control patients (C,
HC) had physiologic proteinuria (<150 mg/24h) and serum creatinine concentration
in reference ranges. Additionally, control patients C had serum K and L concentrations
and rFLC values in the 100% reference interval. Patients with predominant tubular
(T, HT) or glomerular (G, HG) proteinuria were determined by SDS-AGE profil
(Hydragel Proteinuria, Sebia) and by albuminuria or 50% of total proteinuria,
respectively. FLC renal clearance (Clr) was calculated as the ratio of 24h urinary
excretion of K or L FLC (mg/24h) to their respective serum concentration (mg/L).
Results: median (ranges); Mann-Withney test (significance: P<0.05), Spearman
correlations (significance: P<0.05)
Results: Both in patients with H and in those without, K and L FLC urinary excretions
were significantly greater in T and HT and in G and HG patients than in their
respective control patients C and HC; these values were up to 151 (7-574) and 35
(2-261) mg/24h, respectively in G patients, 421 (93-8640) and 96 (34-3270) mg/24h,
respectively in HG patients. Moreover, FLC excretions significantly correlated with
both serum FLC and creatinine concentrations; in addition, in patients with H, they
significantly correlated with Ig concentrations. Ratios of K Clr to L Clr (K Clr
/L Clr) decreased significantly through different states of renal function from C to
T and to G patients and were negatively correlated with serum creatinine and Ig
concentrations. Patients with H, as compared to those without at the same stage of
renal function, had significant and similar increases in K and L FLC excretions (3.5
and 4.0 fold for K and L FLC, respectively). While K Clr/L Clr values were similar in
C and HC patients, they were significantly lower in HT than in T patients and in HG
than in G patients.
Conclusion: This study determined the appropriate reference intervals for patients
with H and/or RI. In all these patients, urinary polyclonal FLC excretions varied
according to renal function. Besides, H was associated with an increase in FLC
excretions values that was independent of renal function. RI progression was
associated with a decrease in K Clr/L Clr values showing that the renal capacity to
clear K faster than L is progressively lost. In addition, this effect worsened in patients
with H suggesting a polyclonal FLC-mediated injury.

A-031
Serum Free Light Chains in Patients With Polyclonal
Hypergammaglobulinemia And/Or Renal Impairment

J. Wils1, B. Legallicier2, V. Brunel1, M. Quillard1, M. Godin2, S. Claeyssens3.

1
Medical Biochemistry, University Hospital, Rouen, France, 2Nephrology,
University Hospital, Rouen, France, 3Medical Biochemistry, University
Hospital - Inserm UMR107 - IRIB, Rouen, France
Background: Serum immunoglobulin-free light chain (FLC) assay is a major
marker in the identification and management of patients with plasma cell dyscrasias.
However, since these patients frequently present renal impairment (RI), it should be
interpreted with caution owing to the dependence of polyclonal FLC values on renal
function. Besides, recent data reported that serum FLC concentrations increased in
patients with polyclonal hypergammaglobulinemia (H); however, data were rare and
renal function was not always assessed. In this study, we assessed the effect of H and/
or RI on serum FLC concentrations.
Methods: Fresh paired samples of serum and 24h urine were analyzed in 270 patients
exempt of monoclonal gammopathy. Patients with H (n=87) had sum of serum Ig
G, A and M ( Ig) concentrations 20 g/L. All patients were classified in 6 groups
according to their renal function and the presence, or not, of H. Control patients (C,
HC) had physiologic proteinuria (<150 mg/24h) and serum creatinine concentration
in reference ranges. Additionally, control patients C had serum K and L concentrations
and rFLC values in the 100% reference interval. Patients with predominant tubular
(T, HT) or glomerular (G, HG) proteinuria were determined by SDS-AGE profil
(Hydragel Proteinuria, Sebia) and by albuminuria or 50% of total proteinuria,
respectively. Serum FLC (Freelite, The Binding Site) were analysed on a BNII
nephelometer (Siemens Healthcare). Results: median (ranges); Mann-Withney test
(significance: P<0.05), Spearman correlations (significance: P<0.05).
Results: Both in patients with H and in those without, serum K and L FLC
concentrations rose significantly through different states of renal function from C to T
and to G patients; in those latter, these values reached 36 (9-115) and 36 (9-106) mg/l,
respectively in G patients, 139 (40-945) and 109 (30-691) mg/l, respectively in HG
patients. Moreover, FLC concentrations and rFLC values were significantly correlated
with creatinine concentrations; in addition, in patients with H, FLC concentrations
significantly correlated with Ig concentrations. Patients with H, as compared to those
without at the same state of renal function, had significant and similar increases in
FLC concentrations (3.5 and 3.0 fold for K and L FLC, respectively, for comparisons

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

of C vs HC patients, T vs HT and G vs HG patients). Alike, patients with H had

a significantly and similar increase in rFLC values than those without (1.2 fold for
comparisons of C vs HC patients, T vs HT and G vs HG patients).
Conclusion: This study determined the appropriate reference intervals for patients
with H and/or RI. In all these patients, serum polyclonal FLC concentrations and
rFLC values shifted to higher values with RI progression and varied according to renal
function. Besides, H was associated with an increase in FLC concentrations and in
rFLC values that were independent of renal function. Therefore, rFLC values should
be interpreted with caution, not only in case of RI, but also in case of H: we showed
that rFLC values between 0.24 to 0.74 should provoke a thorough search for plasma
cell dyscrasias and lymphoproliferative disease.

For the serum panel, the ROC area dropped to 0.85 (vs. 0.95 for the training set); for
the plasma panel, the ROC area dropped to 0.81 (vs. 0.93). Nevertheless, even the
ROC area of 0.85 for the serum panel with clinical sensitivity and specificity of 81%
and 84%, respectively, and the ROC area of 0.81 for the plasma panel with clinical
sensitivity and specificity of 76% and 78%, respectively, should be clinically useful.
Analysis of the combined training and test sets with 100x cross-validation resulted
in a 4-marker serum panel (Flt-3L, EGFR, MMP-3, and NME-2) with an ROC area
of 0.91 and clinical sensitivity and specificity of 88% and 82%, respectively, and a
5-marker plasma panel (Flt-3L, cytokeratin-19, Flt-1, KGF, and HGF) with an ROC
area of 0.91 and clinical sensitivity and specificity of 84% and 83%, respectively.
Conclusion:Using MULTI-ARRAY technology and high quality clinical samples, we
were able to identify promising biomarker panels for early detection of lung cancer
in high-risk individuals.

A-034
A-036

Huaier suppresses proliferation and induces apoptosis in human lung

adenocarcinoma cells via promotion of miR-26b-5p

Z. Lu, T. Wu, W. Chen, S. Liu, H. Lu, H. Wang, X. Huang, Q. Kong.

Central Hospital of Wuhan, Wuhan, China
Background: Aqueous extract of Trametes robiniophila murr(Huaier)has been applied
for cancer complementary therapy in recent decades. Varies studies have reported that
Huaier possess the anti-tumor effects. However, the mechanisms are not completely
elucidated. MicroRNAs (miRNAs) are small (18-25 nucleotides), noncoding
RNAs whose dysregulation have been discovered to involve in tumorigenesis and
development.
Methods and Results: In this study, we found miRNA expression profiles were
altered in Huaier-treated human pulmonary adenocarcinoma A549 cells. MiR-26b-5p,
which is upregulated in the expression profiles and simultaneously downregulated in
both several lung cancer cell lines and patients, was selected for further study. We
then used miRNA mimics or inhibitors to perform gain- and loss-of-function studies
to demonstrate the roles of miR-26b-5p in pulmonary cancer. Moreover, EZH2 was
identified as a target of miR-26b-5p by luciferase reporter assay and by EZH2 knockdown we observed a decrease in cell proliferation and an increase in apoptosis rates of
A549 cells, which was corresponding to the effects of both Huaier treatment and the
transfection of miR-26b-5p mimic. Additionally,-catenin and bcl-2, as the indirect
downstream effectors of EZH2, was found attenuated after Huaier treatment and miR26b-5p overexpression.
Conclusion: Taken together,our findings shed light on the mechanisms that Huaier
might suppress proliferation and induce apoptosis in lung cancer by miR-26b5p-EZH2-mediated approach in lung cancer cells, which provides a new idea for
understanding the anti-tumor effects of Huaier.

A-035
Blood Test for Early Detection of Lung Cancer

M. Stengelin1, H. Pass2, W. Rom2, S. Kumar1, S. Vaithlingam1, A.

Aghvanyan1, D. Roy1, G. Dobrescu1, R. Sivakamasundari1, E. N. Glezer1,
J. N. Wohlstadter1. 1Meso Scale Diagnostics, LLC, Rockville, MD, 2New
York University, New York, NY
Background: Lung cancer is the largest single cause of death from cancer worldwide.
Even though lung cancer often can be treated successfully when detected early,
approximately 90% of patients diagnosed with lung cancer die of the disease. Screening
with low dose CT can reduce mortality, but the positive predictive value of this test is
low, leading to a large number of suspicious but ultimately non-malignant results that
nevertheless require follow-up. Our objective was to develop a simple blood test to
risk-stratify patients at high risk of lung cancer.Methods: We developed multiplexed,
serum/plasma immunoassay panels to measure more than 40 lung cancer-related
biomarkers using a 96-well, 7-spot format and electrochemiluminescence detection.
Due to the high sensitivity of MSDs MULTI-ARRAY technology, these panels were
run with diluted serum or plasma, bringing the total sample volume required to run all
40 assays down to approximately 40 L per replicate. This enabled us to measure all
markers simultaneously in precious, high-quality serum and EDTA plasma samples.
We used samples from early-stage lung cancer patients (drawn before lung cancer
surgery) and from a lung-cancer screening cohort of age-matched heavy smokers who
did not have lung cancer at the time of the blood draw.Results: In a training set of 300
samples, 12 serum and 6 plasma markers had areas under an ROC curve (ROC areas)
of 0.7 or higher. We used a logistic regression model with 100x cross-validation to
develop a multi-marker panel. One serum panel (Flt-3L, EGFR, MMP-3, and NME-2)
and one plasma panel (Flt-3L, cytokeratin-19, MMP-3, Flt-1, KGF, and PlGF) were
selected and tested using approximately 250 additional samples from the same cohort.

Serum MicroRNA Panel as Novel Non-Invasive Biomarker for Early Diagnosis

of Cervical Cancer

Y. Zhang1, N. Ning1, Q. Li2, W. Cui1. 1Peking Union Medical College

Hospital Chinese Academy of Medical Sciences, Beijing, China, 2Duke
University Medical Center, Durham, NC
Background: Currently, pathologic evidence of malignant cells, which typically
requires an invasive strategy such as vaginoscopy and cervical biopsy, or loop
electrosurgical excisional procedure, is referenced as the gold standard in diagnosing
human cervical cancer. And serum tumor biomarkers, such as CEA, AFP, CA125,
SCCAg, have provided some predictive information to tumor diagnosis, but poor
sensitivity and specificity has also limited their clinical applications. We aimed to
identify serum miRNAs for diagnosing cervical cancer.
Methods: Serum miRNA expression was investigated from 348 participants by using
qRT-PCR, including 111 patients with cervical cancer, 115 cervical intraepithelial
neoplasia (CIN) individuals and 122 healthy controls, recruited between July 2012
and November 2013 from Peking Union Medical College Hospital in China. First,
we fully screened the differently expressed 425 miRNAs in 9 serum samples for
diagnosing cervical cancer. A logistic regression model was constructed using a
training cohort (n=72) and then validated using an independent cohort (n=269). Hela
cells stably expressing miR-497 were established to analyze their roles in vivo and
vitro.
Results: We identified a miRNAs panel (miR-124-3p, miR-195, miR-2861, miR-497)
that provided high accuracy in discriminating cervical cancer from healthy controls
(AUC=0.907 and 0.793 for training and validation groups, respectively), from CIN
individuals (AUC=0.960 and 0.963 for training and validation groups, respectively).
This 4 miRNAs can also differentiate CIN from healthy controls (AUC=0.903 and
0.87 for training and validation groups, respectively). Among the 4 up-regulated
miRNAs, miR-497 levels in serum were the most specific one for cervical cancer that
had no significance between ovarian or breast cancer patients and healthy controls.
Forced expression of miR-497 suppressed proliferation and induced apoptosis of
Hela cells (p<0.05) in vitro. Further investigation showed that Hela xenograft mouse
treated miR-497 overexpression was significantly smaller in weight than control
(p<0.05). MiR-497 could exert the effect of tumor growth inhibition in vivo.
Conclusion: Our results have identified a miRNAs panel (miR-124-3p, miR-195,
miR-2861, miR-497) that has considerable clinical value in diagnosing cervical
cancer as a novel noninvasive approach.
Keywords: miRNA panel; cervical cancer; serum

A-038
Can the incomplete serum separation on gel tube vacutainers lead to the
diagnosis of Multiple Myeloma ?

S. Sen1, S. Chakraborty2. 1Calcutta Medical Research Institute (CMRI),

Kolkata, India, 2Peerless Hospital & B K Roy Research Centre, Kolkata,
India
BACKGROUND: Gel tubes have become common in clinical labs and have made
analysis and storage of samples easier; eliminating the need for transferring of serum
into secondary tubes The basic principle is gradient density centrifugation using the
thixotropic property of the gel, where it forms a barrier separating serum from the
cells. Occasionally incomplete separation is seen in some samples where the gel
packed cells fail to go below the gel. There are very few studies conducted on this
phenomenon and some studies have suggested that incomplete separation of serum
occur in patients with paraproteinemia particularly multiple myeloma (MM).

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S9

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

METHODS: Hence we conducted a prospective study on the relationship between
incomplete serum separation on gel tubes and paraproteinemia. This was done to
identify whether incomplete gel separation was associated with increased total
protein. The gel tubes used in our study was BD- SST.
RESULTS: This study was done for a period of two years. Incomplete gel separation
was seen in a total 14 gel tube samples out of a total of 99,850 patient samples (0.010
%).In 4 samples the incomplete separation was corrected on repeat sample collection.
In 10 patient samples we observed incomplete separation even on repeat sampling.
Raised total protein with altered albumin to globulin ratio was seen in those samples.
Serum Protein Electrophoresis (SPEP) confirmed the presence of M bands in all
the 10 cases and subsequently multiple myeloma was confirmed with bone marrow
aspiration. The immunoglobulin subtypes with immunofixation were: Ig A (5/10), Ig
M (4/10) and biclonal type with Ig M and Ig G (1/10).
CONCLUSIONS: Our study shows that incomplete separation on gel tubes is very
commonly associated with paraproteinemia like multiple myeloma. The increase
in paraprotein component possibly increases the viscosity of the sample leading
to inhibition in separation. Clinical laboratorians need to be aware of this and
should estimate the total protein and albumin reflectively in those patients showing
incomplete separation. Patients having increased total protein and altered albumin
globulin ratio should be followed up with SPEP. Clinical correlations and interaction
with treating physicians might lead to early diagnosis in such patients

A-039
Clinical comparisons of two free light chain assays to immunofixation
electrophoresis for detecting monoclonal gammopathy

M. Park, H. Kim, H. Kim, K. Shin, W. Song, H. Kim, H. Kim. Hallym

University College of Medicine, Seoul, Korea, Republic of
Background: Free light chains (FLC) are useful biomarkers for the diagnosis and
monitoring of various plasma cell dyscrasias. Recently, several monoclonal antibodybased assays for serum FLC have become available.
Methods: One hundred fifty seven samples from 120 patients for screening or
monitoring of monoclonal gammopathy (MG) were included in this study. The new N
Latex FLC assays (Siemens Healthcare Diagnostics GmbH, Germany) were compared
with the Freelite FLC assays (The Binding Site Ltd, UK) and immunofixation
electrophoresis (IFE).
Results: The Freelite FLC assay showed significantly wider assay ranges than the
N Latex FLC assay. The correlation coefficients of the two FLC kappa () assays,
lambda () assays, and the / ratio were 0.9792, 0.8264, and 0.9064, respectively.
The concordance rate was 84.7% for the FLC assays, 79.6% for FLC , and 89.2%
for the / ratio. Compared to the results for IFE, the clinical sensitivity, specificity,
and percent agreement of the / ratios were as follows: 72.2%, 93.6%, and 82.8%,
respectively, for the Freelite assay and 64.6%, 100%, and 82.2%, respectively, for the
N Latex FLC assay. Conclusion:
Several differences in dynamic assay ranges were observed between the two FLC
assays. The N Latex FLC assay showed good correlations and concordance with the
Freelite FLC assay. The clinical sensitivity of the / ratio was higher in the Freelite
FLC assays; however, clinical specificity was higher in the N Latex FLC assay.

A-040
Contribution of Fokl polymorphism to disease development and risk prediction
values in bladder cancer cases

O. Baykan1, M. Akgul1, N. Uren2, F. Gerin1, I. Tnay1, E. Ergul1, A. Sazci2,

L. Turkeri1, G. Haklar1. 1Marmara University School of Medicine, Istanbul,
Turkey, 2University of Kocaeli, Kocaeli, Turkey
Bladder cancer is the fourth most common cancer in men. Although smoking is known
to be the most important etiological factor in bladder cancer, 40% of cases remained
to be unknown. In our study we aimed to investigate the contribution of a common
single nucleotide polymorphisms rs2228570 (FokI) in the vitamin D receptor gene to
the formation of urothelial bladder cancer.
Age and gender matched 101 patients diagnosed as urothelial bladder cancer and
109 healthy individuals who has no history of cancer in their first degree relative
were included in the study. Polymerase chain reaction, and restriction fragment
lenght polymorphism techniques were used to determine the polymorphisms. The
frequencies of FokI polymorphism FF, Ff and ff genotypes were 60.4%, 31.7%, 7.9%
in bladder cancer and 44%, 47.7% and 8.3% in controls, respectively (p=0.048). FF
genotype frequences were higher (p=0.018) in patients, while Ff frequences were
lower (p=0.018) compared to controls (Table 1).

S10

Associations between risk factors and cancer were estimated by calculating ORadj using
logistic regression analyses. When smoking status and FokI polymorphism analysed
together, the effect of genotype and allel frequencies on cancer risk prediction were
not statistically significant, however smoking increased bladder cancer risk 7.27
times (ORadj= 7.27; 95% CI= 3.8-13.9; p<0.001) The genotype distributions of the
polymorphisms were in agreement with Hardy-Weinberg equilibrium among the cases
and controls.
Studies investigating the contribution of VDR gene polymorphism and urothelial
cancers were limited, although there were many publications for other cancer types.
Our study is the first for inverstigating this relation in Turkish population. We
demonstrated significant polymorphism in the patient group when compared to the
control, however there was no effect of genotype on cancer occurence. Further studies
which will be planned to reveal the effect of this difference may be beneficial in the
etiopathogenesis of urothelial cancers.

A-041
Enrichment of heterogenous circulating tumor cells by multiplexed
immunomagnetic micro particles

K. Bourcy1, K. Goudy1, L. M. Millner2, M. W. Linder1, R. Valdes1. 1PGXL

Technologies, Louisville, KY, 2University of Louisville, Louisville, KY
Background: Circulating tumor cells (CTCs) are low abundance cells that have
detached from the primary tumor and may produce metastatic lesions. They represent
the greatest threat to a cancer patient. Paradoxically, they may also be an invaluable
source of prognostic information by predicting metastases and chemotherapy
resistance. Significant obstacles and limitations exist in current methods utilized for
the detection and isolation of CTCs. Currently, the only FDA-cleared system for the
detection of CTCs has a number of drawbacks including: i) the requirement that buffy
coat be used, resulting in cell loss; ii) the cells must be fixed, which severely limits the
ability to perform downstream analysis; and, iii) this system uses a single antigen for
the recovery of CTCs, which can miss cells that do not express the target antigen. In
this study we describe the application of a magnetic bead technology for enrichment
of low abundance CTC where we evaluated this technology for enrichment of low
abundance cells from whole blood, potential benefits in enrichment using a multiantigen approach, and viability of CTCs following the enrichment procedure. The
long-term objective of this study is to develop a multi-plex immunomagnetic method
for the isolation and recovery of circulating tumor cells from whole blood, with the
ultimate goal of characterizing circulating tumor cells in breast cancer patients.
Methods: We pre-labeled SKBR cells with carboxyfluorescein succinimidyl ester and
spiked 10 or fewer cells into 1 mL of whole blood. The spiked blood was incubated
with anti-EpCAM or a combination of anti-EpCAM and anti-HER2 conjugated
magnetic microbeads. Blood was applied to a magnetized separation column and
washed 4x with buffer. Magnetically labeled cells are retained in the column while
unlabeled cells pass through. The column was removed from the magnet and labeled
cells were eluted with buffer, sedimented by centrifugation and resuspended in 10 L
of elution buffer. We visualized and quantified cells using a fluorescent microscope
and cell viability was determined based on a Trypan blue exclusion. All procedures
were performed at room temperature.
Results: The average percent recovery from 1 mL of blood using anti-EpCAMconjugated beads was 70.95% 5.27 (mean SEM), interassay CV of 40.68%
(N=30). The average percent recovery from 1 mL of blood using a combination of
anti-EpCAM and anti-HER2 beads improved to 81.32% 3.959 and interassay CV of
24.82% (N=26), p = 0.1308. The data was analyzed using an unpaired t-test. For each
method of cell recovery, 100% of CTCs remained viable.
Conclusion: Based on these data, we concluded that this technology can be adapted
for the purpose of enriching low abundance cells directly from whole blood. Use

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

of a multiplex strategy has the potential to achieve recovery efficiency consistent

with the current state of the art, and improves % recovery over a single antigen
approach. This method consistently yielded recovery of viable cells which will make
this approach uniquely useful for single cell phenotyping studies such as for testing
chemotherapeutic sensitivity.

A-042
Decoding miRNA Expression of Breast Carcinoma Behavior using Next
Generation Sequencing of LCM Procured Cells

J. L. Wittliff1, K. Bramlett2, S. A. Andres1, J. Schageman2, C. Hinahon2,

J. Cienfuegos2, R. Setterquist2. 1University of Louisville, Louisville, KY,
2
Thermo Fisher Scientific, Austin, TX
Refinements in selecting biomarkers for breast carcinoma management require
identification of clinically relevant parameters complementing patient endocrine
status and tissue biopsy content of estrogen receptors (ER), progestin receptors (PR)
and HER-2/neu oncoprotein, which correlate with prognosis and therapy response.
Our objective is to compare miRNA expression profiles of intact tissue sections
from breast cancers with those of laser capture microdissection (LCM)-procured
carcinoma cells. The hypothesis is that miRNA signatures, discerned from LCM
acquired populations of specific cell types, more accurately reflect the molecular
basis of cancer clinical behavior than provided by protein biomarkers of intact
tissue biopsies. miRNAs are 19-24 nucleotide non-coding RNA species regulating
gene expression by inhibiting translation or by triggering degradation of specific
mRNA targets. De-identified frozen tissue biopsies were selected from our IRBapproved Biorepository using criteria in the comprehensive de-identified Database to
standardize the study population (e.g., invasive ductal carcinomas of known grade and
biomarker status). Serial tissue sections containing 55 +/- 23% cancer were prepared
and stained with H & E using established protocols, and carcinoma cells (~ 14000
LCM pulses) were procured non-destructively from an adjacent section. RNA was
extracted from intact tissue sections and LCM-isolated cells using PureLink RNA
Mini KitsTM (Invitrogen), evaluated for integrity (Agilent Bioanalyzer) and analyzed
for miRNA expression using the Ion TorrentTM Next Generation Sequencing System
(Thermo Fisher Scientific). Total RNAs were enriched for small RNA species using
mirVana miRNA isolation kits (Thermo Fisher Scientific) and RNA libraries were
constructed from 5 ng of enriched RNA using Ion Total RNA-Seq LibraryTM kits.
Barcodes were utilized to multiplex libraries for template preparation and sequencing
on two Proton PI chips as two twelve-plex library pools. Resulting sequences were
aligned to mirBASE precursors, and expression levels were calculated by tallying
the number of reads mapping to each individual miRNA precursor. Counts/1M reads
values for each miRNA were normalized to a housekeeping gene to determine relative
miRNA expression. Reads mapping to mirBase were assessed for each carcinoma
preparation. Comparison of the top 20 expressed miRNAs in the intact tissue sections
with those of cognate carcinoma cells procured by LCM, in general, revealed that
smaller defined miRNA gene sets were expressed in isolated populations of carcinoma
cells. Furthermore, miRNA expression patterns of experimental pairs (intact section
vs LCM-procured cells) were highly variable in carcinomas with different grades,
suggesting relationships to disease status. Strikingly, when miRNA gene frequency
plots (transcript abundance vs fold-change) were developed, comparing expression
from intact tissue sections to that of LCM-procured cell population, subsets of miRNA
genes were revealed. Although the limited number of samples analyzed precluded
identification of particular miRNA gene signatures associated with a specific breast
pathology or biomarker status (e.g., ER, PR, HER-2/neu), application of Next
Generation Sequencing of miRNAs using LCM-procured carcinoma cells provides
an innovative approach for decoding miRNAs involved in breast cancer behavior.
Supported in part by a grant from the Phi Beta Psi Charity Trust (JLW & SAA) and a
CTSP Award from the Commonwealth of Kentucky (JLW).

A-044
Novel Approach to Diagnose High Grade of Cervical Lesion: Combination of
HPV E6/E7 and hTERT mRNA Real-Time RT-PCR Assay

S. Park1, H. Wang2, S. Kim3, G. Kim1, D. Lee4, Y. Kim1, H. Kim1, J. Kim1,

S. Ahn1, H. Jin5, K. Park6, H. Lee1. 1Department of Biomedical Laboratory
Science, College of Health Sciences, Yonsei University, Wonju, Korea,
Republic of, 2M&D, Inc., Wonju Eco Environmental Technology Center,
Wonju, Korea, Republic of, 3Institute for Life Science and Biotechnology,
Yonsei University, Seoul, Korea, Republic of, 4Department of Clinical
Laboratory Science, Hyejeon College, Hongseoung, Korea, Republic of,
5
Department of Clinical Laboratory Science, College of Health Sciences,
Catholic University of Pusan, Busan, Korea, Republic of, 6Department of
Pathology, Yonsei University Wonju College of Medicine, Wonju, Korea,
Republic of
Background: Human Papillomavirus (HPV) is a major causative factor of cervical
cancer, which is the third of the most common cancer in women. The Real-HPV-E6/
E7 mRNA multiplex RT-qPCR assay (M&D, Wonju, Republic of Korea) has been
developed and evaluated because E6 oncoprotein inhibits apoptosis by degradation of
cellular tumor suppressor protein p53 and E7 oncoprotein prevents cell cycle arrest
of damaged cells. HPV high-risk types (HPV genotype 16, 18, 31, 33, 35, 39, 45, 51,
52, 53, 56, 58, 59, 66, 68 and 69) are regard to be detectable and significant marker
in high grades of cervical lesion. Human telomerase reverse transcriptase (hTERT) is
also considered as complementary marker that may provide a criteria in high grades
of cervical lesions. Aberrant telomerase activity has been suggested to be critical for
human tumor genesis and related to following mechanism of E6 oncoprotein.
Methods: In this study, HPV E6/E7 oncogene and hTERT are detected by RealHPV-E6/E7 mRNA multiplex RT-qPCR and Real-hTERT mRNA RT-qPCR assay
(M&D), respectively. A total of 545 patients including 18 squamous cell carcinoma
(SCC), 21 high grade squamous intraepithelial lesion (HSIL), 17 atypical squamous
cells-cannot exclude HSIL (ASC-H), 101 low grade squamous intraepithelial lesion
(LSIL), 100 atypical squamous cells of undetermined significance (ASC-US), and
288 normal cytology samples, were enrolled and analyzed by cytological diagnostic
grades, respectively. 39 samples in high grades of cytological diagnosis ( HSIL)
were confirmed as high grades in histological diagnosis.
Results: The positive rates of HPV E6/E7 mRNA RT-qPCR assay were 94.4%
(17/18), 95.2% (20/21), 82.4% (14/17), 46.5% (47/101), 25.0% (25/100), and 1.1%
(3/286) in SCC, HSIL, ASC-H, ASC-US, LSIL, and normal samples, respectively.
Relative hTERT mRNA expression levels were able to distinguish high grade and low
grade of cervical lesion significantly (p <0.001). Relative hTERT mRNA expression
levels in low grades of cervical lesion were dramatically lower than in high grade of
cervical lesion. Notably, 5 high grades of cervical samples ( HSIL) were not detected
by HPV E6/E7 mRNA real-time RT-PCR assay, but those samples were high relative
expression levels with hTERT mRNA real-time RT-PCR assay.
Conclusion: For predicting the outcomes of cervical intraepithelial neoplasia (CIN)
1 or CIN 2 patients, the combined use of HPV E6/E7 and hTERT mRNA RT-qPCR
assay could be a significantly complementary approach for diagnosing high grade
cervical lesions because of the hTERT mRNA expression levels highly increased in
cervical cancer and which was very low in low grade cervical lesions and normal
tissues. Therefore the combined detection of HPV and human factors as a predictive
marker might be a very useful for monitoring of patients who have low grade of
cervical lesions.

A-045
SENTIFIT-FOB Gold latex Fecal Immunoassay Test (FIT) evaluation on
SENTiFIT270 analyzer in CoreLab at the AUSL Modena-Nuovo S.Agostino
Estense hospital in Emilia Romagna Region.

M. C. Anelli1, A. R. Soliera2, N. Conti1, A. Cugini1, T. Trenti2, F. Torricelli2,

R. Corradini2, M. Gramegna1. 1Sentinel CH SpA, Milano, Italy, 2CoreLab,
Clinical Pathology Department, NOCSAE-AUSL Modena, Baggiovara,
Italy
Background: the fecal immunochemical test (FIT) for hemoglobin is considered to
be superior guaiac fecal occult blood test for colorectal cancer (CRC) screening and is
becoming central in international CRC screening programs development. Identify the
most appropriate FIT is now a priority. The focus of this study is two-fold: to assess
the FOBGold Latex reagent FIT on a dedicated instrument with pierceable device; to
compare two quantitative FITs.
Methodology: analytical evaluation: Limit of Blank (LoB) and of Quantitation (LoQ),

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S11

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

(CLSI EP17-A); total, between days and runs imprecision (EP5-A2) on 3 quality
control levels (QC Low, 1, 2) and one hemoglobin spiked pool in buffer (target 50,
77, 309 and around 150 ng/mL respectively); linearity (EP6-A), prozone, on-board
reagent and calibration stability (EP25-A).
Methods: 120 selected samples from frozen anonymous routine (not from CRC
screening program population) residuals, with high positive prevalence, sampled with
Eiken and Sentinel devices (OC-Auto Sampling Bottle3 and SENTiFITpierceTube)
and run with OC-Sensor Diana and SENTiFIT270 respectively; statistical analysis:
concordance table. Analyzer evaluation: one week familiarization; piercing, sample
barcode reading, timing evaluation. Training and rating questionnaire (from 1 to 5
where 1=very poor, 3=neither good nor poor, 5=very good) to 18 technicians/degrees.
Results: LoB 4.1 ng/mL; LoQ 15.4 ng/mL. QC Low total imprecision: CV=6.2%,
SD=3.0; between days CV=3.4%, SD=1.7; between runs CV=0.8%, SD=0.4. QC 1
total imprecision: CV=4.0%, SD=3.4; between days CV=0.2%, SD=0.2; between
runs CV=1.5%, SD=1.3. QC 2 total imprecision: CV=2.1%, SD=7.5; between
days CV=0.9%, SD=3.1; between runs CV:0.8%, SD=2.8. Pool total imprecision:
CV=2.7%, SD=4.2; between days CV=1.2%, SD=1.8; between runs CV=0.4%,
SD=0.7. Accuracy: Recovery QC Low 98%, QC1 111.7%, QC2 113.7%, Pool 102.7%.
Linearity up to 871 ng/mL; prozone checked-up to 50,000 ng/mL; on-board reagent
and calibration stability 33 days. Method comparison: Hamza et al. published a cut-off
value of 117 ng/mL for FOBGold corresponding to 100 ng/mL for OC-Sensor. On this
basis we evaluate the concordance: 99 results were negative and 16 positive with both
methods; 5 results were positive with FOBGold only. Positive rate in the evaluated
samples: OC-Sensor=13.3% and FOB Gold=17.5%. FOB Gold had higher positive
rate than OC-Sensor (24%). SENTiFIT270: Familiarization was done successfully
and the instrument is user friendly and reliable. Piercing and barcode reading on 200
tubes were always correct; time to first result: 14 minutes. Questionnaire rating mean
values for software, sample management, calibration, quality control and maintenance
are 4.1/4.2/4.0/4.1/4.8 respectively.
Conclusion: FOBGold Latex reagent FIT on SENTiFIT270 analyzer shows some
important features: very low LoB, a Low QC at 50 ng/mL, 1250 on-bard test autonomy
and automated maintenance; on-board samples capability should be improved. A
regression study is not appropriate, due to sampling bias (different devices, different
buffers) and different signal origin (different calibration materials and wavelenghts
between OC-Sensor and SENTiFIT270) therefore a concordance table is the only
possible statistical analysis. The discrepant samples cannot be investigated in this
study with anonymous samples selected from routine with high positive prevalence,
as it is clear that a final clinical validation needs colonoscopy check.

A-046
Prostate Specific mRNAs as Potential Specific Markers of Circulating Tumor
Cells and for Detection of Prostate Cancer

W. Zhang, Z. Wang, E. Klein, M. K. Gupta. Cleveland Clinic Foundation,

Cleveland, OH
Background: Currently, prostate cancer (PCa) screening relies on prostate specific
antigen for early-detection of PCa. However, this test suffers with poor specificity
with high rate of negative biopsy stressing the necessity of more specific biomarkers.
Prostate-specific RNA markers can be used to detect circulating cancer cells in
peripheral blood to provide a sensitive and specific alternative to detect PCa. We
performed a pilot study to explore the role of a combined RNA markers including
prostate specific membrane antigen (PSMA), prostate cancer antigen 3 (PCA3) and
TMPRSS2-ERG fusion gene (TM-ERG) in screening for PCa.
Methods: Total RNA was extracted from peripheral blood mononuclear cells
(PBMC) and was analyzed for PSMA, PCA3 and TM-ERG transcripts using
quantitative reverse-transcriptase polymerase chain reaction (qRT-PCR). The assays
were calibrated using either purified RNA from PCa cell line (VCap, CRL-2876) or
synthesized complemented RNA. Analytic sensitivity was tested on RNA extracted
from spiked female blood with VCap. Total of 33 patients with PCa before treatment,
and 19 age-matched male patients with no PCa were analyzed for all three markers.
Results: Based on experiments using VCap spiked female blood PSMA and TM-ERG
mRNA were reliably detected at 1-10 cell/ml respectively. The median (95percentile)
mRNA levels in patients with PCa were significantly higher for all three markers than
benign controls (p 0.001). Furthermore, PCa patients with surgical margin(+) had
significantly higher levels than margin(-) patients. Combination of all three markers
significantly increased the clinical sensitivity and specificity for PCa as evidenced by
logistic regression analysis. Results are summarized in Table 1.
Conclusion: Measurement of PSMA, PCA3 and TM-ERG transcripts extracted
from PBMC increases the detection rate of PCa, especially invasive PCa with high
specificity. Further studies with large cohort are needed for clinical validation and to
assess pathologic stage prior to radical prostatectomy.

S12

Table 1: Quantitation of mRNA for PSMA, PCA3,TM-ERG in Blood

TM-ERG
LRc
PSMA
PCA3a

Benign
Median (95% CI)
PCa
Median (95% CI)
p-valueb
Margin(+)(N=13)
Margin(-)(N=20)
p-value
ROC Analysis
AUC (%)
p-value d
Sensitivity (%)
Specificity (%)
a
PSMA, TM-ERG:
ng/PCR, PCA3:
copies/PCR

0.8 (0.2-1.5) 256 (128-1020)

9.7 (7.0-12)

NA

2.8 (1.6-9.2)
0.0001
17 (1.3-24)
2.2 (1.5-4.0)
0.021

3049 (1947-4988)
<0.0001
4355 (2839-14631)
2016 (832-5003)
0.019

51 (15-81)
0.0001
81 (61-111)
17 (8.0-52)
0.003

NA
NA
NA
NA
NA

83
0.004
76
79

88
0.090
76
90

78
0.001
68
100
d
LR compare to
b
c
MannLogistic regression
individual marker
Whitney test analysis
(McNemar)

96
NA
85
95

A-047
Utility of Stringent Complete Response in routine treatment of Multiple
Myeloma patients with novel agents

J. L. Garca de Veas Silva, C. Bermudo Guitarte, R. Duro Milln, V.

Snchez Margalet, C. Gonzlez Rodriguez. Hospital Universitario Virgen
Macarena, Sevilla, Spain
Background: Normalization of serum free light chains (sFLC) ratio in patients with
Multiple Myeloma (MM) achieving complete response (CR) may define a deeper
degree of response after therapy than that defined by the CR criteria. The stringent
CR (sCR) requires normalization of sFLC ratio and absence of clonal plasma cells in
bone marrow in addition to the criteria for CR (Negative immunofixation of serum and
urine, disappearance of any soft tissue plasmacytomas and <5% plasma cells in bone
marrow). The aim of this study is to evaluate the prognostic utility of sCR in patients
newly diagnosed with MM treated with novel agents in the routine practice.
Methods: Twenty three patients with MM (10 IgG MM, 5 IgA MM, 2 IgD MM and
6 Bence Jones MM) achieving CR after therapy with Bortezomib/Dexametasone
were included in this study. Disease Free Survival (DFS or time after treatment where
disease remains stable) was estimated by Kaplan-Meier method and compared by logrank tests. Cox proportional hazard analysis was performed for multivariate analysis.
Serum free light chains were measured by turbidimetry (Freelite) in a SPA PLUS
analyzer (The Binding Site Group Ltd, Birmingham, UK) and immunofixation was
performed in a HYDRASYS (Sebia, FR) analyzer.
Results: The median follow-up of the patients was 18 months (range 14-31 months).
Eleven patients achieved CR and 12 patients achieved sCR. During the period of study
there were 8 relapses, six in patients achieving CR and two in patients achieving sCR.
The median DFS for patients achieving CR was 18 months and not reached for those
achieving sCR. Patients achieving CR had a DFS rate of 24% compared with 75% for
sCR (p=0.022). Results showed that achieving a sCR was an independent prognostic
factor for survival (HR = 6.57; 95% CI, 1.09-39.80; vs CR; p = 0.039).
Conclusion: The presence of an altered sFLC ratio suggests the existence of a persistent
clonal population that is secreting small amounts of monoclonal protein. Our results
indicate that sCR represents a deeper response state compared with conventional CR
which translates into a longer DFS. Despite the small cohort, analysis of sFLC ratio
was able to identify a group of patients with more favorable prognosis and support its
inclusion in the response criteria for MM patients treated with novel agents.

A-048
Determination of ROMA Score Performance Using the Roche Elecsys HE4 and
CA 125 Immunoassays

M. A. Lasho, A. Algeciras-Schimnich. Mayo Clinic, Rochester, MN

Background: Ovarian cancer is the fifth-leading cause of death in women. Ovarian
cancer symptoms are related to the presence of an adnexal mass and are often vague
and unspecific. Treatment of women presenting with an adnexal mass and at high
risk for ovarian cancer by specialized gynecologic oncologists has been shown to
improve patient outcomes. The risk of ovarian malignancy algorithm (ROMA) is a
calculation that incorporates the patients serum concentrations of cancer antigen 125
(CA125) and human epididymis protein 4 (HE4) in conjunction with the menopausal
status to calculate a predictive probability of finding epithelial ovarian cancer on
surgery in women presenting with an adnexal mass. Women are classified as highrisk or low-risk for ovarian cancer. ROMA has only been validated using HE4 by
enzyme immunoassay (EIA) in conjunction with Abbott ARCHITECT CA125 assay

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

or CA125 by EIA. This study evaluates the use of the Roche Elecsys HE4 and CA125
electrochemiluminescence immunoassay (ECLIA) for the calculation of the ROMA
score.
Methods: Serum samples from 114 premenopausal females (75 benign gynecological
conditions, 39 epithelial ovarian cancer [EOC]) and 93 postmenopausal females
(56 benign gynecological conditions, 37 EOC) were included in the study. Benign
gynecological conditions included cysts, cystadenomas, leiomyomas, myomas, or
fibromas. Epithelial ovarian cancers included stages I through IV (stage I N=19, II
N= 3, III N=45, IV N= 9). HE4 and CA125 were measured using the Roche Elecsys
ECLIA on a Roche Cobas e601 instrument. Serum HE4 and CA125 concentrations
and menopausal status were used to calculate the ROMA score. Equations used to
calculate the ROMA score were as follow: Premenopausal Predictive Index (PI)
= -12.0 + 2.38*LN[HE4] + 0.0626*LN[CA125]; Postmenopausal PI = -8.09 +
1.04*LN[HE4] + 0.732*LN[CA125]; and ROMA score = exp(PI) / [1 + exp(PI)] *
10. Receiver Operating Characteristic (ROC) curve analysis was used to determine
optimal clinical cut-points and clinical specificity and sensitivity.
Results: In premenopausal women, the ROMA ROC curve area under the curve
(AUC) was 0.95. A ROMA score equal or greater than 1.00 yielded 75% specificity
and 95% sensitivity. Using the manufacturer suggested cut-point of equal or greater
than 1.14 yielded 84% specificity and 95% sensitivity. In postmenopausal women, the
ROMA ROC AUC was 0.94. A ROMA score equal or greater than 2.44 yielded 75%
specificity and 95% sensitivity. Using the manufacturer suggested cut-point of equal
or greater than 2.99 yielded 86% specificity and 92% sensitivity.
Conclusion: This study established the performance of ROMA score cut-points using
the Roche Elecsys HE4 and CA125 immunoassays. This information could serve as
guidance for laboratories implementing the ROMA score in clinical practice.

A-049
Comparing the Performance of Newly Developed Heavy Chain/ Light Chain
Immunoassays with Serum Protein Electrophoresis and Nephelometric
Measurements of Total Immunoglobulin for Monitoring Multiple Myeloma
Patients

L. Adie, O. Berlanga, H. Carr-Smith, S. Harding. The Binding Site Group

Ltd, Birmingham, United Kingdom
Background:Both serum protein electrophoresis (SPEP) and total immunoglobulin
(tIg) measurements have been recommended for quantification of monoclonal Ig
(M-Ig). However, SPEP is inaccurate at low (<10 g/L) and due to dye saturations
at high (>20-30 g/L) concentrations of M-Ig. By contrast tIg measurement is an
accurate method but is unable to distinguish between monoclonal and polyclonal
Ig. Newly developed Heavy chain/ light chain immunoassays may provide an
alternative method of quantifying M-Ig concentrations. Here, we compare the
performance of these assays with traditional methods for monitoring MM patients.
Methods: HLC Ig and Ig were quantified in 127 IgG (87 IgG, 40
IgG) and 61 IgA (37 IgA, 24 IgA) MM patient sera. The results were
compared to published normal ranges (IgG: 4.03-9.78 g/L, IgG: 1.97-5.71
g/L, IgG/ IgG: 0.98-2.75; IgA: 0.48-2.82 g/L, IgA: 0.36-1.98 g/L, IgA/
IgA: 0.80-2.04), historic SPEP, immunofixation and tIg concentrations.
Weighted Kappa and Pearson correlation were used to analyse results.
Results: At presentation all 127 IgG and 61 IgA patients had an abnormal HLC ratio
and involved HLC (iHLC) concentrations (median (range) IgG: ratio 56 (6- 1275),
iHLC 32 g/L (14-102); IgG: ratio 0.024 (0.001- 0.329), iHLC 34 g/L (9- 90); IgA:
ratio 233 (10- 6226), iHLC 34 g/L (6- 79); IgA: ratio 0.0119 (0.0003- 0.1181), iHLC
29 g/L (6- 72)). Whilst M-Ig concentrations were measurable by SPEP in all IgG
patients, only 66% (40/61) IgA patients were quantifiable. In all samples, iHLC and
dHLC (involved HLC-uninvolved HLC) concentrations showed a good correlation
with SPEP for IgG (iHLC y=0.83x+1.8, R2=0.87; dHLC y=0.84x+0.48, R2=0.88),
IgA (iHLC y=0.88x+0.88, R2=0.87; dHLC y=0.89x+0.44, R2=0.88) and with tIgA
measurement in IgA patients (iHLC y=0.87x-0.68, R2=0.90; dHLC y=0.88x-1.33;
R2=0.90). During the course of the patients disease, changes in iHLC and dHLC
concentrations reflected the changes in M-Ig measured by SPEP (IgG: iHLC
y=0.87x-0.05, R2=0.83; dHLC y=0.91x-0.06, R2=0.86; IgA: iHLC y=1.31x+0.25,
R2=0.87; dHLC y=1.36x+0.28, R2=0.88) and with tIgA changes in IgA patients
(iHLC y=0.94x-0.03, R2=0.90; dHLC y=0.96x-0.03, R2=0.90). Responses assigned
based on reductions in M-Ig measured by either iHLC, dHLC or traditional methods
showed substantial agreement for IgG and near perfect agreement for IgA patients
using Weighted Kappa analysis (IgG: iHLC vs. SPEP 81% agreement, Weighted
Kappa (95% CI): 0.78 (0.56-1.00); dHLC vs. SPEP 80% agreement, Weighted Kappa
(95%CI): 0.77 (0.55-1.00); IgA: iHLC vs. SPEP/tIgA 89% agreement, Weighted
Kappa (95% CI): 0.92 (0.84-1.00); dHLC vs. SPEP/tIgA 89% agreement, Weighted
Kappa 0.92 (0.84-1.00)). Changes in HLC ratio similarly showed a good comparison

to the assigned responses (IgG: 71% agreement, Weighted Kappa (95% CI): 0.74
(0.56-0.92); IgA: 73% agreement, Weighted Kappa (95% CI): 0.86 (0.81-0.91)).
Conclusion: Responses assigned using reductions in iHLC, dHLC, HLCr or SPEP
showed a good agreement. Furthermore, iHLC, dHLC and HLC ratio were able to
assign responses in 34% of IgA patients that were not quantifiable by SPEP. The HLC
immunoassays provide an alternative method of quantifying M-Ig in patients with
MM.

A-050
Serum MicroRNA Panel as Biomarkers for Early Diagnosis of Colorectal
Adenocarcinoma

C. Wang, G. Zheng, L. Du, L. Wang, X. Zhang, Y. Yang, J. Li, Y. Wang, Y.

Liang. Qilu Hospital, Shandong University, Jinan, China
Background: Due to the high mortality of colorectal adenocarcinoma (CAC), there is
an urgent need to identify new biomarkers with high sensitivity and specificity. The
recent discovery of serum microRNA (miRNA) profile in human cancer has provided
a new auxiliary approach for tumor diagnosis. Our study is the first global analysis of
serum miRNAs based on the normal-colorectal adenoma (CA)-CAC sequence.
Methods: Serum samples were collected from 307 CAC patients, 164 CA patients and
226 healthy controls. We firstly profiled pooled serum of CAC, CA and healthy controls
by Miseq sequencing. The differentially expressed serum miRNAs were chosen as
candidate biomarkers for CAC. Both the candidate reference genes and candidate
biomarkers were validated by the reverse-transcription polymerase chain reaction
(RT-qPCR). The miRNA panel was developed with a logistic regression model and
then validated using an independent cohort. Receiver operating characteristic (ROC)
curves were constructed, and area under the ROC curve (AUC) was used to evaluate
the diagnostic accuracy of the panel.
Results: The Miseq sequencing results revealed 15 differentially expressed miRNAs
in CAC patients compared with controls. Using the selected reference gene of miR191-5p and U6, we identified a 4-miRNA panel (miR-19a-3p, miR-223-3p, miR92a-3p and miR-422a) with a high diagnostic accuracy of CAC. Even in the low
carcinoembryonic antigen (CEA) level group, the diagnostic accuracy of this miRNA
panel was still acceptable (AUC = 0.810). Surprisingly, our results indicated that the
miRNA panel could differentiate stage I/II CAC patients from controls. In addition,
this panel could also differentiate CAC from CA (AUC=0.886).
Conclusions: In the present study, we established a serum miRNA panel with
considerable clinical value in the early-stage diagnosis of CAC.

A-051
Total, free, and complexed prostate-specific antigen concentrations among U.S.
men, 2007-2010

D. A. Lacher, J. P. Hughes. National Center for Health Statistics,

Hyattsville, MD
Background: Screening for prostate cancer using prostate-specific antigen (PSA) is
common but remains controversial. Prostate cancer has been associated with higher
total PSA (tPSA), lower free PSA (fPSA), and lower percent free PSA (fPSA/tPSA
x 100%). More recently, higher complexed PSA (cPSA), bound primarilywith -1antichymotrpsin, has been associated with prostate cancer. The distributions of total,
free and complexed PSA concentrations, percent free PSA and percent complexed
PSA (cPSA/tPSA x 100%), and the free/complexed PSA ratio in men were examined
in the 2007-2010 National Health and Nutrition Examination
Survey (NHANES).
Methods: Total, free and complexed PSA were performed on 3251 men aged 40
years and older who were examined in the 2007-2010 National Health and Nutrition
Examination Survey. NHANES is a cross-sectional, nationally representative, area
probability survey of U.S. non-institutionalized participants. Distributions of the PSA
tests were examined by age, race and ethnicity, and body mass index (BMI) groups. In
addition, percentages of men at total and percent free PSA cut-points were examined.
All PSA tests were log-normal in distribution except percent complexed PSA, which
was normally distributed. The geometric mean (GM) or arithmetic mean (for percent
complexed PSA), standard error (SE) of the mean, and selected percentiles were
determined. Age-adjusted means were used for analysis of PSA tests for race and
ethnicity and BMI groups.
Results: The geometric mean (SE) for tPSA was 0.96 (0.02) g/L with 5.2% of men
4.0 g/L and 1.1% 10.0 g/L. Free PSA had a GM of 0.27 (0.01) g/L. The
GM of percent fPSA was 28.1 (0.3) % and 8.6% of men had 15% percent fPSA.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S13

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

Complexed PSA had a GM of 0.53 (0.01) g/L. The arithmetic mean of percent cPSA
was 56.0 (0.5) % and the free/complexed PSA ratio GM was 0.52 (0.01). Total, free,
and complexed PSA increased with age. Total PSA GM increased from 0.74 g/L for
men 40-49 years to 1.82 g/L for men 80 years and older. Free PSA GM increased
from 0.22 g/L for men 40-49 years to 0.51 g/L for men 80 years and older, while
complexed PSA increased from 0.40 g/L for men 40-49 years to 0.99 g/L for men
80 years and older. The adjusted mean for non-Hispanic white men had lower tPSA
(1.03 g/L) and cPSA (0.56 g/L) than non-Hispanic black men (tPSA 1.25 g/L
and cPSA 0.72 g/L). Hispanic men had higher cPSA (0.64 g/L) than non-Hispanic
white men. Obese men had lower age-adjusted mean total, free and complexed PSA
(0.94, 0.27, and 0.51 g/L, respectively) than men with normal BMI (tPSA 1.21, fPSA
0.32, and cPSA 0.68 g/L).
Conclusion: The free and complexed PSA may provide additional information in
conjunction with total PSA in screening for prostate cancer. Total, free and complexed
PSA increased with age; total and complexed PSA were highest in non-Hispanic black
men; and obese men had the lowest total, free, and complexed PSA.

A-052
Leptin and insulin hormones increase Sam68 expression and phosphorylation
in human breast adenocarcinoma cells

F. Snchez-Jimnez, A. Prez-Prez, C. Gonzlez-Rodrguez, J.

A. Virizuela, V. SANCHEZ-MARGALET. VIRGEN MACARENA
UNIVERSITY HOSPITAL, SEVILLE, Spain
Background: Obesity and insulin resistance are well known risk factors for breast
cancer development in postmenopausal women. High insulin levels, together with
other hormones, such as leptin and cytokines, IGFs, estrogen and EGF, positively
modulate the growth of these tumor cells. All these factors may act through signaling
cascades that lead to the final effect of increasing growth and cell proliferation. Sam68
protein is a member of the signal transduction activator of RNA (STAR) family of
RNA-binding proteins that can interact both with RNA and signaling proteins.
According to this dual role, Sam68 has been involved in different carcinogenic
mechanisms including alternative splicing or cell cycle regulation. Moreover, our
group has previously described the role of Sam68 in the insulin and leptin signaling
pathways as a receptor substrate, and it has been shown to participate in proliferation,
cellular growth and antiapoptotic effects mediated by these hormones in different
cellular types.
Objective: We aim to study the expression of Sam68 and its phosphorylation level
upon insulin and leptin stimulation, seeking for a possible role of Sam68 in leptin and
insulin receptor signaling in human breast adenocarcinoma cells.
Methods: We used the human breast adenocarcinoma cell line MCF7. We
studied leptin-mediated and also insulin-mediated Sam68 phosphorylation by
immunoprecipitation and immunoblot with anti-phosphotyrosine antibodies as well as
polyU affinity precipitation. Quantitative RT-PCR and immunoblot were used to study
the effect of leptin and insulin on Sam68 expression. siRNA was used to downregulate
Sam68 expression and its effects on leptin and insulin activation of MAPK and PI3K
pathways. Phosphorylation of some of the main proteins of these pathways (ERK1/2
and MEK as well as PKB and P70s6K respectively) was tested by using immunoblot
with antibodies against phosphorylated proteins and anti-tubulin as loading control.
Results: Sam68 protein quantity and gene expression were found to be increased under
leptin as well as insulin stimulation, by using 1 nM dose after a 24 hours stimulus.
Moreover, both insulin and leptin stimulation promoted an increase in Sam68 tyrosine
phosphorylation in MCF7 cells and negatively regulated RNA binding of Sam68,
as previously observed in other systems. Sam68 downregulation resulted in lower
activation of MAPK and PI3K pathways under both hormones stimulation. Sam68
was necessary for the complete leptin and insulin phosphorylation of the main
proteins of these pathways.
Conclusion: These results suggest the participation of Sam68 in both leptin and
insulin receptor signaling in human breast cancer cells, where Sam68 could mediate
the trophic effects of these hormones in proliferation and cellular growth. Thus,
Sam68 could be also considered as a future prognostic marker or therapeutic target in
this kind of non genetic breast cancer.

A-053
MiR-28-5p, a potential biomarker for renal cell carcinoma, acts as a tumor
suppressor in renal cell carcinoma for multiple antitumor effects by targeting
RAP1B

C. Zhang1, C. Wang1, C. Wang1, C. Wu1, Q. Yang1, C. Zhang2. 1Jinling

Hispital, Nanjing, China, 2Nanjing Univertisy, Nanjing, China
Background: The mechanisms involved in renal cell carcinoma (RCC) development
and progression remain unclear, and new biomarkers are needed in routine practice
to improve the diagnostic and/or prognostic accuracy. However, there is no standard
serum biomarker to facilitate diagnosis or prognostic stratification in patients with
RCC. There is increasing evidence that microRNAs (miRNAs) are involved in
cancer development and progression and circulating miRNAs have great potential as
biomarkers for diagnosis and prognosis in patients with several types of cancers. Our
purpose was to investigate whether serum miR-28-5p could be a useful biomarker for
the diagnosis of RCC and evaluate the functional significance of the miR-28-5p in
RCC. Methods: This study included 33 RCC patients and 33 healthy controls. First,
we analyzed tissue miR-28-5p levels in tumor tissues and matched normal tissues
from the 33 RCC patients. Second, we investigated the serum miR-28-5p levels in
the 33 RCC patients and the 33 normal controls. TaqMan probe based-RT-qPCR was
used to measure serum miRNA levels. A combination of let-7d, let-7g and let-7i (let7d/g/i) was used as endogenous control for normalizing the data of RT-qPCR. We also
examined the expression level of miR-28-5p in some human RCC cell lines. The CCK8
proliferation, transwell and wound healing assays were used to explore the potential
functions of miR-28-5p in RCC cells. Luciferase reporter assays were employed to
validate regulation of a putative target of miR-28-5p. The effect of modulating miR28-5p on endogenous levels of this target were subsequently confirmed via Western
blotting. Results: MiR-28-5p expression was relatively decreased in RCC specimens
compared with adjacent normal tissues (P<0.01). Consistent with the results from
tissues, serum miR-28-5p levels were decreased in RCC patients compared to controls
(P<0.001). ROC curve analysis showed an AUC of 0.90 (95% confidence interval,
0.85-0.95) and a sensitivity and specificity of 95 and 86%, respectively. MiR-28-5p
was found to be also downregulated in human RCC cell lines A498 and Caki-2 as
compared with normal cell line HK-2. Luciferase reporter assays showed that miR-285p directly regulated RAP1B. In RCC clinical specimens, the expression of RAP1B
protein was significantly higher in cancer tissues than in non-cancerous tissues.
Statistical analysis results indicated that the RAP1B protein level was negatively
correlated to the miR-28-5p expression in RCC tissues (P<0.05). RAP1B protein
was found to be upregulated in A498 and Caki-2 cells, and knockdown of RAP1B
inhibited cell proliferation and migration, suggesting that RAP1B has oncogenic
functions in RCC. Ectopic expression of miR-28-5p could result in increased RAP1B
protein expressions, and inhibited proliferation and invasion of A498 and Caki-2 cells,
while the downregulation of miR-28-5p with the inhibitor had the opposite effect.
The miR-28-5p induced cell proliferation and migration could be rescued by RAP1B.
Conclusion: MiR-28-5p may potentially serve as a novel biomarker for RCC and may
act as a tumor suppressor in RCC progression by inhibiting the RCC cell proliferation
and migration through targeting oncogeneRAP1B. Our findings indicate that targeting
miR-28-5p by a genetic approach may provide a novel strategy for the treatment of
RCC.

A-054
Serum thyroglobulin measurement in autoantibody positive samples by LC-MS/
MS and immunoassay: is positivity rate different between the methods?

M. M. Kushnir1, A. L. Rockwood2, A. W. Meikle3. 1ARUP Institute for

Clinical and Experimental Pathology, Salt Lake City, UT, 2Department
of Pathology, University of Utah, Salt Lake City, UT, 3Departments of
Pathology and Medicine, University of Utah, Salt Lake City, UT
Background: Measurement of thyroglobulin (Tg) in serum and plasma is used to
monitor patients after treatment for differentiated thyroid carcinoma (DTC). A
complicating factor in using Tg as biomarker of the recurrence of DTC is related to the
presence of endogenous anti-Tg autoantibodies (Tg-AAb) in blood of many patients.
Tg-AAb can interfere with immunoassay (IA) measurements and cause false-negative
results. LC-MS/MS methods for Tg are expected to overcome Tg-AAb interference
with measurement of Tg, but there were no studies supporting better utility of the LCMS/MS methods for Tg in Tg-AAb positive samples.
Methods: Recently we developed LC-MS/MS method for measurement of Tg in
serum samples (Clin Chem 2013;59:982-90). The lower limit of quantification is
0.5ng/mL; total imprecision is below 10%. We performed comparison of LC-MS/
MS and Beckman IA using Tg-AAb negative and positive samples; and reviewed

S14

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

historical data on analysis of Tg in Tg-AAb positive samples using Beckman IA

(n=1367) and LC-MS/MS (n=6180) methods, by comparing Tg positivity rates for
the assays. Tg calibrators were standardized between the methods. Measurement of
Tg-AAb was performed by Beckman Antibody-II assay.
Results: In a set of Tg-AAb negative samples Tg concentrations determined with
Beckman IA agreed well with LC-MS/MS (IA=1.00*LC-MS/MS-2.35, r=0.982,
Sy,x=9.52); IA underestimated Tg concentrations in Tg-AAb positive samples. In
a set of Tg-AAb positive samples tested negative for Tg using IA, concentrations
determined by LC-MS/MS method were at or above 0.5ng/mL in 23% of samples.
Positivity rate for different Tg cutoff concentrations and differences in the positivity
rate between the methods are shown in Table.

-Patient 3: Refractory MM with stable disease: abnormal sFLC and HLC ratios with
4.23 g/dL of monoclonal protein.
Conclusion:The inclusion of the Hevylite assay allows the quantitative follow-up
of monoclonal immunoglobulins, particularly interesting when patients achieve
a deepness of response where the standard SPE and IF become negative. Results
obtained with Hevylite assay are in agreement with the other results. This preliminary
data suggests that the new HLC assay may add specificity to the Stringent Complete
Response. More studies are required to establish its prognostic value in the response
evaluation.

A-056

Conclusions: Higher Tg positivity rate in Tg-AAb positive samples was observed

by LC-MS/MS method as compared to the IA; the difference is likely caused by
underestimation of Tg concentrations caused by interference of Tg-AAb with the IA.

Performance of the Roche Elecsys Thyroglobulin (Tg) II immunoassay

Table. Tg positivity rate in Tg-AAb positive samples analyzed by LC-MS/MS and

Beckman Coulter immunoassay.

B. C. Netzel, M. A. Lasho, A. Algeciras-Schimnich. Mayo Clinic,

Rochester, MN

Tg cutoff concentration,
ng/mL
>0.5
>0.6
>1
>2
>5
>7
>10
>15
>20

Percent of samples with concentration above the cutoff

Background: Serum thyroglobulin (Tg) measurement is considered the gold

standard in the follow-up of patients with differentiated thyroid cancer following total
thyroidectomy and radioactive iodine ablation. In athyrotic patients, Tg is an excellent
tumor marker because it is produced exclusively by the follicular cells of the thyroid.
Performance of currently available Tg immunoassays varies due to different assay
sensitivities, standardization against the certified reference material (CRM-457), and
Tg autoantibody (TgAb) interference, among others. In this study we evaluated the
analytical performance of the Roche Elecsys Tg II immunoassay.

LC-MS/MS, % Beckman Access, % Difference, %

37.5
37.1
33.4
26.4
18.3
15.1
12.5
10.1
8.1

37.0
34.8
30.0
23.9
16.6
13.8
11.4
9.2
7.6

0.5
2.3
3.4
2.5
1.7
1.3
1.1
0.9
0.5

A-055
Value of a new biochemical parameter (serum Heavy chain/Light Chain pairs)
in the follow-up of Multiple Myeloma after treatment

R. P. Garay, I. J. Ventura, E. A. Diez, E. Landeta, I. Amariko, A. G. Vicua,

A. S. Izarra. Hospital Universitario Cruces, Bilbao, Spain
Background:Multiple Myeloma (MM) has an incidence of 1-10% of all cancers. The
laboratorial assays usually used for its diagnosis and follow-up are: detection of the
monoclonal protein by Serum Protein Electrophoresis (SPE) and Immunofixation (IF)
in serum and 24h urine, total immunoglobulins levels by nephelometry and serum
Free Light Chains (sFLC). Determination of the bone marrow plasma cells, lytic
lesions by MRI, complete blood count, creatinine and calcium levels are also used.
Recently, a new technique that allows the analysis of immunoglobulin heavy chain/
light chain pairs (Hevylite) has been developed.Objectives:The sFLC determination
has been introduced in 2006 in the Stringent Complete Response subcategory by
the International Myeloma Working Group (IMWG), corresponding to a Complete
Response plus normal FLC ratio and absence of clonal cells in bone marrow. The
reasons for its inclusion are that it is a highly sensitive marker for sFLC and an
excellent indicator of clonality (Durie et. al., Leukemia 2009). A recent study has
further corroborated its correlation with a more stringent level of response with
clinical prognostic impact (Kapoor et. al., JCO 2013). The aim of this study is to
evaluate the utility of the Hevylite ratio (HLCr) together with sFLC in the follow-up
of 3 MM IgG patients under treatment and follow-up.
Methods:sFLC and HLC were measured by turbidimetry (FreeliteTM and Hevylite,
on a SPAPLUS, Binding Site). The monoclonal protein was identified and quantified
by electrophoresis (Capillarys Hydrasys Focusing, Sebia). Samples from 3 MM
patients were analyzed:
-Patient 1: 58 years old woman with an IgG-L type MM stage II-A. 13 samples
collected from June 2011 to July 2013. 1st line treatment (Velcade-DexamethasoneAdriamycin) and TASPE (July 2012).
-Patient 2: 61 years old man with IgG-K MM stage II-A, ISS-2. 14 samples from May
2011 to November 2013. 1st line treatment (Velcade-Dexamethasone-Adriamycin)
and TASPE (May 2012).
-Patient 3: 76 years old woman with IgG-K MM stage II, ISS-2. 14 samples from
March 2013 until January 2014. 1st line treatment (Bortezomib-MelphalanPrednisone) and 2nd line treatment (Cyclosphophamide-prednisone).
Results:-Patient 1: in sCR since December 2012. sFLC and HLC ratios normalized
with absence of monoclonal protein by SPE and IF.
-Patient 2: achieved CR after TASPE, at this time in biochemical relapse. After
treatment, the HLCr normalized in May 2012, the SPE became negative in August
2012 and the IF in November 2012, achieving a CR. The sFLC ratio is abnormal
throughout follow-up, and currently there is a monoclonal protein detected by IF and
SPE.

Methods: The Roche Elecsys Tg II immunoassay (Roche Diagnostics, Indianapolis,

IN) is a quantitative, two step, double antigen sandwich assay standardized against
CRM-457 for the measurement of thyroglobulin in serum and plasma using the
electrochemiluminescence immunoassay ECLIA technology. The assay uses 35
l of sample and has a total assay time of 18 minutes. Testing was performed on
the Roche Diagnostics cobas e411 analyzer. Accuracy was investigated by recovery
studies using CRM-457. Imprecision studies were conducted using Liquichek
Tumor Marker Control (Bio-Rad, Hercules, CA) and Roche PreciControl quality
control (QC) materials. Analyte measurement range (AMR) studies were conducted
by diluting a high concentration Tg serum sample with a negative Tg serum sample.
LOQ studies were conducted using a low concentration Tg serum sample. Method
comparison studies with the Beckman Access Tg assay (Beckman Coulter, Brea, CA)
and an in-house developed LC-MS/MS Tg assay was performed using de-identified
serum samples collected for routine Tg determination.
Results: Average recovery of CRM-457 was 106% (range 101-111%). Inter-assay
imprecision studies produced coefficient of variation (CV) of <6% (range 1.4-5.5%)
at concentrations of 6.4, 21.8, 85 and 188 ng/mL. The AMR of the assay was 0.1
to 500 ng/mL with Passing-Bablock regression fit of y = 0.96x -0.32 (r 2=0.999).
Serial dilutions (x2 to x64) to expand the AMR produced an average recovery of 97%
(range 91-104%). LOQ was determined to be 0.1 ng/mL (CV = 18%). The assay was
compared to the Beckman Access Thyroglobulin assay (N=37, range 0.1-500 ng/mL).
The Spearman correlation coefficient was 0.990 with a slope of 1.28 and intercept of
-0.14 by Passing-Bablock regression fit. Comparison with the in-house Tg LC-MS/
MS assay (N=129, range 0.5-500 ng/mL) produced a Spearman correlation coefficient
of 0.943 with a slope of 1.58 and intercept of -0.69 by Passing-Bablock regression fit.
Conclusion: The Roche Elecsys Tg II shows good analytical performance and
provides reliable Tg measurement for the management of thyroid cancer patients.

A-057
Single Cell Analysis of Heterogeneous Circulating Tumor Cell Populations

L. M. Millner1, K. S. Goudy2, M. W. Linder3, R. Valdes Jr3. 1University of

Louisville, Louisville, KY, 2PGXL Technologies, Louisville, KY, 3University
of Louisville and PGXL Technologies, Louisville, KY
Introduction: Enumeration of circulating tumor cells (CTCs) in blood is used in
breast cancer patients as an independent predictor of outcome. Present methods do
not distinguish subtypes and only detect epithelial-type CTCs. This is significant
because CTCs experience epithelial to mesenchymal transition (EMT), a process that
increases motility, disease progression, and decreases epithelial marker expression.
This process may also change the CTC susceptibility to chemotherapeutics and
eligibility for treatment. For example, patients with overexpression of HER2 are
eligible for treatment with Herceptin. Examining CTCs as a bulk population may mask
individual overexpression of markers used for targeting therapy. Here we describe a
comprehensive method for characterizing the molecular heterogeneity of CTCs which
could play an important role in directing personalized cancer therapeutics.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S15

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

Objective: To develop a method for identifying single cell heterogeneity within a
population of circulating tumor cells.
Methods: As a model, we used the breast cancer cell line MDA-MB-231 (MDA).
This cell line was chosen because it has low expression of EpCAM and may represent
cells that have experienced EMT. MDA cells are typically negative for EpCAM and
HER2 protein expression but do express CD44. Cells were sorted into single cell
populations using DEPArray Technology (Silicon Biosystems) and analyzed by single
cell RT-PCR for 3 targets (EpCAM, ErbB2 (HER2), and CD44) and a housekeeping
gene (ACTB). Using DEPArray technology the mean fluorescent intensity (MFI) of
EPCAM, CD44 and HER2 protein expression was measured on 3,656 individual cells
and MFI signals greater >1000 were considered positive.
Results: ACTB transcript expression as a positive control was confirmed in each
of the eleven cells. EpCAM transcript expression was not detectable in any of the
cells. CD44 transcript expression was observed in 10/11 (91%) and ErbB2 (HER2)
expression was observed in 4/11 (36%) cells. We then measured the mean fluorescent
intensities (MFI) for each of the cell surface target antigens on 3,656 individual cells.
The MFI values [mean SD (range)] were: EpCAM, 52591 (369-870); CD44,
37221598 (1223-13963), and HER2, 562154 (389-6312).
Conclusion: The wide range of MFI signaling for HER2 demonstrates a discrete
subpopulation of cells expressing high levels of HER2 within a population that on
average expresses no or very low levels of HER2. This single cell analysis method
may provide identification of a subpopulation of Herceptin-responsive cells within an
apparently non-responsive group. Single cell subtyping has the potential to facilitate
individually tailored therapies based on each patients heterogeneous CTC profile.

A-058
Prognostic Biomarker Isocitrate Dehydrogenase-1 Mutations in Patients with
Glioblastoma Multiforme

J. Polivka1, J. Polivka2, V. Rohan2, M. Pesta3, T. Repik2, O. Topolcan4.

1
Department of Histology and Embryology and Biomedical Centre,
Faculty of Medicine in Plzen, Charles University in Prague, Plzen, Czech
Republic, 2Department of Neurology, Faculty of Medicine in Plzen, Charles
University in Prague and Faculty Hospital Plzen, Plzen, Czech Republic,
3
Department of Biology, Faculty of Medicine in Plzen, Charles University
in Prague, Plzen, Czech Republic, 4Central Imunoanalytical Laboratory,
Faculty Hospital Plzen, Plzen, Czech Republic
Background Glioblastoma multiforme (GBM) is the most malignant primary brain
tumor in adults with high mortality. Standard therapy (surgery, radiotherapy and
chemotherapy with temozolomide) has only limited effectiveness and the median
survival of patient with GBM is 12.1 - 14.6 months. Recent GBM whole-genome
studies revealed some novel prognostic and predictive biomarkers such as the
recurrent mutations in metabolic enzyme IDH - Isocitrate dehydrogenase (isoforms
IDH1 and IDH2). The distinctive mutation IDH1 R132H was uncovered to be a strong
prognostic biomarker for glioma patients. Therefore we investigated the prognostic
role of IDH1 R132 mutation in our GBM patient cohort.
Methods The IDH1 R132H mutation status was assessed in the Formalin-Fixed
Paraffin-Embedded (FFPE) tumor samples from 44 GBM patients treated in the
Faculty Hospital Plzen between 2008 and 2013. The real-time PCR with TaqMan
mutation detection assays and TaqMan mutation detection IPC reagent kit was used.
The IDH1 R132H mutation status was correlated with the progression free survival
(PFS) and overall survival (OS) of patients using Kaplan-Meier survival analysis and
Wilcoxon test.
Results The IDH1 R132H mutation was identified in 20 from 44 GBM tumor samples
(45,4%). The majority of mutated tumors were secondary GBMs (16 in 18, 89.9%).
Low frequency of IDH1 mutations was observed in primary GBMs (4 in 26, 15.3%).
Patients with IDH1 R132H mutation had longer PFS - 136 vs. 51 days (P<0.021) as
well as OS - 270 vs. 130 days (P<0.024).
Conclusion The prognostic value of IDH1 R132H mutation in GBM patients was
observed in our study. Patients with this mutation had significantly longer PFS and OS
than patients with wild-type IDH1 and suffered more likely from secondary GBMs.
The IDH1 mutation status could be used as a strong prognostic factor for patients with
GBM and should be further studied in larger patient cohort.
Supported by MH CZ - DRO (Faculty Hospital Plzen - FNPl, 00669806) and by the
project ED2.1.00/03.0076 from the European Regional Development Fund.

A-059
New monoclonal antibodies detect all immunoglobulin free light chains in urine
samples from over 13,000 patients

J. Campbell, J. Heaney, Y. Wang, M. Cobbold, M. Goodall, T. Plant, M.

Drayson. University of Birmingham, Birmingham, United Kingdom
A decade or so ago, the first automated assay was launched for the quantitation of
serum and immunoglobulin free light chains (FLC). FLC measurement is now
a fundamental procedure in the diagnosis and monitoring of patients with plasma
cell dyscrasias including multiple myeloma and related subtypes including lightchain-only, oligosecretory and non-secretory myeloma. Despite these advances, the
assay, which uses sheep polyclonal anti-human FLC antibodies, has a number of
well-observed limitations. It has been proposed that monoclonal antibodies (mAbs)
may overcome these limitations. The development of FLC specific mAbs is difficult
because the mAbs must demonstrate specificity for epitopes that are exposed on
FLC but hidden on LC bound to whole immunoglobulin. This is complicated by
the paucity of constant domain epitopes available; which can be further reduced by
polymerisation of FLC, particularly FLC , thus reducing the number of potential
binding sites. Production of mAbs specific for FLC has been described previously but
other groups have either found that their mAbs did not detect FLC from all neoplastic
plasma cell clones tested, or, have not tested sufficient clones to be confident that the
mAbs would detect the FLC from 100% of neoplastic clones. Hence, the purpose of
this study was to prospectively assess the clinical utility of new highly-specific mouse
anti-human FLC mAbs on a large number of consecutive patient samples. Anti- and
anti- FLC mAbs were covalently coupled to different polystyrene Xmap beads
and assayed, simultaneously, in a multi-plex format by Luminex (mAb assay). The
mAbs displayed no cross-reactivity to bound LC, the alternate LC type, or other human
proteins and had improved sensitivity (<1mg/L) over the gold standard for identifying
paraprotein, immunofixation electrophoresis (IFE; approximate sensitivity is 10mgL).
The competitive inhibition format gave a broad calibration curve (up to 437.5 mg/L)
and prevented anomalous results for samples in antigen excess (i.e. high FLC levels).
The mAb assay had no false negatives and identified all monoclonal FLC in 13,090
urine samples tested (22.8% with monoclonal and 9.0% with monoclonal by IFE),
and also detected all samples with polyclonal FLC. In a small cohort of Bence Jones
positive samples (n=100), the mAb assay correlated excellently with densitometry,
the gold standard for quantitating urine FLC. Importantly this shows that the mAbs
are close to the ideal of detecting FLC from all patients and neoplastic plasma cell
clones, and may be the first published mAbs with this clinical utility. Given the overall
effectiveness of the anti-FLC mAbs, further clinical validation is now warranted on
these mAbs in other assay platforms they are incorporated, including Seralite, a rapid
(10-minute) and portable FLC test to be used at the point-of-care.

A-060
Development of RT-qPCR Assays for The Detection of Circulating Tumor Cells
in Breast Cancer

S. Ahn1, S. Park2, H. Wang3, S. Park1, S. Kim4, Y. Kim1, H. Kim1, G. Kim1,

J. Kim1, H. Jin5, S. Kim6, H. Lee1. 1Department of Biomedical Laboratory
Science, College of Health Sciences, Yonsei University, Wonju, Korea,
Republic of, 2Department of Clinical Laboratory Science, College of Health
and Therapy, Daegu Hanny University, Daegu, Korea, Republic of, 3M&D,
Inc., Wonju Eco Environmental Technology Center, Wonju, Korea, Republic
of, 4Institute for Life Science and Biotechnology, Yonsei University, Seoul,
Korea, Republic of, 5Department of Clinical Laboratory Science, College
of Health Sciences, Catholic University of Pusan, Busan, Korea, Republic
of, 6Department of Surgery, College of Medicine, Yonsei University, Seoul,
Korea, Republic of
Background: Cancer cells which become detached from the primary tumors and
enter into the systemic circulation are called circulating tumor cells (CTCs). The
human epidermal growth factor receptor 2 (HER2, also known as erbB2) is crucial
for treatment of breast cancer patients. HER2 is over-expressed in 20 to 30% in
blood samples of breast cancer patients. Especially, HER2 genes over-expression is
associated with a poor clinical outcome. Therefore, the detection of HER2 expressing
CTCs in the blood may have important prognostic and therapeutic treatment
implications. In this study, apart from using HER2, EpCAM, CK-19, Ki-67, and
hTERT were used for detection CTCs in peripheral blood of breast cancer patients.
Furthermore, correlation between HER2 and CTC markers mRNA level in the blood
were determined.
Methods: Human breast carcinoma cell line SK-BR-3, MCF-7 and MDA-MB-231
were used for the development of the assay and the confirmation of HER2, EpCAM,

S16

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

CK-19, Ki-67, hTERT and GAPDH expression. A Total of 188 breast cancer patients
who include 34 ductal carcinoma in situ (DCIS) patients, 93 stage I patients, 58 stage
II patients and 3 stage III patients. A total of 50 healthy donors who did not have a
breast cancer were also enrolled for this study. All blood samples were handled for
extracting total RNA using TRIzol Reagent (Invitrogen, Carlsbad, California, USA)
then the cDNA was synthesized. The mRNA expression levels of HER2, EpCAM,
CK-19, Ki-67, and hTERT relative to GAPDH were measured by RT-qPCR TaqMan
assay.
Results: Among a total of 188 patients, 39 patients (20.7%) displayed an overexpression of HER2 mRNA, while none of the healthy blood donors over-expressed
HER2 mRNA. In 37 out of 39 HER2 positive patients, not only HER2 mRNA, but
also at least one other type of marker was overexpressed at the same time. Among the
149 HER2 negative patients, 114 patients (76.5%) were positive at least one other type
of marker. The HER2 mRNA levels in blood had a correlation with Ki-67 mRNA level
(Pearson r=0.4358, R square=0.2360) and hTERT mRNA level (Pearson r=0.2988, R
square=0.0893) in blood. As the breast cancer stage progress, patients who were overexpressed tumor association markers, such as hTERT, Ki-67, and HER2 were tend to
increasing. On the other hand, expression of CTC epithelial markers, such as EpCAM
and CK-19 not seem to have correlation with stage of cancer.
Conclusion: In conclusion, HER2 expression in the blood occurs concurrently with
CTC markers. In this reason, CTC markers could be used for detection CTCs in blood
of breast cancer patients. The results from this study seems to suggest that detection
of CTCs using CTC markers and HER2 allow for more effective management of and
better prognosis for breast cancer.

A-061
Analytical Evaluation of a Newly Developed ELISA for the Detection of Soluble
Tumour Necrosis Factor Receptor 2 (sTNFRII) in Sera from Patients with
Ovarian Cancer

A. Chacko, P. Ratcliffe, L. Stevenson, J. Lindsay, P. Lowry, R. McConnell,

S. FitzGerald. Randox Laboratories Limited, Crumlin, United Kingdom
Background: There has been significant interest in discovery of biomarkers and
biomarker panels for early detection of precancerous ovarian tumours. Various
screening strategies combining serum-based markers with other clinical parameters
are being tested in on-going trials. To date, no serum-based markers have been proven
to improve diagnostic performance over the standard ovarian cancer marker CA125, either alone or as part of a biomarker panel. Previous studies have suggested
that sTNFRII levels may increase during tumour progression in different cancer
types, including ovarian cancer. This study reports the analytical evaluation of a
new developed enzyme-linked immunosorbent assay (ELISA) for the specific and
sensitive detection of sTNFRII and its applicability to the study of patients with
ovarian cancer. Methods: Sheep were immunized with the extracellular domain of
the sTNFR II recombinant protein. Lymphocytes were collected and fused with
heteromyeloma cells. Hybridomas were screened for immunoreactivity against native
human sTNFRII. Hybridomas which showed strong reactivity to native antigen
and <1% reactivity to cross reactants were selected for cloning to produce stable
monoclonal hybridomas. An optimal antibody pair was identified for development of a
sandwich ELISA platform for detection of sTNFRII in human serum. sTNFRII levels
were determined in sixty one serum samples (34 from ovarian cancer patients and 27
healthy female controls). Statistical analysis was performed and box plots / Receiver
Operating Characteristics (ROC) curves were constructed using GraphPad Prism.
Results: The assay detected native sTNFRII across an assay range of 0-16ng/ml. The
limit of detection,defined as the analyte concentration corresponding to an absorbance
equal to blank mean plus 2xSD (n=20), was 0.324ng/ml. Intra-assay precision data
(n=12) showed recovery at 103.6% 10.4% (10.0% CV) for 2ng/ml, and 86.5%
5.1% (5.9% CV) for 4ng/ml sTNFRII. Patient sera were run at a dilution of 1 in
5 to best match the expected clinical range for sTNFRII, giving an effective assay
range of 1.6-80ng/ml. Median sTNFRII levels were significantly increased in ovarian
cancer patients compared to healthy females (22.7ng/ml v 5.3ng/ml, Mann Whitney
p<0.0001), which gave an area under the curve (AUC) of 0.907 for ovarian cancer
versus healthy females. Conclusion: The results indicate that this newly developed
ELISA is applicable to the specific detection of sTNFRII in human serum. The
significant increase in median sTNFRII levels in serum from ovarian cancer patients,
when compared to controls, indicates that this assay may be suitable for further studies
into the use of sTNFRII as a marker for cancer diagnosis and monitoring.

A-062
Examination of Thyroglobulin and Thyroglobulin Antibody Testing Processes
for an Urban Endocrine Center
S. E. Wheeler, L. Liu, H. Blair, O. Peck Palmer. University of Pittsburgh, Pittsburgh,
PA
BACKGROUND: The American Society of Cancer estimates in 2014 ~62,980
individuals will be diagnosed with thyroid cancer. Patients and physicians require
accurate and timely in-house thyroglobulin results. Specifically, in differentiated
thyroid carcinoma (DTC), the most common thyroid cancer, thyroglobulin (Tg) is
used to assess disease recurrence in patients who have undergone thyroidectomy.
Commercially available Tg immunoassay methods are most common but are
susceptible to Tg antibody (TgAb) interference. In most laboratories TgAb is quantified
and manufacturer cutoffs are used to categorize a specimen as TgAb negative (Tg
analyzed by immunoassay) or TgAb positive (Tg measurement is referred to an
alternate methodology). Common methodologies include the Tg radioimmunoassay
(RIA) method and the recently available LC/MS/MS methodology. Examination of
alternative test options is critical as referring samples to outside laboratories increases
result turnaround time and patient costs.
OBJECTIVE: To determine the optimal testing algorithm for a large endocrine
center, we examined the correlation between the current in-house immunoassay
methodology for Tg measurement with RIA and LC/MS/MS methodologies using
patient specimens categorized with low or high TgAb concentrations.
MATERIALS AND METHODS: Excess samples (n=40; -80C storage) were
obtained from outpatient adults as well as 10 healthy volunteer specimens. Patient
specimens were divided into 2 groups: 20 TgAb <20 U/L specimens; 20 TgAb
>20 U/L specimens. Samples were analyzed for TgAb using a solid phase enzyme
labeled chemiluminescent sequential immunometric assay (Siemens Immulite 2000
XPi, Erlangen Germany; manufacturers cutoff of 20 U/L). Tg was analyzed using
a simultaneous one-step immunoenzymatic assay (UniCel DxI 800 automated
analyzer, Beckman Coulter, CA; all samples), a RIA method (USC Endocrine
Laboratories, CA; used only for samples TgAb >20 U/L), and LC/MS/MS
methodology (Quest Diagnostics, Nichols Institute, CA; all samples).
RESULTS: Overall high correlation was demonstrated between the DXI800 and
LC/MS/MS methodologies (R2=0.99; slope = 1.309). Specimens with low TgAb
(<20 U/L) demonstrated good correlation between the DXI800 and the LC/MS/
MS methodologies (R2=0.99; slope = 1.312). We examined the effects of TgAb
interference (TgAb >20 U/L) and found a good correlation between the DxI800 and
the LC/MS/MS methodology (R2 = 0.97; slope = 1.243). However,Tg measurement
between the RIA and LC/MS/MS methods was lower (R2 = 0.82) particularly for
specimens with Tg concentrations >13ng/mL (>13 ng/mL; R2 = 0.67). TgAb
interference was reflected in the method comparison of the DxI 800 and RIA (R2 =
0.76; slope = 1.216) methods. We observed high variability between TgAb methods
similar to previous studies.
CONCLUSIONS: The high correlation between LC/MS/MS and DXI800 suggests
that both methods may be appropriate for our patient population. However, before LC/
MS/MS testing is placed into routine use our findings warrant validation in a larger
TgAb defined population.

A-063
The upregulation of ICAM-1 mediated by leptin is Rho/ROCK-dependent and
enhances gastric cancer cell migration

C. Wang, Z. Dong, Y. Yang, L. Du, G. Zheng, G. Zheng, W. Li, X. Zhang,

Z. Li, L. Wang, J. Li, H. Liu. Qilu Hospital, Shandong University, Jinan,
China
Background: Gastric cancer (GC) ranks as the second leading cause of cancer-related
death in the world. Adipocytes provide fatty acids for rapid tumor growth, and the
dysfunction of lipid metabolism can lead to the pathogenesis of human GC. Leptin
is an adipokine of the obesity (ob) gene, and our previous study showed that leptin
promote GC cell invasion by AKT/MT1-MMP pathway. However, the exact effect
and the underlying mechanism of leptin in GC metastasis remain unclear. Intercellular
adhesion molecule-1 (ICAM-1) is overexpressed and plays crucial roles in tumor
metastasis. This study aimed to characterize the influence of leptin on ICAM-1
expression in GC and elucidate its underlying molecular mechanism.
Methods: Archived paraffin-embedded GC tissues and matched adjacent normal
gastric tissues were collected from 84 patients who underwent surgery for primary
gastric carcinoma. The expression of leptin and ICAM-1 were detected by
immunohistochemistry, and the correlation of two proteins was further analyzed. The

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S17

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

effect of leptin on GC cell (AGS and MKN-45 cells) migration was measured by
transwell. The level of ICAM-1 at both mRNA and protein were detected by RTPCR and western blot after treatment with leptin. Moreover, the cell surface ICAM1 and sICAM-1 were detected by flow cytometry and ELISA. ICAM-1-siRNA was
designed and transiently transfected in GC cells. RhoA GTPase activity was detected
using the G-LISA RhoA activation assay kit. Correlations of leptin and ICAM-1
expression with clinicopathologic factors were analyzed by Kruskal-Wallis test or
Mann-Whitney U test, as appropriate. Chi-squared test was applied to analyze the
correlation of leptin and ICAM-1 respectively. Other data from experiments were
analyzed by paired Students t-test or one-way ANOVA wherever appropriate. P<0.05
was statistical significance.
Results: Immunohistochemical analysis revealed that leptin (48/84, 57.1%) and
ICAM-1 (54/84, 64.2%) were overexpressed in GC tissues, and they were positively
correlated with each other (P<0.001), as well as with the clinical stage and lymphatic
metastasis. In transwell assay, leptin promoted GC cell (AGS and MKN-45)
migration in a time- and dose-dependent manner. Furthermore, leptin induced GC
cell migration by upregulating ICAM-1 expression (mRNA: 4.060.54-fold for
AGS, P<0.001; 2.560.33-fold for MKN-45, P=0.005. Protein: 3.070.25-fold for
AGS, P<0.001; 2.90.26-fold for MKN-45, P=0.003), and knockdown of ICAM-1 by
small interference RNA (siRNA) blocked this process (AGS: 53.9%3.6%, P=0.020;
MKN-45: 42.79%3.78%, P=0.005). Notably, the surface expression of ICAM1 (AGS, P<0.001; MKN-45, P<0.01), as well as the soluble ICAM-1 (sICAM-1)
(AGS, P<0.05; MKN-45, P<0.01), was also enhanced by leptin. Moreover, leptin
increased ICAM-1 expression through Rho/ROCK pathway, which was attenuated
by pharmacological inhibition of Rho (C3 transferase) at 0.25 g/mL (AGS, P<0.01;
MKN-45, P<0.01) or inhibition of its downstream effector kinase Rho-associated
protein kinase (ROCK) (Y-27632) at 3.3 M (AGS, P<0.01; MKN-45, P<0.01),
suggesting an essential role of Rho/ROCK pathway in this process.
Conclusions: Our findings indicate that leptin enhances GC cell migration by
increasing ICAM-1 expression through Rho/ROCK pathway, which may provide
preliminary experimental clues for the development of new therapies against the
metastasis of GC.

A-064
hCG candidate epitopes for improving the measurement of hCG: results from
the second ISOBM TD-7 workshop

P. M. Hemken1, E. Paus2, C. Sturgeon3, W. Stewart4, J. Skinner1, L.

Harwick1, S. Saldana1, C. Ramsay1, K. Rupprecht1, K. H. Olsen2, J. M.
Bidart5, U. H. Stenman6, P. Berger7. 1Abbott Laboratories, Abbott Park,
IL, 2Oslo University Hospital, Oslo, Norway, 3Royal Infirmary, Edinburgh,
United Kingdom, 4Ninewells Hospital and Medical School, Dundee, United
Kingdom, 5Institut Gustave-Roussy, Villejuif, France, 6Helsinki University
Central Hospital, Helsinki, Finland, 7University Innsbruck, Innsbruck,
Austria
The purpose of this collaboration was to determine specificity profiles and
epitopes recognized by diagnostically relevant antibodies (Abs) directed against
human chorionic gonadotropin (hCG). This provided the basis for improving the
measurement of hCG by harmonization of epitopes of the Abs used and to build broad
assay specificity consensus for use as a tumor marker, pregnancy and pregnancy
related disorders.
Eight companies and research groups submitted 69 Abs directed to hCG and hCGrelated variants. Each of these Abs were characterized in detail by the participants
of the Second International Workshop (WS) on hCG of the International Society of
Oncology and Biomarkers Tissue Differentiation 7 (ISOBM TD-7).
To determine the specificities of the Abs, the First WHO International Reference
Reagents for six hCG variants, hCG, hCGn, hCG, hCGn, hCGcf, and hCG were
used. Seventeen reference monoclonal (m)Abs were used to assign molecular epitope
localizations for the ISOBM-mAbs in the WS. This was performed by comparing
ISOBM-Ab specificity, sandwich compatibility, mutual inhibition profiles and
affinities, to mAbs of known epitope specificities.
The data shows that 48 mAbs recognized hCG, 8 hCG-, and 13 -heterodimerspecific epitopes. Twenty-seven mAbs were of pan hCG specificity. Two of these pan
hCG mAbs had very low cross-reactivity with hLH (<0.1 %; epitope 1), 12 with low
hLH cross-reactivity (<1.0 %; epitopes 2/4), and 13 with high hLH cross-reactivity
(>>1 %; epitopes 3/5). Four mAbs recognized epitopes on hCGcf-only (e.g.,
epitopes 11 and 13) and six mAbs epitopes on the remote hCG-carboxyl-terminal
peptide (epitopes 8 and 9).
For routine diagnostic measurements, methods are used that either detect hCGonly, hCG-only, or hCG together with hCG or hCG together with hCG and
hCGcf. Sandwich assays that measure hCG plus hCG and eventually hCGcf

S18

should recognize the protein backbone of the analytes preferably on an equimolar

basis, should not cross-react with hLH and not be susceptible to blunting of signal
by nonmeasured variants like hCGcf. Such assays can be constructed using pairs
of mAbs directed against the cystine knot-associated epitope 1 in combination with
epitopes 2 or 4 on hCG peptide loops1+3 protruding from the central cysteine
knot.
In summary, the results of this hCG ISOBM TD-7 WS1 in combination with those of
the First WS2 enable recommendations to be made regarding epitope combinations to
be used for the design of immunoassays for hCG and its variants.
References: 1. P. Berger, E. Paus, P. M. Hemken, C. Sturgeon, W. W. Stewart, J. P.
Skinner, L. C. Harwick, S. C. Saldana, C. S. Ramsay, K. R. Rupprecht, K. H. Olsen,
J.-M. Bidart, U.-H. Stenman. Candidate epitopes for measurement of hCG and related
molecules: the second ISOBM TD-7 workshop. Tumor Biol. 2013; 34:4033-4057.
2. P. Berger, C. Sturgeon, J.-M. Bidart, E. Paus, R. Gerth, M. Niang, A. Bristow, S.
Birken, U.-H. Stenman. The ISOBM TD-7 workshop on hCG and related molecules.
Towards user-oriented standardization of pregnancy and tumor diagnosis: assignment
of epitopes to the three-dimensional structure of diagnostically and commercially
relevant monoclonal antibodies directed against human chorionic gonadotropin and
derivatives. Tumor Biol. 2002; 23:1-38.

A-066
High sensitivity detection of residual disease in multiple myeloma using mass
spectrometry

J. R. Mills, D. R. Barnidge, D. Murray, J. A. Katzmann, A. Dizpenzieri, D.

L. Murray. Mayo Clinic, Rochester, MN
Background:Traditionally, detection of an M-protein (monoclonal immunoglobulin)
has been used to diagnose and monitor multiple myeloma (MM). As therapies for
MM have improved, more sensitive methods have been used to define response:
immunofixation electrophoresis (IFX) of serum and urine, normalization of the serum
immunoglobulin free light chain (FLC) ratio, and high sensitivity flow cytometry
to detect clonal plasma bone marrow cells. It is hoped that these more sensitive
approaches will differentiate those patients with minimal residual disease (MRD)
versus no residual disease (NRD), the latter which could mean cure. Flow cytometry
of plasma cells requires bone marrow aspiration, which is inconvenient and expensive
and is potentially limited by sample bias. More sensitive methods to differentiate
MRD from NRD using serum would be advantageous.
Objectives:To develop a high sensitivity method to detect residual M-protein secreted
from malignant plasma cells as a means to monitor minimal residual disease (MRD)
in serum.
Methods:We developed a microLC-ESI-Q-TOF mass spectrometry assay to detect
the presence and accurate mass of malignant-specific monoclonal immunoglobulins,
a method we termed monoclonal-immunoglobulin-Rapid-Accurate-MassMeasurement (miRAMM). Briefly, serum immunoglobulins are enriched and then
reduced with DTT, and samples are separated using an Eksigent-Ekspert LC system
using a Poroshell C3 1x75mm column running at 25L/min with gradient of aqueous
(0.1 % formic acid) to organic (90:10 ACN/IPA) over 24 minutes. The accurate mass
of monoclonal light chains is determined by deconvolution of the mass spectra of
multiply charged ions across the retention times of immunoglobulin light chains.
Serum samples from 14 patients (cohort #1) positive for M-proteins by serum protein
electrophoresis (SPEP) and IFX were collected. Neat and diluted serum (diluted
1:1000 in pooled serum from healthy donors) was analyzed by IFX, SPEP, FLC and
miRAMM. Cohort #2 included pre-and post-therapy serum from 21 MM patients who
had been classified to have achieved stringent complete response (sCR) post; sCR is
defined as an undetectable M-protein by SPEP, a negative FLC ratio, a negative IFX,
and absence of clonal plasma cells by 4-color flow cytometry from bone marrow. The
accurate mass of disease-associated light chains determined by miRAMM pretherapy
was used to test for MRD post-therapy in the sCR patients.
Results:For patients in cohort #1 we were able to identify and determine the accurate
molecular mass of the monoclonal light chains in all cases (14/14; 100%). To further
test the relative sensitivities of each assay, these same patients samples were subject
to 1:1000 dilution into pooled human serum and retested by SPEP, IFX, FLC ratio,
and miRAMM. A monoclonal was detected in 0/14, 2/14 (14%), 0/6, and 13/14 (93%),
respectively. In addition, of 21 patients in cohort #2 who had been classified to have
achieved sCR by conventional means, including 4-color flow cytometry of bone
marrow, 67% had detectable residual monoclonal light chains by miRAMM.
Conclusion:This study demonstrates that miRAMM is a more sensitive approach to
detect MRD compared with current methods. This method provides additional value
in that the accurate mass measurement provides a unique molecular mass identifier of
an individuals malignancy.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cancer/Tumor Markers

Tuesday, July 29, 9:30 am 5:00 pm

A-067
Identification of four molecular subclasses of Luminal Breast Cancer with
different likelihood of recurrence and response to Tamoxifen

R. A. Abbud-Antaki, A. Deluca. Falcon Genomics, Inc., Allison Park, PA

Background: Genomic studies have revealed four main molecular subtypes of breast
cancer with significant heterogeneity within each class. Several genomic assays are
now being used in the clinic for prediction of recurrence in breast cancer patients, and
determination of which patients could be spared from Chemotherapy. The objective
of this study is to develop a reliable clinical diagnostic test for breast cancer patient
classification.
Results: We report on a gene expression-based test for classification of breast
cancer tumors that identifies additional subsets of Luminal breast cancer with
differences in prognosis and response to Tamoxifen therapy. For these studies, we
employed Affymetrix HG-U219 GeneAtlas microarrays to examine gene expression
profiles from breast cancer tissue (n=17) and normal adjacent tissue (n=6). When
we performed Supervised hierarchical clustering using the previously published
intrinsic gene signature, we could not obtain definite classification for each patient
tumor. This has motivated us to come up with a quantitative approach to breast cancer
patient classification.
We first examined expression of genes that are commonly used in the pathological
evaluation of breast cancer tissue. These were: ESR1, PGR, ERBB2, AR, luminal
keratins (KRT 8, 18, 19), basal keratins (KRT 5, 6B, 14), and Cadherins (E-CDH, OBCDH, and P-CDH). We then classified our patient tissues according to luminal, basal
KRTs, E-CDH, and OB-CDH. This gave us 8 different classes with 380 differentially
expressed genes. Only 30 out of the 380 genes were found in the intrinsic gene list.
The genes that were differentially expressed between these classes were not limited to
those related to the luminal and basal phenotypes; other differentially expressed gene
clusters included proliferation and stromal genes.
We then validated the ability of this signature to classify published breast cancer gene
expression datasets (GSE2034, GSE5327, GSE1561, GSE26971, GSE9195, and
GSE6532). Supervised Hierarchical Clustering and Gene Set Enrichment Analysis
using our 380-gene signature revealed that Luminal cancer patients could be further
divided into 4 groups (L1-L4). L1 samples, also known in the literature as Luminal-B,
expressed high levels of proliferation genes and showed the highest likelihood of
recurrence among untreated patients, but responded well to Tamoxifen treatment.
However, the likelihood of recurrence after Tamoxifen for these patients only became
similar to other untreated Luminal groups (L3-L4). However, the likelihood of
recurrence for L2 patients was similar irrespective of Tamoxifen. L3 patients, on the
other hand, showed statistically significant increase in survival after Tamoxifen, but
it was not as good as L4 patients. Some patients had additional genetic abnormalities
that made them susceptible to relapse. L4 patients had a better prognosis, with
significantly improved survival after Tamoxifen.
Conclusions: Our 380-gene classifier is capable of providing comprehensive
classification of breast cancer patients. This test could be used in the clinical setting
for identification of the major molecular classes of breast cancer and further subclassification of luminal patients into 4 classes with different rates of recurrence. This
will allow for the early identification of patients that would not respond to Tamoxifen
therapy.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S19

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

Tuesday, July 29, 2014

Poster Session: 9:30 AM - 5:00 PM
Clinical Studies/Outcomes

A-068
Cystatin C as a marker of renal function in adult Nigerians with Hypertension
and Diabetes

A. O. Tomi-Olugbodi1, O. O. Oladipo2, A. O. Olugbodi3. 1Department of

Pathology, Drexel College Of Medicine, Hahnemann University Hospital,,
Philadelphia,, PA, 2Clinical Chemistry Department of Pathology/
Laboratory Medicine, Staten Island University, NY, 3Department of
Internal Medicine,Reading Hospital and Medical Center,, Reading, PA
Background/Objective: Cystatin C (Cys C) has been suggested to be a better marker
of the glomerular filtration rate (GFR) compared to the widely used serum creatinine.
The aim of this study was to compare the accuracy of Cys C with that of serum
creatinine in the assessment of GFR in patients with hypertension and diabetics.
Method: Twenty hypertensives and twenty diabetics were compared with forty agematched healthy controls. Serum Cys C and serum creatinine were estimated in all
study subjects and compared with the actual GFR as estimated by the Cockcroft and
Gaults algorithm. The strength of significance of correlation was assessed using
Pearson correlation (a p value of< 0.05 was accepted as significant).
Results: The serum Cys C correlated well with serum creatinine (R = 0.757, p<0.01)
and with the estimated GFR (R = -0.733 and - 0.710 respectively; p<0.01). However,
in the ROC analysis, the AUC of serum cystatin C (0.727) was found to be superior to
that of plasma creatinine (0.539).
Conclusions: Serum Cys C and serum creatinine were well correlated in evaluating
GFR in hypertensive and diabetic patients. However, serum Cys C was more closely
correlated with the GFR and may therefore be a more accurate test of renal function in
hypertensive and diabetic patients.

A-069
The best markers for systemic inflammatory response syndrome (SIRS) criteria

K. Saito1, N. Nomura1, K. Ishizuka1, K. Yoshioka1, S. Yuasa1, T. Inaba2,

M. Nakanishi3, N. Fujita2. 1HORIBA Ltd., Tokyo, Japan, 2Department of
Infection Control and Laboratory Medicine, Kyoto Prefectural University
of Medicine, Tokyo, Japan, 3Department of Medical Instrumental Research
and Technology, Kyoto Prefectural University of Medicine, Tokyo, Japan
Background: The present study was undertaken to find out and compare the
usefulness of C-reactive protein (CRP) and hematology parameters in patients with
SIRS. Patient presenting systemic inflammatory response syndrome (SIRS) exhibit
two or more of the following criteria: Temperature greater than 38oC or less than 36oC,
Heart Rate greater than 90 beats/minute, Respiratory rate greater than or equal to 20
breaths/minute or PaCO2 less than or equal to 32 mmHg, WBC greater than or equal
to 12,000/L or less than or equal to 4,000/L, and suspected or proven infection.
Some of these patients were also under treatment of antineoplastic drug. This type of
treatment induces bone marrow suppression and therefore influences the immunity
response to inflammation and infection. Therefore we examined WBC, LYM#, NEU#,
ALY#, LIC#, PLT and CRP as biomarkers in patients with combination of SIRS and/
or antineoplastic drugs in order to evaluate if the bone marrow suppression could
affect the predictive value of the different parameters and analyze which would be
the most valuable
Methods: We examined 51 patients that we classified into distinct 4 groups depending
on the combination of presence or not of SIRS and Antineoplastic drug medication:
5 patients with SIRS(+) and Antineoplastic drug medication (+);
14 patients with SIRS(+) and Antineoplastic drug medication (-).
13 patients with SIRS(-) and Antineoplastic drug medication (+).
18 patients with SIRS(-) and Antineoplastic drug medication (-).The area under the
receiver operator characteristic curve (AUROC) was determined with 4 groups.
Whole blood patient samples were tested for WBC, LYM#,NEU#,ALY#,LIC#,PLT
and CRP using Pentra MS CRP(HORIBA Lid.)

S20

Results: The AUROC value by 14 patients with SIRS(+) and Antineoplastic drug
medication (-) and 18 patients with SIRS(-) and Antineoplastic drug medication (-)
were 0.81 for CRP, 0.65 for PLT and 0.61 for WBC. And also the AUROC value
by 5 patients with SIRS(+) and Antineoplastic drug medication (+) and 13 patients
with SIRS(-) and Antineoplastic drug medication (+) were 0.82 for PLT, 0.8 for CRP
and 0.77 for WBC. The AUROC of both PLT and CRP was 0.87. The AUROC of
both WBC and CRP was 0.80.The ALY# and LIC# did provide useful information
for evaluating that the patient had SIRS or not. The LYM# and NEU# had almost the
same performance as WBC.
Conclusion: Some patients with SIRS(-) based on SIRS criteria also had inflammatory
disease such as cholecystitis and otitis media. We evaluated patients with SIRS(-)
by CRP. The results show, patients with SIRS(-) and CRP(>5mg/L) and including
inflammatory disease such as cholecystitis and otitis media and patients with SIRS(-)
and CRP(<5mg/L) without inflammatory disease. This shows that CRP is an excellent
marker for SIRS criteria. Since WBC is affected by bone marrow suppression
of Antineoplastic drug medication and CRP is not affected by it. The AUROC
combination between CRP and PLT is better than the AUROC combination between
CRP and WBC. Therefore, our conclusion is that CRP and PLT provide better results
as supportive markers for SIRS criteria.

A-070
A case of Glutaric Aciduria type 1 in a Black South African girl

M. J. Turzyniecka. IALCH-NHLS and UKZN, Durban, South Africa

Background: Glutaric Aciduria type 1 is an autosomal recessive inborn error
of metabolism caused by a deficiency of glutaryl-Co enzyme A dehydrogenase
(GCDH). If unrecognised it results in a severe extrapyramidal disease
and mental retardation. It has been well described in North American and
European Caucasian populations but only recently identified as potentially
one of the commonest inherited metabolic disorders in black South Africans.
Methods and Results: We present a case of a 9 month old Black South African girl who
was admitted for investigations of sudden hypotonia following a fall from her mothers
back. There was no family medical history of note. She was delivered at term by normal
vaginal delivery and had normal developmental milestones until the fall. She had no
previous medical history. She had relative macrocephaly (head circumference 46cm
- 97th centile) and was globally hypotonic with hyperreflexia. She had normal renal
and liver function tests, lactate, ammonia, amino acid profile and negative reducing
substances screen. She had high urine glutaric acid at 4061umol/mmol creatinine
(reference: <30) and 3-OH glutaric acid at 49 umol/mmol creatinine (reference: < 2.5).
MRI of brain showed T2 bilateral hyperintensity of the basal ganglia, cerebral peduncles
and central tegmental tracts, symmetrical restricted diffusion of globus pallidus
in keeping with acute on chronic changes of pre-existing Glutaric aciduria type 1.
Conclusion: Macrocephaly is an important clinical sign of many neurometabolic
disorders and one of the earliest signs of Glutaric Aciduria type 1 (GA type 1)
thus warrants screening in all cases of macrocephaly of unknown origin regardless
ethnicity. GA type 1 typically presents with acute cerebral injury precipitated by an
intercurrent infectious illness. It is a preventable and treatable disorder requiring
patients with GA type 1 to low protein (lysine and tryptophan) diet, riboflavin and
L-carnitine supplementation. It has been linked to A293T mutation in the glutarylCoA dehydrogenase gene in black South Africans.

A-071
Assessment Of Serum And Urine Sialic Acid In Sickle Cell Anaemia Patients

s. E. idogun, E. Oriseseyigbemion-Sakpa. university of Benin Teach Hospit,

BENIN, Nigeria
Aim and objectives: To assess serum and urine sialic acid and compare their values
with other biochemical markers of nephropathy in patients with sickle cell anaemia.
Patients and methods: It was a cross sectional descriptive study that involved adult
(18 60 years) sickle cell anaemia patients that were clinically stable without sickle
cell anaemia crisis. The control participants were healthy volunteers that were agedmatched. Urine and blood specimens were obtained from the participants. The urine
albumin analysis was by Lowry method, urine and plasma creatine was by Jaffe,
Kinetic method, while the serum and urine sialic acid was by standard Ehrlich (P
dimethyl amino benzaldehyde) method. Data were analysed using SPSS version 16.
Results Sixty eight of the subjects were sickle cell anaemia patients while 30
respondents served as control. Table 1, shows the plasma and urine results of the
studied analytes in the sickle cell anaemia patients and controls. There was a negative

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

correlation between serum sialic acid and urea (r = -0.15, P = 0.41), creatinine (r -0.17,
P = 0.34) albumin creatinine ratio (r = -0.19, P =0.81), but these were not significant.
However, there was a positive significant relationship between urine sialic acid and
albumin creatinine ratio (r = 0.44, P = 0.01).
Conclusion: Serum sialic acid unlike urine sialic acid does not correlate well with
other well established marker of nephropathy (creatinine and urea). Monitoring
of urinary sialic acid in patients with sickle cell anaemia who are in steady state
is therefore an important adjuvant test in detecting the early onset of sickle cell
nephropathy.
Table 1: Plasma and urine results of the studied analytes in control (non SCA)
and SCA patients
Control (nonSCA anaemia)
respondents
(n = 30)
Plasma urea (mmol/L)
3.330.16
Plasma creatinine (mol/L) 73.982.95
SSA (mmol/L
1.930.67
ACR (mg/mmol)
2.190.10
USA (mmol/L)
0.780.04
USCR (mmol/mol)
60.523.39
ACR = Albumin creatinine ratio

SCA respondents (n
= 68)

p-value

6.430.54
86.706.03
1.580.96

<0.001
0.06
0.04

4.500.24
1.130.05
169.3913.59

<0.001
<0.001
<0.001

A-073
Risk map in three emergency laboratories

E. Gonzalez1, . Salas Garca2, E. Guilln Campuzano1, M. Buxeda

Figuerola1, L. Juan Pereira1, I. Caball Martn1. 1Catlab, Terrassa, Spain,
2
Consorci Sanitari de Terrassa, Terrassa, Spain
Background: Patient Safety is considered one of the key aspects of the quality
policies of Health Systems The objective is to perform an analysis of potential errors
in three emergency laboratories to estimate the impact on patient safety.
Methods: The Clinical Laboratory of Catlab consists of a central laboratory and
three hospital emergency laboratories. It is calculated an estimation of the potential
risks in each of the three emergency laboratories: Laboratory (1) 609794 clinical
analysis/2013, Laboratory (2) 424882 clinical analysis/2013 and Laboratory (3)
256857 clinical analysis/2013.
The Failure Mode and Effect Analysis (FMEA) is the methodology that analyzes
the quality, safety and reliability of the functioning of a system, identifying potential
errors and their causes from their effects. By using the FMEA, it is calculated the
risk priority number (NPR), which is the product of the estimation of the incidence,
gravity and detectability (I*G*D). The NPR will allow us to prioritize the importance
of the possible errors.

USA = Urine sialic acid

We analyze the possible potential errors that can arise in the different processes that
are developed in the laboratories, according to the item that is affected, the causes
and their effects.

USCR = Urine sialic and creatinine ratio

SCA = Sickle cell anaemia

Results:

A-072
Interleukin-18 a Potential Marker of Liver Cirrhosis in Chronic Hepatitis C
Patients.

E. M. Abdalla, S. E. Younes, M. M. Abdo, S. A. M. Ahmed. Suez Canal

University, Ismailia, Egypt
Background: WHO reported that Egypt has a very high prevalence of HCV and a
high morbidity and mortality from chronic liver disease, cirrhosis, and hepatocellular
carcinoma. In patients with liver cirrhosis, liver biopsy is the gold standard method
to establish the diagnosis. But liver biopsy has many disadvantages, it is invasive,
coasty and difficult to be standardized. That is why there has been increasing interest
in noninvasive assessment of liver cirrhosis by the use of alternative serum markers.
IL-18, a proinflammatory cytokine can be synthesized by injured hepatocytes and
increases the susceptibility of liver endothelial cells to undergo apoptosis. An
increased circulating level of IL-18 is likely to play a pathogenic role in patients with
chronic liver disease.
Aim of the work: This study aimed to prove that Interleukin-18 can be used as a
potential marker of liver cirrhosis in chronic hepatitis C patients. IL-18 levels were
measured in 20 HCV infected patients with cirrhosis and compared to 20 non-cirrhotic
HCV patients and 20 healthy controls.
Results: Serum IL-18 levels were significantly higher in cirrhotic (836 pg/ml) and
non-cirrhotic CLD patients (751 pg/ml) than in healthy controls (278 pg/ml). IL-18
level is significantly increased with increase in histological stage of liver cirrhosis
and its concentration may predict the degree of hepatocellular damage. A Positive
correlation founded
between serum IL-18 levels and Child Pugh score suggest that IL-18 can be used as an
additional non-invasive marker for monitoring the degree of liver cirrhosis in chronic
hepatitis C patients and as a monitoring tool to assess response to therapy. Low cut off
value of serum IL-18 at 495 pg/ml (which showed sensitivity of 100%) can be used
as a screening value under which most of cases are probably negative cases for liver
cirrhosis. While, a high cut off value of serum IL-18 at 875 pg/ml (which showed
specificity of 95%) can be used as a diagnostic value above which 95% of cases are
probably positive cases for liver cirrhosis.
Conclusions: IL-18 level as a non-invasive marker can be used for
follow up of chronic HCV patients and assessement of the severity of the disease and
degree of liver cirrhosis instead of liver biopsy which has been proved to be invasive,
coasty and difficult to standardize. Usage of low cut off value of serum IL-18 at 495
pg/ml as a screening value under which most of CLD cases are probably negative for
liver cirrhosis and no need for liver biopsy. Usage of high cut off value of serum IL-18
at 875 pg/ml as a confirmatory test with liver biopsy above which 95% of cases are
probably positive for liver cirrhosis. Usage of IL-18 level between 495 and 875 pg/ml
as an indication for liver biopsy to diagnose cases with liver cirrhosis.

Processes
Laboratory(1)
Laboratory(2)
Laboratory(3)
Strategics
27%
17%
31%
Support
10%
6%
10%
Pre-preanalytical
17%
23%
19%
Preanalytical
17%
24%
10%
Analytical
8%
14%
8%
Postanalytical
12%
9%
19%
Post-postanalytical 9%
7%
8%
Conclusion: There are differences between the pre-analytical errors in Laboratory
(2) and the Laboratories (1) and (3). The Laboratory (3) makes a greater estimation of
the error in the post-analytical processes. It is relevant the importance of the results
of the strategic processes in the Laboratories, especially in the Laboratories 1 and 3.
The results obtained using the FMEA will allow us to implement a posteriori
improvement actions.

A-077
Value of suspecting undiagnosed G6PD deficiency in very severe
hyperbilirubinemia with hepatitis A.

A. Hazra, J. Chakraborty, L. Banerjee. ESIC PGIMSR Manicktala Kolkata,

Kolkata WB India, India
Background: India has a high burden of both hepatitis A and G6PD deficiency
(upto 15%) with no Government funded screening for G6PD. In adults only 2% of
infectious hepatitis cases cross a bilirubin value of 30 mg/dL, and we present a case
with total Bilirubin more than 60 mg/dL.
History: A 16 year old Indian boy who presented with 7 day history of fever, nausea,
vomiting, malaise, anorexia, abdominal pain and yellow urine. On examination he
showed deep icterus, and hepatomegaly. Total Bilirubin at presentation was 17.26 mg/
dL of which about 60% was conjugated. AST was 7300 IU/L and ALT 6700 IU/L and
ALP 640 IU/L. HEV and HAV IgM came back strongly reactive. His Hemoglobin was
already 6.3 g/dL when initially checked after admission and two days after it dropped
further to 5.3 g/dL and his Bilirubin shot up to 66.95 total of which about 70% was
conjugated. next day the transaminases were coming down, with AST 745 mg/dL,
ALT 1751 mg/dL, ALP was down to 350 mg/dL. LDH which was not checked initially
turned out to be 1364 U/L this time, and Reticulocytes 19%. Putting together the
dropping Hb, very severe Bilirubin, and high LDH and Reticulocytosis, an additional
hemolytic etiology was suspected. DAT and a G6PD screen were added on in the
pre-transfusion sample. DAT came out negative but the G6PD screen was strongly
positive for deficiency and a quantitative spectrophotometric G6PD test fetched a
level ~30% of lower reference rage even at this hemolytic phase. The patient was also
screened for Wilsons disease by Serum Ceruloplasmin and 24 hr Urine copper which
all came negative. The patient gradually recovered with transfusion and supportive
therapy and his diagnosis of G6PD in addition to HepatitisA.
Conclusion: G6PD deficiency should be suspected in subjects with a history
of hemolysis with Hepattis A, even when the majority of fraction of bilirubin is

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S21

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

conjugated . We also suggest to vaccinate all G6PD deficient patients for hepatitis
A. Whether to screen blood bag for G6PD before transfusion to Hepatitis patients is
of debate. Note Hepatocye damage would be aggraveted due to hepatocellular G6PD
deficiency since the deficiencys is not isolated to only Erythroid tissue.

A-079
Neutrophil Gelatinase Associated Lipocalin and atherosclerosis: a study of its
association with known cardiovascular risk factors and metabolic syndrome.

M. C. Freire1, I. Bendet1, M. L. Moreira1, L. L. Leite1, A. J. L. Jorge2, M. G.

Rosa2, E. T. Mesquita2. 1DASA, RIO DE JANEIRO, Brazil, 2Universidade
Federal Fluminense, Niteri, Brazil
Background: The Neutrophil Gelatinase Associated Lipocalin (NGAL) is a
glycoprotein involved in the processes of innate immunity and inflammation. Initially
described in neutrophils and epithelial cells, it is also produced by macrophages
and smooth muscle cells in atherosclerotic lesions. It activates proteolytic processes
linked to metalloproteinase 9 (MMP-9), considered to be responsible for vascular
remodeling and rupture of atherosclerotic plaque, leading to clinical presentations of
cardiovascular diseases (CVD).
Objectives: To evaluate the associations of NGAL with parameters known as
cardiovascular risk factors and criteria for metabolic syndrome.
Methods: Observational, cross-sectional study of a sample (n = 219) of outpatiens
assisted by the Family Medicine Program in the city of Niteri, Brazil, aged 45 years
and older. All data and samples were collected between August 2011 and August 2012.
NGAL serum levels were determined by sandwich ELISA (Bioporto Diagnostics),
samples were frozen at -80C until first thawed. We evaluated its associations with
several risk factors.
Results: Of the parameters tested, we observed significant positive correlation
between NGAL means and age, creatinine, leukocyte count, Framingham estimated
risk, and a negative correlation with HDL cholesterol (Table 1).
Discussion: To stablish preventive strategies for CVD is one important goal of health
systems, and it depends on accurate individual risk assessment. There are consistent
data from studies in animals and humans that support the importance of NGAL in
the patho-physiology of atherosclerosis and CVD. The biological effect of NGAL in
atherosclerosis is suggested by the co-localization of NGAL and MMP-9 expression
in human carotid atherosclerotic plaques with increased gelatinase activity in these
tissues. The quest for understanding the role of this new biomarker in predicting
cardiovascular events is of paramount importance, and at this point of the study we
could show its association with HDL and to Framingham risk estimates.

of the molecule glycol-phosphatidyl-inositol (GPI) and partial or inability to express

GPI-anchored proteins. Conventional assays to diagnose PNH present high specificity,
but low sensibility. Flow Citometry has been used to evaluate the expression of GPIanchored proteins in different blood lineages, because it is a highly specific and
sensitive technique, and it is the gold standard for PNH management and diagnose.
Objective: standardize and validate the PNH immunophenotyping assay using
the reagent FLAER (fluorescent aerolysin) for implementation in the routine of a
diagnostic laboratory.
Methodology: A total of 80 peripheral blood samples within 24 hours collection
from population with ages between 20 and 40 years old were used. Two protocols
of surface markers were tested and all the monoclonal antibodies were titrated
(Flaer, CD55, CD64, CD24, CD14, CD16, CD45). The acquisition was performed
on BD FACSCanto II (BD, San Jose, CA) and the data was analyzed in the Infinicyt
(Cytognos S.L., Salamanca -Spain) software.
Results: In the study population, were obtained seven PNH positive samples (9%)
and 73 PNH negatives (91%). In all cases the FLAER reagent presented excellent
performance in detection of neutrophils and monocytes with PNH phenotype, and
there was no conflict between the FLAER staining and GPI-anchored proteins
expression. The possibility to differentiate cells type I, II and III in erythrocyte
population, when the MEM43 clone from the monoclonal antibody CD59-PE is used
was also observed (FI = 2589. 62 negative samples and FI= 954,72 positive samples).
Conclusion: The evaluation of the FLAER reagent performance in detection of PNH
clone through Flow Citometry, in this study, presented clear and precise distinction
between normal cells and deficiency GPI cells, corroborating with the current
literature.

A-081
Effects of Vitamin D Treatments on Arginine Derivatives in patients with Stage
3 and 4 Chronic Kidney Disease

C. Heideloff, J. M. El-Khoury, J. Hyland, J. Simon, S. Wang. Cleveland

Clinic, Cleveland, OH
Background: Vitamin D nutritional status has been linked to chronic kidney disease
(CKD) and cardiovascular disease (CVD) in epidemiologic studies. Arginine
derivatives, especially symmetric dimethylarginine (SDMA) and asymmetric
dimethylarginine (ADMA) have been associated with CKD and CVD. However,
data available in literature regarding the relation between vitamin D blood levels and
arginine derivatives is conflicting. Objective: In this prospective study, our objective
was to examine if vitamin D supplementation and increase in 25-hydroxyvitamin D
blood levels significantly change arginine, ADMA, SDMA, and the ratios of these
biomarkers. Methods: This study was approved by our Institutional Review Board.
A total of 16 CKD patients (stage 3 and 4) with vitamin D deficiency (<20 ng/mL)
and insufficiency (<31 ng/mL) were enrolled in our study. Patients were instructed to
take either vitamin D2 or vitamin D3 at 50,000 IU/month for 6 months if they were
vitamin D insufficient or 50,000 IU/week for a month then 50,000 IU/month for 5
months if they were vitamin D deficient. 25-Hydroxyvitamin D2, 25-Hydroxyvitamin
D3, arginine (ARG), ADMA, and SDMA were measured at baseline and after 24
weeks of supplementation. All analytes were determined by liquid chromatographytandem mass spectrometry methods. Results: Vitamin D supplementation increased
25-hydroxyvitamin D levels from a mean of 24.8 ng/mL at baseline to 39.6 ng/mL
after 24 weeks of treatment (p<0.0001). Arginine, SDMA, ADMA and their ratios
were not significantly different before and after the treatments (p>0.05). Conclusion:
Though vitamin D supplementation in stage 3 and 4 CKD patients significantly
improved vitamin D status it did not significantly affect arginine, ADMA or SDMA
blood levels.

A-084
A study of the parasitic causes of anaemia in children under five(5) years of age
in the Bolgatanga municipality, Ghana

A-080
Standardization of paroxysmal nocturnal hemoglobinuria assay

L. B. Sousa1, M. A. Viana1, R. L. Albuquerque1, E. X. Souto1, M. D. Freire2,

N. Gaburo1. 1DASA, Sao Paulo, Brazil, 2DASA, Rio de Janeiro, Brazil
Introduction: Paroxysmal Nocturnal Hemoglobinuria (PNH) is caused by clonal
expansion of hematopoietic stem cells that present somatic mutations of the
phosphatidylinositol-glycan-class A (PIG-A), resulting of biosynthesis deficiency

S22

C. NKRUMAH1, S. ALHASSAN ANEYIRE2. 1METHODIST HOSPITAL,

WENCHI, Ghana, 2COLLEGE OF HEALTH, KINTAMPO, Ghana
Background: Anemia is a condition affecting children especially those less than 5
years, and pregnant women in the world (WHO, 2000). In tropical and developing
countries, anemia is particularly prevalent with 50% or more of pre-school children
and pregnant women being moderately or severely anemic (Cheesbrough, 2006). It is
against this background that this research is being conducted.
Methods: This study represents the results of a descriptive cross-sectional study to
establish the prevalence, magnitude and causes of anemia among children less than

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

5 years in the Bolgatanga municipality. Simple random sampling technique was used
to obtain specimen from 100 participants following informed consent. Specimens
analyzed included stool (wet preparation with physiological saline) for intestinal
parasites and blood samples for hemoglobin measurement and detection of the
presence of malaria parasites using Rapid Diagnostic Technique (RDT) kits. Actual
measurement of their hemoglobin (Hb) was done using a Sysmex analyzer. Anemia
was defined as haemoglobin concentration less than 11.0g/dl (Cheesbrough, 2006).
Results: The prevalence of anemia among the children was 62%. Of the anemic
children, 35 (56%) showed presence of malaria parasites in their bloodstream whiles
18(29%) had intestinal parasites. 42(66%) of the anemic children had Hb between
10.0 and 11.0g/dl (mildly anemic), 13.0 (23%) had Hb between 7.0 and 9.9g/dl
(moderately anemic), with 7 (11%) of them being severely anemic (Hb below 7g/
dl). All the parasites detected in the stool samples were hookworm. The presence of
these intestinal parasites and malaria parasites in the stool and blood respectively had
a significant association with low Hb (p<0.005). Conclusion: Based on the results
of the study, the prevalence of anemia was high in the study population. Malaria
parasitaemia was a major cause of anemia in our environment followed by hookworm
infestation. The study recommends among other things that the Ministry of Health
should embark on intensive health education programs in the communities on early
detection of malaria and hookworm infestation and enforce their prevention.

A-086
Fecal Biomarker Testing Identifies Exocrine Pancreatic Insufficiency in Patients
with Possible Irritable Bowel Syndrome

J. G. Goepp, T. McBride, E. Fowler, A. Peace-Brewer, D. Landis. Genova

Diagnostics, Asheville, NC
Objective: 1) To use fecal biomarker testing to identify a subset of patients with
symptoms of irritable bowel syndrome (IBS) that might be explained by the presence
of exocrine pancreatic insufficiency (EPI), and 2) to identify within that subgroup
additional fecal biomarkers that might suggest a primary disease process capable of
causing secondary EPI.
Relevance: IBS, long considered a diagnosis of exclusion, is now recognized as
an umbrella diagnosis, identifying a heterogeneous group of patients who may in
fact have one or more of several underlying, treatable diagnoses. EPI is one such
diagnosis, but few studies have examined its prevalence or biological context in the
setting of possible IBS. Because EPI may occur secondary to another gastrointestinal
problem such as inflammation or parasitic infection, fecal biomarkers suggesting
such processes may further clarify the diagnostic picture in patients presenting with
symptoms consistent with IBS. A fecal biomarker panel with appropriate components
may thus represent an important role for clinical laboratory medicine in diagnosis and
management of conditions producing symptoms consistent with IBS.
Methods: One-year analysis of de-identified stool testing data for fecal pancreatic
elastase-1 (FPE1) in all patients for whom stool specimens were submitted for testing
bearing at least one of 13 ICD-9 codes indicating the potential for IBS. FPE1 testing
was conducted using a standard dual monoclonal antibody ELISA technique with
binding sites for two unique epitopes on the human PE-1 molecule. When biomarker
data on other conditions were available on the same specimen, they were analyzed for
significant associations with FPE1 status using Fishers Exact Test, with significance
set at p < 0.05.
Validation: A total of 9112 records were identified that included at least one of the
13 IBS-related ICD-9 codes and also a result of FPE1 testing. FPE1 in the range 200
mcg/g defined normal. Definitions of abnormal for other biomarkers were: calprotectin
>120 mcg/g, eosinophil protein X (EPX) > 7 mcg/g, H. pylori, B. hominis, and other
parasites = present.
Results: Objective 1: Some degree of EPI was suggested in 858 (9.4%) of all records,
with FPE1 <100 mcg/g stool in 289 (3.2%) and FPE1 100 to 199 mcg/g in 569 (6.2%).
Objective 2: FPE1 of <100 was significantly associated with abnormal results for
fecal calprotectin and EPX, as well as with evidence of infection with Blastocystis
hominis, non-Blastocystis parasites, H. pylori, and entamoeba. FPE1 of 100-199 was
significantly associated with parasitic infection as well.
Conclusions: Among patients in whom IBS was a consideration, fecal biomarker
testing revealed that 9.4% had indications of some degree of exocrine pancreatic
insufficiency by FPE1. Fecal biomarker testing was also more likely to suggest
presence of an inflammatory or parasitic gastrointestinal condition in patients with
abnormal FPE1, suggesting that in some cases EPI itself might be secondary to such
underlying processes, and indicating potential lines for further evaluation of possibly
treatable diagnoses. Further prospective studies with definitive patient follow-up are
recommended.

A-087
Association between the Delta Estimated Glomerular Filtration Rate and the
Prevalence of Monoclonal Gammopathy of Undetermined Significance in
Korean Males

T. Jeong, W. Lee, S. Chun, W. Min. University of Ulsan College of Medicine

and Asan Medical Center, Seoul, Korea, Republic of
Background: The prevalence of monoclonal gammopathy of undetermined
significance (MGUS) varies with age, gender, and ethnicity. We investigated the
association between the reduction in the estimated glomerular filtration rate (eGFR)
and the prevalence of MGUS in healthy Korean males.
Methods: We enrolled 723 healthy Korean males who visited the hospital for regular
health checkups. Serum creatinine concentration, serum electrophoresis, serum
immunofixation, and the serum free light chain assay were performed. Data, including
age, date of health checkup, and previous serum creatinine concentrations were
obtained from electronic medical records. We calculated delta eGFR per year and the
prevalence of MGUS was compared based on the delta eGFR per year and age group.
Results: Thirteen (1.8%) of seven hundred and twenty-three participants exhibited
the monoclonal band on serum immunofixation. Prevalence of MGUS by age group
was 0.00% (0/172 for 40s), 1.63% (6/367 for 60s), and 3.80% (7/184 for > 60-years).
The median decrease in delta eGFR per year was 5.3%. The prevalence of MGUS in
participants in their 50s with > 5.3% decline in delta eGFR per year was significantly
higher than those with < 5.3% decrease in delta eGFR per year (3.16% vs. 0.00%; p =
0.049). The prevalence of MGUS in participants in their 50s with > 5.3% decrease in
delta eGFR per year was similar to that of healthy males in their 60s.
Conclusion: Using the rate of reduction in delta eGFR per year in healthy Korean
males who had their serum creatinine level checked regularly may increase the MGUS
detection rate in clinical practice.

A-090
An Update on a Candidate BK Virus DNA Standard Reference Material

J. Harenza1, L. Cook2, P. M. Vallone1. 1NIST, Gaithersburg, MD, 2University

of Washington, Seattle, WA
Background: The polyomavirus, BKV, is widespread in the population due to primary
infection during childhood and largely remains latent in the majority of individuals.
However, illnesses as a result of BKV occur in immunocompromised patients, organ
transplant, HIV/AIDS- infected, or diabetic patients. Two common illnesses resulting
from BKV, hemorrhagic cystitis and nephropathy, occur in transplant patients, as
prescribed anti-rejection medications can weaken the immune system. This often
leads to renal allograft failure. To treat BKV, accurate quantitation of viral load is
necessary for proper adjustment of anti-rejection medications in an effort to improve
an individuals immune function. However, current BKV standards are only available
for the Ia genotype of the virus, and it has been well-documented that using assays
specific to Ia can lead to underestimation of viral load of other genotypes, due to
primer and/or probe binding site polymorphisms. Therefore, we propose a candidate
reference material panel of BKV genotypes to aid in precise measurements of viral
load.
Methods: Viral DNA was extracted from six available clinical isolates and distinct
genotypes (Ia, Ic, III, IV, V, VI) of the BK virus. Single site restriction digestion
was performed to linearize each viral genome and long-range PCR was performed
to amplify entire genomes. Each amplicon was ligated into a plasmid backbone,
transformed into, and subsequently propagated in, XL-10 Gold E. coli cells. DNA was
extracted, purified, and re-linearized to enable sequencing and accurate copy number
measurement. Using the Illumina MiSeq platform, whole genome sequencing was
performed for each genotype. New primer and probe sets, specific to each genotype,
were designed using this sequence information and optimized for quantitative PCR
(qPCR) and digital PCR (dPCR).
Results and Future Goals: We have amplified and sequenced six full-length
BKV genomes. Next-generation sequencing information allowed us to design and
optimize efficient (between 90-110%) primer and probe sets for both qPCR and
dPCR platforms. Homogeneity and stability studies will be performed for each DNA
construct using the newly developed assays, and the materials will be certified for
genome copy number using dPCR.
Conclusion: The availability of a panel of BKV DNA genotypes as potential reference
materials will enable traceability for manufacturers of calibrant materials. This will
ultimately improve upon the consistency and accuracy of viral load quantitation
measurements across laboratories, leading to improved dosing regimens in infected
patients.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S23

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

particles); the report was used to assess its effectiveness for urine culture screening.

A-092
Can Procalcitonin or Lactate help with Sepsis Evaluation in Cancer Patients?

D. R. Mendu, M. Fleisher, S. I. McCash, Q. Xiang, L. Miller, M. S. Pessin,

L. V. Ramanathan. Memorial Sloa-Kettering Cancer Center, New York, NY
Objective: Severity of illness due to sepsis in patients with cancer is a major clinical
complication and contributes to high mortality in this patient population. We
evaluated two biomarkers used to help discriminate between bacterial infection and
noninfectious acute-phase reactions: lactate (LA) and procalcitonin (PCT).
Methods: A cohort of 40 random patients with malignant disease admitted to
Memorial Sloan Kettering Cancer Center was studied. Out of these 40 patients, 10
were considered as controls using the sepsis diagnostic criteria of lactate < 2.0
mmol/l, and negative bacterial cultures: and PCT< 0.5 ng/mL per manufacturers
recommendations. Six out of the forty patients had viral infections. LA and PCT
were measured at the time of admission followed by an additional sample collection
for each patient within 24 hours. Heparinized blood samples were obtained and
the plasma used for LA analysis on the Nova pHOX and for PCT determined by
an Enzyme Linked Fluorescent Antibody analysis on the mini Vidas ( BioMrieuxS,
France). Concordance studies between LA and PCT were done.
Results: Our study demonstrated that elevated LA and PCT levels have only a 50%
concordance. PCT was elevated in 50% of the patients and demonstrated a lack
of concordance with plasma Lactate. Additionally, we observed that the PCT was
elevated in patients who were negative for bacterial infection, based on microbiological
cultures, and did not correlate well with LA concentrations >4 mmol/L. The six
patients with viral infection had significantly elevated PCT levels and the LA levels
were < 4 mmoL/L. In order to examine the possible effect of pre-analytical factors
that may explain the LA/PCT discordance, we analyzed LA stability in five healthy
volunteers at different time points at room temperature. Within 15 minutes, the LA
in heparinized blood increased steadily, and by 30 minutes, LA concentrations were
almost 40 % higher than the initial LA measurement. Conclusions: Our preliminary
findings suggest that PCT levels were consistently high in cancer patients without
sepsis. Our data also indicates a poor correlation between LA and PCT in cancer
patients evaluated for suspected bacterial infection and neutropenia. Increased
LA concentrations could be due to pre-analytical variation, oxygen deficit due to
mitochondrial oxidative phosphorylation, inhibition by interleukins (ILs), tumor
necrosis factor (TNF) or a combination of the latter. These data support the already
known poor specificity of LA as an accurate indicator of sepsis. Further clinical
investigation is also needed to explain elevations in PCT in the absence of bacterial
infection in patients with cancer.
Based on these data, the significance of PCT in cancer patients without infection is
unclear and requires further investigation. Our data also suggests a need to investigate
the relationship between plasma LA and PCT in the clinical evaluation of sepsis in
cancer patients.

A-095
Urinalysis/Microscopy Reflex to Urine Culture on i R I C E L L Complete
Urinalysis Workcells: Are We There Yet?

Results: 801 (49.0%) had positive culture. 1344 (82.3%) specimens were identified as
urine culture candidates. Of the 1,344 that were identified as a candidates, 772 (57.4%)
had positive culture, 547 (40.7%) were negative, 25 (1.9%) were contaminated. 290
(17.7%) specimens were not selected as culture candidates because of the negative
urinalysis; of these specimens 29 (10%) had positive culture and 259 (89.3%) were
negative.
Conclusion: using the candidate for urine culture would have avoided urine culture on
about 15% of the patients but would have missed the diagnosis for 10% of the patients
that were considered as having normal specimens. Urinalysis and microscopy still
lack the sensitivity and specificity to be used alone to diagnose urinary tract infection,
more work needed to establish better parameters and algorithm to be used when
screening patients for UTI using urinalysis.

A-096
Interleukin-6 (IL-6) Combined with Clinical and Demographic Parameters
Predicts Death in Critically Ill Patients with Systemic Inflammatory Response
Syndrome

J. M. Coln-Franco1, S. Litt2, D. A. Anderson2, D. Plummer2, W. D.

Dupont2, T. W. Rice2, A. P. Wheeler2, A. Woodworth2. 1Medical College
of Wisconsin, Milwaukee, WI, 2Vanderbilt University Medical Center,
Nashville, TN
Background: Increasing sepsis severity is associated with significant mortality.
Currently, there are few prognostic biomarkers for septic patients. Objective:
To determine the prognostic utility of 5 inflammatory biomarkers, clinical and
demographic data to predict mortality among Medical ICU (MICU) patients with
systemic inflammatory response syndrome (SIRS). Methods: This retrospective
cohort study enrolled 201 MICU patients with SIRS. Among them, 91 were septic
and 45 died during their admission. TNF, IL-6, IL-10, CRP and LBP were measured
on the Immulite 1000 (Siemens Healthcare Diagnostics) in residual plasma collected
on the day of SIRS and 1 or 2 days prior to meeting SIRS criteria (days -1 and -2).
Procalcitonin (PCT) was measured on the Vidas BRAHMS Assay (Biomeriux).
The prognostic strength of individual biomarkers and logistic regression models
that combined baseline co-variables (SIRS criteria, age, and sex; Model A) and
biomarker concentrations (Model B) were evaluated. The change in serial biomarker
concentrations (days -2, -1 and 0) and the difference in median concentrations for each
day were determined for survivors and non-survivors by two way ANOVA and MannWhitney tests respectively. Results: Table 1 lists the areas under the receiver operating
characteristic curves (AUCs) for individual biomarkers and models. The evolution
of IL-6 over time was significantly different in survivors and non-survivors (timesurvival interaction p=0.02). Median IL-6 concentrations were significantly different
in both groups during all study days (p<0.01, <0.001, and <0.0001 on days -2, -1, and
0 respectively). Conclusions: IL-6 was the best single predictor of mortality. PCT
showed limited prognostic ability. When combined with baseline demographics and
clinical data, IL-6 showed improved ability to predict death. Additional biomarkers
(model AB) did not improve the models prognostic strength. Serial measurement of
IL-6 combined with other clinical parameters accurately predicts mortality in patients
with systemic inflammatory response syndrome.

R. Khoury, A. Gandhi, B. P. Salmon, P. Gudaitis, D. Gudaitis. Aculabs, Inc.,

East Brunswick, NJ
Background: Urinary Tract Infection (UTI) is serious health problem, it is very
common among the geriatric population, it is the second most common type
of infection (second only to respiratory tract infection),and can cause serious
complications and it is a significant cause of morbidity and death, with the expected
death rate as high as 3% in those who develop pyelonephritis. The high mortality
rate is largely due to delayed presentation and the development of bacteremia/sepsis.
Urinalysis (UA) is one of the most frequently requested tests; it is estimated that
between 200 and 300 million UA are performed yearly in the United States. It is an
invaluable tool in the diagnosis of UTI, malignancy, calculi, and detecting systemic
diseases affecting the kidneys.
Methodology: 1,634 specimens were collected for urinalysis and culture from
residents in Long-Term Care Facilities over a period of 2 weeks. Urinalysis and
microscopy was done using iRICELL automated system. All the specimens were
cultured; the culture was done using MicroScan Walkaway96 conventional panels. We
used urine culture as the reference standard, no growth or <10,000 colony-forming
unit/mL were considered negative, cultures with > 50,000 colonies colony-forming
unit/mL were considered positive. The urinalysis instrument generates a urine
culture candidate report when at least one of the five bacteriuria parameter (Nitrite,
Leucocyte Estrase, and microscopy values: white blood cells, bacteria and all small

S24

A-098
Fatigue in rheumatoid arthritis and its relation to interleukin-6 serum level

d. I. hashad, A. Helal, E. Shahine, M. Hassan, R. Abdel Moneim. FACULTY

OF MEDICINE, alexandria, Egypt
Objectives: This work aimed to investigate the occurrence and level of fatigue and
its relation to IL-6 serum level in rheumatoid arthritis (RA) patients. Patients and
methods: The study included 60 patients with RA diagnosed according to the 2010
ACR-EULAR classification criteria for RA and 30 healthy controls. Patients were
included if they were above 18 years and fulfilled a score P6 over 10 of the 2010
ACR-EULAR classification criteria for RA. Disease activity was assessed using
28 joint disease activity score (DAS28), erythrocytes sedimentation rate (ESR)
and C-reactive protein (CRP). Fatigue was assessed using the Bristol Rheumatoid
Arthritis Fatigue Multidimensional Questionnaire (BRAF-MDQ) and serum IL-6
level was measured in patients and controls using enzyme-linked immunosorbent
assay (ELISA) technique. Results: The BRAF-MDQ was significantly higher among
patients (mean =50.6 15.2) than controls (mean= 7.8 3.7) (p< 0.001). Patients
mean IL-6 serum level was 35.05 21.23 pg/ml and 4.72 3.09 pg/ml among control
subjects (p< 0.001). DAS 28 ranged between 4.33 and 7.67. Mean 1st hour ESR was

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

43.57 mm and CRP was positive in 76.7% of patients. Significant correlations were
found between BRAF-MDQ score and serum IL-6 level (r = 0.947, p <0.001), ESR
(r =0.509, p< 0.001) as well as CRP positivity (r =0.411, p=0.005) in RA patients.
Serum IL-6 level correlated with ESR (r = 0.463, p< 0.001) and CRP (r =0.376, p=
0.01) among patients. Conclusion: Fatigue is a common symptom and scores higher
among RA patients than healthy controls and should be measured in all RA patients
with simple fatigue questionnaires matching with different cultures. Fatigue becomes
more prominent as serum IL-6 level increases independently of the disease duration
and activity.

A-099
Ferritin in sera and CSF: Its importance as both predictive and etio-diagnostic
biomarker in ischemic stroke, single center prospective study

The study extends the spectrum of exon 3 mutations in the AR gene.

H. M. Demerdash , O. Mansour , M. Megahed , T. Zytoun . pharos

University, Alexanddria ,Egypt, Alexandria, Egypt, 2Faculty of medicine ,
Alexanddria ,Egypt, Alexandria, Egypt, 3Faculty of medicine , Alexandria
,Egypt, Alexandria, Egypt
1

Background:Iron and ferritin are known to have an important role in stroke as well
as in other disorders. This prospective study was designed to determine whether
determination of CSF ferritin levels might help to estimate the severity and prognosis
of stroke.Methods:Thirty-two patients with a diagnosis of acute stroke due intrinsic
atherosclerotic vessel pathology were included in the study within 24 h from onset
of symptoms. Plasma and CSF ferritin were assayed; and correlated them with the
known marker A1-42 at admission. Clinical status was determined by the Canadian
Stroke Scale at admission and on day 21.
Results:Serum ferritin level was found to be higher in patients with large lesion size
(P < 0.01), deteriorated neurologic status during clinical follow-up (P = 0.03) and
ICAD stroke patients (P < 0.01). CSF and Serum ferritin level were correlated with
neurologic deficit (r = 0.50, P < 0.001). No correlation was found between A1-42 and
ferritin levels (r = 0.07, P = 0.7). Serum ferritin level (P = 0.007; OR = 1.02; 95% CI,
1.01-1.03) and large size of lesion (P = 0.021, OR = 11.92; 95% CI; 1.46-197.12) were
independently associated with stroke due to ICAD pathology, Increased serum ferritin
levels correlate to severity of stroke and the size of the lesion.
Conclusion:our results supported that a raised level of CSF and serum ferritin may
imply a poor prognosis in terms of neurologic deterioration or ICAD induced stroke
patients.

Figure 1. DNA sequence of exon 3 of the androgen receptor(AR) of the patient(A) and
the normal sequence (B). The mutant sequence shows conversion of arginine (AGG)
to tryptophan (TGG).

A-101
Association Between Serum Uric Acid Levels and Non-Alcoholic Fatty Liver
Disease

E. Sertolu1, C. N. Ercin2, G. Celebi2, H. Gurel2, H. Kayadibi3, H.

Genc4, M. Kara5, T. Dogru2. 1Ankara Mevki Military Hospital, Anittepe
Dispensary, Biochemistry Laboratory, Ankara, Turkey, 2Gulhane School
of Medicine, Department of Gastroenterology, Ankara, Turkey, 3Adana
Military Hospital, Department of Medical Biochemistry, Adana, Turkey,
4
Izmir Military Hospital, Department of Gastroenterology, Izmir, Turkey,
5
Etimesgut Military Hospital, Department of Gastroenterology, Ankara,
Turkey
Background: Non-alcoholic fatty liver disease (NAFLD) represents a spectrum of
conditions ranging from simple steatosis (SS) to non-alcoholic steatohepatitis (NASH)
and cirrhosis. It has become one of the most common forms of chronic liver disease
affecting 20-30% of the Western population. Elevated serum uric acid (SUA) levels
have been suggested to be an independent risk factor for the development of NAFLD
in healthy adults. In this study, we aimed to investigate the relationship between SUA
and liver histology in non-diabetic and normotensive patients with NAFLD.
Materials and Methods: A total of 242 male patients, 102 with non-alcoholic
steatohepatitis (NASH) and 140 with non-NASH were included. The histopathological
examination of the cases were carried out according to Kleiner NAFLD activity
score (NAS) combining steatosis, lobular inflammation and ballooning hepatocyte
degeneration. Hyperuricemia was diagnosed as SUA more than 7 mg/dL.

A-100
A Novel Missense Mutation in the Androgen Receptor Gene Causes the
Complete Androgen Insensitivity Syndrome

X. Liu1, Z. Cai2, L. Xie3, L. Ji1. 1Peking Unversity Shenzhen Hospital,

Shenzhen, China, 2Hengyang Blood Center, Hengyang, China, 3Shenzhen
Research Institute of Population and Family Planning, Shenzhen, China
A 22-year-old, phenotypically female individual was presented at hospital
because of primary amenorrhea. The physical examination revealed normal
female external genitalia, sparse pubic hair and developed breasts with
juvenile nipples and pale areolas. A mass was found in the left inguinal
canal. The gynecological examination revealed a short, blind-ending
vagina 7.0 cm in length. Ultrasonography of the pelvis showed no uterus,
fallopian tube, or ovaries. The hormonal profiles were as follows: folliclestimulating hormone (FSH) 23.4 IU/L, luteinizing hormone (LH) 39.6
IU/L, testosterone (T) 23.0 nmol/L, estradiol (E2) 0.177 nmol/L. FSH and
LH were above the upper normal limit. T was elevated compared with the
normal female level and was in the normal range for males. E2 was in the
normal range for females. Cytogenetic analysis from the peripheral blood
lymphocytes revealed a 46,XY karyotype in the patient.
DNA (genomic DNA was extracted from peripheral blood of the patient) sequencing
of polymerase chain reaction (PCR) products revealed a single-nucleotide substitution
(AT) at position 2184 in exon 3, resulting in the conversion of arginine (AGG) to
tryptophan (TGG) at amino acid position 608 in the DNA-binding domain of the AR
(Figure 1). The novel missense mutation of exon 3 in the AR gene, resulting in the
nonfunctional protein, is responsible for the clinical symptoms of CAIS.

Results: The prevalance of hyperuricemia was 33.4%. SUA levels in patients with
NASH were significantly higher than those of non-NASH (p= 0.035). There was also
no difference between the NASH and non-NASH groups in terms of hyperuricemia
prevalence (p= 0.107). Univariate and multivariate analyses both demonstrated that
hyperuricemia had a significant association with younger age (OR, 0.930; 95% CI,
0.884-0.979) (p= 0.005), higher body mass index (OR, 1.173; 95% CI, 1.059-1.301)
(p= 0.002) and hepatocellular ballooning (OR, 1.678; 95% CI, 1.041-2.702) (p=
0.033). Area under the curve for hepatocellular ballooning was OR, 0.626; 95% CI,
0.544-0.708) (p= 0.005) and NAS was OR, 0.579; 95% CI: 0.507-0.652) (p= 0.035).
Hyperuricemia was insignificant in predicting other hepatic necro-inflammatory
changes.
Conclusions: SUA level seems to be a useful metabolic parameter in the differentiation
of NASH and non-NASH. In addition, hepatocellular ballooning, which is considered
to be an earlier predictor of hepatocyte injury, was the unique histological parameter
associated with hyperuricemia. It is also thought to be a biochemical evidence in the
pathogenesis of liver damage. Further studies on the involvement of SUA in NAFLD
will not only expand our understanding of the mechanism of NAFLD, but will also
assist in the eventual development of new prevention and treatment strategies for
NAFLD by modulating the SUA levels.

A-102
The Elecsys Periostin assay as a companion diagnostic for the novel asthma
drug lebrikizumab

J. Sherman1, C. Holweg2, H. Kincaid3, S. Sidobre3, L. Ma3, R. Maciuca2,

A. Geistanger4, J. Matthews2, S. Palme4, K. Yen2. 1Roche Professional

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S25

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

Diagnostics, Indianapolis, IN, 2Genentech, South San Francisco, CA,
3
Covance Central Laboratory Services, Inc, Indianapolis, IN, 4Roche
Professional Diagnostics, Penzberg, Germany
Objective: The Elecsys Periostin assay is being developed as a companion
diagnostic for the anti-IL13 drug lebrikizumab in asthma. The current clinical trial
assay was used to evaluate the precision performance according to CLSI EP 5-A2
under field conditions. The assay was used in the Phase IIb lebrikizumab trials (LUTE
and VERSE) to stratify asthma patients by baseline serum periostin levels.
Methods: The Periostin assay is an automated electrochemiluminescence
immunoassay. The assay requires a total of 18 minutes to perform. It employs two
monoclonal antibodies targeted to Periostin via the sandwich principle. The assay is
used for the in vitro, quantitative determination of periostin in human serum.
Imprecision testing was conducted using a total of eight human serum pools covering
the entire measuring range (10-160 ng/mL). This experiment was carried out at three
external routine laboratories, on three cobas e 601 analyzers. The experimental
setting included two replicates in two runs per day for twenty-one days (EP-5),
using two reagent lots. Using this design, repeatability, intermediate precision and
reproducibility were estimated based on a variance-components model.
It has been shown that high serum Periostin levels ( 50ng/ml) are associated with
benefit from lebrikizumab in asthma (Phase II lebrikizumab trial MILLY, Corren
et.al. NEJM, 2011). LUTE and VERSE were randomized, multicenter, doubleblind, placebo-controlled, parallelgroup clinical trials. Patients were aged 18 years
and older whose asthma remained uncontrolled despite daily treatment with inhaled
corticosteroids (ICS) therapy and a second controller medication. Patients were
randomized in a 1:1:1:1 ratio to receive subcutaneous lebrikizumab at 250 mg, 125 mg
or 37.5 mg dose or placebo, administered every 4 weeks. The primary endpoint was
reduction in exacerbation rate and the main secondary endpoint was percent change in
forced expiratory volume (FEV1). Efficacy analysis for each endpoint was assessed in
subgroups of periostin high (50 ng/ml) and low (<50 ng/ml) patients.
Results: Repeatability for the Periostin assay varied from 1.24% to 1.64% CV,
intermediate precision varied from 2.30% to 2.56% CV and overall reproducibility
varied from 3.10% to 3.95% CV. The precision obtained fulfilled the assay
specification and met the requisite target to segment subjects properly to the correct
periostin stratum.
A total of 463 patients were enrolled in both studies, of which 42% of patients were
classified as periostin-high. Compared with placebo, the exacerbation rate was
reduced by 60% (95% CI: 18, 80) in periostin-high patients (combined dose levels)
and by 5% (95% CI: -81, 47) in periostin-low patients. FEV1 at 12 weeks increased in
the periostin-high group by 9.1% (95% CI: 2.2, 15.9) over placebo, compared with a
2.6% (95% CI: -2.7, 7.9) increase over placebo in the periostin-low group.
Conclusion: The Periostin assay demonstrated good precision and was used
for clinical sample testing to stratify patients in the lebrikizumab Phase IIb trials.
Lebrikizumab treatment reduced the exacerbation rate and increased FEV1 in patients
with uncontrolled asthma on ICS and a second controller, particularly in those who
were periostin-high, confirming the findings from the Phase II trial MILLY.

A-103

an analytical measurement range of 0.7mg/dL to 20 mg/dL and at 940 nm for hs-CRP

with the range of 0.02 mg/dL to 6 mg/dL. Equivalence of these two methods for the
chosen population was evaluated with an allowable total error of 0.6 mg/dL or 25 %,
whichever is higher. Also, 18 samples with ls-CRP values <0.7 md/dL were analyzed
by both methods.
Results: 33 specimens were compared over a range of 0.22 to 23.35 mg/dL by lsCRP. The difference between the two methods was within the allowable error for 31
of 33 (94 %) specimens, with a linear regression equation y=0.886x+0.3713 and the
correlation coefficient (R) of 0.9908.The error index was determined by the ratio of
the difference between two methods (hs-CRP ls-CRP) to the allowable total error
and the average error index was -0.13 with a range of -1.03 to 1.25. For 12 samples
with concentrations < 2.4mg/dL by ls-CRP, the mean bias was -0.16 mg/dL with a
standard deviation of 0.39 mg/dL, out of which 2 samples had error >0.6 mg/dL. For
21 samples with concentrations > 2.4mg/dL by ls-CRP the mean bias was -1% with a
standard deviation of 12%. Also, all the 18 samples with CRP concentrations <0.7mg/
dL by ls-CRP presented values <0.7 mg/dL by hs-CRP method.
Conclusions: CRP concentrations in the patient population of <1 year old infants can
be analyzed by both Beckman Coulter UniCel DxC ls-CRP and hs-CRP methods to
monitor inflammatory processes. Overall, the hs-CRP assay is equivalent to the lsCRP method for the majority of patients (94%). However, in about 17% patients with
CRP < 2.4 mg/dL, the bias may be greater than the allowable total error.

A-105
SDMA Outperforms Serum Creatinine-Based Equations in Estimating Kidney
Function Compared with Measured GFR

D. A. Payto, J. M. El-Khoury, D. R. Bunch, S. Wang. Cleveland Clinic,

Cleveland, OH
Background: Symmetric dimethylarginine (SDMA), a catabolic product of posttranslational modified arginine containing proteins, is an emerging biomarker for
renal function. Recent studies have shown that there is a correlation between SDMA
and renal function in chronic kidney disease (CKD) patients. However, there are no
published reports that have investigated the performance of SDMA versus creatinine
in estimating glomerular filtration rate (GFR) in both CKD and non CKD patients. The
objective of this study was to determine the correlations of measured GFR with SDMA
levels, creatinine levels, eGFR (MDRD), and eGFR (CKD-EPI). Method: EDTA
plasma samples were obtained from 21 male and 19 female adult subjects who were
21-76 years old. These subjects had GFR measured by radioactive iothalamate renal
clearance, ranging from 13-151 mL/min/1.732. Arginine, asymmetric dimethylarginine
(ADMA), and SDMA in EDTA plasma were assayed using a previously validated
LC-MS/MS method. Results: A power regression analysis showed R2=0.7544 and
R2=0.5827 the correlation of measured GFR with SDMA and creatinine, respectively.
A linear regression analysis was used for the correlation of measured GFR with eGFR
by MDRD and eGFR by CKD-EPI with R2=0.6534 and R2=0.7183, respectively.
Conclusion: Correlation of measured GFR with SDMA was higher than those with
creatinine, eGFR (MDRD), and eGFR (EPI) in a population of CKD and non CKD
patients. In Conclusion: SDMA outperforms creatinine and creatinine-based equations
in estimating kidney function compared with measured GFR.

Equivalence between high sensitivity CRP and low sensitivity CRP tests in
infants

S. N. Narla, P. Nelson, D. Brass, B. Horne, Y. Zhu. Medical University of

South Carolina, Charleston, SC
Background: C-reactive protein (CRP) is an acute-phase protein that rises in
response to inflammatory processes. Most clinical laboratories perform two different
assays for measuring CRP, low sensitivity CRP (ls-CRP) and high sensitivity CRP
(hs-CRP). The hs-CRP is usually used in combination with other biomarkers to assess
risk of developing myocardial infarction in patients presenting with acute coronary
syndromes and risk of cardiovascular disease in adults who do not manifest disease
at present. Along with the monitoring of inflammatory process, ls-CRP is also used
in the diagnostic evaluation of sepsis in neonates and infants. CRP values in healthy
infants and neonates vary from a range of 0.2 mg/dL to 1.0 mg/dL. Therefore, for the
diagnosis of sepsis, a concentration >1.0 mg/dL is considered abnormal and serial
testing is frequently performed. The objective of this study is to verify whether the hsCRP method is equivalent to ls-CRP in monitoring inflammatory processes in infants.
Methods: For this study, 33 serum samples with ls-CRP higher than 0.7 mg/dL from a
patient population of <1 year old were collected. These serum samples were analyzed
by both ls-CRP and hs-CRP assays on the UniCel DxC800 Systems by turbidimetric
immunoassays. In both methods, CRP reacts with a specific antibody to form insoluble
antigen-antibody complexes. The turbidity was measured at 340 nm for ls-CRP with

S26

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

between Harmonized and NCEP ATP III (=0.62), moderate between Harmonized &
IDF (=0.54) and NCEP ATP III & WHO (=0.51), fair between NCEP ATP III & IDF
(=0.33) and Harmonized & WHO (=0.37) and slight for WHO & IDF definitions
(=0.26). The most frequent component of MetS was central obesity according to
the WHO (98.8%), IDF (99.9%) and Harmonized (85.6%) definitions respectively.
However, decreased HDL was the most frequent component (93.5%) according to
NCEP ATPIII definition. On the other hand, hypertension was the least frequent
component according to all four definitions (WHO: 66.2%, NCEP ATP III: 72.1%,
IDF: 65% & Harmonized: 67%). The overall diagnostic potential of the NCEP ATP III
definition was the best among others [sensitivity: 90.8%, specificity: 98.9%, positive
predictive value: 99.9% and negative predictive value: 49.9%] while comparing with
Harmonized definition as the gold standard.
Conclusion: The prevalence of MetS in type 2 DM is very high in both genders
irrespective of the definitions used and increases with age thus posing a potential
threat of increased cardiovascular diseases in this group of patients. The patient
education and control of the modifiable risk factors for MetS should be given due
priority by the clinicians in the management of subjects with type 2 DM.

A-107
Serum 25-Hydroxy Vitamin D3 Levels in Patients with Psoriasis

S. Abusoglu1, A. Unlu1, F. Akyurek1, F. T. Akyurek1, G. S. Kurtoglu2.

1
Selcuk University Faculty of Medicine, Konya, Turkey, 2Meram Training
and Education Hospital, Konya, Turkey
Background: Vitamin D has classically been associated with phosphorus-calcium
metabolism and bone physiology. However, the finding of vitamin D receptors at
different sites suggests that vitamin D also has important extraskeletal functions.
Vitamin D plays a pivotal role in modulating dendritic cell function and regulating
keratinocytes and T-cell proliferation. Psoriasis is considered a prototypic T helper
(Th) 17-mediated disease with a putative role played by vitamin D deficiency in
its pathogenesis. The most efficient laboratory approach consists of a well-defined
evaluation of immune response in order to help diagnosis, to monitor evolution, and
to evaluate the effects of individualized therapeutic treatments. Vitamin D has been
found to regulate immune function in a number of inflammatory and autoimmune
disorders. The objective of this study was to find out the serum vitamin D levels in
patients with psoriasis.

A-106
Prevalence of Metabolic Syndrome and its Components according to Different
Definitions among Nepalese Type 2 Diabetic Patients

D. R. Pokharel1, D. Khadka2, M. Sigdel1, N. K. Yadav1, L. B. Sapkota1,

P. Rayamajhi1, S. K. Jha1, A. Yadav1, P. S. Shukla1. 1Manipal College of
Medical Sciences and Teaching Hospital, Pokhara, Nepal, 2Gandaki
Medical College and Teaching Hospital, Pokhara, Nepal
Objective: The objective of this study is to determine the prevalence of the
metabolic syndrome (MetS) and its components according to WHO (1998),
NCEP ATP III (2005), IDF (2005) and newly introduced Harmonized
(2009) criteria and determine their agreement and disparity in diagnosing
MetS among Nepalese patients with type 2 diabetes mellitus (DM).
Methods: Patients with type 2 DM without any other acute or chronic illness were
recruited for the study from the Manipal Teaching Hospital (MTH), Pokhara, Nepal.
Clinical data were obtained by interviewing the patients with structured questionnaire,
anthropometric and blood pressure measurements and biochemical analyses of the
blood samples. The analyzed biochemical parameters included fasting serum glucose,
insulin, homeostasis model assessment of insulin resistance (HOMA-IR), triglycerides
and high density lipoprotein cholesterol. Statistical analysis included usage of
Students t and Chi square tests, kappa statistics and 95% confidence intervals.
Results: 1061 (male: 589, female: 472) type 2 diabetic patients aged between 3089 years were included in the study. The total age adjusted prevalence was 69.9 %,
73.9%, 66.8% and 80.3% according to WHO, NCEP ATP III, IDF and Harmonized
definitions, respectively. The Harmonized definition outperformed other definitions
in diagnosing MetS. Prevalence generally increased with the age and remained
highest in the age range of 50-69 years in both the sexes. It was significantly higher
in females (p=0.000) according to WHO, NCEP ATPIII and Harmonized definitions.
Patients of Dalit community had the highest prevalence according to NCEP ATP III
(p=0.033) and Harmonized (p=0.037) definitions whereas patients of Mongol and
Newar communities outnumbered other communities according to WHO (p=0.001)
and IDF (p=0.037) definitions, respectively. The overall agreement was substantial

Methods: Blood samples were collected and serum was seperated with centrifugation
from 51 healty and 93 patients with psoriasis. Patients with chronic disease and
calcium metabolism disorders were excluded. This study was approved by local
ethic committee. 100 L internal Standard (d6-25-hydroxyvitamin D3) and 1000
L acetonitrile were added to 250 L of serum, calibrator, control for protein
precipitation, vortexed for a minute and centrifuged at 13.000 rpm for 10 minutes.
40 L of supernatant was injected into HPLC analytical column for chromatography.
Mass spectrometric analyses were performed using an Shimadzu LC-20-AD (Kyoto,
Japan) coupled with a ABSCIEX API 3200 triple quadrupole mass spectrometer
(USA) equipped with an atmospheric pressure chemical ionisation (APCI) operating
in positive mode. This methods coefficient of variation and % bias values were
9.35,1.29; 3.81,3.21 and 2.29,2.60 for 10, 32 and 150 g/L, respectively. Statistical
analysis was performed with SPSS v16.
Results: Serum 25-hydroxy vitamin D3 levels were significantly lower in patient
group (10.3 5.6) compared to control group (13.7 7.8) (p=0.004) according to
Mann-Whitney U test. Also, there was no statistically significant correlation between
Psoriasis Area Severity Index (PASI) and serum vitamin D levels (p=0.99) according
to Spearman correlation analysis.
Conclusion: Vitamin D may play an important role of psoriasis pathogenesis. Vitamin
D has been used to treat psoriasis in the topical form with great success. Low levels
of vitamin D may also have important implications in the pathogenesis of psoriasis.
Vitamin D3 acts mainly on the vitamin D receptor to regulate keratinocyte growth and
differentiation, but also
has an influence on immune functions of dendritic cells and T lymphocyte. Vitamin
D deficiency may be common in patients with psoriasis. Screening for vitamin D
deficiency may be useful for comprehensive management.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S27

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

A-108
CEDIA Cyclosporine Applications for the Beckman Coulter AU480, AU680
and AU5800 Analyzers

D. L. Cheng, T. Miwa, C. Wong. Thermo Fisher Scientific, Fremont, CA

Background: Cyclosporine is an immunosuppressant used in the treatment
of autoimmune diseases and the reduction of tissue rejection following organ
transplantations.
Although its mechanism of action is still under investigation, cyclosporine appears
to affect the metabolism of T-helper lymphocytes and T-suppressor lymphocytes,
resulting in the inhibition of the immune system. Although cyclosporine is safely used
over an established narrow range of concentrations, inadequate dosing may lead to
organ rejection, while overuse may lead to a number of adverse effects including
nephrotoxicity, and hepatotoxicity. The probability of adverse effects on users health
increases with drug concentration. Therefore, it is crucial to monitor cyclosporine
levels to achieve optimal immunosuppressive effects in patients.
Methods: The CEDIA Cyclosporine PLUS Assay is based on the enzyme
-galactosidase, which has been genetically engineered into two inactive fragments.
Cyclosporine in the human whole blood sample competes with cyclosporine
conjugated to one inactive fragment for antibody binding site. Once cyclosporine in
the sample binds to antibody, inactive enzyme fragments reassociate to form active
enzymes. The amount of active enzyme results in an absorbance change that is
directly proportional to the amount of cyclosporine in the sample, and is measured
spectrophotometrically. The Beckman Coulter AU480/AU680/AU5800 analyzers are
new applications for the CEDIA Cyclosporine PLUS Assay. Performance measures
of the CEDIA Cyclosporine PLUS Assay on the Beckman Coulter AU480/AU680/
AU5800 analyzers were conducted in low-range (25-450 ng/mL) and high-range
(450-2000 ng/mL) assay studies. Analyzer performance was determined for precision,
linearity, limit of detection, and accuracy. Correlation studies using the AU480/
AU680/AU5800 analyzers were conducted against the reference analyzer, Hitachi
911.
Results: All studies were evaluated using CLSI guidelines. Five levels of
cyclosporine controls were used in the studies. The precision ranged from 7.1 to
3.1%CV for within-run and 12.7 to 8.6%CV for total run. Low range linearity was
measured and confirmed over a range of 9.08-427.85 ng/mL on the AU480/AU680/
AU5800. High range linearity was measured and confirmed over a range of 590.12027.6 ng/mL on the AU480/AU680/AU5800. The limit of detection on the AU480/
AU680/AU5800 yielded 8.3 ng/mL. Accuracy was measured using patient correlation
against the reference analyzer Hitachi 911, which yielded a Demings Regression
for each analyzer. (Low Range Cyclosporine): AU480 = 1.03*(Hitachi 911) - 0.40
(N = 115, r = 1.00), AU680 = 0.97*(Hitachi 911) + 13.00 (N = 100, r = 1.00),
AU5800 = 1.00*(Hitachi 911) - 0.50 (N = 116, r = 0.99). High Range Cyclosporine:
AU480=1.03*(Hitachi 911) + 90.96 (N = 126, r = 0.97), AU680=1.07*(Hitachi 911)
+ 30.23 (N = 126, r = 0.97), AU5800=1.05*(Hitachi 911) + 9.00 (N = 126, r = 0.97).
Conclusion: All measured studies demonstrated acceptable performance, validating
the use of the CEDIA Cyclosporine PLUS Assay on the Beckman Coulter AU480/
AU680/AU5800 analyzers, and will provide an effective monitoring system for
patients receiving cyclosporine therapy.

A-109
Chemical composition of urinary tract calculi assessed in a Caribbean teaching
hospital

L. L. Dilworth, D. McGrowder, M. Tapper, R. Thompson, D. Stennett. The

University of the West Indies Mona, Kingston, Jamaica
Background:Urolithiasis is a heterogeneous and fairly common urological disorder
with an estimated lifetime prevalence of between 5 to 10%, with the risk being higher
in men. Environmental and genetic factors contribute to calculi formation. Research
regarding kidney stones has gained increased attention in light of its complex
molecular genetic basis as well as recent associations between urolithiasis and
cardiovascular diseases. Importantly, the proper management of patients with renal
stone disease involves the analysis of urinary calculi by varied methods. The main
objective of this study was to assess the chemical constituents particularly inorganic
minerals of urinary tract stones observed at the University Hospital of the West Indies
over a 4-year period. The study is validated by the observation that there is a paucity
of data on the types and composition of calculi recovered from patients within this
region.

with deionized water, air-dried and powder obtained via pulverization in an agate
mortar. Qualitative chemical analysis of the stones for calcium, magnesium,
phosphate, oxalate, uric acid, cystine and bicarbonate was done based on standard
methods. Carbonate content was determined by the effervescence test by exposing
the stone powder to concentrated hydrochloric acid. After the mixture was boiled,
the filtrate was used for detection of calcium with ammonium oxalate, magnesium
with potassium phosphate and ammonia, phosphate with ammonium molybdate
and oxalate with calcium chloride. The powder was also boiled with N-potassium
hydroxide and Folins uric acid reagent with sodium cyanide added to the filtrate to
detect uric acid. Cystine was detected with sodium nitroprusside and sodium cyanide.
Results: The incidence of stones from males and females was in the ratio of 1.4:1.
Calcium, the main constituent was present in 96.2% of the stones followed by
phosphate 67.7%, oxalate 56.3%, magnesium 28.2% and uric acid 17.3%. Mixed
calcium phosphate accounted for 42.4% of the stones while there were 25.4% pure
calcium phosphate stones. Mixed calcium oxalate accounted for 43.4% of the stones
while there were 12.9% pure calcium oxalate stones. Mixed uric acid was present in
16.3% and urinary stones containing bicarbonate accounted for 11.8%. One cystine
stone was recorded.
Conclusion: The study revealed that a relatively high proportion of the urinary tract
stones in the sample population consisted of both pure and mixed calcium phosphate,
followed by calcium oxalate and uric acid. The main contributory factor to the
frequency of these stones seems to be hypercalciuria resulting from hypercalcaemia.
Results show that males are more likely to present with urolithiasis. Ongoing studies
are geared at garnering more detailed information on both the composition and
structure of the urinary tract stones using solid state nuclear magnetic resonance
spectroscopy and X-ray diffraction crystallography.

A-110
Screening for Hemoglobin Variants [HVs] While Measuring Hemoglobin A1c
(HA1c).

N. Yadak1, C. Hoang2, D. Moore2, E. S. Pearlman2. 1University of Tennessee

Health Sciences Center, Memphis, TN, 2Veterans Affairs Medical Center,
Memphis, TN
Context: As part of an effort to develop an in-lab protocol for the reporting of HA1c
results in the presence of HVs, we sought to assess the frequency and identities of HVs
in the VAMC population.
Methods: Over seven weeks, 5188 patients were assayed for HA1c using HPLC
[Tosoh; South San Francisco, CA] according to manufacturers directions. Specimens
with an HV flag were referred (Lab Corp; Raleigh, NC) for hemoglobin ID [H-ID].
Estimates of the frequency of AS and AC (8 and 2.8% respectively) among AfricanAmerican [A-A] and the population prevalence of DM (9.5%) are literature derived. A
total of 40,648 HA1c assays were done in calendar year 2012.
Results: A total of 208 (4%) HV flags were generated. The median age was 61 YO
[27, 90], 188 (90.4%) were male and 192 (92.3%) A-A while 6 were of unknown race
(UR). Of these, 181 [87%] had sufficient specimen for H-ID and results included AS
(118/113 A-A, 4 UR), AC (42/40 A-A, 1 UR) while the remaining 21 comprised HPHF
(9), D-LA (3), SC (2), G-Norfolk (1), Kokomo (1), A2 (1) and no HV identified [NVI]
(4). Of a total of 57,680 veterans registered at the VAMC, 35.7% are self-identified
as A-A. Diabetic patients (5480) were assumed to be tested 3 times (on average) in
any 12 month period and the remaining 24,208 assays were attributed to once-tested
patients. Thus in any given 12 month period 29,668 patients out of the total of 57,680
(51.47%) patients would be tested. Requests were assumed to be uniformly distributed
over 52 weeks. Making all these assumptions the expected number of AS/A-A patients
was [29,668][0.357][0.08][7/52] = 114 [actual=113]. A similar calculation yields
an expected 40 instances of AC trait [actual=40]. HPHF and Kokomo specimens
generated an HV flag but not a numeric result for HA1c.
Conclusion: In our population that was 35.7% AA, 4% of patients assayed for A1c
generated an HV flag. More than 90% occurred in AAs and 85% were S and C trait.
Four (2.2%) specimens were NVI. The observed and (crudely) estimated number of
AA patients with S and C trait were in close agreement [chi-sq. = 0.084; p>.9].

Methods: The study was conducted on 288 urinary tract stones sent to the chemistry
laboratory for analysis from both male and female patients. Samples were washed

S28

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

of IL-18 by measurement of serum IFN- and CRP. We also determined the effect of
IL-18 -607 A/C and IL-18-137 C/G polymorphisms on graft function.

A-111
Tubular Damage Is Present In Patients with MGUS and Asymptomatic
Multiple Myeloma Even in the Absence of Impaired Estimated Glomerular
Filtration Rate; Alterations of Neutrophil Gelatinase-Associated Lipocalin and
Cystatin-C in Myeloma Patients Post IMiD- and Bortezomib-Based Regimens

I. Papassotiriou1, D. Christoulas2, E. Kastritis2, G. P. Papassotiriou2, A.

Margeli1, N. Kanellias2, E. Eleutherakis-Papaiakovou2, M. A. Dimopoulos2,
E. Terpos2. 1Department of Clinical Biochemistry, Aghia Sophia
Childrens Hospital, Athens, Greece, 2Department of Clinical Therapeutics,
University of Athens, Medical School, Athens, Greece
Background: We have recently shown that urinary and serum Neutrophil gelatinaseassociated lipocalin NGAL were elevated in the vast majority (90% and 70%,
respectively) of newly diagnosed patients with multiple myeloma (MM), while
serum cystatin-C (CysC), an accurate marker of GFR, was elevated in 70% of them.
However, there is no information for the value of these markers in patients with
MGUS, asymptomatic MM (AMM), as well as in symptomatic MM post treatment.
Patients and Methods: Thus, we measured urinary and serum NGAL and serum CysC
in 40 patients with MGUS; 36 with AMM and 120 healthy controls. Furthermore,
we measured serum NGAL and CysC in 39 newly diagnosed symptomatic MM
patients before and after frontline therapy with novel agents. Serum and urinary
NGAL was measured using an immunoenzymatic technique, while CysC was by
means of immunonephelometry. The estimated GFR (eGFR) was calculated using
the CKD-EPI equation. Patients were divided into the 5 CKD stages of the KDIGO
classification, according to eGFR.
Results: Urinary NGAL was elevated in patients with MGUS (median: 14ng/ml,
range 0.5-31ng/ml) and AMM (22.3ng/ml, 0.9-78ng/ml) compared to controls (5.
ng/ml, 0.7-9.8ng/ml, p<0.001 for both comparisons). Similarly, serum NGAL was
elevated in patients with MGUS (106ng/ml, 74.9-205.5ng/ml) and AMM (94.2ng/
ml, 29.5-306.4ng/ml) compared to controls (63ng/ml, 37-106ng/ml, p<0.01). There
was no difference between MGUS and controls or MGUS and AMM regarding CysC
serum values, indicating that traditional indices of renal function could not detect early
renal damage. However, 22 (55%) patients with MGUS and 24 (66%) with AMM
had higher urinary NGAL values than the higher value of the controls. Similarly,
9 (22.5%) MGUS and 11 (30%) AMM patients had higher levels of serum NGAL
than the higher value in the control group. As expected, patients with symptomatic
MM had elevated serum NGAL and CysC (p<0.001). NGAL strongly correlated with
CysC (r=0.675, p<0.001) and CKD stage (meanSD values for stages 1/2, stage 3
and stages 4/5 were: 9757ng/ml, 14479ng/ml and 205124ng/ml, respectively,
p=0.014). CysC also correlated with CKD stage (0.960.29mg/l, 1.540.32mg/l and
2.511.00mg/l respectively, ANOVA p<0.001). Among patients with eGFR <50ml/
min at baseline (n=22), 4/4 who received bortezomib-based regimens and 5/18 who
received IMiD-based regimens achieved at least minor renal response. After 4 cycles
of therapy, serum NGAL increased in patients who received IMiD-based therapy
compared to baseline (255264ng/ml vs. 147104ng/ml, p=0.021), but not in patients
who received bortezomib (11968ng/ml vs. 159111ng/ml p=0.520), regardless of
myeloma response to treatment.
Conclusions: We conclude that the high levels of urinary and serum NGAL in MGUS
and AMM indicate the presence of subclinical renal damage in these patients early in
the course of their disease, when other markers of renal function, such as sCr or even
the more sensitive CysC indicate that renal function is preserved. Thus, NGAL may
be useful as an early marker that predicts the development of renal damage and the
progression of the disease in these patients. NGAL seems also to increase in patients
with renal impairment who receive IMiD-based regimens.

A-112
Association of Interleukin (IL-18) -607A/C and -137C/G Polymorphisms With
Early Graft Function In Renal Transplant Recipients

H. Akbas, F. Davran, A. Dinckan, H. Kocak, D. Ozel, M. Gultekin, G.

Suleymanlar. Akdeniz University, Faculty of Medicine, Antalya, Turkey
Background: Recent studies have suggested that increased levels of IL-18 in serum
and renal parenchyma may predict acute rejection in patients with renal transplantation.
IL-18 mediates a wide range of inflammatory and oxidative responses including renal
injury, fibrosis and graft rejection. It has been reported that the promoter -607 and
-137 polymorphisms of IL-18 influence the level of cytokine IL-18 expression. This
prospective observational study aimed to assess the relevance of serial postoperative
serum/urine creatinine, cystatin C and IL-18 measurements for monitoring early graft
function in renal transplant recipients and to evaluate the pro-inflammatory property

Methods: This study included 75 renal transplant recipients (28 female, 47 male;
mean age: 38.28 13.03) from living related donors. Blood and urine samples were
collected immediately before and after transplantation at day 7 and month 1. Serum
IFN-, IL-18, creatinine, cystatin C, CRP and urinary IL-18, cystatin C and creatinine
levels were measured. Polymorphisms of the promoter region of the IL-18 gene, IL18607A/C and -137C/G were determined by analysis of real-time PCR/Melting curve.
GFR values were estimated by Modified Diet in Renal Disease (MDRD), Chronic
Kidney Disease Epidemiology Collaboration (CKD-EPI) and some cystatin-C-based
formulas (Larsson, Rule, Hoek). SPSS 20.0 software was used for statistical analysis.
Results: Serum creatinine, cystatin C, CRP, IFN- , IL-18, urine cystatin C levels
and urinary cystatin C/creatinine, urinary IL-18/creatinine ratios were significantly
decreased after transplantation (p<0.005). Serum cystatin C and IL-18 levels were
significantly higher in patients with IL-18-137 GG genotype before transplantation.
While pretransplant levels of serum creatinine and IL-18 were found significantly
higher in patients with IL-18-607 CC genotype, we also observed significantly
higher serum IFN- levels and estimated GFR (MDRD and CKD-EPI) values in CA
genotype (p=0.012). Receiver operating characteristic (ROC) analysis was performed
to quantitate the accuracy of the different markers to detect changes in GFR.
Posttransplant serum creatinine and cystatin C demonstrated a significantly greater
AUC (area under the curve), sensitivity and specificity values than IL-18 and IFN- .
Conclusion: In this study, although we observed significantly differences in serum
IL-18 levels and some GFR markers according to genotypes, the influence of
polymorphisms on early graft function has not been clearly shown. Future larger
studies are needed to confirm the association of cytokine gene polymorphisms with
graft function. Prior to transplantation, screening of genetic predisposition which
may have deleterious effect on graft function could lead to the development of
new treatments for better graft survey and ultimately improve the outcome of renal
transplantation.

A-113
May routine urine analysis reduce the number of unnecessary culture requests

A. Ozturk, Z. Ginis, T. Hanci, Z. Yildiz, M. Yavuz Taslpinar, F. Ucar, G.

Ozturk, A. Yalcindag, E. Alay, G. Erden. Ankara Diskapi Yildirim Beyazit
Research and Training Hospital, Department of Clinical Biochemistry
Ankara, Turkey
Background: Urinary Tract Infection (UTI) is the most common bacterial infection
in the society. Bacterial infections can lead to the leukocytosis. Urine analysis is one
of the most common tests for assessing urinary-tract infections. Urine culture is still
the gold standard for the detection of urinary tract infection, however, it is time- and
labor-intensive and has a high number of unnecessary cultures. The aim of this study
was to evaluate diagnostic performance of infection-related (associated) parameters
of urine preliminary analysis (leukocyte esterase, nitrite, bacteria and leukocyte) in
comparison to urine culture as the reference method and to investigate whether the
presence of UTI cause leukocytosis.
Methods: The electronic database of our laboratory was searched between July
2013 and December 2013. Our hospital is a tertiary central hospital with 671 beds.
Approximately 239.029 urinalyses were requested while the majority of these requests
were from outpatients and emergency patients. A total of 3427 patients (1980 women,
1447 men, mean age 49.2318.42) which were request urinanalysis, CBC and urine
culture on the same day were enrolled in the study. All results were retrospectively
reviewed. Leukocyte esterase and nitrite in chemical analysis were measured with
fully automatic urine analyzer (IRIS iQ200 Diagnostics, USA). Pyuria (WBC) and
bacteriuria in microscopy, leukocyte of CBC (Beckman Coulter LH780, USA)
parameters were compared with urine culture. Diagnostic performance of parameters
for detection of UTI were evaluated.
Results: Urine cultures were positive in 413 patients (%11.9). E. coli was the most
frequently isolated bacteria (70.4%). Ratios of positive dipstick results for leukocyte
esterase and nitrite in culture positive patients were 85% (n=352) and 40 % (n=166)
respectively. The positive microscopy results for bacteria and leukocyte were 31
% (n=127) and 75 % (n=310), respectively. Positive predictive values of leukocyte
esterase, nitrite, pyuria and bacteriuria tests were 69 %, 97 %, 74 %, and 91 % while
negative predictive values were 80 %, 62 %, 75 %, 58 % respectively. The highest
specifity rate was nitrite (99 %). Leukocytosis rate in patients with a positive urine
culture were 23 % (n=96). A relatively high correlation was found between LE and
microscopic WBC count (r = 0.827; P < 0.001).
Conclusions: Routine urine analysis is thought to be reliable in preliminary diagnosis
of UTI and start empirical treatment without waiting for culture results. Also urine

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S29

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

analysis is easy to apply and quick test that could reduce unnecessary requests with
predicting culture results. Dipstick urine analysis and urine microscopy can rule out
UTI in considering most samples have no or unsignificant growth. With a systematic
algorithm, laboratory workload, cost and unnecessary antibiotic prescriptions could
be reduced.

A-115
Enhanced Liver Fibrosis (ELF) Score indicates progressive fibrosis in preclinical stages of Primary Biliary Cirrhosis

D. C. Baldo1, A. Dellavance1, F. Latini2, J. A. Barreto2, N. R. Frana3, L. E.

C. Andrade3. 1Fleury Medicine and Health, So Paulo, Brazil, 2Beneficent
Association of Blood Collection from Sao Paulo, So Paulo, Brazil,
3
Universidade Federal de Sao Paulo, So Paulo, Brazil
Background: Liver fibrosis is a common consequence of most chronic liver diseases.
The assessment of liver fibrosis is usually made by histologic analysis of liver biopsy
samples. Percutaneous liver biopsy has inherent risks to the patient. Primary biliary
cirrhosis (PBC) is a slowly progressive cholestatic disease associated with the
development of cirrhosis and liver failure that may justify liver transplantation. PBC
diagnosis is based on completion of at least two of the following criteria: characteristic
histological findings; elevated alkaline phosphatase serum levels for over six months;
and circulating anti-mitochondria autoantibodies (AMA). AMA is detected in roughly
95% of PBC patients, with a specificity of at least 98%. Occasionally AMA-positive
(AMA) subjects with normal liver enzymes are identified by the characteristic
cytoplasmic pattern in the regular antinuclear antibody indirect immunofluorescence
assay on HEp-2 cells. Biochemically normal (BN) AMA-positive individuals may
represent earlier stages of CBP and may require confirmation by liver histology
analysis. The Enhanced Liver Fibrosis (ELF) score is an alternative approach that
combines three serum markers in an algorithm able to generate a score correlated
with liver fibrosis state (mild, moderate and severe). Aim: To investigate liver fibrosis
status, as assessed by ELF score, in BN AMA-positive individuals along a 4.5-year
follow-up period.
Methods: ELF score was determined for 37 PBC patients, 68 BN/AMA-positive
individuals, and 172 age- and gender-matched blood donors. BN/AMA-positive
individuals and PBC patients had two samples obtained 4.57 (1.37 - 16.04) and
4.75 (1.91 - 8.37) years apart, respectively. ELF score was calculated according to a
pre-established algorithm based on the serum levels of hyaluronic acid, procollagen
III and inhibitor of metalloproteinase determined by chemiluminescence (Siemens
Healthcare Diagnostics).
Results: ELF score was higher in baseline PBC samples (mean 9.63; 95%-CI:
9.09-10.17) than in baseline BN/AMA-positive samples (9.14; 95%-CI: 8.92-9.37)
(p=0.038) and blood donor samples (8.91; 95%-CI: 8.81-9.02) (p<0.001), with
no significant difference between BN and blood donors (p=0.284). There was a
significant increase in ELF score in the follow-up of PBC patients (p=0.010) and BN/
AMA-positive individuals (p<0.001).
Conclusion: PBC patients presented evidence of more severe liver fibrosis than BN/
AMA-positive individuals did. The progression in liver fibrosis status in BN/AMApositive individuals along the years indicates a subclinical inflammatory process in
these individuals. ELF score appears to be a sensitive parameter for evaluation of the
progression of liver involvement in BN/AMA-positive individuals.

A-116
Cardiac troponin I and B-type natriuretic peptide predict clinical outcomes in
stable renal transplant recipients

P. Jarolim1, B. L. Claggett1, M. A. Pfeffer1, A. Ivanova2, M. A. Carpenter2,

A. G. Bostom3, J. W. Kusek4, L. G. Hunsicker5, P. F. Jacques6, L. GravensMueller2, P. Finn1, M. J. Conrad1, S. Solomon1, A. S. Levey7. 1Brigham
and Womens Hospital, Harvard Medical School, Boston, MA, 2University
of North Carolina, Chapel Hill, NC, 3Rhode Island Hospital, Providence,
RI, 4National Institute of Diabetes and Digestive and Kidney Disease,
Bethesda, MD, 5University of Iowa, Iowa City, IA, 6US Department of
Agriculture, Jean Mayer Human Nutrition Research Center on Aging,
Boston, MA, 7Tufts Medical Center, Boston, MA

Outcome Reduction In Transplantation) using the B-type natriuretic peptide (BNP)

and high sensitivity troponin I (hsTnI) assays (both Abbott). CO included all-cause
death, dialysis-dependent kidney failure (DDKF) and CV outcomes. The relationship
between BNP and hsTnI values and CO were assessed via Cox regression models.
Quartiles (Q) of the two biomarkers were included in models adjusted for age,
gender, race, treatment, history of smoking, coronary heart disease, diabetes mellitus,
A/C ratio, eGFR, BMI, blood pressure, lipid levels, graft vintage and donor type.
Combinations of Q4 of BNP and hsTnI (BNP high, hsTnI high) and Q1-3 of BNP
and TnI (BNP low, hsTnI low) were also evaluated. Results: Median concentrations
and interquartile ranges for TnI and BNP were 5.6 (3.3, 10.5) ng/L and 56 (22, 129)
pg/L. Both BNP and TnI levels were associated with age, donor type, history of CHD
and diabetes mellitus, systolic BP, eGFR and urine ACR. Hazard ratios (HRs) for
each fatal and non-fatal endpoint increased significantly with increasing quartiles of
BNP after adjustment and remained significant after adding TnI to the model. HRs
for the BNP/TnI combinations were strongly associated with all CO studied (Table).
Conclusion: BNP is predictive for death, CV as well as renal outcomes in stable renal
transplant recipients. Simultaneous elevation of BNP and TnI is relatively common
and strongly predictive of CO.
Adjusted hazard ratios for combinations of low and high BNP and hsTnI results
BNP low/
BNP low/hsTnI BNP high/hsTnI BNP high/hsTnI
Clinical outcome hsTnI low
high (13%)
low (13%)
high (12%)
(63%)
All-cause
Ref
1.68 (1.03-2.72) 1.97 (1.19-3.26) 3.26 (1.92-5.54)
mortality
Mortality and
Ref
1.21 (0.77-1.91) 1.73 (1.09-2.76) 3.34 (2.04-5.48)
DDKF
DDKF
Ref
0.75 (0.39-1.44) 1.32 (0.70-2.49) 2.64 (1.31-5.29)
CV outcomes
Ref
1.14 (0.67-1.92) 2.17 (1.27-3.69) 2.46 (1.38-4.36)
Mortality and CV
Ref
1.41 (0.90-2.20) 2.12 (1.33-3.36) 2.67 (1.63-4.38)
outcomes

A-117
Results of Sample Testing for the Determination of Reference Intervals
in Apparently Healthy Pediatric Subjects for ADVIA Centaur Systems,
Dimension Vista and Dimension EXL Systems Thyroid Assays

R. H. Christenson1, D. Counts1, B. Plouffe2, S. Gafary2, J. L. Burcham2,

R. Levine2, T. Sullivan3. 1The University of Maryland, Baltimore, MD,
2
Siemens Healthcare Diagnostics Inc., Tarrytown, NY, 3Norwich Pediatric
Group, Norwich, CT
Background: Age-specific reference intervals are necessary for appropriate
interpretation of thyroid hormone measurements in the pediatric population and
may vary due to methodological differences. A challenge for establishing pediatric
reference intervals has been the availability of well-characterized samples from
healthy pediatric subjects. Of the few studies that provide this information, most are
based on specimens from patients who were hospitalized or required medical care.
This study used methodology consistent with CLSI guidelines to pedigree and collect
samples from apparently healthy pediatric subjects presenting for regular well child
care.
Objective: To test well-characterized specimens from healthy pediatric subjects to
establish pediatric reference intervals for various assays and instruments.
Methods: Eight US sites prospectively collected samples from apparently healthy
pediatric subjects, under institutionally approved consent/assent procedures. Subjects
were normal according to CDC weight- and height-based growth charts, were free of
chronic and acute diseases, were not on medication, had no family history of thyroid
dysfunction, no visible or palpable goiters, and were negative for anti-thyroglobuin
and anti-thyroid peroxidase antibodies. Three age subgroups were analyzed with
approximately equal numbers of males and females. Samples were shipped to a central
laboratory and tested in singleton using multiple Siemens immunoassay systems. The
lower and upper reference limits were defined as the 2.5th and the 97.5th percentiles
of the distribution of test results for each of the two older subgroups. For the infant
subgroup, a robust method (Horn and Pesce) was used to calculate the reference
intervals.

Background: Cardiac troponins and natriuretic peptides are becoming established as

predictors of clinical outcomes (CO) in patients with cardiovascular (CV) disease;
however their value in renal transplant recipients has not been established. Methods:
Using a case-cohort design, we tested baseline specimens from 1114 stable renal
transplant recipients enrolled in the FAVORIT study (The Folic Acid for Vascular

S30

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

Results:
System
System
ADVIA Centaur

Assaya Infant
n Children
n Adolescent
Assaya 1mo to
n 2yr to
n 13yr to
<24mo
<13yr
FT3
72
190 <21yr
3.28-5.19
3.34-4.80
3.04-4.65
FT4
0.94-1.44
0.86-1.40
0.83-1.43
ADVIA Centaur
72 1.05-2.07
190 0.86-1.92
T3
1.17-2.39
6.03-13.18 82 5.50-12.10 191 5.50-11.10
Dimension VISTA T4
TSH 0.741-5.24
0.628-3.90
0.438-3.98
FT3
3.34-5.24
3.31-4.88
191 2.91-4.53
Dimension VISTA FT4
82 0.81-1.35
0.88-1.48
0.78-1.33
190 6.0-11.6
T4
7.4-14.3
Dimension EXL
75 6.8-12.5
TSH 0.781-5.72
0.704-4.01 185 0.516-4.13
FT3
3.47-5.29
75 3.35-4.82
185 2.91-4.70
Dimension EXL
FT4
0.93-1.45
77 0.82-1.40
187 0.78-1.34
T4
6.6-13.4
75 5.8-11.8
186 5.4-10.6
a
Assay units: FT3, pg/mL; FT4, ng/dL; T3, ng/dl; T4, g/dL; TSH, IU/mL

A-119

n
n
129

A Hope for Healing Using Amniotic Membrane and Stem Cells

A. O. B. S. Osman1, M. M. S. El Ansary2, H. G. Metwally3, A. A. Gad2,

H. G. Al-Inany2, A. H. B. Elbadawy4. 1Misr University for Science and
Technology, Faculty of Medicine, Cairo, Egypt, 2Cairo University, Faculty
of Medicine, Cairo, Egypt, 3Cairo University,Faculty of Medicine, Cairo,
Egypt, 4Assiut University, Faculty of Medicine, Assiut, Egypt

129
148
148
147
147
147

Conclusion: Pediatric reference intervals were established for ADVIA Centaur,

Dimension Vista and Dimension EXL thyroid assays using rigorously pedigreed
samples. These data will assist with the appropriate interpretation of thyroid
measurements in infants, children and adolescents.

A-118
Evaluation of six different formulas and equations for estimate low density
lipoprotein cholesterol (LDLc) concentration.

O. M. Arroyo Huanaco, S. R. Alcantara Tito, B. J. Snchez Jacinto, S. M.

Flores Toledo, J. C. Jara Aguirre. Universidad Peruana Cayetano Heredia
- Facultad de Medicina Alberto Hurtado - Escuela de Tecnologia Medica,
Lima, Peru
Background: The National Cholesterol Education Program Adult Treatment Panel
(ATP III) identified the serum LDLc concentrations as the primary criterion of diagnosis
and treatment of patients with hyperlipidemia and also as one of the most important
parameter for evaluation of coronary heart disease and cardiovascular risk assessment.
Most of Peruvian clinical laboratory used the Friedewald formula and some LDLc
precipate method for estimate serum LDLc concentration, but the Friedwald formula
used has some limitations especially with extreme low and high triglycerides values
(> 400 mg/dL). The aim of the study was compare six different proposal formulas and
equations: Friedewald, Anandaraja, Chen, Puavilai, Vujovic and Cordova for estimate
value of serum LDLc compared to LDLc precipitate method. Methods: A descriptive
cross-sectional study was conducted and 220 serum samples from adolescent
students of medical technology at the Peruvian University Cayetano Heredia were
collected during the annual clinical evaluation performed on November 2012. Serum
samples on fast conditions for lipid profile were obtained, the measurement of LDLc
cholesterol and HDLc cholesterol were performed by precipitation methods (Wiener
Lab, Argentina), Total Cholesterol (TC) and Triglycerides (TG) were measurement by
enzymatic colorimetric endpoint method (Wiener Lab, Argentina), using a BT 3000
Plus automatic photometer analyzer at the Clinical Laboratory of Cayetano Heredia
Clinic - Universidad Peruana Cayetano Heredia. Linear regression analysis and Blant
Altman plots were performed to evaluate six different formulas and equations for
LDLc estimation and compared to LDLc precipate method, using MS Excel and
MedCalc Statistical Software. TG and LDLc obtained values for the statistical analysis
and clinical classification were divided in levels according to ATP III. Results: Linear
regression and Blant Altman plots showed a significant correlations and agreements,
respectively. The correlation coefficient for serum sample group of TG 149 mg/
dL were to Friedewald, Chen, Puavilai, Vujovic (R2>0.95), Cordova (R2=0.94),
Anandaraja (R2=0.81), respectively. For serum sample group of TG between 150-199
mg/dL were Friedewald, Puavilai (R2=0.88), Chen, Vujovic (R2=0.89) and Cordova
(R2=0.91), whereas TG (200-400 mg/dL) Friedewald, Chen, Anandaraja, Puavilai ,
Vujovic (R2=0.87) and Cordova (R2=0.86). Using Friedewald formula n=18 (8.18%),
Anandaraja n=30 (13.63%), Chen n=44 (20%), Puavilai n=14 (6.36%), Vujovic n=24
(10.91%) and Cordova n=106 (48.18%) patients show a probably misclassification for
hyperlipidemia. Conclusion: Six differents formulas showed significant correlations
and agreements according Lineal regression analysis and Blant Altman plot. The
Puavilai equation/formula showed good performance to estimate LDLc in all
triglycerides levels and shows less misclassification of patients for hyperlipidemia
compared to LDL precipitate method used in this study.

Objectives: Testing a new technique for treatment of chronic non-healing ulcers using
amniotic membrane (AM) alone or in combination with autologous mesenchymal
stem cells (MSCs) as an application of clinical laboratory medicine in regenerative
medicine.Methodology: After Institutional Ethical Committee approval, each patient
signed informed written consent. The study was conducted on 37chronic leg ulcers.
Patients were randomly divided into 4groups; Group I: (control group 11ulcers),
ulcers were treated with conventional wound dressings that were changed day by
day for 8weeks. Group II: (14ulcers) AM was placed in contact with ulcer and held
in place with 2ry dressing; which was changed day by day. Group III: (6ulcers)
autologous BM derived MSCs were injected into ulcer bed and ulcer edges. Group
IV: (6ulcers) Autologous MSCs were injected into ulcer bed and ulcer edges, and
freshly prepared AM was placed in contact with ulcer and held in place with 2ry
dressing.Patients were subjected to:I. Assessment of ulcer healing and wound
measurement:-1. Percentage of the healed wound area and healing rate:Healed
wound area%= (Original wound size-final wound size) /Original wound size100
Healing rate in cm2/day= (Original wound size-final wound size) /Time consumed
to reach final wound size-2. Wound size: Greatest length and greatest width were
measured, and surface area was calculated in cm2. -3. Wound grading: Based on
Pressure Ulcer Scale for Healing (PUSH Tool) Ulcers classified as Mild, Moderate,
and Severe. II- Pain Assessment:On a scale from 1 to10, pain level is classified as
(No pain, Mild, Moderate, & Severe).III- Follow up:-Healing rate and ulcer sizePain Assessment -Taken of AM graft (Day2).-Ulcer images (Day0, and weekly till end
of the study)Statistical Analysis It was performed with SPSS software version15.0
for Windows (SPSS Inc., CR). Results:Significant improvement in patients of groups
II,III, and IV regarding pain &ulcer size .There was significant difference in healing
rate &reduction in ulcer size between group I and (group II &IV) with (P<0.001)
where wounds showed an overall decrease in wound size and improvement in wound
bed with healthy granulations. Group II results showed complete healing of 14ulcers
in14-60days with mean of 33.314.7, healing rate was 0.064 2.22and mean of 0.896
0.646cm2/day with 100% reduction in ulcer size. In group III; 4ulcers (66.7%)
showed complete healing in 45-60days with mean of 50.56.7, while 2ulcers (33.3%)
showed partial healing with healthy granulations. Ulcers healing rate was 0.084
0.787cm2/day, with mean of 0.303 0.254. Reduction in ulcer size ranged from 50.4%
to100%, with mean of
83.9% 24.9%. In group IV- in the combined treatment with AM and autologous
MSCs- 3leg ulcers (50%) showed complete healing in 14- 25days with mean of
17.76.35 and 3ulcers (50%) showed partial healing with healthy granulations. Range
of Ulcers healing rate was 0.3591.23cm2/day with mean of 0.759 0.361. Range of
reduction in ulcer size in group IV was 46.5% 100%, and mean of 78.95%24.2%.
Conclusion:Using AM alone or in combination with autologous MSCs represents
promising simple, safe, effective, and novel therapeutic approach for closing and
healing of persistent non-healing chronic leg ulcer.

A-120
Inflammatory Markers in Polycystic Ovary Syndrome

G. Ozturk1, S. Ozdemir1, O. Ozdemir2, D. Kalkan2, G. Erden1, S. Sezer3,

C. R. Atalay2. 1Ankara Diskapi Yildirim Beyazit Research and Training
Hospital, Department of Clinical Biochemistry Ankara, Turkey, 2Ankara
Numune Research and Training Hospital, Department of Obstetrics and
Gynecology Ankara, Turkey, 3Ankara Numune Research and Training
Hospital, Department of Clinical Biochemistry Ankara, Turkey
Background: Previous studies have demonstrated polycystic ovary syndrome
(PCOS) associated with a proinflammatory state. The clinical spectrum of PCOS
includes components of the metabolic syndrome, such as central obesity, insulin
resistance, dyslipidemia and hypertension. All these disorders are epidemiologically
related to cardiovascular disease, most probably through low-grade intravascular
chronic inflammation. We aimed to investigate whether inflammatory markers,
including C-reactive protein (CRP), procalcitonin and ischemia modified albumin
(IMA) are related and altered in polycystic ovary syndrome.
Methods: A case-control study including fifty women diagnosed with PCOS,

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S31

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

according to Rotterdam criteria, and thirty-three controls, matched for body mass
index (BMI) and age. Serum samples for CRP, procalcitonin and IMA were collected
from these women under spontaneous menstrual cycles between the third and
seventh days and between 08:00 and 10:00 after an overnight fast, but at random
times if they suffered severe oligo- or amenorrhoea. CRP levels were measured
by immunoturbidimetric method (ADVIA 2400 Chemistry System, Siemens
Healthcare Diagnostics Inc.Tarrytown USA). Procalcitonin levels were measured by
sandwich immunoassay method (ADVIA Centaur CP System, Siemens Healthcare
Diagnostics Inc. Tarrytown USA). MA levels were measured by spectrophotometric
method. The p<0.05 was considered as statistically significant.
Results: PCOS patients had increased levels of testosterone and luteinizing hormone
(LH) compared to healthy BMI matched controls. Mean CRP, procalcitonin and IMA
levels were higher in patients with PCOS than controls, but these differences were not
statistically significant.
Conclusion: In PCOS women, plasma levels of CRP, procalcitonin and IMA were
not significantly increased when compared with age- and BMI-matched controls.
Several previous studies suggest that chronic low-grade inflammation in PCOS is an
important risk factor for long-term situation including cardiovascular complications,
metabolic disorders and ovarian dysfunction. The precise mechanisms underlying
these associations require additional studies to clarify the state of the cardiovascular
system in women with PCOS compared with controls in large numbers of patients to
determine the relative contribution of different factors including insulin resistance,
androgen status and BMI.

A-121
Recommendations of U.S. Preventive Services Task Force for Health Screenings
that Include Use of Laboratory Tests

S. Shahangian1, V. A. Moyer2. 1CDC, Atlanta, GA, 2U.S. Preventive Services

Task Force, Chapel Hill, NC
Background: Health screening tests greatly impact the publics health because
they involve testing of asymptomatic populations for diseases or conditions where
interventions may impact health outcomes. U.S. Preventive Services Task Force
(USPSTF) is the leading independent panel of experts that conducts rigorous and
impartial assessments of scientific evidence for effectiveness of clinical preventive
services.Methods: All screening recommendations are rated by USPSTF into 1
of 5 grades: A for strong recommendation with high certainty that net benefit
is substantial, B for recommendation with high certainty that net benefit is
moderate or moderate certainty that net benefit is moderate to substantial, C for
individualized decision making with at least moderate evidence that net benefit is
small, D against screening with moderate to high certainty that net benefit is zero
or harms outweigh benefits, and I is a statement that no recommendation can be
made due to insufficient, poor quality, or conflicting evidence, with the balance of
benefits and harms of screening undetermined.Results: Laboratory screening tests
for 35 diseases/conditions have been included in USPSTFs evaluations for different
target populations, resulting in a total of 60 currently active recommendations for
laboratory testing. Of these, there were 20 A and B recommendations including:
cervical cancer in women aged 30-65 years; chlamydial infection in sexually active
women aged 24 years and older pregnant women at risk; colorectal cancer in adults
aged 50-75 years; diabetes mellitus in adults with sustained hypertension, gonorrhea
infection in sexually active women at increased risk, hepatitis B virus (HBV) infection
in pregnant women; hepatitis C virus infection of persons at high risk and 1-time
screening of those born in 1945-1965; HIV screening of pregnant women, 1-time
HIV screening of adolescents and adults aged 15-65 years; risk of coronary heart
disease in men aged 35 years, as well as both men aged 20-34 years and women aged
45 years who are at increased risk; and syphilis infection in pregnant women. The
most common recommendation grades were D and I (17 each). D recommendations
included: cervical cancer in women aged <21 years, those aged >65 years not at high
risk and with adequate prior screening, those with hysterectomy leading to removal
of cervix and no history of high-grade or cancerous lesion, and human papilloma
virus screening of women aged <30 years; colorectal cancer in adults aged >85 years;
genital herpes simplex virus infection in adults, pregnant women, and adolescents;
gonorrhea infection in men and nonpregnant women at low risk for infection; HBV
infection in men and nonpregnant women; hereditary hemochromatosis in adults; lead
poisoning in children aged 1-5 years at average risk and pregnant women; ovarian
cancer in women excluding those with known genetic mutations that show increased
risk; pancreatic cancer; prostate cancer; and syphilis infection in nonpregnant persons
not at increased risk.Conclusion: While common perception is that screening is
recommended for major and prevalent diseases, the UPSTF either recommended
against (28%) or insufficiently supported (28%) use of more than half of the evaluated
laboratory markers.

S32

A-122
Performance of the Elecsys Vitamin D Assay in a Multicenter Evaluation

A. Algeciras-Schimnich1, A. A. Valcour2, P. Jarolim3, M. Lessig4, J. Kappes5.

1
Mayo Clinic, Rochester, MN, 2LabCorp, Burlington, NC, 3Brigham and
Womens Hospital, Boston, MA, 4NLS Lab Director Nationwide Labs, Fort
Lauderdale, FL, 5Roche Diagnostics Operations, Indianapolis, IN
Background: Increased awareness and investigation into the importance of vitamin
D for many physiological processes has resulted in dramatically increased demand
for vitamin D testing. New automated methods to measure 25-hydroxyvitamin D (25OHD) levels have been developed to help laboratories accommodate this surge. The
fully automated Roche Elecsys Vitamin D assay, which allows for the quantitative
determination of total 25-OHD in human serum and plasma, was recently introduced
to the US market. The Elecsys Vitamin D assay is standardized against a liquid
chromatography-tandem mass spectrometry (LC-MS/MS) method using National
Institute of Standards and Technology (NIST) reference materials for calibration.
The current study evaluated the performance of the Elecsys Vitamin D assay
across several laboratories in the United States to determine its suitability for use in
clinical practice. Functional sensitivity, imprecision, and clinical concordance with
a well-established NIST standardized LC-MS/MS method were assessed. Methods:
Functional sensitivity was assessed using 8 pooled samples covering a range of
0-16 ng/mL, which were tested once a day for 10 consecutive days. Precision was
assessed, following CLSI EP5-A2/EP15-A2, from 5 pooled samples in the 10- to
50-ng/mL range. Deming regression analysis was used to compare the results of the
Elecsys Vitamin D assay obtained at 2 sites on the cobas e411 analyzer with LC-MS/
MS performed at Mayo Clinic. Four sample cohorts were used: subjects referred for
routine vitamin D testing were included in cohort I, while predialysis patients (cohort
II), pregnant women (cohort III), and patients in the intensive care unit (cohort IV)
represented unique subpopulations. The clinical performance of the 2 methods was
evaluated using 20 ng/mL and 30 ng/mL as cutoffs for vitamin D sufficiency. Results:
The functional sensitivity of Elecsys Vitamin D, assessed at 2 sites, was 3.52 ng/mL
and 3.37 ng/mL with coefficient of variation (CV%) of 15% and 16%, respectively.
Elecsys Vitamin D demonstrated consistent reproducibility across lots (N=2), across
the 2 sites tested, with CV% of 5.66% to 7.12% (SD for concentrations below 15 ng/
mL; 1.40 ng/mL). Consistent repeatability (CV% of 1.30% to 5.48% [0.40 to 0.72 ng/
mL]), within site/lot precision (CV% of 1.94% to 7.47% [0.49 to 1.26 ng/mL]), and
within site/across 2 lots precision (CV% of 4.06% to 7.78% [1.00 and 1.10 ng/mL])
was demonstrated. Overall, acceptable consistency between Elecsys Vitamin D assay
at both sites, and LC-MS/MS was observed for regression parameters in both the
routine and specialty cohorts with a slope range of 1.03 to 1.22 and Pearsons r range
of 0.857 to 0.932. The LC-MS/MS sufficient/insufficient classification based on the
20-ng/mL cutoff over cohort I (routine adults) was 67.3% sufficient/32.7% insufficient
compared with 68.8%/31.2% observed with the Elecsys method at 1 site. Using the
30-ng/mL cutoff over cohort I, the LC-MS/MS classified 36.7%/63.3% as sufficient/
insufficient, respectively, and the Elecsys showed agreeable numbers at 38.3%/61.7%.
Conclusion:
The Elecsys Vitamin D assay showed robust technical performance, with good
functional sensitivity and low imprecision. Analysis of multiple cohorts showed high
clinical concordance with NIST standardized LC-MS/MS, supporting the suitability
of the assay for use in clinical practice.

A-123
Homogeneous immunoassay for estradiol based on blue-emitting upconverting
nanophophors and luminescence resonance energy transfer

V. Kale, L. Mattsson, R. Arppe, T. Soukka. University of Turku, Turku,

Finland
Objective: The purpose of this study was to demonstrate a competitive homogeneous
assay for 17-estradiol (E2) based on photon upconversion luminescence. NaYF4:
Yb,Tm upconverting nanophosphor (UCNP) with strong emission at 475 nm under
980 nm excitation was used as a donor in upconversion luminescence resonance
energy transfer (UC-LRET) with small molecular conventional fluorescent dyes
energy acceptors.
Methods: Upconverting nanophosphors (UCNPs; NaYF4: 20% Yb, 0.5% Tm) of 2530 nm in size were synthesised using oleic acid as chelating agent and shape modifier.1
The nanophosphors functionalized silica shell comprising carboxylic acid groups,
which were covalently conjugated to anti-E2 Fab antibody fragments (Fab S16) using
carbodiimide chemistry. E2 conjugated Alexa Fluor 488 (E2-AF488) and fluorescein
isothiocyanate (E2-FITC) dyes were used as acceptor conjugates. The luminescence

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

spectrum of the donor (emission peak around 475 nm) overlapped with the excitation
spectrum of the acceptors. In the homogeneous E2 assay, E2 dilutions in buffer were
first incubated together with the Fab S16 coated upconverting donor, then E2-AF
488 or E2-FITC were added to the reaction and the sensitized acceptor emission was
measured at 565 nm with anti-stokes luminescence plate reader (Hidex Oy, Turku,
Finland) under 980 nm infrared excitation.
Results: By using 0.006 mg/ml UCNP donor with 16 nM E2-AF488 acceptor
conjugate or 0.01 mg/ml UCNP donor with 4 nM E2-FITC acceptor conjugate in the
assay reaction, standard curves with IC50 values of 2 nM and 1.7 nM were obtained,
respectively. The working range of the assay was from 0.80 nM to 3.10 nM E2
concentrations with both the acceptors. Signal levels for the homogeneous assay using
the E2-FITC acceptor conjugate were two times higher than by using the E2-AF488
acceptor conjugate. However, the ratio between maximum and minimum signals, or
the signal-to-background ratio, was 12 for E2-AF488 and 8 for E2-FITC acceptor.
Conclusions: In this study, we introduced a blue emitting upconverting nanophosphor
as an efficient donor in UC-LRET and demonstrated a competitive homogeneous
upconversion based immunoassay for E2. Photon upconversion and luminescence
resonance energy transfer enable sensitive homogeneous immunoassays, which
eliminate the autofluorescence background with simple instrumentation and render
UCPs an attractive reporter technology for clinical diagnostics.
References:1. Kale, V; Soukka, T; Hls, J, Lastusaari, M. J. Nanopart. Res.2013,
15, 1850.

A-124
Enhanced Liver Fibrosis (ELF) Score for the Evaluation of Liver Fibrosis in
Autoimmune Hepatitis

E. Oliveira1, R. M. Perez2, P. M. Oliveira1, A. Dellavance3, L. E. C.

Andrade1, M. L. Ferraz1. 1Federal University of Sao Paulo, Sao Paulo,
Brazil, 2Federal University of Rio de Janeiro, Rio de Janeiro, Brazil,
3
Fleury Medicine and Health, Sao Paulo, Brazil
Background: Evaluation of fibrosis is very important in chronic liver diseases,
in order to define prognosis and therapeutic options and to establish strategies of
follow-up. Liver biopsy has been considered the gold-standard for the evaluation
of fibrosis. However, is an invasive method and prone to sampling error. These
problems have motivated the search of non-invasive and more reliable methods for
detecting and staging liver fibrosis. The ELF (Enhanced Liver Fibrosis) score is based
on an algorithm of three markers: (hyaluronic acid, pro-collagen III e inhibitor of
metalloproteinase 1) and has been evaluated in different etiologies of liver diseases.
However, ELF score has not yet been evaluated in patients with autoimmune hepatitis
(HAI). Aim: to evaluate the performance of ELF score in patients with AIH, using the
liver biopsy as a gold-standard.
Methods: Patients with HAI and a liver biopsy performed in less than 12 months were
included and had a serum sample collected for the ELF score (Siemens Healthcare
Diagnostics). Patients with hepatitis B or C were excluded, as well as patients with
alcohol consumption > 20g/day.
Results: 109 patients were evaluated and 89 were included (90% female, mean age
34y). According to the ELF cut-offs proposed by the manufacturer, 8% of patients had
absent/mild fibrosis, 42% moderate fibrosis and 50% advanced fibrosis. ELF score
overestimated fibrosis in 28% of cases and underestimated in 33%. The accuracy of
ELF (AUROC) for significant fibrosis (F 2) was 0,64 (CI 95%: 0,52-0,77; p= 0,05),
for advanced fibrosis (F 3) was 0,59 (CI 95%: 0,47-0,71; p= 0,17) and for cirrhosis
was 0,66 (CI 95%: 0,54-0,77; p= 0,009). Coeficient of correlation for ELF and fibrosis
was 0,24 (p=0,22), for periportal necroinflammatory activity was 0,28 (p= 0,009) and
for lobular necroinflamatory activity was 0,43 (p<0,001).
Conclusion: ELF score had not a good performance as a marker of liver fibrosis
in patients with AIH. Using proposed cut-offs, there was overestimation as well as
underestimation of liver fibrosis in many cases when comparing to liver biopsy. It is
possible that the biomarkers included in the ELF score could have been influenced by
the intense necroinflammatory process observed among AIH patients. In fact, ELF
score was more intensely associated to lobular activity (p< 0,001) then to liver fibrosis
(p= 0,22). It is interesting to note that 4 patients with acute hepatitis and no fibrosis
had a high ELF score, confirming that intense inflamation is a relevant confounder of
indirect liver fibrosis assessment.

A-125
Organisms cultured from the synovial fluid of infected prosthetic joints.

P. Kilmartin1, S. Gulati2, P. Citrano2, K. Kardos1, C. Deirmengian1. 1CD

Diagnostics, Inc, Wynnewood, PA, 2Citrano Medical Laboratories, Towson,
MD
Objective:Prosthetic joint infection (PJI) occurs after primary joint replacements
at a rate of 1.0-3.0% and after revision joint replacements at a rate of 2.0-20.0%.
Synovasure is a laboratory developed test (LDT) for the detection of alpha-defensins
in the synovial fluid (joint fluid), which have been shown to be elevated in infected
samples. There are two main objectives of this study; first to describe the variety
and distribution of organisms infecting joint replacements nationally, and second to
evaluate for organism-specific differences in the synovial fluid alpha-defensin level.
Methods:1698 synovial fluid specimens from 40 states were sent to our laboratory
for the SynovasureTM alpha-defensin assay and were also cultured using the BacT/
ALERT FAN FA/FN culture bottles for recovery of aerobic and anaerobic organisms
(Biomerieux). The organisms were identified and evaluated for susceptibilities using
the VITEK 2 ID/AST system (Biomerieux), a fully automated system that provides
rapid microbial identification and susceptibility testing.
Results:47 different organisms from 237 culture-positive samples were identified
(distribution summarized in graph). The data indicate that Staphylococcus epidermidis
and Staphylococcus aureus account for 46% of the organisms cultured. Of S. aureus
isolates, 41% were methicillin-resistant. In addition, there were no statistically
significant organism-specific differences in the synovial fluid alpha-defensin levels in
the culture-positive samples.
Conclusion:In a large national survey, we have identified Staphylococcus epidermidis
(33%), Staphylococcus aureus (13%) and Staphylococcus lugdunensis (6%) as the
most common organisms isolated from synovial fluid. The 40% rate of methicillinresistance among S. aureus isolates is quite concerning given the poor outcomes
associated with the treatment of these prosthetic infections. The elevation of alphadefensin in the synovial fluid of infected joint arthroplasties does not appear to be
influenced by the organism type.

A-126
Predictive roles of N-terminal proBNP and high sensitive Troponin T levels in
determining mortality in chronic renal failure patients

S. Ozdem1, V. T. Yilmaz2, G. Uzun1, L. Donmez3, S. S. Ozdem4, G.

Suleymanlar2, R. Cetinkaya2, F. Ersoy2. 1Akdeniz University Medical
School, Department of Biochemistry, Antalya, Turkey, 2Akdeniz University
Medical School, Department of Internal Medicine, Division of Nephrology,
Antalya, Turkey, 3Akdeniz University Medical School, Department of Public
Health, Antalya, Turkey, 4Akdeniz University Medical School, Department
of Pharmacology, Antalya, Turkey
Background: The mortality in chronic renal failure (CRF) patients is higher than that
in normal population. Cardiovascular events are the major contributors in increased
mortality rate in CRF patients since more than 50% of deaths are due to cardiovascular
events. Cardiac troponins are highly specific and sensitive markers of myocardial
damage. Increments in troponin levels in CRF are related to adverse clinical outcomes.
B-type natriuretic peptide (BNP) is a cardiac neurohormone predominantly released
from the ventricles in response to left ventricular volume expansion and pressure
overload. It is released as a preproBNP and is cleaved into proBNP and a signal peptide.
ProBNP is subsequently cleaved into BNP and the inactive N-terminal proBNP (NTproBNP) peptide. Levels of BNP or NT-proBNP are known to be elevated in patients
with left ventricular dysfunction. BNP may serve as an important plasma biomarker
for cardiac stress and ventricular hypertrophy in patients with CRF. In fact, elevated
levels of BNP indicate an increased risk of cardiac events and mortality in patients
undergoing peritoneal dialysis. Markers with capacity to identify CRF patients with
high mortality risk early in the course of disease may help in clinical management of
these patients improving outcomes. In the present study we investigated the predictive
values of NT-proBNP and high sensitive Troponin T (hsTNT) levels separately and in
combination in determining all-cause mortality in CRF patients on peritoneal dialysis.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S33

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

Methods: A total of 51 CRF patients (21 female, 30 male, mean age: 50.9 16.5
years) and 41 healthy control subjects (22 female, 19 male, mean age: 45.7
11.6 years) were included in the study. Serum levels of NT-ProBNP and hsTnT
(ECLIA method) were measured in all study subjects. Optimal prognostic cutoff points for NT-ProBNP and hsTnT were determined. Overall survival rates
were determined by using the Kaplan-Meier method with log-rank test for
comparison among groups with different cut-off points of NT-ProBNP and hsTnT.
Results: CRF patients had significantly higher levels of NT-ProBNP (5382.00
1115.00 vs 77.009.99 pg/mL, p<0.00001) and hsTnT (0.070 0.010 vs 0.004
0.001 ng/mL, p<0.00001) than control subjects. Of the 51 patients enrolled, 24
died and 27 survived during a median follow-up period of 56.4 months (P25-P75:
32.3-57.3). Optimal prognostic cut-off points for NT-ProBNP and hsTnT were
1950 pg/mL and 0.050 ng/mL, respectively. Patients with elevated hsTnT
levels had higher risk than patients with elevated NT-proBNP levels when
compared with the group which had low levels of both markers (HR 4.00 [95% CI,
1.5710.20], P =0.004 and HR 3.31 [95% CI, 1.407.82], P = 0.006], respectively).
Patients with elevated levels of both hsTnT and NT-proBNP had a markedly
increased all-cause mortality risk (HR 10.38 [95% CI, 2.82-38.27], P <0.0001).
Conclusion: Our results have demonstrated for the first time that determination of
both hsTnT and NT-proBNP levels is better at predicting CRF patients with high
mortality risk than determination of either biomarker alone.

A-127
Enhanced Liver Fibrosis (ELF) Score for the Evaluation of Liver Fibrosis in
Shistossomiasis Mansoni

T. B. Medeiros1, A. L. C. Domingues1, E. P. Lopes1, A. Dellavance2, L.

E. C. Andrade3, M. L. Ferraz3. 1Federal University of Pernambuco,
Recife, Brazil, 2Fleury Medicine and Health, So Paulo, Brazil, 3Federal
University of So Paulo, So Paulo, Brazil
Background: Schistosomiasis mansoni has a worldwide distribution and affects
over 70 million people, constituting a major cause of fibrogenic liver disease. The
assessment of periportal fibrosis by ultrasound is the gold-standard for staging
fibrosis in patients with schistosomiasis. However, although simple and non-invasive,
ultrasound examination is not always available in endemic areas. Aim: to correlate the
ELF score with patterns of periportal fibrosis obtained by ultrasonography in patients
with Schistosomiasis mansoni.
Methods: Patients of both genders, aged> 14 years, coming from an endemic area for
schistosomiasis with a history of exposure to contaminated water were included. An
ultrasonography was performed by a single examiner, with unit Siemens Acuson X
150 with 3.5 MHz convex transducer. Periportal fibrosis was classified according to
the six patterns of Niamey (graded from A to F, with A = no liver fibrosis and F = very
advanced fibrosis). Exclusion criteria were the presence of serum markers of viral
hepatitis, ethanol consumption > 210 g/week and steatosis. All patients had serum
samples collected for determination of ELF score on the same day of ultrasonography.
The evaluation of ELF markers (hyaluronic acid, procollagen III and inhibitor of
metalloproteinase 1) was performed by chemiluminescence (Siemens Healthcare
Diagnostics).
Results: 78 patients were evaluated; 50 (64%) were female, mean age 50 14 years.
According to Niamey classification, patients were divided into three groups: 7 patients
(9%) presented pattern A/B of fibrosis, 33 (42%) pattern C/D and 38 (49%) pattern
E/F. Groups showed the following ELF scores, according to the pattern of fibrosis:
A/B = 8.55 0.40, C/D = 9.31 1.23, E/F = 9.52 0.93. Significant difference was
observed between groups A/B and E/F (p = 0.01).
Conclusion: In patients with Schistosomiasis mansoni the ELF score was able to
discriminate patients with patterns A/B (absent or mild fibrosis) of those with E/F
(advanced fibrosis).The test can be an useful tool for the diagnosis and for monitoring
fibrosis progression in patients with Schistosomiasis mansoni in situations where
ultrasound is not available.

A-128
Circulating blood platelet function in acute liver allograft rejection

Y. Wang. Tianjin First Central Hospital, Key Laboratory for Critical Care
Medicine of the Ministry of Health, Tianjin, China
Background: Platelet surface receptor expression is increased in acute vascular
events in various clinical settings. Endothelial injury is associated with increased

S34

adhesion and aggregation of platelets. In the present study, we investigated platelet

aggregation, and markers of platelet activation in liver allograft recipients with acute
rejection.
Methods: The whole blood samples from 20 recipients with biopsy-confirmed acute
rejection (rejection group), 20 recipients with stable graft function (stable group) after
liver transplantation, and 20 healthy volunteers (control group) were collected. To
examine platelet function we measured platelet aggregation tests by using platelet
aggregometer, platelet expression of glycoprotein (Gp)Ib, GpIIIa, GpIIb/IIIa, and
P-selectin (granule membrane protein 140 [GMP140]) under unstimulated and
activated conditions by quantitative flow cytometry, and plasma soluble P-selectin
(sP-selectin) by enzyme-linked immunoadsorbent assay.
Results: When the platelet membrane GpIb, GpIIIa, and GpIIb/IIIa expression of
rejection group, stable group, and control group were compared in a resting state
and in a agonist [thrombin receptor-activating peptide (TRAP)]-activated state,
no significant differences were observed among three groups (P>0.05). Platelet
expression of P-selectin and sP-selectin levels were significantly increased in rejection
group and stable group compared with control group (P<0.05). A significantly
increased sP-selectin level was found in rejection group compared with stable group
(P<0.01), and platelet P-selectin expression in a TRAP activated state (P<0.05).
Platelet maximum aggregation values induced by all agonists (adenosine diphosphate,
arachidonic acid, collagen, epinephrine) were higher in rejection group than stable
group, but the difference did not reach statistical significance (P>0.05).
Conclusion: Acute rejection induces platelet activation without inducing circulation
platelet aggregation, which reveals the state of ongoing pathogenesis of immune and
inflammatory processes.

A-129
Interferences by hemolysis, lipemia and bilirubin on routine parameters on
chemistry analyzers in a pediatrics hospital

S. Agarwal, G. Vargas, G. Buffone, S. Devaraj. Texas Childrens Hospital

and Baylor College of Medicine, Houston, TX
Background: Clinical laboratory assays can be affected by different interferences,
most commonly, hemoglobin, lipids, bilirubin, auto-antibodies, and heterophile
antibodies. Our goal was to study the effect of interference by hemolysis, lipemia
and bilirubin on routinely performed parameters on various chemistry analyzers
including VITROS 5600, BN Prospec, Dimension Xpand, Architect i1000SR, and
Advia Centaur using pediatric samples in the Texas Childrens Hospital laboratory.
Further we wanted to establish cut-off indices above which these interferences
confound analysis of pediatric samples. Methods: We tested the effect of hemolysis
on K+, AST, LDH, TBIL, ALT, CK, Mg++, ALB/TP, ALKP, Fe, Lipase,NH3+ and
phosphorus on VITROS 5600. Similarly for lipemia interference we analyzed HIV1/2,Testosterone, Progesterone, Ceruloplasmin,Haptoglobin,C3 / C4, IgG / IgM / IgA,
Vitamin D,AFP,aHBC,HCG,CK-MB,TSH, Albumin,Ferritin and Glucose. Lastly, for
icteric interference we performed analyses for Estradiol, Folate, ALB/TP, ALT, GGT,
Glucose, Na+, Cortisol, C4, Haptoglobin and HbA1c on the respective analyzers.
Experiments were carried out with test samples from 3 different serum pools spiked
with increasing concentrations of hemolysate (0.75g/l, 1.5g/l, 3.0g/l, 6.0g/l,), 20%
Intralipid (400mg/dl, 1000mg/dl, 2000mg/dl) and commercially available Bilirubin
(100uM, 250uM and 500uM). These were then analyzed on the various instruments
in laboratory: VITROS 5600, Architect i1000SR, Prospec, Xpand, and Centaur. The
Hemolysis-(H), Lipemia-(T) and Icterus- (I) indices were measured and documented
on VITROS 5600. Lastly, for lipemic interference (>2000mg/dl) we further treated the
samples with Lipoclear(1:5 ratio) and then re-analyzed the samples.
Results: We categorized hemolysis by H-index of 101-250 as mild, 251-500 as
moderate, 501-999 as significant and H-index of >1000 as grossly hemolysed.
Similarly, for Lipemia T-index of < 100 was considered mild,101-500 as moderate
and T- index of >501 as severely lipemic. For Icteric evaluation I-index of 0-5 was
characterized as normal, 5.1-9.9 as mild, 10.0 -19.9 as moderate and 20 as grossly
icteric. For significant hemolysis, the main parameters affected and which shouldnt
be reported included AST, LDH, TBIL, and NH3+. All the measured analytes were
compromised by gross hemolysis and these samples should be rejected. None of the
analyzed parameters were significantly affected by the level of Icterus. With regards
to lipemia, mild lipemia (<400mg/dl) did not affect the assays while Ceruloplasmin,
Haptoglobin, C3, C4, and Immunoglobins (IgG, IgA, IgM) were significantly
affected by moderate (400-1000 mg/dl) and severe lipemia (>2000mg/dl), and hence
should not be reported. Vitamin D also showed a significant decrease in moderately
lipemic samples (decrease of 10-15%). However, since most samples are drawn
in a postprandial state, we tested the effect of Lipoclear. Addition of Lipoclear to
moderately lipemic samples significantly attenuated the effect of lipemia interference
on the above mentioned analytes.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

Conclusion: Accurate reporting of pediatric samples for the analytes affected by

common interferences will lead to better clinical interpretation. Our results can be
applied to other laboratories for the analysis and reporting of these parameters.

A-130
Serum Neopterin Levels in Patients with Pulmonary Embolism

S. Abusoglu1, F. Akyurek1, A. Unlu1, Z. Buyukterzi2, E. Kurtipek2. 1Selcuk

University Faculty of Medicine, Konya, Turkey, 2Meram Training and
Education Hospital, Konya, Turkey
Background: Neopterin (D-erythro-1-2-3-trihydroxypropylpterin) is produced from
guanosine triphosphate by activated human monocytes, monocyte derived dendritic
cells, and macrophages. Release and production of neopterin is stimulated mainly
by interferon-c (IFN-c) released by activated Th1- lymphocytes during the cellular
immune response. Pulmonary thromboembolism (PTE), is an extremely common
medical problem. Yet despite its frequency, much remains to be learned regarding
the pathogenic mechanisms that initiate pulmonary thromboembolism. Marked
activation of endothelium, platelets, and leukocytes are key events in thrombogenesis
in pulmonary thromboembolism. The role of clinical laboratories is to give accurate,
fast and reliable results to clinicians. Also, early detection of diseases as pulmonary
thromboembolism is important to take care before the serious complications. The aim
of this study was to determine serum neopterin concentrations as an early biomarker
on patients with pulmonary thromboembolism.
Methods: Blood samples were collected from 41 healthy control and 38 patients
with pulmonary embolism. Patients with chronic disease and inflammatory
disorders were excluded. This study was approved by local ethic committee. Serum
neopterin levels were analyzed with flourometric detection by high perfomance
liquid chromatography. Briefly, 100 L trichloroacetic acid was added to 500 L
serum sample, vortexed and centrifuged at 10000 g for 10 minutes. Supernatant was
injected to C18 chromatographic column on Agilent HPLC system. Chromatographic
detection was perfomed at 353 and 438 nm excitation and emission wavelength. This
methods coefficient of variation and % bias values were 4.21,3.50; 6.89,5.62 and
8.14,6.64 for 0.5, 5 and 10 mol/L, respectively. Statistical analysis was performed
with SPSS v16.
Results: Serum neopterin levels were significantly lower in control group (5.72 2.29
mol/L) compared to patient group (7.83 4.45 mol/L) (p=0.049) according to
Mann-Whitney U test.
Conclusions: Activated T cells and macrophages synthesise and release a number of
cytokines whose main function is immunoregulation. As a marker of cellular immune
response activation depending on IFN- release, neopterin may better reflect the
disease. Our findings suggest that neopterin levels may be used as an immunological
marker in follow-up the disease.

A-131
How non-standardized serum creatinine-based definitions of acute kidney
injury contribute to disparate reported incidence rates of contrast-induced
nephropathy

J. S. McDonald, R. J. McDonald, E. E. Williamson, D. F. Kallmes. Mayo

Clinic, Rochester, MN
Background- Prior controlled studies of intravenous contrast-induced nephropathy
(CIN), acute kidney injury (AKI) resulting from iodinated contrast material
administration, have reported incidence rates ranging from 2% to 40%. This
discrepancy may be partially explained by the lack of a standardized serum creatinine
(SCr)-based definition of AKI. To date, there is little evidence regarding how different
definitions of SCr-defined AKI and baseline renal function affect the incidence of
AKI, and which, if any, of these definitions provide a reliable means to distinguish
between CIN and contrast-independent AKI. In the current study, we examined how
using different SCr definitions of AKI and baseline renal function affect AKI incidence
rates in a propensity score-matched cohort of contrast-enhanced and unenhanced CT
scan recipients.
Methods- All contrast-enhanced and unenhanced abdominal, pelvic, and thoracic CT
scans performed at our institution from 2000-2010 and accompanying pre- and postscan SCr results were identified. Contrast and noncontrast scan recipients of similar
clinical and demographic characteristics were compared following propensity scorebased 1:1 matching. AKI was defined using absolute (SCr 0.5mg/dl and 0.3mg/dl
over baseline), relative (SCr 25% and 50% over baseline), and hybrid definitions
(SCr 0.5mg/dl or > 25%, and 0.3mg/dl or > 50% over baseline). Baseline renal
function was defined by using either the mean SCr result 24 hours or 7 days prior to

scan. The incidence of AKI was compared between groups using Fishers exact test.
Results- Following 1:1 matching, 21,372 patients were identified who underwent
a contrast enhanced (N=10,636) or unenhanced CT scan (N=10,636). Using the six
definitions of AKI, the incidence of contrast-dependent AKI ranged from 2% to 15%
while the incidence of contrast-independent AKI ranged from 1% to 14%. Regardless
of AKI definition, the rates of contrast-independent AKI were statistically similar
to the rates of contrast-dependent AKI (ORs ranged from 0.91 (95% CI 0.66-1.24),
p=.58 to 1.84 (0.91-3.75), p=.08). Higher incidences of AKI were observed if baseline
renal function was defined using the mean 7-day pre-scan SCr result compared to
using the mean 24-hour pre-scan SCr result.
Conclusion- Differing definitions of AKI and baseline renal function contribute to
disparities in the reported incidence of AKI. However, no definition could extricate
CIN from contrast-independent AKI causes in our cohort.

A-133
Intravenous Contrast Material Exposure is not an Independent Risk Factor for
Dialysis or Mortality

J. S. McDonald, R. J. McDonald, E. E. Williamson, D. F. Kallmes. Mayo

Clinic, Rochester, MN
Background Contrast-induced nephropathy (CIN), defined as acute kidney
injury (AKI) occurring immediately after exposure to iodinated contrast material, is
generally reported to be a self-limited phenomenon. However, concern remains that
CIN can cause irreversible nephrotoxicity resulting in dialysis and death, particularly
in patients with specific pre-existing comorbidities including renal failure, diabetes
mellitus, and congestive heart failure. In the current study, we sought to determine
the true incidence of short-term dialysis and mortality following intravenous contrast
administration among individuals with closely matched demographic and clinical
characteristics using propensity score analysis.
Methods - All contrast-enhanced and unenhanced abdominal, pelvic, and thoracic
CT scans performed at our institution from 2000-2010 and related pre- and postscan serum creatinine (SCr) results were identified. Contrast and noncontrast scan
recipients were compared following propensity score-based 1:1 matching to reduce
intergroup selection bias. Patients with pre-existing diabetes, congestive heart failure,
or chronic or acute renal failure were identified as high-risk patient subgroups for
nephrotoxicity. The effects of contrast exposure on the incidence of AKI (SCr
0.5mg/dl above baseline within 24-72 hours of exposure), and dialysis or death within
30 days of exposure were determined using odds ratios and covariate-adjusted Coxproportional hazard models.
Results - 1:1 propensity score matching yielded a cohort of 21,346 patients (10,673
contrast-enhanced exams/10,673 noncontrast exams). Within this cohort, the risk
of AKI (O.R.=0.94 (95% CI: 0.83-1.07), p=.38), emergent dialysis (O.R.=0.93
(0.52-1.65), p=.89), and 30-day mortality (H.R.=0.97 (0.87-1.06), p=.45) was not
significantly different between contrast and noncontrast groups. Although patients
who developed AKI had overall higher rates of dialysis and mortality, contrast
exposure was not an independent risk factor for either outcome (dialysis: O.R.=0.89
(0.40-2.01), p=.78; mortality: H.R.=1.03 (0.82-1.32), p=.63). Similar findings were
observed among the subgroups of patients with renal failure, diabetes, and congestive
heart failure.
Conclusion - Intravenous contrast material administration was not associated with
excess risk of AKI, dialysis, or death, even among patients with comorbidities
reported to predispose them to nephrotoxicity.

A-134
Neutrophil gelatinase-associated lipocalin (NGAL) and Pentraxin -3 (PTX-3)
Levels in Chronic Renal Failure and their relationship with Inflammation

O. Sar1, N. -. Eren1, S. Cigerli1, B. Aslan2, T. Basturk1, M. Sevinc1, A.

Unsal1. 1ili etfal trainig and research hospital, stanbul, Turkey, 2Quality
Management Program-Laboratory Services, Toronto, ON, Canada
Background: Early diagnosis of Chronic Renal Failure (CRF) provides to initiate
the effective treatments that can decelerate the progress of the disease and improve
prognosis. However, lack of early predictive biomarkers presents difficulties in early
diagnosis. Therefore, finding biomarkers that considerably increases in the early
stages of CRF and demonstrates correlation with the disease progress is necessary.
NGAL is a known biomarker of acute kidney injury that is secreted in response to
renal tubular epithelial cells damage. Increased NGAL levels are determined in the
blood and urine following acute renal injury. Pentraxin-3 (PTX-3) is an inflammatory
mediator. In contrast to other members of the pentraxin family such as CRP and

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S35

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

Serum Amiloid Protein, PTX-3 is produced in the inflammation site and shows close
correlation with the level of tissue damage. Inflammation plays an important role
in the progression of the CRF and contributes to increased mortality and morbidity
observed in CRF.
In this study, we investigated serum NGAL and PTX-3 levels and their relationship
with the severity of renal damage in patients with early stage CRF.
Method: 20 (2 male and 18 female) non diabetic Stage 1 and 34 (17 male and 17
female) Stage 2 CRF patients were included in this study. Average age in the Stage
1 CRF group was 42.557.00, and in the Stage 2 CRF group 47.447.47. In both
groups, serum NGAL and PTX-3 were measured by using the ELISA method from
Bioscience Human and Aviscera Bioscience, respectively.
Results: NGAL levels in patients with Stage 2 CRF were higher than that of the
patients with Stage 1 CRF (p<0.01). ROC analysis and diagnostic tests used for the
calculation of a diagnostic cut off value for Stage 2 CRF. For the diagnostic cut - off
value of 1037 pg/ml, sensitivity was 73.55%, specificity 70.00%, positive predictive
value 80.65%, and negative predictive value 60.87%. Area under the curve was
estimated as 73.3% and its standard erreor was 6.9%.In the patients with NGAL levels
higher than 1037pg/ml, the probability of having Stage 2 CRF was found to be 6481
times higher than that of the patients with NGAL levels less than 1037 pg/ml.In Stage
1 and 2 CRF groups , 19.4% and 80.6% of the patients had NGAL levels higher than
1037, respectively. In Stage 2 cases, there was a positive correlation between serum
creatinin and NGAL concentrations (r=0,362; p0.05). In Stage 1 and 2 CRF patient
groups, CRP and PTX-3 concentrations did not show correlation with eGFR values.
There was no statistically significant correlation between NGAL and PTX-3
Conclusion: The serum NGAL concentrations were found to be higher in Stage 2 CRF
than Stage1 CRF group. Our results indicate that NGAL could be used as a strong
and independent diagnostic biomarker for the early stage CRF and a risk for disease
progression. We did not observe statistically significant correlation between serum
NGAL and PTX-3 levels.

A-135
Serum Lipoprotein-Associated Phospholipae A2 and Methylarginine Levels in
Patients with Pulmonary Embolism

A. Unlu1, F. Akyurek1, S. Abusoglu1, Z. Buyukterzi2, E. Kurtipek2. 1Selcuk

University, Konya, Turkey, 2Meram Training and Education Hospital,
Konya, Turkey
Background: Pulmonary thromboembolism (PTE) is a preventable disease with
higher mortality and morbidity characterized by diagnostic difficulties and recurrence
risk. Obstruction of pulmonary arteries is developed from the detached fragments
of thrombus in deep veins of the lower extremities. Plasma lipoprotein-associated
phospholipase A2 (Lp-PLA2), also known as platelet activating factor acetylhydrolase,
is produced by inflammatory cells, co-travels with low-density lipoprotein (LDL), and
hydrolyzes oxidized phospholipids, thereby propagating inflammation and potentially
thrombosis. ADMA is derived from the catabolism of proteins containing methylated
arginine residues. Higher ADMA concentrations have been measured in many
cardiovascular and metabolic diseases. There has been limited markers for laboratory
evaluation of pulmonary thromboembolism. The aim of this study was to find out
the serum methylated arginine and lipoprotein-associated phospholipase A2 levels in
patients with pulmonary embolism.
Methods: Blood samples were collected from 41 healthy control and 45 patients
with pulmonary embolism. Patients with chronic diseases were excluded. This study
was approved by local ethic committee. Serum lipoprotein-associated phospholipase
A2 test was analyzed with a colorimetric kit on automated system (Abbott C16000).
Reported intra- and inter-assay CV values for 171 and 456 ng/mL were 1.3;0.6 and
3.6;4.9, respectively. Method was linear up to 486 ng/mL. Serum methylated arginine
levels were analyzed with liquid chromatography tandem mass spectrometry on
ABSCIEX API 3200 system. Briefly, 1 mL deuterated d7-ADMA containing methanol
was added to 200 L serum sample for protein precipitation, vortexed and centrifuged
on 13000 g for 10 minutes. 40 L supernatant was injected to Shimadzu LC-20AD
HPLC system. Chromatographic seperation was performed on C18 column for 5
minutes. This methods coefficient of variation and % bias values were 15.6,10.2;
9.72,7.88 and 6.45,6.02 for 0.4, 0.8 and 1.6 mol/L, respectively. Statistical analysis
was performed with SPSS v16.

group (0.70 0.17 mol/L) (p<0.001). Serum arginine levels were significantly lower
in patient group (163 81 mol/L) compared to control group (285 127 mol/L)
(p<0.001). Serum citrulline levels were significantly lower in patient group (9,14
6,20 mol/L) compared to control group (21,81 7,26 mol/L) (p<0.001). Serum
arginine/ADMA ratio was significantly lower in patient group (379 145 mol/L)
compared to control group (477 205 mol/L) (p=0.013).
Conclusion: ADMA has been evaluated in several different classes of pulmonary
hypertension. In previous studies, no significant difference was reported between
pulmonary embolism and healthy controls. Also, in this study serum methylated
arginines and phospolipase A2 levels were found to be statsitically lower compared
to control group. This might be due to venous characteristic of this disease. Serum
asymmetric dimethylarginine levels may be affected by several factors. Arginine/
ADMA levels may provide accurate data and be a reliable marker compared to single
ADMA measurements.

A-136
Prevalence of metabolic syndrome among hypertensives in Ghana

C. NKRUMAH1, W. OWIREDU2. 1METHODIST HOSPITAL, WENCHI,

Ghana, 2KWAME NKRUMAH UNIVERSITY OF SCIENCE AND
TECHNOLOGY, KUMASI, Ghana
Background: Metabolic syndrome can be found in approximately one-third of
patients who do not have diabetes but have hypertension. There are numerous
correlations between the metabolic syndrome and hypertension, although this is not
always the case. As metabolic syndrome and hypertension are independent risk factors
for the same disease process, namely cardiovascular disease, it is possible that patients
suffering from both these disease entities may have a compounded risk. Our study will
therefore attempt to determine the prevalence of metabolic syndrome and investigate
the proposed association between these two disease entities.
Methods: This cross-sectional study was conducted at the Hypertension Clinic of the
Department of Medicine, Komfo Anokye Teaching Hospital (KATH), Kumasi, Ghana
between April 2009 and November 2010. Informed consent was obtained from 300
participants consisting of 200 hypertensives (diagnosed by a Consultant Physician
based on WHO International Society of Hypertension Guideline of blood pressure
140/90 mmHg or use of antihypertensive) and 100 apparently healthy normotensives
served as control. Blood samples were collected in the morning after an overnight fast
of at least 12 hours. Serum and plasma were stored at -80oC after centrifugation at
2000g for 5 minutes until assayed. Fasting blood glucose (FBG), Apolipoprotein A-1
(Apo A-1), Apolipoprotein B (Apo B), Total Cholesterol (TC), Triglyceride (TG) and
High Density Lipoprotein (HDL) were measured on an Auto-Analyzer (Flexor junior,
Vital Scientific N.V., The Netherland).
Results: The prevalence of MetS among the hypertensive patients were significantly
higher than the normotensive control (56.5% vrs 9.0%, 54.5% vrs 5.0% and
65.5%vrs15.0%, p<0.001) using NCEP ATP III, WHO and IDF criteria respectively.
Irrespective of the criteria applied, all the components of MetS were significantly
higher among the hypertensive patients as compared to the normotensive control.
Among the hypertensive patients, the highest prevalence of cardiovascular risk factor
was abdominal obesity as measured by WHR (77.0%), followed by reduced HDLcholesterol (74.0%). From the univariate analysis, females were at about 3 times
at risk of developing hypertension as compared to the male counterpart (OR = 2.7;
95% CI = 1.6-4.4; p = 0.0000). Reduced apolipoprotein A1 served as a risk factor
(aOR = 13.4; 95% CI = 1.5-121.4; p = 0.0210) whilst high apolipoprotein A1 protects
the individual from developing hypertension (aOR = 0.1; 95% CI = 0.0-0.2; p =
0.0000). High apolipoprotein B poses about 9 times risk of developing hypertension
as compared to the normal level (aOR = 9.3; 95% CI = 4.2-20.9; p = 0.0000). Both
Impaired fasting glucose and diabetes each pose more than 10 times risk of developing
hypertension as compared to normoglycaemia.
Conclusion: The study demonstrated that, hypertension is more than just elevated
blood pressure in our setting; it is intimately associated with the metabolic syndrome.
There is therefore the need for metabolic screening of all hypertensives and increase
awareness on the critical importance of public health strategies aimed at reducing risk
factors in the entire population. Early detection and treatment of the global risk profile
should thus become a priority.

Results: Serum phospholipase A2 levels were significantly lower in patient group

(158.3 49.8 ng/mL) compared to control group (206.5 38.4 ng/mL) (p<0.001)
according to independent sample t test. Serum asyymetricdimethylarginine levels
were significantly lower in patient group (0.44 0.17 mol/L) compared to control
group (0.60 0.18 mol/L) (p<0.001). Serum syymetricdimethylarginine levels
were significantly lower in patient group (0.54 0.20 mol/L) compared to control

S36

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

A-137
CEDIA Mycophenolic Acid Applications on the Beckman Coulter AU480,
AU680 and AU5800 Analyzers

D. L. Cheng, C. Wong. Thermo Fisher Scientific, Fremont, CA

Background: Mycophenolic Acid (MPA) is metabolized from Mycophenolate Mofetil
(MMF) or Mycophenolate Sodium. It is an immunosuppressant used in the prevention
of tissue rejection for patients who have undergone renal and heart transplants. MPA is
a specific inhibitor of inosine-monophosphate dehydrogenase (IMPDH) - an enzyme
used by B and T lymphocytes for de novo purine synthesis. This repression of B and T
cell proliferation results in the desired immunosuppressive effects.
Methods: The CEDIA MPA Assay is based on the enzyme -galactosidase, which has
been genetically engineered into two inactive fragments. MPA in the human plasma
patient sample competes with MPA conjugated to one inactive fragment for antibody
binding site. Once MPA in the sample binds to antibody, inactive enzyme fragments
reassociate to form active enzymes. The amount of active enzymes results in an
absorbance change that is directly proportional to the amount of MPA in the patient
sample. This change is measured spectrophotometrically for a quantitative value of
concentration. The Beckman Coulter AU480/AU680/AU5800 analyzers are new
applications for the CEDIA MPA Assay. Analyzer performance was determined for
precision, linearity, limit of detection, and accuracy on the Beckman Coulter AU480/
AU680/AU5800 clinical chemistry analyzers, over the range of the assay (0.3-10 g/
mL). Results were measured against the reference analyzer Hitachi 917.
Results: All studies were evaluated using CLSI guidelines. Three levels of MPA
controls were used in the studies. The precision ranged from 6.1-2.2%CV for withinrun and 7.7-2.5%CV for total run. Linearity was measured and confirmed over a range
of 1.8-11.9 g/mL. The limit of detection on the AU480/AU680/AU5800 yielded
0.07 g/mL. Accuracy was measured using patient correlation against the reference
analyzer Hitachi 917, which yielded a Demings Regression for each analyzer: AU480
= 0.992*(Hitachi 917) - 0.096 (N = 107, r = 0.998), AU680 = 0.995*(Hitachi 917)
- 0.043 (N = 107, r = 0.998), AU5800 = 0.993*(Hitachi 917) - 0.089 (N = 107, r =
0.998).
Conclusion: All measured studies demonstrated acceptable performance, validating
the use of the CEDIA MPA Assay on the Beckman Coulter AU480/AU680/AU5800
analyzers, and will provide an effective monitoring system for patients receiving
MMF or Mycophenolate Sodium therapy.

A-138
Fructosamine and Glycated albumin with Risk of Coronary Heart Disease and
Death

E. Selvin1, A. M. Rawlings1, P. Lutsey2, J. Pankow2, L. Kao1, M. Steffes2,

J. Coresh1. 1Johns Hopkins, Baltimore, MD, 2University of Minnesota,
Minneapolis, MN
Background:HbA1c is the standard measure to monitor glucose control and is now
used for diagnosis of diabetes. Fructosamine and glycated albumin are markers of
short-term glycemic control that may add complementary information to HbA1c.
However, the associations of fructosamine and glycated albumin with cardiovascular
outcomes are uncharacterized.
Methods: measured glycated albumin and fructosamine in 11104 adult participants
(792 with a history of diabetes) of the community-based ARIC Study without
cardiovascular disease at baseline (1990-1992). We evaluated the associations of
fructosamine and glycated albumin with incident coronary heart disease and total
mortality. We compared these associations to those for HbA1c.
Results: Baseline HbA1c was highly correlated with fructosamine (Pearsons r =0.82)
and glycated albumin (Pearsons r=0.86). During over two decades of follow-up there
were 1,032 new cases of coronary heart disease and 2,594 deaths. In Cox proportional
hazards models adjusted for traditional cardiovascular risk factors, elevated baseline
levels of fructosamine and glycated albumin were significantly associated with
coronary heart disease and total mortality (Figure). After additional adjustment for
HbA1c, the associations were attenuated but remained significant, particularly at
diabetic levels of fructosamine and glycated albumin. The associations with death
were J-shaped, with an elevation of risk also apparent at the lowest levels of each
biomarker, as has been previously observed for HbA1c.
Conclusion:The acceptance of new measures of hyperglycemia is partly dependent on
establishing their association with long-term outcomes. We found that fructosamine
and glycated albumin were associated with coronary heart disease and mortality and
that these associations were similar to those observed for HbA1c. The elevated risk
of death at very low levels of fructosamine, glycated albumin, and HbA1c deserves
further examination.

A-140
Study of the Association between Endothelial Nitric Oxide Synthase G894T
Gene Polymorphism and Heart Failure Severity

d. I. hashad, S. Marzouk, M. sadaka, R. Zaghlool. FACULTY OF

MEDICINE, Alexandria, Egypt
Objectives: The aim of the study was to investigate the association between
endothelial nitric oxide synthase (eNOS) G894T gene polymorphism and severity of
congestive heart failure (CHF) in a group of Egyptian patients.
Methods: This study was conducted on 30 consecutively selected Egyptian patients
admitted to the Cardiology department at Alexandria Main University Hospital
suffering from non valvular CHF which was documented clinically and by ECG.
Thirty age & sex matched healthy individuals were also included in the study as a
control group. eNOS G894T gene polymorphism was studied using polymerase chain
reaction-Restriction fragment length polymorphism-(PCR-RFLP) Results:. Genotype
distribution of eNOS G894T gene polymorphism among patients revealed 13 (43.3%)
patients of GG genotype (wild type), 11 (36.7%) of GT genotype (heterozygote) and
6 (20%) were of the TT genotype (homozygote for the mutant allele). Among the
controls, 16 (53.3%) were GG genotype, 12 (40%) were GT genotype and 2 (6.7%)
were TT genotype. Thus, no statistically significant difference in genotype distribution
between patients and controls was observed. In addition, the present study showed
no statistically significant difference between different eNOS patients genotypes as
regards NYHA classes, ejection fraction and six-minute walk test. The prevalence
of TT genotype was significantly higher in ischemic cardiomyopathy patients when
compared to those who had dilated cardiomyopathy and also in hypertensive patients
as compared to normotensive patients.
CONCLUSIONS: Although the mutant eNOS G894T allele is associated with both
hypertension and dilated cardiomyopathy in Egyptian heart failure patients, it was not
correlated to disease occurrence or severity. 1

A-141
Seasonal Variation of Vitamin D Deficiency in a Large Rural Health System:
Effect of Assay Change on Deficiency Prevalence

J. B. Jones, H. H. Harrison, D. D. Sargent, J. Labarbera, M. Sneidman.

Geisinger Health System, Danville, PA
25 hydroxy Vitamin D (25OH Vit D) continues to be a commonly ordered test to
assess bone health in the large, rural Geisinger Health System outpatient population in
central Pennsylvania USA (latitude 40-41 degrees N). It is well known that seasonal
variation of 25OH Vit D levels exists in temperate latitudes as sunlight exposure
varies. Our objective is to report practical impact of monthly surveillance of 25OH Vit
D seasonal deficiency/sufficiency prevalence during 2009 to 2013 (N=228,714) using
clinical cutoffs of <20, 20-30, and 30-100 ng/mL to judge Deficiency, Insufficiency,
and Sufficiency, respectively.
Monthly test volumes increased dramatically in 2009 when the test was first
performed in house (2300/month in July 2009 to 4000/month in Jan 2010). Since
then the volume has steadily increased to 5500/month. Our osteoporosis management
program reports that patients are uniformly monitored with 25OH Vit D and are at

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S37

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

steady state as far as population enrollment stands. Steady increases of 25OH Vit
D test volumes (approximately 15%/year) may be attributed to generally increasing
numbers of outpatients rather than significant changes in utilization.
Over the course of July 2009 to December 2013, cyclical seasonal variation was
documented by monthly prevalence plots, with deficiency rates lowest in July-AugustSeptember (16.5%) and highest in Jan-Feb-March (27.7%). Inversely, sufficiency
rates were highest in Jul-Aug- Sept (48.5%) and lowest in Jan-Feb-Mar (40.5%).
Because deficiency rates repeatedly increase from summer to winter by an average of
68%, it is important to advise clinicians to interpret 25OH Vit D levels and prescribe
supplementation taking the month of the year into account. Although common practice
is to supplement 25OH Vit D in patients with deficient or insufficient levels (i.e. < 30
ng/mL), a systematic study of prescribing patterns has not yet been undertaken.
In November 2012, the 25OH Vit D immunoassay was changed from Diasorin to
Roche. Correlation between the two assays using patient specimens (N=152) was
acceptable; Deming slope= 0.963; correlation coefficient=0.8455. Ongoing monthly
deficiency prevalence, although similar in amplitude and periodicity, changed
somewhat with Roche Jan- Feb-Mar 2013 deficiency prevalence trending 11%
higher and Roche Jul-Aug-Sept 2013 deficiency prevalence trending 24% lower than
prior comparable monthly Diasorin results. It is too early to tell if this difference is
physiological , analytical, or coincidental. Although the population surveillance data
are unsorted by patient clinical status, the assay change per se may have an effect on
shifting population outcome data, and should be more extensively studied along with
more traditional specimen correlation.
We conclude that seasonal variation of 25OH Vit D significantly affects deficiency
prevalence and has the potential to change treatment on a population basis. The season
of testing should be factored into treatment and other clinical considerations based on
limited duration cross-sectional studies. Additional studies are needed to determine
if different assays perform differently in populations exposed to different amounts of
sunlight and if these differences may be due to levels of Vit D precursor or binding
protein.

A-142
Evaluation of circulating levels of inflammatory and bone formation markers in
Axial Spondyloarthritis

T. S. Frode, K. R. de Andrade, G. R. W. de Castro, G. Vicente, J. S. da

Rosa, M. Nader, I. A. Pereira. University Federal of Santa Catarina, Brasil,
Brazil
Background: Studies have demonstrated the important role of bone remodelling
and osteoimmunology in the progression of inflammatory lesions in Axial
Spondyloarthritis (SpA) disease.
Objective: This study was conducted to evaluate the inflammatory response by
analysis of the serum levels of pro-inflammatory and new bone formation markers in
patients with SpA who were treated or not treated with anti-tumor necrosis factor-
(anti-TNF-) or non-steroidal drugs (NSAIDs) and to correlate these markers with the
clinical evaluation scores of the disease activity.
Methodology: The serum levels of myeloperoxidase (MPO), adenosine deaminase
(ADA), nitric oxide metabolites (NOx), bone alkaline phosphatase (BAP),
Dickkopf-1 (DKK-1), and osteoprotegerin (OP) were measured in 52 SpA patients
who were treated or not with anti-TNF- or NSAIDs and in 26 healthy controls using
colourimetric and enzyme immunoassays. The activity and the severity of illness in
patients with SpA were assessed using questionnaires (BASMI, Bath Ankylosing
Spondylitis Metrology Index; BASFI, Bath Ankylosing Spondylitis Functional Index;
and BASDAI, and Bath Ankylosing Spondylitis Disease Activity Index). The data
were expressed as the mean standard deviation of the mean (SD) for symmetric
distribution or in percentage. Comparison of clinical parameters and bone biomarkers
between all groups were done with analysis of variance (ANOVA) followed by
Bonferroni post hoc test for multiple comparisons among the groups. P < 0.05 was
considered to be statistically significant. Results: A significant difference between the
controls and the patients without medication was observed in relation to NOx, BAP,
and OP (p < 0.01). When the patients were compared with regard to their treatment,
there were no clinically significant differences between the groups (p > 0.05).
Conclusion: The NOx, BAP, and OP are emerging as an important inflammatory
pathway in Axial Spondyloarthritis. Also the anti-TNF- or non-steroidal drugs
reduce the inflammation and destruction, however these treatment does not modify
the serum levels of these biomarkers.

S38

A-143
Suspected Lassa Fever (LF) Case Outcomes: A Comparison to a Non-Febrile
Population in Sierra Leone

B. L. Brown1, M. L. Boisen1, L. M. Moses2, J. S. Schieffelin2, A. Goba3,

M. Momoh3, M. Fullah3, F. K. Kamara3, L. Kanneh3, S. H. Khan3, D.
Oottamasathien1, D. S. Grant4, D. Beckham5, K. R. Pitts1, J. Barnett5, R. F.
Garry2. 1Corgenix, Inc., Broomfield, CO, 2Tulane University, New Orleans,
LA, 3Kenema Government Hospital, Kenema, Sierra Leone, 4Ministry of
Health and Sanitation, Kenema, Sierra Leone, 5University of Colorado
Anschutz Medical Campus, Denver, CO
Background: Lassa virus (LASV) is the causative agent for Lassa fever (LF), causing
an estimated 100,000 - 300,000 cases annually. The precise factors resulting in fatal
outcome in LF patients are still largely uncharacterized, although hypovolemic shock
is thought to ultimately result in death of afflicted subjects. Additionally, signs of
acute renal failure are consistently noted preceding fatal outcomes. Evaluating the
difference in clinical and laboratory outcomes between LF cases and non-febrile
controls is important to better characterize the clinical presentation of patients with
LASV infection and potentially select patients with a higher pre-test probability of
infection for diagnostic testing.
Methods: This is a case controlled study of patients suspected of LASV infection,
identified in Sierra Leone, West Africa. Cases include both suspected and confirmed
LF patients who had a temperature of 38C for <3 weeks and displayed clinical signs
and symptoms of a LASV infection upon examination. Controls include non-febrile
Sierra Leoneans with a temperature of 37.5C. We measured five outcomes and
compared results between the confirmed LF case group (n = 57) and the non-febrile
control group (n = 118). We measured BMI, Pulse, BUN, Cr, and BUN:Cr ratio and
correlation of clinical parameters with diagnosis and other clinical parameters. Other
outcomes such as the metabolic panel and hemoglobin were collected and evaluated
as well.
Results: Using the Wilcoxon-Mann-Whitney nonparametric test, we found that
confirmed LF cases exhibit elevated BUN (p<.0001) and Cr (p=.0002) measurements
(BUN: 10.5 mM/L, Cr: 216. 3 M/L) compared to non-febrile controls (BUN:
3.4mM/L, Cr: 84.6M/L). Confirmed LF cases also had a significantly higher
respiratory rate (p<.0001) than the non-febrile controls (39.6 breaths per minute
versus 21.6).
Conclusion: Early evaluation of BUN, Cr, and respiratory rate in febrile patients
may aid in selecting patient populations at high risk for LF and improve diagnostic
accuracy.
BMI, Pulse, BUN, Cr, and BUN:Cr compared between confirmed LF cases and
non-febrile controls.
Confirmed LF Cases Non-febrile Controls
p-value
BMI
18.6
21.9
0.0006
Pulse
93.8
81.4
0.0004
BUN mM/L
10.5
3.4
<0.0001
Cr M/L
216.3
84.6
0.0002
BUN:Cr mg/dL 19.2
12.7
0.0008

A-144
A comparative study of the prevalence of Salmonella typhi infection in the
wenchi municipality using the Widal and culture methods

C. NKRUMAH1, M. AMOAH2. 1METHODIST HOSPITAL, WENCHI,

Ghana, 2COLLEGE OF HEALTH, KINTAMPO, Ghana
Background: Salmonella typhi is a Gram negative bacterium that can cause typhoid
fever. Worldwide, typhoid fever is a serious public health problem, with an estimated
22 million cases, resulting in 200,000 deaths. The burden of the disease lies mainly
in developing countries including Ghana where the provision of sanitary conditions
may be inadequate. However, not much research work has been done to explore
the problem confronting citizens and ways to eliminate them. Yearly report from
Methodist Hospitals Laboratory, Wenchi (2012), indicates that 45.8% and 47.5% of
In-patients and Out-patients respectively reacted to the Widal test. The total percentage
of reactive cases in 2012 was 47.1%.This study was then carried out to determine the
prevalence of typhoid fever and propose recommendations on the appropriate way of
diagnosing the disease in Wenchi municipality.
Methods: Non-interventional exploratory study using a purposive sampling for
the selection of respondents in March and April 2013. The study comprised 100
participants who were suspected of having typhoid fever and referred to the laboratory
for diagnosis. Informed consent was obtained from each participant and then blood
and stool specimens were collected. The blood was centrifuged and Salmonella

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

typhi H and O antibody titre determined following serial dilution with the Widal kits
(Salmonella typhi Antigens H and O). After emulsification with physiological saline,
the stool samples were then cultured directly onto Desoxycholate Citrate Agar (DCA).
The plates were then incubated at 37 oC in for 24 hours and observed for bacterial
growth. Culture results were then compared with the Widal test for each participant.
Results: About 41% of participants had high antibody titre to S. typhi H antigen while
12% also reacted to S. typhi O.The majority of the respondents were female (61%) and
had the lowest reaction to the Widal Salmonella typhi H (42.6%) and O (11.5%) test
but with no bacterial growth on the Agar. The males (29%) had the highest reaction to
the Widal Salmonella typhi H (51.7%) and O (17.2%) test and also with no bacterial
growth on the agar. Irrespective of the antibody titre shown in the Widal test, there was
no corresponding bacterial growth by culture.
Conclusion: The high antibody titres recorded by the Widal technique with no
corresponding bacterial growth by culture may point to the existence of circulating
antibodies established during previous exposure. Therefore relying solely on the
Widal test may lead to over diagnosis of the infection and concomitant abuse of
antibiotics which may contribute the emergency of resistant strains in our region.
Proper diagnosis of Salmonella typhi should thus be based on bacteria culture and
sensitivity testing.

A-145
Performance of REBA MTB-XDR to detect Extensively Drug-resistant
Tuberculosis in a High burden country

Y. Lee1, M. Kang2, H. Jung3, S. Choi4, K. Jo5, T. Shim5. 1Division of

Pulmonary and Critical Care Medicine, Department of Internal Medicine,
Busan Paik Hospital Inje University College of Medicine, Busan, Korea,
Republic of, 2YD R&D Center, YD Diagnostics, Yongin-Si, Gyeonggi-do,
Korea, Republic of, 3Deparment of Internal Medicine, Inje University
School of Medicine, Ilsan Paik Hospital, Goyang, Gyeonggi-do, Korea,
Republic of, 4Department of Internal Medicine, Sanggye Paik Hospital, Inje
University College of Medicine, Seoul, Korea, Republic of, 5Department
of Pulmonary and Critical Care Medicine, University of Ulsan College of
Medicine, Asan Medical Center, Seoul, Korea, Republic of
Setting: Multidrug-resistant/extensively drug-resistant tuberculosis (MDR/XDRTB) is a serious problem worldwide. The early diagnosis and treatment of MDR/
XDR-TB is very important. The REBA MTB-XDR (REBA-XDR) line probe assay
has been developed recently for the detection of resistance to ofloxacin, kanamycin,
and streptomycin in Korea.
Objective: The aim of this study was to evaluate the diagnostic accuracy of the
REBA-XDR in detecting XDR-TB in acid fast bacilli (AFB) smear-positive sputum
specimens.

A-146
Circulating Presepsin (Soluble CD14 Subtype) in Patients with Severe Sepsis
and Septic Shock. Data from the Albumin Italian Outcome Sepsis (ALBIOS)
Study

S. Masson1, P. Caironi2, C. Fanizza3, R. Bernasconi1, L. Pirozzi3, A. Noto4,

V. Parrini5, G. Pasetti6, G. Tognoni3, L. Gattinoni2, R. Latini1. 1IRCCS Istituto di Ricerche Farmacologiche Mario Negri, Milan, Italy, 2Fondazione
IRCCS Ca Granda - Ospedale Maggiore Policlinico, Universit degli
Studi di Milano, Milan, Italy, 3Fondazione Mario Negri Sud, Santa Maria
Imbaro, Italy, 4A.O. San Paolo - Polo Universitario, Milan, Italy, 5Ospedale
del Mugello, Borgo San Lorenzo, Italy, 6Ospedale San Giovanni di Dio,
Orbetello Scalo, Italy
Background: The cornerstone of the emergency treatment of severe sepsis and septic
shock is an early, goal- directed therapy. Presepsin, a soluble fragment of CD14 that
participates in the innate recognition of pathogens, has been proposed as a novel
diagnostic and prognostic marker in sepsis.
Aims: To validate presepsin as a prognostic marker in a large, representative cohort
of septic patients.
Methods: Blood samples were collected 1, 2 and 7 days after enrolment in 997
patients with severe sepsis or septic shock enrolled in the multicenter Albumin
Italian Outcome Sepsis (ALBIOS) trial (NCT00707122). Presepsin was measured
in a central laboratory with a chemiluminescent enzyme immunoassay (PATHFAST
Presepsin, Mitsubishi Chemical). The relation between presepsin concentration
and mortality in Intensive Care Units (ICU) or at 90 days was assessed with Cox
proportional hazards multivariable models. Prognostic discrimination was tested with
reclassification metrics.
Results: Concentration of presepsin on day 1 was 946 [492-1887] ng/L (median [Q1Q3]) and increased gradually with the number of organ dysfunctions and the number
of newly developed organ failures in ICU. Circulating presepsin rose in decedents
over 7 days in ICU while it markedly decreased in survivors (p<0.0001 for timesurvival interaction). Presepsin on day 1 independently predicted ICU and 90-day
mortality (Figure). Addition of presepsin measured on day 1 on top of all risk factors
significantly improved prognostic discrimination and correctly reclassified the risk of
90-day mortality (c-statistics from 0.77 to 0.79, p=0.004; integrated discrimination
improvement IDI = 0.03 [0.02-0.04], p<0.0001; continuous net reclassification index
NRI = 0.53 [0.40-0.65], p<0.0001).
Conclusion: Presepsin is a robust predictor of mortality in patients with severe sepsis
and septic shock. It provides incremental information on top of widespread clinical
risk factors and may help in early risk stratification.

Design: We prospectively enrolled 104 patients with AFB smear-positive specimens

between July 2010 and January 2013. Mycobacterium tuberculosis was cultured in all
samples. The performance characteristics were compared between the REBA-XDR
and the conventional drug susceptibility test (DST), GenoType MTBDRsl assay, and
DNA sequencing analysis results. The conventional DST results were considered to
be the gold standard.
Results: Among the 104 specimens, MDR-TB was found in 29.8% (31/104) and
XDR-TB in 7.7% (8/104). The sensitivity of the REBA-XDR in detecting resistance
to ofloxacin, kanamycin, and streptomycin was 66.7%, 90.9%, and 60.0%, and the
specificity was 93.3%, 93.5%, and 85.4%, respectively. The positive predictive
values were 62.5%, 62.5%, and 40.9%, and the negative predictive values were
94.3%, 98.9%, and 92.7%, respectively. The accuracy was 89.4%, 93.3%, and 81.7%,
respectively. Discordant results between REBA-XDR and conventional DST were
found for ofloxacin in 11 samples (10.6%), for kanamycin in 7 samples (6.7%), and
for streptomycin in 19 samples (18.3%). Two of 10 (20%) discordant REBA-XDR
results for ofloxacin resistance and all 6 (100%) discordant specimens for kanamycin
resistance corresponded to the results of the gold standard DNA sequencing test. Five
of 10 (50%) discordant results for ofloxacin resistance and 2 of 6 (33.3%) discordant
results for kanamycin in the MTBDRsl test corresponded to the results of the DNA
sequencing test.
Conclusion: REBA-XDR seems to be a useful kit to rule in XDR-TB in a high-risk
group for drug resistance, especially for the detection of kanamycin resistance.

A-147
Neutrophil-Gelatinase -Lipocalin in adult cardiac surgery patients: beyond AKI

C. Cosma1, M. Zaninotto1, R. Bianco2, M. Kosovi2, G. Gerosa2, M.

Plebani1. 1Department of Laboratory Medicine, University-Hospital,
Padova, Italy, 2Division of Cardiac Surgery, Department of Cardiac,
Thoracic and Vascular Sciences, University, Padova, Italy
Background: Neutrophil gelatinase-associated lipocalin (NGAL) has been largely
described as an early marker of acute kidney injury. We report the association of
urinary NGAL (uNGAL) values with different adverse outcomes in adult cardiac
surgery patients, particularly the need for continuous venous hemofiltration (CVVH),
cardiac mechanical assist devices and low cardiac output syndrome

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S39

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

Materials and methods: fresh urine sample of 137 patients (34 females and 103 males;
mean of age 64 y) undergoing cardiac surgery (coronary artery bypass, artificial
heart valve, heart transplants, complex cardiac surgery) have been collected before,
immediately after surgery and then 24 h and 48 h after. AKI was defined as an increase
in plasma creatinine levels (more than 50% or more than 0.3 mg/dL (26.5 umol/L)
in comparison to the preoperative value during the first 48 h after surgery. uNGAL
has been evaluated using a chemiluminescent microparticle immunoassay (CMIA,
Architect, Abbott Diagnostic) showing during the study a total imprecision (CV)
ranging from 3.85 to 2.35% (20 ug/L and 1218 ug/L, respectively). The reference
ranges adopted are 2-120 ug/L for uNGAL and 53-97 (females), 62-115 umol/L
(males) for plasma creatinine respectively.
Results: Mean uNGAL levels peaked immediately after cardiac surgery (630
890(SD) ug/L), and remained significantly higher 24 and 48 hours after surgery
(565 842 and 293455 respectively, p=0.0003). 24 patients (17%) developed AKI
(mean of basal creatinine values and 48 hours after cardiac surgery 10630.9 and
346.5127.57 respectively, p<0.001)
Accuracy of u NGAL (AUC) as a predictor for AKI immediately after cardiac surgery
and 24 and 48 hours later was 0.775 (95% confidence interval [CI], 0.660 to 0.890),
0.864 (95% CI, 0.788 to 0.940) and 0.838 (95% CI, 0.749 to 0.926), respectively.
uNGAL levels correlated significantly with the need of CVVH, cardiac mechanical
assist devices and low cardiac output syndrome being AUC 0.862 (95% CI, 0.779 to
0.945), 0.913 (95% CI; 0.840 to 986) and 0.910 (95% CI; 0.840 to 0.981), respectively.
Conclusions: Our study confirms the clinical usefulness of NGAL measurement
for the early diagnosis of AKI but underlines that relevant clinical informations on
adverse outcome of patients undergoing cardiac surgery may also be obtained.

Robust measurement of branched chain amino acids on the Vantera Clinical

Analyzer and the clinical association of NMR-measured valine with type 2
diabetes

J. Wolak-Dinsmore, I. Shalaurova, S. P. Matyus, M. A. Connelly, J. D.

Otvos. LipoScience, Inc., Raleigh, NC

A-148
Diabetes, pre-diabetes and incidence of subclinical myocardial injury

E. Selvin , M. Lazo , L. Shen , J. Rubin , J. McEvoy , R. Hoogeveen , C.

Ballantyne2, J. Coresh1. 1Johns Hopkins, Baltimore, MD, 2Baylor College
of Medicine, Houston, TX
1

A-149

Background: Persons with pre-diabetes and diabetes are at high risk for
cardiovascular events. However, the relationships of pre-diabetes and diabetes to
development of subclinical myocardial damage are unclear. Our objective was to
characterize the associations of pre-diabetes, undiagnosed diabetes, and diagnosed
diabetes with 6-year incidence of subclinical myocardial injury, as assessed by a novel
highly sensitive assay for cardiac troponin T (hs-cTnT).
Methods: We measured hs-cTnT at two time points, 6 years apart, among 8692
participants of the community-based Atherosclerosis Risk in Communities (ARIC)
Study without a history of heart disease, silent MI by ECG, or stroke at baseline
(1990-92). The primary outcome was incidence of elevated hs-cTnT (14 ng/L) at 6
years of follow-up.
Results: Cumulative probabilities of elevated hs-cTnT (14 ng/L) at 6 years among
persons with no diabetes, pre-diabetes (HbA1c 5.7-6.4%), undiagnosed diabetes
(HbA1c 6.5%), and diagnosed diabetes were 4.1%, 7.6%, 10.2%, and 16.9%,
respectively. Across these same categories and compared to persons with no diabetes
and HbA1c <5.7% (reference), the adjusted relative risks for incident elevated hscTnT were 1.40 (95%CI 1.13, 1.73), 1.85 (95%CI 1.24, 2.75), and 2.85 (95%CI 2.14,
3.75).
Conclusion: Pre-diabetes and diabetes were strongly associated with the future
development of elevations in troponin T far below the threshold for a diagnosis of
myocardial infarction. The results from this community-based prospective study
provide evidence for a deleterious effect of hyperglycemia on the myocardium,
possibly reflecting a microvascular etiology. Our findings underscore the importance
preventing progression to early hyperglycemic states and development of diabetes.

Background: Metabolomic studies have shown that branched chain amino acid
(BCAA) levels are independently associated with insulin resistance and type 2
diabetes (T2D). NMR technology has been employed for years to measure lipoprotein
particle concentrations in a clinical laboratory setting. However, the information-rich
nature of the NMR spectrum lends itself to the measurement of other clinically useful
metabolites; therefore, assays were developed to quantify the BCAAs, valine, leucine
and isoleucine.
Methods: Proton NMR spectra were collected on fasting serum samples using the
Vantera Clinical Analyzer, a 400MHz NMR platform with automated fluidics sample
handling, data processing and analysis. The NMR spectra were deconvoluted using
proprietary software with models containing reference spectra from serum proteins
and lipoproteins. For method comparison purposes, NMR-measured BCAAs were
compared with mass spectrometry quantified BCAAs collected from the same
serum samples. Valine concentrations were quantified from NMR spectra previously
captured for participants in the Multi-Ethnic Study of Atherosclerosis (MESA)
and multivariable logistic regression analyses were performed to interrogate the
association of valine with the development of T2D.
Results: The levels of valine, leucine and isoleucine quantified in serum samples
using NMR and mass spectrometry, were highly correlated (e.g. valine R2=0.96). For
the valine NMR assay, the coefficients of variation (CVs) for the inter-assay and intraassay precision data were 2.7-5.9%. NMR-measured valine concentrations in MESA
subjects (n=3309), who were non-diabetic and whose fasting plasma glucose was
<110 mg/dL, were strongly associated with incident diabetes and made a statistically
significant contribution to a logistic regression model containing age, gender, race and
glucose (see Table).
Conclusions: Levels of the BCAAs valine, leucine and isoleucine can be obtained
from the same NMR spectra acquired for lipoprotein particle concentrations. Similar
to published BCAA literature, NMR-measured valine is strongly associated with
incident diabetes.

Age
Gender
Race
Glucose
Valine

S40

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

MESA
(n=352/3309)
(incident diabetes/total # subjects)
Wald 2
p
11.4
0.0007
20.8
<0.0001
20.1
0.0002
218.8
<0.0001
32.9
<0.0001

Clinical Studies/Outcomes

Tuesday, July 29, 9:30 am 5:00 pm

A-150
Different genotypes of a functional polymorphism of the TSHR gene are
associated with the development and severity of Graves and Hashimotos
diseases

M. WATANABE, N. Inoue, Y. Katsumata, Y. Hidaka, Y. Iwatani. Osaka

University Graduate School of Medicine, Suita, Japan
Background: The disease severities of autoimmune thyroid diseases (AITDs), such
as Hashimotos disease (HD) and Graves disease (GD), can vary among patients.
ST4 is one of the splicing variants of thyrotropin receptor (TSHR). Increased ST4
transcription may enhance the generating of a shed A subunit, which results in the
production of thyroid-stimulating antibody (TSAb). ST4 expression was higher in
the AA genotype of TSHR rs179247 polymorphism compared to the GG genotype.
Methods: We genotyped this polymorphism in 98 HD patients including 44 patients
developed hypothyroidism before the age of 50 years and were treated with thyroxine
(severe HD) and 33 HD patients over the age of 50 years were left untreated and
demonstrated euthyroid (mild HD), in 112 GD patients including 50 GD patients who
had been treated with methimazole and were still positive for TRAb (intractable GD);
and 33 GD patients in remission (GD in remission), and in 56 healthy volunteers.
Results: The AA genotype was more frequent in GD patients compared to control
subjects (p=0.0201). In contrast, the frequency of the GG genotype was higher in HD
patients compared to control subjects (p=0.0186). The A allele was more frequent in
GD patients compared to the HD patients (p=0.0010). The AA genotype and A allele
were more frequent in intractable GD patients compared to GD patients in remission
(p=0.0024 and 0.0005, respectively). The GG genotype was more frequent in severe
HD patients compared to control subjects (p=0.015). The proportion of patients who
developed GD under 50 years of age was significantly higher in GD patients with the
AA genotype compared to those with the AG/GG genotypes (P=0.0251).

Almost all patients (93%) had higher levels of urinary NGAL than the higher value
of the controls; the respective frequency for the other markers was: 68% for serum
NGAL and serum Cys-C, 50% for urinary Cys-C and only 10% for urinary KIM-1.
All studied markers correlated with eGFR: serum Cys-C (r=-0.758, p<0.001), serum
NGAL (r=-0.627, p<0.001), urinary Cys-C (r=-0.498, p=0.008), urinary NGAL
(r=-0.430, p=0.01) and urinary KIM-1 (r=-0.369, p=0.021). Only serum Cys-C
strongly correlated with the involved serum free light chain (r=0.806, p<0.001).
Urinary NGAL correlated also with urinary Cys-C (r=0.880, p<0.001), serum NGAL
(r=0.503, p=0.002), 24-h proteinuria (r=0.431, p=0.01) and ISS stage (meanSD
values for ISS-1, ISS-2 and ISS-3 were: 3129ng/mL, 4752ng/mL and 408695ng/
mL, respectively; p=0.03). Serum Cys-C correlated also with ISS stage (the values for
ISS-1, ISS-2 and ISS-3 were: 0.850.19mg/L, 0.940.24mg/L and 2.150.98mg/L,
respectively; p=0.01), while urinary Cys-C correlated with 24-h proteinuria (r=0.564,
p<0.001).
Conclusions: Our data suggest that almost all newly diagnosed symptomatic MM
patients have tubular damage as assessed by elevated urinary NGAL suggesting that
renal impairment is present very early in the disease course. Measurement of urinary
NGAL and serum Cys-C offers valuable information for the kidney function of MM
patients and their measurement may help in the identification of patients with high risk
for the development of acute renal function. The value of KIM-1 seems to be very low
in myeloma reflecting the differences in the pathogenesis of myeloma-related renal
dysfunction than toxic acute renal injury of other etiology.

Conclusion: The AA and GG genotypes of the rs179247 polymorphism of the TSHR

gene were susceptible to GD and HD, respectively and were also associated with the
intractability of GD and the severity of HD, respectively.
Summary of results
Genotype ST4 expression Associated with
Development of Graves disease
AA
HIGH
Intractability of Graves disease
Development of Hashimotos disease
GG
LOW
Severity of Hashimotos disease

A-151
Tubular Damage is Ubiquitous in Newly-Diagnosed Patients with Multiple
Myeloma: Comparison of Three Urinary and Two Serum Markers of Kidney
Injury

I. Papassotiriou1, D. Christoulas2, E. Kastritis2, M. Gkotzamanidou2, A.

Margeli1, N. Kanellias2, G. P. Papassotiriou2, E. Eleutherakis-Papaiakovou2,
M. A. Dimopoulos2, E. Terpos2. 1Department of Clinical Biochemistry,
Aghia Sophia Childrens Hospital, Athens, Greece, 2Department of Clinical
Therapeutics, University of Athens School of Medicine, Athens, Greece
Background: Neutrophil gelatinase-associated lipocalin (NGAL) is a protein
overproduced by proximal tubular cells in response to kidney injury, while
kidney injury molecule-1 (KIM-1) is a type 1 transmembrane glycoprotein that is
overexpressed in dedifferentiated proximal tubule epithelial cells after ischemic or
toxic injury. Urinary NGAL and KIM-1 have never been evaluated in MM patients.
Patients and Methods: To assess the value of these molecules in MM, we measured
urinary and serum NGAL, urinary KIM-1, urinary and serum cystatin-C (Cys-C) in
48 newly diagnosed symptomatic MM patients (27M/21F, median age 65 years). The
estimated GFR (eGFR) was calculated using the CKD-EPI equation. (proposed by
the CKD Epidemiology Collaboration and is widely accepted in renal impairment).
Serum and urinary NGAL was evaluated using immunoturbidimetric assay (BioPorto
Diagnostics A/S, Denmark) with a protocol applied in the Siemens Advia 1800
Clinical Chemistry System. Serum and urinary Cys-C was measured on the BN
ProSpec analyzer (Siemens Healthcare Diagnostics, Liederbach, Germany), while
urinary KIM-1 was also measured using an ELISA (R&D Systems, Minneapolis, MN,
USA). For the urinary measurements, a 24h urine collection was used.
Results: The median values (range) for the studied markers in MM patients and in
120 healthy controls were: for urinary NGAL 36ng/ml (0.5-2512ng/ml) vs. 5.3ng/ml
(0.7-9.8ng/ml), p<0.001; for serum NGAL 162ng/ml (53-576ng/ml) vs. 63ng/ml (37106ng/ml), p<0.001; for urinary KIM-1 1.1ng/ml (0.13-4.87ng/ml) vs. 1.3ng/ml (0.15.3ng/ml), p=0.345; for urinary Cys-C 0.05mg/l (ND-13.9mg/l) vs. non-detectable,
p<0.01; and for serum Cys-C 1.0mg/l (0.4-3.2mg/l) vs. 0.7mg/l (0.3-0.9mg/l), p<0.01.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S41

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

Tuesday, July 29, 2014

Poster Session: 9:30 AM - 5:00 PM
Endocrinology/Hormones

A-152
Assessing macroprolactin interference in prolactin assays after polyethylene
glycol precipitation in two automatized platforms

R. R. Rodrigues, R. C. Basso, R. C. A. Sanso, P. Osorio, C. F. Pereira.

Diagnostico da America (DASA), Barueri, Brazil
Background: Prolactin serum levels over 30 ng/ml in the absence of pregnancy
and postpartum breastfeeding are indicators of hyperprolactinaemia which could
impact in more complexes and expensive diagnostic protocols. Seric Prolactin
are classified in three main forms: monomeric which is the predominant form;
dimer, also known as big prolactin; and the high molecular weight form which is
usually known as macroprolactin or big-big prolactin (bbPRL). It is known that
macroprolactinaemia may correspond to approximately 20 - 25% of cases of
hyperprolactinemia and it is a common disorder in a healthy population. Thus, the
investigation of macroprolactinaemia as the main cause of hyperprolactinaemia would
avoid clinical investigation of prolactinoma and other diseases. The reference test
for detecting macroprolactin is gel filtration chromatography, but the test based on
polyethylene glycol (PEG) is simpler and cheaper and has been validated in 1999 by
Olokoga and Kane. All commercially available prolactin immunoassays have a crossreactivity level with macroprolactin. The purpose of this study was to validate the
PEG precipitation test using Siemens Healthcare Diagnostics ADVIA Centaur System
when compared to Abbott Architect.
Methods: 46 patient samples presenting levels over 30 ng/mL previously dosed were
re-tested in ADVIA Centaur and Abbott Architect after PEG treatment. Analysis of
mean, standard deviation and correlation were calculated. Moreover, a cutoff of 60%
was established to determine the presence of bbPRL.
Results: The samples were measured before extraction and the results were: 69.29
ng/mL average in ADVIA Centaur and 94.42 ng/mL in Architect with a standard
deviation of 30.12 and 42.10 respectively and R2 of 0.88. 60% was regarded as cutoff
for bbPRL screening. When comparing the ADVIA Centaur with Architect, a relative
sensitivity of 100% and a Relative Specificity of 83.33% with positive predictive
value of 97.56% and negative predictive value of 100% were obtained.
Conclusion: The Prolactin assay varies according to the elected methodology for
Macroprolactin detection. Correlation results between compared instruments were
satisfactory. Moreover, ADVIA Centaur is capable of dosing the three main forms
of seric prolactin. Among 46 samples only one sample showed doubtful result for
Architect and positive for ADVIA Centaur. As we can observe macroprolactin is a
major interference source and may lead to diagnostic errors and processing errors
involving patients with hyperprolactinemia. Samples present greater dispersion
measurements prior to PEG precipitation for Architect results and the difference
between the values of pure samples can be related to low interference that ADVIA
Centaur system presents, being considered as a positive point once risk in releasing
high results decreases. This provides a smaller number of high results which can
generate inadequate diagnostic or request a further test to make the diagnosis. In
conclusion, we can confirm that the evaluation of bbPRL methods, such as PEG
precipitation is still necessary, even in trials that have low reactivity for macroprolactin
as the ADVIA Centaur.

A-153
Vitamin D status in healthy and rheumatoid arthritis groups.

S. Kang, M. Lee, J. Yang, M. Kim, W. Lee. KyungHee University Hospital

at Gangdong, Seoul, Korea, Republic of
Introduction: Vitamin D is important for maintenance of calcium homeostasis and
bone metabolism. Its association with chronic and inflammatory diseases including
rheumatoid arthritis (RA) has been suggested while consensus on the optimal level of
vitamin D is yet to be made. The aims of this study are 1) to assess vitamin D status
among healthy and RA groups from Korean population; 2) to evaluate biochemical
markers of their relationship with vitamin D level; and 3) to determine an alternative
cutoff level and assess the relationship with RA.

S42

Materials and methods: The study includes 346 healthy individuals and 222
diagnosed RA patients during winter (January-February) and summer (July-August).
C-reactive protein (CRP), adjusted calcium (aCa), erythrocyte sedimentation rate
(ESR), parathyroid hormone (PTH), -isomerized C-terminal telopeptides (-CTx)
and serum 25(OH)D levels were measured. Each groups were classified by vitamin
D status and related markers into subgroups for comparison. Statistical analyses
were made using t-test, ANOVA, chi-squared test and logistic regression analysis.
Segmented linear regression analysis was used to determine an alternative cutoff.
Statistical significance was determined at P<0.05.
Results: 25(OH)D level was lower in female, younger age group, high PTH subgroup
and during winter. Vitamin D insufficiency/deficiency was highly prevalent among
healthy group (95.7%) and RA group (98.2%) but the distribution of vitamin D status
was not different. 25(OH)D andshowed a negative correlation with PTH (P<0.01,
r=-0.29), however the increase of PTH level was mostly within reference range and
increase of PTH beyond reference range level in vitamin D deficient subgroup was
rare (2.3%). Despite vitamin D deficiency, -CTx and aCacalcium levels were not
different among vitamin D status subgroups and no correlation with 25(OH)D or PTH
was found. Comparison between healthy and RA group also revealed similar seasonal
differences, RA group had lower 25(OH)D level during summer, although the
distribution of vitamin D status was not different. The alternative cutoff 22.08 ng/mL,
classified healthy and RA group as vitamin D insufficiency/deficiency in 85.54% and
90.54% respectively. The prevalence of vitamin D insufficiency/deficiecy was low
but the difference between healthy and RA group was not found. Logistic regression
analyses showed that 25(OH)D level was not associated with RA, and ESR and CRP
as markers of disease activity.
Discussion: Most healthy individuals are categorized as vitamin D insufficiency/
deficiency under the currently used conventional criteria for vitamin D status. The
level of vitamin D deficiency from this study did not result in increment of PTH,
-CTx or aCa greater than reference range. This finding is suggestive that the status
of vitamin D deficiency using the current criteria may not be clinically practical. In
contrast to previous reports on the relationship of vitamin D and chronic inflammatory
diseases, application of lower cutoff in this study also did not exhibit association with
RA and correlationor with markers of disease activity. Future investigation of vitamin
D should be conducted through a large, randomized controlled trial and focus on
deciding the optimal vitamin D level in correlation with clinically meaningful results,
regarding calcium homeostasis and bone metabolism.

A-154
Evaluation of Hb A1c bias and precision across eight platforms in the presence
of Hb AS and Hb AC

J. M. Rhea1, M. Richter-Roche1, A. Woodworth2, N. Korpi-Steiner3,

J. Miller4, D. Koch5, R. Molinaro1. 1Emory University, Atlanta, GA,
2
Vanderbilt University, Nashville, TN, 3University of North Carolina,
Chapel Hill, NC, 4University of Louisville, Louisville, KY, 5Grady Memorial
Hospital, Atlanta, GA
Introduction: Changes in serial Hb A1c results are due to changes in the clinical
condition or to inherent sources of biologic and analytic variation. Results of the
2012 CAP GH2-B proficiency survey which contained an Hb AS sample, suggest that
samples containing an Hb variant may impact assay precision for some methods, and
potentially reference change values (RCVs). Objective: To calculate the effect of Hb
variants on RCV for Hb A1c by determining the precision of each method. Methods:
Seven different NGSP-certified Hb A1c platforms were used to measure imprecision
and bias using patient samples containing Hb AA, Hb AS, or Hb AC. Precision was
determined following CLSI EP05-A2, and bias by calculating the percent difference
between each test method and the comparative method (NGSP Secondary Reference
Lab). RCVs were calculated using the standard formula. Results: Differences in
imprecision and bias were observed, and were typically greater in samples containing
an Hb variant. RCVs for all methods except two at were 0.5% Hb A1c; overall,
RCVs were slightly increased in the presence of Hb AS and Hb AC at 0.8%.
Conclusions: The total analytical error for the majority of assays was significantly
greater in samples containing Hb AS or Hb AC, and may indicate a need for including
proficiency samples containing the most common Hb variants, especially when assays
are used to measure Hb A1c in populations with a high prevalence of Hb variants.
In addition, while the clinical relevance of increased was beyond the scope of this
study, the change in RCV suggests a difference in how serial Hb A1c results may be
interpreted.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

Conclusion There are statistically significant and consistent difference between the
two methods compared. In extreme glycaemia (outside normal range) this difference
is even greater. Our data show that if the glucose levels at the bedside (strips) were
held on average 10 mg / dl of hypoglycemia target to be achieved in the plasma assay
by enzymatic method, we can increase exam security without compromising its
reliability and possibly reduce clinical patient discomfort.

A-159
Laboratory process improvement through adjustments in calibrations of
immunoassays

A. L. Camilo, D. Waltrick, C. F. A. Pereira. DASA, So Paulo, Brazil

Background: Workflow optimization identifying the specific processes and
improvements not only improves operational efficiency, it also reduces error and
contains costs. Because system capabilities are not the only considerations when
selecting the optimal system for the DASA laboratory, created customized workflow
scenarios to help maximize your productivity.
Objective: The objective of this study is to show that continuous monitoring and
standardization of quality system improved the flow in large laboratory routines,
increase productivity and yield of reagents reducing the turnaround time on the
pre-analytical (TAT), impacting on operational costs, reducing interfering in quality
control.
Methods: Based on an extensive statistical analysis on the performance of systems
ADVIA Centaur XP (Siemens Healthcare Diagnostics), were established changes in
calibration routine for Progesterone test, setting 6 equipments in which the calibration
for this analyte would be performed every seven days and for other six systems
calibration would be performed every twenty-eight days, number of replicates of the
calibrator was changed from 3 to 5 according to the instructions of reagents after a
period of four months. The performance analysis of the test and the variation in assay
constitute evaluated.

A-155
Is it possible to make insulin tolerance test (ITT) better?

M. P. Campagnoli1, A. Menna Barreto1, M. Freire2, M. Scharf1. 1DASA,

Curitiba, Brazil., CURITIBA, Brazil, 2DASA, Curitiba, Brazil., Rio de
Janeiro, Brazil
Background The diagnosis of growth hormone (GH) deficiency in children with
growth retardation is complex. Because of the pattern of pulsatile secretion of GH,
isolated determinations have no value. So, the functional stimulus tests for evaluation
of GH secretory reserve are fundamental for the diagnosis. The Insulin Tolerance Test
(ITT) is considered to be the gold standard test for diagnosis in children over 2 years
of age. This test consists on the administration of intravenous insulin and sampling a
multi amostral sequence of glucose and GH, and sometimes, cortisol.
The given insulin should be able to reduce 50% of fasting glucose or reaching below
the normal range, promoting the stimulus needed to boost the production and release
of GH and cortisol .
Bedside blood glucose control is done, during the lab test, through the measurement of
plasma glucose with blood glucose strips. Decisions like, correction of blood glucose
as well as administering another dose of insulin, are based on these values .
We observed that in many patients glucose measured at the bedside shows superior
values to subsequently assayed by enzymatic method (hexokinase).
This study aims to determine the difference between plasma glucose measured by
blood glucose strips (bedside) and enzymatic method and propose new targets for
bedside blood glucose to increase the security of ITT.
Methods From March 2013 to January 2014 we conducted 423 ITT. In all these,
measurements of plasma glucose were performed using blood glucose strips (bedside)
and enzymatic method. The blood sampling was done at the same time for both
methods. The results were matched and compared statistically.
Results: There is statistically significant difference between the plasma glucose levels
measured by blood glucose strips and enzymatic method . The value assayed by the
enzymatic method is smaller than the other (bedside). The median difference between
all values is -9.624mg/dl (-10.11 to -9.137)
Regarding extreme glycaemia (under 60 mg / dL and greater than 99mg/dl) , the
difference between the two methodologies is constant , statistically significant and
even higher. The enzymatic method is -10.87 mg / dL (-11.0 to 10.64) lower than the
equivalent sample at the bedside method (strips). At standard deviation (SD) we found
a deviation of + / - 21.75mg/ dL.

Results: The number of disparate control represented 15.4% with a confidence

interval from 14.6 to 16.2%. In 8,862 processed control results, 1,363 were outside
2SD range. After implementation and improvement processes in the system, in
between January and April 2012, the disparate controls number decreased to 7.8%
with confidence interval of 7.2% to 8.4%. In 7,931 only 615 control results were
outside 2SD range. The number of test calibrations decreased from 600 to 420 in the
respective time periods and processed control number decrease 10%.
Conclusion: The preanalytical phase may cause the inaccuracy of the results when
there is a long time for the release of the operating systems in laboratories of large
routines. The results suggest that variations in the time interval between calibrations
can impact the quality of the result, interfere with the performance of quality control
and increase the release time of the equipment for routine, compromising the
productivity of clinical laboratory routine.

A-160
Performance evaluation of novel C-peptide immunoassay reagent using a fullyautomated immunoassay analyzer

Y. Watanabe, I. Sato, N. Hayashi, J. Saegusa, S. Kawano. Department of

Clinical Laboratory, Kobe University Hospital, Kobe, Japan
Background: C-peptide is co-secreted with insulin in equimolar amounts from
pancreatic cells. Assessment of endogenous insulin production with C-peptide
immunoassay requires sufficient sensitivity and high specificity. In this study, we
evaluated the analytical performance of newly developed C-peptide reagent.
Methods: The ST AIA-PACK C-Peptide II reagent* on Tosoh AIA-2000 fullyautomated immunoassay analyzer is an enzyme immunoassay which is performed
entirely in a single cup. C-peptide in the sample is bound with monoclonal
antibody immobilized on magnetic beads and alkaline phosphatase-labelled
monoclonal antibody. After 10 minutes incubation at 37 , the beads are washed
to remove unbound materials and are then incubated with a fluorogenic substrate,
4-methylumbelliferyl phosphate. The amount of enzyme-labelled monoclonal
antibody that binds to the beads is directly proportional to the C-peptide concentration
in the sample. A standard curve is constructed using the Calibrator Set and unknown
concentration of C-peptide is automatically calculated using this curve. In this study
we evaluated the precision, functional sensitivity, interference, recovery and crossreactivity of this new reagent toward human proinsulin. Method comparison, against
ARCHITECT C-peptide immunoassay based on chemiluminescent immunoassay,
was evaluated with clinical specimens from patients. Correlation of the C-peptide

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S43

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

concentrations between serum and EDTA plasma samples was also studied.
Results: The standard curve extended from 0.02-30ng/mL for serum. Within-run and
between-run coefficients of variation ranged from 2.5% to 3.5% and from 2.5% to
3.6%, respectively. Based on the imprecision profile, functional sensitivity (at 10%
CV) of this reagent was 0.017ng/mL. Assay correlation with ARCHITECT C-peptide
immunoassay was determined: y=0.97x+0.02, r=0.997, (n=50 x; ARCHITECT,
y; AIA-2000). There was a good correlation between serum and EDTA plasma
concentrations of C-peptide: y=0.96x-0.07, r=0.996, (n=220 x; EDTA plasma, y;
serum). The cross-reactivity against purified human proinsulin was below 0.4%.
Conclusion: ST AIA-PACK C-Peptide II reagent using a fully-automated analyzer,
a novel enzyme immunoassay for detecting C-peptide, exhibited extremely
low cross-reactivity with proinsulin, and showed high sensitivity for C-peptide.
Our results demonstrated that this reagent is a reliable method for the rapid
and accurate quantification of C-peptide in clinical laboratories, and turns out
a useful tool for both the screening and the management of diabetic patients.
*) This product has not been approved by the FDA yet.

correctly classify the type of DM based upon the clinical phenotype has, however,
recently been challenged. Since the appropriate classification has important
implications with regard to treatment options, expected outcomes and genetic
counseling, a systematic, cost-effective algorithm to assist in the initial classification
of DM is needed.
Objective: To evaluate the use of an auto-antibody algorithm to classify new onset
diabetes patients and its use in curtailing testing costs
Methodology: Data from children (<18 yrs of age) hospitalized at CCMC from
Jan 2010 - Oct 2012 with new-onset DM was analyzed. In contrast to T2D, T1D is
an autoimmune disease (AD) characterized by the presence of >1 diabetes-related
antibodies (DR-Ab). Historically the initial evaluation, including DR-Ab testing, has
been left to the discretion of individual Pediatric Endocrinologists. Other Abs are
often measured to assess concurrent autoimmune diseases that commonly occur in
individuals with DM, such as Hashimotos thyroiditis and celiac disease. Inclusion
Criteria: 1) Age < 18 yrs at diagnosis; 2) New-onset DM; 3) Onset Jan 2010-Oct 2012.

A. Fortunato1, M. Caputo2, P. Garofalo3, C. Marchetti1, R. Castello4.

1
Laboratorio di Chimica Clinica ed Ematologia Ospedale S. Bortolo,
Vicenza, Italy, 2Laboratorio Chimica Clinica e Microbiologia, Ospedale
Orlandi, Bussolengo (VR), Italy, 3UOC Endocrinologia AOOR Villa
Sofia-Cervello, Palermo, Italy, 4UOC Medicina generale indirizzo
endocrinologico, AOUI, Verona, Italy

Results & Conclusions: The American Diabetes Association classifies DM into T1D,
T2D, gestational diabetes, and diabetes due to other causes. While the majority of
those <18 yr of age have T1D, the number with T2D is increasing. Individuals with
T2D are often obese. With the exponential increase in the number of children who
have become overweight/obese, classifying DM based on a childs phenotype has
become problematic. In children with overt signs/symptoms of DM the presence of >
1 DR Ab is generally considered sufficient evidence of auto-immunity (i.e. T1D). In
our study, subjects were routinely tested for 2 DR-Abs (GAD65 97.9%; ICA 95.9%).
Since 73.3% of subjects were positive for GAD65, additional testing for ICA increased
cost w/out additional benefit. While not included in the present study, additional
screening tests are sometimes also requested for celiac and thyroid disease. For those
whose initial screening was positive (celiac 12.5%; thyroid 16.1%), eliminating
further testing would have helped reduce cost.

Background: total testosterone level measurement is the most requested one among
steroid hormones assays. Unfortunately, the diagnostic accuracy at low concentrations
of the most common immunoassays proved to be insufficient. In 2007 the Endocrine
Society recommended the determinations of testosterone in children and in women
has to be done only with one reference method (extraction, chromatography and
determination by mass spectrometry). Due the method related difficulties in most of
the laboratories the testosterone determinations are still done by immunoassays.

Ab testing to help classify children with new-onset DM may be enhanced with use of
an algorithm, especially if it includes reflex testing. Reflex tests are tests automatically
performed by the laboratory if the initial test requested fails to meet preset criteria.
Subsequent tests can generally be performed w/out need for additional samples and
may consist of > 1 sequential tests. Although there is a charge for additional tests, if
the likelihood of the criteria being met with the initial sample is high, reflex testing has
the potential to reduce medical cost.

Samples and methods: we measured testosterone with three different fully automated
immunoassays present in most of the clinical labs and repeated the determinations
both with a commercial RIA and LC-MS/MS method. The latter one, considered
the reference method, has been done in the Perkin Elmer labs (Turku, Finland), with
updated equipment, by skilled personnel and determinations carried out in replicated.
The serum samples were collected from 70 patients, male and female in pediatric age.
The obtained concentrations by LC-MS/MS, considered as reference, ranged from
11 to 110 ng/dL.

Quality control management improving immunoassays systems in the clinical

laboratorial routine

A-161
Low testosterone concentrations: only mass spectrometry?

Results: the distribution of the concentrations obtained with the methods used should
be noted that, although the averages and medians of the concentrations obtained with
the LC-MS/MS method are less, the differences are not such as to distort the clinical
information can be obtained: the 3 automated methods show values ranging from 10
to 134 ng/dL with correlations coefficients respectively to LC-MS/MS ranging from
0,829 to 0,934; whereas the RIA method shows a higher concentrations dispersal,
values ranging from 20 to 149 ng/dL and a worse correlation to the reference method.
(r=0,705).
Conclusions: the position of the scientific community on total blood testosterone
measurement at low concentrations is critical to the use of direct immunoassays
because without the necessary diagnostic accuracy, and recommends the use of
methods that are not within the reach of general laboratories. Our results, although
preliminary, open an interesting perspective on the possibility of arriving at a
reasonable future to employ even the immunoassays, certainly more feasible, as an
aid to diagnosis of common and important endocrine syndromes of the woman and
the child.

A-162
Classification of Children with New-Onset Diabetes Mellitus Using an AutoAntibody Algorithm

S. Prakash1, V. Leung-Pineda2, S. Suzuki1, J. Radack2, J. Dallas2, P. S.

Thornton2, D. P. Wilson2. 1University of North Texas Health Science Center,
Fort Worth, TX, 2Cook Childrens Medical Center, Fort Worth, TX
Background: Historically the diagnosis of Type 2 diabetes mellitus (T2D) in children
has relied on a typical clinical phenotype. The ability of experienced clinicians to

S44

A-163

D. Waltrick, A. L. N. Camilo, C. F. A. Pereira. DASA, So Paulo, Brazil

Background: Automated analyzers provide several advantages on processing
immunoassay methods. The literature demonstrates that laboratory errors can be
associated with pre-analytical (30.6%), analytical (31.6%) and post-analytical
(30.8%) processes or even due to combined processes (6%). Errors in the analytical
phase are commonly related to lack of preventive maintenance, inappropriate Quality
Control (QC) management, and improper handling of samples or reagents. Errors
due to analytical problems have been significantly reduced over time, but there is
evidence that this interference may have a serious impact on patient results, especially
for immunoassays.
Objective: The aim of the study was to identify the potential causes for quality control
variability in immunoassays to improve the laboratory routine productivity when
adopting best quality control practices.
Methods: 32,760 quality control points of Immunoassay Plus QC Lot 40240 (BioRad)
were collected during eight months using 14 ADVIA Centaur Systems (Siemens
Helathcare Diagnostics), in the clinical laboratories. The statistics evaluated were:
Coefficient of variation (CV), Standard deviation (SD), observed mean and outliers.
Data was compared to those reported on Biorad International Quality Control Program.
All assays that presented better or equal statistic results were considered acceptable.
Assays with higher CV than the reference, Biorad worldwide report, were submitted
to further technical investigation and corrective actions. After critical analysis of the
first four months data, some improvements were implemented, such as: a new plan
of the preventive maintenance, increasing periodicity from once per quarter to once
each three months matching laboratory number of tests; a new definition of mean and
standard deviation for each QC target level for each assay. Once improvements were
implemented, quality control data was collected during the following four months.
Results: In the first four months period, before implementation of proposed
improvements, the CV mean was 18.8%, 5,402 tests were spent in calibrations,
16,841 tests were used in QC material analysis and reagents profitability was 96.4%.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

Thereafter, data of a new four months period were collected, showing CV mean
decreased to 10.7%; 4,708 tests were spent in calibrations; 15,919 tests were used in
QC material analysis and reagents profitability raised to 98.0%.

statistical analyses were performed using the software SPSS 15.0 (SPSS inc.,
Chicago, Il, USA) program. For all statistical tests, two-tailed p value <0.05 indicated
the statistical significance of the results.

Conclusion: Through statistic analysis, it was possible to identify the importance

on studying clinical lab routine in order to implement a best plan for preventive
maintenance and appropriate rules of the quality control management. The new
definition of a single target value for all ADVIA Centaur Systems was useful
for reducing time to evaluate daily internal quality control and also to ensure the
commutativity between the systems. This project brought financial and nonfinancial
gains, such as lower consumption of reagents, reduced downtime and reported results
confidence.

Results: There were no differences in age, gestational week and BMIs between the
patients with Hyperemesis gravidarum and control subjects. Serum ghrelin and
prealbumin levels were significantly lower in patients with Hyperemesis gravidarum
than the control group (p<0.05). Serum obestatin and nesfatin1 levels were not
statistically different between the two groups.

A-164
A Comparison of CVD risk in newly diagnosed hypothyroid and type 2 diabetes
mellitus subjects using Framingham risk score sheet

P. Purohit, P. Sharma. All India Institute of Medical Sciences, Jodhpur,

India
Background Thyroid disorders and type 2 diabetes are known for their association
with CVDs owing to their effect on derangement of the lipid metabolism. However
there are no studies to document a comparative CVD risk in these two disorders.
Aim We aimed to compare the various CVD risk parameters in thyroid disorder and
type 2 DM subjects at the time of diagnosis.
Material and methods The study participants were 150 hypothyroid and 180 type 2
DM subjects reporting for the first time to our endocrinal clinics. The patients were
selected on the basis of symptomatology, a TSH >5IU/ml and a FBS > 126 mg/dl.
All participants were evaluated for BMI, Blood Pressure, serum Insulin, HOMA - IR,
Lipid profile, apo -B and A1. CVD risk was assessed using the Framingham risk score.
Statistical analysis was done using the students -t test to assess significance.
Results At diagnosis the hypothyroid and T2DM subjects presented with raised BMI
(p>0.001), hypertension (SBP 132.98 17.40 v/s 132.60 12.18 [NS]; DBP 86.52
9.82 v/s 88.79 8.02 [NS]), insulin resistance (30.6316.18 v/s 17.29 15.61,
p<0.0001 ), gross dyslipidemia, with the T2 diabetic subjects showing significantly
raised total cholesterol (231.15 22.19 v/s 213.50 38.95, p<0.0001), triglycerides
(197.35 35.31 v/s 187.91 39.12, p<0.0001), reduced HDLc (33.56 2.67 v/s 42.99
4.70, p<0.0001) and significantly reduced apo B (154.47 12.87 v/s 175.58 34.56,
p<0.0001 ) and apo A1 (96.94 8.55 v/s 139.76 17.40, p<0.0001). The CVD risk
ratios T.chol/HDLc 6.93 v/s 5.03 and apoB/apoA1 were 1.60 v/s 1.29. The ten year
risk of CVD in the T2DM subjects was 25% and in the hypothyroid subjects was 13%.
Conclusion: The present study concludes a significantly raised CVD risk in T2DM as
compared to hypothyroid subjects at diagnosis.

A-165
Serum ghrelin, obestatin and nesfatin1 levels in pregnant women with
hyperemesis gravidarum

G. Ozturk1, S. Erdinc2, F. Ucar1, Z. Ginis1, G. Erden1, N. Danisman2.

1
Department of Clinical Biochemistry, Diskapi Yildirim Beyazit Training
and Research Hospital, Ankara, Turkey, 2Zekai Tahir Burak Womens
Health Education and Research Hospital, Ankara, Turkey
Backgrounds: Hyperemesis gravidarum, which affects 0.3-2.3% of pregnancies, is
defined as excessive vomiting during pregnancy and usually starting at 4 to 5 weeks
gestation , which may lead to severe outcomes including weight loss, dehydration,
ketonemia, ketonuria, fasting acidosis, alkalosis due to hydrochloric acid loss and
hypokalemia. Although, their exact cause is unknown, various metabolic and
neuromuscular factors have been implicated in the pathogenesis of Hyperemesis
gravidarum. The aim of this study was to investigate the levels of prealbumin, total
ghrelin , nesfatin 1, obestatin in Hyperemesis gravidarum.
Methods: A total of 40 pregnant women with Hyperemesis gravidarum and 38
pregnant women who were perfectly healthy and whose pregnancy had a normal
course were included in this study. After an 8-12 hour overnight fast, blood samples
were collected into plain tubes for obtaining serum. Blood samples were centrifuged
at 2.500 g for 15 min at 4 oC within 30 min of collection, and serum samples were
stored at -80oC until analysis. Measurements of ghrelin (Phoneix, USA) obestatin
(Biovendor,Czech Republic) and nesfatin1 (Phoneix, USA) were performed in an
EPOCH system (BioTek Instruments, Inc, USA) using the commercially available
enzyme-linked immunosorbent assay kit in accordance with the manufactures
instructions. Prealbumin levels were measured by spectrophometric method. All

Conclusions: Hyperemesis gravidarum is a disease of severe nausea, vomiting, and

anorexia in early pregnancy resulting in dehydration and weight loss. Prealbumin is
an indicator to assess nutritional status, so our data also suggests that prealbumin
levels are decreased in patients with Hyperemesis gravidarum. Ghrelin is involved
in stimulation of appetite, control of energy balance, and gastric motility. Ghrelin
administration increases food intake through central mechanisms. One possible
explanation might be that the decreased levels of ghrelin in Hyperemesis gravidarum
may be a mechanism to lose of appetite and the energy balance of the Hyperemesis
gravidarum pregnant women.

A-166
Evaluation of the Impact on IGF-I Control of Pharmacological Treatment with
Octreotide LAR isolated compared to Association with Cabergoline in Patients
with Acromegaly

L. F. A. Nery, C. M. Arajo, R. H. Jcomo, S. S. S. Costa, L. A. Naves.

Laboratrio Sabin de Anlises Clnicas, Braslia, Brazil
Background: Discrepancies concerning GH and IGF-1 levels in acromegaly patients
can occur in patients submitted to pharmacological treatment. Objectives: The aim of
this study was to compare the efficacy in decrease and normalization of IGF-I values
of Octreotide LAR treatment isolated and associated with cabergoline, based on IGF1 determinations.
Methods: This is a case series study that enrolled 34 patients with confirmed diagnosis
of acromegaly recruited from outpatient clinics of the Neuroendocrine Unit of the
University Hospital of Brasilia. All of them received the diagnosis of acromegaly
confirmed by clinical findings suggestive of the disease, elevated GH and agematched IGF-I levels, GH not suppressible by the oral glucose load and evidence of
pituitary adenoma on CT or MRI. The statistical analysis was performed using SPSS
17.0 software. Values are expressed as the mean standard deviation. The values of
IGF-I are presented both as absolute values as percentage values of the upper normal
limit normal range of IGF-I (% ULNV-IGF-I).
Results: The cohort was composed by 15 men and 19 women; mean age 54 (27-74)
years old, divided in two groups, group 1, treated by Octreotide LAR (OC-LAR)
30 mg/month, and group 2 treated by Octreotide LAR (OC-LAR) 30 mg/month
associated with cabergoline 2,0 mg/week (OC-LAR + CBG). Mean serum IGF-I and
% ULNV-IGF-I pretreatment were significantly higher in the group OC-LAR + CBG.
Those variables decreased 6 and 12 months after treatment in both groups, and the
values of OC-LAR+CBG group became inferior to OC-LAR group. However, no
significant difference was found between the OC-LAR and OC-LAR + CBG group
neither 6 nor 12 months after treatment.
Conclusion: OC-LAR + CBG association resulted in a significantly higher decrease
of IFG-I, both 6 and 12 months after treatment, compared to those treated with OCLAR

A-167
ADVIA Centaur Vitamin D Total Assay*: Expected Vitamin D Values in a
Healthy Pediatric Population

H. Mindicino1, J. Freeman1, R. Christenson2, B. Plouffe1, A. Woods3, M.

Zgela1. 1Siemens Healthcare Diagnostics, Tarrytown, NY, 2University
of Maryland School of Medicine, Baltimore, MD, 3Siemens Healthcare
Diagnostics, Berkeley, CA
Background: Vitamin D is a hormone involved in the intestinal absorption of calcium
and the regulation of calcium homeostasis. It is a key regulator of bone metabolism.
Vitamin D3 is derived from skin exposure to sunlight while food supplements contain
either Vitamins D2 or D3. Vitamin D deficiency is caused by a lack of exposure to
sunlight and by insufficient dietary intake. Measurements of serum or plasma levels of
the metabolite 25-hydroxyvitamin D (25[OH]D) are the best indicators of nutritional
Vitamin D status.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S45

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

Objective: The goal of the study was to test healthy pediatric donor specimens in
order to establish pediatric observed values for the ADVIA Centaur Vitamin D Total
assay*.
Methods: Serum samples were obtained from donors with ages ranging from 1 year
up to 21 years. Donors resided in 8 regions, geographically distributed across the
continental U.S. Donations occurred throughout one calendar year. All donors were
free of chronic or active diseases, and were not receiving any prescription medications
within 7 days of donation. All samples were assayed for iPTH and TSH on the Siemens
IMMULITE 2000 Immunoassay System. Samples with abnormal iPTH or TSH
levels were excluded. The remaining samples were assayed for total vitamin D Total
on the Siemens ADVIA Centaur. The ADVIA Centaur Vitamin D Assay used was
aligned to the ID-LC/MS/MS 25(OH)vitamin D Reference Measurement Procedure
(RMP), the reference procedure for the Vitamin D Standardization Program (VDSP).
Results: After all inclusion criteria were met, 227 samples were assayed for Vitamin
D Total. Values were calculated for each age group, by season, and by geographic
location. The lower and upper reference limits were estimated as the 2.5th and the
97.5th percentiles of the distribution of test results for each age group. The n, mean,
median, 2.5th and the 97.5th percentiles for each population sub-group were:
Northern: n=119, mean=23.55, median=22.17, 25th% = 12.36, 75th = 38.92
Southern: n=108, mean=26.33, median=25.47, 25th%= 9.70, 75th% = 49.16
Summer: n=136, mean=26.64, median=24.86, 25th% = 12.46, 75th% = 46.58
Winter: n=91 , mean=22.23 , median=22.10 , 25th% =9.70 , 75th% = 32.38
1yr-3yr: n=22 , mean=23.34 , median=24.41 , 25th% = 13.98, 75th% =32.45
3yr-12yr: n=114 , mean=24.99 , median=23.10 , 25th% =12.46 , 75th% =45.96
12yr-21 yr: n=91 , mean=25.09 , median=23.95 , 25th% =8.16 , 75th% =45.83
All Donors: n=227 , mean=24.87 , median=23.37 , 25th% =11.36 , 75th% =45.83
Conclusion: Vitamin D levels were consistent across the 3 age groups with no
apparent changes with age. There was no statistical difference between those
receiving and not receiving vitamin supplements. Vitamin D levels were statistically
higher for the southern region versus northern region, and for summer (maximum
sunlight) versus winter (minimum sunlight). There was extensive overlap in ranges
for all sub-populations and the expected range for the entire population can applied to
any of the sub-populations.
* This version of the ADVIA Centaur Vitamin D Total assay is not available for sale
in the U.S.

Our laboratory validation protocol will help any laboratory personnel from any part
of the world to validate & establish reference interval based on their own population
demographic variation.

A-169
Glutamate Decarboxylase Antibody Positivity in Diabetics

L. Ong, M. Tan, S. Saw, S. Sethi. National University Hospital, Singapore,

Singapore
Background: Glutamate decarboxylase antibody(GAD) testing is useful in
identifying patients with latent autoimmune diabetes in adults (LADA), and some
studies showed that higher antibody titres were associated with specific phenotypes.
We looked at patients with positive GAD and reviewed their laboratory and clinical
features.
Methods: All GAD tests performed in 2013 were included, and positive results were
analyzed with respect to demographic parameters, indication for testing (evaluation of
diabetes mellitus or neurological signs or symptoms), presence of other autoantibodies
or autoimmune diseases. GAD was performed using radioimmunoassay using the
CentAK kit and a positive result was defined as 0.9U/L. Statistical analysis was
done using SPSS Version 17.0.
Results: There were 454 GAD requests in 2013, with 75(16.5%) positive cases. The
median age was 39.3 years old, with female: male ratio of 0.55.
In patients with positive GAD, the median age was 43.4 years old, with female: male
ratio of 1.06. There were 45 Chinese, 10 Indians, 9 Malays and 10 from other ethnic
groups. 94.6% were requests to exclude LADA, and 21.3% had concurrent requests
for anti-islet cell antibody, with 50% positivity. 4 patients had co-existing autoimmune
diseases (myasthenia gravis, thyroid disease with positive thyroid peroxidase
antibody, vitiligo and pernicious anemia).
GAD titres showed a Gaussian distribution with a left skew and peak at 0-25U/L.
There was no association between GAD titres and age or C peptide levels.
Conclusion: GAD levels were predominantly ordered to evaluate diabetes mellitus in
younger patients and there was an association with anti-islet cell antibody and other
autoimmune diseases. Further studies may be performed to determine the clinical
significance of high GAD titres in diabetics.

Product availability varies from country to country and is subject to local regulatory
requirements.

A-168
Enigma behind Thyroid Function Tests: Harmonization Efforts

B. Das. Kokilaben Dhirubhai Ambani Hospital & Medical Research

Institute, Mumbai, Maharashtra, India
Background: Thyroid function tests (TFTs) form a very important set of tests in a
pathology laboratory; a tool that clinicians and patients alike depend on to pin down
the symptoms for treatment and relief However, it is this very set of tests that have
come in question. First, what is the normal and acceptable range (upper & lower limits)
of TSH in a particular population has been debated in different scientific fora A second
problem which the doctors and technicians have to grapple with is the variability of
test-results in itself; even a broadly similar set of instruments and methods can give
up to 40% more or less values in TSH levels. In our laboratory, we validated thyroid
function test and established reference range of TSH for Indian population.
Methods: In our laboratory, validation of thyroid function test were done [with
particular reference to Thyroid Stimulating Hormone (TSH)] by verifying analytical
accuracy and precision, and Analytical measurement range (AMR) as well as sigma
metrics. We have also verified the reference range for Indian Population. We have
screened 800 subjects. 630 healthy subjects were chosen in the study group for
reference interval verification.
Results: In our laboratory, we have seen, high degree of analytical accuracy between
two instruments (r2 = 0.985). Within Run (Repeatability) Precision and Within
Laboratory Precision were comparable with the manufacturers claim. Our obtained
reference range (0.62 - 4.22 micro IU/ml) was within that of the manufacturers (0.35
- 4.94 micro IU/ml). AMR was also verified with C.V. 1.70%, 1.89% and 2.51%, for
control sera. Our obtained reference range (0.33 - 4.90 micro IU/ml) was correlating
with that of the manufacturers (0.35 - 4.94 micro IU/ml).
Conclusion: In our laboratory, we have verified thyroid function tests in our hospital
set up. However, standardization of TSH and other thyroid function test is still a
formidable challenge, due to the lack of proper reference intervals and standardized
measurement procedures.

S46

A-170
New therapeutic effect of metformin due to increased levels of FGF21 ?

D. Stejskal1, R. Ochmanova2, M. Uvirova3. 1Agel Research and Training

Institute, branch Novy Jicin, Laboratories Agel, Novy Jicin, Czech
Republic, 2Agel Research and Training Institute, branch Prostejov, Central
Moravian Hospital Trust, Department of laboratory medicine, Prostejov,
Czech Republic, 3Agel Research and Training Institute, branch Ostrava,
CGB Laboratory, Ostrava, Czech Republic
Background:: Fibroblast growth factor 21 (FGF21) is an endocrine hormone that
exhibits anti-obesity and anti-diabetes effects. Recently was presented that metformin
in patients with type 2 diabetes modulates FGF21 expression and blood concentration.
Results indicate that metformin induced expression of FGF21 through an ATF4dependent mechanism by inhibiting mitochondrial respiration independently of
AMPK and its concentration in blood. AIM: Studying the effect of metformin on the
concentration of FGF21 in serum in type 2 diabetes patients.
Methods:The study was approved by the Ethics Commission of the Hospital
ternberk. Study was monocentric, prospective and randomized. A total of 108
individuals were recruited for our study (18 healthy controls (HC), 18 T2D
individuals without anti-diabetes therapy (W), 18 T2D individuals with diabetes
monotherapy with derivate of sullfonylurea (SU),18 T2D individuals with diabetes
monotherapy 500 mg metformin/day (M5), 18 T2D individuals with diabetes therapy
1000 mg metformin/day (M10), 18 T2D individuals with diabetes therapy 1500 mg
metformin/day (M15). Anthropometric (height, weight, BMI, waist circumference
(WC)), clinical (systolic and diastolic pressures) and laboratory fasting analyses were
performed. Serum samples were separated in a cooled centrifuge and immediatelly
analyzed for total cholesterol, HDL-cholesterol, LDL-cholesterol, triglycerides,
glucose, high sensitivity CRP, creatinine, uric acid, AST, ALT (all Siemens, Advia
1800). FGF21 serum level was determined by a commercially available ELISA kit
(Biovendor, DS2, Dynex) in serum samples stored at -80o C.
Results:The study analysed 108 subjects, of which 18 were in good health while
100 probands suffered from T2D. In-defined subgroups, we found no significant
differences by gender in FGF21 concentration. Healthy individuals had the lowest

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

values of FGF21, in other subjects are increased by the value of the diagnosis, the
type of therapy and dose (HC 84.2 ng/l vs W 111.6 vs. SU 158.6 vs. M5 269.7 vs.
M10 404.1 vs. M15 558.7, P <0.01). Changes remained significant after adjustment
for age, sex and BMI. Serum glucose levels fluctuated in subgroups below 8 mmol/l.
Conclusion:: in a randomized prospective study, we for the first time confirmed
the recently presented hypothesis that metformin leads to the rise of FGF21. The
new finding was the fact that this happens regardless of gender, weight and age of
probands. FGF21 induction by metformin might explain a portion of the beneficial
metabolic effects of metformin

A-171
Pre-analytical assessment of AMH stability in human serum using a wellcharacterized midpro-mature immunoassay

A. Kumar1, B. Kalra1, A. S. Patel1, G. Savjani1, E. E. Eklund2, G. LambertMesserlian2. 1Ansh Labs, Webster, TX, 2Women and Infants Hospital,
Providence, RI
Background: The aim of the study was to assess the stability of AMH in human
serum using a well-characterized midpro-mature immunoassay.
Relevance: AMH is a homodimeric glycoprotein composed of two 55 kDa N-terminal
and two 12.5 kDa C-terminal homodimers, non-covalently linked by disulfide
bridges. Recently, there have been concerns related to AMH stability in serum/plasma
and complement interferences affecting the end result. This has generated numerous
debates and publications related to reproducibility of AMH measurements and impact
of pre-analytical sample handling. To date, no publication has clearly stated if the
AMH variability is related to process (pre-analytical) or the assay. The AMH in
female serum is mostly pro-mature associated form. The kinetics of association of pro
and mature is rapid. Assay design that includes stable epitope antibodies and is not
impacted by molecule association or complements will generate reproducible results.
Methods: A prospective study (n=16) was designed in which serum samples were
tested within 3 hrs of draw, aliquoted and stored at room temperature (RT), -20C,
2-8C and re-assayed at 7, 10, 24, 48 and 168 hours. Multiple samples were thawed
up to 4 cycles and measured at two independent sites. A well-characterized two-step,
ELISA (Ansh Labs, US AMH, AL-105) was used to measure AMH levels in 25 L of
sample in < 3 hours. The assay is specific for human and measures pro-mature AMH
complex. The assay is calibrated (0.09-19.4 ng/mL) against standardized recombinant
human AMH.
Results: No significant changes were observed when samples were stored at RT,
2-8C and -20C. The median AMH concentration (16 serum samples, range 0.34-20
ng/mL) measured at 7, 10, 24, 48 and 168 hrs were 5.2, 4.9, 5.1, 4.9, 5.0 ng/mL at RT,
5.0, 5.0, 4.7, 4.3, 4.4 ng/mL at -20C and 4.8, 4.8, 5.0, 4.7, 4.9 ng/mL at 2-8C. The
average CV on multiple runs at RT, 2-8C, -20C was 8.7%, 6.9%, 9.3%, respectively.
Total imprecision (all data points) on stored samples and two controls were 9.3%,
4.8% and 3.1%, respectively. Freeze thaw analysis on two independent sample sets
showed that AMH levels were stable over 4 thaw cycles, with median levels of 4.5,
4.7, 4.8, 4.7 ng/mL obtained at site 1 (n=4) and 1.2, 1.2, 1.3, 1.3 ng/mL, respectively
at site 2 (n=5).
Conclusion: AMH as a biomolecule is very stable. This finding helps to resolve the
uncertainty related to AMH sample stability and reliability of AMH measurements
that have been debated. This study demonstrates that well-characterized assays and
good pre-analytical methods will produce reliable and reproducible results.

A-172
Performance of Thyroid Hormone Assays on Mindray CL-2000i
Chemiluminescence Immunoassay System

Methods: Imprecision studies were performed according to the CLSI EP5-A2

guideline, and have been evaluated using two samples with low and high concentration.
The within run imprecision was performed by measuring each sample for 20 times.
The total imprecision was evaluated by measuring each sample continuously for 20
days with the same lot of reagent. The imprecision was expressed as coefficient of
variation (CV%). Patient samples from healthy subjects, hypo and hyperthyroidism
were freshly collected from the clinical laboratory of our hospital. The comparison
studies were performed using CL 2000i and the reference methods in our laboratory,
Siemens ADVIA Centaur and Beckman Coulter DxI 800.
Results: The CVs of TSH, FT4, FT3, TT4, and TT3 are in a range from 1.09 - 6.16%
for within run, and from 3.22 - 9.16% for total imprecision. TT4 shows relative higher
CVs for both within run and total imprecision comparing to other parameters, but
within the manufacturers claim ( 10%). The comparison studies indicated slopes
for the five thyroid hormone parameters ranged from 0.748 to 1.051 and the intercepts
from -0.48 to 0.6. All of the five assays displayed high interassay correlation (r2 >
0.92). TSH displayed the highest correlation between CL 2000i system and Centaur
XP system (slope = 0.993; r2 = 0.971), while TT4 showed the lowest agreement
between CL 2000i system and the Centaur XP (slope = 1.051; r2 = 0.921).
Conclusion: The imprecision was highly acceptable for all the five parameters of
thyroid hormones tested. The method comparison between CL-2000i system and
the reference methods evidenced high concordance. Therefore, the parameters of
CL-2000i system are well suited for the detection of thyroid hormones in clinical
laboratories.

A-173
Linearity study of fertility assays assuring the quality control requirements

D. Waltrick, A. L. Camilo, C. F. A. Pereira. DASA, So Paulo, Brazil

Background:In the fertility trials, each step in the reproduction process is evaluated
to monitor gynecological health. There are basic exams requested for each phase of
the process. To certify released results accuracy, it is important to ensure the linearity
of the tests.A Brazilian laboratory implemented an Easy Linearity Curve (ELC) tool
to monitor the immune-hormone analytical systems, establishing a self-inspection
program to verify the efficiency and accuracy of procedures and results. The use of a
tool that checks assay linearity provides additional safety and reliability of the results.
This study aims to use the statistic tool to monitor the AMR of prolactin, progesterone,
LH, FSH, testosterone and estradiol.
Materials and methods:We selected samples from the laboratory routine, with
concentrations within assay linearity range for each test. Samples were tested in Advia
Centaur XP (Siemens Healthcare Diagnostics) using a chemiluminescent method
and analyzed with the ELC tool.
Results: AMR studies were carried out according to CLSI Guideline EP6-A. Results
are demonstrated in the table below.
Discussion:The tests showed satisfactory linearity results with different samples. In
this study, the sample pool was not used, because all the results from this dilution test
presented high percentage of recovery, above the manufacturers reference, due to
a matrix effect. Because of that, we established the utilization of different samples,
respecting the expected concentrations of the sample pool, if diluted samples were
prepared. The coefficients of second and third degree regression are statistically equal
to zero, at 5% of the significance level.
Conclusion:The fertility trials tested in this study presented Assay Linearity Range
as established by the manufacturer, within CLSI Guideline EP6-A and Total Error
Laboratorys target. Thus tests were approved by the Quality Control Management
in the laboratory.
Table 1 Assay

Y. Qiu , J. Hao , C. Kang , P. Feng , C. Yang , S. Yuan , Y. Zhang , F. Xia .

1
Nanfang Hospital, Southern Medical Univeristy, Guangzhou, China,
2
Nanshan Hospital, Huazhong University of Science and Technology,
Shenzhen, China, 3Shenzhen Mindray Bio-Medical Electronics Co., Ltd.,
Shenzhen, China
1

Prolactin
Progesterone
LH
FSH
Estradiol
Testosterone

Assay Linearity
Range
0.3 - 200 ng/ml
0.21 - 60 ng/ml
0.07 - 200 mUI/ml
0.3 - 200 mUI/ml
11.8 - 3000 pg/ml
10 - 1500 ng/dl

Obtained
CV %
4.9
3.1
3.2
4.5
3.5
3.2

CV
Target
%
5.75
8.71
4.78
5.5
4.53
8.3

Obtained
TE %
12.65
15.47
10.48
8.9
12.6
16.54

TE max
Target %
21.1
25
19.8
17.1
21.6
25

Linear Regression

R2

1.013x + (-4.058)
1.038x + (-0.67)
0.988x + (-0.947)
1x + (-0.835)
1.029x + (15.194)
1.086x + (-30.634)

0.99505
0.99821
0.99699
0.99934
0.99797
0.99356

Background: Thyroid hormones are among the first-line tests subject to

international collaborative investigations, aiming at assessing the key performances
and comparability of results between the available tests. Mindray CL-2000i
Chemiluminescence Immunoassay System (CL-2000i) is a recently launched
automatic immunoassay system. We have evaluated the performance of the
system with serum thyroid-stimulating hormone (TSH), free thyroxine (FT4), free
triiodothyronine (FT3), total thyroxine (TT4), and total triiodothyronine (TT3).

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S47

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

A-174
Calcitriol and its free and bioavailable fractions are better markers than 25
hydroxy vitamin D for monitoring vitamin D status during Pregnancy

R. Pandian1, J. Pandian1, Z. Seres2, A. Elias1. 1Pan Laboratories, IRVINE,

CA, 2Immunodiagnostic system LTD, Bolden, United Kingdom
Objective To compare bioavailable 25-hydroxy vitamin D(25OH-D)concentrations
with those of bioavailable calcitriol
(1, 25 di Hydroxy vitamin D) in sera of pregnant women.
Relevance Pregnancy is associated with major changes in calcium homeostasis
with critical roles played by vitamin D and its metabolites.Current practice favors
monitoring of vitamin D status in pregnancy using 25OH-D concentrations in serum.
Calcitriol is the active form of vitamin D.The level of 25OH-D and calcitriol vary due
to changes in maternal serum vitamin D binding protein(DBP).Serum concentration
of bioavailable and free vitamin D are not influenced by DBP. Therefore, we measured
total, bioavailable and free fractions of 25OH-D or calcitriol in serum samples of
pregnant and non-pregnant women. Correlation between total 25OH-D and its
fractions (bioavailable and free) and calcitriol and its fractions with PTH and CTX(Cterminal collagen degradation product) was explored in order to determine which
compound or compounds were optimal markers of vitamin D status in pregnancy.
Methods Bioavailable 25OH-D or bioavailable calcitriol are fractions, not bound to
DBP.They are the combined fractions of albumin bound and free fractions of 25OH-D
or calcitriol.To obtain the bioavailable fraction, total vitamin D
(25OH-D or calcitriol)was quantitated using immunoassays(IDS, Phoenix,AZ).
DBP was quantitated by an ELISA using reagents from R&D systems. Albumin
was quantitated by a calorimetric method.Using the affinity constants of 25OH-D
and calcitriol for DBP, and the affinity constants of 25OH-D or calcitriol for serum
albumin, bioavailable 25OH-D,bioavailable calcitriol,free 25OH-D and free calcitriol
were calculated.PTH and CTX assays were performed in pregnancy serum samples
using IDS kits.Pregnant serum samples (n=54) were collected between 27 to 38 weeks
of pregnancy.
Results Total 25OH-D was significantly lower in pregnant women despite a
significant increase in DBP (276 15 Vs 410 30ug/ml).Bioavailable and free 25OHD levels were lower than normal(n=55),and were in the deficiency or insufficiency
range although the levels of PTH and CTX were in the normal range.The correlation
between PTH with total 25OH-D was poor(r2= 0.3).There was also poor correlation
between PTH and bioavailable or free 25OH-D (r2= <0.5). Calcitriol was high in
the pregnancy samples (127.5 15.5 pg/ml)compared to non-pregnant samples (36.2
5.6 pg/ml) and the DBP-corrected bioavailable and free calcitriol was twofold
higher than non-pregnant controls.Calcitriol and its fractions (bioavailable and free of
calcitriol) correlated well with serum PTH and CTX (r2 = >0.9).
Conclusion Current assessment of vitamin D status in pregnant women by measurement
of 25OH-D does not adequately reflect calcium homeostasis in pregnancy.25OH-D or
its fractions do not correlate with PTH or CTX. On the other hand, calcitriol correlates
well with PTH and CTX in pregnancy when determined either as total, bioavailable or
free calcitriol.The data indicate that bioavailable or free calcitriol are the best markers
for determining vitamin D status in pregnant women.

A-175
Analytical Measurement Range (AMR) Monitoring for Immune-hormone
System

A. L. Camilo, D. Waltrick, C. F. A. Pereira. DASA, So Paulo, Brazil

Background: Linearity tests verify if a method is able to provide linear results,
proportionally to the analytes concentration. It is extremely important to ensure the
accuracy of the results released. A Brazilian laboratory established a self-inspection
program to verify the efficiency and accuracy of procedures and results through
the implementation of the Easy Linearity Curve (ELC) statistic tool to monitor an
immune-hormone system. ELC checks assay linearity and provides additional safety
and reliability to their results. This study aims to use the statistic tool to monitor the
AMR of ten immune-hormone parameters.
Materials and methods: Samples were selected from the laboratory routine, electing
concentrations within assay linearity range for each test. Then, samples were tested
for Insulin-like Growth Factor 1 (IGF-1), Insulin-like Growth Factor Binding Protein
3 (IGFBP-3), Growth Hormone (GRH), Prolactin (PRL), Sex Hormone Binding
Globulin (SHBG), C-Peptide (Pep-C), Homocysteine (HCY), Adrenocorticotropic
Hormone (ACTH) and Intact Parathyroid Hormone (iPTH) on IMMULITE 2000
System (Siemens Diagnostics Healthcare) and results analyzed with the ELC tool.

S48

Results and Discussion: AMR studies were carried out according to CLSI Guideline
EP6-A. Results are shown in the following table bellow.Tested Samples showed
satisfactory linearity results for each parameter. The coefficients of second and third
degree regression are statistically equal to zero, at 5% of the significance level. All
results analyzed by ELC, presented Assay Linearity Range as established by the
manufacturer, within CLSI Guideline EP6-A and Total Error Laboratorys target.
Conclusion: We conclude that the implementation of a self-verification program as
ELC can increase efficiency and accuracy of procedures and results; aiding laboratory
on accomplishing Quality Control Program requirements and guaranteeing safety and
reliability of their released results.
Table 1 Assay
IGFBP3
IGF1
GRH
PRL
SHBG
Pep C
HCY
ACTH
IPTH

CV Lab
Assay Linearity Obtained
max Target
Range
CV %
%
0.1 - 16 ug/mL
1.40
6.3
20 - 1600 ug/dL 2.80
4.7
0.05 - 40 ng/mL 3.20
4.6
0.5 150 ng/mL
2.50
5.9
2 - 180 nmol/L
3.20
6.2
0.1 - 20 ng/mL
2.40
8.3
2 - 50 umol/L
7.70
6.4
5 - 1250 pg/mL 3.94
4.6
3 - 2500 pg/mL 6.73
13.0

Obtained TE max
Linear Regression
TE %
Target %

R2

14.6
8.8
12.6
14.7
13.2
11.2
15.3
9.1
20.4

0.99377
0.99995
0.99634
0.99509
0.99837
0.99956
0.99986
0.99844
0.98744

17.50
14.90
20.00
21.10
21.10
20.80
17.70
11.82
30.20

1.09x + (-0.236)
0.985x + (2.037)
0.982x + (-0.246)
0.93x + (3.362)
0.946x + (1.952)
1.03x + (-0.044)
1.053x + (-0.145)
0.986x + (3.936)
1.042x + (51.56)

A-176
Clinical Applications Of Adiponectin Measurements In Type 2 Diabetes
Mellitus - Screening, Diagnosis and Marker of Diabetes Control

O. A. Mojiminiyi, N. Abdella. Faculty of Medicine, Kuwait University,

Kuwait, Kuwait
Background: Adipose tissue-derived adiponectin has pleiotropic protective effects
with suppression of inflammatory and metabolic derangements that may result in
insulin resistance, metabolic syndrome, Type-2 diabetes (T2DM) and cardiovascular
disease. No study has evaluated the potential clinical significance of adiponectin
measurements that may be useful in routine practice. The aim of this study was to
evaluate adiponectin as a screening tool and diagnostic marker of T2DM and diabetic
control.
Methods: Fasting adiponectin, insulin and glucose and HbA1c were determined in
575 subjects with undiagnosed diabetes but with family history of T2DM. To evaluate
adiponectin as a marker of DM control, we studied 376 patients with known T2DM
duration of 12.4 8.1 years. Clinical and anthropometric data were recorded and
subjects were classified on the basis of the degree of adiposity, insulin resistance (IR)
using the homeostasis model assessment, and achievement of target HbA1c levels <
53mmol/mol. Using standard cut off values for glucose and HbA1c, receiver operating
characteristic curve (ROC) analysis was used to examine the diagnostic performance
characteristics for undiagnosed DM.
Results: In undiagnosed subjects, adiponectin was significantly lower in subjects
with IR (7.0 vs 8.5 g/mL) and diabetic subjects (7.4 vs 8.6 g/mL) compared
with those without. 73 of 575 subjects were found to have T2DM. Binary logistic
regression showed that the odds ratio of T2DM as predicted by adiponectin was 0.88
[95% confidence interval 0.80-0.96; p = 0.007]. At cut-off points of 7.5 g/mL, the
diagnostic sensitivity and specificity of adiponectin for T2DM were 88% and 51%
respectively. Using the ADA glucose and HbA1c diagnostic criteria as reference, the
area under the adiponectin ROC curve for diagnosis of DM was 0.740 (95% CI 0.570
- 0.910). In known T2DM subjects, those with good control (HbA1c < 53mmol/mol)
had significantly higher adiponectin (8.5 vs 7.1 g/mL) compared to subjects with
poor control.
Conclusions: Adiponectin levels are associated with better glycemic control and
could be useful adjunct for screening for IR and T2DM. Therapeutic measures that
increase adiponectin levels might be valuable targets for improving diabetes control
and decreasing complications.

A-177
Performance of GH stimulation tests at a private laboratory in Brazil: Is insulin
tolerance test still the best?

P. B. M. C. Araujo, K. Marca, L. Santos, J. Duarte, A. C. Bispo, M. Freire.

Diagnsticos da America S.A., Rio de janeiro, Brazil
Background: Growth hormone deficiency (GHD) is the most important endocrine
cause of short stature in childhood. Once growth hormone (GH) secretion is pulsatile,
diagnosis of GHD is made with a combination of clinical assessment, IGF1 and

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

IGFBP3 levels, and GH stimulation tests. In Brazil, the most common GH stimulation
tests are insulin tolerance test (ITT), clonidine and glucagon. The aim of this study was
to determine the performance of those GH provocative tests and to make comparisons
among them and with IGF1 levels.
Methods: Retrospective assessment of GH stimulation tests, performed during the
year of 2013 in children between 4 to 18 years old, including 141 ITT (mean age
11,72 2,58 yrs, 72,3% male), 285 clonidine tests (mean age 10,41 2,91 yrs, 76,5%
male) and 42 glucagon tests (mean age 7,00 3,4 yrs, 66,7% male). The mean dose of
each medication administered was respectively 0,074UI/kg, 0,138 mg and 0,568 mg
and the mean glucose nadir at ITT was 27,95 8,02. Comparison among tests showed
statistical difference regarding peak stimulated GH (GH > 5,0 ng/mL) between ITT
and glucagon (p<0,01) and ITT and clonidine (p<0,01), but not between glucagon and
clonidine tests (p=0,1014). Peak stimulated GH happened at 26,24% (mean GH 10,62
5,78), 82,10% (mean GH 11,88 5,79) and 71,43% (mean GH 11,63 7,38) from ITT,
clonidine and glucagon tests, respectively. GH peaks concentration was at time of
hypoglycemia at ITT (70,6%), at 60 minutes after stimulation with clonidine (60,3%)
and at 2h after stimulation with glucagon (63,3%). Levels of IGF1 did not correlate
with GH answer to the stimulation at insulin (p=0,6165), clonidine (p=0,4914) and
glucagon tests (p=0,5551).
Conclusion: Our finding that clonidine and glucagon tests presented a better rate of
response compared with ITT represents a new scenario for GH provocative tests, once
ITT has been considered the gold standard test for GHD investigation. Maybe it could
be explained by the fact that in a private laboratory environment, we can not let the
patient recover spontaneously, which could increase side effects. Instead, the recovery
from hypoglycemia is done with oral glucose, which may impair the GH peak, finding
reported by Yeste and cols. Hence, clonidine and glucagon tests emerge as a reliable
and safer alternative to ITT. Surprisingly, IGF1 levels did not correlate with rates of
GH response to the tests, which highlights the importance of a clinical/laboratory
combined evaluation.

A-180
A Fully-Automated 1,25-Dihydroxy Vitamin DXp Assay on the IDS-iSYS
Automated System

J. Tran1, D. Bautista2, Z. Seres3, L. Cornaut4, T. Gundlach3, A. Rousseau4,

H. Griesser3. 1Immunodiagnostic Systems Ltd, Westminster, CA,
2
Immunodiagnostic Systems Ltd, San Clemente, CA, 3Immunodiagnostic
Systems Ltd, Boldon, United Kingdom, 4Immunodiagnostic Systems Ltd,
Pouilly, France
1,25-Dihydroxyvitamin D (1,25D) is one of the major regulators of calcium
metabolism. Due to its lipophilic nature and low circulating concentration, the
measurement of 1,25D concentration levels has been labour intensive and technique
dependent in addition to multiple equipments required for the sample purification
procedure. We reported the results of fully automated IDS-iSYS 1,25 VitDXp assay.
IDS-iSYS 1,25 VitD assay purifies 1,25D in human sera utilising the anti-1,25D
antibody coated magnetic particles in cuvette 1. After incubation, the magnetic
particles are washed and 1,25D is eluted. The eluate is transferred to cuvette 2 where
the immunoassay procedure is performed utilising the IDS-iSYS 1,25-Dihydroxy
Vitamin D test. The purified 1,25D competes with 1,25D-Acridinium (1,25D-ACR)
for a limited amount of biotinylated anti-1,25D antibody sites. Bound complexes are
captured via streptavidin-coated magnetic particles. Following washing, the bound
1,25D-ACR is measured where signal generated is inversely proportional to the 1,25D
concentration in the sample. Below are the preliminary analytical performance of the
IDS-iSYS 1,25 VitDXp (iSYS XP125) assay:
Xp

Performances
Functional sensitivity
Inter-assay precision

Results
<12.0pg/mL
16.0% (23.5pg/mL)
8.6% (59.4pg/mL)
9.6% (79.3pg/ml)
6.9% (141pg/mL)
Linearity
92-109%
Method comparison against IDS- iSYS XP125 = 0.87 x (iSYS 125D) +6.4pg/mL
iSYS 125D (n = 81)
r = 0.95
Combining the innovative on-board sample purification procedure with the already
proven IDS-iSYS 1,25 Dihydroxy Vitamin D test, the IDS-iSYS 1,25 VitDXp delivers
accurate results for patient care while enhancing the clinical laboratory 1,25D testing
efficiency.

A-181
POST-FIRE STRESS REACTIVITY IN VOLUNTEER FIREFIGHTERS

M. A. LEGUIZAMON1, S. UCA-Villarrica2, A. Lird3, C. Mora2, M.

Cabral4. 1Laboratorio Central del Hospital de Clnicas. Facultad de
Ciencias Mdicas. Universidad Nacional de Asuncin. San LorenzoParaguay, ASUNCION, Paraguay, 2Universidad Catlica Nuestra
Seora de la Asuncin, Campus Guaira. Facultad de Medicina.
Ctedra de Bioqumica. Guair-Paraguay, VILLARRICA, Paraguay,
3
Laboratorio Central del Hospital de Clnicas. Facultad de Ciencias
Mdicas. Universidad Nacional de Asuncin. San Lorenzo-Paraguay,
SAN LORENZO, Paraguay, 4Instituto de Investigaciones en Ciencias de la
Salud., ASUNCION, Paraguay
Firefighters belong to a population especially sensitive to stress factors. The
responsibility for people sharpen the stress feeling due to the need of pushing
oneself, sometimes more that one can. The physical and emotional stress, just like
diseases, could increase the levels of cortisol because during the normal response
to stress, the hypophysis secretes more corticotropin. Objective: To determine the
stress level, the analysis of the level of cortisol in serum in Volunteer Firefighters
of Caaguaz, before and after a 12-hour duty is presented. Material and methods:
observational and longitudinal design. Subjects: Male and female volunteer
firefighters of any rank and in activity from Caaguaz-Paraguay that accepted blood
drawing and do not present disease of the suprarenal glands like Cushing Disease.
The determinations performed were levels of cortisol before and after a 12-hour duty.
Results: Thirty two firefighters participated, 81% were men and 34% were aspirants.
Age range was 18 to 33 years. In the first simple, there was a mean of 10.585.2 ug/
dl and in the second 12.425.3 ug/dl. (p=0.0001). There was an increase of 15.43%
in the aspirants, 22.29% in the firefighters and 4.34% in the instructors. In the others
positions, such as driver and members of the board of directors there was an increase
in the level of cortisol of 18.39%. Conclusions: the rank of the firefighters as well as
the years of experience in the community service influence considerably in the stress
level as the aspirants presented more variation in the level of cortisol in serum after
the stress situation.
Keywords: cortisol, stress, firefighters.

A-182
Evaluation of a direct HbA1c Assay on a fully automated Chemistry Analyzer
versus two common HPLC methods.

D. Rampersaud1, H. Lee1, W. Walters2. 1Clinitox Diagnostix Inc,

Mississauga, ON, Canada, 2Pointe Scientific, Canton, MI
Pointe Scientific, Inc. (now part of MedTest) has developed a latex-enhanced
immunoturbidimetric assay that directly measures the % HbA1c in whole blood.
This new method was adapted to a Roche Hitachi 917 clinical chemistry analyzer
in a fashion that offline sample pretreatment was not required. The objective of this
study was to evaluate the performance of this direct HbA1c assay on the Hitachi 917
chemistry analyzer versus a Tosoh G8 and a Bio Rad Variant II (HPLC methods) at a
large commercial lab.
Samples with normal hemoglobin that were run on the Tosoh G8 versus the Hitachi
917 showed a correlation of R=0.987, with a regression equation of y = 1.18x 0.80.
(n=172). Additional samples containing various hemoglobin variants were compared
to the results from the Biorad and Tosoh systems. Utilizing all samples (normal and
variants) resulted in a correlation of R=0.958, with a regression equation of y = 1.15x
0.77. (n=256). High and low Biorad QC materials were run for Day-to-Day and
Within Run precision. Results are shown below:
Day to Day Precision
Within Run Precision
Average
4.83
9.21
Average
4.75
9.27
Std Dev
0.144
0.166
Std Dev
0.165
0.082
%CV
2.97
1.80
%CV
3.47
0.89
The data shows that for normal samples, performance of the Pointe Scientific
HbA1c assay is comparable to the Tosoh HPLC system. When Biorad and Tosoh
variant sample results were included, the correlation was minimally affected. These
results demonstrate that this assay can be a viable solution for large volume testing
environments where HPLC systems may not be capable of handling a very high
volume of samples.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S49

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

A-183
Analytical evaluation of a glycated protein method on the Siemens ADVIA 1800

T. Higgins, K. Rodriguez Capote, K. Tovell, D. Holmes, J. Dayton.

DynaLIFEDx,, Edmonton, AB, Canada
Background HbA1c has become the gold standard for assessment of glycemic control
in individuals with diabetes mellitus. However the test cannot be used if the individual
has a homozygous hemoglobin variant or a condition in which there is a rapid turnover
of red blood cells. In these cases glycated protein has been proposed as an alternative
assay. We wished to evaluate the analytical performance of the Diazyme Glycated
Protein assay, establish a reference interval and to compare the glycated protein
results against HbA1c results.
Methods The Diazyme Glycated Protein assay was programmed onto a Siemens
ADVIA 1800 analyzer using a program supplied by the reagent manufacturer.
Glycated protein was analyzed on samples that had serum albumin, total protein and
HbA1c results available. Samples with a bisalbumin were also analyzed for glycated
protein by 2 methods.
Results The within run imprecision was calculated at 1.2% and 0.4% respectively
at glycated protein concentrations of 181.2umol/L and 684.2 umol/L. At the same
concentrations the between run imprecision was 2.3% and 1.2% respectively.
Correlation of glycated protein against HbA1c gave a regression equation of y
(glycated protein) = 56.41(HbA1c)-35.72, (n=155, r2=0.91) The reference range was
calculated on 42 samples that had both glucose and HbA1c within the respective
reference intervals and was 165 (90% CI 145 to 186) to 367 (90%CI 340 to 386)
umol/L. This compares with the manufacturers suggested reference interval of 100
to 295 umol/L. Using the ratio glycated albumin /albumin the reference interval was
10.5 (90%CI 9.8 to 11.4) to 18.5(90%CI 17.5 to 19.4). Linearity was established from
10 to 1150 umol/L. Comparison of Diazyme Glycated Protein results on bisalbumin
samples with a glycated albumin method, Lucica GA-L, gave a correlation equation of
y(Diazyme)=-0.5 (Lucica GA-L) +6.2 (n=25,) r2=1, two tailed T test p=0.14).
Conclusion The analytical performance of the Diazyme Glycated Protein was
satisfactory although the reference interval obtained was higher than that suggested by
the manufacturer. Comparison against a glycated albumin method for bisalbunemia
samples was good although the absolute values are different. The effect of albumin
variants needs to be further evaluated.

A-184
Hypercalcemia with Normal PTH: A Diagnostic Puzzle?

O. Krasnozhen-Ratush, L. Bilello, S. Weinerman, L. K. Bjornson. HofstraNSLIJ Scool of Medicine/North Shore-LIJ Health System, Manhasset, NY
Introduction: The two major causes of hypercalcemia are primary hyperparathyroidism
and malignancy, where patients generally have elevated PTH results in the former and
suppressed PTH in the latter condition. However, there is a smaller group of patients
who present with mild hypercalcemia and normal PTH that can be a diagnostic puzzle
if clinical history and condition are unremarkable.
Objective: To determine the prevalence of hypercalcemia with normal PTH in a
large endocrinology practice in an integrated health system and to determine what
effect correcting (adjusting) the calcium concentration for albumin would have on the
classification of hypercalcemia.
Results/Discussion: 3,958 calcium results were retrieved by a computer search of the
data base from the endocrinology faculty practice at North Shore-LIJ Health System
in Long Island, NY from January 2013 through January 2014, where 155 (3.9%) of
these results have been classified as hypercalcemic, i.e., above 10.5 mg/dL. This study
focused on calcium, PTH and albumin results without any patient history or other
clinical information. Within this original hypercalcemic group a subgroup of 42 (
27.1% ) patients were identified with normal PTH results. For this subgroup albumin
results were also retrieved and a corrected calcium concentration was calculated
according to a standard textbook equation; Ca (corrected) = Ca (measured) + 0.8 x
(4.0 Albumin), where the calcium and albumin units of measure are mg/dL and g/dL,
respectively. From the original subgroup of 42 patients, 18 (42.9%) were reclassified
to normocalcemia based on corrected calcium results whereas 24 (57.1%) remained
in the hypercalcemia classification. While the correction of calcium for albumin is
not routinely performed or reported in most laboratories, its primary application has
been for patients with hypoalbuminemia and not for patients with normal or high
normal albumin as in this subgroup. However, the fact that a corrected calcium
concentration reclassifies nearly half of these patients as normocalcemic gives rise
to questions regarding the original classification of hypercalcemia based on total
calcium concentrations. While measurement of free (ionized) calcium is considered

S50

a more accurate assessment of calcium status than total calcium, most laboratories
primarily perform total calcium measurements for outpatients since this assay is easily
automated and provides rapid, cost effective results. Outreach physicians do not often
order free calcium because it is more expensive, has special requirements for blood
collection and is usually not required for diagnosing calcium abnormalities. However,
corrected calcium results are an estimation and free calcium measurements should be
performed if the total calcium measurement is in question.
Conclusions: We have investigated a subgroup of hypercalcemic patients that have
normal PTH results and have found that nearly half (42.9%) are reclassified as
normocalcemic when the total calcium concentration is corrected for albumin. Since
corrected calcium concentrations are an estimation it is suggested that measurement
of free calcium may provide a more accurate assessment of calcium status for some
patients in this subgroup.

A-185
Validation of a productive platform for HbA1c testing

N. Z. Maluf, A. L. N. Camilo, C. Rosin, C. F. d. Pereira, O. V. P. Denardin.

DASA, Barueri, Brazil
Background: The current guidelines for diabetes have recommended the use of
HbA1c testing for diagnosis as well for monitoring of diabetes mellitus type 2. In
HbA1c testing has the important role of diagnosis, there is the need for accurate
and precise methods for quantification as recommended by the National Glycated
Hemoglobin Standardization Program (NGSP). The purpose of exceeding the
expectations of customers makes the labs have to consolidate their growth strategies
and strengthen its activities in order to deliver results quickly, with quality and
updated methodologies. Large routines demand faster methods which are efficient
and have unquestionable quality results. DASAs routine faces an average of 917,000
tests per month where we need to ensure that the time spent for tests processing do
not generate a negative impact on laboratory routine, while guaranteeing that the
integrity of the samples and the results of high quality promote the positive impacts
within established time to deliver results. This study aims to validate a platform which
improves productivity, reduces the time for analysis of samples and the number of
installed equipments, enables the allocation of resources and maintains acceptable
correlation with the methodology currently used for HbA1c testing.
Methods: HbA1c for Siemens ADVIA 2400 Chemistry is a turbidimetric assay
with range from 2.9% to 15.4%. To assess the assays correlation, we analyzed
100 whole blood samples collected in EDTA K2 tubes. The samples were divided
according to the three following ranges: 4 to 6%, 6 to 14% and >15%. The results
were compared with results obtained from Tosoh G7 platform. To compare the
productivity, time to analyze the samples and TAT (Turn Around Time) were used 5
days of LIS data for routine performed at Tosoh G7 and a second period held on
ADVIA 2400 Chemistry.
Results: The comparative analysis of the results revealed a correlation coefficient
(r) of 0.98 for HbA1c in ADVIA 2400 Chemistry and a linear regression equation
y = 0.9808 x + 0.2838 and R2 = 0.9579. When we verified LIS data analysis, the
total processing time was 10.6 hours for a routine of 7,422 samples on five platforms
Tosoh G7 and 4 hours for a routine of 8,550 samples on one platform ADVIA 2400
Chemistry . The TAT for ADVIA Chemistry 2400 was 06 hours and 33 minutes
and the Tosoh G7 was 20 hours and 32 minutes.
Conclusion: In the present study, Siemens HbA1c assay for the ADVIA Chemistry
System has a proper correlation with the results of the Tosoh G7 . Thus, it ensures
that the migration of this assay to tested platform causes no significant difference
in results and clinical conduct. Furthermore, ADVIA Chemistry 2400 presented
higher throughput and a 68% reduction of TAT compared to the Tosoh G7 .

A-186
The correlation of Fasting and Post Prandial Plasma Glucose with Estimated
Average Glucose Levels

S. Ozdemir, N. Ozcan, M. Yavuz Taslipinar, F. Ucar, Z. Ginis, A. Guneyk,

A. E. Arzuhal, E. Bulut, G. Erden. Ankara Diskapi Yildirim Beyazit
Research and Training Hospital, Department of Clinical Biochemistry
Ankara, Turkey
Background: Association of fasting plasma glucose (FPG) and post prandial plasma
glucose (PPG) on hemoglobin glycation is still controversial. Estimated average
glucose (eAG) is a value calculated by hemoglobin A1c (HbA1c) that represents
the integrated values for glucose over the preceding 8-12 weeks. The A1C-Derived
Average Glucose (ADAG) study showed a linear relationship between HbA1c and

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

eAG in both diabetic and non-diabetic patient populations. The aim of the present
study was to investigate the correlation of fasting and post prandial plasma glucose
with estimated average glucose levels.
Methods: This retrospective study includes 7887 subjects with HbA1c data(within
a 4 months period). Three subgroups were created: group 1 with FPG (n=4596;
2218 diabetic, 2478 non-diabetic), group 2 with PPG (n=2775, 1683 diabetic, 1092
non-diabetic), and group 3 with FPG and PPG (n=516; 280 diabetic, 236 nondiabetic). The equation published by the ADAG study was used to obtain eAG with
the following formula: eAG mg/dL = 28.7x HbA1c(NGSP,%)-46.7 [eAGmmol/
L=1.59x HbA1c(NGSP,%)-2.59]. HbA1c values were measured by boronate affinity
chromatography methods (Trinity Biotech, Premier Hb9210, Ireland). Glucose was
measured by glucose oxidase method (ADVIA 2400 Chemistry System, Siemens
Healthcare Diagnostics Inc.,Tarrytown USA). The correlation coefficients and their
significance were calculated using the Pearson test.
Results: In all subjects for group 1 and 2, FPG and PPG showed a strong positive
correlation with eAG (r=0.817, p<0.01 and r=0.853, p<0.01 respectively). There was
a positive correlation between FPG and eAG in group 1(diabetic subgroup r=0.643; p<
0.01 and nondiabetic subgroup r=0.397; p<0.01). A positive correlation was observed
between postprandial glucose and eAG in group 2 (diabetic subgroup r=0.762; p<
0.01 and nondiabetic subgroup r=0.357; p<0.01). It was also found out that eAG had
positive correlations with FPG and postprandial glucose in group 3 (diabetic subgroup
r=0.725; p< 0.01 and nondiabetic subgroup r=0.581; p<0.01; diabetic subgroup
r=0.632; p< 0.01 and nondiabetic subgroup r= 0.255; p<0.01 respectively).
Conclusion: eAG values obtained from HbA1c were highly correlated with FPG and
PPG values. Thus, we may suggest that blood glucose expressed as eAG improves the
understanding of blood glucose monitoring.

A-187
Development of quantitative Estradiol assay for fully automated analyzer
LUMIPULSTM G1200

Y. Kitamura, Y. Ishii, H. Murakami, S. Kojima, M. Tomita, S. Sato, K.

Aoyagi. Fujirebio.Inc, Tokyo, Japan
Background: Estradiol (E2) is one of a female steroid hormone which is produced in
ovarian tissue. Mainly it is used for monitoring of ovarian hypo-function or infertility
treatment. In this time, we developed new reagent (Lumipulse G E2-III) which has
excellent correlation with ID-GC/MS and reference materials (IRMM) and improved
cross reactivity to some drugs or E2 derivative. Lumipulse G E2-III is one-step
immunoassay, and E2 in specimen samples and ALP-labeled E2 competitively react
with anti E2 monoclonal antibody coated on the micro particles. It is finally detected
based on CLEIA technology. Here we show the excellent fundamental performance
with fully automated chemiluminescence analyzer LUMIPULSE G1200.
Methods: The sample types used for this study were serum or Heparin-Li. Correlation
with ID-GC/MS, commercial competitive kit, matched pair correlation between
serum and plasma, cross-reactivity to drugs, within-run precision, limit of detection
(LoD) and limit of quantification (LoQ) were evaluated following recommendation
from CLSI documents EP-5, EP-7, EP-12, EP-14 and EP-17. All evaluations were
executed with LUMIPULSE G1200 (FUJIREBIO INC.).
Results: Correlation with ID-GC/MS with 25 specimen samples was excellent (slope:
1.04, regression: 1.00) and the measurement value in Lumipulse G E2-III calibrators
was traceable to BCR577 reference materials. The significant correlation with the
commercial available kit with 79 specimen samples was observed (Cobas, slope: 0.93,
regression: 1.00, Centaur, slope: 1.06, regression: 1.00). Correlation between serum
and heparin-Li with 56 matched pair samples was excellent (slope: 1.03, regression:
1.00). Within-run precision % CVs for our assay ranged from 1 to 3% when 3 different
conc. of samples were tested, the calculated LoD is at 15 pg/mL and the LoQ ranged
from 13 pg/mL to 17 pg/mL. As a result of evaluation with total 43 kinds of drugs and
E2 derivatives, cross reactivity with almost all cross reactants were < 0.1 %.
Conclusion: These results demonstrated that improved Lumipulse G E2-III is a
precise and highly useful for measuring E2 in serum and heparin-Li. Also this assay is
perfectly traceable to ID-GC/MS and reference materials.

A-188
Plasma total homocysteine, high - sensitivity C- reactive protein and thyroid
function in metabolic syndrome patients in Port Harcourt, Nigeria.

C. G. Orluwene1, M. O. Mommoh2. 1University of Port Harcourt Teaching

Hospital, Port Harcourt, Nigeria, 2NNPC Zonal Industrial Hospital, Port
Harcourt, Nigeria
Aims and Objectives: Metabolic syndrome is assuming epidemic proportions in
most populations of the world and so are its co-morbidities. Metabolic syndrome,
hypothyroidism, total plasma homocysteine and C-reactive protein are independent
risk factors for cardiovascular disease. There could be an association between these
risk factors. The aim of our study was to examine the levels of thyroid hormones,
total plasma homocysteine and C-reactive protein in metabolic syndrome patients
and assess the possibility of an association between these four risk factors for
cardiovascular disease.
Patients and Methods: A total of 93 subjects were recruited for this study. (48
with metabolic syndrome and 45 as the control group). Basic demographic data,
components of the metabolic syndrome, thyroid hormones, high-sensitivity C-reactive
protein and total plasma homocysteine were estimated for all subjects using standards
methods. Appropriate statistic was used for data analysis.
Results: Components of the metabolic syndrome, thyrotropin, total plasma
homocysteine and high sensitivity C-reactive protein were significantly higher in the
study group (P<0.05) while free thyroxine and high density lipoprotein cholesterol
were significantly lower in the study group (P<0.05). The predominant type of thyroid
dysfunction in the study group was sub-clinical hypothyroidism (87.5% in the study
group compared to 18% in the control; P<0.05). Logistic regression showed significant
association between sub-clinical hypothyroidism, total plasma homocysteine and
high-sensitivity C-reactive protein in the study group.
Conclusion: There was a strong association between metabolic syndrome, subclinical hypopthyroidism, total plasma homocysteine and high-sensitivity C-reactive
protein. Females are at an increased risk of this association.
Keywords: Metabolic syndrome, Thyrotropin, Sub-clinical hypothyroidism, total
plasma homocysteine, high-sensitivity C-reactive protein.

A-189
An Evaluation of Three Novel Biomarkers: Total-sLHCGR, LH-sLHCGR and
hCG-sLHCGR.

M. Jansen1, C. F. Rosmalen1, S. Banerjee2, A. E. Chambers2. 1NBCL B.V.,

Nijmegen, Netherlands, 2Origin Biomarkers, Lanarkshire, United Kingdom
Background: Researchers in the field of human reproduction need better research
tools for the care of pregnant woman (e.g., prenatal diagnosis) and women who wish
to become pregnant (e.g., fertility treatment). The assays for the soluble receptor for
both human chorionic gonadotropin (hCG) and human luteinizing hormone (hLH)
and the same soluble receptor bound to either hCG or hLH could be used as such
research tools. The receptor is called the soluble LH/hCG receptor or sLHCGR. Three
different ELISA assays have been developed and based on the type of HRP-labeled
detector antibody employed, the user can quantitate Total sLHCGR, a complex of
LH with sLHCGR or a complex of hCG with sLHCGR in a standard microtiter plate
format. Methods: The levels of Total-sLHCGR, LH-sLHCGR or hCG-sLHCGR
were measured using sandwich ELISA. The three ELISA formats are the same;
samples were diluted 5-fold, incubated for 10 minutes in a microtiter plate coated
with a monoclonal capture antibody directed against the receptor. Next the HRPlabeled conjugate was added and incubated for 1.5 hour to generate the sandwich.
After a wash step the substrate was added and after 20 minutes the reaction was
stopped and read at 450 nm using a microplate reader. The optical densities from the
reader are directly proportional to the amount of Total-sLHCGR, hCG-sLHCGR or
LH-sLHCGR present in the sample. Results: The analytical sensitivity for the assays
are 0.01, 0.004, and 0.02 Pmol/mL for the Total-sLHCGR, hCG-sLHCGR and LHsLHCGR, respectively. The within assay precision was done at three different levels
for the three assays (n=16). For hCG-sLHCGR the CV% are 4.4, 3.5, and 3.5% at 2.1,
11.7 and 18.0 Pmol/mL, respectively. For LH-sLHCGR the CV% are 3.2, 3.7, and
3.5% at 2.1, 11.2 and 28.9 Pmol/mL, respectively. For Total-sLHCGR the CV% are
7.4, 6.5, and 9.0 % at 3.5, 4.3 and 11.5 Pmol/mL, respectively. Inter-assay precision
was determined for the three assays (n=4) and were for the hCG-sLHCGR 2.7, 7.7 and
4.2% at 2.2, 12.7 and 19.1 Pmol/mL, respectively. For LH-sLHCGR the CV% were
3.5, 3.6, and 2.9% at 2.1, 11.6 and 30.3 Pmol/mL, respectively. For Total-sLHCGR the
CV% were 8.9, 8.6, and 11.0% at 3.7, 5.2, and 11.8 Pmol/mL, respectively. Sample
recoveries for the three assays are between 80.6 and 101.7% Conclusion: Three

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S51

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

simple, fast ELISA assays for the detection of Total-sLHCGR, hCG-sLHCGR and
LH-sLHCGR in serum or plasma were evaluated and were shown to be precise and
sensitive for the detection of these novel biomarkers.

A-190
Iatrogenic Vitamin D Toxicity in an Infant: Clinical Relevance of Vitamin D
Metabolic Profiling

H. Ketha, H. Wadams, M. Fahse, A. Lteif, S. Grebe, R. J. Singh. Mayo

Clinic, Rochester, MN
Background Public concern over vitamin D (vitD) deficiency has led to widespread
use of over the counter (OTC) vitD supplements, containing up to 10,000 IU
(400IU=10g ). Overzealous use of such supplements can result in vitD toxicity. Infants
are particularly vulnerable to toxicity associated with vitD overdose. Mutations in the
CYP24A1 gene have been shown to cause reduced serum 24,25-dihydroxyvitamin
D (24,25(OH)2D) to 25-hydroxyvitamin D 25(OH)D ratio (<0.02), elevated serum
1,25-dihydroxyvitamin D (1,25(OH)2D), hypercalcemia, hypercalciuria and
nephrolithiasis. Additionally, studies in infants have shown that C3 epimer of 25(OH)
D can contribute upto 9-61.1% of the total 25-OH)D. Therefore, measurements of
parathyroid hormone (PTH) and vitD metabolites 25(OH)D, 1,25(OH)2D, 3-Epi-25hydroxyvitamin D (3EPI-25(OH)D) and 24,25(OH)2D are useful to diagnose and
manage hypercalcemia due to vitD overdose or due to CYP24A1 mutations.
Relevance to Clinical Laboratories Measurement of vitD metabolites 25(OH)D,
1,25(OH)2D, 3EPI-25(OH)D and 24,25(OH)2D is of clinical value for differentiating
between genetic vs iatrogenic causes of hypercalcemia.
Case A significantly underweight four month old female presented with a three day
history of emesis, diarrhea, lethargy and dehydration. The medical work-up revealed
hypercalcemia, hypercalciuria and nephrocalcinosis. She had been exclusively breastfed and had been given OTC vitD supplementation at a higher dose than recommended
on the supplements label. 25(OH)D, 1,25(OH)2D, 3EPI-25(OH)D, 24,25(OH)2D and
the vitD content of the OTC preparation were measured by LC-MS/MS.
Results Nephrocalcinosis was confirmed by ultrasound studies. Serum cacium (SCa)
was 18.7 mg/dL (ref range: 9-11 mg/dL) and PTH was < 6pg/mL (ref range: 15-65 pg/
mL) at presentation. Urine calcium was 157 mg/dL with a calcium to creatinine (Cr)
ratio of 2618 mg/g of Cr (ref range <2100 mg/g). The vitD content of the supplement
was threefold higher (6000 IU of D/drop) than listed on the label (2000IU). Combined
with the gross overdosing, this was estimated to have resulted in a daily vitD dose of
50,000 IU for two months. The SCa gradually decreased upon calcitonin injection on
day 3, but trended upward again (15.5, 13.3, 12.1, 10.4, 9.8, 10.7, 10.9, 10.3 and 11.6
mg/dL on days 1, 3, 5, 7, 11, 14, 20, 25 and 40 respectively). The 25(OH)D decreased
slowly from 294 ng/mL on day 1 to 257, 227, 197, 189, 138, 124 and 84 ng/mL on
days 3, 5, 7, 11, 14, 20, 25, and 40 respectively. Serum 3EPI-25(OH)D levels were
44 % to 29 % of the 25(OH)D levels(126-36 ng/mL). Serum 1,25(OH)2D levels were
elevated. The ratio of 24,25(OH)2D to 25(OH)D was 0.11-0.14 (ref range: 0.07-0.18).
Discussion Genetic cause of hypercalcemia, hypercalciuria and nephrocalcinosis
could be ruled out on the basis of normal 24,25(OH)2D to 25(OH)D ratio. Clear
warning labels regarding maximum allowable doses of OTC vitD supplements are of
value from a public health perspective.
Conclusion Infants receiving high dose OTC vitD supplementation are vulnerable
to vitD toxicity. Vitamin D metabolic profiling is of value for evaluating such cases
and to exclude certain genetic causes. Our study also emphasizes the need for stricter
regulation of vitD content in OTC supplements and prominent warning labels
regarding maximum allowable daily doses.

A-192
Insulin resistance and secretory functions in Pre-diabetics and newly diagnosed
diabetics of north-west India: role of adipocyte mediators

hyperglycaemic patients attending our diabetic clinic. The study subjects were
grouped as-Group I: Healthy control (n = 56), Group II: Pre-diabetics (n=39), Group
III: Diabetics (n= 124). All subjects were evaluated for waist to hip ratio (W: H), body
mass index (BMI), fasting blood glucose, insulin, HOMA-IR, HOMA-beta, NEFA
and lipid profile. Statistical analysis was done using ANOVA and Multiple logistic
regression analysis.
Results: The diabetic subjects had significantly raised W: H and BMI as compared to
the pre-diabetics (0.880.06; 27.444.62) and controls (0.870.06; 24.24.34) with
the F value 14.64 (p<0.001) and 5.98 (p=0.003) respectively for W: H and BMI. Age
adjusted base line characteristics according to BMI and W: H quintiles for predicting
risk of type 2 Diabetes showed significant trends across quintiles for total cholesterol
(TCH), triglycerides (TG), HOMA-IR and HOMA-beta. Similarly Age adjusted base
line characteristics according to NEFA quintiles for predicting risk of type 2 Diabetes
showed significant trends across quintiles for BMI, W: H, lipid profile, insulin,
HOMA-IR and HOMA-beta. Finally Multiple logistic regression analysis in newly
diagnosed type 2 diabetic subjects with family history (Negative v/s Positive) as a
dependent variable showed the strongest risk due to raised NEFA (OR 3.83), followed
by HOMA-IR (OR 1.38), TCH (OR 1.35), WC (OR 1.7) and TG (OR 1.13).
Conclusion: We conclude that the insulin secretory rates and IR in pre-diabetics and
newly diagnosed type 2 diabetics are associated with BMI, W: H and NEFA. W: H
and NEFA can prove to be a strong predictor of type 2 DM even with a negative
family history.

A-193
Comparison of Aldosterone and Renin Determination by Conventional
Radioimmunoassay and by Automated Chemiluminescente Immunoassay

E. Zeruya, Y. Sharabi, H. Kanety, R. Hemi. Sheba Medical Center, RAMATGAN, Israel

Background: Renin-angiotensin-aldosterone system plays a paramount role
in water homeostasis and electrolyte balance and in the regulation of arterial
pressure; Aldosterone, major mineralcorticoid secreted by adrenal cortex, keeps
water and salt balance helping in maintaining blood pressure. Renin, a proteolytic
enzyme synthesized primarily by the juxtaglomerular cells of the kidneys converts
angiotensinogen to angiotensin I, which is then cleaved by angiotensin converting
enzyme to form angiotensin II. Angiotensin II increases blood pressure directly
through vasoconstriction and indirectly by stimulating secretion of aldosterone.
Measurement of plasma renin and aldosterone provides a marker of renin-angiotensinaldosterone system activity.
Objective: To evaluate analytical performances of commercially available automated
immunoassays for Aldosterone and direct renin (DiaSorin LIAISON) and to compare
LIAISON results to currently used radioimmunoassay (RIA).
Results: 43 plasma samples were measured for Renin activity (PRA) and Renin
concentration (PRC) by RIA (RADIM) and LIAISON, respectively. In addition,
57 serum samples were analyzed for Aldosterone levels by RIA (SIEMENS) and
LIAISON. High correlation was found between the two systems both for Aldosterone
and Renin (R=0.95 and R=0.96, respectively, Figure 1 A&B). Satisfying clinical
concordance of 95% and 100% was found between both LIAISON Aldosterone and
Renin assays vs. RIA, respectively. Assays performance (precision and linearity) was
tested and found compatible with the manufacturers declaration.
Conclusion: LIAISON Aldosterone and Direct Renin assays are precise, accurate
and reliable and can be used as an alternative to the conventional RIA which is
technically demanding and laborious. Our data suggests that the LIAISON Direct
Renin assay can be a good diagnostic and practice replacement for the RIA PRA assay,
although biochemically, they measure different parameters. Setting normal ranges for
the Renin/Aldoseterone ratio on the LIAISON is required in order to diagnose and
monitor hypertensive patients.

P. Sharma, P. Purohit. All India Institute of Medical Sciences, Jodhpur,

India
Background: Diabetes Mellitus (DM) is a group of metabolic diseases characterized
by hyperglycemia resulting from defects in insulin secretion, insulin action, or both.
There has been a dearth of studies evaluating the natural course of insulin secretion in
pre-diabetics progressing to diabetes in north western India.
Aim: To examine the natural history of insulin secretory dysfunction and insulin
resistance during the development of diabetes and to examine the role of adipocyte
mediators non-esterified fatty acids (NEFA) in development of type 2 DM.
Material and Method: The study was conducted on newly diagnosed, untreated

S52

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

AMR self-management, solving linearity studies issues, ensuring additional safety
and reliability of released results, and Laboratorys Quality Control Improvement.
Table 1 Assay

Assay Linearity
Range

Obtained
CV %

VB12
FOL
FER
BNP
Cor
Pep C
Ins

45 - 2000 ng/mL
0.35 - 24 ng/mL
0.5 - 1650 ng/mL
2 - 5000 pg/mL
0.2 - 75 ug/dL
0.05 - 30 ng/dL
0.5 - 300 mU/L

4.13
2.43
5.92
2.63
3.97
5.14
4.19

CV Lab
max Target
%
7.50
12.00
7.10
2.67
10.50
8.30
10.60

Obtained TE max
Linear Regression
TE %
Target %

R2

14.35
23.43
5.38
9.59
13.76
12.70
12.74

0.99759
0.96698
0.99532
0.99672
0.99941
0.99360
0.98842

30.00
39.00
16.90
12.43
29.80
20.80
32.90

0.958x + (50.874)
0.854x + (1.557)
1.109x + (-32.411)
1.038x + (-76.888)
1.057x + (-1.134)
0.985x + (-0.173)
0.993x + (4.202)

A-197
Analytical performance of the Insulin-like growth factor I (IGF-I) assayin two
Immunoassays Systems that used different standardization procedures

M. L. Moreira, M. D. C. Freire. Diagnsticos da Amrica, Duque de

Caxias, Brazil
Background: Insulin-like growth factor I (IGF-I) is intended for use in the
diagnosis and monitoring of children who have growth-related disorders and adults
with acromegaly.Large variability exists among different IGF-I assays owing to
differences in calibration, antibody specificity, isoform recognition. The World Health
Organization (WHO) Expert Committee on Biological Standardization (ECBS)
established criteria for standardization and evaluation of IGF-1 assays(Recombinant
IGF-1, coded 02/254) and for the content monitoring of therapeutic products.
Objective:This study intends to evaluate the analytical performance of the IGF-I assay
on Siemens IMMULITE 2000 IGF-I (Recombinant IGF-1, coded 87/518) and Liaison
Analyser IGF-I Diasorin (Recombinant IGF-1, coded 02/254).
Methods:We tested59 patients serum samples from DASA, Rio de Janeiro, with
concentrations within the range of 25 ng/mL to 1039 ng/mL, in both analytical
platforms.The IMMULITE 2000 IGF-I is a 2-cycle, sequential immunometric assay
withcalibration range of 20-1600 ng/mL.In the assay procedure, prediluted patient
sample is needed to reduce interference from binding proteins before analysis. The
photon output is proportional to theconcentration of the analyte.The Liaison Analyser
IGF-I is a 1-cycle, immunometric assay with calibration range of IGF-I 10-1500 ng/
mL.
Results:Comparison between IMMULITE 2000 and Liaison assaysyielded
a correlation coefficient of 0.97, with linear regression of x(IMMULITE
2000)=0.792(Liaison) + 30.48ng/mL. Moreover, means were 313 and 278 ng/mLfor
IMMULITE 2000 and Liaison Analyser, respectively. Test t-student was 1.049
(expected is 1.98) and estimated total error (TE) was 20.2% which is lower than the
allowed TE (22.35%) in all levels of decision-making practice.

A-196
Linearity study to define the Analytical Measurement Range (AMR) for
immunoassays

A. L. Camilo, D. Waltrick, C. F. A. Pereira. DASA, So Paulo, Brazil

Background:Linearity studies are crucial to assure analytical range accuracy and to
validate reagents performance in clinical laboratory. Due to difficulties to evaluate
linearity results, a Brazilian laboratory implemented a new tool (Easy Linearity
Curve, ELC).ELC established self-inspection program, immune-hormone analytical
systems monitoring and quality control requirements accomplishment. We described
linearity data for nine different immunoassays: VB12, Folato(FOL), Ferritin(FER),
BNP, Cortisol (Cor), C-Peptide(Pep C) and Insulin(Ins).
Materials and methods: Samples were selected from laboratory routine,
with concentrations within assay linearity range for each test, performed by
chemiluminescence on Advia Centaur XP (Siemens Healthcare Diagnostics).
Linearity was evaluated by calculating the recovery of repetition performed. The
coefficients of second and third degree regression are statistically equal to zero, at 5%
of the significance level. In addition, they were analyzed with ELC tool to validate
immunoassays AMR according to CLSI Guideline EP6-A.
Results and Discussion:Tests showed satisfactory linearity results for different
samples. They are summarized in the table below.The sample pools were not used,
because all the results from this dilution test presented high percentage of recovery,
higher than manufacturers references, due to a matrix effect (data not showed).The
main difficulty in linearity studies is to obtain samples in the ideal analytical range.
ELC uses an algorithm that suggests samples dilutions to contemplate the whole
AMR for each parameter. In this study, for each analyte, five different samples were
selected to simulate ideal concentrations covering the linear range.
Conclusion:Based on our results, we do not recommend the use of sample pools in
linearity studies of hormones.The presented protocol permits laboratory autonomy to

Conclusion:Preliminary assessment of these clinical evaluation results indicates that

the IMMULITE 2000 IGF-I immunoassay is a precise method for measuring IGF-I
in serum across a wide range of clinically relevant concentrations and shows good
correlation to the Liaison IGF-I assaydespiteof using different standardization.

A-198
Diiodothyropropionic acid interferes with TT3 and FT3 measurements on
common Immunoassay Platforms for Thyroid Function Panel.

X. Yi, S. Refetoff, E. K. Y. Leung, K. T. J. Yeo. The University of Chicago,

Chicago, IL
Background: Monocarboxylate transporter 8 (MCT8) is a thyroid hormone-specific
cell membrane transporter. MCT8 deficiency produces in young males an unusual
pattern of thyroid hormone abnormality with elevated serum triiodothyronine
(TT3) and causes severe neuropsychiatric defect. A thyroid hormone analogue,
diiodothyropropionic acid (DITPA) was found to enter cells independently of MCT8.
Therefore, this compound was tested in children with MCT8 deficiency and found
to normalize their thyroid function tests and improve their nutritional status. A
problem in the follow up of DITPA treatment is the interference of DITPA in the
routine laboratory measurement of TT3. This necessitates the measurement of DITPA
by LC-MS/MS to correct for its interference in TT3 determination by immunoassay.
The objective of this study was to evaluate the possible interference of DITPA on
commercial thyroid assays available from four in-vitro diagnostic companies.
Method: Pooled human serum was collected and stock DITPA (1mg/ml) was added
to create a serum set containing 3 -75 g/dL DITPA. This sample set was then assayed
for TT3, TT4, free T3, free T4, and TSH on the Roche Elecsys, Siemens IMMULITE,
Siemens ADVIA Centaur, Siemens Dimension EXL, Siemens Dimension RXL,

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S53

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

Beckman Access, and Abbott Architect platforms, where available, respectively. All
samples were analyzed in duplicate and the respective assay values in the DITPAtreated samples were expressed as the difference above the baseline sample. In
addition, to investigate if the interference of DITPA could be overcome by excess
TT3, serum samples at a constant concentration of 20 g/dL DITPA were spiked with
increasing amount of TT3 and measured on Roche Elecsys TT3 method. Results: At
75 g/dL DITPA, the overestimation above the baseline endogenous values were:
Elecsys: TT3, 193 ng/dL, FT3, >3000 pg/dL; Access: TT3, 745 ng/dL, FT3, 2405
pg/dL; IMMULITE: TT3, 146 ng/dL; Architect: TT3, 33 ng/dL, FT3, 96 pg/dL,
respectively. Minimal interference was observed for ADVIA Centaur TT3 assay at
all concentration of DITPA tested. TT4, FT4, and TSH tests were not affected by
DITPA on all the immunoassay platforms tested. It is also interesting that with excess
TT3, the DITPA interference on the Elecsys TT3 declined from 208% to 114% above
baseline. Conclusion: DITPA significantly interferes with several commercial TT3
and FT3 assays with the exception of the Siemens ADVIA Centaur TT3 assay. This
suggests that the TT3 antibody reagents used in the Elecsys, Access, Immulite and
Architect cross-react significantly with DITPA. For patients undergoing treatment,
the ADVIA Centaur TT3 assay is suitable for monitoring response to DITPA therapy.

A-199
Gender Differences in the interactions between Adipokines and the Insulin-Like
Growth Factor-I System in a Metabolically High Risk Population

O. A. Mojiminiyi, N. Abdella. Faculty of Medicine, Kuwait University,

Kuwait, Kuwait
Background: Accumulating evidence indicate important roles for the insulin-like
growth factor (IGF)/IGF-binding protein (IGFBP) system in metabolic homeostasis.
Despite potential molecular mechanisms that link obesity and insulin resistance with
the IGF system, the pathophysiological metabolic interactions between adipose tissue
derived adipokines and the IGF-I system remain unknown due to conflicting reports
in the literature. In this study, we test the hypothesis that gender differences could be
responsible in part for the conflicting reports on the associations of some adipokines
with the IGF system.
Methods: Fasting adiponectin, resistin, leptin, leptin receptor (sOB-R), insulin,
glucose, total IGF-I, IGFBP-3 and full lipid profile were determined in 590 (238M and
352F) first-degree relatives of patients with Type 2 Diabetes Mellitus. Sex hormone
binding globulin (SHBG), oestradiol (E2), testosterone (T), were also measured. Free
androgen index (FAI), Free leptin index (FLI), bioavailable IGF-1 (BIGF1), betacell function (%B), insulin sensitivity (%S) and insulin resistance (IR) (Homeostasis
Model Assessment) were calculated. The data were analysed using simple and
multivariate regression analyses.
Results: There are significant differences in mean (SEM) BIGF1 between males
(87.6 (9.1)) and females (67.7 (4.6)). There were also significant gender differences
in adiponectin, leptin, sOB-R, FLI, %S and IR. There were no gender differences in
resistin and IGFBP3. Significant gender differences were found in the correlations
of BIGF1. The following showed significant correlations with BIGF1 in females
but not in males: adiponectin, sOB-R, FLI, SHBG, glucose, insulin, %S, IR, waist
circumference, BMI, Apo B, total cholesterol, triglycerides and LDL-cholesterol.
Males and females showed similar correlations of all other variables with BIGF1.
Correlations with sex hormones (E2, T, SHBG, FAI) were not significant in males
and females. Multivariate linear regression analysis showed that age, BMI, WC,
adiponectin, FLI were significant determinants of BIGF1 in females but not in males.
Age was the only significant determinant of BIGF1 in males.
Conclusions: There are significant gender differences in the metabolic interaction
between adipokines and the IGF-1 system. Despite the putative links with obesity, sex
steroids do not play a role in the gender differences.

A-200
LC-MS/MS detection of increased Androstenedione levels in patients receiving
Danazol therapy

V. Gounden1, D. El-Maouche2, B. R. Stolze1, R. Muniyappa3, S. J. Soldin4.

1
National Institutes of Health ( Clinical Center), Bethesda, MD, 2National
Institutes of Dental and Craniofacial Research, National Institutes of
Health, Bethesda, MD, 3National Institute of Diabetes and Digestive and
Kidney Diseases, National Institutes of Health, Bethesda, MD, 4National
Institutes of Health ( Clinical Center) andDepartment of Medicine,
Georgetown University, Washington, District of Columbia, Bethesda, MD
Background : Danazol is a 17-ethinyl testosterone derivative. Danazol has long been

S54

used in the management of endometriosis, however its reported immunomodulatory

effects such as reducing interleukin-1 and TNF- have led to the use of danazol in
management of immune related conditions such as aplastic anemia. The side effects
associated with danazol are largely due to its androgenic effects. Danazol has been
reported to act as an interference in the immunoassay measurement of various
androgens (including androstenedione), resulting in falsely elevated values for these
hormones. As a consequence measurement of these hormones in patients receiving
danazol is best performed by liquid chromatography tandem mass spectrometry (LCMS/MS). Here we report eight cases of significantly elevated androstendione (AND)
levels following LC-MS/MS measurement in patients receiving danazaol for aplastic
anemia.
Method: LC-MS/MS measurement: AND is measured as part of a steroid hormone
panel. Samples were prepared as per previously published method (Guo T et al.
Simultaneous determination of 12 steroids by isotope dilution liquid chromatographyphotospray ionization tandem mass spectrometry. Clin Chim Acta 2006;372:76-82
and Mendu DR et al Clin Chem 2011 Abstract E-57, pA212). An Agilent 6460
triple quadropole mass spectrometer (Agilent, USA) equipped with an atmospheric
pressure photoionization was used, employing isotope dilution with deuterium labeled
internal standard for each analyte. Quantitation by multiple reaction monitoring
(MRM) was performed in the positive ion mode. The quantifier MRM transition for
androstenedione used was 287.2> 97.1 and the qualifier MRM of 287.2 >109 was
used to confirm. The ratio of response of the two MRMs used, ranged from 5658%. Retention times (RT) for AND were 6.540 -6.556 minutes for patient samples
and 6.529- 6.554 minutes for the accompanying internal standard (IS). The ratio of
androstenedione RT to internal standard RT for each specimen run was 1.00, thus
confirming the identity of the peak.
Results: A total of 8 adult patients (female n=5; male n=3) were identified with
increased AND values at either 6 or 12 month follow- up post danazol initiation.
Baseline AND values for the female patients ranged from 38 - 176 ng/dL ( reference
interval 17- 175) and at 6 or 12 month follow up values increased markedly and
ranged from 8128-33703 ng/dL. For the male patients baseline AND values ranged
from 105- 240 ng/dL (reference interval 25-125) and at follow up values increased to
range from 5609 to 17325 ng/dL. Similar increases were not observed for the other
androgens measured. Of interest two of the three males had elevated LC-MS/MS
estradiol levels above the reference interval on follow-up while all the female patients
had estradiol concentrations that remained within the appropriate estradiol reference
interval during therapy. Whilst on therapy patients responded well and side effects of
therapy were reported to be minimal.
Conclusion: One of the important advantages of MS analysis is the greater specificity
of over immunoassay based testing In the above described cases use of two MRM
transitions (quantifier and qualifier) enabled the laboratory to confirm the presence of
elevated AND and exclude the likelihood of an interference.

A-201
Evaluation of the effect of elevated Fetal Hemoglobin (HbF) on three HbA1c
Assays Methods in Marshfield Clinic system.

A. KHAJURIA, B. Scheibe. Marshfield Clinic, Marshfield, WI

Background: Accurate measurement of HbA1c is crucial for decision making in
diabetic control and diagnosis. Elevated levels of HbF are reported to falsely decrease
the HbA1c results. There are many clinical conditions presenting with elevated HbF
and prevalence of elevated HbF can be as high as 7 to 8% in a diabetic population
and Clinicians may be unaware of potential interference with HbA1c results. At a
glycemic control target of 6.5% the critical difference between two results within a
subject should not exceed ~0.4%. It is therefore, crucial that laboratories are aware, to
what extent HbF interference affects HbA1c results.
Methodology: Following commercial assays were evaluated; TOSOHG8 (HPLC),
Dimension EXL 200 & DCA 2000 (Immunoassay). Two whole blood EDTA patient
pools as Normal (5 to 6%) and Abnormal (7 to 8%) were prepared and incubated with
varying concentration of HbF (5 to 40%) by mixing umbilical cord blood with known
HbF levels (estimated by G8). The effect of HbF interference was then evaluated
relative to control pools. Percent decrease in HbA1c greater than 5% was considered
a significant change.
Results: TOSOH G8 did not show any interference with up to 25% HbF concentration.
Dimension & DCA 2000, however exhibited a dose dependent interference with HbF
(>5%) at 5% HbF concentration.
Conclusions: Elevated HbF can be identified in Ion-exchange HPLC but not in
Immunoassays. Laboratory professionals should make clinicians aware of potential
interference from elevated HbF levels on a particular method that could adversely
affect HbA1c results. Clinicians can then make informed decisions if HbA1c results
appear discrepant related to patient history and glucose hemostasis.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

loss among pre-dialysis and dialysis-dependent CKD patients. We hypothesized that
in CKD high levels of total serum P1NP would correlate with high bone formation
rate (BFR) measured by biopsy.
Methods: In 22 patients (male=8; female=14; mean ageSD 6811 years,) with
CKD stages 2-5D, fasting morning blood was collected within 6-months of double
tetracycline double-labeled transiliac bone biopsy. Estimated glomerular filtration
rate (eGFR) was by MDRD formula. Total serum PINP (mono- and multi-meric) was
measured by Electrochemiluminescence immunoassay (Elecsys 2010 analyzer, Roche
Diagnostics, Indianapolis, IN). Intra- and inter-assay precision were 1.1% and 5.5%
respectively (Reference range = 20-100 ug/L). BFR (m3/ m2/day ) was determined
histomorphometrically in trabecular, endocortical and intra-cortical bone from biopsy
using American Society for Bone and Mineral Research (ASBMR) criteria. Data are
presented as meanSD. P1NP and BFR were log transformed prior to analyses and
relationships were determined by Pearson correlations.

A-202
VALIDITY OF A SEMI-QUANTITATIVE METHOD FOR
MICROALBUMINURIA SCREENING IN DIABETES MELLITUS

J. D. Santotoribio, C. Caavate-Solano, A. Garca de la Torre, F. ArceMatute, S. Prez-Ramos. Puerto Real University Hospital, Cdiz, Spain
INTRODUCTION Urine Albumin is an early marker of chronic kidney disease
(CKD) in diabetic patients. The excretion of recent urine albumin greater than 20
mg/L (microalbuminuria) is considered as a predictor of diabetic CKD. The aim of this
study is to analyse the validity of a semi-quantitative method for microalbuminuria
screening in diabetes mellitus.
METHODS Recent urine microalbuminuria of diabetic patients were determined
by two methods: 1. Semi-quantitative: Colorimetric method using the strip H13
in DIRUY H-800 PLUS (RAL). The content of microalbuminuria is inversely
proportional to the quantity of the color of the reagent pad. The instrument measures
the color change of the reagent pad on a scale of 0 to 4000.
2. Quantitative: microalbuminuria was measured by immunoturbidity in COBAS
C311 (ROCHE DIAGNOSTIC). Patients were classified into two groups according
to the quantification of microalbuminuria: positive (microalbuminuria > 20 mg/L) and
negative (microalbuminuria < 20 mg/L). Statistical analysis was determined using
receiver operating characteristic (ROC) techniques by analysing the area under the
ROC curve (AUC).
RESULTS We analyzed 469 diabetic patients between 27 and 85 y.o. (mean age =
56.3), 82 patients (17.5%) had a positive microalbuminuria and 387 patients (82.5%)
were negative. The AUC was 0.985 (p < 0.0001). With a cut-off color scale less
than 1305 determined by the test strip, we obtained a sensitivity of 100% and a
specificity of 86.3%. With these results, it would only be necessary the quantification
by immunoturbidimetry the samples with a value lower than 1305. In this case it
would not have been necessary to measure microalbuminuria in 334 samples of the
469 studied, getting a saving of 71%.
CONCLUSIONS The semi-quantitative method by test strip, can be used as screening
for microalbuminuria in diabetic patients with a sensitivity of 100%. Microalbuminuria
would only be measured in samples with positive test strip.

A-203
Procollagen of type-1 N-terminal propeptide levels by Elecsys assay correlates
with bone formation rate in Chronic Kidney Disease

S. Cremers, D. Dempster, H. Zhou, E. Dworakowski, M. Kamanda-Kosseh,

T. Nickolas. Columbia University Medical Center, New York City, NY
Background: Renal osteodystrophy is a common metabolic bone disorder due to
chronic kidney disease (CKD) that is associated with high risk of bone loss, fracture
and death. It is characterized principally by a spectrum of abnormal bone turnover
ranging from extremely low (adynamic bone disease) to high (osteitis fibrosa cystica).
Bone turnover assessment is an absolute required for treatment. However, bone
turnover is currently determined by tetracycline double-labeled transiliac crest bone
biopsy with histomorphometry, an invasive, and not widely available procedure. Noninvasive turnover determination in CKD is controversial because most bone turnover
markers (BTMs) are renally cleared; thus, in CKD they may be non-specifically
elevated and not reflect accurately remodeling activity. Procollagen of type-1
N-terminal propeptide (P1NP) is a biochemical marker of bone formation, it circulates
in mono- and multi-meric forms and multimeric P1NP is cleared non-renally. Our
group reported that higher circulating total levels of P1NP predicted incident bone

Results: Five patients were on hemodialysis and mean eGFR in pre-dialysis patients
was 3617 mL/min. Mean BFR at trabecular, endocortical and intracortical regions
were 0.0180.031, 0.0190.035 and 0.0300.038 respectively and there were no
significant differences in BFR between pre-dialysis and hemodialysis patients.
Mean PINP for the total, pre-dialysis and hemodialysis cohorts were 332585ug/L,
9851ug/L and 1125878ug/L respectively. P1NP levels were significantly greater
in hemodialysis compared to pre-dialysis (p=0.004) and there were no significant
relationships between P1NP and eGFR among pre-dialysis patients. There were
significant, moderate and direct associations between PINP and BFR in the three
envelopes (R2 0.41, 0.34 and 0.34, all p<0.05 for trabecular, endocortical and
intracortical bone, respectively).
Conclusion: These data suggest that measurement of total serum PINP by the Elecsys
assay correlates well with BFR in CKD. Larger studies are needed in CKD populations
to validate these data, and to determine whether P1NP predicts future fracture and can
be used to guide treatment to protect against bone loss and fracture.

A-204
The contribution of angiotensin II-dependent oxidative stress to megalin
expression in the renal cortex during the normoalbuminuric stage of diabetes
mellitus in the rat.

Y. Kurosaki1, Y. Yamada2, P. K. Carmines3, T. Tsukushi4, A. Imoto1, T.

Ichikawa2, M. Yokoba1, S. Obata4, H. Ikenaga5, Y. Aoki6, T. Takenaka7, T.
Ichikawa1, M. Katagiri1, N. Ishii1. 1Kitasato University School of Allied
Health Sciences, Sagamihara, Japan, 2Kitasato University Graduate
School of Medical Sciences, Sagamihara, Japan, 3University of Nebraska
College of Medicine, Omaha, NE, 4Kitasato University Hospital,
Sagamihara, Japan, 5Ikenaga Clinic, Ohtawara, Japan, 6Kanagawa Health
Service Association, Yokohama, Japan, 7International University of Health
and Welfare, Sanno Hospital, Minato-ku, Japan
Background: Renal albuminuria can result from impaired albumin handling by
the glomerulus or the renal tubules. Megalin plays a critical role in proximal tubular
albumin reabsorption, and altered megalin expression or function can contribute to
renal tubular albuminuria. Elevated glucose reabsorption during acute hyperglycemia
results in impaired tubular handling of protein and proximal tubular damage (Enzyme
Protein 8:243-50, 1994-1995). Moreover, renal albuminuria during diabetes mellitus
(DM) is associated with decreased megalin expression in the proximal tubule
(Diabetes 56:380-388, 2007). During the early stage of diabetes mellitus (DM) in
rats, prior to development of albuminuria, the renal cortex exhibits oxidative stress
that can be suppressed by renin-angiotensin system (RAS) inhibition (Clin Sci
24:543-52, 2013); however, it is not known whether these events arising during the
normoalbumuric stage of DM influence megalin expression and protein excretion.
Objective: The goal of this study was to evaluate impact of oxidative stress
suppression, achieved by angiotensin II receptor blocker (ARB) treatment, on
proximal tubular megalin expression during the normoalbuminuric stage of DM in
rats.
Methods: Four groups of rats were examined: 1) STZ group (n=5): rats studied 2
wks after induction of DM by streptozotocin injection (STZ, 65 mg/kg,i.p.), 2) Sham
group (n=5): rats receiving the STZ vehicle, 3) STZ+TLM group (n=4): STZ rats
treated with telmisartan (TLM, an ARB; 10 mg/kg/day in chow for 2 wks), and 4)
Sham+TLM group (n=4): TLM-treated Sham rats. In each rat, blood glucose, blood
pressure, glomerular filtration rate (GFR) were measured, as well as urinary albumin
levels and activity of N-acetyl--D-glucosaminidase (NAG; a proximal tubulederived enzyme) in urine. Further, we measured renal cortical 3-nitrotyrosine (3-NT)
production (oxidative stress marker) by HPLC and megalin expression (Western blot
analysis).

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S55

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

Results: Blood glucose levels were higher in STZ and STZ+TLM groups than in
Sham and Sham+TLM groups (P<0.05), confirming development of DM; however,
blood pressure and urine albumin level did not differ among groups. GFR and urinary
NAG activity (an index of proximal tubule damage) were increased in the STZ group
compared with Sham (each P<0.05), and these changes were prevented by TLMtreatment (each P<0.05 STZ vs. STZ+TLM). Renal cortical 3-NT production in the
STZ group was 70% greater than in the Sham group (Sham, 34.03.0 pmol/mg protein;
STZ, 58.42.1 pmol/mg protein; P<0.05 vs. Sham); however, this phenomenon was
completely suppressed by TLM treatment (STZ+TLM, 35.53.5 pmol/mg protein;
P<0.05 vs. STZ). Renal cortical megalin expression was elevated in the STZ group
(30389% of Sham; P<0.05); however, the enhanced expression of megalin in the
STZ group was not evident in the STZ+TLM group (14961% of Sham; P<0.05 STZ
vs. STZ+TLM).
Conclusions: These observations demonstrate that increased renal cortical megalin
expression accompanies oxidative stress during the early stage of DM, prior to
development of albuminuria. The ability of ARB treatment to prevent the DM-induced
elevation of megalin implicates the renin-angiotensin system in this phenomenon,
perhaps through an oxidative stress-dependent mechanism.

A-205
Analytical validation of the new Roche Thyroglobulin II
eletrochemiluminescent immunoassay.

M. C. Batista, P. R. S. Ferreira, C. E. S. Ferreira, A. C. L. Faulhaber, C. L.

P. Mangueira. Albert Einstein Hospital, So Paulo - SP, Brazil
Background: Thyroglobulin (Tg) is a serum marker of thyroid cancer. It is
thus critical to measure it accurately in the low range of 0.1 to 1.0 ng/mL level.
Unfortunately, this is not possible with most commercial kits, including the traditional
Roche eletrochemiluminescent Tg assay. One exception is the Beckman-Coulter
Tg chemiluminescent assay (Brea, USA), which has a limit of quantitation (LOQ)
of 0.1 ng/mL and, for this reason, is considered by many the reference method.
In this study, we evaluated the analytical specifications of the new Roche Tg II
eletrochemiluminescent immunoassay (Mannheim, Germany).
.Methods: All serum samples selected for this protocol were routine clinical
specimens previously assayed for Tg in our lab using the Beckman-Coulter Access
II assay. They were all rerun within one week in Roche E170 Modular Analytics
using the new Tg II assay. This is an eletrochemiluminescent immunometric assay
using 2 monoclonal antibodies to form a sandwich complex with Tg. Three samples
with low, normal or high Tg levels were run in Modular E170, all in duplicate in the
morning and afternoon for 5 days, and then used to calculate intra and interassay
variation and LOQ (lowest concentration with CV < 20%). Linearity and analytical
measuring range (AMR) were evaluated in Modular E170 by mixing a low and
high Tg sample in different proportions. Correlation studies were performed in 46
samples assayed in both Access II and Modular E170. Results were analyzed with EP
Evaluator 11.0 software (Data Innovations, South Burlington, USA).
Results: Intra and interassay CV were 2.3% and 5.1% at 0.28 ng/mL; 1.5% and 2.7%
at 2.2 ng/mL; and 0.7% and 2.3% at 45.9 ng/mL, respectively. LOD was set at <
0.2 ng/mL, based on the low CV found at this level. The assay was linear at 0.21493.3 ng/mL, which was defined as the AMR. Modular E170 results were compared
with Access II using Deming regression over a range of 0.20 to 34.4 ng/mL. The
correlation coefficient was 0.96, average error index 1.50 (range -3.71 to 6.41),
slope 1.47, intercept -0.46 and standard error of estimate 3.17. The methods were
not considered equivalent within allowable total error of 21.9% (biologic variation
database, Ricos C 2012).
Conclusion: The new Roche Tg II assay exhibited an excellent precision and linearity
down to 0.2 ng/mL. Although results correlated well with the Beckman-Coulter Tg
assay, they were not considered equivalent to this reference method.

A-206
Prediabetic Importance of Serum Zinc Alpha Glycoprotein and Ghrelin Levels
in Subjects Classified According to Oral Glucose Loading Test and Fasting
Glucose Levels

E. Kurtulu, D. Konukoglu. Istanbul University, Cerrahpasa Medical

Faculty,Biochemistry Department, Istanbul, Turkey
Background: Increasing evidence suggests that the postprandial state and fasting
hyperglycemia are a contributing factor to the development of Diabetes Mellitus.
It has been recently suggested that an adipokine, zinc-2-glycoprotein (ZAG), may
also have a protective role in the prevention of obesity and its associated disorders.

S56

ZAG has been proposed to play a role in the pathogenesis of insulin resistance and
suspected to be related with Type 2 Diabetes. Ghrelin a peptide hormone secreted
mainly by the stomach, increases appetite and food intake. It has been suggested
that Ghrelin hormone plays role in insulin secretion and glucose metabolism. In
the present study we determined serum ZAG and Ghrelin levels, and evaluated
whether the relationship between serum ZAG and ghrelin levels in prediabetic stages.
Methods: Subjects were categorized according to WHO criteria as Controls (n:23,
women:13, men:10, mean ages:55,6 7,7 years ) , Impaired Fasting Glucose (IFG;
women:29, men:23, mean ages: 55,1 7,0 years), Impaired Glucose Tolerance (IGT;
women:26, men: 20,mean ages: 59,1 8,4 years) and Diabetic Glucose Tolerance (
DGT; women:15,men: 15, mean age 59,9 11,1 years) in our study. There was no any
difference in Body Mass Index and plasma lipids levels (total cholesterol, triglyceride,
HDL and LDL -cholesterol) between groups. Subjects patients did not use any
medication or vitamin pills. Baseline serum ZAG and Ghrelin levels were determined
by ELISA. Serum inslin levels were determined by chemiluminesans assay. The
homeostasis model assessment of insulin resistance (HOMA-IR) was calculated.
Results: Serum ZAG levels in Control group were found to be significantly higher
than in DGT, IGT and IFG groups (p<0,005, p<0,001 and p<0,001 respectively). IFG
group have significantly lower serum ZAG levels than both DGT and IGT groups
(p<0,005 and p<0,001). Serum Ghrelin levels in IGT was significantly higher than in
IFG, DGT and Controls (p<0,001, p<0,001 and p<0,001 respectively). Subjects with
IFG have significantly higher serum ghrelin levels than Controls (p<0, 05). There was
a significant negative correlation between ZAG and HOMA-IR (r:-0,321, p<0,001),
as well as Ghrelin levels.(r=-0,530,p< 0,001). A positive correlation were obtained
from serum ZAG and 2-hours post challenged plasma glucose levels (r=0,187, p<0,
05). Conclusion: The results of our study suggest that ZAG and Ghrelin are involved
prediabetic stages and their levels can be important in the regulation of glucose
metabolism. ZAG and ghrelin was found to be effective in the opposite direction.
Also, we thought that basal ZAG levels can be a predictive marker for the 2-hours
post challenged glucose levels. The present workwas supported by the Research Fund
of Istanbul University. Project No. 29822

A-207
Angiotensin II-dependent oxidative stress and increased hypoxia-inducible
factor-1 expression in the renal cortex during the normoalbuminuric stage of
diabetic mellitus in the rat.

Y. Yamada1, Y. Kurosaki2, P. K. Carmines3, T. Tsukushi4, A. Imoto2, T.

Ichikawa1, M. Yokoba2, H. Ikenaga5, Y. Aoki6, T. Takenaka7, T. Ichikawa2,
M. Katagiri2, N. Ishii2. 1Kitasato University Graduate School of Medical
Sciences, Sagamihara, Japan, 2Kitasato University School of Allied
Health Sciences, Sagamihara, Japan, 3University of Nebraska College of
Medicine, Omaha, NE, 4Kitasato University Hospital, Sagamihara, Japan,
5
Ikenaga Clinic, Otawara, Japan, 6Kanagawa Health Service Association,
Yokohama, Japan, 7International University of Health and Welfare, Sanno
Hospital, Minato-ku, Japan
Background: Hypoxia-inducible factor-1 (HIF-1) is composed of and subunits,
with HIF-1 considered to be a master regulator of the hypoxic response. HIF-1
and oxidative stress associate with the progression of diabetic nephropathy (Clin Exp
Pharmacol Physiol 33:997-1001, 2006). We previously reported that renal cortical
oxidant production is already increased during the early stage of diabetes mellitus
(DM), prior to development of albuminuria, and that renin-angiotensin system (RAS)
inhibition suppresses development of renal oxidative stress under these conditions
(Clin Sci 124:543-52,2013). However, the relationship between oxidative stress and
the hypoxic response in the kidney during normoalbuminuric stage of DM has not
been established.
Objective: The goal of this study was to determine the effect of oxidative stress
suppression by RAS inhibition (treatment with an angiotensin II receptor blocker;
ARB) on HIF-1 expression in the renal cortex during the normoalbuminuric stage
of DM.
Methods: Four groups of rats were examined: 1) STZ group (n=5): rats studied
2 wks after induction of DM by streptozotocin injection (STZ, 65 mg/kg, i.p.), 2)
Sham group (n=5): rats receiving the STZ vehicle, 3) STZ+TLM group (n=5): STZ
rats treated with telmisartan (TLM, an ARB; 10 mg/kg/day in chow for 2 wks), and
4) Sham+TLM group (n=4): TLM-treated Sham rats. In each rat, blood glucose,
blood pressure and glomerular filtration rate (GFR) were measured. Production of
3-nitrotyrosine (3-NT; an oxidative stress marker) in the renal cortex was measured by
HPLC, and HIF-1 expression was quantified by the Western blot analysis.
Results: Blood glucose levels were significantly higher in STZ rats than in Sham
rats, and was unaffected by TLM (similar to other treatments that suppress the RAS).
Blood pressure did not differ among groups. Compared with the Sham group, GFR

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

was increased in the STZ group (P<0.05), and this was prevented by TLM treatment
(P<0.05 STZ vs. STZ+TLM). Renal cortical 3-NT production in the STZ group
was 70% greater than in the Sham group (Sham, 35.23.4 pmol/mg protein; STZ,
59.61.4 pmol/mg protein; P<0.05 vs. Sham); however, this phenomenon was
completely suppressed by TLM treatment (STZ+TLM, 37.33.5 pmol/mg protein;
P<0.05 vs. STZ). The STZ group also displayed an increase in renal cortical HIF-1
expression (25726% of Sham; P<0.05); however, the DM-induced increase in HIF1 expression was not evident in the STZ+TLM group (15711% of Sham; P<0.05
STZ vs. STZ+TLM).

Results:AMR studies were carried out according to CLSI Guideline EP6-A.Results

are demonstrated in the table below.

Conclusions: An increase in renal HIF-1 expression accompanies oxidative stress

during the normoalbuminuric stage of DM in the rat, and both of these phenomena
are prevented by ARB. These observations indicate that the hypoxic response arises
in the renal cortex early during the course of DM, and that this occurs either directly
or indirectly under the influence of the RAS, possibly secondary to the presence of
oxidative stress.

Conclusion:The thyroid trials tested in this study presented Assay Linearity Range
as established by the manufacturer, within CLSI Guideline EP6-A and Total Error
Laboratorys target.Thus, tests were approved by the Quality Control Management
in the laboratory.

Discussion: The tests showed satisfactory linearity results with different samples. In
this study, the sample pool was not used, because all the results from this dilution test
presented high percentage of recovery, above the manufacturers reference, due to
a matrix effect. Because of that, we established the utilization of different samples,
respecting the expected concentrations of the sample pool,if diluted samples were
prepared.The coefficients of second and third degree regression are statistically equal
to zero, at 5% of the significance level.

Table 1 Assay
TSH3UL
T4
FT4
T3
FT3

A-208
Evaluation of TSH Levels in Rio de Janeiro State/Brazil Neonatal Screening.

M. F. M. C. Pinheiro, C. M. M. Oliveira, S. V. L. Argolo, O. F. Silva Filho.

DASA, Rio de Janeiro, Brazil
Background:Congenital Hypothyroidism (CH) is the most common congenital
endocrine disorder and it is a leading preventable mental retardation. Its incidence
ranges from 1:2000 to 1:4000 live births in iodine-sufficient countries. In Brazil,
the screening for CH is mandatory by law and usually done by TSH determination
on dried blood spot on filter paper samples collected by heel puncture. Diagnostic
confirmation is required dosing TSH and free T4 in serum. The objective of this study
was to evaluate the distribution of TSH levels in newborns blood samples from Rio
de Janeiro state and the frequency of CH confirmed patients. We also compared our
results to those described by the manufacturer.
Methods:We evaluated 18,609 dried blood spots on filter paper samples for TSH of
newborns from Rio de Janeiro state over one year period (2013). The range of age
was 3 to 30 days of life. We used an automatic immunofluorimetric system, GSP
Neonatal hTSH kit (Wallac Oy, Turku, Finland). The cutoff value for TSH was 10.0
mUI/L, children with levels above these limits were recalled for confirmation with
serum TSH and FT4.
Results:The most of TSH levels, 85.3%, were less than 2.0 mUI/L. The percentiles 95
and 99 were 3.0 and 4.7 mUI/L, respectively. Comparing to the percentiles described
by the manufacturer, 7.9 and 10.7mUI/L, our results were much lower. We found 9
samples (0.05%) above the TSH cutoff. All these patients underwent measurement
of serum TSH and FT4. Congenital Hypothyroidism was confirmed in 7 (1:2658)
patients. Their initial TSH filter paper level ranged from 12.8 to 279.0 mUI/L, with
mean 115.0 mUI/L and median 42.8 mUI/L.
Conclusion:Our data are indicative that, in this Brazilian population, the distribution
of TSH levels in newborns blood filter samples were lower to those presented by
the manufacturer, based on European individuals studies. This reinforces the need
for each laboratory to evaluate the TSH levels in its specific population. Using 10.0
mUI/L as the TSH cutoff we found, in this Brazilian population, a CH incidence
(1:2658) similar to others iodine-sufficient regions.

A-209
Linearity study of thyroid assays assuring the quality control requirements

D. Waltrick, A. L. Camilo, C. F. A. Pereira. DASA, So Paulo, Brazil

Background:The ability to quantitate circulating levels of thyroid hormones is
important in evaluating thyroid function. It is especially useful in the differential
diagnosis of primary (thyroid) from secondary (pituitary) and tertiary (hypothalamus)
hypothyroidism.To certify released results accuracy, it is important to ensure
the linearity of the tests.A Brazilian laboratory implemented an Easy Linearity
Curve(ELC) tool to monitor the immune-hormone analytical systems, establishing
a self-inspection program to verify the efficiency and accuracy of procedures and
results. The use of a tool that checks assay linearity provides additional safety and
reliability of the results. This study aims to use the statistic tool to monitor the AMR
of TSH,T4, FT4, T3 and FT3.
Materials and methods: We selected samples from the laboratory routine, with
concentrations within assay linearity range for each test. Samples were tested in Advia
Centaur XP (Siemens Healthcare Diagnostics) using a chemiluminescent method
and analyzed with the ELC tool.

CV Lab
Obtained
max Target
CV %
%
0.008 - 150 uIU/mL 4.92
9.70
0.3 - 30 ug/dL
2.38
5.10
0.1 12 ug/dL
9.68
5.26
10 800 ng/dL
6.15
5.95
0.2 2 ug/dL
5.72
4.00
Assay Linearity
Range

Obtained TE max
Linear Regression R2
TE %
Target %
6.93
8.98
11.79
13.10
14.00

23.70
10.55
15.00
20.00
30.00

0.990x + (0.312)
1.003x + (-0.017)
0.975x + (0.057)
1.028x + (-9.808)
1.003x+ (0.031)

0.99973
0.99983
0.99684
0.99778
0.99074

A-210
A REVIEW OF 312 GROWTH HORMONE STIMULATION TESTS
PERFORMED AT A REFERENCE LABORATORY (DASA-RJ) IN RIO DE
JANEIRO - BRAZIL

Y. Schrank, M. C. Freire, R. Fontes, D. M. V. Gomes, O. F. Silva. DASA laboratory medicine, rio de janeiro, Brazil
Background: There is still no consensus on the stimulation test considered the gold
standard for the diagnosis of GH deficiency. The optimal criteria for a definitive test of
growth hormone function are not met by any single stimulus. Lack of standardization
of GH response to each type of stimulus, poor reproducibility and lack of correlation
between the response to the test and growth are some of the various limitations of
these tests. The aim of our study was to examine the main tests of GH stimulus applied
in our environment so as the response to these tests.
Materials and Methods: We did a retrospective review of 312 patients submitted to
GH stimulation test in a period of 12 months. A test was considered responsive when
peak GH >5ng/mL.
Results: GH stimulus with Clonidine was the most requested test .The mean age of
our patients was 10,2 years, and male:female ratio was 2,4:1. Most of patients were
therefore male. Interestingly, however, among patients submitted do the Glucagon
simulation test, the majority were female. The greatest GH peak was seen with
Glucagon stimulus. No significant complications were observed with the applied tests.
STIMULUS
CLONIDINE
INSULIN
GLUCAGON
PIRIDOSTIGMIN

N (%)
202
(63,7%)
85
(26,8%)
28 (8,8%)
2 (0,7%)

Mean
AGE
(years)

GENDER
(M/F ratio)

RESPON- mean GH
SIVE
peak
(%)
(ng/mL)

GH peak
(time)

10,1

2,9

73

9,5

60 and 90

11,6

2,1

50

7,0

60

6,5
11,5

1,2
1

68
0

10,7
0,05

120
-

Conclusions: The high number of tested patients so as the high rate of GH response
to Clonidine suggests that this is the preferred screening test by test prescribers. On
the other hand, the Insulin stimulation seems to be preferably reserved to confirm
the diagnosis of GH deficiency, since the number of patients submitted to these test
was significantly smaller and the percentage of responders was only 50%. The use of
Insulin as a second-line test in the investigation of GH deficiency is easily explained
by the justified fear of potential complications associated with this test. Moreover,
Glucagon stimulation is the preferred screening test in children under 6 years.

A-211
Strategy to Improve Diabetes diagnosis in Primary Care: Preliminary Results
and Evaluation.

M. Salinas1, M. Lopez-Garrigos1, A. Asencio2, M. Gutierrez1, J. Lugo1, C.

Leiva-Salinas3. 1Hospital Universitario San Juan, San Juan, Spain, 2CAP
Mutxamel, Mutxamel, Spain, 3University of Virginia, Charlottesville, VA
BACKGROUND: With the introduction of HbA1c as a tool to diagnose diabetes,
a strategy was designed, established and evaluated in consensus with general

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S57

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

practitioners (GPs) to detect diabetic patients through an opportunistic study to
improve HbA1c requesting, and to ascertain if previous HbA1c demand was
appropriate to detect diabetes.
METHODS: The laboratory decided to approach the GPs to design a strategy
that would improve the diabetes diagnosis efficiency: Laboratory Information
System (LIS) automatically would add HbA1c to every sample from primary care
patients older than 45 years, without an HbA1c request in the previous three years
and glucose results between 100 and 126 mg/dl. If results were above 6.4% , LIS
recommended a second request in 3-6 month period. In a last meeting the strategy was
approved, established March 1st 2013 and evaluated after a 6 month period. HbA1c
was measured using a Variant II Turbo Hemoglobin Testing System (Bio-Rad
Laboratories, Madrid, Spain).
RESULTS: 412 HbA1c were added automatically, causing 39 HbA1c values above
6.4%. After medical record review, 6 HbA1c results above 6.4% were justified. In
one case it was impossible to contact because a change of residency. 3 patients were
diagnosed as diabetic without the need of a second request. To eleven patients a
second HbA1c was requested, being 4 diagnosed as diabetic patients (HbA1c >6.4%)
and in seven patients the second HbA1c result did not confirm diabetes. Despite their
abnormal HbA1c results, until now, to 17 patients has not been requested a second
HbA1c to confirm/discard the illness. Results are showed in figure.The cost of adding
the 412 HbA1c was 535.6 US dollars. At this moment, each case of the seven diabetes
diagnosed represented a cost of 76.5 US dollars.
CONCLUSION: Our proposed opportunistic screening to discover diabetes seems
cost-effective. Hba1c was previously under requested.

analytes were 101-107%, 99-106% and 92-104% for testosterone, androstenedione

and DHEA, respectively. Linearity was shown in dilution series (mean R2 was >0.999
for all analytes). This method tested negative for interference from DHEA-sulphate,
estrone, 17-estradiol, androsterone, 17-hydroxy progesterone, dihydrotestosterone,
epi-testosterone, cortisone and cortisol and did not show ion suppression. The method
was shown to be suitable for serum as well as EDTA and heparin plasma.
The present testosterone method compared well (y = 1.000 x + 0.035 nmol/L; r =
0,9982) to another ID-LC-MS/MS method for testosterone concordant with a
published reference method (Bui et al. 2013). In the near future, the present method
will also be compared to another LC-MS/MS method for androstenedione and DHEA.
Conclusion: We developed a sensitive and accurate method to measure serum
testosterone, androstenedione and DHEA levels in one run.

A-213
Glucagon Quantification: Comparison of Radioimmunoassay and Sandwich
ELISA methods

M. Raines, C. King, S. Christensen, P. Amin, D. Schranz. Pacific

Biomarkers, Seattle, WA
Background: Accurate and robust measurement of glucagon is important in
understanding glucagons role in glucose metabolism and homeostasis as well as
its role in the pathology of type 2 Diabetes and other metabolic diseases. It is also
commonly used in clinical studies as a surrogate marker of drug efficacy. Competitive
RIA methods, like the ALPCO Glucagon RIA (A), have been the gold standard for
measuring glucagon but can be limiting due to the short shelf-life, long ordering lead
times, long assay times, and large sample volume required (1 mL). The objective of
this study is to evaluate the performance of the R&D Systems Glucagon Quantikine
ELISA (R) and the Mercodia Glucagon ELISA (M) relative to the FDA cleared
ALPCO Glucagon RIA.
Methods: All three assays were validated for precision, linearity, recovery, sensitivity,
and normal glucagon ranges using fasting plasma samples collected in either K2EDTA
or P800 tubes, frozen and stored at -70 C. A set of fasting and nonfasting P800 plasma
samples were also used to directly compare glucagon results of all three assays.

A-212
Measurement of serum testosterone, androstenedione and
dehydroepiandrosterone (DHEA) levels using Isotope-Dilution LiquidChromatography Tandem Mass Spectrometry (ID-LC-MS/MS)

R. Bttler1, F. Martens1, M. Ackermans2, M. Blankenstein1, A. Heijboer1.

1
VU university medical center, Amsterdam, Netherlands, 2Academic
Medical Center, Amsterdam, Netherlands
Background: The adrenal and gonadal androgens testosterone, androstenedione and
dehydroepiandro-sterone (DHEA) play an important role in sexual development and
fertility as well as in several other processes.
Methods: We developed a method to assess serum testosterone, androstenedione and
DHEA levels in one run using Isotope-Dilution Liquid-Chromatography Tandem Mass
Spectrometry (ID-LC-MS/MS). Sample preparation consisted of addition of internal
standards (13C3-testosterone, 13C3- androstenedione and 2H6-DHEA) and a liquid-liquid
extraction using hexane-ether. The samples were analyzed on an Acquity 2D UPLC
system (Waters), equipped with a C4 column (Waters) and a Kinetex Fluorophenyl
column (Phenomenex), and a Xevo TQ-S tandem mass spectrometer (Waters). The
three analytes were baseline separated in a total run time of 9 minutes. The calibration
curves ranged from 0.10 to 26 nmol/L for testosterone and androstenedione, and from
0.96 to 78 nmol/L for DHEA.
Results: The intra-assay CVs were <4.0%, <4.6% and <7.0% for testosterone,
androstenedione and DHEA, respectively. The inter-assay CVs were <6% for
testosterone and <8% for androstenedione and DHEA. At the lower concentrations
inter-assay CVs were 15%, 7.0% and 9.3%, for testosterone (0.08 nM),
androstenedione (0.47 nM) and DHEA (1.18 nM), respectively. Recoveries of spiked

S58

Results: Precision for all three methods was acceptable with the intra-assay precision
being less than 5% for all three assays and inter-assay precision being 3.8 - 11.9% (A),
4.9 - 9.2% (R), and 4.3-8.1% (M). Dilutional linearity was acceptable up to 40-fold
(A), 16-fold (R), and 16,000-fold (M) dilutions for the three assays. The Mercodia
ELISA was the most sensitive with a lower limit of quantitation (where the % CV
is equal to 20%) of 1.5 pmol/L (M) versus 8.6 (A) and 9 pmol/L (R) for the two
other kits. The most striking difference in the three assays was in the glucagon values
observed in apparently healthy donor samples. The mean glucagon values for normal
samples analyzed using the ALPCO RIA method were much higher (39.3 pmol/L)
than the R&D systems (28.8 pmol/L) and Mercodia (9.1 pmol/L). Although the R&D
systems ELISA had mean normal glucagon values that were more in line with those
obtained for the ALPCO RIA, Deming Regression Analysis using the same set of
P800 plasma samples yielded a correlation coefficient of 0.6445 and slope of 2.174,
while the Mercodia ELISA had a correlation of 0.9093 and a slope of 0.606 when
compared with the ALPCO RIA. The majority of the twenty samples analyzed with
the R&D systems ELISA yielded glucagon values 30-50% lower than the ALPCO
RIA method, but there were three samples that had glucagon values that were higher
in the R&D Systems ELISA than the ALPCO RIA which resulted in poor correlation.
The Mercodia ELISA glucagon values were consistently lower than the ALPCO RIA
values with greater biases observed for samples less than 10.0 pmol/L. Potential crossreactivity with other glucagon-related molecules is speculated and currently being
investigated as it may account for some of the differences observed between the three
assays.
Conclusion: The Mercodia Glucagon ELISA may be a suitable alternative to the
ALPCO Glucagon RIA method especially when sample volumes are limiting and
better sensitivity is required.

A-214
Associations of Leukocyte Telomere Length with Cardiometabolic Risk Factors
and Circulating Biomarkers of Inflammation and Oxidative Stress

R. M. AlKhaldi, O. Mojiminiyi, F. Al Mulla, N. Abdella. Kuwait University,

Kuwait, Kuwait
Background: Telomeres are TTAGGG sequences at the end of chromosomes
necessary for and chromosomal integrity which upon reaching critical length,cell
become senescent or otherwise dysfunctional. However, telomerase a reverse

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

transcriptase enzyme prevents telomere exhaustion and chromosomal instability.

Telomeres and telomerase were linked to aging & associated diseases namely obesity,
diabetes type 2 [T2DM] and cancer. We hypothesize that shortened telomere length
would be associated with cardio-metabolic risk factors, and that this relationship
might be mediated by obesity related metabolic changes.
Methods: Indices of obesity (Body Mass Index [BMI], Waist Circumference [WC],
Waist to height Ratio [WHtR]), glycated Hemoglobin [HbA1c%], lipid profile, fasting
glucose, serum human Telomerase Reverse Transcriptase [hTERT], total adiponectin,
Insulin, Myeloperoxidase [MPO], Malondialdehyde [MDA], Total Oxidative stress
status [TOS] and Leukocyte Telomere Length [LTL] were measured in 225 T2DM
patients and 245 age and sex matched controls. Insulin resistance [IR] was estimated
using Homeostasis Model Assessment [HOMA] calculator.
Results: T2DM patients had significantly (p<0.0001) lower LTL compared to controls
[(MeanSD:2.10.2) vs. (MeanSD:4.10.1)] respectively. Levels of hTERT were
higher in controls compared to T2DM patients [(MeanSD: 32.98.9 ng/mL) vs.
(MeanSD: 21.44.7 ng/mL)]. Spearmans rank correlation coefficients showed
that LTL correlated negatively with age [r = -0.2, p=0.009], BMI [r = -0.3, p=0.006],
WC [r = -0.3, p<0.0001], and Insulin [r = -0.2, p=0.03]. The significance of these
correlations disappeared after adjusting BMI but not age and/or sex. Additionally,
LTL correlated negatively and strongly with WHtR [r = -0.5, p=0.004], and HbA1c%
[r = -0.6, p=0.003]. These significant correlations were not affected by BMI, age or
sex. Multivariate regression analysis showed that LTL negatively associated with BMI
[=-0.7, p=0.005], WC [=-5.7, p=0.004], HOMA-IR [=-1.1, p=0.003], MPO [=0.6, p<0.0001], MDA [=-0.1, p=0.04], TOS= [=-2.2, p<0.0001]. hTERT showed
similar trends in relation to BMI [=-0.2, p=0.004], WC [=-1.4, p=0.006], HOMA-IR
[=-1.3, p=0.007], MPO [=-0.6, p<0.0001], MDA [=-0.42, p=0.002], TOS[=-0.3,
p=0.007]. On the other hand, LTL and hTERT were associated significantly and
positively associated with adiponectin [=3.1, p=0.02; =1.5, p=0.003] respectively.
Using binary logistic regression analysis, higher BMI was associated with higher risk
of telomeres shortening [OR=2.4, p=0.008]. Higher WC and WHtR were associated
with higher risk of telomere shortening [(OR=2.4, p=0.001); (OR=1.9, p=0.002)]
respectively. Other obesity related factors such as IR, hyper-insulinemia and hypertriglyridemia [(OR=7.7, p<0.0001), (OR=1.2, p<0.0001), (OR=1.3, p=0.01)] were
also associated with higher risk of short telomeres. Higher levels of adiponectin were
associated with lower risk of telomere shortening [OR=0.7, p=0.004]. Additionally,
shorter telomere length were associated significantly with higher risk of T2DM
[OR=7.5, p=0.003]. Higher hTERT levels though were associated with lower risk of
T2DM [OR=0.8, p=0.008].
Conclusions: Our results demonstrate the link between telomere biology,
cardiometabolic risk factors, and T2DM in the Kuwaiti population which has not
been studied before. Metabolic changes such as the dys-regulation of adipokines
(such as adiponectin), dys-lipidemia, hyper-insulinemia, IR and obesity associated
inflammatory process, could play a role in mediating telomere shortening. Since,
obesity and T2DM are increasing at epidemic pace in Kuwait; telomere attrition &
telomerase levels could be potential cardio-metabolic risk markers of obesity and
T2DM.

A-215
Associations of Common TERC Single Nucleotide Polymorphisms with
Telomere Length, Human Telomerase Reverse Transcriptase and Obesity
Related Factors

R. M. AlKhaldi1, O. Mojiminiyi1, F. Al Mulla1, N. Abdella2. 1Pathology

Department, College of Medicine, Kuwait University, Kuwait, Kuwait,
2
Medicine Department, College of Medicine, Kuwait University, Kuwait,
Kuwait
Background: Quantitative trait locus studies have mapped putative loci that
probably be involved in the regulation of leukocyte telomere length [LTL] to human
chromosomes 3p26.1, 10q26.13, 12q12.22 and 14q23.2. The strongest associations
with LTL were reported for SNP rs12696304 and rs16847897 near TERC on 3q26. It
is unclear though whether this locus identified in Europeans, American, and Chinese
exerts a similar effect on LTL in other populations. Additionally, the effect of such
SNPs on serum levels of human telomerase reverse transcriptase (hTERT) has
not been explored before in any population. The aim of this research to: study the
influence of TERC SNPs on LTL, levels of [hTERT], indices of obesity and explore
the potential associations with type 2 diabetes mellitus [T2DM].
Methods: In a study on 225 T2DM patients and 245 age and sex matched
controls, we used Allelic Discrimination (AD) genotyping to determine near
TERC SNPs (rs12696304 and rs16847897). Fasting [hTERT], adiponectin, Insulin,
Myeloperoxidase [MPO], and [LTL] were also measured. Body Mass Index (BMI),

and waist circumference (WC) were also recoded and subjects were classified on
the basis of the degree of obesity. Body fat percentage (BF%] was measured using
Bioimpedance analysis [BIA]. Insulin resistance [IR] was assesed using [HOMA-IR]
calculator.
Results: [C/C] genotype of SNP rs16847897 was significantly associated with
telomere shortening [OR=1.6, p=0.004] and lower levels of hTERT [OR=0.4,
p=0.006]. Nevertheless, [C/C] genotype was significantly associated with higher BMI
[OR=2.2, p=0.006], WC [OR=23.4, p=0.007] and BF% [OR=2.0, p=0.005]. However,
[C/C] genotype SNP rs16847897 was associated with hypo-adiponectemia [OR=0.6,
p=0.006]. We found that [G/G] genotype of SNP rs12696304 was significantly
associated with shorter telomeres [OR=1.5, p=0.004], lower levels of hTERT
[OR=0.7, p=0.006] and hypo-adiponectemia [OR=0.5, p=0.008]. [G/G] genotype of
SNP rs12696304 was associated with higher anthropometric measures such as BMI
[OR=1.2, P=0.006], WC [OR=5.3, P=0.004] and BF% [OR=1.9, p=0.003]. Binary
logistic regression showed that; [C/C] genotype of SNP rs16847897 and [G/G]
genotype of SNP rs12696304 were significantly associated with higher T2DM risj
[OR=1.7, p=0.004]. Carriers of haplotype [CG] had significantly higher (p<0.0.0001)
BMI compared to the other two identified haplotypes [CC] and [GG] [BMICG 30.88.2
Kg/m2 vs. BMICC 26.98.4 Kg/m2 and BMIGG 28.75.3 Kg/m2]. Similar trends were
observed for WC and BF%. Additionally, telomere lengths were significantly the
shortest and hTERT levels were the lowest in( [CG], LTL: 0.8.01; hTERT: 21.85.5
ng/mL])haplotypes compared to the other haplotypes([CC], LTL: 1.030.1; hTERT):
23.76.9 ng/mL] and ([GG], LTL: 1.50.1; hTERT: 28.15.4 ng/mL]). On the other
hand, levels of MPO were significantly higher in haplotype ([CG], MPO: 6.61.7
ng/mL) compared to other two haplotypes [CC],( MPO: 3.90.4 ng/mL]) and
([GG], MPO: 4.10.4 ng/mL]). We also found that [CG] haplotype was associated
significantly with higher risk of T2DM [OR=1.5, p=0.006] and IR [OR=2.6, p=0.03].
Conclusions: We provide insights into genetic determination of a structure that is
critically involved in genomic stability. Given the importance of telomeres in nuclear
and cellular function and the central role of telomere length in determining telomere
function; our findings could have broad relevance for both normal and pathological
age associated processes.

A-216
Do anesthesia provider personnel working indoors have lower Vitamin D levels?

G. Erden1, S. Ozdemir1, G. Ozturk1, A. I. Erden2, D. Kara2, S. Isik3, J. Ergil4,

C. Vural5, A. E. Arzuhal1. 1Ankara Diskapi Yildirim Beyazit Research and
Training Hospital, Department of Clinical Biochemistry Ankara, Turkey,
2
Hacettepe University Faculty of Medicine, Department of Anesthesiology
and Reanimation, Turkey, 3Ankara Numune Research and Training Hospital,
Department of Endocrinology Ankara, Turkey, 4Ankara Diskapi Yildirim
Beyazit Research and Training Hospital, Department of Anesthesiology
and ReanimationAnkara, Turkey, 5Ankara Numune Research and Training
Hospital, Department of Anesthesiology and Reanimation Ank, Turkey
Background: There has been an increasing awareness to vitamin D deficiency.
Some of the causes of vitamin D deficiency are reduced skin synthesis (sunscreen,
skin pigment etc), decreased absorption, increased catabolism, heritable disorders,
decreased synthesis of 1,25 dihydroxyvitamin D, acquired disorders. One approach
for clinicians to decide which patients demand screening laboratory testing is to
consider serum testing in patients at high risk for vitamin D deficiency. As the most
well-known source of vitamin D is known as sun exposure, anesthesia providers
and anesthesia support personnel working indoors (in operating rooms) might be
considered at increased risk of vitamin D deficiency due to limited sun exposure.
This study aimed to investigate whether there was a higher vitamin D insufficiency
or deficiency rate among anesthesia personnel working indoors when compared with
personnel working outdoors.
Methods: 125 volunteered anesthesia (provider and support) personnel and 55 control
(outdoor workers in marketplace) subjects were included in this study. All of the
individuals were apparently healthy Turkish citizens of Ankara, Turkey (39 North,
32 East).The study was performed at the end of the winter (February 15-March
15 2013). Socioeconomic status, daily diet, vitamin D supplementation, periods of
exposure to sunlight, the use of sunscreen, regular physical activities, family history
of bone fractures and osteoporosis, and the clothing style in all of the individuals were
asked about in a questionnaire. People with high BMI, chronic disease such as asthma,
type 1 diabetes mellitus, hypertension, history of cardiac, kidney or liver disease,
those taking calcium, vitamin D or multivitamin supplements were excluded from
the study. Serum levels of total 25 hydroxyvitamin D (25-OHD) were measured by
a chemiluminescent immunoassay (CLIA) method using an autoanalyzer (LIAISON
DiaSorin, Italy ). 25-OHD levels were categorized as follows: Deficient: <10 ng/

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S59

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

mL ; Insufficient 11-29 ng/mL and Adequate: >30 ng/mL. Data were tabulated and
subjected to analysis using the Statistical Package for Social Science (SPSS) version
17.0.
Results: 74.4% of indoor anesthesia personnel and 76.6% of outdoor workers had
serum 25-OHD concentrations <10 ng/mL. 20.8% of anesthesia personnel and 23.4
% of outdoor workers had serum 25-OHD concentration levels 10-20 ng/mL. 4.8 of
% anesthesia personnel had serum 25-OHD concentration levels 21-30 ng/mL. There
was no significant difference in the mean serum 25-OHD level between two groups
(Anesthesia group: 8.984.89 ng/mL, Control group: 8.212.64 ng/mL ) ( p>0.05).
Conclusion: This study in Ankara suggests that significant proportions of the study
populations had very low vitamin D levels at the end of winter. Vitamin D deficiency/
insufficiency is common among indoor and outdoor workers. Anesthesia personnel
do not have a significant higher Vitamin D deficiency/insufficiency risk. As we have
seen, UV irradiance is not the only determinant of vitamin D status. Individuals
living at lower latitudes in relatively sunny environments are also at risk of vitamin D
insufficiency. Vitamin D supplementation may be suggested in all groups in Ankara,
including those with the highest sun exposure.

A-217
Increased Cortisol and NADPH Production in Magnesium Deficient
Hepatocytes: Implicated in the Onset of Insulin Resistance and Obesity.

C. Voma1, A. Romani2. 1Cleveland State University, Cleveland, OH, 2Case

Western Reserve University, Cleveland, OH
Most of the clinically quantifiable liver functions take place within the hepatocytes
(80% of liver cells). The current western diet is approximately 35% deficient in
magnesium (Mg2+). Subnormal Mg2+ concentrations have been reported in both
diabetes and obesity, but no clear-cut cause-effect mechanism has been stated to
elucidate the onset of these pathological conditions in Mg2+ deficiency. At the cellular
level, Mg2+ is highly concentrated within organelles including the endoplasmic
reticulum (ER), in which 15-20 mM [Mg2+]Total has been measured. Hexose 6-phosphate
dehydrogenase (H6PD), the reticular counterpart of the cytosolic G6PD, is the main
NADPH generating enzyme within the ER of the hepatocyte and is regarded as an
ancillary enzyme in pre-receptor glucocorticoid activation. In the present study, we
report that by modulating glucose 6-phosphate entry into the ER of HepG2 cells,
Mg2+ also regulates the oxidation of this substrate via H6PD. This regulatory effect is
dynamic as glucose 6-phosphate entry and oxidation can be rapidly down-regulated by
the addition of exogenous Mg2+. In addition, HepG2 cells growing in low Mg2+ show a
marked increase in H6PD mRNA and protein expression. Metabolically, these effects
on H6PD are important as this enzyme increases intra-reticular NADPH production,
which favors fatty acid and cholesterol synthesis. Under Mg2+ deficient conditions,
exposure of HepG2 cells to cortisone results in a marked production of cortisol via the
NADPH-dependent 11-HSD1, thus eliciting high intra-hepatic active glucocorticoid
concentrations, which in turn affects hepatocyte metabolism. Obesity is a known risk
factor for type 2 diabetes. However, the degree of obesity varies greatly in people
with type 2 diabetes. Not all type 2 diabetic patients are overweight or obese, and
not everyone who is overweight or obese will necessarily develop type 2 diabetes,
suggesting the involvement of other mechanisms in the development of the pathology.
11--OHSD1 has been implicated as one of these auxiliary mechanisms, as it would
lead to cortisol-based insulin resistance at least in certain patients.
HepG2 cells were grown in the presence of 0.6 (deficient) or 1.0 mM (physiological)
[Mg2+ ]o and analyzed for NADPH and 11-HSD1-mediated cortisol production.
The mRNA expression level of H6PD, G6Pase, and 11-HSD1 were analyzed by
RT-PCR while protein expression was assessed by Western Blot analysis. Under our
experimental conditions, insulin responsiveness - assessed as pAKT level by Western
Blot analysis was decreased by approximately 25% while cortisol production was
increased and associated with an increased expression of PEPCK, a key enzyme
in gluconeogenesis activation. Taken together, our results indicate that the ~60%
increase in NADPH production via H6PD in Mg2+ deficient cells resulted in increased
cortisol production and a decreased insulin responsiveness. In addition, these Mg2+
deficient cells showed 3 to 4 fold increase in H6PD and 11-HSD1 mRNA and protein
expression.
Our results support the hypothesis that Mg2+ deficiency increases H6PD activity and
expression, setting the conditions for increased production of cortisol and decreased
hepatic insulin responsiveness.

S60

A-218
Comparison of IFA and RIA based assays for measuring adrenal autoantibody
response

O. Zhukov, J. Popov, R. Vergara, D. Keno, L. Liu, J. Nakamoto, S. J.

Naides. Quest Diagnostics Nichols Institute, San Juan Capistrano, CA
Background: The diagnosis of autoimmune Addisons disease (ADD), a primary
adrenal insufficiency, depends on demonstrating inappropriately low cortisol
production and high titers of adrenal cortex autoantibodies (ACAs) or 21-hydroxylase
(21-OH) autoantibodies. ACA titers are determined using an immunofluorescence
assay (IFA), while 21-OH autoantibodies are detected with a radioimmunoassay
(RIA). In IFAs, response against 21-hydroxylase (21-OH) accounts for majority
of the immunoreactivity, but antibodies against two other steroidogenic enzymes
(17 hydroxylase [17-OH] and SCC [P450cSCC]) also contribute. A sensitive and
convenient RIA is available to measure anti 21-OH antibody using recombinant
125
I-labelled 21-OH expressed in yeast. Discrepancy between ACA IFA and 21-OH
RIA test results was reported among individuals with endocrine autoimmune diseases
often associated with adrenal insufficiency. We evaluated concordance between ACA
IFA and 21-OH RIA results in a large set of samples received for routine adrenal
antibody testing. Methods: De-identified residual specimens (n=280) originally
submitted to Quest Diagnostics Nichols Institute (San Juan Capistrano, CA) for
routine adrenal antibody testing by either IFA (n=140) or RIA (n = 140) were re-tested
with both assays. Also included in the analysis were 1) results for an additional 264
specimens submitted for both testing with assays (ACA IFA and 21-OH Ab RIA), and
22 sera positive for mitochondrial antibodies. The ACA IFA was performed using
monkey adrenal tissue (MarDx Diagnostics, Inc., Carlsbad, CA) and FITC-labeled
goat antibodies to human IgG (Inova Diagnostics, Inc., San Diego, CA). Anti 21OH testing was performed with a commercial RIA (Kronus Inc., Star, ID) employing
125
I-labeled 21-OH produced in yeast. Anti-mitochondrial antibodies (AMAs) were
detected by IFA using rat kidney tissue (MarDx Diagnostics) and by Quanta Lite
ELISA (Inova Diagnostics) employing recombinant antigen (MIT3). Results: The two
assays yielded concordant results in 460 (83%) of the 554 samples, including 328 with
negative and 132 with positive results, 94 samples were discordant. Samples with
low positive results were the main contributors for IFA/RIA discordance: among 55
RIA+/IFA- samples, 49 showed RIA values close to the cut off and only 6 samples
had values >10 U/mL. This discrepancy could be due to the fact that some of 21-OH
epitopes recognized in RIA were hidden or not present in IFA substrate. Among
29 RIA-/IFA+ samples, most (18/29) had titers of 1:10 and only 4 had titers >1:40.
The presence of other antibodies, either to steroid-producing cell antibodies or antimitochondrial antibodies, could cause these discrepancies. Presence of mitochondrial
antibodies was identified in one sample. Conclusions: Low-positive samples are the
major contributors to discrepancies between IFA and RIA results on adrenal antibody
testing. The presence of anti-mitochondrial antibody may interfere with adrenal
antibody testing by IFA.

A-219
Extreme Physical Stress Stimulates Bone Marrow-derived Circulating Stem/
Progenitor Cells that Mediate Tissue Repair: Possible Clinical Implications

I. Papassotiriou1, E. Goussetis2, K. Skenderi3, M. Tsironi4, G. Chrousos5.

1
Department of Clinical Biochemistry, Aghia Sophia Childrens
Hospital, Athens, Greece, 2Stem Cell Transplant Unit, Aghia Sophia
Childrens Hospital, Athens, Greece, 3Laboratory of Nutrition and Clinical
Dietetics, Harokopio University, Kallithea, Greece, 4Department of
Internal Medicine, University of Peloponnesus, School of Nursing, Sparta,
Sparta, Greece, 5Department of Pediatrics, University of Athens Medical
School, Athens, Greece
Background: Autologous progenitor cells represent a promising option for regenerative
cell-based therapies. Endothelial progenitor cells (EPCs) participate in vascular repair
and angiogenesis, while circulating bone marrow-originated fibrocytes represent
multipotent cells mediating tissue repair and remodeling after injury. Aging and
cardiovascular risk factors, such as diabetes, however, affect circulating endothelial
and bone marrow-derived progenitor cells, limiting their therapeutic potential. The
Spartathlon ultradistance foot race (246Km continuous, prolonged, brisk exercise
for up to 36h), is associated with profound physical strain, which renders it an ideal
model of prolonged severe physical stress. The runners endure dramatic systemic and
inflammatory changes, as their immune system functions intensively to cope with
heart and skeletal muscle and other organ damage secondary to excessive physical
strain. We hypothesized that this type of exercise might stimulate release of EPCs and
other bone marrow-derived cells.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

Athletes and Methods: We investigated the effect of physical stress on the number
of circulating EPCs and fibrocytes, along with circulating molecules indicative of
endothelial dysfunction and adipose tissue-derived proteins, in 20 Spartathlon
athletes before, at the end and at 48 h post-race. The EPCs were obtained by culturing
peripheral blood mononuclear cells (PBMC) under endothelial cell conditions
(EndoCult) and were measured as colony-forming units (CFUs). Circulating
fibrocytes were cultured from PBMCs in IMDM medium supplemented with IL-3
and M-CSF and identified as CD(45+)CD(14+)CD(34low)Collagen-I(+) fibroblastic
cells. We also determined the plasma levels of endothelial dysfunction molecules E-,
L- and P-selectins, soluble Intercellular Adhesion Molecule-1 (sICAM-1), soluble
Vascular Cell Adhesion Molecule-1 (sVCAM-1), and thrombomodulin (TM),
along with adipose tissue-derived proteins leptin, adiponectin (ADPN), lipocalin-2
(NGAL), Retinol Binding Protein-4 (RBP-4), Plasminogen Activator Inhibitor-1
(PAI-1), Macrophage Migration Inhibitory Factor (MIF), IL-8 and Macrophage
Chemoattractant Protein 1 (MCP-1) by means of immunoenzymatic techniques.
Results: Circulating EPCs increased by nearly ten-fold in peripheral blood at the
end of the Spartathlon race (from 4815 cells/ml to 46436 cells/ml) and they
remained increased (42028 cells/ml) even at 48h post-race (p>0.5). Plasma levels of
endothelial dysfunction molecules showed different patterns of responses: E-selectin,
sICAM, sVCAM and thrombomodulin were increased significantly at the end of the
race and returned to pre-race levels 48 h post-race, (p>0.6). Similarly, the adipose
tissue-derived proteins NGAL, IL-8 and MCP-1 showed significant increases at the
end of the race and returned to pre-race levels 48 h post-race, (p<0.5).
Conclusions: Our study demonstrates that acute inflammatory tissue damage induced
by exhausting exercise increases EPCs but not fibrocytes. Given the ability of EPCs
to promote angiogenesis and vascular regeneration and the association of fibrocytes
with tissue fibrosis after persistent inflammation, we conclude that this kind of cell
mobilization may serve as a physiologic repair mechanism in acute inflammatory
tissue injury and a source of potential cell therapies in the near future. Furthermore,
this study shows different patterns of adipose tissue-derived protein response to the
systemic effort and inflammatory changes.

A-220
CDC Standardization Programs- Testosterone, Estradiol, and Vitamin D

J. C. Botelho, H. W. Vesper. Centers for Disease Control and Prevention,

Atlanta, GA
Laboratory measurements are critical in patient care and public health decision
making. However, the accuracy and reliability of these measurements prevent
appropriate detection, treatment and prevention of diseases. The aim of CDC
Standardization Programs is to standardize clinical measurements which ensure that
accurate and comparable measurements are obtained regardless of the measurement
procedure, location, and time. To achieve this goal, the CDC Standardization
Programs are providing a comprehensive range of services and programs such as
Reference Services, Standardization-Certification Programs, and Accuracy-based
Quality Assurance Monitoring Services for testosterone (T), estradiol (E2), and
vitamin D [25(OH)D].
As part of the Reference Laboratory Services the CDC has established higher
order reference measurement procedures for T, E2, 25(OH)D2 and 25(OH)D3 in
serum using LC-MS/MS. These measurement procedures are traceable to primary
reference materials and to JCTLM certified reference measurement procedures. Using
these reference methods, CDC assigns target values to sera used in its certification
programs and by outside partners such as clinical and research laboratories, assay
manufacturers, and proficiency testing providers. These materials are used for method
comparisons, calibration, and trueness controls. CDC Standardization
CDC Standardization-Certification Programs are operating for T and E2 with the
Hormone Standardization (HoSt) Program and total 25(OH)D with the Vitamin D
Standardization-Certification Program (VDSCP). In both of these programs, quarterly
blinded challenges are performed. Bias and imprecision assessments using established
protocols and final assessment are made using criteria derived from biological
variability. At present, 17 participants are enrolled in the HoSt-T Program (established
in 2010) and 23 in VDSCP (established in 2013). Participants include clinical,
academic, and pharmaceutical laboratories as well as manufacturers. Approximately
85% of participants have met the established criteria. Successful laboratories are
published on the CDC website (http://www.cdc.gov/labstandards/hs.html). Over the
past 4 years the CDC has provided 97 calibration verification serum sets to requestors
and has had 85 enrollments in HoSt and VDSCP, which include many reenrollments.
While participation has increased the success rate of participants has continued to
improve as well. The testosterone HoSt Program has increased success rates by
participants over the past 3 years from 79% in cycle 1 to 100% in cycle 3.

CDC Hormone Standardization Programs are endorsed and supported by key

stakeholders such as the Partnership for Accuracy in Hormone Testing (PATH) and its
affiliated organizations (i.e., AACC, The Endocrine Society, and American Urology
Association). Furthermore, it collaborates with these organizations to further improve
testing for other hormones such as thyroid hormones.

A-221
Quantifying Insulin-like Growth Factor-1: Inter-assay Variation Remains an
Issue

H. Li1, D. Bailey2, L. Martin2, T. Randell2, K. Garen2, G. Waite2, A. S.

Ptolemy2. 1Gamma-Dynacare Medical Laboratories, Brampton, ON,
Canada, 2Gamma-Dynacare Medical Laboratories, London, ON, Canada
Background: As the main mediator of the somatotropic effects of growth hormone
(GH), an accurate measurement of human insulin-like growth factor-1 (IGF-1) is
required for the diagnosis and management of GH secretion disorders. However, the
standardized measurement of IGF-1 continues to suffer from inter-assay variability,
which may lead to inaccurate patient case decision making. In early 2013, the receipt
of a vendor notification stating that the IGF-1 reagent lots deployed in our laboratory
positively shifted patient median values prompted us to validate and ultimately
deploy an alternate vendors IGF-1 platform for patient testing. Recently, the primary
vendor resumed IGF-1 reagent supply, triggering a secondary validation of these
reformulated lots. These two studies specifically examined the inter-assay variability
of IGF-1 measurements and the relative analytical performance of each test.
Methods: Linearity, intra- and inter-day precision, accuracy and sample carry-over
were validated for IGF-1 measurements using the IDS-iSYS (Immunodiagnostic
Systems) and the reformulated Immulite 2000 (Siemens Healthcare Diagnostics)
assays, respectively. Patient correlation studies between the IDS-iSYS and the
original and reformulated Immulite 2000 reagents were also respectively performed.
Results: The IDS-iSYS and reformulated Immulite 2000 assays had linear ranges of
10 to 1200 ng/mL (R^2 = 0.998, slope= 0.973) and 20 to 1700 ng/mL (R^2= 0.999,
slope=1.02), respectively. At IGF-1 concentrations of 30.8, 249.1, 830.3 ng/mL and
45.0, 67.5 and 227.0 ng/mL the intra- and inter-day precision (%CV, N=20) of the
IDS-iSYS and reformulated Immulite 2000 assays did not exceed 4.6% and 7.5%,
respectively. The relative error (%RE) of the IDS-iSYS and Immulite 2000 methods
respectively ranged from -8.4% to 1.5% and -3.5% to 7.5% for these precision
studies. No significant carry-over was observed on either platform. Patient sample
comparisons between the IDS-iSYS and the original Immulite 2000 formation showed
significant bias (Deming regression: y = 0.739x+35.87, N=94, R^2=0.988). This
method bias was exacerbated at IDS-iSYS derived IGF-1 concentrations >300 ng/mL
(Deming regression: y = 0.616x+114.88, N=15, R^2=0.981, IDS-iSYS range= 256.3
to 770.1 ng/mL, Immulite 2000 range = 307.0 to 1128.0 ng/mL), relative to lower
IGF-1 concentrations (Deming regression: y = 0.854x+14.73, N=79, R^2=0.976,
IDS-iSYS range= 42.8 to 290.7 ng/mL, Immulite 2000 range = 31.5 to 351.0 ng/
mL). Interestingly, this bias was less significant when patient results obtained with the
reformulated Immulite 2000 reagents were correlated to those obtained with the IDSiSYS (Deming regression: y = 1.082x-9.9, N=60, R^2=0.989, Immulite 2000 range =
25.0 to 352.0 ng/mL, IDS-iSYS range=31.0 to 352.0 ng/mL), although fewer samples
with IGF-1 concentrations >300 ng/mL were included in this cohort.
Conclusion: Although the reformulated Immulite 2000 and IDS-iSYS IGF-1 assays
offer acceptable analytical and clinical performance, a significant bias was noted
with the original Immulite 2000 formulation. This difference was observed despite
both Immulite formulations being traceable to the reference standard NIBSC 1st IRR
87/518. The IDS-iSYS assay is traceable to NIBSC 02/254. Laboratories should be
aware that inter-assay IGF-1 variability must be carefully examined and its impact on
the diagnosis and management of GH deficiency and acromegaly considered when
testing platforms are changed.

A-222
Development of a Biochip Based Immunoassay for Quantification of Total Beta
hCG in Serum

R. Duncan, A. McFall, S. Brockbank, A. Chacko, C. Richardson, P. Lowry,

R. McConnell, S. FitzGerald. Randox Laboratories Limited, Crumlin,
United Kingdom
Background: Human chorionic gonadotropin (hCG) is a member of the
glycoprotein hormone family. It is heterodimeric and the alpha-subunit of hCG
(92 amino acids) is identical to that of LH, FSH and TSH. The beta-subunit of
hCG (145 amino acids) comprises the unique component of hCG and accounts

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S61

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

for the biological activity of this hormone. hCG interacts with the LH/hCG
receptor and stimulates and maintains the corpus luteum after fertilization so
it will not degenerate. The corpus luteum of pregnancy produces increasingly
greater amounts of estrogen and progesterone for an additional ten weeks until the
placenta takes over the secretion of these steroid hormones.This study reports the
development of a biochip based immunoassay for the determination of total beta
hCG in serum. This represents a new analytical tool for the detection of pregnancy.
Methods: A sandwich chemiluminescent biochip based immunoassay applied to the
Evidence Investigator analyser was employed. The capture antibody was immobilised
and stabilized on the surface of the biochip and detector antibody was conjugated
to HRP. Chemiluminescent signal was detected by digital imaging technology. The
intensity of the signal is proportional to the analyte concentration in the sample. A
correlation study was conducted using a commercially available immunoassay.
Results: The assay was target specific showing <1% cross reactivity with FSH, LH
and prolactin and <1% recovery of hCG. The limit of detection was 0.914 mIU/
ml for an assay range 0-2500 mIU/ml and the limit of blank was 0.388 mIU/ml.
The intra-assay precision (n=23), expressed as %CV, was <7.5%. In the correlation
study 80 serum samples were tested and the following linear regression equation
was achieved versus another available immunoassay: y=1.5138x-185.07; r=0.984.
Conclusion: This evaluation indicates applicability of the developed biochip based
immunoassay for the detection of total beta hCG in serum. This represents a new
analytical tool for the detection of pregnancy in test settings.

A-223
Performance Characteristics of Six Automated 25-Hydroxyvitamin D Assays:
Mind Your 3s and 2s

S. P. Wyness1, J. A. Straseski2. 1ARUP Institute for Clinical and

Experimental Pathology, Salt Lake City, UT, 2University of Utah School of
Medicine, Salt Lake City, UT
Objective: Analyze the performance of 6 automated total 25-hydroxyvitamin D
assays using 25(OH)D2/D3 and 25(OH)D3 only samples.
Methods: Access 2 and DxI 800 (Beckman Coulter*), ARCHITECT i2000SR (Abbott
Diagnostics), ADVIA Centaur XP (Siemens), Liaison XL (DiaSorin) and Modular
E170 (Roche Diagnostics) assays were evaluated for imprecision, method comparison
and concordance. Imprecision used commercial control material tested in duplicate
twice daily for 5 days. Method comparisons used residual serum samples with
endogenous D2 and D3 (n=50) or D3 only (n=86). Comparisons with all 136 samples
were intended to simulate real-world laboratory testing. Results were compared to an
in-house LC-MS/MS method (traceable to NIST SRM 972) using Passing-Bablok
regression and Bland-Altman bias plots. Acceptability criteria were coefficient of
variation (CV) <10% and bias <15.8%.
Results: Imprecision was acceptable for all assays except E170 and Centaur (both CV
11%). Regression analysis of all samples in comparison to LC-MS/MS demonstrated
under-recovery for ARCHITECT, DxI, E170 and Liaison assays (slopes 0.868,
0.983, 0.912, 0.834) while Access and Centaur over-recovered (slopes 1.013, 1.030).
All correlation coefficients were below 0.95. Compared to D2/D3 samples, E170
and Centaur showed the greatest improvement in slope without D2 while Liaison
was unaffected. Also, E170 under-recovered with D2/D3 and over-recovered in the
absence of D2. Access, Centaur and DxI assays exhibited the opposite effect. Constant
bias for all samples ranged from -3.3 (Centaur) to 1.7 ng/mL (ARCHITECT).
Intercepts improved without D2 present for all assays except ARCHITECT and
E170. Centaur constant bias improved the most in the absence of D2. Testing all
samples, Centaur had the lowest overall bias (2%) and E170 (20%) and Liaison (22%)
exceeded acceptable criteria. Testing D2/D3 samples, DxI and Access had the lowest
bias (4%); ARCHITECT (26%), E170 (36%) and Liaison (29%) exceeded acceptable
criteria with these samples. In the absence of D2 the Liaison still exceeded this limit
(18%), the ARCHITECT had the lowest bias (1%) and E170, Centaur and Access
were comparable to each other (8-9%). All assays over-recovered when analyzing
vitamin D deficient samples (<20 ng/mL, n=31), with E170 (20%) and Liaison (19%)
exceeding bias criteria. Concordance with LC-MS/MS at 20 ng/mL ranged from 77%
(Centaur) to 89% (DxI). ARCHITECT, E170 and Liaison concordance improved
without D2.
Overall, Access and DxI had slopes close to 1 and acceptable bias for all sample
groups. Liaison had the lowest slopes and was not affected by D2. While ARCHITECT
slope and intercept were not greatly affected by D2, bias and concordance improved
without D2 present. E170 and Centaur assays were most affected by D2, based on
improvements in slope, intercept or bias when D2 was absent.
Conclusions: It is important to consider the effects of D2 and D3 on individual
assay performance. Assessing performance using total vitamin D may mask possible
interferences in supplemented populations.

A-224
Quantitation of Anti-Mllerian Hormone by the AnshLabs picoAMH ELISA
Assay

J. Lu1, C. Holt2, M. Choi2, K. Kalp2, J. Straseski3. 1ARUP Institute

for Clinical and Experimental Pathology, Salt Lake City, UT, 2ARUP
Laboratories, Salt Lake City, UT, 3Department of Pathology, University of
Utah School of Medicine, Salt Lake City, UT
Background: Anti-Mllerian hormone (AMH) is responsible for regression of the
female ductal system during embryonic development. It is produced by the male testes
until puberty and the female granulosa cells until menopause. A highly sensitive AMH
assay is ideal for investigation of infertility, menopause, ovarian reserve or monitoring
granulosa cell tumors post-therapy. The picoAMH ELISA kit (Ansh Labs, Webster,
TX, USA) is a new quantitative immunoassay that detects ultra-low concentrations of
AMH in human serum. Here, we describe the analytical performance of the picoAMH
assay.
Methods: Imprecision studies used manufacturers controls (65 and 185 pg/mL) and
patient pools (201 and 404 pg/mL) assayed in duplicate or triplicate once daily for
10 days. Dilution imprecision was tested using 1:10 and 1:100 dilutions of serum
pools assayed in triplicate once daily for 5 days. Limit of blank (LOB) and limit of
detection (LOD) were assessed using the blank and 6.3 pg/mL calibrators. Linearity
was determined by serially diluting a high AMH serum sample in blank calibrator to
create 5 samples tested in duplicate. Recovery was evaluated by adding the highest
2 calibrators to patient samples (84 and 202 pg/mL). Temperature stability was
determined by storing 2 specimens (99 and 301 pg/mL) ambient for 24 hr, 4C for 7
days and -20C for 3 weeks. Effects of up to 3 freeze/thaw cycles were studied. Method
comparison of 57 samples in the range of 80-181,000 pg/mL was performed using
the Beckman AMH Gen II ELISA as the comparator method. Gender/age-specific
reference intervals were established using fresh or biorepository serum specimens (6
mos-71 yrs, n=1,273).
Results: Imprecision and dilution imprecision studies showed total CVs 6.3 and
8.7%, respectively. LOB and LOD were 0.81 and 3.11 pg/mL, respectively. The
assay was linear to 696 pg/mL. Recovery ranged 76-101%. AMH differed <18% at all
storage temperatures and <5% after 3 freeze/thaw cycles. Deming regression of the
method comparison yielded y = 0.999 x - 0.226, R = 0.99. Eleven gender/age-specific
reference intervals were established using non-parametric and robust statistics.
Conclusions: The AnshLabs picoAMH ELISA demonstrated performance close to
the manufacturer claims and excellent correlation with the comparator method. This
method lowered the LOD from the current 80 pg/mL to 3 pg/mL. We report gender/
age-specific reference intervals for this assay that will be useful for clinical practice.

A-225
Insulin and leptin signaling in placenta from gestational diabetic subjects

A. Prez-Prez1, P. Guadix1, J. Maym2, J. L. Dueas1, C. Varone2,

C. Gonzlez-Rodrguez1, M. Fernndez-Snchez3, V. SANCHEZMARGALET1. 1VIRGEN MACARENA UNIVERSITY HOSPITAL,
SEVILLE, Spain, 2University of Buenos Aires, Buenos Aires, Argentina,
3
IVI, SEVILLE, Spain
Background: Insulin and leptin receptors are known to share signaling pathways, such
as JAK2/STAT-3 (Janus kinase 2/signal transduction and activator of transcription
3), MAPK (Mitogen activated protein kinase) and PI3K (phosphoinositide 3-kinase).
Both positive and negative cross-talk have been previously found in different
cellular systems. Gestational diabetes (GDM) is a pathophysiological state with high
circulating levels of both insulin and leptin. We have previously found that these
three signaling pathways are activated in placenta from GDM patients to promote
translation, involving the activation of leptin receptor. Now, we tested the hypothesis
that both leptin and insulin receptors might contribute to this activation in a positive
way that may become negative when the system is overactivated.
Methods: To answer this question we studied the activation of leptin and insulin
receptors in placenta from GDM and normal pregnancies by Western blot. Besides,
we performed in vitro studies with insulin and leptin stimulation of trophoblast
explants to study PI3K and MAPK signal transduction pathways by Western blot
using specific antibodies of phosphorylated proteins. Bands were scanned and data
analyzed by Anova followed by Bonferronis post test.
Results: We have found that both leptin and insulin receptors are activated in placenta
from GDM. In vitro stimulation of trophoblast explants with both leptin and insulin at
submaximal doses (0.1 nM) potentiated the activation of PI3K and MAPK signaling,
whereas preincubation with maximal concentrations of insulin (10 nM) and further

*Assays pending US FDA approval

S62

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

stimulation with leptin showed negative effect. Similarly, trophoblastic explants from
GDM placenta, which presented high signaling levels, had a negative signaling effect
when further incubated in vitro with leptin.
Conclusions: Insulin and leptin receptors have positive effects on signaling,
contributing to high signaling levels in placenta from GDM, but insulin and leptin
have negative effects upon overstimulation.

A-226
Ultrasensitive Luteinizing Hormone Assay on the MesoScale Discovery
Platform

N. Vidal-Folch, S. K. Grebe, R. J. Singh, A. Algeciras-Schimnich. Mayo

Foundation, Rochester, MN
Background: Luteinizing Hormone (LH) or Lutropin is a glycoprotein composed
of and subunits secreted by the anterior pituitary gland after stimulation by
gonadotropin-releasing hormone (GnRH). In children, LH is measured as an aid in
the diagnosis of gonadal disorders such as central precocious puberty (CPP) and
delayed puberty. Current automated immunoassays are sensitive to about 0.1 IU/L.
These levels are appropriate for adults, but clinical studies suggest that basal LH
reference ranges in pre-pubertal children are below the lower limit of detection of
automated assays and a more sensitive assay is needed. Objective: Develop an
ultrasensitive LH assay for pediatric patients using the electrochemiluminescence
multi-array technology from MesoScale Discovery (MSD). Methods: The LH
assay is a sequential two-site electrochemiluminescence laboratory developed test.
A monoclonal biotinylated LH capture antibody is added to a streptavidin coated
plate, incubated for 30 min and washed to remove unbound antibody. Sample is
added and incubated overnight at 4C. After washing, a SULFO-TAG labeled
detection antibody is added. After 2 h the plate is washed and counted on the MSD
sector imager reader. The assay is calibrated against the WHO International Standard
reference material from NIBSC (2nd IS 80/552). The performance characteristics of
the assay were established over two different reagent lots and included determination
of imprecision, limits of quantification and detection, linear measurement range
and dilution linearity, spike recovery, interferences, sample stability and a method
comparison with the Beckman Access LH assay. Results: Intra-assay and intra-assay
imprecision on patient samples (0.03-25 IU/L) ranged from 1.3% to 3.1% and from
5.8% to 7.9%, respectively. The limit of detection was 0.004 IU/L and the limit of
quantitation was 0.02 IU/L (15% CV). The linear measurement range was 0.02 to 28
IU/L. Average dilution linearity and spike recoveries were 92.9% (range 86-122%)
and 101% (range 94-104%), respectively. The assay is not affected by hemolysis (up
to 1000 mg/dL hemoglobin), lipemia (up to 1000mg/dL triglycerides) or bilirubin (up
to 5mg/dL bilirubin). Repeat measurements showed <20% variability for serum and
serum-separator tubes up to 14 days ambient or refrigerated and through 3 freeze/thaw
cycles. Method correlation using Passing-Bablok regression fit against the Beckman
Access LH assay was y=0.9119x-0.04042 and r=0.982 (N=200) in the concentration
range of 0.3 - 28 IU/L. Conclusion: We have developed an ultrasensitive LH assay
useful in the diagnosis of gonadal disorders in pediatric patients. The assay provides
accurate results with a 10-fold improvement in functional sensitivity over existing
automated assays.

A-227
Preanalytical validation of a serum normetanephrine, metanephrine and
3-methoxytyramine assay

O. M. Itkonen, U. Turpeinen, N. Tohmola, E. Hmlinen. Helsinki

University Central Hospital, HUS Helsinki, Finland
Background: Analysis of metanephrine (MN), normetanephrine (NMN) and
3-methoxytyramine (3MT) in plasma or serum has recently replaced the urinary assay
in many laboratories for the diagnosis of pheochromocytoma. The aim of this study
was to validate preanalytical factors of serum MN, NMN and 3MT assays.
Methods: We used samples from apparently healthy adult volunteers to study sample
stability (n=25), sampling device (n=13), postprandial effect (n=7), intra-individual
within-day variation and diurnal variation (n=7). Samples (200L) with [2H3]-labeled
internal standards were extracted with Oasis WCX Elution plates (Waters), washed
with water, methanol and 0.1% formic acid in acetonitrile and eluted with 2 x 50 L
of 2% formic acid in 95% acetonitrile-5% water. The eluent (25 L) was analyzed
by liquid chromatography-tandem mass spectrometry (LC-MS/MS) employing an
Agilent 1200 liquid chromatograph (Agilent Technologies), a 4000 QTRAP mass
spectrometer (AB Sciex), and an Atlantis HILIC Silica 50x2.10 mm column (Waters).
The LOQ of the assay was 0.025 nmol/L, the intra-assay CV was <7.2%, the inter-

assay CV was <8.3%, and the linear range 0.025-5 nmol/L for MN, NMN and 3MT.
Paired t-test was performed by Analyse-it for Microsoft Excel 2003.
Results: Serum NMN and MN were stable (concentration changed <20%) for at least
7 days at room temperature and at +4C, for 12 weeks at -20C. NMN was stable
during 1 and MN at least during 4 freeze-thaw cycles. No valid stability data of serum
3MT could be obtained because the concentrations were below the detection limit in
the majority of our samples. NMN and 3MT concentrations were lower (p0.032)
in samples drawn into Li-heparin plasma tubes (mean 0.41 and 0.03 nmol/L,
respectively, Venosafe 60 USP U Lithium Heparin tube, Terumo) than in samples
drawn into glass tubes (0.49 and 0.05 nmol/L, respectively), clotting catalyzator tubes
(0.47 and 0.04 nmol/L, respectively) and SST II Advance gel tubes (0.47 and 0.04
nmol/L, respectively). All serum tubes were from Vacutainer. On contrary, MN was
the highest in Li-heparin plasma (0.18 nmol/L), but the difference was significant
only as compared to serum drawn into catalyzator tubes (0.17 nmol/L, p=0.0165).
A regular breakfast meal had no effect on serum NMN, MN or 3MT concentrations
(p<0.075 for all). There was no difference (p>0.066) in NMN and 3MT concentrations
in samples drawn at 8 a.m. (0.48 and 0.03 nmol/L, respectively), noon (0.51 and 0.04
nmol/L, respectively) and 4 p.m. (0.45 and 0.04 nmol/L, respectively). However, MN
concentration was 0.16 nmol/L at 8 a.m., 0.17 nmol/L at noon and 0.19 nmol/L at 4
p.m. (p=0.0304). The mean intra-individual within-day variation of NMN, MN and
3MT was 13% (range 7%-23%), 13% (range 3%-13%) and 22% (range 9%-36%),
respectively.
Conclusions: To minimize assay variation due to preanalytical factors, we suggest that
samples be transported to the laboratory at room temperature but stored frozen. Only
1 freeze-thaw cycle should be allowed before analysis, serum instead of Li-heparin
plasma should be used, sampling should occur before noon and no fasting before
sampling is required.

A-228
Functional Sensitivity of Five Automated Estradiol Immunoassays

W. E. Owen1, J. A. Straseski2. 1ARUP Institute for Clinical and Experimental

Pathology, Salt Lake City, UT, 2Department of Pathology, University of
Utah Health Sciences Center, Salt Lake City, UT
Background: Estradiol (E2) is a steroid hormone produced primarily by the ovaries
with small amounts produced in the testes and adrenal cortex. E2 measurement is
used for assessing sexual development, fertility disorders, gynecomastia, estrogenproducing tumors and hyperplasia in the adrenal cortex. E2 is also used in monitoring
fertility therapy for patients undergoing in vitro fertilization. Imprecision and methodto-method differences, especially at clinically important low concentrations, continue
to be problematic for E2 immunoassays. We studied the functional sensitivity (FS) of
five automated E2 immunoassays.
Methods: We evaluated the ARCHITECT i2000 (Abbott), DxI 800 (Beckman),
ELECSYS E170 (Roche), and ADVIA Centaur and IMMULITE 2000 (Siemens)
immunoassays. Five pools of different concentrations were each prepared by
combining serum samples with comparable E2 concentrations as determined by LCMS/MS. Pools were aliquotted and stored frozen (-70 0C) until testing. Imprecision
was evaluated over 12 days using two lots of reagent and two calibrations. Five
aliquots per pool (one per method) were thawed per day and assayed once per run, one
run per day, two days per week, and three weeks per reagent lot (total=12 replicates).
FS was determined by fitting a power function to the imprecision data using Excel.
Results: The FSs (ng/L) for ARCHITECT i2000, DxI 800, ELECSYS E170,
ADVIA Centaur, and IMMULITE 2000 were determined to be 3, 39, 11, 30, and 22
respectively. All methods met manufacturers claims except ADVIA Centaur (12.5
ng/L). Mean concentrations per pool are summarized in Table 1.
Conclusions: The ARCHITECT i2000 and ELECSYS E170 showed the best
performance with FSs below 20 ng/L. However, it has been suggested that FS of 5
ng/L or lower are needed for clinical usefulness. Additionally, these immunoassays did
not provide comparable mean E2 concentrations for the serum pools tested. Further
harmonization of E2 immunoassays is required, particularly at lower concentrations.
Pool

Comparison of pool E2 cencentrations (ng/L) and FS (ng/L) by method

LC-MS ARCHITECT
ELECSYS ADVIA
DxI 800
IMMULITE 2000
/MS
i2000
E170
Centaur

1
2
3
4
5

15.4
36.0
101.0
202.0
651.0

25.8
38.0
86.1
160.3
524.8

19.1
39.3
93.1
215.6
738.8

15.5
33.6
92.3
207.2
782.3

29.5
41.0
89.8
183.3
631.1

25.5
41.6
88.8
210.3
634.9

FS

0.5

39

11

30

22

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S63

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

aminotransferase (ALT) which was measured by enzymatic colorimetric method.

Serum ALT level > 30 u/l was considered elevated.

A-229
Development of a sensitive Dried Blood Spot Anti-Mullerian Hormone (AMH)
ELISA

A. Kumar1, B. Kalra1, A. S. Patel1, G. Savjani1, L. Lechtenberg2, C.

R. Garcia2. 1Ansh Labs, Webster, TX, 2University of Pennsylvania,
Philadelphia, PA
Background: The aim of this study was to develop a highly sensitive and simple dried
blood spot human AMH ELISA to assess ovarian reserve.
Relevance: AMH has been reported to be strongly associated with age, antral follicle
counts (AFC), FSH and has emerged as a clinically useful biomarker of ovarian
reserve. Recently, there have been concerns related to AMH stability in serum/plasma
and complement interferences affecting the end result. This has generated numerous
debates and publications related to reproducibility of AMH measurements and impact
of pre-analytical sample handling. Dried blood spot specimens stability makes it a
practicable alternative to venous blood. It opens new possibilities in AMH testing,
such as comparison of historical to current patient results; simplified blood sampling
for patients in remote locations or for those who are homebound. Instead of traveling
to a clinic to get blood drawn, a blood spot sample can be taken at a convenient site
and mailed to a laboratory. This technology will be especially useful for monitoring
ovarian function of physically challenged cancer patients undergoing chemotherapy.
Methods: A three-step, sandwich-type enzymatic microplate assay has been
developed to measure AMH levels in two 7.9 mm dried blood spot disc in less than 6
hours. The assay measures human AMH and uses stabilized recombinant human AMH
as calibrators (7-1000 pg/mL). This method uses a drop of whole blood collected on
filter paper from a simple finger stick. The sample is eluted from the dried blood spot
in an extraction solution and is added directly to the well. The assay measures the bioessential AMH and does not exhibit interference by hematocrit in the extracted spot.
Results: Ansh Labs DBS AMH ELISA (AL-129), when compared to Ansh Labs US
AMH ELISA (AL-105) using 56 matched serum and dried blood spot samples in
the range of 62-18443 pg/mL yielded a correlation coefficient of 0.98 (p < 0.0001)
and a slope of 0.96 with an intercept of -7.56 pg/mL. DBS AMH ELISA (AL-129)
when compared to Ansh Labs picoAMH ELISA (AL-124) using 65 matched serum
and dried blood spot samples in the range of 5-5240 pg/mL yielded a correlation
coefficient of 0.99 (p<0.0001) and a slope of 1.02 with an intercept of -4.7 pg/mL.
Serial dilution of seven extracted dried blood specimens (5000-11000 pg/mL) in the
sample diluent showed an average recovery of 87-105%. Total imprecision, calculated
on 3 controls over 40 runs, 2 replicates per run, was 5.84% at 22.58 pg/mL, 3.15%
at 86.51 pg/mL and 4.34% at 373.18 pg/mL. The functional sensitivity of the assay
calculated at 20% CV was 3.9 pg/mL.
Conclusion: A highly simplified, sensitive, specific and reproducible dried blood spot
AMH assay has been developed to assess ovarian reserve in females of reproductive
age. The DBS results are comparable to serum based assays. The specimen stability,
ease and low cost of collection and transportation makes it a very attractive sample
type for epidemiologic and other research studies.

A-230
Nonalcholic Steato-hepatitis (NASH) in Type 2 Diabetes: Serum Body
Fat-normalized Plasma Leptin Level is a Predictor of Serum Alanine
Aminotransferase Level

H. S. Chaudhury1, S. Akter2, S. Murshed2, M. K. Rahman2, L. Ali2.

1
International Medical College, Dhaka, Bangladesh, 2Biomedical Research
Group, BIRDEM, Dhaka, Bangladesh
Background: Leptin is a multifunctional hormone which may be involved in the
pathogenesis of type 2 diabetes mellitus (T2DM) and its complications. Anti-steatotic
function of leptin is well demonstrated in animal studies. Nonalcholic steato hepatitis
(NASH) is common complication of metabolic syndrome and T2DM. In present study
we investigated the role of leptin in the development of NASH in type 2 diabetes.
Methods: A total of 119 subjects were included. In Group I (Case), 59 newly
diagnosed T2DM and in Group II (Comparison), 60 age (year, 364 vs. 354, MSD)
and Body Mass Index (BMI)-matched (kg/m2, 24.03.1 in Group I vs. 23.62.0 in
Group II) healthy control subjects were included in this observational study. Plasma
insulin (fasting and 30 min post-glucose) and leptin were estimated by Enzyme
Immunoassay. Insulin secretory capacity (HOMA-B%) and insulin sensitivity
(HOMA-S%) were calculated by Homeostasis Model Assessment using HOMACIGMA software. Fasting serum non-esterified fatty acid (NEFA) was measured by
enzymatic-colorimetric method. Hepatocellular damage was assesed by serum alanine

S64

Results: The diabetic subjects showed highly significant -cell dysfunction and also
insulin resistance as evident from HOMA B% [20.2(4.2-89.6) in diabetic vs. 78.4(35.5365.7) in control, p<0.001] and HOMA S% [84.6(39.1-226.4) vs. 118.8(22.0-3573.0),
p<0.004]. Serum fat normalized leptin level was found significantly lower in diabetic
subjects [ng/ml, 5.44 (0.65-34.7)] compared to controls [8.35 (1.36-55), p=0.012].
Diabetic subjects had higher prevalence of elevated serum ALT compared to control
subjects (33% vs 62%, p <0/01). The fat-normalized serum leptin level inversely
correlated with insulin secretory dysfunction. The serum ALT level was correlated
with fat-normalized serum leptin level (r= -0.224, p=0.016), serum triacylglycerol
(r= 0.372, p<0.001) and phase 1 insulin secretion (r= -0.213, p=0.024). The total
NEFA level in the diabetic subjects was higher than control (mmol/l, 0.6520.196 vs.
0.420.15, p<0.001).
Conclusion: The data suggest that a) Low plasma leptin in type 2 diabetes mellitus
subjects is associated with insulin secretory dysfunction; and b) Elevated serum
ALT in diabetic subjects is associated with lower level of fat-normalized leptin
and decrease in phase 1 insulin secreation. c) The fat-normalized serum leptin is a
predictor serum ALT level.

A-231
Predicted decrease of plasma 1,5-anhydroglucitol (AG) in presence of inhibitors
of glucose reabsorption (SGLT2 inhibitors): potential utility of AG as a primary
marker of drug effect

D. Fortuna, L. J. McCloskey, D. F. Stickle. Jefferson University Hospitals,

Philadelphia, PA
Background: Renal reabsorption of glucose under conditions of normoglycemia is
essentially 100%. Drugs that inhibit renal glucose reabsorption, via inhibition of
the main renal glucose transporter (sodium-glucose transporter-2, or SGLT2), have
recently been approved for use in the U.S. for treatment of Type 2 diabetes (e.g.,
Invokana). Renal reabsorption of plasma 1,5-anhydroglucitol (AG, 1-deoxyglucose),
an unregulated, non-metabolizable glucose analogue derived from diet, is normally
>99%; normal plasma AG represents a balance between slow rates of input (5 mg/
day) and excretion. In diabetes, plasma AG is often substantially decreased due to
accelerated renal loss that occurs when glucose concentration is high enough to
saturate reabsorption capacity (viz., under conditions of glucosuria). Correspondingly,
plasma AG is likely to be directly affected by drugs that inhibit glucose reabsorption.
Our objective was to examine this potential effect using an established mass balance
model for AG and varying the AG reabsorption fraction according to the same degree
of the effect on glucose reabsorption caused by the new reabsorption inhibitors.
Methods: We used a two-compartment AG mass balance model previously described
(Am J Physiol Endocrinol Metab 1997;273:E821-E830). If displaced from steadystate, changes in body total AG (T) are given by dT/dt = ki - T (GFR fe)/(1+K)/V,
where ki = AG input rate (5 mg/day), GFR = glomerular filtration rate (nominally
100 mL/min), K is the ratio between tissue and plasma compartments (K = 2.1),
V is the plasma volume (nominally 3 L), and fe is the fractional excretion (0 to 1)
of filtered AG. In normoglycemic steady-state, fe is <0.01 (= fe(ss)). According to
literature, fractional excretion of glucose in presence of target blood concentrations
of the SGLT2 inhibitor dapagliflozin is 35%-50%. We assumed an equivalent effect,
fe = 0.35, for AG in the presence of inhibitor. Using normal, steady-state plasma AG
= 21 ug/mL as an intial condition, we calculated the model-predicted time course of
changes in plasma AG following a step increase in fe from fe(ss) to fe = 0.35.
Results: The model predicts an exponential decrease in plasma AG when fe is
increased. For fe = 0.35, plasma [AG] transitions rapidly from a normal value (AG =
21 ug/mL) to a new steady-state (AG = 0.9 ug/mL) within approximately 24 hours,
with t1/2 = 3.1 hours. According to the model, the same t1/2 would be operative
for any starting plasma AG upon initiation of the same degree of inhibition of AG
reabsorption. Low GFR will slow this effect (e.g., t1/2 = 6.8 hours for GFR = 45 mL/
min) but will not affect the eventual net % change in plasma AG.
Conclusions: SGLT2 inhibitors are predicted to produce a rapid and substantial
decrease in plasma AG. The effects of SGLT2 inhibitors would obviate the usual
intent of AG measurement, which is to verify increasing plasma AG as a marker of
improvement in glycemic control. Conversely, however, plasma AG measurement
might potentially be useful in SGLT2 therapy precisely because it might (is predicted
to) act as a direct marker for successful inhibition of reabsorption of hexoses.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

Objective: The goal of this study was to evaluate the impact of systemic RAS
inhibition (AngII receptor blocker;ARB), and the attendant suppression of renal
oxidative stress, on the renal tissue RAS during the normoalbuminuric stage of DM.

A-232
Rapid and Cost-effective Determination of Plasma Renin Activity in Human
EDTA Plasma by Two Dimensional Liquid Chromatography-Tandem Mass
Spectrometry

T. Guo, B. Yue. NMS Labs, Willow Grove, PA

Background: Determination of plasma rennin activity (PRA) is a critical part in
evaluation of primary aldosteronism, which accounts for at least 2% of hypertensive
patients. PRA assay is used to quantitatively assess the capacity of renin to generate
angiotensin I (Ang I) from angiotensinogen. PRA has traditionally been measured by
radioimmunoassay (RIA) with fair sensitivity and consistency when optimal sample
handling and incubation conditions were applied. However, RIAs lack specificity
due to potential cross-reactivity between the Ang I antibody and other endogenous
peptides, have a limited dynamic range, and require laborious sample handling.
Several liquid chromatography tandem mass spectrometry (LC-MS/MS) based assays
were reported to measure PRA using either off-line or on-line solid phase extraction
(SPE). The off-line SPE based cleanup is labor-intensive and time consuming;
while on-line SPE is not cost effective. We developed a rapid and cost-effective two
dimensional LC-MS/MS assay to measure PRA in human plasma.
Methods: An API-5000 triple-quadrupole mass spectrometer (AB-Sciex) coupled
with electrospray ionization (ESI) source and Shimadzu HPLC system was used
to quantify Ang I in human plasma after incubation. Plasma samples with doublelabeled degradation standard (DS) were incubated in a water bath at 37 oC for three
hours. Labeled internal standard (IS) were spiked prior to protein precipitation with
acetonitrile. After centrifugation, the supernatants were transferred into injection
vials. 70 L extracted sample were injected onto a ZORBAX-C8 columns as the
first dimension and a Synergi Polar-RP column with a water/acetonitrile/formic acid
gradient as the second dimension. The ESI source was operated in positive ion mode
with ionspray voltage at 4000 V and heater temperature at 400 oC. Quantitation by
multiple reaction monitoring (MRM) analysis was performed. Two ion pair transitions
were monitored for the analyte and its IS/DS.
Results: The assay was validated as linear over the range from 0.15 to 150 ng/mL for
Ang I. The lower limit of quantiation (LLOQ) was 0.15 ng/mL for Ang I. Within-run
CVs were < 3.0% for all three levels of QC samples tested. Between-run CVs were
4.14% for low QC, 3.59% for mid QC, and 4.3% for high QC Samples, respectively.
Recoveries ranged from 69% to 81% for Ang I. Mean carryover was <0.13% for Ang I
in 3 runs. No interference was observed. Preliminary comparison with a validated LCMS/MS method for PRA was assessed as follows: y = 1.06x + 0.03 (r=0.999, n=6).
Conclusion: LC-MS/MS method offers specificity superior to that of RIAs. This 2D
LC-MS/MS method can rapidly measure PRA in human plasma within 3.5 minute.
Compared to off-line SPE, the advantages of this method include simplicity, high
throughput, and low cost. Thus, it can be routinely employed in a clinical environment.

A-233
Impact of angiotensin II receptor blockade on the renal cortical tissue reninangiotensin system during the normoalbuminuric stage of diabetic mellitus in
the rat.

N. Ishii1, Y. Kurosaki1, P. K. Carmines2, Y. Yamada3, T. Tsukushi4, A.

Imoto1, T. Ichikawa3, M. Yokoba1, H. Suzuki5, H. Ikenaga6, Y. Aoki7, T.
Takenaka8, T. Ichikawa1, M. Katagiri1. 1Kitasato University School of
Allied Health Sciences, Sagamihara, Japan, 2University of Nebraska
College of Medicine, Omaha, NE, 3Kitasato University Graduate School
of Medical Sciences, Sagamihara, Japan, 4Kitasato University Hospita,
Sagamihara, Japan, 5Kitasato junior college of health and hygienic
sciences, Minamiuonuma, Japan, 6Ikenaga Clinic, Otawara, Japan,
7
Kanagawa Health Service Association, Yokohama, Japan, 8International
University of Health and Welfare, Sanno Hospital, Minato-ku, Japan
Background: The circulating renin-angiotensin system (RAS) produces changes in
plasma angiotensin II (AngII) levels as a mechanism for regulating blood pressure
and maintaining fluid and electrolyte homeostasis. The renal tissue RAS can function
independent of the circulating RAS. One key component of the renal tissue RAS
is the (pro)renin receptor ((P)RR), which is able to bind either renin or prorenin.
When bound to the (P)RR, prorenin catalyzes formation of angiotensin I from
angiotensinogen (similar to the action of renin). Increased renal (P)RR expression has
been reported to contribute to development of diabetic nephropathy (DN) through a
pro-inflammatory mechanism (Clin Exp Pharmacol Physiol37:277-82,2010). Renal
oxidative stress has also been implicated in DN, and RAS inhibition suppresses renal
cortical oxidant production even during the early, normoalbuminuric stage of DM
(Clin Sci124:543-52,2013).

Methods: Four groups of rats (n=5 per group) were examined: 1) STZ group: rats
studied 2 wks after induction of DM by streptozotocin injection (STZ, 65 mg/
kg,i.p.), 2) Sham group: rats receiving the STZ vehicle, 3) STZ+TLM group: STZ
rats treated with telmisartan (TLM, an ARB; 10 mg/kg/day in chow for 2 wks), and 4)
Sham+TLM group: TLM-treated Sham rats. In each rat, blood glucose, blood pressure
and glomerular filtration rate (GFR) were measured. We quantified the following
parameters in renal cortex: 3-nitrotyrosine (3-NT) production (an oxidative stress
marker; by HPLC), AngII levels (by RIA), (P)RR expression, and expression of both
angiotensin type-1 and type-2 receptors (AT1R and AT2R by western blotting).
Results: Similar to previous reports, blood glucose levels were higher in STZ
and STZ+TLM groups than in Sham and Sham+TLM groups. Blood pressure did
not differ among groups. GFR was increased in STZ group compared with Sham
(P<0.05), and this was prevented by TLM-treatment (P<0.05 vs.STZ+TLM). Renal
cortical 3-NT production was increased in STZ compared with Sham (P<0.05);
however, TLM suppressed this phenomenon (P<0.05vs.STZ+TLM). Renal cortical
AngII levels did not differ among groups. In contrast, STZ rats showed significant
increases in renal cortical (P)RR (32352% of Sham;P<0.05) and both of 42 and
58kDa AT1R expression (2866% and 2285% of Sham, respectively;P<0.05). These
changes were prevented by TLM treatment (P<0.05vs.STZ+TLM), although TLM
did not alter either parameter in the Sham group. Renal cortex AT2R expression was
elevated in the STZ group (1556% of Sham;P<0.05), and further increased by TLMtreatment (18210% of Sham;P<0.05).
Conclusions: During the normoalbuminuric stage of DM in the rat, the renal cortex
exhibits upregulation of major components of the intrarenal RAS (AT1R, AT2R and
(P)RR) without a change in tissue AngII levels. The DM-induced changes in AT1R
and (P)RR expression are prevented by systemic AT1R blockade, and may arise via
oxidative stress. These observations indicate that the renoprotective effects of ARB
may involve not only an antioxidant effect but also effects that rely on suppression of
the intrarenal RAS.

A-234
Comparison of Immunoassays to Mass Spectrometry for Free and Total
Testosterone in Men, Women, and Children

S. L. Laulu1, J. A. Straseski2. 1ARUP Institute for Clinical and Experimental

Pathology, Salt Lake City, UT, 2Department of Pathology, University of
Utah, Salt Lake City, UT
Background: Circulating testosterone may be bound to albumin, sex hormone
binding globulin (SHBG) or remain free. Measurement of these various forms of
testosterone provides an overall assessment of androgen status and aids diagnosis of
several conditions in men, women, and children. Study objectives were to compare
5 commercially available immunoassays to mass spectrometry for free and total
testosterone in adults and children.
Methods: Residual serum samples from men (n=150), women (n=100), boys (n=25),
and girls (n=25) were obtained after completion of clinical testing for total testosterone
(TT) by liquid chromatography tandem mass spectrometry (LC-MS/MS). Free
testosterone (FT) was determined in men using equilibrium dialysis (ED)/LC-MS/
MS. All samples were further tested for TT and SHBG by the Abbott ARCHITECT
ci8200, SIEMENS ADVIA Centaur and IMMULITE 2000, Beckman Coulter DxI,
and Roche Modular E170. Albumin was measured using the Abbott ARCHITECT
ci8200 and Roche c702. FT was calculated using the Vermeulen equation. For women,
boys, and girls, calculated FT by immunoassays was compared to calculated FT using
TT by LC-MS/MS.
Results: Comparisons using Deming regression for TT and FT in men and women are
provided (Table). For boys and girls, slopes for TT ranged from 0.72 (IMMULITE) to
1.14 (ARCHITECT) and 0.84 (IMMULITE) to 1.76 (DxI), and slopes for FT ranged
from 0.82 (IMMULITE) to 1.25 (ARCHITECT) and 1.06 (ARCHITECT) to 1.18
(DxI), respectively. Overall, more immunoassays under-recovered in men and women
and over-recovered in boys and girls for both TT and FT. The average of absolute
percent bias was highest in boys for both TT (92.4%) and FT (98.1%) compared to
men (8.7% and 10.8%, respectively).
Conclusions: Consistent biases were not observed amongst methods and populations
evaluated. Challenges with accurately measuring testosterone appear to remain in
some immunoassays, but not all.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S65

Endocrinology/Hormones

Tuesday, July 29, 9:30 am 5:00 pm

Total Testosterone
Method

Slope

Intercept R

1.07
0.99
0.76
0.95
0.81

-0.87
-41.0
34.3
-7.83
25.3

0.985
0.906
0.970
0.985
0.922

1.04
0.89
0.73
0.93
0.71

0.64
7.15
23.9
-1.87
10.2

0.995
0.943
0.927
0.985
0.970

Men (comparison to
LC-MS/MS)
ARCHITECT
Centaur
DxI
E170
IMMULITE
Women (comparison
to LC-MS/MS)
ARCHITECT
Centaur
DxI
E170
IMMULITE

S66

Free Testosterone
% Bias Method
Men (comparison
of calculated
FT to ED/LCMS/MS
8.6
ARCHITECT
-12.3
Centaur
-8.8
DxI
-7.6
E170
-6.2
IMMULITE
Women
(comparison to
calculated FT
using TT by LCMS/MS)
20.8
ARCHITECT
22.0
Centaur
81.4
DxI
-9.9
E170
-3.3
IMMULITE

Slope

Intercept

% Bias

0.98
1.00
0.82
0.86
0.82

0.66
-0.85
0.25
0.45
0.42

0.954
0.831
0.970
0.959
0.906

8.2
-12.5
-14.1
-8.4
-10.9

1.06
0.93
0.85
0.94
0.76

0.00
0.10
0.30
-0.02
0.12

0.990
0.933
0.917
0.990
0.970

21.0
22.7
82.6
-9.9
34.3

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

metallo, and threonine) exist with various mechanisms of action. In our study we
aimed to evaluate the protective role of different protease inhibitors on the degradation
of parathormone (PTH), insulin, and prolactin in blood samples.

Tuesday, July 29, 2014

Poster Session: 9:30 AM - 5:00 PM
Factors Affecting Test Results

A-235
Thrombin-Mediated Degradation of Parathyroid Hormone in Rapid Serum
Tubes

S. L. Laulu1, J. A. Straseski2, R. L. Schmidt2, J. R. Genzen2. 1ARUP

Institute for Clinical and Experimental Pathology, Salt Lake City, UT,
2
Department of Pathology, University of Utah, Salt Lake City, UT
Background: Measurement of parathyroid hormone (PTH) is important for the
clinical assessment of parathyroid disease and calcium homeostasis. Previous studies
have demonstrated decreased stability of PTH in serum versus plasma specimens,
although the precise mechanism for this difference has not been established. If
thrombin activation during clot formation is responsible, the effect should be
exacerbated in tubes containing additional thrombin. Exogenous thrombin is a
constituent of Vacutainer Rapid Serum Tubes (RST; BD Diagnostics, Franklin
Lakes, NJ), which include bovine thrombin to induce accelerated clot formation. As a
known thrombin cleavage site exists in the 84 amino acid human PTH polypeptide, we
hypothesized that ex vivo PTH cleavage in RST tubes might occur. Such a possibility
has analytical and clinical implications, as intact PTH diagnostic tests are sandwich
immunoassays that use paired capture/detection antibodies to the N- and C-terminal
regions of PTH.
Methods: 1) Screening Study: Initial experiments were conducted to determine
whether measurement of analytes with potential thrombin cleavage sites would be
affected by RST tubes. Aliquots from previously collected serum specimens were
incubated for 24 hrs in either an RST or a Serum Separator Tube (SST) before
analysis on a cobas e602 immunoanalyzer (Roche Diagnostics, Indianapolis, IN). A
subset of replicate experiments was performed on an ARCHITECT i1000SR (Abbott
Diagnostics, Lake Forest, IL). 2) Fresh Collections: To verify findings in freshly
collected specimens, three tubes of fresh blood were drawn from each of 10 healthy
normal donors. Specifically, we collected one Plasma Separator Tube (PST), one
SST, and one RST from each donor. After clotting and centrifugation, intact PTH
assays on the cobas e602 were performed at multiple time points on aliquots held
at ambient temperature or 4C. Confirmatory experiments were conducted using
exogenous thrombin and the direct thrombin inhibitor hirudin.
Results: In screening studies, PTH results were lower after specimen incubation
in RSTs versus SSTs. These findings were confirmed in additional time-course
experiments and on a separate immunoassay platform. In fresh collections, RST
and SST specimens stored at room temperature also showed a decrease in PTH
results (versus PST specimens) beginning at our earliest measurement time-point
(approximately 1 hr post-collection). The magnitude of this decrease was more
prominent in RSTs versus SSTs. A similar (but smaller and slower) decrease in
PTH results in serum specimens was seen in aliquots stored at 4C. Further studies
confirmed that the decrease in measured PTH was thrombin-mediated, as it was
blocked by hirudin.
Conclusion: The present study demonstrates that thrombin is responsible for the
decrease in PTH results observed in RSTs. It is presumed that endogenous human
thrombin, activated during clot formation, may be responsible for the smaller
decreases observed in SSTs. As there is an incomplete understanding of which
additional polypeptides may possess bovine thrombin cleavage sites, these studies
provide a simplified screening strategy that laboratories can use when evaluating
RSTs for assays at their own institutions.

A-236
The effect of different protease inhibitors in blood samples taken for
parathormone, insulin, and prolactin analysis.

O. Baykan, A. Yaman, F. Gerin, O. Sirikci, G. Haklar. Marmara University

School of Medicine, Istanbul, Turkey
Background: Proteolytic degradation by proteases can affect peptide hormone levels,
which becomes especially important when there is a lag time between sampling and
analysis. This should be taken into consideration when the samples are conveyed
from peripheral to central laboratories. Five group of proteases (serine, cysteine, acid,

Methods: Blood samples (n=10) were collected from healthy volunteers into
vacutainer tubes with gel seperator (Becton Dickinson, NJ, USA) with a) no additive,
b), 1% protease inhibitory coctail (PIC) (Sigma) which inhibits serine, cysteine, and
acid proteases, and aminopeptidases added immediately after blood sampling, c) PIC
added after centrifugation (30 min after sampling) d) aprotinin which inhibits serine
proteases (500 KIU/mL) (Sigma) added immediately after blood sampling and e) K2
EDTA (1.8 g/L). The samples were allowed to clot for 30 min and centrifuged at 1300xg
for 10 min. Then, each batch of sample either stored at 4C or at room temperature
(RT) until analyzed at 6, 24, 48, and 72 hours and compared against baseline values.
Insulin, PTH, and prolactin levels were measured with electrochemiluminescence
immunoassay in modular E170 (Roche Diagnostics, Germany) analyzer. The
desirable bias values were taken from the Westgard QC database.
Results: All parameters remained within desirable bias limits when stored at 4C
until 72 hours. PTH exceeded desirable bias limits when stored at RT longer than
24 hours. PIC addition before or after centrifugation inhibited protease associated
PTH degradation. Since the PIC amount was less when added after centrifugation, a
more economic application became possible. Insulin stored at RT decreased higher
than desirable bias limits after storing longer than 6 hours and only EDTA preserved
insulin at RT. Addition of PIC before centrifugation led to hemolysis which enhanced
the insulin degradation through proteases. Prolactin remained to be stable under each
condition.
Conclusion: These results shows that when the samples are conveyed between
different locations, the preservation of peptide hormones should be kept in mind.
Although this can be achieved with various protease inhibitors, storing the samples at
4 C from sampling until analysis, will work equally well.

A-238
Pre-analytical factors effecting Alzheimers disease biomarker stability in CSF

D. R. Kozo, I. Baburina, J. B. Courtney, K. Williams, C. N. Baldasano, S.

J. Salamone. Saladax Biomedical, Inc., Bethlehem, PA
Background: Evaluation of cerebrospinal fluid (CSF) biomarkers in Alzheimers
disease (AD) is becoming increasingly important to improve the reliability of antemortem disease diagnosis to ensure proper patient management. Development of
highly precise assays for amyloid1-42 (A42) and tau protein has allowed investigators
to better characterize pre-analytical sample handling factors that may affect clinical
interpretation. A prospective collection study was designed to model different CSF
handling scenarios for storage conditions at the clinical site, followed by shipping
to and then storage and handling at the testing site. Objective: Model CSF handling
scenarios to identify conditions that would not significantly impact the determination
of A42 and tau protein. Methods: CSF was prospectively collected from 46 healthy
individuals under an IRB approved protocol. A 10mL CSF aliquot was collected
by lumbar puncture in the L3/L4 or L4/L5 interstitial space. One milliliter aliquots
were stored and shipped from the collection site at: -80C, -20C, 4C and 25C.
Using an immunometic chemiluminescent assay, 6 randomly selected patient samples
were used to investigate storage, shipping and handling conditions, including thaw
conditions, post thaw handling, and freeze/thaw cycles. Additional studies examined
the influence of varying the storage tube manufacturer and the tube type and lot for
a single manufacturer. Results: When stored at -20C and 25C, A42 values were
from 5 to 20% lower compared to the control stored at -80C. For aliquots stored at
ambient temperature for 48 hours, the A42 values were 40% lower. There was little
to no effect on tau concentrations across the range of storage temperatures. Aliquots
initially stored at -80C or ambient temperature showed the greatest decreases in
measured A42 levels (10 to 15%) following a second freeze/thaw cycle, whereas
no loss of tau protein was observed following a freeze/thaw cycle. Thawing CSF
samples at ambient temperature versus a water bath did not affect concentrations of
either biomarker. Results were consistent between patients. Storing samples for 2h at
ambient temperature post thaw resulted in less than 10% loss for either marker. The
Eppendorf Lobind tubes allowed transfer of CSF and storage at 4C for up to six
days without significant loss of either A42 or tau protein biomarkers. No significant
difference in either A42 or tau protein concentrations was observed between
different lots and sizes of these tubes. Conclusion: The measured level of A42 was
significantly affected by CSF storage, shipping and handling conditions, although
there was no appreciable impact on tau. The utility of these markers in routine testing
can be improved by using conditions at the collection and testing sites that minimize
analyte losses. Immediate storage of freshly collected CSF at -80C and shipment on
dry ice gave the best A42 recovery. These conditions will also preserve tau protein.
Also, it was found that use of Lobind tubes did not affect the measurement of either
marker.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S67

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

A-239
Using Factorial Design-of-Experiments (DOE) to Investigate Interactions
among Pre-Analytical & Analytical Factors in the Laboratory

C. Hild, R. Paulsen, K. Whigham. Aegis Sciences Corporation, Nashville,

TN
Loss of cannabinoids from solution during processing and analysis has been shown.
Roth et al (1996) showed the impact of storage and kinetic conditions on THCCOOH loss due to analyte binding to surfaces. Stout, Horn, and Lesser (2000) show
the contribution of storage temperature to THC-COOH loss in urine specimens. In
addition to the factors commonly investigated, we investigate the interactive effects
of storage, sample and processing factors on THC-COOH loss. The first behavior
investigated is a consistent low bias of 10-12%; the second behavior studied is a less
frequent, 30-60% loss of analyte encountered in some patient samples. Erratic losses
of 30-60% was observed in less than 0.2% of specimens tested. Both pre-analytical
as well as analytic factors, and the interactions amongst these factors, were theorized
causes of these observed losses. In order to understand the causes of the bias as well as
the outliers in measured concentrations, six handling, storage, and processing factors
were investigated.
Methods The investigation of factor effects and interactive effects on the THCCOOH loss during storage, handling, and processing conditions is studied using a
sequence of investigative studies. The impact of dilution method on THC-COOH loss
was investigated across four analytical chemists. Each chemist prepared replicate
dilutions by two different methods on a non-glucuronidated solution fortified to 100
ng/mL concentration. The first method prepared a dilution at a 2 mL exact volume;
the second method removed 500 mL of excess solution after dilution preparation.
Graphical analysis of variance (ANOVA) was used to investigate the impact of the
dilution method on percent loss.
The outlier losses in THC-COOH observed during urinalysis by GCMS were studied
using a blocked, fractional factorial design of experiment (DOE). These six factirs
were: initial specimen state (thawed vs. never frozen), storage conditions (light vs.
dark), vortexing, pipette transfer, glucuronidated vs. non-glucuronidated specimens,
and temperature (42oF vs. Room Temperature). 64 samples were tested across all
levels of the factors. Impact on THC-COOH concentrations were evaluated through
a least squares screening analysis that included both main effects and all pair-wise
interactions.
Conclusions 1. The dilution method had a statistically significant effect on % loss.
The dilution preparations that required removal of excess solution resulted in 40-50%
loss of THC-COOH. This effect reproduced across all four chemists who performed
dilutions.
2. The experiment showed three factors and two interaction effects as statistically
significant. Non-glucuronidated specimen exhibited greater losses across the 64
samples with a maximal loss of 40%. Non-glucuronidated samples lost 20-30% more
THC-COOH than did the gluceronidated samples. Transfers had the greatest effect on
non-glucuronidated samples and had an additive effect with vortexing. Exact-volume
pipetting reduced the impact of vortexing.
The control of kinetic and handling conditions is essential in order to reduce the
loss of THC-COOH during normal processing of urine specimens. The relationships
amongst the sample handling factors with state of conjugation provide insight into
causes of THC-COOH loss in GC-MS analysis.

A-240
Developing a Cutoff for Urinalysis of Bloody Urine

S. Albahra, K. Born, R. Martinez, W. B. Furmaga. UTHSCSA, San Antonio,

TX
Background: When bloody urine is submitted to the lab for testing, the most common
procedure is to spin it down and test the clear supernatant. Clarity of supernatant
is visually determined by technologist which caries a subjective bias. Additionally,
plasma contents may significantly alter the results of testing, despite passing a visual
inspection. In our project we were looking for objective criteria that define bloody
urine which is acceptable for laboratory testing.
Methods: Both, hemolyzed whole blood (n=11) and fresh, non-hemolyzed blood
where mixed with urine at known decreasing concentrations. After spinning down,
the supernatants were tested by dipstick and by chemistry analyzer. The precipitants
were submitted for microscopic examination.
Results: We found that blood has a significant impact on urinalysis even when it
accounts for 0.8% of the urine volume. In this mixture, dipstick showed (+3) blood

S68

concentration, and the total protein as well as albumin level went up by 448% and
2240% respectively, when compared to the concentration in the urine. Significant
differences were also noted in microscopic examination. Similarly, the mixture of
hemolyzed whole blood and urine at 0.8% concentration showed an increase of
1762% in the total protein and 4560% in the albumin. The mixture of urine with blood
at 0.08% concentration showed (+2) blood concentration by dipstick and an increased
concentration of total protein by 160% and albumin by 700%. Microscopic evaluation
was affected to a lower extent and the rest of the analytes tested within clinically
acceptable ranges.
Conclusion: We have demonstrated that dipstick can be used as a method to evaluate
acceptability of bloody urine for laboratory testing. A blood concentration of (+2),
which correspond to 0.08% of blood volume within the urine, may consist of
acceptable cut off for most of the analytes with exception of total protein, albumin and
RBC/WBC on microscopic evaluation. If the blood concentration within the blood/
urine mixture exceeds 0.08% the specimen is not acceptable for the majority of testing
such as total protein, albumin, creatinine, specific gravity, and microscopic analysis.

A-241
The Effect of Hemolysis and Lipemia on 23 Analyte Values Measured on an
Abbott c16000 Chemistry Analyzer

K. Doyle1, P. Bach2. 1University of Utah, Salt Lake City, UT, 2Intermountain

Healthcare, Murray, UT
Background: Medical laboratory test values may become erroneously elevated
or decreased due to interfering substances or endogenous contamination such
as hemolysis, icterus, or lipemia (HIL). Spectrophotometric-based assays are of
particular concern as these can be affected by interferents that absorb incident light
or inhibit transmitted light. The objective of this study was to evaluate the effects
of hemolysis and lipemia on several analytes as determined on an Abbott c16000
chemistry analyzer with particular attention as to how interference could affect low,
medium, or high baseline analyte values and how any observed changes could alter
the clinical treatment of a patient.
Methods: Hemolysate was prepared from discarded whole blood specimens and
Intralipid was obtained from the pharmacy. For each of 23 analytes to be measured,
residual serum or plasma samples containing low, medium, and high concentrations
of analyte were pooled to obtain a sufficient volume at a desired concentration.
Hemolysate or Intralipid was added to these samples prior to analysis to obtain a
final concentration of 65, 320, 1180 mg/dL hemoglobin (Hgb) or 50, 250, 1000 mg/
dL triglyceride, respectively. Analyte concentrations and indices for hemolysis and
lipemia were compared to unmodified specimens.
Results: Of the 23 analytes tested, 18 were partially or significantly altered by the
presence of hemolysate or lipemia. Deviation from baseline values was caused either
from increased analyte in hemolysate or Intralipid; or by interference of the spectral
quantification of the analytes. Results of significant clinical interest are summarized
below:
Interferent

Measured Analyte Significance of Interference

LDH
+ 1.2 U/L for every 1 mg/dL of Hgb
AST
+ 5.7% of the Hgb value (U/L AST, mg/dL Hgb)
Mg
+ 0.3-0.4 mg/dL per 100 mg/dL Hgb
K
+ 0.3 mmol/L per 100 mg/dL of Hgb
Hemolysate P
+ 0.2 mg/dL per 100 mg/dL of Hgb
Total protein
+ 0.3 g/dL per 100 mg/dL Hgb
Lactic acid
+ 1.0 mmol/L per 1000 mg/dL of Hgb
Direct bilirubin
Both Hgb and analyte concentration dependent
Triglycerides
+ 5 mg/dL per 100 mg/dL of Hgb
Mg
+ 0.1 mg/dL per 100 mg/dL of Intralipid
Total Protein
+ 1 g/dL per 300 mg/dL of Intralipid
Intralipid
Lactic acid
+ 1.1 mmol/L per 1000 mg/dL of Intralipid
Direct bilirubin
Both lipid and analyte concentration dependent
Triglycerides
Directly proportional
Conclusion: Our results provide a basis for predicting changes in analyte
concentrations coordinate with hemolysis and lipemia indices on an Abbott c16000
chemistry analyzer, thereby enabling laboratory personnel to (1) assess the quality of
the sample; (2) mitigate inaccurate test results; and (3) provide clinicians direction in
interpreting results of specimens harboring interferents.

A-242
How to reduce TAT delays in a large Molecular Biology Laboratory

V. T. Alstico, L. C. Scarpelli, V. D. T. Niewiadonski, A. Alfieri, C. F. A.

Pereira, N. Gaburo Jr.. DASA, Sao Paulo, Brazil
Background: Quality can be defined as the ability of a product or service to satisfy
the needs and expectations of the customer. Turnaround time (TAT) is one of the signs

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

of laboratory service and is often used as indicator of laboratory performance. To

improve the services quality, the TAT should be constantly monitored and the outliers
should have their causes investigated to avoid further delays.
Objective: The aim of this study was to measure and to reduce TAT delay in a large
Molecular Biology laboratory.
Methods: Statistical study was performed during the period from January to
December of 2013. In this period, the total number of tests reported and the number
of tests release with delay were verified, based on reports obtained from our internal
Laboratory Information System (LIS -Motion). TAT goals were determined according
to the complexity of each test. For tests based on manual procedures; the goal
was defined as 95% of the tests released monthly within the TAT established. For
automated tests, the goal was defined as 98%. From July to December, the following
actions were taken: meeting with the team to present the TAT data and mapping
the process to identify the main causes of delays, such as pre-analytical (errors in
patients input information, lack of documents needed for consistency analysis of
the results) and analytical causes (routine days not well established, lack of trained
labor, turnover). Additionally, some improvement alterations in the routine were done:
training, define routines priority, changing platform from manual to automated. The
data was monitored monthly using histograms for each test involved.
Results: Every month an average of 27.000 results are reported at DASAs Molecular
Biology laboratory. From January to June the average percentage of delays was
0,89%, according to a report generated by internal management system (Motion). In
the second semester, (July to December) the percentage of TAT delays was 0,49%. We
observed that changing manual tests to automated platforms and team training were
the most significant factors for reducing TAT delays.
Conclusion: In our experience, we were able to reduce the laboratory TAT delays in
45% by implementing strategies to improve the process and making people committed
to the services quality.

A-243
Bias, imprecision and uncertainty evaluation for some immunoassays

F. Yesildal1, T. Ozgurtas1, M. Serdar2, I. Kurt1. 1Gulhane Military Medical

Academy, Ankara, Turkey, 2Acbadem University, Faculty of Medicine,
Ankara, Turkey
Background: Bias (B), imprecision (I) and uncertainty (U) are important parameters
in evaluating the analytical performance of a laboratory. Internal quality control (IQC)
data can be used as an indicator of imprecision while external quality assessment
(EQA) data can be used to detect biases. We can calculate the uncertainty of a
measurement, using these data and other uncertainty resources. The purpose of
this study was to evaluate the analytical performance of immunoassays in our core
laboratory according to Fraser criteria.
Methods: B, I and U of 12 immunoassays were calculated according to EURACHEM/
CITAC Guide, by using the IQC and EQA data of the period between January 2013 and
December 2013. Cortisol, Estradiol, FSH, hCG, Insulin, LH, Prolactin, Parathormone,
FreeT3, FreeT4, TSH and Testosterone assays were used for evaluation of analytical
performance of immunoassay autoanalyzers - ADVIA Centaur XP (Siemens
Healthcare Diagnostics Inc., USA). Desirable specifications for allowable total error
(TE), I and B were used in evaluation according to Fraser criteria.
Results: Cortisol, estradiol, inslin, LH, prolactin, and TSH met the desirable
specifications for B, I and TE. hCG, FreeT4 and testosterone performance were the
worst of all and met none of the desirable specifications. Additionally, FSH and free
T3 had imprecision problem while parathormone had inaccuracy problem (Table
1). Conclusion: Immunoassays are difficult tests to achieve standardization. There
are many factors affecting the immunoassays and the amount of measurand is very
small, so accuracy and precision are very important. When it is not possible to meet
the optimum and desired specifications, minimum specifications may be considered.
Uncertainty does not mean error or doubt about the measurement, but it is about the
confidence of the result of measurements. Each laboratory should make this evaluation
to determine whether there are systematic or random errors.

Table 1: Evaluation of each immunoassay according to Fraser criteria.

Calculated Comment Calculated Comment Calculated Comment
Uncertainty (TE)
Bias
(B)
Imprecision (I)
Cortisol
27,84
PASS
10,52
PASS
9,11
PASS
Estradiol
17,31
PASS
6,16
PASS
6,08
PASS
FSH
18,37
FAIL
7,37
PASS
5,48
FAIL
hCG
38,01
FAIL
16,20
FAIL
9,94
FAIL
Insulin
23,73
PASS
9,55
PASS
7,04
PASS
LH
16,78
PASS
6,78
PASS
4,95
PASS
Prolactin
14,83
PASS
5,70
PASS
4,74
PASS
PTH
28,95
PASS
12,64
FAIL
7,04
PASS
Free T3
12,72
FAIL
3,65
PASS
5,20
FAIL
Free T4
22,10
FAIL
8,13
FAIL
7,48
FAIL
TSH
16,19
PASS
5,36
PASS
6,07
PASS
Testosterone 32,29
FAIL
12,56
FAIL
10,14
FAIL

A-244
Sample recollection, an experience in a hospital where samples are collected by
nurses, So Paulo, Brazil.

L. R. Almeida. DASA, So Paulo, Brazil

Background In recent years, there has been increasing interest in quality improvement
and patient safety activities in healthcare. The clinical laboratory has a leader in the
field of healthcare quality management with a focus on analytical quality born of its
scientific background and was one of the first areas to use quantitative statistical control
methods. However laboratories are now being asked to widen their focus to consider
activities outside their immediate control. Accreditation agencies are increasingly
requiring laboratories to go beyond analytical quality and take responsibility for the
pre- and post-analytical (or extra-analytical) phases where most errors arise. Blood
sample collections are performed by venipuncture for the implementation of this
procedure it is required that the professional is qualified, have technical training and
practical experience in nursing. During the analytical process it is identified that the
sample cannot be analyzed due same pre-analytical interference, the most frequent
causes are inadequate, insufficient material, hemolysis and clotted. Consistent final
result does depend only by the analytical process, a well collected sample is also
important. When the samples are collected by nurses it is expected a high number of
recollection because nurses are more used to the patient care, laboratory practices is
not part of the experience.
Methods: The laboratory as an outsource service follows the number of recollection
monthly and to keep the total number of blood recollection according with standards
recommendations of 2,0% is challenging once the procedure is hold by the nurse
team. All recollection we analyzed considering only reasons related with the technics
applied during the procedure. The causes of recollection are, inadequate material,
insuficiente, hemolysis and clotted samples.
Results: In 2013 the laboratory of a private hospital received 414.902 solicitation of
laboratorial tests, 0,7% of this total were recollected, 45% were from hemolysis, 38%
clotted, 12% insuficiente material and 3% inadequate material. 91.428 patients were
attended in this period.
Conclusion: Nurses that Work in the laboratory are more used to blood collection
than the team that work with patient care, so the laboratory must follow the procedure
done by the hospital and is responsible for improving these number of recollection by
stablishing a training program , and also avaliate if the program is effective because
the main reason of following these data is to reduce the impact to the patient.

A-245
Stability and clinical usefulness of thyroid hormones, TSH, GH, IGF1, BP3,
SDHEA, and cortisol results in serum frozen at -20C after long- term storage.

E. Kutasz, J. M. Lazzati, M. Maceiras, V. Zaidman, S. Rodriguez, M.

Rubinstein, E. A. Chaler. Hospital de Pediatra Garrahan, Buenos Aires,
Argentina
Background: Stored samples are commonly used. It would be necessary to verify
if samples stored for a given time at -20C in a serum bank maintain stability for
subsequent clinical interpretation.
We believe that the definition of stability should be different for long-term or shortterm storage conditions.
The concept of uncertainty (U) is useful for long-term storage conditions because
it considers all sources of error in a result. Error may be associated with method
precision (CVa%: analytical coefficient of variation) and with method accuracy (bias).
Collaborative Peer Group data are used to calculate the bias of the results. Therefore

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S69

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

U is calculated according to the following formula: U= 2(CVa)2 + biasmedia%2 +
(CVgroup/nlab)2. Where CVgroup is group coefficient of variation and nlab number of
laboratories in the Peer group.

by EDTA contamination due to the wrong order of draw. Spurious hypocalcemia

needs to be immediately recognized and appropriately interpreted in order to avoid
misdiagnosis and unnecessary intervention.

The reference change value (RCV: Zx 2 x (CVa2 + CVi2) Where Cvi is biological
coefficient of variation) shows a significant change if two consecutive results are
outside this range, which may be interpreted as a change in patient status.
Objective: To assess stability and/or clinical usefulness of the results of thyroid
hormones, TSH, GH, IGF1, BP3, SDHEA, and cortisol in serum stored at -20C for
8 months.
Materials: Statistically representative numbers of sera from pediatric patients were
evaluated according the following process T0: measured between 1 and 2 hours postextraction. T1: after 8-month storage at -20C. .
Results were considered stable and/or clinically useful if T1 was within the 95%CI
for U and RCV, respectively.
Results:
Analyte (n)
TSH (40)
T3 (40)
T4 (40)
fT4 (40)
GH (27)
IGF1 (27)
BP3 (27)
CORTISOL (21)
SDHEA (27)
Stable: S

U % (95%CIT1)
11.81 (NS)
14.39 (S)
12.18 (NS)
7.75 (NS)
15.82 (NS)
15.59 (NS)
There is not peer group
16.42 (NS)
17.27 (NS)

RCV % ( 95%CIT1)
58.6 (CU)
34.7 (CU)
28.3 (CU)
34.0 (CU)
20.0 (NCU)
20.8 (NCU)
23.0 (CU)
30.0 (CU)
28.6( CU)

Not Stable: NS
Clinically Useful: CU
Not Clinically Useful: NCU
Conclusions: Under these storage conditions only T3 remained stable and the results
could be used for reference range and/or diagnosis-related group values. GH and
IGF1 were not stable, and their results were not clinically useful. The remaining
results were not stable but were clinically useful.

A-246
Factual or fictitious hypocalcemia?

M. Chen, B. C. Handy, E. A. Wagar, Q. Meng. MD Anderson Cancer

Center, Houston, TX
Background: Calcium plays very important physiological functions. Measurement
of serum calcium helps to identify many clinical disorders. Accurate results play a
pivotal role in patient management. Unrecognized hypocalcemic emergencies can lead
to significant morbidity or death. On the other hand, misdiagnosed hypocalcemia can
result in inappropriate management and significant impact on patient care. Spurious
hypocalcemia is not uncommon in clinical laboratory practice, and distinguishing
false hypocalcemia from true hypocalcemia is essential. We report here two typical
cases of hypocalcemia.
Methods: Total calcium and ionized calcium were measured from the original sample
and redrawn blood samples to verify the results.
Results: Case 1 was a 19-year-old man with acute leukemia who underwent
leukophoresis in which acid-citrate-dextrose formula A (ACD-A) was used as the
anticoagulant. His initial blood chemistry results were as follows: serum calcium
7.9 mg/dL (normal range 8.4-10.2), ionized calcium 1 mmol/L (1.13-1.32), serum
magnesium 2.0 mg/dL (1.8-2.9), serum phosphorus 8.3 mg/dL (2.8-4.6), blood urea
nitrogen 25 mg/dL (8-20), and serum creatinine 1.26 mg/dL (0.70-1.30). The ionized
calcium measured by point of care testing was 1.04 mmol/L (1.12-1.32). Repeat
measurements from specimens collected next day showed consistently low total
and ionized calcium: serum calcium 6.6 mg/dL (8.4-10.2) and ionized calcium 0.84
mmol/L (1.13-1.32) on chemistry analyzer. Case 2 was a 10-year-old boy with multiple
endocrine neoplasia type 2 whose calcium level was initially unmeasurable. Calcium
measurement was repeated several times from this sample and the results were the
same. Magnesium and iron also were reported as undetectable from this sample, while
potassium level was greater than 10.0 mmol/L. Specimen was recollected into a serum
separator tube, and analysis of this sample revealed a calcium level of 9.4 mg/dL and
potassium level of 4.5 mmol/L.
Conclusion: The low total and ionized calcium results in Case 1 are suggestive
of factual hypocalcemia. The low calcium is presumed to be due to the chelation
of calcium by citrate in the ACD-A during dialysis. Citrate-induced hypocalcemia
is a major side effect of dialysis. Case 2 is clearly a fictitious hypocalcemia caused

S70

A-247
Laboratory investigation of spurious hyperkalemia

M. Chen, B. C. Handy, E. A. Wagar, Q. Meng. MD Anderson Cancer

Center, Houston, TX
Background: Pseudohyperkalemia is a laboratory artifact that is induced during
the process of specimen collection, transportation, and preparation. The most
common causes for spurious hyperkalemia include in vitro hemolysis, excessive
tourniquet time or fist clenching during phlebotomy, contamination with potassium
ethylenediaminetetra-acetic acid (K2EDTA), and specimens collected from patients
with hematological disorders such as leukocytosis and thrombocytosis. Factitious
hyperkalemia occurs frequently in laboratory practice. Here we report two
representative cases that presented with plasma potassium greater than 10.0 mmol/L,
illustrating the investigation process for hyperkalemia.
Methods: Various types of peripheral blood specimens were collected. Repeat
measurements of potassium were performed on different analyzers including Vitros
Fusion 5.1, blood gas analyzer IL Premier 3000, and POCT. The results of these
analyses were compared.
Results: Case 1 was an 84-year-old woman with chronic lymphocytic leukemia
(leukocytes 181,300/L). Her plasma potassium levels were persistently elevated
(>10.0 mmol/L). Repeated measurements of potassium on Vitros Fusion 5.1 showed
a potassium level >10.0 mmol/L in lithium-heparin plasma with or without gel but a
potassium level of 3.6 mmol/L in serum. A sample collected into a lithium-heparin
balanced syringe and analyzed by blood gas analyzer IL Premier 3000 showed a
potassium level of 3.6 mmol/L, while a sample collected simultaneously in lithiumheparin showed a level of 11.1 mmol/L on Premier 3000. The patients potassium
level as determined by POCT was 2.6 mmol/L. Case 2 was a 10-year-old boy with
multiple endocrine neoplasia type 2 who presented with a potassium level greater than
10.0 mmol/L. Surprisingly, the levels of calcium, magnesium, and iron in this sample
were undetectable. In another specimen collected separately in a serum separator tube,
the potassium was reported as 4.5 mmol/L and the calcium 9.4 mg/dL.
Conclusions: Severe hyperkalemia is a potentially life-threatening condition;
immediate recognition and appropriate interpretation are critical. Case 1 is clearly
a pseudohyperkalemia attributed to heparin-induced cell lysis in leukocytosis,
whereas case 2 is confirmed due to contamination of K2EDTA during specimen
collection not following the order of draw. Venous blood collected in a lithiumheparin balanced syringe and analyzed by blood gas analyzer is recommended for
potassium measurement in patients with leukocytosis. Standard operating procedures
and order of draw should be followed for specimen collection. An investigation
algorithm needs to be developed and the laboratory should follow the algorithm when
pseudohyperkalemia is suspected.

A-248
Case Report: Multiple False Electrolyte Abnormalities In A Patient with
Primary Biliary Cirrhosis Due To Extreme Hypercholesterolemia

M. S. Farooqi1, L. D. Racsa1, K. Jessen2, S. Bosley2, S. Nadkarni2, I. A.

Hashim1. 1University of Texas Southwestern Medical Center, and Parkland
Health & Hospital System, Dallas, TX, 2Parkland Health & Hospital
System, Dallas, TX
Background: Primary biliary cirrhosis (PBC) is an inflammatory, likely autoimmune,
liver disease marked by the destruction of intrahepatic bile ducts. Common features of
PBC include hyperbilirubinemia, pruritus, and elevation of plasma lipids, especially
total cholesterol (TC). 96% of symptomatic patients are reported to have TC
concentrations greater than 200 mg/dL, with an average of 377 mg/dL and a range of
77 mg/dL to 1035 mg/dL. We report a case of PBC where the finding of hyponatremia
led to the discovery of a plasma TC level > 2000 mg/dL.
Methods: All measurements were made on patient plasma utilizing a Roche cobas
c501 instrument using manufacturer-supplied reagents and instructions unless
otherwise specified.
Case Report: A 43 year-old female with PBC was admitted to the hospital after
outpatient laboratory tests showed hyponatremia. Her complaints on admission
included blurry vision, nausea, and significant pruritus. Physical exam was remarkable
for scleral icterus; no xanthomas were noted. Relevant abnormal laboratory results

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

were: sodium, 121 mmol/L (reference range: 135-145); potassium, 3.0 mmol/L (3.65.0); chloride, 87 mmol/L (98-109); and total bilirubin, 10.0 mg/dL (0.2-1.3). She
was started on intravenous (IV) fluids (0.9% sodium chloride). Subsequently, a lipid
panel was ordered. Her previous lipid panel from 1.5 years ago showed a plasma
TC concentration of 322 mg/dL (120-199). Her current lipid panel showed a TC
concentration of 2156 mg/dL; triglycerides, 226 mg/dL (50-150); and HDL, 37 mg/
dL (45-65). The LDL was unreportable. This was the highest total cholesterol value
measured by our laboratory. An investigation took place to determine if this was an
erroneous result. There was no evidence of contamination. The sample appearance
was clear and not grossly viscous or lipemic. The hemolysis index of the sample was
3, icterus index, 11, and lipemic index, 73. No significant interference is expected
below indices of 700, 14, and 2000, respectively, per the manufacturer. Furthermore,
serial dilution of the specimen indicated no interferences. A second sample from the
patient was obtained and showed a TC value of 2415 mg/dL; triglycerides, 299 mg/
dL; and HDL, 42 mg/dL (LDL, unreportable). Treating these as accurate values,
we investigated whether the patients sodium concentration was falsely low due to
hyperlipidemia. Direct ion-selective electrode measurement of the initial sample
on a Radiometer ABL825 FLEX analyzer showed a sodium concentration of 141
mmol/L (137-145); potassium, 4.4 mmol/L (3.6-5.5); and chloride, 105 mmol/L (101111). Serum lipoprotein electrophoresis done at a reference laboratory confirmed the
elevated TC level at 2295 mg/dL and found the major component to be lipoprotein X.
Conclusion: This case is important due to the degree of hypercholesterolemia, lack
of lipemic sample appearance, and its link to multiple false electrolyte abnormalities.
To our knowledge, there is only one other report of a PBC patient with a cholesterol
level > 2000 mg/dL, and it, too, was associated with pseudohyponatremia significant
enough to prompt clinical action. In this case, the clinical team stopped treatment with
IV fluids upon learning of the patients hyperlipidemia, and the patient was eventually
discharged.

A-250
Biological Variation as a goal for comparability of thyroid stimulating and freethyroxine hormones measurements in patients

J. Diaz-Garzon, V. Parrillas, P. Fernandez-Calle, R. Mora, M. Duque, J.

Iturzaeta, R. Gomez-Rioja, A. Buno. Hospital Universitario La Paz,
Madrid, Spain
Background: Laboratory testing of serum thyroid stimulating hormone (TSH) and
free-thyroxine (FT4) are essential for the assessment of thyroid function. However,
changes of methodologies could limit the application of clinical guidelines depending
on the standardization status. The IFCC Working Group for Standardization of
Thyroid Function Tests reports no significant differences in healthy individuals
within TSH assays (when calibrators are traceable to WHO international standards).
They used 10% as harmonization goal that is considered the current state-of-theart of immunoassay comparability. The same occurs with FT4 quantification if
methodologies are traceable to reference method.
Objective: To assess the comparability of patients results between two TSH and FT4
immunoassays using Biological Variation (BV) criteria as comparability goal.
Material and methods: Following CLSI EP9-A2IR protocol, TSH and FT4 from 40
randomly selected patient serum samples were performed in two analysers, Elecsys
(Roche) and Advia Centauro XP (Siemens) by duplicate. Both TSH methods were
traceable to second and third WHO international standards (81/565 and 80/558)
respectively. FT4 methods were traceable to different standards: Elecsys was traceable
to the reference measurement procedure based on equilibrium dialysis and Centauro
was traceable to internal standard using United States Pharmacopoeia material.
Analytical imprecision of all methods were within desirable limits based on BV. BV
desirable bias was set as analytical goal: TSH 7.82 % and FT4 3.34 %. Confidence
interval at 95% (CI) of the predicted difference was compared to laboratory allowable
bias at medical decision levels (MDL).
Results: Table shows the main results of the study.
TEST
Units
MDL
(Assayed range)
TSH IU/mL 0.27
4.20
(0,37-4,71)
FT4 ng/dL
0.90
1.80
(0,87-1,63)

Allowable BV
desirable bias
-0.10 to 0.19 0.02
-0.86 to -0.55 0.33
0.06 to 0.13 0.03
-0.24 to -0.13 0.06

Difference CI

Goal achieved

0.05
-0.71
0.1
-0,18

Not
Not
Not
Not

Conclusions: Despite standardization efforts, applying biological variation criteria,

comparability of TSH and FT4 results could not be guaranteed in patients. When a

change of methodology occurs, laboratory should always perform studies in order

to alert clinicians about the possible non interchangeability of patients results with
possible impact in their follow-up.

A-251
PATHFAST Presepsin (sCD14-ST) Sample Matrix Evaluation

E. Spanuth1, M. Preusch2, R. Thomae3, E. Giannitsis4. 1DIAneering Diagnostics Engineering & Research, Heidelberg, Germany, 2Department
of Internal Medicine III, Intensive Care Unit, University Hospital
Heidelberg, Heidelberg, Germany, 3Mitsubishi Chemical Europe,
Dsseldorf, Germany, 4Department of Internal Medicine III, University
Hospital Heidelberg, Heidelberg, Germany
Background Soluble CD14 is released from monocytes during activation by TLR4specific inflammatory reaction against infectious agents yielding presepsin (sCD14ST). Presepsin demonstrated powerful diagnostic and prognostic information in
critical ill patients with infectious-inflammatory diseases. The objective of our study
was to examine the suitability of different sample types for presepsin determination.#
Methods Whole blood samples were collected from 20 patients in serum, lithium
heparin, K2 EDTA and Na3 citrate blood collection tubes (S-Monovette, Sarstedt,
Germany). The patients were hospitalized with different degrees of infectiousinflammatory conditions at the intensive care unit (ICU).Serum and plasma samples
were prepared from each whole blood sample by centrifugation (20 minutes at
2500 x g) and separation of plasma/serum within 2 hours after blood drawing. The
native serum and plasma samples were measured using the PATHFAST Presepsin
assay. For estimation of the inflammatory status C-reactive protein (CRP) values
were determined in the heparin plasma samples using the cobas CRP assay (Roche
Diagnostics).
Results No considerable differences of the presepsin concentrations measured in the
different sample matrices were observed. The samples spanned a significant portion
of the measurement range of the test. EDTA plasma samples ranged from 161 to 7400
pg/ml. The measurement of presepsin in the different sample matrices revealed CVs
between 2.11 and 12.34 % showing a high concordance between the samples. The
results are summarized in Tab. 1.
Tab. 1: Presepsin concentrations (ng/L) measured by using PATHFAST
Presepsin in different sample matrices
N
Mean
Median
Minimum Maximum IQR
Serum
20
1715
618
169
7251
273-2453
Heparin
20
1749
635
182
7335
277-2610
plasma
Na-citrate
20
1614
559
194
6997
265-2225
plasma
EDTA
20
1715
632
161
7400
223-2373
plasma.
Conclusion The results demonstrated excellent comparability between the different
sample matrices across the full measurement range suggesting that PATHFAST
Presepsin may be useful for risk stratification of critical ill patients from inflammatory
diseases in the emergency room and point-of-care setting.

A-252
Reference Intervals of Many Common Chemistry Tests are Affected by
Advancing Age in Adults

A. K. Fzry1, Y. Qiu2, M. La2, G. S. Cembrowski2. 1Alberta Health

Services, Edmonton, AB, Canada, 2University of Alberta Hospital,
Edmonton, AB, Canada
Objectives: More than 40 million Americans are over 65 years of age and their number
will grow rapidly over the next 40 years. Current laboratory reference intervals have
age-specific ranges for some tests but only for infants, children, and non-senior adults.
Not enough is known about the effect of aging on many tests and geriatric reference
intervals are not widely available. This study aims to determine the effect of advanced
age in adults on chemistry reference intervals.
Methods: We compiled the results of common chemistry tests and pertinent
demographic data of 55 to 85 year old US volunteers sampled in 4 cycles of the US
National Health and Nutrition Examination Survey (2005-2006, 2007-2008, 20092010, 2011-2012). Reference interval diagrams were constructed with the 97.5, 95,
90, 50, 5, and 2.5th percentiles plotted against age intervals: 55-60 years (n=64 males,
n=45 females), 60.1-65 years (53 M, 72 F), 65.1 to 70 years (50 M, 63 F), 70.1 to 75
years (50 M, 65 F), 75.1 to 80 years (51 M, 57 F), and 80.1 to 85 years (94 M, 83 F).

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S71

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

Only non-Hispanic whites with a waist circumference less than 105 cm (male) or 100
cm (female) were included in this study.
Results: The Figure shows a very interesting reference interval diagram. The
increase in BUN with age is clearly apparent. We observed similar patterns for
creatinine, potassium, osmolality, and total bilirubin. In contrast, we observed inverse
relationships with age for albumin, ALT, and cholesterol. We found only minimal agedependent relationships for ALP, AST, calcium, chloride, globulins, total CO2, iron,
LD, sodium, phosphorus, and total protein.
Conclusions: Advancing age in the adult population affects many common chemistry
tests. Our results suggest that it will be necessary to include age-appropriate reference
intervals when reporting test results for geriatric patients.

biochemical cascade of the CTNI assay. The stability of the blank ensured a much
higher reproducibility of the results and reduced the need for frequent calibrations
of the assay.
Recommendations for the maintenance of the water purification systems are described
to ensure a consistent supply of clean water to feed clinical analyzers, in-line with
CLSI C3-A4 guideline.

A-254
Comparison of RBC hemolysis according to plasma and serum separator tubes
among outpatient specimens

D. Ko, D. Won, T. Jeong, W. Lee, S. Chun, W. Min. University of Ulsan

College of Medicine and Asan Medical Center, Seoul, Korea, Republic of
Background: In our laboratory, we use plasma separation tubes (PST) for
chemistry analysis to obtain outpatient results rapidly. If lactate dehydrogenase (LD)
measurement is required, serum separation tubes (SST) are used. PST can be used
immediately after centrifugation with no need for clotting, and a greater volume of
blood can be obtained. Although many studies have compared the results of blood
chemistry analysis using PST and SST, there has been no evaluation of hemolysis
using these types of tube. Herein, we compare the hemolytic index obtained using
PST and SST, and apply this to quality control in the laboratory.
Methods: We analyzed the hemoglobin index of outpatients visiting the Asan Medical
Center (Seoul, Korea) from January to December 2012. The hemolytic index, a
quantitative serum index, was scored from 0 to 10 according to the concentration
of hemoglobin (0-5.0 g/L) using the Toshiba-200FR automated instrument (Toshiba
Medical Systems Co., Tokyo, Japan). Hemolytic index was classified by sample tube
type (PST or SST), and significant hemolysis was defined as a hemolytic index of 2
or greater. In cases of significant hemolysis, electronic medical records were reviewed
to identify the cause.

A-253
Keeping Bacteria Under Control to Minimize Impact on Assays and Maximize
Analyzer Uptime

M. Gauthey Baraou, S. Mabic. Merck Millipore, Saint-Quentin-en-Yvelines,

France
Low bacteria count in pure water is particularly critical in clinical analyzers, because
bacteria can generate numerous interferences in biochemistry and immunochemistry
assays. The objective here is to describe several of those issues on assays and analyzer
maintenance, and provide solutions to avoid bacterial contamination in water
supplying the analyzer.
Typical impacts of bacterial contamination on assays include unstable calibrations,
high absorbance of blanks, reference drifts, and errors on mean patient values. Those
effects are observed, for instance, on the Arsenazo calcium assay, the potassium
potentiometric assay, as well as immunoassays involving alkaline phosphatase (e.g.
CTNI, fluorescence 6-MUP- and AMPPD-based assays). Those effects generated
by the typical bacteria strains identified in clinical analyzers (Ralstonia pickettii,
Sphingomonas paucimobilis, Caulobacter crescentus) result from proteins and small
organic acids released by bacteria.
Bacteria also have an effect on maintenance of the analyzer: an incomplete rinsing
generates interferences with the 340 nm detection of assays using NAD/NADH
chemistry and the frequency of sanitization can be significantly increased. Maintaining
low bacteria level in the analyzer and the water purification system supplying the
instrument can reduce downtime and minimize the risks of false results.
Some of the issues mentioned above and resulting from poor design can be avoided
by selecting key purification technologies, such as ultrafiltration and germicidal 254
nm UV lamps. A reduction of two logs is observed on bacteria levels of water treated
by UV mercury lamps (ca 100 cfu/mL to < 1 cfu/mL, using CLSI recommended
bacteria culture procedure). Blank variation in the CTNI assay was shown to decrease
over a 30 day period, from a range of 18 to 52 mAU to a range of 20 to 26 mAU
by adding an ultrafiltration step to the water treatment process. The ultrafiltration
removes the alkaline phosphatase that is released by bacteria, and interferes with the

S72

Results: Significant hemolysis was found in 0.66% (1,128 of 171,519) of the total
specimens, 0.68% (1,051 of 154,886) of PST specimens, and 0.46% (77 of 16,633) of
SST specimens. The mean hemolytic index in PST was 0.18 (SD: 0.43), which was
significantly greater than that in SST (0.14, SD: 0.37) (P<0.001). The proportion of
significant hemolysis was also higher in PST than in SST (P=0.001). The cause of
significant hemolysis was identified in 48.1% (543 of 1,128) of the specimens; the
causes of in vivo hemolysis were chemotherapy and prosthetic valve. The remaining
hemolysis specimens (51.9%, 585 of 1,128) may have been due to complex sampling
errors. Hemolysis of unknown cause was particularly common in pediatric samples.
Conclusion: The incidence of hemolysis was slightly higher using PST compared
to SST, although both were <1%. We conclude that PSTs are thought to be more
useful than SST in outpatient testing because they offer more rapid turnaround and
can contain a greater sample volume in our laboratory.

A-255
The effect of storage at room temperature on the measures of lactate and
pyruvate in cerebrospinal fluid with and without blood contamination

L. Zhang1, J. Gabler2, D. Payto3, M. Na1, S. Wang3. 1Cleveland State

University, Cleveland, OH, 2ThermoFisher Scientific, San Jose, CA,
3
Cleveland Clinic, Cleveland, OH
Background: Measurement of pyruvate and lactate, intermediates of carbohydrate
metabolism, in cerebrospinal fluid (CSF) is important in evaluating disorders of the
central nervous system. However, the stability data of these analytes in CSF is scarce
in literature, especially for pyruvate. Objective: To assess the in-vitro stability of
pyruvate and lactate at room temperature (RT) in CSF specimens with and without
blood contamination. Method: Blood contaminated and non-contaminated CSF
specimens were collected for this study. Separate pools were made for contaminated
and non-contaminated samples. Unspiked pools were used for low concentration
levels (0.09mM pyruvate and 1.9mM lactate) and high levels were spiked at 0.22mM
pyruvate, 3.5mM lactate. Pools were stored at RT over 120h while duplicate samples
were aliquoted at each of 6 time points (0h, 3h, 6h, 24h, 52h, 120h). Samples were
deproteinized (12% trichloroacetic acid) and analyzed by enzymatic assays for
both lactate and pyruvate. Results: Pyruvate concentrations were constant up to
6h in non-contaminated CSF samples at both concentrations (<5% decrease). The
concentrations of pyruvate progressively decreased after 6h, reducing up to 45% after
120h and showing a greater instability at the low concentration level. In contaminated
samples (hemolytic index=38), pyruvate concentrations showed no significant change
through 120h (<3% decrease) at both high and low levels. At low concentration,

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

lactate concentration was constant through 48h in both contaminated and noncontaminated samples. High lactate concentration remained constant through 6h.
Conclusion: Pyruvate and lactate at physiological concentrations in CSF specimens
with and without blood contamination remain constant at RT up to 6h.

A-256
Performance Evaluation of Olympus AU2700 Plus by Six-Sigma Using Three
Different Internal Quality Control Materials

C. Ylmaz Demirtas1, M. Kocabyk1, F. Gucel2, S. Elbeg1, O. Gulbahar1.

1
Gazi University Medical Faculty, Ankara, Turkey, 2Etlik Zubeyde Hanm
Gynecology Training and Research Hospital, Ankara, Turkey

Table 1: Calculated six-sigma values of three different QC materials.

QC material
Six-sigma >6
Six-sigma <3
Glucose, creatinine, uric
acid, total bilirubin, ALT,
ALP, total cholesterol,
Beckman Coulter
triglyceride, calcium,
Sodium, chloride, and total protein
level 1
potassium, magnesium,
amylase, creatine kinase,
and GGT
Creatinine, uric acid,
total bilirubin, AST, ALT,
Total protein, sodium, chloride,
Beckman Coulter
ALP, triglyceride, LDH,
level 2
and GGT
magnesium, amylase, and
creatine kinase
Uric acid, total bilirubin,
triglyceride, calcium,
ALT, LDH, and sodium
BioRad level 1
potassium, magnesium,
and amylase
Uric acid, ALP,
triglyceride, potassium,
Urea, total bilirubin, AST, and
BioRad level 2
chloride, magnesium,
sodium
amylase, and creatine
kinase
Uric acid, total bilirubin,
total cholesterol,
Glucose, urea, AST, ALT, LDH,
Randox level 1
triglyceride, magnesium,
sodium, and chloride
amylase, and creatine
kinase
Albumin, uric acid,
ALP, total cholesterol,
Randox level 2
triglyceride, magnesium, Glucose, LDH, and sodium
amylase, and creatine
kinase

A-257

Background:The analytical performances of the instruments in the medical

laboratories should be satisfactory. Six-sigma level is a measurement of quality in
the evaluation and comparison of performance. We evaluated the sigma levels of the
parameters in our comprehensive metabolic panel in a single instrument by using
different internal quality control (QC) materials.

Bilirubin oxidase resolves bilirubin interference in a colorimetric

acetaminophen assay

Methods:The chemistry instrument evaluated was Olympus AU2700 Plus (BeckmanCoulter, Brea, CA, USA) and the QC materials provided were Beckman Coulter
(Beckman-Coulter, Brea, CA, USA), BioRad (Bio-Rad Laboratories, Hercules, CA,
USA) and Randox (Randox Laboratories, North Ireland, UK). Our comprehensive
panel included glucose, creatinine, uric acid, total bilirubin, alanine aminotransferase
(ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), total
cholesterol, triglyceride, calcium, potassium, magnesium, amylase, creatine kinase,
gamma-glutamyl transferase (GGT), uric acid, total bilirubin, direct bilirubin, lactate
dehydrogenase (LDH), sodium, chloride, albumin and total protein. We measured
internal QC materials at two levels for 20 days consecutively. If a measurement
was outside two standard deviation range then a rerun preformed. Analytical
reproducibility assessed using the CLSI EP-5 protocol. CLIA criteria were the basis
for the allowed total error values except for two parameters, which are GGT and
direct bilirubin.

Background: Acetaminophen is a pain reliever found in over-the-counter and

prescription medications. Acetaminophen toxicity is the most common cause of
drug overdose and acute hepatic failure in the US. Measurement of acetaminophen
levels in blood, once a suspicion of toxic ingestion has been established, is crucial
for risk assessment, treatment and management. However, samples with high levels
of bilirubin (icteric), which are very common in individuals with drug-induced
hepatotoxicity, interfere with acetaminophen measurement in most commercial
clinical assays and can lead to false positive results. Bilirubin oxidase (BOx) is an
enzyme that catalyzes the oxidation of bilirubin to biliverdin. Objective: In this
study, our objective was to explore the potential use of BOx as an additive to samples
with high bilirubin concentrations to resolve acetaminophen interference. Methods:
In experiment 1, pooled non-icteric heparinized plasma samples were spiked with
acetaminophen and aliquoted. Aliquots were then spiked with acetaminophen-free
samples containing different amounts of bilirubin. Acetaminophen and Icteric index
(I-index) were measured at baseline and after a 30min incubation step with BOx
(1 U/ml) at 37C. In experiment 2, highly icteric patient samples (n=9, I-index>8)
were analyzed before and after incubation with BOx (2 U/ml) at room temperature
for 30min and 60min. All samples were analyzed using the Roche Integra 800, for
which the manufacturer reports interference to acetaminophen with an I-index>4.
Assay limit of detection is 15ug/mL. Results: See table. Experiment 1 demonstrated
that BOx effectively removes bilirubin without affecting acetaminophen. Experiment
2 demonstrated that under normal lab conditions, BOx may be used to resolve false
positive acetaminophen results due to elevated bilirubin. Data for 60min incubation
not shown; all acetaminophen levels at 60min <15ug/mL. Conclusion: BOx can be
used to resolve false positive acetaminophen results due to bilirubin interference. This
suggests a potential role for BOx to resolve bilirubin interference in other clinical
chemistry assays.

Results:Our data is expressed in the table with two cutoff values: six-sigma of more
than six and less than three.
Conclusion:Health services aim zero error but being under the influence of many
variables it is difficult to achieve this goal. Six-sigma methodology is useful in the
planning of laboratory quality control process. The sigma level calculated for each
test may provide insight about the quality. We found that the calculated sigma levels
using different QC materials are similar in some parameters but different in others. We
support the idea of using appropriate internal quality control material for each test and
this procedure can guide in reducing the cost of poor quality.

A. Abbadi, J. M. El-Khoury, E. Z. Reineks. Cleveland Clinic, Cleveland,

OH

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S73

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

Experiment 1

Before BOx incubation

AcetamiI-index
nophen
(ug/mL)
Baseline Non-icteric 22
Sample 1 1.4
22
Sample 2 2.4
22
Sample 3 4
22
Sample 4 7.7
23

After BOx incubation

(30 min at 37C)
AcetamiI-index
nophen
(ug/mL)
Non-icteric 21
0.5
20
0.7
21
0.9
20
1.4
19

Experiment 2

Patient 1
Patient 2
Patient 3
Patient 4
Patient 5
Patient 6
Patient 7
Patient 8
Patient 9

Before BOx
incubation
AcetamiI-index nophen
(ug/mL)
21.6
<15
15.8
<15
9.2
<15
39.5
29
40.2
24
8.9
<15
17.8
<15
16.4
<15
14.7
<15

After BOx incubation

(30 min at RT)
AcetamiI-index nophen
(ug/mL)
8.8
<15
7.4
<15
6
<15
15.2
<15
12.7
<15
4.4
<15
5.5
<15
8.6
<15
6.1
<15

A-258
Case Report: Paternity case with three genetic incompatibilities

D. Moratori1, J. P. G. Batista1, M. Malaghini2, L. C. Scarpelli1, R. B.

Moraes1, O. Fernandes1, N. Gaburo Jr.1. 1DASA, Sao Paulo, Brazil, 2DASA,
Paran, Brazil
Background: The occurrence of germ line mutations at microsatellite loci may
cause problems in ascertaining non-fatherhood status in paternity testing. A mutation
event of DNA markers caused by loss or gain of repetitive units is quite a common
phenomenon in forensic practice and is becoming more frequent due the increased
number of paternity tests performed worldwide.
Objective: To describe a case that revealed a potentially erroneous exclusion from
paternity due to the presence of three genetic inconsistences between the alleged
father and the child.
Methods: The case was performed during the routine of paternity test at DASA. The
DNA from mother, child and alleged father were obtained from blood samples in FTAtreated cards. The punching process was automated with BSD600 Duet equipment
(Life Technologies, Foster City, CA). A total of 47 loci were analyzed. The STRs
(Short Tandem Repeats) markers amplification was performed using AmpFlSTR
Identifiler Direct PCR Amplification Kit (Life Technologies). Additional STRs were
performed using Powerplex 16 HS System (Promega Corporation, Madison, WI),
NGM Select (Life Technologies) and others 26 in house loci. The detection was
performed at 3130xl Analyzer (Applied Biosystems, Foster City, CA). Paternity index
(PI) was calculated including the mutation rates according to American Association of
Blood Banks (AABB) report.
Results: Out of 47 analyzed loci, three genetic inconsistencies between the alleged
father and the child were detected: two mutations at STRs D2S1338 e D2S1776 and
the presence of null allele at STR TPOX. The inconsistencies at STRs D2S1338
and TPOX were confirmed by 2 different commercial kits, while the STR D2S1776
was confirmed by co-amplification. The combined paternity index in the case was
19.897.397.674.899,85 which corresponded to final probability of paternity value
higher than 99.9999%.
Conclusion: The overall number of cases performed in our paternity laboratory, since
2005, is more than 38.000. This is the first case that we have found three genetic
inconsistences in a non-exclusion paternity investigation. This case emphasizes
the requirement that an exclusion from paternity must be based on calculating the
appropriate statistical estimations in every case.

A-259
Identification of key meta data to enable safe accurate and effective
transferability of biological variation data.

W. A. Bartlett1, F. Braga2, A. Carobene3, A. Cokun4, R. Prusa5, P.

Fernandez-Calle6, T. Rraas7, I. Leimoni8, S. Sandberg9. 1Blood Sciences,
Ninewells Hospital & Med School. Biological Variation Working Group,
European Federation of Clinical Chemistry and Laboratory Medicine
(EFLM), Dundee, United Kingdom, 2Luigi Sacco Univeristy Hospital,
Biological Variation Working Group, EFLM., Milan, Italy, 3Diagnostica
e Ricerca, San Raffaele Spa, Biological Variation Working Group, EFLM.,
Milan, Italy, 4Acibadem University, School of Medicine, Glsuyu, Maltepe.
Biological Variation Working Group, EFLM., Istanbul, Turkey, 5Charles
University and University Hospital Motol. Biological Variation Working
Group, EFLM., Prague, Czech Republic, 6Hospital Universitario La Paz.
Biological Variation Working Group, EFLM., Madrid, Spain, 7Norwegian
Quality Improvement of Primary Care Laboratories (NOKLUS),
Haraldsplass, Hospital., Bergen, Norway, 8Euromedic S.A. Biological
Variation Working Group, EFLM., Athens, Greece, 9Haukeland University
Hospital. Biological Variation Working Group, EFLM., Bergen, Norway

S74

Background: Biological variation data (BVD) are reference data with many
applications in laboratory medicine. Appropriate transfer of BVD across populations
and through time requires the user to have -knowledge of the characteristics of the
population from which the data were derived -an understanding of how the data were
derived and -an appreciation of the uncertainty that surrounds the reported estimates.
As a consequence an estimate of within and between subject biological variations
should be transmitted and adopted for use only if accompanied by a set of meta data
that sufficiently characterises the BVD in those contexts. The Biological Variation
Working Group (BVWG), set up by the European Federation of Clinical Chemistry
and Laboratory Medicine (EFLM), have undertaken work to identify a candidate
minimum data set (MDS) to accompany published indices of within and between
subject biological variations to enable this issue to be addressed.
Methods: The BVWG considered and discussed the content of published literature
and web based databases to identify the key meta data to accompany BVD to enable
safe accurate and effective transfer and application across populations and through
time.
Results: Key meta data were identified under six main headings. Those are, with
example subheadings: Target - analyte and measurand, sample matrix, method characteristics. Population
characteristics - demographics, state of well being, physical/physiological
characteristics, medication Study Characteristics - study duration and design, power
of study to detect BVD indices, model assumptions, and statistical approach. Data
Characteristics - indices of biological variability, confidence intervals, tests for model
assumptions. Publication Details - links to the original publication. Data rating
- new concept to be developed to indicate the quality of the BV data against a set
of key criteria. Conclusion: Published reviews of the literature describing BVD for
albuminuria, haemoglobin A1c, C-reactive protein and liver enzymes all indicate a
high degree of heterogeneity in the approach to derivation and reporting of BVD.
Published BVD are of varying quality, often poorly characterised and consquently
applied inappropriately into clinical laboratory practice. This highlights the need for
generation of recognised standards for these important data sets.
The working group believe that availability of a standardised minimum data set, as
proposed above, will enable users to be more objective in the transfer of published
BVD into their local and wider practice. This will prove challenging to deliver,
and require mechanisms to facilitate the extraction of meta-data from publications
for attachment to the BVD to enable onward transmission and transferability (e.g.
incorporation into databases). This will require further development of the concept
of a BVD data archetype incorporating internationally accepted coding systems (e.g.
SNOMED, LOINC) and vocabularies.
The MDS has been identified as part of a larger programme of work being undertaken
by the BVWG aimed at developing a critical appraisal checklist for papers containing
BVD. This will enable the creation of the data rating concept included in the MDS.

A-260
Evaluation of Technopath Controls on the ARCHITECT Family of Instruments

D. Brault1, A. Croce2, L. Lennartz3, M. Orth4, J. Shih5. 1Hopital Tenon, Paris,

France, 2Ospedale Civile Sondrio, Sondrio, Italy, 3Abbott Laboratories,
Wiesbaden, Germany, 4Vinzenz von Paul Klinikem, Marienhospital,
Stuttgart, Germany, 5Abbott Laboratories, Abbott Park, IL
Introduction: Quality controls are important for laboratories to ensure that released
results meet the required quality in regard to accuracy and precision. Consolidation of
controls is an important trend in laboratories to simplify QC testing. Recently, multiconstituent controls (MCCs) have been introduced by Technopath Manufacturing
Ltd (Ballina, Ireland) that cover a wide range of clinical chemistry and immunoassay
analytes.
Objective: The goal of this study was to evaluate the performance of the Multichem
S Plus, Multichem IA Plus and Multichem U controls on the ARCHITECT family
of instruments. Precision and accuracy compared to the target value were evaluated.
Methods: Three European sites (Paris, France; Stuttgart, Germany; Sondrio, Italy)
used the three controls for a minimum of thirty days in parallel with the labs routine
QC controls. Testing was performed on the ARCHITECT c8000, c16000, i1000SR
and i2000SR instruments. Data presented here are from the following serum clinical
chemistry analytes: (A)ALT, (A)AST, total bilirubin, chloride, total cholesterol,
creatinine (enzymatic and picrate), glucose, potassium, total protein, sodium,
triglycerides and urea; the following immunoassays: CA 19-9, CEA, total PSA, free
T3, free T4, TSH, troponin-I, total beta HCG, testosterone, estradiol and FSH; and
on the following urine assays: chloride, creatinine (enzymatic and picrate), glucose,
potassium, sodium and urea. The Multichem S Plus and IA Plus are serum based with
three control levels; the Multichem U is prepared from human urine with two control

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

levels. All data were collected via AbbottLink for automated data retrieval. Means,
standard deviations and ranges were calculated for all controls. Assay reagent lots and
calibrator lots varied across the sites and within the sites.
Results: The results from twelve frequently performed clinical chemistry assays
were analyzed. The %CV for the 12 assays with the Multichem S Plus control
ranged from 0.42 to 4.71% at the individual sites. The %CV for the 6 assays with
the Multichem U control ranged from 0.50 to 5.24% at the individual sites. For both
controls, the majority of the CVs were less than 2%. The results from these eleven
frequently performed immunoassays were analyzed. The %CV for the 11 assays with
the Multichem IA Plus control ranged from 1.82 to 14.94% at the individual sites;
however the majority of the CVs were less than 5%. Overall little variation was seen
instrument to instrument, site to site or reagent lot to reagent lot.
Conclusions: The Technopath S Plus, IA Plus and U controls demonstrated similar
performance to the routine internal laboratory quality controls. The use of these MCCs
reduce the number of controls required for the analytical quality control testing of
both clinical chemistry and immunoassay analytes with no compromise on quality.

Equivalency Testing of Serum and Plasma on the Siemens Dimension Vista

1500 System at University Hospitals (UH) Case Medical Center

C. Schmotzer1, T. Barker1, D. Yakubek2, M. B. Smith2. 1University Hospitals

Case Medical Center, Cleveland, OH, 2Siemens Healthcare Diagnostics,
Newark, DE
Background: UH Case Medical Center has 1032 beds and processes 6000 samples
a day and 5.2 million tests per year. Consistent and rapid turnaround time (TAT)
requires efficient processing and analysis of samples. Using plasma will eliminate
clotting time and decrease pre-analytical processing aiming to shorten TAT.
Objective: To demonstrate the equivalency between serum (x) and plasma (y) on
assays performed on the Dimension Vista System.
Methodology: Serum and lithium heparinized plasma samples were obtained from a
single draw in individual patients in the hospitalized population and tested in parallel.
Least squares regression analysis was performed on single determinations of serum
(x) and corresponding plasma (y) samples. The criteria used for assessing equivalency
were a slope of 0.90 to 1.10, an intercept that is clinically insignificant, a correlation
coefficient greater than 0.95 and average percent bias (plasma-serum) less than 10%
or the CLIA Total Allowable Error, whichever was greater.
Results: 46 assays (n=53-59 serum and corresponding plasma samples) evaluated
met the above criteria. Table 1 lists data observed on representative assays of different
method principles. The minimum and maximum slope, correlation coefficient and
average percent bias observed for 46 assays are 0.92-1.05, 0.976-1.000 and -8.7 to
5.3%, respectively. Intercepts for these assays were clinically insignificant.

Assay
1.
ALB
2.
A1AT
3.
ALTI
4.
ASL
5.
AST
6.
B2MIC
7.
FERR
8.
FT4
9.
GLU
10.
hsCRP
11.
IGG
12.
PBNP
13.
K

Units

Slope

Intercept

Correlation
n
Coefficient

Low

High

Average
% Bias
(plasmaserum)

g/dL

0.995

-0.037

0.977

56

2.1

3.9

-1.7

mg/dL

0.947

6.875

0.983

57

116

376

-2.0

U/L

0.977

-0.005

0.998

58

10

168

-2.3

U/mL

0.962

0.855

0.997

59

13

342

-2.9

U/L

1.020

-0.780

0.998

58

232

-0.3

mg/L

0.992

0.014

1.000

58

1.21

33.19

-0.4

ng/mL

0.978

-1.096

0.999

58

21.2

3108

-2.4

ng/mL

0.991

0.013

0.993

58

0.67

1.97

0.1

mg/dL

1.053

-0.010

0.991

53

69

208

5.3

mg/L

1.009

0.360

0.997

55

0.16

164.2

1.6

mg/dL

0.991

-4.038

0.997

57

307

2605

-1.4

pg/mL

0.986

18.027

1.000

56

25

14637

0.1

mmol/L 0.988

-0.072

0.976

56

3.2

4.6

-3.0

A-262
25-Hydroxyvitamin D2: Prevalence and Impact on 25-Hydroxyvitamin D
Measurement

Y. Zhang, T. Kwong, M. Stolla. Strong Memorial Hospital, ROCHESTER,

NY
Background: This study is to evaluate the distribution of 25OH-D2/D3 in a general
patient population in western New York to provide insights into the current prevalence
and the trend of vitamin D2 usage, and its impact on the accuracy of vitamin D
measurement.

A-261

Table 1

Other assays tested for serum vs. plasma equivalency: 1) ALP, 2) BUN, 3) CRP, 4)
CL 5) CHOL, 6) CKMB, 7) C3,
8) C4, 9) CKI ,10) DBI, 11) ECREA, 12) FT3, 13) GGT, 14) HAPT, 15) HDLC,
16) IGA ,17) IGM,18) IRON, 19) LDI, 20) LDLC, 21) LIPL, 22) MYO, 23) PHOS,
24) PREALB, 25) RF, 26) NA, 27) TBI, 28) TIBC, 29) TP, 30) TRF, 31) TRIG, 32)
TSH, 33) URCA
Conclusions The observed data on 46 assays support the equivalency of serum and
plasma on the Dimension Vista System based on the predefined evaluation parameters
to demonstrate serum and plasma equivalency.

Methods: 25OH-D2 and 25OH-D3 results measured by in-house LC-MS/MS

method at Strong Memorial Hospital at the University of Rochester Medical Center
from June 2009 to December 2012 were retrospectively analyzed. The results were
retrieved through the laboratory information system using specific parameters to
include patients age, gender, 25OH-D2, 25OH-D3, and total 25OH-D data. All
the results during that period of time were included except research samples and
the ones with missing data. The mean, standard deviation, overall distribution and
monthly prevalence of 25OH-D2, 25OH-D3, and total 25OH-D were calculated using
Statistica 10 Software. The significant levels were assessed using two tailed student T
test in EXCEL. The limit of quantification (LoQ) for 25OH-D2 and 25OH-D3 were
&lt 4ng/mL and &lt 6ng/mL, respectively. Total 25OH-D level of 60 ng/mL was
considered potentially harmful.
Results: A total of 266,269 samples from were included for analysis. The percentage
of samples with 25OH-D2 levels above assay LoQ decreased from 32% to 17% over
the course of the study period. The percentage of samples with 25OH-D2 levels higher
than those of 25OH-D3 decreased from 21% to 12% (see figure). Sixty-seven percent
of the samples with 25OH-D2 levels above LoQ had serum concentrations of 25OHD2 higher than those of 25OH-D3.
Conclusion: 25OH-D2, despite its gradual decrease in local prevalence, was still
present in 17% of the patients tested in December 2012, many of whom had 25OH-D2
levels higher than those of 25OH-D3. To achieve accurate vitamin D measurement,
clinical laboratories should assess the accuracy of their assays for both isomers, and if
necessary, determine their local prevalence of 25OH-D2.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S75

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

dL (mean=217+/-144); baseline labile HbA1c ranged from 1.1 to 6.2% (mean=2.5+/1.6%). Linear regression analysis of the Days 4 and 7 HbA1c values versus those
on Day Zero indicates only a 3 percent proportional difference on Day 4 with an
observed slope of 1.03. The latter change could result in an increase of up to 0.18%
at the 6.1% cutoff. However, on Day 7, a proportional difference of 8 percent was
observed (slope=1.08) suggesting that older stored samples show a consistent positive
proportional bias amounting to as much as 0.5% at our current reference range cutoff
of 6.1. The Day 4 and 7 regression lines show a good fit (Day 4,Syx=+/-0.37,r2 =0.991;
Day 7, Syx=+/-0.74,r2= 0.967).
Conclusion: We conclude that HbA1c can be performed on whole blood samples
stored at room temperature for up to 4 days, irrespective of ambient glucose levels,
with no statistically significant change in levels. However increases of up to 3% from
baseline should be expected. Test results for HbA1c on whole blood samples stored
for 7 days show an unacceptably high degree of proportional error of 8 percent.

A-264
The message content of quality control rules

M. H. Kroll, C. Garber, C. Bi, S. Suffin. Quest Diagnostics, Madison, NJ

Background: We assessed the ability of three quality control (QC) rules (R4S [R4],
13S [3S] and 22S [2S]) to provide signal compared with noise and discriminate between
imprecision (random error) and bias. QC rules should identify unacceptable runs and
discriminate specifically between bias (B) and random error (standard deviation,
SD). Comparing the probability of signal with that of noise quantifies the message
content. Appropriate corrective action depends on signal, not noise, and requires
discrimination between errors types.

A-263
Stability Studies of Hemoglobin A1c (HbA1c) Based on Specimen Storage

K. A. VandenHeuvel, K. Morrison, P. Akl, K. E. Blick. Un of OK Health

Sci Ctr, Oklahoma City, OK
Background: Hemoglobin A1c (HbA1c) is useful for the assessment of glycemic
control in patients with known or suspected diabetes and, in some cases, for the
diagnosis of diabetes. While freshly collected samples are preferred, in some instances
it is necessary to test for HbA1c on stored, whole blood samples. Published studies
suggest that both storage time and temperature can affect the HbA1c stability,
therefore we investigated the effect of room temperature storage on HbA1c up to
seven days, considering variables between specimens including levels of ambient
glucose and labile HbA1c.

Methods:We calculated the number (N) of expected combinations of quality

control messages for each rule based on the number of control materials. We
calculated baseline probabilities (P) of rule activation using the cumulative
distribution function (CDF), varying bias (B) from 0 to 1.25 and SD from 0.5 to
1.5, assuming the following allowable errors: ErrorTotal=3, ErrorBias=0.75; and
ErrorSD=1.
P(R4)=2{CDF(B,SD,-2)}{1-CDF(B,SD,2)};
P(3S)={CDF(B,SD,3}+{1-CDF(B,SD,3)};
P(2S)={CDF(B,SD,-2)}2+{1-CDF(B,SD,2)}2.
P(Noise)
=P(RuleActivation|B=0.75,SD=1);
P(Total)=P(rule-activation|B>0.75,SD>1);
P(Signal)=P(Total)-P(Noise)
P(CombinationActivation)=N*BaselineProbabil
ity=Nj*Pj(Signali); j=[R4,3S,2S],i=[SD,Bias]. Results:The number of message
combinations (N) is n/2 for R4, n for 3S, and 3(n/2-1)+1 for 2S for n-controls;
for 4-controls: N=2 for R4, N=4 for 3S, and N=4 for 2S. The figure shows the
CombinationActivation probabilities, for n=4 controls, as a function of bias or SD.
Given bias=1 or SD=1.33, CombinationActivation probabilities are as follows:
PR4(Noise)=0.0012, PR4(SignalRandom)=0.012 (<3% messages), PR4(SignalBias)=0.000;
P3S(Noise)=0.049 (21% of 3S messages), P3S(SignalRandom)=0.142 (77% of 3S Signals),
P3S(SignalBias)=0.042; P2S(Noise)=0.045 (25% of 2S messages), P2S(SignalRandom)=0.077
(58% of 2S Signals), P2S(SignalBias)=0.056.
Conclusion:The 3S and 2S rules impart significant noise (22% of messages). R4
specifically identifies random error but represents < 3% of messages. 3S and 2S
discriminate poorly between random and bias errors, with only 77% of messages
representing random errors for 3S and 58% for 2S. The quality control rules
demonstrate inadequacies in ascertaining run acceptability and discriminating
between imprecision and bias.

Methods: Twenty-five patients were included in the study; samples collected from
each patient included whole blood (EDTA) and plasma (heparin). Patients were
selected to reflect a wide range of HbA1c and glucose values: HbA1c; 4.8-14.7 % and
glucose; 95-558 mg/dL. The whole blood samples were stored at room temperature
(20C) then analyzed for HbA1c by HPLC (Variant II Turbo) on days Zero, 4, and 7;
whole blood glucose levels were measured on the Abbott i-Stat while baseline plasma
glucose levels were also measured on the UniCel DxC800 (Beckman Coulter).
The HbA1c values obtained on days Zero, 4, and 7 were compared by two statistical
methods: 1) the means for days 4 and 7 were compared to day Zero levels using the
two-tailed paired Students t-test, and 2) linear regression analysis to assess the degree
of change in regression line slope over the incubation period.
Results: The HbA1c values obtained on Day Zero ranged from 4.8 to 14.7%
(mean=9.1 +/- 3.6%) representing a wide spectrum of normal and abnormal values;
values obtained on Days 4 and 7 ranged from 5.4 to 15.7% (mean=9.1+/- 3.7%) and
5.3 to 15.9% (mean=9.0+/- 3.9%) respectively. The Students t-test indicates that the
Day 4 and Day 7 means are not statistically different from that observed on Day Zero
(p=0.86, Day 4; p=0.57,Day 7). Baseline ambient glucose ranged from 95 to 558 mg/

S76

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

A-265
Comparison of Uncertainty Values Between Core and Emergency Laboratories
for Routine Biochemical Parameters

S. Tapan1, M. Uyanik1, E. Sertoglu2, I. Kurt1. 1Glhane School of Medicine,

Department of Clinical Biochemistry, Ankara, Turkey, 2Ankara Mevki
Military Hospital, Anittepe Dispensary, Department of Biochemistry,
Ankara, Turkey
Background/Aim: Most regulatory authorities that use International Organization
for Standardization (ISO) Standards to assess laboratory competence require an
estimate of the uncertainty of measurement of assay test results. According to ISO
15189, the uncertainty of measurement is a parameter associated with the result of a
measurement that characterizes the dispersion of the values that could be reasonably
attributed to the measurand. ISO 15189 requires that The laboratory shall determine
the uncertainty of results, where relevant and possible. In this study, we aimed to
calculate the uncertainty of measurement of our core and emergency laboratory, in
terms of routine biochemistry tests.
Materials and Methods: Study was conducted at the Laboratory of Department
of Medical Biochemistry in Gulhane School of Medicine, Ankara, Turkey. Internal
and External Quality Control (IQC and EQC, respectively) results of two Olympus
AU2700 (device 1 and 2 in core laboratory) and one Olympus AU 640 (device 3
in emergency laboratory) autoanalyzers were assessed in the scope of this study. 24
parameters were evaluated between a period of January 2013-December 2013. The
calculations were derived from the EURACHEM/CITAC and AACB guides. Then we
assessed the analytical performances and uncertainty of measurements according to
CLIA, Rilibak, Fraiser and Six Sigma.
Results: Considering the evaluation criteria; the calculated uncertainty according
to AACB of three devices met the expectations for in all parameters. However, Clresults of device 1 and 3 were determined above CLIA while K+ results of device 1
and Na+ results of device 3 were above Rilibak criteria when calculated according to
EURACHEM. Moreover, calculated total error (TE) rates of Na+, K+, Cl-, calcium and
total protein were above Fraser TE rates. Cl- and urea in all devices; total protein,
albumin, Na+, K+ and calcium in device 3; Na+, creatinine and ALP in device 2 were
under 3 sigma, according to six sigma.
Conclusion: This study shows that the Core Biochemistry Laboratory of Gulhane
School of Medicine Hospital results met the expectations within the appropriate limits
of total error defined by CLIA, Rilibak, Fraser in most of the parameters. However, in
emergency laboratory, most results were found to be close to the upper limit or above
the criteria. As expected, its quite difficult for emergency laboratories, that work for
24 hours a day and 7 days a week, to meet all criteria due to shift and personnel
changes. Establishing uncertainty of measurement guidelines for clinical laboratories
can accelerate the pace of understanding both for clinicians and clinical laboratory
practitioners.

A-266
Variables impacting the analytic false positive rate for Cardiac Troponin T on
the Roche cobas e411

A. M. Wockenfus, C. D. Koch, K. J. Hartung, B. S. Karon. Mayo Clinic,

Rochester, MN
Background: We have observed rare occurrences of analytic false positive results
for cardiac Troponin T (cTnT) using plasma separator tubes (PST) and the STAT
Troponin T assay on the Roche cobas e411 immunoassay analyzer. Internal and
external (assay manufacturer) discussions identified several potential sources of error:
cellular debris, intermittent carryover, electrical supply fluctuation, and inadequate
sample centrifugation. PST has historically been used for cTnT testing because it
allows for faster processing than serum tubes. Previous studies have demonstrated
PST specimens contain more cellular debris. This increases the likelihood of microclot formation in the analyzer sample path; which could negatively impact the assay
accuracy. The goal of our study was to identify and minimize variables associated with
analytic false positive cTnT results while maintaining acceptable turnaround time.
Methods: Our Emergency Department (ED) patients receive a panel consisting of
cTnT measurement at 0, 3, and 6 hours after presentation. To capture analytic false
positive results, all samples with 0 hour cTnT 0.04 ng/mL, or 3 or 6 hour samples
demonstrating a change of 0.03 ng/mL from 0 hour value, were repeated. All samples
were repeated in this manner between 2007 and 2012. An analytic false positive was
defined as a sample that upon repeat analysis demonstrated a change in results of +/0.04 ng/mL for cTnT values 0.10 ng/mL or +/- 20% for values >0.10 ng/mL. Initial
attempts to reduce analytic false positives included adjusting centrifugation times

from 3 to 5 minutes; and manufacturer-performed instrument and measuring cell

performance verification when false positives were noted. To address relatively higher
concentrations of cellular debris in PST samples, a rapid clot serum tube (RST) was
implemented mid-2012. RST tubes fully clot within 5 minutes, drastically reducing
the clot time (up to 30 minutes) of traditional serum tubes. Finally, the electrical power
source for the analyzer was stabilized and a redesigned measuring cell was installed,
both at the recommendation of the manufacturer.
Results: Neither increasing centrifugation time, performance verification after false
positives, nor installation of the redesigned measuring cell significantly impacted the
rate of false positives observed (1 in 2200). A slight decrease in analytic false positives
was observed when switching from PST to RST (1 in 2200 vs. 1 in 2375, respectively).
The most significant decrease in false positive rate was observed when the electrical
power was stabilized while using RST. Under these conditions false positive rate
decreased to approximately 1 in 5000. In order to determine if this decrease would
also be observed with PST; RST vs. PST was re-evaluated with electrical stabilization.
RST continued to show a decreased false positive rate.
Conclusion: The combination of electrical power supply stabilization and
implementation of RST decreased the rate of analytic false positive cTnT results on
the cobas e411 to approximately 1 in 5000. This facilitated rapid analysis of cTnT
samples while eliminating the need to repeat positive results for patients presenting
to the ED.

A-267
Effects of 49 Different Rare Hb Variants on HbA1c Measurement in Seven
Methods

R. L. Schmidt1, S. L. Laulu2, S. E. Hanson3, C. L. Rohlfing3, R. R. Little3.

1
University of Utah School of Medicine and ARUP Laboratories, Salt Lake
City, UT, 2ARUP Institute for Clinical and Experimental Pathology, Salt
Lake City, UT, 3University of Missouri School of Medicine, Columbia, MO
Background: Hemoglobin A1c (HbA1c) is recommended for routine monitoring of
long-term glycemic control in patients with diabetes and for use in diabetes diagnosis.
Previous studies have shown interference from the four most common heterozygous
Hb variants (HbAS, HbAE, HbAC, and HbAD) with some HbA1c assay methods.
Here we examine analytical interference from 49 different less common variants with
7 different HbA1c methods using various method principles (ion-exchange HPLC,
boronate affinity HPLC, immunoassay, and enzymatic).
Methods: Hb variants were screened using the Bio-Rad Variant II beta thal short
program, confirmed by alkaline and acid electrophoresis, and identified by sequence
analysis. The Trinity ultra2 boronate affinity HPLC method and Roche Tina-quant
immunoassay were used as primary and secondary comparative methods, respectively,
since these methods are least likely to show interference from Hb variants. Other
methods included were the Tosoh G7 and G8, Bio-Rad D-10 and Variant II Turbo, and
Diazyme Enzymatic. Identified variant samples (n=88) and a group of non-variant
(AA) samples (n=92) were analyzed by each method; there were multiple samples
for some variants. Samples with HbF >10% (based on G8 %HbF) were excluded.
To eliminate any inherent calibration bias, results for each method were adjusted
using regression verses the ultra2 with AA samples. When results from the two
comparative methods matched within 10%, all other methods calibration-adjusted
results were compared and judged to be acceptable if within 10% of the ultra2 result.
For five variants (one sample each), results from the two comparative methods were
discordant and this could not be explained by an amino acid substitution that would
affect binding of the Tina-quant antibody; these results were excluded from further
evaluation.
Results: Almost all variant samples were recognized as such by the ion-exchange
HPLC methods by the presence of abnormal peaks or results outside the reportable
range. For most variants (80%), interference (>10% difference from ultra2 or no
result reported) was seen with one or more of the ion-exchange methods. Following
manufacturer instructions for interpretation of chromatograms usually, but not always,
prevented reporting of inaccurate results. For 14 variants studied, the Diazyme results
were inaccurate; this method does not detect the presence of variants and all results
are reported.
Conclusions: In order to insure that accurate HbA1c results are obtained for all
patients, it is important to know if a patient has a hemoglobin variant and how that
variant can affect their HbA1c results. Laboratories must be cautious about reporting
results when the presence of a variant is suspected. Manufacturers may need to
clarify and/or tighten their criteria for accepting results and update their software to
flag potential variants in an effort to reduce the likelihood of incorrect HbA1c results
being reported.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S77

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

A-268
Evaluation of HbA1c measurement in Trinidad and Tobago

R. R. Little1, C. L. Rohlfing1, D. E. Goldstein1, P. Ladenson2, M. Rastogi3.

1
University of Missouri School of Medicine, Columbia, MO, 2Johns
Hopkins University School of Medicine, Baltimore, MD, 3Sangre Grande
Hospital, Sangre Grande, Trinidad and Tobago
Background: HbA1c is recommended for routine monitoring of long-term glycemic
control in patients with diabetes and also for diabetes diagnosis. The prevalence of
diabetes in Trinidad and Tobago is over 12%. Although HbA1c testing is offered
by public and private sector laboratories, no HbA1c proficiency testing (PT) has
previously been performed. Therefore, Johns Hopkins Medicine International and the
Diabetes Diagnostic Laboratory (DDL) at the University of Missouri organized a pilot
HbA1c Proficiency Testing Study for 7 key laboratories as part of the Trinidad and
Tobago Health Sciences Initiative.
Methods: Sets of 10 samples containing blinded duplicates were created from five
whole blood pools with HbA1c levels from 5.1 to 9.3% HbA1c and shipped to
participating laboratories. To assess within-day imprecision, the pooled estimate of
the SDs between the duplicates (Sp) was calculated; 0.229 was the acceptable limit
based on the current NGSP HbA1c standardization program monitoring criterion.
To assess accuracy, each laboratorys results were compared to an NGSP Secondary
Reference Laboratory (SRL9 using Tosoh G8).
Results: One laboratory reported results as IFCC %HbA1c; these were aligned to
NGSP using the NGSP/IFCC master equation [NGSP%=0.915(IFCC%)+2.15].
Methods used by the laboratories included Roche Tina quant on Cobas Integra, Cobas
6000 and Hitachi 911, Sebia Minicap, and Axis-Shield NycoCard. All laboratories
except the two using the NycoCard showed acceptable imprecision. All results from
three laboratories (one each using Sebia Minicap, Roche Cobas 6000, and Roche
Cobas Integra) were within 6% of the NGSP SRL assigned values; most results from
the two laboratories using the NycoCard were outside 6%.
Conclusion: Because inconsistent HbA1c results can negatively impact patient care,
it was recommended that all laboratories report NGSP % HbA1c and those using the
NycoCard use a more precise method. Laboratories are instituting changes based on
these findings and a future study will re-assess performance.

subjects (59 males and 61 females; 19-75 years) and G6PD determined by kinetic
spectrophotometry using commercially available reagents (Trinity Biotech USA)
at 37C. For the Hoffman method, 20,736 G6PD test results were extracted from
the laboratory information system and linear regression performed over the linear
portion of the cumulative frequency distribution. The LLRI and upper limit of
the reference interval (ULRI) were calculated as LLRI=2.5(slope)+intercept and
ULRI=97.5(slope)+intercept. The effects of pre-analytical sources of error were
investigated using samples submitted to the laboratory for G6PD testing and included
the method of vortex mixing, incubation time for RBC lysis, and analytical dwell
time. The study was approved the University of Utah Institutional Review Board.
Results: G6PD activities in 120 reference subjects ranged from 9.0-14.7 U/gHb
(median=11.6). The non-parametric reference interval was 9.9-14.1 U/gHb. G6PD
activities in the 20,736 clinical subjects ranged from 0.3-86.2 U/gHb (median=13.2).
6.1% of the samples were less than the current LLRI of 7.0 U/gHb. The Hoffman
technique yielded a reference interval of 9.9-16.6 U/gHb. Applying this LLRI to
the clinical subjects identified 8.9% as deficient. The method of vortex-mixing was
evaluated in 40 samples and did not significantly affect G6PD activity. Compared to
the reference technique (single-tube with single-tube vortex mixer) which produced
a mean G6PD activity of 13.0 U/gHb, the use of a tube rack with multi-tube vortex
mixer or a tube rack with single-tube vortex mixer produced mean activities of 12.9
and 13.2 U/gHb, respectively (p>0.24). The incubation time for RBC lysis after
vortex-mixing did affect G6PD activity. Compared to the reference time of 5 minutes,
a 3-minute incubation had no effect (mean difference 0.2 U/gHb, p=0.05) but 15and 30-minute incubation times produced significantly lower results (mean difference
-1.0 and -2.1 U/gHb, respectively; p<0.0001). Analytical dwell time (40 minutes) had
no significant effect on G6PD activity as determined from 70 aliquots of a whole
blood sample that produced a CV of 1.5% at a mean of 15.6 U/gHb (r=0.12, p=0.32).
Conclusions: The G6PD LLRI should be 9.9 U/gHb as determined empirically and by
the Hoffman method. The incubation time for RBC lysis is a variable that should be
controlled when performing G6PD testing.

A-270
Diurnal Variation of analytes: an underestimated pre-analytical factor in
clinical chemistry

M. Abou El Hassan, B. Hoffman. University of Toronto, Toronto, ON,

Canada
Objectives: Diurnal variation is a well known cause of biological variation. We set
out to identify analytes affected by diurnal variation, the amplitude of the within day
difference and whether collection instructions specified the impact of collection time
on test interpretation.
Design and Methods: We identified analytes reported to undergo diurnal variation
through a review of Tietzs Textbook of Clinical Chemistry and through a search
of Pub Med to capture recent citations. We checked if the time of collection, and
its potential impact on test interpretation, was specified in the posted collection
procedures issued by two large commercial labs in Ontario.
Results: Our search identified 21 analytes affected by diurnal variation. As expected,
cortisol, the archetypal diurnally changing analyte, was the most cited. Hormones
represented the major class of identified analytes, but fasting plasma glucose (FPG),
MR-proANP, and hepcidin were other noteworthy inclusions. Diurnal amplitude
ranged from a low of 1.1 for FPG (still sufficiently large to degrade test performance
in ruling diabetes in or out using the cutpoint of 7.0 mM) to a maximum of 65-fold for
melatonin. Of the 14 analytes listed in posted collection instructions, the importance
of the time of collection was specified for only 5 (36%).

A-269
Identifying G6PD Deficiency: Defining the Lower Reference Limit and
Evaluating Pre-analytical Sources of Error

Conclusions: The confounding effect of diurnal variation on test interpretation is

underappreciated given that cautions regarding when analytes should be collected
were issued in collection instructions only 36% of the time.

B. Suh-Lailam, D. Grenache. University of Utah, Salt Lake City, UT

Background: Glucose-6-phosphate dehydrogenase (G6PD) protects red blood cells
(RBCs) from oxidative damage by regenerating NADPH. G6PD deficiency is the
most common X-linked enzymopathy that results in hemolytic anemia triggered
by oxidative stress. Diagnosis of G6PD deficiency is facilitated by determining its
activity in RBCs and interpreting the result against the lower limit of a reference
interval (LLRI). Despite the use of similar methods, the LLRI for G6PD varies
considerably between laboratories (range, 4.6-8.8; N=4). The objectives of this study
was to 1) validate the LLRI of the G6PD reference interval used by our laboratory
(7.0-20.5 U/gHb) and 2) to identify pre-analytic sources of error that may influence
the accuracy of G6PD test results.
Methods: A G6PD reference interval was determined empirically and by the Hoffman
method. For the empirical approach, whole blood was collected from 120 reference

S78

A-271
Can platelet aggregation testing by Multiplate be influenced by one minute
tourniquet application?

G. Lima-Oliveira1, G. Lippi2, G. L. Salvagno3, S. Gaino3, G. Poli3, M.

Gelati3, G. Picheth1, G. C. Guidi3. 1Federal University of Parana, Curitiba,
Brazil, 2University of Parma, Parma, Italy, 3University of Verona, Verona,
Italy
Background: The study of the role of platelets in the pathogenesis of ischemic
vascular diseases and the monitoring of antiplatelet drugs require reliable plateletfunction tests. Presently several techniques are in use to measure platelet aggregation.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

Assessment of platelet function by multiple electrode aggregometry (Multiplate

Roche, Germany) is common in laboratory practices. Indeed, sample collection
and processing are essential steps for quality assurance, so that the fulfillment of
standardized preanalytical conditions are key factors in maintaining patient safety.
The collection of diagnostic blood specimens for Multiplate is traditionally performed
by medical staff using a tourniquet for evidencing veins. The Clinical and Laboratory
Standards Institute H03-A6 document, recommends the use of the tourniquet for
localizing suitable veins and the tourniquet time can not in any case be extended over
60 sec. This study was aimed to assess the impact of 60 sec tourniquet application on
platelet function evaluated by multiple electrode aggregometry (Multiplate).
Methods: Ten volunteers, after 12-hours fasting, were maintained seated during 15
minutes to eliminate possible interferences from both lipemia and blood distribution
due to different posture. After this time frame, 6mL of blood were collected by
venipuncture with a 20G straight needle directly into two 3.0mL Hirudin Blood Tube
for Multiplate analysis (proprietary vacuum tube 06670105001, Roche Diagnostics
GmbH, Penzberg, Germany) from two different procedures: Procedure I (no stasis) - a
radial vein was localized on right forearm by a subcutaneous tissue transilluminator
device without tourniquet, to prevent interference from venous stasis. Procedure II
(stasis) - a radial vein was localized on the left forearm by tourniquet application
during 60 sec prior to venipuncture. To eliminate any potential interference due to
either the contact phase or the tissue factor all 1st tubes from each volunteer for both
procedures were discarded. All diagnostic blood specimens were mixed gently and
carefully by five times inversion, as recommended by the manufacturer. All samples
were processed for assessment of platelet function (ADP-test, ASPI-test, COLtest, RISTO-test and TRAP-test) by multiple electrode aggregometry (<15
min after collection) on the same Multiplate (Roche Diagnostics GmbH, Penzberg,
Germany). In the Multiplate Test Cell, activated platelets adhere to and aggregate on
the sensor wires. This leads to an increased resistance between the sensor wires, which
in continuously recorded and expressed via the area under the curve in arbitrary units.
The significance of differences between samples was assessed by paired Students
t-test after checking for normality by the DAgostino-Pearson omnibus test. The level
of statistical significance was set at p<0.05.

Table1: Stability of ethanol with various volumes at room temperature

Base Line 1h
2h
3h
100 L of P* Mean SD (mg/dL) 143 95 126 84 107 72 91 60
R (%)
100
88
76
66
(NS)
100 L of P* Mean SD (mg/dL) 143 95 144 95 144 95 144 96
R (%)
100
101
101
101
(ST)
3 mL TV (0.5 Mean SD (mg/dL) 142 94 141 94 137 92 131 87
mL P*)
R (%)
100
99
96
92
(NS)
4 mL TV (1
Mean SD (mg/dL) 153 100 150 98 147 95 143 94
mL P*)
R (%)
100
98
95
94
(NS)
5 mL TV (2
Mean SD (mg/dL) 142 73 141 72 138 71 136 71
mL P*)
R (%)
100
99
97
96
(NS)
TV: Total Volume, P*: Plasma, R: Recovery, ST: Stopper, NS: No-Stopper

A-273
Evaluation of a method to detect prozone/hook effect/antigen excess
phenomenon for Free Light Chain quantification using a simple pooling
protocol

V. De Guire1, F. Courjal2, R. Leblanc1, M. Malvaso1, M. Pich1, H. Harbec1,

A. Romro Ospina1, J. Amyot1, N. Ttreault1. 1Maisonneuve-Rosemont
Hospital, Montreal, QC, Canada, 2The Binding Site Group, Birmingham,
United Kingdom
Background: Antigen excess is an important issue that can lead to underestimation
of patients results and misdiagnosis. With patient samples which can range from
less than 1mg/L to over 100 000mg/L, serum free light chain (FLC) measurement is
especially prone to this interference. Some institutions choose not to implement on
site testing for FLC, due to the lack of an instrument platform which can automatically
detect antigen excess and in which concerns over workflow interruptions and
personnel costs inhibit implementation of an antigetn excess protocol.

Results: All platelet functions tested showed lower values from samples collected with
tourniquet than without tourniquet. Significant differences (p<0.05) were observed for
RISTO-test (-16%), ADP-test (-10%) and TRAP-test (-8%) determination in samples
collected after 60 sec tourniquet application.

Methods: We propose a methodology based on sample pooling for a fast and cost
effective method to detect of antigen excess phenomenon for serum free light chain
quantification. Our strategy was evaluated on the Immage (Beckman) and on the BNII
(Siemens) nephelometers using the Freelite assay (The Binding Site).

Conclusion: The significant variations after 60 sec tourniquet application suggest

that platelets could undergo a sort of pre-activation by venous stasis, thus decreasing
the sensitivity to subsequent activation by agonists. Furthermore, this effect can
compromise clinical results interpretation and jeopardize patient safety. In conclusion
tourniquet application should be avoided during phlebotomy for Multiplate analyses.

Results: First, we evaluated the sensitivity of our strategy using a pool of 15 mixed
samples spiked with different concentrations of kappa or lambda free light chain.
Secondly, patients with antigen excess ranging from 1192 to 8500 mg/L of kappa
free light chain were efficiently detected using our strategy. Finally, a preliminary
evaluation of the size of the pools was conducted using pools ranging from 15 to 42
different samples. Based on the precision of the method, our preliminary data suggest
that the number of samples in a pool can reach up to 43 different patients to detect an
antigen excess of 1192 mg/L.

A-272
Impacts of sample volume and stopper on the stability of ethanol in lithium
heparin plasma

S. N. Narla, D. Brass, S. Dean, B. Horne, Y. Zhu. Medical University of

South Carolina, Charleston, SC
Background: It is recommended that ethanol samples should be analyzed immediately
upon opening the tube, but there is no evidence about how long the ethanol is stable
in real specimen tubes without stoppers. The objective of this study is to evaluate the
stability of ethanol with different sample volumes in unstoppered and stoppered tubes
at room temperature (RT).
Methods: Heparin plasma samples with ethanol 10 mg/dL were selected for this
study and stored at -20C until analysis. Ethanol was analyzed on UniCel DxC800
Systems by an enzymatic assay. To determine the impact of stopper on ethanol at RT,
30 plasma samples with a volume of 100 L stored in 5 mL plain tubes at RT with
and without stoppers and ethanol was measured after 1, 2 and 3 hours of baseline
measurement. To determine the influence of sample volume on ethanol stability, 30
heparin separator tubes each with a total volume (including cells, gel, and plasma)
of 3, 4 and 5 mL with plasma volume of 0.5, 1 and 2 mL respectively were stored
at RT without stoppers and ethanol was measured after 1, 2 and 3 hours of baseline
measurement. The allowable total error (ATE) for ethanol was 25%.
Results: The average ethanol concentrations in samples with different volumes with
and without stoppers and recoveries were shown in table1.
Conclusions: Ethanol concentrations are within ATE for samples stored for at least 3
hr without stopper at RT with a minimum of 0.5 mL plasma in original heparin plasma
separator tubes and for samples stored in plain tubes with stoppers at RT for at least 3
hr with a minimum volume of 100 L, whereas ethanol in this low volume of sample
in plain tubes is unstable for more than 2 hr at RT without a stopper.

Conclusion: We described and validated a strategy based on sample pooling for

detection of antigen excess on free light chain measurement. This approach could be
suitable for any laboratory measuring serum free light chain that does not currently
have an instrument platform for detection of antigen excess.

A-274
Performance Evaluation of Rapid Test Strips for the Identification of EDTAContaining Specimens

K. L. Scholes1, S. L. Laulu2, A. T. Ence1, H. L. Logan1, R. L. Parker1, F.

G. Strathmann3, J. R. Genzen3. 1ARUP Laboratories, Salt Lake City, UT,
2
ARUP Institute for Clinical and Experimental Pathology, Salt Lake City,
UT, 3University of Utah, Department of Pathology, Salt Lake City, UT
Background: Most laboratory assays have specimen acceptability requirements.
Proper identification of ethylenediaminetetraacetic acid (EDTA) containing specimens
is frequently necessary when investigating mislabeled specimens, suspected primary
tube mixtures, and aliquots. The purpose of this study was to evaluate the performance
of commercially available EDTA test strips using biological specimens.
Methods: Studies included human sera and plasma specimens, bovine calf serum,
and phosphate-buffered saline. Specimens were applied to test strips (Quantofix
EDTA; Macherey-Nagel) using a pipet (drop mode). These test strips are embedded
with bismuth nitrate, xylenol-orange, and citric acid to detect chelating agents such
as EDTA. Test strips were blotted to remove excess fluid; visual evaluation was made
at 15 seconds. Reactions were scored on a scale from red (no EDTA), orange (low
EDTA), to yellow (EDTA). A NODE+Chroma color detector (Variable, Inc) on a
custom 3D-printed test strip channel was also used to capture quantitative RGB values

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S79

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

on a Bluetooth-linked iPhone 5, thus permitting additional statistical analysis and
graphical display. Potentially reactive chelating agents tested included alpha lipoic
acid (ALA), deferoxamine (DEF), 2,3-dimercapto-1-propanesulfonic acid (DMPS),
dimercaptosuccinic acid (DMSA), K2EDTA, Na2EDTA, ethylene glycol tetraacetic
acid (EGTA), and penicillamine (PEN). Results were compared to K+, Ca2+, Mg2+,
and serum indices (Roche cobas 8000), as well as a sodium tetraphenylborate method
of K+ detection. Finally, supernatant bismuth concentrations were evaluated using
inductively coupled plasma mass spectrometry (ICP-MS).

width, reticulocytes, white blood cells count and differential, including neutrophils,
lymphocytes, monocytes, eosinophils and basophils, platelet count and mean platelet
volume, glucose, total cholesterol, high-density lipoprotein cholesterol, triglycerides,
total protein, albumin, C-reactive protein, urea, creatinine, uric acid, alkaline
phosphatase, amylase, pancreatic amylase, aspartate aminotransferase, alanine
aminotransferase, g-glutamyl transferase, lactate dehydrogenase, creatine kinase, total
bilirubin, direct bilirubin, phosphorus, calcium, magnesium, iron, sodium, potassium,
chloride, lipase and hemolysis index.

Results: Test strips detected EDTA in specimens taken from the following primary
tube types: royal blue (Na2EDTA), tan (K2EDTA), lavender (K2EDTA), and pink top
(K2EDTA). Test strips did not falsely detect EDTA in serum tubes including gold
(serum separator), red (serum), and orange top (thrombin / serum separator), although
three discordant red top serum results - due to pour off and/or aliquot errors - were
identified using EDTA test strips in preliminary experiments. Test strips did not
falsely detect EDTA in most other plasma specimens, including light green (lithium
heparin / plasma separator), green (sodium or lithium heparin), and light blue top
(sodium citrate), although one type of gray top tube (potassium oxalate with sodium
fluoride) produced an orange low EDTA reaction. EDTA test strip reactivity was
observed with a variety of chelating agents (order of reactivity: DMSAEGTA>K
EDTA=Na2EDTA>DMPS>PEN; ALA and DEF were non-reactive at 10 mM),
2
although in general reactivity was only observed at concentrations higher than might
be expected during chelation therapy. No pH effect was observed between pH 2-12.
Lipemia did not interfere with test strip performance. Marked hemolysis caused an
orange appearance to otherwise yellow reactions from EDTA-containing specimens,
while visible icterus caused a slightly brown appearance to normally red reactions
from non-EDTA specimens. ICP-MS demonstrated that bismuth levels were higher
in specimens which had test strips dipped into the solution, arguing that a dip mode
method may interfere with certain clinical assays.

Results: No fibrin filaments or microclots were observed in any samples. Significant

differences (P<0.01), were found only for: a) erythrocytes (0.5%) and haematocrit
(1.1%) when M1 was compared with M2; b) alanine aminotransferase (-3.0) when
M1 was compared with M3; c) erythrocytes (-0.9%) and haematocrit (-0.8%) when
M2 was compared with M3.

Conclusions: EDTA test strips reliably detected the presence of EDTA in clinical
specimens. Indeterminate or low EDTA orange results may require further
investigation. Dip mode can produce analytical complications for assays which
measure (or are interfered by) test strip constituent reagents.

A-275
Lack of tube mixing can be validated as regards ISO-15189 standard preliminary validation

G. Lima-Oliveira1, G. L. Salvagno2, G. Lippi3, G. Brocco2, M. Pantani2, W.

Volanski1, F. G. Rego1, G. Picheth1, G. C. Guidi2. 1Federal University of
Parana, Curitiba, Brazil, 2University of Verona, Verona, Italy, 3University
of Parma, Parma, Italy
Background: Gentle and careful mixing by inverting 5 times is standardized as good
laboratory practice by manufacturers and endorsed by Clinical Laboratory Standard
Institute. Correct mixing after blood collection is claimed to be important. Nowadays
laboratory quality managers have evidenced that phlebotomists do not mix vacuum
tubes as recommended by manufacturers. As regards ISO-15189 international
standard, all necessary improvements and potential sources of nonconformities, either
technical or concerning the quality management system, shall be identified and all
laboratory process shall be validated. The aim of this study was to evaluate whether
it is really necessary to mix both K2EDTA- and lithium heparin-vacuum tubes
immediately after blood collection.
Methods: Blood collection: Samples from 100 volunteers were drawn in three 3.0mL
vacuum tubes containing 5.9mg K2EDTA and in three 3.5mL vacuum tubes with
52.5USP U of lithium heparin and gel separator. To eliminate any potential interference
due to either contact phase or tissue factor, ~2mL of blood were preliminarily collected
in a discard tube without additive. Blood collection was accurately standardized,
including the use of needles and vacuum tubes of the same lot.
Processing: All vacuum tubes (one from each kind of additive) were processed using
3 different methods. Method 1: Gold Standard (M1): All specimens were mixed gently
and carefully by inverting 5 times as recommended; Method 2: Rest time (M2): All
specimens remained 5 min in upright position, followed by gentle careful mixing by
inverting 5 times; Method 3: No mix (M3): All specimens were left in upright position
without mixing afterwards. The influence of the primary tube mixing procedure was
evaluated for routine hematology- and clinical chemistry-testing by paired t-test.
Laboratory testing: All samples were processed for routine hematological
testing immediately after collection (<15min) on the same Sysmex XE-2100D,
Automated Hematology Analyzer (Sysmex Corporation, Kobe, Japan). Routine
clinical chemistry was performed on the same Cobas 6000 c501 module
(Roche Diagnostics GmbH, Penzberg, Germany). The parameters tested included
erythrocytes, hemoglobin, hematocrit, mean corpuscular volume, RBC distribution

S80

Conclusion: This preliminary evaluation has shown that K2EDTA- and lithium
heparin- tube mixing after collection with evacuated system appears unnecessary.
Moreover, this outcome indicates that not mixing vacuum tubes should not viewed as
a non conformity for quality system and in conclusion the lack of tube mixing can be
validated as regards ISO-15189 standard.

A-276
Comparability of urine total protein assays in patients with monoclonal
proteinuria

C. D. Giesen, K. S. Lockington, T. S. Larson, J. C. Lieske. Mayo Clinic,

Rochester, MN
Background: Urine contains various proteins including albumin, Tamm Horsfall
protein, and low molecular weight proteins. Several different reagents and instrument
platforms are used for urinary total protein measurement, with no established gold
standard. All methods appear to measure albumin consistently, yet detection of nonalbumin proteins appears more variable. This study compared the performance of five
different reagent kits using urine from patients with and without kidney disease and
also with known monoclonal M-spikes.
Methods: Urine samples submitted for random urinalysis (RUA) and monoclonal
protein electrophoresis testing were analyzed using four pyrogallol red urine protein
assays (Pointe Scientific Microprotein Reagent Set, Canton MI; Quantimetrix
QuanTtest Red, Redondo Beach CA; Wako Diagnostics Autokit Micro TP, Richmond
VA; and Siemens Healthcare Diagnostics Total Protein_2 (Urine), Tarrytown NY),
and one benzethonium chloride assay (Roche Diagnostics Total Protein Gen. 2,
Indianapolis IN) on the Roche cobas 6000 c501. Pyrogallol red assays were all
performed using identical instrument settings, while the benzethonium chloride assay
was performed per manufacturers instruction. The Wako pyrogallol red assay served
as the reference assay for the analysis. Samples were electrophoresed and stained
on a SPIFE 3000 system (Helena Laboratories, Beaumont TX). M-spike quantitation
was performed using a Helena Laboratories Quick Scan 2000. The Quick Scan 2000
scans and calculates the relative percentage of each electrophoretically separated
protein fraction based on its staining intensity. The original urinary concentration of
the M protein band is obtained by multiplying its relative percentage on the gel by the
urinary total protein concentration.
Results: Among RUA samples with varying levels of proteinuria (1- 2632 mg/dL;
mean=130 mg/dL; median=52 mg/dL), Passing and Bablok regression analysis
revealed good correlation between all methods and the reference (Wako) assay
(Pointe: y=1.15x+1.90, n=100; Quantimetrix: y=0.97x+0.39, n=325; Siemens:
y=0.99x+5.04, n=25; Roche: y=1.01x+4.31, n=100). In the urine samples with a
known M-spike but total protein content >25% albumin, linear correlations with the
Wako assay were still reasonably good (Pointe: y=1.0273x+7.8232 R2=0.9985, n=18;
Quantimetrix: y=0.9549x-3.6582 R2=0.9974, n=22; Siemens: y=1.0114+4.1414 R2=1,
n=4; Roche: y=0.9496x+11.007 R2=0.9946, n=18). Urine samples with M-spike
concentrations <100 mg/24 hours also compared reasonably well, regardless of the
albumin percentage (Pointe: y=1.0731x+3.0911 R2=0.9564, n=44; Quantimetrix:
y=0.9101x+0.4658 R2=0.9781, n=54; Siemens: y=0.9059x+6.181 R2=0.9843,
n=10; Roche: y=1.0416x+3.0244 R2=0.884, n=44). However, among samples with
M Spikes 100 mg/24 hours and albumin 25%, comparisons between assays
were quite variable (Pointe: y=1.1156x+10.509 R2=0.8207, n=52; Quantimetrix:
y=0.2386x+18.907 R2=0.6697, n=66; Siemens: y=0.9536x+15.617 R2=0.9735, n=14;
Roche: y=1.2409x+15.676 R2=0.7977, n=52).
Conclusions: All urine protein assays in this study compared well for testing random
urinalysis samples obtained from a population of patients with and without kidney
disease in whom albumin is usually a major protein species. However, these total
protein assays performed quite variably when testing urine samples that contained
large quantities of monoclonal protein, particularly those with <25% albumin and

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Factors Affecting Test Results

Tuesday, July 29, 9:30 am 5:00 pm

a calculated M-spike greater than 100 mg/24 hours. In particular, urine total protein
assays can vastly underestimate protein excretion in certain patients with large
M-spikes.

A-277
Elevated Cerebrospinal Fluid Total Protein caused by Povidone-Iodine
(Betadine) Interference

V. Gounden, Z. Zhao. National Institutes of Health ( Clinical Center),

Bethesda, MD
Background: The measurement of cerebrospinal fluid (CSF) total protein (TP) is
useful in the diagnosis of meningitis and the detection of other inflammatory diseases.
The laboratory was contacted by a clinician concerned with a clinically discrepant
elevated CSF TP. The specimen received for chemistry testing was the first tube of
the CSF collection. The CSF TP was 417 mg/dL (reference interval 15- 45 mg/dL)
and CSF glucose was 87.2 mg/dL (reference interval 40-70 mg/dL). CSF cell counts
were: red blood cells 6/mm3 and white cell count 1/ mm3. Another specimen from
the patient (tube 2) taken at the same time and submitted to hematology was analyzed
for CSF TP. The CSF TP result of tube 2 was 20 mg/dL and was confirmed on repeat
analysis.
Objectives: To investigate falsely high CSF TP results, suspected to be caused by
preparation of collection site by povidone-iodine (Betadine) solution.
Design and Methods: We performed interference studies to determine the effect of
Betadine, iodine only and povidone only on the CSF TP concentration. A CSF diluent
was prepared by pooling clear CSF specimens to which reagent grade water was
added to give a final ratio of 90%:10% v/v CSF to water. An initial sample of the CSF
diluent and either the povidone-iodine or iodine only or povidone only solution was
prepared to give a final ratio of 90%:10%v/v CSF to solution. This initial sample was
then serially diluted with the prepared CSF pool to give different final concentrations
of povidone-iodine/iodine/ povidone. CSF TP was then measured for these prepared
samples using the laboratory routine chemistry analyzer, the Siemens Dimension
Vista (Siemens Healthcare Diagnostics, Tarrytown, USA). This assay involves the
reaction of protein in the sample with the pyrogallol red (PGR) sodium molybdate
complex to form a bluish-purple colored complex, which absorbs at 600 nm. Further
CSF TP measurements were made on the Siemens Dimension Xpand using the
PGR method. In addition to this CSF samples were analysed for TP at a reference
laboratory using the modified biuret method. To investigate if the positive interference
was a result of the iodine in the Betadine solution and its direct absorbance at the same
wavelength as the PGR- protein complex, spectrophotometric absorbance studies
were also performed.
Results: The experimental data of the interference confirmed a positive interference
for the PGR assay when Betadine containing povidone or povidone alone were added
to a CSF sample. The Betadine solution did not interfere the modified biuret method
for TP. The false positive TP interference was clinically significant (> 10% change)
even at Betadine concentrations (as low as 0.0025%) where the CSF sample appeared
clear on visual inspection. Spectrophotometric studies of the Betadine solution and
patients sample showed no absorbance at 600nm.
Conclusions: Low levels of povidone-iodine contamination of CSF specimens can
lead to clinically significant positive interference for TP results. Alternate iodine
solutions not containing povidone should be used for preparation of sites for CSF
collection.

A-278
Extending the Time Restriction on Transit Time for Lactate Measurement

L. Racsa1, C. Eggert2, M. Mohamed2, P. Kutscher2, I. Hashim1. 1University

of Texas Southwestern Medical Center, Dallas, TX, 2Parkland Health &
Hospital System, Dallas, TX
Background: Plasma Lactate is useful for assessing tissue perfusion in critically ill
patients. When using sodium fluoride/potassium oxalate (FlOx) as a preservative, test
reagent manufacturers require that blood specimens be processed within 15 minutes
from collection. This transit time limit, however, cannot always be met, particularly
when the laboratory is distant from patient care, as the case in our institution. 3% of
lactate samples arrive late into the laboratory and are thus rejected. In an attempt to
reduce the number of rejected lactate specimens, we examined the effect on lactate
values following extended transit time to 30 minutes.
Methods: Lactate samples were collected in triplicate from 50 patients (with prior
physician orders for lactate) and from 50 normal volunteers. One sample set was kept

at room temperature (RT) and processed at 15 minutes from collection according

to manufacturers instruction. Of the remaining two sample sets, one was kept at
RT and the other on ice (4C) for 30 minutes before lactate analysis. Lactate was
measured using Roche Cobas LACT2. Accuracy was determined with lactate values
obtained at 30 minutes (both RT and 4C) compared to the standard 15 minutes.
Intra-assay precision was obtained using 10 aliquots from the same samples spanning
the analytical measuring range. Inter-assay precision studies were performed on 20
patients.
Results: Lactate values ranged from 0.6 to 25.4, and from 0.7 to 2.1 mmol/L in patients
and volunteers, respectively. There was good correlation between RT lactate values
measured at 15 and 30 minutes (r= 0.9993, bias 0.04). There was good correlation
(r=0.9992, bias -0.06), between RT lactate values measured at 15 minutes and those at
30 minutes (4C). There was good correlation (r=0.9993, bias -0.01) between RT and
4C lactate values measured at 30 minutes. Intra-assay imprecision for RT samples
processed at 15 minutes ranged from 1.6%, 1.0%, 1.1%, and 0.9% at lactate levels of
1.3, 5.5, 9.0 and 12.4 mmol/L respectively. For RT samples processed at 30 minutes,
imprecision ranged from 0%, 1.2%, 1.0%, and 0.8% compared with 4C samples, 0%,
1.3%, 0.9%, and 1.1%, at lactate levels of 1.3, 5.5, 9.0 and 12.4 mmol/L respectively.
Inter-assay imprecision for RT samples processed at 15 minutes ranged from 0.06.7%; RT samples process at 30-minutes ranged from 0.0-13.1% compared with 0.010.2% for samples kept at 4C. All precision studies met our acceptance criteria of
<14%.
Conclusion: The performance of the assay for samples kept for 30 minutes either
at RT or 4C was not significantly different from those kept at RT and processed at
processed at 15 minutes. Our institution changed the lactate processing protocol to 30
minutes, leading to a reduction of lactate order rejections from 3% to less than 1%.

A-279
Measurement of Serum Total Calcium Using the 5-nitro-5-methyl-BAPTA
Method in the Presence of Four Gadolinium-based Contrast Agents

C. A. Wittwer, L. J. Ouverson, D. R. Block, N. A. Baumann. Mayo Clinic,

Rochester, MN
Background: Some gadolinium-based contrast agents administered to patients
undergoing MRI procedures are known to interfere with the widely-used
o-cresolphthalein complexone (o-CPC) dye method used to measure serum total
calcium. Patient samples with serum calcium <7.0 mg/dL using this method were
routinely re-tested in the clinical laboratory with an Arsenazo III method to avoid
reporting falsely decreased calcium results due to gadolinium interference. Roche
Diagnostics recently introduced a calcium reagent formulation which utilizes a
different ion-selective indicator, 5-nitro-5-methyl-BAPTA (NM-BAPTA).
Objectives: (1) Verify the manufacturers claim that gadolinium-containing MRI
contrast media at therapeutic concentrations does not interfere with serum total
calcium measurement utilizing the NM-BAPTA method. (2) Evaluate laboratory
efficiencies gained by implementing a serum calcium method that is not affected by
gadolinium-based contrast agents.
Methods: Gadolinium-based contrast agents commonly used at Mayo Clinic were
added to five residual patient serum pools with varying concentrations of calcium
(6.5-10.3 mg/dL). Gadodiamide, gadobenate dimeglumine, gadoxetate disodium and
gadofoveset trisodium were each added to serum pools yielding final concentrations
of 0.1, 0.25, 0.5, 1.0, and 2.0 mmol/L contrast agent. Total serum calcium was
measured utilizing the o-CPC and NM-BAPTA methods on a Roche Cobas c501 and
c701 (Roche Diagnostics) and the Vitros 350 Arsenazo III method (Ortho Clinical
Diagnostics). Results obtained were compared to control serum pools with de-ionized
water added to account for dilution. Absolute differences between total calcium
concentrations in samples with and without contrast agent were calculated. The
number of patient samples with calcium <7.0 mg/dL, cost of maintaining a secondary
method and turn-around time were assessed over a one year time period.
Results: Addition of gadodiamide at final concentrations of 0.1, 0.25, 0.5, 1.0 and 2.0
mmol/L lowered serum calcium results by an average (mg/dL) of 0.3 (range 0.1-0.5),
0.5 (range 0.2-0.8), 0.8 (range 0.4-1.3), 1.4 (range 0.9-1.9) and 2.4 (range 1.7-3.0)
mg/dL, respectively, using the o-CPC method. There were no differences between
calcium results in control and gadodiamide-containing samples using the Arsenazo
III or NM-BAPTA method (average absolute difference 0.0 mg/dL, range 0.0-0.3
mg/dL). In samples containing gadobenate dimeglumine, gadoxetate disodium
and gadofoveset trisodium, serum calcium results differed by <0.2 mg/dL between
contrast-added and control samples (average absolute difference 0.0 mg/dL, range
0.0-0.2 mg/dL) at all concentrations of contrast agent using the o-CPC, NM-BAPTA,
and Arsenazo III methods. During 2013, 732 patient samples were re-tested using the
Arsenazo III method costing >$1,200 in additional reagent, calibrators and quality

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S81

Tuesday, July 29, 9:30 am 5:00 pm

control. Turn-around time for reporting of calcium results was 25-35 minutes longer
for specimens requiring re-testing.
Conclusions: The Roche Diagnostics NM-BAPTA method for measuring serum
total calcium does not show clinically significant interference in the presence of four
gadolinium-based contrast agents. By implementing a calcium method that is free
from gadolinium interference, the laboratory improved quality and reduced the risk
of reporting falsely decreased serum calcium results due to gadolinium interference.
Efficienies gained included eliminating a secondary Arsenazo III calcium method,
reducing patient re-testing, and eliminating the reporting delays and costs associated
with re-testing.

S82

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Factors Affecting Test Results

Hematology/Coagulation

Tuesday, July 29, 9:30 am 5:00 pm

oral anticoagulants and to verify the function of protein synthesis in the liver. This
study aims to evaluate the analytical performance of factors II, V, VII and X measured
in Siemens BCS XP System in comparison with Stago STA Compact System
performance.

Tuesday, July 29, 2014

Methods: In a normal population, 18 samples (n = 18) were collected with citrate

as anticoagulant. Within 4 hours after collection, samples were analyzed by
coagulometric method simultaneously on Siemens BCS XP System, using optical
detection and Siemens deficiency factor plasmas and on STA-Compact using
mechanical detection. In order to establish correlation between methodologies, we
evaluate coefficient of correlation, slope and intercept.

Poster Session: 9:30 AM - 5:00 PM

Hematology/Coagulation

A-280

Results: Statistical data is summarized in the table below.

Clinical Utility of Hematological Parameters to Predict Sepsis Prior to Clinical

Presentation in Medical Intensive Care Unit Patients

S. W. Njoroge, M. Thompson, J. M. Colon-Franco, S. Litt, D. A. Anderson,

T. W. Rice, A. P. Wheeler, A. Woodworth. Vanderbilt University School of
Medicine, Nashville, TN
Introduction: Sepsis is characterized by pathogen invasion into the bloodstream and
the hosts response to this invasion. Early detection of sepsis allows for rapid initiation
of therapy, decreases morbidity and mortality, and reduces healthcare expenses.
Objective: To investigate the diagnostic utility of hematological parameters to predict
the onset and severity of sepsis in Medical Intensive Care Unit (MICU) patients prior
to clinical presentation of systemic inflammatory response syndrome (SIRS).
Methods: This retrospective cohort study employed 125 MICU patients with SIRS.
These patients were identified by software that scans electronic medical records and
alerts investigators when a MICU patient meets SIRS criteria. Several hematological
parameters were quantitated on the Sysmex XE 5000 analyzer from specimens
collected 24 or 48h prior to the patient meeting SIRS criteria. Procalcitonin was
quantified in residual plasma using the Vidas B.R.A.H.M.S PCT assay (bioMerieux,
Inc). The diagnoses of non-infectious SIRS, sepsis and sepsis severity were blindly
adjudicated by 2 MICU physicians as: SIRS (n=70) and Sepsis (n=55; severity: sepsis
n=7, severe sepsis n=22, and septic shock n=26). Receiver operator characteristic
(ROC) curves were generated to evaluate the diagnostic utility of these hematological
parameters to predict sepsis.
Results: Areas under the ROC curves (AUC) for each parameter to predict sepsis
or severe sepsis/septic shock are listed in table 1. Four hematological parameters
WBC, ANC, % neutrophils, and IGC, were significantly different between septic
and non-septic patients and between patients with early/no sepsis vs. severe sepsis
or shock. Additionally, WBC, ANC, and % neutrophils showed significantly
improved diagnostic strength to predict sepsis when compared to a sepsis biomarker,
procalcitonin.
Conclusions: Certain hematological parameters measured before onset of overt
symptoms of systemic inflammation accurately predicted patients who developed
sepsis, severe sepsis and shock. The diagnostic utility of these markers may be
improved by combining them into logistic regression models.
Table 1: Diagnostic Utility of Selected Hematological Parameters Measured 24
- 48 Hours Prior to Onset of SIRS to Predict Sepsis and Sepsis Severity
Early/No Sepsis vs. Severe
Sepsis vs. No Sepsis
Sepsis and Shock
Predictor
AUC 95% CI
P value AUC 95% CI
P value
Procalcitonin (PCT) 0.67 [0.57-0.76] 0.001 0.69 [0.60-0.79] <0.001
White Blood Cell
0.7 [0.61-0.79] <0.001 0.68 [0.58-0.77] <0.001
Count (WBC)
Red Blood Cell
0.52 [0.42-0.62] 0.74
0.53 [0.43-0.64] 0.54
Count (RBC)
Immature
Granulocyte
0.60 [0.50-0.70] 0.07
0.61 [0.51-0.71] 0.04
Percentage (IG-%)
Immature
Granulocyte Count 0.66 [0.57-0.76] 0.001 0.66 [0.57-0.76] 0.002
(IGC )
Neutrophils (Neut %) 0.72 [0.63-0.81] <0.001 0.74 [0.65-0.83] <0.001
Absolute Neutrophil
0.71 [0.62-0.80] <0.001 0.71 [0.62-0.80] <0.001
Count (ANC)
Platelet Count
0.52 [0.41-0.62] 0.74
0.52 [0.42-0.63] 0.64

Analytical performance comparison for deficient plasmas

Assay
n Regression eq.
r
Mean SD Selection range
Factor II
18 y=0.3996x + 66.885
0.64 94
18 67% - 133%
Factor V
18 y=0.6791x + 13.734
0.94 109
23 72% - 153%
Factor VII
18 y=1.2714x 35.975
0.91 120
24 78% - 217%
Factor X
18 y=0.7043x + 20.576
0.90 123
21 87% - 155%
Conclusion: We observed good correlation between platforms for the V, VII and X
factors. For Siemens BCS XP, matrix of factor II is of human origin, providing better
performance and low correlation with Stago platform on which the matrix is a mixture
of human serum and bovine plasma.

A-282
Plasma cell myeloma with rare presentation

L. B. Zuppani1, E. H. J. T. Dauar1, S. S. Rodrigues1, O. Fernandes1, C.

L. Oliveira1, O. F. D. Dias Neto2, J. Tabacof2, D. P. Jardini3, F. B. F.
Carvalhares3. 1DASA, So Paulo, Brazil, 2Clnica Oncolgica Paulista, So
Paulo, Brazil, 3Hospital Paulistano, So Paulo, Brazil
Background: Plasma cell myeloma is a multifocal plasma cell neoplasm associated
with an monoclonal immunoglobulin called M-protein in serum and/or urine. The
disease spans a clinical spectrum from assyntomatic to aggressive forms and due to
deposition of abnormal immunoglobulin chains in tissues. The diagnosis is based on
a combination of pathological, radiological and clinical features. Over 90% of cases
of mieloma occur in patients 50 years old or older. The median age at diagnosis is 70
years. Only 2% of cases begin before the age 40.
Objective: The objective of this case is to bring awarness to atypical presentations of
plasma cell myeloma that may hinder its diagnosis. In the present case the difficulty in
diagnosis is due to the atypical age at onset, absence of monoclonal immunoglobulin
in the protein electrophoresis, hypocelularity of bone marrow without typical elements
of the pathology and presence of histiocytes and plasma cells with cytoplasmatic
inclusions.
Clinical case: CLMS, 38 year old woman, addmited to Hospital Paulistano due to
metrorrhagia and pelvic pain. CBC: hemoglobin 5.6 g/dl, platelets: 86,000/ mm3.
Radiographic study: lytic lesions in the hip. Blood marrow count : hipocelularity
of all myeloid cells; 1.6% of plasma cells with abnormal morphology; presence of
cells with crystallized cytoplasmic imunoglobulin with a lamellar pattern. Protein
electrophoresis with absence of M-protein. Blood Immunofixation with presence
of monoclonal kappa without correspondence with heavy chain IgA, IgG or IgM.
Urinary Immunofixation: presence of monoclonal kappa without correspondence with
heavy chain IgA, IgG or IgM. Bone marrow biopsy with imunohistochemistry report:
CD68 clone PG-M1 positive in nomerous histiocytes with crystallized cytoplasmic
imunoglobulin. Kappa positive in most plasma cells with large cytoplasm. CD 138
positive in about 40-50% of cells. FDG-PET: hypermethabolism in the lithic lesions
of the iliac crest. Increased FDG uptake in the bone that corresponds to bone marrow
activity.
Conclusion: The enlarged lamelar cytoplasm of plasma cells and histiocytes, probably
correspond to the cytoplasmic accumulation of immunoglobulin. The identification of
these cells as plasmocites, associated with the imunohistochemistry, FDG-PET and
blood and urinary immunofixation allowed the diagnosis of Plasma Cell Myeloma
of Kappa chain.

A-281
Comparison of coagulation factors assays in two automated platforms

S. TUFIK, M. C. FERES, M. C. DE MARTINO, M. P. MIRANDA, S. C.

F. DA SILVA. AFIP, SAO PAULO, Brazil
Background: The determination of coagulation factors II, V, VII and X in the plasma
is indicated to diagnosis congenital or acquired deficiency factor, to distinguish
dysproteinemias, to aid in therapeutic monitoring of concentrated prothrombin and

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S83

Hematology/Coagulation

Tuesday, July 29, 9:30 am 5:00 pm

Table I. Nutritional composition of standardized meal

A-283
Validation of Thrombin-Antithrombin III Complex by Enzyme-Linked
Immunosorbent Assay in Humans, Non-Human Primates, and Canine Citrated
Plasma to Support Pre-clinical and Clinical Coagulation Studies

C. Phanthalangsy1, J. Stejskal2, R. Giovanelli1, J. Wisniewski1, S. Ramaiah2.

1
Pfizer, Groton, CT, 2Pfizer, Andover/Cambridge, MA
Background: The conversion of prothrombin into active thrombin is a significant
event within the coagulation cascade. Thrombin is primarily inhibited by antithrombin
(ATIII) in which results in a stable inactive proteinase/ inhibitor complex. The
concentration of thrombin-antithrombin III complex (TAT) can be measured, and
represents a sensitive clinical biomarker for the diagnosis of thrombotic disease in
the coagulation cascade.
Methods: Siemens Enzygnost TAT micro immunoassay (Catalog #:OWMG15)
was validated in human, non-human primate (NHP), and canine citrated plasma.
The concentration of TAT was measured in-vitro quantitatively through a sandwich
enzyme immunoassay and used two different antibodies directed against thrombin
and ATIII, respectively. The first incubation step consists of TAT binding to
peroxidase-conjugated antibodies that are attached to the surface of the microtitration
plate against thrombin. The second incubation step consists of a reaction in which the
enzyme-conjugate antibodies are bound to the free ATIII determinants. Any unbound
constituents and excess enzyme-conjugated antibodies are removed by a series of
washes after each incubation. The enzymatic reaction between hydrogen peroxide and
chromogen is stopped by the addition of diluted sulphuric acid. This results in a color
intensity change which is proportion to the concentration of TAT. TAT concentrations
are measured photometrically through the SPECTRAmax 384plus reader within the
kit standard concentration range of 2 to 60 g/L. For higher TAT concentrations, the
sample was diluted using Sekisui TAT deficient material (Catalog #: 203) for humans
and normal TAT levels (<4 g/L) for NHP and canine.
Results: Intra-assay was established through 5-10 replicates and the %CV was 4.6
for human, 10.8 for NHP, and 8.3 in canine. Inter-assay precision was established
through a minimum of 3 separate assay runs and the %CV was 1.50 for human,
15.5 for NHP, and 11.4 in canine. The limit of blank was established by analyzing
10 replicates and determined to be 0.135 g/L for human and 1.3 g/L for NHP. The
lower limit of quantitation was determined as 2.36 g/L with a %CV of 11.28 in
human and 2.10 g/L with a %CV of 15.7 in NHP. The upper limit of quantitation
was determined as 60.15 g/L and %CV of 4.58 in human and 60.0 g/L and %CV of
13.4 in NHP. Dilutional linearity was determined using samples near the upper limit
of the assay calibration curve, and spiked recovery was determined using samples
spiked with several concentrations of kit calibrators. All results were within 20% of
the expected value. Sample freeze/ thaw stability was performed using samples with
concentrations near the lower and middle limits of the assay calibration curve and
consist of 4 freeze/thaw cycles. All results were within 20% of the expected value.
Sample stability was performed for human and canine samples frozen at -80 C.
Human samples were stable up to and including 3 months, and canine samples were
stable up to and including 2 weeks.
Conclusion: All outlined criteria for the validation of TAT in human, NHP, and canine
were met and used to support pre-clinical and clinical coagulation studies.

A-284
Inevitable is fasting time for coagulation laboratory tests - Preliminary
evaluation

G. Lima-Oliveira1, G. L. Salvagno2, G. Lippi3, E. Danese2, G. Poli2, M.

Gelati2, M. Montagnana2, C. Recchi2, M. Saggiorno2, G. Picheth1, G. C.
Guidi2. 1Federal University of Parana, Curitiba, Brazil, 2University of
Verona, Verona, Italy, 3University of Parma, Parma, Italy
Background: Errors in the preanalytical phase generate further work or additional
investigation that may cause unnecessary procedures for patients. This study was
aimed to evaluate the inevitability of fasting time for coagulation laboratory tests.
Methods: The first blood sample was collected from 10 healthy volunteers at fast
(12h). Immediately after blood collection, the volunteers consumed a standardized
meal (Table 1).

S84

The following blood samples were collected 1, 2 and 4 hours after the end of the
meal. Each phase of sample collection was standardized, including use of needles
and vacuum tubes of the same type and lot. Coagulation tests included the following:
activated partial thromboplastin time (aPTT sec), prothrombin time (PT sec),
fibrinogen (mg/dL), antithrombin III (AT %), protein C (PC %) and protein S (PS %).
The significance of differences between samples was assessed by paired Students
t-test after checking for normality by the DAgostino-Pearson omnibus test. The level
of statistical significance was set at p < 0.05.
Results: One hour after food intake, variations were observed for PT (-2.3%, P=0.45)
and AT (1.9%,P=0.04). Two hours after meal differences were observed for aPTT
(-4.0%, P=0.03) and PT (-3.8%,P=0.57). Statistically significant increases could
be observed four hours after the meal only for AT (3.0%,P=0.02). The results of
fibrinogen, PC and PS were not influenced by the meal at any time points.
Conclusion: Significant variations of aPTT, PT and AT coagulation laboratory tests
after a standardized meal were observed. In conclusion these preliminary outcomes
had shown that the fasting time should be carefully considered when performing these
tests, in order to prevent spurious results and reduce laboratory errors especially in the
therapeutic monitoring.

A-285
Mean platelet volume in patients with pre-eclampsia

W. Lee, S. Cho, E. You, Y. Jeon, H. Lee, T. Park. Kyung Hee Medical

school, Seoul, Korea, Republic of
Background Larger platelets have greater haemostatic efficiency than smaller ones
by producing larger amounts of vasoactive and prothrombotic components. Mean
platelet volume (MPV) is a useful marker indicating alteration of platelet activity
which shows association with various inflammatory diseases. Pre-celampsia (PE) is a
disease characterized by endothelial damage, elevated intravascular platelet activation
while increased MPV in the second of third trimester of pregnancy has been reported.
The aim of this study is to evaluate possibility of MPV as a marker of recovery from
PE or eclampsia after delivery.
Methods Twenty-one pre-eclamptic and one eclamptic women who gave birth at
Kyung Hee Medical Center during January 2011 to June 2012 were include in the
study. The results of white blood cells (WBC) count, hemoglobin (Hb) concentration,
platelet (PLT) count, and MPV were obtained using an automated hematologic
analyzer, Advia 2120 (Siemens Diagnostics, Tarrytown, NY, USA). Medical records
were analyzed retrospectively. Postpartum laboratory data of 16 patients with existing
serial results of the day of delivery, a day after and two days later were analyzed
retrospectively.
Results Nineteen patients had preterm delivery and three had term delivery. Major
comorbidities of diabetes mellitus (DM) and myoma were presented in eight patients.
Both MPV and platelet count were shown to have decreasing tendency over the
observed period, although statistical significance was not shown (P=0.152 and
P=0.327, respectively). A ratio of MPV/PLT was slightly increased without statistical
significance (p=0.222). Longitudinal serial data of six postpartum days in 5 preeclampsia women showed definitely decreasing tendency.
Conclusions MPV has shown decreasing tendency during postpartum periods
following PE or eclampsia. These results demonstrate reduced platelet activity and
decreased maternal intravascular systemic inflammation after resolution of pregnancy.
This study provides evidence that MPV is able to reflect the recovering state from PE
or eclampsia in postpartum periods. Most of the previous
studies were focused on increased MPV at pregnant period of PE. Meanwhile, we
carefully suggest MPV could be supportive surrogate marker as recovery from
inflammatory state after delivery in PE based on these results of our study. Large scale
prospective studies are required to affirm these findings and elucidate the underlying
mechanism to induce changes of MPV in PE or eclampsia.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Hematology/Coagulation

Tuesday, July 29, 9:30 am 5:00 pm

A-286

A-288

Capillary electrophoresis identification and laboratory evaluation of a complex

protein finding in a patient serum

Targeted Metabolomic Profiles Are Strongly Correlated With Metabolic

Alterations In Patients With Sickle Cell/Beta Thalassemia Disease

L. B. Porto, P. R. C. Souza, L. L. Leite, Z. R. M. Cardoso, C. F. A. Pereira.

DASA, RIO DE JANEIRO, Brazil

I. Papassotiriou1, F. Panetsos2, T. Livadara2, A. Tzivaras3, E. Voskaridou4.

1
Department of Clinical Biochemistry, Aghia Sophia Childrens
Hospital, Athens, Greece, 2Analytical Chemistry Laboratory, Bioiatriki,
Athens, Greece, 3Blood Transfusion Center, Aghia Sophia Childrens,
Athens, Greece, 4Thalassemia Center, Laikon General Hospital, Athens,
Greece

Background: A 74-years old female was evaluated by her primary care physician, who
ordered basic routine tests. All results were within the reference intervals, except the
serum electrophoresis, which revealed two unusual intense bands expressed in the
gamma region.
The aim of our study was to better characterize the finding according to the guidelines
in order to provide more information to the physician.
Methods: Laboratory analysis was conducted by performing a high resolution
capillary electrophoresis using the Capillarys (Sebia) for protein identification. In
order to define abnormal protein type, immunofixation of Immunoglobulins and
and light chain, antisera was performed using agarose gel and reagents with the
Hydrasys Electrophoresis System (Sebia). Subsequently, the direct measurement to
define the level of immunoglobulins and light chains was performed by nephelometry
using BN II (Siemens Healthcare Diagnostics).
Results: Immunofixation with Immunoglobulins and and light chain antisera
showed the presence of two monoclonal bands against IgG and an IgA compatible
with a biclonal gammophaty. The direct nephelometric measurement revealed a
normal IgG level of 1010 mg/dl, a high IgA level of 849 mg/dl, a normal IgM level of
60.5 mg/dl, a high level of light chain level of 416.0 mg/dl and light chain level of
228.0 mg/dl. The / ratio of 1.82 was within the reference interval.
Conclusion: In this particular case, the initial diagnosis was apparently a polyclonal
distribution, and although not initially requested, the use of complementary tests like
quantification and immunological identification, by immunoglobulins profile, were
important to help the physician to streamline the diagnosis, monitor and stratify the
risk of this biclonal pathology.

A-287
Mean platelet volume (MPV) in patients with chest pain

G. Lobreglio, F. Sicuro, M. Renis. Azienda ASL Lecce Presidio Ospedaliero

Vito Fazzi, Lecce, Italy
BACKGROUND AND OBJECTIVES. Variation in platelets size is indicative of
change in platelet function; mean platelet volume (MPV) is directly related to platelet
activation in vivo, which plays a central role in the pathogenesis of many vascular
diseases. Recent studies have shown conflicting results about the relationship between
increasing MPV and cardiovascular disease. This work was aimed at investigating if
there is a difference in mean platelet volume between patients with acute myocardial
infarction (AMI) and subjects with other cardiac diseases or non-cardiac chest pain,
owing to platelets activation in acute coronary syndromes.
METHODS. The study included 5927 consecutive patients with acute non traumatic
chest pain presenting to the emergency department of a medium-sized hospital
over a period of 30 months, from January 2011 to June 2013. The median age of
study participants was 61 years (range, 47 to 93); 3051 (51,5%) were male, 2876
(48,5) were female. EDTA anticoagulated whole blood and serum samples were
obtained on admission from all patients. According to the Universal Definition of
Myocardial Infarction, AMI was diagnosed in patients showing a rise and/or fall of
cardiac troponin I (cTnI) above the diagnostic threshold for MI (0,30 ng/mL), typical
changes of electrocardiogram, imaging evidence of new loss of viable myocardium or
presence of an intracoronary thrombus. MPV was measured in the course of complete
blood count with Sysmex XE-2100 automated hematology analyzer using the Hydro
Dynamic Focusing and Direct Current Detection. cTnI was measured by Architect
i System with chemiluminescent microparticle immunoassay (CMIA) from Abbott
Laboratories.
RESULTS AND CONCLUSIONS. AMI was diagnosed in 652 (11%) patients with
chest pain (364 men, 288 female). Mean MPV measured at hospital admission was
11.2 0,7 femtoliter (fL) in patients with AMI and 10.4 0,5 fL in all other subjects,
with a mean difference of 0.8 fL (95% CI 0.6 1.1; P < 0.05). These results show
that MPV is higher in patients with acute myocardial infarction compared to those
with other cardiac diseases or non-cardiac chest pain, suggesting a substantial role of
platelets activation in the formation of the thrombus that occludes the culprit coronary
arteries. Therefore, MPV is a potentially useful biomarker of platelet activity in the
setting of coronary atherothrombotic events.

Background: The complex pathophysiology of Sickle Cell Disease (SCD) makes

unlikely that a single therapeutic agent will prevent or reverse all SCD complications.
Metabolomic analysis might help in the characterization of the endogenous and
exogenous effects of potential new treatments. Metabolites are small molecules that
are chemically transformed during metabolism and provide a functional readout of
cellular state. Metabolites serve as direct signatures of biochemical activity and are
therefore easier to correlate with phenotype. The metabolome is typically defined
as the collection of small molecules produced by cells and offers a window for
interrogating how mechanistic biochemistry relates to cellular phenotype. There are
very few reports providing comprehensive measurements of metabolites present in
blood and almost no reports on metabolites changes associated with SCD. In this
context we aimed to detect and to quantify targeted metabolites abnormalities in
patients with Sickle Cell/beta thalassemia disease (HbS/Thal) and their implication
in pathways that might be of interest to prevent vaso-occlusion and/or to monitor the
effects of new therapies on SCD.
Patients and Methods: Thirty adult patients with HbS/Thal were enrolled in the
study, while 20 apparently healthy individuals matched for age served as controls.
Targeted metabolome analyses based on aminoacids and carnitines were performed
after extraction from dry blood spots (DBSs) on filter paper using High Performance
Liquid Chromatography followed by tandem Mass Spectrometry (LC/MS/MS),
with derivatization (AB SCIEX 5500 triple quadrupole and QTRAP LC/MS/
MS Systems, Framingham, MA, USA) with reagents provided by Chromsystems
Instruments & Chemicals, Germany. The injection volume was 10 L and the analysis
lasted ca. 2 min.
Results: The main results of the study showed that: a) fifty metabolites were separated
in patients and controls samples, b) mapping the results of analyses, patients with
HbS/Thal compared to controls had 17 metabolites with significantly lower
concentration, 10 metabolites with comparable concentration and 23 metabolites with
significantly higher concentration, c) L-arginine and L-ornithine concentrations were
significantly lower in patients HbS/Thal compared to controls, 9.32.6 vs 14.73.7
moles/L, (p<0.01), and 116.015.0 vs 211.219.5 moles/L, (p<0.06) and d)
carnitine, acetylcarnitine and propionylcarnitine correlated significantly positive with
reticulocyte production index (p<0.001).
Conclusions: Our findings showed significant alterations in whole blood metabolome
of patients with HbS/Thal. Also we demonstrated the very important metabolic
abnormality of Nitric Oxide biosynthesis pathway due to the low concentration of
the aminoacids serving of substrates in this cycle and disturbances in carnitine and
acylcarnitines homeostasis. These abnormalities in the metabolome reflected the
hemolysis, inflammatory process and pulmonary hypertension observed in these
patients.

A-289
Validation of Fluorescence in Situ Hybridization assay for detection of
rearrangements involving the Mixed Lineage Leukemia gene

F. K. Marques, W. G. Teixeira, P. C. ngelo, E. C. C. Mateo, A. C. S.

Ferreira. Instituto Hermes Pardini, Vepasiano, Brazil
Background: The Fluorescence In Situ Hybridization (FISH) testing has become
an important tool of clinical practice in many laboratories dealing with neoplastic
deseases. The FISH is more sensitive than conventional chromosomal analysis
because can detect a specific alteration in both dividing and nondividing cells and
small population of abnormal cells. Although the performance of the most FISH
probes has been evaluated by the manufacturer prior to marketing, they also must
be validated prior implementation of assay for clinical use. Clinical laboratories
must independently adopt protocols for verify the performance of the assay.
Rearrangements involving the Mixed-Lineage Leukemia gene (MLL) represent 10%
of abnormalities detected in acute leukemias, which in many cases represent poor
prognosis. Thus the rapid identification of these rearrangements is needed to guide
prognosis and treatment.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S85

Hematology/Coagulation

Tuesday, July 29, 9:30 am 5:00 pm

Objetive: To validate FISH assay for detection of rearrangements involving the MLL
gene following recommendations from the American College of Medical Genetics
(ACMG).
Methods: We use the MLL dual-color breakapart probe manufactured by Cytocell.
In the familiarization phase the analysts should become familiar with the probe
labeling, testing probe strategy and result reporting. Metaphase cells obtained from
5 karyotypically normal male blood samples were used to localize the probe and
determine its analytical sensitivity and specificity. To establish a reference range
(normal cutoff) were estimate the false positive rate from 20 uncultured bone marrow
samples that would be unlikely to harbor a MLL rearrangement. Two analysts score
500 interphase cells (250 per analyst). All MLL probe signal patterns were recorded.
The normal cutoff for each signal pattern was calculated using the BETAINV and
CRITBINOM function available in Microsoft Excel.
Results: The MLL breakapart probe presents two portions of the gene, differentially
labeled: the proximal 5 labeled with green fluorophore, the distal 3 labeled with red
fluorophore. A typical result of using this probe should show 2 fusions of green and
red signals (2F). The separation of the fusion signals indicates the presence of MLL
rearrangement. The clinical validation of FISH showed rearrangement involving MLL
gene, in agreement with conventional karyotyping. The probe demonstrated 100%
specificity and analytical sensitivity. Among the 20 bone marrow samples analyzed,
six had one or two false positive cells, which have been designated by the abbreviation
1F1R1G (one fusion, one red, one green). MLL atypical probe signal patterns were
also found: 1F, 3F, 1F1G and 2F1G. The cutoffs obtained with BETAINV function
was validated for counting 200 cells. The signal patterns and respective normal
cutoff are 1F1R1G (2.3%), 1F(4.4%), 3F(1.4%), 1F1G(3.0%) and 2F1G(1,4%). The
BETAINV function does not allows a cutoff of zero and show minimal change with
increasing cell score. It can minimize the problem of false positive signals that may
be due to probe random loss or gain of signals, diffuse probe signals in cells with less
condensed chromatin, probe size and design, sample type and others.
Conclusion: The FISH for MLL gene rearrangements with Cytocell probe was
considered approved.

A-291
Thrombocytopenia In Children with Malaria

C. E. Lekpor1, W. Ababio2, R. Ansah3, S. Amankwah4, F. A. Botchway5.

1
Department of Medical Laboratory, Kwame Nkrumah University Of
Science and Technology, Kumasi, Ghana, 2Hematology Department, Korle
Bu Teaching Hospital, Accra, Ghana, 3University of Ghana, Accra, Ghana,
4
Chemical Pathology Department, University of Ghana Medical School,
Accra, Ghana, 5Department of Child Health, Korle Bu Teaching Hospital,
Accra, Ghana
Background: The objective of this study was to access the occurrence and severity of
thrombocytopenia in children with malaria at the Child Health department of Korle
Bu Teaching Hospital between January and June 2013.
Methods: It was a retrospective study, done at the Child Health Department of Korle
Bu Teaching Hospital. Data regarding positive cases of malaria <
12 years attending the Out Patient Department and admitted at the Emergency Room
between January 2012 and June 2012 were obtained. Patients were further assessed
for thrombocytopenia and its severity. Data were analyzed by Chi square test using
SPSS version 13.0.
Results: A total of 131cases were included in the study with a mean age of presentation
of 8 years. Plasmodium falciparum was identified in 119 (90.8%) patients while
Plasmodium malariae in 6 (4.5%) patients and Plasmodium ovale in 3 (2.3%) with 3
(2.3%) cases or mixed infections of Plasmodium falciparum and Plasmodium ovale.
Thrombocytopenia was observed in 93(71%) cases, of which 40(31%) cases had
mild, 56(43%) cases moderate and 34(26%) cases had severe thrombocytopenia.
Thrombocytopenia was equally found in falciparum, malariae and ovale infections
with no significant difference in severity between falciparum, malariae and ovale
species.
Conclusion: Thrombocytopenia is frequently seen in malaria and it is not dependant
on type of malaria. In any acute febrile illness, thrombocytopenia should alert one to
the possibility of malaria.

A-292
Hemophilia C with a Cys482Trp Mutation in the F11 Gene

J. Yoo1, B. Kim2, J. Choi2. 1NHIS Ilsan Hospital, Goyang, Korea, Republic

of, 2Yonsei University College of Medicine, Seoul, Korea, Republic of
Background: Coagulation factor XI (FXI) is a member of the contact pathway and
is either activated intrinsically by coagulation factor XII (FXII) or by thrombin, which
is produced by an extrinsic pathway and plays an important role in hemostasis. Factor
XI deficiency, also known as hemophilia C, is a predominantly autosomal recessive
genetic bleeding disorder. The F11 gene encodes the FXI protein, and mutations in
the F11 gene have been found in patients with FXI deficiency. We report the first case
of a heterozygous mutation (Cys482Trp) in the F11 gene, resulting in a mild FXI
deficiency in a Korean patient.
Methods: A 14-year-old male patient with intermittent epistaxis was admitted to
the hospital because of increased epistaxis frequency. The patient did not have any
abnormal medical history that would have indicated bleeding tendency, apart from
being treated for allergic rhinitis. The blood test results were as follows: leukocyte
count, 5.6 109/L; hemoglobin level, 159 g/L; and platelet count, 215 109/L. The
blood coagulation test showed normal prothrombin time but prolonged activated
partial thromboplastin time (aPTT; 45.9 seconds, reference range, 20.9-35.0 seconds).
The level of von Willebrand factor was within the reference range. The activity levels
of factors II through XII were within the reference ranges, whereas FXI showed
slightly decreased activity (26%; reference range, 60-140%). The patients genomic
DNA was extracted from the collected whole blood sample by using the Easy-DNA
Kit (Invitrogen Corporation, Carlsbad, CA, USA). Exons 7, 8, 11, and 13 of the F11
gene were amplified by polymerase chain reaction (PCR). Direct sequencing of the
amplified regions was performed by using the ABI Prism 3500dx automated genetic
analyzer (Applied Biosystems, Foster City, CA, USA), the same primers that were
used to amplify the 4 exons of the F11 gene by PCR, and the Big Dye Terminator
Cycle Sequencing Ready Reaction Kit (Applied Biosystems). To determine whether
the patient had a DNA mutation, Sequencher software (Gene Codes Corporation, Ann
Arbor, MI, USA) was used to compare the patients DNA sequences to the reference
DNA sequence (GenBank accession number, NM_000128.3). The guidelines of
the Human Genome Variation Society (HGVS) were used to identify the sequence
variations.
Results: Direct sequencing of the proband demonstrated a heterozygous mutation,
c.1500C>G (p.Cys482Trp) of exon 13 in the F11 gene.
Conclusion: The Cys482Trp missense mutation identified in our patient had been
previously reported in England, but has never been reported in Korea. Owing to
the small number of FXI deficiency cases with associated mutations that have
been reported in Korea to date, further studies are warranted to contribute to the
development of a database that would help clarify the distribution of mutations
present in the Korean population. Such a database would also facilitate studies aimed
to clarify the possibility of a founder effect for F11 gene mutations. Furthermore,
these efforts would help improve the molecular and genetic diagnostic strategies used
for Korean patients.

A-293
The utility of flow cytometry in the diagnosic of hemolytic anemias

B. Alvarez, A. Gonzlez, D. Velasco, N. Del Amo, M. Medrano, L. Conejo,

A. Vlagea, J. Alonso, J. Villarrubia, R. Guilln, F. Cava Valenciano.
Laboratorio Central BRSalud, Madrid, Spain
Background: Hereditary spherocytosis (HS) is the most common inherited anaemia in
Northern Europe and North America. The red cells in HS show abnormal morphology
attributable to a deficiency or dysfunction of one of the red cell cytoskeleton protein
(spectrin, ankyrin, band 3 and/or protein 4-2). Recently is used the flow cytometric
test that measures the fluorescence intensity of red cells labelled with the dye eosin5-maleimide (EMA), which reacts covalently on the first extracellular loop of band
3 protein. By this technique a decreased expression of Band 3 in the red membrane
surface results in a lower mean fluorescence intensity of EMA. The aim of our study
is to differentiate between HS and immune and nonmembrane-associated haemolytic
anaemias, such as autoimmune hemolytic anemia (IHA) and Thalassemia (TL).
Method: The EMA-binding test was performed as described by King et al. with minor
modifications in 150 controls, 44 HE, 17 IHA and 15 Thalassemia patients, using a
BD FACSCanto II flow cytometer.
Results: Significant difference in EMA mean fluorescence intensity (MFI) results was
obtained between HE and all of the groups studied (control group (p<0.001); IHA (P=

S86

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Hematology/Coagulation

Tuesday, July 29, 9:30 am 5:00 pm

p<0.001), and TL patiens ( (p<0.001)). There were differences between IHA and TL
patients (p<0.022) while we dont find significant differences between TL patiens vs
control group (p=0.533). See figure 1.
Conclusion: Measuring the fluorescence intensity of EMA labeled red cells by flow
cytometry could be a powerful tool in the study of hemolytic anemias being a method
available in most haematology laboratories. In our case has been shown to be effective
for the discrimination of hereditary spherocytosis and IHA versus normal controls
and other hemolytic anemias. We found a specific pattern of EMA expression in IHA
probably due an increased level of reticulocytes in these patients.

Conclusion: The results indicate applicability of this immunoturbidimetric assay to

the determination of haptoglobin in serum samples. The inclusion of a new ready to
use reagent leads to a simplified procedure and a reduction of handling errors prior to
analysis. This is of value for application in clinical laboratories.

A-295
Free Protein S correlation between two systems

S. TUFIK, M. C. FERES, M. C. DE MARTINO, M. P. MIRANDA, S. C.

F. DA SILVA. AFIP, SAO PAULO, Brazil
Background: Protein S is a vitamin K plasma protein synthesized by the liver. During
blood coagulation, protein S forms a complex with APC which bonds to phospholipids
surface and speeds up the proteolytic inactivation catalyzed by APC of the factors Va
and VIIIa. The free portion of protein S corresponds to 40% of total protein S. A total
or acquired protein S deficiency is associated with a venous thromboembolism risk.
There are three types of hereditary deficiencies: type I involves a reduction in total
and free protein S levels; type II, a rare form, presents fall in protein S activity and
normal levels of total and free protein S; and finally, type III presents normal levels
for total protein S and low levels for the free form. Acquired protein S deficiency
may be caused by liver dysfunction, nephrotic syndromes and vitamin K deficiency,
pregnancy, treatment with L-asparaginase, estrogen therapy, virosis and disseminated
intravascular coagulation. This study aims to evaluate the free protein S analytical
performance on Siemens BCS XP System compared to obtained by Elite Pr IL
system to decide on the introduction of these tests in routine.
Methods: Fresh and frozen blood samples were dosed (n=32), collected with citrate
as an anticoagulant. The sample range selection situates between 40% and 122%.
Plasma was analyzed within two hours after unfreeze simultaneously in Siemens
BCS XP, using Siemens INNOVANCE Free PS-Ag, and IL Protein S, applying
the turbidimetric method through optical detection for both systems. Correlation
coefficient, slope and intercept were evaluated to examine the correlation degree
between the methodologies of this study.
Results: Comparison resulted in a regression equation y = 0.908x + 17.35 and a
correlation coefficient r = 0.83. The simple concordance index is 84%, concordance
chance 0.63%, kappa index of 0.57, classified as moderate concordance. From
five discordant samples, four are very close of a normality value and dont cause,
therefore, clinical impact.

A-294
Development of An Immunoturbidimetric Assay for the Determination of
Haptoglobin Incorporating a New Ready to Use Reagent

S. Baxter, S. McElhatton, M. Rodriguez, J. Campbell, S. FitzGerald.

Randox Laboratories Limited, Crumlin, United Kingdom
Background: The primary function of the plasma protein haptoglobin is to bind to
free haemoglobin thereby preventing haemoglobin-driven oxidative tissue damage,
the renal excretion of iron and the subsequent kidney damage following intravascular
hemolysis. The plasma levels of this protein are reduced during episodes of hemolysis
and the measurements are used in the diagnosis of haemolytic anaemia. Haptoglobin
is also a positive acute-phase protein with immunomodulatory properties, the levels
of this protein are elevated in inflammatory, infectious processes and in malignancies.
This study reports the development of an assay for the determination of haptoglobin
in serum samples, which incorporates a new ready to use reagent leading to a
simplified procedure and a reduction of handling errors prior to analysis. The assay
is applicable to a variety of automated analysers. This is of value for application in
clinical laboratories.
Methods: The assay is immunoturbidimetric, the sample containing haptoglobin reacts
with anti (human) haptoglobin antibody; insoluble complexes are formed allowing
quantitative measurement at 340 nm. The amount of complex formed is proportional
to the concentration of haptoglobin in the sample. The reagent is liquid and ready to
use. The assay is applicable to a variety of automated systems. Accelerated stability of
the reagent was assessed: three lots of reagent in the final packaging were nonstressed
(stored at 20C to 80C) or heat-stressed (at 300C, 370C and 400C for 2, 4, 8, 13, and
26 weeks) and run in parallel. Within-run precision was assessed by testing serum
samples at defined medical decision levels, 2 replicates of each sample were assayed
5 times for 1 day. Correlation studies were conducted using a commercially available
assay system.
Results: Initial evaluation assigned 56 weeks real time stability for the reagent. The
assay showed a sensitivity of 0.19 g/L and was linear up to 3.66 g/L. Prozone was not
observed to a concentration of 4.65 g/L. The within-run precision for levels 0.85 g/L
and 2.85 g/L, expressed as %CV, was 1.0% and 1.2% respectively. In the correlation
study 95 serum patient samples were tested and the following linear regression
equation was achieved: y = 1.03x - 0.04; r = 0.999.

Conclusion: We conclude the Siemens INNOVANCE Free PS-Ag analytical

performance on Siemens BCS XP is comparable to Elite Pr IL. Thus, these tests
were approved for laboratory routine implementation.

A-296
A Dual Monoclonal Antibody Chemiluminescent ELISA for the Detection of
Hepcidin-25

J. Chapo, A. Jones, K. R. Pitts. Corgenix, Inc., Broomfield, CO

Background: Iron is essential for all metazoans. Hepcidin is the peptide hormone
that helps maintain plasma iron homeostasis and is an important clinical biomarker.
It is first synthesized as an 84-amino acid precursor which is then proteolytically
processed into a 60-amino acid pro-peptide and subsequently into the mature and
bioactive 25-amino acid peptide. Hepcidin-25 regulates iron bioavailability by
binding to the iron exporter ferroportin, causing its internalization and degradation
from the surface of duodenal enterocytes, placental trophoblasts, macrophages and
hepatocytes. Dysregulation of plasma iron homeostasis contributes to the anemia
associated with chronic diseases such as infection, rheumatoid arthritis, and various
cancers. As hepcidin-25 is a key regulator of iron homeostasis, a reliable method to
detect hepcidin-25 levels in a blood sample would greatly improve our understanding
of how to impact the anemia of chronic disease. Here we report the development of a
robust immunoassay to measure hepcidin-25 in a blood sample.
Methods: A monoclonal antibody pair (capture and detection) was characterized and
optimized for assay development. The capture antibody was coated onto opaque white
96 well plates while the detection antibody was conjugated to horseradish peroxidase
(HRP) using standard immunochemistry methods. Antibody coat concentrations and
conjugate detection concentrations were systematically titrated to achieve optimal
sensitivity and dynamic range. Numerous iterations of coating and blocking buffers
were assayed to further enhance assay manufacturability and consistency. Finally, inprocess testing of linearity and precision was conducted to ensure robust performance
to the end-user.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S87

Hematology/Coagulation

Tuesday, July 29, 9:30 am 5:00 pm

Results: The assay is specific for hepcidin-25, and does not cross-react with either
hepcidin-22 or hepcidin-20. Linearity testing was performed over multiple kit lots
and demonstrated a linear range of 60-6000pg/mL, spanning both normal and disease
state levels of hepcidin-25 previously reported. Precision of the assay ranges from
5.0-14.0%CV, with sample recoveries falling between 20%. The assay was used to
test a set of normal and non-hematologic cancer samples. We found that, while a range
of hepcidin-25 levels was observed in non-hematologic cancer samples, the assay was
able to differentiate between normal and disease state sample populations.
Conclusions: Unlike other hepcidin assays, ours utilizes paired monoclonal antibodies
that facilitate the direct detection of hepcidin-25 in a sandwich ELISA configuration.
The assay also takes advantage of chemiluminescent detection to improve the
sensitivity and stability of signal. Taken together, the data support the utility of this
monoclonal-based chemiluminescence sandwich ELISA for the specific detection of
hepcidin-25.

A-297
Evaluation of the Newly-Developed Serum Ferritin Measurement System, Point
ReaderTM and Point StripTM Ferritin

R. Kubota1, K. Kanamori2, Y. Kawai3, N. Sakai1, K. Shiba2. 1Saitama

Prefectural University, Saitama, Japan, 2Bunkyo Gakuin University, Tokyo,
Japan, 3USHIO INC., Tokyo, Japan
Background: Serum ferritin is widely measured to evaluate the iron levels in the
body. Point ReaderTM (USHIO, Tokyo, Japan) and Point StripTM Ferritin (ASKA,
Kanagawa, Japan), a newly-developed serum ferritin measurement system, is based
on the principle of immunochromatography. This system measures serum ferritin
concentrations in five minutes and may allow detection of the early iron-deficiency.
Therefore, the objective of this study was to evaluate the performance of this new system.
Methods: Unidentified residual serum specimens with known ferritin concentrations
were used. Serum ferritin concentrations were measured using Point Reader and
Point Strip Ferritin according to the manufacturers instructions. The absorbance
of phosphate buffered saline was measured to ascertain the sensitivity. The
reproducibility of the test was determined by taking 8 replicate measurements of the 3
standards that had low, medium, or high levels of ferritin (12.0, 39.7, and 78.8 g/L,
respectively). For the correlation test, serum ferritin was measured using N-assay LA
Ferritin (NITTOBO, Tokyo, Japan) on TBA-40FR Accute (TOSHIBA, Tokyo, Japan).
Results: The minimum detection limit of the serum ferritin using Point Reader and Point
Strip Ferritin was determined to be 10.0 g/L. In addition, the measurement range of this
system was 10.0-100 g/L as determined from linearity testing. This system has a 3.8%
reproducibility in a low control, 6.3% in a medium control, and 6.1% in a high control.
We examined the correlation between serum ferritin concentrations obtained using
this system (y) and those obtained using TBA-40FR Accute (x) (n = 107). The linear
regression equation was y = 1.07x + 0.84, and the correlation coefficient (r) was 0.956.
Conclusions: The newly-developed Point Reader and Point Strip Ferritin measurement
system is simple and provides fast measurement of serum ferritin concentrations. This
system can be used to assess the iron deficiency during pregnancy as well as in infants
and blood donors.

A-298
Mathematical Model-based Estimation of Red Blood Cell Clearance Identifies
Low Iron States in Patients with Normal Complete Blood Counts

H. Patel, J. Higgins. Massachusetts General Hospital: Center for Systems

Biology, Boston, MA
Introduction The healthy human hematologic system is held in a state of dynamic
equilibrium by the carefully regulated processes of (1) cell production, (2) cell
maturation in the peripheral circulation, and (3) cell clearance. The resulting steady
state is routinely quantified by measurements such as hematocrit (HCT), hemoglobin
(HGB), or red blood cell count (RBC). Diseases and other conditions perturbing one
of these processes may trigger compensation in the others. For instance, decreasing
iron availability will eventually lead to iron deficiency and anemia, but prior to the
development of anemia, a decrease in RBC production may be compensated by a
delay in RBC clearance. This compensatory delay in clearance will maintain the
steady state -- and will also confound our ability to discover the problem because the
HCT, HGB, and RBC will remain normal. We hypothesize that an estimate of the rate
of the underlying RBC clearance process would serve as a useful screening biomarker
of decreasing iron availability, helping to identify seemingly healthy patients whose
iron levels are close to abnormal or are already abnormally low. These patients should
have iron levels checked and may need to be followed more closely.
Methods We used a mathematical model of RBC population dynamics to infer RBC
maturation and clearance rates for patients from routine CBC and reticulocyte counts
performed on an Abbott CELL-DYN Sapphire automated hematology analyzer. We
first established a normal range for the estimated RBC clearance threshold, defined the
probability of clearance as a function of an RBCs volume and hemoglobin content.
We identified patients whose CBCs were entirely normal according to existing CBC
indices but whose RBC clearance thresholds fell below the 5th percentile of the normal
range. We then measured ferritin levels for these patients at the time of their entirely
normal CBC.
Results We found that among patients with an entirely normal CBC and a low
modeled RBC clearance threshold, 5% had abnormally low ferritin levels, and 19%
had ferritin levels that were within 5% of the lower limit of the normal range. Low
and low-normal ferritin was 4x more prevalent in this low clearance group than
in a control group whose RBC clearance threshold was normal. We also found a
statistically significant relationship between the estimated RBC clearance threshold
and the ferritin level, with lower clearance threshold associated with a lower ferritin
level.
Conclusions Iron deficiency anemia is an early sign of a number of important
diseases, such as colon cancer, gastric cancer, celiac disease, and more. Our results
suggest that RBC clearance threshold is often reduced, possibly as compensation to
maintain desired steady states. This compensation confounds our ability to detect
the illness by measuring HCT or HGB. But by using a mathematical model of in
vivo RBC population dynamics, we can estimate a patients clearance threshold and
identify patients likely to have low ferritin levels. These patients would benefit from
direct iron level assessment and further revaluation to identify the underlying cause of
the decreased iron availability.

A-300
Distribution of hemoglobinopathies and thalessemias in a Northern Alberta
population

T. Higgins, k. Rodriguez Capote. DynaLIFEDx,, Edmonton, AB, Canada

Background DynaLIFEDx is the sole laboratory performing hemoglobinopathy
and thalassemia investigations for a catchment area of 2 million people in northern
Alberta. Samples for hemoglobinopathy and thalassemia testing come from physician
request, immigration medicals, and red cell exchange programs for sickle cell anemia
patients and as a reflex test generated by the presence of a hemoglobin variant noted
on HPLC analysis for HbA1c. We wished to know the frequency of thalassemia and
hemoglobinopathies in our catchment area.
Methods EDTA-anti-coagulated blood samples were analyzed by high performance
liquid chromatography (HPLC) using the thalassemia program on the Bio-Rad
VARIANT II. Hemoglobin variants found on HPLC were analyzed by electrophoresis
at alkaline and acid pH. A complete blood count (CBC) and ferritin tests were
requested as part of the hemoglobinopathy/thalassemia investigation.
Results 5562 thalassemia and hemoglobinopathy investigation were physician
requested, 1054 HbS screens were requested and 702 hemoglobin variants fortuitously
found on HbA1c analysis were analyzed in the period January 1, 2013 to December
31, 2013. 3438 were interpreted as normal and 532 were classified as iron deficient

S88

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Hematology/Coagulation

Tuesday, July 29, 9:30 am 5:00 pm

on the basis of their hematology indices (hemoglobin >120g/l, mean cell volume >
80fL), absence of a hemoglobin variant, replete iron status and calculated Mentzer
Index and Discriminant Factor. 322 and 370 were classified as or thalassemiaa
trait respectively. 357 were classified as S trait, 121 as E trait, 63 as D Punjab trait
and 62 as C trait. 22 were classified as Homozygous S, 4 as homozygous HbE and 2
as homozygous HbC and 8 as SC disease. 10 were classified as thalassemia trait
and 11 as H disease. The remainder was classified as unusual hemoglobin variant or
thalassemia.
Conclusion There is a wide diversity of hemoglobinopathies found in northern
Alberta. 11% of the hemoglobinopathies were found as a reflex to HbA1c testing.

A-301
Pediatric Reference Intervals for New Reticulocyte and Platelet Parameters

M. OLeary1, P. Horn2, B. Ward3, P. Steele2, D. Wright4. 1Texas Childrens

Hospital and Baylor College of Medicine, Houston, TX, 2Cincinnati
Childrens Hospital and University of Cincinnati, Cincinnati, OH,
3
Cincinnati Childrens Hospital, Cincinnati, OH, 4Abbott Diagnostics,
Santa Clara, CA
Objective: Advances in algorithm development for the Abbott CELL-DYN Sapphire
has introduced several new red blood cell parameters describing cell-by-cell size and
hemoglobin content for circulating reticulocytes (MCVr, MCHr, CHCr). Reported
clinical applications include monitoring RBC production kinetics, distinguishing
anemia of chronic disease from iron-deficiency anemia (IDA), and detecting
pediatric IDA and erythropoietin-associated functional iron deficiency. Additionally,
the reticulocyte channel measures the percentage of platelets that contain residual
RNA, reported to correlate with megakaryocytic activity. Potential applications for
reticulated platelet (rPLT) values include distinguishing production from consumption
and predicting recovery in thrombocytopenic patients. The objective of this study was
to determine reference intervals for five pediatric age groups.
Methods: 762 K3-EDTA whole blood surplus samples from previously ordered
CBCs were analyzed to collect the reticulocyte channel information. The IRB waived
consent for the study samples. Children with clinical and/or laboratory evidence of
RBC or platelet abnormalities were excluded. The five age groups included; Group
1 = 1 month - 1 year; 2 = 1 - 3 years; 3 = 3 - 6 years, 4 = 6 - 12 years; and 5 = 12 18 years. Reference intervals were calculated (SAS/STAT Software) as per the CLSI
guidelines C28-A3C. The non-parametric calculation was used for groups with >120
samples, and the robust method was used for two groups with <120 samples. The
RBC indices and platelet count were included in the reference interval calculations to
provide a comparison to their reticulated counterparts.
Results: The 95% reference interval limits are in Table 1. Ninety percent confidence
intervals for the ninety-five percent upper and lower reference interval limits were
calculated, as well (data not shown).
Conclusions: Our results suggest age related trends for the new reticulocyte and
reticulated platelet parameters.

A-302
The evaluation of affecting factors on platelet reference ranges in the
population of Northeastern China

S. Y. Wu1, X. Shi2, J. Sun3, F. Hassouna1, C. Wang3, J. Niu2. 1Confirmatrix

Clinical Laboratory, Lawrenceville,, GA, 2Internal Medicine, First
Affiliated Hospital, Jilin University, Changchun, Jilin, China, 3Geriatrics ,
First Affiliated Hospital, Jilin University, Changchun, Jilin, China
Objective: The purpose of this study is to evaluate the clinical laboratory variables,
their correlation to platelet count (PLT), the reliability of established reference ranges
and their applications in clinical research and diagnostic test.
Methods: This study was carried out in Chang Chun city and its suburban area in
China. 3800 cases were selected during year of 2010 to 2012. Inclusive criteria
were as following: 1. No medications within one month; 2. No medical history of
thrombosis and hemorrhagic disease. Exclusive criteria were: 1. Female patients are
at menstrual period; 2. Female who are pregnant; 3.Patients who have medical history
of liver disease or hematologic diseases. Patients who met the inclusive criteria were
put in questionnaire survey for past medical history and social history. 5 ml of blood
was drawn for clinical chemistry analysis including liver function test (to exclude
unknown hepatic disease), blood glucose test and lipid panels. Hematology variables
include platelet count (PLT), mean platelet volume (MPV), plateletcrit (PCT), and
platelet distribution width (PDW). SPSS software has been used for ANAOV analysis.
P<0.05 was used for statistical significant testing in this study.
Results: This study showed PLT ranged from 147 -363 K/l. MPV were from 6.8-10.5
fl. PCT were from 0.13%-0.29%. PDW were from 15,5% -18.4%. These variables
were affected by the cholesterol and triglyceride level, particularly in the group of 50
years or older. The correlations analysis indicated that the higher the cholesterol, the
higher the PLT, MPV, and PCT. The average platelet count was higher in female than
male. Smoking, alcoholic, BMI, blood pressure and blood sugar have not effects on
platelet variables.
Conclusion: Based on these case studies and the analysis of clinical chemistry
variables, we concluded that diet, blood pressure and social behavior do not have
impact on platelet values. Age and high cholesterol cause elevated platelets count.
This supports the previous study that high cholesterol causes high coagulation
diseases including stroke, myocardia thrombosis and deep vein thrombosis.

A-303
Measurement of Serum Ferritin Concentrations in a Japanese Young
Population Using a Newly-Developed System, Point ReaderTM and Point StripTM
Ferritin

K. Kanamori1, R. Kubota2, Y. Kawai3, N. Sakai2, K. Shiba1. 1Bunkyo Gakuin

University, Tokyo, Japan, 2Saitama Prefectural University, Saitama, Japan,
3
USHIO INC., Tokyo, Japan
Background: Iron-deficiency anemia is usually diagnosed by measuring the level of
hemoglobin. However, even if the hemoglobin level is within the reference range, the
serum ferritin concentrations may still be <12 g/L resulting in a latent iron-deficiency
state without anemia. Blood donation has a significant impact on the iron levels of the
body, especially in blood donors with latent iron deficiency. Therefore, the objective
of this study was to evaluate the serum ferritin concentrations in a Japanese young
population, using the newly- developed serum ferritin measurement system, Point
ReaderTM (USHIO, Tokyo, Japan) and Point StripTM Ferritin (ASKA, Kanagawa,
Japan).
Subjects: Serum and blood samples were obtained from 36 male and 108 female
students (age, 20-24 years) from Bunkyo Gakuin University, Tokyo, Japan. The
ethical committee of Bunkyo Gakuin University approved this study, and informed
consent was obtained from all subjects.
Methods: Serum ferritin concentrations were measured using Point Reader and Point
Strip Ferritin based on the principle of immunochromatography. The hemoglobin
levels were measured using XE-2100 (Sysmex, Hyogo, Japan).
Result: The serum ferritin concentrations of the young Japanese population were
shown in histogram. The serum ferritin concentrations in the 36 male students ranged
from 20.2 g/L to 179.1 g/L, while that of 12 (11.1%) female students were <10
g/L. Overall, the serum ferritin concentrations of the male students were observed to
be higher than those of the female students. 23 (21.3%) female students showed latent
iron deficiency state without anemia.
Conclusions: Out of 108 female students, the 21.3% with latent iron deficiency are at
a significant risk of developing iron deficiency anemia from blood loss, such as that

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S89

Hematology/Coagulation

Tuesday, July 29, 9:30 am 5:00 pm

occurring during blood donation. Estimation of the hemoglobin level alone in blood
donors may not be adequate; therefore, the estimation of serum ferritin concentrations
may also be necessary to detect pre-clinical iron deficiency.

A-306
Acute erythroleukaemia with atypical presentation

L. B. Zuppani1, S. S. Rodrigues1, O. Fernandes1, C. L. Oliveira1, J.

Tabacof2, O. F. Dias Neto2, D. P. Jardini3, E. H. J. T. Dauar1. 1DASA, So
Paulo, Brazil, 2Clnica Oncolgica Paulista, So Paulo, Brazil, 3Hospital
Paulistano, So Paulo, Brazil
Background: Acute erythroid leukaemias are a group of acute leukaemias
characterized by the predominance of erythroid population. It is divided into two
groups based on the presence or absence of a significant myeloid component:
Erytroleukaemia (Erythroid/ myeloid) and Pure erythroid leukaemia. Erytroleukaemia
(Erythroid/ myeloid) is diagnosed when the presence of erythroid precursors in the
bone marrow is 50% of the nucleated cell population and 20% of myeloblasts in the
non-erythroid cell population. Pure erythroid leukaemia is diagnosed when 80% of
cells in the bone marrow are commited with the erythroid lineage (undifferentiated or
proerythroblastic in appearance) and there is no significant myeloblastic component.

A-304
Prognostic Value of Modest Increases of Plasma D-dimer Concentration in
Patients with Previous History of Myocardial Infarction

H. Naruse, J. Ishii, R. Okuyama, T. Hashimoto, F. Kitagawa, T. Kuno, T.

Ishikawa, S. Matsui, Y. Ozaki. Fujita Health University, Toyoake, Japan
Background: D-dimer can be considered as a global marker of the turnover of crosslinked fibrin and of activation of the hemostatic system. We prospectively investigated
whether modest increases of plasma D-dimer levels might be relevant to prognosis in
patients with a previous history of myocardial infarction. Methods: We studied 606
consecutive patients with a previous history of myocardial infarction (median age, 65
years; 508 males) and estimated glomelurar filtration rate (eGFR) 15 ml/min/1.73
m2. Blood samples for measurements of D-dimer, total plasminogen activator
inhibitor-1 (tPAI-1) and high-sensitive C-reactive protein (hsCRP) were obtained at
enrollment. Among these patients, 65% had hypertension, 36% had diabetes, 35% had
chronic kidney disease, and 74% had history of coronary revascularization. Results:
During a median follow-up period of 43 months, there were 120 cardiovascular events
including 20 cardiovascular deaths. Comparably, patients who had cardiovascular
event were older (median, 67 vs. 65 years, p = 0.01), had higher levels of D-dimer
(0.65 vs. 0.44 g/ml, p < 0.0001), and displayed a lower level of eGFR (62 vs. 67 ml/
min/1.73 m2, p = 0.02) and left ventricular ejection fraction (LVEF, 50 vs. 53 %, p <
0.0001) than those who did not. On Cox regression analyses multivariate including
12 clinical and angiographic variables, D-dimer levels were independently associated
with cardiovascular events (relative risk: 1.92 per 10-fold increment, p = 0.02).
Clinical characteristics and cardiovascular mortality and cardiovascular event rates
according to tertiles of D-dimer were shown in Table. Conclusion: Modest increases
of D-dimer may be independently associated with adverse outcomes in patients with
a previous history of myocardial infarction. Measurements of D-dimer may be useful
for the risk stratification of adverse outcomes in this population.

Clinical case: MCA, a 71 year old woman was admitted to the orthopedic clinic
at Hospital Paulistano with bone pain. Her first blood count showed mild anemia
(hemoglobin: 10g/dl). On the course of her hospitalization she presented a severe
drop in hemoglobin levels, so a bone marrow count was performed. Bone marrow
examination (04/16/2013): Slightly hipocelular bone marrow with 18% of cells
from erythroid lineage (5% of those cells were proerytroblasts with morfological
alterations). Presence of megaloblasts, binucleated red blood cells and altered
cytoplasmic features. Neutrophilic series showing assyncronic maturation, 1% of
myeloblastic components and megakaryocytic serie without morphological alteration.
Iron stains: 30% of ringed sideroblasts. Cytogenetic analysis: 45, XX, deletion of
the long arm of chromosome 5, terminal deletion of the long arm of chromosome 7,
monossomy of chromosome 19 and isochromosome of the long arm of chromosome
21. Bone Marrow examination 07/03/2013: Bone marrow with 50,4% of cells of
erythroid lineage with predominance of early forms and 21,6% of myeloid blasts.
Conclusion: This case is an example of a rapidly progressive acute erytroleukaemia,
that was initially diagnosed as a myelodysplasic syndrome with 5% of proeriytroblasts
and absence of a myeloblasts. The cytogenetic abnormalities found in this patient
are present in myelodysplasic syndromes (MDS) as well as in erytroleukaemia.
Although there are no specific chromosomic abnormalities, the -5/del, -7/del (both
present in this case) and chromosome 8 trissomy are the most common cytogenetic
abnormalities found in erytroleukaemia. It is known that the presence of complex
cytogenetic abnormalities (more than 3 chromosomal abnormalities) is the only
statistically significant independent variable that adversely affects survival in the
acute erythroid leukaemia group. The present report shows a case that began with
a bone marrow count of MDS with 5% proerytroblasts and absence of myeloblasts
and rapidly changed its morfological characteristics to a erytroleukaemia. This case
highlights the necessity of a biomarker that can preciselly differentiate between these
two diseases, so that the appropriate treatment can be initiated without delay.

A-307
Evaluation of the Alcor Scientific iSED Erythrocyte Sedimentation Rate
Analyzer

J. Dohnal, D. Skale. NorthShore University HealthSystem, Evanston, IL

OBJECTIVE: This study evaluates the performance of the Alcor Scientific iSED
analyzer. The analyzer measures erythrocyte sedimentation rate using the principles
of hemorheology in a photometric based method.
METHODS: Split sample comparisons were performed over a 5-week time period
using the manual Westergren Sedimentation Rate method manufactured by LP Italiana
Spa as the predicate device. During the study period two levels of Alcor Scientific
control material were run each day that sample analysis was performed.
RESULTS: CVs for the commercial control material run during the study were
29.1% at a value of 3 mm/hour and 7.1% at a value of 35 mm/hour. Studies using
fresh patient samples yielded similar results and were maintained over time once
the instrument was placed into clinical service. There was no evidence of sample
carryover. Correlation between the predicate method results and the iSED analyzer
results using patient samples was iSED = 0.7006 x (manual method) + 5.113 with an
R2 value of 0.7412. 217 patients were used in the study (61 males and 156 females).
The clinical interpretation of the results for the two methods on all 61 male samples
and 150 of the 156 female samples correlated. No pattern or explanation for the 6
female patients with disparate clinical interpretations could be found. For the patient
samples the concordance coefficient was 0.9423 (substantial agreement) and the
weighted kappa coefficient was 0.9424.

S90

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Hematology/Coagulation

Tuesday, July 29, 9:30 am 5:00 pm

CONCLUSIONS: The analytical performance of the iSED analyzer was acceptable

with CVs at least as good as those listed in the package insert. The instrument is
easy to operate and requires a short training period. Maintenance is minimal and
mechanical performance is acceptable. Although some disparate values were found
between the two methods no systematic biases were noted. Instrument repairs are
affected via a central service center, so instruments needing repairs must be shipped
out to the manufacturer. Loaner instruments are not available at the time of this
abstract submission. The downtime associated with instrument repair is unacceptable
being in the range of 14 - 20 days. Unless a laboratory has multiple instruments in
operation, this represents a significant drawback that needs to be addressed by the
vendor.

A-308
Elevated Serum Levels of Von Willebrand Factor Antigen are Associated
with Poor Prognosis In Patients with Symptomatic Waldenstroms
Macroglobulinemia

I. Papassotiriou1, E. Kastritis2, E. Terpos2, N. Simos1, E. EleutherakisPapaiakovou2, N. Kanellias2, K. Pantelaki1, M. Mazarakis1, M.

Gavriatopoulou2, M. Roussou2, M. A. Dimopoulos2. 1Department of
Clinical Biochemistry, Aghia Sophia Childrens Hospital, Athens,
Greece, 2Department of Clinical Therapeutics, University of Athens,
School of Medicine, Athens, Greece
Background: Waldenstroms Macroglobulinemia (WM) is an uncommon malignancy
which, is characterized by the infiltration of the bone marrow by lymphoplasmacytic
cells, which produce a monoclonal IgM. Recently, it was shown that high levels of
von Willebrand Factor antigen (vWF:Ag), are associated with adverse prognosis
in patients with symptomatic WM and it was suggested that vWF:Ag levels may
reflect interactions between lymphoplasmacytic cells and other cells of their
microenvironment such as mast cells and endothelial cells. We aimed to evaluate the
prognostic importance of vWF:Ag levels of patients with symptomatic previously
untreated WM, in order to validate vWF:Ag as a possible prognostic marker for
progression free (PFS) and overall (OS) survival.
Patients and Methods: The analysis included 42 patients with symptomatic WM,
and 19 healthy controls of matched gender and age. Anemia (250U/L and 58% had
serum albumin <3.0g/dL. Median serum IgM was 3340mg/L (range 246-9563mg/
dL). According to IPSSWM, 22% had low, 43% intermediate and 35% high risk
WM, respectively. Reasons to initiate therapy included cytopenias in 42% of patients,
B-symptoms in 15%, hyperviscosity in 12%, neuropathy in 10% and other reasons
in 21%. Primary therapy based on rituximab was given in 93% of the patients and
54% achieved at least 50% reduction of IgM. vWF:Ag levels were measured in
serum collected before initiation of therapy by means of a latex particle-enhanced
immunoturbidimetric assay (ACL Top 3G, Instrumentation Laboratory, USA).
Results: Median serum level of vWF:Ag was 101U/dL (mean 132.5U/dL, range 19.9399.0U/dL) and were slightly higher compared to the serum levels of healthy controls
(median 85.0U/L, mean 85.0 U/L, range 48.0-124.0U/L). However, 6/42 (14%) had
vWF:Ag levels median value was more frequent in patients with beta2-microglobulin
>3.0 mg/L (p=0.006) and less frequent in patients with low (11%) vs. patients with
intermediate (59%) or high (62%) risk IPSS (p=0.036). There was no correlation of
vWF:Ag levels with IgM levels or with the extent of bone marrow infiltration or
with other manifestations of the disease. Median follow up of symptomatic patients
was 4 years. Patients with vWF:Ag levels within the upper quartile (i.e. vWF:Ag
200.0U/dL) had a median progression free survival of 12 months vs. 63 months of
patients with vWF:Ag <200.0U/L (p<0.001), while the median survival for patients
with vWF:Ag 200.0U/dL was 37 months (4-year survival 29%) vs. 4-year survival
of 97% for patients with vWF:Ag <200.0U/L (p<0.001).
Conclusion: In conclusion, the serum levels of vWF:Ag provide significant prognostic
information in patients with symptomatic WM and patients with levels 200.0IU/dL
have a very poor prognosis compared to patients with lower levels. vWF:Ag measured
in the serum, may become an important prognostic marker in WM and needs further
validation.

A-309
Full Automation of Heparin Induced Thrombocytopenia ELISA Assay on
Dynex DSX ELISA platform.

C. Cronin, F. Lucas, V. Ricchiuti. University of Cincinnati Medical Center,

Cincinnati, OH
Background: Heparin induced thrombocytopenia (HIT), with or without thrombosis,
is an important cause of morbidity and mortality in patients treated with unfractionated
heparin. HIT is invariably characterized by antibodies specific to platelet factor
4 (PF4) and heparin complex. Our goal was to fully automate a HIT ELISA assay
(Immucor, Inc., Norcross GA) with the Dynex DSX instrument (Dynex Technologies,
Chantilly, VA) for application in clinical laboratories to achieve a less labor intensive
approach and high through-put HIT testing.
Method: 37 patient samples were analyzed side by side using the manual HIT ELISA
and the automated HIT ELISA on Dynex DSX. Patient samples were added to micro
wells coated with platelet factor 4 (PF4) and complexed to polyvinyl sulfonate
(PVS) as a PF4:PVS complex. If an antibody was present and recognizes a site on
the PF4:PVS complex, binding would occur. Unbound antibodies were washed away
and an alkaline phosphate labeled anti-human globulin reagent (Anti-IgG, A, M) was
added to the wells and incubated. The unbound anti-IgG/A/M was washed away, and
the substrate PNPP (p-nitrophenyl phosphate) was added. After an incubation period,
the reaction was halted with a stopping solution and the optical density (OD) was
measured by a spectrophotometer at an absorbance of 405 nm using a reference filter
of 490 nm. The only changes in the procedure during the programming of Dynex DSX
software from the manual HIT ELISA protocol were the removal of the pre-soaking
step prior to pipetting reagent and the programming of a longer incubation time which
was extended from 30 to 35 minutes. 20 known negative samples for HIT and 17
known positive samples for HIT were analyzed for the presence of PF4 antibodies
by the fully automated method on the Dynex DSX, and then manually run by hand.
An optical density (OD) greater than 0.4 is used as positive cut-off for PF4 (HIT)
antibodies detection.
Results: The qualitative comparison of results obtained with the manual HIT assay
and the automated HIT assay showed 100% agreement using the 0.4 OD cut-off.
The correlation using the OD from samples was excellent (y(automated)=0.9x(man
ual)-0.02, r=0.96) with a bias of -0.1. The precision (between days) using the Dynex
DSX was above manufacturer expectation (CV<20%) as CV at low level (OD0.3)
was 19.5% (n=8) and at high level (OD1.8) CV was 5.2% (n=8). Precision using
manual procedure was 5.5% (n=6) at OD0.3 and 16.2% (n=6) at OD1.8. The DSX
instrument uses disposable pipette tips and performs all pipetting steps for reagents,
controls, patient samples as well as washing and rinsing steps. This method is less
labor intensive and improves the turnaround time.
Conclusion: The fully automated DSX instrument for HIT testing showed an excellent
correlation with the manual procedure. The HIT assay on the Dynex DSX takes less
than two hours. It is a fully walk-away solution once the instrument is loaded with
reagents and patient specimens. To our knowledge, this is the first fully automated
clinical laboratory HIT assay has been described using the Dynex DSX ELISA
instrument.

A-310
New policy reduces manual slide reviews on platelet tests

B. R. Morgan1, D. L. Clark2. 1Sacred Heart-St. Marys Hospitals; Ministry

Medical Group, Rhinelander, WI, 2Sacred Heart-St. Marys Hospitals,
Rhinelander, WI
Implementation of a new platelet manual slide review policy in a small hospital/
clinic laboratory serving an oncology service can produce a modest reduction in slide
reviews related to platelet count flags, improve turnaround time, and save an estimated
$3/slide review (6 min/case). The policy at this institution for performing a Wright
stained slide review on specimens tested using a Sysmex XE-2100 Hematology
Instrument was <100,000/uL and/or a Sysmex platelet flag (platelet clumps CLP,
platelet abnormal distribution PAD, platelet abnormal scattergram PAS). Absolute
delta check values set at 20,000/uL (<100,000/uL) and 50,000/uL (101,000 - 106/
uL) reflexed a review of clinical history, but not slide review. The goal of the policy
revision was to decrease slide reviews related to platelets on blood samples from
patients with known histories of low platelet counts. Real time review of LIS/EHR
patient information was a necessary component for the new platelet slide review
policy and designed as follows:
No flags- report result

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S91

Hematology/Coagulation

Tuesday, July 29, 9:30 am 5:00 pm

Slide review performed first time for thrombocytopenia (<100,000/uL), CLP, PAS or
PAD; always for CLP flag
If TCP <100,000/uL and/or PAD, PAS flags present on hemogram:
-Previous slide review performed, same platelet flag, and patient result review
consistent with clinical history, finalize result
-Previous slide review performed and patient result review inconsistent with clinical
history, perform slide review, then finalize result
-Previous slide check performed and different platelet flag, perform slide review, then
finalize result
Three months platelet count data collected after implementation of the policy yielded
22 patient platelet test results (60% oncology service patients), 0.4% of 5564 platelet
count tests showing platelet flags. Nineteen patients had previously documented
platelet counts <100,000/uL within a 4 year time span archived by the LIS. Eightysix subsequent slide checks were avoided by implementation of the policy. On EHR
review of patient records, no significant clinical findings appeared to have been
missed by the new policy. Median slide check time, defined as start to finalizing result,
was 6 min/case; $3/platelet slide check (savings based on $30 labor and benefits).
If this sample reflects our patient population and provider ordering patterns, annual
savings are modest at approximately $1000/yr. More important are the improved
turnaround time of at least 6 min for these specimens, significant given that average
turnaround time for platelet count at this laboratory without flags (receipt to verify)
is 10min, and decreased technologist disruption by eliminating unnecessary platelet
slide checks. Implementation of a new platelet manual slide review policy modestly
reduced the number of manual slide reviews performed on patients showing repetitive
blood counts with thrombocytopenia <100,000/uL, abnormal platelet distribution or
scattergram flags.

A-313
Can Neutrophil/Lymphocyte Ratio be Used in the Differential Diagnosis of
Abdominal Pain of Appendicitis and Familial Mediterranean Fever?

O. Baykan, A. Yaman, G. Gokulu, D. Altinok, O. Sirikci, G. Haklar.

Marmara University School of Medicine, Istanbul, Turkey
Backround: The neutrophil/lymphocyte ratio (NLR) is proven to be associated
with some conditions such as chronic inflammation in cardiovascular diseases,
malignancies, ulcerative colitis, gangrenous appendicitis, and amyloidosis. Familial
Mediterranean Fever (FMF) is an inherited disorder which is common among
individuals of Mediterranean descent. Both FMF and appendicitis may manifest
with abdominal pain which is hard to differentiate. This retrospective study aimed
to evaluate the ability of the NLR to predict acute appendicitis pre-operatively and to
differentiate cases with abdominal pain due to FMF and acute appendicitis.
Methods: A hundred patients who were admitted to emergency unit of Marmara
University Pendik E&R Hospital with abdominal pain were included. Fiftysix patients
had the diagnosis of appendicitis, were treated operatively and proven as appendicitis
histopathologically. FMF group was formed by 44 patients who were monitored at
the Rheumatology Clinic with colchicine treatment and without any other diseases.
Results: Median ages (25-75 percentile) of appendicitis and FMF groups were 12
yrs (10-15,75) and 13 yrs (9-17), respectively which were not different significantly
(p=0.691). Complete blood counts were measured by LH 780 (Beckman Coulter,
USA) and NLR was obtained from the records. Mean NLR values (25-75%) were
significantly higher for appendicitis group when compared to the FMF group [2.54
(1,5-4,8) vs. 9.34 (5,32-15,45)] (p<0.001) (Fig 1). The diagnostic performance of
NLR to differentiate acute appendicitis from FMF was assessed with a ROC plot
and cut-off value of NLR was >4.97. (AUC:0.842, 95%CI: 0.763-0.920, p<0.001)
(sensitivity 80% and specificity 77%).
Conclusion: According to the results of our study, NLR of 4.97 seems to be a reliable
parameter in discriminating abdominal pain of acute appendicitis from FMF attack.
NLR is cost effective, readily available, and can be calculated easily.

S92

A-315
Reticulated Platelets - Towards A Standardized Approach: Results From
Apparently Disease-free Subjects In 3 Countries

N. LLEWELLYN-SMITH1, B. HEDLEY2, S. LANG3, D. ROSENFELD3,

N. McNAMARA3, M. KEENEY2. 1Abbott Diagnostics, Santa Clara,
CA, 2London Health Sciences Centre, London, ON, Canada, 3Liverpool
Hospital, Liverpool, Australia
Background Despite obvious potential as a sensitive marker of thrombopoietic
activity, the lack of a reference method and standard materials for Reticulated
Platelet (RP) counting have hampered widespread adoption. Here we applied a new,
highly standardized method (described in a separate poster) to assess typical values
in healthy (apparently disease-free subjects) at 3 international sites. Our goal was
to test whether 3 different laboratories, using the same protocol on 3 different flow
cytometers, could recover statistically similar results in apparently disease-free
(normal) subjects.
Methods The 3 sites (instruments) were 1) Abbott Hematology, California
(Accuri-C6), 2) Liverpool Hospital, Australia (FACSCanto) and 3) London Health
Sciences, Canada (Gallios). K2EDTA specimens were collected from 71 consented
voluntary participants in the Abbott staff donor program, lab staff or surplus CBC
specimens from subjects without any hematological abnormality. Briefly: fresh (<
8 hours post draw) whole blood was incubated with the CD61-APC/CD41-APC
monoclonal antibody mix and stained with TO before fixation with formaldehyde.
A second tube, with PBS in place of TO, served as the control. During analysis, a
positive marker was drawn on a TO histogram (gated on the platelet cluster) of the
control tube to include 0.1% of the negatives. The instrument settings and this marker
were held fixed for the acquisition and analysis of the tube containing the TO-stained
platelets. Basic statistics (means, medians, SD and CV) were generated in MS Excel
2010, Analysis of Variance (ANOVA) was used to assess the mean results (males,
females, pooled) across the 3 sites.
Results The mean RP results from the 3 sites were as follows: California 5.2% (SD
3.3%, n=21), Canada 4.8% (SD 0.7%, n=21) and Australia 4.6% (SD 3.0%, n=29).
Mean for the pooled 71 subjects was 4.9% (SD 2.7%, Range 1.7-17.1%). Differences
in the means between the 3 sites were not statistically significant (p > 0.05). At one site
(California) the mean female RP count was higher (6.1%) than the mean male value
(4.1%) though the differences were not statistically significant.
Conclusion We have demonstrated that a standardized flow cytometric protocol, at 3
geographically distinct sites, on 3 different instruments, yielded comparable results
in apparently disease free subjects. The major shortcoming of all nucleic acid based
methods is the non-specific uptake of the nucleic acid dye by platelet dense and alpha
granules. We were able to minimize this by using a dilute TO solution. To further
mitigate non-specific dye uptake, we used a simple and objective control strategy
(negative control tube) to differentiate specific from non-specific staining. We believe
this method is simple and robust enough to form the basis for development of a
potential reference method. Further studies to define the ideal dye concentration and
characterize inter-lab performance are planned.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Hematology/Coagulation

Tuesday, July 29, 9:30 am 5:00 pm

A-316
Proteomic profiling of platelets in acute ischemic stroke patients

O. Cevik1, T. Baykal2, G. Somay3, A. Sener4. 1Cumhuriyet University

Faculty of Pharmacy, Sivas, Turkey, 2Medipol Univesity School of Medicine,
Istanbul, Turkey, 3Hisar Intercontinental Hospital, Neurology, Istanbul,
Turkey, 4Marmara University Faculty of Pharmacy, Istanbul, Turkey
Background: Platelets are important in the pathogenesis of stroke and ischemic stroke
is high level of mortality. Especially antiplatelet agents are useful for prevention and
treatment in this patients. Platelets are so easily activated by different stimulants in the
circulation which can be affected by activation and apoptosis.
Methods: Using LC-MS based protein profiling, we examined and correlated the
proteomic response of stroke patients platelets. Patients were admitted to neurology
department who had ischemic stroke within 24 h. Stroke is defined as rapidly
developing clinical symptoms/signs of cerebral dysfunction lasting more than 24 h
without any cause other than a vascular abnormality. Venous blood was drawn from
all patients who had not taken any antiplatelet drugs for the prior 14 days, and was
collected into acid citrate dextrose containing tubes for platelet isolation. Platelets
tryptic peptides were analyzed in triplicate on the LC-MS/MS system (UPLC-ESIqTOF-MS). Tandem mass data extraction were done with ProteinLynx Global Server
v2.5 (Waters) and searched with the IDENTITYE algorithm against the reviewed
database of homo sapiens (www.uniprot.org). Protein Identification detected with
PLGS 2.3 and Quantification of the protein expression changes was done with
Progenesis LC-MS software V4.0 (Nonlinear Dynamics). In addition, ELISA was
used to quantify pro-inflammatory cytokines as TNF- in the plasma.
Results: Proteomic profiling of platelets obtained from the stroke patients resulted
in identification of 500 proteins. Of these, 83 proteins were found to be differentially
expressed in patient as compared to control. These differentially expressed proteins
were involved in various processes such inflammatory response, cellular movement,
immune cell trafficking, cell-to-cell signaling and interaction, hematological system
development and function and nucleic acid metabolism. Plasma levels of TNF-
increased in the stroke patients (29.12 1.37 pg/ml) compared with control subjects
(5.16 2.84 pg/ml).
Conclusion: This is the first report providing a global proteomic profile of platelets
from stroke patients. Our data provide an insight into the proteins that are involved
as platelets inflammation response during ischemic stroke. Inflammation caused by
stroke changed to platelet cellular proteins interactions in patients.

A-317
Investigating lipid effects on protein C and thrombin-activatable fibrinolysis
inhibitor activation by thrombin/thrombomodulin complex

V. Gruzdys, X. Sun. Cleveland State University, Cleveland, OH

Background: Thrombomodulin (TM), an endothelial membrane protein, plays
central roles in maintaining haemostasis and preventing inflammation by increasing
protein C (PC) and thrombin activatable fibrinolysis inhibitor (TAFI) activation
by thrombin. APC selectively inactivates coagulation factors Va and VIIIa, thus
preventing excessive coagulation. TAFI modulates inflammatory mediators
and reduces plasmin generation. During cell apoptosis, membrane lipids, such as
phosphatidylserine and ethanolamine, are enzymatically flipped to the cell surface.
Endothelial cell damage often involves lipids changes. The lipid membrane facilitates
TM-enhancement of protein C activation by thrombin, however, lipid effect on TAFI
activation is unknown. We studied the effects of different phospholipids in generating
APC and TAFI by the TM/thrombin complex in an effort to better understand
thrombotic and inflammatory processes for possible implications in diagnosis and
treatment monitoring.
Methods: To study the capacity of TM/thrombin to generate APC and TAFIa,
liposomes containing TM were generated by extrusion after TM was mixed with a
swelled lipid solution of different compositions in Tris-HCl buffer. Liposomal TM
was separated from free form by size exclusion chromatography using Sepharose
CL-4B. Liposomal TM concentration was verified by first separating lipid and
protein components followed by protein content determination by Bradford assay.
APC and TAFI were generated by incubation with liposomal TM and thrombin in
otherwise identical conditions. Concentrations of APC and TAFI were determined
by hydrolysis of Spectrozyme PCa and hippuryl-arginine, respectively.
Results: It was shown that incorporation of TM into phoshatidylcholine vesicles
increased generation of APC (26.4 pmol 0.7 pmol) in comparison to free TM (18.2
pmol 0.7 pmol). Phosphatidylethanolamine had decreased APC generation (10.6
pmol 0.6 pmol), phosphatidylserine had a large increasing effect (35.9 pmol

1.3 pmol) while APC generation by thrombin alone and with phosphatidylcholine
vesicles (but no TM) was below detection range. In addition, free TM was added to
phosphatidylcholine liposome solution and APC was generated (20 pmol). This data
suggested that separate lipid components do not increase APC level as much as the
complex between lipids and TM does.Increase in TAFI generation was observed after
TM was reconstituted in phosphatidylcholine (8.34x104 U, where 1 U = hydrolysis of
1 mol of substrate per min.) and phosphatidylserine containing (9.18x104 0.3x104
U) vesicles when compared to free TM (6.46x104 U). Phosphatidylethanolamine
reconstitution resulted in slight elevation of TAFI, albeit high variation was observed
between measurements (7.54x104 1.6x104 U).
Conclusion: We Investigated lipid effects on protein C and thrombin-activatable
fibrinolysis inhibitor activation by thrombin/thrombomodulin complex. We found
that phosphatidylserine-bound TM dramatically increases the generation of APC and
TAFI in liposome-based systems. This is contrary to phosphoethanolamine, which
had a reducing effect on APC generation. This study suggests possible significance of
the effects of cell membrane lipids on hemostatic balance and inflammation and could
have possible implications related to damaged endothelium situations such as septic
shock. Future studies will test this observation with human endothelial cell lines.

A-318
Establishment of reference intervals for HbF and HbA2 in a Northern Alberta
population

T. Higgins, K. Rodriguez Capote. DynaLIFEDx,, Edmonton, AB, Canada

Background DynaLIFEDx is the sole laboratory performing hemoglobinopathy
and thalassemia investigations for a catchment area of 2 million people in
northern Alberta. Samples for hemoglobinopathy and thalassemia testing come
from physician request, immigration medicals and as a reflex to the presence
of a hemoglobin variant noted on HPLC analysis for HbA1c.A review of our
data showed that 17% of HbF results exceeded the manufacturers reference
interval of <1.0% so it was necessary to reevaluate the reference interval
for Heft. A reference interval for HbA2 was calculated at the same time.
Method 6616 thalassemia and hemoglobinopathy investigation requests were
analyzed in the period January 1, 2013 to December 31, 2013. 3306 were on patients
older than 2 years and interpreted as normal by a Clinical Chemist based on
their hematology indices (hemoglobin >120g/l, mean cell volume > 80fL), absence
of a hemoglobin variant, replete iron status and calculated Mentzer Index and
Discriminant Factor EDTA-anti-coagulated blood samples were analyzed by high
performance liquid chromatography (HPLC) using the thalassemia program on the
Bio-Rad VARIANT II. Total imprecision (CV%) for HbF was 2.2% and 1.2 % at
levels of 2.17% and 9.54% respectively. For HbA2 the total imprecision (CV %) was
3.7% and 2.1% at levels of 2.82% and 5.74% respectively. The statistical software
package MedCalc version 11.4.2.0 for Windows was used to analyze the data.
Results 3306 individuals, median 31 years, range 3 to 92 years were included in
the reference interval calculation. As recommended by CLSI a non-parametric
method was used to calculate the reference interval as the KolmogorovSmirnov test showed that the distribution of HbF and HbA2 values did not
follow a normal distribution. Reference Intervals are HbA2 Lower limit (90%
CI) 2.3% (2.20 to 2.34), Upper limit(90% CI) 3.4% (3.3 to 3.45); HbF Lower
limit (90% CI) 0.2%( 0.20 to 0.23) and Upper limit(90% CI) 1.8% (1.80 to 1.90).
Conclusion The appropriate reference intervals for the Northern Alberta population
using the thalassemia program on the Bio-Rad VARIANT II was up to 1.8% for HbF
and up to 3.4% for HbA2.

A-319
Protein C Antigen And Activity In Patients With Sickle Cell Anaemia

S. E. Idogun1, O. E. Ihenachor2. 1university of Benin Teach Hospit, BENIN,

Nigeria, 2University of Benin Teaching Hospital, BENIN, Nigeria
BACKGROUND: Sickle cell disease (SCD) is characterized by vaso-occlusive
events and organ damage which form important causes of morbidity and mortality.
Protein C is a naturally occurring anticoagulant that has been demonstrated to be
deficient in sickle cell anaemia patients and the impact of hypercoagulable states and
thrombosis on sickle cell disease still remains uncertain.
AIM AND OBJECTIVES: The objective of this study was to determine the protein
C (PC) levels of the sickle cell patients and to explore the relationship between protein
C levels and vaso-occlusive events as well as related complications.
SUBJECTS AND METHODS: A cross-sectional study comprising sixty one sickle
cell subjects and thirty healthy control subjects. Protein C antigen (quantitative) and

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S93

Hematology/Coagulation

Tuesday, July 29, 9:30 am 5:00 pm

activity (qualitative) levels were assayed using enzyme linked immunosorbent assay
and the Protac (photometric) method respectively. Haematological parameters were
measured with the haematology automated analyzer. Data were analyzed using SPSS
version 16.
RESULTS: The adult SCD subjects had lower values for both PC antigen (68.6
16.0 %; p=0.004) and activity (49.0 13.0 %; p=0.005) when compared to the adult
control group (PC antigen = 84.8 18.0 %; PC activity = 60.3110.37 %). Similarly
the paediatric SCD subjects had lower PC antigen (54.9 14.9 %; p=<0.001) and
activity (48.0 13.1 %; p=0.006) levels compared to their control group (PC antigen
= 79.6 16.7 %; PC activity = 58.6 8.0 %). However there was no significant
association between PC levels and SCD complications or vaso-occlusive events (p
>0.05).

Conclusion: There is need for a strengthened antenatal care system with increased
awareness of the problem among communities most affected by malaria. Preventative
strategies including regular chemoprophylaxis, intermittent preventative treatment
with antimalarials and provision of insecticide-treated bed nets should be implemented
as well as integration of malaria control tools with other health programmes targeted
to pregnant women and newborns.

A-321
Risk Stratification and Progression follow-up of MGUS Patients: value of the
sFLC and Heavy Chain/Light Chain Pairs markers

CONCLUSION: There is functional PC deficiency in SCD patients. This supports

a hypercoagulable state in the patients. However there was no significant association
between PC levels and SCD

J. Jimenez1, L. Campos2, T. Pais2, N. Barbosa2, C. H. Larramendi1.

1
University Hospital Severo Ochoa, Madrid, Spain, 2The Binding Site,
Barcelona, Spain

complications or vaso-occlusive events. Thus protein C level in SCD patients may not
be a good prognostic marker for disease severity.

Background: MGUS is the most frequent MG, usually considered a benign MG,
and defined by a serum monoclonal protein (MP) <3 g/dL, less than 10% of plasma
cells in the bone marrow, no related organ or tissue impairment and no evidences of
other B-cell proliferative disorder. The rate of progression towards Multiple Myeloma
(MM) is of 1-2% however, the designation itself highlights the current difficulty in
precisely identifing patients that will progress towards malignant MG. Based on the
MP size and type and the serum Free Light Chains (sFLC) ratio at diagnosis it is
possible to stratify the patients according to probability of progression allowing a
better management of the MGUS patients, according to the IMWG guidelines. We
have been studying the frequency of specific immunoglobulin heavy/light chain pairs
alterations in a cohort consisting of MGUS patients with different risk of progression.
To validate the significance of our preliminary findings we have increased the cohort
in all risk groups and included the follow-up of patients that have progressed to MM.

PROTEIN C SERUM LEVELS

Variables

Adult
SCD
Controls
(n = 30) (n = 15)

Protein C Antigen 68.6

Mean SD (%)
16.0
Protein C Activity 49.0
Mean SD (%)
13.0
*t-test

P-value

Paediatric
SCD
Controls
(n = 31) (n = 15)

P-value

84.8 18.0 0.004

54.9
14.9

79.6 16.7 <0.001

60.3 10.0 0.005

48.0
13.1

58.6 8.0

0.006

A-320
Thrombocytopenia in Plasmodium falciparum Parasitized Pregnant Women

F. A. Botchway1, W. Ababio1, P. Bannerman- Williams2, H. Salifu3, C.

E. Lekpor4. 1Department of Child Health, Korle Bu Teaching Hospital,
Accra, Ghana, 2Department of Maternity, Korle Bu Teaching Hospital,
Accra, Ghana, 3Morehouse School of Medicine, Atlanta, GA, 4Department
of Medical Laboratory, Kwame Nkrumah University Of Science and
Technology, Kumasi, Ghana
Background: Malaria infection during pregnancy is a major public health problem in
tropical and subtropical regions of the world. Hematological changes associated with
malaria in pregnancy are not well documented, and have focused predominantly on
anemia.
The aim of this study was to determine the impact of Plasmodium falciparum
parasitaemia on the platelet count of pregnant women at the maternity department of
Korle Bu Teaching Hospital, Ghana
Methods: This case control study evaluated the effect of malaria parasitemia on
the platelet count of Sixty (60) plasmodium parasitized pregnant subjects. Sixty
non- malaria parasitized pregnant women and sixty non-pregnant and non-malariainfected subjects served as control. 1.0 ml of blood sample was taken into EDTA
bottle for Full Blood Count using Mindray BC 5300. Thick and thin film prepared
and stained with Giemsa for the determination of P. falciparum parasite species and
density using light microscopy. A platelet count of 100 109 /L was the threshold at
two standard deviations below the mean for healthy Ghanaian pregnant women used
to indicate thrombocytopenia. Differences in platelet counts were compared based
on malaria species and parasitemia in matched non-pregnant and pregnant women.
Results: The mean platelet counts (109 /L) were significantly lower in pregnant
subjects with an episode of Plasmodium falciparum malaria 101.3 9.2 109 /L
compared to non-parasitized and healthy non-pregnant controls (245.09 23.10 and
260 50.5 109 /L) respectively. Platelet count values were 102.5 9.58 109 /L and
116.3 15.7 109 /L for the primigravidae and multigravidae respectively. (= 10.36;
P = 0.05). Parasite density was significantly higher among Plasmodium falciparum
parasitized primigravidae compared to multigravidae 2140 (1628-2652) parasites/L
in primigravidae women compared to 1816 (1420-2212) parasites/L in multigravid
women. The mean parasite count in Plasmodium falciparum parasitized subjects was
2610 224 parasites/L, 95% confidence interval (2082-3108). Malaria parasite
was found to exert a significant reduction in platelet density in parasitized subjects.
This reduction was more pronounced in primigravidae and multigravidae. An inverse
relationship was established between parasite density and platelet count (y =-0.020
+86.2, r =-0.3).

S94

Methods: 308 samples from 248 MGUS patients were included, both newly and
previously diagnosed. All patients were risk stratified according to the IMWG
guidelines serum M-protein levels and type by SPE and IF, and sFLC and HLC by
nephelometry (FreeliteTM and Hevylite, respectively). Total immunoglobulin levels
were also determined to establish the frequency of classic immunoparesis (BNII.
Siemens).
Results: The Hevylite assay allows the determination of an imbalance on the
immunoglobulins of the same isotype (i.e., IgG-K/IgG-L ratio) identifying the
presence of a MP in serum. It also allows to observe the immunoparesis within the
same isotype of immunoglobulin (i.e., suppression of the uninvolved HLC pair IgG-L
in a IgG-K monoclonal gammopathy). The frequency of HLC immunoparesis is
significantly superior to the classic immunoparesis for low-intermediate (p<.0005) and
intermediate-high risk groups(p<.001). Besides, the frequency increases in the higher
risk-of-progression groups, although only significantly for the HLC immunoparesis
type(p<.01 HLC vs p<.27). In IgM cases the differences between classic and HLC
immunoparesis did not reach significance, possibly due to the size of the population.
Furthermore, for IgG MGUS patients, both the HLC ratios and the uninvolved HLC
immunoglobulin levels show a significant trend towards more extreme values as
the risk of progression increases. 3 IgG MGUS patients with low-intermediate risk
progressed towards MM. For these particular cases an abnormal sFLC ratio was the
only established criteria that indicated the risk for progression: 1) While normal at
presentation, the HLC ratio became abnormal during follow-up, initially due to the
suppression of the uninvolved HLC pair; 2) both FLC and HLC ratios were abnormal
prior conversion and those abnormalities became more extreme with active MM; 3)
HLC normal at presentation, the patient evolved into an oligosecretory disease.
Conclusion: The sFLC ratio is a relevant indicator of risk for progression in this
population of MGUS patients. The frequency and distribution of HLC alterations
(both ratio and HLC-immunoparesis) within the specific MGUS risk-groups is
suggestive of its utility as a marker for progression.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

Tuesday, July 29, 2014

Poster Session: 9:30 AM - 5:00 PM
Immunology

A-323
Evaluation of pre eclampsia markers in pregnant women with chronic
hypertension and pregnant women without hipertension

C. E. Ferreira1, P. R. S. Ferreira1, L. G. Oliveira2, R. C. Sanches1, A. C.

L. Faulhaber1. 1Hospital Albert Einstein, So Paulo, Brazil, 2Hospital Vila
Nova Cachoeirinha, So Paulo, Brazil
Background: The relationship between two biomarkers, soluble tyrosine kinase
(sflt1) and placental growth factor (PIGF) has been described as useful for precocious
identification of pregnant women at risk for pre eclampsia
Methods: 20 hypertensive pregnant women were included among the twenty- and
thirty-sixth weeks of pregnancy and a control group of non hypertensive pregnant
women was the comparison group. All serum samples were frozen at -80C , and
processed simultaneously. The markers were processed Cobas 6000 analyzer from
Roche Diagnostics.The cut- off ratio for sFlt-1/PLGF exclusion of preeclampsia
described in the package insert of the kit is 33 , with 95% sensitivity and 94%
specificity . Values greater than 85 are suggestive of disease.
Results: In the group of pregnant women without hypertension, the average ratio
was 22 and in the hypertensive group was 284 . There was a significant difference
between the groups evaluating the paired t-test ( p < 0.0001 ) . Group of hypertensive
pregnant women showed no relation to the lower cut- off suggested for exclusion of
pre -eclampsia . In the group of healthy patients only 2 patients showed higher values
compared to the deleting 33 does not pre- eclampsia are possible. In the group of
chronic hypertensive pregnant women 70 % ( 13/20 ) had higher values at 85 .
Conclusion: In our sample, the use of markers of preeclampsia was found to be of
the most value to clinical practice in the evaluation of patients with potential risk for
progression to pre - eclampsia . Clinical trials are necessary to follow up in this group
of patients

A-324
Seasonal frequence of the most requested specific IGE

N. Z. Maluf, S. A. D. Mendona, L. B. Faro, F. R. M. Abreu, M. D. C.

Freire. DASA, Barueri, Brazil
Background: Allergic asthma, rhinitis, and atopic eczema are among the commonest
causes of chronic ill health in the world. Asthma is one of the most common chronic
conditions affecting both children and adults, yet much remains to be learned of its
aetiology. Although genetic predisposition is clearly evident, gene-by-environment
interaction probably explains much of the international variation in prevalence rates
for allergy and asthma. In our laboratory the ten most requested specific IgE are:
IgE specific for dust and mites (HX2), IgE specific to fungi (MX1), IgE specific for
epithelial animals (EX1), IgE specific for baby food (FX5), IgE specific for seafood
(FX2), IgE specific for D. farinae (D2), IgE specific to egg white (F1), Milk specific
IgE (F2), IgE specific to soybean (F14), IgE specific to D. pteronyssinus (D1) and
IgE specific for cocoa (F93). Analyse the seasonal frequence of the most requested
specific IgE.

A-325
A Rapid and Effective Tool for Monitoring Monoclonal Antibody Production

B. Bhullar, C. Hui, M. Hariharan, J. Hewitt, N. Vats. MedMira Laboratories,

Halifax, NS, Canada
Background: Monoclonal antibodies (Mab) are used in a variety of fields from
diagnostics to therapeutics. We examined the effectiveness and utility of tools created
with MedMira Miriad RVF Toolkit to assess Mab post-production functionality.
Methods: Six hybridoma supernatants, developed against unique peptide sequences
derived from HIV and HCV proteomes, were obtained from a supplier. Miriad RVF
Toolkit was used as per manufacturer instructions, with 0.5L (1mg/ml) of each
peptide or mouse IgG (control) spotted on to the cartridges using a micropipettor.
Cartridges contained multiple test spots, one spot per peptide and one control spot
(see Fig 1 for illustrative example for HIV peptides). Two sets of cartridges, one each
for HCV and HIV peptides, were prepared and allowed to dry at room temperature
for 30 minutes. Samples of neat or PBS diluted hybridoma supernatants were added
during procedural steps.
Results:Reactive results, shown in Fig 1, were scored on a one to three grading scale,
three being the highest. Increasing dilutions resulted in decreased reactivity. Two
antibodies HCV MDL-1 and 3 became non-reactive at a 1:100 dilution, all others
yielded reactive results at the 1:100 dilution. For example, anti-gp 120 antibodies
were reactive down to a 1:100 dilution as illustrated in Fig 1. The specificity of the
antibodies was evidenced by lack of cross reactivity of each Mab applied to cartridges
containing multiple spots.
Conclusion: Testing with the Miriad RVF Toolkit was completed in less than one
hour following receipt of antibodies. Miriad RVF Toolkit therefore represents a tool
that can be used to efficiently assess production levels, functionality, and specificity of
Mab. The time to results also allows Miriad RVF Toolkit to be used as an in-process
monitoring tool during production. Results can be documented by recording of visual
interpretation or by photographs as illustrated in Fig 1.

Methods: We analyse the frequence of the IgE specific for dust and mites, IgE specific
to fungi, IgE specific for epithelial animals, IgE specific for baby food, IgE specific
for seafood, IgE specific for D. farina, IgE specific to egg white, Milk specific IgE,
IgE specific to soybean, IgE specific to D. pteronyssinus and IgE specific for cocoa
throughout the year 2012. The informatics system has given the data and for the
statistical analysis we used the dispersion
Results: The results of this data are on the graph.
Conclusion: We conclude that there is no seasonal fluctuation in the incidence of IgE
specific studied.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S95

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

A-327
What is the Optimal Threshold for an ANA ELISA?

J. Parikh1, U. Smith2, L. Ingram2, E. S. Pearlman2. 1University of Tennessee

Health Sciences Center, Memphis, TN, 2Veterans Affairs Medical Center,
Memphis, TN
Context: ELISA ANA screening [Seradyne; Indianapolis, IN] was brought inhouse at the VAMC in 2013 using an automated Triturus processor [Grifols;
Los Angeles, CA]. Rheumatologists had previously indicated dissatisfaction
with excess false positive screens and initial validation studies had suggested
improved specificity using a threshold optical index [OI] of 2.5 in place of the
manufacturers suggested value of 1.0. We evaluated this change using clinical data.
Methodology: Reference lab [RL] confirmation (Lab Corp; Raleigh, NC) was obtained
in 34(group 1) and 23 (group 2) consecutive specimens (33 and 21 patients) with OI
values 2.5 and between [1.0, 2.5) respectively with adequate specimen volume. RL
confirmatory testing entailed either a panel of specific antigens (if ordered) or a lab
ordered IFA. Positive results on one of more panel antigens or an IFA titer 1:80 was
considered confirmatory. The electronic health record was reviewed for clinical data.
Results: Median age was 64 [27-91] YO with11 (20%) females [twice the female
proportion in our general population]. 26 (76.5%) and 9 (39%) specimens confirmed
in groups 1 and 2 respectively [chi sq. = 8.16, p<0.005]. Review of the 7 patients (9
specimens) with false negative screens using the higher threshold revealed 2 with
hepatitis C and 2 who were 80 YO. The appearance of aberrant autoantibodies in
these instances is well known. The fifth patient was status post renal transplant for
IgA nephropathy. This patient had two specimens sent 4 months apart with OIs of
2.91 and 1.46 and IFA titers of 1:640 (homogeneous pattern)] and 1:320 (speckled)
respectively. A multi-antigen panel done with the second specimen was negative. The
sixth patient had been diagnosed with Rheumatoid arthritis with high titer of anti-CCP
antibodies. The last patient had been diagnosed with SLE elsewhere in the 1990s based
on renal biopsy. RL ELISA screens in 2008 and 2010 were negative. It did not appear
that results of ANA screening affected clinical management in any of these patients.
Conclusions: Our study, although small, suggested that improvement in specificity
with a decreased number of workups for false positive ANA screens was possible
through adjustment of the OI threshold without negative clinical consequences.

A-328
Evaluation of an Anti-Streptolysin O assay for use on the Binding Site Next
Generation Protein Analyser

S. Ayub, A. Kaur, M. Solanki, S. J. Harding, P. J. Showell. The Binding Site

Ltd, Birmingham, United Kingdom

A-326
Usefulness of highly sensitive on-chip immunoassay for fucosylated fraction of
alpha-fetoprotein in patients with hepatocellular carcinima

K. Young Jae. Samsung Changwon Hospital, Changwon, Korea, Republic

of
Background:Alpha-fetoprotein(AFP) has been used as a diagnostic marker for
hepatocellular carcinoma(HCC), and fucosylated fraction of AFP(AFP-L3) has been
proposed as a marker for HCC.
Methods: We evaluated performance of the micro-total analyzer system(-TAS), onchip immunoassay analyzer for AFP-L3. The linearity, precision and carry-over rate
of -TAS were evaluated, and we compared the AFP-L3% levels between patients
with early HCC and control group with benign liver diseases.
Results:The linearity was good(R2=0.9995) and coefficient of variation(CVs)
of between-day precision in high and low concentration were 0.2% and 0.18%,
respectively. AFP-L3% levels were higher in patients with early HCC than in
control(13.4%16.9% versus 4.6%3.4%). The sensitivity and specificity with
AFP-L3% were 57% and 67% at a cut-off value of 5%, and 43% and 83% at a cut-off
value of 7%, respectively.
Conclusion: -TAS showed good performance of linearity, precision and carry over
rate, and AFP-L3% could be a suitable serologic marker for evaluating early HCC.

S96

Elevated blood serum concentrations of human Anti-Streptolysin O (ASO) can be used

to provide serologic evidence of past or present infection by -Haemolytic Streptococci
bacteria. Increasing serum concentration of ASO antibodies are produced in response
to Streptolysin-O exotoxins secreted by the bacteria. Measurement of ASO levels in
serum can be used as an aid in the diagnosis of diseases such as glomerulonephritis,
rheumatic fever, bacterial endocarditis, tonsillitis, and scarlet fever. Here we evaluate
an ASO assay designed for the quantitative in vitro measurement of human ASO in
serum using the Binding Sites next generation protein analyser. The instrument is a
continuous throughput, bench top turbidimetric analyser capable of automatic sample
dilutions up to 1/10,000 and having a throughput of up to 120 tests per hour. Analyser
precision is promoted by single-use cuvettes, whilst the user interface is enhanced
through bi-directional communication capability, primary sample ID and fully bar
coded reagent management systems. Evaluation of the assay on the next generation
protein analyser demonstrated an overall assay time of 12 minutes which was read
at end point. The assay auto dilutes a single serum based calibrator to produce a
measuring range between 50 - 800.00 IU/mL at the standard (1/10) sample dilution.
Samples outside the standard range auto re-dilute to neat (1/1) or a secondary dilution
(1/20) as appropriate. Results which are still outside the measuring range following
auto dilution are reported as <5.00 IU/mL, or >1600 IU/mL. Total precision studies
performed at 5 different levels across the measuring range were assessed in duplicate
over 21 working days, using a single kit lot on three analysers. Levels assessed were
at 731 IU/mL (SD = 33.696, %CV = 4.6%), 427 IU/mL (SD = 16.403, %CV = 3.7%),
82 IU/mL (SD = 4.575, %CV =5.8%), 236 IU/mL (SD = 9.285, %CV =3.9%), and
151 IU/mL (SD = 6.586, %CV = 4.5%). The assay gave a linear response over the
measuring range of 50 - 800.00 IU/mL at the standard 1/10 sample dilution and over
a range of 5.00 IU/mL- 80.00 IU/mL at the minimum 1/1 dilution. A linear regression
of y = 0.981x - 16.2 and R2 = 0.997 was demonstrated at 1/10, whereas a regression
of y= 1.032x - 0.411 and R2 = 0.998 was demonstrated at 1/1 dilution. No significant
interference was seen when the assay was challenged with haemoglobin (500mg/

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

dL), bilirubin (200 mg/L) and chyle (1500 FTU). Comparison comprising normal
and clinical samples (n =121) was carried out against the Binding Site SPA PLUS
analyser, covering between 52.000 IU/mL - 822.000 IU/mL. Analysis by PassingBablok regression demonstrated a linear fit of y = 0.96x + 0.92. We conclude that the
ASO assay for the Binding Site Next generation protein analyser is reliable accurate
and precise and shows good agreement with existing assays.

Conclusion: We conclude that the IgG avidity test for toxoplasmosis may be an
important tool for the interpretation of IgG/IgM results in pregnant women.
Avidity
Number of samples
>60%
398
30-60%
11
<30%
4
Total
413
Table 1- IgG avidity detected in the studied samples.

%
96,4
2,7
1,0
100

A-329
A-331

Cytokine and IFN--Induced Chemokine mRNA real-time PCR Taqman Probe

Assay After Mycobacterium tuberculosis Specific Antigen Stimulation in Whole
Blood from Infected Individuals

Neopterin ELISA kit: Analytical evaluation and verification of the reference

interval in native population in Argentina.

H. Kim1, S. Kim2, J. Cho3, S. Hong1, Y. Kim1, G. Kim1, S. Ahn1, S. Park1, D.

Lee4, H. Jin5, S. Park6, S. Cho7, H. Lee1. 1Yonsei University, Wonju, Korea,
Republic of, 2Institute for Life Science and Biotechnology, Seoul, Korea,
Republic of, 3Daegu Health College, Daegu, Korea, Republic of, 4Hyejeon
College, Hongseoung, Korea, Republic of, 5Catholic University of Pusan,
Busan, Korea, Republic of, 6Daegu Hanny University, Daegu, Korea,
Republic of, 7Yonsei University, Seoul, Korea, Republic of
Background: Recently, the interferon gamma (IFN-) release assay (IGRA) was
introduced as an alternative immunodiagnostic method to the tuberculin skin test
(TST) for detecting latent tuberculosis infection (LTBI). However, IGRAs are
known to have limited sensitivity and cannot differentiate between active pulmonary
tuberculosis (PTB) disease and LTBI. Numerous cytokines and regulator factors
have been implicated in the pathogenesis and control of Mycobacterium tuberculosis
(MTB) infection. Therefore, additional cytokines including T helper 1 (TH1)type and T helper 2 (TH2)-type cytokines and chemokines associated with MTB
infection may improve the performance of IGRAs. Methods: In the present study,
a molecular diagnostic method using the real-time RT-PCR TaqMan assay, which
is able to quantitate mRNA expression levels, was developed for eight human targets
(IFN-, TNF-, IL-2R, IL-4, IL-10, CXCL9, CXCL10, and CXCL11) and evaluated
with three different patient groups (active PTB, LTBI, and healthy non-TB groups).
Results: Results revealed that positivity of TNF-, IL-2R, and CXCL10 in the active
PTB group was 96.43%, 96.43%, and 100%, respectively. The positivity of IL-2R
and CXCL10 in the LTBI group was 86.36% and 81.82%, respectively. Statistical
results revealed that TNF- and CXCL9 (both p<0.0001) were the best individual
markers for differentiating between the three different MTB infection groups. For
optimal sensitivity, the simultaneous detection of multiple targets was attempted.
The combination of IFN-, TNF-, and IL-2R and the combination of TNF-, IL2R, CXCL9, and CXCL10 showed the best performance for detecting active PTB
(both 100% positivity) and LTBI (86.36% and 81.82% positivity). Conclusion: These
results imply that the combination of suitable single markers is very useful for the
efficient diagnosis of MTB infection and the differentiation of MTB infection status.

A-330
Toxoplasma gondii IgG avidity among Brazilian IgG+/IgM+ women

V. V. Frade, I. C. B. Almeida, D. Panisa, V. T. Alstico, C. F. A. Pereira, N.

Gaburo. DASA, Sao Paulo Brasil, Brazil
Background: Toxoplasma gondii is prevalent in Brazil and and IgG and IgM is
routenly requested for pregnant women. The IgG avidity test analyzes the bonding
strength of the antigen-antibody complex after treatment with urea. In the acute phase
of the disease, the binding of antigen-antibody complex is easily dissociated because
the IgG avidity is reduced. This test maybe used as a supplemental assay for samples
that are IgM and IgG positive in order to indicate recent infection.
Objective: To describe the results obtained by the IgG avidity test in women that is
IgG and IgM positive for Toxoplasma gondii.
Methods: We have selected 413 samples from our routine that were IgM positive
with Toxo IgM kit (Roche, Mannheim, Germany) and presented IgG level higher
than 50 IU/mL by the Toxo IgG kit (Roche). The Toxoplasma gondii avidity test
was performed using TSI Toxok -G - Plus kit (DiaSorin, Saluggia, Italy) which is
an ELISA test. The avidity index was calculated by the ratio of the optical density of
the urea buffer treated sample divided by optical density of the sample without urea
treatment. Results below 30%, were considered recent infection (less than 12 weeks),
results higher than 60 % infection were considered more than 12 weeks of infection.
Between 30-60% the result was considered inconclusive.

L. E. DAgostino1, F. D. Ventimiglia2, J. A. Verna1, J. J. Bruno1, A. L.

Capparelli3. 1Laboratorio de Anlisis Clnicos e Inmunolgicos, La Plata,
Argentina, 2Universidad Nacional de La Plata, La Plata, Argentina,
3
INIFTA- Universidad Nacional de La Plata, La Plata, Argentina
Background:
Neopterin,
2-amino-4-hydroxy-6-(D-erytrhro-1,2,3trihydroxypropyl)-pteridine, is a low-molecular compound synthesized from
guanosine triphosphate (GTP) by macrophages. The clinical utility is like specific
marker for the activation of cellular immunity. Measurement of plasma neopterin
concentrations has been proposed as a test for monitoring the activity of infectious,
autoimmune and malignant diseases.
Objective: The purpose of this study was to evaluate GenWay Neopterin ELISA
kit performance and verification of the reference interval in native population in
Argentina.
Materials and methods: Plasma samples were obtained from 38 healthy volunteers
(aged 17-60 years). Neopterin was determined by ELISA GenWay Biotech Inc.
The verification of the reference interval was calculated using the CLSI guidelines
C28-A3. Bias and imprecision, were calculated using the protocol EP-10 from CLSI,
with three different standard concentrations (nmol.L-1): 4; 12 and 37. The results were
compared with the specifications of folate, due to neopterin desirable specifications
are not yet established. Intra-assay and inter-assay variability were determined in
ten normal plasma samples and ten samples with elevated neopterin concentrations
from rheumatoid arthritis patients (RA), QC plasma provided by the manufacturer and
compared with literature data. The criteria for acceptable performance were the target
value plus or minus 2 standard deviations. To study behavior among different sources
of specimens, variance analysis ANOVA was used.
Results: The limits of references values of neopterin were verified (< 10.0 nmol.L-1).
The lower and higher limits were 1.36 (percentile 2.5) and 9.93 (percentile 97.5)
nmol.L-1 respectively. Bias and imprecision, encountered in the standard concentrations
were: 13.6 % and 11.2%; 9.8 % and 3.1%; 7.6% and -1.4% respectively. Intra-assay
variability in normal and elevated samples was 7.9% and 11.4% respectively; in
QC samples (normal and elevated) were 9.8% and 6.1% respectively. Inter-assay
variability for the same group of samples was 8.2%; 13.8%; 10.4% and 7.8%
respectively. The lot to lot CVs for the two samples used for Quality Control of
three lots of the neopterin ELISA kit were 13.1% and 7.4% respectively. Total error
allowable (TEa) for neopterin was calculated for the three standard concentrations:
33.64%; 19.27%; 13.94%, TEa for folate is 39.0%.Variance analysis had not shown
statistical significance among the processed samples.
Conclusions: The verification of neopterin reference interval on healthy volunteers
in native population is consistent with literature data. This interval allowed
distinguishing between healthy and RA patients. The precision was verified by the
good intra and inter-assay coefficients of variation. The lot to lot CV was < 13.5%.
Although desirable specifications for neopterin are not yet established, we used
folate in terms of comparison; folate is a pterin derivative. We found that the mean
neopterins TEa was 22.3% and the highest value obtained, 33.64 %, was smaller
than the folate. We conclude that the results with GenWay ELISA on imprecision and
bias were acceptable when were compared with the well-known folates desirable
specifications. Variance analysis had not shown statistical significance between
neopterin measurements among different sources. In summary, GenWay ELISA
kit characterized here is valuable for clinical applications, is simple and rapid for
neopterin determinations in serum or plasma.

Result: The results are shown on Table 1. Only 1% of the females were considered
recent infected and 2.7% inconclusive

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S97

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

Materials and Methods:

A-332
Assessment of a Particle Immunofiltration Assay [PIFA] for Antibodies
Associated with Heparin Induced Thrombocytopenia (anti-HIT)

C. Morris1, K. Ballard2, C. Finch Cruz2, E. S. Pearlman2. 1University of

Tennessee Health Sciences Center, Memphis, TN, 2Veterans Affairs Medical
Center (VAMC), Memphis, TN
Background: The Department of Surgery at VAMC queried the clinical laboratory about
the feasibility of implementing a rapid assay for anti-HIT. The only assay satisfying
their specifications was a PIFA [Akers Biosciences; Thorofare, NJ] The question was
how to evaluate this assay given the low incidence of true positive (TP) cases of HIT as
indicated by a positive result on a functional assay e.g. the serotonin release assay [SRA].
Methods: To circumvent the low incidence of TPs we pooled serum specimens
from five patients that had been received for Vitamin D assay. PIFA was negative on
the pooled serum that was then used to serially dilute the positive control material
supplied by the manufacturer between 1:2 and 1:32. The diluted control material was
run using PIFA and the dilutions and the negative pooled serum were then sent to a
reference lab (RL) [Lab Corp; Raleigh, NC] where the material was tested using an
ELISA assay with a threshold index of 0.4. The entire procedure was repeated using
a different lot of positive control material (total of 12 determinations sent to the RL).
Results: Of the 12 simulated specimens, 5 were positive by both PIFA
and ELISA and 4 negative by both methods [overall concordance=75%]
Three specimens were positive by PIFA and negative by ELISA.
All discordant specimens had ELISA indices >0.3 but less than 0.4.
Conclusions: The study was small and SRA results were not available on the simulated
samples. However it would appear that FPs already known to be a problem when
employing ELISA would be increased when using PIFA as the latter apparently begins
generating positive results when the ELISA index is >0.3. A negative PIFA result is
however concordant with ELISA. PIFA may therefore be usable as the first step in an
algorithm insofar as a negative result would suggest that anti-HIT is unlikely and a
positive result could be referred for additional testing.

A-333
New immunological assays for diagnosis of Schistosoma mansoni for clinical
acute and/or chronic forms

R. F. Q. Grenfell1, V. Silva-Moraes1, D. Harn2, P. Z. Coelho1. 1Fundao

Oswaldo Cruz, Belo Horizonte, Brazil, 2University of Georgia, Athens, GA
Background: Control constraints of schistosomiasis include the lack of diagnostic
methods with high sensitivity. We initiated a prospective study in southeast Brazil
in order to develop sensitive diagnostic methods for Schistosoma mansoni infection,
with 4 endemic areas together with 80 travelers infected in a freshwater pool.
Methods: Sera, whole blood, urine and saliva samples from the patients were used for
the standardization of innovative diagnostic methods. Comparisons were performed
with eggs in feces, IgG titers, encephalomyelitis by NMR and clinical symptoms.
The new methods used were immunochromatography (dipstick), Immunomagnetic
Separation and ELISA with highly purified monoclonal antibodies.
Results: We could diagnose acute patients 10 days post-infection, also more than
95% of positive cases from chronic and low endemicity patients. New methods for
IgG detection using purified glycoprotein or recombinant protein or peptides (10
aminoacids) were superior to conventional ELISA.

Autoimmune neurological antibodies Anti-NMDA, Anti-VGKC, Anti-AQP4, AntiHu, Anti-Yo and Anti-Ri requested from January 2010 to December 2013 in our
institution were reviewed; only the first positive antibody result was considered for
patients with multiple requests. For Anti-AQP4, cerebrospinal fluid oligoclonal bands
(OCB) and pleocytosis information were reviewed as well.
Statistical analysis was performed using SPSS Version 17.0.
Results:
There were 443 requests of neurological antibodies for 203 patients, of which
three most common presenting complaints were seizures, altered mental state and
encephalitis.
The median age was 48.4 years old, with male to female ratio of 1.09, and 118
Chinese, 19 Malays, 17 Indians and 49 belonging to other ethnic groups. The median
age and ethnic breakdown for each antibody is shown in the table.
7.6% of these patients were positive for any of the neurological antibodies. AntiAQP4 had the highest seropositivity at 23.4%, followed by Anti-VGKC(9.9%), AntiNMDA(9.2%), Anti-Yo(3.9%), Anti-Hu(1.3%) and Anti-Ri(0%). No patients were
positive for more than 1 antibody.
Amongst patients with positive Anti-AQP4, they are older (55.6 versus 36.9 years
old in negative patients) and less likely to have a positive OCB or pleocytosis. Young
females were more likely to be positive for Anti-NMDA compared to older females or
males, as shown in the table.
Indian ethnicity was positively associated with Anti-VGKC positivity.
Conclusion:
Our data showed that autoimmune neurological antibody positivity were not
uncommon, however this could be due to requesting bias. Further studies will be
helpful to identify their prevalence in patients with different neurological complaints.
Type of neurological autoantibody and
Patients with Patients with
All patients
characteristics
positive result negative result
Anti-Aquaporin 4 Antibody (AQP4)
Number of patients (%)
47
11 (23.4)
36 (76.5)
Median age (years)*
44.6
55.6
36.9
Ratio of chinese:malay:indian: other
22:5:6:14 8:0:0:3
14:5:6:11
ethnic groups
Anti-N-methyl-D-aspartate Antibody
(NMDA)
Number of patients (%)
120
11 (9.2)
109 (90.8)
Median age (years)*
42.5
26.5
44.9
Ratio of chinese:malay:indian: other
71:13:6:20 5:1:0:5
66:12:6:25
ethnic groups
Anti-Voltage gated K channel (VGKC)
Number of patients (%)
81
8 (9.9)
73 (90.1)
Median age (years)
53.8
50.1
54.1
Ratio of chinese:malay:indian: other
53:3:8:17 3:1:3*:1
50:2:5:16
ethnic groups
Anti-Yo
Number of patients (%)
76
3(3.9)
73 (96.1)
Median age (years)
57.3
56.7
57.5
Ratio of chinese:malay:indian: other
53:5:6:12 2:1:0:0
51:4:6:12
ethnic groups
Anti-Hu
Number of patients (%)
77
1 (1.3)
75 (97.4)
Median age (years)
57.3
73.4
57.2
Ratio of chinese:malay:indian: other
53:7:4:12 0:1:0:0
53:6:4:12
ethnic groups
* Difference between patients with positive antibody compared to those with
negative antibody is significant with p value <0.05.

Conclusion: Best results were seen for recombinant protein with 100% of sensitivity.
Data showed 100% of sensitivity of chronic patients and 98% of acute patients.
Financial support: Fapemig, CNPq, Capes, Fiocruz, Fiotec, PDTIS (Brazil). Fulbright,
NIH, University of Georgia (USA).

A-334
Audit of Autoimmune Neurological Antibodies requesting in a Singapore
institution

M. Tan, L. Ong, B. Saw, A. Tan, Y. Tan, M. Voong, S. Saw, S. Sethi.

National University Hospital, Singapore, Singapore
Background: There is increased testing of autoimmune neurological antibodies for
evaluation of neurological diseases, but there is limited information on its prevalence
in our local population. We aimed to determine the neurological antibody requesting
patterns and seropositivity rates in patients presenting with neurological disorders in
our institution.

S98

A-335
Deficiency of CD16 in polymorphonuclear neutrophils. Myelodisplastic
disorder, paroxysmal nocturnal hemoglobinuria or neutrophilic FcGRIIIB gene
deficiency?. A case report.

B. Alvarez1, A. Gonzlez1, A. Lafuente2, E. Ruiz2, L. Conejo1, M.

Medrano1, A. Vlagea1, J. Alonso1, N. Del Amo1, C. Gonzlez1, D. Velasco1,
J. Villarrubia1, R. Guilln1, F. Cava Valenciano1. 1Laboratorio Central
BRSalud, Madrid, Spain, 2Hematology department. Hospital del Tajo,
Madrid, Spain
Background CD16 antigen, the human Fc receptor III (FcgRIII), is a protein
constitutively expressed on polymorphonuclear neutrophils (PMNs), monocytes
and NK-cells. CD16 participates in phagocytosis and in antibody-dependent
cellular cytotoxicity. It exists in 2 different forms encoded by 2 nearly identical
linked genes, FcgRIIIA and FcgRIIIB that generate alternative membrane anchored
molecules: FcgRIIIA (50-65 kd) is a transmembrane form expressed on NK cells

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

and macrophages; FcRIIIB (48 kd) is anchored through a phosphatidylinositol (PI)

linkage and expressed only on neutrophils.
Abnormal CD16 expression in cell surface might indicate myelodysplastic disorder
or indicate an acquired clonal hematologic disorder such as paroxysmal nocturnal
hemoglobinuria (PNH).
In the present study we report a case of total CD16 deficiency on PMNs.
Methods We studied the peripheral blood from a 72-year-old male that only presented
neutropenia. A 6-color flow cytometry tube was used including the markers CD16,
CD24, CD14, FLAER, CD33 and CD45. The cells were run on a 3-laser FACSCanto
II (BD) with FACSDiva software (BD) and analyzed using Infinicyt software
(Cytognos).
Results Immunophenotypic analysis of peripheral blood sample from patient showed
a lack of CD16 expression on neutrophils, but NK cells were CD16+. The GPIanchored proteins such as FLAER, CD24, and CD14 and all other myeloid antigens
were expressed normally in neutrophils and monocytes.
Conclusion These results suggest a neutrophilic FcGRIIIB gene deficiency and reject a
myelodysplastic disorder and paroxysmal nocturnal hemoglobinuria. Previous studies
indicate deficiency of CD16 does not compromise the host defence. Apparently, the
other receptors for IgG, CD32 and CD64, can compensate for the lack of CD16 Eur J
Clin Invest. 2004 Feb;34(2):149-55.

A-336
Diagnostic efficacy evaluation of IL-1RI, IL-1 and CDK2 in peripheral blood
and synovial fluid with rheumatoid arthritis

X. Huang, X. Chen, J. Zhuang. Second Affiliated Hospital of Guangzhou

University of Chinese Medicine, Guangzhou, China
Background:To explore the diagnostic efficacy of interleukin 1 receptor typeI(IL1RI), interleukin 1(IL-1 ) and cyclin dependent kinase 2(CDK2) in peripheral blood
and synovial fluid with rheumatoid arthritis (RA).
Methods: There were selected 94 cases with RA patients in rheumatology outpatient
and inpatient department, 40 cases with systemic lupus erythematosus (SLE) patients
in outpatient department, 20 cases with acute upper respiratory tract infection(AURTI)
patients in emergency department, 20 healthy persons. All subjects were eligible for
inclusion criteria. All subjects were drew vein blood 3 ml besides 1 ml knee joint
fluid from 24 patients with active RA patients. The level of IL-1RI, IL-1 and CDK2
were detected in serum and synovial by quantitation ELISA, then the three items were
evaluated diagnostic efficacy.
Results: There were all significant differences among experimental groups on either
IL-1RI, IL-1 or CDK2 by square variance analysis(P<0.001), respectively. On IL1RI, there were significant differences between RA patients group and RA patients
joint synovial fluid group or SLE patients group or acute upper respiratory tract
infection group and control group (P<0.001), respectively. On CDK2, there were
significant differences between RA active phage and RA relieve phage, between RA
patients group and RA patients joint synovial fluid group or SLE patients group or
acute upper respiratory tract infection group or control group (P<0.001) , respectively.
On IL-1, there were significant differences between RA active phage and RA relieve
phage, between RA patients group and RA patients joint Synovial fluid group or
healthy people group (P<0.001), respectively.
Compared RA patients active phage with RA relieve phage, area under curve of ROC
in CDK2 was largest, followed IL-1 andIL-1RI. Compared RA patients group with
SLE patients group, area under curve of ROC in CDK2 was largest, followed IL1RI. Compared RA patients group with acute upper respiratory tract infection patients
group, area under curve of ROC in IL-1RI was largest, followed CDK2. Compared
RA patients group with control group, area under curve of ROC in CDK2 was largest,
followed IL-1RI.
Conclusions:IL-1RI had low diagnostic efficiency next to CDK2, but it could
efficiently differentiate RA and acute upper respiratory infection (AURI). CDK2 had
higher diagnostic efficiency, which could efficiently differentiate active phase and
relieve phase of RA, and differentiate RA and SLE, but had low diagnostic efficacy
next to IL-1RI, in differentiating RA and AURI. CDK2 plus IL-1RI plus IL-1
paralleling joint diagnosis may increase diagnostic value of RA. CDK2 plus IL-1RI
plus IL-1 tandem joint diagnosis may increase early diagnostic value of RA.

A-337
Evaluation of an IgG3 assay for use on The Binding Site Next Generation
Protein Analyser

K. A. Samuels, J. R. Kerr, L. D. Southan, A. Kaur, S. J. Harding, P. J.

Showell. The Binding Site Ltd, Birmingham, United Kingdom
Assays for measurement of IgG subclasses in serum are routinely used in many
immunology laboratories in the diagnosis of IgG deficiencies. Abnormal levels of
one or more subclass may be associated with conditions including anaphylaxis,
autoimmune- and gut diseases as well as hypo- and hyper-gammaglobulinaemia.
Deficiency of IgG3 has been reported to be associated with viral infections of the
urinary tract. Here we describe the evaluation of an IgG3 assay for use on the Binding
Sites next generation protein analyser. The instrument is a random-access bench top
turbidimetric analyser capable of a wide range of on-board sample dilutions (up to
1/10,000) and throughput of up to 120 tests per hour. Precision is promoted by singleuse cuvettes which are automatically loaded and disposed of, whilst the utility is
enhanced through host interface capability, primary sample ID and bar coded reagent
management systems. The instrument automatically dilutes a single calibrator to
produce a calibration curve with a measuring range of 55 - 2200 mg/L at the standard
1/20 sample dilution. Samples lower than the bottom of the standard measuring range
are automatically retested at the lower 1/2 sample dilution, providing a measuring
range of 5.5-220 mg/L. Precision studies (CLSI EP05-A2) were performed at five
levels, over 21 days with 2 runs per day. Antigen levels of 8.9mg/L, 117.8 mg/L,
159.85mg/L, 258.3 mg/L and 1774.5mg/L were assessed for total and within run
precision over 3 reagent lots on 3 analysers. The coefficients of variation were 12.1%
and 8.4% for the 8.9mg/L sample, 8.3% and 5.1% for the 117.8mg/L sample, 9.9% and
8.2% for the 159.8mg/L sample, 6.6% and 3.8% for the 258.3mg/L sample and 5.8%
and 2.0% for the 1774.5mg/L sample respectively. Linearity was assessed by assaying
a serially-diluted patient sample pool across the width of the measuring range (552200mg/L) and comparing expected versus observed results. The assay showed a high
degree of linearity when expected values were regressed against measured values (y=
1.0136x - 11.368, R =0.9971). No significant interference (within 10%) was observed
on addition of bilirubin (20mg/dL), haemoglobin (500mg/dL) or chyle (1500 FTU)
when spiked into a sample with known IgG3 concentration. Correlation to the Binding
Site IgG3 assay for the SPA PLUS was performed using normal and clinical samples
(n=102, range 102.10-1836.90 mg/L). Good agreement was observed between assays
when anlysed by Passing-Bablok regression; y=1.07x -20.80. We conclude that the
IgG3 assay for the Binding Site next generation protein analyser is reliable, accurate
and precise and shows good agreement with existing assays.

A-338
Validation of a multiplex electrochemiluminescence assay for quantitation of
synovial fluid cytokines and establishing reference interval in non-infected
arthroplasty patients

X. Zhang, B. Gibson, S. J. Frangiamore, A. Saleh, M. J. Grosso, M. F.

Kovac, T. W. Bauer, E. T. Ricchetti, J. P. Iannotti, T. M. Daly. Cleveland
Clinic, Cleveland, OH
Background: Periprosthetic joint infection (PJI) is a severe complication following
arthroplasty. Emerging studies suggest that proinflammatory cytokines in synovial
fluid are promising markers for PJI diagnosis. However no assay has been validated
for measurement of cytokines in synovial fluid and baseline levels have not been
established in non-infected artroplasty patients. We developed and validated an
electrochemiluminescence (ECL) multiplex assay for quantification of 9 cytokines
in human synovial fluid and established reference intervals in artroplasty patients.
Method: The human ultrasensitive cytokine assay (Meso Scale Discovery.
Rockville, MD) was modified to measure IL-1, IL-2, IL6, IL-8, IL-10, IL-12p70,
GM-CSF,IFN-, TNF- in synovial fluid. The assay was validated following CLSI
guidelines. Patients who underwent primary arthroplasty or arthroscopic rotator cuff
repair in hip, knee or shoulder, were enrolled prospectively. These patients were
known not have infections and thus were ideal controls for PJI. Synovial fluid was
collected intraopratively. Results: The intra- and inter-assay imprecision (n=20)
for all the cytokines was less than 15%. Assay accuracy was evaluated by spiking
recombinant cytokines into synovial fluid. Percent recovery was within 10020%.
The assay sensitivity and linear range was summarized in the table. No difference
was observed between different genders or the two surgical groups. However, patients
>=70 years old had significantly higher synovial IL-6 and IL-8 levels than other age
groups. Knee and shoulder joints showed similar cytokines levels while hip synovial
fluid contained significant increases in IL-6 and IL-8(Fig1). Reference intervals were
established for synovial cytokines of knee/shoulder joints in non-infected arthroplasty

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S99

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

patient (<70 years old, see table). Conclusion: The ECL assay provides a reliable
method for quantitation of multiple cytokines in less than 25L synovial fluid. IL-6
and IL-8 are the major cytokines in synovial fluid. Age and joint specific reference
interval was necessary for diagnosis of PJI using synovial cytokines.

Results: Precision showed %CVs of 2.6-5.9 over the range of 34.18 to 3799.79
pg/mL. LoB and LoQ were 0.35 pg/mL and 3.11 pg/mL, respectively. Method
comparison showed a correlation coefficient of r = 0.98 and a slope = 0.98. The 99th
percentile value of 166 Japanese apparent healthy subjects was 22.4 pg/mL. This
value was equivalent to the 99th percentile value reported in the package insert from
Abbott Laboratories. The multiple regression analysis revealed that male sex and age
were independent factors for increased cardiac troponin-I levels. Cardiac troponin-I
levels were significantly higher in males and positively associated with the age.
Conclusion: The ARCHITECT STAT high sensitive Troponin-I assay demonstrated
good analytical performance and improved imprecision at low concentration in
comparison to the conventional ARCHITECT STAT Troponin-I assay. ARCHITECT
STAT high sensitive Troponin-I meets the definition of high-sensitivity troponin
reagent proposed by the IFCC Task Force. The 99th percentile value which was
established by the manufacturer may be used in Japan, but it would be necessary to
consider the effects of age and gender.

A-340
A Lyophilized Quality Control Material for Human Allergen Specific IgE
Testing.

Z. Bradic, J. Cole, A. Riviere. Bio-Rad Laboratories, Inc., Irvine, CA

In vitro measurements of circulating IgE antibodies specific for allergens (sIgE), such
as pollens, foods, drugs, venoms, insects, mites, molds and epidermals, are useful for
the diagnostic assessment of a patients allergies. Lyphochek Allergen sIgE Control
Negative and Panel A is a lyophilized, human serum based third party quality control
material to monitor the precision of laboratory testing procedures for specific IgE
antibodies.
Procedure: Control material was tested for fifteen sIgE on various testing platforms
for recoveries, precision, opened vial stability, and accelerated temperature studies for
shelf-life prediction.
Results: The control material was assayed to determine sIgE recovery values using
several test platforms. The Negative control results were all below the detection limit
of the test method used. The table below lists results for the sIgE present in the Panel
A control.

A-339
Evaluation of the immunoassay reagent kit for high sensitive troponin-I
(ARCHITECT STAT high sensitive Troponin-I) with fully-automated
chemiluminescent immunoassay analyzer, and the clinical trials in Japan using
medical checkup examinees

I. Tanaka1, K. Sakamaki2, T. Kasumi2, K. Nagashima2, M. Akuzawa2,

K. Nakajima3, Y. Shimomura3, N. Nagano3, Y. Ando4, H. Kinukawa1, T.
Yoshimura1. 1Abbott Japan Co., Ltd., Chiba, Japan, 2Hidaka-kai, Hidaka
Hospital, Health Care Center, Gunma, Japan, 3Hidaka-kai, Hidaka
Research Center, Gunma, Japan, 4Hidaka-kai, Gunma, Japan
Background:Troponin-I is widely used as an aid in the diagnosis of acute myocardial
infarction because of its cardiac specificity. In Japan, the measurement of troponin-I
belongs to class-I in Guidelines for Management of Acute Coronary Syndrome without
Persistent ST Segment Elevation(Japanese Circulation Society (JCS) 2007)and ClassII in Guidelines for Treatment of Chronic Heart Failure (JCS 2010). Use of a high
sensitivity troponin assay is recommended due to improved analytical performance.
An assay is considered to be high sensitive if the %CV value is 10% or less at the 99th
percentile and if at least 50% of normal subjects are detectable. The goal of this study
was to evaluate the analytical performance characteristics of ARCHITECT STAT high
sensitive Troponin-I assay on the fully automated chemiluminescent ARCHITECT
analyzer and to determine the 99th percentile upper reference limit and the factors
which affect troponin-I in Japanese populations undergoing routine health checks.
Methods:The ARCHITECT STAT high sensitive Troponin-I assay is a double
monoclonal antibody sandwich assay. This assay has an assay time of approximately
18 minutes and an analytical range of 0.0 to 50,000.0 pg/mL. Precision was
evaluated with 3 different sera in duplicate, twice a day for 5 days (n=20). The limit
of blank (LoB) and limit of quantitation (LoQ) were evaluated with 20 consecutive
measurements of calibrator A (zero concentration) and 5 different patient sera in
duplicate for 5 days. Dilution linearity was determined by diluting 3 serum samples
with the dedicated diluted solution. Correlation was performed with 50 samples
spanning the range of 32-35,389 pg/mL on ARCHITECT STAT Troponin-I assay (the
conventional troponin-I assay). The 99th percentile of Japanese healthy population
and the factors which affect troponin-I were evaluated with 283 patients undergoing
routine health checks.

S100

Open vial stability studies showed that the sIgE in the control is stable for a minimum
of 28 days at 2-8C. Accelerated temperature studies predicted a shelf-life of over 3
years when stored in lyophilized form at 2-8C.
Conclusion: Lyphochek Allergen sIgE Control, Negative and Panel A levels are
suitable to monitor the precision of laboratory testing procedures for sIgE in human
serum or plasma.

A-341
Effect of CD95 on inflammatory response in rheumatoid arthritis fibroblast-like
synoviocytes

X. Huang. Second Affiliated Hospital of Guangzhou University of Chinese

Medicine, Guangzhou,Guangdong, China
Background:Many CD95-expressing cells dont always undergo apoptosis after
stimulation with CD95 ligation. To investigate the role of expression of CD95 (Fas/

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

Apo1) on inflammatory response in fibroblast-like synoviocytes(FLS) obtained

from rheumatoid arthritis(RA) and to evaluate the role of phosphatidylinositol
3-kinase(PI3K)/protein kinase B(PKB or Akt) pathways within this process.
Methods: The expression levels of CD95 were measured by immunohistochemistry
and reverse transcription polymerase chain reaction(RT-PCR). Apoptotic cells were
detected by in situ apoptosis detection (TUNEL) assay. The RA-FLS were treated with
agonistic anti-CD95 antibody or CD95 siRNA, then the proliferation was detected
by CCK-8,and mRNA level of inflammatory cytokines were detected by RT-PCR.
After the RA-FLS were treated with agonistic anti-CD95 antibody, the total Akt and
pAkt protein expression was analyzed by western blot, and the changes mentioned
above were observed while incubated with the PI3K inhibitor LY294002. Results: A
significant increase of CD95 antigen was found in RA compared with osteoarthritis
(OA) samples, while apoptosis in RA synovial tissue were not obvious. Low
concentrations of agonistic anti-CD95 antibody could promote RA-FLS growth and
IL-6 mRNA expression, while high concentrations could induce apoptosis and both
of these phenomenons were inhibited by CD95 siRNA. Agonistic anti-CD95 antibody
could stimulate the expression of pAkt, and PI3K specific inhibitor LY294002 could
induce opposite change.
Conclusion: Stimulation of CD95 could promote RA-FLS proliferation and
inflammation, and activation of the PI3K/Akt signaling pathway might be the
potential mechanism.

A-342
Quantitative detection of Plasmodium falciparum Histidine Rich Protein 2 in
saliva

up to 120 tests per hour. Precision is promoted by single-use cuvettes which are
automatically loaded and disposed of, whilst the utility is enhanced through host
interface capability, primary sample ID and bar coded reagent management systems.
The instrument automatically dilutes a single calibrator to produce a calibration curve
with a measuring range of 0.1928 - 6.17 g/L at the standard 1/10 sample dilution.
Precision studies (CLSI EP5-A2) were performed at five levels in duplicate over 21
working days. Five antigen levels were assessed for total, within-run, between-run and
between-day precision, using one lot of reagent on three analysers. The coefficients of
variation were 3.2%, 1.0%, 1.2% and 2.8% for the 5.3g/L sample, 2.8%, 0.7%, 1.0%
and 2.5% for the 4.1g/L sample, 3.1%, 1.1%, 1.1% and 2.7% for the 3.5g/L sample,
2.4%, 0.9%, 0.9% and 2.1% for the 2.3g/L sample and 3.5%, 2.1% 1.3% and 2.5%
for the 0.34g/L sample. Linearity was assessed by assaying a serially-diluted patient
sample pool across the width of the extended measuring range (0.184 - 6.363 g/L)
and comparing expected versus observed results. The assay showed a high degree of
linearity when expected values were regressed against measured values (y=1.002x
- 0.01903, R = 0.9998). No significant interference was observed on addition of
bilirubin (20mg/dL), haemoglobin (500mg/dL) or chyle (1500 formazine turbidity
units) when spiked into samples with known Alpha 2-Macroglobulin concentrations at
the standard sample dilution. Correlation to the Binding Site Alpha 2-Macroglobulin
assay for the SPA PLUS analyser was performed using normal and clinical samples
(n=146, range 0.233-5.554g/L). Good agreement was demonstrated by PassingBablok regression; y=1.03x - 0.01g/L. We conclude that the Alpha 2-Macroglobulin
assay for the Binding Site Next generation protein analyser is reliable, accurate and
precise and shows good agreement with existing assays.

A-345

F. A. Botchway1, C. E. Lekpor2, R. Ansah3. 1Department of Child Health,

Korle Bu Teaching Hospital, Accra, Ghana, 2Department of Medical
Laboratory, Kwame Nkrumah University Of Science and Technology,
Kumasi, Ghana, 3University of Ghana, Legon, Ghana
Background:Malaria is a global health priority with a heavy burden of fatality and
morbidity. Improvements in field diagnostics are needed to support the agenda for
malaria elimination. Saliva has shown significant potential for use in non-invasive
diagnostics, but the development of off-the-shelf saliva diagnostic kits requires
best practices for sample preparation and quantitative insight on the availability of
biomarkers and the dynamics of immunoassay in saliva. This study measured the
levels of the PfHRP2 in patient saliva.
Methods:Matched samples of blood and saliva were collected between March and
August, 2011 from forty patients at the ER and OPD of the pediatric unit of Korle Bu
Teaching Hospital. Parasite density was determined from thick-film blood smears.
Concentrations of PfHRP2 in saliva of malaria-positive patients were measured
using a custom chemiluminescent ELISA in microtitre plates. Forty negative-control
patients were enrolled. Saliva samples were stabilized with protease inhibitor
Results:Of the forty patients with microscopically confirmed P. falciparum malaria,
thirty seven tested positive for PfHRP2 in the blood using rapid diagnostic test kits,
and forty for PfHRP2 in saliva. All negative-control samples tested negative for
salivary Pf HRP2. The ELISA greed with microscopy with 100 % sensitivity and
100 % specificity. Salivary levels of PfHRP2 ranged from 15 to 1,162 pg/mL in the
malaria-positive group.
Conclusion:Saliva is a promising diagnostic fluid for malaria when protein
degradation and matrix effects are mitigated. Systematic quantitation of other
malaria biomarkers in saliva would identify those with the best clinical relevance and
suitability for off-the-shelf diagnostic kits.

A-343
Evaluation of an Alpha 2-Macroglobulin assay for use on the Binding Site Next
Generation Protein Analyser

S. L. Stone, A. Kaur, A. J. Alvi, S. J. Harding, P. J. Showell. The Binding

Site Ltd, Birmingham, United Kingdom
Alpha 2-Macroglobulin is a 725 kDa protein which, due its high molecular weight,
is distributed almost exclusively in the intravascular pool. Increased levels of Alpha
2-Macroglobulin are associated with nephrotic syndrome, liver cirrhosis and diabetes
mellitus. Measurement of Alpha 2-Macroglobulin can also aid in the diagnosis of
blood-clotting or clot lysis disorders. Here we describe the evaluation of a serum
Alpha 2-Macroglobulin assay for use with the Binding Sites next generation protein
analyser. The instrument is a random-access bench top turbidimetric analyser capable
of a wide range of on-board sample dilutions (up to 1/10,000) and throughput of

Evaluation of the performance characteristics of a new assay for Ferritin on

Toshiba 2000FR

A. Shimakawa1, M. Kano2, R. Tachibana2, Y. Ito2. 1JA Niigata Kouseiren

Sado General Hospital, Niigata, Japan, 2Denka Seiken Co.,Ltd., Tokyo,
Japan
Background:
Ferritin has been a well known marker over the past 20 years as its level in blood
is abnormally increased with certain disorders. Furthermore, more recently high
sensitivity is increasingly required to diagnose the iron deficiency anemia, for which
a lower threshold is set at 12 g/L or below. On the other hand, serum Ferritin levels
may become very high in some disorders such as hemochromatosis, hemosiderosis,
etc. Thus, assays for Ferritin levels in serum and plasma are required to have a wide
assay range with good prozone (high dose hook effect) tolerance and with a high
sensitivity for accurate and reliable measurements. We evaluated the FERNX, a new
latex particle-enhanced turbidimetric immunoassay for serum and plasma Ferritin
levels with ultra-sensitivity and good prozone tolerance, available from Denka
Seiken Co., Ltd. We evaluated the FERNX comparing to the current reagent and also
monitored the on-use stability at lower end.
Methods:
We carried out the study on a TBA-2000FR automated clinical chemistry analyzer
(Toshiba). The performance data was compared to the current reagent available
from Denka Seiken by testing lower detection limit, precision, linearity, prozone,
interferences and correlation. The on-use stability was assessed by testing lower
detection limit every week with the reagent which was set on the analyzer during the
whole study.
Results: The FERNX showed the correlation coefficient 0.99 against the current
reagent with 170 clinical samples. No interferences were observed with hemolysis
(hemoglobin: 487 mg/dL), icterus (Bilirubin: 20 mg/dL) or lipemia (1410 FTU)
with both reagents. CVs (SDs) of the new assay from within-run imprecision with
3 different samples (10 g/L, 60 g/L and 310 g/L) were smaller than the current
reagent. The new assay showed prozone tolerance better than the current reagent.
Recoveries by the new assay were higher than the upper limit of the measuring range
(1,000 g/L) with samples containing actual Ferritin concentrations up to 50,000
g/L, on the other hand recoveries by the current reagent were lower than the upper
limit with the samples with actual Ferritin concentrations at 6,000 g/L or higher.
Lower detection limit was 3-fold smaller than the current reagent (2 g/L and 6 g/L,
respectively). The lower detection limit remained at the sample level (2-3 g/L)
during the on-use stability study up to 5 weeks.
Conclusion: The FERNX showed better performance compared to the current reagent.
The FERNX showed excellent prozone tolerance and also excellent sensitivity even in
the on-use stability study. The FERNX, a new latex particle-enhanced turbidimetric
immunoassay reagent for serum and plasma Ferritin, is useful for diagnosis of wide
variety disorders including the iron deficiency anemia.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S101

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

A-346
Development of a latex-enhanced immunoturbidimetric assay for the
measurement of MMP-3 levels on automated clinical chemistry analyzers

A. Hirayama, C. Kaneko, M. Asami, K. Okuyama, S. Kitahara, T. Shimizu,

Y. Nakamura, H. Takahashi, N. Fujita, H. Yago, K. Saito. SEKISUI
MEDICAL CO., LTD, Ryugasaki, Ibaraki Pref., Japan
Matrix metalloproteinase-3 (MMP-3) is a proteolytic enzyme produced by synovial
cells and which participates in joint destruction. The concentration of serum MMP-3
in normal subjects is reported to range from approximately 17.3 to 121 ng/mL, while
patients with rheumatoid arthritis (RA) display significantly increased serum MMP3 levels as the condition worsens. It has been demonstrated that measuring serum
MMP-3 levels is a useful marker for evaluating inflammatory activity, prognostic
outcome and therapeutic effect in RA.
We have developed a new method for measurement of serum and plasma MMP-3
levels for use in clinical chemistry analyzers. This method is based on latex-enhanced
immunoturbidimetry, using anti-human MMP-3 mouse monoclonal antibodies. The
concentration is determined by measuring the change in absorbance that results from
agglutination of latex particles.
The reagents are supplied ready-to-use, and the assay can be completed within 10
min. Using a Roche/Hitachi 917 auto analyzer, 2.4 L of human serum or plasma
was mixed with 120 L of the first buffer solution and incubated for 5 min at 37C.
Subsequently, 40 L of the second reagent, which contains the monoclonal antibodycoated latex particles, was added and the absorbance was monitored at 570 nm/800
nm (main/sub wavelengths) for 5 min.
The lower detection limit for MMP-3 was 10 ng/mL, and the upper quantitation
limit was 1,600 ng/mL. No prozone effect was observed in MMP-3 samples of
concentrations from 1,600 through 2,500 ng/mL. The within run C.V. (n=10) at 100
ng/mL, 200 ng/mL, and 400 ng/mL was 1.2 %, 0.6 %, and 0.9 %, respectively. The
between run C.V. (n=10) at 100 ng/mL, 200 ng/mL, and 400 ng/mL was 1.7 %, 2.1
%, and 2.6 %, respectively. Interference studies showed no effect from bilirubin,
hemoglobin, rheumatoid factor (RF), or chyle at concentrations of 20 mg/dL, 500 mg/
dL, 500 IU/dL, and 2,000 formazin turbidity units, respectively.
Comparison of our assay kit with the approved IVD reagent, the principle of which
is enzyme immunoassay, yielded a correlation coefficient of 0.980 and an equation of
Y (present method) = 0.95X (the ELISA kit) + 10.53 (n = 115 serum samples). We
concluded that this assay reagent provides an accurate, precise, and simple method for
routine measurement of MMP-3 levels in serum and plasma samples.

A-347
Development of a Duplex Assay for the Simultaneous Detection of AntiThyroglobulin and Anti-Thyroid Peroxidase Autoantibodies Employing Biochip
Array Technology.

T. McFadden, S. Deegan, C. Richardson, S. Brockbank, J. Lamont, R.

McConnell, S. FitzGerald. Randox Laboratories Limited, Crumlin, United
Kingdom
Background: Autoantibodies (aAb) are established clinical markers of autoimmune
disease; in particular, they aid in the diagnosis of autoimmune thyroid disease (AITD)
and distinguish it from other forms of thyroiditis. AITD causes cellular damage and
alters thyroid gland function by humoral and cell mediated mechanisms. A characteristic
feature of AITD is the production of aAb to thyroglobulin (Tg) and thyroid peroxidase
(TPO), key regulatory proteins in the synthesis of the hormone thyroxine. Elevated
serum levels of Tg aAb and TPO aAb have been shown to be associated with chronic
thyroiditis such as Hashimotos thyroiditis (which results in hypothyroidism) and
Graves disease (which results in hyperthyroidism). TPO aAb are considered a more
sensitive marker of thyroid autoimmunity; however, depending on the patient, TPO
aAb may be low while Tg aAb are elevated, thus dual measurement of TPO aAb and Tg
aAb will facilitate a more accurate diagnosis of thyroid autoimmunity. The aim of this
study was to develop a duplex detection system that will allow for the simultaneous
detection of Tg aAb and TPO aAb from a single serum sample using biochip array
technology. This represents a useful analytical tool for applications in clinical settings.
Methods: Human TPO and Tg proteins were immobilized to discrete testing
regions (DTR) on a biochip surface using an indirect sandwich assay format.
Chemiluminescent signal from each DTR was detected by digital imaging
technology on the Evidence Investigator analyser. The multi-analyte calibrators
for the standard curve were developed from serum samples containing high levels
of TPO aAb with low levels of Tg aAb and vice versa. The calibrators were then
standardised using NIBSC reference material 65/93 and 66/387 for Tg aAb and

S102

TPO aAb respectively. A correlation study was conducted with a cohort of 236
clinical serum samples, which were assessed for both Tg aAb and TPO aAb,
using the biochip array technology and commercially available immunoassays.
Results: Cross-reactivity and interference testing demonstrated that each individual
assay was specific for its target analyte. Both assays demonstrated high sensitivity
with detection levels of 0.08 IU and 0.002 IU for Tg aAb and TPO aAb respectively.
Mean %recovery for the reference material was 107% with a %CV of 5.22 for Tg
aAb and 103% with a %CV of 18.63 for TPO aAb. Correlations to the assigned
values resulted in a correlation coefficient of 0.968 and a slope of 0.9372 for Tg
aAb and a correlation coefficient of 0.918 and a slope of 0.8446 for TPO aAb.
Conclusion:This study reports on the development of a clinical diagnostic product
for the simultaneous measurement of TPO aAb and Tg aAb in the detection of AITD.
Using biochip array technology, this duplex assay simultaneously measures levels
of both Tg and TPO aAbs from a single sample, offering advantages over current
diagnostic tools which use individual tests for the measurement of these aAbs. This
newly developed assay uses low sample volume and will provide a highly sensitive
and specific test for the detection of each analyte in a clinical setting.

A-349
Comparison of the AESKU HELIOS IFA system with another ANA Screening
Method

K. Tokoro1, M. Thun1, B. Larida1, T. Matthias2, B. Rabin3. 1AESKU.INC,

Oakland, CA, 2AESKU.KIPP Institute, Wendelsheim, Germany, 3University
of Pittsburgh Medical Center, Pittsburgh, PA
Background: This study compares the AESKUSLIDES HEp-2 cell line processed
on the fully automated HELIOS IFA analyzer (AESKU.SYSTEMS) to the NOVA
Lite HEp-2 cell line (INOVA) processed on the PhD system (Bio-Rad Laboratories).
Methods: 82 de-identified serum samples were tested on the HELIOS automated
analyzer at AESKU, Oakland and processed on the PhD system at University of
Pittsburgh Medical Center (UPMC). Base dilutions were performed at 1:80 and
titrated to an end-point dilution of 1:1280. Several discrepant positive samples were
subsequently run in the BioPlex 2200 system using the BioPlex 2200 ANA screen
(Bio-Rad Laboratories). Two independent clinical laboratory scientists evaluated the
results at both locations.
Results: Qualitatively, 80/82 (97.6%) HELIOS results were concordant with the PhD
system. Titration results also correlated well: 19 samples were double negative, 10
were borderline (1:80 or negative), 41 within 1 titer, 7 within 2 titers, 4 within 3 titers,
and 1 within 4 titers Borderline outcomes were treated as comparatively equal when
paired results were 1:80 and negative.
5 out of 7 samples within two titers of the PhD/INOVA results had higher HELIOS
end-point titrations (range 1:320-1:1280). 3 out of 4 samples with a distance of 3
titers had higher HELIOS end-point titrations. Clinical data and BioPlex results of
these 3 samples indicated autoimmune hepatitis or Sjgrens syndrome and an SSA or
Centromere result >8 (BioPlex cutoff <1.0). Two additional samples with no clinical
history of connective tissue diseases (1:320/negative and 1:80/1:1280, PhD:HELIOS)
had a BioPlex result of negative and SSA/SSB >8 respectively. The clinical and
diagnostic accuracy of the AESKUSLIDE/HELIOS reagent system is favorable based
on higher end-point titrations and confirmatory data.
Conclusion: Many laboratories allow a comparative titer discrepancy of 1 dilution as
a diagnostic convention when determining precision. Therefore, the combination of
AESKUSLIDES and HELIOS analyzer has a higher sensitivity and specificity than
the INOVA/PhD system. AESKU ANA IFA reagent systems are designed to be more
clinically relevant to disease state individuals and are therefore more diagnostically
significant.

A-350
Evaluation of an IgG1 assay for use on the Binding Site Next Generation
Protein Analyser

E. L. Freeman, J. R. Kerr, L. Southan, A. Kaur, S. J. Harding, P. J. Showell.

The Binding Site Ltd, Birmingham, United Kingdom
Assays for measurement of IgG subclasses in serum are routinely used in many
immunology laboratories in the diagnosis of IgG deficiencies. Abnormal levels of
one or more subclass may be associated with conditions including anaphylaxis,
autoimmune- and gut diseases as well as hypo- and hyper-gammaglobulinaemia.
Reduced IgG1 levels are often indicative of general immunodeficiency. Here we
describe the evaluation of an IgG1 assay for use on the Binding Sites next generation
protein analyser. The instrument is a random-access bench top turbidimetric analyser

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

capable of a wide range of on-board sample dilutions (up to 1/10,000) and throughput
of up to 120 tests per hour. Precision is promoted by single-use cuvettes which are
automatically loaded and disposed of, whilst the utility is enhanced through host
interface capability, primary sample ID and bar coded reagent management systems.
The instrument automatically dilutes a single calibrator to produce a calibration curve
with a measuring range of 1500-36,000mg/L at the standard 1/10 sample dilution,
with sensitivity of 150mg/L. High samples were remeasured at a dilution of 1/40
with a measuring range of 6000-144,000mg/L. Precision studies (CLSI EP5-A2) were
performed at nine levels in duplicate over 21 working days and were assessed for
total, within-run, between-run and between-day precision, using one lot of reagent
on three analysers. The coefficients of variation were 3.2%, 1.7%, 2.4% and 1.1% for
the 522 mg/L sample, 5.5%, 2.6%, 2.5% and 4.2% for the 2871mg/L sample, 3.0%,
1.6%, 1.4% and 2.1% for the 3083mg/L sample, 3.3%, 1.8%, 1.8% and 2.1% for the
4869mg/L sample, 5.5%, 1.2%, 5.4% and 0% for the 7179mg/L, 4.1%, 1.5%, 3.6%
and 1.4% for the 12,131mg/L sample, 3.4%, 1.2%, 3.0% and 1.1% for the 14,542mg/L
sample, 5.4%, 2.0%, 3.6% and 3.5% for the 13,847mg/L sample and 4.7%, 3.1%,
2.5% and 2.5% for the 28,132mg/L sample respectively. Linearity was assessed by
assaying a serially-diluted patient sample pool across the width of the measuring
range (1500-36,000mg/L) and comparing expected versus observed results. The
assay showed a high degree of linearity when expected values were regressed against
measured values (y= 0.97x + 101.70, R = 1.00). No significant interference (within
10%) was observed on addition of bilirubin (20mg/dL), haemoglobin (500mg/dL)
or chyle (1500 formazine turbidity units) when spiked into a sample with known
IgG1 concentrations when run using the minimum sample dilution. Correlation to
the Binding Site IgG1 assay for the SPA PLUS was performed using 142 samples
(range 1064 - 16,822mg/L). Good agreement was observed between assays (mean
6273mg/L; range 1064 - 16,822mg/L v mean 6199mg/L; range 1055-15,177g/L),
Passing-Bablok regression; y=1.05x - 228.61. We conclude that the IgG1 assay for
the Binding Site next generation protein analyser is reliable, accurate and precise and
shows good agreement with existing assays.

A-353
Examining the Frequency of Autoantibodies in the Brazilian Population in 2013
Using the Multiplex Technique

F. B. Flumian, F. O. Paiva, D. C. B. Carvalho, M. A. Rigone, T. N. Carvalho,

O. P. Denardin, N. Gaburo. DASA, Sao Paulo, Brazil
Background: Autoantibodies are antibodies that react against the organisms own
components. The appearance of a large number of autoantibodies is a pathological
condition that can occur in a number of diseases, such as diabetes mellitus type 1,
systemic lupus erythematosus, Sjgrens syndrome, Hashimotos thyroiditis, Graves
disease and rheumatoid arthritis, among others.Objective: Describe the frequency of
autoantibodies in Brazilian samples using the Luminex technology.
Methods: In the period from January to December 2013, 42,249 serum samples
of patients from all over Brazil were analyzed by DASAs Manual Immunology
laboratory. The samples were tested for the following autoantibodies: dsDNA, SSA,
SSB, Sm, RNP, Scl-70, Jo-1, Centromere B and Histone, using the Athena MultiLyte ANA II Plus Test System (Zeus Scientific, Raritan, NJ) and the Luminex 200 IS
equipment, version 2.3 (Luminex, Austin, TX). The results were interpreted using the
Luminex 200 software, version 2.3, according to the reference value indicated by the
kits manufacturer.
Results: Of the samples tested for autoantibodies using the Luminex method, 27,884
(66%) presented a negative result and 12,674 (30%) presented a positive result, while
1,689 (4%) presented an inconclusive result. Among the samples that presented a
positive result, the most prevalent autoantibody observed was SSA, present in 28.3
% of the samples analyzed. SSA is related to diseases such as Sjgrens syndrome,
Systemic Lupus Erythematosus, Rheumatoid Arthritis, SS/SLE Overlap Syndrome,
Subacute Cutaneous LE (SCLE), Neonatal Lupus and Primary Biliary Cirrhosis,
among others. Graph 1 represents the autoantibodies detected and their respective
frequencies in the population studied.
Conclusion: The Luminex xMAP technology enables the detection of multiple
analytes in a single reaction well, reducing operating costs and physical space,
and may be widely applicable to various processe Graph 1 - Frequency of the
autoantibodies detected.

A-351
Evaluation of an IgG2 assay for use on the Binding Site Next Generation
Protein Analyser

L. D. Southan, E. L. Freeman, A. Kaur, S. J. Harding, P. J. Showell. The

Binding Site Ltd, Birmingham, United Kingdom
Assays for measurement of IgG subclasses in serum are routinely used in many
immunology laboratories in the diagnosis of IgG deficiencies. Abnormal levels of
one or more subclass may be associated with conditions including anaphylaxis,
autoimmune- and gut diseases as well as hypo- and hyper-gammaglobulinaemia.
In particular, reduced production of IgG2 in children is associated with recurrent
infection. Here we describe the evaluation of an IgG2 assay for use on the Binding
Sites next generation protein analyser. The instrument is a random-access bench top
turbidimetric analyser capable of a wide range of on-board sample dilutions (up to
1/10,000) and throughput of up to 120 tests per hour. Precision is promoted by singleuse cuvettes which are automatically loaded and disposed of, whilst the utility is
enhanced through host interface capability, primary sample ID and bar coded reagent
management systems. The instrument automatically dilutes a single calibrator to
produce a calibration curve with a measuring range of 194-7000mg/L at the standard
1/10 sample dilution, with sensitivity of 2.96mg/L. High samples were remeasured at
a dilution of 1/40 with a measuring range of 776-28,000mg/L. Precision studies (CLSI
EP5-A2) were performed at five levels in duplicate over 21 working days and were
assessed for total, within-run, between-run and between-day precision, using one lot
of reagent on three analysers. The coefficients of variation were 4.8%, 1.5%, 2.4% and
3.9% for the 80mg/L sample, 5.7%, 3.3%, 1.6% and 4.4% for the 390mg/L sample,
3.5%, 1.0%, 1.3% and 3.0% for the 1828mg/L sample, 2.6%, 1.0%, 1.4% and 2.0%
for the 2982mg/L sample and 2.9%, 1.2%, 1.4% & 2.3% for the 5855mg/L sample
respectively. Linearity was assessed by assaying a serially-diluted patient sample pool
across the width of the measuring range (194-7000mg/L) and comparing expected
versus observed results. The assay showed a high degree of linearity when expected
values were regressed against measured values (y= 0.9612x + 0.0298, R = 0.9878).
No significant interference (within 10%) was observed on addition of bilirubin
(20mg/dL), haemoglobin (500mg/dL) or chyle (1500 formazine turbidity units) when
spiked into a sample with known IgG2 concentrations when run at the minimum
sample dilution. Correlation to the Binding Site IgG2 assay for the SPA PLUS was
performed using 143 samples (range 386 - 8031mg/L). Good agreement was observed
between assays when analysed by Passing-Bablok regression; y=0.99x - 0.02. We
conclude that the IgG2 assay for the Binding Site next generation protein analyser is
reliable, accurate and precise and shows good agreement with existing assays.

A-354
Evaluation of a Cystatin C assay for use on the Binding Site Next Generation
Protein Analyser

H. Johnson, F. Murphy, S. J. Harding, P. J. Showell. The Binding Site Ltd,

Birmingham, United Kingdom
Cystatin C has been shown to be superior to creatinine as a marker of glomerular
filtration rate and is increasingly used in the diagnosis of renal dysfunction. Here we
describe the evaluation of a Cystatin C assay for the Binding Sites next generation
protein analyser. The instrument is a random-access bench top turbidimetric analyser
capable of a wide range of on-board sample dilutions (up to 1/10,000) and throughput
of up to 120 tests per hour. Precision is promoted by single-use cuvettes which are
automatically loaded and disposed of, whilst the utility is enhanced through host
interface capability, primary sample ID and bar coded reagent management systems.
The assay is programmed to automatically dilute a single calibrator to create a 6-point
calibration curve with a measuring range of 0.33-7.09mg/L at the standard 1/10
sample dilution, with sensitivity of 0.33mg/L. High samples are automatically remeasured at the 1/20 sample dilution with a measuring range of 0.67-14.18mg/L with
a total assay time of 11 minutes. Within-run precision was assessed by running 5

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S103

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

samples at concentrations across the measuring range twenty times with a single kit
lot, on one analyser. The coefficients of variation for these 5 levels were as follows;
2.10% for the sample at 0.57mg/L, 1.21% at 1.22mg/L, 1.13% at 2.52mg/L, 1.08%
at 3.69mg/L and 2.36% at 6.14mg/L. Linearity was assessed using a high Cystatin C
sample which was serially diluted across the width of the curve at the 1/10 sample
dilution (range 0.24-8.4mg/L) and tested in triplicate at each dilution. Results
were compared to expected values and demonstrated good linearity when results
were analysed by linear regression; y=1.003x-0.0327, R2=0.9998. No significant
interference (within 6%) was observed when serum pools of known Cystatin C
concentrations were spiked with bilirubin (20mg/dL), haemoglobin (5g/L), Intralipid
(2000mg/dL) or triglyceride (1000mg/dL) and tested at the standard sample dilution.
Comparison was made to the Cystatin C assay for use on the Binding Site SPA PLUS
analyser by comparing 115 normal and clinical serum samples (range 0.53-8.54mg/
L). Good agreement was observed when the data was analysed by linear regression
analysis; y=1.0175x - 0.2179, R2=0.9917. We conclude that the Cystatin C assay for
the Binding Site next generation protein analyser is reliable, accurate and precise and
shows good agreement with existing assays.

A-355
Evaluation of a liposome-based CH50 assay for use on the Binding Site Next
Generation Protein Analyser.

J. R. Kerr, K. L. Sharp, A. Kaur, C. Rowe, S. J. Harding, P. J. Showell. The

Binding Site Ltd, Birmingham, United Kingdom
The complement cascade is made up of around 20 serum proteins that form part
of the innate immune system. A major function of complement is to lyse bacteria
through formation of the membrane attack complex (MAC). The complex interactions
of the complement cascade mean that functionality of the MAC cannot necessarily
be inferred by apparently normal levels of any single component. Measurement of
complement activity is therefore desirable, however existing methods involving
antibody-sensitised erythrocytes are limited by the need to create appropriate serum
dilutions and by the instability of the erythrocytes. These issues have been overcome
through the use of liposomes encapsulating glucose-6-phosphate dehydrogenase
(G6PDH) in place of erythrocytes. On addition of sample, antibodies in the reagent
combine with dinitrophenyl on the liposomes. The resultant complex activates
complement in the sample, which lyses the liposome, releasing G6PDH to react
with glucose-6-phosphate and NAD in the reagent. The change in absorbance can
be measured turbidimetrically and is proportional to the complement activity in the
sample. Complement activity has been correlated with the active stage of systemic
lupus erythematosus, rheumatoid arthritis, cryoblogulinemia-vasculitis, some forms
of nephritis, and inherited deficiencies of the complement system. Here we describe
evaluation of a CH50 assay for use on the Binding Sites next generation protein
analyser. The instrument is a random-access bench top turbidimetric analyser capable
of a wide range of on-board sample dilutions (up to 1/10,000) and throughput of
up to 120 tests per hour. Precision is promoted by single-use cuvettes which are
automatically loaded and disposed of. The analyser is programmed to construct a six
point calibration curve from a single, lyophilized serum-based calibrator. The standard
curves are validated by assay of control fluids. The assay range was 12.5 - 100 U/mL
using a 1/2 sample dilution. The assay showed a high degree of linearity when serially
diluted serum samples were assessed using a weighted linear regression analysis of
measured values against expected values (y = 1.01x + 0.26 r2= 0.998, range 9.987
- 102.230 U/mL). Precision was assessed at five levels across the measuring range.
Coefficients of variation for within run and total precision respectively were 20 U/
mL - 4.8% and 8.3%, 31 U/mL - 3.7% and 5.8%, 42 U/mL - 1.5% and 4.4%, 52 U/mL
- 1.4% and 3.5%, and 80 U/mL - 2.1% and 4.4%. No significant interference (within
10%) was observed upon addition of haemoglobin (500 mg/dL), bilirubin (20 mg/
dL), ascorbic acid (50 mg/dL) or Intralipid (250 mg/dL) to samples with known
CH50 values. Good agreement was observed with the Binding Site SPA PLUS CH50
assay when normal and deficient samples (n=28, range 23.091 - 66.143 U/mL) were
compared and analysed by Passing-Bablock regression: y=0.96x - 0.24. We conclude
that the CH50 assay for the Binding Site next generation protein analyser measures
complement activity rapidly, precisely and accurately and shows good agreement with
existing assays.

S104

A-356
Evaluation of the urine utility of a multipurpose albumin assay for use on the
Binding Site Next Generation Protein Analyser

H. Johnson, F. Murphy, S. J. Harding, P. J. Showell. The Binding Site Ltd,

Birmingham, United Kingdom
Albumin measurement in urine is routinely performed for the diagnosis of kidney
disease in conjunction with other laboratory and clinical findings. Here we describe
the evaluation of a multipurpose albumin assay for use on the Binding Sites next
generation protein analyser with respect to the measurement of urine. The instrument
is a random-access bench top turbidimetric analyser capable of a wide range of onboard sample dilutions (up to 1/10,000) and throughput of up to 120 tests per hour.
Precision is promoted by single-use cuvettes which are automatically loaded and
disposed of, whilst the utility is enhanced through host interface capability, primary
sample ID and bar coded reagent management systems. The assay is programmed to
automatically dilute a single calibrator to create a 6-point calibration curve with a
measuring range of 11-332mg/L at the standard 1/1 sample dilution, with sensitivity
of 11mg/L. High samples are automatically re-measured at the 1/50 sample dilution
with a measuring range of 550-16600mg/L and a total assay time of 10.5 minutes.
Precision studies (CLSI EP5-A2) were performed testing five urine concentrations in
duplicate over 21 working days. Each antigen concentration was assessed for total,
inter-lot and inter-instrument precision using three reagent lots on three analysers.
The coefficients of variation were 3.9%, 0.97% and 1.24% for the 23.0mg/L sample,
3.2%, 0.26% and 0.50% for the 39.0mg/L sample, 2.1%, 0.87% and 0.31% for the
143.4mg/L sample, 3.2%, 1.21% and 1.23% for the 275.1mg/L sample and 2.6%,
0.94% and 0.84% for the 1490.2mg/L sample. Linearity was assessed using a pool
of urine samples spiked with purified human albumin. The fluid was serially diluted
and results were compared to expected values. The assay showed good linearity when
observed results were compared to expected results and analysed by linear regression;
y=0.9933x-0.7688, R2=0.9995. No significant interference was observed (<3%)
when urine samples of known albumin concentration were spiked with bilirubin
(20mg/dL), haemoglobin (25mg/dL), ascorbic acid (20 mg/dL) or total protein
(100mg/dL). Comparison was made to the albumin urine assay for use on the Siemens
BNII analyser by comparing clinical urine samples (n=71, range 12.8-1334.8mg/
L). Good agreement was observed when the data was analysed by linear regression;
y=1.048x + 7.3231, R2=0.9959. We conclude that the urine albumin assay for the
Binding Site next generation protein analyser is reliable, accurate and precise and
shows good agreement with existing assays.

A-357
Patient sensitization profile by ImmunoCAP Solid Phase Allergen Chip (ISAC)
in a large Brazilian laboratory.

E. X. Gomes, A. Almeida, F. Prieto, D. Akiyama, E. Taira, N. Maluf, C.

Pereira. Dasa, Barueri, Brazil
Background: Most atopic patients have positive test results to numerous allergens and
the true cause of symptoms can be difficult to identify due to an inconclusive medical
history regarding the role of different allergens and reactions. ImmunoCAP ISAC
is a microarray technology developed by allergy specialists to allergy specialists for
investigation of multisensitized patients. It can also reveal unexpected sensitizations,
cross reactive components and risk assessment, providing physicians with relevant
information to handle allergen avoidance and individualized patient management.
Objective: The aim of this study was to evaluate the allergy sensitization to some
of the important molecular allergens and evaluate the incidence of Latex and Insect
Venom. Samples were collected at SP and RJ DASA laboratories from 05 Jun 2012
to 30 Jan 2014. All 66 patients were tested with a panel of 112 different allergens
components summing up 7,392 results with ImmunoCAP ISAC (Phadia AB, Thermo
Fisher).
Methods: ImmunoCAP ISAC is a semi-quantitative test and results are reported in
ISAC Standardized Units (ISU) giving indications of specific IgE antibody levels. The
allergen components are spotted in triplets and covalently immobilized to a polymer
coated slide. IgE antibodies from the patient sample bind to the immobilized allergen
components. The complex is detected by a fluorescence labeled anti-IgE antibody
which is measured with a laser scanner. Results:
Results for components were evaluated using Phadia Microarray Image Analysis
(MIA) software. 27,3% of the reports were found negative (66,7% women/ 33,3%
men) and 72,7% of the reports were found positive (56,3% women / 43,7% men).
Shrimp Tropomyosin (nPen m1) and Grass Pollen (nCyn d 1) had almost similar
positive frequency (16%) and Ovomucoid (nGal d 1) was 50% less compared to

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

the first two ones (8,33%). Taking patient as calculation basis, the higher incidence
was found for insect venom components (15.58%), followed by Latex (8,33%) and
Peanuts (2,03%)
Molecular Allergens: % positive
Shrimp Tropomyosin nPen m1 16
Ovomucoid nGal d 1 8,33
Grass Pollen nCyn d 1 16,67
Patients: % positive

against moesin N1-297 in the standards whereas an isotype-matched ITP serum did
not have any impact on the detection of this autoantibody. Similarly, C471-577 terminal
polypeptide in the presence of both low (0.5ug/ml) and high (2.5ug/ml) concentrations
of the blocking autoantibody against moesin C471-577 terminal portion. We propose that
autoantibodies against moesin N1-297 and C471-577 may be specific serum biomarkers for
clinical diagnosis and differentiation of ITP from non-immune thrombocytopenia and
other hematologic diseases.

A-359

Latex 8,33
Insect Venom 15,58
Peanuts 2,03
Conclusion: A high positivity of Tropomyosin allergens, which might indicate cross
reactivity between mites, cockroach and shrimp is important to handle shrimp allergy,
since these patients might also react to mites. Patients with sensitization to peanuts
components rAra h 1 rAra h2 are among the group with high risk to severe allergic
reaction to peanut. Ovomucoid is an allergen which indicates that egg symptoms might
persists after childhood as well as indicates severity. Positivity to pollen allergens
indicates that pollen allergy might be underestimated in our population. The results
to Latex and Insect venom show the importance of this toll, since other sensitizations
may be found in parallel providing more patient information.

A-358
Autoantibodies directed against moesin N1-297 /C471-577 are specific serum
biomarkers for immune thrombocytopenic purpura (ITP)

Y. Zhang1, H. Mao2, A. Zhao3, F. Du2, J. Qian1, J. Yang1, J. Wei3, X. Liu2, Z.

Sha2, Y. Han1, W. Situ1, C. Li1, L. Sheng1, Y. Xu2, Z. Shou2, S. Vardanyan4,
B. Bonavida5, J. R. Berenson4, H. Chen4. 1Shanghai Kexin Biotec Co.,
Ltd, Shanghai, China, 2The First Peoples Hospital of Guiyang, Guiyang,
China, 3Department of Obstetrics and Gynecology , Ren Ji Hospital
Affiliated to Shanghai JiaoTong University School of Medicine, Shanghai,
China, 4Institute for myeloma and bone cancer research, West Hollywood,
CA, 5Department of Microbiology, Immunology & Molecular Genetics,
David Geffen School of Medicine, UCLA, Los Angeles, CA
Moesin is a member of the ezrin/radixin/moesin (ERM) protein family, and is
localized in the cytoplasmic side/end of membrane in filopodia and other microextensions on the cell surface. Moesin consists of three domains: N-terminal (N)
membrane-binding domain, -helical (), and positive charge C-terminal domain.
In comparison to other cell types, human platelets demonstrate high expression of
moesin but not ezrin or radixin. Phosphorylation of threonine558 in carboxy-terminal
actin-binding domain of moesin is associated with the activation of platelets. We have
cloned three polypeptides of human moesin: N1-297 terminal, 298-470 helix domain
and C471-577 terminal as well as investigated autoantibodies against moesin in patients
with ITP. Following informed consent, serum samples from patients with ITP (n=77),
patients with non-immune thrombocytopenia or other hematologic diseases (n=47),
and gender-matched healthy control subjects (n=50) were evaluated. The titers of
moesin autoantibodies were significantly elevated in the sera from patients with ITP
compared with healthy subjects (Mean of autoantibodies titers = N1-297 0.515 vs 0.155;
P =0.0001). The levels of moesin autoantibody against C471-577 in ITP patients were
also markedly higher than healthy subjects (Mean of autoantibody titer = C471-577 0.430
vs 0.103, P < 0.0001). In contrast, the titer of autoantibody against moesin 298-470
helix domains was similar in ITP and healthy subjects. When an autoantibody cutoff value of 2D in normal control subjects (n=50) was assigned, the serum levels
of moesin autoantibodies in ITP patients were found to be elevated in 91% (70/77)
for N1-297, 72% (56/77) for C471-577 but only 1.3% (1/77) for 298-470 . Patients with
other hematologic diseases, including non-immune thrombocytopenia, anaphylactic
purpura, multiple myeloma, pure red cell aplasia, and myelodysplastic syndrome
were all negative for moesin autoantibodies. Regardless of the cause of ITP, the
autoantibodies against moesin N1-297 or C471-577 were significantly high than among
patients with non-immune-related thrombocytopenia and healthy subjects. We
used Western blot analysis to confirm the presence of moesin autoantibodies in ITP
patients. The results showed that only ITP patients serum specifically recognized the
moesin N1-297 and C471-577 terminal domains as well as commercial moesin antibodies.
In contrast, the antoantibodies from ITP patients serum did not recognize 298-470 helix
domains. We also confirmed using Western blot analysis that serum from patients with
thrombocytopenia with other hematologic diseases and healthy subjects did not show
the presence of autoantibodies against moesin N1-297 or C471-577 terminal. To further
confirm that the autoantibodies against moesin N1-297 or C471-577 terminal domains were
present in ITP patients, antigen competitive inhibition assay was accessed. This N1-297
terminal polypeptide (0.5ug/ml or 2.5ug/ml) blocked the detection of autoantibodies

Analytical evaluation of third-generation allergen-specific IgE assay

3gAllergy Measurement by Automated Immunoassay System
IMMULITE 2000 XPi

E. Hamada, M. Maekawa. Hamamatsu University School of Medicine,

Hamamatsu, Japan
Introduction: With the changes in the living environment including diet and air
pollution, allergic diseases have been expanded and diversified. Accordingly,
laboratory testing for any kinds of allergy has become important and important.
Recently, a new generation of allergy blood testing, 3gAllergy, coupling with
IMMULITE has been launched. The full automated random-access multiparameter
luminescence immunoassay system, IMMULITE 2000 XPi is based on a solid phase
two-site chemiluminescent enzyme immunoassay (CLEIA) and provides quantitative
determination of allergen-specific IgE. Here we evaluated analytical performance of
the IMMULITE 2000 XPi measurement system of 3gAllergy.
Specimens: We used serum and plasma samples collected from our inpatients/
outpatients and the employees who volunteered, as well as control samples
commercially available. This study has been approved by the ethical committee in
Hamamatsu University School of Medicine.
Methods: In this study, we evaluated the basic performance and compared
IMMULITE 2000 XPi Immunoassay System and its dedicated reagents (Siemens
Healthcare Diagnostics, USA) with the ImmunoCAP (Phadia AB, Sweden).
Results: 1. Dermatophagoides pteronyssinus (D1), House dust mites (H1) and
Japanese cedar (T17) are the major allergen in Japan. The 3gAllergy reagents showed
the following performance.
1) Within-run precision The CVs(%) for control samples measured 20 times with
7 cconcentrations in sequence were 1.7 to 5.98% (D1), 2.7 to 8.7% (H1) and 4.3 to
6.1% (T17).
2) Linearity Linearity observed in high concentration range and in low concentration
range indicated favorable results, respectively. However, a downward trend were
observed in the range of more than 250 IUA/mL.
3) Minimal detection limit By use of sequential dilution of the specimen with low
concentration, the detection limit were shown to be 0.02 IUA/mL.
2. Fourteen tests for Allergen (mite, cedar, alder, cypress, orchard grass, ragweed,
mugwort, dog, cat, egg white, ovomucoid, egg yolk, cows milk and peanuts)
1) Correlation We evaluated correlation with immuno-CAP for the 14 allergens. The
class concordance rates and measurement correlation of 14 tests were comfortable.
However, slightly higher tendencies in 3gAllergy are observed in the egg white,
ovomucoid and egg yolk.
Conclusion: The basic performance of IMMULITE 2000 XPi of 3gAllergy
and correlation with immunoCAP was satisfactory. The slight higher tendencies in
3gAllergy for egg white, ovomucoid and egg yolk were considered due to the higher
sensitivity, and then patients with low titer of allergen-specific IgE could be discovered
and followed. 3gAllergy is a third-generation allergen-specific IgE assay, delivering
fast and accurate results to help enhance the quality of care and service provided to the
patient. In addition, the device has some excellent properties, for example, offering
excellent precision, accuracy and lot-to-lot reproducibility, a wide range of test menu,
easy handling and maintenance, reducing workloads of manual sample sorting with
the innovative sample rack system and time for reporting. The properties could lead
to improved patient care.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S105

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

A-360
Droplet Digital PCR (ddPCR) for Quantitative Analysis of Treg-specific
Demethylation Region (TSDR) in peripheral blood compared to Flow
Cytometry (FACS)

J. Schmitz1, S. Gtze1, R. Andag1, J. Brockmller2, O. Kollmar3, E.

Schtz4, M. Oellerich1. 1University Medical Center Gttingen, Department
of Clinical Chemistry, Gttingen, Germany, 2University Medical Center
Gttingen, Department of Clinical Pharmacology, Gttingen, Germany,
3
University Medical Center Gttingen, Department of General, Visceral
and Pediatric Surgery, Gttingen, Germany, 4Chronix Biomedical,
Gttingen, Germany
Background: Regulatory T cells (Tregs) are crucial for immune homeostasis and
regulate several immune responses, especially tolerance induction. In transplanted
patients they may be useful for monitoring immune status, whereas the value for
tolerance prediction of peripheral Treg determination is still debated. Several methods
for Treg detection are in use, either based on analysis of FoxP3 protein expression by
FACS analysis or the determination of a specific de-methylated region in the FOXP3
gene (TSDR).
Methods: The TSDR of the FoxP3 gene de-methylation was determined with the
QX100 Droplet Digital PCR System (Biorad) using methylation specific assays.
DNA was extracted after CD4+ T-cell isolation with Dynabeads CD4, followed by
bisulfite conversion. We analyzed Treg numbers by FACS compared to the TSDR
ddPCR in 60 blood samples (31 from male, 29 from female) from 38 patients during the
first 13 months after liver transplantation (LTx). Peripheral blood mononuclear cells
(PBMC) were isolated with Lymphoprep (Axis-Shields) and frozen at -80C. After
thawing the cells were stabilized in RPMI 1640/FCS/Penicillin overnight at 37C,
and subsequently stained. Treg were identified as CD4+CD25+CD127lowFoxP3+/
CD4+ cells in a BD FACSCANTO II (BD Biosciences). For the CD25 staining an
antibody (clone M-A251) recognizing a different epitope than the therapeutically used
Basiliximab was chosen to eliminate interference.
Results: Compared to an earlier conventional qPCR method (LightCycler480), the
precision was substantially improved with the TSDR ddPCR (6.3% CV) between runs
at a level of 1.3% de-methylated TSDR compared to 30% with qPCR using the same
control sample. FACS analyses were based on a minimum count of 30,000 CD4+
cells and showed an imprecision of 6.1% (CV) at 2.6% Tregs. In LTx patients TSDR
ddPCR showed higher values (5.2%2.6) compared to the FACS assay (1.4%1.0).
Noteworthy, if FoxP3+/CD4+ cells were considered, resulting values were still lower
(1.9%1.5) compared to TSDR. No significant correlation was found between Treg
percentages determined by both methods (r=0.218, p=0.095).
Discussion: The de-methylated FoxP3-TSDR is highly specific for natural Treg cells. It
has already been proven that human non-regulatory T cells conserve their methylation
status after being activated (imprinting). Hence TSDR ddPCR provides a more robust
measure than the analysis of protein expression by the use of antibodies in FACS.
The precision of ddPCR is comparable to FACS analyses. Our new ddPCR assay is
able to measure all TSDR T-cells, even those, which are undetectable by FACS assay
such as latent Tregs, which have lost their phenotypic FoxP3 expression. Another
reason for the deviation and the lack of agreement might be due to the complexity
of the flow cytometric analyses in particular if intracellular staining is needed. In
addition in critically ill patients high background signals are common, but hard to
compensate for. This new ddPCR assay allows for specific Treg quantification with
adequate precision, good reproducibility, fast turn around, low costs and only requires
20% of blood compared to FACS. Further studies will show if the exact ratio of Treg/
CD4 cells measured by ddPCR is useful as a biomarker for tolerance assessment in
transplanted patients.

A-361
The Association of Serum Free Light Chain Levels with Markers of Renal
Function

nephelometer (Siemens Diagnostics, Germany). / sFLC ratio was calculated. Serum

creatinine levels were analysed by modified Jaffe method in Cobas 8000 analyser.
GFR was estimated by the CKD-EPI (Chronic Kidney Disease Epidemiology
Collaboration) equation. Patients were assigned to 2 groups depending on their eGFR
values: 60 mL/min/1.73m2 (group 1, n=103) and >60 mL/min/1.73m2 (group 2,
n=410). Data were expressed as median and min.-max. All statistical analyses were
done with SPSS version 20.0 and a significance level of 0.05 was considered.
RESULTS: Serum kappa FLC median value was 36.4 (5.62-16000) mg/L, serum
lambda FLC was 21.7 (4.91-8770) mg/L, / sFLC ratio was 1.33 (0.01-3258), serum
creatinin was 1.56 (0.63-7.21) mg/dL in group 1. Both of Lambda sFLC and / sFLC
ratio were correlated with eGFR (r=-0.318, r=0.198, p<0.05, respectively). We did not
find any significant correlation between / sFLC ratio and eGFR in group 2.
CONCLUSIONS: We examined the association between polyclonal sFLC
concentrations and renal function. Our preliminary findings suggest that serum
Lambda FLC might be considered useful marker of predicting renal function.
Prospective studies are needed to clarify the usefulness of these parameters for
identifying renal failure due to a monoclonal gammopathy.

A-362
Evaluation of a Caeruloplasmin assay for use on the Binding Site Next
Generation Protein Analyser

J. K. Robinson, F. Murphy, A. Kaur, S. J. Harding, P. J. Showell. The

Binding Site Ltd, Birmingham, United Kingdom
Caeruloplasmin is synthesised in the liver and has a major role in copper metabolism,
carrying approximately 95% of the total copper in serum. Decreased levels of
Caeruloplasmin can be caused by hereditary disorders of copper metabolism, for
example; inability to transport oxidised copper (Cu2+) from the gastrointestinal
epithelium into the circulation (as in Menkes disease), or the inability to insert Cu2+
into the developing Caeruloplasmin molecule (as in Wilsons disease). Dietary
copper insufficiency, including malabsorption, also reduces serum Caeruloplasmin
concentrations. Here we describe the evaluation of a serum Caeruloplasmin assay
for use on the Binding Sites next generation protein analyser. The instrument is a
random-access bench top turbidimetric analyser capable of a wide range of onboard sample dilutions (up to 1/10,000) and throughput of up to 120 tests per hour.
Precision is enhanced by single-use cuvettes which are automatically loaded and
disposed of, whilst the utility is enhanced through host interface capability, primary
sample ID and bar coded reagent management systems.The instrument automatically
dilutes a single calibrator to produce a calibration curve with a measuring range of
0.03-0.82g/L at the standard 1/10 sample dilution, with sensitivity of 0.03g/L. High
samples are automatically re-measured at a dilution of 1/20, with an upper measuring
range of 0.06-1.64g/L. Precision studies (CLSI EP5-A2) were performed at five levels
in duplicate over 21 working days. Five antigen levels were assessed in duplicate,
twice daily for total, within-run, between-run and between-day precision, using one
lot of reagent on three analysers. The coefficients of variation were 9.4%, 1.5%, 5.5%,
7.4% and 8.8% for the 0.061g/L level, 7.6%, 1.6%, 3.4%, 6.6% and 6.2% for the
0.156g/L level, 5.6%, 1.7%, 2.3%, 4.8% and 2.9% for the 0.247g/L level, 5.7%, 1.7%,
2.1%, 5.0 and 3.7% for the 0.442g/L level and 6.4%, 2.0%, 1.8%, 5.9% and 3.7%
for the 0.867g/L level respectively. Linearity was assessed by assaying a seriallydiluted patient sample pool across the width of the measuring range (0.03 - 0.82g/L)
and comparing expected versus observed results. The assay showed a high degree
of linearity when expected values were regressed against measured values (y=0.96x
+ 0.00, R = 0.999). No significant interference (within 10%) was observed on
addition of bilirubin (20mg/dL), haemoglobin (500mg/dL) or chyle (1250 formazine
turbidity units) when spiked into a sample with known Caeruloplasmin concentrations
and measured using the minimum sample dilution. The assay was compared to the
Caeruloplasmin assay for the Binding Site SPA PLUS analyser by running clinical
samples (n=36). Good agreement was demonstrated by Passing-Bablok regression;
y = 1.00x + 0.01. We conclude that the Caeruloplasmin assay for the Binding Site
next next generation protein analyser is reliable, accurate and precise and shows good
agreement with existing assays.

H. Akbas, B. Karatoy Erdem, F. Davran, V. T. Yilmaz. Akdeniz University,

Faculty of Medicine, Antalya, Turkey
BACKGROUND: The kidney is often affected in plasma cell dyscrasias, usually
due to the effects of nephrotoxic monoclonal free light chains. Renal failure due to
a monoclonal gammopathy may be detected by the highly sensitive serum free light
chain (sFLC) ratio yet missed by electrophoretic assays. The aim of this study was to
assess sFLC levels in relation to markers of renal function.
METHODS: 513 patients were included in this study. sFLC levels were measured
by Freelite (The Binding Site Group Ltd, Birmingham, UK) assay using the BNII

S106

A-363
Identification of an IgD biclonal gammopathy

K. Lockington, L. L. Eckelkamp, J. A. Katzmann. Mayo Clinic, Rochester,

MN
Background: Monoclonal abnormalities are confirmed by immunofixation
electrophoresis (IFE) which identifies the immunoglobulin heavy chain and/

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

or light chain type. Routine IFE includes antisera to gamma, alpha and mu heavy
chain and kappa and lambda light chain. When a free light chain is detected without
a corresponding heavy chain, IFE is performed with antisera to delta and epsilon
heavy chains to rule out a monoclonal IgD or IgE. We do not routinely reflex to
anti-delta and epsilon IFE when we detect a free light chain in the presence of an
intact immunoglobulin with the same light chain type. We recently initially reported a
monoclonal IgG lambda plus monoclonal free lambda that eventually was confirmed
as a biclonal IgG lambda and IgD lambda. Because the presence of a monoclonal IgD
protein is often associated with either multiple myeloma or primary amyloid, this IgD/
IgG biclonal gammopathy triggered an evaluation of our anti-delta and epsilon reflex
process for free light chains.
Methods: Protein electrophoresis is performed on Helena agarose gels and IFE is
performed with Sebia reagent kits. IFE reflex testing uses BioWhittaker antiserum for
IgD and Binding Site antiserum for IgE.
Results: During a 1 month period we detected 1245 patients with a serum
monoclonal protein. Twenty-eight of the samples (2.2%) were also tested with IgD
and IgE antisera. Of these 28 patients, 19 had a monoclonal free light chain and the
remaining 9 were eventually reported as having a biclonal gammopathy of 2 intact
immunoglobulins with differing light chain type. In addition, there were 26 patients
with an intact monoclonal immunoglobulin plus a monoclonal free light chain of the
same type as the intact immunoglobulin. As per the laboratory protocol, these samples
were not reflexed to the expanded IFE.
Conclusion: Our protocol to reflex monoclonal free light chains to anti-delta and
epsilon IFE does not include free light chains that are associated with an intact
monoclonal immunoglobulin with the same light chain. This protocol resulted in
2.2% of positive IFE tests being reflexed. If all new monoclonal free light chains are
reflexed (regardless of the presence of an intact immunoglobulin with the same light
chain), our reflexed IFE testing would approximately double.

A-364
Clinical Evaluation of the New BioPlex Celiac IgA and IgG Kits

M. Snyder1, J. A. Murray1, I. Absah1, S. Paul2, J. Sama2, X. Guo2. 1Mayo

Clinic, Rochester, MN, 2Bio-Rad Laboratories, Hercules, CA
Background: Serologic testing, specifically for tissue transglutaminase (TTG) and
deamidated gliadin peptide (DGP) antibodies, is increasingly being relied upon
to establish a diagnosis in patients with suspected celiac disease (CD). Therefore,
diagnostic accuracy of CD-specific autoantibodies is critically important. Although
monitoring patients with CD who have instituted a gluten-free diet (GFD) is routine
practice, how responses compare between various serologic markers is not always
clear. This study compares performance of bead-based multiplex immunoassays
(MIAs) for TTG and DGP antibodies to that of currently available individual enzyme
immunoassays (EIAs) for diagnosis and monitoring.
Methods: Retrospective samples were obtained from healthy controls (n=210),
inflammatory disease controls (n=101), and patients with CD and related diseases
(total n=105), including adult patients with biopsy-proven CD (n=96; 7 with
IgA deficiency), pediatric patients with CD (n=6), and patients with dermatitis
herpetiformis (DH) (n=3). An additional retrospective CD cohort was included
(n=10), consisting of paired samples pre- and post-GFD (time on GFD 1.6 to 52.6
months). MIA testing was performed on BioPlex 2200 Celiac IgA and IgG kits (BioRad); EIA testing was performed on QuantaLite R h-tTG and Gliadin IgG II IgA and
IgG kits (INOVA Diagnostics). All testing was performed according to manufacturers
instructions.
Results: The specificity of TTG-IgA, TTG-IgG, DGP-IgA, and DGP-IgG by MIA
in healthy donors ranged between 96.2% and 99.5%. These were not significantly
different from the EIA specificities (97.6% to 99.5%). For the disease control cohort,
TTG-IgA and DGP-IgA showed similar specificities between MIA (97% and 100%)
and EIAs (100% for both). However, the EIAs had lower specificities for TTG-IgG
and DGP-IgG at 92% and 96%, compared to 100% and 99% for the MIA. In the CD/
DH cohort, excluding IgA deficient patients, MIAs and EIAs demonstrated identical
sensitivities at 92% for TTG-IgA and 86% for DGP-IgA. For TTG-IgG and DGP-IgG,
the total CD/DH cohort was analyzed, including IgA deficient patients. The sensitivity
of DGP-IgG was 83% and 79% for MIA and EIA; for TTG-IgG, overall sensitivities
were 41% and 48% for MIA and EIA. If only IgA deficient patients were analyzed, a
sensitivity of 57% was obtained for DGP-IgG by both methods. However, TTG-IgG
by MIA had a higher sensitivity than EIA (71% vs 29%). Among patients on a GFD,
for individuals positive for DGP-IgA at baseline, 25% remained positive by both MIA
and EIA. For those patients positive for TTG-IgA, 11% and 14% remained positive
on MIA and EIA, respectively. In contrast, 62.5% and 71.4% of patients positive for
DGP-IgG remained positive by MIA and EIA.

Conclusion: The diagnostic accuracy of autoantibody serology for diagnosis of CD

appears comparable between MIA and EIA methods. The only significant differences
identified between the methods were for TTG-IgG and DGP-IgG. These results also
confirm that TTG-IgA displays the best combination of sensitivity and specificity for
CD, as demonstrated by previous studies. Further, the IgA isotypes appear to have
the most utility for monitoring, based on conversion to a negative serology following
implementation of a GFD.

A-365
Diagnostic utility of autoantibodies and HLA-DRB1 Shared Epitope in patients
with recent onset Rheumatoid Arthritis

J. L. Garca de Veas Silva, B. Hernndez Cruz, C. Gonzlez Rodriguez,

V. Snchez Margalet, F. Navarro Sarabia. Hospital Universitario Virgen
Macarena, Sevilla, Spain
Background: Rheumatoid arthritis (RA) is a systemic inflammatory autoimmune
disease characterized by chronic polyarthritis. Anti-cyclic citrullinated peptide (antiCCP) antibodies have diagnostic value in RA but the role of shared epitope (SE) is
unclear. We assessed the diagnostic value of anti-CCP antibodies and SE in patients
with symptoms of arthritis in their first visit to the rheumatologist.
Methods: We measured anti-CCP antibodies with QUANTA LiteTM enzyme-linked
inmmunosorbent assay (ELISA) kit for the detection of IgG anti-CCP3 (Cyclic
Citrullinated Peptide 3) antibodies in patient sera (cut-off value, 40 UI/ml). SE was
determined with GenID Reverse Hybridization kit for detection of SE in HLADRB1 alleles kit in patient plasma. They were tested for 211 patients with suspected
rheumatoid arthritis. The American College of Rheumatology (ACR) criteria for RA
were fulfilled for 106 patients. These patients were diagnosed of RA. The other 105
patients were diagnosed with other rheumatic disease. We also determined rheumatoid
factor (RF) with BAYER FR IgM immunoturbidimetric assay for ADVIA 2400 (cutoff value, 20 UI/ml). We determined the diagnostic value (sensibility, specificity and
likelihood ratios) for anti-CCP antibodies, SE and RF. We determined the area under
the curve (AUC) for anti-CCP antibodies and RF. Statistical analyses were performed
using IBM SPSS Statistics version 19 for Windows (New York, USA).
Results: Sensitivity of anti-CCP antibodies, SE and RF for RA were 66,0%, 72,6%
and 81,1%, and specificity were 96,2%, 38,1% and 76,2%, respectively. The AUC
for anti-CCP antibodies was 0.875 with 95% CI of 0.828 to 0.922 and the AUC for
RF was 0.864 with 95% CI of 0.815 to 0.913. The positive likelihood ratio for antiCCP antibodies, SE and RF were 17.37, 1.06 and 3.41 respectively. The negative
likelihood ratio for anti-CCP antibodies, SE and RF were 0.35, 0.72 and 0.25. AntiCCP antibodies were positive in 30.0% of RF-negative RA patients. Anti-CCP
antibodies and RF were positive in 60.4% of total RA patients and were negative in
13.2% of total RA patients.
Conclusion: the anti-CCP antibodies have a very good diagnostic value in their first
visit to the rheumatologist. However, the SE doesnt have diagnostic value in our
population.

A-366
4-Phenyl butyric acid attenuate apoptosis via inhibition of endoplasmic
reticulum stress against diabetic cardiomyopathy

J. Xu1, M. Cui2, S. Shi1, Q. Zhou1. 1First Hospital of Jilin University,

Changchun, China, 2Clinical Medical College of Jilin University,
Changchun, China
Background: Diabetic cardiomyopathy (DCM) is a major cause of death of patients
with diabetes. It is known that apoptosis has been considered to play a critical role
in DCM. Our recent studies have demonstrated the important role of endoplasmic
reticulum stress (ER stress) in diabetes-induced cardiac cell death. The aim of this
study was to investigate cardiac protection by 4-Phenyl butyric acid (PBA), a low
molecular weight compound that acts as a chemical chaperone to enhance protein
folding and ameliorate ER stress, in the development of DCM.
Methods: At 2 weeks, and 2 and 5 months after diabetes onset with Type 1 diabetic
mouse model induced with multiple low-dose streptozotocin, cardiac remodeling and
dysfunction were determined using echocardiography and hemodynamic evaluation
and cardiac fibrosis was detected by Picric acid-Sirius red staining; ER stress
signal pathway and apoptosis were detected by western blotting assay and terminal
deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay.
Mechanism of cardiac protection by PBA was used by cell culture with embryonic rat
heart derived cells (H9c2).

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S107

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

Results: Apoptotic cells and CHOP, the activated form of caspase-3 and caspase-12
delineated that diabetes mainly induced cardiac cell death at the early stage of diabetes
(2 weeks), but not in the late stages (2 and 5 months). However, there was no apoptotic
cell death in the hearts of diabetic mice treated with PBA. In parallel with apoptotic
effect, significant up-regulation of the ER chaperones, including phosphorylated eIF2
(p-eIF2), GRP78, GRP94 and cleaved ATF6 proteins were significantly increased in
the heart of diabetic mice at 2 weeks after diabetes onset. However, none of these
increased ER chaperones in the hearts of diabetic mice were observed in the heart
of diabetic mice treated with PBA. Pre-exposure of H9c2 cells to PBA significantly
prevented high glucose or tunicamycin-induced ER stress and apoptosis, while same
pretreatments did not have any effect on normal H9c2 cells.
Conclusion: These results suggest that ER stress exists in the diabetic heart, which
may cause the cardiac cell death. PBA can attenuate diabetes-induced cardiac cell
death via suppression of cardiac ER stress and associated apoptotic effects. This study
could provide important theoretical support for prevention and treatment of DCM.
The study also might open up a new way for the drugs in research and development
for diabetic cardiovascular complications.
Acknowledgments: These studies were supported in part by Grants from National
Science Foundation of China (no. 81000310), Jilin Science and Technology
Development Program (no. 20100124), Norman Bethune Program of Jilin University
(no. 2012223).

A-367
Vitamin D Receptor Polymorphisms and HLA-Class II Genotypes Among
Lebanese with Multiple Sclerosis - A Pilot Study

R. T. Daher, R. A. Mahfouz, N. M. Karaky, F. A. Jaber, B. I. Yamout.

American University of Beirut, Beirut, Lebanon
Background: Multiple sclerosis (MS) is an autoimmune disease with multifactorial
etiology. Previous studies showed that HLA-DRB1*15 allele is a major genetic risk
factor for MS in other populations possibly through regulation of vitamin D receptor
(VDR) complex. In this study, we investigated the HLA class II genotypes and VDR
gene polymorphism among a group of Lebanese MS patients and controls.
Methods: Fifty MS patients (remitting/relapsing, aged: 19-74 years,
male:female=1:2.1) were selected for this study, based on the Expanded Disability
Status Scale. The controls included: 49 healthy subjects (aged: 15-59 years,
male:female=1:2) and 51 neurologic patients other than MS (Non-MS, aged 1370 years, male:female=1:1.12). After a thorough history, blood in EDTA tube was
collected. Extracted genomic DNA was used for molecular analysis of VDR genotypes
(ApaI, TaqI and BsmI) and HLA class II typing (low resolution HLA-DRB1/3/4/5)
(Luminex, San Diego, CA). Differences between groups were evaluated using Mann
Whitney-U test. Chi-square test was used for association between various categorical
variables. (P<0.05 statistically significant*)
Results: All determined variables were not statistically different between healthy and
non-MS patients (p>0.05); therefore both were combined into one control group for
analysis. Results are summarized below.
HLA-D
Age(years)
RB1*
VDR Gene
(MeanSD)
15(%)
ApaI
AA(%) Aa(%) aa(%)
MS (n=50) 42.813.5 11(22)
15(30) 26(52) 9(18)
Control
33.712.7 8(8)
37(37) 48(49) 14(14)
(n=99)
P value
<0.001*
0.017* >0.05 >0.05 >0.05

A-368
Frequency of antinuclear antibodies (ANA) by indirect immunofluorescence in
Brazilian samples

F. B. Flumian, F. O. Paiva, D. C. B. Carvalho, M. A. Rigone, T. N. Carvalho,

L. R. Galhardo, C. R. Costa, L. B. Campanholi, E. A. Garcia, O. Fernandes,
N. Gaburo Jr.. DASA, Sao Paulo, Brazil
Background: Antinuclear antibodies (ANAs) are considered a hallmark of
autoimmune rheumatic diseases (ARDs), and the indirect immunofluorescence assay
(IFA) on HEp-2 cells is the standard method for ANA detection. During the past
decade, as the demand for ANA testing increased, new automated methods have arisen
for screening/detection of ANAs.Objective: To describe the frequency of positive
ANAs and the most common patterns found in Brazilian samples, following the IV
Brazilian Consensus on FanHep2 in 2013.
Methods: During the 2013 year, 192.593 serum samples were screened to ANA
detection by the Immunology department of DASA, using IFA technique. In the first
semester, a semi-automated technique was used, with slides preparation performed
by Quanta-Lyser 240 equipment (Inova Diagnostics, San Diego, CA) and manual
reading. In the second semester, the Integrated Laboratory (Inova Diagnostics)
system was implemented. This system is composed by Nova View equipment and
Quanta-link software. It allows the performance of full automated process, from slides
preparation to reading. During both phases, the initial dilution of samples was 1/160.
Results: We found that 73% of the analyzed samples showed negative result and 27%
showed positive result for ANA. The most frequent pattern found among positive
samples was Speckled Nuclear Fine Dense - PSNFD, followed by Speckled Nuclear
Fine, as described at Table 1. Some patterns were found in less than 1% of the positive
samples as NUMA 1 (Pattern Nuclear Fine Speckled with Mitotic Apparatus), Fine
Dense Speckeld Cytoplasmitic, Cytoplasmatic with Isolated Dots, NUMA2 (Mitotic
Apparatus Type Mitotic Fuse), Nuclear Type Nuclearmembranous, Polar Speckled
Cytoplasmatic, Mitotic Apparatus type Intercellular Bridge, Cytoplasmic Fibrilar,
Fine Speckled Cytoplasmatic, Mitotic Apparatus type Centriole and Pleomorphic
Nuclear Speckled (PCNA).
Conclusion: Our findings demonstrated that 35.43% of the positive samples presented
PSNFD pattern, which according to previous reports are rarely found in ARDs.
Table 1. ANAs patterns detected. Legend: *Other patterns: frequency less than 1%.
Pattern
Number of samples
%
Nuclear Fine Dense Speckled (PSNDF)
18609
35.43
Nuclear Fine Speckled
13832
26.33
Nuclear Quasi-Homogeneous Speckled
5179
9.86
Nuclear Homogeneous
2752
5.24
Nucleolar
2423
4.61
Nuclear Coarse Speckled
2419
4.61
Nuclear Centromere
2121
4.04
Nuclear Coarse Speckled Reticuladet
1401
2.67
Reticular Specled Cytoplasmatic
1364
2.60
Nuclear dots
588
1.12
*Other patterns
1838
3.51
Overall
52526
100

A-369

Groups

TaqI
TT(%) Tt(%) tt(%)
20(40) 24(48) 6(12)

BmsI
BB(%) Bb(%) bb(%)
7(14) 22(44) 21(42)

31(31) 47(48) 21(21) 20(20) 51(52) 28(28)

>0.05

>0.05

>0.05

>0.05

>0.05

>0.05

Frequency of HLA-DRB1*15 was significantly higher in MS patients compared to

controls. None of the VDR genes differed between the two groups. Odds ratio (OR)
for MS in the presence of DRB1*15 allele is 3.21 (p= 0.016; 95%CI= 1.20-8.59).
Cosegregation of HLA-DRB1*15 and VDR genotypes indicated a slightly increased
risk for MS in the presence of A-allele (OR=3.40; p= 0.022; 95%CI= 1.14-10.19).
Similarly, combination of DRB1*15 with b-allele resulted in even higher OR of 4.22
although not statistically significant (p= 0.08; 95%CI= 0.75-23.89).
Conclusion: Our results confirm that HLA-DRB1*15 is a strong predisposing factor
for MS in Lebanese patients. Furthermore, the interaction between specific VDR
alleles and HLA polymorphism may synergistically influence the susceptibility to MS.

S108

THE SERUM LEVELS OF DIKKOPF-1 (DKK-1) IN AXIAL

SPNDYLOARTHRITIS (AXSPA) ARE RELATED TO DISEASE DURATION

E. Melguizo1, V. NAVARRO-COMPN2, C. GONZALEZ-RODRIGUEZ1,

F. NAVARRO-SARABIA1, R. ARIZA-ARIZA1. 1Virgen Macarena
University Hospital, Sevilla, Spain, 2LA PAZ University Hospital,
MADRID, Spain
Background: Tumor necrosis factor (TNF) alpha is responsible for induction of dkk-1
which down-regulates bone formation. Therefore, it was expected that TNF-blocker
therapy would inhibit radiographic progression in patients with axSpA but this effect
has not been observed yet. Nevertheless, most of the studies have included patients
with long disease duration and it is unknown whether or not this effect would be the
same in patients with an early stage of the disease.
Objectives: To investigate if disease duration influences on the serum levels of
dkk-1 in patients with axSpA. Methods: Observational study including consecutive
patients with axSpA according to ASAS criteria visiting a tertiary hospital between
January 2011 and June 2013. All patients were receiving NSAIDs and none of them
was under biologic therapy. The following characteristics were recorded at one visit:
Demographic (age, gender), symptoms duration, HLA-B27, disease activity indices
(BASDAI, CRP, ESR) and function (BASFI). Blood samples to determine dkk-1
serum levels by enzyme immunoassay were collected at the same visit too. Patients
were classified as early axSpA (symptoms duration 5 years) and established axSpA

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

(>5 years) and the characteristics enumerated above were compared between both
groups. Univariate and multivariate linear regression models were employed to
identify the characteristics related to dkk-1 serum levels.
Results: Thirty one patients with early axSpA and 21 patients with established disease
were included. Patients with early axSpA were younger (32.6 9.3 vs 41.0 10.2
years; p<0.01), had lower degree of disease activity (BASDAI: 4.6 2.7 vs 6.6 1.9;
p<0.01 and ESR: 7.7 9.2 vs 18.1 15 mmHg; p<0.05) and worst function (3.2 2.9
vs 5.8 2.5; p<0.01) compared with patients with established disease. Serum levels
of dkk-1 were significantly higher in patients with early disease (25.9 11.5 vs 13.9
13.5; p<0.001 ng/dl). No statistically significant differences were found between both
groups for the rest of characteristics. In the univariable analysis, symptoms duration
and BASDAI were inversely related to dkk-1 levels (std : -0.435; p<0.01 and Std :
-0.283; p<0.05, respectively). However, only the relationship with symptoms duration
remained statistically significant in the multivariable analysis (std : -0.415; p<0.01).
Conclusions: Serum Dkk-1 levels in patients with axSpA depend on disease duration,
being higher in patients with recent onset of the disease. The effect of TNF-blocker
therapy on radiographic progression may be different in patients with an early stage of
the disease compared with patients with established disease.

A-370

were serially diluted across the width of the standard curve and results were compared
to expected values. No significant interference (3.03%) was observed when CSF
and serum samples of known albumin concentration were spiked with bilirubin
(20mg/dL), haemoglobin (500mg/dL) or Chyle (1500 FTUs). Comparisons were
made to both the serum albumin and CSF albumin assays for use on the Siemens
BNII analyser by comparing clinical samples. The main assay characteristics are
summarised in the table below:
Assay
Initial sample dilution
Initial range
Maximum sample dilution
Maximum range
Sensitivity
Assay time (mins)
Total precision (concentration)(C.V)

Inter-kit precision
(concentration)(C.V)

Inter-instrument precision
(concentration)(C.V)

Impact of genetic (HLA-DRB1 Shared Epitope) and environmental (Smoking)

factors in the presence of anti-CCP antibodies in Rheumatoid Arthritis

J. L. Garca de Veas Silva, B. Hernndez Cruz, C. Gonzlez Rodriguez,

V. Snchez Margalet, F. Navarro Sarabia. Hospital Universitario Virgen
Macarena, Sevilla, Spain
Background: Reumatoid Arthritis (RA) is a multifactorial autoimmune disease
where environmental and genetic factors interact in the etiology of the disease and
development of anti-cyclic citrullinated peptide antibody. The aim of this study is
to investigate whether HLA-DRB1 shared epitope (SE), tobacco exposure (TE)
and smoking dose (SD) are associated with the presence of anti-cyclic citrullinated
peptide (anti-CCP) antibodies in Spanish patients with RA.
Methods: A cohort of 106 patients with early diagnosed RA was studied. Anti-CCP
antibodies and rheumatoid factor (RF) were measured at diagnosis and HLA-DRB1
genotyping was performed for SE. TE was categorized as never or ever. Smoking
dose (SD) was categorized in pack-years with a cut-off of 20 pack-years. Contingency
tables and models of logistic regression were used to calculate the association between
SE and smoking with the presence of anti-CCP.
Results: In univariate analysis, SE (OR=2.68; 95% CI 1.11 to 6.46), TE (OR=2.79;
95% CI 1.12 to 6.97), SD (OR=6.04; 95% IC 1.68 to 21.74) and the presence of
RF (OR=8.73; 95% IC 2.84 to 26.80) were associated with the presence of antiCCP antibodies. In logistic regression analysis, only SE-TE interaction (OR=7.083;
95% IC 1.01 to 49.50) and presence of RF (OR=3.07; 95% IC 1.26 to 7.49) were
independently associated with the presence of anti-CCP antibodies.
Conclusion: SE-TE interaction and the presence of RF were significantly and strongly
associated with the presence of anti-CCP antibodies.

Serum
1/200
2200-66400mg/L
1/200
2200-66400mg/L
11mg/L
10.5
(3.7g/L) (2.9%)
(13.0g/L) (3.0%)
(28.5g/L) (2.88%)
(37.0g/L) (2.6%)
(54.4g/L) (3.3%)

CSF
1/1
11-332mg/L
1/10
110-3320mg/L
11mg/L
10.5
(145.5mg/L) (5.92%)
(281.5mg/L) (5.37%)
(439.9mg/L) (3.62%)
(593.1mg/L) (3.52%)
(975.2mg/L) (8.08%)

(3.7g/L) (0.6%)

(145.5mg/L) (1.15%)

(13.0g/L) (0.4%)
(28.5g/L) (1.0%)
(37.0g/L) (0.6%)
(54.4g/L) (0.6%)

(281.5mg/L) (2.89%)
(439.9mg/L) (1.36%)
(593.1mg/L) (1.88%)
(975.2mg/L) (1.87%)

(3.7g/L) (1.1%)

(145.7mg/L) (0.77%)

(13.0g/L) (0.5%)
(28.5g/L) (0.5%)
(37.0g/L) (0.5%)
(54.4g/L) (1.4%)
106
19,868-55,847 mg/L

(282.3mg/L) (1.59%)
(441.8mg/L) (2.30%)
(594.6mg/L) (1.30%)
(977.8mg/L) (1.41%)
62
33.9-961.9 mg/L

Comparison Sample no.

Range
Passing and Bablock
Y=1.01x + 1010.77 Y= 1.08x +14.88
analysis
R2
0.9541
0.9782
Linearity
Linear regression
y=0.9978x +589.88 y=1.0x-4.72
R2
0.9996
0.9995
We conclude that the low level albumin assay for the Binding Site next generation
protein analyser is reliable, accurate and precise and shows good agreement with
existing albumin assays.

A-372
Multiplex Assay of Circulating Inflammatory Biomarkers in Patients with
Stroke

J. Polivka1, J. Polivka2, V. Rohan2, P. Sadilkova3, O. Topolcan4. 1Department

of Histology and Embryology and Biomedical Centre, Faculty of Medicine
in Plzen, Charles University in Prague, Plzen, Czech Republic, 2Department
of Neurology, Faculty of Medicine in Plzen, Charles University in Prague
and Faculty Hospital Plzen, Plzen, Czech Republic, 3Department of
Histology and Embryology, Faculty of Medicine in Plzen, Charles
University in Prague, Plzen, Czech Republic, 4Central Imunoanalytical
Laboratory, Faculty Hospital Plzen, Plzen, Czech Republic
Background: Neuroinflammation is involved in the pathophysiological mechanisms
of stroke. However, the role of inflammatory blood biomarkers in relation to the
clinically relevant information remains unclear. This study determines the association
of various circulating inflammatory biomarkers with stroke severity, classification
into different stroke subtypes and patient 3-month outcome.

A-371
Evaluation of the CSF and serum utilities of a multipurpose albumin assay for
use on the Binding Site Next Generation Protein Analyser

H. Johnson, F. Murphy, S. J. Harding, P. J. Showell. The Binding Site Ltd,

Birmingham, United Kingdom
The integrity of the blood-brain barrier may be compromised in certain neurological
conditions. Measurement of albumin in CSF and serum, and calculation of a ratio,
is widely used as a diagnostic tool in these circumstances. Here we describe the
evaluation of the serum and CSF utilities of the low level albumin assay for the
Binding Sites next generation protein analyser. The instrument is a random-access
bench top turbidimetric analyser capable of a wide range of on-board sample dilutions
(up to 1/10,000) and throughput of up to 120 tests per hour. Precision is promoted
by single-use cuvettes which are automatically loaded and disposed of, whilst the
utility is enhanced through host interface capability, primary sample ID and bar coded
reagent management systems. The assay is programmed to create a 6-point calibration
curve from a single serum based calibrator. Total, inter-lot and inter-instrument
precision was assessed for both sample types by running serum and CSF pools at
five concentrations across the respective curves. Each sample was run in duplicate,
twice daily for 21 days across three analysers. Linearity was assessed using pools
of normal serum and CSF samples spiked with purified human albumin. Both fluids

Methods: 215 stroke patients (Large artery (LAA), n = 93; Cardioembolic (CE),
n = 47; Lacunar (LAC), n= 33; Cryptogenic (CR), n = 7 and Hemorrhagic stroke
(HS), n = 35) were prospectively evaluated in the Faculty Hospital Plzen between
2012 and 2013. Stroke severity (National Institutes of Health Stroke Scale; NIHSS)
was measured at hospital admission. Functional outcome (modified Rankin Scale;
mRS) was assessed after 3 months. A multiplex panel of 14 biomarkers (IL1, IL6,
IL10, IL12, MCP-1, OPG, OPN, VEGF, MMP1, MMP2, MMP7, MMP9, 25-OHVitamin D, 1,25-OH-Vitamin D) was assessed in plasma samples (collected within
4 hours from symptom onset) by Luminex xMAP technology. The associations
of circulating inflammatory biomarkers with the stroke severity, classification into
different stroke subtypes and patient outcome were evaluated by Spearmans rank
correlations and Wilcoxon test.
Results: Positive correlations with stroke severity (NIHSS at baseline) were found for
IL6 (r = 0.15, P = 0.02), IL10 (r = 0.16, P = 0.014) and MMP9 (r = 0.14, P = 0.03).
The worse 3-month outcome (mRS) was correlated with IL6 (r = 0.14, P = 0.036) and
blood leukocytes (r = 0.14, P = 0.038). Higher plasma levels of IL6 (P = 0.02), IL10
(P = 0.006) and MMP9 (P = 0.029) and startlingly lower cholesterol (P = 0.029) were
found in patients with more severe stroke at baseline (NIHSS > 10). Patients with

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S109

Immunology

Tuesday, July 29, 9:30 am 5:00 pm

worse 3-month outcome (mRS 3) had higher plasma level of MMP9 (P = 0.04),
glucose (P = 0.025), higher NIHSS at baseline (P < 0.0001), lower cholesterol (P =
0.0082) and trend to higher IL6 (P = 0.084). The biomarkers that varied by the stroke
classification were OPG (osteoprotegerin, P = 0.018), IL10 (P = 0.015), MMP2 (P
= 0.0004), cholesterol (P = 0.0066), NIHSS at baseline (P < 0.0001) and 3-month
mRS (P < 0.0001). Patients with ischemic (LAA + CE + LAC) in comparison to
hemorrhagic stroke had lower NIHSS at baseline (P = 0.0192), mRS at 3 months (P
= 0.003), OPN (osteopontin, P = 0.016), OPG (P = 0.0087) and MMP2 (P = 0.0004).
Conclusion: In our study, various inflammatory circulating biomarkers correlated
with stroke severity, such as IL6, IL10 and MMP9. IL6 and blood leukocytes
correlated with 3-month outcome. Lower level of OPN, OPG and MMP2 were found
in ischemic in comparison with hemorrhagic stroke patients and should be further
studied as diagnostic biomarkers of stroke classification. Patients with hemorrhagic
stroke had more severe neurological deficit at baseline and worse 3-month outcome.
Circulating inflammatory biomarkers should be further examined in patients with
stroke.
Supported by MH CZ - DRO (Faculty Hospital Plzen - FNPl, 00669806) and by the
project ED2.1.00/03.0076 from the European Regional Development Fund.

A-373
Evaluation of an IgM assay for use on the Binding Site Next Generation Protein
Analyser

S. Amin, F. Murphy, S. J. Harding, P. J. Showell. The Binding Site Ltd,

Birmingham, United Kingdom
IgM is the first immunoglobulin produced in a primary immune response.
It constitutes around 10% of the total serum immunoglobulin, and is
largely in a pentameric form. Pentameric IgM is able to flex into a crab
or staple formation when binding to antigen and activates the classical
pathway of complement. Here we describe the evaluation of a serum IgM
assay for use on the Binding Sites next generation protein analyser. The
instrument is a random-access bench top turbidimetric analyser capable of
a wide range of on-board sample dilutions (up to 1/10,000) and throughput
of up to 120 tests per hour. Precision is promoted by single-use cuvettes
which are automatically loaded and disposed of, whilst the utility is
enhanced through host interface capability, primary sample ID and bar
coded reagent management systems. The instrument automatically dilutes
a single calibrator to produce a calibration curve with a measuring range
of 0.2 - 7.5g/L at the standard 1/20 sample dilution, with sensitivity of
0.1g/L. High samples are remeasured at a dilution of 1/400 with an upper
measuring range of 4.0 - 150.0g/L. The assay time is 10.5minutes and is
read at end point. Precision studies (CLSI EP5-A2) were performed at
eight levels in duplicate over 21 working days. Antigen levels of 0.19g/L,
0.28g/L, 0.38g/L, 1.44g/L, 1.65g/L, 2.68g/L, 5.11g/L and 10.21g/L were
assessed for total, within-run, between-run and between-day precision,
using one lot of reagent on three analysers. The coefficients of variation
were 6.2%,1.4%, 2.3% and 5.5% for the 0.19g/L sample, and 5.6%, 1.5%,
3.8% and 3.8% for the 0.28g/L sample, 5.6%, 1.2%, 3.7%, 4.1% for the
0.38g/L sample, 4.0%, 2.0%, 3.4% and 0.0% the 1.44g/L sample, 3.4%,
1.6%, 2.6% and 1.4% for the 1.65g/L sample, 3.6%, 1.2%, 3.0%, 1.6% for
the 2.68g/L sample, 4.0%, 2.2%, 3.0% and 1.3% for the 5.11g/L sample and
5.6%, 2.5%, 3.6% and 3.4% for the 10.21g/L sample respectively. Linearity
was assessed by assaying a serially-diluted patient sample pool across the
width of the measuring range (0.198 - 7.662 g/L) and comparing expected
versus observed results. The assay showed a high degree of linearity when
expected values were regressed against measured values; y=0.9953x
+ 0.0235, R = 0.9993. No significant interference (within 10%) was
observed on addition of bilirubin (20mg/dL), haemoglobin (500mg/dL)
or chyle (1500 formazine turbidity units) when spiked into a sample with
known IgM concentrations and measured at the minimum sample dilution.
Correlation of this assay with the equivalent assay for the Binding Site
SPA PLUS was performed using both normal and clinical samples (n=115,
range 0.135 - 60.424 g/L). Good agreement was demonstrated by PassingBablock regression; y=0.98x - 0.02 g/L. We conclude that the IgM assay
for the Binding Site next generation protein analyser is reliable, accurate
and precise and shows good agreement with existing assays.

S110

A-374
Antibodies-to-Infliximab: Assay Development and Correlation with Infliximab
Concentrations in Serum Samples of Treated Patients

M. A. V. Willrich, J. G. Balsanek, P. M. Ladwig, D. R. Barnidge, D. L.

Murray, M. R. Snyder. Mayo Clinic, Rochester, MN
Background: Infliximab (IFX) is a chimeric therapeutic monoclonal IgG1 kappa
antibody targeting tumor necrosis factor- and is FDA-approved for treatment of
several inflammatory disorders. Patients undergoing therapy may form antibodiesto-Infliximab (ATIs), which can reduce circulating drug concentrations. Measurement
of IFX concentration and ATIs is useful for guiding therapy in patients who have lost
clinical response. Here we present the development of an electrochemiluminescent
immunoassay (ECLIA) for detection of ATIs. Methods: A bridging ECLIA was
developed in which the ATI forms a link between biotin- and sulfo-tag-labeled IFX
(MesoScale Discovery LLC). An 8-point standard curve was established using a highaffinity human IgG1 ATI (AbD Serotec), and an acid-dissociation step was performed
to disrupt immune-complexes. Residual sera with physician-ordered IFX and ATI
were evaluated (n=37) and compared to results from a reference laboratory (Esoterix).
Serial sera from patients on IFX were collected at trough levels (n=36), 48-72 hours
post-infusion (n=33) and 28-32 days post-infusion (n=33). IFX was measured using
clonotypic heavy chain peptides by LC-MS/MS. Results: The analytical measurable
range for the ECLIA ATI assay was established as 19.5-2,500ng/mL (R2=0.9911).
Limit of quantitation was defined as 19.5ng/mL; results >19.5ng/mL were classified
as positive. Using residual serum samples, the ECLIA showed an overall qualitative
concordance of 86.4% to a commercial method. In the cohort of serial samples,
22% (8/36) of trough samples, with IFX concentrations of 8.58.8g/mL, were
positive for ATI; in 7/8 samples the ATIs persisted 28-32 days after IFX infusion
(IFX concentration 1511g/mL). Samples obtained 48-72h after IFX infusions were
negative, suggesting that high concentrations of IFX (7740g/mL) interfere with
ATI measurement. Presence of ATIs was associated with lower concentrations of IFX
at trough (8.88.9g/mL in ATI negatives vs. 3.34.8g/mL in positives, p=0.0038),
48-72h post-infusion (7841g/mL vs. 4712g/mL, p=0.0357) and at 28-32 days
after (157g/mL vs. 710g/mL, p=0.0048). Out of the 8 positives measured at
trough, 7 had IFX<5g/mL. In the residual sera cohort used for method comparison
samples with IFX<5g/mL (n=18) showed 100% concordance for ATI by the two
methods. In contrast, samples with IFX5g/mL (n=19), had a concordance of 73.7%.
In the presence of detectable trough levels of IFX, significance of ATIs is unclear.
Based on these data, we propose an algorithm for assessment of patients showing a
loss of response to IFX. Initial testing would be performed for quantitation of trough
IFX by LC-MS/MS with a reflex to ATIs in cases when IFX concentrations are
<5g/mL. Conclusions: Using a newly-developed ECLIA bridging method, we have
demonstrated that the presence of ATIs correlated with lower IFX concentrations and
that the majority of ATIs were found in patients with trough IFX<5g/mL. In addition,
interference by endogenous IFX may be an issue, specifically when assessed at peak
levels. An algorithmic approach starting with the quantitation of IFX measured at
trough with a reflex to ATI could provide guidance to clinicians in identifying patients
who might respond to increased doses of IFX compared to patients for whom another
biologic agent might be more appropriate.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

green chemistry technique that combines sampling/sample preparation in a single step,
has shown to be a powerful tool for the determination of multiple prohibited drugs in
complex matrices. However, in some circumstances the quantification capabilities of
SPME are constrained by the instrumental limits of detection and quantification of the
LC-MS/MS systems.

Tuesday, July 29, 2014

Poster Session: 9:30 AM - 5:00 PM
Mass Spectrometry Applications

A-375
Tandem mass method for disaccharide units of urinary glycosaminoglycans
from MPS patients

C. Chuang, H. Lin, C. Tsai, S. Lin. Mackay Memorial Hospital, Taipei,

Taiwan
Background: Identification of acid mucopolysaccharide by liquid chromatography/
tandem mass spectrometry method (LC-MS/MS) of predominant disaccharide units of
glycosaminoglycans (GAGs) (chondroitin sulfate, CS; dermatan sulfate, DS; heparan
sulfate, HS) after methanolysis is validated and applicable for mucopolysaccharidosis
(MPS) phenotype determinations.
Methods: A total of 76 urine samples were collected and analyzed, including 9 MPS I
patients, 13 MPS II, 7 MPS III, 8 MPS VI, and 39 normal controls. Urinary GAG was
first precipitated by Alcian blue method followed by a treatment of 3N HCl methanol.
The protonated species of the methylated disaccharide products were detected by
using a multiple reaction monitoring experiment. Internal standards, [2H6] CS, [2H6]
DS and [2H6] HS, were in-house prepared by deuteriomethanolysis of CS, DS and HS.
Results: The within-run and between-run precisions were good, and the recoveries
were 94.3% for DS and 95.1% for HS. Linearity of DS and HS was calculated
individually and the correlation coefficients (r) were 0.9914 for DS and 0.9935 for
HS, respectively. One particular disaccharide for each GAG was selected, in which
the parent ion and its daughter ion after collision were m/z 426.1236.2 for DS (m/z
432239 for dimmers derived from [2H6] DS) and m/z 384.2161.9 for HS (m/z
390.4162.5 for the [2H6] HS dimmer). The results were correspondent well when
comparing with the two-dimensional electrophoresis method. The quantities of DS
and HS were determined, which were varied from one MPS phenotype to the others,
and the results can be used to evaluate the severity of MPS subgroups, as well as the
amelioration of follow-up after enzyme replacement therapy (ERT).
Conclusion: The modified LC-MS/MS method for MPS phenotype determination is
specific, sensitive, validated, accurate and applicable for simultaneous quantifications
of urinary DS and HS. This method can help to make correct diagnosis of MPS
patients and evaluate the effectiveness of ERT.

A-376
Improving detection limits of prohibited substances and therapeutics by Solid
Phase Microextraction (SPME) coupled to LC-MS/MS

F. Ruparelia1, J. Pawliszyn2, N. Reyes-Garcs2, G. Augusto Gmez-Ros2,

B. Bojko2. 1IONICS Mass Spectrometry Group Inc., Bolton, ON, Canada,
2
University of Waterloo, Waterloo, ON, Canada
This work presents the enhancement achieved in limits of detection for a set of
prohibited substances and therapeutics by coupling a completely automated thin-film
SPME analytical protocol to a powerful LC-MS/MS system developed by IONICS
Mass Spectrometry.
LC-MS/MS analyses were performed using a triple quadrupole MS 3Q-320
(IONICS, Bolton, Ontario, Canada). Complete separation was achieved using a
pentaflourophenyl column (Phenomenex, Torrance, CA, USA) and a ternary gradient
of water, methanol and acetonitrile (0.1% formic acid in all mobile phases). Sample
preparation was performed using a Thin Film Microextraction (TFME)-Concept
autosampler. The automated procedure consisted of the following steps: TFME preconditioning (samples were simultaneously incubated at 30 C) (30 min), extraction
(75 or 90 min from pooled urine samples spiked at different concentration levels),
washing step (10 s), and desorption in a proper solvent for 20 or 60 min. HLB and
C18 coatings in thin-film SPME configuration were prepared in-house following a
protocol developed in our laboratory.
In order to preserve sporting ideal and ensure fair play game, the World Anti Doping
Agency (WADA) has banned performance-enhancing substances in competitive
sports. In most cases, despite the use of extensive and cumbersome sample preparation
protocols, the extraction and detection of these compounds in complex matrices such
plasma, urine, and blood, can be a challenge. Solid phase microextraction (SPME), a

This work presents a SPME method coupled to an effective MS/MS system

towards the quantification of compounds with low minimum required performance
level (MRPL) values set by WADA. Analytes selected in this study (boldenone,
cannabidiol, cannabinol, dihydrotestosterone, fentanyl, fluoxymestrone, methyltestosterone, nandrolone, testosterone and 19-norandrosterone) were of particular
interests here as they did not meet the MRPL levels with the SPME-LC-MS protocol
previously developed in our group. Analysis of neat standards (prepared in methanol
at a concentration range of 0.05 to 100 pg/L) performed in a triple quadrupole MS
IONICS 3Q-320 demonstrated allowed meeting the MRPL values required by WADA,
and selected compounds displayed outstanding limits of quantification. For instance,
instrumental LOQs of 33 and 200 ag/L were achieved for fentanyl and testosterone,
respectively, with a correlation coefficient equal or better than 0.999. In addition,
therapeutics such as pancuronium and rocuronium exhibited similar instrumental
LOQs using the same MS analyzer. These findings provide an opportunity to expand
the applicability of SPME to quantify low concentrations not only for ex-vivo analysis
but also for in-vivo applications in which the temporal resolution, sensitivity and
accuracy provided by SPME is highly desired.

A-377
Enhanced Resolution and Matrix Interference Reduction for the Analysis of
Vitamin D Metabolites

C. Aurand, D. Bell, T. Ascah-Ross. Supelco, Bellefonte, PA

Analysis of Vitamin D metabolites has continued to be a topic of interest in recent
publications, primarily as biomarkers for possible disease states and vitamin
sufficiency. While Vitamin D is present in two forms, current ELISA methods cannot
distinguish D2 and D3 forms of the vitamin metabolites resulting in a reporting
of total 25-hydroxyvitamin D. In this study, an LC/MS method for the analysis of
Vitamin D metabolites is expanded to include dihydroxy metabolites along with the
epi-homologs. Chromatographic resolution is utilized for the quantitation of hydroxy
and dihydroxy Vitamin D2 and D3 metabolites including the isobaric epimers. In
addition, sample preparation techniques are evaluated for the impact of biological
matrix ionization effects.
Chromatographic method development consisted of screening C18, Cyano,
Phenyl Hexyl and pentyl fluorophenyl (F5) stationary phase. Method development
experminets resulted in conditions for the direct quantitation of isobaric metabolites
25 hydroxyvitamin D3, 3-epi-25 hydroxyvitamin D3 1- hydroxyvitamin D3 along
with 25 hydroxyvitamin D2, 3-epi-25 hydroxyvitamin D2. In addition, human serum
samples were processed using standard protein precipitation techniques along with
novel phospholipid depletion plates for the comparison of matrix interference impact.
. The unique combination of the selectivity of the F5 separation along with the novel
sample preparation technique allow of a robust and accurate LC/MS method for
quantitation of all the associated Vitamin D metabolites

A-378
Reference interval determination for the ratios of L-arginine (ARG), symmetric
dimethylarginine (SDMA) and asymmetric dimethylarginine (ADMA)

J. M. El-Khoury, D. R. Bunch, B. Hu, S. Wang. Cleveland Clinic,

Cleveland, OH
Background:
Symmetric
dimethylarginine
(SDMA)
and
asymmetric
dimethylarginine (ADMA), metabolic products of methylated L-arginine (ARG)
containing proteins, play an important role in regulating nitric oxide production.
Recently, SDMA and ADMA have been extensively evaluated as biomarkers of renal
and/or cardiovascular diseases with indication of potential use of the ratios of ARG/
ADMA and SDMA/ADMA. More specifically, the ratio ARG/ADMA has been shown
to be an independent predictor of mortality in patients with dilated cardiomyopathy,
while the ratio SDMA/ADMA has been investigated as a biomarker of hypertension in
rats and was found to be a better predictor for disease activity and progression than the
individual parameters alone. However, reference intervals (RIs) for the ratios ARG/
ADMA, ARG/SDMA and SDMA/ADMA have not been reported in the literature.
Our objective in this study was to determine RIs for these ratios in a healthy adult
population using a liquid chromatography-tandem mass spectrometry (LC-MS/MS)
method. Methods: Collection of blood samples for RI determination was approved

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S111

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

by the Institutional Review Board. In brief, EDTA whole blood samples (n = 51) were
collected from healthy adults (39 females) identified via a questionnaire, aged 19-64 y
(38.8 12.6), after a minimum of 8 hour fasting. A published LC-MS/MS assay was
used for the analysis of ARG, SDMA and ADMA. EP Evaluator release 10 was used
for statistical analysis. Results: The central 95% RI for ARG/ADMA, ARG/SDMA
and SDMA/ADMA were 108 to 247, 95 to 261, and 0.64 to 1.27, respectively. Based
on EP evaluators determination a transformed parametric method for ARG/ADMA
and a parametric method for ARG/SDMA and SDMA/ADMA were used. Histograms
are shown in the figure. Conclusion: RIs for ARG/ADMA, ARG/SDMA and SDMA/
ADMA were determined using a well-defined healthy population by an LC-MS/MS
method. These values are important in defining the clinical utility of these parameters.

fragmentation specificity. Moreover, ion mobility-derived collision cross sections

provided orthogonal physicochemical data that were used with retention time,
accurate mass and MS/MS data to increase confidence of metabolite identification.
Data was collected using both negative and positive ionization mode in the dataindependent acquisition mode with an alternate low and elevated collision energy
method to acquire both precursor and product ion information in a single analytical
run. Lipidome and metabolome data were fused and mined using multivariate
statistical and pattern-recognition tools. Initial observations were confirmed using
more targeted approaches for quantification. Pathway analysis was then used to
incorporate the novel molecular information into the known biochemical pathways.
The results obtained were further integrated with clinical data to generate testable
hypotheses on the functional significance of the abnormalities observed in AD. Our
preliminary results reveal novel molecular alterations in AD and a unique lipidome
and metabolome biosignature that differentiates the brains from individuals with AD
compared from those from control subjects.

A-380
Pain Management Drug Monitoring in Urine using HPLC-MS/MS

J. K. Wolken, T. Doran, P. Smith, A. Iwanski, D. Wiebe. University of

Wisconsin Hospital and Clinics, Madison, WI

A-379
A unique brain lipidome and metabolome biosignature in Alzheimers Disease

G. Astarita1, G. Paglia2, J. Langridge3, S. Lai3. 1Georgetown University/

Waters Corporation, Milford, MA, 2Center for Systems Biology, University
of Iceland, Reykjavik, Iceland, 3Waters Corporation, Milford, MA
Alzheimers disease (AD) is the most common cause of adult dementia, but the
cause of this inexorable neurodegenerative disease remains still elusive. Alterations
in both lipid and polar metabolites biochemical pathways have been associated with
AD. Here we conducted an unbiased investigation of the underlying biochemical
alterations in AD human tissues. We used an integrated lipidomics and metabolomics
approach to survey frozen brain tissue samples from clinically characterized AD
patients and age-matched controls. Lipids and polar metabolites were extracted
using a biphasic, liquid-liquid extraction procedure. Polar metabolites were separated
using a hydrophobic interaction liquid chromatography (HILIC), whereas lipids
using an integrated microfluidic device packed with reversed phase C18. TravellingWave ion mobility mass spectrometry was used to improve peak capacity and CID

S112

Background: With prescription drug abuse reaching epidmic proportions,

clinicians are seeking solutions to monitor aptients on long-term pain management
prescriptions. These monitoring programs need to be sensitive to low level amounts
of pain management therapeutics in a patient population correctly adhering to
physicians orders while screening for non-prescribed pain medication and drugs
of abuse. The University of Wisconsin Hospital and Clinics (UWHC) Toxicology
group has developed a clinical test which will monitor 35 drug substances
with minimal sample preparation in urine in a 10 minute HPLC-MS/MS run.
Methods: The following compounds are monitored in the following catagories: Drugs
of abuse: 6-MAM (heroin metabolite), amphetamine, benzoylecgonine (cocaine
metabolite), MDA, MDMA (ecstasy), and methamphetamine. Benzodiazepines:
alprazolam, hydroxyalprazolam, 7-aminoclonazepam, diazepam, nordiazepam,
lorazepam, midazolam, oxazepam, and temazepam. Opioids: buprenorphine,
norbuprenorphine, codeine, fentanyl, norfentanyl, hydrocodone, hydromorphone,
meperidine, normeperidine, methadone, EDDP (methadone metabolite), morphine,
naloxone, naltrexone, oxycodone, oxymorphone, tapentadol, N-desmethyltapentadol,
tramadol, and N-desmethyltramadol. Patient urine is mixed 1:1 with an internal standard
which consists of 8 deuterium labeled compounds (a subset of the 35). This is injected
onto a BiPh Restek column, and using a Agilent 1200 HPLC the compounds are eluted
into a AB Sciex 4000 triple quadruploe mass spectrometer. Each compound has 2 ion
transitions for identification and quantitation. The AMR for most of the compounds are
25-1000 ng/ml with most of the benzodiapizines having an AMR of 10-1000 ng/ml.
Results: The within run CV range for the compounds at the low end cutoff
(25ng/ml for most compounds) was between 1.5-10.1% where the day to day
precision at the same low cutoff was 4.0-15.5%. Carryover has not been shown
to be an issue for any drug substance despite some extremely high patient
samples being analyzed. Comparison of methods to an in-house GC-MS method
showed increased sensitivity for most compounds, most notably 6-MAM,
morphine, and oxycodone. Assayed College of American Pathologist (CAP)
Toxicology surveys from 2011 and 2012 showed 100% qualitative agreement
for all 30 survey samples for the compounds in the Pain Management Profile.
Conclusion: The new pain panel method shows very little limitation in real world
patient samples. Even with minimal sample preparation steps very few issues have
arisen with interferences, ion suppression, and retention time shifts. We advise our
clinicians that the panel results are only a snapshot in time. The dose, time of dose,
state of hydration, and individual metabolism all play a part in the concentration of
drug substances in the urine. Intermittant testing over time is recommended to help
compile a clearer picture of the patients compliance.

A-382
Determination of Tacrolimus and Sirolimus in whole blood by liquid
chromatography electrospray ionization tandem mass spectrometry

M. E. R. Diniz, L. S. Vilhena, T. M. C. C. Barbosa, B. P. Paulo, E. C. C.

Mateo, A. C. S. Ferreira. Instituto Hermes Pardini, Vespasiano, Brazil
Tacrolimus and Sirolimus are immunesuppressive drugs used in organ transplantation,
exhibit narrow therapeutic ranges and adverse effects are common. Methods based
on Liquid Chromatography-tandem Mass Spectrometry (LC-MS/MS) are considered

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

the gold standard in therapeutic monitoring of these drugs. So a simple, rapid and
sensitive LC/MS/MS method has been developed and validated for the determination
of Tacrolimus and Sirolimus in whole blood using Ascomycin as the internal standard
(IS). In this process, 100 L of whole blood samples containing the internal standard
were treated by liquid-liquid extraction using ethyl acetate and subjected to LCMS/MS analysis using positive electrospray ionization (ESI+). Chromatographic
separation was performed on a Symmetry C18 column (3.5 m 4.6 x 75 mm) and
mobile phase acetonitrile:methanol:water (80:10:10, v/v/v) with 0,1% of formic acid
and 0,02% of Ammonium hydroxide 25% at 400 L/min. The MS/MS detection
was conducted by monitoring the fragmentation ions of 821.7768.5 (m/z) for
Tacrolimus, 931.5864.5 (m/z) for Sirolimus and 809.7756.7 (m/z) for Ascomycin.
Ammoniated adducts of protonated molecules were used as precursor ions for all
analytes. The method had a chromatographic running time of approximately 4 min.
The linear analytical range of the procedure was between 1.0 and 51.0 ng/mL for
Tacrolimus and 2.0 and 52.0 ng/mL for Sirolimus. The medium range of recovery
for the Tacrolimus was 98.1-103.2% over a interval of 2.0-37.5 ng/mL and for the
Sirolimus 97.1-106.1% over a range of 2.0-39.0 ng/mL. The intra and inter-day
precision was less than 6.8% for Tacrolimus and 10.8% for Sirolimus. In conclusion,
the LC-MS/MS method has been developed successfully for the quantitative analysis
and therapeutic monitoring of these immunosuppressive drugs.

A-383
Evaluation of high performance liquid chromatography and liquid
chromatography-tandem mass spectrometry methods for 25 (OH) D3 assay

B. Omer, F. A. Aydin, P. Mikailova, S. Genc. Istanbul University Istanbul

Medical Faculty, Istanbul, Turkey
Background: Growing evidence about the role of vitamin D on health in different
fields of medicine emerged the necessity of establishing more reliable and accurate
25 (OH) vitamin D3 assessment methods. This study was designed to compare
performance characteristics of two different 25 (OH) vitamin D3 assessment methods.
Methods: 25 (OH) D3 vitamin levels were quantified using two methods as follows:
high-performance liquid chromatography (HPLC) with Thermo Finnigan TSP
(Florida-USA) and liquid chromatography-tandem mass spectrometry (LC-MS/
MS) with Zivak Technologies ZinMass-200 LC-MS/MS System with APCI
ionization source (Istanbul-TURKEY) as the reference method. Assay performance
characteristics were performed according to the National Committee for Clinical
Laboratory Standards (NCCLS) .The comparison studies were done using randomly
chosen 306 plasma samples from routine clinical samples submitted for 25 (OH) D3
vitamin measurement.
Results: The LC-MS/MS assay had within-run coefficient of variation (CV) 7.3%
and between-day CV 6.0% for low control (22.43.3 ng/mL), for high control (82.3
11.8 ng/mL) within-run CV 7.2% and between-run CV 9.6%. HPLC method had
within-run CV 6.9% and between-day CV 12.5% for low control, and for high control,
within-run and between-run CVs were 5.6 and 8.7 respectively. The linearity studies
showed good correlations between expected and obtained values for both methods.
When the relationship between the results obtained from HPLC and LC-MS/MS
assays was investigated in 306 subjects, the HPLC assay showed an acceptable
correlation with the LC-MS/MS (y=1.054x-1.98, R=0.9752).
Conclusion: With good precision and accuracy, HPLC system revealed an acceptable
correlation with LC-MS/MS for 25 (OH) vitamin D3 assay.
Figure 1: Comparison of plasma 25 (OH) vitamin D3 levels measured by HPLC vs
LC-MS/MS.

A-384
Elimination of false positives and false negatives for the screening of
amphetamines, PCP, and benzoylecgonine by Agilent RapidFire

D. Ames. IQ Laboratories, Fort Wayne, IN

This research shows a more accurate alternative to using an immunoassay for the
screening of amphetamines, PCP and benzoylecgonine. Immunoassays are the typical
method used to screen for amphetamines, PCP, and benzoylecgonine. However,
immunoassays are known to produce false positives, false negatives and require the
use of expensive reagents. By utilizing an Agilent RapidFire 300 High-throughput
System we eliminated false positives and false negatives. False positives and false
negatives were eliminated because RapidFire uses an Agilent QQQ triple quad mass
spectrophotometer as the detection method rather than enzymes or antibodies. The
use of a triple quad mass spectrophotometer as the method of detection also allows
for selectivity that is not possible with immunoassays. In this study on RapidFire
PCP, benzoylecgonine, amphetamine, methamphetamine, MDMA, MDA, MDEA,
methylphenidate and ritalinic acid were screened for and on the DxC 600i the reagents
from Beckman Coulter for amphetamines, PCP, MDMA and benzoylecgonine were
used. Patients were screened on a Beckman Coulter DxC 600i and then on an Agilent
RapidFire 300 High-throughput System connected to an Agilent 6460 QQQ triple
quad mass spectrophotometer. Those patients found to be positive on either system
were then analyzed by LC-MS/MS for further confirmation. The use of RapidFire
allowed us to eliminate false positives and false negatives and increase selectivity.

A-386
Determination of urinary ethyl glucuronide and ethyl sulfate by LC/MS/MS for
clinical research

L. Cote. Agilent Technologies, St-Laurent, QC, Canada

Background: Liquid chromatography triple quadrupole mass spectrometry (LC/MS/
MS) is ideally suited for the rapid analysis of multiple analytes. A highly sensitive
and specific LC/MS/MS analytical method has been developed for the quantitation
of urinary ethyl glucuronide (EtG) and ethyl sulfate (EtS). A dilution procedure and a
solid phase extraction (SPE) procedure are evaluated and compared based on ease of
use, analyte recovery and post-extraction cleanliness.
Methods: A dilution procedure and a solid phase extraction (SPE) sample preparation
procedure were developed and compared for the simultaneous extraction of ethyl
glucuronide and ethyl sulfate in urine. Calibrators were created by spiking synthetic
urine (Surine-Cerilliant) with various concentrations of EtG and EtS standards
(Cerilliant). The chromatographic system consists of a Polaris 3 C18-Ether column
coupled with a guard column and a mobile phase comprised of acetonitrile and water
containing 0.1% formic acid. Quantifier and qualifier transitions were monitored.
EtG-D5 and EtS-D5 internal standards (Cerilliant) were included to ensure accurate
and reproducible quantitation. Urine controls (UTAK Laboratories) were used and
samples were kindly supplied by collaborators.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S113

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

Results: The separation of EtG and EtS from isobaric interferences is especially
critical; without proper separation by retention time, impurities present in both
compounds can cause interferences with one another and lead to inaccurate
quantitation. The described method achieves the required functional sensitivity and
is capable of quantitating analytes over a sufficiently wide dynamic range with a
single injection. All analytes displayed excellent linearity from 25 to 10000 ng/ml.
All calibration curves displayed an R2 > 0.9993. Back calculated accuracies for all
calibrators ranged from 93% to 120% for the dilution procedure and from 96 to 113%
for the SPE procedure. Commercially available quality control material was used to
test the reproducibility of this method. Measurements were repeated on three separate
days to assess interday reproducibility and CVs were found to be below 6%.
Conclusion: A method has been developed for quantifying ethyl glucuronide (EtG) and
ethyl sulfate (EtS) in urine for clinical research. Two sample preparation procedures
consisting of a simple dilution from urine and SPE are shown. Chromatographic
separation of all analytes and interferences with conditions compatible with LC/MS/
MS have been developed. Typical analytical method performance results are well
within acceptable criteria.

A-387
A Quantitative Determination of Methadone and its Metabolite (EDDP)in Dry
Blood Spot by LC-MS/MS

J. Ye, H. Qiao, E. Majdi, L. Cousins. IONICS Mass Spectrometry Group

Inc., Bolton, ON, Canada
Background: Critically ill children routinely receive opioids for analgesia and
sedation to reduce pain and stress, facilitate ventilation, and avoid secondary
complications. The typical course of treatment often induces tolerance, and withdrawal
symptoms may be precipitated if the drugs are discontinued abruptly. Withdrawal
symptoms are not only unpleasant but can be life-threatening and may prolong the
need for mechanical ventilation and potentially extend hospitalization. Methadone is
a synthetic opioid receptor agonist widely used in the treatment of severe pain and in
maintenance treatment of opioid addicts. It is approved by the U.S. Food and Drug
Administration for detoxification treatment of opioid addiction in adults, but does not
have a label for pediatric use. Therefore it is necessary to monitor the dosage and to
avoid the abuse. The present study provides a much simplified approach whereby a
novel, highly sensitive LC-MS/MS triple quadrupole mass spectrometer is used to
measure directly the methadone and EDDP concentrations through the dried blood
spot (DBS) samples.
Methods: Fresh human whole blood was spiked with different concentrations of
methadone and EDDP. 30 ul of the blood were spotted onto Whatman DMPK-C
cards. Cards were air dried for about 2 hours. 6 mm spot were punched placed into a
vial containing 100ul IS (0.01ng/mL) working solution. Samples were vortexed for 3
minutes and centrifuged for 5 minutes at 4000 rpm. The supernatants were transferred
into HPLC vials for analysis. An IONICS 3Q 220 mass spectrometer was used. This
instrument is equipped with heated coaxial flow ion source and Hot Source-Induced
Desolvation interface, with a multi-orthogonal channel and laminar flow sampling.
The samples were injected using Shimadzu UFLC XR system. Sample was loaded to
a Chromolith-RP18E column (100X3 mm, 3 um) with a gradient method at 0.5 mL/
min. Mobile Phases were A (0.1% formic acid 100% H2O) and B (0.1% formic acid
in 100% ACN). The total LC run time is 3.5 minutes.
Results: Calibration curve of the neat methadone and EDDP solution showed good
linearity over a range of 0.0025-10 ng/mL with correlation values of R2=0.9997 and
0,9998, respectively. In DBS extraction the calibration curves for methadone and
EDDP over a range of 0.1 to 100 ng/mL were created. The curves also showed good
linearity with weighting factor of 1/x for both analytes. The correlation values were
R2=0.997 and 0.998 for methadone and EDDP, respectively. At LLOQ of 0.1 ng/
mL, a good accuracy of 108% and 99% and CVs of 9% and 8.3% were obtained for
methadone and EDDP, respectively.
Conclusion: The results in this study show a fast, accurate, and precise LC-MS/
MS method with IONICS 3Q 220 mass spectrometer for quantifying methadone
and EDDP in DBS samples. The LLOQs for both samples are 0.1 ng/mL with
good precision and accuracy. The sample preparation procedure is simple and rapid
without SPE and LLE extraction. Therefore, the LC-MS/MS method in this study has
confirmed its clinical applicability and can be used in routine bioanalysis, especially
for methadone level monitoring.

S114

A-388
Increased Throughput for the Analysis of delta-9-THC in Oral Fluids using
Triple Quadrupole Mass Spectrometry coupled to Automated Dual-Channel
HPLC

K. McCann, A. Szczesniewski. Agilent, Santa Clara, CA

Background: Many forensic labs are interested in improved sample throughput to get
better utilization of the testing instrumentation. Through this work we demonstrated
the ability to increase mass spectrometer productivity through the automated use of
a dual channel high performance liquid chromatography (HPLC) system. A newly
developed software interface intelligently determines the timing of all HPLC
components and coordinating the analytical utilization of the mass spectrometer.
Methods: The integrated LC/MS/MS system is comprised of a triple quadrupole
mass spectrometer coupled to a configurable HPLC system, all controlled by a single
software application. For the purposes of this work, the HPLC system consists of a
high-capacity autosampler, two binary pumps, two HPLC columns, two temperaturecontrolled column compartments and one switching valve. To operate the system,
a standard data file collected by LC/MS/MS is loaded into the software. The data
analysis method is extracted from the data file and a window of interest is specified
using the data files chromatogram. Based on that information, the software
automatically coordinates all timing related to running the HPLC system.
Results: The analysis of delta-9-THC is performed in many forensic toxicology labs
analyzed by LC/MS/MS where sample throughput is a major concern. An established
LC/MS/MS method for the analysis of this analyte from oral fluids was used for testing
the capabilities of this new instrument. The standard method uses an autosampler,
binary pumps, HPLC column and temperature-controlled column compartment. With
a runtime of 5 minutes, the analytes of interest reach the mass spectrometer between
approximately 2 minutes to 4 minutes. Hence, more than 50% of the data collected by
the mass spectrometer is of no interest.
The standard method utilizes what is considered a single HPLC stream. The new
HPLC system mirrors certain components of this single stream system to provide
a second stream, operating in parallel to the first stream. By loading the standard
method and window of interest into the automation software, the software is able to
determine the most efficient method of injecting and analyzing a list of delta-9-THC
samples without any user configuration necessary. By staggering injections on parallel
streams and switching between the two streams at the appropriate time, throughput of
the integrated expanded system can double the throughput achieved with the standard
method, without any sacrifice to the quality of quantitation.
Conclusion: Fully automated software controlling a completely integrated LC/MS
system consisting of two parallel LC streams has been developed and implemented
in the analysis of 9-THC. No special method development is required; the user
supplies a standard method and defines a window of interest, allowing the software
to determine all necessary timing and coordination of the analysis. Throughput for
this method has been doubled through the use of an Automated Dual-Channel HPLC.

A-389
Quantitative Analysis of Acetaminophen and Salicylic Acid in Urine by
Rapidfire Coupled with Triple Quadrupole Mass Spectrometry

F. Mbeunkui, S. Sullivan, R. B. Dixon. Physicians Choice Laboratory

Services, Rock Hill, SC
Background: A rapid and quantitative method for the direct quantitation of
acetaminophen and salicylic acid in biological matrices is warranted in clinical and
toxicological chemistry. Compliance monitoring and therapeutic drug monitoring
of these compounds can inform clinicians when assessing potential side-effects of
the drugs including hepatocellular damage or tinnitus. The predominate method
for measuring acetaminophen and/or salicylate in biological matrixes is by enzyme
immunoassay (EIA). EIA techniques and reagents are labeled and limited to use for
serum testing. The purpose of this assay is to provide specific drug and metabolite
quantitative data in alternative biological matrices.
Methods: Rapidfire is an automated sampling instrument that injects the sample onto
a small cartridge packed with stationary phase. The instrument is programmed to load
the sample onto a cartridge then rinse with aqueous mobile phase. After rinsing the
sample, the Rapidfire diverter valve switches the flow path to the mass spectrometer. A
highly organic mobile phase then elutes the purified compounds which are ionized by
electrospray ionization prior to detection by the triple quadrupole mass spectrometer.
The panel detects acetaminophen, acetaminophen-glucuronide, and salicylic acid.
Deuterium labeled internal standards for acetaminophen and salicylic acid are utilized
for isotope dilution. Quantitation is enabled by calibration of the system with 7 levels

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

of matrix spiked standards constructing a standard curve prior to sample analysis.

Results: We have developed a quantitative Rapidfire mass spectrometry (RFMS)
method for the analysis of acetaminophen and salicylic acid in urine. The sample
preparation method is simple. Samples are hydrolyzed, then diluted with an
internal standard solution prior to loading on the RFMS. The analytical procedure
is approximately 10 seconds from sample to sample. Due to the high variability of
matrix affects in patients undergoing chronic pain or mental health treatment, it is
necessary to have a robust analytical method. To that end, we have investigated a
wide range of samples with variable creatinine concentrations. The calibration range
of the method is 0.5-200 g/mL for acetaminophen and 1-400 g/mL for salicylic
acid providing a wide analytical measurement range. External QC materials were also
tested with acceptable imprecision and accuracy (<15%). Real patient samples have
been analyzed and this procedure demonstrates a robust and reproducible method for
this assay.
Conclusion: This technique provides rapid and quantitative analysis of acetaminophen
and salicylic acid from biological specimens with minimal sample preparation.

A-390
Determination of Insulin-Like Growth Factor-1 in serum by HRAM LC-MS for
clinical research

L. Cote1, K. McCann2, C. Chu2. 1Agilent Technologies, St-Laurent, QC,

Canada, 2Agilent Technologies, Santa Clara, CA
Background: High Resolution Accurate Mass (HRAM) Liquid Chromatography
Mass Spectrometry (LC-MS) is ideally suited for the rapid analysis of biomolecules.
A highly sensitive and specific HRAM LC-MS method has been developed for the
quantitation of Insulin-Like Growth Factor-1 (IGF-1) in serum. This method uses a
simple sample preparation combined with an online sample cleanup procedure coupled
to a high resolution accurate mass quadrupole time-of-flight mass spectrometer.
Methods: An efficient sample preparation procedure was developed for the extraction
of IGF-1 in serum. Calibrators were created by spiking clean serum with various
concentrations of IGF-1. The chromatographic system consists of a C8 extraction
column coupled with a high resolution, 300 angstrom pore size analytical column
and a mobile phase comprised of acetonitrile and water containing 0.2% formic acid.
Quantifier and qualifier transitions were monitored and Rat IGF-1 internal standard
was included to ensure accurate and reproducible quantitation.
Results: Online sample cleanup and chromatographic separation of a sample is
achieved in less than three minutes. The separation of Rat IGF-1 and Human IGF1 is especially critical since these compounds share common interferences. Without
proper separation by retention time, impurities present in both compounds can cause
interferences with one another and lead to inaccurate quantitation. The described
method achieves the required functional sensitivity and is capable of quantitating IGF1 over a sufficiently wide dynamic range. This method displayed excellent linearity
from 5.63-990 ng/mL. All calibration curves displayed an R2 > 0.999. Back calculated
accuracies for all calibrators ranged from 84% to 105% and showed intra-day CVs
below 8%. Separately prepared incurred samples were used to test the accuracy and
reproducibility of this method. Measurements were repeated in triplicates and on three
separate extractions to assess intra- and inter- day reproducibility and CVs were found
to be below 10%.
Conclusion: A robust method for quantifying Insulin-Like Growth Factor-1 in serum
with excellent reproducibility and accuracy has been developed.

A-391
Fast Determination of Serum Methylated Arginines By Liquid
Chromatography Tandem Mass Spectrometry

A. Unlu, S. Abusoglu, F. Akyurek. Selcuk University, Konya, Turkey

Background: NG,NG-dimethyl-l-arginine, or asymmetric dimethylarginine is a
naturally occurring amino acid that circulates in plasma. Free ADMA, and related
amino acids, NG-monomethyl-l-arginine, as well NG,N_G-dimethyl-l-arginine are
produced normally by all cells from hydrolysis of proteins containing methylated
l-arginine residues. SDMA shares the pathway for cell entry with arginine.
Endogenous methylarginines are important potentially modifiable molecules that
may be associated with impaired synthesis of nitric oxide. The aim of this study
was to implement a fast, accurate and simple mass spectrometric method for serum
methylated arginine.
Methods: 100 L of internal standard in methanol were added to 200 L of serum
and centrifuged at 13.000 rpm for 10 minutes. Supernatant was evaporised with N2

flow at 65 C. Derivatisation step was performed dissolving the dried extract in 200
L of a freshly prepared butanol solution containing 5% (v v1) acetyl chloride and
kept at 65 C for 30 min. The solvent was removed by evaporation under nitrogen
flow at 65 C. The samples were dissolved in 200 L of watermethanol (90:10, v v1)
containing 0.1% (v v1) formic acid and 40 L injected into system. Multiple reaction
monitoring was performed with a continuous infusion of a 50 M solution of each
analyte. Recovery test was calculated as average of measured value/expected value
ratio (%). Limit of detection and quantification were determined by a signal to noise
ratio of 3:1 and 10:1, respectively. Inter- and intra-assay precision were evaluated by
analysis of ten replicates of C1, C2 and C3, daily for 3 days and expressed as mean,
SD and CV%.
Results: This methods intra-assay CV and % bias values were 15.6,10.2; 9.72,7.88
and 6.45,6.02 for 0.4, 0.8 and 1.6 mol/L ADMA, respectively. Calibration curves in
serum were obtained using concentrations of ADMA, SDMA, NMMA at 0.2, 0.4, 0.8,
1.6, 3.2, 32 M and of Arg and Cit at 1, 25, 50, 100, 250 M. The linearity of calibration
curves in plasma was estimated by the coefficients of correlation (r2), which ranged
from 0.987 to 0.999. The standard curves for serum asymmetric dimethylarginine was
linear within the range of 0.2-32 mol/L. The equation for calibration was y=0.943x +
7.469 and R2=0.992. Total run time was 5 minutes. Recovery was found to be between
90-105%. Limit of detection and limit of quantification were 0.1 and 0.25 mol/L
for ADMA, respectively. The intra- and inter-assay CV values were below 20% for
SDMA; LNMMA, arginine and citrulline.
Conclusion: Satisfactory characterization, stability of the label during chromatography
as well as mass spectrometry, standardization of commercially available as well
as of self-synthesized stable-isotope labeled analogs of analytes, and final added
concentration of the internal standard in the matrices being analyzed is essential and
crucial for reliable
quantitative analysis. Data from calibration curves and method validation reveal that
the method is accurate and precise. The short and fast run time, the feasibility of high
sample throughput and the small amount of sample required make this method very
suitable for routine analysis in the clinical setting.

A-392
Comparison of Q-TOF Acquisition Modes for Quantitative Analysis of
Tetrahydrocannabinol Metabolite

H. Park1, N. Chindarkar2, R. Fitzgerald2. 1Sungkyunkwan University School

of Medicine, Seoul, Korea, Republic of, 2University of California-San
Diego, San Diego, CA
Background & Objectives:
In clinical chemistry laboratories, unit resolution tandem mass spectrometry (MS)
platform is considered the gold standard for quantitative analysis of small molecules.
High resolution mass spectrometry (HRMS) is also a viable option with higher
specificity and retrospective full scan analysis capability. The previously limited
quantitative capability of HRMS has improved, such that both high performance
screening and acceptable quantitation may now be feasible. We evaluated the
quantitative performance of a quadrupole-time of flight (Q-TOF) analyzer for urine
11-nor-9-carboxy tetrahydrocannabinol (THC-COOH). Three Q-TOF acquisition
modes MSE, Q-TOF, and Q-TOF with enhanced product ion sensitivity (EPIS) were
optimized and compared to unit resolution tandem MS.
In the Q-TOF mode, the quadrupole selects a precursor mass followed by collision
cell fragmentation and TOF full scan analysis of product ions. Q-TOF-EPIS mode
is the same except product ion sensitivity is increased by limiting the TOF scan
range around a set m/z value. MSE works in dual scan mode to collect TOF full scan
data with low collision energy (precursor ion scan) and high collision energy/ramp
(fragment ion scan). The quadrupole is RF only in MSE mode.
Experimental: Waters Acquity UPLC with 1)Xevo G2 Q-TOF and 2)Xevo TQ-S
were used with a Phenomenex (Kinetex, C8, 2.6m, 50x2.1mm) column. Mobile
phases A/B were 5mM ammonium formate (pH 3)/0.1% formic acid in water/
acetonitrile respectively (flow 0.4mL/min). The gradient was 30-90% B in 4min.
During solid phase extraction 2mL of urine, internal standard (THC-COOH-d3) and
diluent were applied on Phenomenex Strata-X-Drug B cartridges. Eluates were dried
and reconstituted in 100uL of mobile phase of which 10uL was injected on UPLC.
Calibration was carried out by using 5 calibrators over a range of 12.5-200ng/mL
spiked in drug free urine. All MS were used in negative mode, with optimized settings
of cone voltage (40V), capillary voltage (2.5V), desolvation temperature (500C) and
desolvation gas (1000L/h).
Results and Conclusion:
Each Q-TOF mode was first optimized for ionization of THC-COOH. We compared
the signal intensity of THC-COOH in each mode by acquiring data by infusing pure

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S115

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

THC-COOH in methanol and by analyzing patient urine specimens that screened
positive for THC metabolites by immunoassay. These patient specimens and a CAP
PT sample were analyzed for this study. The concentration range of 13.2-175.8ng/
mL was observed among patient specimens. We found excellent correlation of THCCOOH concentrations among all TOF modes and the unit resolution tandem MS.
Q-TOF-EPIS was 2.4 fold more sensitive than Q-TOF mode. Of the TOF modes,
MSE provided the maximum sensitivity for quantitation (m/z 343.2 for MSE and
343.2/299.2 for Q-TOF-EPIS). MSE mode was approximately 2.5 fold more sensitive
than the Q-TOF-EPIS mode but was less sensitive than unit resolution tandemMS. Our data indicate that quantitative analysis on the Xevo G2 Q-TOF compares
favorably with traditional unit resolution MS but is less sensitive. Future experiments
will determine if the high mass resolution achievable with the TOF can be used to
shorten chromatographic run times without sacrificing specificity.

A-394
Ultrafast, high-throughput quantitative analysis of creatinine in serum by laser
diode thermal desorption coupled to tandem mass spectrometry

J. Lacoursiere, P. Picard, A. Birsan. Phytronix, Quebec, QC, Canada

Background: Serum cretatinine is used as an important indicator of renal health.
Elevated level of creatinine in serum is a key clinical biomarker of impaired kidney
function in humans. In some clinical or pre-clinical studies, only small amounts (L)
of serum are available for creatinine analysis. Several analytical methods are available
for creatinine detection and quantification, but require large serum volume or take
more than 3 minutes per sample analysis. An alternative approach to the analysis of
creatinine enabling high- throughput analysis and needing low serum volumes was
developed, using the laser diode thermal desorption (LDTD) coupled to a tandem
mass spectrometer (MS/MS).
The objective of this work is to analyze multiple real patient samples using the
LDTD-MS/MS and cross validate the results against known traditional methods for
the detection and quantification of creatinine in serum.
Methods: Sample preparation consisted of a protein precipitation extraction by adding
190 L of IS solution (2000 ng/mL d3-Creatinine) in acetonitrile to 10 l of serum.
After vortex-mixing and centrifugation (2 min at 14000 rpm), a 2 L aliquot was
deposited in a LazWell plate, and allowed to dry at room temperature. Creatinine is an
endogenous compound and calibration curve cannot be made into the serum matrix.
Standards were prepared into water and treated similarly to real sample. A 2l aliquot
of the final extracts was deposited into Lazwell plate and dried completely before
analysis. The LDTD laser power was ramped from 0 to 65% of maximum power
in 6 seconds and maintained 1 second at 65%. The mass spectrometer is operated
in negative ionization MRM mode, monitoring transitions 112 -> 41 for creatinine.
Results: A simple high-throughput protein precipitation method for creatinine
analysis in serum was developed and validated. The optimization of instrumental
parameters and a method application will be presented. The method demonstrated
a linear dynamic range over two orders of magnitude, between 0.04 and 4 mg/dL.
Standard curves of creatinine spiked serum extracted using this method shows good
linearity (R2 between 0.9995 to 0.9992 over the quantification range. Three levels
of QC samples were prepared by spiking known creatinine levels in serum for the
validation tests. The endogenous concentration of creatinine was determined before
the QC spiking addition. Sum of endogenous concentration and spiked concentration
were used as nominal concentration. The accuracy measured of QC samples ranged
from 92.9 to 99.9%. The quality controls had a precision error (% CV) of less than
15 % for inter- and intra-day assay. The use of negative ionization mode gave better
signal intensities and eliminated interferences from the serum extract giving no false
positives .The cross validation study with a traditional method for the analysis of
creatinine confirmed the effectiveness of this new analytical approach using the
LDTD-MS/MS. Analysis time (8 sec per sample) as well as sample throughput are
significantly improved.
Conclusion: The LDTD-MS/MS method is an effective tool for the quantitation of
creatinine at a rate of 8 seconds per sample to support preclinical and clinical studies.

A-396
Analysis of Pain Panel Medications in Urine on Raptor Biphenyl by LC-MS/
MS

F. Carroll. Restek Corporation, Bellefonte, PA

The Restek Biphenyl has been the column of choice for clinical diagnostic and Pain
Management drug screening testing because of its ability to provide highly retentive,
selective, and rugged reversed-phase separations of drugs and metabolites. By

S116

bringing the speed of Superficially Porous Particles to the Biphenyl family, Resteks
Raptor Biphenyl provides clinical labs with an even faster option for a wide
variety of clinical assays. Drug screening applications can be difficult to optimize
and reproduce due to the limited selectivity and ruggedness of the analytical column.
The Raptor Biphenyl has been engineered to be rugged and selective with a pain
management analyses that can be performed with a 5-minute cycle time and complete
isobaric resolution. The Raptor Biphenyl beats popular competitor methods in both
selectivity and performance.
Comparison analyses were performed on the Raptor Biphenyl 2.7m, 50 x 3.0mm
and competitor phenyl hexyl and C18 columns. Each manufacturers own optimized
method conditions were used in this evaluation. A pain panel drug standard was
prepared in diluted urine and injected for assessment of retention and resolution. The
ruggedness of the Raptor Biphenyl column was tested by performing a minimum
of 2500 injections of a minimally diluted urine standard on a single column with the
guard cartridge changed every 1000 injections. Retention time and response were
monitored throughout the experiment. Analysis for both experiments were performed
on a Shimadzu UFLC-XR HPLC equipped with an ABSCIEX 4000 LC-MS/MS using
electrospray ionization in positive ion mode.
The Raptor Biphenyl displayed increased retention over the competitor phenyl
hexyl and C18 columns. By improving the separation of the early-eluting compounds
such as morphine, oxymorphone, and hydromorphone from hydrophilic matrix
interferences, ion-suppression was decreased resulting in an increase in sensitivity.
In addition, isobaric compounds such as morphine and hydromorphone display
increased resolution and response. The Raptor Biphenyl has also proven to be
rugged under high through-put conditions. In the first lifetime experiment a pain panel
drug standard was diluted in urine 6x and filtered through 0.2 m PVDF Thomson
filter vials. The filtered matrix standards were injected on a single column with the
replacement of the guard cartridge every 1000 injections. Under these conditions, the
column lasted through 3000 injections when the study was ended. A second lifetime
study was executed for 2500 injections using a new column and the same interval
for guard cartridge replacement however the matrix standards did not receive any
filtration. All 2500 injections were completed without a drastic change in response
or peak shape. The Raptor Biphenyl 2.7m, 50 x 3.0mm column has proven to
withstand over 2000 injections of matrix samples regardless of filtration. It is the
recommendation of Restek that with guard cartridge changes every 500 injections the
column can last up to 3000 injections or beyond.

A-397
A reduced workflow SPE- LC-ESI-MS/MS method to distinguish healthy from
elevated concentrations of metanephrine and normetanephrine in patient
plasma samples

F. A. Kero1, J. Ye2, T. Enzweiler1, V. Vandell1, E. Gairloch1, E. Majdi2, H.

Qiao2, L. Cousins2. 1Biotage, Charlotte, NC, 2Ionics, Bolton, ON, Canada
Background: Adrenal neuroendocrine tumors known as pheochromocytomas induce
excessive production of catecholamines in mammalian blood and urine. The salient
metabolites, metanephrine and normetanephrine, are routinely screened as biomarkers
for this condition in both matrices. The bottleneck of these analytical methods has
traditionally been laborious sample preparation methods that mitigate the variability
in matrix inherent with patient samples. Additional issues include the complexity of
the measurement that challenges detectors that lack sensitivity and robustness. This
report details a load, wash, elute weak cation exchange solid phase extraction
procedure amenable to both plasma and urine samples. The extracts are subsequently
injected into an LC-MS/MS system. The preliminary sample preparation method was
developed at the Biotage US Applications lab (Charlotte, NC).The method was then
transferred to Ionics (Bolton, ON, Canada) to facilitate the nmole/L measurements of
the selected biomarkers by laminar flow tandem mass spectrometry. The SPE-LC-ESIMS/MS method parameters were first optimized using pooled mixed gender plasma.
A set of patient samples (n=32) was later supplied by the Mayo Clinic (Rochester,
MN) that had previously been analyzed by a validated referee method. The population
represented measured values across a range of clinical relevance.
Methods: Plasma samples were processed using a Biotage PPM96 positive pressure
manifold. Plasma samples (100 microliters ) were diluted with 50 mM ammonium
acetate (300 microliters). The plasma samples were then loaded onto a Biotage
EVOLUTE EXPRESS 30mg WCX 96 well plate. The plates did not require
conditioning or equilibration steps. The samples were sequentially washed with
50/50% MeCN/MeOH and H2O. The analytes were eluted with 47.5/47.5/5% MeCN/
MeOH/formic acid. Samples were then dried down using a Biotage SPE Dry nitrogen
evaporator. The reconstituted extracts were analyzed using a Shimadzu LC system in
tandem with an Ionics 3Q 220 triple quadrupole mass spectrometer.
Results: Pooled plasma samples yielded >80% recovery for both analytes. Analyte

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

suppression was determined to be <10%. The patient data obtained by this reduced
workflow method compared well to historical data obtained by the validated referee
method. The method LOQ was 0.1 nmole/L for normetanephrine and 0.05 nmole/L
for metanephrine.
Conclusion: It is anticipated that this time saving and sensitive SPE-LC-ESI-MS/
MS method will have significant impact in population screening strategies for these
metabolites.

A-399
Sensitive LC-MS/MS assay for detecting testosterone in female, pediatric and
male serum

C. Grote, S. Erfurth. PeaceHealth Laboratories, Springfield, OR

Introduction: Testosterone (Te) measurements are widely used to assess steroid
hormone status in children and adults of both sexes. Immunoassays are often not
sensitive enough to determine the lower Te concentrations found in children and
females. An LC-MS/MS method was developed that provides a single assay with the
necessary accuracy (R>0.97), precision (CV<10%) and sensitivity (LOQ<5 ng/dL)
to routinely measure low concentrations of Te in children and females and higher Te
concentrations in males.
Method: Serum (200 L) extraction involves protein precipitation with 0.1 M
zinc sulfate, SPE with Bond Elut Plexa columns and derivatization with 25%
hydroxylamine. The chromatographic system consists of a Zorbax Eclipse Plus-C18
guard column, Zorbax Eclipse XDB-C18, 2.1x30 mm, 1.8 micron analytical column
and a mobile phase comprised of A: 0.1% formic acid and 1 mM ammonium formate
in water and B: 0.1% formic acid in acetonitrile. MRM transitions for qualifier and
quantifier ions were monitored and a deuterated Te internal standard was added to
each calibrator and specimen. Calibrators (1.0 to 2000 ng/dl Te) were prepared in 2%
BSA. LC-MS/MS instrument included an Agilent 1260 HPLC Series SL binary pump,
vial plate auto sampler with thermostat, temperature controlled column compartment
and an Agilent 6460 QQQ with JetStream Technology and ESI source.
Validation: Ion suppression was evaluated by injecting extracted serum from
two low (13 and 53 ng/dL) serum pools while monitoring sixteen phospholipid/
lysophospholipids and Te MRM transitions. No phospholipid interference was
observed at the Te retention time. Structurally related compounds (androstenedione,
nandrolone, cortisol, corticosterone, 11-deoxycortisol, DHEA, progesterone, 5-alphadihydrotestosterone, 17-alpha-hydroxyprogesterone, 17-alpha-methyltestosterone
and aldosterone) were evaluated for interference. An isotopic molecular ion of
androstenedione monooxime, an isobar of Te, increased the concentration of Te by
0.8%. DHEA, an isomer of Te, produced a product ion transition of Te at a retention
time of 0.81 minutes and does not interfere at physiologic concentrations. The average
absolute and extraction recovery determined with pooled serum at 13 and 53 ng/dL
was 100.4% and 78.2% respectively. Intraday and interday imprecision (CV) using 5
serum pools between 14 and 1031 ng/dL were < 4% and < 6% respectively. The LOQ
determined by analyzing Te calibrators was 1 ng/dL. Extracted Te calibration curves
from 1.0 to 100 ng/dL using a linear curve fit and 1.0 to 2000 ng/dL using a quadratic
curve fit showed correlation coefficients of R=0.9996 and R=0.9998 respectively.
Correlation of Te serum levels for females and children (n=88) with a national
laboratory showed an R=0.974 and y=0.975x - 1.342 and combined serum levels for
children and adults of both sexes (n=171) showed R=0.984 and y=1.069x - 1.495.
Conclusions: The method described is a highly sensitive (LOQ=1 ng/dL) and specific
LC-MS/MS method suitable for analysis of serum Te in females, children and males.
The formation of a Te oxime derivative allows use of a short 30 mm C18 column
with a total run time of 3.0 minutes. The analytical protocol is free of cross reactivity,
interference from structurally related steroids and phospholipids.

A-400
Application of an Immunocapture-LC-MS/MS Insulin Analogue Method to
Clinical and Postmortem Insulin Investigations

S. L. Wong1, G. Van Der Gugten2, D. T. Holmes2. 1University of British

Columbia, Vancouver, BC, Canada, 2St. Pauls Hospital, Vancouver, BC,
Canada
Background: Synthetic insulins, or insulin analogues, are routinely prescribed in
type 1 and advanced type 2 diabetes mellitus. Detection and quantitation of synthetic
insulins are useful in confirming accidental or intentional overdoses of insulin
analogues, as well as in the workup of aberrant insulin results.
In most laboratories, insulin concentration is determined by immunoassay. Insulin

immunoassays exhibit variable cross-reactivities to synthetic insulins, are prone to

interferences from insulin precursors/degradation products/autoantibodies, and fail to
provide information on the type(s) of analogue(s) present. Given these limitations,
we have developed an LC-MS/MS method for the identification of 5 commonly
prescribed pharmaceutical insulins.
Methods: Twenty-five microliters of 500 U/mL of bovine insulin (internal standard),
and 5 L of 5 g/L dextran sulfate + 0.5 M MgCl2, were added to 1 mL of patient serum
or calibrator (insulin analogues in 20% acetonitrile) in Eppendorf LoBind tubes. The
mixtures were incubated with 500 L of paramagnetic beads coated with monoclonal
mouse anti-insulin antibodies at room temperature for 1 hour. The beads were washed
3 times with 1 mL 0.01M PBS, and the insulin analogues extracted with 2 x 100 L
1% acetic acid into a BSA-treated 96-well plate. Chromatographic separation was
achieved with an ACE 300 C18 column (5 x 2.1 mm, 5 m ID) with a run-time of
8.5 minutes. The samples were analyzed on an AB SCIEX QTRAP 5500 system in
positive ESI mode, with MRM transitions monitored for regular insulin (qualifiers
m/z 1162/226, 1162/652; quantifier m/z 1162/345), lispro (qualifier 969/217;
quantifier 1162/217), aspart (qualifiers 972/226, 1166/219; quantifier 972/136),
detemir (qualifiers 987/454, 1184/357; quantifier 1184/454), and glargine (qualifiers
1011/1164, 1011/1179; quantifier 867/136). Calibration curves were constructed with
linear regression using 1/x weighting.
Results: The method demonstrated good linearity over a concentration range of
5-200 U/mL, with R >0.997 for all insulin analogues. Analytical recoveries were
between 90.3% and 113.4%. Approximate LLOD and LLOQ were 3.5 U/mL and 5
U/mL, respectively. Within-run CVs ranged from 3.2% to 14.8%. The utility of the
method was shown in a series of case studies: (a) postmortem investigations of deaths
secondary to suspected insulin overdose (b) clinical workup of a type 1 diabetic patient
who presented with hypoglycemia, and questions of whether excess synthetic insulin
was administered deliberately, and (c) confirmation of insulin concentration in an
insulinoma patient with discrepant insulin results from 4 commercial immunoassays.
Conclusion: We have developed a robust LC-MS/MS assay for the quantitation of
5 popular insulin analogues. It is valuable in the workup of insulin-related clinical
and forensic cases, and has overcome some of the limitations exhibited by current
commercial insulin assays.

A-401
Ultra-sensitive simultaneous LC-MS/MS quantification of human insulin,
glargine, lispro, aspart, detemir and glulisine in human plasma using 2D-LC
and a novel high efficiency column: method development and application in an
overdose case

E. E. Chambers1, K. Fountain1, T. Rosano2, C. Kim2, J. Desemone2. 1Waters

Corporation, Milford, MA, 2Albany Medical College, Albany, NY
Background: Recombinant human insulin and its analogs represent the primary
treatment for insulin dependent Type I and Type II diabetes patients. Many
these insulin formulations have or will be coming off patent shortly, generating a
tremendous interest in quantitative methods for pharmacokinetic and bioequivalence
assessments. In addition, there is interest in insulin quantification from an antidoping and forensic perspective. Historically, insulins have been analyzed using
radioimmunoassay (RIA) or enzyme-linked immunosorbent assay (ELISA). LCMS analysis of insulins is needed due to the many shortcomings of these ligandbinding assays, chiefly: lack of standardization, cross-reactivity, limited linear
dynamic range, and sample preparation time. Hybrid assays (affinity capture + LCMS) have been the most effective though they lack the simplicity and throughput
required for routine testing and bioanalysis. This work provides single method
for quantification of intact human insulin and 5 insulin analogs in human plasma.
This investigation solves both the selectivity and sensitivity problems encountered
for accurate quantification of insulins in plasma since the former is not possible
with conventional assays and the latter with conventional LC-MS/MS. We then
retrospectively apply the method to a unique case of insulin glargine and aspart
overdose which required prolonged dextrose infusion to prevent hypoglycaemia.
Methods: Blood samples were collected when the patient was admitted and
approximately every 5 hours thereafter until symptoms no longer presented.Plasma
samples are prepared using protein precipitation to remove high abundance proteins,
followed by mixed-mode strong-anion exchange SPE to selectively eliminate
closely related interferences and provide orthogonal ity. The multidimensional
LC system includes at-column-dilution and trap and back elute components
which improve sensitivity (through increased loading) and selectivity (cleanup achieved during the trapping phase.) Insulins are separated with formic acid
in water (A) and ACN (B) on a superficially-porous, charged surface column
packed with sub-2 m particles using a linear gradient from 15 to 40% B.
Results: Method LODs of 50-200 pg/mL were achieved for each insulin. Average

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S117

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

accuracy for standard curve points was 99-100%. Average inter-and intra-day
accuracies for QCs samples were 98 and 94 %, respectively. Average inter- and intraday precisions were 7.5 and 5.3 %, respectively. Matrix factors were calculated in 6
sources of human plasma and CVs of matrix factors for all analogs were <15 %. In
addition, the presence of artificially high human insulin did not affect quantification of
any of the analogs. Samples from two over-dose incidents were quantified. Dextrose
infusion was required for 110 and 96 hours in the two cases.M1 metabolite and
aspart were detected up to 90 and 22-29 hours, respectively, after admission. Higher
levels of glargine M1 metabolite correlated to higher rates of dextrose infusion.
Conclusion: This method represents a single, simple method for the simultaneous, direct
quantification of intact human insulin and analogs in human plasma which achieves
detection limits in the 50 pg/mL (8.6 fmol/mL) range. This assay was successfully applied
to quantify glargine, its metabolite, and aspart in samples from two overdose cases.
Disclaimer: This method is intended for clinical research use only, not for use in
diagnostic procedures

A-402
Analytical and Clinical Validation of an LC-MS/MS Method for Urine
Leukotriene E4: a Marker of Systemic Mastocytosis

J. W. Meeusen, A. Lueke, A. Gray, L. J. Donato, J. H. Butterfield, A. K.

Saenger. Mayo Clinic, Rochester, MN
Background: Systemic mastocytosis (SM) is a disorder that results in excessive
accumulation of clonally derived mast cells in various tissues. When triggered, these
mast cells release large amounts of histamine, prostaglandins and leukotrienes. This
release of signal molecules causes intermittent spells with symptoms of itching,
flushing, lightheadedness, tachycardia, gastrointestinal distress, or even loss of
consciousness. Diagnostic criteria include the presence of mast cells in tryptasestained biopsy sections (typically bone-marrow) plus one of the following: abnormal
mast cell morphology; KIT Asp816Val mutation, CD25 positive mast cells, or serum
tryptase > 20 ng/ml. Urine concentrations of N-methyl histamine (NMH) and 11-beta
prostaglandin F2 (BPG), the primary metabolites of mast cell derived histamine and
prostaglandin, can aid in screening, reduce unnecessary biopsies and guide therapy.
However, NMH and BPG lack sensitivity. Leukotriene E4 (LTE4) is the primary stable
metabolite of total cysteinyl leukotrienes. We hypothesized that secretion of LTE4
would be increased in SM and could be used alone or in combination with NMH and
BPG to optimize screening for SM. Here we describe a novel liquid chromatographytandem mass spectrometry assay to accurately and precisely quantitate LTE4 in urine
and outline its clinical utility in SM screening.
Methods: D5-labeled internal standard was added to urine specimens and followed
by an acetonitrile precipitation before injection into a Turboflow MAX mixed-mode
anion exchange column with subsequent chromatographic separation using a C8 2.5m analytical column. LTE4 was measured in negative ion mode using an AB Sciex
API 5000. All LTE4 concentrations were normalized to urine creatinine (enzymatic
method, Roche).
Results: Intra-assay precision (%CV) determined in pooled urine specimens ranged
from 2.6% to 5.0% at mean LTE4 concentrations of 41, 631, and 1452 pg/mL (n=20).
Inter-assay %CV determined over 20 days ranged from 6.9% to 8.2% at mean LTE4
concentrations of 44, 445, and 1380 pg/mL. Linearity was determined between 0-2000
pg/mL and the mean recovery from mixing studies performed in triplicate was 100%
(y=1.01x-2.28, r2=0.9985). Limits of detection and quantitation were determined at 2
and 8 pg/mL, respectively. The normal reference range of <104pg/mg creatinine was
determined based on the 95th percentile in a cohort of 128 healthy individuals.
Clinical performance was determined in 409 patients referred for clinical evaluation.
SM was diagnosed in 66 (16%) patients according to World Health Organization
criteria. Clinical sensitivity was 53% for BPG (>1000ng/24h) and 71% for NMH
(>200g/g creatinine) in our cohort of 409 symptomatic patients. Sensitivity improved
to 86% with a specificity of 68% when BPG or NMH were both considered. Including
LTE4 (>104pg/mg creatinine) improved the SM diagnostic sensitivity to 97%, with
minimal change in specificity (61%).
Conclusion: We have developed a sensitive and precise LC-MS/MS assay for
quantitation of LTE4 in urine. This assay has significant potential utility as a useful
screening marker of SM, greatly improving screening sensitivity when used in
combination with other biomarkers of mast cell activation. Incorporating LTE4
into a panel including BPG and NMH provides a much needed screening tool for a
complicated disease with non-specific symptoms and invasive confirmatory testing.

A-403
Addition of solid phase extraction to opiate sample preparation for UPLC-MS/
MS

H. F. Mohammad1, P. J. Bates1, C. A. Hammett-Stabler2. 1University of

North Carolina Hospitals, Chapel Hill, NC, 2Department of Pathology and
Laboratory Medicine, University of North Carolina, Chapel Hill, NC
Background: Opiate testing has increased in parallel with the rise in patients seeking
pain management care. These often prescribed drugs must be monitored due to their
potential toxicity. Since qualitative immunoassay screening methods fail to distinguish
specific opiates/opioids, clinical laboratories have turned to liquid chromatography
coupled with tandem mass spectrometry (LC/MS/MS). Sample preparation is a
critical step in LC/MS/MS analysis. Currently our samples are prepared using dilute
and shoot (D/S) after digestion with -glucuronidase. The goal of this study was to
assess the inclusion of an additional step involving solid phase extraction (SPE) in the
sample preparation.
Methods: Drug free urine spiked with buprenorphine, norpubrenorphine, codeine,
hydroxycodone, oxycodone, morphine, hydromorphone, oxymorphone, and
6-monoacetylmorphine (Cerilliant Corp) at known concentrations and previously
tested patient samples were used in this study. In brief: 100 l of urine was incubated
with -glucuronidase at 50 C for 1 h, and centrifuged. For the D/S method, 100
l of the enzyme-digested sample was diluted 1:1 with water prior to analysis. For
the SPE method, Oasis MCX Elution plates (Waters Corp, Milford, USA) were
pre-conditioned with 200 l methanol and 200 l 5% formic acid. 100 l of the
enzyme-digested sample was applied and washed with 5% methanol in 5% formic
acid solution. Samples were then eluted with 100 l of 5:90 methanol:acetonitrile
containing 5% of NH4OH solution. The elutes were dried under N2 and reconstituted
with 100 l water and analyzed. Opiate measurements were obtained using UPLC/
MS/MS (ACQUITY UPLC/MS, Waters Corp) as described previously (1). Data were
analyzed using EP Evaluator software (Data Innovations, LLC).
Results: The addition of the SPE step increased total analysis time by 10-15 minutes.
We did not achieve a significant change in assay performance in terms of the
analytical measuring range and linearity. The response curves for each of the nine
drugs and metabolites were linear from 50-1000 ng/ml. However, recovery improved
for all analytes; but particularly for 6-monacetylmorphine, hydroxymorphone, and
norbuprenorphine at the lower concentrations where we saw improvement from
71-87% to >94% with the addition of SPE. Method comparison data using linear
regression demonstrated good agreement with all correlation coefficients (R2) > 0.95.
Conclusion: The addition of SPE to our sample preparation increases cost by ~$3/
sample and turnaround time by 10-15 minutes. However, we saw improved recovery
across all analytes which may be attributed to a reduction in interferents from the
sample matrices. This in itself is desirable as cleaner samples extend column life and
reduce maintenance. Thus we believe the benefits of this additional step outweigh the
costs and time.
1. Bates PJ, et al. Simultaneous detection of nine opiates, including buprenorphine and
norbuprenorphine in urine using UPLC-MSMS. www.msacl.org

A-404
Using intact immunoglobulin light chains to quantitate rituximab by mass
spectrometry

J. R. Mills, D. R. Barnidge, A. B. Girtman, P. M. Ladwig, D. L. Murray, M.

R. Snyder. Mayo Clinic, Rochester, MN
Background: The therapeutic monoclonal immunoglobulin rituximab is used to
deplete CD20+ B-cells in the treatment of lymphoproliferative and autoimmune
disorders. In autoimmune diseases, the exact mechanism by which B-cell depletion
modulates disease activity is unknown and instances of clinical relapses can occur
in the absence of detectable B-cells. The lack of methods to measure rituximab has
restricted clinical studies aiming to better understand how drug levels relate to disease
relapse.
Objective: Recently we demonstrated that microLC-ESI-Q-TOF mass spectrometry
is capable of detecting residual levels of monoclonal light chains in human serum.
We hypothesize this technology would be able to quantitate rituximab light chains
in patient sera as a diagnostic tool, eliminating the need for tryptic digestion. Here
we compare intact light chain quantification using microLC-ESI-Q-TOF MS to our
current method using proteotypic SRM quantification on a triple quadrupole MS.
Methods: Rituximab intact light chain (iLC) quantification was performed on
immunoglobulin-enriched serum reduced with DTT, separated on an Eksigent Ekspert
liquid chromatography system, and analyzed on an ABSciex 5600 Triple TOF mass

S118

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

spectrometer. The peak area for the rituximab iLC (molecular mass - 23,034 Da)
was found by integrating the molecular mass peak observed after deconvolution of
the summed mass spectra from the rituximab elution time. The therapeutic mAb
infliximab was added to each sample prior to Ig-enrichment as an internal standard.
For the proteotypic quantitation, peptides unique to rituximab heavy chain (HC) and
light chain (LC) variable regions were quantified by SRM from ammonium sulfatecrashed serum that was reduced, alkylated and digested with trypsin at 37oC for 12h.
A proteotypic peptide from horse IgG was used as an internal standard and stable
isotope labeled peptides were added to monitor retention times. Tryptic peptides were
separated using a Thermo TLX-2 system then analyzed on an ABSciex API 5000
triple quadrupole mass spectrometer. Linearity, LOD, LOQ, intra-assay precision
were assessed for both assays using rituximab spiked into human serum. For both
methods, 6-point standard curves were generated [0-100 ug/mL] by spiking known
amounts of rituximab into pooled human serum.
Results: Linearity was established by performing serial dilutions in human serum
(100-1.0ug/mL; R2=0.99). The rituximab iLC molecular mass peak area was
detectable above the polyclonal immunoglobulin background with an LOD of 1.2 ug/
mL and an LOQ of 2 ug/mL. Intra-assay precision was 6.7% at 100 ug/mL and 16.7%
at 2ug/mL. We have established rituximab HC and LC proteotypic peptides have an
LOQ ~2-fold lower. A method comparison using weighted linear regression between
Intact and light or heavy chain peptides were comparable ([slope =0.99, y-intercept
=-0.04 and R2>0.99] and [slope =0.99, y-intercept =-0.03 and R2>0.99], respectively).
Conclusion:Measurement of rituximab iLCs based on accurate mass assessment is
a viable analytical approach. Quantitation of iLCs compares well to a proteotypic
peptide approach, although differences in the analytical sensitivity of the methods
may exist. Further studies using this methodology are warranted to understand how
rituximab levels correlate with disease relapse.

A-405
Optimization and evaluation of an isotope dilution liquid chromatography
tandem mass spectrometry method for the determination of total cholesterol in
human serum and a comparison with field methods

H. Lee1, S. Park1, E. Ahn1, S. Park1, G. Lee2. 1Korea Research Institute of

Standards and Science, Daejeon, Korea, Republic of, 2Chungnam National
University, Daejeon, Korea, Republic of
BACKGROUND: Abnormal cholesterol levels are strongly associated with
cardiovascular disease because these promote atheroma development in arteries.
Lowering total cholesterol reduces the risk of coronary heart disease. For the
reference intervals of cholesterol levels in blood are narrow, accurate methods of
analysis for cholesterol in serum are essential, and the establishment of a reference
and a definitive method has been needed to evaluate the field method. Isotope
dilution gas chromatography-mass spectrometry has been accepted as a definitive
method. Although ID GC-MS is considered to be highly accurate, it contains complex
procedures such as derivatization. We describe simple ID LC/MS/MS as another
method for the determination of total cholesterol in serum and its optimization
conditions here. In order to compare with field method, proficiency testing programs
were performed on the base of this method.
METHODS: Human serum samples were obtained from pooling of healthy human
serum free from HIV, HCV, and HBV. Cholesterol-d4 was used as an internal standard.
0.1 mL of serum sample was taken into an amber vial. An appropriate amount of
isotope standard solution was spiked into the sample vial to make a 1:1 weight ratio.
After adding 0.6 mL of an aqueous 8.6 mol/L KOH solution and 4 mL ethanol, we
heated the solution in thermomixer for 3 h at 70 to hydrolyze the cholesterol esters.
After the solution cooled, we added 5 mL of water and 10 mL of hexane, shook the
tube for 5 min, separated the hexane phase, evaporated the solvent under vacuum, and
dissolved the residue with 1 mL of ethanol. Aliquot of the ethanol solution was filtered
and then analyzed by LC/MS/MS. The LC column was C18, and kept at 50 during
the chromatographic run. The mobile phase was methanol containing 0.1% acetic
acid, and the flow rate was 0.3 mL/min. Serum samples were distributed to about 180
clinical laboratories in korea for the comparison with field methods.
RESULTS: The optimum volume of KOH solution for hydrolysis of cholesterol ester
was about 5% of total sample mixture, and reaction time was 3 h. The optimized
ID LC/MS/MS method was verified through the measurement of NIST SRM 909b
and the participation in key comparison, and showed good agreement with the
SRM values. The pooled serum samples were certified by this method, and used as
materials for the proficiency testing programs. Expanded uncertainty of certification
was about 2% within the 95% confidence interval. Proficiency testing programs of
field laboratories have shown some discrepancies of 6.2% CV and 5.3% CV in total
cholesterol results among the laboratories.

CONCLUSION: An optimized ID LC/MS/MS method was proposed as another

method for the determination of total cholesterol in serum. This method was
verified through comparison with the NIST SRM. We developed the two levels of
total cholesterol CRM on the basis of this method and used them as materials for
proficiency testing programs. Through the proficiency testings, it was possible to view
the state-of-the-art of total cholesterol measurement by field laboratories.

A-406
LC-MS/MS Method For The Detection Of Free Thyroxine And Free TriIodothyronine Using The Ionics 3Q 320 Triple Quadrupole Tandem Mass
Spectrometer

V. Gounden, Z. Yuan, S. J. Soldin. National Institutes of Health, Bethesda,

MD
Background: The majority of routine clinical laboratories perform free thyroxine
(FT4) and free tri-iodothyronine (FT3) measurements on immunoassay (IA)
platforms. These IAs are affected by changes in binding protein concentrations, have
a weak inverse linear log relationship to TSH in hypo- and hyperthyroid individuals
and have poor performance at the upper & lower values of the reference intervals. The
gold standard for free thyroid hormone analysis involves preparation of sample using
equilibrium dialysis. This is a time consuming and technically difficult technique.
However liquid chromatography-tandem mass spectrometry (LC-MS/MS) following
ultrafiltration of the sample at 37 C (method as previously described by Gu et al.
Clin Biochem. 2007;40;1386-1391) has been shown to perform better than IA in
the above described circumstances and involves a simpler more convenient sample
preparation than equilibrium dialysis. Our objective was to improve on the sensitivity
of this initial method. Here we describe our 3rd generation LC-MS/MS method with
improved sensitivity over the initial mass spectrometry method.
Method: Sample preparation was performed by ultrafiltration of 500ul of serum
using a 30-kDa centrifugal filter (Centrifree YM-30, Millipore). Following addition
of sample to filtering device samples were centrifuged in a temperature controlled
centrifuge at 1113g for 30 minutes at 37 C. 150 ul of the ultrafiltrate was added
to 450ul of methanol containing deuterium-labeled internal standards (IS) for FT4
and FT3 and centrifuged. 350 uL of the supernatant was further diluted with water,
vortexed and 200 uL injected into LC-MS/MS. FT4, FT3 were detected by electrospray
ionization in negative mode with the following transitions: FT4 775.6>126.7 and FT3
649.9>126.7. LC-MS/MS setup consisted of a Shimadzu UFLCXR HPLC system
interfaced to a Ionics 3Q 320 triple quadrupole tandem MS. Chromatographic
separation was performed using a Poroshell 1.7m C18 column (100mmx2.1mm)
with a gradient mobile phase (A: 2% Methanol in water containing 0.01% acetic acid;
B: 98% Methanol, at a flow rate of 0.5mL/min. Run time per injection was 13 minutes.
Results: The method described displayed good linearity over a concentration range of
0-25 pg/ml (FT3) and 0-5 ng/dL (FT4) with r2 >0.995. Between day precision CVs
for across the concentration range were: FT3 4.8-8.8 %; FT4 7.5-7.8%. Lower limit
of Quanititation (LLOQ) at signal to noise ratio(S/N) =10 was 0.2 pg/ml for FT3 and
0.05 ng/dL(S/N=20) for FT4.
MS MS comparison r values with our first generation method were 0.87 and 0.82 for
FT4 and FT3 respectively.
Conclusion: The sensitivity of the 3rd generation FT4/FT3 method described above is
greatly enhanced due to improvements in mass spectrometer and column technology.
LLOQ is now 10 fold lower than that found for previous FT4/FT3 methods.

A-407
A Simple and Robust Targeted Quantitative Method for Insulin and its
Therapeutic Analogs

E. E. Niederkofler1, S. Peterman2, A. Leber1, K. Antwi1, T. Schroeder3, U.

A. Kiernan1, K. A. Tubbs1, B. Krastins2, A. Prakash2, J. Frahm1, M. Lopez2.
1
Thermo Fisher Scientific, Tempe, AZ, 2Thermo Fisher Scientific BRIMS,
Cambridge, MA, 3Thermo Fisher Scientific, Somerset, NJ
Background The measurement of Insulin has been identified as a paramount
metric in clinical research, drug development, forensic toxicology, and sports doping
applications. Conventional insulin assays are plagued by the inability to differentiate
endogenous insulin from exogenous insulin analogs. The use of LC-MS can overcome
this shortcoming; however, the LC-MS methods to-date lack the analytical sensitivity
demanded by the field. Therefore, a highly-analytically selective sample interrogation
workflow is required to address the complexity of plasma samples and, ultimately, for
accurate and analytically sensitive LC-MS detection and quantification. To meet these

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S119

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

requirements, a Mass Spectrometric Immunoassay (MSIA) method was developed
for the high-throughput, analytically-sensitive quantification of insulin and its analogs
from human plasma.
Methods Both neat and plasma samples containing a mix of insulin and its analogs
at various concentrations were analyzed. A heavy version of insulin was used as an
internal reference standard and spiked into each sample prior to target selection. In a
60-minute protocol (per 96 samples) the Thermo Scientific MSIA D.A.R.T.S
pipette tips derivatized with a pan-anti insulin antibody were used for insulin target
selection. After affinity enrichment, MSIA detection and quantification was achieved
in a 10 minute LC-MS method on a Thermo Scientific Ultimate 3000 LC system
coupled to a Thermo Scientific Q Exactive mass spectrometer. Full MS scans
were acquired, thus enabling the full characterization and quantification of multiple
insulin analogs from a single sample.
Results - One of the primary limitations to current insulin assays is the inability to
distinguish between endogenous and exogenous insulin analogs. The immobilized
insulin pan-antibody in the MSIA D.A.R.T.S pipette tips recognizes a common
epitope region in the beta chain that is conserved across all of the analyzed variants,
which allows the capture and detection of all variants from the sample. Further,
utilizing full scan MS mode in the analysis stage of the MSIA workflow enables
simultaneous detection of multiple insulin analogs and the ability to screen for
unsuspected insulin analogs post-acquisition. Accurate intact mass and fragmentation
of the insulin analogs confirmed the identity of each variant.
An additional limitation to high-throughput targeted quantification of insulin and
its analogs are inefficient sample preparation protocols that result in their lack of
analytical sensitivity and robustness. Using the MSIA Insulin workflow described
above, we achieved an LLOQ of 15 pM (87 pg/mL) and an LOD of < 7.5pM (~47 pg/
mL) for the intact variants in plasma. Further, reproducibility studies demonstrated
inter- and intra-day CVs of < 3% and spike and recovery resulted in recoveries of
96-100%. In addition to the improved analytical sensitivity, the MSIA workflow
significantly reduces the background matrix. The reduced complexity affords shorter
LC gradients, and, therefore, shorter LC-MS analysis times. Altogether these results
demonstrate the high analytical sensitivity, reproducibility, and robustness of the
MSIA Insulin workflow in clinical research methodology.
Conclusion A robust clinical research methodology incorporating antibody-directed
target selection from a complex matrix with highly-analytically sensitive LC-MS
detection was developed for the qualitative and quantitative simultaneous analyses
of multiple insulin analogs.

A-408
Development and Validation of a High Performance Liquid Chromatography
Tandem Mass Spectrometry 9 Steroid Panel using Minimal Sample Volume

B. Stolze1, V. Gounden1, J. Gu1, S. J. Soldin2. 1National Institutes of Health

( Clinical Center), Bethesda, MD, 2National Institutes of Health ( Clinical
Center) and Department of Medicine, Georgetown University, Washington
DC, Bethesda, MD
Background and Objectives: Steroid profiles play a critical role in the evaluation
of endocrine disorders. High performance liquid chromatography tandem mass
spectrometry (HPLC-MS/MS), with superior sensitivity, specificity, and simultaneous
multi-analyte quantitation capabilities, is the preferred method for steroid analysis.
Our first generation steroid profile by HPLC-MS/MS simultaneously measured 9
steroids in 18 minutes using 760L of serum. Our second generation reduced the
sample volume to 200L with a 1 mL injection. Our objective was to improve on
our initial methods whilst reducing sample volume and run time. Here we describe
our third generation steroid profile assays, quantifying 9 steroids with a run time
of 11.5 minutes using a 50uL sample volume and 100uL injection volume, while
providing better sensitivity, specificity, and a ten-fold lower limit of quantitation due
to improvements in mass spectrometric and column technology.
Method: An Agilent 6490 triple-quadrupole MS coupled with an Atmospheric
Pressure Photoionization (APPI) source and Agilent 1200 Infinity series HPLC were
used employing isotope dilution with deuterium labeled internal standard for each
analyte. 50uL of serum were deproteinized by adding 75L of acetonitrile containing
internal standards. After centrifugation, 75L of supernatant was diluted with 250uL
of water and a 100L aliquot was injected onto a Poroshell 120 EC-C18 column.
After column washing the steroids were eluted using a methanol gradient as follows:
80% A (methanol: water 2:98, v/v) for 3 minutes, 50% B (methanol: water 98:2, v/v)
to 58% B over 3 minutes, 58% B to 90% B over 1 minute, holding at 90% B for 1.5
minutes, and finally 90% B to 20% B in 0.01 minutes. Quantitation for all 9 analytes
was performed in positive MRM mode. Instrument parameters were as follows: gas
temperature 325 oC, vaporizer 400 oC, gas flow of 11 L/min, nebulizer 60psi, and
capillary 4000V.

S120

The MRM for each analyte and compound dependent parameters are listed below:
Cortisol 363.3/121.1 Collision energy (CE) 26; Cortisone 361.2/163 CE
22; 11-deoxycortisol 347.3/97.1 CE 30; Corticosterone 347.2/121.2 CE 22;
17-hydroxyprogesterone 331.2/109.1 CE 30; Progesterone 315.3/109.1 CE 26;
Testosterone 289.1/109 CE 22, Androstenedione 287.1/97.1 CE 18, 21-deoxycortisol
347.4/311.3 CE 13,
Results: Within-day CVs ranged from 2.4-9.5% and between-day CVs from 3.0-9.9%.
Method comparison analysis was performed using split sample analysis of 20- 75
serum samples. MS to MS comparison studies yielded r-values between 0.943 and
0.997 with recoveries from 90-105%. Regression analysis slope and intercept values
for all steroids in the panel were as follows: slope range 0.89-1.1; intercept range
-0.3 to 6.4
Conclusions: Our method measures 9 steroids in 11.5 minutes with minimal sample
volume and preparation. This method is advantageous in a clinical environment
because of simple sample processing, increased sensitivity, and high-throughput.. The
low sample volume used permits assessment of steroid status in neonates and infants
thereby optimizing early diagnosis of endocrinopathies. The low limits of quantitation
make this method ideal for measurement of androgens and estrogens in women and
prepubertal children.

A-409
Quantitation of 1,25-dihydroxyvitamin D using solid-phase extraction and
fixed-charge derivitization in comparison to immunoextraction

N. Chan, E. Kaleta. Houston Methodist Hospital, Houston, TX

Background: Quantitation of 1-25 dihydroxyvitamin D (DHVD) has been a
difficult task due to the relative concentration of this metabolite with respect
to 25-hydroxyvitamin D, and well known cross-reactivity of immunoassays.
Immunoextraction (IE) techniques have been well characterized, but can be costly due
to antibodies required for extraction. As an alternative to IE, solid-phase extraction
(SPE) coupled to enhanced ionization with fixed-charge derivitization has been
made commercially available. This work describes the validation of the SPE method
directly comparing against IE with traditional triazole-dione derivitization.
Methods: DHVD was extracted by SPE and IE preparations prior to LC-MS/MS. SPE
was performed using AmplifexTM C1000 and S500 cartridges with diisopropyl ether
and hexane/isopropanol extraction, and AmplifexTM Diene reagent derivitization (AB
Sciex, Framingham, MA). Immunoextraction was performed using ImmunoTube
1,25(OH)2 Vitamin D extraction kits (ALPCO), with derivitization using 9mmol/L
4-phenyl-1,2,4-triazole-3,5-dione (PTAD, Sigma). LC was performed for quantitation
using an Acuity UPLC BEH C18 column (Waters), A: H2O, 0.1% formic acid and
B:acetonitrile, 0.1% formic acid, from 37%B to 51%B over a 3.5 min linear gradient.
MS/MS was performed using an AB Sciex 5500 Q-Trap with multiple reaction
monitoring (MRM) for DHVD2 and DHVD3. Clinical validation was performed for
each laboratory developed test, including accuracy, intra- and inter-assay precision,
reportable range, reference range, sensitivity and specificity.
Results: Extraction of DHVD by SPE and AmplifexTM derivitization demonstrated
similar assay performance to the traditional IE followed by traditional PTAD
derivitization.
SPE-Amplifex
IE-PTAD
DHVD2
DHVD3
DHVD2
DHVD3
Linear Range
4-200 pg/mL 4-200 pg/mL 4-200 pg/mL 4-200 pg/mL
Intra-assay Precision
12 pg/mL
7.1
5.5
11.4
8
60 pg/mL
11.4
9.2
8.9
6.1
Inter-assay Precision
12 pg/mL
13.30%
7.20%
8.80%
12.80%
60 pg/mL
4.00%
5.50%
6.10%
8.20%
Limit of Detection
1.9 pg/mL
2.7 pg/mL
2.7 pg/mL
1.7 pg/mL
Limit of Quantitation 4 pg/mL
4 pg/mL
4 pg/mL
4 pg/mL
The SPE method requires more technologist time (1hr vs 30 min), but requires equal
derivitization and LC time, making overall assay time comparable. Cost analysis
shows the SPE method to be lower cost than IE by avoiding expense associated with
antibody extraction.
Conclusion: Work demonstrated that SPE shows comparable analytical performance
to IE, showing promise for utility as a clinical method for sensitive measurement of
DHVD.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

A-410
Development of an Ultra Pressure Liquid Chromatography -Tandem Mass
Spectrometry Method for Pain Management Drugs in Urine

A. A. Ejilemele1, S. Sharma1, D. Patel1, J. Petersen2, R. Patel1. 1Universal

Clinical & Research Laboratories, Webster, TX, 2University of Texas
Medical Branch, Galveston, TX
Background: Provision of adequate pain relief is an important standard of care in
the management of chronic pain and the use of opioids is one of the mainstays of
the management plan for this category of patients. These medications provide partial
analgesia and maintain or improve function. However, their use must be monitored by
regular urine screening in order to monitor compliance, identify diversion as well as
the concomitant use of drugs of abuse.
Methods: The UPLC-MS/MS quantitative method detected 17 drugs and 5
metabolites. Calibrators were prepared in drug free urine using certified reference
materials from Ceriliiant at five levels ranging from 50 -1000 ng/mL for 20 high
concentration analytes and 5 - 100 ng/mL for 2 low concentration analytes. Quality
control samples were also prepared in drug free urine at three target levels (125 ng/
mL, 375 ng/mL and 750 ng/mL for the 20 high analytes and 12.5 ng/mL, 37.5 ng/
mL and 75 ng/ml for the two low concentration analytes). Preparation of sample/
standards/quality control for analysis required the dilution of 100 L with 850 L
of diluent (0.1% formic acid in 100% methanol) and 50 L internal standard. The
internal standard contained 22 analytes at a concentration of 1000 ng/mL. The 22
analytes were separated on a Waters Acuity TQD instrument using a BEH C18
1.7m 2.1 X 50 mm column and binary mobile phase ( A: 0.1% formic acid in 5 mM
ammonium acetate; B: 0.1% formic acid in 100% Methanol) within 7.5 minutes and
a flow rate of 0.3 mL/min. Analytes were detected on the tandem mass spectrometer
in a positive ion mode. Chromatographic peaks of each analyte were acquired using
quantitating and confirmatory ion transitions at cone voltages and collision energies
specific to each compound. Accuracy was tested using recovery experiments.
Results: All analytes had linear calibration curves (r2 >0.950 for the 20 high
concentration analytes; r2 > 0.999 for the 2 low concentration analytes). The within
run coefficient of variation (CV) for the low, medium and high QC ranged from 1.9 12.2% for the high concentration analytes and 1.6 - 12.4% for the low concentration
analytes at three levels. The between run CV for the low, medium and high QC
ranged from 0.05 - 10.8% for the high concentration analytes and 0.1 - 8.6% for the
low concentration analytes. Matrix effect, carryover and interference were minimal.
Samples could be diluted a minimum of eight fold and still remain linear. Analysis of
24 positive and 50 negative CAP samples as well as 10 spiked samples gave excellent
correlation with expected concentrations (r2 > 0.97 for all analytes).
Conclusion: We developed a rapid, linear, accurate and sensitive UPLC-MS/MS
method for the measurement of 17 pain management medications and 5 metabolites
which is suitable as a screening and confirmatory method.

A-412
Evaluation of Q-Exactive coupled with liquid chromatography for Total
Testosterone and Dehydroepiandrosterone Quantification in Serum

G. Altawallbeh1, D. R. Bunch2, J. Gabler3, S. Wang2. 1Cleveland

State University, Cleveland, OH, 2Cleveland Clinic, Cleveland, OH,
3
ThermoFisher Scientific, San Jose, CA
Background: Clinically, total testosterone (TT) and dehydroepiandrosterone (DHEA)
are measured in men and women for androgen abnormalities, and in pediatrics for
cases of delayed or precocious puberty. Historically, immunoassays were most often
used to measure TT. However, studies have shown that immunoassays overestimate
the serum TT at the lower concentrations typically found in females and pediatrics.
Liquid chromatography tandem mass spectrometry (LC-MS/MS) has been identified
as the gold standard technology for steroid determination. Most of the published
LC-MS/MS methods for TT and DHEA analysis are performed on triple quadrupole
MS. Recently, a high resolution quadrupole-Orbitrap (Q-Exactive) MS is available
and may offer improved specificity. Purpose: To evaluate the bench-top Q-Exactive
MS coupled with LC for the quantification of low levels of TT (2.5 ng/dL) and DHEA
(20 ng/dL) in serum and to investigate possible interferences from blood collection
tubes. Methods: Charcoal stripped serum was spiked with testosterone and DHEA
then serially diluted at 10 concentrations to demonstrate linearity. Female serum
specimens were obtained in BD vacutainer tubes with serum separator (SST) and
without serum separator (Non-SST) and compared to check for interference. All
samples (200 L) were spiked with testosterone-d3 internal standard (25 L; 225
ng/mL) and extracted with methyl tert-butyl ether. The supernatant was evaporated

at 40C under a stream of nitrogen then derivatized with hydroxylamine (50 L;

100 mg/mL). Methanol (50 L) was added then it was incubated for 30 min at room
temperature before injection of 50 L. The analysis was carried out on an LC-QExactive system using an Accucore C18 column (50 x 2.1 mm, 2.6 m). Results:
The coefficient of variation for the linearity study was <15% for both DHEA and
testosterone. DHEA was linear from 4-4000 ng/dL using mass transition 304.2>253
with a correlation of R2=0.9877. Testosterone was linear from 0.12120 ng/dL and
had an R2=0.9706 for 304.2>111.6 and R2=0.9727 for 304.2>123.6. The SST to NonSST comparison demonstrated interference with testosterone for mass transition
304.23>111.57 in SST which confirms published findings. Accurate mass was unable
to eliminate the interference, however there was no interference for mass transition
304.2>123.6. Testosterone and DHEA were separated both chromatographically and
with unique mass transitions post-derivatization. Conclusions: The Q-Exactive MS
coupled with LC can be used to quantify TT and DHEA at very low concentrations.

A-413
A Sensitive and Rapid Liquid Chromatography-Tandem Mass Spectrometry
Method for Quantification of Arginine Derivatives

D. A. Payto, C. Heideloff, S. Wang. Cleveland Clinic, Cleveland, OH

Background: Arginine (Arg) is the substrate of nitric oxide synthase for the production
of nitric oxide. Arginine can be methylated to form asymmetric dimethylarginine
(ADMA) and symmetric dimethylarginine (SDMA) through the activity of
methyltransferases. ADMA is an endogenous inhibitor of nitric oxide synthase and
a biomarker for endothelial function. SDMA is a biomarker for renal function that
has been shown to outperform creatinine based equations for determining estimated
glomerular filtration rate (GFR) in predicting kidney function when compared
to measured GFR (mGFR). Objective: To develop a sensitive and rapid liquid
chromatography tandem mass spectrometry (LC-MS/MS) method for measurement
of Arg, ADMA, and SDMA in plasma. Methods: EDTA plasma (50l) and 50L
of internal standard (IS) solution (1.8M Arg-IS as L-arginine:HCL [U-13C6, 9799%], 0.5M of ADMA-IS as ADMA:HCl:H2O [2,3,3,4,4,5,5-d7, 98%] and SDMAIS as [NG,NG-Dimethyl-L-arginine-d6]) were vortex mixed. 1% ammonium acetate
in methanol (300L) was added to the mixture, vortex mixed and centrifuged.
Supernatant (50L) was mixed with 150L of 1% formic acid in acetonitrile and 10L
was analyzed on an LC-MS/MS system using a Polaris Si-A analytical column. Total
chromatographic run time was 3.5 minutes. Multiple Reaction Monitoring (MRM)
transitions were 175.00-70.40, 203.05-46.50, 203.00-172.10 for Arg, ADMA, and
SDMA respectively. Results: Matrix effects were shown to be compensated by the
deuterated internal standards through a mixing study. No carryover was observed up
to 822.1M, 8.9M, and 9.3M for Arg, ADMA, and SDMA respectively. Analytical
Measurement Range (serial dilution of a spiked patient pool), analytical recovery,
and CV (based on CLSI EP10-A3 guidelines) are shown in table 1. Conclusion:
This validated LC-MS/MS method offers sensitive and rapid quantification of Arg,
ADMA, and SDMA in EDTA plasma.
Table 1: Method Validation Data
Arginine
Analytical Measurable Range 7.40-1022.3 M
Analytical Recovery (%)
87.6-114.9
Total CV (%)
8.2-10.4
Intra-Assay CV (%)
5.7-10.3

ADMA
0.09-9.54 M
89.3-114.0
6.8-9.1
4.6-8.2

SDMA
0.09-11.50 M
100.2-106.6
6.4-8.8
4.4-6.1

A-414
Flow injection-tandem mass spectrometry for inborn error metabolism research
using a meta calculation software

J. Lai1, B. Hart1, C. De Nardi2, M. Kozak1, K. Van Natta1, B. Duretz3,

D. C. Kasper4. 1Thermo Fisher Scientific, San Jose, CA, 2Thermo Fisher
Scientific GMBH, Dreieich, Germany, 3Thermo Fisher Scientific, Villebon,
France, 4Medical University of Vienna, Vienna, Austria
Background: Since bacterial inhibition method was first developed for research of
inborn error metabolism in 1960, the technology has changed drastically from EIA,
RIA, FIA, ELISA to LC and Tandem MS over the past 50 years.
Tandem MS allows for higher quality results compared to the old approaches.
However, manual data processing in the post-analytical phase still remains a common
cause of errors in the total testing processing.
This research describes a method of flow injection-tandem MS in analyzing donor
samples for the quantitation of amino acids and acylcarnitines with a meta calculation
software.
Methods: Samples were extracted from dried blood spot cards; the internal standards

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S121

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

were added during the extraction procedure and extracted samples were derivatized
prior to injection onto an LC-Tandem MS system. QC samples were added to the
batch.
The flow injection was conducted using a LC with open-tube providing an automated
sample introduction to a Tandem MS (Thermo Scientific) without chromatographic
separation. The Tandem MS used Selected Reaction Monitoring scanning for the
detection of amino acids and acylcarnitines. This beta version software is developed
for an automatic calculation of mass ion ratio and user defined formulas uaing data
files generated from Tandem MS.

Comparison of the manual and automated extraction techniques within our laboratory
was described by the Deming Equations y = 1.01x + 0.01 and y = 0.97x + 0.17 for
testosterone and androstenedione, respectively. Bland Altman mean bias between the
manual and automated methods was shown to be < 2.5% for both testosterone and
androstenedione.
Conclusion: We have successfully quantified serum testosterone and androstenedione
using both manual and automated SPE with UPLC/MS/MS for clinical research
purposes. The method demonstrates excellent linearity, precision and accuracy.
For Research Use Only, Not for Use in Diagnostic Procedures.

Results: A total of 41 samples and 779 analytes were processed.

The comparison result of this sample set shows that over 99% of concentration
calculations (Analytes and Formulas) are within 10% of bias. Over 87% of Formulas
Ratios are within 10% of bias. Table below shows comparison between software
calculations (One-Step) and manual calculations (Multiple-Steps).
Conclusion: This offline automated data processing tool shows a good agreement with
manual calculation process, and it can process both concentration and user defined
formulas. This meta calculation software improves time effectiveness by eliminating
manual calculation process and removing transcription errors in post-analytical phase.
Comparison Between Software Calculation and Manual Calculation

Type of
Calculations

Analyte
Concentration
Formula
Concentration

Analyte/
Formula
C0, C8, C14,
C14:1, C16,
Cit, Met, Orn,
Phe, Tyr

Number
Bias%
(N)
410
< 20%
407
< 10%
381

F1=C0 + C14:1 41

F2=(Orn - Phe)/
Tyr
Formulas Ratios F3=(C8 + C14:1
- C16)/ (Orn
+ Tyr)

< 5%
< 5%

82
80
72

< 40%
< 20%
< 10%

61

< 5%

R2

Linearity
Equation

0.9986

Y = 0.0593 + 0.81 to
0.9993 X
199.15(ng/mL)

0.9992

Y = -0.7703 43.98 to 167.04

+ 0.9996 X (ng/mL)

0.9966

Y = 0.0039 +
-0.98 to 2.25
0.9983 X

Value Range

A-415
Analysis of serum testosterone and androstenedione for clinical research using
either manual or automated extraction

D. Foley1, B. Keevil2, L. Calton1. 1Waters Corporation, Wilmslow, United

Kingdom, 2University Hospital of South Manchester, Wythenshawe, United
Kingdom
Background: Here we evaluate a UPLC/MS/MS method used to measure serum
testosterone and androstenedione enabling investigation of metabolic dysfunction for
clinical research purposes. An analytically selective method was developed using a
mixed-mode Solid Phase Extraction (SPE) sorbent in 96-well plate format. Either
manual or automated extraction was employed, providing flexibility in sample
preparation options depending on the laboratory environment.
Methods: Certified testosterone and androstenedione reference material purchased
from Cerilliant (Round Rock, TX) were used to create calibrators and QC material
in stripped pooled serum purchased from Golden West Biologicals (Temecula, CA).
Hormone Standardization (HoSt) testosterone certification program samples from the
CDC (Atlanta, GA) were used to provide an initial assessment of analytical method
bias. A set of serum samples (University Hospital of South Manchester, UK) were
analyzed using the newly developed method and an independent LC/MS/MS method
for testosterone and androstenedione and results were compared. This same set of
serum samples was used to show equivalence of the manual and automated extraction
techniques. All samples were pre-treated with ammonia, zinc sulphate and methanol.
SPE was carried out with a Waters Oasis MAX Elution 96 well plate to reduce ion
suppression and concentrate the samples without the need for evaporation. Automated
extraction was performed using the Waters Offline Automated Sample Preparation
Station (OASPS). Using an ACQUITY UPLC I-Class system, samples were injected
onto a 2.1 x 50 mm Waters ACQUITY UPLC HSS C18 SB column using a water/
methanol/ammonium acetate gradient and quantified with a Waters Xevo TQD Mass
Spectrometer.
Results: The method was shown to be linear from 0.05 - 15 ng/mL for testosterone
and androstenedione. Coefficients of variation (CV) for total precision and
repeatability on 5 separate days for low (0.15 ng/mL), mid (1.0 ng/mL) and high (10
ng/mL) QC samples were all < 6% (n = 30) for both testosterone and androstenedione
using manual or automated extraction. Comparison with the values assigned to HoSt
testosterone certification program samples analyzed with this method was described
by the Deming equation y = 1.07x - 0.03 and Bland Altman mean bias was shown
to be < 5% for testosterone. Comparison with samples previously analyzed by the
independent LC/MS/MS method were described by the Deming equations y = 1.06x
+ 0.03 and y = 1.00x - 0.09 for testosterone and androstenedione, respectively.

S122

A-416
Alternative Calibration Strategies for LC-MS Based Analysis of Broad
Reportable Range Analytes

C. Crutchfield, M. T. Olson, W. Clarke. The Johns Hopkins University

School of Medicine, Baltimore, MD
Background: Traditional LC-MS calibration strategies are associated with increased
cost and time to result relative to response factor (Rf) based calibration strategies.
Most reports using Rf based calibration have been applications where the AMR is
less than two orders of magnitude. Testosterone requires an AMR of ~3 full orders of
magnitude to encompass both male and female reference ranges (Females: 2-45 ng/
dL; Males: 250-1,100 ng/dL). Accurate determination of the analyte:internal standard
ratio is problematic as the ratio diverges from unity particularly when A/IS = <0.1,
>10. We wanted to examine the response factor performance of using two non-isobaric
isotopically labeled internal standards, Testosterone-13C3 and Testosterone-2H5, placed
at different concentrations in the same solution.
Methods: Testosterone calibrators at 2, 5, 10, 20, 50, 100, 200, 500, 1000 & 2000
ng/dL were prepared with internal standard concentrations of 20 ng/dL Testosterone13
C3 and 200 ng/dL Testosterone-2H5. QC material was prepared with testosterone at
2.2 ng/dL, 75 ng/dL, & 1,800 ng/dL. Measurements were performed using a Prelude
SPLC coupled to Thermo TSQ Vantage mass spectrometer equipped with a HESIII probe using reverse phase chromatography. Response factors were generated as
n=12 replicate measurements (Testosterone-2H5 Rf = 1.03 + 0.09; T-13C3 Rf = 0.71
+0.17, 95% CI). Complete linear regression included all calibrators in the set with 1/
x2 weighting; constrained linear regression excluded the bottom 3 and top 3 calibrators
for the High IS and Low IS, respectively.
Results: The concentration of the internal standards influenced performance
with respect to the recovery and CV of the QC materials. Interestingly, the higher
concentration internal standard performed as well as the low concentration internal
standard. The table includes a complete comparison.
Conclusion: These data demonstrate the feasibility of applying RF based calibration
to testosterone analysis; however, optimization of internal standard placement and
resolution of the disparity of IS response factors warrant further investigation.
Calibration Strategy
Complete
Constrained
Linear Regression Linear Regression
Low IS High IS Low IS High IS
18%
19%
28%
8%
High QC Bias
CV 7.0%
4%
7%
4%
Bias
5%
12%
14%
8%
Mid QC CV 7%
4%
7%
3.8%
Bias
11%
7%
9%
235%
Low QC CV 10%
6%
11%
2%

Response Factor
Low IS
39%
7%
24%
7%
13%
7%

High IS
20%
4%
12%
4%
-7%
7%

A-417
Comparison of Voriconazole Levels Using LC-MS/MS and HPLC.

M. Hinsdale1, A. Veenis2, S. Wong1, E. Palavecino1. 1Wake Forest Baptist

Medical Center, Winston Salem, NC, 2Bridgewater College, Bridgewater,
VA
Background: Voriconazole is a triazole antifungal agent used for treatment of
invasive fungal infections. Large variations in voriconazole pharmacokinetics may
be associated with decreased efficacy or with toxicity and therefore monitoring
of voriconazole levels is highly recommended. The objective of this study was to
investigate the performance of a newly developed method for measuring voriconazole
levels using LCMS/MS compared to standard HPLC methodology.
Material and Methods: Serum samples from patients receiving voriconazole therapy
were collected according to our institution standards protocols. Aliquots from each
sample were tested by HPLC at a reference laboratory and by the LCMSMS method
developed on site and validated according to standard recommendations using for

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

separation an Aria two channel HPLC with a Cyclone (50 x 0.5mm) column from
ThermoFisher for online cleanup and a Hypersil Gold (50 x2.5 mm 3um particle
size) column from ThermoFisher for the separation. The detection was accomplished
by a ThermoFisher Vantage tandem quadrapole mass spectrometer The LCMS/MS
protocol was the following: A 100 l aliquot of each serum sample, control, and
calibrator was added to 300ul of extraction mixture (50ng/ml voriconazole D3 in
MEOH). The mixture was vortexed for 20 seconds and then centrifuged at 12000 rpm
for 5 minutes. The supernatant of this mixture was diluted 1:1 with water and 100 l
of this supernatant was injected into the column. Once the sample was introduced
into the LCMS/MS system, automated turbo-flow analysis was followed by LCMS/
MS. The mass spectrometer was run in HESI mode with positive polarity. The spray
voltage was 4500v and the vaporizer temperature was 400oC and the sheath gas
pressure was 20 psi and the N2 gas pressure was 100 psi. We detected fragments in
atomic mass units of 281.2 and 224.1 from voriconazole 350.1.The internal standard
voriconazole D-3 amu 353.15 yielded fragments 284.2 and 130. Runtime was (6.3
min) with a detection window of 3min/sample. Performance of the LCMS/MS method
for detecting voriconazole levels in 21 clinical serum samples was compared with that
of the HPLC method.
Results: The LCMS/MS method for voriconazole was linear over the analytical range
of 0.25 to 6 mcg/mL and r2= 0.9958. This study found that the LCMS/MS is precise
with an intra- and inter-assays coefficients of variation of <6% and <4% respectively.
The correlation between the LCMS/MS method and the standard HPLC was very
good with r2= 0.9711 (y= 0.833x + 0.4724).
Conclusions: The LCMS/MS method is a rapid and accurate method for measuring
voriconazole levels and compared well with the values obtained by standard HPLC
procedure. This method is an efficient tool for monitoring voriconazole levels in
serum samples from patients receiving voriconazole therapy.

A-418
Unified LC-MS/MS Assays for Therapeutic Drug Monitoring and Clinical
Trials

D. A. Svinarov, L. Kassabova. Medical University of Sofia, Sofia, Bulgaria

Background: Triple quadrupole LC-MS/MS with its unique selectivity and sensitivity
for quantification has become unbeatable tool to resolve analytical challenges in the
field of therapeutic drug monitoring (TDM), clinical toxicology, steroid analysis,
vitamin D, thyroid hormones and many newer biomarkers. Pharmacokinetic assays for
clinical trials (CT) are another major application of this technique. This study presents
an unified analytical system for TDM and CT based on LC-MS/MS. Methods: Twelve
assays were combined in a common analytical system via unified sample preparation
and chromatography: analytes and internal standards were extracted from 50-100 l
of human whole blood (WB) or plasma (PL) with respective organic solvent, isocratic
separation was performed on a C18 analytical column with a mobile phase consisting
of 85% aqueous methanol with 0.005 mM ammonium acetate and 0.1% formic acid.
Electrospray positive ionization and selected reaction monitoring were used to follow
the respective predominant transitions. Mass chromatograms were collected and
processed by specialized software, and linear regression was performed to determine
analyte concentrations. Validation strategy was strictly adhered to industrial guidance.
Results: For each analyte selectivity was assessed with 6 individual sources of WB
or PL with matrix effect in the range 90112%; extraction recoveries of 7093%;
stability: freeze-thaw was determined for three cycles of 24 h; post-preparative was
documented for 3672 h at 8oC, short-term - at ambient temperature was proven for
624 h in the dark and for 26 h at daylight; stock solution and long term in WB or PL
- for 14 months at -20oC. List of analytes, application profile, and rest of validation
characteristics are as follows:
Matrix
PL
PL
PL
PL
WB
WB
PL
PL
WB
PL
PL
WB
WB

Compound
Accuracy Precision Linearity Range Application
Alprazolam
11%
6%
0.1 24 g/L
CT
Amlodipine
10 %
9%
2014 400 ng/L CT
Clarithromycin
4%
5%
0.4 1725 g/L CT
Clopidogrel
7%
10 %
5 2160 ng/L
CT
Cyclosporine A
11 %
7%
10 2000 g/L TDM, CT
Everolimus
11 %
10 %
1 45 g/L
TDM, CT
Fexofenadine
6%
8%
0.8 322 g/L CT
Galantamine
10 %
11 %
0.2 8 g/L
CT
Indapamide
8%
8%
0,2 79,0 g/L CT
Midazolam
12 %
5%
0,1 100,0 g/L CT
Sildenafil
4%
7%
0.4 740 g/L CT
Sirolimus
11 %
10 %
1 40 g/L
TDM, CT
Tacrolimus
11 %
10 %
1 42 g/L
TDM, CT
25-Hydroxyvitamin
PL
11 %
7%
1 150 g/L
TDM
D
Conclusion: With validation according to current industrial requirements, a
throughput of 100200 samples per working day and immediate method switching,
this unified system provides convenience and optimal versatility for a single LC-MS/
MS instrument.

A-419
Simultaneous Analysis of Multiple Azole Antifungal Drugs in Plasma for
Clinical Research using a simple Protein Precipitation Extraction Protocol

B. J. Molloy1, B. Keevil2, L. J. Calton1. 1Waters Corperation, Wilmslow,

United Kingdom, 2University Hospital of South Manchester, Manchester,
United Kingdom
Background: Here we evaluate a UPLC/MS/MS method to simultaneously measure
the azole antifungal drugs voriconazole, fluconazole, itraconazole, ketoconazole and
posaconazole in plasma for clinical research purposes to determine pharmacodynamic
and pharmacokinetic properties to understand possible drug-drug interactions. A
simple, sensitive, precise and robust analytical method was developed using a simple
protein precipitation protocol and state of the art UPLC/MS/MS technology.
Methods: Voriconazole, fluconazole, itraconazole, ketoconazole and posaconazole
were purchased from Sigma-Aldrich Company Ltd (Dorset, UK) and were used to
create calibrators and QC material in pooled plasma obtained from Sera Laboratories
International (West Sussex, UK). Stable labelled forms of all analytes were used as
internal standards and were purchased from Toronto Research Chemicals (Toronto,
Canada).
Linearity, precision, analytical sensitivity, carry-over and matrix effects were all
assessed. The method was also compared to an independent LC/MS method for the
measurement of voriconazole, and the effect of potential interferences on the method
was assessed. All samples were prepared by precipitation with a solution of the
internal standards in methanol. Using an ACQUITY UPLC I-Class system, diluted
samples were injected onto a 2.1 x 30 mm Waters CORTECS UPLC C18 column
employing a water/methanol/ammonium acetate/formic acid gradient for separation
and quantified with a Waters Xevo TQD Mass Spectrometer.
Results: The method was shown to be linear from 0.060 - 9.5 mg/L, 0.062 - 10.2
mg/L, 0.055 - 10.4 mg/L, 0.048 - 8.8 mg/L and 0.067 - 10.2 mg/L for voriconazole,
fluconazole, itraconazole, ketoconazole and posaconazole respectively. Total
precision and repeatability was assessed over five days with five replicates per day
and is expressed as coefficient of variation (CV). For all analytes the CV for the low
QC (0.25 mg/L) was 5.6% and the CV for the mid (3.5 mg/L) and high (7.5 mg/L)
QC was 3%. Comparison of the results with those for the same samples previously
analyzed by an independent LC/MS/MS method for voriconazole was described by
the Deming equation y = 0.95x + 0.05.
Conclusions: We have successfully quantified voriconazole, fluconazole,
itraconazole, ketoconazole and posaconazole in plasma using a simple protein
precipitation extraction protocol with UPLC/MS/MS for clinical research purposes.
The method demonstrates good linearity, precision, analytical sensitivity and lack of
significant matrix effects.
For Research Use Only. Not for use in Diagnostic Procedures

A-420
Development and Validation of a Dried Blood Spot Method for
25-Hydroxyvitamin D

Z. Wu, D. Bass, K. Urek, J. A. Maggiore. Doctors Data, Inc., Saint

Charles, IL
Background: With sustained consumer and professional demand for Vitamin
D testing, we sought to develop a convenient, precise, and accurate method for
25-hydroxyvitamin D analysis of a dried blood spot (DBS) sample. Single use and
self-retracting bloodletting devices enable the self-collection of capillary blood from
lay users. Filter paper collection and transport media have become highly standardized
and are increasingly used for the analysis of several analytes in the clinical laboratory.
The employment of Liquid Chromatography-Tandem Mass Spectroscopy (LC-MS/
MS) serves to expand the DBS offerings in clinical laboratories for the reliable
analysis of micronutrients. Combining these components provides the basis for the
Doctors Data, Inc. (DDI) DBS method for 25-Hydroxyvitamin D.
Methods: Capillary blood samples are self-collected using SurgiLance sterile
lancets, and spotted onto PerkinElmer 226 Spot Saver Cards and permitted to dry.
Cards are desiccant packaged and shipped via US or International postage to DDI
Laboratory.Two 6-millimeter spots are punched from homogeneous blood spots,
and extracted using a methanol-rich solvent solution which also contains deuterated
internal standards for 25-hydroxyvitamin D2 and D3. Extracts are further processed
and purified using solid phase extraction, eluted, and prepared for injection through a
C-18 analytical column on an Agilent 6460 LC-MS/MS System. Results are read from
a 5-point calibration curve, derived from certified standards for 25-hydroxyvitamin

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S123

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

D2 and D3. Analytical precision, linearity, recovery, accuracy, interference, and
stability were assessed.
Results: For 25-hyroxyvitamin D3, intra-assay precision coefficients of variation
(CV) (n=24) at 35.3 and 74.8 ng/mL were 2.7% and 1.7%, respectively. Interassay CV (n=24) at the same levels were 3.2% and 3.5%, respectively. For
25-hydroxyvitamin D2, the intra-assay CV (n=24) at 15.9 and 76.8 ng/mL were 3.7%
and 2.6%, respectively. Inter-assay CV (n=24) at the same levels were 5.0% and
3.2%, respectively. For assay linearity (n=5), 25-hydroxyvitamin D3 was confirmed
linear between 2.0 and 224.0 ng/mL, with recovery between 98.9% and 105.1%;
25-hydroxyvitamin D2 was confirmed linear between 0.5 and 76.5 ng/mL, with
recovery between 98.9% and 104.2%. Volunteers provided DBS and paired serum
samples allowing sample matrix comparison. Least-squares regression analysis
comparing Total 25-hydroxyvitamin D values in serum to DBS (n=46, range 7.5 - 92.6
ng/mL) yielded a correlation coefficient (R2) of 0.972, y = 0.957x + 8.48; standard
error of estimate = 3.58. Both forms of vitamin D demonstrated one-year stability
when collection cards are stored desiccated in sealed Ziploc bags at ambient (25oC)
temperatures or lower. No detectable analytical interference from hemoglobin was
apparent. Of the first 2000 DBS samples submitted to DDI for Vitamin D testing,
99.1% of DBS cards received contained blood spots of sufficient quantity and quality
to permit processing and analysis.
Conclusion: The analytical method developed and validated by DDI for DBS
25-hydroxyvitamin D testing provides a precise and accurate means of determining
Vitamin D status. The collection system for this method has proven to be wellaccepted by lay users, while the transportation system provides extended stability
to preserve sample integrity to facilitate shipping from remote locations to a central
laboratory for analysis.

A-421
Development of an Assay for Methotrexate and its Metabolites
7-hydroxymethotrexate and DAMPA in Serum by LC-MS/MS

R. Schofield1, K. Murata1, L. V. Ramanathan1, M. E. Grace2, M. Fleisher1,

M. S. Pessin1, D. Carlow1. 1Department of Laboratory Medicine, Memorial
Sloan Kettering Cancer Center, New York, NY, 2Department of Pathology,
Mount Sinai School of Medicine, New York, NY
Background: Methotrexate (MTX) is a folic acid antagonist that is widely used as an
immunosuppressant and chemotherapeutic agent. After high dose administration of
MTX serum levels must be monitored to determine when to administer leucovorin,
a folic acid analog that bypasses the enzyme inhibition caused by MTX and reverses
its toxicity. Patients in renal failure who are given high-dose MTX are often given
carboxypeptidase-G2 (CPDG2) to reverse the effects of MTX. CPDG2 is an enzyme
that converts MTX into glutamate and 4-amino-4-deoxy-N-methylpteroic acid
(DAMPA) that are much less toxic. DAMPA cross-reacts in immunoassays rendering
them unsuitable for monitoring patients given CPDG2 therapy.
Objective: The objective was to develop a very sensitive and specific assay for
MTX by LC-MS/MS that had no cross-reactivity with DAMPA or other metabolites,
including the major metabolite 7-hydroxymethotrexate (7-OH MTX). The assay
needed to be relatively simple to allow its use in a clinical laboratory. In addition, the
assay needed to be able to accurately measure the levels of 7-OH MTX and DAMPA
to support clinical trials utilizing CPDG2 and related compounds.
Methods: Serum samples were prepared by protein precipitation using methanol
containing deuterated MTX as internal standard. LC-MS/MS analyses were performed
on a Thermo Scientific TLX-2 HPLC system (TurboFlow technology) interfaced
to a TSQ Quantum Ultra mass spectrometer operated in the positive ion ESI mode.
Chromatographic separation was achieved using a Cyclone-P TurboFlowcolumn and
a Hypersil Gold C8 analytical column. The HPLC gradient elution was 20-80% of 10
mM ammonium formate/0.1% formic acid in methanol over 1.8 minutes. Calibrators
(7) were prepared in blank human serum.
Results: The LOQs of MTX and DAMPA were 10 nmol/L and for 7-OH MTX it was
20 nmol/L. The analytical measurement ranges for MTX, 7-OH MTX and DAMPA
were 0-1000 nmol/L; the calibration curves were linear over the AMR with correlation
coefficients R2 0.995. Dilutions of 10, 100 and 1000-fold were validated giving a
clinically reportable range of 0-106 nmol/L. The accuracy of MTX was evaluated by
comparison to a dihydrofolate reductase (DHFR) enzymatic inhibition assay, Abbott
TDx immunoassay (Abbott Laboratories, Abbott Park, IL), and an alternate LC-MS/
MS assay. The slopes of the linear regression curves comparing the 4 assays were
all +/- 1% with excellent correlation coefficients. MTX recoveries at concentrations
spanning the AMR were between 98 and 103%. The accuracy of 7-OH MTX and
DAMPA was evaluated using recovery experiments; recoveries of 7-OH MTX and

S124

DAMPA at five different concentrations spanning the entire AMR were between
98.8% and 105.1%. Within-day and between-day (N=10) CVs at concentrations
spanning the AMR were less than 10% for all three analytes.
Conclusion: We have developed a simple, accurate and sensitive assay to measure
MTX levels in serum by LC-MS/MS. Unlike immunoassays this assay shows no
cross-reactivity with either DAMPA or 7-OH MTX and can be used in the setting of
CPDG2 therapy. In addition, the assay accurately measures the levels of 7-OH MTX
and DAMPA to support clinical trials utilizing CPDG2 and related compounds.

A-422
Use of complementary scanning methods by LC-MS/MS in the detection of
urinary synthetic glucocorticoids in patients being investigated for Cushings
syndrome

N. K. Djedovic, S. J. Rainbow. North West London NHS Trust, Middlesex,

United Kingdom
Introduction: Liquid chromatography tandem mass spectrometry methods (LC-MS/
MS) are now routinely being used for the analysis of steroids in clinical biochemistry
laboratories. The majority of methods use multiple reaction monitoring mode (MRM),
which confers the highest specificity. The purpose of this study was to assess the
usefulness of complementary LC-MS/MS scan methods such as precursor ion (PI) and
neutral loss (NL) used in doping analysis studies for the open detection of synthetic
glucocorticoids and their metabolites in patients under investigation for Cushings
syndrome. PI scanning may be used to detect steroids which share a product ion, while
NL scanning detects analytes with a common loss irrespective of the parent ion mass.
Method: The ionization and fragmentation behaviour of eight synthetic glucocorticoids
(prednisolone, methylprednisolone, betamethasone, dexamethasone, triamcinolone
acetonide, fluocinolone acetonide, beclomethasone dipropionate and fluticasone
propionate) and two endogenous glucocorticoids (cortisol and cortisone) was assessed
in positive and negative electrospray on an API 3000 tandem mass spectrometer
equipped with a TurboIonSpray source. Common fragments and neutral losses were
identified, and MRM, NL and PI scan methods were developed. In MRM mode, two
m/z transitions were monitored for each analyte in positive electrospray: 363.3 >
97.3/121.2 for cortisol, 361.3 > 121.3/162.9 for cortisone, 361.2 > 147.1/279.1 for
prednisolone, 375.2 > 161.0/279.0 for methylprednisolone, 393.0 > 147.1/237.1
for betamethasone and dexamethasone, 435.4 > 212.9/339.2 for triamcinolone
acetonide, 453.0 > 121.2/337.1 for fluocinolone acetonide, 521.4 > 279.1/337.1 for
beclomethasone dipropionate, and 501.2 > 121.3/293.0 for fluticasone propionate.
In PI mode, four fragment ions were monitored in positive electrospray: m/z 121.0,
147.0, 275.0 and 279.0. In NL mode, two neutral losses were monitored in negative
electrospray: m/z 76.0 and 104.0. Samples were analysed after liquid-liquid extraction
with dichloromethane of 500 L urine spiked with internal standard (d4-cortisol).
Chromatographic separation was achieved using an Agilent 1100 system HPLC
system and BDS Hypersil C8 column (50 x 2.1 mm, 3 m).
Results: In order to assess the precision and sensitivity of each method, a urine sample
was spiked with a mixture containing the selected corticosteroids at two different
concentrations: 5.00 and 50.0 nmol/L. In MRM mode, the limit of detection (LOD)
was 5.00 nmol/L, while, in PI and NL modes, the LOD was 50.0 nmol/L. Inter- and
intra-assay precision (n = 10) was less than 15% at the LOD. Interference from
isobaric compounds was detected using the branching ratios established for each
compound. Dexamethasone and betamethasone were resolved mathematically. Patient
samples containing synthetic glucocorticoids (methylprednisolone, prednisolone,
betamethasone and dexamethasone) were identified by MRM, and metabolites
of these compounds were detected using the PI and NL modes. Negative samples
from patients were analysed using the established methods to identify endogenous
metabolites.
Conclusion: We have developed an MRM method specific for certain synthetic
glucocorticoids and scanning methods for the potential detection of other exogenous
glucocorticoids based on structural similarities. This additional information should
improve patient management.

A-423
Analysis of plasma catecholamines and metanephrines by mixed-mode SPE and
HILIC LC/MS/MS

J. Danaceau, E. Chambers, K. Fountain. Waters Corporation, Milford, MA

Background: In clinical research, elevated concentrations of urinary catecholamines
can be used in conjunction with their O-methylated metabolites (metanephrines) to
indicate the presence of conditions such as pheochromocytomas, neuroblastomas,

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

ganglioblastomas and ganglioneuromas. However, these compounds (in particular,

norepinephrine, epinephrine, and dopamine) can be a challenge to analyze via
reversed-phase LC/MS/MS due to their high polarity. As a result, many research
laboratories still analyze this panel using ion-pairing reagents and ECD detection.
While reversed-phase LC/MS/MS has been used successfully, challenges still exist
due to ion-suppression from matrix components, insufficient retention, and inadequate
separation of normetanephrine and epinephrine. This work describes a single
extraction and analysis method for monoamine neurotransmitters and metanephrines
from human plasma.
Methods: 250 L plasma samples were pretreated with 50 mM NH4CH3COO, and
loaded onto pretreated wells of mixed-mode Elution SPE plates. SPE wells were
then washed with 20 mM NH4CH3COOH and 50:50 ACN:IPA and eluted with 2 x 25
L aliquots of 85:15 ACN:H2O with 2% formic acid. HILIC-based chromatographic
separation was achieved using an UHPLC silica-hybrid amide column. MPA and
MPB consisted of 30 mM NH4COO dissolved in 95:5 H2O: ACN and 15:85 H2O:
ACN, respectively. Compounds were detected by MRM in ESI positive ionization
mode.
Results: All compounds eluted within 2.0 minutes, with baseline separation between
normetanephrine and epinephrine enabling their unambiguous identification and
quantification. Recoveries ranged from 45-90% and averaged 76%. Matrix effects
were less that 25% for dopamine and norepinephrine and under 10% for the remaining
analytes. Calibration curves were linear from to 10-2000 pg/mL for dopamine, 3-MT,
metanephrine and normetanephrine, and from 50-10,000 pg/mL for epinephrine and
norepinephrine. Calibration curves for all compounds had R2 values of 0.999 or
greater. %CV and bias values for quality control samples were less than 10% for all
analytes at even the lowest QC concentration (40 pg/mL).
Conclusion: This combination of mixed-mode sample preparation and HILIC
chromatography results in a rapid, robust method with excellent linearity, accuracy,
and precision that is suitable for measuring even the lowest endogenous concentrations
of these compounds.
For Research Use Only, Not for Use in Diagnostic Procedures.

A-424
Simultaneous Detection of 60 Pain Management Drugs and Metabolites in
Urine with A High Performance Liquid Chromatography - Tandem Mass
Spectrometry (HPLC-MS/MS) Method

Z. Cao, E. Kaleta, P. Wang. Houston Methodist Hospital, Houston, TX

Background: For chronic pain management, there is a growing need to closely
monitor patients taking pain medications for compliance and illicit drug use. In recent
years, LC-MS/MS based methods that are highly sensitive, specific and cost-effective
have been reported. However, most of these existing methods are limited to common
drugs, such as opiates, benzodiazepines, amphetamines or designer drugs. Therefore,
a method monitoring multiple drugs and drug classes is necessary, especially for
patients with chronic pain that are frequently prescribed multiple medications. The
objective of this work was to develop an HPLC-MS/MS method simultaneously
detecting 60 drugs and metabolites from the following groups: opiates, synthetic
opioids, benzodiazepines, stimulants, anticonvulsants and opioid antagonists.
Methods: Sixty drug standards and 37 deuterated internal standards were monitored
using scheduled multiple reaction monitoring (sMRM) on an AB Sciex QTRAP
5500 mass spectrometer with electrospray ionization in a positive ion mode.
Reversed-phase HPLC separation was performed using a KinetexTM Phenyl-Hexyl
column (50x2.6 mm, 2.6 m particle size) (Phenomenex, CA) with a binary mobile
phase (A: 10 mM ammonium formate in water; B: 0.1% formic acid in methanol) by
gradient (5-90% mobile phase B) with a 0.6 mL/min flow rate. Four-level calibrators
were prepared in drug-free urine (Bio-Rad, CA) in a range of 5-1000 ng/mL (with
exceptions of fentanyl: 0.2-25 ng/mL; gabapentin and pregabalin: 10-2000 ng/mL).
All internal standards were prepared at 50 ng/mL in mobile phase A. For sample
preparation, 50 L patient urine or calibrator and 50 l internal standard mixture were
diluted with 400 l mobile phase A before LC injection.
Results: In this lab developed HPLC-MS/MS assay, all analytes were
chromatograpically resolved. Without additional sample clean-up, none of the
analytes was affected by ion-suppression with this dilute-and-shoot method. All
calibration produced linear calibration curves (R2 >0.940), with the within-run
coefficient variations of 2-33%. We tested 21 patient urine samples previously
screened positive by the Alere Triage TOX Drug Screen assay. Using LC-MS/
MS cut-offs consistent with other reference laboratories, our HPLC-MS/MS assay
was 35/40 (88%) in agreement with the positive Triage results in the following drug
classes: opiates (11/11), amphetamine (7/7), methamphetamine (5/5), cocaine (8/8)
and benzodiazepines (4/9). The 5 samples missed by our LC-MS/MS assay were
positive for benzodiazepine in the Triage assay: one was prescribed alprazolam; four

were prescribed lorazepam. In urine, lorazepam is mostly present as its metabolite

lorazepam-glucuronide, which was detected by Triage but not included in the LCMS/MS assay. The HPLC-MS/MS assay also identified some blinded-spiked drugs
not detected by the Triage assay in 21/22 (95%) urine samples. The identities of these
drugs were confirmed by comparing to standards.
Conclusion: We presented an HPLC-MS/MS method that can simultaneously detect
and quantify 60 pain management drugs and metabolites with complete separation.
Investigation of the cause of false negative in alprazolam, and further optimization and
validation of the assay is ongoing. This work provides a solid foundation for further
development of this method into a robust quantitative assay for clinical workflows.

A-425
Steroid Ionization Efficiency as a Function of Derivation using Multiple
Derivation Reagents

D. R. Bunch, S. Cahalan, S. Wang. Cleveland Clinic, Cleveland, OH

BACKGROUND Ionization plays a crucial role for mass spectrometry
measurements. Blood steroid measurements are important for the diagnosis of various
endocrinological disorders. Conditions such as congenital adrenal hyperplasia would
benefit from simultaneous measurement of multiple steroids with high sensitivity.
However, some of these compounds have inefficient LC-MS/MS ionization, which is
required to achieve the sensitivity for clinical use. The aim of this study was to explore
the impact of derivatization reagents and reaction time on ionization efficiency for
multiple steroid compounds. METHODS Steroids (pregnenolone, cortisone, cortisol,
aldosterone, testosterone, 17-hydroxylprogesterone, progesterone, 11-deoxycortisol,)
at 3.6 M in methanol were derivatized using 50 L of hydroxylamine, methoxylamine,
ethoxyamine, and 2-hydrazinopyridine each at 100 mg/mL at room temperature
and QAO reagent (AB Sciex, Framingham, MA) per the manufacture instructions.
Aliquots were collected at 0, 15, 30, 45, and 60 min and analyzed in real time.
Each time course was performed in triplicate and each sample was spiked with 10
L of reserpine (521 mg/mL; 609 m/z), which was used as an ionization internal
standard. The precursor masses were collected for 150 scans (0.1 s each). Ionization
was normalized to the reserpine peak. RESULTS Due the large amount of data a
representative figure is presented (Figure 1) for aldosterone. The impact of derivation
differed significantly using different reagents. The areas of the circles represent
the coefficients of variation for the triplicate measurements. The y-axis is the
ionization normalized to reserpine and time zero. The x-axis is the time course for
the experiments. CONCLUSION Ionization was greatly increased through derivation
with these reagents. Hydroxylamine and QAO derivations produced the highest and
most consistent responses.

A-426
Identifying four serum peptides as biomarkers for T2DM early diagnosis by
MALDI-TOF MS

Q. Meng1, W. Yan1, Q. Ma2, W. Xu3, X. Liu4, F. Wang4, W. Wang5. 1The PLA

301 Hospital, Beijing, China, 2Peking University, Beijing, China, 3Shenyang
Pharmaceutical University, Shenyang, China, 4Bioyong Technology Co.
Ltd., Beijing, China, 5Capital Medical University, Beijing, China
Background: Currently, there is no ideal serum biomarker for the early diagnosis
of type 2 diabetes mellitus (T2DM). Established diagnostics for T2DM include oral
glucose tolerance (OGT), fasting blood glucose (FBG) level, and hemoglobin A1c

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S125

Mass Spectrometry Applications

Tuesday, July 29, 9:30 am 5:00 pm

level, all of which are markers for the late stages of the disease. The aim of this
study was to apply magnetic bead fractionation coupled with matrix-assisted laser
desorption/ionization-time of flight mass spectrometry (MALDI-TOF MS) to screen
serum samples from patients with T2DM and healthy controls to screen and identify
T2DM specific peptides.
Methods: We (1) performed a discovery screen for peptide differences in serum
proteomic profiles using magnetic-bead enrichment; (2) used a model based on a
genetic algorithm (GA) to distinguish between the serum peptide profiles of patients
with T2DM from healthy controls, to establish a training set, and to validate an
independent test set; and (3) identified the most promising protein/peptide biomarkers
of T2DM using linear ion trap (LTQ)-Orbitrap-MS. Patients selected in the study
were pathologically diagnosed with T2DM, with FBG levels > 7.0 mmol/L and 2-h
OGT (75 g) > 11.1 mmol/L. A total of 306 patients (141M/165F) and 330 healthy
volunteers (83M/247F) were recruited and divided into training sets (206/230)
and test set (100/100). Serum samples were collected before meals, prepared, and
fractionated using weak cation exchange magnetic beads (MB-WCX) according to
the manufacturers instructions (Bioyong Tech, Beijing, China). The resultant samples
were diluted, spotted onto a ClinTOF target and performed the MALDI-TOF-MS
measurements by calibrated ClinTOF instruments (Bioyong Tech, Beijing, China).
All spectra in this research were analyzed using BioExplorer (Bioyong Tech,
Beijing, China) to subtract baseline, normalize spectra, and determine peak m/z values
and intensities in the range of 1,000 to 10,000 Da.

spike and recovery experiment and yielded recoveries ranging from 97.9-102%
for 3 QC levels. The lower level of quantitation was 0.2 ng/mL. Specificity was
evaluated and no interference was observed for serum spiked with hepcidin-20 and
hepcidin-22 at 200 ng/mL each. Dilution linearity was verified to be acceptable up to
8x. The reference interval was verified to be 3.1-43.5 ng/mL for males, 1.1-25.7 ng/
mL for pre-menopausal women, and 2.0-46.9 ng/mL for post-menopausal women.
Stability was established for up to 2 days at ambient temperature and up to 4 days at
refrigerated temperature (2-8C). Freeze-thaw stability was established for 4 cycles at
both -70C and -20C. Long-term frozen stability was established for up to 4 months
at both -70C and -20C.
Conclusion: We have developed and validated a sensitive and specific UHPLCMS/MS method for the quantitative measurement of hepcidin in clinical serum
samples. The method is capable of quantitating hepcidin from 0.2-100 ng/mL. The
method utilizes solid phase extraction for sample preparation. The chromatography
is carried out on a reverse phase sub-2 m particle size column using ultra-high
pressure liquid chromatography. The total chromatographic run time is 7 minutes.
The mass spectrometry is carried out on an AB Sciex QTRAP 5500 instrument. The
predominant precursor ion for hepcidin was determined to be the quintuple charged
species, [M+5H]5+.

Results: Using LTQ-Orbitrap-MS detection, the sequences of seven diagnostic

peptides with m/z values of 1691.7, 1778.7, 1865.5, 2022.1, 2210.3, 2929.3, and
4093.2, which were used to establish the GA model, were found to represent four
different proteins. Four peaks (1691.7 m/z, 1778.7 m/z, 1865.5 m/z, and 2022.1
m/z) were identified as complement C3f, which is cleaved from C3b by factor I and
enters the alternative complement pathway to promote the generation of iC3b. One
peak (2210.3 m/z) was identified as the kininogen-1 isoform 1 precursor. Two peaks
(2929.3 m/z and 4093.2 m/z) were identified as the fibrinogen alpha chain precursor.
An 1473.3 m/z peak was not recognized by this assay, but we identified this peak as
transthyretin according to previous results.
Conclusion: A diagnostic model was generated using a genetic algorithm which may
discriminate T2DM patients from healthy subjects. Four peptides were derived from
complement Cf3 (1691.7 m/z, 1778.7 m/z, 1865.5 m/z, and 2022.1 m/z), kininogen-1
isoform-1 precursor (2210.3 m/z), fibrinogen alpha chain precursor (2929.3 m/z
and 4093.2 m/z), and transthyretin ( 1473.3 m/z). The presence of these peptides at
elevated levels and our laboratory findings may provide new biomarkers for the early
detection of T2DM.

A-427
A Sensitive and Specific Ultra-High Pressure Liquid Chromatography - Tandem
Mass Spectrometry Method for the Quantitation of Hepcidin in Human Serum

D. Chu1, C. Hedin1, D. Chollet2, D. Bertelson1, E. Ellis1. 1Covance,

Indianapolis, IN, 2Covance, Geneva, Switzerland
Background: Hepcidin is a 25-amino acid peptide hormone produced in the liver and
is considered to be the central regulator of iron metabolism. It is a promising biomarker
for the diagnosis and monitoring of iron metabolism disorders such as anemia,
hypoxia and inflammation. Until recently, the assays for measuring hepcidin have
lacked precision, accuracy, and specificity. The objective of this study was to develop
and validate a sensitive and specific ultra-pressure liquid chromatography-tandem
mass spectrometry (UHPLC-MS/MS) method for the quantitative determination of
hepcidin in human serum samples.
Methods: Calibrators were created by spiking charcoal stripped human serum with
hepcidin concentrations ranging from 0.2 to 100 ng/mL. Human serum samples (0.2
mL) were combined with labeled internal standard (13C18,15N3-hepcidin), and
extracted using a 96-well format solid phase extraction (SPE) plate (Waters Oasis
HLB). Hepcidin and its internal standard were analysed using a Waters UPLC
system coupled to an AB Sciex QTRAP 5500 mass spectrometer in MRM mode.
Chromatographic separation was performed using a Waters Acquity reverse phase
column (1.7M, 2.1x100mm). The mobile phases consisted of 0.5% acetic acid in
water (mobile phase A), and 0.05% acetic acid in methanol:acetonitrile (50:50, v:v)
(mobile phase B). The linear gradient started at 20% B and ramped up to 100% B over
6 minutes, followed by 1 minute of re-equilibration. Hepcidin and its internal standard
were detected by positive electrospray ionization with the following transitions:
hepcidin m/z 558.7693.7 and internal standard m/z 562.8697.1.
Results: The method described displayed good linearity over a concentration range
of 0.2-100 ng/ml with r2 >0.99. Intra-day and inter-day precision for all 3 QC levels
showed CVs =<3.5% and =<5.9%, respectively. Accuracy was evaluated using a

S126

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Animal Clinical Chemistry

Tuesday, July 29, 9:30 am 5:00 pm

A-430

Tuesday, July 29, 2014

Modulation of the inflammatory response induced by carragenan in a murine

model by Effect of Jungia sellowii Less.

Poster Session: 9:30 AM - 5:00 PM

T. S. Frode, M. Nader, G. Vicente, J. S. da Rosa, T. C. Lima, A. M. Barbosa,

M. W. Biavatti. University Federal of Santa Catarina, Brasil, Brazil

Animal Clinical Chemistry

Background: Jungia sellowii Less. is a native plant from Brazil used in traditional
medicine to treat inflammatory diseases.

A-428
Estimated complete blood count (CBC) reference ranges for aged adult male
rhesus monkeys (Macaca mulatta) as measured on the Beckman Coulter HmX
analyzer

N. M. Nwaoduah, C. D. Lang, J. S. Daviau, L. J. McCloskey, D. F. Stickle.

Thomas Jefferson University, Philadelphia, PA
Introduction: The clinical laboratory was asked by veterinary services to provide
CBCs for routine checkups on a small number of aged adult male rhesus monkeys.
As we had no reference values for this population, results were analyzed as if from a
control group representing a well population. We report these results obtained using
the Beckman HmX analyzer, and compare reference range estimates to available
literature.
Methods: Whole blood K-EDTA samples were collected from 12 male rhesus
monkeys (age: 23.1 1.3 years; age range: 21 to 25 years). CBC analysis was
performed within 3 hours of collection using the Beckman Coulter HmX analyzer.
Results were analyzed according to expectation of a normal distribution by correlation
of data with the normal distribution predicted by the calculated mean (x) and standard
deviation (s). Outliers were defined as samples deviating from x by more than 3s when
excluded from the group. Parametric reference ranges were defined by convention as
the central 95% of results (x-2s to x+2s).
Results: shown in Table. Despite low sample number, distributions for all CBC
component tests were consistent with normal distributions (expected r^2 for n= 12:
r^2>0.9). Calculated uncertainty in the width of the estimated reference ranges was
20% for n=12. Mean values for WBC, MCV, MCHC and platelets were substantially
different from literature data (|z|>2). These differences may be due to the substantially
advanced age of the animals (life expectancy of 4 years in the wild).
Conclusions: Reference ranges for CBC component tests using the Beckman Coulter
HmX analyzer were estimated based on a small dataset for an aged adult population
of male rhesus monkeys. Datasets for each test were well characterized as normal
distributions. Estimated reference ranges for four CBC component tests (WBC,
MCV, MCHC and platelets) for this population differed substantially from available
literature.

Objective: The aim of this study was to evaluate the anti-inflammatory effect of
the crude extract(CE) from Jungia sellowii Less. its derived aqueous fraction(Aq),
and isolated compounds, succinic acid(SA) and lactic acid(LA) on leukocytes,
exudation, myeloperoxidase(MPO) and adenosine-deaminase(ADA) activities
and nitric oxide(NOx), interleukin-1(IL-1), tumor necrosis factor-(TNF-)
and interleukin-17A(IL-17A) levels, using a murine model of pleurisy induced by
carrageenan(Cg,1%).
Methodology: Fresh Jungia sellowii Less leaves were extracted with ethanol/water
to obtain the CE, which was partitioned with solvents of increasing polarity, yielding
a residual Aq fraction. The compounds, SA and LA, were isolated from this fraction
and their structures were determined by nuclear magnetic resonance(1H NMR). Swiss
mice were used throughout the experiments (Brit.J.Pharmacol.183.811-19.1996).
The study was approved by Committee for Ethics in Animal Research of Federal
University of Santa catarina (protocol: PP00757). Different groups of animals (n=5)
were treated with CE(10-50mg/kg), Aq fraction(1-25mg/kg), SA(0.5-2.5mg/kg) or
LA(0.5-2.5mg/kg) administered by intraperitoneal route, 0.5h prior to the intrapleural
injection of Cg to analyze the effect of the herb on leukocytes and exudation. A
group of animals received a gingival injection of Evans blue dye(25mg/kg) 10min
before herb treatment to evaluate the exudation. The Evans blue dye was measured
by colorimetric assay on enzimaimmunoassay (ELISA) plate reader. The leukocytes
were determined on veterinary automatic counter. Other groups of animals were
pretreated (0.5h) with CE(25mg/kg), Aq(5mg/kg), SA(1mg/kg) or LA(1mg/kg) to
evaluate the effect of the herb on MPO and ADA activities, NOx , IL-1, TNF-,
and IL-17A levels. The MPO and ADA, and NOx, were analysed in accordance with
methods described by Giusti and Galanti, 1984; Rao et al., 1993, and Green et al.,
1982, respectively. The IL-1, TNF-, and IL-17A levels, were determined using
commercially available ELISA kits. All the inflammatory parameters were analyzed
after 4h of pleurisy induction. Statistical differences between groups were determined
by ANOVA complemented by Newman-Keuls test. Values of p<0.05 were considered
significant.
Results: The herb inhibited leukocytes (CE:42.82.9% to 66.85.8, Aq:47.45.0%
to 60.75.2, SA:24.96.8% to 54.42.9% and LA:31.24.3% to 66.35.3%),
neutrophils: (CE:40.33.4 to 65.86.0%, Aq:45.95.3% to 59.85.2,
SA:25.66.3% to 53.23.3%, and LA:31.84.2% to 66.25.2%), and exudation
(CE:31.23.8 to 51.43.3%, Aq:41.42.7 to 73.43.2%, SA:15.02.6% to
42.94.8%, and LA:23.42.9% to 52.63.4%)(p<0.05). Additionally, this plant
inhibited MPO (CE:60.11.6%; Aq:67.51.1%; SA:58.83.9%; LA:65.92.8%),
ADA
(CE:45.22.3%;
Aq:63.95.8%;
SA:37.56.0%;
LA:64.46.7%),
NOx (CE:40.71.1%; Aq:70.40.8%; SA:73.82.6%; LA:76.51.4%), IL1 (CE:78.32.0%; Aq:74.21.5%; SA:24.61.2%; LA:14.91.3%), TNF-
(CE:61.93.4%; Aq:55.12.7%; SA:82.42.3%; LA:63.32.7%), and IL-17A
(CE:64.06.4%; Aq:54.32.6%; SA:41.934.0%; LA: 21.25.4%)(p<0.05).
Conclusion: J.sellowi less.showed an important modulation of the inflammatory
response induced by carrageenan into the mouse pleural cavity by inhibiting the
leukocytes content and the degree of exudation. These inhibitory effects were
associated with the decrease of MPO and ADA activities and NOx, IL-1, TNF-
and IL-17A levels.

A-431
Effects of the IL-10 gene deficiency on Mouse liver function

J. DU, L. JI, D. ZOU. Peking University Shenzhen Hospital, Shenzhen,

China
Background: Interleukin (IL)-10 is an important immunoregulatory cytokine
produced by many cell populations. Numerous investigations suggest that IL-10
plays a major role in chronic liver diseases. Our aim is to investigate the effect and
mechanism of IL-10 deficiency on mouse liver.
Methods: Using the automatic biochemical analyzer to analyze serological biomarkers
of liver function between IL-10 gene knockout mice and IL-10 wild type control.
The pathological morphological changes were observed with the light microscope.
The levels of iNOS and IL-1genes in liver tissues were determined by real-time

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S127

Animal Clinical Chemistry

Tuesday, July 29, 9:30 am 5:00 pm

fluorescence quantitative PCR and enzyme-linked immunosorbent assay(ELISA).
Results: Compared with the wild type, the serum levels of ALB,TP,TBIL and DBIL
of IL-10 deficient mice were significantly decreased(P<0.05), no obvious differences
were found in AST,ALT and liver pathological morphology(P>O.05). The expression
of iNOS and IL-1 genes, the serum levels of iNOS and IL-1 were significantly
higher in IL-10 deficient mice than in wild type mice (P<0.05).
Conclusion: Endogenous IL-10 deficient mice can significantly decreased serum
ALB and BIL. The effect may be related to the upregulated expression of iNOS and
IL-1.

A-432
Analytical Evaluation of an Assay Kit Incorporating New Ready to Use Liquid
Stable Reagents for the Determination of Glucose, Through Conversion by
Hexokinase, in Different Biological Fluids

K. Gilmore, S. Baxter, Z. Burnside, K. Piper, S. McElhatton, M. Rodriguez,

J. Campbell, S. FitzGerald. Randox Laboratories Limited, Crumlin, United
Kingdom
Background: Carbohydrates provide the human body with glucose, a simple sugar
used as source of energy by the cells. The body must maintain proper glucose
levels to ensure that a person remains healthy. Glucose determination is useful in
the diagnosis and monitoring of carbohydrate metabolism disorders (i.e. diabetes
mellitus, hypoglycaemia, pancreatic islet cell carcinoma) as well as in research
and drug discovery processes. This study reports the evaluation of an assay for the
determination of glucose by hexokinase-mediated reaction in serum, plasma, urine,
cerebrospinal fluid (CSF) samples. This assay is applicable to automated systems and
incorporates new ready to use liquid stable reagents, which facilitates the application in
test settings by simplifying the experimental procedure and reducing handling errors.
Methods: The assay involves a series of steps, initiated by the conversion of
glucose to glucose-6-phosphate by hexokinase. The glucose-6-phosphate is then
oxidized by glucose-6-phosphate dehydrogenase, causing the reduction of oxidized
nicotinamide adenine dinucleotide (NAD) to reduced nicotinamide adenine
dinucleotide (NADH). The absorbance of NADH is measured as endpoint reaction
at 340/410 nm. The assay is applicable to a variety of analysers. The reagents are
liquid stable and ready to use. On-board and calibration stabilities were tested by
storing two lots of reagents uncapped on the analyser for a period of 60 days. Withinrun and total precision were assessed by testing serum samples at defined medical
decision levels, 2 replicates of each sample were assayed twice a day for 10 days.
Correlation studies were conducted using a commercially available assay system.
Results: The reagents presented an on-board stability of 60 days and calibration
frequency of 60 days. The assay was linear from 4 mg/dL up to 700 mg/dL for serum,
plasma, urine and CSF. The within-run and total precision for different concentration
levels, expressed as %CV, was < 2.0. In the correlation studies 99 serum patient
samples, 88 plasma samples (lithium heparin), 87 plasma samples (potassium EDTA),
51 urine samples and 113 CSF samples were tested and the following linear regression
equations were achieved: y = 1.001x + 0.3; r = 1.0 (serum, range 5-676 mg/dL), y =
1.001x + 0.2; r = 1.0 [plasma (lithium heparin), range 5-686 mg/dL), y = 1.002x +
0.0; r = 1.0 [plasma (potassium EDTA), range 5-676 mg/dL), y = 0.989x - 0.3; r =
1.0 (urine, range 4-664 mg/dL), y = 1.005x - 0.1; r = 1.0 (CSF, range 20-654 mg/
dL). Conclusion: The results of this evaluation indicate that this assay is applicable
to the determination of glucose in different biological fluids. This assay kit exhibits
good correlation with existing commercial assay systems for all the analysed matrices.
Furthermore, it incorporates liquid stable reagents, which simplifies the experimental
procedure and reduces handling errors.

A-433
Effects of Chronic Ozone Exposure on the Oxidant-Antioxidant System of
Brain Tissue

m. ozler, N. Ersoz, t. topal, A. H. Apaydin, S. Oter, A. Korkmaz. Gulhane

military medical faculty, Ankara, Turkey
Background: Ozone treatment entails exposing a body cavity or circulation system to
a mixture of oxygen and ozone, and has been used in conjunction with other therapies
to treat various pathologies. Although repeated ozone treatment has been purported
to stimulate an antioxidant response, research characterizing long-term oxidantantioxidant homeostasis has not yet been reported. As a therapeutic treatment for
peritoneal adhesion, rats were exposed to ozone by one of two regimens. During this
study, we evaluated oxidant and antioxidant parameters from brain tissues retrieved
from these rats.

S128

Methods: For this study, 24 Sprague-Dawley male rats were separated into three
groups. The adhesion model for this study was established by making an incision
in the cecum of the rats, followed by suturing. Two groups were administered ozone
treatment for 15 days; however, each group was subjected to a different regimen: one
group was treated with ozone immediately following surgery, whereas the other group
was treated 24 h post-surgery. After 15 days, and while anesthetized, surgery was
performed to open the abdomen in order to evaluate the adhesion site and to excise
the brain tissue. Malondialdehyde (MDA), superoxide dismutase (SOD), carbonilized
protein (PCO), and glutathione peroxidase (GSH-Px) levels were measured in the
excised brain tissues.
Results: Brain tissues from rats immediately treated with ozone following surgery
exhibited higher levels of SOD activity compared with the other two groups. By
contrast, the MDA levels observed in this group were significantly lower. There was
no difference between groups in terms of PCO levels. We determined that both groups
exposed to ozone (i.e., 0 h and 24 h post-surgery) exhibited significantly higher GSHPx activities in comparison with the control group.
Conclusion: Our findings indicate that long-term ozone treatment supports the
antioxidant system in brain tissue. Furthermore, it should be noted that the time it
takes to receive treatment is critical, as quicker ozone treatment more effectively
stimulated an antioxidant response.

A-435
Evidence of the anti-inflammatory properties of Ageratum conyzoides L. in a
murine model of pleurisy induced by carrageenan

T. S. Frode, S. V. G. Vigil de Mello, B. M. Facchin, A. B. G. Luz, C. F.

Bosi, D. W. Rosa, M. Biavatti. University Federal of Santa Catarina,
Brasil, Brazil
Background: A.conyzoides L. is used in Brazilian folk medicine as analgesic and
anti-inflammatory agent. The aim of this study was to evaluate the anti-inflammatory
effect of the Crude Extract(CE), and its derived fractions: Ethanol(EtOH)
and Hexane(HEX), and isolated compounds: Methoxy Nobiletin(MeONOB),
1,2-Benzopyrone(BP) and Eupalestin(EP) from A.conyzoides on: leukocytes,
exudation and myeloperoxidase(MPO) and adenosine-deaminase(ADA) activities,
and nitrate/nitrate(NOx) and cytokines (TNF-alpha and IFN-gamma) levels in a
murine carrageenan-induced pleurisy.
Methodology: The aerial parts of A.conyzoides were air-dried at 50C, crushed and
stored at 8C. The CE was prepared by maceration with ethanol, concentrated in rotary
evaporator. The EtOH and HEX fractions were obtained by, extraction of CE, with
different solvents of increasing polarity: ethanol and n-hexane. EtOH was partitioned
with solvents of increasing polarity: n-hexane, dichloromethane and ethyl acetate.
The dichloromethane fraction obtained from EtOH was chromatographed on silica gel
flash column using dichloromethane and methanol gradient as eluent, being collected
100 fractions. Fractions 27-34 resulted in BP and fractions 35-69 yielded a mixture of
methoxylated flavonoids were re-chromatographed on silica gel flash column using
dichloromethane and methane gradient as eluent to obtain EP and MEONOB. Swiss
mice were used though the experiments (Br.J.Pharmacol.183.811-19.1996). This
study was approved by the local Ethical Committee (protocol: PP00757/CEUA/2012).
Different groups of animals (n=5/group) were treated with CE(10-200mg/kg),
EtOH(5-25mg/kg), HEX(25-50mg/kg), MeONOB(2.5-10mg/kg), BP(2.5-10mg/kg)
or EP(1.0-10mg/kg) administered by intraperitoneal route, 0.5h before carrageenaninduced pleurisy(Cg,1%) administered by intrapleural route(i.pl.). The inflammation
was analyzed after 4h. The leukocytes were analyzed using an automatic counter.
A group of animals was previously challenged with Evans blue dye (25mg/kg, i.v.)
to evaluate the exudation. The doses of CE(50mg/kg), EtOH(10mg/kg), HEX(50mg/
kg), MeONOB(5mg/kg), BP(5mg/kg) or EP(5mg/kg) administered 0.5h before were
selected to evaluate the effect of the herb on MPO and ADA activities, and NOx level
which were analyzed by colorimetric assays. Analysis of cytokines were conducted
using mouse inflammation cytometric bead array kit (BD Biosciences) only in EtOH
and its isolated compounds. Statistical differences were determined by ANOVA
and Student-Newman-Keuls post-hoc analysis. Values of p<0.05 were considered
significant.
Results: CE(50-200mg/kg), EtOH(10-25Mg/kg), HEX(50mg/kg), MeONOB(510mg/kg), BP(5-10mg/kg) or EP(5-10mg/kg) inhibited leukocytes(CE:35.06.8 to
79,31.8%; EtOH:57.06.6 to 75.60.3%; HEX:56.77.5%; MeONOB:58.52.2
to 60.06.2%; BP:59.63.2 to 60.95.8%; and EP:64.71.6 to 69.51.4%),
neutrophils(CE:37.47.1 to 80.11.7%; EtOH:68.32.2 to 77.00.2%;
HEX:54.38.1%; MeONOB:63.12.0 to 63.85.5%; BP:64.35.1 to 71.92.2%;
and EP:69.71.8 to 74.10.8%), exudation(CE:26.52.8 to 68.22.9%;
EtOH:58.53.9 to 63.32.9%; HEX:33.32.4%; MeONOB:51.37.2 to 54.32.7%;
BP:40.83.6 to 44.13.6%; and EP:39.98.1 to 42.16.9%), MPO(CE:34.75.6%;

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Animal Clinical Chemistry

Tuesday, July 29, 9:30 am 5:00 pm

EtOH:31.65.3%;
HEX:29.64.2%;
MeONOB:28.62.3%;
BP:22.42.9%;
and EP:18.31.3%), and ADA activities(CE:70.56.3%; EtOH:71.04.9%;
HEX:72.14.8%; MeONOB:67.32.1%; BP:27.48.4%; EP:54.63.7%) and NOx
level (CE:79.17.2%; EtOH:63.212.9%; HEX:71.29.4%; MeONOB:55.513.9%;
BP:53.814.1%; EP:80.80.3%). Also EtOH and its isolated compounds inhibited
TNF-alpha(CE:24.71.6%; EtOH:25.94.5%; MeONOB:26.62.2%; BP:31.51.3%;
and
EP:21.45.4%)
and
IFN-gamma(CE:15.71.6%;
EtOH:11.01.4%;
MeONOB:13.60.5%; BP:6.42.0%; and EP:11.00.6%)(p<0.05).
Conclusion: A.conyzoides presented important anti-inflammatory properties not only
by inhibiting leukocytes migration but activated neutrophils. This effect was also
associated with the decrease of exudation and NOx and pro-inflammatory enzymes
(MPO and ADA). This effect appears to be mainly related to the EtOH fraction and
its isolated compounds: MeONOB, BP, and EP which inhibited all the inflammatory
parameters, including TNF-alpha and IFN-gamma.

A-436
Effects of methanolic leaf extract of African mistletoes (Loranthus micranthus)
on male sexual function in streptozotocin-induced diabetic Wister rats

B. O. Idonije1, O. O. Festus1, H. B. Osadolor2. 1Ambrose Alli University,

Ekpoma, Nigeria, 2University of Benin, Benin, Nigeria
Background: The leaves of African mistletoes (Loranthus micranthus) have been
shown in traditional African setting to improve sexual function in diabetic males but
data on scientific proofs of this therapeutic action of these leaves is scanty hence this
study. In this study, the effect of methanolic extract prepared from the leaves of L.
micranthus on serum testosterone levels, sperm count and motility in diabetic male
wistar rats was studied.
Method: The animals were randomly divided into four (4) groups made up of six
(6) rats each and diabetes was induced in the rats by the administration of alloxan
(100mg/kg) for 7 days. Group A served as the control (untreated diabetes), groups B
and C were treated with 150mg/kg and 300mg/kg respectively of the extract while
group D received the 100mg/kg of the standard antidiabetic drug (chlorpropamide).
The duration of substance administration was fourteen days. On the fifteenth day,
all the animals were lightly anaesthetized with ether and their blood collected for
testosterone analysis. The rats were further dissected and the caudal epididymis of
each incised and seminal fluid collected for sperm count and motility tests.
Results: Table 1 showed diabetes to decrease male sexual functions (group A) when
compared with standard reference value. Also, significant increase (p<0.05) in the
level of serum testosterone, sperm count and sperm motility which was dose depended
was showed with the extract administration (groups B and C) compared with control
(group A). Chlorpropamide treated rats (group D) also showed a significantly
increased male sexual function compared with the control, however, mistletoe was
more potent.
Conclusion: From the findings of this study, we suggest that leaf extract of African
mostletes be studied in detail so as to know its therapeutic dose for it possible use as a
therapeutic agent in the treatment of male infertility secondary to testosterone/sperm
abnormalities.

C-benzylated monoterpenes, aromatic oils, flavanones, C-benzylated flavanones, and

C-benzylated dihydrochalcones. Traditionally, a decoction of the root is used in the
treatment of many diseases, including diabetes. However, the use of this plant extracts
in the treatment of diabetes have not been scientifically validated. In this study, we
determined some blood analytes in streptozotocin-induced diabetic rats administered
aqueous or ethanolic extract of the root of Uvaria chamae. Methods: Thirty six
(eighteen adult normal and eighteen streptozotocin-induced diabetic rats) Sprague rats
were administered aqueous or ethanolic extract (300 mg/kg body weight) of Uvaria
chamae for 35 days [6 rats per group, average body weight (265.23 7.20 g)]. The six
groups were composed as follows: Healthy rats receiving de-ionized water (Normal
Control); Normal rats receiving aqueous extract (Normal plus Aqueous Extract);
Normal rats receiving ethanolic extract (Normal plus Ethanolic Extract); Diabetic
rats receiving de-ionized water (Diabetic Control); Diabetic rats receiving aqueous
extract (Diabetic plus Aqueous Extract); and Diabetic rats receiving ethanolic extract
(Diabetic plus Ethanolic Extract). Diabetes was induced using a single injection
of streptozotocin (Sigma-Aldrich, 60 mg/Kg body weight in 0.05 M-citrate buffer,
pH 4.5) intraperitoneally. Animals were euthanized by decapitation on day 35 after
commencement of the feeding trial. Blood was collected for assays. Results: There
was a significant (p<0.05) decrease in blood glucose level in the treated diabetic groups
compared to the diabetic control. We also noted significant (p<0.05) increase in BUN
in the diabetic control compared to the normal control. The administration of aqueous
or ethanolic extract to the diabetic rats did not restore the level of BUN to that of normal
control group. The diabetic groups administered aqueous or ethanolic extract showed
increasing trend in the level of MCV toward the normal control group compared
to the diabetic control. The levels of MCHC and WBC were significantly (p<0.05)
lower in the diabetic groups administered aqueous or ethanolic extract compared to
diabetic control. The levels of RBC, Hgb, PCV, platelets, monocytes and granulocytes
were not significantly (p>0.05) altered among the groups. We noted reducing trend in
the levels of IL-6 and IL- in the diabetic groups administered aqueous or ethanolic
extract compared to the diabetic control. However, serum creatinine level was slightly
elevated in the diabetic group administered ethanolic extract. Conclusion: Overall, the
consumption of aqueous or ethanolic extract of Uvaria chamae lowers blood glucose
level which may be beneficial in the management of diabetes. The increasing trend
in MCV level due to the administration of ethanolic or aqueous extract may protect
against the development of anemia that is associated with diabetes. The inflammatory
cytokine (IL-6) normally up-regulated in diabetes was depressed by the aqueous or
ethanolic extract administration. However, the increased serum creatinine level is
indicative of the potential adverse effect of the ethanolic extract on renal function.

A-438
Validation of Automated Immunoglobulin A, G, and M in Non-human Primate
Serum to Support Pre-Clinical Toxicology Studies

S. E. Wildeboer, J. H. Bock, J. L. Wisniewski, R. P. Giovanelli. Pfizer,

Groton, CT
Background: Immunoglobulins are glycoprotein molecules that are produced during
an immunogenic response, and are divided into five classes, based on the differences
in the amino acid sequence found on the heavy chains. Immunoglobulin A (IgA),
Immunoglobulin G (IgG) and Immunoglobulin M (IgM), are the three most common
immunoglobulins found in serum. Measurement of immunoglobulin concentrations
are used to aid in the diagnosis of immune response or abnormal protein metabolism.
Here, we validated automated human IgA, IgG and IgM assays for use in pre-clinical
primate studies.
Methods: Siemens IgA (02194102), IgG (02193432), and IgM (02193483) assays are
PEG-enhanced immunotubidimetric assays performed on the Siemens Advia 1800.
Serum samples are pre-diluted by the instrument, then mixed with specific antiserum
to form a precipitate that can be measured turbidimetrically at 340/694 nm. The
measured absorbance is then compared to an established calibration curve (Advia
Chemistry Liquid Specific Protein Calibrator, 07711199) and the concentration is
determined and reported as mg/dL. Bio-Rad Liquid Assayed Multiqual (Level 1:
694, Level 2:695, Level 3: 696) quality control (QC), AUDIT MicroCVTM Protein
Linearity (K702M-5), and colony primate (Macaca fascicularis) serum were analyzed
to test the performance and dynamic range of the assays.

A-437
Effects of Uvaria chamae Extracts on Blood Glucose,Inflammatory
Markers,Hematological and Renal Status in Streptozotocin-induced Diabetic
Rats

F. E. Olumese1, I. O. Onoagbe1, F. O. Omoruyi2. 1University of Benin, Benin

City . Nigeria, Benin, Nigeria, 2Texas A&M University, Corpus Christi,
USA, Corpus Christi, TX
Background: Uvaria chamae is a medicinal plant that is used in some regions
of the world in the treatment of diabetes, and as an antifungal, antimalaria,
and bacteriostatic herb. The chemical constituents of Uvaria chamae, include

Results: Intra-assay precision testing was performed using 4 primate serum samples,
and 2 levels of QC. Samples were analyzed a minimum of 5 replicates in a single
assay run. All samples demonstrated %CV 3.6. Accuracy and inter-assay precision
testing was performed using 3 levels of QC run in triplicate for 5 runs. Mean and
imprecision were calculated and fell within the manufacturers established 2SD
range, and demonstrated %CV 3.0. Three primate serum samples were analyzed in
duplicate over 4 separate assay runs, demonstrating %CV values 3.6. Commercially
available linearity standards were analyzed and demonstrated reportable assay

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S129

Tuesday, July 29, 9:30 am 5:00 pm

analytical ranges of 31.4 - 669.4, 140 - 3009, and 17.4 - 393.8 mg/dL for IgA, IgG and
IgM, respectively, confirming the manufacturers stated analytical range. Dilutional
linearity was performed with 3 primate serum samples with high concentrations.
Samples diluted starting at a 1:2 dilution in 0.9% saline and analyzed in duplicate.
Dilutional linearity was established as 1:16, 1:8, and 1:4 for IgA, IgG and IgM,
respectively. Spike recovery was performed with pooled primate serum spiked with
10%, 7.5% and 5% of the highest level of linearity material. All spiked sample results
were within 20% of the expected value. Sample frozen stability and freeze/thaw
stability (-80C) was performed with eight primate serum samples with varied IgA,
IgG and IgM results. The samples were assayed neat then analyzed after 1, 3, and 6
months and after 1, 2, and 3 freeze/thaw cycles. All samples had appropriate percent
recovery (80 - 120%).
Conclusion: The Siemens IgA, IgG, and IgM assays met all outlined criteria for
validation and are appropriate for use in non-human primate serum samples to support
pre-clinical toxicology studies.

S130

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Animal Clinical Chemistry

Automation/Computer Applications

Wednesday, July 30, 9:30 am 5:00 pm

Wednesday, July 30, 2014

Poster Session: 9:30 AM - 5:00 PM
Automation/Computer Applications

B-001
Reference values study for the urinalysis parameters measured in the sysmex
uf1000

N. Z. Maluf, C. F. d. Pereira, E. V. V. Oliveira, L. L. Jaques, S. B. Ferreira,

R. Balherini, M. d. Menezes, S. D. S. Vieira. DASA, Barueri, Brazil
Background: The urinalysis department of our Laboratory implanted the Sysmex
UF1000 for the urinary cells analysis. The designed new sample flow includes the
Roche Urisys 2400 for biochemical analysis and UF1000 for the urine sediment
analysis. The methodology used by this equipment is flow cytometry. It performs
the analysis and counting of erythrocytes, leukocytes, epithelial cells (EC), cylinders,
crystal, mucus, sperm, bacteria (Bact) and yeasts. Currently, all the literature reference
values for these parameters are based on the analysis by optical microscopy, which was
the previous methodology used in the lab. Thus, we need to review and standardize the
references values of these parameters in our laboratory. The objetive is to evaluate and
standardize the reference values of the UF1000 measured parameters.
Methods: We evaluated 100 healthy individuals with no disease and medications or
vitamins, to standardize the reference values for erythrocytes, leukocytes, cylinders,
epithelial cells and bacteria. These individuals are FTEs from the laboratory and we
ask them to participate in this study. For the statistical analysis we use dispersion
graph and Gaussian distribution.
Results: After analyzing the results and statistics through the ABC curve, we found the
results that is on the table. We also noted that the RBC values between 13.4 to 28.0 x
10/L should be reassessed and compared with optic microscopy.
Conclusion: We concluded that the reference values for the UF1000 were higher than
the obtained by the optical microscopy methodology. It uses larger amounts of sample
and counts a greater number of cells, what allows it to be more accurate. After this
study and with the continuous experience of monitoring patients in the last months, the
new reference values got a good acceptance by the Lab professionals and physicians.
Parameters
RBC
WBC
EC
CASTS
BACT

Value - UF1000
13,3x10 /L
31,5x10/L
3,6/L
1,07/L
26,4x10/L

Results: The performance of the QMS Tacrolimus Assay was evaluated on the
Beckman Coulter AU480/AU680/AU5800 analyzers. All studies were evaluated
using CLSI guidelines. On the AU480 and AU5800, four levels of Tacrolimus controls
were used in the studies. The precision ranged from 4.3 %CV to 4.2 %CV for withinrun and 7.2 %CV to 4.8 %CV for total run. Linearity was measured and confirmed
over a range of 1.0 ng/mL to 28.7 ng/mL. The functional sensitivity was observed at
1.0 ng/mL. Patient correlation studies: AU480=1.0(AU680) - 0.08 (N=107, r=1.00),
AU5800=1.02(AU680) + 0.23 (N=108, r=1.00). On the Beckman Coulter AU680
analyzer, three levels of Tacrolimus spikes and patient pools with lowest concentration
at 2.9 ng/mL and highest at 25.0 ng/mL were tested twice per run, two runs per day
for 20 days. The precision ranged from 1.8 %CV to 4.9 %CV for within-run and 3.9
%CV to 7.5 %CV for total run. Linearity was measured and confirmed over a range of
0.4 ng/mL to 30 ng/mL. The functional sensitivity was observed at 0.9 ng/mL. Patient
correlation studies: AU680=1.14(LC-MS/MS)+0.50 (N=266, r=0.97).
Conclusion: All measured studies demonstrated acceptable performance, validating
the use of the QMS Tacrolimus Assay on the Beckman AU480/AU680/AU5800
analyzers, and will provide an effective monitoring system for patients receiving
Tacrolimus therapy.

B-004
Comparative study between ELISA and Chemiluminescence (CLIA) methods
for the analysis of ENA-screening and specific ENA.

E. Melguizo1, C. Gonzlez-Rodrguez1, G. vila Garca1, . FernndezHermida2, P. Falc-Pegueroles2. 1Virgen Macarena University Hospital,
Sevilla, Spain, 2Menarini Diagnostics, S. A., Barcelona, Spain
Introduction: The anti-cellular antibodies are autoantibodies directed against a variety
of cellular structures (DNA, ribonucleoproteins ...). The group of specific antibodies
directed against specific cellular proteins, anti-Ro/SSA, anti-La/SSB, anti-Sm, antiRNP/U1RNP, anti-Scl-70/topoisomerase I and anti-Jo-1 / histidyl-tRNA synthetase
are clinically important in patients with autoimmune diseases (Sjgrens syndrome,
systemic lupus erythematosus (SLE), scleroderma, dermatomyositis and polymyositis
among others).
Objetive: Our aim was to analize the degree of agreement between chemiluminescence
(CLIA) Zenith-RA from Menarini Diagnostics (Florence, Italy) and the habitual
ELISA from Inova Diagnostics (San Diego, USA) for anti-ENA screening, anti-Ro/
SSA, anti-La/SSB, anti-Sm, anti-RNP/U1RNP, anti-Scl-70 and anti-Jo-1.
Material y method:
Serum samples from 496 patients with positive anti-cellular antibodies (title 1/160
or higher) were selected. ENA screening tests for specific antibodies were measured
and in positive results, specific antibodies (anti-SSA, anti-SSB, anti-Sm, anti-RNP,
anti-Scl-70 and anti-Jo-1) were measured by ELISA (INOVA diagnostics) and CLIA
(Menarini, Zenit RA). Samples we classified as positive or negative according to the
manufacturer cut-offs (20 U/mL for ELISA and 10 U/mL for CLIA assays, except
CLIA ENA-screening wehere cut-off is 1) and the agreement degree was obtained
using SPSSv19 statistical program.

B-003
QMS Tacrolimus Assay for the Beckman Coulter AU480, AU680, and AU5800
Clinical Chemistry Analyzers

C. Wong, D. Cheng, A. Thao, L. Ye. ThermoFisher Scientific, Fremont, CA

Background: The objective of this study is to evaluate the performance of Beckman
Coulter AU480/AU680/AU5800 clinical chemistry analyzers for the quantitative
determination of tacrolimus in human whole blood used in the management of
kidney, heart, and liver allograft patients receiving tacrolimus therapy. Monitoring
for tacrolimus is important for effective use to prevent allograft rejection following
organ transplantation. The measurement of tacrolimus concentrations in whole
blood in conjunction with other laboratory data and clinical evaluation can optimize
immunosuppressive effect and minimize adverse side effects for patients.
Methods: The QMS Tacrolimus assay is a liquid stable particle-enhanced
turbidimetric inhibition immunoassay. The assay is based on competition between
free tacrolimus in the sample and tacrolimus derivative coated onto a micro-particle
for anti-tacrolimus antibody binding sites. The tacrolimus-coated micro-particle
reagent is rapidly agglutinated in the presence of anti-tacrolimus antibody reagent and
the rate of agglutination is inversely proportional to the tacrolimus concentration in
the sample. The rate of absorbance change is measured photometrically and is directly
proportional to the rate of agglutination of the particles. A concentration-dependent
classic agglutination inhibition curve can be obtained to determine the tacrolimus
concentration in the sample. The assay consists of two reagents and an extraction
solution for sample pretreatment. The calibrators contain tacrolimus in the human
whole blood matrix at concentrations of 0, 2, 5, 10, 20, and 30 ng/mL.

Results: The results are shown in table 1. It was not possible to calculate the
Kappa index for anti-Jo-1 since all samples were negative by CLIA. The rest of
determinations show a good correlation between the two methods, and showed a good
classification of patients with systemic autoimmune disease.
Conclusions: Both methods show a good degree of agreement in the analysis of
specific anti-ENA. Given the advantages of CLIA techniques in front of ELISA
(master curve for each lot of calibrators and controls, linearity and continuous access
of samples) it could be a valid option for the analysis of specific anti-ENA in the
clinical laboratory.
cut-offs and degree of agreement (measured by kappa index) ENA screening and
specific ENAs.
CutCut-offELISA
kappa ELISA CLIA
offCLIA
ELISA + CLIA +
(UI)
index (UI)
ENA-screening 20
1
0.769 295
328 201
167
anti-SSA (Ro) 20
10
0.947 82
85
126
123
anti-SSB (La) 20
10
0.914 133
153 75
55
anti-RNP
20
10
0.876 145
165 63
43
anti-Sm
20
10
0.917 190
193 18
15
anti-Scl-70
20
10
0.976 200
196 8
12
anti-Jo-1
20
10
201
208 7
0

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S131

Automation/Computer Applications

Wednesday, July 30, 9:30 am 5:00 pm

B-005
Comparative evaluation of four assays for the automated determination of
glycated hemoglobin

N. Z. Maluf, C. F. d. Pereira, C. Rosin, M. R. Ribeiro, L. K. Tirado, E. F.

Moreira, A. N. Camilo, F. S. Fukuoka. DASA, Barueri, Brazil
Background: Recently, it has been discussed, among endocrinologists, pathologists
and general physicians, which is the most accurate method for automated
determination of glycated hemoglobin (HbA1c). Attempting to answer this question,
in this study, we compared HbA1c results obtained from two distinct methodologies
in four different analyzers to evaluate their performance and the impact in clinical
monitoring of diabetes mellitus.
Materials and methods: We selected 73 samples, 16 with a percentual 5% of
HbA1c, 20 borderline samples (upper normal range result between 5.5 and 7%), 37
samples with high percentual of HbA1c (7%) to perform on the following analyzers
for HbA1c determination: TOSOH HLC-G7, Bio-Rad VARIANT II, both based on
the principle of High-Performance Liquid Chromatography (HPLC), ROCHE Cobas
6000 and Siemens ADVIA 2400, both based on turbidimetric methods. Among the
selected samples, it is possible that some presented anomalous hemoglobin.
Results: The correlation among the analyzers is summarized in the table below.
Conclusion: There was good agreement among results of HbA1c when comparing
different analyzers and distinct methodologies. Our study suggests that determination
of HbA1c for clinical monitoring of diabetes mellitus can be performed using any of
the automated assays systems evaluated.

Methods: During 2012, an Excel program was developed in the lab in order to
determine the percentage of the reported panic values from each laboratory. The
introduction of a simple periodic report in an easy and automated manner revealed
several problematic laboratories on one hand, and on the other, increased the
awareness among laboratory staff and their commitment to report panic values. This
new parameter was chosen to be one of the quality criterions in laboratory surveys.
After implementation of this program and increased awareness among laboratory
staff, panic values reporting by each lab increased gradually and steadily to >80% in
2012 with a continuous increase in 2013.
Conclusions: The availability of the report and the ability of the managers and staff
to present it quickly improved quality, and allowed real-time monitoring of failures
in reporting critical results. This report allowed us to know exactly- where, when and
who did not report critical results and to address each problem. Results during 20122013 indicated that this led to a fundamental change in the conduct of the lab staff and
their commitment to report the PANIC values in real-time.
As an outcome of this project, and due to its importance, this feature will be
implemented in all laboratories of Clalit Health Care Services as a new module in
the LIMS software. This implementation will take into account the labs experience
and knowledge. Such a module will allow control of the rules for complex alarms,
managing alerts via pop-up windows, and producing statistics reports for different
sectors in a convenient and flexible way.

B-007
Performance Evaluation of Siemens Dimension EXL 200 Integrated Chemistry
System for a Regional Medical Center.

A. KHAJURIA, B. Robeson, J. Hafermann, G. Palmer, T. Walters.

Marshfield Clinic, Marshfield, WI
Background:The Dimension EXL 200 Integrated Chemistry System offers
LOCI advanced chemiluminescent technology and automated, productivity features
for the smaller-sized laboratory. Chemistry and immunoassay integration allows
simultaneous processing to maximize workflow efficiency. Manufactures design their
analytical systems to achieve certain performance characteristics for intended use
and all instruments from a particular vendor may not have equivalent performance
specifications. In a clinical laboratory risk management should begin with the
verifications of those characteristics based on the performance goals to assess the
suitability for the intended use.
Methods: A test menu of 22 analytes was evaluated on Dimension EXL 200 for
precision, reportable range, accuracy, detection limit and analytical specificity.
The patient comparisons were done with another Dimension EXL 200 to have
method comparability for the analytes. Assay performances were evaluated against
performance criteria established in our laboratory. The statistical analysis was done
by Analyse-it and method performance was calculated as Sigma metrics value
[(Tea-Bias)/CV]. A Sigma 3.0 value is the minimum performance and 6.0 Sigma is
considered world class performance.
Results: Method performances evaluated on Sigma scale against our established
quality performance criteria showed Glucose meeting a Sigma >6.0, five assays
were <3 and rest varied between >3 and 5.5 Sigma. Chloride, BUN, creatinine and
triglycerides showed <3 Sigma at low concentrations. HbA1c assay was unable to
meet the performance goal requirement of 6% proposed by CAP & NGSP.

B-006
Reporting Critical Laboratory Values (PANIC) - Identifying Problems and
Improving Processes

Conclusion: Estimation of performance of a method is greatly influences by the

stability of method under routine operating conditions and may not always mimic the
controlled testing environment of vendors. Sigma metric performance assessment tool
is a good measure to evaluate the performance of analytical processes. The data shows
that most of methods on Dimension EXL 200 perform in the scale of 3 to 5.5 Sigma.

D. Weinstein, E. Neumark, J. Goldman, T. Tohami, A. Shalom, E. Weiss, G.

Rashid. Meir Medical center, Kfar Saba, Israel
Background: Reporting critical laboratory test values to hospital wards in real-time is
essential to provide immidiate treatment to critically ill patients. It is challenging for
the laboratory staff due to the workload and the differences in critical values between
regular departments and specialized departments, like dialysis, oncology, etc.
A routine monitoring of panic values reported by phone documented on the LIMS
(Laboratory Information Management System) revealed a 40-60% gap between
the number of results that should have been reported in real-time and the actual
implementation. Therefore, the laboratory management initiated a major improvement
project in order to minimize this gap.
Purpose: To increase reporting of real-time critical lab values in order to improve the
quality of patients care. Improvement targets were reporting at least 80% of critical
laboratory results in 2012 and at least 90% in 2013.

S132

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Automation/Computer Applications

Wednesday, July 30, 9:30 am 5:00 pm

The resulting IICC standard is documented in an IHE Laboratory Analytical Workflow
(LAW) profile. This profile provides the following capabilities, most of which are not
supported by the LIS2 (ASTM) standard:

B-008
Diagnostic paths - towards computational evidence

1) Support for Immunoassay, Chemistry, Hematology, and Microbiology testing

Z. Liniger, G. Lindner, G. M. Fiedler, A. B. Leichtle. Inselspital - Bern

University Hospital, Bern, Switzerland

2) Unique identification of each order request at the test or test panel level

Background: For the establishment of rapid, efficient, and financeable laboratory

diagnostics in an ICD-10-funded hospital environment, defined diagnostic paths
are of ever-growing importance. Usually they are set up as hierarchical trees or
flow charts, leading to altered diagnostic suggestions depending on the outcome of
previous tests (step-by-step diagnostic schemes). These guidelines are designed to
focus diagnostic efforts on the most selective analytes, thereby avoiding unnecessary
testing and providing a test panel sufficient to cover the most important side diagnoses.
By overall reducing the number of recommended tests, they can help to improve
cost effectiveness - especially in a health care compensation system that includes all
diagnostic testing in a flat charge (as e.g. the recently introduced Swiss DRG system).

4) Selection of query as the default mode

Diagnostic paths arise from different sources: single publications, recommendations

of the different medical associations, and from the (inter-)national societies of
Laboratory Medicine. However, all of these recommendations bear a severe drawback:
they are agreements of experts in the field and therefore reflect opinions, not evidence.
Methods: Along with the rapid evolution of computational tools for parallel, GPUbased and grid- or cloud-based computing and the development of powerful statistical
strategies an appealing resolution for this aforementioned lack of evidence emerges:
To utilize already collected laboratory data, to merge it with diagnostic information
and to derive which analytes are selective and likewise superior for the setup of a
diagnosis. The conventional diagnostic path recommendations can be replaced by
diagnosis-specific models inferred from the laboratory and classification data, a
process that does not imply prior expert knowledge in terms of opinions which
parameter to measure, but that is based only of the numerical evidence contained in
the dataset and that delivers disease probabilities, assisting the physician to balance
decisions for further diagnostic and therapeutic procedures. In a proof-of-principle
study we utilized laboratory data and diagnostic information of n>15000 patients of
the Inselspitals Department of Emergency Medicine and computationally determined,
which lab tests are indispensable and which ones are unnecessary for establishing the
top ICD-10 coded diagnoses via the estimation of posterior inclusion probabilities
with bootstrapped confidence intervals of a set of lab tests.
Results: Our results clearly show the feasibility of our new approach: For myocardial
infarction e.g. our algorithm without any prior knowledge of the disease nor any
pathophysiological basis suggests a panel of lab test similar to current guidelines solely based on computational principles, our patient population, and laboratory data
already generated thereof.
Conclusion: In a highly digitalized hospital environment, the present lack of evidence
for diagnostic paths is unjustifiable: All diagnosis-related classification of all patients
- hospitalized or out-patient - are electronically registered, and usually all lab tests are
also electronically available. These data, stored away and laid untouched for decades,
could improve and streamline diagnostic testing and implicitly generate benefit for the
patients - the tools therefor are ready.

B-009
New Instrument Interface Standard to Enable Improved Interoperability with
Integrated Information Systems.

J. B. Jones1, E. Heierman2, R. Bush3, E. Olson4. 1Geisinger Health

System, Danville, PA, 2Abbott Diagnostics, Irving, TX, 3Orchard Software,
Indianapolis, IN, 4Siemens Healthcare, Raleigh, NC
The In Vitro Diagnostic (IVD) Industry Connectivity Consortium (IICC) has worked
with several standards organizations to develop a new interoperability (i.e. instrument
interface) standard that provides plug-n-play connectivity between IVD analyzers and
IT systems, eliminating the need for unique analyzer interfaces. Currently, the Clinical
and Laboratory Standards Institute (CLSI) LIS1 and LIS2 specifications (also known
as ASTM) provide limited guidance on the structure and content of the data being
exchanged in instrument interfaces. These older standards are highly flexible and have
been implemented in many different ways thus creating barriers to integration and
interoperability. In addition laboratories must validate these unique interfaces every
time a new analyzer is installed, and often encounter lengthy implementation cycles
before they can go live.

3) Improved query for orders

5) Simplified order download
6) Ability for an analyzer to accept/reject orders
7) Improved device identification for test logging
8) Contributing substance identification for test logging
9) Basic and enhanced message interface to support IVD instrument rule evaluation
10) Support for LOINC to identify test requests and observations
11) Unique identification of runs
12) Support for hematology images, graphs, and plots
13) Support for transmission of raw values
To confirm that the LAW profile could support plug-n-play connectivity, vendors
representing seven IVD analyzer and three IVD IT systems have participated in the
2012 European IHE and 2014 North American IHE Connectathons. The IHE team
defined ten test cases representing major LAW scenarios and focused on immunoassay
and clinical chemistry orders. Each IVD analyzer tested interoperability with each
IVD IT system through the execution of the test cases. The testing was monitored
by IHE independent representatives. Each IVD IT system used the same interface
implementation to communicate with each of the seven instruments.
All 13 testing events were successfully completed, allowing all participating vendors
to register an IHE Integration Statement documenting that their implementation
successfully integrated with the other vendors through the use of the LAW profile.
This new IICC instrument interface standard is now available for adoption by
IVD instrument vendors (ivdconnectivity.org). Its use should greatly simplify
interoperability between different IT systems in the more integrated healthcare
continuum that is currently evolving under federal guidelines of Meaningful Use.

B-010
Measuring Reproducibility of analysis in a proficiency testing (PT) scheme
using modified control materials. A novel approach using big data analysis.

O. Panagiotakis1, D. Rizos2, K. Makris3, A. Haliassos1. 1ESEAP Greek

Proficiency Testing scheme for Clinical Laboratories, Athens, Greece,
2
Hormone Laboratory, Aretaieion Hospital, Medical School, University of
Athens, Athens, Greece, 3Clinical Biochemistry Department, KAT General
Hospital, Kifissia, Greece
The evaluation of reproducibility in proficiency testing (PT) schemes is based on the
analysis of the same sample multiple times during a cycle. This can be tempting for
participants to detect the replicated samples and to report already known target values.
In order to overcome this possibility ESEAP designed a study involving the analysis
of two replicates of a sample and two other samples, derived from the initial, but
diluted or concentrated by 10% by modification of the serum volume lyophilized per
vial.
The aim of our study was to investigate the interference of samples modification to
the estimation of reproducibility.
The 290 laboratories of ESEAP from Greece and Cyprus participated in this blind
study during the 2013 cycle of Clinical Chemistry scheme. We excluded the results
from the laboratories that havent reported all four samples, from those excluded from
the normal bias analysis and from the laboratories that reported a method change
during this cycle. We evaluated 4 results from 209 laboratories for 21 different
parameters, the two as received (samples 1&2) and the other two after correction with
the appropriate factor for the dilution (sample 3) or concentration (sample 4) and we
calculated the mean value for each sample. We evaluated the 6 possible comparison
couples for each parameter using t-test. Our results are presented at the followinf
table:

IICC established partnerships with the CLSI (Clinical Laboratory Standards Institute),
IHE (Integrating the Healthcare Enterprise), and HL7 (Health Level 7) standards
organizations in order to leverage existing work, accelerate the creation of a plug-nplay standard, and promote worldwide adoption.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S133

Automation/Computer Applications

Wednesday, July 30, 9:30 am 5:00 pm

B-012

Simoa HD-1: a fully automated digital immunoassay analyzer capable of single

molecule counting, sub-femtomolar sensitivity, and multiplexing

D. H. Wilson, C. W. Kan, A. J. Rivnak, T. G. Campbell, T. Piech, D. M.

Rissin, P. Patel, E. Frew, G. Provuncher, A. Schroff, R. E. Meyer, M.
Fishburn, L. Chang, D. R. Fournier, D. C. Duffy. Quanterix Corporation,
Cambridge, MA

A pattern (in bold) emerged for the majority of the analytes showing that the first
unmodified and the diluted sample, as also the two unmodified and the concentrated
and the second unmodified showed statistically non-significant difference in contrast
to the other three combinations that where statistically different.
Our data show that the possible interference due to the modification of the samples is
smaller than the uncertainty of measurements of the identical samples (1&2) thus our
approach can be used for the estimation of reproducibility.

B-011
Rapid Consistent Turnaround time (TAT) of Lab Results through an Innovative
Centrifugation Protocol

J. Baker1, S. Johnson1, R. Wonderling2. 1MultiCare Health System, Tacoma,

WA, 2Abbott Diagnostics, Abbott Park, IL
More than a billion laboratory tests are performed each year in the United States
influencing 64% to 74% of the medical decisions. Labs have to provide accurate
results and as quickly as possible for the STAT tests. So, turnaround time (TAT) of test
results is a critical component in patient care.
MultiCare Health System (Tacoma, WA) laboratory has been providing excellent TAT
for several critical lab tests over the past several years. We have recently achieved
a Received to Result mean TAT time of 23.1 minutes (N=2743, SD=7.2 minutes,
median=21 minutes) using Abbott ARCHITECT Enzymatic Creatinine test as a
proxy for our Comprehensive Metabolic Panel (CMP) This test requires the longest
analytical time in a CMP, thus equating to 90% of CMP results reported in 30 minutes.
In case of cardiac assays, 90% of all Troponin results are reported within 35 minutes
and 85% of all BNP results are reported within 35 minutes.
Centrifugation time is a major bottleneck and becomes the rate-limiting step and
greatly reduces the laboratory specimen throughput. To improve TAT, we switched
from the conventional slow spin procedure (6 minutes spin at 3500 RPM) to the
new fast spin procedure (1 minute spin at 12,000 RPM). There was no noticeable
difference in the results obtained after a fast spin compared to the slow spin when we
tested 41 different patient samples in 16 different tests. After this off-automation track
high speed centrifugation step for STAT samples, the samples are placed on the TLA
(Total Lab Automation) track system for processing as usual. Thus the STAT TAT
target is met with the advantages of TLA.
Manufacturers of blood collection tubes typically recommend slower speeds and
times for centrifugation. Using a fixed rotor refrigerated centrifuge (Beckman Coulter
Allegra Model X30R and F1010 rotor assembly) allows faster speeds and shorter
times for separation of plasma from cells. Gel barrier plastic collection tubes are
highly resistant to breakage or failure during centrifugation at this higher speed and
no statistical change in hemolysis is present. However, balance tubes in the centrifuge
should be replaced monthly as a precautionary step as they show visual bending of the
tube after this period of centrifugation.
Most laboratories are constantly pressured to deliver results more quickly. A viable
option to improve sample handling in the laboratory is total laboratory automation
which has been shown to dramatically improve laboratory TAT and clinical throughput.
However, even when using automation, centrifugation is still a major bottleneck for
STAT specimens and becomes a rate limiting step. Performing off-automation track
centrifugation of STAT specimens using this Innovative Centrifugation Protocol (with
2 centrifuges each with 6 tube capacity) has allowed MultiCare Health System to
achieve quick and consistent turnaround times for stat chemistry results.

S134

Objective: The aim of this work was to develop the next generation immunoassay
analyzer capable of several orders of magnitude greater sensitivity than current
best-in-class conventional immunoassay systems. The technology utilizes single
molecule array (Simoa) technology to usher in fully automated digital immunoassay
and multiplexing capability to the clinical laboratory. Simoa technology isolates
individual paramagnetic beads in arrays of femtoliter-sized wells and detects
single enzyme-labeled proteins on these beads using sequential fluid flows in
microfabricated polymer array assemblies for ultra-sensitive signal measurements.
These array assemblies have been incorporated into a low cost disk consumable. The
array approach for assay signal quantification allows for rapid digital data acquisition
and high throughput, enabling development of a fully automated system for low-cost
measurement of clinically relevant biomarkers with high precision and unprecedented
sensitivity across a broad dynamic range.
Methods: Detection of single molecules using Simoa has been reported previously.
In brief, proteins are captured on antibody-coated paramagnetic microbeads (2.7-mm
diameter) and labeled with single enzymes, followed by partitioning single beads into
arrays of femtoliter-sized wells and sealing the arrays in the presence of a fluorogenic
substrate. We developed a low cost disk consumable that enables standard fluidics
handling instrumentation to load and seal assay beads into the arrays using only fluidic
flow. Beads with single enzyme label molecules are isolated in single wells in the
presence of a substrate, and fluorescent product is allowed to build up within the
40 femtoliter confines of the wells. The fluorescence signal quickly concentrates in
such a small volume, allowing detectable signal from a single enzyme label in only
30 seconds. Depending on the analyte concentration, hundreds to many thousands of
single molecule signals are counted simultaneously using a fluorescence microscope
optical system and image analysis software. Next we integrated this array and imaging
module together with a standard fluidics-handling platform that performs sequential
cuvette processing of paramagnetic bead-based ELISA reagents. The reagents employ
antibody-coated capture beads, biotinylated detector antibodies, and streptavidin-galactosidase as the signal enzyme. A standard bead-based immunoassay is performed,
and then the beads are transferred to the Simoa module for signal development and
digital quantification.
Results: Prototype single-plex digital immunoassays were developed for PSA,
Troponin, IL-6, and A42. A prototype cytokine 6-plex was also developed. LoDs
ranged from 0.002 to 0.05 pg/mL. The LoQ of the PSA assay was estimated as
0.037 pg/mL. These sensitivities ranged to over 1000-fold greater than conventional
immunoassay. Imprecision for the prototype assays was evaluated over 10 runs across
five days in a CLSI format. CVs were generally less than 10%. Spike recovery and
linearity met standard criteria for acceptability. The system throughput is 68 tests/
hour, and over 4 logs of dynamic range were demonstrated. The prototype 6-plex
gave equivalent precision and sensitivity performance to single-plex versions of the
same assay.
Conclusion: The data indicate we have developed a next generation fully automated
immunoassay analyzer capable of orders-of-magnitude greater sensitivity than
conventional state-of-the-art immunoassay systems.

B-013
Capability Analysis for Procalcitonin Assay Performed with the miniVidas
Method.

V. M. Genta1, R. Murray1, E. Ravago1, B. Boston1, S. Spingarn2, W. Tang1.

1
Sentara Virginia Beach General Hospital, Virginia Beach, VA, 2Sentara
Norfolk General Hospital, Norfolk, VA
Background: Procalcitonin blood values have been linked to both increased risk
and severity of sepsis. Consequently, in-hospital assays assure availability of
timely results for prompt medical intervention. In our institution we evaluated the
capability of the miniVidas method with phase1 (short-term) and phase2 (long-term)
precision verification studies. Methods: After a miniVidas instrument (IVD1210422,
BIOMERIEUX) was installed by the manufacturers representative, a phase1 study
was performed by assaying two levels of QC material (level1 lot#1031940; level2
lot#103150, BIOMERIEUX) with five independent assays for five consecutive days.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Automation/Computer Applications

Wednesday, July 30, 9:30 am 5:00 pm

Phase2 study was performed by assaying two levels of control material once a day for
100 days. The observations were transferred to Minitab (Version 16, Minitab Inc.)
statistical software. The observations were analyzed with descriptive, exploratory,
inferential and diagnostics univariate and multivariate statistical techniques. For
the process capability analysis the UCL and LCL were calculated using the mean of
phase1 study +/-(0.5xtotal error allowed by CAP). Results: Phase1 study. Descriptive
statistics: Level1 mean=17, s=0.7, C.V.=3.9%, min=16, Q1=16.8, median=17,
Q3=17.8, max=19; Level2 mean=1.57, s=0.05, C.V.=3.2%, min=1.4, Q1=1.55,
median=1.58, Q3=1.59, max=1.64. For level1 histogram, normality plot and
Anderson-Darling (A-D ) test (level1 P=0.4) showed quasi-normal distribution; for
level2 A-D test showed non-normality (P=0.005) due to a possible outlier (obs. #14
= 1.4 ng/mL), the histogram and the normal probability plot showed a quasi-normal
distribution. Tests for equality of variances showed discordant results; for both levels
Bartletts test P<0.05, Levenes test P>0.5. Estimates of daily s with Bonferronis
95% C.I. showed that the values of s for level1 of day 2 and for level 2 of day 14 were
larger than the others and had larger 95% C.I. Parallel boxplots and ANOVA with
Tukeys multiple comparisons showed that while for level 1 there were statistically
significant differences between daily means (P=0.004), for level 2 there were no
statistically significant differences(P=0.06). However, the maximum mean difference
for level1 = 1.4ng/mL was not significant for either QC or clinical practices. The
plots of the autocorrelation function showed statistically significant autocorrelation
for the first two observations only. The plots of Hotellings T-square and generalized
variance did not show either parallelism or non-randomness. The individual point
QC charts for level 1 and 2 were constructed using the estimates of mean and s (for
s = s/0.98, corrected for bias) with LCL= mean-3s, UCL= mean+3s. Phase 2 study:
Level1 mean=16.5, s=0.8, C.V.=4.8%, Level2 mean=1.6, s=0.07, C.V.=4.3%. The
individual points charts for both levels of contol did not show trends, shifts,outliers
or autocorrelation. There were no statistically significant differences between either
means or s for Phase1 and Phase2 studies (P>0.05).The capabilities indexes Cp
(level1=2.4, level2=2.3) and Cpk (level1=2.8, level2=2.7) were similar indicating
centering of the mean, their values (> 2) indicated acceptable six-sigma performance.
Conclusion: These studies showed that the phase1 study design was adequate to
estimate mean and s for the individual points QC charts.Furthermore, phase2 studies
indicated that the methods reproducibility and capability were adequate to monitor
the variability within the total error specifications. Finally, appropriate statistical
software was essential for the analysis of the observations.

B-015
Improvement of work processes in the laboratory after introduction of the
Automate 1250 System

E. Neumark, E. Weiss, J. Goldman, A. Shalom, G. Rashid. Meir Medical

center, Kfar Saba, Israel
Background: In early 2012, the automated Automate 1250 system (Beckman
Coulter) was introduced to the Chemistry, Immunology and Endocrinology
Laboratories at Meir Medical Center. This computerized robotic system sorts and
archives laboratory samples. It is designed to treat pre-analytical processes (such
as decapping, sorting and if necessary aliquoting samples, followed by samples
distribution to designated workstations) and post-analytical processes (archiving
samples). Prior to introducing the system, work processes in the 3 laboratories were
mapped. Immunology Laboratory processes were different from other laboratories
since most of the blood samples that arrived were aliquoted to several tubes in order
to run on different analyzers. This resulted in delays and results were only available
1-3 days after receiving the samples, depending on how often the equipment was
operated. We present in this work the improvement that was made in the Immunology
Laboratory.
Aim: This study investigated several workflows to decrease immunology TAT (Turn
Around Time). This was accomplished by using the Automate 1250 to route samples
for immediate testing on analyzers that operated daily in the original tubes, and then to
send the aliquoted samples to analyzers that operated less frequently.
Results: Analyzing for several months how the Immunology tests were processed,
the various tests requested for each blood sample, and their distribution between
analyzers, was the basis for the improvement in the lab. If the analyzer on which
the most tests were performed was run daily, more than 50% of the samples would
not be aliquoted, and the test results would be available within a day. Due to these
observations, Immunology Laboratory staff began working on a method whereby they
analyzed samples immediately upon their arrival in the first round of the Automate
1250. All the original samples were sent to the analyzer that ran most of the tests. This
change in procedure affected other processes in the lab, which were not directly linked
to the Automate 1250 system, and all together it resulted in a significant decrease in
turnaround times:
Before the improvement: Average result availability before the use of Automate
1250 was 2.7 days _ 17% longer than the specified result time.

B-014

After the improvement: Average result time decreased to 1.1 days, only a 2%
deviation in the specified result time.

Performance Evaluation of three URiSCAN series for routine urinalysis

K. Ko, M. Kwon, H. Woo, H. Park. Kangbuk Samsung Hospital,

Department of Laboratory Medicine, Seoul, Korea, Republic of
Background: Urinalysis is one of the important diagnostic screening tools in clinical
practice. Correct urinalysis results offer information of the renal and genitourinary
system. The URiSCAN devices (Yeongdong Diagnostics, Seoul, Korea) are one
of the most commonly used urine analyzers in Korea. Our study aimed to evaluate
the analytical performance of the three URiSCAN devices for routine urinalysis in
comparison with Roche urine analyzers.
Methods: A total of 1,273 urine specimens were enrolled in this study between
June 2013 and November 2013. We performed urinalysis using three URiSCAN
devices; Optima, Pro II and Super+, and compared to other urine analyzers (Roche
Diagnostics, Switzerland); Urisys 1100, Cobas u411 and Urisys 2400. Each Roche
analyzer was selected with consideration of complexity of each URiSCAN device.
The results of three analyzers for blood, bilirubin, urobilinogen, ketone, protein,
nitrite, glucose and leukocyte were considered concordant if they were within 1
grading difference in comparison with the results by Roche analyzers. Moreover, the
screening of leukocytes and erythrocytes using both systems were compared with
microscopic examinations.
Results: Good correlation between three URiSCAN devices and their corresponding
methods were observed (range of correlation coefficient: 0.602 to 0.989, p<0.001).
Overall agreement rates for eight test items were acceptable: 84.9% - 100% for
Optima vs. Urisys 1100; 96.8% - 100% for Pro II vs. cobas u411; and 99.3% 100% for Super+ vs. Urisys 2400. The sensitivity and specificity of the URiSCAN
Optima were 62.7% and 95.4% for leukocytes, 91.4% and 78.1% for erythrocytes;
for URiSCAN Pro II were 79.6% and 86.4% for leukocytes, 62.2% and 96.9% for
erythrocytes; for URiSCAN Super+ were 82.5% and 87.4% for leukocytes, 92.9%
and 83.8% for erythrocytes.
Conclusions: The three URiSCAN devices showed high agreement rates with the
corresponding Roche urine analyzers and microscopic examination. Therefore, these
three URiSCAN series would be useful for clinical laboratory performing in routine
urinalysis.

Conclusions: Using the Automate 1250 system increased Immunology Laboratory

efficiency, resulting in shorter TAT times and improved service. TAT time reduced
from 2.7 days before the introduction of the new system to the lab to 1.1 days, affecting
60% of the samples. This resulted in safer patient care and improved service quality.
Following the above results, the use of the Automate 1250 system was extended to
other laboratories.

B-016
QMS Everolimus Assay for the Beckman Coulter AU480, AU680, AU5800
Clinical Chemistry Analyzers

C. Wong, D. Cheng, H. Naqvi, K. Yokobata. ThermoFisher Scientific,

Fremont, CA
Background: The objective of this study is to evaluate the performance of the
Beckman Coulter AU480/AU680/AU5800 clinical chemistry analyzers with the
QMS Everolimus Assay for the quantitative determination of everolimus in human
whole blood used in the management of organ allograft transplant patients receiving
everolimus therapy. Monitoring for everolimus is important for effective use to
prevent allograft rejection following organ transplantation. The measurement of
everolimus concentrations in whole blood in conjunction with other laboratory data
and clinical evaluation can optimize immunosuppressive effect and minimize adverse
side effects for patients.
Methods: The QMS Everolimus assay is a liquid stable homogeneous particleenhanced turbidimetric inhibition immunoassay. The assay is based on competition
between drug in the sample and drug coated onto a microparticle for antibody binding
sites of the everolimus antibody reagent. The everolimus-coated microparticle reagent
is rapidly agglutinated in the presence of the anti-everolimus antibody reagent and
in the absence of any competing drug in the sample. The rate of absorbance change
is measured photometrically. When a sample containing everolimus is added, the
agglutination reaction is partially inhibited, slowing down the rate of absorbance
change. The concentration-dependent classic agglutination inhibition curve can be

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S135

Automation/Computer Applications

Wednesday, July 30, 9:30 am 5:00 pm

obtained with the maximum rate of agglutination at the lowest everolimus concentration
and the lowest agglutination rate at the highest everolimus concentration. The assay
consists of ready-to-use reagents, calibrators (value-assigned concentrations at 0, 1.5,
3, 6, 12, and 20 ng/mL) and controls (value-assigned concentrations at 4, 8, and 15
ng/mL).
Results: The performance of the QMS Everolimus Assay was evaluated on the
Beckman Coulter AU480/AU680/AU5800 analyzers. All studies were evaluated
using CLSI guidelines. Three levels of everolimus controls were used in the studies.
The precision ranged from 7.1%CV to 6.4%CV for within-run and 10.5%CV to
8.1%CV for total run. Linearity was measured and confirmed over a range of 1.5 ng/
mL to 20 ng/mL. The least detectable dose on the AU480/AU680/AU5800 yielded
0.3 ng/mL. Patient correlation studies: AU480=0.90(Hitachi 917) + 0.18 (N=107,
r=0.99), AU680=0.94(Hitachi 917) + 0.0 (N=100, r=0.99), AU5800=0.99(Hitachi
917) + 0.13 (N=106,r=0.98).
Conclusion: All measured studies demonstrated acceptable performance, validating
the use of the QMS Everolimus Assay on the Beckman Coulter AU480/AU680/
AU5800 analyzers, and providing an effective monitoring system for patients
receiving everolimus therapy.

B-017
Error Rate Testing for the Accelerator p540 Preanalytical Sample Processor
Vision System

A. DeFrance, B. Smith, K. Reed, Y. Smith. Abbott Laboratories, Irving, TX

Introduction: The ACCELERATOR p540 is a fully automated perianalytical sample
processor that performs sample loading, identification, decapping, aliquoting, and
sorting operations. A Vision System was added to the p540 aliquoter unit to identify
cap color, if present.. Parameters link the cap color to the specimen type. Middleware
determines compatibility of specimen type for the tests ordered. Depending on that
determination the tube can either be further processed, or moved to an error location
subject to user preference.
Methodology: The vision system library contains 35 common cap colors.
Additionally, it can be trained to read other cap colors. The vision system consists of
a digital camera that acquires images of racks as they are presented to the aliquoter.
This picture is available for operator viewing. The cap color as detected by the camera
and the pixel data for each cap is sent to the p540 software. The software analyzes the
data and determines whether the cap matches a trained color. In the error rate test, 12
different caps were tested on two different systems. These caps consisted of a mixture
of rubber, plastic and screw caps of various colors. Some caps were part of the library
and some were trained.

identify the correct color. However, it is a simple process to train any cap color. This
ability to train any cap color keeps the error rate low and allows for changes in the
color manufacturers caps without having to reconfigure the camera.

B-018
Utilising Information Technology (IT) to improve work processes at the Satellite
Laboratories

S. Lim, C. Yeo. Singapore General Hospital, Singapore 169608, Singapore

Background: Singapore General Hospital (SGH) Clinical Biochemistry Laboratory
has under its charge; 9 satellite laboratories that are situated in polyclinics (publicfunded primary healthcare facilities). These satellite laboratories offer phlebotomy
services and some onsite laboratory tests, although the bulk of the specimens are
sent back to the central laboratory for analysis. HbA1c (Glycated Haemoglobin
A1) constitutes a significant 38% of the onsite testing repertoire. The objective of
this study was to utilise information technology to improve the otherwise manual
processes for handling HbA1c results and to achieve a standardised workflow across
the 9 satellite laboratories. Laboratory turnaround time (TAT) for HBA1c was used as
a performance indicator to measure the success of this endeavour.
Methods: HbA1c analysers were connected to the Laboratory Information System by
the first quarter of 2012. Standard procedures for specimen registration and processing
and result verification were instituted at the laboratories in the last quarter of 2012
and autoverification (AV) of HbA1c test results was piloted at 2 laboratories in July
2013. TAT, defined as the time taken from registration at the lab to result verification.
Results: Significant improvement in TAT was observed at all the satellite laboratories
after the introduction of LIS connectivity and a standardised workflow. By the end of
2012 >90% of HbA1c results were completed within 30 minutes, a betterment over
the pre-study rate of 50% completion within 30 minutes. This improvement was also
sustainable from 2013. AV was piloted in July 2013 at 2 of the satellite laboratories to
explore its effect on the TAT. Data showed that TAT slipped initially due to operator
unfamiliarity but improved in the following months, enabling both laboratories to
complete >80% HbA1c testing within 15 minutes.
Conclusion: Online connectivity of the analysers and standardisation of work
processes have definitely improved the efficiency of onsite HbA1c testing at the
satellite laboratories. Auto-verification of test results, with careful planning, can
also improve TAT. Our next steps will be to sustain the work process improvements,
implemented AV for the remaining 7 satellite laboratories and study the impact of our
enhanced practices on the overall operations of the polyclinics.

B-019

Results: The table below summarizes the p540 Vision system error rate testing.
p540 Error Rate Testing
System 1
System 2
Trained/ Original
Error Trained/ Original
Error
Cap Type
Retrain
Retrain
Library
error rate
rate Library error rate
Rate
Greiner:
Trained
0%
No
0% Trained 0%
No
0%
lavender
Greiner:
Library
100%
Yes
0% Library 100%
Yes
0%
white
Terumo:
Library
12%
Yes
0% Library 8%
Yes
0%
red
Terumo:
Library
20%
Yes
0% Library 0%
No
0%
green
BDPlastic: lt Library
0%
No
0% Library 0%
No
0%
green
BDPlastic:
Library
0%
No
0% Library 0%
No
0%
gold
Sekisui:
Library
0%
No
0% Library 0%
Yes
0%
tan
Sekisui:
Library
16%
Yes
0% Library 36%
Yes
0%
gray
BDRubber:
Library
0%
No
0% Library 0%
No
0%
red
BDRubber:
Trained
0%
No
0% Trained 0%
No
0%
blue
Sarstedt:
Library
100%
Yes
0% Library 80%
Yes
0%
orange
Sarstedt:
Library
0%
No
0% Library 4%
Yes
0%
lavender
Conclusion: The error rate of the p540 vision system is low. As seen in the retrain
column, if the cap color exists in the library the system may or may not initially

S136

Implementing the Integrated Laboratory tool for ANF tests in a highfunctioning laboratory

F. B. Flumian1, R. Mansur1, R. J. Marani1, A. Cintra2, F. O. Paiva1, G.

Lakos3, R. Vrunski2, M. D. Freire4, N. Gaburo1. 1DASA, Sao Paulo, Brazil,
2
Werfen Medical, Sao Paulo, Brazil, 3INOVA Diagnostics, San Diego, CA,
4
DASA, Rio de Janeiro, Brazil
Background: The gold standard method used to screen for antinuclear factor (ANF)
is Indirect Immunofluorescence (IIF) on HEp-2 cells. However, this methodology
presents some problems, such as subjectivity in interpreting the standard reading,
a lack of trained technicians and the tests low standardization. On the other hand,
ANF diagnosis has great clinical importance, which justifies the increasing demand,
year after year. Given this scenario, laboratories processing a large number of ANF
screening tests on a daily basis require alternatives that meet this high test volume
while not compromising diagnosis quality and accuracy.
Objective: To present the implementation of Integrated Laboratory system and
the Quanta-link software that integrates the Quanta-Lyser and NOVA View
equipment with the laboratorys LIS.
Methods: The validation of the equipment used to prepare the HEp-2 slides
(QUANTA-Lyser) utilized 40 samples, 20 positive and 20 negative, and the positive
samples were from different titers and standards. Quanta-Lysers reproducibility
was verified using 4 positive samples and 4 negative samples. The samples were
replicated 5 times in the same session. Both the AFN automated reading equipment,
NOVA View, and the training to read AFNs were used every day. Therefore, 200
AFNs/day were visualized, with comparative readings between the NOVA View
reading equipment and the manual microscope reading, which was carried out by a
team of 8 technicians trained to read AFN.
Twenty one collaborators from different areas of the company were involved in

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Automation/Computer Applications

Wednesday, July 30, 9:30 am 5:00 pm

configuring the systems, infrastructure and participated in the technical-diagnostic

process in order to make this validation possible.
Results: The implementation process lasted 5 months. In the first phase, the
presentation of the project and the installation and validation of Quanta-Lyser were
carried out, followed by the installation, calibration, training and validation of NOVA
View. In the second phase, the implementation of the DAPI reading and the learning
curve was carried out, as well as the implementation and configuration, training and
validation of QUANTA Link. In the third and last phase, the training and validation
of the system as a whole was developed. When compared to the previous method
used, the results found were: a 2.64% reduction in repetitions, a 10% reduction in false
positive results, a 70% reduction in
the use of paper and a 53.6% reduction in the release time of negative samples and
7.1% reduction for positive samples. Qualitative results can also be highlighted:
Improved standardization of the results release, greater precision in diagnosis,
significant improvement in the traceability of samples, reduction in the manual entry
of results, ensuring reduction in possible transcription errors, and reduction in the
causes of repetitive strain injury.
Conclusion: Implementing the Integrated Laboratory was a great challenge
for DASAs manual immunology sector, as it was a project that involved many
collaborators from different sectors and companies, and broke paradigms in the
traditional diagnosis of AFN. It provided important gains for the departments costs
and productivity, with an important gain in analytic quality. The project also led to the
integration of various departments within the company and produced knowledge that
can be used in future projects

B-020
Evaluation of Calibrator and System Stability for Beckman Coulter Access 2
System

N. Akbas1, A. Algeciras-Schimnich1, N. A. Baumann1, D. R. Block1, J. R.

Budd2, S. Gaston2, G. G. Klee1. 1Mayo Clinic, Rochester, MN, 2Beckman
Coulter, Chaska, MN
Introduction: It is standard practice in clinical laboratories to calibrate assays prior
to use and recalibrate at the pre-established calibration expiration time. All automated
assays have a time frame for calibration curve stability set by manufacturers which are
monitored with quality control and assurance systems to ensure calibration integrity
in clinical labs.
Objective: In this research, we investigated the stability performance of calibrators
and the system for three automated assays (human luteinizing hormone (hLH), total
triiodothyronine (TT3) and vitamin B12) performed on the Beckman Coulter Access
2 analyzer.
Methods: Thirty data sets were collected over forty-three days for six levels of
calibrator by testing as unknown and analysis of results was compared to acceptance
criteria. Each data set consisted of four replicates of calibrator levels S0, S1, S4,
and S5 and three replicates of calibrator levels S2 and S3. Three levels of serum
quality control pools for the same three assays also were analyzed for comparison.
In addition, one replicate for three levels of quality control materials (QC1 through
QC3) were performed. The maximum number of days before the calibration exceeds
the defined calibration limits were calculated using a time regression analysis program
(provided by Beckman Coulter) for calibrator levels and QC levels. Two separate
models (percent change versus day model and concentration change versus day
model) were used for analysis. The initial calibration curve was constructed using
total number of twelve calibrator measurements. Percent coefficient of variation (%
CV) for each calibrator and QC levels were calculated for each data set.
Results: Time regression analysis of hLH showed stability that was beyond the
manufacturers stated stability limit (28 days) for all calibrator and QC levels using
both analysis model. The observed CVs were less than 3.5% for all calibrator and
QC levels except QC3 which had CV of less than 5.5%. Results of analysis showed
extended stability compared to manufacturers suggested stability (14 days) for TT3
at all levels of calibrators and QCs using both analysis models. TT3 assay had CVs of
less than 5% for higher levels of calibrator (S2 through S4) and QC (QC2 and QC3)
while CVs were less than 10% for lower levels (S1 and QC1). Most of the results for
vitamin B12 were beyond the recommended 21 day stability: 6 of 7 analyses using
percent versus day model and 4 of 7 analyses using concentration versus day model
had greater stability. The observed CVs were less than 7% for calibrator levels and
less than 16% for QC levels.
Conclusions: Results indicated that measured stability was increased for all three
assays when data was analyzed using percent change as opposed to the concentration
change model. hLH and TT3 assays were stable longer than the manufacturers
recommendations. Results of four analyses for vitamin B12 did not meet the

manufacturers suggested stability limit. It is important to recognize the current

research study includes both calibrator and system stability in a manner not typical of
the manufacturers intended use.

B-021
A multicenter study on the performance of Grifols Erytra, a fully-automated
high throughput analyzer, for Kell grouping in US population

J. Roback1, S. Barclay1, J. M. Moulds2, G. Denomme3. 1Emory University

Hospital, Atlanta, GA, 2LifeShare Blood Centers, Shreveport, LA,
3
BloodCenter of Wisconsin, Milwaukee, WI
Background: After the antigens of the ABO and Rh blood groups, Kell antigen (K)
is the next most immunogenic. The Kell locus is highly polymorphic and gives rise
to many antigens. The most important ones, K and Cellano (k), are produced by two
major codominant allelic genes. Exposure to K antigen can stimulate an IgG type
antibody that can trigger transfusion reactions and hemolytic disease of the fetus and
newborn. However, pre-transfusion K antigen typing is not routinely performed in
the US. The objective of this study was to test Erytra (Grifols, Barcelona, Spain), a
fully-automated high throughput analyzer for pre-transfusion testing, to establish its
performance for K grouping versus comparative method.
Methods: K typing was performed on 2669 samples selected from routine workload
representing a very diverse population of donors and patients (46% and 54%,
respectively) in 2 US sites. The 8-column DG Gel 8 ABO/Rh+Kell cards (Grifols)
containing monoclonal antisera were used for test procedures. Comparative method
was traditional tube testing. Positive Percent Agreement (PPA), Negative Percent
Agreement (NPA) and OPA (Overall Percent Agreement) between the Erytra and
the comparative method were calculated at the 95% confidence level. At least a 99%
concordance was considered acceptable.
Results: Of the 2669 tests performed, the Erytra detected 222 K positive and 2447
K negative. The concordances with the comparative method were 99.09% for PPA
with a 95% lower confidence bound (LCB) of 96.75%; 99.84% for NPA (99.58%
LCB); and 99.78% for OPA (99.20% LCB). Of the 6 discrepancies found (0.22 %), 4
samples were positive by the Erytra and negative by the reference system (all of them
were confirmed as true positives by the Erytra), and 2 samples were negative by the
Erytra and positive by the reference system (all of them were concluded to be true
negative in favor of Erytra). Further investigation revealed that the 2 false negatives
were apparently due to clerical or technical error in performance or predicate test.
Conclusion: The Erytra test performance in the Kell determination with its DG Gel
8 cards was safe and effective, consistently obtained the expected results in all the
repetitions and was substantially equivalent to the FDA-licensed reagents and FDA
cleared instruments used in the study. The Erytra system can be acceptable for K
antigen detection by routine pre-transfusion tests in US centers.

B-022
Workflow efficiency of Erytra in a hospital transfusion service environment

L. C. Winn1, S. Leonard1, C. Snchez2, S. Sol2. 1Grifols, Emeryville, CA,

2
Diagnostic Grifols, S.A., Parets del Valls, Spain
Background: Automated blood grouping systems for hospital blood banks,
transfusion services and donor centers should demonstrate high loading capacity and
self-organization to provide maximum processing power. The Erytra (Grifols) is the
newest generation fully automated blood bank system for blood group determination
and pre-transfusion compatibility testing using the gel agglutination technique that
has been recently approved by the FDA. The objective of the study was to assess the
workflow efficiency of the Erytra in a hospital transfusion service environment in
terms of turn-around times from process start to finish, efficiency and advantages of
the 8-column cards and the ease of use and acceptance by the laboratory staff.
Methods: Patient samples collected in ACD, EDTA or sodium citrate were tested
for ABO/Rh+K (Grifols DG Gel 8 ABO/Rh + Kell card) and antibody screening
(Grifols DG Gel Anti-IgG card and Search Cyte 0.8%) using the Erytra automated
blood banking system. The Erytra was loaded with increasing number of samples. The
following performance metrics were assessed: time to first result (TTFR), turnaround
time from first result to last result (cadence), manual hands on time required and
walk-away time. STAT (urgent) samples could be inserted at different times during
the regular testing. Operators followed the Package Insert Instructions for Use for the
Grifols DG Gel 8 cards, Red Blood Cell Reagents and for the Erytra system. For the
ease of use and acceptance evaluation, the following activities were tracked: Set-up,
QC, sample preparation, sample sort and loading, routine testing, post-run procedures,
consumables used, IT/data review, space requirements, and ergonomics.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S137

Automation/Computer Applications

Wednesday, July 30, 9:30 am 5:00 pm

Results: Consecutive loads of 96, 120, 144, 168 and 192 samples for the ABO/
Rh+Kell typing and antibody screening in the Erytra gave similar values of cadence
from 42 to 46 samples per hour. This means that increasing loads had no negative
effect on the Erytra performance. STAT samples did not modify the cadence of the
routine samples. Most valued feature of Erytra by the staff was related to cards
and reagents continuous loading and traceability tracking. Cards and reagents were
considered easy to load, thus shortening time to start processes. In addition, software
allowed accurate and timely notification of cards tracking and reagent status. Being
a see-through instrument was also a valued feature since its well lighted interior
provided a clear view of all operating processes.
Conclusion: High workflow efficiency Erytra was demonstrated through its increasing
performance with increasing sample loads. Erytra was particularly valued for its ease
of loading and tracking cards and reagents, which lead to efficiency and time savings.

B-023
Optimization of Sample Workflow with Total Laboratory Automation: The
experience of a 4,5 million tests/month Clinical Laboratory

S. V. L. Argolo, M. L. Moreira, C. F. A. Pereira, O. F. Silva. DASA, Rio de

Janeiro, Brazil
Background: Central laboratories today face significant challenges: increasing
number of samples and tests; the need to find ways to reduce costs; limited technical
staff; to minimize laboratory errors; and, to provide results in lower turnaround times
(TAT). Laboratory automation can be the solution for most of these issues, but in an
operational cenario of 4 to 5 million tests/month, the developed solution has to be
carefully and individually studied.
Objective: To evaluate the improvements in efficiency and productivity obtained
by the implementation of a high volume laboratory automation in a large clinical
laboratory.
Methods: We installed, with support from Roche Diagnostics Brazil, two FlexLab
System 3.6 from Inpeco, included in each system an Input Output Module, Rack
Output Module, Bulk Input Module and Decapper Module. In the first one, were
connected the following platforms: 2 Cobas 775, 3 Modular EEE, 2 UniCel DxI
and 1 ImmunoCap 1000. In the second were connected the following platforms: 1
Cobas 775, 1 Cobas 777, 1 Cobas 8000 EEE, 3 Modular EEE, 2 UniCel DxI and 1
ImmunoCap 1000. We then compared pre versus post implementation data, regarding
number of tests processed by tube, TAT, and number of tests by technician.
Results:There was a significant decrease of time consuming in load platforms and
all the actions related to it, increasing productivity, reduction of number of tubes
collected with important gain in costs and a decreased TAT for most parameters. We
had an increase in number of tests per tube by 34%, saving around 48.000 dollars per
month. The productivity per technician was increased in 15% and TAT was reduced,
in average, by 32%.
Conclusion: At the time we started the studies to implement this solution, we were
processing, at the serum area, 2,6 million tests/month using equipments from different
suppliers, and also used a sorter to manage the destination of the tubes. All equipments
were loaded manually by the technicians. This new technology implemented at
our large clinical analysis laboratory allowed us to improve TAT, productivity and
reduce production costs, and to deal with almost 3,2 million tests/month. With the
new platform we were able to absorb an increase of 20% in number of tests, without
increases in personnel costs. Reorganization of the laboratory using automation can
be the solution to support the growing of large laboratories.

B-024
Comparison between the determination of glycated hemoglobin through
automation system and front-loaded on a clinical chemistry analyzer

N. Z. Maluf, L. K. Tirado, M. D. C. Freire. DASA, Barueri, Brazil

Background: Automation systems provide cost efficiency and increased productivity,
by integrating test menu and automating manual steps. However, some assays present
specificities on sample handling. The determination of Hemoglobin A1c (HbA1c),
a diabetes marker, requires mixing the blood samples immediately before testing.
Generally, accurate results will be achieved if the sample is tested within 10 minutes
after mixing. If the sample sits for too long, the red blood cells will settle, only plasma
will be aspirated and false lower results could be reported. In this context, laboratories
discuss if performing HbA1c through automation system would be appropriated. Here,
we compare HbA1c results when samples are front-loaded on chemistry analyzer with
samples loaded on automation system, considering the 10-minute limitation.

S138

Methods: 166 blood samples were tested for HbA1c using the automated pretreatment
kit on Siemens ADVIA Chemistry 2400. The samples were first mixed and loaded
on a STAT tray on Siemens ADVIA LabCell (100 capped tubes and 66 decapped).
Samples were run in batch, without any other samples on the track. The aspiration
time of the samples were obtained from CentraLink Data Management System. Then,
these samples were mixed and front-loaded on Siemens ADVIA Chemistry 2400. The
test was performed in triplicate, mixing the samples between each run. The time to
aspirate all the samples in the tray was measured
Results: The correlation data is summarized in the table below
Conclusion: Correlation results between compared measurements were satisfactory.
However processing through automation system took longer than the manufacture
recommendations (10 min. after mixing). Our study suggest that is possible to process
HbA1c tests in batch on ADVIA LabCell under ideal circumstances of sample
homogenization and system monitoring.
Table 1 - Linear regression and correlation coefficients between automation vs.
front-loaded measure
Total
Aspiration
Correlation coefficient
Sample Handling N
Linear Regression
Time
(R)
(min)
capped tubes
100 19
1,0516x-0,7496
0,9434
deccaped tubes
66 12
0,9782x-0,1449
0,9726

B-025
Manual Verification of Aldolase Reference Materials and Validation of an
Aldolase Assay on an Automated Chemistry Analyzer

J. Wiencek1, E. Olson2, K. Lembright2, S. Lawson2, S. Wang2, E. Reineks2.

1
Cleveland State University, Cleveland, OH, 2Cleveland Clinic, Cleveland,
OH
Background: Aldolase (ALD) is a useful marker of muscle damage and is still
frequently ordered in conjunction with creatine kinase. A commercial ALD kit (Roche
Diagnostics, Indianapolis, IN) is available as a manual method. The test principle of
this kit utilizes the disappearance of NADH UV absorbance in a coupled reaction.
Although we were interested in employing the reagents of this kit for use on an
automated platform, initial calibration materials from third parties were found to
be unreliable for quality control(QC) and calibration purposes. Using a validated
spectrophotometer and the manual method, we verified a set of QC and calibration
materials. These materials, along with method comparison studies, were important in
transferring the ALD test to an automated platform.
Methods: A Beckman DU800 spectrophotometer (DU800) was first used to asses
ALD QC (Roche) following their kits established protocol. Patient samples (n=16)
and a third party calibrator were further assayed and compared to values obtained by
an outside reference laboratory. After this analysis, the kit was modified to work with
the P800 automated analyzer (Roche). Calibration of the analyzer was required to
establish a k-factor in its system. The average value of replicate ALD measurements
(n=6) determined by the DU800 was assigned to the calibrator to be used in this
procedure. Once this parameter was defined, studies were performed to determine
linearity, precision and method comparison of ALD activity in patient samples (n=42).
Results: Roche QC (PreciNorm and PreciPath) assayed on the DU800 fell within
the recommended ranges of ALD activity of 12.4 0.8 and 24.4 1.8 units/L,
respectively. Deming regressions of 16 patients analyzed with both the DU800 and
a reference laboratory showed a slope of 0.998, an intercept of -0.12 units/L, and a
correlation coefficient of 0.9632. The labeled value of a third party calibrator stated
the products lot contained 18.7 units/L of ALD; however the DU800 and a reference
laboratory reported values of 21.2 and 21.7 units/L, respectively. Using this data, the
k-factor was established on the P800 and the automated method was determined to be
linear over a range of 1.8 to 53.1 units/L. Also the analysis of inter-day precision over
20 days revealed a CV of 4.6% for PreciNorm and 2.2% for PreciPath. Final method
comparisons between patient samples (n=42) with the P800 and a reference laboratory
showed a Deming regression with a slope of 1.070 an intercept of -0.19 units/L, and a
correlation coefficient of 0.9136.
Conclusion: We established confidence in the third party calibration material and
vendor QC materials by verification on a manual platform. Our data established that
the manual Roche ALD kit can be adapted to an automated method valid for routine
clinical service.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Automation/Computer Applications

Wednesday, July 30, 9:30 am 5:00 pm

B-026
Automated Band Neutrophil Counts by the CellaVision DM96 Digital Blood
Imaging System

H. Broome1, H. Bengtsson2. 1University of California, San Diego, La Jolla,

CA, 2CellaVision AB, Lund, Sweden
Background: Early recognition of sepsis is critical for instituting life-saving
therapy. Recent studies shows that the neutrophil band count remains one of the most
predictive, readily-available tests for positive blood cultures (1, 2). CellaVision DM96
images blood smears and pre-classifies nucleated cells into leukocyte differential
categories including band neutrophils. Before posting results from the DM96, clinical
laboratory scientists (CLS) must confirm or re-classify the cells in the differential
categories (3). The objective of this study was to compare pre-classification and postreclassification results of DM96 blood smears from emergency department patients to
evaluate the reliability of the DM96 reliably to screen for increased band neutrophils.
Methods: From a total of 180 consecutive complete blood counts plus differentials
performed on ED patients, 64 smears were selected that had DM96 pre-classification
band counts ranging from 0% to 48%. After two CLS independently reclassified the
DM96 results from these 64 blood smears, the pre-classified and post-reclassified
results were analyzed for correlation and agreement using post-reclassification band
percentages as the reference method.
Results: Correlation coefficient (R2) = 0.66 with a linear regression equation of: preclassification band % = 1.06 x (average post-reclassification band percentage) + 0.09.
Using a cutoff of 25% bands for the pre-classification DM96 and 15% bands for the
average of the two CLS post-reclassification results, the positive agreement was 71%,
and the negative agreement was 100%. There were two discrepant values with high
band % by the DM96 pre-classification (26%; 28%) but normal by the two CLS postreclassification (0.9% and 1.7%; 6.1% and 0.9%).
Conclusion: This preliminary study suggests that DM96 pre-classification band counts
reliably screens for increased band counts using a cutoff of 25%. Additional studies
need to be done with more samples with high band counts to determine the optimum
cutoff and to determine the sensitivity, specificity, and predictive value of the DM96
pre-classification band percentage. These results suggest the DM96 pre-classification
probably is more sensitive than specific since the two discrepant samples had higher
DM96 pre-classification results and since pre-classification showed a positive bias
compared to post-reclassification. High sensitivity with a reasonable specificity is
acceptable for a screening test.
References:
1. Chase M et al. Predictors of bacteremia in emergency department patients with
suspected infection. American Journal of Emergency Medicine. 2012;30:1691-1697.
2. Drees et al. Bandemia with Normal White Blood Cell Counts Associated with
Infection. American Journal of Medicine. 2012 125:1124.e9-1124.e15
3. Kratz et al. Performance Evaluation of the CellaVision DM96 System. American
Journal of Clinical Pathology 2005; 124:770-781.

B-027
Automated review of laboratory information system quality assurance (QA)
reports using text analysis

B. P. Jackson, J. M. Toohey, M. S. Berntsen, L. J. McCloskey, D. F. Stickle.

Jefferson University Hospitals, Philadelphia, PA
Background: Laboratory accreditation requirements include regular review of results
reporting. At our hospital, a quality assurance (QA) report generated by the laboratory
information system (LIS) is printed and reviewed daily to verify proper reporting
of all critical results, linearity failure results, and delta checks. In most cases, the
review process is entirely algorithmic -- there is a fixed series of steps and rules by
which most results either pass or fail review. Such a process was amenable to
automation. Our objective was to produce a computer program to read and assess
all pertinent text elements of the QA report so as to eliminate manual review of as
many results as possible.

program verified that results were reported as appropriate for the individual analyte
(e.g., not reported as > for analytes that should have been repeated on dilution).
For delta checks, results were excluded from further consideration by a variety of
rules, such as if the time between the LIS-generated delta check results exceeded 72
hours. Results not meeting acceptance criteria were assigned a non-verified code
and automatically printed for manual review. The results of automated review were
recorded in a summary text file for all results, and both the original LIS QA report and
the summary text file were archived in electronic form.
Results: After VB program development, a validation period of one month was
used to compare manual and automated reviews of the LIS QA reports. With
program refinement, automated review contained no comparison errors. Following
this period, automated review was adopted as the routine procedure for QA review.
Prior to adoption of automated review, printed QA reports were 20-30 pages, and
approximately 30 minutes were required daily for QA review. After adoption,
summary printed reports for non-verified results were 1 page maximum, and less
than 10 min total was required for daily QA review. In projection, it is estimated that
automated QA review will save more than 3 man-weeks per year in labor. Moreover,
automated QA review is a definitively green undertaking, in that it will eliminate
printing and archiving of more than 7000 pieces of paper per year.
Conclusions: Automated review of LIS QA reports was accomplished using a custom
computer program performing text analysis. The program was successful in assessing
common pass/fail criteria to an extent that greatly reduced manual effort and costs
associated with daily results review.

B-028
Characterization and stability of time-of-day patterns of running averages as
potential inputs to patient-based quality control algorithms: examples for basic
metabolic panel analytes in a university hospital

B. P. Jackson, L. J. McCloskey, D. F. Stickle. Jefferson University

Hospitals, Philadelphia, PA
Background: Regular patterns of time-of-day (TOD) variation of running means of
patient data are potentially useful as inputs in patient-based quality control (PBQC).
Characterization of TOD patterns and their stabilities are initial steps to determine
whether use of patterns might improve PBQC. For 7 analytes of a Basic Metabolic
Panel (BMP) at a university hospital, we mathematically characterized average TOD
patterns (t = 0-24 hours) of running means of patient data for one month intervals, and
assessed the stability of such patterns across successive months data.
Methods: Successive one-month datasets (M1, M2) for patient measurements were
obtained for BUN, Ca2+, Cl-, CO2, creatinine (Cr), K+ and Na+. Running means of
length 30 (A30) were calculated across M1 and M2, with data restricted to samples
within each analytes reference range. Independently for both M1 and M2, average
A30 as a function of time-of-day (TOD) was calculated for 48 half-hour intervals
across 24 hours. From these data, TOD-dependence of A30 was characterized
mathematically using low-frequency ( 12/day) Fourier transforms to produce
continuous, smooth functions F1(t) and F2(t). Stabilities of TOD patterns were
evaluated by correlation (r2) between F1(t) and F2(t).
Results: An example TOD pattern is shown for K+ in Figure. The TOD-dependence
of A30 for K+ was relatively stable, as demonstrated by high correlation between
F2(t) and F1(t) (r2 = 0.929). Across the remainder of BMP analytes, there was marked
variation in apparent cross-month stability of patterns as assessed by r2: Cr (0.299),
BUN (0.765), Na+ (0.811), Cl- (0.849), CO2 (0.870), Ca2+ (0.973).
Conclusions: Methods presented here exemplify an approach to characterization
of TOD patterns in running means and assessment of their stability. The stability
assessment results provided a logical order among BMP analytes for further evaluation
of the potential utility of TOD patterns as inputs to PBQC algorithms at our institution.

Methods: Programming was conducted using Visual Basic (VB). The LIS QA
report for a given day was produced as a text file sent to a computer hard drive. The
executable VB program (QADR.exe, named for QA Data Reduction) read the QA
report so as to extract and group cases of critical results, linearity failures, or delta
checks. Information content related to assessment of each condition was extracted
from the original report as defined by non-null characters in fixed locations in the
original text file. For critical results, the program verified whether results were
called back according to standardized comment codes. For linearity failures, the

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S139

Automation/Computer Applications

Wednesday, July 30, 9:30 am 5:00 pm

B-030

Laboratory Automation with the Beckman-Coulter Power Express produces

improved turnaround times for stat and routine chemistries

C. P. Yeo, D. S. L. Lee, C. S. P. Chiong, J. S. Y. Wong, P. R. Chincholkar, K.

S. L. Lim, C. H. C. Tan, W. Y. Ng. Singapore General Hospital, Singapore,
Singapore

B-029
Optimization of Sample Work-Flow and Testing Efficiencies with a Paradigm
Shift in Automated Systems for Clinical Molecular Diagnostics Laboratories

C. Newhouse1, T. W. Myers1, E. Smith1, S. Boyle1, V. Namasivayam1, A.

Nelson1, D. Schoener1, S. L. Moseley2, P. Rodriguez2, M. Lewinski1, J.
Osiecki1, S. Roehrig1. 1Roche Molecular Systems, Inc., Pleasanton, CA,
2
Roche Diagnostics Corporation, Inc., Indianapolis, IN
Introduction: Similar to trends faced by clinical chemistry laboratories 20 years ago,
todays molecular diagnostics laboratories are increasingly challenged with balancing
volume growth across an expanding menu of tests, while maintaining meaningful
productivity and efficiency gains. The challenges currently faced by clinical molecular
diagnostic labs extend from complexities in tubes standardization; increasing demand
for testing from multiple sample types; fluctuations in demands for specific tests;
subjectivity in result reporting to increased demands on labs to streamline workflows
with less hands-on time for applications. With the constant pressure of these variables,
there is potential for work-flow inefficiencies to increase costs and potential risk for
human error associated with result generation.
Objective: To detail a paradigm shift in design features for automation in clinical
molecular diagnostic testing. A description of how innovative integrated analyzer
design features could address the current challenges faced by labs responding to
volume growth in molecular testing.
Methodology: Design features of two fully automated, integrated Real-Time PCR
testing systems, the cobas 6800 System* and cobas 8800 System* will be
detailed. These systems include automation of sample transfer, processing and target
detection, including onboard assay specific reagent cassette storage and handling.
A dedicated lane for urgent bypass testing accommodates the need for sample
prioritization. The cobas 6800 System and cobas 8800 System can accept multiple
primary and secondary tube types with no pre-sorting or batching of tubes or racks.
Through universal sample preparation and PCR profiles (cobas omni process) the
instrumentation can process samples for 3 different assays simultaneously for detection
of HIV, HBV, HCV, and CMV. By taking up to 3 aliquots from one specimen, up to
3 assays can be run from a single patient sample. Sample input parameters include
200uL and 500uL EDTA plasma for testing of HIV-1. A user defined channel for lab
developed tests provides flexibility and benefits of automation using the cobas omni
process. The instrumentation requires very little user interaction and maintenance.
Results & Conclusion: The first 96 results are available in less than 3.5 hours
with an additional 96 results every 90 minutes for cobas 6800 System (or 30
minutes for cobas 8800 System). The cobas 6800 System can report up to 384
patient results in 8 hrs. The cobas 8800 System accommodates higher volumes
with up to 960 patient results in 8 hrs. User interactions are limited to loading and
unloading and periodic removal of waste, reducing hands on time and risk for human
error. Collectively, the combination of the new system design and assay design
enhancements provide a level of automation that approaches systems used in clinical
chemistry, facilitating a paradigm shift for routine clinical molecular diagnostics
laboratories and advancements for both laboratorians and patients.
*The cobas 6800 System and cobas 8800 System are in development and not
available for sale in the US.

S140

Background The Singapore General Hospital (SGH) Clinical Biochemistry

Laboratory, with its need to provide 24/7 coverage for stat and routine chemistry
tests to SGH and its affiliated cancer and heart centers, has constantly strived to
achieve optimal performance and service delivery. In 2007, the laboratory introduced
laboratory automation in the form of the Beckman-Coulter Power Processor
Laboratory Automation System (PP LAS). With its 2007 workload of 5.94 million
tests increasing at a rate of 5-7% a year, the laboratory quickly outgrew its LASs
capacity. In 2013, in tandem with plans to move into new purpose-built facilities,
design considerations were made to assimilate further refined workflow and specimen
routing strategies and to install the latest generation of LAS - now coined the
Beckman-Coulter Power Express (PE). Method The LAS inlet was placed next to
the pneumatic tube station which receives deliveries from Emergency, Intensive Care
and other campus sites. The PE LAS with three centrifuges, decappers, aliquoters
and two 5000-tube refrigerated stockyards was linked to 2 DxI800s, 1 AU680 and 2
AU5822s. Turnaround time (TAT) was defined as that between specimen arrival in the
laboratory and result release (sans auto-verification) to EMR. TAT for stat specimens
from Emergency and Intensive Care was the key performance indicator for the
laboratory. Results and Discussion Pre-PE-LAS TAT for stat specimens was 95% in
45 minutes and 70% in 35 minutes; a good performance achieved through expending
much effort into manually moving stat specimens ahead of routine specimens on the
LAS tracks. With PE LAS, TAT was sustained but with significantly less manual
effort. Data also showed that the PE LAS provided faster turnaround for routine test
orders from the specialist outpatient clinics and wards, hitting >80% in 45 minutes (a
vast improvement over previous <20% in 45 minutes). The improvement in TAT and
workflow could be attributed to: (1) priority identification design layout of the stat/
routine laboratory that places the PE LAS close to the specimen reception area, (2) an
improved LAS and the much higher capacity of the online analysers. The important
features of the Power Express LAS are: (a) a dynamic inlet that allows seamless input
of up to 4 types of specimens (stat, routine, pre-centrifuged and archival) without the
need to initiate the pause or standby mode, (b) ability to support up to 4 centrifuges
(we installed 3 and have space for a fourth) and therefore significantly reducing the
occurrence of oft-cited bottlenecks at the LAS centrifuges, (c) 4-laned specimen
transportation track that allows stat specimens to bypass routine specimens, (d)
online RFID specimen identification that circumvents previous issues with barcode
readers and (e) online high-throughput analysers. Conclusions With close proximity
to specimen reception, workflow refinements and favourable new features of the latest
Beckman-Coulter Power Express LAS, quicker result reporting has been shown. In
addition, provision for a fourth centrifuge and high capacity of the online analysers
permit further growth in the coming years.

B-031
Look before you leap: Developing optimized automated rule sets for reporting
hemolyis, icterus and lipemia based on a priori outcomes analysis

J. M. Boyd1, R. Krause1, G. Waite2, W. Hui2, E. Yadzi2, D. Wilmik2, I. Seiden

Long1. 1Department of Pathology and Laboratory Medicine, University of
Calgary and Calgary Laboratory Services, Calgary, AB, Canada, 2Gamma
Dynacare Medical Laboratories, Brampton, ON, Canada
Objective: The new CLSI C56A guideline directs labs to generate their own policies
for hemolysis(H), icterus(I) and lipemia(L) reporting by automated methods. Here
we describe a process to review and optimize reporting policies for test results with
known interference prior to implementation.
Methods: Table 1 describes proposed commenting and cancelling policies generated
using CLSI C56A vs baseline policies from two high volume community laboratories.
Identical Proposed Rules were applied to the test result data sets from each lab.
Test Data: Lab A, Roche Modular platform, IBM AS400 Custom LIS reporting
541236 tests, for 72 analytes over 1 week. Lab B, Roche Cobas C8000 with Cerner
Millennium/Roche PSM reporting 391515 tests for 45 analytes over 2 weeks.
Any test receiving >5 flagged results was considered for generation of Optimized
Rules including analyte level and clinical significance. Number of Commented and
Cancelled tests was counted for each lab and rule set.
Results: Optimized rules were only required for hemolysis. Comment and cancellation
test counts at baseline were compared to optimized rules for each lab. Hemolysis

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Automation/Computer Applications

Wednesday, July 30, 9:30 am 5:00 pm

rules for Lab A: baseline(comment:3656, cancel:NA) vs optimized(comment:155,

cancel:19) reduced hemolysis flagging by 96%, while for Lab B: baseline(comment:0,
cancel:277) vs optimized(comment:107,cancel:2) reduced result cancellation 99%.
Icterus rules for Lab A: baseline(comment:492) vs. optimized(comment:0) reduced
flagging 100%, while in Lab B: baseline(comment:0) vs optimized(comment:13)
increased flagging. For lipemia Lab A: baseline(comment:2383) vs
optimized(comment:24) reduced flagging 99%, while in Lab B baseline(comment:74)
vs optimized(comment:193)increased flagging 62%.

mmol/L) was used to determine the performance of the algorithm to detect inaccurate
results, which were identified by the difference between observed and predicted
potassium.

Conclusions: Implementation of identical rule sets in Labs A and B indicated that the
outcomes of automated HIL reporting are significantly lab dependent . This process
of testing and optimization of HIL reporting rules prior to implementation by a priori
outcomes analysis demonstrates the clear benefit of impact assessment for reporting
policies with automated HIL rule sets.

Conclusions: The model described herein represents a powerful new quality tool
whereby predicted concentrations could be used to prevent reporting grossly inaccurate
potassium results. This is particularly valuable given the clinical importance of
potassium and abundance of preanalytical considerations. Routine application of the
model (e.g. as a autoverification rule) could prevent reporting of inaccurate potassium
results due to unforeseen preanalytical factors.

Table 1: Baseline, proposed and optimized reporting rules

Baseline Rules
Proposed Rules
Optimized Rules
Test result reported,
Test result reported,
Manual visual
Automated HIL
Lab inspection of
detection &
application of rules,
A
sample with
Test result reported,
application of HIL
comment on direction
Automated HIL
flag, no comment
of interference up
detection and
to H of 600, cancel
application of
H>=600, only when
Hemolysis
rules,
comment
Test result not
CLINICALLY
on
direction
of
reported, Automated
SIGNIFICANT*.
interference
up
to
Lab HIL detection
* defined as a degree
H of 600, cancel
B
and application
of interference
H>=600
of rules with test
sufficient to generate
cancellation
a test result outside of
the normal range for
the test
Test result reported, Test result reported,
Manual visual
No cancellation,
Lab inspection of
Automated HIL
A
sample with
Same as proposed
detection and
Icterus
application of HIL application of
rules
flag, no comment rules, comment
Test result reported,
Lab
no flag, no rules in on direction of
B
interference
place
Test result reported,
Manual visual
Lab inspection of
Test result reported,
A
sample with
no cancellaton,
application of HIL Automated
flag, no comment HIL detection
Same as proposed
Test result reported,
Lipemia
with comment
rules
Automated
AFTER Manual
HIL detection
Lab
ultracentrifugation
with comment
B
process applied
AFTER Manual
ultracentrifugation
process applied

Results: A comparison of the predicted and accurate potassium concentrations is

shown below. The most important predictors were creatinine, urea, sodium, anion gap,
and WBC. Based on simulations, the algorithm detected an error of 0.5 mmol/L with
a sensitivity of 78% and specificity 74% and an error of 1.0 mmol/L with a sensitivity
of 86% and specificity of 96%.

B-033
Differentiation between glomerular and non-glomerular hematuria by an
automated urine sediment analyzer

P. V. Bottini, B. D. Andreguetto, K. Krempser, J. Lauand, C. R. Garlipp.

Univesity of Campinas, Campinas, SP, Brazil

B-032
Identification of Erroneous Potassium Results Using a Laboratory DataDerived Machine Learning Algorithm

C. R. McCudden, M. Henderson. The Ottawa Hospital, Ottawa, ON,

Canada
Background and Objectives: Potassium is a critically important analyte with many
preanalytical considerations. While laboratories have procedures to avoid reporting
erroneous results, it is difficult to identify if a given potassium result is accurate. The
objectives of this study were: 1) to use commonly available laboratory data to develop
a machine learning algorithm to predict potassium concentrations 2) determine the
performance of the algorithm for error detection.
Methods: LIS data was used to develop a random forest regression model to predict
potassium results. The prediction algorithm was trained using known, accurate
potassium concentrations with commonly available hematology (complete blood
count) and biochemistry (basic metabolic panel) data. Data included 2876 result sets
(80% for model training, 20% for testing) from 1642 patients with encounters at 174
different hospital/clinic locations over a two-year period. Accurate potassium was
defined as values where blood gas (GEM4000) and chemistry analyzer (Vista 1500)
values were within +/- 0.4 mmol/L. Error simulation (introduced bias from 0.5-2.0

Background: Differentiation between glomerular and non-glomerular hematuria by

observation of the erythrocytes morphology using phase-contrast microscopy has been
very well established for almost 35 years. However, it is a time-consuming and laborintensive procedure that requires skilled personnel. Some years ago, an automated
urine sediment analyzer based on the KOVA method with on-screen review of the
images was introduced. The aim of this study was to evaluate the performance of this
image based automated sediment analyzer (UriSed, also called sediMAX in some
countries) as an alternative to the phase-contrast microscopic analysis of erythrocytes
morphology.
Methods: We studied 312 urine samples with hematuria (erythrocytes>5/hpf).
Samples were analyzed by UriSed and all the images reviewed by an experienced
analyst. Parallelly the urine samples were centrifuged (10 mL, 5 minutes, RCF = 400)
and the sediment (0.5 mL) was placed on a slide and examined under a coverslip by
phase-contrast microscopy. Erythrocytes morphology was analyzed by both methods
by different observers. Based on the presence of codocytes and/or acanthocytes,
samples were classified as non-glomerular (absence of codocytes or acanthocytes)
and glomerular (presence of codocytes, acanthocytes or both). Kappa correlation was
used to assess the agreement between both methods.
Results: Our data showed an agreement of 97.4% between erythrocytes morphology
analyzed by both methods (kappa=0.9484, p<0.001). From 312 samples, 140 of
them (45%) presented isomorphic erythrocytes and hematuria was classified as nonglomerular by both methods whereas in 164 samples (52.5%) we observed the presence
of codocytes and/or acanthocytes by phase contrast microscopy and by UriSed being
classified as glomerular hematuria. Only 8 samples (2.5%) had discordant results. Five

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S141

Wednesday, July 30, 9:30 am 5:00 pm

of them revealed the presence of codocytes by phase contrast microscopy which were
not displayed on UriSed. On the other hand, 3 samples classified as non-glomerular by
phase contrast microscopy presented codocytes on UriSed images.
Conclusion: UriSed is a precise and accurate alternative to the gold standard phasecontrast microscopy that allows a better workflow and may significantly improve
turnaround time.

B-034
Moving Patient Averages: A Pilot Study Using Error Simulation

R. Ybabez, J. Boysen, A. Zblewski, B. Hendrix, C. Wittwer, L. Rossini,

S. Chanakarnjanachai, L. Duncan, D. Block, N. Baumann. Mayo Clinic,
Rochester, MN
Background: Robust quality control (QC) processes in clinical laboratories are
required to ensure stable operation of analytical systems and to provide reliable test
results. There is potential added value in using real-time patient data to supplement
traditional statistical QC. The concept of monitoring moving averages of patient
results has been discussed for decades, and recently middleware programs that can
calculate moving averages continuously and in real time have become commercially
available. The objectives of this study were to evaluate Moving Averages software
by: (i) collecting moving average patient data for three common analytes (calcium,
chloride and creatinine) in a high volume clinical laboratory, (ii) configuring analytespecific protocols and (iii) testing the protocols using simulated systemic errors.
Methods: All patient data were generated using Roche Cobas 8000 reagents and
analyzers. Moving Averages (Data Innovations, Inc) protocols were configured for
calcium, chloride, and creatinine. For each analyte, the mean of patient results and
Sp (standard deviation of the patient population) were calculated over a 2 week
time period. Sa (standard deviation of the analytic method) was obtained from the
standard deviation (SD) of in-use QC at concentrations near the patient population
mean concentration. Sp/Sa was calculated and power function charts (Cembrowski,
GS et al. (1984)) were used to estimate the appropriate number of patient test results
to average (N). Exclusion criteria were applied to calcium and creatinine protocols
to exclude values >4SD from the mean and results from patients in dialysis units
were excluded for creatinine. Systemic 2SD errors were simulated by analyzing
consecutive patient samples with results approximately 2SD above the patient mean.
The concentrations of analyte in the simulated error samples were: calcium, 9.5-9.8
mg/dL; creatinine, 1.3-1.4 mg/dL; chloride 104-106 mmol/L. The number of patient
samples with simulated errors needed to trigger an error warning was calculated.
Results: Sp/Sa and suggested number of patient test results to average (N) were 9.8
and N=200 for calcium, 4.6 and N=50 for chloride, and 6.4 and N=70 for creatinine.
Error simulation showed that 2SD error warning thresholds were triggered after 140
patients for calcium, 50 patients for chloride, and 47 patients for creatinine. A second
simulation study was performed using protocols with different N for each analyte
(calcium, N=100; chloride, N=30; creatinine, N=30) and yielded 2SD error warnings
after 75, 30, and 26 patients for calcium, chloride and creatinine, respectively.
Conclusion: Moving Averages may help laboratories detect systemic errors in
real-time. Three analytes (calcium, chloride, and creatinine) were used to evaluate
the program and optimize parameters for detecting a simulated 2SD error. Errors
simulated using preliminary estimates of the number of patient test results to average
(N) suggested that a smaller N (fewer patient results averaged) may allow errors to be
detected earlier and before large numbers of patient results are affected. The ability to
detect a shift in patient results on a continuous basis and in real time may complement
existing QC processes in the laboratory.

S142

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Automation/Computer Applications

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

plasmids were designed that harbor unique DNA sequences that can be amplified by
the same primer pair sets designed to amplify the individual targets of CT, NG, and
TV. Hybridization of the control amplicons can be detected and distinguished from
hybridization of target amplicons on the integrated DNA array.

Wednesday, July 30, 2014

Poster Session: 9:30 AM - 5:00 PM
Infectious Disease

B-035
Comparative Study of Granada and ChromoID StreptoB media for
identification of Group B Streptococcus

N. Z. Maluf, D. C. Santos, C. L. Chacon, J. Silva, C. F. A. Pereira. DASA,

Barueri, Brazil
Background: BIOMRIEUX StreptoB Granada and ChromoID media are used to
identify group B Streptococcus. Microbiological screening of these bacteria is crucial
since it can lead to neonatal septicaemia in pregnant women. Generally, it appears
in the first 24 hours of life and often results in fulminant septicaemia, meningitis, or
pneumonia associated with high morbidity and mortality. The most effective strategy
to reduce the incidence of GBS in newborns is a pre-natal tracking of all pregnant
women between 35-37 weeks of gestation time. This is done using culture methods
to determine the necessary intrapartum antibiotic prophylaxis. The objective of this
study was to compare the effectiveness of the Granada medium at identifying Group
B Streptococcus in samples of vaginal and perianal openings.
Methods: A 100 samples of vaginal and perianal openings of women with 35-37
weeks of gestation were collected and transported in Stuart and AIMES transport
at room temperature. The procedure was carried according to the package insert.
After 24 hours of incubation at 35 C, macroscopic analyses in the broth were
carried out to detect any orange coloring, and if so, samples were positive for group
B Streptococcus. Those samples were seeded in the ChromoID StreptoB biphasic
Granada broth to compare both tests and verify which one provides more effective
and precise results. Inoculated plates were incubated in micro-aerobic culture systems
at 35 C for a duration of 12 hours.

Results:Several different clinical reference laboratories provided us with

approximately 100 diverse specimens (vaginal swabs, endocervical swabs, and urine
specimens), previously tested using FDA-cleared devices in their facilities. Evaluation
of the same samples with the Rheonix CT/NG/TV CARD assay yielded similar
results. Moreover, since the FDA-cleared test was only able to test for the presence
of CT and/or NG, a number of samples were also found to be co-infected with TV. In
addition, the use of the chimeric plasmid controls yielded positive signals on all runs,
thus confirming that each step of the fully automated assay was properly conducted
by the unattended system. Furthermore, to confirm that the DNA arrays were properly
orientated in the CARD, spotting controls that display signals were placed on the
DNA array in a defined pattern. In order to assure that the proper organisms were
detected, the imaging software was designed to only accept results that displayed
the proper spotting control orientation, thus confirming that the microorganism(s)
detected were correctly scored.
Conclusion:The ability to analyze specimens in a fully automated, sample in_result
out format, will allow detection and identification of three sexually transmitted
infections to be performed by individuals of varying skill level. The automatic
performance in an unattended manner of all sample preparation, DNA purification,
amplification, end-point detection, analysis and readout functions makes the platform
suitable for central lab, point-of-care, as well as non-traditional healthcare settings.
Clinical studies intended to gain FDA clearance are expected to be undertaken in 2014

B-038
A Novel Enzyme-Linked Immunosorbent Assay for the Detection of
Nontreponemal Antibodies in the Sera of Patients with Syphilis.

A. R. Castro1, M. R. Shukla2, J. Deustch1, K. Karem1, H. C. Mody3. 1Centers

for Disease Control and Prevention, Atlanta, GA, 2Arlington Scientific,
Inc., Springville, UT, 3Arlington Scientific, Inc., Sprinville, UT

Results: More than half of the 100 samples analyzed showed incompatibilities
between the results, especially the Granada ID medium presenting an average of less
than 45% positivity when comparing to the ChromoID StreptoB.

Background: We describe an enzyme-link immunosorbent assay (EIA) for the

detection of nontreponemal antibodies. This assay is ideal for the automation of high
throughput screening of sera

Conclusion: In the presented data, three types of tones were considered: strong,
intermediate, and weak. A weak-colored orange still resulted in a clinical evaluation
of the sample and the patient. If the result returned positive, the patient was returned
to isolation and identified as positive for group B Streptococcus. The Granada ID
test was shown unreliable if done alone. In order to achieve better accuracy, this test
should be done with another identification technique, such as the camp-test or even
Chromo ID (STRB). The latter presents excellent specificity and is relatively easy to
execute. Its only limitation is that it takes12 hours longer than Granada ID to achieve
results.

Methods: The nontreponemal (cardiolipin) antigen was chemically modified and it

was attached covalently to amine functionalized micro titer plates.

B-037
Automated Sample-to-Results Analysis of Clinical Specimens for SexuallyTransmitted Infections

Results: A total of 1,006 banked serum samples were evaluated and the results
compared to a quantitative rapid plasma regain (RPR) test. The accumulative reactive
concordance of the nontreponemal EIA was 93.3% when the RPR titer of the sera was
1:1, 96.2% at 1:2, 98.5% at 1:4, 99.3% at 1:8 and 100% at 1:16. The nonreactive
concordance was 100%. Also 50 samples with known stages of syphilis and 158 from
diseases other than syphilis were included.
Conclusion: These results indicate that the nontreponemal EIA test can be used
for the screening of large volume of samples using the traditional syphilis testing
algorithm of screening with a nontreponemal test and confirming the results with a
treponemal test.

B-039

G. Spizz1, C. McGuire1, K. Kowitski1, W. Hungerford1, R. Yasmin1, R. A.

Montagna2. 1Rheonix, Inc., Ithaca, NY, 2Rheonix, Inc., Grand Island, NY
Background:The global burden of sexually-transmitted diseases (STDs) is
considerable with an estimated 340 million new cases occurring each year. Although
many of these new cases could potentially be effectively cured with modern antibiotic
therapy, the early stages of the infections can often go unnoticed. Females are
disproportionately affected, in whom untreated STDs can proceed to disabling pelvic
inflammatory disease which, in turn, can lead to infertility, infant mortality and infant
blindness. Complications in untreated males, although rarer, can proceed to urethritis,
epididymitis, as well as infertility. In order to streamline testing, we have developed a
fully automated molecular detection system to simultaneously detect N. gonorrhoeae
(NG), C. trachomatis (CT) and T. vaginalis (TV) in an unattended manner from a
variety of specimens.
Methods:An injection molded disposable CARD (Chemistry & Reagent Device)
was developed that, when inserted into the EncompassMDx workstation, can
automatically lyse cells, extract and purify DNA, multiplex PCR amplify rRNA
genomic targets in NG and TV and cryptic plasmid DNA of CT. In order to confirm that
all steps of the assay were performed correctly by the system, three separate chimeric

Evaluation of the Alere Determine HIV-1/2 Ag/Ab Combo Kit for the Rapid
Determination of Antibody/Antigen Status in STAT Specimens in a Busy
Metropolitan Hospital Setting

M. Eskandari, S. Barden, K. W. Simkowski, Y. Posey, E. Sykes. Department

of Clinical Pathology, Beaumont Health System, Royal Oak, MI
BACKGROUND: A limited number of 4th Generation (4G) HIV tests is currently
available to detect both HIV-1 and HIV-2 antibodies and the p24 antigen (HIV-1).
Our hospital system currently uses the Alere Clearview STAT-PAK point-of-care
device for HIV-1/2 Antibody STAT testing (needle stick; Labor & Delivery cases).
The Abbott ARCHITECT 4G immunoassay (IA) with reflex to BioRad Multispot
HIV-1/HIV-2 Rapid Test kit is used for routine testing. When the Multispot does not
confirm a reactive ARCHITECT result, serum is referred for further HIV-1 RNA
qualitative testing.
OBJECTIVE: The Alere 4G Determine HIV-1/2 Ag/Ab Combo device was recently
approved by the FDA. The ARCHITECT reports a signal-to-cutoff ratio for combined
HIV-1/2 antibodies and p24 antigen (HIV-1), whereas the Determine distinguishes
the HIV-1/2 antibody result from the HIV-1 p24 antigen result. We were, therefore,

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S143

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

interested in evaluating the Determine as a replacement for the current Clearview
method for STAT testing.
STUDY DESIGN: 111 serum samples from 106 patients, previously tested for HIV
status, were re-tested by the ARCHITECT/Multispot and then by the Clearview and
Determine rapid tests. Reactive samples had been stored frozen for up to 18 months,
however, there was no significant difference between original and repeat results. Most
non-reactive samples had been refrigerated for up to 5 days. ARCHITECT algorithm
results and chart review were used for definitive HIV diagnosis.
RESULTS: Two samples from a HIV-2 antibody positive patient were reactive by all
methods. Sensitivity and specificity for HIV-1 testing are shown:
Test Systems

Sensitivity (%)

Specificity
(%)
77

100
Abbott ARCHITECT 4 Generation IA (n=109)
Alere Determine HIV-1/2 Ag/Ab Combo kit (n=109)
HIV-1 Antibody
90
93
p24 Antigen
17
98
p24 and/or Antibody
98
92
Alere Clearview HIV-1/2 STAT-PAK (n=108)
83
100
BioRad Multispot HIV-1/HIV-2 Rapid Test (n=109) 83
100
CONCLUSIONS: The Alere Determine HIV-1/2 Ag/Ab Combo kit had an overall
sensitivity of 98% for detection of HIV-1 p24 antigen and/or antibody while the
Clearview, which does not detect HIV-1 p24 antigen, had a sensitivity of 83%.
Therefore, the Determine Combo would be an improvement over the current method
for the detection of HIV-positive patients for STAT testing.
th

B-040
Evaluation of Analytical Sensitivity and Workflow of the VERSANT Hepatitis
C Virus Genotype 2.0 Assay (LiPA)

A. Lal, P. Lau, H. Huang, D. Monga, U. Vajapey, R. Nandkeshwar,

G. Kritikos, J. Surtihadi, G. Gorrin. Siemens Healthcare Diagnostics,
Berkeley, CA
Background: The VERSANT HCV Genotype 2.0 Assay (LiPA) is a reverse
hybridization line probe assay that uses sequence information from both the 5
untranslated region (UTR) and the core region to accurately distinguish between HCV
genotypes 1 to 6 and subtypes 1a and 1b. Prior studies have shown that the assay
can genotype 96% of HCV samples with 99.4% accuracy.(1) Assay steps have been
automated to improve efficiency and decreased time to results. This study evaluates
assay workflow and analytical sensitivity.
Methods: The VERSANT HCV Genotype 2.0 Assay (LiPA) is run in three steps:
extraction, amplification, and genotyping. Viral RNA is extracted from plasma or
serum using the VERSANT Sample Preparation 1.0 Reagents. The 5 UTR and core
regions of HCV are amplified using RT-PCR and the VERSANT HCV Amplification
2.0 Kit (LiPA). Biotinylated amplicons are hybridized to immobilized oligonucleotide
probes on nitrocellulose strips and visualized using reagents in the VERSANT HCV
Genotype 2.0 Assay (LiPA) Kit. Processed strips are interpreted using the optional
LiPA Scan software to yield the HCV genotype. Assay intermediates from each
step can either be processed immediately or stored at defined conditions. Analytical
sensitivity was evaluated using one specimen for each genotype (1a, 1b, 2, 3, 4,
5, and 6) diluted separately in serum and plasma. Dilution series were prepared at
concentrations ranging from 50 to 2000 IU/mL, and each target concentration was
tested in multiple replicates and runs with multiple reagent lots on different days.
These data are analyzed using a regression method with probit link function.
Results: Automation of extraction and strip processing allows for simultaneous
processing of 94 samples. Extraction and loading of the PCR plate have been
optimized with the VERSANT kPCR Sample Preparation module, a fully automated
instrument for isolation and purification of nucleic acids using magnetic-bead
extraction technology. Genotyping has been optimized on the automated Auto LiPA
48 Genotyping Instrument (Strip Processor), which can process up to 46 samples and
2 controls per run. Assay times for 94 samples are 3.5 hours for extraction, 4 hours for
amplification, and 4 hours for genotyping (with two Auto LiPA 48 processors), which
includes a hands-on time of 2 hours. Initial assessments of analytical sensitivity,
measured as the limit of detection for individual genotypes/subtypes, was less than
or equal to 500 IU/mL. Further assessments are underway to confirm the analytical
sensitivity.
Conclusion: The VERSANT HCV Genotype 2.0 Assay (LiPA) is a sensitive and
reliable HCV genotyping assay. Automation of the VERSANT HCV Genotype 2.0
Assay (LiPA) workflow results in higher throughput, improved efficiency, and a
decreased time to results.

References: 1.Verbeeck J, Stanley MJ, Shieh J, Celis L, Huyck, E, Wollants E,

Morimoto J, Farrior A, Sablon E, Jankowski-Hennig M, Schaper C, Johnson P, Ranst
MV, Brussel MV. J Clin Microbiol. 2008;1901.
VERSANT HCV Genotype 2.0 Assay (LiPA) is CE-marked in Europe and for
research use only (RUO) in the U.S.

B-041
Quantitation of IFN- and IFN- induced chemokine mRNA expression levels
in active pulmonary tuberculosis patients for effective monitoring of anti-TB
therapy

H. Kim1, S. Kim2, J. Cho3, S. Hong1, Y. Kim1, G. Kim1, S. Ahn1, S. Park1, J.

Kim1, D. Lee4, H. Jin5, S. Park6, S. Cho7, H. Lee1. 1Yonsei University, Wonju,
Korea, Republic of, 2Institute for Life Science and Biotechnology, Seoul,
Korea, Republic of, 3Daegu Health College, Daegu, Korea, Republic of,
4
Hyejeon College, Hongseoung, Korea, Republic of, 5Catholic University
of Pusan, Busan, Korea, Republic of, 6Daegu Hanny University, Daegu,
Korea, Republic of, 7Yonsei University, Seoul, Korea, Republic of
Background: Tuberculosis (TB) that is mainly caused by Mycobacterium tuberculosis
(MTB) remains a major global health problem, with approximately 9 million new TB
cases annually, leading to 1.4 million in 2011. The results of several previous studies
showed numerous cytokines have been implicated in the pathogenesis, diagnosis and
control of MTB infection. Especially, interferon gamma (IFN-)-specific response
to the MTB specific antigens can be used as biomarker for differentiation of active
TB and latent tuberculosis infection (LTBI). However, there is an urgent need of
prognosis markers for tuberculosis (TB) to determine the response to therapy and
improve treatment strategies. Methods: In this study, the messenger RNA (mRNA)
expression levels of IFN- and IFN- induced chemokines (MIG, IP-10 and I-TAC)
were quantitatively measured by using real-time RT-PCR. For effective anti-TB
therapy monitoring, blood sampling, MTB specific Ag stimulation, and molecular
assay were performed with a total of 32 active PTB patients at the time of diagnosis
(before therapy) and after therapy completion (6 months later). Results: The target
genes (IFN-, MIG, IP-10 and I-TAC) mRNA expression levels were significantly
changed and showed a statistical significance at the time of therapy completion from
the initial diagnosis active PTB (IFN-; p=0.0387, MIG; p<0.001, IP-10; p<0.001,
I-TAC; p<0.001). Conclusion: In conclusion, data show that the analysis of IFN-
and IFN- induced chemokine mRNA expression levels after MTB specific Ag
stimulation could provide useful information during the anti-TB therapy of active
PTB patients group.

B-043
Proficiency Test for Laboratory Identification of Corynebacterium diphtheriae in
Health Care System at Lower Northern Thailand

R. Wiriyaprasit, V. Srisakul, D. Boonyod, P. Ratanakasetsin. Clinical

Pathology, Regional Medical Sciences Center 2 Phitsanulok, Phitsanulok,
Thailand
Background: Diphtheria is an acute, communicable infectious disease of upper
respiratory tract caused by toxigenic lysogenized strains of Corynebacterium
diphtheriae. The incident in Thailand often found in June to February. The
immunized vaccination can only reduce the violation of diphtheria, but cannot be
used for extremely protection from re-emerging of diphtheria. Therefore, the rapidly
accurate identification of Corynebacterium diphtheriae for diagnosis is important
for prevention and control. Rehabilitation, practical training laboratories were not
enough for confidential results. The proficiency test (PT) program was an assurance
implement to use for evaluation of efficiency and quality of laboratory.
Methods: Eighteen hospitals in nine provinces of Lower Northern from Thailand
were enrolled in the PT program. The accuracy Identification of Corynebacterium
diphtheriae was evaluated by statistically from three different PT samples which were
homogeneous and stable and were prepared from the same matrix of throat swab in
Amies transport medium .
Results: One hundred percentages (16/16) of PT results have been reported and
returned to PT provider within the period prescribed. Accuracy identification of the
three PT samples was calculated and revealed for 87.5 % (14/16). However, there was
13.5% (2/16) represented inaccuracy with uncorrected results. The satisfaction of PT
program was rated for 90.0 %.
Conclusion: The PT program for quality laboratory identification of Corynebacterium
diphtheriae in health care system at Lower Northern from Thailand was the effectively
implement for the sixteen- laboratory participants .

S144

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

B-044
Screening on Sexually Transmitted Infections among pregnant women

O. Aliyeva. Skin-venereal health center, Ust-Kamenogorsk, Kazakhstan

Background:STI constitute a serious threat to reproductive health of the person in
the form of possible complications or increase in risk of transmission of HIV.50
- 60% of STI in female organism proceed without symptoms, causing serious
consequences such as pelvic inflammatory disease, tubal infertility,ectopic pregnancy.
Children and pregnant women are especially vulnerable concerning STI.The low
health index of women of reproductive age and the complicated pregnancy period
because of transferred STI, lead to the birth of newborns with a low and very low
mass of a body, and is at the bottom of 15% of cases of early neonatal mortality.
Prenatal screening of pregnant women can prevent development of the listed above
complications. In Kazakhstan protection of motherhood and the childhood is one of
the priority directions of a strategic course of development of health care. According
to the resolution of the president of Kazakhstan Republic About Motherhood and
Childhood Protection from 12.12.2003, one of additional expenses for rendering
the state volume of free medical care is inspection of pregnant women on pre-natal
infections, congenital anomalies and STI.
Methods:Within the program - screening in clinic laboratory of Skin - venereal
specialized health center (East - Kazakhstan region) since 2004 to 2012 examination
of pregnant women on DNA STI identification - Chl.trachomatis, Ureaplasma spp.,
Mycoplasma hominis, Trichomonas vag., Gardnerella vag. Patient material: the fence
of a material for researches of pregnant women was spent from an urethra and vagina
by disposable urogenital probe in special transport medium. Used method nucleic
acid amplification tests (NAATs) - Polymerase chain reaction (PCR),AmplySens
(Russia).
Results:From 2004 to 2012 were surveyed 19116 pregnant women. Screening was
carried out during different terms of pregnancy. The majority of women passed
screening on early terms. The percentage of detect ability on 5 infections has been
analyzed. On the first place on detect ability is conditionally pathogenic flora
- Ureaplasma spp. - 41,8 %, Gardnerella vaginalis - 32%, Mycoplasma hominis 17%. Among pathogenic flora on first place on detect - Chl.trachomatis - 10% of
and Trichomonas vag. - 4%. The share of mono-infection has made - 38%, mixed
- infection - 62%. The highest percentage of occurrence of different commensal
belongs to the association - Ureaplasma spp - Gardnerella vag. - 45%, Ureaplasma
spp. - Mycoplasma hom. - Gardnerella vag. - 17%, Chl.trachomatis - Ureplasma spp.
- 8%, Chl.trachomatis - Trichomonas vag. - 3%.
Conclusion:Factors which strengthen potential pathogenicity are: violation of the
immunological reactivity. Change of hormonal background. All these factors promote
development of diseases of a small pelvis with massive colonization of the urogenital
tract. Prenatal screening of pregnant women can prevent the development of premature
placental abruption, uterine inertia fetal hypoxia, placentation abnormalities of the
fetus in STI.

B-045
Traditional clinical laboratory tests for Dengue fever diagnosis in a childrens
hospital, So Paulo, Brazil.

L. R. Almeida. DASA, So Paulo, Brazil

Background Dengue fever (DF) and dengue hemorrhagic fever (DHF) the more
severe form of dengue illness. Dengue viruses are transmitted though the bite of
an infected mosquito Aedes aegypti is the primary mosquito vector, however other
species can also be vectors of Dengue virus. Illness caused by dengue viruses can
range from nonspecific febrile illness, as the most DF cases, to more severe illness
with bleeding, thrombocytopenia, and plasma leakage in cases of DHF. Dengue
incidence and prevalence are rising in endemic areas of the tropical and subtropical
regions. Dengue infections occur in more than 100 countries in the Asia-Pacific
region, the Americas, the Middle East, and Africa, and cases of infection continue
to rise worldwide.3-5Approximately 50 million infections are estimated to occur
each year.3 Dengue incidence rates are increasing mainly in tropical and subtropical
regions of the world, and in the Americas, a dramatic increase of cases has been
reported during the last decades.
Methods: We establish a classification of the results according with the clinical
laboratory findings by comparing the results of the tourniquet test and the platelets of
samples from all children attended on a Childrens Hospital in So Paulo, SP, Brazil.
Results: 249 tests were analyzed from January to December 2013, March, April and
May had the highest number of tests, we divided the classification of the results in
four groups, group A negative tourniquet test with normal platelets count, group B

negative tourniquet test with low platelets count (thrombocytopenia), group C positive
tourniquet test with normal platelets count and group D positive tourniquet test
with low platelets count (thrombocytopenia). Group A had the highest combination
201 tests, negative tourniquet test with normal platelets, this result shows that for
all suspected cases we have a low incidence of positivity, group B 24 combination
(negative tourniquet test with low platelets count), group C 16 combinations (positive
tourniquet test with normal platelets count) and group D 4 combination (positive
tourniquet test with low platelets count (thrombocytopenia).
Conclusion: Traditional tests for diagnose Dengue Fever is not the gold standard,
we expected to have a high number of combination os group D (positive tourniquet
test with low platelets count (thrombocytopenia) which shows more evidence to
diagnose DF however group A showed us that both tests are still very much solicited
by clinical and is not helpful for the diagnose. This way laboratory did not contribute
to the DF diagnose though Its available other tests that can substitute and be helpful
such as serology (had number of tests during this period), rapid test and also PCR.
There are currently more advanced tests for the diagnosis of dengue fever of which
confer higher sensitivity and specificity and contrite to more accurately diagnosing
the disease.

B-046
A Multiplex Real-time PCR Assay for the Rapid Detection of the mecA Gene
with Staphylococci Directly from Positive Blood Cultures

J. Kim1, H. Kim1, H. Wang2, S. Kim3, Y. Kim1, S. Park4, E. Choi4, G. Kim1,

S. Ahn1, S. Park1, H. Jin5, S. Park6, Y. Kim7, Y. Uh4, H. Lee1. 1Department
of Biomedical Laboratory Science, College of Health Sciences,
Yonsei University, Wonju, Korea, Republic of, 2M&D, Inc., Wonju Eco
Environmental Technology Center, Wonju, Korea, Republic of, 3Institute for
Life Science and Biotechnology, Yonsei University, Seoul, Korea, Republic
of, 4Department of Laboratory Medicine, Yonsei University Wonju
College of Medicine, Wonju, Korea, Republic of, 5Department of Clinical
Laboratory Science, College of Health Sciences, Catholic University of
Pusan, Busan, Korea, Republic of, 6Department of Clinical Laboratory
Science, College of Health and Therapy, Daegu Hanny University, Daegu,
Korea, Republic of, 7Department of Internal Medicine, Yonsei University
Wonju College of Medicine, Wonju, Korea, Republic of
Background : Sepsis causes increasing morbidity and mortality, particularly in
elderly, immunocompromised patients, and it represents one of the greatest challenges
in intensive care medicine. Staphylococci are the most commonly isolated organisms
accounting for almost 50% of sepsis. In addition, Methicillin-resistant Staphylococcus
aureus (MRSA) is the most prevalent cause of sepsis and is recognized as a major
nosocomial pathogen. This study aimed to evaluate a newly designed multiplex
real-time PCR assay capable of the simultaneous detection of mecA, S. aureus, and
coagulase-negative staphylococci (CoNS) in blood culture specimens.
Methods : The Real-MRSA and -MRCoNS multiplex real-time PCR assay
(M&D, Republic of Korea) uses the following TaqMan probes which were labeled
with different fluorophores (FAM, HEX, and Cy5, respectively): 16S rRNA for
Staphylococcus species, the nuc gene for S. aureus, and the mecA gene for methicillin
resistance. For blood culture, two or three pairs of culture bottles for aerobes or
anaerobes were incubated in the BacT/Alert 3D (bioMrieux, Marcy, France),
BACTEC 9240 system (Becton Dickinson Diagnostic System, Spark, MD, USA),
or the BACTEC FX (Becton Dickinson) blood culture systems for 5 days after
inoculating with blood drawn from the patient at the bedside. The identification of
organisms and antimicrobial susceptibility tests (ASTs) were conducted by the
microplate method, the MicroScan system (Siemens Healthcare Diagnostics,
Sacramento, CA, USA), and the Vitek 2 system (bioMrieux, Durham, NC, USA).
Results : The multiplex real-time PCR assay was evaluated using 118 clinical
isolates from various specimen types and a total of 350 positive blood cultures from a
continuous-monitoring blood-culture system (CMBCS). Cycle threshold (CT) values
were used to determine the limit of detection. The detection limit of the multiplex
real-time PCR assay was 103 CFU/mL for each gene target. The results from the
multiplex real-time PCR assay for the three targets were in agreement with those of
conventional identification and AST methods except for one sample. The sensitivities
of the multiplex real-time PCR kit were 100% (166/166), 97.2% (35/36), and 99.2%
(117/118) for 16S rRNA, nuc, and mecA genes, respectively, and the specificities for
all three targets were 100%.
Conclusion : The Real-MRSA and -MRCoNS multiplex real-time PCR assay
is very useful for the rapid and accurate diagnosis of staphylococcal blood stream
infections (BSIs). Moreover, the multiplex real-time PCR assays may provide the
essential information to accelerate therapeutic decisions for earlier and adequate
antibiotic therapy based on detection of the mecA gene.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S145

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

B-047
Field Test of the Dynex M Multiplexed Assay System in the Democratic
Republic of Congo Using Dried Blood Spots

S. Higgins1, A. Karmali1, M. Poncheri1, A. Fusellier1, P. Mukadi2, J.

Ngamboli2, N. Kavira2, R. Doshi3, N. Hoff3, E. Okitolonda-Wemakoy4,
J. Muyembe-Tamfum2, A. Rimoin3. 1Dynex Technologies, Inc., Chantilly,
VA, 2Institut National de Recherche Biomdicale, Kinshasa, Congo,
Democratic Republic of the, 3UCLA Fielding School of Public Health,
Los Angeles, CA, 4Kinshasa School of Public Health, Kinshasa, Congo,
Democratic Republic of the
Background:The Dynex Technologies, Inc. M multiplex chemiluminescent
immunoassay platform was selected as the processing platform for an MMRVT
immunity assessment in support of the 2013 Democratic Republic of Congo
Demographic Health Survey (DRC-DHS). Within five months and in collaboration
with University of California, Los Angeles, Fielding School of Public Health (UCLAFSPH) Dynex was able to deliver a fully functional automated processing system,
reagents and adequate assay plates to process 10,500 dried blood spot (DBS) samples
to Kinshasa, DRC.
Methods:Polystyrene beads coated separately with antigen to Measles, Mumps,
Rubella, Varicella-Zoster Virus and Tetanus were immobilized within 54-well M
assay strips with 10 beads per well. Three separate within-well positive control beads
were coated with horseradish peroxidase, total human IgG, and polyclonal goat antihuman IgG. Two negative control beads were coated with MRC-5 and E6 cell lysate.
423 dried blood spots were anonymously collected from children visiting Kinshasa
health centers during the pilot study and more than 8,500 samples collected during
the nationwide principal DHS survey. Positive control DBS were made using a
5-donor pool of normal defibrinated serum. Negative control DBS were made from
pooled normal IgG-stripped serum. Each DBS was extracted into 1ml of PBS, 0.5%
tween20, 5.0% dried milk and processed on a modified Dynex automated DS2
ELISA processing system. Optimization runs in the DRC included examination
of DBS spotting order, DBS extraction time, two different anti-human IgG-HRP
conjugates, PBS vs. Tris-NaCl wash, room temperate vs. 37C sample/conjugate
incubation temperatures, and 30 vs. 60 minute sample/conjugate incubation times.
Duplicate DBS reference sets were made using a 32 previously-characterized plasma
samples and a 7-point 4-fold dilution series of pooled positive control into negative
serum. The duplicate reference sets were processed independently in Kinshasa and
the Dynex labs.
Results:During the initial optimization of the M testing platform all samples were
tested in replicate and gave excellent concordance of clinical calls regardless of
processing variables used. Currently, all 423 pilot samples and 1000 DHS samples
have been processed. Sensitivity and specificity of the M system in Kinshasa was
shown to be equivalent to that at Dynex based on extraction of the 32-member
reference set as well as to fresh dilutions of the control sera. Extraction of the 7-point
DBS calibration series in the DRC shows an equivalent assay response to the same
set extracted in the Dynex labs. Samples tested in replicate during optimization runs
in Kinshasa gave 92% concordance of clinical calls regardless of processing variables
used, with discrepant results found within the indeterminate range.
Conclusion:As shown by the speed of assay development, having been deployed to a
substantially resource-limited environment, and agreement of replicates regardless of
processing conditions the Dynex M multiplex immunoassay system has shown itself
to be a very robust assay platform with excellent sensitivity and specificity. The use of
this system in conjunction with DBS processing offers a very cost-effective automated
multiplexed immunoassay processing system in challenging environments.

B-048
Laboratory diagnosis of viral respiratory tract infections in a Childrens
Hospital in So Paulo, Brazil, one year study

L. R. Almeida. DASA, So Paulo, Brazil

2009, Influenza B, Parainfluenza1, 2, 3 and 4, Syncycial Respiratory Virus (SRV)

A and B, Adenovirus, Bocavirus, Metapneumovirus, Coronavirus, Enterovirus and
Rhinovirus.
Results: 1394 respiratory samples were tested by the respiratory virus panel and 67%
of the tests were positive for at least one virus. The months with higher positivity
were from March to July corresponding to the beginning of autumn and winter,
respectively, in Brazil. April showed the major positivity, 87% when compared with
the other months. The most frequent viruses identified in this period of time were SRV
36%, Bocavirus 14%, Metapneumovirus 7,2% and Adenovirus 6,3%. The samples
were collected from children aged 0 to 14 years and the positivity was higher in young
children under 2 years old with 80% of the positive samples. Influenza A H3N2 was
detected in two samples during the year 35 samples were positive for H1N1. Most of
he results were provided to the physician in two days after collection.
Conclusion: The viral molecular panel detected a wide range of respiratory virus
with high sensitivity, including more than one virus in the same sample. The rapid
result, two days, is important for the etiologic diagnosis of respiratory infections and
infection control measures for the patients admitted to the hospital.

B-049
The First Isolates of the Emerging New Delhi Metallo--lactamase in a
Laboratory in Rio de Janeiro, Brazil

A. Chebabo1, M. G. Quiles2, T. T. Rocchetti2, L. C. Fehlberg2, E. J. U.

Kusano1, D. Thielmann1, R. M. G. Pereira1, O. Fernandes1, A. C. Gales2,
A. C. C. Pignatari2. 1DASA, Rio de Janeiro, Brazil, 2Laboratrio Especial
de Microbiologia Clnica. LEMC/ALERTA. Departamento de Medicina.
Universidade Federal de So Paulo, So Paulo, Brazil
Background: Antimicrobial resistance is a growing global challenge to human
health. The emerging New Delhi metallo--lactamase (NDM), an acquired class B
carbapenemase has gain public attention due to its extended hydrolysis of -lactams
including carbapenems. Objective: We report the first isolates of this emerging
resistance mechanism in our hospitals in Rio de Janeiro, Brazil. Methods: Between
September and October 2013, four carbapenem-resistant strains, three Enterobacter
cloacae and one Providencia rettgeri were isolated in our clinical Microbiology
Lab from distinct patients hospitalized at different hospitals located at two distinct
cities of Rio de Janeiro. MIC was determined by CLSI broth microdilution method.
Specific primers were used for PCR detection of blaNDM, blaKPC, blaGES, blaSPM,
blaGIM, blaSIM, blaCTX-M, blaSHV, blaTEM, blaOXA-48, armA, rmtA, rmtB,
rmtC, rmtD, rmtG, npmA followed by DNA sequencing. Clonal relatedness among E.
cloacae isolates was examined by PFGE. Plasmid extraction was performed by Kieser
protocol. Conjugation with E. coli J53 (LacZ Nalr Rifr) and hybridization with
specific probes were used to determine transfer of carbapenem resistance. The species
identification was confirmed by MALDI-ToF MS and 16sRNA DNA sequencing.
Results: Of the four isolates, three were from public hospital and one from a private
hospital. Enterobacter cloacae were isolated in blood, ascitic fluid and rectal swab
and one Providencia rettgeri was isolated in urine. All strains showed higher resistant
rates to carbapenems and to broad-spectrum cephalosporins and P. rettgeri was
resistant to polymyxin B, as expected. blaNDM-1 and blaTEM-1 were identified in
all isolates. All E. cloacae also produced blaCTX-M-15, and showed different PFGE
pattern. blaKPC-2 was identified in one E. cloacae isolate (isolate E134) and armA
gene was detected in E. cloacae gentamicin and amikacin resistant (isolate E133).
Conjugation of blaNDM-1 was achieved in 2/4 isolates, E. cloacae (isolate E134)
and P. rettgeri (isolate E132). The hybridization revealed that blaKPC-2 was located
in the E. cloacae chromosome. The genetic location of blaNDM-1 is being confirmed
among the isolates evaluated. Conclusion: Those isolates in Rio de Janeiro, showed
the importance of correct molecular study of isolates that express carbapenenem
resistance in the clinical Microbiology Lab, as, although the blaNDM-1 has been
previously reported in Enterobacteriaceae clinical isolates in our country, this study
constitutes the first one that identified the co-association of blaNDM-1 and blaKPC-2
in E. cloacae. The results can lead to improve infection control measures to avoid its
spread in hospital environment.

Background Viruses are recognized as the major cause of respiratory tract infections,
particularly in children. Emerging virus, such as Influenza H1N1, Metapnenumovirus
and Bocavirus are detected by Molecular Biology methods, with high sensitivity
and specificity. Frequently more than one virus are detected in the same sample and
considered responsible for these infections.
Methods: Data from the results of laboratory tests for viral respiratory infections
were collected, from January to December 2013, for patients attended a childrens
hospital in the city of So Paulo Brazil. The test utilized was the RT-PCR Microarray:
CLART Pneumovir virus panel that detects Influenza A, Influenza A H1N1 strain

S146

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

controls, was not observed in our patient group. This discrepancy might be caused
by low vitamin D levels. We advocate measuring vitamin D levels in HBV infected
patients; furthermore, we suggest vitamin D supplementation in deficient individuals.

B-050
New materials for Hepatitis A and Hepatitis B IgM immunoassay calibration
and quality control

A. Doty1, L. OCallaghan2. 1EMD Millipore, Billerica, MA, 2Merck

Millipore, Cork, Ireland
Purpose: We propose that human monoclonal antibodies can be used as replacements
to human serum as controls and calibrators in diagnostic assays for Hepatitis A IgM or
Hepatitis B core IgM immunoassays . These in-vitro produced, standardized products
offer an unlimited and consistent supply of antibody for calibration and quality control
of infectious disease immunoassays. Relevance: Despite advances in public health
and medicine, infectious diseases are persistently counted as a significant cause of
human illness and economic loss. Assays developed for diagnosis and monitoring
infectious diseases require robust, stable and readily available control and calibration
materials. Traditionally manufacturing of these vital materials has depended upon
discovering source plasma units from naturally infected individuals. This material
is increasingly difficult to find. We present data on a new source of material for the
manufacture of controls and calibrators for Hepatitis A Virus VP1 IgM or Hepatitis
B virus core p22 IgM antibodies from immortalized human lymphocytes. Method:
Human lymphocytes from individuals expressing the antibody of interest are isolated
from fresh whole blood by Ficoll. Following in vitro immortalization with EpsteinBarr virus, primary B cells expressing the antibody of interest are fused with a
hybridoma partner. Following an extended growth period hybridomas are screened
to determine if antibody is being secreted. Monoclonality is assessed by the limiting
dilution method. Stability of hybridoma cell lines is assessed and validated through
extended cell culture & passage. Results: Recovery parallel to human serum, antibody
specificity, lack of cross reactivity and performance in several analytical techniques
are shown. Conclusion: These new materials can be used in the formulation of both
calibrators and positive controls by manufactures of diagnostic kits for the detection
of IgM antibodies to Hepatitis A or Hepatitis B. :

B-051
Vitamin D and Vascular Endothelial Growth Factor Levels in Hepatitis B
Infection

A. F. Tuncel, H. Pasaoglu, O. Gumusay, G. Yilmaz, S. Ozenirler. Gazi

University School of Medicine, Ankara, Turkey
Background: In addition to its well-known effect on calcium metabolism, vitamin
D has various roles such as regulation of inflammatory processes, immune response,
cell proliferation and differentiation. Hepatitis viruses can cause inflammatory
liver disease and vitamin D deficiency was reported in patients with hepatitis C
virus (HCV) infection. Vitamin D deficiency is a common finding in chronic liver
disease patients but there is not much data on serum vitamin D levels of hepatitis B
virus (HBV) infected patients. Angiogenesis may be observed during inflammatory
processes and vascular endothelial growth factor (VEGF), which is an important
mediator in angiogenesis, is found to increase in mesentery and liver tissue in cirrhotic
patients. There are different findings on the effect of vitamin D on VEGF expression
in viral hepatitis patients and we aimed to evaluate the relationship between these two
markers in HBV infected patients.
Methods: The study included 57 patients with HBV infection and 19 age-matched
healthy controls. Serum 25-OH vitamin D levels were measured by Advia Centaur
XP chemiluminescence assay (Siemens AG, Germany). Serum VEGF levels were
measured by enzyme linked immunosorbent assay (R&D Sytems, MN, USA). Two
pathologists evaluated liver tissue samples from HBV infected patients. All data
were analyzed using MYSTAT version 12 (SYSTAT, CA, USA). Data is presented as
mean standard deviation. Spearmans rho and Mann-Whitney U tests were used as
appropriate. A test result of p<0.05 was considered statistically significant.
Results: Mean serum vitamin D levels were lower in HBV infected patients by 3.79
ng/mL (27%) compared to the controls (p<0.037). Serum VEGF levels did not show
significant difference between groups. There was no correlation between VEGF and
vitamin D levels in patient population, however, control group showed an inverse
correlation between these markers (r2=0.228, p<0.039).
Conclusion: Vitamin deficiencies are common in various viral hepatitis types and we
showed that the serum vitamin D levels of HBV patients were lower than controls.
Vitamin D is suggested to decrease viral replication so maintaining normal vitamin D
levels might be beneficial in viral hepatitis. Both HBV and HCV genes can lead to an
increase in the expression of VEFG. Vitamin D and its analogs modulate angiogenesis
in viral hepatitis and suggested to be a regulatory factor of VEGF production. The
negative correlation, which is found between serum vitamin D and VEGF levels in

B-054
Use of the MagArray Immunoassay System as a Platform for Pathogenic
Escherichia coli Detection

L. M. Clotilde, A. Juang, H. Yu, L. Carbonell. MagArray Inc, Sunnyvale,

CA
Background: In recent years, pathogenic Escherichia coli have been causing
numerous foodborne outbreaks leading to mild to bloody diarrhea, hemorrhagic
colitis, hemolytic uremic syndrome and even death of patients. Many foods with short
shelf life are related to the public before a negative testing for E. coli is confirmed.
Currently, the process by which regulatory agencies screen for pathogenic E. coli in
foods takes over 3 days. The MagArray immunoassay system is a low-cost chip-based
platform capable of simultaneously detecting up to 80 different analytes in as little
as 10 min. The reduction in detection time of pathogenic E. coli can contribute to a
faster recall of contaminated foods and can therefore limit the number of individuals
ingesting the contaminated food and decrease the total cost of lost productivity and
treatment. The objective of this study was to demonstrate on MagArray platform the
simultaneous detection of two main types of pathogenic E. coli (i.e., O157 and O145)
in ground beef with high sensitivity.
Methods: MagArray chips were first spotted with E. coli O145 and O157 antibodies.
The chips were then blocked and ground beef samples were spiked with E. coli
O145 and O157 for incubation. After incubating with detection antibodies, magnetic
particles were then applied to generate signals. Different concentrations of E. coli
were spiked to establish the standard curve and determine assay sensitivities.
Results: In this 2-plex immunoassay in ground beef, detection of E. coli at a
concentration as low as 2 cfu/l was demonstrated. More specifically, for 2 cfu/uL of
E. coli O145 and E. coli O157, the inter-run CVs were less than 10% for both types.
And the results were compared and agree well with samples spiked to pure buffers.
This sensitivity of detection was achieved using a 30-min assay. And the results
showed that assay sensitivity is minimally affected by changing assay media from
pure buffer to ground beef.
Conclusion: The MagArray technology demonstrated that it can provide exceptional
sensitivity with reasonable reproducibility for simultaneous detection of both E. coli
O145 and O157 in ground beef. Thus we believe this technology provides a good fit
for detecting multiple E. coli serogroups. This assay not only accelerates identification
of pathogenic E. coli, but also holds the potential to help regulatory agencies to
quickly issue a product recall for contaminated foods.

B-055
Real-time PCR TaqMan assay for Rapid Screening of Sepsis using Positive
Blood Cultures

J. Kim1, H. Kim1, H. Wang2, S. Kim3, Y. Kim1, S. Park4, E. Choi4, G. Kim1,

S. Ahn1, S. Park1, S. Park5, H. Jin6, Y. Kim7, Y. Uh4, H. Lee1. 1Department
of Biomedical Laboratory Science, College of Health Sciences,
Yonsei University, Wonju, Korea, Republic of, 2M&D, Inc., Wonju Eco
Environmental Technology Center, Wonju, Korea, Republic of, 3Institute for
Life Science and Biotechnology, Yonsei University, Seoul, Korea, Republic
of, 4Department of Laboratory Medicine, Yonsei University Wonju College
of Medicine, Wonju, Korea, Republic of, 5Department of Clinical Laboratory
Science, College of Health and Therapy, Daegu Hanny University, Daegu,
Korea, Republic of, 6Department of Clinical Laboratory Science, College
of Health Sciences, Catholic University of Pusan, Busan, Korea, Republic
of, 7Department of Internal Medicine, Yonsei University Wonju College of
Medicine, Wonju, Korea, Republic of
Background : Sepsis is a lethal medical condition that results from a harmful or
injurious host response to infection. Rapid detection of pathogens in blood from septic
patients is essential for adequate antimicrobial therapy and prognosis of patients.
The aim of this study was the acceleration of detection and discrimination of Gram
positive (GP)-, Gram negative (GN)-bacteria and Candida species in blood culture
specimens by molecular methods.
Methods : The Real-Sepsis real-time PCR kit (M&D, Wonju, Republic of Korea)
uses the following TaqMan probes: the bacterial 16S rRNA gene for pan-GP, panGN and fungal 18S rRNA gene Candida species, respectively. For blood culture, two
or three pairs of culture bottles for aerobes or anaerobes were incubated in the BacT/

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S147

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

Alert 3D (bioMrieux, Marcy, France), BACTEC 9240 system (Becton Dickinson
Diagnostic System, Spark, MD, USA), or the BACTEC FX (Becton Dickinson)
blood culture systems for 5 days after inoculating with blood drawn from the patient
at the bedside. The identification of bacteria and antimicrobial susceptibility tests
(ASTs) were conducted by the microplate method, the MicroScan system (Siemens
Healthcare Diagnostics, Sacramento, CA, USA), and the Vitek 2 system (bioMrieux,
Durham, NC, USA). For identification of Candida species, a VITEK-2 (bioMrieux)
YST ID CARD was used.
Results : The Real-time PCR TaqMan assay was evaluated using a total of 62 bacterial
reference strains representing 39 of GP, 23 of GN species and 25 fungal reference
strains. Subsequently, it was evaluated with 115 clinical isolates, 256 positive blood
culture specimens and 200 negative blood culture specimens, and results were
compared to those of conventional identification method. The overall sensitivity of
the real-time PCR TaqMan assay was 99.6% and the specificity was 89.5%.
Conclusion : The Real-Sepsis real-time PCR assay could not only differentiate
bacterial and fungal from viral and other pathogens, but also can classify Gram
staining with a much shorter turnaround time than the gold standard culture method.
Furthermore, it could have an important impact on choosing the appropriate antibiotic
therapy based on simultaneous detection and discrimination GP-, GN-bacteria and
Candida species.

B-057
In vitro antimicrobial susceptibility of clinical and environmental strains of
Burkholderia pseudomallei from Brazil

T. P. G. Bandeira1, M. G. Castelo2, E. T. Oliveira3, F. M. P. Ventura3, V.

P. G. Bandeira4, R. S. N. Brilhante5, M. F. G. Rocha6, J. A. Sales6, G. M.
M. Guedes5, O. V. P. Denadin3, C. F. A. Pereira3, D. C. C. Branco5, J. J. C.
Sidrim5. 1DASA; Postgraduate Program in Medical Microbiology, Federal
University of Ceara, Fortaleza, Ceara, Brazil, FORTALEZA, Brazil, 2UFC,
FORTALEZA, Brazil, 3DASA, FORTALEZA, Brazil, 4Christus College,
School of Medicine, FORTALEZA, Brazil, 5Postgraduate Program in
Medical Microbiology, Federal University of Ceara, Fortaleza, Ceara,
Brazil,, FORTALEZA, Brazil, 6Postgraduate Program in Veterinary
Science, State University of Cear, FORTALEZA, Brazil
Background: Burkholderia pseudomallei, the causative agent of melioidosis, is
intrinsically resistant to a wide range of antimicrobial agents [1]. Ceftazidime is the
drug of choice for treating melioidosis, although carbapenems are indicated for severe
infections. Following this initial treatment, an eradication phase is recommended,
consisting of prolonged oral therapy with trimethoprim-sulfamethoxazole (SXT)
combined with doxycycline or amoxicillin-clavulanic acid (AMC) for up to 6 months
[2]. The aim of this study was to determine the antimicrobial susceptibility of clinical
and environmental B. pseudomallei from Brazil.
Methods: Ten clinical strains of B. pseudomallei were included in this study, obtained
from the DASA central laboratory at Fortaleza, Cear, and the others environmental
strains were obtained from bacterial collection of Federal University of Cear.
Identification of B. pseudomallei was confirmed using an automated VITEK
2 system, bioMrieux, followed by sequencing of the 16S-23S spacer region. For
antimicrobial susceptibility assay, five antimicrobial agents were tested by the
microdilution technique according to CLSI guidelines [3].
Results: All MICs determined by the broth microdilution from 20 strains of B.
pseudomallei in this study were distributed in MIC50 and MIC90 and as the percentage
of sensitivity. The percentage of sensitivity for doxycycline, imipenem and sulfametol
/ trimethoprim were 100% each and amoxicillin / clavulanate and ceftazidime was
80% and 90% respectively (Table 1).
Conclusion: The current results were compatible with those previously reported in
the literature [4,5] and corroborate those of Jenney et al. [4]. The susceptibility of the
tested strains appears to be independent of the origin of the isolates (environment or
clinical cases).
In conclusion, this work provides knowledge on the antimicrobial susceptibility of B.
pseudomallei from Brazil, serving as a guide for the selection of appropriate empirical
therapy, thus contributing to better medical care in addressing melioidosis.
Antimicrobials
Amoxicilin/
clavulanate
Ceftazidime
Doxycycline
Imipenem
Trimethoprimsulfamethoxazole

S148

MIC (g/ MIC

90
mL) 50
(g/mL)
8/4
18/8

4/2 - 32/16

Susceptibility
(%)
80

4
0
1

16
1
1

2 - 16
0,25 - 0,5
0,125 - 1

90
100
100

0,125/2,375 2/38 100

Range

B-058
The performance of a highly sensitive chemiluminescent enzyme immunoassay
for HBsAg.

S. Yamauchi, K. Yamaguchi, S. Kojima, H. Sugimoto, Y. Imai, T. Saito, T.

Higuchi, K. Moriyama. Fujirebio Inc., TOKYO, Japan
[Background] HBsAg is an envelope protein of HBV and is continuously secreted
into blood as not only a portion of HBV but also a secretion protein during HBV
persistent infection. HBsAg level in blood is correlated with intra-hepatic covalently
closed circular DNA (ccc DNA) and is useful as an indicator of HBV persistent
infection, especially during treatment with anti-viral drugs.
We have developed a fully automated highly sensitive chemiluminescent enzyme
immunoassay for HBsAg (new CLEIA system) which has 10 fold higher sensitivity
than a commercially available HBsAg kit. We evaluated the basic performance for the
new CLEIA system and here we report the results.
[Methods] The highly sensitive chemiluminescent enzyme immunoassay was run on
the fully-automated CLEIA system LUMIPLUSE G1200 (FUJIREBIO INC.).
[Results] The CV for within-run reproducibility was 0.5-3.3% and for between-run
reproducibility it was 0.5-1.4%. The quantitation limit was 5 mIU/mL (0.005 IU/mL).
The new CLEIA system could detect 1-3 bleeds earlier than the commercial HBsAg
kit in seven seroconversion panels among nine. The correlation coefficient and the
slope with the commercial HBsAg kit were 0.92 and 1.30, respectively.
[Conclusion] The new CLEIA system has good reproducibility and high sensitivity.
And it shows good correlation with a commercial HBsAg kit. In addition, the new
CLEIA system can be easily operated to complete an assay in 30 min.
The new CLEIA system is considered quite useful for the routine quantification of
HBsAg.

B-059
Utilization of Serum Total Bile Acids for the Prediction of HCV Active Infection
in Anti-HCV Antibody Positive Patients

B. Chung, G. Guzman, M. Jin. University of Illinois Hospital and Health

Sciences System, Chicago, IL
Background: Conventional liver function tests do not correlate with active hepatitis
C infection or response to treatment. Recent studies have suggested that bile acids
(BA) in the blood may be elevated in patients with detectable HCV RNA levels. It
has also been shown that bile acids increase HCV RNA replication and has been
suggested as a possible etiology for poor response to IFN therapy in patients with
specific genotypes.
Methods: Total BA levels from blood samples were measured on the Beckman DxC
using the Diazyme Total Bile Acids Assay. Conventional liver function tests using
Beckman reagents were performed on the Beckman DxC. BA levels of 30 anti-HCV
antibody positive patients with detectable HCV RNA and 30 without detectable HCV
RNA were compared to 24 healthy controls. Reference range includes results < 10
umol/L. Mean values, 95% confidence intervals, and p-values from independent
sample Students unpaired t-test were calculated using MS Excel.
Results: Mean total BA values for controls and HCV RNA negative patients fell
within the reference range while mean values for HCV RNA positive patients were
elevated. Mean total BA levels were statistically significantly higher in patients with
detectable HCV RNA levels versus controls and patients with undetectable HCV
RNA levels (p-values: control, 0.09; HCV RNA-, 0.05). No statistically significant
difference was observed for liver function test values between the 3 compared groups.
Conclusions: Currently, quantitative HCV viral load testing is employed to monitor
treatment response but remains costly and is not practical for surveillance in chronic
hepatitis C patients. Our preliminary findings suggest that total BA levels may
distinguish between active and chronic hepatitis C infection. For physicians needing a
non-invasive and less cost-prohibitive method for monitoring of recurrence or active
infection in their hepatitis C patients, bile acid testing may prove a viable tool in their
arsenal

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

B-062
Comparison of clinical performances among Roche Cobas HPV, RFMP HPV
Papillo Typer and Hybrid Capture 2 assays for detection of high-risk types of
human papillomavirus

S. Yu1, K. Ko1, M. Kwon1, E. Lee2, H. Woo1, H. Park1. 1Kagnbuk Samsung

Hospital, Seoul, Korea, Republic of, 2Green Cross Reference Laboratory,
Yongin, Korea, Republic of
Background: High-risk types of human papillomavirus (HR-HPV) is an important
cause of cervical cancers. Current cervical cancer screening guidelines suggest that
early detection of HPV-16 and HPV-18 may prevent the progression of cervical
cancer. We evaluated and compared three HPV DNA tests, Roche Cobas HPV (Roche
Molecular Systems Inc., Pleasanton, CA), RFMP HPV Papillo Typer (GeneMatrix
Inc., Yongin, Korea) and Hybrid Capture 2 (HC2; Qiagen, Gaithersburg, MD, USA).
The HC2 has been recommended for use as a reference test, Roche Cobas HPV
specifically identifies HPV-16 and HPV-18 with concurrently detecting other 12 HRHPV types and RFMP identifies 74 HPV genotypes.

B-061
Prevalence of Tuberculosis in Sao Paulo diagnosed by Laboratory tests in the
period 2011-2013

M. C. Feres, R. Bini, N. A. Raphael, S. G. Oliveira, P. G. Campana, D.

Mazzotti, A. A. Lino de Souza, S. Tufik. Associacao Fundo de Incentivo a
Pesquisa, Sao Paulo, Brazil
Background: According to data presented by the World Health Organization
(WHO) in 2010 were diagnosed and reported 6.2 million cases of tuberculosis (TB)
worldwide, with 5.4 million new cases, representing 65% of the estimated cases for
the same year. Countries like China and India account for 40% of cases, and Brazil is
among the 22 countries which account for 82% of TB cases worldwide.

Methods: A total of 861 cervical swab specimens from women over 30 years of age
were classified into groups of high grade squamous intraepithelial lesion (HSIL) and
non-HSIL according to cervical cytology results and analyzed by Roche Cobas HPV,
RFMP HPV Papillo Typer and HC2. The results of direct sequencing or Linear array
(LA; Roche Molecular Systems Inc., Pleasanton, CA) HPV genotyping test were
considered true when three assays presented discrepancies.
Results: Concordance rates between Roche Cobas HPV vs. RFMP, RFMP vs. HC2,
and HC2 vs. Roche Cobas HPV were 94.5% (814/861), 94.2% (811/861), and 95.8%
(825/861), respectively. In 71 specimens with discrepant results, concordance rates
between each assay and direct sequencing or LA were as follows: Roche Cobas
HPV, 35.2%; RFMP, 93.0%; HC2, 25.4%. Clinical sensitivities and specificities for
detecting HSIL were 80.3% and 95.8% with Roche Cobas HPV, 83.6% and 95.1%
with RFMP and 90.2% and 94.8% with HC2.

Combating TB 2011 - 2015 follows the overall plan proposed by the WHO. Your goal
is to dramatically reduce the burden of disease by 2015. The main objective to reduce
TB are:1) reduce the incidence of TB in HIV / AIDS and the incidence of HIV in TB
patients, prevent and control - multidrug-resistant TB and strengthen actions to meet
the needs of poor and vulnerable populations, 2) strengthen the health system based
on primary care, 3) engage all providers of health services, and 4) enable and promote
research and others.

Conclusion: Roche Cobas HPV, RFMP and HC2 showed high agreement rates each
other. Although Roche Cobas HPV and RFMP showed lower clinical sensitivity in
detecting HSIL compared to HC2, they would be clinically useful since both provide
HPV genotypes.

The Plan also has, as main targets: to reduce the incidence and mortality of TB until
2015 compared to 1990 and eliminate TB as a public health problem until 2050. With
this goal it becomes increasingly important to accurate and early diagnosis of this
disease and laboratory testing and higher efficiency are of great importance in this
context.

Prevalence of fungal bloodstream infections in a tertiary University Hospital in

Brazil - Comparative analysis between two periods in the last decade

Objective: To evaluate the percentage of positive diagnosis of TB for each specific

laboratory test for this disease for the period 2011 to 2013, retrospectively analyzing
the database of a Laboratory Oversize working in So Paulo, Brazil

Background: In the last decades, the growing population of immunosuppressed hosts

has dramatically increased. Therefore, the prevalence of nosocomial fungemia has
increased throughout the world and mortality from this disease is high. The objective
of the study is to identify the etiology of fungal bloodstream infections in a Tertiary
University Hospital in Belo Horizonte, Brazil, comparing two periods in the last
decade.

Material and Methods: The authors retrospectively analyzed 44931 results of

laboratory tests ordered for diagnosis and monitoring of TB originated from 42
ambulatory care of Associao Fundo de Incentivo a Pesquisa- Afip, from Sao
Paulo- Brazil, (January2011-December2013). Laboratory tests analyzed were:
Adenosine deaminase (ADA), Bacillus Koch (BK), BK-automated culture/Bactec ,
Mycobacterium tuberculosis PCR.
Results: The test results of ADA, BK. And BK- automated culture/Bactec from of
years (2011, 2012 and 2013) with their respective percentages of positivity were:
ADA for 2183 (23.5%), 2105 (21.9%), 2233 (48.8%), search for BK , 7219 (9.5%),
7563 (10.1%), 7400 (10.4%); BK-automated culture/Bactec 1926 (6.0%) 1653 (5.5%)
1535 (8.0%), and Mycobacterium tuberculosis PCR, 48 (4.2%) 79 (6.3% ) 95 (7.4%).
The data were presented as percentage of positive prevalence of the most requested
in the affiliated units of the Afip laboratory. However, these methods have limitations
and many of them are carried out in more than one sample.
Conclusion: We note that the ordering patterns of these tests remained constant
over the years. The PCR method showed a small increase in use and an increase
of positivity but the general results showed that the most requested examination
for diagnosis of TB is a direct search in lamina and culture of BK. Analyzing the
percentages of positivity of the Afip tests performed, we conclude that from 2011
to 2013 there was an increase in cases of TB of nearly 0.5 to 1.5%, which is very
worrying that we may achieve the governments goal to reduce TB cases until 2015.

B-063

L. S. Vasconcellos, L. L. Castro, D. H. Bcker, P. H. O. Mouro, J. R.

Romeiro. Universidade Federal de Minas Gerais, Belo Horizonte, Brazil

Methods: we retrospectively analyzed the results of all blood cultures processed in

the hospital, between two periods: from 2001 to 2003 and from 2011 to 2013. For
each triennium were reported the number of blood cultures collected, the number
of positive cases, the percentage of fungemia and all identified fungal species.
The samples were observed in the laboratory routine carried out by incubation in
BacT ALERT (bioMrieux). The positive samples were subcultivated for species
identification through morphologic and biochemical assays.
Results: From 2001 to 2003, 34.822 blood culture were performed and 5,510 (15.8%)
positive. Fungi were isolated in 229 (16.4%) cases. From 2011 to 2013, the number
of blood cultures increased to 55,052, but the number of positive samples decreased
to 4,873 (8.9%). Fungal bloodstream infections increased to 290 (6.00%) cases.
Candidemias were predominant: 97.38% (2001-2003) and 91.72% (2011-2013). The
isolated species are shown in Table 1.
Conclusions: The prevalence of fungemia increased in the last decade. Candidemia
was responsible for more than 90% of the cases. Non-albicans Candida species
increased and C. albicans decreased. Others species of fungi increased too.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S149

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

Prevalence of fungal species on bloodstream infections in University Hospital in the
last decade
Species
2001-2003
2011-2013
Total
N
%
N
%
N
%
Candida albicans
91
39.74 84
28.97 175
33.72
Candida glabrata
1
0.44
5
1.72
6
1.16
Candida
3
1.31
3
1.03
6
1.16
guilliermondii
Candida kefyr
0
0
2
0.69
2
0.39
Candida krusei
0
0
9
3.10
9
1.73
Candida
60
26.2
76
26.21 136
26.20
parapsilosis
Candida spp
29
12.66 24
8.28
53
10.21
Candida tropicalis 39
17.03 63
21.72 102
19.65
Cryptococcus
3
1.31
7
2.41
10
1.93
neoformans
Cryptococcus spp 3
1.31
3
1.03
6
1.16
Fusarium sp
0
0
8
2.76
8
1.54
Trichosporon spp 0
0
6
2.07
6
1.16
Total
229
100
290
100
519
100

B-064
Distribution of HIV genotypes among Brazilian regions

P. Y. Nishimura, D. Nifoci, B. C. Vilanova, V. D. T. Niewiadonski, O. P.

Denardin, N. Gaburo Jr.. DASA, Sao Paulo, Brazil
Background: Lung cancer is the most prevalent life-threatening cancer worldwide
with more than 80% being non-small cell lung cancer (NSCLC). Detection of
mutations of EGFR gene is critical for predicting the response to therapy with tyrosine
kinase inhibitors (TKIs), such as gefitinib and erlotinib, in patients with NSCLC.
Patients that are EGFR mutants have constitutive TK activity and, therefore, a greater
sensitivity to anti-EGFR inhibition.
Objective: To describe the EGFR mutations frequency found in lung adenocarcinoma
samples, using pyrosequencing method.
Method: Thirty samples of lung adenocarcinoma were analyzed from January 2013 to
December 2013. The test was performed on formalin-fixed, paraffin-embedded tumor
specimen, after the selection of the specimen region to be analyzed by a pathologist.
The DNA was extracted using the Qiaamp FFPE Tissue kit (Qiagen, Hiden,
Germany). Concentration of DNA sample was measured spectrophotometrically
using a NanoDrop spectrophotometer (NanoDropTechnologies, Wilmington). Codons
719, 768, 790, 858, 861 and exon 19 were amplified by PCR using the EGFR Pyro
kit (Qiagen, Hiden, Germany). Successful and specific amplification of the region
of interest was verified by visualizing the PCR product on capillary electrophoresis
using Qiaxel DNA Screening Kit (Qiagen, Hiden, Germany). Preparation of singlestranded DNA was done using PyroMark Q24 vacuum workstation (Qiagen, Hiden,
Germany) according to the manufacturer instructions. The pyrosequencing reaction
was analyzed on the Pyro Mark Q24 (Qiagen, Hiden, Germany)
Results: The frequency of EGFR mutations found is presented on Table 1. All
mutations together represent only 27% of the samples.
Conclusion: The results are consistent with previous studies and reports. The singlepoint mutation L858R (CTG> CGG) on exon 21 and the frame deletions on exon 19
represents the majority mutations found in Brazilian lung adenocarcinoma samples,
although most samples showed no mutation at the target regions.
Table 1. Frequency of EGFR mutations found in lung adenocarcinoma samples.
Results
Wild type
2235del15 (exon 19)
2236del15 (exon 19)
2237_2255>T (exon 19)
2239_2248>C (exon 19)
CTG>CAG (L861Q)
CTG>CGG (L858R)

Frequency
73%
3,3%
3,3%
3,3%
3,3%
3,3%
10%

B-065
Soluble CD14-subtype, a possible new biomarker increases in septic patients
plasma from pediatric department.

H. Yamaguchi, S. Fukuoka, H. Oto, T. Soga, M. Inoue, S. Kitazawa, Y.

Umeda, S. C. Kimura. Showa University Northern Yokohama Hospital,
Yokohama City, Japan

bacterial sepsis (Shozushima T, et al. J Infect Chemother 2011;17:764-9), however,

there have been limited reports on pediatric patients. In order to clarify the significance
of P-SEP as a marker of septic disease in children, we conducted a study of serum
P-SEP concentration in pediatric patients with febrile diseases.
Methods: Forty-eight children (29 males, 19 females, 0.6 to 152 months after
birth, mean age 2.43 years old) admitted to our hospital were enrolled. Plasma
was obtained within 24 hours after blood withdrawal. P-SEP was assayed using
PATHFASTTM chemiluminescent enzyme linked immunoassay system (Mitsubishi
Chemical Medience Corporation, Tokyo, Japan). This automatic analyzer enables to
get results within 20minutes. Procalcitonin, white blood cells and C-reactive protein
concentration were assayed simultaneously. The ethic committee of Showa University
Northern Yokohama Hospital approved this study.
Results: P-SEP concentration was 442 plus minus 301 ng/L (mean and SD) in patients
whose blood culture was positive on admission (n=4). For example, staphylococci
were detected with blood culture from a 30 months-old female patient. Her P-SEP
concentration was 866 ng/L on admission, then decreased after antimicrobial
treatment to 235 ng/L when she was discharged. P-SEP concentrations were 191 plus
minus 47 in viral infections (n=9), 313 plus minus 90 ng/L in Kawasakis disease
(n=6). On the other hand, cases with blood culture negative but urine and/or sputum
culture positive showed 349 plus minus 202 ng/L (n=9). Other culture negative
patients (n=20) showed 267 plus minus 132 ng/L.
Discussion: P-SEP has been reported to be an indicator of prognosis in adult septic
patients (Masson S, et al. Crit Care 2014;18:R6), and critically ill preterm newborns
(Mussap M et al. J Matern Fetal Neonatal Med 2012;25:51-3). Though statistically
not significant, plasma P-SEP was higher in septic children compared to those without
bacterial infections. Reference interval of plasma P-SEP concentration in adults under
the age of 70 is ranged 201 to 457 ng/L (Chenevier-Gobeaux C, et al. Clin Chim Acta
2014;427:34-6). Our study suggests reference interval in children is likely to be lower
than that in adults. More study is required to confirm the results.
Conclusion: Increased plasma concentration of P-SEP was observed in pediatric
patients with bacterial sepsis. P-SEP could be a possible biomarker of sepsis in
pediatric patients.

B-066
Analysis of blaKPC gene from Hodge Test screening confirming KPC enzyme
resistance

S. S. Michelotto. LANAC, Curitiba, Brazil

Background:The Klebsiella Pneuminiaie Carbapenemase (KPC) is responsible for
human infections, especially in hospital environment. It is an enzyme produced by
Gram-Negative bacilli and its detection in bacterial isolates confers resistance to the
carbapenem antibiotics and furthermore inactivates penicillins, cephalosporin and
monobactams.The transmission, in hospital environment, occurs through the contact
between secretion from infected patients. The objective of this study was to evaluate
the correlation of positive results on the Hodge Test with detection of the blaKPC gene
confirmed by Molecular Biology (Life Technologies - Real Time PCR System 7300).
Methods: 105 cultures from samples of LANAC Laboratory of different materials
were selected and the resistance profile was observed to multiple antibiotics
(especially carbapenems) using the method of disk diffusion and automated system
MicroScan WalkAway - Siemens Healthcare Diagnostic during 2011 and 2012.
Results:Multidrug resistant strains are not new or specific for Klebsiella species. In
100 of 105 analyzed cultures K. pneumoniae was isolated (95.2%). 98% of these
isolates showed concordance between Hodge Test and the results obtained by Life
Technologies - Real Time PCR System 7300 (blaKPC gene detectable). 2% of the
isolates were indeterminate by Hodge Test. The others microorganisms isolates on
the analyzed cultures were E. cloaceae (2,8%); Enterobacter sp (0.95%) and Proteus
vulgaris (0,95%). In only two samples the correlation between the tests were not
confirmed.
Conclusion: In this study was observed 98% of correlation between Hodge Test and
the detection of blaKPC gene by Molecular Biology ( Life Technologies - Real Time
PCR System 7300). The determination of the Minimum Inhibitory Concentration
(MIC) using the equipment MicroScan WalkAway - Siemens obtained an excellent
performance in the correlation between positive Hodge Test and detectable blaKPC
gene by Molecular Biology (Life Technologies - Real Time PCR System 7300).
Hodge Test and microbiology system automation seems to be an excellent alternative
for clinical laboratory routine with high sensitivity and lower cost.

Background: Soluble CD14-subtype, named presepsin (P-SEP) is a fragment of

CD14 peptide produced through phagocytosis of microorganisms by neutrophils.
Increased serum concentration of P-SEP was reported in adult patients with severe

S150

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

B-067
Ziehl-Neelsen staining as an aid in screening for diagnostic of systemic fungal
infection.

C. F. A. Pereira1, D. P. Silva1, L. G. S. Carvalho1, A. M. Cervo2. 1DASA Group, Cascavel, Brazil, 2UNIPAR, Cascavel, Brazil
Background: Mycology is still an area that has little clinical importance, although
the number of susceptible patients to fungal infection arises through the years. Every
single day, in laboratory cytology routines, sputum samples are collected for Acid-fast
stain (Ziehl-Neelsen method) for the differential staining procedure to members of
the genera mycobacteria (M. tuberculosis, M. leprae), bacteria (Nocardia) and fungus
(Cryptosporidium,). Its known that many systemic fungal infections are similar to
other common pulmonary diseases, and the differential diagnostic is difficult. We
have tried, through a visual screening of compatible fungal structure, to identify
medically significant fungi, to an additional specific Mycosel agar culturing step.
Although Ziehl-Neelsen is not intended to staining of fungal genera, we thought if it
could also be used for the primary identification of fungal pathogens.
Methods: The test was performed with routine sputum samples from Laboratrio
Alvaro (DASA group), collected by spontaneous or induced expectoration and kept
under refrigeration of 2 - 8C. These samples were primary intended to Ziehl-Neelsen
staining procedure for identification of M. Tuberculosis. After staining and visual
observation of fungus structure, the cytologist is capable of reporting if a fungal
infection is or is not present in the sample. After visual inspection, 150 potential
positive samples were selected. Mycosel Merck culture medium was prepared by
dilution as described in technical data sheet. This medium is specific for isolation of
pathogenic fungi. After inoculation, samples remained in an incubator set to 35 C
for approximately 30 days. After the incubation time, each grown fungal structure was
identified by slide morphological observation, in which Cotton blue staining (specific
for examination of fungal colonies) was applied.
Results: From 150 samples, we had no growth in 14% (21/150), 86% (129/150)
were positive, where 55% (37/150) corresponding to C. albicans yeast, 9% (6/150)
to C tropicalis, 5% (3/150) C. glabrata, 5% (3/150) of yeast and hyphae C. albicans,
1% (1/150) C. parapsilosis, 1%(1/150) Nigrospora and 1% (1/150) of Candida
Krusei. Although there is significant positivity for Candida genera, it cant be easily
implicated in systemic fungal infection, as opposed to 9 % (6/150) of fungal, normally
associated to pulmonary disease. Our final finding was 4% of Histoplasma sp, 3%
Aspergillus sp, 2% of Paracoccidioides sp, fungus that are morphologically classified
as positive for severe pulmonary disease.
Conclusion: This study provides evidence of the presence of etiologic agents of
severe pulmonary fungal disease in sputum of patients originally submitted to AcidFast staining. The simple screening to fungal structure in Ziehl-Neelsen stained slides,
have shown to be applicable, simple and effective to directing potential positive
samples to further culturing in selective medium for isolation of pathogenic fungi.
This new procedure can be meaningful in evaluating TB like suspect patients not just
on the basis of symptoms, clinical signs, but providing another reliable screening tool.

B-068
Development of a Point-of-Care Diagnostic for Ebola and Sudan Virus
Detection

D. Oottamasathien1, A. Jones1, M. Boisen1, P. Kulakosky2, R. Wilson2,

L. Branco2, E. Ollman Saphire3, R. Garry4, K. R. Pitts1. 1Corgenix, Inc.,
Broomfield, CO, 2Autoimmune Technologies, New Orleans, LA, 3The
Scripps Research Institute, La Jolla, CA, 4Tulane University, New Orleans,
LA
Background: Viral hemorrhagic fevers are serious, often fatal illnesses characterized
by high fever, damage to the vascular system, and multi-organ failure. Because of
their rapid progression, the ability to detect and distinguish hemorrhagic fevers is
paramount to treatment and survival. To this end, we report on the characterization
of rapid point-of-care diagnostic tests for Ebola (EBOV) and Sudan (SUDV) virus
detection.
Methods: Using recombinantly produced proteins, we generated a library of
polyclonal and monoclonal antibodies recognizing EBOV and SUDV GP, NP, and
VP40. Antibodies were initially tested using a multivariate approach with each
antibody being tested for use as both capture and detection capability. Testing was
performed over a range of plate coating concentrations, nitrocellulose dot blots
and stripings, HRP and gold conjugation conditions, and sample dilution ratios.
The pairings were further optimized by testing the EBOV and SUDV plate coating
concentrations and HRP-conjugate dilutions against various sample dilution titrations

to determine the conditions that favored sensitivity and signal. Testing was performed
using chosen antibody pairings for the EBOV and SUDV ELISA to confirm that the
pairings are optimal by running dose-response curves of both purified EBOV VP40 or
SUDV VP40 antigen spiked into a normal human serum control matrix, determining
the signal to noise ratio, linear range, LOD, LOQ, and LOB. Candidate antibody
pairings identified during the antibody screening process for use on the lateral flow
immunoassay (LFI) format were conjugated to gold nanoparticles and striped onto
nitrocellulose using the Biodot XYZ dispenser. Testing was performed using purified
antigen spiked into normal human serum control matrix.
Results: As demonstrated by ELISA, we found polyclonal antibody pairings against
EBOV and SUDV VP40 to be the most reliable in detecting purified recombinant
protein in spiked samples. Pairings exhibited limits of detection as low as 10ng/mL,
and limits of quantitation ranging from 10-100ng/mL, suggesting that these critical
reagents possess the ability to detect low amounts of EBOV and SUDV protein.
Pairings migrated to the LFI test strips showed the ability to detect both EBOV and
SUDV VP40 recombinant protein in a dose-dependent manner down to 100ng/mL
within 10 minutes. Importantly, this dose-dependency was distinguishable with the
naked eye, indicating the utility of this rapid test in environments lacking conditioned
power and/or significant medical training.
Conclusions: We have developed and characterized prototype ELISA and LFI tests
capable of detecting EBOV and SUDV proteins in sample matrix. In the ELISA
format, multiple pairings were able to detect EBOV and SUDV VP40 antigens in
spiked matrix with acceptable sensitivity, suggesting that with further optimization
and ELISA test for the detection of EBOV and SUDV is within reach. In the LFI
platform, two pairings showed the ability to detect EBOV and SUDV antigens in
a concentration-dependent manner. Importantly, this concentration dependency was
discernible without the aid of any instrumentation, suggesting a path forward for the
optimization of a rapid, point-of-care test that can be used in austere environments.

B-069
Heparin Binding Protein for Discrimination of Infected and Non-infected
Critical Ill Patients from Cardiovascular Conditions - Results of a Pilot
Evaluation

E. Spanuth1, M. Preusch2, A. Vasishta3, E. Giannitsis4. 1DIAneering Diagnostics Engineering & Research, Heidelberg, Germany, 2Department
of Internal Medicine III, Intensive Care Unit, University Hospital
Heidelberg, Heidelberg, Germany, 3Axis-Shield Diagnostic Ltd., Dundee,
United Kingdom, 4Department of Internal Medicine III, University Hospital
Heidelberg, Heidelberg, Germany
Background Heparin binding protein (HBP) is an inflammatory mediator released
into the circulation during neutrophil activation. HBP has been shown to contribute
diagnostic information to the differentiation of viral and bacterial infections. Bacterial
infection is the most important trigger for the development of sepsis. Especially
in critical ill patients the early detection of infection is necessary for appropriate
treatment. We thought to investigate whether HBP is able to detect bacterial infection
in critical patients from cardiovascular conditions admitted at the intensive care unit
(ICU).
Methods 20 patients admitted at the ICU with severe cardiovascular conditions were
included. 12 patients developed additional infectious diseases of whom 4 patients
developed sepsis. Serum HBP concentrations were measured using the Heparin
Binding Protein EIA (Axis-Shield Diagnostics Ltd. Dundee). C-reactive protein
(CRP) was determined using the cobas assay (Roche Diagnostics).
Results The discrimination of HBP and CRP concentrations between patients with
(n=12) and without infection (n=8) was examined by Mann-Whitney independent
sample test. The results are displayed in the table.
Tab. 1: HBP and CRP values in ICU patients with cardiovascular conditions with
and without additional infectious diseases
Without infection, n=8 With infection, n=12
p value
Median (IQR)
Median (IQ)
HBP, g/L 60 (39-95)
145 (121-238)
0.0087
CRP, mg/L 72 (34-124)
159 (99-206
0.0136
The determination of HBP in serum provided a higher significance level
for differentiation between patients with and without infectious diseases
compared to CRP. These results could be confirmed by ROC analysis yielding
area under the curve (AUC) values of 0.854 and 0.833 for HBP and CRP,
respectively. Logistic regression analysis with HBP and CRP as independent
variables revealed an AUC value of 0.906 demonstrating that the simultaneous
determination of HBP and CRP provided additional diagnostic information.
Conclusion HBP allows highly significant discrimination between infected and non-

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S151

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

infected critical ill patients with cardiovascular complications admitted at the ICU
which was superior compared to CRP. Additionally, simultaneous determination of
HBP and CRP showed higher diagnostic efficacy than both markers alone.

B-070
Diagnostic Evaluation of Focus Diagnostics Simplexa Dengue real-time
polymerase chain reaction (RT-PCR) detection and typing of dengue virus

I. C. Ibraim, G. J. Cruz, E. C. C. Mateo, A. C. S. Ferreira. Instituto Hermes

Pardini, Vespasiano, Brazil
Dengue is the most important arthropod-borne viral infection of humans and the
incidence of dengue has grown dramatically. Dengue virus (DENV) infection affects
over 40% of the worlds population. Worldwide, an estimated 2.5 billion peopleare at
risk of infection.
Dengue viruses belong to the genus flavivirus within the Flaviviridae family. The
virus group consists of four serotypes (DENV-1, DENV-2, DENV-3 and DENV-4)
that manifest with similar symptoms. DENVs produce several syndromes that are
conditioned by age and immunological status.Laboratory confirmation of dengue
infection is crucial as the broad spectrum of clinical presentations can make accurate
diagnosis difficult.
Dengue can be diagnosed by isolation of the virus, by serological tests, or by
molecular methods.Seroconversion of IgM or IgG antibodies is the standard for
serologically confirming a dengue infection. Viral antigens also provides evidence
of infection and virus isolation provides the most specific test result.The RT-PCR and
other PCR-based techniques have become a primary tool to detect virus in the early
course of illness. In addition, molecular testing allows the monitoring of outbreaks
by detecting the emergence of new serotypes, thus permitting the implementation of
control measures.
Therefore, the aim of the project was to evaluatethe diagnostic accuracy of the
commercially available Focus Diagnostics Simplexa Dengue real-time polymerase
chain reaction (RT-PCR) assay for the in vitro detection and typing of dengue virus
serotypes 1, 2, 3 and 4 and compare with results obtained from serology.
The RNA of 37 IgM and/or IgG positive samples were extracted using the QIAamp
RNA Viral Kit (Qiagen, Germany) according to the manufacturers recommendations.
The amplification of the extracted RNA used bi-functional fluorescent probe-primers
and reverse primers. The assay amplifies four serotype specific regions: dengue 1
(NS5 gene), dengue 2 (NS3 gene), dengue 3 (NS5 gene) and dengue 4 (capsid gene).
An RNA internal control was used to monitor the extraction process and to detect
RT-PCR inhibition. A positive control for all four serotypes was added during the
extraction and RT-PCR reaction.
We tested 37 serologically positive samples. In order to compare serologywith RTPCR, any samples positive for IgM and/or IgG were considered a positive diagnosis
of dengue. Of the 37 samples tested from patients with positive serology, 8 (21.62%)
were found positive by RT-PCR, with Ct values ranged between 30.3 to 39.6. All
positive RT-PCR samples were IgM positive and were negative for IgG. The assay did
not detected viral RNA in positive IgG sample (32.43%).
Serological assays are commonly used for diagnosis of dengue infection, as they are
relatively inexpensive and easy to perform. However, the detection of antibodies in a
dengue-infected person is only possible after 4-5 days of disease onset. One advantage
of the RT-PCR assay is the ability to detect and serotype viral RNA early in dengue
illness, which is important to diagnosing acute infection and provides the opportunity
to impact patient management.

B-071
Analytical Reactivity and Preliminary Performance Results of the BD MAX
QS Vaginal Panel*

N. Paquette1, V. Brochu1, J. Cormier1, S. Tremblay1, C. Lehouillier1, S.

Morasse1, F. Harel1, M. Tremblay1, D. Cantin1, L. Belley-Montfort1, M.
Farell2, S. Sajo Beqaj3, C. Roger-Dalbert1. 1BD Diagnostics, Qubec, QC,
Canada, 2Planned Parenthood, Houston, TX, 3Pathology Inc, Torrance, CA
*The BD MAX QS Vaginal Panel is not available for sale or use in the U.S

aimed to (1) challenge the assay with a wide range of strains in an analytical reactivity
(inclusivity) study, (2) evaluate the capacity of the assay to detect targets during a
co-infection and potentially support the rationale for patient treatment decisions and
(3) collect preliminary results from clinical specimens tested with the BD MAX QS
Vaginal Panel on the BD MAX System.
Methods: An analytical reactivity study was performed in the presence of simulated
vaginal matrix, on a minimum of 5 strains for each cultivable organism (58 strains
total), originating from 12 countries. The capacity to detect co-infection was tested
using two combinations i.e. low load of Trichomonas vaginalis (TV), Candida
glabrata and Candida krusei in the presence of a high load of Candida albicans and
low load of TV in presence of high load of C. glabrata. Vaginal swabs collected
from women with vaginal symptoms were characterized using various reference
methods and were then tested with the BD MAX QS Vaginal Panel. In Pouch
TV test was used as the reference method for TV while culture followed by BD
Phoenix identification was used for Candida species and the Nugent Score was
used as the reference method for BV. Amsels Criteria were used to provide a final
result for specimens with intermediate Nugent Score. The preliminary performance of
the assay for detection of trichomoniasis, Candida species associated with VVC and
BV was established using a Receiver Operating Characteristic (ROC) curve analysis.
The diagnosis of BV was determined using an algorithm based on PCR parameters
for the detection of five BV associated markers (Lactobacillus species, Gardnerella
vaginalis, Atopobium vaginae, BVAB-2 and Megasphaera-1).
Results: The assay identified all strains tested for each analyte in the inclusivity
study. Simulated co-infection studies demonstrated the capacity of the assay to detect
low loads of a specific organism in the presence of high load of another organism.
The preliminary assay performance results (sensitivity/specificity) based on analysis
of 771 characterized clinical samples were as follows: TV (94.4%/100%), Candida
species (86.8%/94.8%), Bacterial Vaginosis (91.9%/86.2%).
Conclusion: The BD MAX QS Vaginal Panel demonstrated high levels of detection
for BV, trichomoniasis and Candida species associated with VVC simultaneously
from vaginal specimens.

B-072
A fast and sensitive (13)--D-glucan microfluidic assay for the diagnosis and
treatment monitoring of invasive fungal infections

R. Kapoor, I. Yamaguchi, W. Chang. Wako Life Sciences, Inc., Mountain

View, CA
Incidence of invasive fungal infections (IFIs) is on the rise in recent decades with
increasing morbidity and mortality rate in many critically ill patients. Diagnosis is
often difficult with conventional methods, including biopsies (risk of complications),
imaging (nonspecific and limited use for early detection), and culture (slow, high rate
of false negative). Without proper diagnosis and treatment, IFI patients may die within
weeks.
(13)--D-glucan (BDG) is a major cell wall component of pathogenic fungi (e.g.
Candida, Aspergillus, Fusarium, Acremonium and Pneumocystis), and the increase
of BDG concentration in blood has been correlated to fungal infection in patients.
Currently, BDG assays are based on the recognition of BDG by coagulation factor
G from horseshoe crab amebocyte lysate. We have used a recombinant -glucan
recognition protein (rBGRP) that contains the binding domain of factor G to develop
a liquid-phase binding electrokinetic analyte transport assay (LBA-EATA) for BDG
in serum.
Our LBA-EATA assay takes advantage of some inherent features of a micro total
analysis system (TAS), including shorter reaction time, low reagent consumption,
and minimal reagent and sample handling. In a microfluidic chip channel, the complex
of BDG bound by DNA-rBGRP to speed complex migration and fluorescent dye
conjugated rBGRP for detection is concentrated by isotachophoresis (ITP) to enhance
detection sensitivity. The concentrated complex is subsequently separated from noise
signals in another part of the chip channel by capillary zone electrophoresis (CZE) and
detected by laser induced fluorescence (LIF). The ITP-CE process is completed within
3 minutes, and the assay can detect BDG in clinical specimens in the low picogram
per milliliter range (~ 10pg/ml). This sensitivity is sufficient to differentiate BDG in
healthy human population (10-40 pg/mL) and should be capable of detecting the early
onset of fungal infections when used in conjunction with other diagnostic methods.

Background: The BD MAX QS Vaginal Panel is an automated qualitative

in vitro diagnostic test for the direct detection of Candida species associated with
vulvovaginal candidiasis (VVC), trichomoniasis, and bacterial vaginosis (BV) from
vaginal swabs in women with clinical symptoms of vaginitis/vaginosis. The test
utilizes real time PCR for the detection and identification of organisms. This study

S152

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

B-073
Multicenter Evaluation of Mindray Fourth-generation CL-2000i HIV Ag/Ab
Combination Assay

S. Xu1, X. Han2, L. Li3, H. Wang4, L. Zhu5, L. Wu5, F. Xia5. 1Department of

Cell Biology, National Institutes of Food and Drug Control, Beijing, China,
2
Key Laboratory of AIDS Immunology of Ministry of Health, Department
of Laboratory Medicine, The First Hospital of China Medical University,
Shenyang, China, 3Department of Laboratory Medicine, The Sixth Peoples
Hospital of Shenyang, Shenyang, China, 4Key Laboratory of Infectious
Diseases, The Third Peoples Hospital of Shenzhen, Shenzhen, China,
5
Shenzhen Mindray Bio-Medical Electronics Co., Ltd., Shenzhen, China
Background: The Centers for Disease Control and Prevention (CDC) recently
proposed to use fourth-generation HIV immunoassays for screening. Mindray CL2000i HIV Ag/Ab Combination assay (CL-2000i) is a newly developed fourthgeneration HIV assay that simultaneously detects HIV p24 antigen and antibodies
to HIV-1 (groups M and O) and HIV-2. The objective of this study is to evaluate the
performance of CL-2000i via a multi-center study in three clinical trial centers and
National Institutes for Food and Drug Control (NIFD) of China on well-characterized
specimens.
Methods: The evaluation was performed on 635 HIV-infected and 1793 HIVuninfected specimens at three clinical trial centers and NIFD. HIV-infected specimens
were either confirmed with nucleic acid amplification testing (NAT), or repeatedly
reactive by other chemiluminescence immunoassays and clinical diagnostics. Positive
samples of antibodies to HIV-2, HIV-1/O, p24 antigens, and seroconversion panels
are obtained commercially. All samples were tested by CL-2000i in comparison with
ARCHITECT.
Results: The sensitivity of CL-2000i was 100% for antibodies of HIV-1 (635/635;
95% confidence interval: 99.40 - 100.00%). All the positive samples of the following
analytes were reactive: antibodies to HIV-1 Group O (5/5) and HIV-2 (30/30), and
HIV p24 antigen (23/23). The specificity of the assay was 99.83% (1790/1793; 95%
confidence interval: 99.51 - 99.94%). Testing of 13 HIV-1 seroconversion panels
indicates a comparable power of detecting acute HIV infection between CL-2000i
and ACHITECT. In each of 3 seroconversion panels, CL-2000i can detect one more
positive sample than ACHITECT, equivalent to at least 2 days earlier detection. This
was attributed to the high sensitivity of HIV-1 p24 antigen (< 0.25 IU/mL, the most
sensitive p24 assay in the market). One HIV antibodies negative sample determined
by third-generation HIV EIA assays are strongly reactive with both CL-2000i and
ACHITECT, indicating the power of detecting HIV p24 antigen by the fourthgeneration HIV Ag/Ab combination assays.
Conclusion: CL-2000i exhibits high sensitivity and specificity, and the ability of early
HIV detection. It can detect all the available known antibody positive samples for
HIV-1/M, HIV-1/O, HIV-2, and HIV p24 antigen. It is well suited for screening of
early HIV infection.

B-074
Use of an Integrated Molecular Diagnostic Platform with a Diverse Array of
Specimen Types To Address Laboratory Automation Needs

P. Rodriguez1, S. L. Moseley1, J. Osiecki2, M. Lewinski2. 1Roche

Diagnostics Corporation, Inc., Indianapolis, IN, 2Roche Molecular
Systems, Inc., Pleasanton, CA
Introduction: Molecular methods have revolutionized the way clinical labs identify
the presence of microorganisms in patient samples, monitor viral responses to therapy,
and characterize genetic disorders. With the demand on test menu expansion, and
common limitations in lab space, sample flexibility, budget for new instruments
and time for operator training, molecular laboratories require greater platform
consolidation. In addition to this, automated systems for PCR reduce tech time,
deliver faster and reliable results, minimize handling and processing errors, and help
improve confidence in reporting patient results. For these reasons, the cobas 4800
System was developed.
Objective: To review the key design features and testing solutions of an innovative
molecular diagnostic platform which address the automation and integration
challenges faced by molecular diagnostic laboratories.
Methodology: The cobas 4800 System is an automated PCR system, which offers
consolidation of Womens Health, Oncology and Microbiology* testing on a single
platform. The system is configured with two units - the cobas x 480 instrument for
sample preparation/PCR set up and the cobas z 480 analyzer for amplification and
detection. On the cobas x 480 instrument, multiple primary and secondary sample

types can be loaded and automatically scanned to allow for positive sample ID
tracking. Pipetting channels have built-in liquid-level detection and monitoring of all
pipetting steps to ensure quality sample processing. A robotic hand transfers specimens
to multiple incubation positions to assist with lysis, washing and elution of nucleic
acid. The cobas 4800 system has built-in enzymatic and engineered contamination
controls to prevent sample-to-sample or run-to-run carryover contamination. All
processes are controlled by intuitive software, which guides the operator through
initiating a run, monitoring of the instrument and run status. To increase efficiency a
new run can be started in parallel with the amplification and detection of a previous
run. User defined workflow software provides open-mode capabilities on the cobas
z 480 allowing the laboratory to design customized applications that fit their needs,
offering the potential for further platform consolidation.
Results and Conclusions: The cobas 4800 System test offerings cover a broad range
of disease state biomarkers and a diverse array of specimen types. For instance,
testing for high-risk HPV and HPV 16/18 genotyping can be done on cervical
specimens collected in PreserCyt either before or after cytology processing in order to
accommodate sample workflow. C. trachomatis and N. gonorrhoeae infection can be
assessed from male and female urine specimens, endocervical swabs, self-collected
or clinician-collected vaginal swabs and cervical specimens collected in PreservCyt
solution, which supports CDC recommendations for testing with a wide range of
specimen types. BRAF and EGFR mutation testing requires formalin-fixed paraffin
embedded tissue sections, including those already mounted on a glass slide. Other tests
currently in development include KRAS*, MRSA/SA*, HSV-1/2* and C. difficile*.
Collectively, the cobas 4800 System is an innovative molecular diagnostic solution
that addresses the increasing demand for test integration and sample flexibility.
*The cobas MRSA/SA Test, cobas HSV 1 and 2 Test, and the cobas Cdiff Test are
currently in development and not available for sale in the US.

B-075
Metabolomics approach to predict disease severity in influenza virus infection

C. Lin1, S. Lin2. 1Department of Laboratory Medicine, Chang-Gung

Memorial Hospital, Tao-Yuan, Taiwan, 2Research Center for Emerging
Viral Infections, College of Medicine, Chang Gung University, Tao-Yuan,
Taiwan
Background:Influenza A virus spread on a worldwide scale and infects a large
proportion of the human population. Early diagnosis and treatment can have an
important role in preventing the development of long-term complications or in
interrupting transmission of the infectious agent. A promising approach for the
predicting disease progression of influenza infection is targeting host factors that
affect disease outcome. Measuring metabolites represents the dynamic metabolomics
status of living system. Therefore metabolites levels can be regarded as the ultimate
response of biological systems to virus infection. This proposal is using metabolomics
strategy to decipher the disease progression after pathogenic influenza A infection.
Methods: Three different strains of influenza A viruses were used to investigate how
these different pathogenicity of influenza viral strains affect the metabolism of the
host. The three different influenza A H1N1 strains are A/Taiwan/141/2002 (141), A/
Taiwan/126/2009 (swine-origin influenza virus, SOIV) and A/PR8/34 (PR8). These
three strains have the same antigenicity in their hemaglutinin and neuraminidase
(H1N1). PR8 is a high-pathogenicity, SOIV is defined as moderate-pathogenic strain
and 141 is defined as mild-pathogenic strain. Female C57Bl/6 animal (6-12 weeks)
were anesthetized with Isoflurane and then infected by intranasal application of 200
PFU of viruses. Mice were monitored and weighted daily. Infected and nave mice (3
mice per group) were sacrificed on day 7 after infection. Bronchoalveolar lavage fluid
(BALF) samples were collected and apply to liquid chromatography MS/MS assay
based metabolomic analysis (AbsoluteIDQTM180 kit). Dissected mouse lung were
fixed and stained with hematoxylin and eosin for pathologic evaluation.
Results: Animals infected with PR8 had 23% weight loss and heavy leukocyte
infiltration was observed in lung. Principal component analysis was used to analyze
the correlations between the metabolites concentration in samples obtained from
nave mice and mice after different strains of influenza infection. It shows significant
difference in these four groups of BALF samples. All the amino acid concentrations
were dramatically elevated in PR8 infected mice reflected the extremely active
immune response. Long chain acylcarnitine were accumulated when the mice was
infected with PR8. Short chain acylcarnitines help the body produce energy and help
increase circulation. Acetylcarnitine (C2) was thought to be a more bioavailable
form for cells and can induce weight loss. It was significant elevated in PR8 infected
BALF. Sphingomyelins were significantly increased in 141 infected BALF; there was
no significant difference between the nave and the SOIV and PR8 infected groups.
Conclusion: Amino acids concentrations in mice BALF are correlated with the

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S153

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

severity of influenza infection. Combination of multiple markers of amino acids,
acylcarnitines, sphingomyelines can help to predict the severity of the infection.
The metabolic profiling could be a useful method applied to diagnose patients with
H1N1infection and can predict the disease severity.

B-076
Evaluation of the Dynex M MMRV Multiplex Immunoassay Panel vs. Three
Commercial Test Kits

S. Higgins, A. Karmali, M. Poncheri, C. Randall, A. Fusellier. Dynex

Technologies, Inc., Chantilly, VA
Background:Multiplex analysis of clinical samples offers significant advantages
in terms of sample usage and processing time but acceptance has been hindered by
high initial costs, lack of full automation, or both. We have developed M, a robust,
fully automated and cost effective chemiluminescent multiplex immunoassay system
to address these shortcomings. Here we present performance comparisons of the M
multiplex panel vs. two qualitative singleplex and one multiplex immunoassay system
for antibodies against Measles, Mumps, Rubella and Varicella-Zoster Virus (MMRV).
Methods:Polystyrene beads separately coated with antigen to MMRV immobilized
within 96-well M assay strips. Positive and negative control beads were coated
with goat anti-human IgG and combined MRC-5 and E6 cell lysate. Samples were
32 previously characterized human plasma and 7-point 4-fold dilution series of a
5-donor pool of normal serum and run on a modified Dynex DS2 automated ELISA
processing system. Identical samples were processed using conventional ELISA kits
and a commercially available multiplex assay designed for the Luminex Model 200.
Results:Sensitivity and specificity of the M system was calculated independently
for each kit using the 32-member reference panel. For each commercial system any
sample that tested above the negative cut off was considered as a True Positive (TP),
and any sample that fell below was considered a True Negative (TN). Sensitivity was
calculated as TP/(TP+FN). Specificity was calculated as TN/(TN+FP). Dilution series
of pooled positive serum shows M to possess greater assay-assay reproducibility than
any of the three kits that were examined.
Conclusion:The Dynex M multiplex immunoassay system shows excellent
correlation in both sensitivity and specificity vs. commercial ELISA and multiplex
kits across all analytes in an MMRV panel.
Sensitivity and Specificity of Dynex M vs. Commercial Kits
Predicate System Measles Mumps Rubella Varicella
Sensitivity, % SinglePlex 1
100.0
89.1
100.0
100.0
SinglePlex 2
85.0
89.1
90.0
100.0
Luminex
87.5
89.1
82.0
95.5
Specificity, % SinglePlex 1
100.0
100.0 100.0
100.0
SinglePlex 2
100.0
100.0 100.0
92.3
Luminex
100.0
100.0 100.0
100.0

B-077
Workflow Efficiency Through the Use of Mixed Batch Testing for Microbiology
Applications on the cobas 4800 system

M. Lewinski1, P. Rodriguez2, S. L. Moseley2, J. Osiecki1. 1Roche Molecular

Systems, Inc., Pleasanton, CA, 2Roche Diagnostics Corporation, Inc.,
Indianapolis, IN
Introduction: System flexibility for molecular diagnostic testing is becoming
increasingly important for clinical laboratories as space constraints and staffing
shortages impact efficiency. The ability to simultaneously run multiple assays on
a single instrument can reduce turn-around time, improve workflow, and have a
positive impact on job satisfaction for laboratory technologists. Mixed batch testing
streamlines laboratory processing through the use of automated sample extraction and
amplification with identical parameters optimized for multiple applications, allowing
users to mix samples and tests being included in a single run on the same instrument
system. The purpose of this study was to determine the impact of mixed batch
testing for Clostridium difficile (Cdiff), Methicillin-resistant Staphylococcus aureus
(MRSA), and Herpes simplex virus (HSV) when evaluated on the cobas 4800 system
compared with 3 other configurations of commercially available systems.
Methods: Batches of specimens for MRSA, Cdiff, and HSV were tested with
molecular diagnostic systems in the most efficient possible configuration; system A
- BDmaxTM Cdiff and MRSA and BD Viper HSV, system B - GeneXpert XVI Cdiff
and MRSA, and BD Viper HSV, system C - GeneXpert Infinity 48 Cdiff and MRSA,
and BD Viper HSV. Batch sizes of 46, 22, and 6 (including controls) MRSA, Cdiff,
and HSVspecimens, respectively, were assessed with the cobas MRSA/SA Test*,
cobas Cdiff Test*, and cobas HSV 1 and 2 Test* to reflect sample numbers that

S154

would be processed on a typical day in a medium sized clinical laboratory. Hands-on

time is defined as the labor elements associated with each system/process required
to start and finish a testing run that require a manual interaction. Automation time is
defined as the time during testing where the operator has no manual interactions with
the samples. Turn-around time is the actual clock time from start to finish to complete
a testing cycle.
Results: Mixed batch testing on the cobas 4800 system showed improvement
for hands on time of 3, 5, and 6-fold less than comparator systems A, B, and C,
respectively, when processing the same number of specimens. Automation time for
comparator systems A, B, and C was 2.5, 2.4, and 2.2 fold higher than what was
observed with the cobas 4800 system running mixed batch testing. Evaluation of
each configuration showed mixed batch testing improved workflow by reducing turnaround time by 70%, 39%, and 33% over method A, B, and C, respectively.
Conclusions: The system flexibility the cobas 4800 system allows for mixed batch
sample testing for MRSA*, C.diff* and HSV* on a single system which can provide
superior workflow efficiency for the increasing demands of the clinical laboratory.
* The cobas MRSA/SA Test, cobas HSV 1 and 2 Test, cobas Cdiff Test and the
cobas KRAS Test are currently in development and not available for sale in the US

B-078
Next-Generation Sequencing for Hepatitis B Genotype and Resistance Testing
in a Clinical Microbiology Laboratory

C. F. Lowe1, L. Merrick1, T. Mazzulli2, C. H. Sherlock1, G. Ritchie1.

1
Providence Health Care, Vancouver, BC, Canada, 2Mount Sinai Hospital,
Toronto, ON, Canada
Objective: Comparison of hepatitis B (HBV) genotyping and resistance testing
utilizing an in-house developed PCR assay and next-generation sequencing with a
line-probe assay.
Introduction: HBV is one of the most common causes of cirrhosis and hepatocellular
carcinoma worldwide. Antiviral therapy has been associated with a delay in disease
progression. In addition, HBV genotypes have been associated with different rates
of development to advanced liver disease and responses to interferon based therapy.
Newer diagnostic modalities such as next-generation sequencing (NGS) can provide
both genotype and resistance testing in one assay, and have the potential for increased
sensitivity and detection of resistant HBV subpopulations.
Methods: 80 clinical patient samples (plasma) were retrospectively studied. Genotype
(n=50) and resistance (n=80) were previously characterized by line-probe assay
(INNO-LiPA HBV DR Assay, Version 2/3 and INNO-LiPA HBV Genotyping Assay;
Innogenetics, Gent, Belgium). An in-house developed assay for hepatitis B genotype
and resistance testing was studied using the GS Junior (454 Life Sciences, Branford,
Connecticut). DNA was extracted using the MagNA Pure LC 2.0 (Roche Diagnostics,
Mannheim, Germany). PCR amplified a 418bp amplicon of the polymerase region
(codons 143 - 281). Amplicons were then sequenced on the GS Junior following
the manufacturers protocols. A third party bioinformatics software company (ABL
TherapyEdge, Luxembourg) provided support in the interpretation of genotype and
resistance profile of the HBV based on EASL Clinical Practice Guidelines. Sanger
sequencing of the PCR amplicons was performed using the 3730 DNA Analyzer
(Applied Biosystems, Foster City, USA) for discrepant genotype results between the
line-probe assay and NGS.
Results: 50 samples were compared to the INNO-LiPA HBV Genotyping. There
was concordance in 47/50 samples (A=3,B=24,C=14,D=5,E=1). Sanger sequencing
for the 3 discrepant samples (NGS = B,C,C vs. INNO-LiPA=E,D,B, respectively)
confirmed the results of the next-generation sequencing assay. Resistance testing for
80 samples included mutations at the following loci: M204V/I, L180M, A181T/V,
N236T, V173L, T184G and S202I/G. No resistance mutations were detected by lineprobe assay in 61 samples. Five of these samples had a mutant subpopulation (% of the
virus population with a base pair mutation at known resistant loci) detected by NGS:
2 samples with M204I (1.6%; 2.5%), 1 sample with M204I(100%)/L180M(2.5%),
1 sample with V173L (2.5%) and 1 sample with A181T (3.9%). 19 samples with
resistance mutations detected by line-probe were also confirmed by NGS.
Conclusions: Utilizing an in-house developed assay with a novel PCR targeting the
polymerase region of HBV, genotyping and resistance testing for the most significant
mutations can be performed with 1 PCR and 1 NGS reaction. NGS can potentially
provide clinicians with increased sensitivity, earlier detection and detailed analysis
of resistance profiles, as well as accurate detection of genotype. As a result, nextgeneration sequencing may become more accessible to incorporate into clinical
microbiology laboratories for hepatitis B genotyping and resistance testing.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

responsible for septicemia and antimicrobial resistance are significantly important
for appropriate antimicrobial therapy. This study aimed to evaluate the PCR-reverse
blot hybridization assay (PCR-REBA) capable of the identification of pathogens and
antimicrobial susceptibility test (AST) in blood culture specimens.

B-079
An Immunoturbidimetric Assay for Hyaluronic Acid

J. A. Chapo, B. Hurley, I. Muncy, K. R. Pitts. Corgenix, Inc., Broomfield,

CO
Background: Hyaluronic acid (HA), also known as hyaluronate or hyaluronan,
is a linear glycosaminoglycan - a high molecular weight polysaccharide with an
unbranched backbone composed of alternating sequences of -(1-4)-D-glucoronic
acid and -(1- 3)-N-D-acetylglucosamine moieties. Each dimer is referred to as one
unit and has a molecular weight of approximately 450 Da. The HA molecule can
vary in length from less than 10 to more than 1,000 units. HA is mainly produced
by fibroblasts and other specialized connective tissue cells. It is a major constituent
of connective tissue matrix (proteoglycan) and participates in various cell-to-cell
interactions. HA is widely distributed throughout the body and can be found as a free
molecule in plasma and synovial fluid. In plasma, the half-life of the HA molecule
has been estimated to be about 5-6 minutes. HA is found in synovial fluid in high
concentrations and is responsible for normal water retention and articular lubricant.
Synovial HA may pass into plasma via the lymphatic system. In circulation, HA
levels are maintained by an efficient receptor-dependent removal mechanism present
in sinusoidal endothelial cells (SEC) of the liver and by the enzymatic action of
hyaluronidase. Because the liver plays a central role in maintaining HA homeostasis,
increased plasma levels of HA may serve as a sentinel for hepatic inflammation,
fibrosis and cirrhosis. We report here the development of an immunoturbidimetric
method for detecting HA in a blood sample.
Methods: An R1 reagent/reaction buffer was developed and optimized to augment the
specific agglutination reaction of the coated microparticles with hyaluronic acid in
the serum samples. An R2 reagent/coated polystyrene microparticles was developed
using functionalized polystyrene microparticles covalently coated with HA binding
proteins using standard conjugation techniques. Iterative combinations of R1 and
R2 reagents were systematically tested to achieve consistent linearity and precision.
Numerous iterations of coating and blocking buffers were assayed to further enhance
assay manufacturability and consistency. Finally, in-process testing of linearity and
precision was conducted to ensure robust performance to the end-user.
Results: The assays limit of detection (LOD) was determined to be 11.36ng/mL;
limit of blank (LOB) was found to be 6.68ng/mL; limit of quantitation (LOQ) was
determined to be 20.00ng.mL. Rigorous precision testing demonstrated the assays
consistency of the course of 20 operating days with a 5.3% overall percent coefficient
of variation. Assay linearity was between 25ng/mL to 750ng/mL for samples tested.
The overall average percent recovery was 103.5% and lot to lot values showed no
statistical difference (p= 0.736) across both the medical decision range and the range
above. Real time stability concluded the assay can reliably and consistently measure
samples over the course of at least 12 months, with a deviation of less then 10% of
the mean for each group.
Conclusions: The data presented herein highlight the robust performance of this
immunoturbidimetric assay. In summary, these data demonstrate the overall
performance of the assay was consistent with a predicate HA-ELISA and that values
obtained will be consistent both over time and from lot to lot.

B-080
PCR-Reverse Blot Hybridization Assay for Identification of Pathogens causing
Sepsis from Positive Blood Cultures

H. Kim1, J. Kim1, H. Wang2, S. Kim3, Y. Kim1, S. Park4, E. Choi4, G. Kim1,

S. Ahn1, S. Park1, S. Park5, H. Jin6, Y. Kim7, Y. Uh4, H. Lee1. 1Department
of Biomedical Laboratory Science, College of Health Sciences,
Yonsei University, Wonju, Korea, Republic of, 2M&D, Inc., Wonju Eco
Environmental Technology Center, Wonju, Korea, Republic of, 3Institute for
Life Science and Biotechnology, Yonsei University, Seoul, Korea, Republic
of, 4Department of Laboratory Medicine, Yonsei University Wonju College
of Medicine, Wonju, Korea, Republic of, 5Department of Clinical Laboratory
Science, College of Health and Therapy, Daegu Hanny University, Daegu,
Korea, Republic of, 6Department of Clinical Laboratory Science, College
of Health Sciences, Catholic University of Pusan, Busan, Korea, Republic
of, 7Department of Internal Medicine, Yonsei University Wonju College of
Medicine, Wonju, Korea, Republic of

Methods : The PCR- REBA, REBA Sepsis-ID (M&D, Wonju, Republic of Korea)
was designed to contain a total of 25 probes including 6 Gram-positive bacteria (GP)
specific probes, 8 Gram-negative bacteria (GN) specific probes and 5 Candida species
specific probes with a pan-bacteria, a pan-GP and a pan-GN probes. In addition, it
includes mecA ,vanA and vanB probes for detection of antibiotic-resistant bacteria.
For evaluation of the REBA Sepsis-ID, a total of 300 positive blood culture bottles
from BACTEC FX (Becton Dickinson Diagnostic System, Spark, MD, USA)
or BacT/ALERT 3D (bioMrieux, Marcy, France) were used. The identification
of organisms and antimicrobial susceptibility tests (ASTs) were conducted by the
microplate method, the MicroScan system (Siemens Healthcare Diagnostics,
Sacramento, CA, USA), and the Vitek 2 system (bioMrieux).
Results : The correct agreement rates between conventional identification and AST
methods and PCR-REBA for GP, GN, Candida and polymicrobials were 94.5%,
97.3%, 100% and 91.7%, respectively. Of 92 methicillin-resistant Staphylococcus
species, mecA gene was detected in 90 (97.8%) samples and vanA gene was correctly
detected in one (100%) sample which was identified to vancomycin-resistant
Enterococcus (VRE) by phenotypic examination.
Conclusion : Newly developed REBA Sepsis was a rapid and accurate molecularbased method for simultaneous rapid detection of causative agents and antimicrobial
resistant genes in positive blood cultures even though there was a limitation for
evaluation with negative cultures such as nonviable after exposure to antibiotics or
small amount bacteria.

B-081
Assessment of Chlamydia trachomatis and Neisseria gonorrhoeae with the
cobas CT/NG v2.0 Test on the cobas 4800 system: Infection prevalence in
pregnant women enrolled in a large multicenter clinical trial

J. Osiecki1, P. Rodriguez2, S. L. Moseley2, J. Duncan1, M. Lewinski1.

1
Roche Molecular Systems, Inc., Pleasanton, CA, 2Roche Diagnostics
Corporation, Inc., Indianapolis, IN
Background: Miscarriage, pre-term delivery, low birth weight, and morbidity in the
neonate are potential consequences when pregnant women become infected with
sexually transmitted diseases. In an effort to identify women with infection, treatment
guidelines recommend screening pregnant women for Chlamydia trachomatis (CT)
and Neisseria gonorrhea (NG) on the first prenatal visit. This study was performed
to determine the frequency of CT and NG infection observed in pregnant women
enrolled in a large clinical trial study population.
Methods: This multicenter retrospective cohort analysis was performed with archived
specimens collected during the VENUS clinical trial and prospective specimens
collected during the VENUS II clinical trial, to characterize the clinical performance
of the cobas CT/NG v2.0 Test on the cobas 4800 system. As recommended by the
FDA, Patient Infected Status (PIS) was determined for each enrolled participant using
two FDA-cleared nucleic acid amplification tests (NAATs) as comparator assays.
PIS was defined as positive when results from NAATs with different target regions
generated positive results with collected samples. Diverse settings in the United States
served as specimen collection sites and included obstetrics-gynecology practices,
family planning clinics, and STD clinics.
Results: Of 6,035 eligible participants, 6,004 subjects were evaluated for CT and/
or NG for primary anlaysis. PIS determined 365 women and 92 men were infected
with CT, and 122 women and 67 men were infected with NG. Of the female patients
evaluated, 6.93% (365/5265) were found to be positive for CT infection and 1.75%
(92/5265)were positive for NG according to PIS outcomes. Alternatively, 8.4%
(17/202) of eligible pregnant women were positive for CT, where 1.48% (3/202) of
pregnant women were considered positive for NG by PIS.
Conclusion: Screening of pregnant women for CT and NG with the cobas CT/NG
v2.0 Test on the cobas 4800 system compared to two additional NAATs during the
VENUS clinical trial revealed comparable rates of infection for CT and NG between
pregnant women and non-pregnant women in the general female population.

Background : Sepsis is a lethal medical condition leading to the systemic

inflammatory response to infection. Sepsis is the 10th leading cause of death in the
United States, accounting for 6% of all deaths and an estimated 135,000 patients die
each year of sepsis-associated complications in Europe. Early detection of pathogens

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S155

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

invasive infection with CRE. Infection with CRE has limited therapeutic and high
morbidity and mortality.

B-082
Performance evaluation of the Access HCV Ab PLUS assay on the UniCel DxI
800 system

H. LE GUILLOU-GUILLEMETTE1, D. NOGUES2, A. DAVID2, R.

FALCOU-BRIATTE3, F. LUNEL-FABIANI1. 1Laboratoire de virologie,
Dpartement des Agents infectieux et Pharmaco-toxicologie CHU Angers,
ANGERS, France, 2BIO-RAD, STEENVOORDE, France, 3BIO-RAD,
MARNES LA COQUETTE, France
Background: The Access HCV Ab PLUS assay (Bio-Rad) is a chemiluminescent
microparticle immunoassay (CLIA) for the qualitative detection of antibodies to the
hepatitis C virus in human serum or plasma. The purpose of these studies was to
evaluate diagnostic performance.
Four serological automated assays were compared: Access HCV Ab PLUS assay
on UniCel DxI 800 Immunoassay system (Beckman Coulter Inc.), Architect antiHCV assay on Architect I2000SR analyser connected to the APS system (Abbott
Diagnostics), Elecsys Anti-HCV II assay on MODULAR ANALYTICS E170
system or Cobas e601 system (Roche Diagnostics) and ADVIA Centaur HCV assay
on ADVIA Centaur XP system (Siemens).
Methods: The Access HCV Ab PLUS assay is a two-step indirect antibody detection
format.
First study: 659 fresh samples tested for HCV diagnosis in the routine virology
laboratory at the University hospital of Angers were prospectively tested, 199
frozen positive samples from a retrospective data collection of patients sera and 2
commercial panels were tested on Access HCV Ab PLUS and Architect anti-HCV
assays. Two other commercial panels were tested only on Access assay. For Architect
assay, data from the supplier were used.
Second study: the specificity was estimated by testing 500 non-selected fresh serum
samples from a routine laboratory, 3 commercial panels plus one anti-HCV low titer
performance panel were tested on Access HCV Ab PLUS, Elecsys anti-HCV II and
ADVIA Centaur HCV assays. Another panel was used on Access and Elecsys assays.
For ADVIA Centaur assay, data from the supplier were used.
Results: First study: On unselected routine samples, the agreement between the two
assays was equal to 99.5%. The clinical specificity was 98.9% (95% CI: 97.8-99.6%)
and 99.7% (95% CI: 98.9-100%) for Access HCV Ab PLUS and Architect anti-HCV
assays, respectively. The clinical sensitivity for all positive samples was 100% for
both assays. The sensitivity on seroconversion samples showed performance in
accordance with the state-of-the-art for Access and Architect assays.
Second study: 488 of the 500 non-selected samples were true negative. The clinical
specificity was 99.6% (95% CI: 98.53-99.95%) for Access HCV Ab PLUS, Elecsys
anti-HCV II and ADVIA Centaur HCV assays. The concordance between the 3
assays was 99.20%. The clinical sensitivity from 12 positive samples was 100% for
all assays. Using 4 commercial seroconversion panels and one anti-HCV low titer
performance panel, Access HCV Ab PLUS and Elecsys anti-HCV II assays showed
equivalent performance and detected earlier than ADVIA Centaur HCV assay.
Conclusion: The performance of the Access HCV Ab PLUS assay on the UniCel DxI
800 immunoassay system was excellent in terms of specificity and sensitivity. The
clinical specificity was slightly better with Architect anti-HCV assay as compared to
the other assays. The clinical sensitivity on true positive samples was 100% for all
assays. The seroconversion sensitivity was better on UniCel DxI 800, Architect and
Modular than on ADVIA Centaur. Adapted for high throughput routine testing, the
Access HCV Ab PLUS assay performed on UniCel DxI 800 immunoassay system is
fully suited for the screening of HCV infection in diagnostic laboratories.

B-084
Carbapenem-Resistant Enterbacteriaceae (CRE) in Geriatric Population:

R. Khoury, P. Patel, B. P. Salmon, A. Gandhi, P. Gudaitis, D. Gudaitis.

Aculabs, Inc., East Brunswick, NJ
Background: CRE refers to Carbapenem-Resistant and/or Carbapenemase-producing
Enterobacteriaceae. These are families of bacteria that are resistant to several classes
of antibiotics including one of the carbapenen group. Carbapenem antibiotics are
used to treat infection caused by gram negative bacteria as the last line of treatment.
Multidrug-resistant gram negative bacteria especially CRE are becoming the new
super bug in the Long-Term Care Facilities where most of the residents are elderly,
frail and are on multiple medications. The most common enterobacteriacae are
Klebsiella species and Escherichia col.; over 40% mortality has been reported with

S156

Methodology: 35,330 specimens were collected for culture from residents in LongTerm Care Facilities over a period of 6 months. All positive cultures were subcultured
and then identified using MicroScan Walkaway 96 conventional panels, the isolate
was considered CRE if it was resistant to one or more of the carbapenem, with
Ertapenem nonsusceptibility being the most sensitive indicator of carbapenemase
production. Statistical analysis were done using Analyse-it.
Results: 18,569 (52.6%) specimens were positive, 320 patients had CRE positive
culture, the most common source was urine 250 cases (78.1%) followed by wound 42
cases (13.1%), respiratory 16 cases (5.0%) ), rectal swab 5 cases (1.6%) and blood 4
cases (1.3%). Majority of the cases were reported in August (23.8 % of all cases) and
the lowest was in November (11.3 % of all cases); we noticed an increase in the cases
in January which was due to respiratory infections. The most common organism was
Klebsiella Pneumoniae (ESBL or MDR), followed by E. CLOACAE MDR, E. Coli,
and SERRATIA MARCESCENS.
Conclusion: CRE incidence is high in the long-term care facilities, facilities should
follow the CDC recommendation to implement the detect and protect strategies.
Prompt implementation of infection prevention and control measures requires close
collaboration between clinical laboratory, infection prevention staff, physicians, and
nurses. Early detection and implementing infection control and prevention will help
reducing the transmission to other residents in addition to identifying the risk factors
for CRE. Cautious and appropriate use of antimicrobial therapy for the treatment of
suspected infections in residents of long-term care facilities are very important to
prevent the occurrences.

B-086
Reduced Methicillin-resistant Staphylococcus aureus infections rate after the
three-year implementation of a Rapid Molecular Screening in Intensive Care
Unit

P. Stano, M. Avolio, R. De Rosa, M. L. Modolo, A. Camporese. S. Maria

degli Angeli Hospital, Pordenone, Italy
Background: Previous studies have suggested that Methicillin-resistant
Staphylococcus aureus (MRSA)-colonized patients are at higher risk for acquiring
MRSA infections compared with non-colonized patients, while in the hospital, thus,
MRSA-carriers should be monitored closely. Molecular screening methods have the
advantage of high sensitivity and rapid turnaround times (TATs), assuring a rapid
delivery of results and, consequently, improving infection control procedures and the
clinical impact. Here, we assess the effects of implementation of a specific MRSA
bundle, based on rapid molecular screening for MRSA and decolonization, on the
prevalence of MRSA infection at our ICU, over a study period of three years.
Methods: The study was conducted over two time periods, before and after
implementation of a specific MRSA bundle. A total of 431 and 886 nasal screening
swabs were obtained from ICU patients, respectively before and after the bundle
implementation and analyzed by the molecular test Xpert MRSA (Cepheid). The prebundle period (from April 2009 through December 2010), has been used to assess the
rate of ICU colonization and to evaluate the more appropriate measures to be applied
in MRSA-carriers, thus screening results did not activated any preventive measures
in patients colonized. Later, an MRSA bundle was implemented from January 2011
through December 2013 at our ICU (post-bundle period). The bundle consisted of
rapid molecular screening for MRSA nasal carriage, at the ICU admission, contact
precautions and nasal decolonization (mupirocin 2% ointment three times-a-day
for five days) for patients colonized with MRSA. Clinical samples from patients
suspected as being infected with MRSA were tested by standard laboratory culture
procedures. The results of rapid nasal screening were available to physicians within 2
hours from specimen receipt.
Results: In the pre-bundle period, 9 patients (2%) developed a generalized MRSA
infection, but during the three years that followed the bundle implementation (postbundle period), MRSA infection rates declined from 2 % to 0.2% (2 patients) with a
total MRSA infection decrease of 100% in the third year post-intervention. On the
contrary, MRSA colonization rates at admission increased from 7,1% in the prebundle period to 8,6 % in the post-bundle period. Overall, during the three years postintervention, the relative risk reduction, absolute risk reduction and relative risk were
as follows: 0.9 (95% confidence interval: 0.58-0.98), 0.26 (95% confidence interval:
0.1-0.33) and 0.09 (95% confidence interval: 0.014-0.4), respectively. Moreover, the
risk of MRSA infections among colonized patients, already reduced in 2011 (Relative
Risk 0.18, 95% confidence interval: 0.008-1.1) and 2012 (Relative Risk 0.12, 95%
confidence interval: 0.006-0.8), compared with the pre-bundle period, dropped

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Infectious Disease

Wednesday, July 30, 9:30 am 5:00 pm

dramatically in 2013 (Relative Risk 0.000, 95% confidence interval: 0.000-0.6), with
no case of MRSA infection reported.
Conclusion: The present study showed that a strategy of active surveillance based on
rapid molecular screening for MRSA, immediately after admission, rapid reporting
and prompt nasal decolonization, resulted in a significant decline in MRSA infections
rate in our ICU over the three years post-bundle period. Real time PCR demonstrated
a superior sensitivity to culture and rapid TATs, allowing a better management of
MRSA-carriers who will more likely develop MRSA infections.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S157

Lipids/Lipoproteins

Wednesday, July 30, 9:30 am 5:00 pm

Wednesday, July 30, 2014

Poster Session: 9:30 AM - 5:00 PM
Lipids/Lipoproteins

of atherosclerosis and thrombotic diseases. This study intends to assess the level of
lipoprotein a in diabetic patients.
Methods: The study included 204 patients with type 2 diabetes mellitus and 204 age
and sex matched controls. Lipoprotein a levels were measured and comparison was
done between the lp(a) levels in diabetic patients and control.
Result: Mean serum Lp(a) levels in diabetes mellitus patients was 44.235.8 mg/dl,
which was significantly higher when compared to control group (mean 21.111.2 mg/
dl, p < 0.05).
Conclusion: The result of present study indicates that levels of Lp(a) are increased in
patients with type 2 diabetes mellitus.

B-087
A study of the difference in the Request of laboratory lipid metabolism tests in
Primary Care in Spain

C. Hernando de Larramendi1, M. Lopez-Garrigos2, L. Maiz3, L. Rabadan4,

M. D. Calvo5, A. Rodriguez-Rodriguez6, C. Gallego7, M. HerranzPuebla8, V. Poncela9, M. J. Baz10, M. J. Martinez-Llopis11, T. Avello12, I.
Llovet13, C. Lorenzo14, M. Lopez-Hoyos15, M. J. Zaro16, L. Lopez-Yepes17,
M. Salinas2. 1Hospital Universitario Severo Ochoa, Leganes, Spain,
2
Hospital Universitario San Juan, San Juan de Alicante, Spain, 3Hospital
Universitario Lucus Augusti, HULA, Lugo, Spain, 4Complejo Asistencial
de Soria, Soria, Spain, 5Hsopital Clinico de Valladolid, Valladolid, Spain,
6
Complejo Asistencial de Palencia, Palencia, Spain, 7Hospital Rafael
Mendez, Lorca, Spain, 8Hospital Universitario de Getafe, Getafe, Spain,
9
Hospital Universitario de Burgos, Burgos, Spain, 10Hospital de Llerena,
Badajoz, Spain, 11Hospital de Denia, Denia, Spain, 12Hospital San Agustn,
Aviles, Spain, 13Hospital Verge de la Cinta, Tortosa, Spain, 14Hospital
Santa Barbara, Puertollano, Spain, 15Hospital Universitario Marques de
Valdecillas, Santander, Spain, 16Hospital Don Benito, Villanueva, Spain,
17
Hospital Virgen del Castillo, Yecla, Spain
BACKGROUND: To compare the inter-practice variability in lipid metabolism
laboratory tests requested by General Practitioners (GPs) in Spain, according
geographic and hospital characteristics, using appropriateness indicators, to try to
ascertain the degree in requesting appropriateness.
METHODS: We obtained the number of serum cholesterol (Chol), HDL-cholesterol
(HDL-chol) and tryglicerides (Tryg) requested by GPs for the year 2012 from 76
laboratories at different hospitals from diverse regions across Spain. Every patient
seen in any primary care center (PCC) of any of these 76 health departments,
regardless of the reason for consultation, gender or age, was included in the study.
Two types of appropriateness indicators were calculated: every test requests per
1000 inhabitants and ratio of related tests requests (HDL-chol/Chol, Tryg/Chol). The
indicators results obtained in different location and for type of management were
compared.
RESULTS: In total GPs requested 16013622 laboratory lipid metabolism tests in year
2012 in a Spanish population (17679195 inhabitants) that is almost half of the whole
country population. Chol, HDL-chol and Tryg per 1000 inhabitants indicators results
ranged from 106.3 to 550.7; 20.4 to 417.5 and from 94.0 to 439.2 respectively. The
variability of HDL-chol/Chol, Tryg/Chol indicators results was also considerable, and
ranged from 0.19 to 1.00 and from 0.54 to 1.00 respectively.
There were significant differences according to hospital setting in tests requests per
1000 inhabitants. In rural location Chol, HDL-chol and Tryg were higher. However,
no significant differences according to hospital setting in related tests requests
indicator results were detected.
In relation to institution management, no significant differences were obtained.
DISCUSSION: The high variability observed is difficult to explain by differences in
patient case mix between regions.
CONCLUSION: There is a need to design and establish strategies from laboratory in
consensus with requesting clinicians to improve lipid metabolism tests appropriate
use and hence clinical decision making.

B-089
Lipoprotein A Levels in Individuals with Type 2 Diabetes Mellitus Attending
Tribhuvan University Teaching Hospital ,Nepal

L. Shrestha. IOM, Nepal, kathmandu, Nepal

Background: Type-2 diabetes mellitus is a common disease, affecting a large
proportion of individuals worldwide. It is also recognized as an independent risk factor
for cardiovascular diseases. Various markers for assessing risk for cardiovascular
diseases are used in clinical laboratory. One of the emerging marker in this regard is
lipoprotein a, elevated level of which is regarded to be associated with increased risk

S158

B-091
High-fat diet and lipid profile.

P. D. Perris, I. Fernandez, M. C. Sanahuja, M. C. Mambrin, N. H.

Slobodianik, M. S. Feliu. School of Pharmacy and Biochemistry, Buenos
Aires, Argentina
The importance of diet in maintaining health is widely accepted and recognized. Diet
lipid profile is important to prevent chronic diseases and improve the quality of life
of individuals. The objective is to analyze the effect of high-lipid diet from different
sources, on triglycerides (TG), total cholesterol (TC), noHDL cholesterol and fatty acid
profile in serum of growing rats. Weanling Wistar rats were fed during 10 days with
40% dietary fat provided: by butter (B group); by olive oil (O group) and by high oleic
oil (AO group). Control group (C) received normocaloric diet according to AIN93.
Diets fatty acid profiles were determined by gas chromatography (GC); 6/3 and
unsaturated/saturated (PUFA/SFA) ratios of diets were calculated. Serum levels of TG
and TC were determined by enzymatic-colorimetric method and fatty acid profile was
determined by GC. The statistical analysis used Bartletts test, followed by one-way
analysis of variance (ANOVA) and Dunnett as post test (*p<0.01). RESULTS: diets:
6/3 ratio: B=5,6/1; O=49,6/1; AO=86/1; C=9/1; PUFA/SFA, B=0.06; O=1,36;
AO=0,72; C=3,89. Serum (meanSD mg/dL) TG B=113.031.2*; O=77.612.1;
AO=67.015.9 C=59,114,8; TC B=89.210.1*; O=73.17.3; AO=71.010.6
C=62.113.6 Fatty acids profile expressed as area%SD were:
B
O
AO
C
Palmitic
21,22,5
15.71,7
13.50.7
17.31,4
Oleic
19.15.2 *
22.05.1 *
33.04.8 *
10,6 2,0
Linoleic
8.91.8 *
11.82.8 *
8.91.0 *
19.03,5
-Linolenic
0.40.1*
0.50.2*
0.30.1*
1.20.3
Araquidonic
6.361.45
8.151.97
9.481.73
8.592.15
EPA
0.930.82
0.670.30
0.880.19
0.830.44
DHA
1.220.30
0.740.20*
0.830.18*
1.250.24
TG and TC levels in B group were statistically higher compared to C. Experimental
groups showed higher serum oleic acid levels with lower -linolenic and linoleic acids
levels compared to C. This fact would exacerbate the route of the 9 family and
decreases essential fatty acids. It seems that the sources of dietary lipids provoked
changes in serum fatty acid profile levels, but not in response to the high fat percentage.
Supported by UBACyT 20020120200068

B-093
Comparison of a Direct Enzymatic Assay and Polyacrylamide Tube Gel
Electrophoresis for Measurement of Small Dense Low-Density Lipoprotein
Cholesterol

P. Srisawasdi1, S. Vanavanan1, M. Rochanawutanon1, J. Kerdmongkol1, M.

H. Kroll2. 1Department of Pathology, Faculty of Medicine, Ramathibodi
Hospital, Mahidol University, Bangkok, Thailand, 2Quest Diagnostics,
Mandison, NJ
Background: Small-dense low density lipoprotein cholesterol (sdLDL-C) has been
linked to the progression of cardiovascular disease. We compared two methods
for determination of sdLDL-C: a direct enzymatic method (sdLDL-EX) and
polyacrylamide tube gel electrophoresis (sdLDL-PGE). In addition, we evaluated the
associations of these lipid measures with other atherosclerosis-related markers.
Methods: A total of 242 outpatients (age more than or equal to 19 years old) were
recruited. All blood samples (excluding those with triglycerides over 400 mg/dL)
were analyzed for lipid profile with sdLDL-PGE (Quantimetrix Lipoprint, CA)
and sdLDL-EX assay (Denka Seiken, Japan) an enzymatic-surfactant-based assay.
We also evaluated the following atherosclerosis-related markers: apolipoprotein A-I
(apoA-I), apoB, glucose, hemoglobin A1c (HbA1C), high-sensitivity C-reactive
protein (hsCRP), creatinine, cystatin C, and vitamin D. The sdLDL-PGE method

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Lipids/Lipoproteins

Wednesday, July 30, 9:30 am 5:00 pm

separates the intermediate density lipoprotein (IDL) particles into three midbands
(MID-A to C) and the LDL particles into seven subfractions (LDL-1 to 7); the sdLDLPGE result is calculated as the sum of cholesterol concentrations from LDL3 to LDL7.
Results: The mean age of the patients (58 males and 184 females) was 54.5 years. The
regression equation between the sdLDL-PGE (x) and sdLDL-EX assay (y) was ymg/dL
= 0.748x +26.14, r = 0.713. The sdLDL-EX assay yielded higher measured sdLDL-C
concentrations than the sdLDL-PGE assay (33.06+12.79 vs. 9.3+12.19 mg/dL, P
<0.001); however, the absolute difference between two methods did not significantly
correlate with the average sdLDL-C concentration (R2 = 0.005, P = 0.290). sdLDL-C
as measured with the sdLDL-EX assay exhibited significant positive correlations
with VLDL, MIDC, MIDB, and LDL2 (all P <0.001), which have been suggested as
atherogenic lipoproteins, but did not correlate with the less atherogenic lipoproteins
MIDA (P = 0.891) and LDL1 (P = 0.604). The sdLDL-EX and sdLDL-PGE methods
yielded similar patterns of correlation between sdLDL-C and atherosclerosis-related
markers: positive correlations with TG, TC, LDL-C, apoB, glucose, and HbA1C but
inverse correlations with HDL-C and vitamin D. However, there was no significant
correlation between sdLDL-C and apoA-I, hsCRP, or creatinine levels with either
method.
Conclusion: The direct enzymatic assay for sdLDL-C correlated well with the assay
based on polyacrylamide gel electrophoresis. The enzymatic method appears to
measure cholesterol in a boarder range of atherogenic lipoprotein particles than PGE,
and thus contribute in directing specific interventions of cardiovascular prevention.
Because the direct enzymatic assay can be automated, it can be used as a routine
method to assess small dense LDL cholesterol.

B-094
Performance Evaluation of a New Ready-to-use Liquid Triglycerides Assay on
the High-Throughput ADVIA Chemistry Systems

P. Datta1, J. Dai2. 1Siemens Healthcare Diagnostics, Tarrytown, NY,

2
Siemens Healthcare Diagnostics, Glasgow, DE
Background: Measurement of serum triglycerides is an important component to
determine lipid status of a patient. Increased serum triglycerides are risk factors
for cardio vascular disease. Serum triglycerides are also used in the diagnosis and
treatment of patients with diabetes mellitus, nephrosis, liver obstruction, other
diseases involving lipid metabolism, or various endocrine disorders. The automated
ADVIA Clinical Chemistry Systems currently have a Siemens serum triglycerides
assay that requires manual mixing of two reagent components before use on the
system. An improved assay, Triglycerides_2 (TRIG_2), using ready-to-use liquid
reagents, is under development. Furthermore a new concentrated reagent (TRIG_c)
was developed to be automatically diluted on-system to provide larger number of
tests per kit for high-volume users. The objective of this study was to evaluate the
performance of both new assays on the ADVIA Chemistry Systems.
Methods: In the ADVIA Chemistry TRIG_2 and TRIG_c assays, sample is diluted
and reacted with a single reagent for 5 minutes. The lipase in the reagent hydrolyzes
triglycerides into glycerol, which is then converted into glycerol-3-phosphate by
glycerol kinase, the latter then being oxidized to H2O2 by glycerol oxidase. The H2O2
is colorimetrically (at 505 nm) detected by a Trinders reaction. The triglyceride
concentration in a sample is determined from a linear calibration curve using Siemens
ADVIA Chemistry Calibrator. The performance evaluation in this study included
precision, interference, linearity, and correlation with a commercially available
triglyceride (TGL) assay run on the Dimension XPAND system. Data were collected
for all ADVIA Chemistry Systems (ADVIA 1200, ADVIA 1650, ADVIA 1800, and
ADVIA 2400), which use the same ADVIA Chemistry TRIG_2 or TRIG_c reagent
packs, calibrators, and commercial controls.
Results: The imprecision (total %CV) of the new ADVIA Chemistry assays with
two-level commercial controls and two serum pools ranging from ~90 to ~500 mg/
dL (n = 80) on all ADVIA Chemistry Systems (1200/1650/1800/2400) was 2.1%
(for both TRIG_2 and TRIG_c). The analytical ranges of the new assays are from
10 - 550 mg/dL (extendable to 1100 mg/dL by auto-dilution). The assays correlated
well with the Dimension TGL assay: TRIG_2 = 0.94 [TGL] + 4.4 and TRIG_c =
0.93 [TGL] + 4.3 (r = 0.99, n = 101; sample range: 20-540 mg/dL for both). The
new assays demonstrated no interference at a triglycerides level of ~150 mg/dL with
unconjugated or conjugated bilirubin (up to 15 mg/dL), hemoglobin (up to 500 mg/
dL), and ascorbic acid (up to 3 mg/dL). Minimum on-system stability for both was 60
days (with reagent blanking every 14 days).
Conclusion: The data demonstrates good performance of the TRIG_2 and TRIG_c
assays on the high-throughput ADVIA Chemistry Systems from Siemens Healthcare
Diagnostics.* Under development. Not available for sale in the USA.

B-095
Comparison of equations for the calculation of LDL-C in hospitalized patients

T. S. Pillay1, J. Martins1, L. M. Murray1, S. A. S. Olorunju2. 1University

of Pretoria/NHLS, Pretoria, South Africa, 2Medical Research Council,
Pretoria, South Africa
Background: Prediction of cardiovascular disease (CVD) mortality is dependent
on the calculation of low density lipoprotein-cholesterol (LDL-C). The Friedewald
equation is the most widely used formula to calculate LDL-C but is less accurate in
patients with comorbidities and extreme lipid values. Several novel formulae have
been reported to outperform the Friedewald formula over a wide-range of lipid levels.
Methods: This study was a retrospective evaluation of lipid profiles in 14219
patients in South Africa, from 1 January 2013 to 30 June 2013. We evaluated four
formulae (Friedewald, Chen, de Cardova, Hattori) and compared these to our direct
measurement of LDL-C, (total cholesterol) TC, (HDL-cholesterol) HDL-C using
Beckman reagents and instruments (Beckman Coulter). Linear regression and ROC
analysis were performed.
Results: Average age of the population was 52 years (39% male, 61% female,
mean LDL-C 2.9 mmol/l1.15 SD). Directly measured LDL-C highly correlated
with non-HDL (r=0.93; 95% CI 0.926-0.933). The de Cardova formula showed a
high correlation with directly measured LDL-C (r=0.90 p<0.001), comparable to
Friedewald calculated values for directly measured LDL-C (r=0.95 p<0.001). The
de Cardova formula was favorable in some ranges of HDL, TC and the lowest TG
range (r=0.97 p<0.001) but performed least well compared with three other LDL
calculations (AUC=0.8331). The Chen formula performed better than Friedewald
(AUC=0.9049). The Hattori formula outperformed all formulae including Friedewald
over various ranges of lipid values (AUC=0.9097, Figure 1).
Conclusions: We confirm that the de Cardova formula could replace directly
measured LDL-C if validated in the laboratory. In contrast to several recent findings,
we show favorable correlations of this formula with Friedewald at extreme TG values.
However, the Hattori formula appears to be the best for application in hospitalized
patients, even at extreme lipid value Figure 1 ROC analysis of Friedewald, Chen,
de Cardova(nwLDL-C), Hattori calculations

B-097
Comparison of Lipoprotein (a) methods using two commercially available
immunoturbidimetric methods on an automated chemistry analyzer.

E. L. Ryan, J. C. Ritchie. Emory Univerity School of Medicine, Atlanta, GA

Background: Lipoprotein(a) is an LDL-like particle that can vary in size and
components based upon the variable number of kringle IV repeats present in
apolipoprotein(a). Plasma Lipoprotein (a) levels are static in an individual and can
help with assessing an individuals risk of developing atherosclerotic plaques and
coronary artery disease. Previously we ran the Diasorin SPQ II immunoturbidimetric
method on a Roche Cobas Fara.
Objective: To compare two commercially available immunoturbidimetric userdefined methods for measuring Lipoprotein (a) on an automated chemistry analyzer to
the method used by our reference laboratory.
Methods: Method#1 (Kayama Biomedical Company K-Assay Lp(a)
immunoturbidimetric assay) and Method #2 (Pointe Scientific, Inc. Lp(a)
immunoturbidimetric assay) were evaluated sequentially on the same Beckman
UniCel DxC 800 Synchron chemistry analyzer. Each method was performed per
manufacturers instructions. The methods were evaluated for accuracy, precision,
linearity and patient sample comparisons were performed. The medical decision point
of 30mg/dL was used to assess patient sample results between the assays. We used 37
frozen Lithium Heparin plasma samples ranging from below the reportable range for
either assay up to above the reportable range for the current method.
Results: Both assays were linear from across the reportable range with the Method#2
showing a negative bias at values above 90mg/dL (ranging from 2.4-9 mg/dL).
Accuracy was within the allowable error limit of 10mg/dL for both methods. Assay
variation was assessed using the same material and similar between methods, %CV
were 2-5% for Method#1, and 3-12% for Method #2. Recovery of QC material was
lower in Method#2 than in Method#1. The patient samples results were similar with a
mean biases of 4.97mg/dL for Method #1 and -7.02mg/dL for Method #2. In method
#1 two samples had a bias > 10mg/dL, both were above 60mg/dL. Method #2 had
seven samples with a bias> 10mg/dL, four between 30 and 60 mg/dL and three above
60 mg/dL.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S159

Lipids/Lipoproteins

Wednesday, July 30, 9:30 am 5:00 pm

Conclusion: Both assays were comparable to the reference method. Method #1
yielded results for the QC material and samples more similar to historical values.

B-098
Lipid hydroperoxides in apolipoprotein E-containing high-density lipoprotein

S. Mamiya, A. Yoshimoto, M. Sato, R. Ohkawa, M. Tozuka. Tokyo Medical

and Dental University Analytical Laboratory Chemistry, Graduate School
of Health Care Sciences, Tokyo, Japan
Background: Higher levels of high-density lipoprotein (HDL)-cholesterol have been
associated with lower risk of coronary heart disease. It is noteworthy that the low
level of HDL-cholesterol remains predictive risk of cardiovascular disease (CVD)
even when low-density lipoprotein (LDL)-cholesterol concentration has been kept to
low level by the treatment. Some cholesteryl ester transfer protein (CETP) inhibitors
are currently undergoing clinical evaluation. In fact, CETP inhibitors increase HDLcholesterol levels; however, the mortality rate of CVD has not largely changed. It
indicates that the HDL levels do not necessarily reflect its functions. So it is important
to estimate the quality of HDL together with its quantity. It is well known that CETP
inhibitors raise apolipoprotein (apo) E-containing HDL, a minor subpopulation of
HDL. Therefore we investigated the qualitative evaluation of apoE-containing HDL,
especially antioxidant ability estimated by containing lipid hydroperoxides (LOOH)
level.
Methods: HDL (1.063<d<1.210 g/mL), isolated from serum obtained from 10 healthy
volunteers by ultracentrifugation, was separated into apoE-containing and apoEdeficient HDL by Heparin-Sepharose choromatography. The compositions of HDL
were determined by Lowrys Method (protein) and enzymatic test kits (cholesterol,
phospholipids, and triglyceride). Before and after the oxidation by CuSO4, the
concentration of LOOH was measured by the ferrous oxidation in Xylenol Orange
(FOX) assay using triphenylphosphine to get higher specificity for LOOH. Nondenaturing gel electrophoresis was performed for analyzing distributions of apoE and
particle sizes. Surface charge was characterized by lipoprotein electrophoresis using
agarose-gel.
Results: The particle size of apoE-containing HDL, developed by CBB-R250 staining
and immunoblotting using anti-apoE antibody, was larger than apoE-deficient HDL.
The relative electrophoretic mobilities were obviously small in apoE-containing HDL
on agarose gel electrophoresis pattern. Concentrations of LOOH in apoE-containing
HDL and apoE-deficient HDL were 22.3 and 1.2 nmol/mg protein, respectively. After
the oxidation by CuSO4, LOOH levels were increased to 153.8 and 294.0 nmol/mg
protein in apoE-containing HDL and apoE-deficient HDL, respectively.
Conclusion: ApoE-containing HDL which is known to increase by CETP inhibitors
is different in the particle size and surface charge from apoE-deficient HDL. ApoEcontaining HDL has extremely higher level of LOOH than apoE-deficient HDL. This
suggests that apoE-containing HDL might protect LDL from early and weak oxidation
due to its higher susceptibility to oxidation or its rapid ability to accept LOOH from
oxidized LDL. However, the total capacity of antioxidant ability per unit protein mass
could be smaller than apoE-deficient HDL, suggesting that the increase of apoEcontaining HDL did not simply reflect the increase of antioxidant ability.

B-099
LDL Subfractions Analysis in Pro-atherogenic Dyslipidemia

N. Muniz, E. Nunez. Quantimetrix Corp., Redondo Beach, CA

Background: Early recognition of pro atherogenic risk factors is important for
prevention and treatment of coronary artery disease (CAD). The NCEP ATP III
guidelines identified LDL cholesterol (LDL-C) as the primary target for CAD therapy
and risk assessment. New ACC/AHA guidelines replaced traditional lipid risk factors
with a 10 year ASCVD risk calculator weighing heavily on non-lipid risk factors,
ignoring a large body of evidence clearly recognizing specific dyslipidemic profiles
with increased CAD risk. Numerous studies clearly demonstrate that small dense
LDL, VLDL remnants and IDL are independently atherogenic while large buoyant
LDL, HDL and possibly large VLDL may not be. Exclusion of such evidence could
result in patient misclassification possibly leading to under or overtreatment of
individuals. In this study, pro atherogenic lipoprotein subfractions were measured
using the Quantimetrix Lipoprint LDL system, (Quantimetrix Corporation, Redondo
Beach, CA). The test yields critical information for early detection of individuals at
risk or with existing CAD and allowing for a more individualized implementation of
treatment.

S160

Objective: Demonstrate the benefit of measuring the atherogenic LDL subfractions

with the comprehensive analysis on the Lipoprint LDL system and assist clinicians in
identifying, stratifying and customizing treatment for those at risk.
Methods: Lipid profiles for a total of 273 recruited subjects were determined by
testing their total cholesterol, triglycerides, LDL-C and HDL-C using standard
clinical methods. Subjects were segregated into two groups, normolipedemic and
dyslipidemic based on ATP III desirable lipids status. Cholesterol levels in the
lipoproteins subfractions, large VLDL, Mid-C (VLDL remnants), Mid-B (large IDL),
Mid-A (small IDL), LDL-1 and LDL-2 (large buoyant LDL), LDL-3 to LDL-7 (small
dense LDL) and HDL were also measured in both groups using the Quantimetrix
Lipoprint LDL system, a linear polyacrylamide gel electrophoresis method. Results
from the traditional lipid profile were compared to the lipoprotein subfraction profiles
obtained by Lipoprint.
Results: The lipid test results, mean and range, for the 273 study subjects were: total
cholesterol 196 (104 - 319) mg/dL, triglycerides 96 (25 - 345) mg/dL, LDL-C 117
(58 - 215) mg/dL and HDL-C 55 (26 - 137) mg/dL. Out of the 273 total subjects,
133 (49%) were classified normolipidemic according to the ATP III lipid guidelines
while140 (51%) had at least one parameter outside the recommendations. LDL
subfractions analysis by the Lipoprint system revealed that 17 (13%) out of the
133 previously classified normal subjects had cholesterol levels outside the 95 %
confidence interval range for a given LDL subfraction. Of the 141 dyslipedemic
subjects, 69 (49%) had a normal LDL subfraction distribution. Lower levels of large
buoyant LDL-1 were observed in many of the dyslipidemic subjects.
Conclusions: Clinical studies identify small dense LDL, VLDL remnants and
IDL subfractions independently associated with increased CAD risk above other
lipid factors. Measurement of these highly atherogenic lipoprotein subfractions as
demonstrated by the Lipoprint system could be a better predictor of CAD risk than
measurement of other traditional lipid risk factors.

B-100
Cell Surface Re-Engineering by Lipid Anchoring Approach

P. Vabbilisetty, X. Sun. Cleveland State University, Cleveland, OH

Background: Many of the biological processes such as cell-cell adhesions,
extracellular/ intracellular communications occurring on the cell surface are governed
and guided by the cell surface receptors. Cell surface is a platform for introduction of
various biomolecules such as proteins, carbohydrates, etc., that may further improve
the potentiality of the cells. Lipidation of cell membranes is one such approach which
plays an important role in many biological applications such as drug/gene delivery
and serves as biomimicking models. Introduction of chemoselective functional groups
via bio-orthogonal copper-free click chemistry at the cell surface further facilitates for
many cellular modifications and enables for rapid and efficient cell surface labeling.
Lipids can be introduced on to the cell membrane in different forms such as liposomes,
micelles for efficient delivery of drug/gene or other bioactive molecules of interest.
Objective: To study and evaluate the potential anchoring effects of phospholipid
(DSPE-PEG2000-DBCO) and cholesterol (CHOL-PEG2000-DBCO) based lipids on
cell surface for cell surface re-engineering.
Methods: To investigate the lipid anchor incorporation effects on cell membranes,
different concentrations of biotin conjugated anchor lipids were prepared by reacting
N3-Biotin with anchor lipids namely DSPE-PEG2000-DBCO and CHOL-PEG2000
-DBCO via Copper free click chemistry for 1hr at RT, PBS buffer pH 7.4. The
obtained conjugated anchor lipids at varying concentrations were incubated with
raw 267.4 cells for different incubation times ranging from 5- 20 mins at 37C. The
lipid conjugated cells were further labeled with Streptavidin-FITC for 5 mins and the
effects were examined using Confocal microscopy and Flow cytometry.
Results: Confocal microscopy and flow cytometry data suggests that 5 mins of
incubation of the anchor lipid (CHOL-PEG2000 -DBCO, 5 M) with cells is
enough to see its incorporation into the cells; however a higher fluorescent intensity
signal was observed at 20 mins indicating more lipid incorporation. The confocal
microscopy data clearly depicts the intact incorporation of CHOL-PEG2000 -DBCO
-biotin conjugate into the cell membrane without any internalization. In comparison at
20 mins, for (DSPE- PEG2000 -DBCO, 5 M) there was decreased fluorescent signal
and seen from both the cell membrane and cytoplasm indicating internalization of this
conjugated lipid. This data shows the different effects of phospholipid and cholesterol
based anchor lipids on cell membrane and thus can be used for introduction or
transport of various drug/gene/carbohydrates/biomolecules for cell surface reengineering purposes.
Conclusion: CHOL-PEG2000-DBCO was shown to rapidly incorporate into the
cell membrane within 5 mins (20 mins being the optimum incubation time) and it
was visibly evident that there is no internalization of the lipid into the cytoplasm,

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Lipids/Lipoproteins

Wednesday, July 30, 9:30 am 5:00 pm

unlike the DSPE-PEG2000-DBCO. Moreover the fluorescent signal intensity for

DSPE based anchor lipid was very weak when compared to CHOL anchor lipid. This
comparative study of lipid anchoring via copper free Click Chemistry suggests that
they can serve for potential in vivo cell surface re-engineering applications.

B-101
CORRELATION BETWEEN GLYCATED HEMOGLOBIN AND SERUM
LIPIDS IN TYPE 2 DIABETICS IN EASTERN LIBYA.

N. Naseb1, A. R. Said2, S. Shakila2, S. D. Kolla3, L. T. Peela4, S. J.

Dhoipode2, J. R. Peela2. 1Department of Biochemistry, Faculty of
Medicine,University of Benghazi, Benghazi, Libyan Arab Jamahiriya,
2
Faculty of Medicine,Quest International University Perak, Ipoh, Malaysia,
3
Department of Biochemistry,Rangaraya Medical College,, Kakinada,
India, 4Great Eastern Medical School, Srikakulam, India
Background: Diabetes mellitus with its accompanying vascular complications is on
a rise globally and a similar trend has been observed in Libya too. Early detection,
effective monitoring and timely management is important to control this growing
problem. Various studies have shown dyslipidemia as a significant risk factor of
atherosclerosis leading to vascular complications in diabetes mellitus. HbA1c as a
marker of long term glycemic control in diabetics is an established fact. The present
study is aimed at correlating HbA1c with lipid parameters in blood to understand its
role as a marker of dyslipidemia.
Materials and methods: Sixty subjects in the age group ranging from 40 to 70 years
have been recruited from Seventeenth February Teaching Hospital, Al- Baida for the
study, twenty controls with no history of diabetes, twenty recently diagnosed diabetics
under treatment and twenty old cases of diabetes mellitus who have suffered an
episode of coronary artery disease (CAD) or cerebrovascular accident (CVA). Venous
samples were drawn after an overnight fast for glucose, glycated hemoglobin, total
cholesterol, triacylglycerol and HDL cholesterol and these tests were performed using
authenticated kits and Cobas integra 400 analyzer. LDL cholesterol was calculated
using Friedwalds formula.
Results: Slightly high levels of fasting blood glucose (p=0.03), Glycated hemoglobin
(p=0.02), high levels of triglycerides (p<0.001) and low levels of HDL cholesterol
(p<0.001) were observed in diabetics under treatment when compared with controls,
but total cholesterol and LDL cholesterol showed no difference. However, diabetics
with complications showed higher levels of fasting blood glucose (p<0.0001),
glycated hemoglobin (p<0.0001), total cholesterol (p<0.001), triglycerides (p<0.001),
low density lipoprotein cholesterol (p<0.001), and low HDL cholesterol (p<0.0001).
There is positive correlation between glycated hemoglobin and serum triglycerides
and inverse correlation with HDL cholesterol in both the diabetic groups, though not
strong in recently diagnosed diabetic patients under treatment. A strong correlation
was observed between glycated hemoglobin and total cholesterol and LDL cholesterol
only in the diabetic group with coronary artery and cerebrovascular complications.
Conclusion: The present study has shown a significant correlation between glycated
hemoglobin and serum total cholesterol and LDL cholesterol; the dyslipidemia
risk factors causing atherosclerosis in diabetic mellitus not under control. Hence
measurement of glycated hemoglobin is significant as a dual marker not only to
monitor long term glycemic control but also in predicting dyslipidemia in type 2
diabetes mellitus.

B-102
CARDIOVASCULAR RSK FACTORS N PATIENTS WITH
HEMODIALYSIS:PARAOXONASE AND HYPERHOMOCYSTEINEMIA

M. DEMIR1, K. KSE2, C. YAZICI2, A. ZAHN2, B. TOKGZ2, C.

UTA2. 1MEDCALPARK HOSPTAL, ANTALYA, Turkey, 2ERCYES
NVERSTY MEDCAL FAKULTY, KAYSER, Turkey
Increased risk of cardiovascular disease (CVD) has been recognized as an important
cause of morbidity and mortality in chronic renal failure (CRF). Hyperhomocysteinemia
has also been accepted as an independent risk factor for CVD. Paraoxonase (PON1)
is an enzyme with antioxidant activity, which circulates in plasma attached to HDL.
The aim of this study was to investigate the risk of CVD in CRF by depending on
lipid/lipoprotein profiles, homocysteine and PON1 activity in plasma of hemodialyzed
patients.
MATERIAL and METHODS
Subjects: Total of 42 patients undergoing hemodialysis (HD), and 43 healthy
volunteers were included the study. The clinical data of patients and the controls are
summarized in Table 1.

Table 1:Clinical features of the study groups

Control
Hemodialysis
Number of participants
43
42
Sex (M/F)
21/22
21/21
Age (years)
24-61
16-64
Mean age(months)
34,710,7
36,913,3
Dialysis duration (months)
8-108
Mean dialysis duration (months) 36,622,2
Methods: Fasting blood samples obtained from study groups were drawn into
anticoagulant-free tubes and centrifuged at 2,000 g for 10 min. Serum samples were
used for the measurements of triglyceride (TG), total cholesterol (TC), LDL-C,
HDL-C, and homocysteine levels,arylesterase activity and also paraoxonase activity.
RESULTS
There was no significant difference in age and sex distribution between the study
groups.When compared to controls increased plasma TG but decreased TC,HDL-c
and LDL-C levels were observed HD patients, even though all lipid/lipoprotein levels
were in normal laboratory range.Significantly decreased PON1 and arylesterase
activities, but increased homocysteine values were found in HD patients than those
of control
DISCUSSION Patients with CRF, especially HD patients, experience excessively
high cardiovascular morbidity and mortality.increased TG, but decreased TC, HDL-C
and LDL-C levels , and also decreased PON1 and arylesterase activities observed
in HD patients in the present study were in agreement with many previous reports.
That increased homocysteine levels in the present study have supported the previous
studies , and homocysteine may be suggested as an independent risk factor for
CVD in HD patients.n conclusion, as reflected by decreased PON1 and increased
homocysteine levels, CVD risk might be increased in HD pat ents.

B-103
The usefulness of non-HDL cholesterol in less resourced countries

P. Nsiah1, E. F. Laing2, B. A. Eghan2. 1University of Cape Coast., Cape

Coast, Ghana, 2Kwame Nkrumah University of Science and Technology,
Kumasi, Ghana
Background: Common characteristic features of diabetic dyslipidemia are the
elevation of plasma triglycerides and triglyceride-rich VLDL cholesterol, reduced
HDL cholesterol, and an increased number of small dense LDL cholesterol particles
(1). Although LDL cholesterol is not typically elevated in patients with diabetes, the
changes in LDL cholesterol composition that can accompany the disease make the
LDL cholesterol exceptionally atherogenic (2,3).
Aim: This study primarily aims to determine and compare the power and influence
of non-HDL cholesterol and LDL cholesterol, in predicting coronary heart disease
among diabetic versus non-diabetic adult Ghanaians.
Methods: A cross-sectional study was performed on 302 subjects who consisted
of 154 previously diagnosed diabetes patients and 148 non- diabetics. BMI, WC,
blood pressure (BP), total cholesterol, high-density lipoprotein cholesterol (HDL),
low-density lipoprotein cholesterol (LDL), triglycerides (TG), fasting glucose, highsensitivity C-reactive protein (hs-CRP), adiponectin and resistin were measured.
Results: The mean age was 52.2 (8.7). There was a higher (negative) correlation
between adiponectin and non-HDL cholesterol (r=-0.5756; p<0.0001) than adiponectin
and LDL cholesterol (r=-0.5152; p<0.0001). Higher positive correlation was observed
with resistin and non-HDL cholesterol (r=0.5756; p<0.0001) than resistin and LDL
cholesterol (r=0.5494; p<0.0001). Similar patterns of correlation were observed with
hs-CRP, blood pressure and 10 year cardiovascular disease risk.
Conclusion: This study contributes to the existing body of literature by suggesting
that easily calculated non-HDL cholesterol is superior to LDL cholesterol in
cardiovascular disease risk assessment. The use of non-HDL would also be more
practical and reliable target for lipid lowering therapy in less resourced country like
Ghana
References: 1. Sniderman AD, Scantlebury T, Cianflone K: Hypertriglyceridemic
hyperapoB: the unappreciated atherogenic dyslipoproteinemia in type 2 diabetes
mellitus. Ann Intern Med 135:447-459, 2001
2. Turner RC, Millns H, Neil HAW, Stratton IM, Manley SE, Matthews DR, Holman
RR, for the United Kingdom Prospective Diabetes Study Group: Risk factors for
coronary artery disease in non-insulin dependent diabetes mellitus (UKPDS 23). BMJ
316:823-828, 1998
3. Lu W, Resnick HE, Jablonski KA, Jones KL, Jain AK, Howard WJ, Robbins DC,
Howard BV: Non-HDL cholesterol as a predictor of cardiovascular disease in type 2
diabetes: the Strong Heart Study. Diabetes Care 26:16-23, 2003

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S161

Lipids/Lipoproteins

Wednesday, July 30, 9:30 am 5:00 pm

B-104
Comparison of the Vantera NMR Analyzer Tests for Lipid and Lipoproteins
to Conventional Enzymatic Assays on the Siemens Dimension Vista Chemistry
Analyzer

and lower HDL2-C/HDL3-C ratios (p < 0.05) than younger group. Consequently,
only sex difference was found in HDL2-C, whereas both sex and age differences were
identified in HDL3-C.
Conclusion: Our analysis suggests that HDL subclasses can identify the CHD risk
more accurately than total HDL-C.

M. L. Sampson, V. Gounden, R. Shamburek, A. T. Remaley. National

Institutes of Health, Bethesda, MD
Objective: Traditional lipid cardiovascular biomarkers are largely based on the
cholesterol content of the major lipoprotein fractions. Lipoproteins, however, have
a polydisperse size distribution and their association with cardiovascular disease has
been shown to vary not only based on their cholesterol content but also on their size
and particle count. The Vantera analyzer uses NMR technology to quantify not only
lipid concentrations but also size and particle counts of lipoproteins. In this study, we
evaluated several lipid and lipoprotein assays on the Vantera and compared the results
to conventional enzymatic assays for lipids and lipoproteins, as measured on the
Siemens Vista. Methods: The Vantera has 3 FDA approved tests: Triglycerides (TG),
HDLc and LDL particle number (LDLp). We measured these tests, as well as several
other lipid and lipoprotein test parameters, in 450 patients with a wide variety of lipid
disorders by the Vantera and Vista. LDLc was calculated by the Friedewald equation,
using total cholesterol (TC) from either the Vantera or the Vista as indicated. Results:
The results from the two assay systems show good correspondence (refer to Table
below). Similar calculated LDLc results were also obtained, using either Vantera TG
and HDLc plus Vista TC (LDLc1) or all Vantera parameters, including TC (LDLc2),
when compared to LDLc calculated with Vista parameters. In addition, we compared
the measured LDLc on Vantera (mLDLc) with the calculated LDLc from Vista.
TC

TG

HDLc

LDLc1

LDLc2

mLDLc

Deming
0.96
0.89
0.88
1.13
1.09
1.17
Slope
Intercept 2.71
1.69
6.42
-0.23
3.30
3.59
R2
0.93
0.92
0.93
0.93
0.85
0.87
We also validated the lipoprotein particle count and size measurements for the major
lipoprotein classes by showing that the calculated lipoprotein core volumes matched
the measured core lipids, namely cholesteryl esters and TG (R2 = 0.91). Conclusions:
The Vantera NMR analyzer generates comparable lipid and lipoprotein results to
traditional methods, while at the same time yielding additional cardiovascular risk
information related to lipoprotein particle count and size.

B-105
Establishment of Reference Range for HDL Subfractions in Japanese
Population with a New Automated Assay

Y. Ito, N. Satoh, T. Ishii, J. Kumakura. Denka Seiken Co.,Ltd., Tokyo, Japan

Background: Measurement of high-density lipoprotein (HDL) 2 and HDL3
subfractions might be more useful for evaluating coronary risk than total HDLcholesterol (C). However, methods of measuring HDL2 and HDL3 are quite laborious
for general clinical use and thus so far it has been difficult to establish reference
ranges. Recently, we have succeeded in establishing a fully automated homogeneous
assay for HDL3-C. We carried out a study to establish a reference range for HDL
subfractions in Japanese population using our novel homogeneous assay for HDL
subfractions.
Methods: HDL-C and HDL3-C were measured by our homogeneous assays on a
Hitachi 917 automated clinical chemistry analyzer (Hitachi). HDL2-C was calculated
as the difference between total HDL-C and HDL3-C. Subjects were recruited in
Japan. 670 volunteers who did not have a history of CAD/CHD were invited to
participate. Subjects were partitioned according to the following four parameters:
male, female, young (male: equal or younger than 45 y, female: equal or younger than
55 y) and old (male: older than 45 y, female: older than 55 y). Data from the stratified
subjects were analyzed for their distribution characteristics with the Shapiro-Wilk
test, and estimated values were computerized. Based on distribution characteristics
and independency, an appropriate test was chosen to compare the distribution of the
variables among sub-groups.
Results: The Shapiro-Wilk test revealed that all sub-groups were nonparametric.
Therefore, the Wilcoxon rank-sum test was used to compare the distribution of
variables among sub-groups for HDL3-C, HDL2-C, total HDL-C, and HDL2-C/
HDL3-C ratio. Females had significantly higher HDL3-C (p < 0.01), HDL2-C (p <
0.001), total HDL-C (p < 0.001), and HDL2-C/HDL3-C ratios (p< 0.001) compared
to males. Regarding age, HDL2-C (p = 0.72) and total HDL-C (p = 0.62) did not show
any significant differences, whereas older group showed higher HDL3-C (p < 0.001)

S162

B-106
Revising the Lipid Profile (LP): Evaluation of the Apo-A1 and Apo-B Assays on
the Vitros-5600 [V-5600] Analyzer

R. Pillappa1, C. Hoang2, D. Moore2, E. S. Pearlman2. 1University of

Tennessee Health Sciences Center, Memphis, TN, 2Veterans Affairs Medical
Center, Memphis, TN
Backgroundt: In view of recent literature we are considering revision of the VAMC
LP to consist only of apo-B, apo A1 and a calculated ratio, The initial step was to
assess the properties of the apo-A1 and apo-B assays on our V-5600 analyzers (OCD:
Parsippany, NJ).
Methods: On day 1 apo-A1 and apo-B were assayed once on each of 3 V-5600
analyzers in 36 specimens on which an LP had been requested. On the same day
aliquots were sent to a referral laboratory [RL] (ARUP; Salt Lake City, UT) for assay
of apo-A1 and apo-B using immuno-nephelometry and after 24 hour refrigerated
storage the apo-A1 and apo-B assays were re-run on one of our V-5600 analyzers. All
in-house assays were done according to manufacturers instructions.
Results: The TC and TG concentration ranges were [99, 298 mg/dL] and [56, 746 mg/
dL.] respectively. Data suggested that median within machine, between day CVs for
apo-A1 (1%) and apo-B (0.65%) were significantly less than between machine, within
day CVs [2.17 and 1.36%; p<.002 for both analytes, Mann-Whitney test]. Betweenmachine within day CVs were an increasing function of concentration for apo-A1
and fit by the equation: CV (%) = 0.78 + 0.000089 [apo-A1]2 while for apo-B the
corresponding CV has a minimum at a concentration of 100 mg/dL. Assuming RL
apo-A1 and apo-B assays to be done the following day we compared the second day
V-5600 assays (Y) to the RL assay (X) using Deming regression with results: Y =
24.42 (14, 34.8) + 0.716 (0.64, 0.79) X and Y = 7.38 (1.78, 13) + 0.994 (0.935, 1.05) X
for apo-A1 and apo-B respectively. With respect to classification into accepted risk
categories [<80, 80-119 and 120 mg/dL.] the V-5600 and RL apo-B assays place 32/36
(89%) patients into the same risk category and the remaining 4 are one category apart.
Conclusions: The current data indicates acceptable precision parameters for the
V-5600 apo-A1 and apo-B assays. Significant proportional and constant bias in the
V-5600 apo-A1 assay relative to the RL assay indicates that results for this assay
remain method dependent. This appears less problematic for the apo-B assay where
only constant bias is noted. V-5600 and RL Apo-B assays are in satisfactory agreement
with respect to risk classification..

B-108
Determination of reference intervals for LDL and HDL cholesterol subclasses
and their clinical relevance in acute coronary syndrome

H. Ookura, H. Takahashi, R. Suzuki, Y. Kadosaka, M. Yoshika. Kansai

Medical University, Hirakata, Osaka, Japan
Background:To evaluate the performance of a newly developed assay system
in quantifying small-dense LDL cholesterol (sdLDL-C) and HDL3 cholesterol
(HDL3-C), we collected blood from 2041 control subjects during an annual health
check-up conducted at the Kansai Medical University.
Methods:Exclusion criteria included hypertension (systolic blood pressure 160
mmHg or diastolic blood pressure 100 mmHg), a high body mass index (BMI 30
kg/m2), or a decrease in the estimated glomerular filtration rate (60 ml/min), HbA1c
(6.6%), LDL (40-160 mg/dL), AST ( 60 U/L), ALT ( 70 U/L), or triglyceride (TG)
(>250 mg/dL) levels. As a result, samples from 1412 subjects (average age 34 10
years; 293 men and 1119 women) were used for further analysis. Twenty samples
collected from acute coronary syndrome (ACS) patients during their first visit to our
emergency department were used to examine the clinical utility of the novel assay.
LDL-C, HDL-C, sdLDL-C, and HDL3-C were simultaneously measured using
a homogenous assay system (Denka-Seiken Co., Ltd.). Large buoyant (lb) LDL-C
levels were calculated by subtracting sdLDL-C values from LDL-C values. Similarly,
HDL2-C levels were calculated by subtracting HDL3-C from HDL-C. Samples
were centrifuged within an hour of collection, and the serum was stored at 4C until
measurement, which was done within 8 hours of collection. Reference intervals were

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Lipids/Lipoproteins

Wednesday, July 30, 9:30 am 5:00 pm

determined for 4 groups (based on gender and age). This was determined using the
normalized data generated from the Box-Cox power transformation model, and the
95% confidence interval was calculated using a parametric method.

of atherosclerotic diseases. This might be a reason for LAB activity reflects the
progression and the risk of atherosclerotic disease better than oxLDL concentration
determined by anti-oxLDL antibody.

Results:Coefficient of variance values for the intra- and inter-assay variation in

LDL-C, HDL-C, sdLDL-C and HDL3-C were less than 6.13%. LDL-C, sdLDL-C,
lbLDL-C, HDL-C, HDL2-C, and HDL3-C values positively correlated with age in
both men and women. LDL-C and lbLDL-C in men, and LDL-C, sdLDL-C, and lb
LDL-C in women positively correlated with BMI, but HDL-C and HDL2-C in both
groups negatively correlated with this parameter. sdLDL-C and HDL3-C in men,
and HDL-C, HDL2-C, and HDL3-C in women positively correlated with alcohol
consumption, while this negatively correlated with lbLDL-C in men.

Hypertriglyceridemia is a major contributor of high small dense LDL level in

patient with metabolic syndrome

The sdLDL-C reference interval was the lowest in younger women (13-22 mg/dL)
and the highest in older men (16-53 mg/dL). The lb-LDL-C interval was the highest
in older women (49-119 mg/dL), and the lowest in younger women (43-111 mg/dL).
Reference intervals for HDL2-C and HDL3-C were the highest in older women (2977 mg/dL and 19-36 mg/dL, respectively), and the lowest in young men (22-66 mg/
dL and 17-30 mg/dL).

Background: Patients with metabolic syndrome (MetS) have shown higher small
dense low density lipoprotein cholesterol (sdLDL-C) level than healthy controls.
However, which component of MetS made the largest contribution to an increase
in sdLDL-C has not fully determined. We aimed to determine major contributing
component of MetS to high sdLDL-C concentration and sdLDL-C/LDL-C ratio.

In ACS patients, both sdLDL-C and lb-LDL-C values were higher while HDL2-C and
HDL3-C values were significantly lower than those in the control group. In relation
to ACS, a multiple logistic regression analysis revealed significant odds ratios in age
(1.3), lbLDL-C/sdLDL ratio (26.4) and sdLDL-C/HDL2-C ratio (269.7).
Conclusion:These findings suggest that the LDL-C and HDL-C subclasses may
be useful biomarkers in predicting cardiovascular complications in patients with
atherosclerosis.

B-109
Overexpression Of Del-1, An Oxidized LDL Blocking Protein, Suppressed
Atherogenesis In Mice Without Lowering Oxidized LDL Concentration, But
with Reducing LOX-1 Ligand Containing ApoB Activity (LAB).

A. Kakino, A. Nakano, Y. Fujita, T. Sawamura. National Cerebral and

Cardiovascular Center, Suita, Japan
Background: Oxidized LDL (oxLDL) is implicated in the pathogenesis of
atherosclerosis. However, measurement of circulating oxLDL concentration often
fail to provide the accurate state of atherosclerosis or the risk of atherothrombotic
diseases such as myocardial infarction and ischemic stroke. To solve this problem,
we have devised a novel ELISA assay system based on the binding of modified LDL
to an oxidized LDL receptor LOX-1, rather than determining the concentration of a
specific epitope of anti-oxLDL antibody. LOX-1 binding activity of apoB-containing
lipoprotein was designated as LAB. LAB well predicted the risk of coronary artery
disease and ischemic stroke (Inoue, Clin Chem 2010), and reflected the intimal
thickening of carotid artery (Okamura, Atherosclerosis 2013).
Aim: To understand the reason why the receptor-based assay has been superior to the
antibody-based assay in evaluating the progression of atherosclerosis and the risk of
atherosclerosis-related diseases.
Methods and Results: We found that Del-1 selectively bound to oxLDL but not
to native LDL, leading to the inhibition of the uptake of DiI-labeled oxLDL (DiIoxLDL) via oxLDL receptors including LOX-1, SR-A, CD36, and SR-B; but not
to the inhibition of DiI-labeled native LDL uptake via LDL receptor expressed in
COS-7 cells. We also found that Del-1 inhibited DiI-oxLDL uptake by cultured
human umbilical vein endothelial cells (HUVEC) and THP-1-deribed macrophages.
Furthermore, Del-1 suppressed oxLDL-dependent signal transduction in LOX1 expressing CHO cells and in HUVEC. Del-1 also suppressed oxLDL-induced
secretion of endothelin-1 in HUVEC.
To examine in vivo effects of Del-1 on atherogenesis, we established Del-1 transgenic
mice (Del-1Tg), and fed their males high-fat diet along with control wild-type
mice (WT) (n=6 each) for 20 weeks from the age of 24 weeks. Oil red O-positive
atheromatous area at aortic roots dramatically decreased in Del-l Tg compared
with WT (3.11.4 vs. 17.72.0 % of aortic roots area, P<0.001). Reflecting the
antiatherogenic effects, plasma LAB activity was significantly decreased in Del-1Tg
compared with WT mice (13.34.3 vs. 106.220.1 ng/ml, P<0.05), while oxidized
LDL concentration determined by conventional antibody-based assay did not
differ between Del-1Tg and WT (698.934.4 vs. 741.946.0 nmol/ml). Other lipid
parameters except triglycerides were not different. Plasma tryglyceride concentration
in Del-1Tg was slightly lower than that in WT.
Conclusion: In the presence of Del-1, an oxLDL blocking protein, the oxLDL
concentration determined by conventional anti-oxLDL antibody-based assay
dissociates from atherogenicity, while LAB well associated with it. This might
be a reason why circulating LAB activity better reflects the state and the risk

B-110

S. Lee, Y. Cho, J. Kim. Yonsei University College of Medicine, Seoul,

Korea, Republic of

Methods: Four hundred and forty seven subjects (225 men; 222 women) with MetS
were randomly selected from the Korean Metabolic Syndrome Research InitiativesSeoul cohort study. Age and sex-matched 360 healthy controls (181 males and 179
females) were also randomly selected from the same cohort.
Results: When we compared means of sdLDL-C concentration between subgroups
divided according to whether subjects met or not met each MetS component in
patients with MetS (Table 1), significant difference in sdLDL-C concentration was
found only between subgroups divided according to whether subjects met or not
met triglyceride (TG) criteria. For healthy control, there were significant differences
in sdLDL-C concentration according to the presence or absence of TG and waist
circumference components. We observed similar pattern for sdLDL-C/LDL-C ratio.
Pearson correlation analysis showed total cholesterol, LDL-C, and TG showed
relatively strong correlations with sdLDL-C concentration (r = 0.730, 0.508 and 0.543
respectively for men; 0.748, 0.692 and 0.653 respectively for women), whereas only
TG maintained a strong correlation with sdLDL-C/LDL-C ratio (r = 0.789 for men
and 0.745 for women). In multiple regression analysis, we found TG level was a
significant determinant of sdLDL-C concentration and sdLDL-C/LDL-C ratio.
Conclusion: Among five MetS components, only the abnormal TG level worked
as a differing factor of sdLDL-C concentration and sdLDL-C/LDL-C ratio, and
which results were reproducible in both genders with or without MetS. Our results
also supported a hypothesis that atherogenic effect of hypertriglyceridemia could be
partially mediated by elevated sdLDL related to high TG.
Table 1. Differences of sdLDL-C between subgroups divided by whether met or not met each component

Components
Patients with
metabolic syndrome
Waist circumference
Triglyceride
High density
lipoprotein
cholesterol
Blood pressure
Fasting blood sugar
Healthy controls
Waist circumference
Triglyceride
High density
lipoprotein
cholesterol
Blood pressure
Fasting blood sugar

Male (n = 406)
Component - Component + p value

Female (n = 401)
Component - Component +

p value

48.66 18.9
30.44 11.22

48.77 18.93 0.9691 43.08 12.55

51.78 18.18 < 0.0001 33.06 11.93

44.88 18.06
47.72 17.3

0.5891
< 0.0001

49.95 20.26

47.69 17.6

0.3709

46.75 20.86

43.71 15.64

0.2344

50.44 19.09
49.78 17.65

48.39 18.87 0.5400

46.95 20.85 0.2796

46.76 16.3
44.34 16.49

43.77 17.74
45.05 18.72

0.2473
0.7673

35.55 15.66
29.7 12.01

42.72 18.33 0.0463 26.81 12.58

48.23 15.72 < 0.0001 24.54 9.63

33.24 17.17
47.44 16.96

0.0147
< 0.0001

35.71 16.46

41.13 13.39 0.1195

27.21 13.33

31.35 15.05

0.1137

36.09 15.96
36.53 16.14

38.65 17.36 0.4559

32.17 19.11 0.6437

28.1 13.76
28.00 13.72

26.86 13.72
-

0.7384
-

B-111
Comparison of Lipoprotein Profile Analysis by Nuclear Magnetic Resonance
(NMR) and Agarose Gel Electrophoresis.

V. Gounden, M. L. Sampson, R. Shamburek, A. T. Remaley. National

Institutes of Health, Bethesda, MD
Objective: Lipoprotein analysis by agarose gel electrophoresis is useful for identifying
rare genetic lipid disorders and confirming the level of the major lipoprotein fractions.
The NMR Vantera method can also be used for quantifying lipoprotein subfractions,
size and particle counts but the two methods have not been compared. Methods:
Serum samples from 250 patients with a wide variety of lipid disorders were analyzed
on the Sebia lipoprotein gel electrophoresis system and the Vantera NMR system.
The major lipoprotein fractions (alpha-HDL, pre-beta-VLDL and beta-LDL) were
quantified by densitometric scanning of the electrophoresis gel; these values were
divided into tertiles and compared to the NMR particle count (HDL-P, VLDL-P and
LDL-P). The speed of migration of each fraction on electrophoresis was characterized

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S163

Lipids/Lipoproteins

Wednesday, July 30, 9:30 am 5:00 pm

and catergorized into tertiles (fast, medium, slow), then compared to lipoprotein
particle size (HDL-Z, VLDL-Z and LDL-Z). Data was analyzed by ANOVA to
identify differences between the methods.
Results: Results from the particle count by NMR analysis showed the same trend as
did quantification of the major bands by lipoprotein electrophoresis.
Electrophoresis Grouping Tertiles
NMR particle count (mean)
low
medium
high
HDLp (mol/L)
p<0.001 27.80.9
33.10.9
36.20.9
VLDLp (nmol/L)
p<0.001 26.94.4
36.44.4
84.84.4
LDLp (nmol/L)
p<0.001 105662.1
115460.7
149062.6
Although ANOVA showed a significant correlation (p<0.006) between size of
lipoproteins with migration speed by electrophoresis, the results were not always
consistent between the three fractions, particularly for VLDL. Using a novel method
for displaying NMR results, we also show that the pattern of NMR results typically
overlap with the interpretation of the lipoprotein phenotype based on electrophoresis
for disorders, such as Type I, IIa, III hyperlipidemia and LCAT Deficiency.
Conclusions: Because the analysis by NMR and electrophoresis depend on different
physical properties of lipoproteins, there was not always a complete concordance
of the results. However, in general NMR analysis of lipoproteins usually leads to
the same clinical interpretation of the lipoprotein phenotype as electrophoresis
and provides additional information that may be potentially useful for assessing
cardiovascular risk.

B-112
A serum oxidized high-density lipoprotein marker and its association with the
smoking status in males

K. Kotani1, S. Mashiba2, M. Ueda2, N. Taniguchi1, T. Yamada1. 1Jichi

Medical University, Shimotsuke-City, Tochigi, Japan, 2Ikagaku Co. Ltd.,
Kyoto, Japan
Background: High-density lipoprotein (HDL) particle, whose major protein is
apolipoprotein A-I (apoA-I), protects against atherosclerosis via its anti-oxidant
properties. The oxidative modification of apoA-I is associated with dysfunctional
HDL. Cigarette smoking, a major atherosclerotic risk factor and a precipitating
condition leading to oxidative stress, is often accompanied by low HDL-cholesterol
levels in the circulation. However, the adverse effects of smoking on atherosclerosis
remain incompletely understood with regard to dysfunctional HDL, and easy
biomarkers for smoking-related HDL modifications are needed. A new assay we
developed for oxidized apoA-I, oxHDL, may be a suitable marker, since we have
found high oxHDL levels under some oxidative stress conditions. The aim of this
study was to investigate the association between the oxHDL levels and the smoking
status in males.
Methods: A total of 260 Japanese males (mean age, 61 years) were consecutively
recruited from general health check-ups. The subjects who had a history of
cardiovascular disease, had been diagnosed with metabolic syndrome or received
lipid-modulating drugs were excluded. The smoking status was self-reported. Clinical
data, including serum lipid levels, were obtained from subjects in a fasted state. The
serum oxHDL levels were quantified using a sandwich ELISA system, which utilizes
monoclonal antibodies prepared by immunization with H2O2-oxidized human apoA-I.
Results: The mean/median levels of the relevant variables were as follows: lowdensity lipoprotein cholesterol, 3.0 mmol/L; triglycerides, 1.0 mmol/L; HDLcholesterol, 1.5 mmol/L; oxHDL, 221 U/mL and oxHDL/HDL-cholesterol ratio,
3.9. Compared to non-smokers (n = 188), current smokers (n = 71) tended to exhibit
higher oxHDL levels (217 versus 239 U/mL) and lower HDL-cholesterol levels (1.6
versus 1.5 mmol/L), while current smokers showed significantly higher oxHDL/
HDL-cholesterol levels (3.9 versus 4.1, p < 0.05). The difference in the oxHDL/HDLcholesterol levels remain significant after adjusting for age, body mass index, blood
pressure, other lipids and glucose levels. Moreover, a significant inverse correlation
(r = 0.3, p < 0.05) was found between the oxHDL/HDL-cholesterol levels and the
Brinkman index (number of cigarettes smoked per day number of years of the habit)
in current smokers.
Conclusion: The present findings suggest that smoking may independently and
oxidatively modify HDL particles, thus leading to dysfunctional HDL in males.
The oxHDL/HDL-cholesterol ratio may therefore be useful for assessing the
atherosclerotic burden in relation to the smoking status.

B-113
Effects of body weight on low-density lipoprotein and high-density lipoprotein
subclasses assessed by homogenous small-dense low-density lipoprotein and
high-density lipoprotein 3 cholesterol assays

H. Takahashi, H. Ookura, R. Suzuki, Y. Kadosaka, M. Yoshika. Kansai

Medical University, Hirakata, Osaka, Japan
Background: The incidence of overweight has been increasing in epidemic
proportions, and it adversely increases the prevalence of most cardiovascular
(CV) diseases. Obesity results in diverse changes in laboratory parameters such as
triglyceride (TG) levels, high- and low-density lipoprotein cholesterol (HDL-C and
LDL-C, respectively) levels, and glucose metabolism due to insulin resistance. Of
these, HDL-C and LDL-C levels play a key role in the development of CV diseases;
as such, they are very valuable biomarkers for the prediction of future CV events.
Both lipoproteins have subclasses: small dense LDL (sdLDL) and large buoyant
LDL (lbLDL) and HDL2 and HDL3, which may be superior biomarkers to LDL and
HDL, respectively. However, assaying these subclasses using ultracentrifugation,
electrophoresis, or high-performance liquid chromatography is not easy, requires
significant time, and commonly produces inaccurate, quantitative results.
Methods: A homogenous assay system for sdLDL-C and HDL3-C has recently
become available, and it is reported to have favorable performance (Clinical Chemistry
57:57-65, 2011; Clin Chim Acta. 427:86-93, 2014). In the present study, using these
homogenous assay systems, we explored the relationship between body weight and
LDL and HDL subclass concentrations. The levels of lbLDL-C and HDL2-C were
calculated by subtracting those of sdLDL-C and HDL3-C from the levels of LDL-C
and HDL-C, respectively. Data are expressed as means standard deviation.
The enrolled subjects were women aged 34 10.5 years (20-64 years) (n = 1,276)
working in our hospital, and informed consent was obtained at the annual health
checkup during blood sampling.
Results: Body mass index (BMI) was significantly correlated with both lbLDL-C (r =
0.2691, p < 0.01) and sdLDL-C (r = 0.3223, p < 0.01). Similarly, waist circumference
(cm) was significantly correlated with lbLDL-C (r = 0.2638) and sdLDL-C (r =
0.3315). In contrast, BMI was negatively correlated with HDL2-C (r = -0.2715,
p < 0.01) but not HDL3-C (r = -0.0173). Waist circumference was significantly
correlated with HDL2-C (r = -0.2771) but not HDL3-C (r = -0.0686). Multiple
regression analysis with age, systolic blood pressure, alanine aminotransferase, TG,
cystatin-C, C-reactive protein, and HDL-C and LDL-C subclasses as independent
variables revealed that these parameters were independently correlated with BMI.
The lbLDL-C and sdLDL-C values were similarly independently correlated with
BMI. Both HDL2 and HDL3 were independently correlated with BMI, but the t-value
of HDL2 was much greater than that of HDL3. When waist circumference was
used as the dependent variable, it was also independently correlated with the above
parameters. The t-value of HDL2-C (8.231) was the highest, followed by those of
lbLDL-C (4.549), sdLDL-C (2.931), and HDL3-C (2.361).
Conclusion: The present findings are in contrast to those of earlier studies in which
sdLDL-C increased and HDL3-C decreased with weight gain and changes in both
lbLDL-C and HDL2-C were more closely related to body weight differences. These
findings also indicate that these LDL-C and HDL-C subclasses assessed using the
homogenous method may be novel predictive markers for atherosclerosis.

B-114
Effect of SAA on the structure and measurement method of HDL

M. Sato1, R. Ohkawa1, T. Kameda1, A. Yoshimoto1, S. Okubo2, Y. Yatomi2,

M. Tozuka1. 1Analytical Laboratory Chemistry, Graduate School of Health
Care Sciences, Tokyo Medical and Dental University, Tokyo, Japan,
2
Department of Clinical Laboratory, The University of Tokyo Hospital,
Tokyo, Japan
Background: Serum amyloid A (SAA), which is one of the acute phase proteins, is
commonly found in high-density lipoprotein (HDL) in the circulation. SAA is known
to become a major HDL protein component in the acute phase due to a displacement
of apolipoprotein AI (apoAI). Consequently, this remodeling could affect HDL
metabolism, however the actual influences have not been fully elucidated. In this
study, we focused on the structural differences between SAA containing HDL and
normal HDL. In addition, the effect of attached SAA on the values of HDL-cholesterol
(HDL-C) measurement was estimated.
Methods: HDL (d=1.063-1.210 g/mL) isolated from the patients with or without
inflammation was characterized by agarose gel electrophoresis for analyzing the
surface charge and by nondenaturing gel electrophoresis for analyzing particle size

S164

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Lipids/Lipoproteins

Wednesday, July 30, 9:30 am 5:00 pm

and distribution of apoAI and SAA. HDL-C concentrations of the patients with
various SAA levels were analyzed by two methods. One is the homogeneous method
using -cyclodextrin sulfate which is commonly used in clinical laboratories, and the
other is the ultracentrifugation method as a reference method. SAA was measured by
latex agglutination-turbidimetric immunoassay method.
Results: The increase of serum SAA levels induced the decrease of HDL mobility on
agarose gel electrophoresis patterns. In the nondenaturing gel electrophoresis, HDLs
obtained from the patients with low serum SAA levels were separated to two distinct
particles, HDL2 and HDL3. On the other hand, HDLs obtained from the patients with
high serum SAA levels indicated two kinds of typical patterns; one was characterized
as the additional band at the intermediate particle size between HDL2 and HDL3,
and the other was characterized as two bands extremely larger size than HDL2 and
smaller size than HDL3. SAA was identified in the additional band for the former and
in the larger band for the latter. HDL-C concentrations measured by the homogeneous
method were highly correlated with those by the ultracentrifugation method in both
the patients with low (SAA8 g/mL, n=94) and high (8<SAA4762 g/mL, n=154)
SAA levels. Although no significant difference was observed in the regression lines of
both groups, the ratios of HDL-C concentrations obtained by the ultracentrifugation
method to those by the homogeneous method showed a tendency to be higher in the
patients with acute inflammation.
Conclusion: Our data indicated that the large amount of SAA attached to HDL during
inflammation and changed in the surface charge and the particle size of HDL. However,
no definite relevance between serum SAA level and HDL particle size was observed.
A good correlation between the homogeneous method and the ultracentrifugation
method could be explained by the assay principle of the homogeneous method used,
in which total cholesterol is measured after inhibition of enzymatic reaction against
lipoproteins (mainly VLDL and LDL) except HDL by -cyclodextrin sulfate. It
suggests that the values obtained by the homogeneous method used here are probably
not affected by a change in the structure of HDL.

B-115
Accuracy based proficiency test for triglyceride in South Korea

Y. Park1, G. Kwon1, J. Lim1, J. Kim1, S. Koo1, H. Park2, C. Cho2, K. Lee2.

1
Chungnam National University Hospital, Daejeon, Korea, Republic of,
2
Korea Centers for Disease Control & Prevention, Chungcheongbuk-do,
Korea, Republic of
Background: When laboratory test results are not standardized, a different result
may be obtained for the same clinical sample. In this study we performed two trials
of accuracy based proficiency test for triglyceride measurement among 53 candidate
laboratories and assessed current performance of the routine measurement for
triglyceride in Korea. .
Methods : A total of 6 levels of commutable frozen serum pools were prepared as
secondary reference materials for triglyceride measurement according to CLSI 37-A.
Test results from ReCCS were regarded as target reference values for group using
glycerol blanking method, and values from CDC for the other group using method
without glycerol blanking. For each trial, 3 levels of pooled serum were sent to
participating laboratories and imprecision, bias and total error for each trial were
calculated.
Results: The bias of 18 laboratories (34%) using enzymatic method with glycerol
blank ranged from -7.62% to -1.47%, and that of 35 laboratories (66%) using
enzymatic method without glycerol blanking ranged from -9.09% to 1.67%.
Coefficient variations (CVs) ranged from 3 to 5% for each level of reference materials
but did not show significant difference between the two groups. When total error
(15%) was used for acceptability criteria, all of the results from 53 laboratories were
acceptable. However, when inaccuracy criteria (5%) is used, unacceptable rates for
the 1st and 2nd trial were 33% and 50% in the group using glycerol blank method, 15%
and 63% in the group using without glycerol blanking method respectively.
Conclusions : Through accuracy based proficiency test, comparison to the target
value determined by a reference measurement procedure allows both an absolute
and relative performance yardstick for laboratories using different measurement
procedures. And data obtained from this proficiency test made a footstep for further
national laboratory standardization for triglyceride.

Reference

Without
glycerol
blanking
(N=35)
With
glycerol
blanking
(N=18)

Table 1. Accuracy based proficiency test for triglyceride among 53 participating laboratories
TG(mg/dL)
CFS 11302 CFS 11303 CFS 21301 CFS 12-1-1 CFS 12-2-3 CFS 12-2-4
CDC
186.38
139.55
231.19
154.73
90.01
239.43
(total glyceride)
ReCCS
175.20
128.50
221.90
144.40
87.10
236.60
(triglyceride)
Mean(n=35)
Max
Min
Range
(Max-Min)
SD
%CV
Mean(n=18)
Max
Min
Range
(Max-Min)
SD
%CV

181.37
197.17
170.83

133.28
143.83
121.83

227.78
248.17
214.83

148.03
164.17
140.83

83.15
93.00
78.17

231.46
255.67
220.00

26.33

22.00

33.33

23.33

14.83

35.67

6.02
3.3
177.63
187.67
161.00

4.51
3.4
128.42
138.83
117.00

7.46
3.3
225.60
246.00
205.00

4.86
3.3
147.65
159.33
135.67

3.09
3.7
79.18
85.50
74.17

7.63
3.3
233.48
251.17
211.33

26.67

21.83

41.00

23.67

11.33

39.83

7.80
4.4

5.77
4.5

11.01
4.9

6.01
4.1

2.73
3.4

9.35
4.0

B-117
The association between lipoprotein(a) levels and coronary heart disease risk in
different ethnic groups: results from the Multi-Ethnic Study of Atherosclerosis

J. Cao1, B. T. Steffen1, G. Warnick2, J. P. McConnell2, W. Guan1, J. H.

Stein3, J. D. Kaufman4, W. S. Post5, M. Y. Tsai1. 1University of Minnesota,
Minneapolis, MN, 2Health Diagnostic Laboratory, Richmond, VA,
3
University of Wisconsin, Madison, WI, 4University of Washington, Seattle,
WA, 5John Hopkins University, Baltimore, MD
Background Elevated plasma lipoprotein (a) (Lp(a)) levels are established as a risk
factor of coronary heart disease (CHD) in populations mainly of European and African
ancestry. The objective of this study was to examine the association of lipoprotein (a)
(Lp(a)) levels with coronary heart disease (CHD) events in 4 racial/ethnic groups in
the Multi-Ethnic Study of Atherosclerosis (MESA).
Methods The MESA consists of 6,814 individuals without clinical evidence
of CHD (aged 45-84 years) at the initial recruitment. Individuals taking lipidlowering medication at baseline or with unavailable samples were excluded from
this analysis resulting in a remaining sample of 4,387 who were followed for 8.5
years. Incident CHD was defined as the first occurrence of myocardial infarction,
resuscitated cardiac arrest, CHD death, or definite angina. Lp(a) mass concentration
was measured with an isoform-insensitive turbidimetric immunoassay (Denka
Seiken, Japan) at the Health Diagnostic Laboratory Inc. (Richmond, VA) with interassay coefficients of variation less than 5%. Statistical analyses were conducted using
Stata (version 12.1, Stata Corp, College Station, TX) and R. Tukey-Kramer HSD
was used to test differences between groups. Since residuals analyses suggested a
non-linear relationship between CHD risk and Lp(a) as a continuous measure, we
dichotomized Lp(a) into < and > median groups. Cox regression was used to test for
association between Lp(a) and CHD events, adjusting for age, gender, race, diabetes,
hypertension, smoking status, high-density lipoprotein cholesterol (HDL-C), lowdensity lipoprotein cholesterol (LDL-C) and log-triglycerides.
Results The study population was composed of 28.7% African American (AA,
n=1,257), 12.4% Chinese American (CA, n=546), 22.7% Hispanic (HS, n=996), and
36.2% Caucasian (CU, n=1,588) participants. The distributions of Lp(a) levels in
all ethnic groups were left-skewed, and AA had a significantly higher level of Lp(a)
compared to the 3 other ethnic groups (p < 0.05). The median levels of Lp(a) in the
4 ethnic groups were: AA 35.1 mg/dL, CA 12.9 mg/dL, HS 13.1 mg/dL, and CU
13.0 mg/dL. The numbers of CHD events were: AA 66, CA 18, HS 49, and CU 105.
Using the Lp(a) median level of the entire cohort (17.8 mg/dL), Lp(a) > median was
associated with a
significantly higher CHD event rate than Lp(a) < median (hazard ratio (HR) = 1.42, p
= 0.0095) following adjustment for age, gender, race, diabetes, hypertension, smoking
status, HDL-C, LDL-C and log-triglycerides. In analyses stratified by ethnic groups,
Lp(a) level above the group-specific median was associated with a significantly higher
incidence of CHD in CU (HR = 1.55, p = 0.03), but not in the 3 other ethnic groups (p
= 0.08, 0.63 and 0.74 for AA, CA and HS, respectively).
Conclusion In MESA participants not on lipid-lowering medications at recruitment,
we found that elevated Lp(a) levels were associated with increased risk of CHD
independent of traditional CHD risk factors. When stratified by race and using racespecific Lp(a) median levels as the cutoff, the association was only significant in CU.
However, the power of subgroup analysis may be limited by the number of events.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S165

Lipids/Lipoproteins

Wednesday, July 30, 9:30 am 5:00 pm

B-118
New Enzymatic Method for Sphingomyelin Measurement verified by Mass
Spectrometry

H. Kuwata1, T. Kimura2, K. Miyauchi3, Y. Katayama1, N. Kayahara2, T.

Honda4, H. Hidaka5. 1Research Laboratories, Kyowa Medex Co., Ltd.,
Shizuoka, Japan, 2Research and Development Division, Kyowa Medex
Co., Ltd., Tokyo, Japan, 3Customer Support Group, Quality Assurance
and Regulatory Department, Kyowa Medex Co., Ltd., Shizuoka, Japan,
4
Department of Laboratory Medicine, Shinshu University, School of
Medicine, Nagano, Japan, 5Department of Biomedical Laboratory
Sciences, Shinshu University, School of Medicine, Nagano, Japan
Background: Serum sphingomyelin (SM) can help to predict the development of
coronary arterial diseases. However, no convenient and specific assay for measuring
SM in serum is available for routine laboratory practice. We previously developed the
new assay for SM using an enzymatic method in combination with a monoglycerolipase
and two types of phospholipase D. In this assay, phosphatidylcholine (PC) and
lysophosphatidylcholine are eliminated in the first step and the remaining SM is
measured in the second step. To validate this assay, we correlated the measurement
values with the data by mass spectrometric analysis.
Methods: We prepared 47 sera at Shinshu University Hospital and measured their SM
content using an enzymatic method on a Hitachi-7170 autoanalyzer. We also analyzed
the lipid extract of the same sera using a TripleTOF 4600 mass spectrometer (AB
SCIEX). N-heptadecanoyl-sphingosylphosphorylcholine (sphingomyelin with C17:0
fatty acid; SM 17:0, Matreya) was used as an internal standard. Lipid extraction was
performed according to Folchs method. The amount of each identified SM species
was determined by the difference of their fatty acid forms. We correlated results of the
enzymatic method with that determined by mass spectrometry.
Results: The total amount of SM in the 47 samples ranged from 0.249 to 0.945
mmol/L (mean : 0.497 mmol/L) in the enzymatic method and 0.241 to 0.870 mmol/L
(mean : 0.417 mmol/L) in the mass spectrometric method. Identified SM species
by the mass spectrometry were SM 16:0, 16:1, 18:0, 18:1, 20:0, 22:0, 22:1, 24:0,
24:1, 24:2. SM 16:0 was the most abundant (0.1910.101 mmol/L, 46.28.3 %) and
24:0 was the second abundant (0.06320.0434 mmol/L, 15.14.0 %) among the SM
species in the 47 samples. The other species each represented less than 10% of the
total SM. Within-run coefficients of variation (CVs) at 0.422 and 0.756 mmol/L in
pooled sera were 1.55% and 1.45% for the enzymatic method and 13.5% and 5.26%
for the mass spectrometry, respectively. We found a high correlation between values
of each SM species measured with the enzymatic method (X) and that determined
by the mass spectrometry (Y). Correlation coefficients and regression equations
were as follows; SM 16:0, r=0.897, Y=0.367X+0.649; SM 16:0+SM 24:1, r=0.912,
Y=0.520X0.249; all identified SM species, r=0.950, Y=0.894X1.95. Conversely,
the correlation coefficient for PC was low (PC 16:0/18:2 + PC 18:0/18:2, r=0.593).
These results suggest that the proposed enzymatic method can measure most of SM
species in serum with high specificity and accuracy.
Conclusion: Our new enzymatic method can measure SM in serum with high
specificity and accuracy, and therefore is very useful in clinical practice.

B-119
Effects of myeloperoxidase-modified HDL on reverse cholesterol transport and
monocytic migration

T. Kameda, Y. Usami, N. Sagae, R. Ohkawa, M. Tozuka. Tokyo Medical

and Dental University, Tokyo, Japan
Background: Myeloperoxidase (MPO) is one of the biomarkers for acute coronary
syndromes. In the advanced atherosclerotic lesions, MPO, mainly secreted from
macrophages, is known to induce the oxidized apolipoprotein AI (apoAI), such as
2-chloro- or 2-nitro-tyrosyl apoAI and apoAI-AII heterodimer. These products would
be expected to give us the different informations from MPO activities in plasma,
likely more specific to cardiac disease. We previously revealed that the plasma levels
of apoAI-AII heterodimer in patients urgently hospitalized for the treatment of acute
myocardial infarction were significantly higher than those of the healthy subjects. In
the present study, we investigated the effect of MPO oxidation on the antiatherogenic
properties of HDL, such as the cholesterol efflux capacity and the inhibition activity
of monocytic migration.

13-acetate. Then macrophages were loaded with acetylated LDL and [3H]-cholesterol
for 24 h. After 18 hours equilibration, cholesterol efflux was assessed in the media in
the presence of HDL, MPO treated HDL, or no acceptor for 4 h. The radioactivity in
the medium and the total cell-associated radioactivity were determined by scintillation
counting. The cholesterol efflux was calculated as a percentage: [3H]-cholesterol
in medium/ ([3H]-cholesterol in medium + [3H]-cholesterol in the cells) x 100. 3)
Evaluation of THP-1 cell migration; THP-1 cell migration assays were performed
with 8 m pore size inserts on the PET membranes. HUVEC (human umbilical vein
endothelial cell) was stimulated by LPS at 37 C for 16 h in the presence of HDL or
MPO treated HDL. The lower compartments of chemotaxis chamber were filled with
supernatant of HUVEC cultured medium. THP-1 cells were placed in upper chamber
and incubated for more than 24 h at 37 C. The THP-1 cells migrated in the lower
chamber were counted using an inverted microscope. The THP-1 migration ability
is defined by the percentage of migrated THP-1 number against the original number.
Results: ApoAI-AII heterodimer in HDL was apparently increased by the incubation
with MPO, which was confirmed by SDS-PAGE under reducing condition followed
by CBB R250 staining and immunoblotting using anti-apoAI and anti-apoAII
antibodies. No difference in the cholesterol efflux capacity was observed between
HDL and MPO treated HDL. In the THP-1 migration assay, the presence of HDL
indicated the significant reductive effect (44%) against LPS stimulation of HUVEC.
This effect was reduced to 34% by the treatment of HDL by MPO.
Conclusion: MPO oxidation did not largely affect the cholesterol efflux capacity
of HDL. However, MPO oxidation partially impaired HDL property to inhibit the
monocytic migration, suggesting that the oxidation of HDL by MPO would affect the
progression of atherosclerotic plaque. It means that the products, such as apoAI-AII
heterodimer, induced by MPO oxidation might be available as a biomarker to reflect
a progression of atherosclerosis.

B-121
Increases in Large HDL Particles are associated with Improved
Cardiopulmonary Fitness by Exercise-based Cardiac Rehabilitation in Patients
with Acute Coronary

S. Koba1, Y. Ito2, Y. Yokota1, T. Hirano1, M. Shoji1, Y. Kobayashi1. 1Showa

University School of Medicine, Tokyo, Japan, 2Denka Seiken Co.,Ltd.,
Tokyo, Japan
Background: Exercised-based cardiac rehabilitation (CR) can increase HDLcholesterol (HDL-C). However it remains unclear how elevated HDL-C and changes
of HDL subfractions are correlated with the improvement of exercise tolerance in
acute coronary syndrome (ACS) patients participated with CR.
Methods: Concentrations of cholesterol and apolipoproteins (Apo) in HDL
subfractions separated by heparin-Mn precipitation method were measured at the
onset of ACS and at the end of the 6-month CR program in patients (45 men and 6
women) aged of 64.3 11.8 years. Cardiopulmonary exercise tests were performed
at the beginning and the end of the CR program. All patients received successful
percutaneous coronary intervention on admission, and then started to take statins.
Results: Serum levels of HDL-C and ApoA1, and concentrations of cholesterol and
ApoA1 in large HDL fraction (HDL2) were significantly increased by CR (42.7 mg/
dl 14.1 mg/dl to 47.4 mg/dl 14.2 mg/dl, 128.8 mg/dl 23.4 mg/dl to 139.2 mg/
dl 26.0 mg/dl, 25.3 mg/dl 12.4 mg/dl to 30.3 mg/dl 13.6 mg/dl, 67.8 mg/dl
20.3 mg/dl to 79.4 mg/dl 24.5 mg/dl, respectively), while cholesterol and ApoA1
in small HDL fraction (HDL3) were not changed. Moreover HDL2-C / HDL-C ratio,
and HDL2-ApoA1 / ApoA1ratio were significantly increased by CR. In addition,
Spearmans rank correlation coefficient analysis revealed that only % increases of
HDL2-C were significantly associated with % increased of peak oxygen consumption
(VO2) and % increases of VO2 at anaerobic threshold ( = 0.439, p = 0.007, = 0.382,
p = 0.020), while neither HDL-C nor HDL3-C were associated with them.
Conclusion: CR can markedly increase the number of HDL2 particles, which is
significantly associated with the improvement of cardiopulmonary fitness. These
results suggest that CR is very useful therapy for the reverse cholesterol transport, and
the secondary prevention.

Methods: 1) Oxidation of HDL by MPO; HDL (1.063<d<1.210 g/mL) was incubated

with phosphate buffer (pH 7.4) containing hydrogen peroxide, diethylenetriamine
pentaacid, L-tyrosine, and MPO for 24 h at 37 C. 2) Evaluation of cholesterol efflux;
THP-1 cells were differentiated into macrophages by addition of phorbol 12-myristrate

S166

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Lipids/Lipoproteins

Wednesday, July 30, 9:30 am 5:00 pm

B-122
A Novel Method Using Cation-exchange and Heparin Affinity Columns
Arranged Tandemly to Determine ApoE-containing HDL-cholesterol in Unpretreated Whole Serum

S. Usui1, T. Ikeda1, R. Shinohata1, M. Okazaki2, S. Ikeda1, S. Kusachi1.

1
Okayama University Graduate School of Health Sciences, Okayama,
Japan, 2Professor Emeritus of Tokyo Medical and Dental University,
Skylight Biotech Inc., Akita, Japan
Background: Measurement of HDL-related subclasses has been suggested to be
more useful for evaluating coronary artery disease risk than total HDL. Clinical
significance of apolipoprotein E-containing HDL (apoE-HDL) has not been clarified.
Development of a reliable and rapid assay system to measure serum apoE-HDL levels
is thus essential.
Aims: We developed a high-performance liquid chromatography (HPLC) equipped
with cation-exchange and heparin affinity columns to measure apoE-HDL-cholesterol
(apoE-HDLc) levels in untreated whole serum, and studied the analytical performance
for serum apoE-HDLc determination. Separation characteristics of the system and
their isolated lipoprotein fractions were also studied.
Methods: An un-pretreated whole serum sample was injected into two tandemly
connected columns and eluted by a step-wise gradient manner, as shown in Fig.1.
Non-HDL lipoproteins are bound to the cation-exchange column, and unbound HDLs
next enter the heparin column. The heparin column retains apoE-HDL but no other
HDLs (apoE-deficient HDL).
Results: Our developed HPLC system completed the assay within 16 mins and
separated lipoproteins in un-pretreated serum into 3 peaks on the cholesterol pattern.
Each peak corresponded specifically to apoE-deficient HDL (first peak), apoE-HDL
(second peak), and non-HDL (third peak). The present HPLC system provided
acceptable small within-day imprecision values; 1.3% CV (n=8) and good acceptable
linearity with the serially diluted pooled serum, up to 18 mg/dL for apoE-HDLc.
The system was not affected by triglycerides at concentrations up to 450 mg/dL in
apoE-HDLc measurement. The apoE-HDLc levels of healthy volunteers determined
by the present HPLC system were 5.31.6 mg/dL (n=26), which accounted for
approximately 6-11% of total HDLc.

Methods: A total of 242 outpatients were scored into six groups, based on their number
of MetS components (from 0 to 5 variables) defined by the NCEP ATP III criteria,
modified for the Asian cutoff for waist circumference. Blood samples were analyzed
for lipid profile, LDL subclass (Quantimetrix Lipoprint, CA), and atherosclerosisrelated markers: apolipoprotein A-I (apoA-I), apoB, glucose, hemoglobin A1c
(HbA1C), high sensitive C-reactive protein (hsCRP), creatinine, cystatin C, and
vitamin D. The PGE method separates the intermediate density lipoprotein (IDL)
particles into three midbands (MID-A to C) and the LDL particles into larger-buoyant
LDL (lbLDL; LDL1 and LDL2), small-dense LDL (sdLDL; LDL3 to LDL7), and
HDL; sdLDL was calculated as the sum of LDL3 to LDL7.
Results: The mean levels of triglycerides, glucose, and HbA1C rose with increasing
MetS score, whereas those of HDL-cholesterol decreased. However, the concentration
of total cholesterol, LDL-cholesterol, non HDL-cholesterol, apoAI, hsCRP, and
vitamin D did not trend with increasing MetS score (all P >0.170). Using PGE, the
mean concentrations of VLDL, MIDC, MIDB, LDL2, and sdLDL positively correlated
with increasing MetS score, but those of MIDA and LDL1 inversely correlated,
similar to the pattern observed for HDL. Using backward stepwise logistic regression,
MIDC, MIDB, MIDA, LDL1, LDL2, and sdLDL were considered the independent
variables. LDL1 and sdLDL [regression coefficient = -0.033 and 0.054, odds ratio =
0.968 (95% CI, 0.943-0.994) and 1.055 (95% CI, 1.023-1.089), respectively] were
identified as being significantly associated with MetS (P <0.02). In the logistic model,
the sdLDL/LDL1 ratio showed the strongest association with MetS and demonstrated
an odds ratio of 5.544 (2.030-14.542, 95% CI). For predicting MetS, the area under
the ROC curve of the sdLDL/LDL1 ratio had the greatest diagnostic value (0.700),
followed by VLDL (0.694), sdLDL (0.689), HDL (0.669), LDL1 (0.648), MIDC
(0.605), and MIDB (0.596), which showed good discrimination power for MetS (P
<0.010), whereas the value of MIDA (0.572) indicated a poor power (P = 0.055).
Conclusion: Respective subpopulations of IDL and LDL particles can vary in their
ability to identify MetS. These variations may partially explain why a quantitative
assessment using absolute LDL-cholesterol concentrations, as typically measured
in conventional practice, is poorly associated with MetS. We show that the ratio of
sdLDL/lbLDL is strongly associated with the metabolic syndrome (high odds ratio
and highest area-under the curve). It may be a potentially important tool to maximize
the effectiveness of risk assessment for cardiovascular disease.

Conclusion: Our developed HPLC system with cation-exchange and heparin affinity
columns showed a rapid and precise determination of apoE-HDLc levels in unpretreated serum. The present system is thus useful in both clinical settings as well
as for lipid research.

B-123
Small Dense LDL Ratio Associates with the Metabolic Syndrome

P. Srisawasdi1, S. Vanavanan1, M. Rochanawutanon1, J. Kerdmongkol1, M.

H. Kroll2. 1Department of Pathology, Faculty of Medicine, Ramathibodi
Hospital, Mahidol University, Bangkok, Thailand, 2Quest Diagnostics,
Madison, NJ
Background: Low-density lipoprotein (LDL) cholesterol is often not an effective
predictor of cardiovascular risk because of the variability of the cholesterol content
within lipid particles. We investigated the association of lipoprotein subclasses,
classified with a polyacrylamide tube gel electrophoresis (PGE) method, with scoring
for metabolic syndrome (MetS).

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S167

Management

Wednesday, July 30, 9:30 am 5:00 pm

Wednesday, July 30, 2014

Poster Session: 9:30 AM - 5:00 PM
Management

B-124
Critical Values Reporting: The Search For an Effective Solution.

R. Khoury, P. Gudaitis, B. P. Salmon, A. Gandhi, J. Camoia, D. Pasquale,

D. Gudaitis. Aculabs, Inc., East Brunswick, NJ
Background: Critical laboratory result according to Dr. George Lundeberg is: A
laboratory test result that represents a pathophysiologic state at such variance with
normal as to be life-threatening unless something is done promptly and for which
some corrective action could be taken. Although his description of critical was more
than 30 years ago, critical vales received more attention when the Clinical Laboratory
Improvement Amendments Clinical Laboratory implemented the importance and the
concept of critical values. CLIA, CAP, State agencies and JCAHO require laboratories
to have written procedure for reporting critical value, defining critical results,
documentation and read back, and turnaround time for reporting critical results to the
care giver. The determination of critical cutoff depends on the population, setting, and
the feedback from the caregivers.

Collection of data: Precision and accuracy were evaluated for each chemistry analyte.
Within-instrument precision was calculated by measuring quality control materials
over one month. Accuracy was estimated by calculating observed bias (mean of the
actual data minus the expected mean).
Results: The AU680 results are summarized on the accompanying Method Decision
Chart. The majority of methods evaluated on both the AU680 and AU5800 were
classified as Good or better with many achieving Six Sigma performance. The latter
group should be easy to control with cost-effective QC practices. In contrast, a few
analytes do not achieve desired status and therefore may require more rigorous QC
practices.
Conclusions: Most clinical chemistry methods function at world-class specifications
using the Beckman Coulter AU680 and AU5800. The methods that fail to achieve
such performance demand analytical improvement from the method research
community. The data presented here help the clinical laboratory profession choose
which analytes to focus on for method improvement. Additionally, optimal QC
practices for laboratory production are implied.

Design: data from 79,328 tests ordered on 13,579 Specimens that were collected
from residents in Long-Term Care facilities were collected and included results
from: chemistry, hematology, coagulation, therapeutic drug monitoring, and cardiac
markers. Critical result calls and any problem with reporting the results to the
caregiver (because the patients are resident in Long Term Care Facilities the results
are given to the nurse in charge of that patient) were documented for every critical test.
Statistical analysis was done using Analyse-it.
Results: 1,784 (2.8%) critical results were documented; the majority of the critical
samples were for chemistry followed by hematology and coagulation. 616 (34.5%)
values were reported immediately to the caregiver by the technologist who performed
the test due to the severity of the results; 1168 values were reported by our call-in
department, 22% of the calls were unsuccessful due to either no one answering the
calls, nurses are busy/unable to take the calls, or nurses refuse to provide her/his name
to be documented in addition to read back the result. All unsuccessful calls were
followed by another call until the results were given to the appropriate caregiver.
Most of the unsuccessful calls were between 10-11 AM, followed by 5-6 PM, and 7-8
PM. which coincide with morning meetings, passing medication, lunch/dinner break.
Conclusion: the majority critical values are reported to the caregiver without any
delay; more than one fifth of the results took longer time to relay the result due to
inability to reach the care giver. Implementing reliable communication systems
will help improving the critical values reporting, such as immediate electronic
notification via fax/online, automated notification system, default phone number or
alternate caregiver in case of inability to reach the facility. Auditing record will help
identifying any inadequate documentation, weakness, and the facilities with the most
unsuccessful call to work with them to solve the problem.

B-126
Six Sigma Reagent Performance Quality Metrics for Beckman Coulter AU
Clinical Chemistry Instruments

D. D. Koch1, D. N. Greene2, S. Westgard3. 1Emory University, Atlanta, GA,

2
Permanente Medical Group, Berkeley, CA, 3Westgard QC, Inc., Orange,
CT
Background: Clinical chemists should be aware of the analytical strengths and
limitations of their laboratory methods. Six Sigma reagent performance quality
metrics are an excellent way to quantify the precision and accuracy of analytical
methods. However, the literature is scant with publications describing these metrics
for todays clinical chemistry instruments. Data for 26 clinical chemistry analytes on
two AU platforms (AU680 and AU5800) were used to calculate sigma metrics. We
display the performance of these methods on Method Decision Charts, a popular tool
for portraying Six Sigma metrics.
Quality Goals: Quality goals to evaluate the performance of the tested methods
were generally derived from CMS CLIA PT goals, but also included criteria from the
Biologic Variation Database and from RCPA.

S168

B-127
The Empower project - integrated tool to evaluate the quality and effectiveness
of laboratory testing.

L. A. C. De Grande1, D. Stckl2, K. Van Uytfanghe1, L. M. Thienpont1.

1
Laboratory for Analytical Chemistry, Faculty of Pharmaceutical Sciences,
Ghent University, Gent, Belgium, 2STT-Consulting, Horebeke, Belgium
Background Classical external quality assessment (EQA) is well established,
however, the need for integrated EQA-services recently emerged, among others to
empower clinical laboratories for future tasks, e.g., contribution to the development
and implementation of global health-care policies. Also, ISO 15189 accreditation
requires laboratories to identify and monitor quality indicators. From this perspective,
we developed the Empower project.
Methods The project comprises 4 pillars: (i) master comparisons with panels of
frozen single donation sera, (ii) virtual EQA-1 and (iii) -EQA-2 based on data readily
available in the laboratory, i.e., patient- and internal quality control (IQC) results,
and (iv) conceptual/statistical education about analytical quality. The pillars (i) to (iii)
are conducted across laboratories and manufacturers. The master comparisons aim at
participation of 20 laboratories per manufacturer, and encourages the latter to include
also their in-house laboratories. It is essential that the participants use homogeneous
systems, i.e., instrument, calibrator and reagent from the same manufacturer. Virtual
EQA-1 requires the laboratories to calculate and send the daily medians of the
results for outpatients. We plot the moving median of the collected data in time. The
participants can consult a graphical user-interface to monitor the mid- to long-term
stability of their performance in comparison to other peer group laboratories. Virtual
EQ-2 part (not operational yet, except in a few laboratories) plans a similar approach,
but on the basis of the daily IQC means.
Results The value of the master comparisons is in showing the intrinsic quality
of assays as performed under field conditions, the performance of the individual
laboratory within its peer group and the calibration fix-point and interchangeability of
results between laboratories/manufacturers. Monitoring of patient- and IQC-data gives
evidence about the mid- to long-term analytical variation of testing in the individual

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Management

Wednesday, July 30, 9:30 am 5:00 pm

laboratory, backed-up by information on peer group performance. It also enables to

uncover biases between different instruments in a laboratory, as well as occurrence of
shifts/drifts. If the cause of the aberration can be identified, the problem can be solved
either by the laboratory or the manufacturer. For example, an unacceptable lot-tolot variation may require factorizing by the laboratory or fundamental improvement
of the lot stability by the manufacturer. The observation of biases may also help
laboratories to understand the sometimes fluctuating flagging frequency of results.
Conclusions The Empower project reflects on the mid- to long-term analytical
stability of laboratory/assay performance and enables to uncover all major bias
components/sources. From this perspective, we believe it is a new integrated tool for
modern quality management. Its major asset is that it works with data generated from
commutable samples, and linked to observations in daily IQC practice. It strengthens
the position of laboratories in their claims from manufacturers, facilitates the dialogue
at the laboratory-clinician interface, and is a tool for the discipline to derive realistic
quality specifications. On longer term, it might establish a constructive relationship
between laboratories, manufacturers, clinicians and health policy makers, so that
together they can build towards a common understanding about manageable quality
of performance to the benefit of the patient.

B-128
The relevance of a computerized system for temperature and humidity control
in a large Brazilian laboratory

the submission of records in audit processes. DataNet also keeps the traceability of all
users activities, following the regulation from The Food Drugs Administration, Title
21, Code of Federal Regulations Part 11.

B-129
Managing Good Internal Quality Control by Adopting Risk Analysis
Framework

J. J. Jeyaraman, Z. Sakiman, E. B. Tan. Sunway Medical Centre Berhad,

Petaling Jaya, Malaysia
Background: Internal Quality Control (IQC) is the heart of quality assurance
and plays a pivotal role in not only ensuring accurate and reliable patient results
but also ensuring high standards of quality in materials, method performance and
manpower. SUNMED lab has implemented newly designed Analytical Quality
Control (AQC) strategy and it could actually improve overall assays monitoring and
performance. However, the question on whether the newly designed AQC strategy
alone is it effective enough for managing good analytical quality? Our objectives
are to determine whether applying Risk Analysis Framework could actually reduce
analytical and the probability of medical error, comply with accreditation standards
and further improve customers outcomes.
Methods: We have adopted Six-Step Risk Analysis Framework and came up with
Quality Control Plan (QCP). We did the following:

R. C. Ferreira, F. T. C. Sakamoto, L. G. Pereira, C. F. d. Pereira, D. Q.

Moraes. Diagnosticos da America, Barueri, Brazil

Firstly we identified the potential failures in incorrect test results.

Background: The controls and records of temperatures and humidity are very
important in many areas. In clinical laboratories, the temperature control of samples
and reagents are fundamental for analytical quality. This activity can be complex and
difficult when involves several areas and many instruments as refrigerators, freezers,
ultra freezers, climate chambers and acclimatized rooms, resulting in 60 points of
control in our DASA SP Central Lab. This Lab is currently processing around 4.5
million tests / month, with huge sample flow and local reagents stock representing
a large monetary amount. This reality demanded a technical solution. The team of
equipment management (SELAB) opted for a system based on thermo transmitter
technology, using radio frequency, software for data analysis and creating an
emergency flow.

The identified risks were evaluated and prioritized using criticality matrix (for
example staff competency, IQC and test algorithm)

Objectives To enumerate the main features and benefits of a computerized system for
temperature and humidity control as: Capability to monitor sixty points of temperature
or humidity; Standardization of our monitoring procedure - change from manually to
computerized Registration and safety in storage of data in compliance with FDA title
21, CFR part 11; Analyze equipments performance; Easy submission of records in
audit processes as PALC, CAP, ISO.
Methodology The laboratory acquired the DataNet System produced by Fourier
Systems. This solution provides an intelligent network sensor system with assurance
of reception and no data loss. The ZigBee is a standard-based protocol built around
the IEEE *802.15.4 wireless protocol, providing the network infrastructure required
for wireless and low power network applications. Basically the system consists of
thermo transmitters, receiver, server and an integrated phone. The thermo transmitters
were fixed next to the equipments. With a thirty minutes interval the measurements
are sent to server. Using the software, the ranges of alarm and pre-alarms are defined
for all points to be monitored. Together with The Department of Analytical Quality,
different ranges were set up. When the temperature reaches the control limits, the
software sends an email containing information about the area and temperatures that
are outside the acceptable range.
Results: In February of 2012, the system was on line, monitoring 24 hours these
instruments:
5 climate chambers;
6 ultra freezers (range: -80 to -60C);
15 refrigerators (range: 2 to 8C);
15 freezers (range: -30 to -8C);
7 heated chambers (various intervals);
12 acclimatized room (range: 15 to 25C)
Two year after implementation several risk situations were avoided for samples and
reagents through the alerts emailed to areas and staff working in the lab.
Conclusion: The system has shown a good level of reliability and has helped us to
avoid problems through tendency of temperature, shown in graphs. The daily registers
allow us to identify low performance equipment. During this period there werent any
problems with samples or reagents.
The Company reached a high level of security for the temperature data. Now is easy

After that we estimated the risk using the probability, severity and detectability.

In addition to that we identified control plans to reduce the proritized risks

We further implemented the mitigation plans into QCP (for example staff competency,
gap analysis, training strategies, revised test algorithm and AQC strategy)
We also reviewed the system for effectiveness of QCP
Results: The results showed
Significant risk reduction from the average criticality rating of 35 (unacceptable)
down to 12 (acceptable); hence reducing the probability of medical error.
Marked improvement in the ISO 15189 audit nonconformance from 17 in the year
2010 to 3 in Jan 2013 assessment.
Improved overall customer satisfaction rate by 15% in the 2013 against the year
2011.
Conclusion: By adopting Six-Step Risk Analysis Framework and implementing it
in QCP enables our SUNMED lab not only to mitigate but also assist in preventing
possible hazard or risk that may occur before incorrect results are reported to health
care providers and clinical actions being taken.

B-130
Head, Shoulders, Knees and Toes... Baby Steps to Specimen Nomenclature &
Informatics

R. Merrick1, C. Johns2, P. Banning3. 1Vernetzt, LLC, Sacramento, CA,

2
LabCorp, Burlington, NC, 33M Health Information Systems, West Linn,
OR
Objective: Applying informatics to laboratory reporting involves complex standards
and terminologies. Accurate portrayal of the specimen is crucial to appropriate
processing at the front end, and correct clinical interpretation by the provider.
Adding layers of computerization; secondary use cases of tumor registries, public
health reporting, laboratory assay research; US federal Meaningful Use mandates in
future years intensifies the needs. Work has been accomplished to date to create a
standardized specimen cross-mapping table using the SNOMED CT terminology. The
Specimen Cross-Mapping Table (CMT)s goals are: 1) Give guidance to SNOMED
CT encoding for Specimen related terms in the context of HL7 v2.x messaging
using the Specimen (SPM) segment, 2) Identify gaps in the existing standard
nomenclature, 3) Map the existing HL7 specimen terminology to SNOMED CT and
4) Provide references about common collection methods and 5) Indicate specimen
preferences for specific laboratory domains.
Methods: Local specimens terms from partner labs were collected and mapped to
a single term, called the Public Health Interoperability Project (PHLIP)-preferred
term. For each PHLIP-preferred term a definition is provided as well as a link to a
description of the collection method, where possible. Each PHLIP-preferred term is
then mapped, either one-to-one or one-to-many, to the HL7 defined SPM fields, using

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S169

Management

Wednesday, July 30, 9:30 am 5:00 pm

SNOMED CT concepts from the appropriate hierarchies for each applicable SPM
field. HL7 terms were also mapped to the PHLIP-preferred term, facilitating a HL7
to SNOMED CT crossmapping. The resulting Specimen-CMT, at this time focused
on human sample types, is now being reviewed by professional associations for the
accuracy of the definitions as well as to provide validation about preferred specimen
types for each laboratory section (microbiology, chemistry, pathology, etc).
Results: The Specimen-CMT has been incorporated into a simple mapping tool for
ease of mapping local codes to standards and across standards. Several new terms
have been submitted to SNOMED CT and guidance on use of specific vocabulary
forwarded to message guide authors.

cover 3 major areas, Auto verification range, Delta check (%) and Integrity Check.
Auto verification ranges were derived based on Reagent Assay Inserts, international
guidelines, and laboratory experience. Delta check values were calculated as %RCV.
Integrity checks were developed to ensure consistency, e.g. serum indices (HIL),
contamination, etc.
Validation: Analysis results were processed through Instrument Manager based on
rules developed. The same data set was compared with manual review and release.
Adjustments to rule parameters were made to ensure auto verification closely matched
manual review by reaching 0 % of false auto released results.
Results:

Conclusion: This work is ongoing at the Laboratory Messaging Community of

Practice (LMCoP), comprised of laboratory and standard experts from public health
laboratories at the state and federal level, national clinical laboratories, the National
Library of Medicine and professional organizations, to improve and expand beyond
human samples to animal and environmental domains. The goal is to provide a starter
set of specimen related vocabulary for electronic data systems in the health care arena.

B-131
Test Overutilization is an Issue That Should be Solved: Folate Studies Taken
From Internal to External Laboratory Reduced the Number of Orders by 76%!

E. Serin, A. Koro, I. Topcu, M. Gokce, T. Aksoy, A. Dogru, T. Kisaarslan.

Istanbul Research and Education Hospital Clinical Biochemistry Lab.,
Istanbul, Turkey
Background:The most frequent reason for the heavy load of the routine clinical
laboratories is overutilization of tests. Some of the requested parameters are not used
by the clinicians in their daily practice. In addition, test turnaround times are gradually
shortened by the laboratories to help the clinicians speed up their patient management.
This study evaluated whether prolongation of the test turnaround times will affect the
ordering frequency of the test by the clinicians.
Methods:The number of folate, vitamin B12 and ferritin test requests were counted
in ten-day periods throughout January and February 2014. Between January 21st
and February 2nd, folate requests were sent to an external laboratory because of the
shortage of folate in the distributors stocks. The supplier was not able to provide the
folate test for about two months and our laboratory consumed its own stocks. Then
we took the statistical numbers of requests of folate, vitamin B12 and ferritin for
this ten-day period(TP2), before(TP1), and after this ten-day period(TP3). One-way
ANOVA and Tukey post-hoc tests were used to compare the numbers of requests of
the three parameters between these three ten-day periods. Statistical significance was
set at p<0.05.
Results:There was a statistically significant difference between the three ten-day periods
for folate orders as determined by one-way ANOVA(F(2.27)=173.745,p<0.001).
A Tukey post-hoc test revealed that the number of test requests was significantly
different for folate orders between TP1(350.4046.14), TP2(84.3011.13) and
TP3(146.4033.06)(p<0.001), whereas there was no statistically significant difference
between the ten-day periods for B12(p=0 .224) and ferritin(p=0.155).
Conclusion:This study indicated that there is an obvious excess of folate test request
in our hospital. There was a remarkable difference between the essentially and the
routinely ordered numbers of folate tests. In contrast, no such difference was observed
for vitamin B12 and ferritin test requests. This remarkable gap between request
numbers of parameters of even the same diagnostic panel such as anemia suggests
that a limitation given to a test request may result in prevention of overutilization.

B-132

Conclusion: Adoption of auto verification successfully reduced the tedious manual

review process. 85.3% of test results were auto-released, and only on the remaining
14.7% required staff manual review. Of all these samples, almost 99.9% were released
correctly compared to our manual review. Only 0.1% was held for review when they
could not be auto-released. There is no impact on patient safety as manual check is
required and no results were falsely released.

B-133
Improving Workflow in the SGC Phlebotomy Area Utilizing Front Line Coworkers and Lean Tools

J. K. Witt, B. Barksdale. Mercy, Smith-Glynn-Callaway, Springfield, MO

Background: The project addressed workflow in the clinic phlebotomy area utilizing
lean and six sigma techniques. Workflow consisted of 6 phlebotomists working in
9 phlebotomy rooms. Phlebotomists collected 350 specimens a day with errors in
collecting/receiving of the specimens in the LIS of 75 specimens a month. The area
design was disjointed and inefficient, without defined roles for the phlebotomists
in collecting/receiving and calling patients. No computers were in the phlebotomy
rooms for collecting/receiving patient specimens.Two computers were located in the
main phlebotomy area. The 6 phlebotomists used these two computers, resulting in
bottlenecks, inventories and delays in specimen processing, inefficient movement,
lost specimens and chaos.
SGC Mercy clinic monitored unreceived specimen errors for 12 months. All errors
are a source of patient, physician and co-worker dissatisfaction as well as an increase
in cost and inefficiency.
Data was gathered to understand the current process and all phlebotomists were
invited to participate. The data was placed in a 5w2h, general process, process activity,
spaghetti and flow charts. Suggestions for improvement were also placed in charts
for comparison.
Methods: Frontline co-worker focus groups, observation, VOC, VOP and flowcharts..
Results: The phlebotomists identified the need of computers in the main rooms and
the need of better defined job roles. Unreceived errors improved by approximately
60%.
Conclusion: Involving front line co-workers to investigate and improve processes
led to the 60% improvement in collecting/receiving errors. Value adding efficiency
increased 27%. The phlebotomists are encouraged to make suggestions for further
workflow improvements as well as debrief on each months issues and problems.
Quality improvement is ongoing.

Developing, Implementing, and Validating Auto verification in the Medical

University Laboratory in Thailand

S. Vanavanan, P. Srisawasdi, N. Teerakanjana, N. Kumproa, K. Khupusup,

P. Kitpoka, S. Wongwaisayawan. Pathology Department, Faculty of
Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand
Objective: The purpose of this study is to build auto verification rules and evaluate
against manual verification using historical data. Rules are optimized to match the
expected outcomes from manual review.
Relevance: Instrument Manager provides a flexible middleware platform for
implementing auto verification using its sophisticated rules engine, freeing the
laboratory staff of other activities.
Methodology: Analysis results over a week were used to evaluate reliability of rules
developed for 36 Chemistry and Immunology tests. These rules were developed to

S170

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Management

Wednesday, July 30, 9:30 am 5:00 pm

B-135
Impact of Technology in Improving the Quality of Pre-analytical Phase of
Laboratory Investigations in a Tertiary Care Oncology Center.

B-134

C. V. Hingnekar, H. K. V. Narayan. Tata Memorial Hospital, Mumbai, India

Differences in laboratory requesting patterns in emergency department in

Spain.

A. Andrade-Olivie1, M. Lopez-Garrigos2, J. L. Barbera3, J. L. QuilezFernandez4, J. L. Ribes5, J. M. Gonzalez-Redondo6, J. Sastre7, J. V. GarciaLario8, J. Asensio9, J. I. Molinos10, J. Molina11, J. R. Martinez-Ingles12, J.
Diaz13, L. Navarro-Casado14, L. Martin-Martin15, M. Salinas2. 1Hospital
Xeral-Cies, CHU, Vigo, Spain, 2Hospital Universitario de San Juan de
Alicante, San Juan de Alicante, Spain, 3Hospital de Manises, Valencia,
Spain, 4Hospital Universitario Reina Sofia de Murcia, Murcia, Spain,
5
Hospital de Manacor, Manacor, Spain, 6Hospital Santiago Apostol,
Miranda de Ebro, Spain, 7Hospital Virgen de los Lirios, Alcoy, Spain,
8
Hospital Virgen de las Nieves, Granada, Spain, 9Hospital Universitario
Infantil Nio Jesus, Madrid, Spain, 10Hospital Sierrallana, Torrelavega,
Spain, 11Hospital Comarcal de la Marina, Villajoyosa, Spain, 12Hospital
General Universitario Santa Lucia, Cartagena, Spain, 13Hospital Francesc
de Borja, Gandia, Spain, 14Complejo Hospitalario Universitario, Albacete,
Spain, 15Hospital General de La Palma, Santa Cruz de Tenerife, Spain
Background:Compare laboratory requiring patterns in patients admitted to emergency
department (ED), in 76 Hospitals in Spain. Methods:20 tests ordered by ED physicians
during 2012 were examined in a cross-sectional study. Data were collected from
laboratory databases and indicators that measured every test request per 1000 ED
admissions and related test requesting ratios, were calculated. Results:Table shows
mean, median, range and variability index (Percentil90/Percentil10) of every indicator
result. The frequency of ordering the stat tests ranged from 9.8 to 466.2 per 1000 ED
patients admissions. Procalcitonin and NT-proBNP were only measured in 61 and
49 stat laboratories respectively. Total proteins were measured in every ED. Albumin
was measured in one. Also, lipase instead of amylase in one. Conclusion:Considerable
variability exists in the use of stat laboratory test by physicians in 76 ED. Variability
between centers was extremely high, especially in the less requested tests, despite
clear indications of such request in emergency setting, indicating that can be often
determined as a matter of routine or out of habit in some areas. These large variations
included tests that are clearly redundant, as urea/creatinine and AST/ALT. What is
really surprising is the high demand for some tests in some ED, such as procalcitonin
and NT-proBNP, compared to the absence of measurement in other settings. The high
variability of indicator results shows a probable stat abuse and misuse, a dangerous
issue in Emergency setting. Requests not justified may lead to delays in testing for
patients who have truly life-threatening conditions. Appropriateness indicators can
be applied across a spectrum of laboratories, being useful for comparing requesting
patterns. There is a need to unify demand by optimizing the use of appropriate tests,
through interdepartmental communication to achieve a good use of diagnostic testing,
on which many emergency clinical decisions are based.

Background: This is a study to assess the impact of technology in improving the

processes in the preanalytical phase of laboratory investigations as compared to
the erstwhile manual method. The objective of this study is to assess the impact in
terms of turnaround time, elimination of transcription and labeling errors resulting
in enhanced patient satisfaction. The technologies such as Laboratory Information
System for Requisitioning and Reporting, Smart card for patient identification and
payments, Patient Summoning System, Automatic Phlebotomy tube Labeler (APTL)
and Pneumatic Tube System have been adopted as part of the Automation.
Methods: A comparative study based on objective and subjective parameters was
carried out to find out the efficacy of the automated system over erstwhile manual
operations. 29 staff members from different sample collection areas and lab personnel
who had experienced both the systems were interviewed and their responses were
tabulated.
Results: The comparative data after automation reveals that there is a significant
reduction in sample labeling time (75%), reduction in transcriptional errors (81%) and
reporting errors (100%) as compared to the manual era. Sample collection per day has
increased by 166%. Similarly with the use of pneumatic tube system the time required
in delivering the samples has decreased (93%) so has the breakage due to mishandling
(10%). The need for repeat sample collection also decreased (66%). By installation of
Patient summoning system the average patient waiting time is reduced by 73%.The
issues of patient identification and traffic has improved.
Conclusion: The results show that with the introduction of integrated automated
systems working in tandem there has been a significant reduction in Turn around
Time and elimination of errors. Staff efficiency has improved and so has the quality of
care resulting in improved patient satisfaction. It is evident therefore that technology
driven management can improve the quality of patient care, however the challenge of
perpetual innovation and maintenance of such systems remain.

B-136
Recommendations for New QC Rules Based on Precision from 2012 Data.

d. S. plaut1, N. Lepage2, T. Taylor3, W. McLellan4. 1unaffiliated, Plano,

TX, 2Childrens Hosp. Eastern Ontario, Quebec, ON, Canada, 3Ingalls
Hospital, Chicago, IL, 4unaffiliated, Hollywood, FL
Objectives: To show that data based on current precision indicate that new QC rules
be seriously considered.
Relevance: Current data indicate a significant increase in precision since the first QC
rules were proposed by CAP (1974) and AACC (1981).
Methodology: Two CAP (1988 and 2012) quantified precision. Our survey of 35
analytes from chemistry, hematology and hemostasis indicated a decrease of 58%

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S171

Management

Wednesday, July 30, 9:30 am 5:00 pm

in CV (average; range 22-83). We translated the 2012 SD into a new set of rules to
detect both systematic and increased random errors. Our translations indicate these
rules to be quite effectively: 1 4SD, 2 3SD and R 5 SD with cumulative sum for
certain analytes.
Results: The table shows representative changes.
Conclusions:
1) The improvements in precision indicate that QC rules be reevaluated. These
changes significantly reduce false rejects while increasing error detection and lower
patient risk.
2) Each analyte should be assessed to determine the proper rule(s).
Conclusion:
Analyte

CAP 1988
Mean
Cholesterol
330mg/dL
Hemoglobin
19.5gm/dL
Prothrombin Time 19.0 sec.

Representative Data
CAP 2012
%CV
Mean
5.6
217mg/dL
3.8
15.1gm/dL
6.4
11.5 sec

%CV
1.0
1.7
2.8

% Decrease
87
41
56

B-139
Changes in Primary Care Requesting patterns in a two year period in Spain

M. Salinas1, M. Lopez-Garrigos1, M. Ortuo2, M. Graells3, M. GarciaCollia4, M. Yago5, M. Frutos6, N. Esta7, N. Fernandez-Garcia8, P. GarciaChico9, A. Rodriguez-Piero10, R. Franquelo11, R. Gonzalez-Tamayo12,
S. Pesudo13, V. Granizo14, V. Villamandos15, V. Perez-Valero16. 1Hospital
Universitario San Juan, San Juan, Spain, 2Hospital Universitario La
Ribera, Alzira, Spain, 3Hospital General Universitario de Alicante,
Alicante, Spain, 4Hospital Ramon y Cajal, Madrid, Spain, 5Hospital
de Requena, Requena, Spain, 6Hospital Universitario Nuestra Seora
de la Candelaria, Tenerife, Spain, 7Hospital Dr Peset, Valencia, Spain,
8
Hospital Universitario Rio Hortega, Valladolid, Spain, 9Hospital General
Universitario de Ciudad Real, Ciudad Real, Spain, 10Hospital Universitario
de Mostoles, Mostoles, Spain, 11Hospital Virgen de la Luz, Cuenca, Spain,
12
Hospital de Torrevieja, Torrevieja, Spain, 13Hospital La Plana, Valencia,
Spain, 14Hospital Universitario de Guadalajara, Guadalajara, Spain,
15
Hospital Santos Reyes, Aranda del Duero, Spain, 16Hospital Regional de
Malaga, Malaga, Spain
BACKGROUND: To compare Primary Care Requesting patterns between two
different years in Spain, using appropriateness indicators, to try to ascertain if better
demanding behaviours are achieved along years.
METHODS: 36 and 76 laboratories for the year 2010 and 2012 respectively from
diverse regions across Spain filled out the number of 29 tests requested by GPs.
Two types of appropriateness indicators were calculated. Every test requests
per 1000 inhabitants of the following: alkaline phosphatase (ALP), alanine
aminotransferase (ALT), aspartate aminotransferase (AST), calcium (CA), cell
blood counts (CBC), C-reactive protein (PCR), cholesterol (CHOL), creatinine
(CREA)erythrocyte sedimentation rate (ESR), ferritin (FERR), phosphate (FOSF),
gammaglutamiltranspeptidase (GGT), glucose (GLUC), HDL-cholesterol (HDL),
glycated hemoglobin (HBA1), iron (IRON), lactate dehidrogenase (LDH), prostate
specific antigen (PSA), thyrotropin (TSH), total bilirubin (TBIL), triglycerides (TG),
urate (URAT), urea (UREA), urinalysis (URIA). And ratio of related tests requests.
Free PSA/PSA (FPSA/PSA), aspartate aminotransferase/alanine aminotransferase
(AST/ALT), direct bilirubin/total bilirubin (BILD/BIL), folate/B12 vitamin (FOL/
B12), free thyroxin/thyrotropin (FT4/TSH), urea/creatinine (UREA/CREA)
RESULTS: In spite of the differences observed, CBC, glucose and urate were less
requested and TSH more requested in the second period, no significant differences
were found in any of the studied tests.
DISCUSSION: Our results suggest that test requesting in Primary Care in Spain have
not varied in a two year period. At least achieving targets in related tests requesting
ratios, as AST/ALT, is necessary. The showed figures can be used as a pillar foundation
to be based in ulterior interventions to achieve appropriate requesting.
CONCLUSION: A more active Clinical Laboratory behavior is necessary to lead to a
better laboratory tests appropriate Primary Care requesting.

S172

B-143
Environmental resource management of an ISO 14000 certified clinical
laboratory in Brazil

A. L. T. Arajo, N. B. C. S. Almeida, J. A. R. Vaz, S. S. S. Costa, L. F. A.

Nery. Sabin Laboratory of Clinical Analysis, Braslia, Brazil
Background: Our clinical laboratory, with headquarters in Brasilia - DF, Brazil,
makes 1.7 mi exams per month and it has an environmental resource management
system, with ISO 14000 certification, that controls several processes in order to
reduce environmental impact in our activities. The purpose of this study is to know
the environmental impact that the laboratorys activities had in the ecosystem in 2013
(environmental performance) and to share the monitoring methods, treatments and
control used in a health care services institution.
Methods: We measured the consumption of natural resources (water, electric energy
and paper), fossil fuel for sample transportation, plastic bags, biomedical and solid
waste, paper and electronic waste recycling, printing savings through exam results
checked online (labs website). In short, the biomedical waste is taken to a waste
treatment facility and eliminated using pyrolysis, the effluent is treated through an
oxidative process using ferric chloride and sodium hypochlorite solution and the
environmental monitoring is performed through measurement of chemical oxygen
demand (COD) and biochemical oxygen demand (BOD).
Results: The laboratorys environmental performance in 2013 can be seen in the table
below:
Environmental performance - 2013
Water consumption (M)
Electric energy consumption (kW)
Paper consumption - white paper (unit)
Paper consumption - recycled paper for reports (unit)
Printed pages (unit)
Waste sorting (recycling) (Kg)
Fuel consumption (liters/year)
Plastic - plastic bags for waste (unit)
Plastic - Oxo biodegradable plastic bags (unit)
Plastic - plastic cup (unit)
Recycled electronic waste (Kg)
Chemical residue treatment (Liters/day)
Biomedical waste treated (Kg)
Exam results checked online
Printing and paper savings (unit)

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

5497
2.087.898
12.920.000
4.191.662
21.126.179
4.038
91.744
338.7
956.053
3.001.800
824
2000
203.166,25
794.614
6.356.912

Management

Wednesday, July 30, 9:30 am 5:00 pm

Conclusion Identifying the environmental performance of the clinical laboratory

allowed us to see the impact of the activities in the ecosystem. This knowledge
encourages all workers to engage in different processes, the involvement of service
users, and it spreads the environmental resource management practices of the ISO
14000 certification. The environmental performance is assessed every semester and
the results are available in the companys website and in the sustainability report
published by the United Nation Global Compact every year, in their website.

B-144
Critical laboratory tests values notification according to the Internacional
Patient Safety Goals by Joint Comission in a general Private Hospital in So
Paulo, Brazil.

L. R. Almeida. DASA, So Paulo, Brazil

Background: A laboratory critical value refers to an extremely abnormal laboratory
test result which may be life threatening if treatment is not initiated immediately. The
number of critical values results will influence laboratory workload, clinical care and
patient treatment. International Paciente Safety Goals (IPSG) has been introduced in
Brazil recently to reach the Joint Comission acreditation standards. The notification
of the critical laboratory tests results, recommended by the goal number 2 (improve
effective communication) of IPSG, is one of the most challenging surveys for the
clinical laboratories.
Methods: Critical laboratory tests results of a 300 bed general hospital were analyzed
from August 2013 to December 2013 using the laboratory information system. The
five main tests were selected to understand the Hospital profile according with the
tests results and to establish a monthly follow up of the notifications registered on the
laboratory system according with the TJC recommendation.
Results: From August to December 2013, 5834 tests from a total of 441.571 (1,3%)
fulfilled the definition of a critical value. The five more representative tests that showed
critical results were: Prothrombin time with 580 (11%), Partial thromboplastin time
with 580 (11%), pH with 285 (5%), Troponin with 216 (4%) and K with 118 (2%).
The notification of these values was improved along the months reaching 87% in the
last month of this series. The final objective is to reach 100% notification according
to the recommendation of TJC.
Conclusion: IPSG is very important to improve the patient safety and can also be
used to improve the laboratory procedures and the communication with the hospital
staff. The results we have obtained in this study were compared with the ones from
another Hospital that has already reached the second period of accreditation by TJC.
We noticed that goal number 2 from the IPSG was improved among the months in
that hospital, reaching 100% of successful communication of the critical values. The
main driver that supported this goal was a very effective continuous education of the
laboratory team.

B-146
Quality Assessment and Management in Clinical Diagnostic Laboratory
Medicine

J. Kalra1, R. Sachdeva2, N. Kalra2. 1University of Saskatchewan and Royal

University Hospital, Saskatoon, SK, Canada, 2University of Saskatchewan,
Saskatoon, SK, Canada
Background: Laboratory results play a major role in guiding decisions in patient
management. In laboratory medicine, meaningful, accurate and precise measurements
are essential for diagnosis, risk management and treatment. Various strategies have
been adopted to reduce laboratory errors including internal quality control (QC)
procedures and external quality assessment (QA) programs. Autopsy services also
play a major role in contributing to clinical knowledge, medical education and quality
assurance programs. The purpose of this study was to assess Proficiency Testing (PT)
programs and to determine the rate of concordance and discordance between clinical
diagnoses and post-mortem findings.
Methods: Our clinical laboratory participated in two external PT programs: Provincial
Health Metrx (mandated by College of Physicians and Surgeons of Saskatchewan)
and College of American Pathologists (CAP) survey. The clinical laboratory tests
commonly used to manage patients were examined for the period of one year for
any discrepancies. We also retrospectively reviewed the records of the medical and
autopsy charts for all the deceased adult in-patients admitted during 2002 to 2004 in
the hospitals of Saskatoon Health Region (SHR). A total of 3416 in-patient deaths
were registered during the study period. In accordance with selection criteria, 158
cases were included in this study. The mean age of subjects was 66.6 15.3 years with

a range of 16-94 years. The study group consisted of 92 males (58.2%) and 66 females
(41.8%) with an average length of stay at the hospital of 12.9 10.9 days. In addition,
we evaluated the impact of diagnostic modalities such as Computerized Tomographic
Scanning (CT) and Magnetic Resonance Imaging (MRI) on clinical diagnoses.
Results: In the Health Metrx PT, 5 groups of 43 analytes were analyzed and from 598
tests, 5 discrepancies resulted, yielding a total discrepancy rate of 0.84%. In the CAP
survey PT, 12 groups of 58 analytes were analyzed and of 431 tests, 3 discrepancies
resulted, yielding a total discrepancy rate of 0.70 %. In both surveys the remaining
tests had no discrepancies. The autopsy results showed that the concordance rate
between clinical and autopsy diagnosis was 70.9%. The discordance rate was 24%
and in 5.1% of the study population a conclusive clinical or autopsy diagnosis was not
finalized. CT scans and MRI were found to be confirmatory or diagnostic in 85% and
93% of the autopsy patients in which these modalities were used.
Conclusion: The quality in the clinical laboratory is maintained in a satisfactory
manner, meet the performance criteria and requirements set up by provincial
regulatory agencies. It is prudent to monitor, promote and enhance quality services for
our patients. The study confirmed that the concordance and discordance rates between
clinical diagnosis and post-mortem findings in SHR are consistent with those reported
in the literature. Also, despite the technical advances in diagnostic modalities,
diagnostic discrepancies remain prevalent in the present day health care system. The
study also emphasizes the value of PT programs and autopsies as an effective quality
improvement and educational tool with a strong impact on quality management.

B-148
Sigma metrics impact on analytical performance in a chemistry area of a
reference clinical laboratory A case study at the clinical laboratory of Hospital
Pablo Tobon Uribe the hospital with soul

A. PORRAS1, A. M. BEDOYA2, S. LOPERA2, S. JARAMILLO2, D.

RIOS2. 1QUIK, BOGOTA, Colombia, 2HOSPITAL PABLO TOBON URIBE,
MEDELLN, Colombia
Sigma metrics impact on analytical performance in a chemistry area of a reference
clinical laboratory A case study at the clinical laboratory of Hospital Pablo Tobon
Uribe the hospital with soul Key words: analytical performance, sigma metrics.
Background: The most frequently method used to validate analytical runs in clinical
laboratories is the Levy Jennings (LJ) graph, but this tool can control only the stability
of the process, not the size of the analytical errors. Lack of control on the size of the
analytical error, could produce misleading or misinterpretation, resulting in wrong
laboratory results. Controlling analytical errors contributes to increase the reliability
and clinical utility of the results. Analytical error control is time-consuming and
demands effort, causing discouragement sometimes. The sigma metrics allow control
of analytical errors for a big group of analytes with different total allowable errors,
contributing to the improvement of analytical performance in clinical laboratories.
Methods: At the Clinical Laboratory of Hospital Pablo Tobon Uribe in Medelln,
Colombia on June, 2011, the sigma metric approach was implemented to control the
performance of 25 analytesof clinical chemistry by awareness concepts, personnel
training, using a specialized software, implementing connectivity for quality control
data and showing new graphing tools such as performance per percentiles, sigma
metric and total error integrated with Levy Jennings in a specialized software .
For the initial diagnosis, the sigma metric was measured using a variety of quality
control specifications as biological variability, RilliBAK, CLIA and the state of the
art by percentiles. After this, multiple corrective actions were implemented and
quality specifications were initially selected, but later on were changed to increase
the constraint with the objective to monitor analytic imprecisin. Results: From
June 2011 to October 2013,the performance improvement of the analytes, controlled
through sigma metricwas considerable. The decrease of the imprecision, bias and
analytical total error was observed in the majority of the analytes, the percentage of
the analytes with sigma metric >5 increased from 68% to 79.4% and the percentage
of the analytes with sigma metric <1.96 decreased from 14% to 2.94%. The trend of
performance improvement is clearly observable during these two years as a result of
the homogenization of knowledge, process standardization, continuous monitoring,
rigorous analysis and corrective actions. Conclusions:The implementation of sigma
metricsand its graphs shows the errors as errors per million of opportunities that
exceed the total error allowable defined. Additionally the behavior of analytes is
displayed, through a quick glance, contributing to a decrease in analytical errors.
The diversity of situations that occur at clinical laboratories, such as differences
in workloads, stabilities of reagents, adjustment or calibration times and clinical
metrological requirements,influencesthe analytical performance. These situations
can be controlled by sigma metric, providing a standard measure that allows for the
monitoring of the performance of a large amount of analyte, even despite the diversity.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S173

Management

Wednesday, July 30, 9:30 am 5:00 pm

Collaboration between ER and Lab to Improve Troponin Turnaround Times

instrument CV and bias combinations have sigma values [(TEa - bias)/CV] ranging
from 3 to 6. We determined the maximum value for E(Nuf) and area under the E(Nuf)
curve for the 4 different cases:

A. M. Boelstler. St. John Hospital & Medica Center, Detroit, MI

1. QC rule means centered on instrument means and QC rule limits based on

instrument SDs

B-149

Purpose: The purpose of this project is to improve the care of Acute Coronary
Syndrome (ACS) patient by developing a collaborative relationship with the
Emergency Room (ER) and laboratory staff in order to decrease the turnaround time
(TAT) of door to troponin result. The goal is door to troponin result in less than 30
minutes. Early diagnosis and medical management of patients with ACS improves
the overall outcome in patients presenting to the ER with a complaint of chest pain.
Cardiac markers are used in the diagnosis and risk assessment of patients with chest
pain.
Design: A multidisciplinary team was formed to create a process and guidelines for the
ACS patient that presents to the emergency room. This team worked collaboratively to
initiate change within the triage area to reduce draw times, result times, and decrease
hemolysis rates.
Participants: A Chest Pain Committee encompassing physicians, registered nurses
(RN), emergency room technicians (ERT) , phlebotomists, and lab representatives.
Methods: Multiple meetings were held between the lab and the ER to understand each
others processes in relation to the entire procedure from door to troponin result. Each
area worked together, not in silos, to improve the time for each step. Well-defined
guidelines were developed for the initial treatment of the ACS patient in triage. Based
on these guidelines, the triage RN quickly determines if the patient meets Cardiac
Markers guidelines and directs the ERT and phlebotomist accordingly. A phlebotomist
is stationed in triage at all times. If cardiac marker guidelines are met, the cardiac
markers will be drawn by the phlebotomist while the ERT is performing an EKG.
The ERT or RN will be responsible for the lab draw if multiple draws are needed. The
introduction of a Mint Green lab tube was added specifically for Troponin levels in
the ED. Troponin specimens from this tube can be processed immediately and help us
to meet the <30 min Troponin TAT goal. Following the initiation of this process, data
was collected and interpreted to determine the success of our process improvement.
Results/Outcomes: Emergency Room and Lab developed a collaborative relationship
of professionalism and partnership. Data was received that exemplifies how we have
improved the care of the ACS patient.
Order to draw times has decreased from 55 minutes to 8 minutes.
Draw to received specimen has decreased to 3 minutes.
Our door to Troponin result time is averaging 47 minutes with data collection
continuing.
Troponin hemolysis rates have decreased from 15% to 1%
Implications: A collaborative Team Approach is being utilized in the care of the ACS
patient to improve both patient and associate satisfaction. The care of the ACS patient
has improved significantly as evidenced by our data interpretation. Data continues to
be collected to evaluate the ongoing effectiveness of our process improvement.

B-150
Designing QC Rules for Multiple Instruments: Should a QC Rule be Centered
on Individual Instrument Means or on a Fixed Mean? Should a the limits be
based on Individual Instrument SDs or on a Fixed SD?

L. S. Kuchipudi, J. C. Yundt-Pacheco, C. A. Parvin. Bio-Rad Laboratories,

Plano, TX
Background: Objective - Compare performance of QC strategies centered on the
individual instrument mean versus a fixed mean and limits based on individual
instrument SD versus a fixed SD when applied to multiple instruments testing the
same analyte.

2. QC Rule means centered on a fixed mean and QC rule limits based on instrument
SDs
3. QC rule means centered on instrument means and QC rule limits based on a fixed
SD
4. QC rule means centered on a fixed mean and QC rule limits based on a fixed SD
In each of the above cases we design the QC rules so the overall false rejection rate
across the 4 instruments = 0.0097, which is the false rejection rate for the 1:3s/2:2s/
R:4s QC rule.
Results: Tables of results are computed for the simulations. In general, the maximum
E(Nuf) and the area under the E(Nuf) curve were lowest for rules using a fixed mean
and fixed SD for the instruments.
Conclusion: Using a fixed mean and fixed SD for the QC rule had the best
performance. The fixed mean appears to balance the risk of reporting unreliable results
across multiple instruments while the fixed SD allocates more false rejection rate to
poorer performing instruments resulting in a lower overall risk of reporting unreliable
results when individual instruments have moderate to good process capability (3-6
sigma).

B-151
Evaluating the Reproducibility of Analysis in the Clinical Laboratories.
Results from a proficiency testing (PT) scheme and comparison with biological
variability.

D. Rizos1, O. Panagiotakis2, K. Makris3, A. Haliassos2. 1Hormone

Laboratory, Aretaieion Hospital, Medical School, University of Athens,
Athens, Greece, 2ESEAP Greek Proficiency Testing scheme for Clinical
Laboratories, Athens, Greece, 3Clinical Biochemistry Department, KAT
General Hospital, Kifissia, Greece
The quality criteria for the proficiency testing (PT) schemes in Laboratory Medicine
include the evaluation of the accuracy of the participating laboratories, the use of
human origin commutable testing materials (sera) that minimize matrix effects, the
evaluation of linearity of assays by using at least two samples per distribution and the
estimation of the reproducibility of measurements.
In order to fulfill these criteria, ESEAP (the Greek PT scheme in Laboratory Medicine)
introduced the measurement of reproducibility using the analysis of the same sample
four times during a yearly cycle. The quality goal for the participating laboratories,
is the reproducibility to be at least 50% of the biological variability for each analyte.
(Ref: Ricos et al. Current databases on biologic variation: pros, cons and progress.
Scand J Clin Lab Invest 1999;59:491-500, revision 2014).
In this study we used results from the 290 laboratories of ESEAP in Greece and
Cyprus and we evaluated them against the proposed limits derived from the biological
variability for each of the analytes included in the clinical chemistry scheme.
We excluded the results from laboratories that havent reported all the four samples,
from the laboratories that where excluded from the normal bias analysis (elimination
in two-passes of all results > or <2.5SD of the consensus mean value) and the
laboratories that reported a method change during this cycle. We finally processed
4 results from 209 laboratories and for 21 different parameters and we calculated the
mean value of reproducibility for each parameter.

Relevance - Using a fixed mean and SD appears to be a common practice when

multiple analytic units evaluate the same analyte. The comparative efficacy of this
approach has not been formally evaluated. We compare the expected number of
unreliable final results reported due to the occurrence of an out-of-control condition,
E(Nuf), when the QC rule means are centered on the instrument means versus a fixed
mean and when the QC rule limits are established based on the instrument SDs versus
a fixed SD.
Methods: We consider 4 analytic units in the laboratory evaluating the same analyte.
We assume each instrument evaluates 2 QC levels using a 1:3s/2:2s/R:4s QC rule
every 100 patient examinations. The fixed mean is set to the average of the instrument
means. The fixed SD is set so the overall false rejection rate across the 4 instruments is
0.0097. We investigate a range of means and SDs for the 4 instruments. The resulting

S174

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Management

Wednesday, July 30, 9:30 am 5:00 pm

These results are presented at the following table:

Measured
Biological
Proposed limits
Analyte
Reproducibility
Variability (%)
(%)
(%)
Glucose
5.60
2.80
2.31
Urea
12.10
6.05
3.36
Creatinine
4.16
5.95
2.98
Sodium
1.42
0.60
0.30
Potassium
4.60
2.30
1.78
Total Protein
2.55
2.75
1.38
Albumin
2.85
3.20
1.60
Cholesterol
5.95
2.98
2.68
HDL- Cholesterol
7.30
3.65
3.57
Triglycerides
19.90
9.95
2.89
Uric Acid
8.60
4.30
2.82
Total Bilirubin
21.80
10.90
4.14
Calcium
2.55
2.10
1.05
Phosphate
8.15
4.08
2.63
Magnesium
3.60
3.60
1.80
Iron
26.50
13.30
4.49
SGOT (AST)
12.30
6.15
3.13
SGPT (ALT)
19.40
9.70
5.30
-Glutamyl Transferase 13.40
6.70
4.59
Creatinine Kinase
22.80
11.40
8.90
Amylase
8.70
4.40
3.08
Our results show that all the currently used methods outperform the proposed
quality goals, except for the cases of Sodium and Calcium as also as of Creatinine,
Total Protein, Albumin and Magnesium where the results are below the biological
variability but not below the quality goal of 50% of the biological variability.

B-152
Relational Data Modeling Approach to Demonstrate Value of the Clinical
Laboratory in Healthcare Systems

S. Chekuri, B. Clements, B. Brimhall. University of Mississippi, Jackson,

MS
Background: Adoption of electronic health records has will likely expand over the
next several years due to incentives and penalties in the American Recovery and
Reinvestment Act. Furthermore, there is growing interest in mining and analyzing
this data to evaluate costs, practice patterns, and clinical effectiveness. We evaluated
the impact of the ability to efficiently link laboratory data to external sources for large
hospital projects. We also developed an appropriate data model to facilitate such data
integration.
Methods: We extracted and linked clinical and operational data from existing sources
(EPIC Systems, Cerner CoPath) which were also joined to detailed financial
data (Allscripts) for each department as well as the hospital system as a whole as
well as external data documents (e.g., external price proposals). Linked data were
used prospectively to evaluate the pro forma return on investment (ROI) for hospital
projects involving the laboratory as well as the impact of such projects on workflow
and quality measures when applicable. Linked data were employed retrospectively
to evaluate the actual ROI and impact on clinical operations. We specifically looked
at two large recent projects (2013-2014): a proposal to purchase a new microbiology
system (prospective) and the results of a project to streamline radiology patient flow
using POCT (retrospective). In each case, we examined the impact of absent, or poor,
linking of laboratory to external data sources on each project. Based on our work we
constructed a relational data schema with Navicat software to connect disparate
clinical, operational, and financial data sources.
Findings: The first project was an evaluation to purchase a MALDI-TOF instrument
for bacterial identification. Without efficient links to external data sources we had to
rely on cost savings resulting from lower reagent costs. With only this information
projected measures of ROI were: payback period = 20.6 years, net present value
(NPV) = ($228,519), and modified internal rate of return (MIRR) = -22.34%. None of
these measures argues financially to pursue the project and one would have to rely on
non-financial factors to make the case for purchase. We then re-evaluated the project
including of cost savings resulting from earlier adoption of appropriate treatment of
patients with sepsis (MSDRGs 371-373) and earlier discharge (based on actual costs
and volumes at our hospital system). With this linked information, measures of ROI
were: payback = 12.62 weeks, NPV = $5,471,954, and MIRR = 80.79%. All of these
measures are highly favorable for the project. For the second project we evaluated the
impact of a joint laboratory/radiology project on net revenue for radiology contrast
studies. After implementation of the POCT intervention, the annual patient volume
for these studies increased by 906 resulting in a net contribution of $410,776. As a
project the actual payback period was 22.1 days. Patient satisfaction measures were
also demonstrably better.
Conclusion: Linking laboratory data and data from external sources allows laboratory
professionals to 1) provide optimal data for making clinical and operational changes,

2) credibly demonstrate value to hospital administrators, and 3) foster collaborative

projects throughout the hospital system.

B-153
Comparison of critical results frequency for point-of-care testing (POCT) vs.
STAT core laboratory testing among critical care unit patients: a management
review regarding POCT utilization

S. K. Mardekian, L. J. McCloskey, D. F. Stickle. Jefferson University

Hospitals, Philadelphia, PA
Background: A point-of-care analyzer (EPOC, Epocal, Ottawa) was recently
introduced into a critical care unit (CCU) at our institution to measure blood gases
(pH, pO2, pCO2), electrolytes (Na, K, ionized Ca), metabolites (glucose, lactate),
and hematocrit (HCT) in select patients. Because of cost considerations, POCT
was originally restricted for use in patients meeting conditions either of use of
extracorporeal oxygenation (ECMO), use of a ventricular assist device (VAD), need
for resuscitation (code), or known hemodynamic instability. In a setting where STAT
testing is frequently warranted, selective use of POCT raised the issue of whether
uniformity of standard of care was compromised when POCT was not universally
deployed. We performed a six-month review of data to compare frequency of critical
values between CCU samples for which POCT was utilized (POCT) and those CCU
samples for which STAT core laboratory testing (CORE) was performed instead, to
assist in management evaluation of whether lesser restriction on use of POCT should
be advocated in this setting.
Methods: POCT and CORE laboratory test results for CCU patients for a six-month
interval were obtained from electronic records. Glucose was excluded from the
analysis because of overlap with extensive use of separate POCT glucose analyzers.
The number and percentage of critical results among POCT and CORE samples were
determined for 8 analytes as follows (analyte (critical ranges)): pH (<7.2 or >7.6);
pO2 (<40 mm Hg); pCO2 (<20 mm Hg or >70 mm Hg); Na (<120 mmol/L or >160
mmol/L); K (<2.5 mmol/L or >6.5 mmol/L); iCa (<3.3 mg/dL or >6.5 mg/dL); lactate
(>3.3 mmol/L); HCT (<20).
Results: POCT data comprised 261 panels of 8 analytes (2088 measurements) from
64 patients. CORE data comprised 11711 measurements from 206 patients. 58.6% of
all 261 POCT panels had one or more critical results, with an overall critical result
percentage of 11.9% among the 8 analytes. CORE test results had an overall critical
result percentage of 3.1%. However, the ratio (R) of absolute number of critical results
for POCT/CORE was 249/367 (R=0.67). This low value for R is likely to be an upper
limit (that is, R is at most 0.67), in consideration of the fact that POCT results are
bundled within a panel rather than ordered independently (i.e., some number of POCT
critical results are likely to be an overcount due to untargeted repeat testing).
Conclusions: CCU use of POCT demonstrated high preselection for critical results.
However, absolute numbers of critical results in the CCU were substantially greater
from among CORE (non-POCT) results (R<1). Given R<1, and in consideration
of the difference in expected turn-around-times between POCT and CORE testing
(nominally, <10 min for POCT vs. up to 60 min for certain CORE analytes), use of
CORE testing in the CCU might be regarded as a lesser service to patients relative to
use of POCT. These findings supported a management recommendation that use of
EPOC POCT in the CCU should be less restricted in order to maintain a more uniform
standard of care.

B-154
Monitoring the Quality of Results from the ARCHITECT Analyzer Using Six
Sigma Metrics Generated by EP Evaluator Software

S. Cox1, S. Westgard2, A. Carrillo3, R. Wonderling3. 1Saint Francis Health

System, Tulsa, OK, 2Westgard QC, Inc., Madison, WI, 3Abbott Diagnostics,
Abbott Park, IL
Quality of lab results has a direct impact on patient care. Six Sigma is a universal
benchmark to measure quality. We used Sigma metrics as a quality indicator
to objectively quantitate the performance of 26 chemistry assays on the Abbott
ARCHITECT c8000 and c16000 instruments. We calculated Sigma metrics using the
equation: Sigma Metric = (TEa - Bias observed) / CV observed Total allowable error (TEa)
was obtained from either CLIA or the RICOS database, our bias compared to our peers
was derived from the BioRad Unity database and CVs were calculated periodically
from our lab results. We have been monitoring Sigma metrics quarterly for over a
year using the EP Evaluator software. This software is user-friendly and allows the
user to input specific TEa goals. We currently compare sigma metrics between 3
analyzers to monitor performance across the instruments. Using laboratory Quality

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S175

Management

Wednesday, July 30, 9:30 am 5:00 pm

Control (QC) data generated for the chemistry assays on the ARCHITECT c8000 and
c16000 instruments, we observed on average, 97% of the assays were greater than 4
Sigma and 71% were greater than 6 Sigma. None of the assays were less than 3 Sigma
and 3% of the assays were between 3 to 4 Sigma. As an example, Sigma metrics for
26 chemistry assays from a quarterly monitoring check on one of the ARCHITECT
c16000 instrument is shown in Table 1. In conclusion, the Abbott ARCHITECT
instruments provide high quality results for 97% of the chemistry assays. Our next
step is to implement streamlined Westgard rules to assist us in reducing the amount of
QC checks currently performed in our lab. We plan to further automate the monitoring
of Sigma metrics when EP Evaluator would have the capability to automatically
download the bias values from the BioRad Unity database.

Background: Turnaround time (TAT) is one of our laboratorys key performance

indicators. We noticed that majority of our Full Blood Count (FBC) tests that exceeded
our 1 hour TAT target were related to manual slide reviews. There seemed to be little
correlation among the complexity of cases or seniority of staff.

Table 1. Six Sigma Metrics assays on one of the ARCHITECT c16000 analyzer (April-June 2013)
Sigma
TEa% Bias% CV% Sigma Test
TEa% Bias% CV5
Metric
Albumin
10
0.4
2.2
4.5
Creatinine
37.28
-0.3
1.8 20.4
Alk.Pkos. 30
2.9
2.0
13.7
GGT
22.7
-7.9
1.9 8.0
ALT
20
5.3
1.7
8.6
Glucose
10
-0.2
1.4 7.1
Amylase
30
-1.3
1.1
25.4
HDL
30
0.7
2.2 13.4
AST
20
2.4
2.4
7.3
Lipase
37.88
-1.4
2.2 16.6
Bili D
45
4.1
5.0
8.1
Magnesium
25
-2.1
3.5 6.6
Bili T
30.69
1.0
4.1
7.2
Phosphorus
10.11
-0.2
2.1 4.7
Calcium
8.29
-2.2
1.4
4.4
Potassium
12.71
0.7
1.4 8.8
Chloride
5
0.4
0.9
5.3
Total Protein 10
0.0
1.3 7.9
Cholesterol 10
-0.1
0.7
13.5
Sodium
3.25
0.7
0.7 4.4
D-LDL
12
-1.4
2.5
4.3
Triglycerides 25
-1.4
1.4 17.4
CK
30
2.2
1.0
27.2
Urea (BUN)
9
-1.7
2.0 3.7
CO2
25
0.4
4.3
5.7
Uric Acid
17
-1.9
1.3 11.4

Results: In August to October 2012, the TAT for FBC is 96.4% (78 cases out of
2175 exceeded TAT target). In November 2013 to January 2014, the TAT for FBC
is 99.2% (22 cases out of 2846 exceeded TAT target). We applied chi square test and
demonstrated a p-value (2-tail) of <0.0000001.

Test

B-156

Method: We reviewed all manual slide reviews for April 2012 (n=651) and grouped
the degree of difficulty into easy, moderate and difficult categories. In each of the
categories, we further analyzed the time taken by our staff based on their job grades:
New hire, Junior Medical Technologist, Medical Technologist and Senior Medical
Technologist. In July 2012, we proposed a recommended slide reading time based
on the complexity of cases and the seniority of staff. All cases exceeding the new slide
review time target were reviewed and retraining given to concerned staff.

Conclusion: We were able to objectively guide staff competency and set slide
reading time according to case complexity and seniority of staff, and propose specific
and targeted retraining. More importantly, we realized a definite and sustained
improvement in our FBC TAT, ultimately providing better patient care.

B-158
Evaluation of Quality OptimiZer quality management software (Awesome
Numbers Inc.) to minimize patient risk, reduce clinical cost and implement EP
23 recommendations.

E. L. Ryan, R. J. Molinaro. Emory Univerity School of Medicine, Atlanta,

GA

Z. C. Brooks1, R. Gerz1, N. Lepage2, D. Plaut3. 1AWEsome Numbers Inc.,

Worthington, ON, Canada, 2Department of Pathology and Laboratory
Medicine, University of Ottawa, Ottawa, ON, Canada, 3Davids Consulting,
Plano, TX

Background: Laboratory test orders placed on specimens after the initial

orders have been placed (addedon orders) can tax resources and are
impacted with reporting delays which may impact patient care. These
added-on orders disrupt normal workflow and can also impact staffing
needs.

Objectives to evaluate the efficacy of Quality OptimiZerTM software to assess analytical

process quality, verify clinical effectiveness of Q.C. processes and recommend a
comprehensive Q.C. strategy, to quantify the impact of OptimiZer Q.C. processes on
patient risk, clinical and laboratory costs, and to compare Quality OptimiZer reports
to EP 23 recommendations.

Managing Emergency Department Add-on Laboratory Orders

Methods: We examined 31 days of Emergency Department (ED) orders at two

institutions to explore differences between initial ordering patterns and the add-on
orders to determine if automation storage and laboratory processes could decrease
these delays. We assessed time to results for both initial orders and add-on orders
on Troponin, Magnesium, and Lipase orders. Time to result for add-on orders was
defined as the time from order electronically placed to results sent electronically, since
the specimen had already been received. Time to result for initial orders was defined
as the time from specimen receipt to results sent electronically. We compared time
to result for add-on orders at the two institutions to determine if automated storage
decreased reporting delays.
Results: Over the 31 days, there were 35,840 ED orders at the two institutions and
627 (1.7%) of these were addon orders. Most of the laboratory add-on orders were
chemistry tests (62%). The following three tests totaled 45% of the addon orders:
Troponin, Magnesium, and Lipase. For each of these three tests, add-on orders were
significantly delayed. One institution had refrigerated storage on the automation line
(Hospital #1) while at the second institution sample storage was off-line (Hospital
#2). For Troponin, an off-line test at both institutions, no difference was observed.
For Magnesium and Lipase, both on-line tests, there was an additional 0.36 to 2.21
hour delay in result reporting for add-on when the sample storage was off-line. For
example, Magnesium addon median time to result at Hospital #1 was 0.57hr (IQR
0.38, 1.07) while at Hospital #2 it was 2.37hr (IQR 1.00, 3.38) p<0.0001. The second
approach to improving addon time to results was a cost-benefit analysis of adding
Magnesium to all comprehensive metabolic panels and only reporting the result in
instances where it was ordered. Depending on the institution, it was estimated $100
- $200 a month in unbillable results would be incurred in order to eliminate the
approximately 1-2 hour delay in result reporting..
Conclusion: In our assessment, the addition of automated storage is the optimal
approach to decrease reporting delays of addon test orders.

B-157
Strategies to Improve Staff Competency and Turnaround Time in Haematology
Slide Review

Relevance: Ineffective quality control practices expose patients to the risk of incorrect
or delayed diagnosis and/or treatment. CLSI EP23 requires labs to ensure test result
quality is appropriate for clinical use; validate the ability of the QC procedures to
detect medically allowable error; and assess potential costs both in terms of the
patients well-being and financial liability.
Methodology: We examined analytical processes, Q.C. processes, patient volumes and
costs from two laboratories x two instruments x five analytes. For each Q.C. sample,
we gathered : A. The four numbers required to evaluate analytical process quality:
1. Measured mean; 2. Measured SD; 3. Peer mean; 4. TEa limit, and B. The three
numbers that determine Q.C. process effectiveness: 5. Q.C. Chart assigned mean; 6.
Assigned SD; and 7. Q.C. rule(s) Quality OptimiZer: - rated analytical process quality
based on Total Error and Margin for Error recommended a 5-part Q.C. strategy simulated a shift that would cause 5% of results to fail TEa limits. - compared the
effectiveness of current and recommended QC processes to detect this significant shift
- quantified patient risk, clinical and laboratory costs of each QC process. We selected
one sodium control to illustrate the importance and interaction of the seven numbers
required to manage quality.
Results For the selected sodium example: - Laboratory practice for all controls on
all tests was to use a 1-2 and 9x rule as warnings, and 1-3, 2-2, 2/3-2, R4 rules as
rejects. - The assigned mean was 0.7 SD below the measured mean; the SD was
assigned at 2.1 x the measured SD. - The OptimiZer Q.C. strategy would detect a
clinically-significant change sooner, prevent risk to 1330 patients, save 66 patients
from clinically-misleading results and result in a net saving of $1,062.00. Quality
OptimiZer reports satisfied EP 23 recommendations to: - ensure test result quality is
appropriate for clinical use, - determine statistical limits that will identify unacceptable
changes in performance of the measuring system, - prove effectiveness of quality
control - quantify patient risks and costs of control quality, - implement and modify
a 5-part Q.C. strategy.
Conclusions OptimiZer Q.C. processes met EP23 requirements, decreased patient
risk, and reduced clinical costs. Error detection is impeded by the common practice
of assigning mean and SD values from inappropriate sources and using outdated Q.C.
rules. Laboratory quality would benefit from increased staff focus on clinical quality
and the interaction of the seven numbers that drive and assess that quality.

L. Lam, J. Lee, L. Tan, C. Soon. Alexandra Hospital, Singapore, Singapore

S176

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Management

Wednesday, July 30, 9:30 am 5:00 pm

and the variability of TEa from different sources can cause inconsistent estimates
of Sigma for the same anlayte. Architect analyzers demonstrated generally high
Sigma values and comparable analyzer-to-analyzer performance. Sigma metrics are
a valuable means for comparing the analytical quality of two or more analyzers to
ensure the comparability of patient test results.

B-159
Generation of statistical protocols derivated from sigma metric in a clinical
laboratory

A. PORRAS1, O. MARTINEZ2, M. F. LEON2, M. ISAZA2, M. ROMERO1,

J. FANDIO1. 1QUIK, BOGOTA, Colombia, 2COLSANITAS, BOGOTA,
Colombia
OBJECTIVE: To demonstrate the usefulness of the sigma metric to generate
statistical protocols to control the analytical performance of measurement systems in
clinical laboratories.
KEY WORDS: statistical protocols, sigma metric, clinically useful results, analytical
performance.
BACKGROUND: Historically, the clinical laboratory has used different strategies for
evaluating the performance of analytical runs since Levy Jennings graphs, statistics
rules on the same graphs and analytical performance limits (CLIA, RilliBAK,
Biological variation, etc.). However, standardizing the selected strategy into a daily
routine becomes a challenge for the clinical laboratory.Recently, some laboratories
in Colombia have invested resources and efforts for the implementation of the
sigma metrics as an indicator of measurement systems efficiency, with the benefit
of including the automatic generation of statistics from the sigma metric protocols,
which has helped to increase the safety of delivering clinically useful results.
METHODS: 1,192 statistical protocols, corresponding to 437 analytes from
chemistry, hormone and coagulation areas from a network of 12 clinical laboratories,
were designed over a period of 12 months with specialized software. At the end
of this period the protocols were ordered and grouped in ranges according to their
sigma metric, defining the protocols most commonly used in each range and finally
generating algorithms to select the protocol to implement.
RESULTS: 6 sigma metric ranges were obtained. The most commonly protocols
obtained for each range of sigma metric were <3 sigma: 13s/22s/R4s/41s/8X (55%)
and 13s/22s(2of32s)/R4s/31s/6x (43%) of cases, for the range 3 to 3.9 :13s/2of32s/R4s/31s/6x
(38%) and 13S/22S/R4S/41S/8X (25%), for the range 4.0 -4.6: 13s/22s/R4s/41s (34%),
13s/22s(2of32s)/R4s/31s/6x (38%), 13s/22s(2of32s)/R4s/31s/8x (25%), for the range 4.6-4.9:
12.5s (72%), 13s (23 %) y 13.5s (5%), for the range 5.0 -5.9: 12.5s (39%), 13S (51%) and
13.5S (10%) and finally for >6: 13.5s (94%).
CONCLUSIONS The use of statistical protocols contributed to the improvement of
performance in laboratory tests achieving an increase of 93% to 96% in the number of
measurements with sigma metrics > 1.96 over a 12 month period. These protocols were
automated in correlation with the sigma metric, thereby decreasing the investment of
time, false rejections and false acceptations produced by inadequate protocols and
poor timing. Additionally, the retrospective analysis led to the conclusion that for
analytes with < 4.7 multi-rules protocols are required and for > 4.7 a single rule
is applicable.

B-160
Sigma metrics to assess analytical quality: Importance of allowable total error
(TEa) target.

D. Armbruster1, S. Westgard2. 1Abbott Laboratories, Lake Villa, IL,

2
Westgard QC, Madison, WI
Background. Six Sigma metrics were used to assess the analytical quality of automated
clinical chemistry tests in a large clinical laboratory and examine the impact of
different allowable total error (TEa) goals on the metric. Clinical laboratories are
challenged to maintain the highest analytical quality but it is difficult to measure it
objectively and quantitatively. Methods. The Sigma metric is estimates quality based
on the traditional parameters used in the clinical laboratory: allowable total error
(TEa), bias, and precision. Sigma metrics were calculated for 41 clinical chemistry
and immunoassay tests, including serum and urine matrices, on five ARCHITECT
c16000 chemistry analyzers. Controls at two analyte concentrations were tested to
calculate precision, bias was estimated as the difference between observed results
and the control target values, and Sigma metrics were calculated using three different
TEa targets (Ricos biological variability, CLIA, and RiliBK) using the following
equation: Sigma = (TEa bias)/CV (all values expressed as %) Results. Sigma metrics
varied with the analyte concentration, TEa target, and between/among analyzers.
Sigma values identified assays that are analytically robust and require minimal QC
rules and those that exhibit more variability, requiring more complex QC rules.
Analyzer-to- analyzer variability was also assessed using Sigma metrics. As an
example, Fig. 1 demonstrates the effect of the TEa target on the Sigma metrics for
albumin.Conclusions. The Sigma metric is an efficient means to measure quality and
optimize QC rules based on observed quality. Lack of TEa targets for some analytes

B-161
Performance of Analysis Teams of Blood Collection after Training in PreAnalytical phase and Phlebotomy

A. G. L. Guimares, J. P. D. Padilha, S. V. Paulino, J. Z. Costa, J. B. L.

Filho, A. C. S. Ferreira. Instituto Hermes Pardini, Vespasiano, Brazil
Background: Pre-analytical conditions are the key factors in maintaining the high
quality of tests results. They are necessary to accurate reproducibility of laboratory
tests for clinical diagnosis. In research at privates clinical laboratories, we evaluated
the impact of some pre-analytical drivers before and after training, in the team of
blood collection.
Objective: Analyze the points of improvement during the process of phlebotomy, and
proposing a tool for continuous improvement that can aggregate in the preparation of
this professionals working in this area. Promote the training of these phlebotomists
and assess how this continuing education can impact not only on the quality of the
tests report, but also the performance of these professionals to the client.
Methods: Pr-Analytical errors are attributed to a lack of patient preparation, care,
collection and identification. Their becoming frequent because, of the low training
and the existence of different degrees of involvement of several people in the process.
The study conducted quantification in percentage of errors (points of improvement)
committed in the collection procedure in three clinical pathology laboratories, in the
South, Midwest and Northeast regions of Brazil, before and after the theoretical and
practical training in venipuncture, following the most current practices described in
the literature. The data were divided into three phases: pre-collecting, collection and
pos-collection.
Results: The request of the patients identity, verification of the test request form,
preparation of client and antisepsis had opportunities of improvement raging from
6-31% before training and 1-7% after the training. The blood collection in a closed
system totaled 75% of the punctures before the training and 86% in the post training,
had 13% and 44% respectively of opportunity for improvement. The time using
tourniquet and homeostasis showed 17% and 60% of errors before training, versus
1% to 4% after. The overall mean compliance before training was 75% and after the
educational activity was 98%.
Conclusion: We conclude that the improvement of the team after the training was
considerable and has an evident gain in the pre-analytical process and the quality of
the request report result.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S177

Management

Wednesday, July 30, 9:30 am 5:00 pm

B-162
Disclosure of Medical Errors: An Approach towards Improving Quality in
Laboratory Medicine

J. Kalra1, J. Steeg2, N. Kalra2, R. Sachdeva2. 1University of Saskatchewan

and Royal University Hospital, Saskatoon, SK, Canada, 2University of
Saskatchewan, Saskatoon, SK, Canada
Background: The quality of healthcare is an emerging concern worldwide. Despite
the advancement in medical field, adverse events resulting from medical errors
are relatively common in healthcare system. We have previously reported a nonpunitive,no-fault model for reporting medical errors in clinical laboratory medicine.
There are several barriers to disclosure including the risk of legal action faced by the
physicians and often strained physician-patient relationship. It also jeopardizes the
opportunity to enhance quality improvement in health care, as many medical errors
are the result of systemic problems that are difficult to detect unless the errors are
reported. However, an appropriate disclosure is vital to overcome these barriers and
manage the consequences of an adverse event.
Methods: In order to analyze the progress made in the area of medical error disclosure
and to understand the rationale for effective error disclosure policies, we reviewed
and evaluated various error disclosure initiatives across Canada and other parts of the
world (Australia, New Zealand and United States of America).
Results: The majority of provincial regulatory bodies in Canada have adopted some
form of disclosure policy. However, these Canadian provincial initiatives remain
isolated because of their non-obligatory nature and absence of federal or provincial
laws on disclosure. In Australia, disclosure policy integrates the disclosure process
with risk management analysis towards investigating the critical events. In New
Zealand, in any adverse event, patients are rehabilitated and compensated through
a no-fault state funded compensation scheme. This disclosure model supports the
health care providers and strengthens the policy of honest disclosure. The United
States Joint Commission on Accreditation of Healthcare Organizations mandated
an open disclosure of any critical event during care to the patient or their families.
By following an open disclosure policy, patients autonomy can be preserved and
malpractice claims can be reduced effectively. The complexities of medical error
disclosure to patients present ideal opportunities for medical educators to probe how
learners are balancing the ethical complexities involved in error disclosure with other
related fields.
Conclusion: The correction of flaws in the healthcare system and the subsequent
protection of patients health should be the industrys top priority, rather than applying
punitive measures to physicians and health care providers who make inevitable errors.
We believe that the disclosure policies can provide framework and guidelines for
appropriate disclosure which can lead to improved quality care and more transparent
practices in clinical laboratory medicine and overall healthcare system. We suggest
that disclosure practice can be improved by creating a uniform policy, centered on
honest disclosure and addressing errors in a non-punitive manner.

B-163
Improvement of pneumatic tube system specimen submission from the
Emergency Department to the Core Laboratory using Lean Six Sigma tools

C. Henemyre-Harris, L. Fujimoto, R. Dotson, J. Waynick, M. Angeles, A.

Brinson. Tripler Army Medical Center, Tripler AMC, HI
Background: Pneumatic tube systems (PTS) can eliminate the need for clinical staff
to leave the ward to deliver specimens and can save time in transporting specimens
to the laboratory. However, once the PTS carrier arrives in the lab, the onerous for
submission compliance rests heavily on lab staff. Implementation of a hospital-wide
PTS resulted in many specimens arriving in the lab without proper order entry. Review
of the PTS usage records indicated that the Emergency Department (ED) submitted
the greatest number of specimens overall and the greatest number of specimens
missing test orders. In addition, ED staff felt that delays in specimen processing were
increasing the overall wait time for ED patients. This management project focused
on strengthening the relationship between the ED and Core Laboratory to both meet
the College of American Pathologists requirement for orders for all lab tests and to
decrease the specimen processing time for ED samples.

identified the extra steps both lab and ED staff took when samples were missing
orders. A root cause analysis was used to identify critical factors relating to the delay
in processing ED samples. Solutions were selected and a new process was defined.
Pilot studies tested the proposed solutions to ascertain improvement significance and
full-implementation potential. Finally, a control/response plan was initiated to sustain
the gains obtained in this management project.
Results: The test orders defect (missing + partial orders) rate of ED specimens
submitted by PTS to the Core Lab was 16.8%. Reworking of ED specimens missing
orders between the ED and Core Lab staff accumulated to one full-time equivalent
(FTE) ED clerk/medical laboratory technician. Value-stream mapping identified some
quick wins in the submission process such as a change in the default setting for label
printers in the ED. The root cause analysis generated three solutions: 1) lab order
entry granted to ED medical support assistants, 2) new employee orientation revised
to ensure new ED staff received the proper laboratory information system (LIS)
signature class, 3) rainbow specimen draw SOP initiated for the ED. These solutions
reduced the specimens submitted by PTS without orders defect rate down to 3.5%.
Conclusion: The LSS structure to this management project provided a forum for open
communication and buy-in from end-users in both the ED and Core Lab. Leadership
support from both departments fostered a culture of change which resulted in a cost
avoidance of one FTE, an enhanced process for specimen submission, and improved
CAP compliance for lab test orders. Lessons learned from this management project
can be applied to other clinics and wards in the medical center.

B-164
Inventory Automation for the Lab and Cost Savings Using an Inventory
Management System

S. Cox1, A. Carrillo2, R. Wonderling2. 1Saint Francis Heath Sysem, Tulsa,

OK, 2Abbott Diagnostics, Abbott Park, IL
Todays laboratories share similar inventory management problems spending too
much time ordering and managing inventory. Low volume usage products, like
calibrators, expire unnoticed requiring urgent orders with overnight shipping expense.
Manual check-in and check-out processes lead to errors. Physical inventories are
required to reconcile discrepant levels. Managing expiration dates and sequestering
new reagent lots requires constant vigilance. Most laboratories are also resourcelimited and technical staff is needed for complex testing issues.
We introduced the Abbott Inventory Manager at Saint Francis Health System to
automate our inventory management. Each item is assigned a Serialized Global
Trade Item Number (SGTIN) tag that is tracked using Radio Frequency Identification
(RFID) technology. This system combines the use of RFID, electronic connectivity
with Abbott and software management to track inventory starting from Abbotts
warehouse to the hospital. Electronic Data Interchange (EDI) connectivity with
Abbott enables our lab to access product catalog, create purchase orders, receive
order acknowledgement and advance shipping notices. The software monitors onhand stock levels, lot numbers, expiration dates with critical alerts and triggers
auto-replenishment suggestions. It configures user roles and approval levels and also
maintains an audit trail. Since Inventory manager is an open system, we are able to
include non-Abbott items as part of the inventory.
Post implementation, our hands-on time of unpacking, labeling and logging-in our
inventory has been reduced by 60% and our error rate has declined from 27% to 1.1%.
Table 1 reflects the annual lab staff labor savings of $21,606.47 as documented during
pre and post implementation. We anticipate, after one year, to see additional savings
due to reduction of expiring reagents and emergency overnight orders.
In conclusion, automation of inventory management in our lab has minimized manual
intervention of labeling, counting, tracking and ordering inventory. We reduced our
error rate and realized both time and cost savings.

Objective: To assess and improve the EDs process for submission of specimens to
the Core Lab via PTS.
Methods: The define-measure-analyze-improve-control (DMAIC) tollgates of a Lean
Six Sigma (LSS) project were used to identify and implement solutions for the ED to
submit specimens with electronic orders. Data captured by lab staff estimated the error
defect rate for ED specimens missing test orders. Process and value-stream mapping

S178

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Management

Wednesday, July 30, 9:30 am 5:00 pm

and WB-TBil was 20% (0.5 mg/dL). Conclusion: Careful workflow analysis and reengineering of transport and analytical process for key laboratory tests significantly
reduce median overall TAT to a ~20 min that will facilitate timely delivery of
chemotherapy.

Table 1. Lab Staff Labor Savings Pre and Post Implementation of the Inventory
Management System
Dept.

Activity

Receiving
Dock
Receiving
Dock

Physical
Count
SAP Goods
Receipt
Receive
Lab
Product into
Lab
Consume
Lab
Products
SAP
Lab
Decrement
Physical
Lab
Inventory
Check/Order
Lab
Products
Yellow
Lab
Sticker
Generation
Transport to
Receiving
Main Lab
Dock
(CC-IA)
Transport to
Receiving
Lab Storage
Dock
(Heme)
Total Annual Hours and
Dollars

Pre IMS
Post IMS
Annual Annual Annual Annual
Labor Labor Labor Labor
Hours Dollars Hours Dollars

Savings
Annual
Dollars

57.4

$969

8.2

$138

$831

86%

12.2

$206

18.3

$309

($103)

-50%

306.5

$9,195 121.4

$3,642

$5,553

60%

T. K. Jackson1, H. K. Lee2. 1Geisel School of Medicine at Dartmouth,

Lebanon, NH, 2Dartmouth-Hitchcock Medical Center, Lebanon, NH

27.0

$809

22.5

$675

$135

17%

44.9

$1,347 12.0

$359

$988

73%

261.7

$7,850 24.9

$746

$7,104

91%

65.5

$1,965 33.2

$996

$969

49%

Background: Our laboratory tests patient specimens for latent tuberculosis

using a test kit to measure cell mediated immune response. Latent
tuberculosis is non-communicable and asymptomatic, but can develop into
active tuberculosis at a later time and therefore is important to diagnose to
prevent disease development.

204.3

$6,130 0.0

$0

$6,130

100%

19.4

$328

19.4

$328

$0

0%

6.9

$116

6.9

$116

$0

0%

B-166
The Effect of Patient Immune Status on QuantiFeron Test Results: An
Investigation

The test is a two-part process. The patients blood is drawn into a series of three tubes:
a tube containing tuberculosis antigen, a tube containing nothing, which serves as a
negative control, and a tube containing mitogen, which is a non-specific stimulator of
T-cells to trigger a gamma interferon response and serves as a positive control. The
tubes are incubated at 37oC for 16-24 hours and then the plasma is analyzed using an
ELISA test. There is a software calculation against a standard curve that determines
IU/mL of the nil, TB antigen and mitogen tubes. The results are then analyzed
compared with a variety of permutations to determine if the results are positive,
negative, or indeterminate. Infectious disease providers are interested in tracking the
individual IU/mL results that the software uses in the interpretation and therefore the
nil, TB antigen minus nil, and mitogen minus nil results are reported in addition to the
qualitative interpretation.

1,218.5 $28,915 254.76 $7,308.53 $21,606.47 75%

B-165

New lots of kits are checked for similar reactivity with specimens that have already
been tested with the kit currently being used prior to patient testing.

Re-engineering Critical Laboratory Testing for Timely Chemotherapeutic

Management.

X. Yi, R. Wolsky, C. VanSlambrouck, D. Mika, J. Leanse, E. K. Y. Leung,

C. Nabhan, K. T. J. Yeo. The University of Chicago, Chicago, IL
Background: Delivery of cytotoxic therapy is a complex multifaceted process
that involves harmonized collaboration between all systems involved. Laboratory
assessment is an essential component of assuring that patient care is efficient,
timely, and accurate. As the volume of oncology patients is increasing, providers and
patients are faced with long wait times until laboratory results are available before
chemotherapy can be safely administered. Optimizing laboratory turn-around time
(TAT) assures timely delivery of chemotherapy, which would subsequently translate
into improved outcomes and satisfaction. In this study, we aimed to investigate how
to reduce the laboratory TAT for key laboratory tests so as to optimize the timely
administration of chemotherapy. Method: The TAT can be affected by several factors
including phlebotomy, clinical laboratory distribution, and test operation. We collected
time data in each step in this process, including the time from specimen collection to
specimen receiving in the central laboratory (Col-Rcv), specimen receiving to result
release (Rcv-Res), and the overall TAT from specimen collection to result release
(Col-Res). Results: The median TATs before our process re-engineering were: 31 min
(Col-Rcv), 40 min (Rcv-Res) and 74 min (Col-Res) for a comprehensive metabolic
panel (CPNL), and 34 min (Col-Rcv), 9 min (Rcv-Res) and 50 min (Col-Res) for a
complete blood count (CBC). After reconfiguring the specimen transfer and analytical
workflows to have specimens tubed directly to the central laboratories and assayed
immediately, the CBC showed a significant reduction in the median (Col-Res) TAT
of 20 min (p<0.0001). For CPNL, by processing all specimens as STAT samples, the
median (Col-Res) TAT was significantly reduced from 74 min to 54 min (p< 0.0001).
To investigate if we can further improve on the (Col-Res) TAT for chemistry tests, we
engaged our clinical colleagues who identified 2 key CPNL analytes, total bilirubin
(TBil) (to monitor liver toxicity) and creatinine (Creat) (to monitor renal toxicity),
that, if we are able to deliver the results within the same timeframe of the CBC, would
allow for faster decision-making for drug infusions. We evaluated the suitability of
the whole blood ABL 8000 analyzer (Radiometer, OH) for this purpose and found that
both whole blood TBil (WB-TBil) and Creat (WB-Creat) can be routinely resulted
within 2 min of sample introduction. Correlation studies (Passing-Bablok and BlandAltman plots) between whole blood and plasma samples from this cancer population
showed: [ABL WB-TBil] = 2.00 [Roche plasma TBil] 0.60, (range 0.1-2.1 mg/
dl, n=164) and [ABL WB-Creat] = 1.08 [Roche plasma Creat] 0.04 (range 0.4-3.1
mg/dl, n=166). There was 90% concordance of WB-TBil results when compared to
plasma TBil results <1.1 mg/dL. The CV for WB-Creat was <10 % (0.43 mg/dL),

It was noticed in June of 2013 that, although the lot-to-lot check had been acceptable,
the newer lot of kit being used was yielding more indeterminate and low-reactive
results. It was important to determine if this was due to a problem with the testing kit
or the immune status of the patients.
Methods: All results over a nine month period were tracked. Any patient with an
indeterminate or low-reactive result (mitogen minus nil of less than 1.0) was checked
for a diagnosis suggesting immunocompromised status.
Results: From May 2013 until January 2014, a total of 1648 samples were tested. Of
those, there were 44 (2.67%) indeterminate results and 14 (0.85%) negative results
due to low mitogen reactions. Case histories of these patients revealed that in fact, they
were immunocompromised and the test results were concordant with clinical findings.
All but one result were explained by the patients being immunocompromised due to
various disease states. The one that could not be explained was a pre-employment
sample with no charted information.
Conclusion: In conclusion, it has been determined that the increase in indeterminate
results that was observed was due to immunocompromised patients and not suboptimal test kits. For patient safety and quality of results, it has been decided that we
will continue our practice of tracking the indeterminate and low-reactive results to
ensure that the immune status of the patient is the cause and not a problem with kit or
sample tube integrity, particularly because infection disease providers are using the
calculated values in their clinical decision-making.

B-167
Use of a Decision Matrix and Positivity Rates to Substantiate Test Scope:
Propoxyphene - A Case in Point

L. Labay, S. Noel, M. Welsh. NMS Labs, Willow Grove, PA

Background: Any toxicology test ordered should have value to the individual
requesting the test and to the patient. It is, therefore, critical to optimize analytical
scopes so that relevant testing is performed. Otherwise, there is consumption of
resources with little value added. Since reference laboratories often serve wide
client bases with competing needs, it is advantageous to have an objective means of
determining if an analyte should be maintained in a particular test. A decision matrix
(DM) was created to quantify the considerations involved, and used to evaluate the
necessity of maintaining propoxyphene in limited-scope immunoassay panels.
Propoxyphene was approved by the US FDA in 1957 for use as an opioid analgesic, but
was withdrawn in Nov 2010 due to risk of serious or fatal heart rhythm abnormalities.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S179

Management

Wednesday, July 30, 9:30 am 5:00 pm

The drug was frequently prescribed, and associated with abuse, overdoses and lethal
outcomes. While past toxicological relevance cannot be argued, is it appropriate to
now remove propoxyphene from immunoassay tests or should the practice continue?
To address this question, the DM was used in conjunction with a multi-year positivity
rate evaluation.
Methods: The toxicology panels evaluated were immunoassay based (ELISA or
EMIT) and used for several matrices. DM assessment criteria with the associated
factors were: a) legal availability in the US [yes:5; no:1], b) legal world-wide
availability [yes:5; no:1], c) length of time unavailable [>10 yrs:1; 5-10 yrs:2; 2-4
yrs:3; 1-2 yrs:4; <1 yrs:5], d) illicit drug sources [yes:5, no;1], e) current positivity
rate [75-100%:5; 50-74%:4; 25-49%:3; 1-24%:2; <1%:1] and f) other testing means
within the laboratory [yes:1; no:5]. The factors for propoxyphene were entered for
each matrix type to obtain the Total Score (TS). Using the DM, the minimum and
maximum possible TS are 6 and 30, respectively. The recommendations based upon
the TS are: 6-9 (remove from scope), 10-19 (monitor) and 20-30 (maintain in scope).
An evaluation of positivity rates over a 3.5 year period was also performed.
Results: The TS for propoxyphene was 8 out of 30 in blood, serum/plasma, urine,
tissue, hair, stool and meconium, and 9 out of 30 for fluid. Positivity rates by matrix
type by year, from 2010 to mid-2013, were: blood (2.03%, 0.32%, 0.18%, 0.10%);
serum/plasma (1.79%, 0.21%, 0.13%, 0.0%); urine (2.65%, 0.56%, 0.28%, 0.22%);
tissue (1.48%, 0.86%, 0.0%, 0.0%); fluid (3.86%, 1.62%, 1.56%, 2.32%); hair
(1.19%, 0.81%, 0.0%, 0.0%); stool (7.5%, 0.0%, 0.0%, 0.0%); meconium (19.62%,
22.45%, 9.53%, 0.0%).
Conclusion: Using the above determinations, our working group made two
recommendations: 1) remove propoxyphene from the immunoassay panels and
2) treat propoxyphene in a similar fashion to most other drugs by using a broadspectrum screening approach (TOF or GC/MS) or a directed analysis when warranted
by case history. The severe decline in positivity rates suggests that the continual need
to monitor the drug is no longer required. When the challenge of assessing scope
relevancy arises, it is prudent to consider all available data including positivity rates
and to employ a DM to substantiate any operational decisions.

B-168
Comparison of Inpatient and Outpatient Genetic Test Utilization in a Pediatric
Tertiary Care Center

(cost savings of $550 vs. $430 per order reviewed). The error rate, which was
calculated as a percentage of tests that were either cancelled or modified because an
incorrect test was ordered, was not significantly different, at 6.8% for inpatient and
5.1% for outpatient genetic test orders (p=0.52).
Conclusion:The rate of test approval and proportion of positive results was similar
between inpatient and outpatient genetic tests orders, although cost savings per
inpatient order was higher. The order error rate of greater than 5% for both patient
groups, however, suggests that maximizing the proportion of genetic test orders under
review would prevent errors in diagnosis and patient management.

B-169
QC Rules for High Sigma-Metric Processes

C. A. Parvin. Bio-Rad Laboratories, Plano, TX

Background: Objective: Define a simple, yet effective QC rule for high sigma-metric
processes and determine the minimum sigma-metric threshold for the rule.
Relevance: High sigma-metric processes are generally easy to quality control. A
simple effective QC rule that can be confidently applied in all high sigma-metric
situations would be valuable.
Methods: QC rejection limits are defined as f*TEa (a fraction of the allowable
total error specification for the analyte). Probabilities of false rejection, Pfr, and
critical systematic error detection, Ped(SEc), were derived assuming either 2 or 3 QC
concentration levels are evaluated. SEc is defined as an out-of-control condition that
would produce 5% patient results containing measurement error exceeding TEa. QC
rule rejection limits were required to provide Pfr < 0.01 and Ped(SEc) > 0.90.
Results: Validation: Mathematically derived probabilities were validated by computer
simulation. Conclusion: The range of values for f that meet the false rejection and
error detection constraints when 2 QC concentration levels are evaluated as a function
of sigma-metric value is shown in the figure. A QC rule with rejection limits given
as 0.6*TEa provides as least 90% critical systematic error detection with a false
rejection rate less than 1% for any process with sigma-metric value greater than 5.3.
The false rejection rate of the QC rule will decrease and the error detection probability
of the rule will increase for sigma-metric values greater than 5.3.

P. C. Mathias1, J. H. Conta2, D. L. Sternen2, M. L. Astion2, J. A. Dickerson2.

1
University of Washington, Seattle, WA, 2Seattle Childrens Hospital,
Seattle, WA
Background: Increasing rates of send out testing represent a growing financial burden
for hospital laboratories. In the pediatric tertiary care setting, genetic testing for rare
inherited diseases is costly, and many such tests are infrequently ordered, which
increases the probability of an order error. In addition, these tests are occasionally
ordered on inpatients, so their cost is not reimbursed. In an effort to better utilize
resources, some medical centers are actively monitoring test utilization for tests
fulfilling certain criteria such as cost. The goal of this work was to analyze the data in
a laboratory utilization database and compare genetic tests ordered in both inpatient
and outpatient settings to determine if inpatient genetic test orders should be treated
differently than outpatient orders.
Methods: A laboratory-generated test utilization database maintained by faculty
and genetic counselors containing over 750 orders for genetic tests over a period
of more than 2 years was analyzed. Send out tests were recorded in the database
if they met the following criteria: 1) cost > $1000, 2) multiple genetic tests on one
requisition, 3) request to send to nonpreferred laboratory, or 4) request to send out
a test performed in-house. At the time of order, patient demographics (including
inpatient vs. outpatient), test information, ordering provider specialty, and test cost
were recorded. After review by the appropriate staff, the approval or modification of a
test (including indications for doing so), cost savings, and the results of the tests were
also entered into the database. Based on the indication for ordering the test, results
were classified as positive, negative, uncertain significance, or pending. For tests
cancelled or modified after review, orders determined to be non-indicated or ordered
incorrectly were recorded. Data on inpatients and outpatients was compared and the
proportions of test approvals, positive results, cost, cost savings, and error rates were
compared, with tests for equality of proportions applied when relevant.
Results: Data from 147 inpatient and 632 outpatient genetic test orders was analyzed.
The rate of test approval without modification was similar for both groups, at 66% for
inpatient orders and 71% for outpatient orders (p=0.27). The proportion of positive
results was also similar between the two groups, at 29% for inpatients and 27% for
outpatients (p=0.78). The mean cost for inpatient genetic tests reviewed was $660
higher and the mean cost savings per order reviewed was $120 higher in inpatients

S180

B-170
Total Laboratory Automation (TLA) to Improve Efficiencies: Before and After
study at a 2,941 Bed Medical Center in Taiwan.

P. Liu, M. Shieh, W. Liu, C. Lin, M. Lee. Department of Pathology and

Laboratory Medicine Central Lab, Taipei Veterans General Hospital,
Taipei, Taiwan
Objective: To assess the increase in efficiency of TLA by implementing the
ACCELERATOR APS to replace multiple separate work processes including better
Turn Around Time goal achievement, optimized workflow, reduction in FTE, reduced
consumables and increased space utilization.
Relevance: ACCELERATOR APS is an automated track system that manages
pre-analysis, analysis, as well as post-analysis operations. It centralizes all these
operations.
Methodology: This study analyzed one months data of % that achieved the TAT
goal and consumables usage of 29 clinical chemistry assays for STAT outpatient

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Management

Wednesday, July 30, 9:30 am 5:00 pm

sample from laboratory information system. The number of working steps and staff
requirements were obtained by workflow observation and space require, all metrics
performed before and after implementation.
Validation: The facility has a TAT goal of 60 minutes for 29 STAT Chemistry assays
for outpatient sample. The percentage of TAT achievement and consumables usage
were analyzed using LIS data. Working steps and personnel required were collected
via workflow observations. Space utilization is from the calculations done by
engineers.
Results:
Before
After
% of increase and
Metrics
Implementation Implementation decrease
No. of tests
27,866
32,286
15.8%
% Achieved TAT goal
96
99.3
3.3%
Working steps
21
5
-76.0%
Personnel
10
4
-60.0%
Sample tube usage
26.058
25,608
-1.7%
Aliquot tip and cup usage 5,211
0
-100%
Space required (sq. meter) 1,397
597
-57.2%
Conclusion: The implementation of the ACCELERATOR APS to combine preanalysis, analysis and post-analysis in one platform, resulted in an improvement
of laboratory efficiencies in all key metrics. This was achieved despite the 15.8%
increase in testing volume observed after implement of the ACCELERATOR APS.
3.3% increase in tests achieving their TAT goal
76%, 60%,1.7% and 52.7% are reduction in working steps, personnel, sample tubes
and space required respectively
100% elimination of aliquot tips and cups due to consolidate testing.

B-171
Use of Technological Advances to Improve Laboratory Turn Around Times

P. B. Kyle, K. B. Brown. University of Mississippi Medical Center,

Jackson, MS
Background: Timely laboratory results can be essential for quality patient care. As
such, laboratory turnaround times (TAT) are often cited as performance measures
and quality indicators. We sought to tabulate changes in TAT observed after the
implementation of new processes and technologies into our laboratory. These included
personalized productivity reports, status display monitors, result autoverification,
preanalytical automation, and the use of plasma instead of serum.
Methods: The TAT of basic chemistry profiles were determined before and after each
change was implemented. The TATs of samples analyzed from 0700-1530 weekdays
only were tabulated as per institutional review board approval at the University of
Mississippi Medical Center, Jackson, MS. Each experimental group included over
4000 samples collected during a given eight week period. A minimum of two weeks
were allowed to pass in order for technologists to adapt to each change, whereas
three months were allowed for the acclimation to new instrumentation. After
baseline measures were collected, the first change involved daily dissemination of
individualized performance reports. Each report included the number of samples
analyzed and average TAT of STAT and routine results of each technologist. Secondly,
an LCD monitor was installed within view of the technologists. The LCD monitor
displayed the status of all pending samples. Next, the results of autoverification was
recorded where only normal patient results were autoverified. The TAT changes
associated with preanalytical automation and use of plasma were also tabulated.
Results: The greatest reduction in TAT, 13.6 minutes, was obtained after the
implementation of preanalytical automation. Reductions of 6.81, 6.65, and 5.99
minutes, respectively, were noted with implementation of autoverification, the
status display monitor, and productivity reports. Autoverification resulted in the best
cost:benefit ratio at no cost, whereas autoverification was the most costly at hundreds
of thousands of dollars.
Conclusion: Preanalytical automation afforded the greatest reduction in TAT, but
also came with significant monetary and space requirements. Laboratories can
significantly reduce their TAT via several methods inexpensive methods. Solutions
include autoverification, personnel productivity reports, LCD displays, and alternate
sample collection tubes.

B-172
Eliminating non-value added pre-analytic processes to improve laboratory
turnaround time and patient safety for emergency room patients

N. I. Rush1, A. Planker2, M. K. Zimmerman1. 1Indiana University School of

Medicine, Indianapolis, IN, 2Indiana University Health, Indianapolis, IN
Background: Our hospital system has been looking at various ways to eliminate
non-value added processes to improve patient care and decrease costs. One cause of
emergency department (ED) patient dissatisfaction is extended length of stay (LOS)
from delays in test ordering to test result availability time. A joint ED and laboratory
team sought to decrease LOS by utilizing value stream mapping (VSM). VSM is
a lean manufacturing method to analyze and improve the flow of information and
materials across a process. We focused on an ED with no stat lab: all specimen testing
outside of a very limited ED point-of-care menu was sent via pneumatic tube to the
main lab. In the main lab, we utilized VSM to optimize the flow of ED specimens
from pneumatic tube receipt to automated chemistry or hematology line log-in time.
Methods: The measuring timeframe criteria for the laboratory portion of the
project were established. The tests tracked were the basic metabolic panel, complete
metabolic panel, hepatic panel, lipase, and complete blood count. Measurements were
(1) laboratory pneumatic tube station receipt to stat specimen log-in (SL) time, (2)
SL to automated chemistry line log-in (CL) or hematology line log-in (HL) time.
Percent of ED relabeled (%R) tubes was determined by manually tracking relabeled
specimens over 24 hour periods. Specimen processing personnel motion was observed
and spaghetti charts were made. After corrective interventions, times for (1) and (2),
%R, and spaghetti chart walk paths were reassessed.
Results: The initial state showed that SL was 2:00 minutes, CL was 13:20 minutes,
HL was 05:20 minutes, and %R was 46%. Major causes of pre-analytical phase delay
were lack of prioritization for stat specimens, issues with label placement on specimen
tubes, and work interruptions and non-value added motion by specimen processing
personnel. The intervention for ED personnel was training on correct label positioning
techniques using visual aids to demonstrate the spatial relationship between specimen
label, automated line bar code reader, and specimen puck. The specimen processing
area interventions included establishment of a STAT bench, establishment of a separate
Problem bench to eliminate unnecessary interruptions, and designated specimen tube
runners. After intervention, SL was 00:48 minutes (60% decrease); CL was 09:20
(30% decrease); HL was 03:44 (30% decrease); and %R was 0 (100% decrease). A
comparison of the initial state spaghetti chart to that of the confirmed state showed an
87% decrease in walk paths. The projected annual monetary savings for the laboratory
portion was $33,396.
Conclusions: Small yet significant interventions implemented via this VSM project
by ED and laboratory personnel improved the specimen flow from the ED to the
laboratory automated lines. The interventions in the current study decreased preanalytical phase delays, improved patient safety and satisfaction, and resulted in
potential monetary savings.

B-173
Quantifying of the Cost of Unnecessary Clinical Laboratory Testing for
Hospital Systems and Healthcare Payers

S. Chekuri1, A. Mandel2, C. Hightower2, B. Brimhall1. 1University of

Mississippi, Jackson, MS, 2CareCore National, LLC, Bluffton, SC
Background: Diagnostic laboratory testing represents the largest source of structured
medical information in most healthcare systems. Testing with little or no diagnostic
value is a significant source of unnecessary cost to the health care system. Using data
available from two hospital systems (>15 hospitals) and four large regional payers
(>6 million covered lives) we evaluated testing patterns for several diagnostic test
scenarios to quantify the cost of unnecessary testing.
Methods: Using available evidence-based medical knowledge we determined test
scenarios for which at least one test was largely unnecessary. We aggregated 14
test scenarios into 7 diagnostic categories: thyroid (e.g., concurrent free T3 and/
or free T4 given TSH within the reference range), liver function (e.g., concurrent
GGT and alkaline phosphatase given ALKP within the reference range), suspected
pancreatitis (e.g., concurrent serum amylase and lipase), vitamin D status, iron status,
general inflammation (e.g., concurrent CRP and erythrocyte sedimentation rate), and
myocardial injury (e.g., concurrent troponin and CKMB).
Combining laboratory, operational, and detailed financial data we calculated costs
based on available cost accounting categories for each hospital system. We also
queried claims databases for four US regional payers and extrapolated ordering

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S181

Management

Wednesday, July 30, 9:30 am 5:00 pm

patterns from hospital data. Hospital data and payer data covered at least 12 months
(24-48 months in most cases) during the period 2008-2013 for hospitals and 20092012 for payers.
Findings: The mean annual volume of unnecessary testing in the hospital systems was
270,517 tests representing variable plus labor costs of $1,824,805 and total costs of
$2,514,822. For payers, the mean annual volume of the same unnecessary testing was
1,261,817 tests worth $30,497,028 in net paid out claims (see table).
Conclusion: The cost of unnecessary testing is substantial to both hospitals and payers.
There is great opportunity for laboratory professionals to quantitatively demonstrate
value by improving test utilization patterns.
Annual volume and cost of unnecessary testing for hospitals (a) and payers (b)
Diagnostic
Variable
Volume(a)
Total Cost(a) Volume(b) Paid Claims(b)
Category
Cost(a)
Thyroid
109,209
$645,389
$913,858
536,097
$12,252,830
Liver
17,542
$80,989
$109,361
30,963
$458,420
Function
Suspected
23,179
$118,793
$164,338
172,798
$3,606,089
Pancreatitis
Vitamin D
4,610
$175,018
$229,000
33,864
$2,166,743
Status
Iron Status 27,292
$145,967
$202,811
145,186
$4,503,433
General
31,108
$194,863
$298,743
196,372
$1,868,772
Inflammation
Myocardial
63,077
$463,786
$596,711
146,537
$5,640,741
Injury
TOTAL
270,517
$1,824,805 $2,514,822 1,261,817 $30,497,028

B-174
Sample Size Requirements for QC Lot Cross-Over Studies

C. A. Parvin. Bio-Rad Laboratories, Plano, TX

Background: Objective: Determine the number of QC results required in a cross-over
study to provide a good estimate for the mean (and possibly SD) for a new lot of QC
material.
Relevance: Establishing the mean concentration for a new lot of QC material is an
important laboratory activity. A major concern is that a poorly estimated QC mean will
lead to a QC rule that gives too many false rejections. Designing cross-over studies
that provide good estimates with a minimum number of replicates is important.
Methods: Simulations were employed to estimate the influence of cross-over study
sample size on a QC rules false rejection rate when the QC rule is based on estimates
obtained from the cross-over study. Three cases are considered; 1) the QC mean is
estimated, but the SD is known, 2) the QC mean is estimated and SD is computed
as estimated mean * known CV, and 3) both QC mean and SD are estimated. The
outcome metrics evaluated are the expected probability of false rejection, E(Pfr),
when cross-over study estimates are used in the QC rule, and the probability the false
rejection rate is not greater than twice the false rejection rate of the QC rule with
known QC mean and SD. One million simulations were employed to obtain estimates.
Results: The table shows some of the results.
Conclusion: If the SD or CV are known and only the QC mean is estimated, then a
cross-over study based on N=10 results can provide a QC rule that has at least an 80%
chance that the false rejection rate of the rule will be no greater than twice the false
rejection rate using the true QC mean and SD. If SD must be estimated from the crossover study, then at least 60 QC results are required.
1-3s/2-2s/R-4s QC Rule
(Using known mean and SD, Pfr = 0.01)
Case
N
SD known
7
SD known
10
SD known
14
SD known
21
CV known (10%) 7
CV known (10%) 10
CV known (10%) 14
CV known (10%) 21
SD estimated
20
SD estimated
40
SD estimated
60
SD estimated
80

S182

E(Pfr)
0.017
0.015
0.013
0.012
0.018
0.015
0.013
0.012
0.025
0.016
0.014
0.013

B-175
Investigation on Causes and Impact of Specimen Rejection in Clinical
Chemistry Laboratory

M. Chen, R. Phipps, R. Del Guidice, E. Wagar, Q. H. Meng. University of

Texas MD Anderson Cancer Center, Houston, TX
Background: Accurate results are critical to patient management and
quality care. Errors occurring in the pre-analytical phase account for up
to 90% of total laboratory errors. Of those, inappropriate specimen due to
quality and quantity account for 70%. Pre-analytical specimen integrity and
quality are extremely important to the final result reported by the laboratory.
Specimen rejection is conducted to ensure accurate identification and
quality of samples as well as the accurate results. Unlike quality control
systems widely applied to ensure the quality of the analytical phase, quality
improvement intending to reduce pre-analytical errors as part of total
quality management has not been fully implemented and achieved. This
study was conducted to investigate the frequency, causes, and impact of
specimen rejection in clinical chemistry laboratory.
Methods: Rejected chemistry specimens in our laboratories recorded in LIS during
a 4-month period were collected. The number of rejected specimens, reason, and
location for rejection were analyzed.
Results: Of the 245,058 chemistry specimens received to the laboratories during the
data collection period, 647 (0.26%) were rejected. The most common reasons for
specimen rejection were contamination (IV fluid or TPN) (227, 35.1%), unacceptable
specimen (wrong collection tube, unlabeled, mislabeled, or inappropriately labeled
specimen) (n=179, 27.7%), quantity not sufficient (QNS) (n=98, 15.1%), hemolysis
(n=61, 9.4%), clot (n=60, 9.3%), and patient ID error (n=14, 2.2%). The most
commonly affected analytes due to specimen rejection occurring after collection
were glucose (n=192, 8.82%); calcium (n=153, 7.0%), magnesium (n=148, 6.8%),
potassium (n=138, 6.3%), creatinine (n=101, 4.6%), and BUN (n=96, 4.44%).
Inpatient areas had the most rejections occurring after collection (45.3%), followed by
outpatient areas (24.6%), adult Intensive Care Units (ICUs) (17.2%), and ER (11.0%).
When compared with the respective frequency with which they collect specimens,
laboratory personnel (phlebotomists) submitted significantly fewer rejected specimens
than other in-hospital personnel groups (at a rate of 0.22% vs. 1.34% for floors often
collected by other in-hospital personnel groups. Total recollected specimens (n=230,
0.09%) during a 4-month period added an average of 108 minutes delay (from
recollect order placed to results completed) to the turnaround time per test. The cost
for total of 230 specimen recollection was $5510.80 USD.
Conclusions: Specimen rejection criteria should be followed and specimen rejection
should be monitored on a regular basis. Those frequent factors that are associated
with rejection and have a great impact on patient results and patient care should
be identified. Actions and education should be taken for quality improvement.
Efforts should be made to standardize laboratory manuals and procedures as part
of continuous quality improvement program. Policy and procedures specifically to
specimen requirement, collection, transportation, and preparation should be strictly
followed.

P(Pfr< 2 * 0.01)
0.75
0.86
0.93
0.98
0.73
0.83
0.90
0.96
0.55
0.71
0.81
0.86

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Electrolytes/Blood Gas/Metabolites

Wednesday, July 30, 9:30 am 5:00 pm

Wednesday, July 30, 2014

Poster Session: 9:30 AM - 5:00 PM
Electrolytes/Blood Gas/Metabolites

B-176
Validation of a spectrophotometric assay for the measurement of pyruvate in
whole blood.

S. Albeiroti, J. M. El-Khoury, S. Wang. Cleveland Clinic, Cleveland, OH

Background: Pyruvate plays a major role as an intermediate in carbohydrate and
amino acid metabolism. Clinically, an elevated Lactate-to-Pyruvate ratio can be
indicative of several diseases including disorders of mitochondrial metabolism
and inborn errors of metabolism. In our laboratory, pyruvate is measured by a
spectrophotometric assay. NADH used in the assay was prepared freshly before
each test by adding Trizma buffer to a 2 mg NADH bottle from Sigma Aldrich.
Inconsistency in absorbance values among different NADH bottles provided by the
manufacturer has been a major challenge in our lab. Objective: To develop a pyruvate
assay that requires shorter processing time and results in more consistent absorbance
values. Methods: NADH stock (0.225 mg/ml) was prepared in 1.5 mmol/L Trizma
and stored at 4C. Deproteinized blood sample was prepared by adding 2 mL whole
blood to 4 mL cold 12% trichloroacetic acid. The mixture was mixed, incubated for
10 minutes at 4C and centrifuged for 7 minutes at 2000 g. 2 mL of the supernatant
was mixed with 1 mL 0.225 mg/ml NADH in a 1-cm cuvette. Using Beckman DU
800 spectrophotometer, absorbance was measured at 340nm. 50 L of 1000 units/mL
LDH was then added and absorbance was measured after 10 minutes of incubation.
Results: NADH stock solution was stable at 4C over a period of 60 days with an
inter-assay precision of 1.8% for the pyruvate measurements in prepared QCs. QCs
used were 0.07 mmol/L pyruvate in 5% BSA in saline. The assay had an analytical
measurement range of 0.04-0.35 mmol/L with recoveries ranging from 89% to 105%.
Average within-run precision was <1%. Pyruvate results of patient samples obtained
by this assay (n=12) were comparable with results obtained by the original method
(r=0.9891, slope=0.986, intercept=-0.0008, SEE=0.0058). Conclusion: The modified
spectrophotometric assay for the measurement of pyruvic acid in whole blood is
accurate and robust.

B-177
Validation of creatinine test using the standard (SRM967) as reference for two
methodologies

A. L. Camilo, F. S. Fukuoka, C. F. A. Pereira. DASA, So Paulo, Brazil

Background: The ADVIA Creatinine_2 is used for in vitro diagnostic
determination of human serum, plasma (lithium, heparin) and urine
creatinine activity on the ADVIA Biochemical Systems (Siemens
Healthcare Diagnostics). These measurements are used in renal diseases
diagnosis and treatment, and also in renal dialysis monitoring. Objective:
This study aims to perform the validation of Creatinine_2 (CREA_2)
ADVIA assay on ADVIA 2400 equipment using as reference the standard
(SRM967) of the National Institute of Standards and Technology (NIST).
Moreover, the study verifies if results obtained by the method had deviation
no greater than the analytical specifications.
Methods: The Creatinine_2 (CREA_2) method is based on the reaction of picric acid
with creatinine in an alkaline medium, as described in the original Jaffes procedure.
The assay range is 0.1 - 25.0 mg/dL (8.84 - 2,210 mol/L) to serum samples and 1.5
- 300 mg/dL (133 - 26,250 mol/L) to urine samples.
Results: Data show the total imprecision of measurements using control material and
prepared pools. The within-run assay obtained CV=0.74% to 2.6%, and the total assay
obtained CV=2.55% to 5.01% on samples serum with creatinine concentrations of

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S183

Electrolytes/Blood Gas/Metabolites

Wednesday, July 30, 9:30 am 5:00 pm

0.48 to 5.81 mg/dL. The within-run assay obtained CV=1.38% to 3.05%, and total
assay obtained CV=2.95% to 6.84% on samples urine with creatinine concentrations
of 0.73 to 235.12 mg/dL. Dilution linearity of multiple serum samples demonstrated
means recoveries of 95.46% - 97.41%. Linearity test by applying eleven (11)
points of dilution factors using the Standard Reference aqueous solution (SRM967)
of the National Institute of Standards and Technology (NIST) at the following
concentrations: 1.0 mg/dL and 30.0 mg/dL and the regression linear equation to
this study was y = 0.956x + 0.057 (R2=1). The comparative results between ADVIA
Creatinine_2 and LABTEST Creatinine (Labtest Diagnstica S.A) were performed in
53 serum samples and 40 urine samples; all known results were within the creatinine
analytical measurement range. The obtained correlation coefficient to serum and urine
samples was 0.999. The obtained linear regression to serum samples was 1.01384 to
the ADVIA Creatinine_2 and -0.005066 to the LABTEST Creatinine. The obtained
correlation coefficient to urine samples was 0.999. The linear regression was 0.987 to
ADVIA Creatinine_2 and -0.52 to LABTEST Creatinine.
Conclusion: The results show that ADVIA Creatinine_2 test presents the total
error estimate less than the allowable total error in all levels of clinical decision.
The estimating bias are less than or statistically equal to the defined EQA Bias in a
significance level of 2.5%. The evaluated results show that ADVIA Creatinine_2 test
is an accurate method to measure creatinine serum and urine samples through a wide
range of clinically relevant concentrations and it also shows equivalent performance
to LABTEST Creatinine assay.

B-178
Serum creatinine determined by Jaffe and enzymatic method, in regular, icteric
and hemolyzed samples

Table 1. Comparison between serum creatinine measured by Jaffe and enzymatic

methods
Mean
Median
SD
(mg/dL) (mg/dL)
All samples
(n=221)
CREA_2
Jaff
ECRE_2
Enzymatic
Regular
samples
(n=106)
CREA_2
Jaff
ECRE_2
Enzymatic
Icteric
samples
(n=69)
CREA_2
Jaff
ECRE_2
Enzymatic
Hemolyzed
samples
(n=57)
CREA_2
Jaff
ECRE_2
Enzymatic

M. Friedrich. Hospital Universitrio, Universidade de So Paulo, So

Paulo, Brazil
Background: Serum creatinine is an important clinical marker for renal
clearance. The Jaffe reaction remains the cornerstone of most current routine
methods, after continuous refinements attempting to overcome inherent
analytical interferences and limitations. With the recent introduction
of the reporting of estimated glomerular filtration rate (eGFR), interlaboratory agreement of serum creatinine results has become an important
international priority. The aim of this study was to compare analytical
performance and practicability of the enzymatic method and Jaffe method
for serum creatinine for routine use and to compare the effects of some
common interfering substances on both methods.
Materials and Methods: We assessed 221 serum samples obtained for routine
clinical care: 106 regular samples (without interfering substances), 69 icteric and 57
hemolyzed samples. 11 samples were both icteric and hemolyzed. Serum creatinine
was determined both by kinetic Jaffes and enzymatic method on Siemens ADVIA
Chemistry 1800. The ADVIA Chemistry Creatinine_2 (CREA_2) assay is a Jaffe,
alkaline picrate, kinetic method and the ADVIA Chemistry Enzymatic Creatinine_2
(ECRE_2) assay is a Creatininase method. We analyzed the agreement between the
two methods and determined the mean difference between them.
Results: Comparison of Jaffe (X) and enzymatic (Y) measurements of serum
creatinine levels reveals significant correlation with or without the presence of
interfering substances. However, mean differences between enzymatic to kinetic
Jaffes methods were higher for icteric and hemolyzed samples, as shown in the table
below.
Conclusions: Serum creatinine can be overestimated by Jaffes method in the
presence of interfering substances, such as hemoglobin and bilirrubin. Enzymatic
method is less affected by interferences so it is a better method to measure creatinine.

S184

Lowest Highest
value
value
R
(mg/dL) (mg/dL)

3.66

1.95

4.65 0.21

30.25

3.52

1.77

4.77 0.25

31.66

4.85

2.46

5.90 0.21

30.25

4.79

2.27

6.10 0.25

31.66

2.52

1.8

2.62 0.46

15.17

2.31

1.58

2.60 0.51

15.05

2.62

1.69

2.66 0.37

14.20

2.41

1.46

2.66 0.45

14.34

Linear
Regres- Bias
sion

y=
0.997 1.023x - -3.82
0.21

y=
0.998 1.025x - -1.23
0.18

y=
0.992 0.991x - -8.33
0.18

y=
0.993 0.997x - -8.01
0.20

B-179
The reference value of non esterified fatty acids determined by enzymatic
method in healthy population

Y. Huang, T. Wang, X. Du, J. Jian, J. Kang, F. Chen, A. Zhu. Beijing Strong

Biotechnologies, Inc., Beijing, China
Background: Increased circulating levels of nonesterified free fatty acids (NEFA) have
been observed in such hyperinsulinemic states as obesity, impaired glucose tolerance,
diabetes, and dyslipidemia where they have been causally linked to the development
of insulin resistance and hyperinsulinemia.We created liquid enzymatic method for
measuring NEFA. Appropriate reference ranges allow for effctive utilization of an
assay. The aim of the current work was to establish NEFA reference ranges in healthy
population.
Methods: Colleted normal serum specimens for referenc range determinations from
101 healthy individuals, age from 20 to 62 years, who attended Beijing Strong
Biotechnologies for annual health check-up. None of these specimans contained
Diabetes and hypertersion. Specimens containing high TG and high GLU were
expected. The ration of female/male was 70/31, Blood was collected in the morning
after fasting for 12 hours. Blood sample were stored on ice. Centrifugation was carried
out within one hour. The serum samples were assayed on Hitachi 7180 within 2 hours
of cellection. The outliers was excluded using Grubbs statistics mehtod.
Results: NEFA displayed a normal distribution for this reference population both
male and female (Figure 1). The reference interval for NEFA using liquid enzymatic
method provided different range between males and females, the range for females
(n=70) was 0.02 to 1.02mol/L and for males (n=31) was 0.10 to 0.78mol/L
Conclusions: The reference range determined in this healthy population was gender
dependent with higher levels for females.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Electrolytes/Blood Gas/Metabolites

Wednesday, July 30, 9:30 am 5:00 pm

at 4 mg/mL with other proprietary excipients. One part of Reagent A was slowly
added to four parts of Reagent B under very controlled processing conditions. The
resulting COD precipitate was centrifuged and the supernatant was decanted. The
pellet was re-suspended in water then spiked into the Quantimetrix Dip&Spin and
QuanTscopics urine sediment control formulations for evaluation under standard
brightfield microscopy and on the IRIS IQ200 automated urine sediment analyzer.

B-180
Hydroxocobalamin Interference with Carboxyhemoglobin (COHb)
Measurements

P. V. A. Pamidi, H. Yim. Instrumentation Laboratory, Bedford, MA

Background: Hydroxocobalamin (OHCbl), a vitamin B12 analog, is known to
interfere with CO-Oximetry measurements (1). Recent study by Livshits et al (2)
refers to two carbon monoxide poisoning patients and claimed that the CO-Oximeter
reported falsely low carboxyhemoglobin after hydroxocobalamin therapy. The paper
by Livshits et al went on to suggest that the false reading by the CO-Oximeter may
lead to incorrect diagnosis and delay of appropriate treatment. However, assessment
in out lab prior to this publication (1) showed a smaller interference at normal COHb
levels. In this study, impact of hydroxocobalamin at high COHb levels is evaluated.
Methods: Blood samples collected from healthy donors were used to prepare COHb
levels (25 -50 %). COHb samples spiked with hydroxocobalamin (1 g/L) were
measured on 3 GEM Premier 4000 analyzers. The effect of OHCbl interference on
carboxyhemoglobin is evaluated using the measured difference between the unspiked
and the spiked samples. To emulate the treatment conditions reported in Livshits
reference, blood samples were tonometered with 100% oxygen.
Results: Carboxyhemoglobin data shown in table below confirmed that accuracy is
slightly affected by the presence of OHCbl (1-2 units). However, with or without
OHCbl, the reduction in COHb is mainly triggered by the oxygen treatment. In
addition, spiked samples were appropriately flagged by the iQM software.
Conclusions: Blood samples spiked with 1 g/L OHCbl showed a small interference
(1-2%) on carboxyhemoglobin consistent with our prior data in reference 1. Based on
these results, the dramatic reduction in the COHb level as reported by Livshits is not
due to interference from hydroxocobalamin and oxygen treatment would likely have
caused reduction in COHb levels.
References: 1) P.V.A. Pamidi, et al , Clin. Chim. Acta, 401, 2009, 63-67.
2) Z. Livshits, et al, New ENGLAND J. MED 367: 1270 - 1271, SEPTEMBER 27,
2O12

Results: Brightfield microscopy at 400x magnification showed a majority of

COD crystals with near perfect octahedral, bi-pyramidal morphology at various
sizes ranging from 5-20 M in diameter in a single preparation. Calcium oxalate
monohydrate (COM) crystals in an ovoid morphology were also apparent to a lesser
degree in the same preparation. Different preparations made under non-optimal
processing conditions resulted in very large COD crystals with complex fourarmed cruciform, stellate, or cloverleaf morphologies. The optimized COD crystal
preparation in both the Dip&Spin and QuanTscopics formulations were accurately
identified and characterized by the IRIS IQ200 analyzer.
Conclusion: This novel method for synthesizing COD crystals with the native
octahedral morphology identical to the most commonly observed crystal type found
in human urine makes for an ideal urine sediment control. Automated urine sediment
systems like the IRIS IQ200 and 77 Elektronika UriSed analyzers have been
particularly challenging as they do not always recognize the presence of crystals
in some third party control formulations. The new Quantimetrix Dip&Spin and
QuanTscopics formulations with the improved COD crystals are easily characterized
by both standard microscopy and automated microscopy methodologies making for
an excellent control solution for the clinical lab.

B-182
Reference Intervals for Random Urinary Calcium and Magnesium

V. Gounden, C. Y. Park, T. Prabhala, K. Spaid, S. J. Soldin. National

Institutes of Health ( Clinical Center), Bethesda, MD
Background: To estimate the rate of excretion of urinary constituents, a 24 hour
sample is required. However these are cumbersome for the patient and often are not
collected accurately. The use of the analyte ratio to urine creatinine accurately reflects
the 24 hour excretion as creatinine is excreted at a constant rate throughout the day. At
the Department of Laboratory Medicine, NIH Clinical Center reference intervals for
urine calcium:creatinine and urine magnesium: creatinine ratios based on the hospital
population were previously unavailable. Thus it was difficult to interpret random urine
results for calcium and magnesium and this required that a timed or 24 hour specimen
be collected for accurate interpretation.
Objectives: To determine the reference intervals for urinary calcium:creatinine
and urinary magnesium : creatinine ratios using the Hoffmann method (Hoffmann
RG Statistics in the practice of medicine JAMA 1963;185:864-73) at the National
Institutes of Health Clinical Center

B-181
Demonstration of In-Vitro Synthesized Calcium Oxalate Dihydrate Crystals
with Native Octahedral Morphology for Use in Urine Sediment Controls

B. Fernndez, C. Grandjean, M. Ghadessi, M. Ban. Quantimetrix, Redondo

Beach, CA
Background: The microscopic analysis of urine sediment is routinely performed to
screen for the presence of red and white blood cells, epithelial cells, microorganisms,
casts, and a wide range of crystals. Crystals, regardless of type, result from the
precipitation of urine solutes out of solution. When crystals are found in freshly
voided urine, they indicate formation in-vivo and are usually clinically significant.
Factors contributing to crystal formation include the concentration of the solute in
the urine, pH, and flow rate of urine through the renal tubules. Super-saturation can
arise from dietary excess, dehydration, or from certain medications. Oversaturation
of urine with crystals can lead to stone formation causing pain and an increased
incidence of urinary tract infections. Nearly 80% of all urinary stones are calcium
compounds, especially calcium oxalate. Calcium oxalate dihydrate (COD) crystals in
an octahedral morphology are the most commonly observed crystals in human urine.
Urine sediment control materials, used by clinical labs to validate the processing,
centrifugation, and microscopy of urine samples, would therefore ideally contain the
native octahedral form of COD crystals.
Objective: To synthesize calcium oxalate dihydrate (COD) crystals with the native
octahedral morphology for use in urine sediment control formulations.
Methods: Two reagents were prepared: Reagent A was an aqueous solution of
sodium oxalate at 1mg/mL and Reagent B was an aqueous solution calcium chloride

Methods: Data was collected for a total of 159 individuals between the ages of 4-73
years. This study population consisted of 131 healthy outpatients seen at the NIH
Clinical Center and 28 normal volunteers. Data analysis was performed on urine
samples collected between May and July 2013. Outpatient data for urine calcium,
magnesium and creatinine results as well as demographic details were obtained
from the laboratory information system (SoftLab, SCC Soft Computer ,FL). The
Dimension XPand chemistry analyzer (Siemens Diagnostics,Tarrytown NJ) was used
to measure the concentrations of urine creatinine, calcium and magnesium following
the manufacturers guidelines. All analytes are measured by spectrophotometric,
bichromatic rate technique.
Urine creatinine, magnesium and calcium values were converted to SI units (mmol/l)
before further data analysis. Statistical analyses were performed on Microsoft Excel.
The ratios were ranked from smallest to largest values. Then, we developed a percent
cumulative frequency chart and plotted the cumulative percent frequencies against
the log of these ratios/values. Using the Hoffmann approach, we analyzed the linear
portions of the curve and calculated the line of best fit. We calculated the 2.5th and
97.5th percentile values as the new reference intervals of these analytes.
Results: The following reference intervals were found employing the Hoffman
approach. Urine calcium:creatinine ratio was 0.11 - 1.03 mmol/mmol and for urine
magnesium: creatinine the ratio was 0.20-0.76 mmol/mmol.
Conclusions:The availability of reference intervals for urine calcium: creatinine and
urine magnesium:creatinine ratio allow for random specimens to be better utilized
for clinical and diagnostic purposes. This is also more convenient for the patient
and health care providers. An interesting question for future studies revolves around
possible gender and racial disparity of these reference intervals and the influence of
seasonal changes on these ranges.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S185

Electrolytes/Blood Gas/Metabolites

Wednesday, July 30, 9:30 am 5:00 pm

B-183
Reference range for sodium, potassium and chloride in single spot urine
samples

A. C. C. S. Botelho1, W. Pedrosa1, L. S. Vasconcellos2, S. M. Eloi-Santos2.

1
Hermes Pardini, Belo Horizonte, Brazil, 2Universidade Federal de Minas
Gerais, Belo Horizonte, Brazil
Background: The determination of urinary electrolytes is of great importance in
the investigation of systemic metabolic disorders. Since renal excretion daily rate
is not uniform during the day, the 24-hour urine is considered the gold standard
biological sample. However, its replacement by a fresh urine sample could reduce
patient discomfort and the possibility of pre-analytical interferences. Nevertheless the
findings in the literature are conflicting. This study aimed to estimate the pattern of
excretion of sodium, chloride and potassium over 24h and define reference ranges for
the periods presenting good correlation with the 24-hour urine.

the 60 mL/min decision limit. The difference between the predicted and actual CKDEPI discordance at the 60 mL/min decision limit was not statistically significant (chi
sq= 0.03, p=0.86). The Jaffe method performed with greater precision at four of the
five concentrations in the 20 day precision profile. At concentrations of 0.28, 0.79,
1.21, 2.73, and 5.08 mg/dL, the CVs of the Jaffe method were 3.0%, 1.4%, 0.8%,
0.8% and 0.8%, respectively. The corresponding CVs of the enzymatic method were
2.9%, 1.7%, 1.7%, 1.3%, and 1.2%.
Conclusion: At our institution the Jaffe method generally had greater precision
than the enzymatic method. Discrepancies in the CKD-EPI eGFR based on the Jaffe
method did not result in a statistically significant increase in disease reclassifications
at the 60 ml/min decision limit in an outpatient population. Studies are needed to
characterize the relative rate of interference in additional populations.

Method: Timed sequential urine samples were collected from 41 healthy adults, aged
18-60 years in the following periods: 6am-9am (after breakfast), 9am-12pm, 12pm3pm, 3pm-6pm, 6pm-9pm and 9pm-6am. Sodium, potassium and chloride were
measured by Ion Selective Electrode method and creatinine by Kinetic Colorimetric
method (Roche P- Modular). The values of each sample were correlated with
those obtained from the 24-hour urine. The urine samples from the best correlated
periods were used to establish the reference ranges, following Clinical and Laboratory
Standards Institute (CLSI). For that, samples of 120 healthy subjects were used and
confidence interval was 95% and the significance level of 0.05.
Results: For sodium and chloride, high positive correlation was seen in all period
samples. However, the best correlation was obtained with samples collected between
6am-9am, after breakfast (sodium: r=0.6185, p=0.0028; chloride: r=0.5787, p =
0.0060). For potassium measurement, the best correlation was with urine collected
between 6pm-9pm (r=0.5824, p=0.0001). The reference ranges are shown in Table 1.
Conclusion: It is possible to replace the 24-hour urine sample by spot urine collected
in predetermined periods: 6am-9am for sodium and chloride and 6pm-9pm for
potassium.
Reference range for sodium, potassium and chloride in single spot urine samples
Time of
Lower limit (mEq/g
Higher limit (mEq/g
collection creatinine)
creatinine)
Na
6am-9am 24 (90% CI 15.0 to 33.7) 300 (90% CI 229.8 to 309.8)
Cl
6am-9am 19 (90% CI 12.8 to 36.6) 287 (90% CI 275.2 to 337.3)
K
6pm-9pm 11 (90% CI 10.1 to 14.1) 91 (90% CI 74.2 to 108.6)
Electrolytes

B-184
A Comparison of Creatinine Measurement by the Jaffe and Enzymatic Methods
in an Outpatient Population

A. H. Adams, J. A. Straseski, C. M. Lehman, R. L. Schmidt. University of

Utah and ARUP Laboratories, Salt Lake City, UT
Background: Serum creatinine (SCr) concentrations and estimated glomerular
filtration rates (eGFR) are widely used for the evaluation of renal function. The Jaffe
and enzymatic methods are the most common methods for creatinine measurement.
The Jaffe method is generally less expensive than the enzymatic method but is more
susceptible to interferences. Significant savings could be obtained if populations could
be identified where the interference rate of the Jaffe method is acceptably low. Most
studies on Jaffe interferences have used spiked samples and the rate of interferences
in defined patient populations has not been well characterized. The interference rate
is likely to vary by patient population. The objective of this study was to compare
the creatinine and eGFR results from Jaffe and enzymatic creatinine methods in an
outpatient population.
Methods: This study analyzed 545 unique, randomly selected, outpatient samples
over a period of 45 days. Samples were analyzed using both the Jaffe (kinetic alkaline
picrate, Abbott Laboratories) and enzymatic (Creatininase, Abbott Laboratories)
methods using an Abbott Architect c8000. eGFRs were calculated using the CKD-EPI
equation. A 20 day precision study, following CLSI guidelines, was also performed
that evaluated both creatinine methods at concentrations of 0.28, 0.79, 1.21, 2.73, and
5.08 mg/dL.
Results: Orthogonal (Deming) regression showed no significant difference between
the Jaffe and enzymatic methods. The slope was 1.006 (95% CI: 0.998, 1.103) and the
intercept was -0.005 (95% CI: -0.015, 0.006). The average difference (bias) between
the methods was -0.007 mg/dL. The Bland-Altman (BA) limits of agreement (LOA)
for the creatinine difference were -0.139 and 0.136 mg/dL. The Bland-Altman limits
of agreement for the CKD-EPI eGFR were -10.3 and 10.4 mL/min. 3.1% (17 of 543)
of CKD-EPI eGFR discrepancies resulted in a change of classification with respect to

S186

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Molecular Pathology/Probes

Wednesday, July 30, 9:30 am 5:00 pm

Wednesday, July 30, 2014

evaluate such changes, the results of 35 patients who carried out themicrodeletions of
the Y chromosome were both analyzed on agarose gel and incapillary electrophoresis.
The Kappa statistic was used to compare the results.

Poster Session: 9:30 AM - 5:00 PM

RESULTS: Results accordance obtained between the two techniques fromthe 35

samples using Kappa statistics was perfect 1,0 (0,669 to 1,0 CI 95%). There were no
statistically relevant difference (p < 0,001) among compared methodologies.

Molecular Pathology/Probes

CONCLUSIONS:We concluded that capillary electrophoresis could be used as an

alternative method for the study of Y chromosome microdeletions, providing a more
efficient and secure assay.

B-185
Case Report: Molecular and Cytogenetic characterization of a 46,XX male

B-189

V. D. T. Niewiadonski , D. C. Abreu , A. F. Cobacho , C. S. Rodrigues ,

O. P. Denardin1, N. Gaburo Jr.1. 1DASA, Sao Paulo, Brazil, 2DASA, Rio de
Janeiro, Brazil
1

Background: The Y chromosome evolves from an auto chromosome and accumulates

male-related genes including sex-determining region of Y-chromosome (SRY) and
several spermatogenesis-related genes. 46,XX subjects carrying the determining SRY
gene usually have a completely male phenotype.
Objective: To identify Y chromosome material in an azoospermic male with a 46,
XX karyotype.
Method: A 39 years old male was tested to micro deletions of Y chromosome in
order to investigate the origin of an infertility characterized by azoospermia. He did
not report any previous familiar history of infertility and was never submitted to any
infertility treatment before. The isolated DNA was submitted to a PCR reaction to
amplify the following regions of Y chromosome according to the protocol previous
described by Simoni et al (2004). The patient was also submitted to a karyotype
analysis which shown presence of two X chromosomes. To confirm the presence of
Y or part of chromosome, fluorescence in situ hybridization (FISH) was performed
using LSI SRY/CEP X probe set (Vysis).
Results: PCR amplification of DNA was detected using QIAxcel DNA Screening
Kit (Qiagen, Hilden Germany) and showed the presence only of the sex-determining
region of the Y chromosome (SRY) and the absence of others target regions of Y
chromosome (AZFa, AZFb and AZFc). FISH analysis showed an X chromosome
containing SRY gene sequence on the top of the short arm. This Y chromosome gene
was not visible by conventional cytogenetic analysis which shown presence of two
X chromosomes.
Conclusion: Molecular and FISH techniques were very useful for detecting and
locating Y sequences in this particular case, allowing an accurate diagnosis and
correct management of the patient. Testing new Y chromosome markers in XX males
will make it possible to narrow the breakpoints further in each case and to establish
correlations with the clinical features, identifying the Y regions implicated in the
definition of the phenotype.

B-188
Validation of molecular testing of the Y chromosome microdeletions in DNA
analyser: A case of laboratory automation

F. S. V. Malta, R. L. A. Bottrel, M. S. Goncalves, G. C. Lopes, A. G. Assini,

E. C. C. Mateo, A. C. S. Ferreira. Instituto Hermes Pardini, Vepasiano,
Brazil
BACKGROUND: Infertility is a medical condition that affects 10-15 % of couples
seeking to have children.Large proportion of cases of idiopathic male infertility
is due to the presence of microdeletions in the Y chromosome genes related to
spermatogenesis. Located on the long arm of the chromosome Y, the 3 regions
known as azoospermia factors ( AZFa , AZFb and AZFc ) are found fully or
partially deleted in azoospermic patients or patients with severe oligozoospermia.
Our laboratory offers the test for Y chromosome microdeletions by multiplex PCR
followed by electrophoresis on 2% agarose gel stained with ethidium bromide. Once
it is a manual method occasional errors may occur in prepare the gel or pipetting the
samples. Furthermore, ethidium bromide is a mutagenic reagent and therefore the
represents a risk to both the operator and the environment.
OBJECTIVE: The objective of this study is to standardize the diagnosis of Y
chromosome microdeletions through a more efficient and secure methodology.
METHODS: For this study, the capillary electrophoresis was chosen to substitute
the agarosegel electrophoresis. Modifications were made in the PCR reactions, the
primers were labeled with different dyes and the settings in the analysis software
GeneMapper ID-X v. 1.1.1 were modified to optimize theanalysis of this genetic
testing in different fluorescence channels in the 3730 DNA Analyzer.In order to

Integrative analyses of Hippo pathway components in human cancer genome

T. Yu, Y. Chang, J. Bachman, Z. Lai. Pennsylvania State University,

University Park, PA
Background: The Hippo signaling pathway regulates cell proliferation and apoptosis
to control tumor growth and organ size. Cancer genome projects found more somatic
mutations in Hippo pathway than candidate approach. However, it is elusive for
their damaging degree and their association with other mutated genes in the signal
transduction network. In here we have analyzed the somatic non-synonymous
mutations of the Hippo pathway components from cancer genome database.
Methods: Hippo pathway components mutation data are downloaded from COSMIC
(Catalogue Of Somatic Mutations In Cancer) and TCGA (The Cancer Genome
Atlas). The conservation of the mutated residues and the 3D structure remodeling
are evaluated by ClustalW2 and CPH models 3.2. All the mutation pathogenicity are
determined by PolyPhen2, Mutation Assessor, SIFT and Provean. We also integrated
knowledge of the mutated sites from literature using candidate mutagenesis, or large
scale phosphoproteomics approaches. The carcinoma genome of ten tissues with top
mutated frequencies of Hippo Pathway components are downloaded from TCGA data
portal. Significant concurrent and mutually-exclusive relationships (p-value<0.01)
were revealed among the mutations of the Hippo components, its interactome, and
cancer census genes. The unbiased screening selected a list of genes with strong
associations with Hippo components. The significantly mutated genes were identified
from 57, 57, 28 and 25 patients carried predicted damaging mutations of LATS1,
LATS2, STK3 and STK4, and compared with the ones found in 19, 33, 32, and 27
patients with synonymous or predicted neutral mutations. Survival analyses were also
performed for the patients with Hippo pathway alterations.
Results: The mutation spectrum of Hippo Pathway components has been explored.
Core Tumor suppressors(LATS1/2 and STK3/4) (5%) have higher damaging mutation
frequency than core oncogenes (YAP1/WWTR1 and TEAD2/4) (1%) in the Hippo
pathway. 67% out of total 652 mutations and 36 predicted damaging mutations sit in
the domain and residues, respectively, reported by literature. Hippo interactome test
found that PSMD2, WWP1, NEDD4 involved in the degradation of tumor suppressors
in Hippo pathway are found concurrent mutated with those tumor suppressors. Cancer
census gene comparison testing found that XPO1, TCEA1 and SUZ12 are concurrent
with Hippo pathway components, while PIK3CA, TP53 and STK11 are mutually
exclusive. The unbiased screening identified CENPC1, ORC are concurrent with
Hippo pathway components, while CDKN2A and RHOA are mututully exclusive.
Patients with up-regulation of oncogenes has worse survival rate than the patients
without.
Conclusions: The analyses for Hippo Pathway mutations in cancer patients provided
comprehensive reference for physicians and researchers to further investigate the
mutations. The genomic analyses shed light on elucidating mechanisms of Hippo
pathway and the concurrence and mutually exclusivity with its associated genes
in cancer, which could aid in the development of a cancer biomarker panels for
prognoses and treatments.

B-191
Comparative Genomic Hybridization arrays (aCGH) Technique as a diagnostic
tool for 22q11.2 microduplication syndrome.

E. C. Mateo, C. B. Campos, C. P. Melo, A. C. S. Ferreira. Instituto Hermes

Pardini, Vepasiano, Brazil
Background: At the moment, it`s uncertain whether the 22q11.2 microduplication is
a natural genetic variant-we are all different-or whether it`s a real syndrome whose
effects can be highly variable. The 22q11.2 syndrome is related to development delay,
intellectual disability, slow growth leading to short stature and poor muscle tone. All
patients underwent examinations of karyotype and in all cases were normal. Most
people have a 22q11.2 microduplication that is about <3 Mb in size. Most of the 30-40

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S187

Molecular Pathology/Probes

Wednesday, July 30, 9:30 am 5:00 pm

genes in the <3Mb stretch of duplication have not been fully characterized. However,
researchers believe that the duplication of one particular gene, known as TBX1, is
responsible for many of the typical symptoms of the syndrome. aCGH is a molecular
cytogentics technique, able to identify unbalanced chromosomal changes (gains or
losses of genomic material) through the general analysis of the entire genome in a
single experiment. Objective: The objective of this study was to show the use of the
technique Comparative Genomic Hybridization arrays (aCGH) as a molecular tool for
the diagnosis of microduplication 22q11.2. Methods: 3 samples of peripheral blood of
pediatric patients referred to the Institute for study of loss or gain of genomic material
were used. The samples were collected in tubes 4ml EDTA and stored at 4 C until
processed. For DNA extraction was used QiampDNA Blood Mini Kit (Qiagen). After
extraction the samples were quantified and evaluated reasons for A260/A280 and
A260/A230 in NanoDrop2000/2000c. The samples had concentrations above 50 ng/
uL and A260/A280 (1.8 to 1.9) and A260/A230 (1.5 to 1.9) were among the reasons
for standard values. gDNA was digested with restriction enzyme labeled with the
cyanines 3 (Cy3-reference) and 5 (Cy5-patient) and purified by SureTag Complete
DNA Labeling Kit (Agilent). After purification, the samples were again quantified to
measure the incorporation of Cyanine 5 (Cy5). All samples had A260/A280 reason to
1.8, DNA concentration above 420.0 ng/uL, the incorporation of Cy5 was above 10.3
pmol/uL, the specific activity was above 24 pmol Cy5/g DNA and yield ug DNA was
above 8.9. The OligoaCGH/ChIP-on-chip Hybridization Kit (Agilent) and Human
Cot-1 DNA (Agilent) kit was used to perform hybridization on the slide containing
probes corresponding to 180,000 genes. After 24 hours of hybridization at 65 C
the slides were washed with AgilentOligoaCGH/ChIP-on-Chip 1 and 2 Wash Buffer
Kit (Agilent) and acetonitrile reagent (Sigma). After extraction of data through the
Software ScanControle Software Fecture Extraction, these were analyzed in Agilent
CytoGenomics Edition Software 2.7.8.0. Results: All samples met the standards of
quality required for the analysis. The technical benchmark DerivativeLR_Spread
(DLRS) was <0.30. Samples showed gain of genomic material (279 Kb) in the
22q11.2 region. Conclusions: This study shows the importance in examining the
aCGH technique for the diagnosis of patients with 22q11.2 syndrome who had normal
results in the karyotype.

B-192
Genotyping of rs12979860 and rs8099917 single nucleotide polymorphisms in
HCV infected Brazilian patients.

V. D. T. Niewiadonski1, P. G. C. Trojano1, P. L. Oliveira1, M. D. Freire2,

N. Gaburo Jr.1. 1DASA, Sao Paulo, Brazil, 2DASA, Rio de Janeiro, Brazil
Background: Recent studies have demonstrated the role of the interleukin 28B
(IL28B) polymorphisms in predicting treatment response, spontaneous clearance
and sustained virologic response for patients with Hepatitis C virus (HCV) infection.
Two main single nucleotide polymorphism (SNP) were identified in proximity
to interleukin IL28B: rs12979860 (homozygous for the major C allele are more
likely to respond to treatment than those who were homozygous for the alternative
nucleotide (T) and rs8099917 (strong predictor of sustained virologic response (SVR)
to pegylated interferon and ribavirin for subjects with homozygous T/T genotype than
subjects with G/G genotype).
Objective: To describe the frequency of the IL28B C/T SNP for rs12979860 and that
of the T/G SNP for rs8099917 in a cohort of Brazilian patients with HCV infection.
Methods: Blood samples from 305 HCV infected Brazilian patients were analyzed
from January 2013 to December 2013. DNA isolation was performed with the
automated platform Qiasymphony SP (Qiagen, Hilden, Germany) according
manufactured instructions. Genotyping for rs12979860 and rs8099917 were
performed using TaqMan SNP Genotyping Assays (Life Technologies, Foster City,
CA) at Viia 7 Real Time PCR System (Life Tecnhologies). The results were analyzed
using TaqMan Genotyper Software version 1.0.1.
Results: Carriers of rs12979860 CT genotype predominated (160/305, 52.5%),
homozygotes for allele C were 90/305 (29.5%) and the remaining homozygotes for
IFN-resistant allele T were 55/305 (18.0%). As for the rs8099917 SNP, genotypes
were distributed as follow: 175/305 (57.4%) carried the rs8099917 TT genotype,
whereas 110/305 (36.0%) carried GT and 20/305 (6.6%) the GG genotype. The coprevalence of genotypes is shown at Table 1.
Conclusion: The frequencies found were consistent with previous studies. Testing for
rs12979860 and rs8099917 has become an important strategy to predict the patient
treatment outcome.

S188

Table 1- Combined genotype frequencies of the interleukin (IL)28B single nucleotide

polymorphisms rs1297860 and rs8099917

rs12979860
CT
CC
TT
Total

rs8099917
GT
n (%)
81 (26.6)
4 (1.3)
25 (8.2)
110 (36.0)

GG
n (%)
2 (0.6)
0 (0.0)
18 (5.9)
20 (6.6)

TT
n (%)
77 (25.3)
86 (28.2)
12 (3.9)
175 (57.4)

Total
160 (52.5)
90 (29.5)
55 (18.0)
305 (100)

B-193
Validation of Hybridizing Genomic Comparative array (aCGH) technique to
screen 180.000 genes in diseases Postnatal.

C. B. Campos, C. P. S. Melo, A. C. S. Ferreira, E. C. C. Mateo. Instituto

Hermes Pardini, Vepasiano, Brazil
Background: aCGH technique has been widely used in postnatal diagnosis of patients
with normal karyotyping and disturbs like autism, neuropsychomotordevelopmental
delay, facial dimorphism, mentaldevelopment delay, low weight and low height,
among others. Because of its capacity to detect microdeletions and microduplications,
aCGH is an important tool to investigate such clinical cases. Objective: The objective
of study was validation and implements the Hybridization Genomic Comparative array
(aCGH) technique for postnatal diagnosis of normal karyotyping patients. Methods:
Samples from 12 patients were analyzed in duplicate by Hermes Pardini Institute and a
reference laboratory. The samples were collected in tubes 4ml EDTA and stored at 4
C until processed. For DNA extraction was used QiampDNA Blood Mini Kit (Qiagen).
After extraction the samples were quantified and evaluated reasons for A260/A280
and A260/A230 in NanoDrop2000/2000c. The samples had concentrations above
50 ng/uL and A260/A280 (1.8 to 1.9) and A260/A230 (1.5 to 1.9) were among the
reasons for standard values. gDNA was digested with restriction enzyme labeled with
the cyanines 3 (Cy3-reference) and 5 (Cy5-patient) and purified by SureTag Complete
DNA Labeling Kit (Agilent). After purification, the samples were again quantified to
measure the incorporation of Cyanine 5 (Cy5). All samples had A260/A280 reason to
1.8, DNA concentration above 420.0 ng/uL, the incorporation of Cy5 was above 10.3
pmol/uL, the specific activity was above 24 pmol Cy5/g DNA and yield ug DNA was
above 8.9. The OligoaCGH/ChIP-on-chip Hybridization Kit (Agilent) and Human
Cot-1 DNA (Agilent) kit was used to perform hybridization on the slide containing
probes corresponding to 180,000 genes. After 24 hours of hybridization at 65 C
the slides were washed with AgilentOligoaCGH/ChIP-on-Chip 1 and 2 Wash Buffer
Kit (Agilent) and acetonitrile reagent (Sigma). After extraction of data through the
Software ScanControle Software Fecture Extraction, these were analyzed in Agilent
CytoGenomics Edition Software 2.7.8.0. Results: All samples met the standards of
quality required for the analysis. The technical benchmark DerivativeLR_Spread
(DLRS) was <0.30. Results for the same patient were compared between both
laboratories and all of them had the same sensibility, specificity and reproducibility.
Four patients had at least one pathogenic alteration detected by aCGH. One had many
LOH regions detected showing that this patients parents came from the same ancestor.
The other seven had alterations considered non-pathogenic. Pathogenic alterations
detected were a deletion on the chromosome 4 short arm (4p16.3p16.2), that overlaps
the Wolf-Hirschorn syndrome region; a duplication on the chromosome 9 short arm
(9p24.3p24.2), related to developmental delay; a deletion on the chromosome X long
arm (Xq27.2), related with mental development delay and obesity; a duplication on
the chromosome 7 long arm (7q36.3), also related with mental development delay;
and a deletion on the chromosome 15 long arm (15q11.2-q13.1), that overlaps PraderWilli and Angelman syndromes region. Conclusions: Concluding, almost half of
the patients presented a pathogenic alteration that was not detected by conventional
karyotyping proving the importance of aCGH as a complementary diagnosis tool.

B-194
Comparison of Invivoscribes T Cell Receptor Gamma Gene Rearrangement
Assay 2.0 vs TCRG Gene Clonality Assay (developed by the Euroclonality,
previously BIOMED-2 Group)

B. Peh, J. Iqbal, L. Oon. Singapore General Hospital, Singapore, Singapore

Background Histology or cytology, supplemented with immunohistology or flow
cytometric immunophenotyping has been used to discriminate between malignant
and reactive lymphoproliferations. However, in some of the cases, the diagnosis
is difficult. The diagnosis of lymphoid malignancies can be supported by clonality
assessment as all cells of a malignancy have a common clonal origin. Gene
rearrangement analysis used to be performed by Southern Blot-based techniques
which is very reliable, but is increasingly replaced by PCR techniques because of

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Molecular Pathology/Probes

Wednesday, July 30, 9:30 am 5:00 pm

the greater efficiency and sensitivity of PCR. These gene rearrangements generate
products that are unique in length and sequence for each cell. Therefore, PCR assays
can be used to identify lymphocyte populations derived from a single cell by detecting
the unique V-J gene rearrangements present within these antigen receptor loci. We
evaluated the T cell receptor gamma gene clonality assay (Euroclonalitys primer)
and T cell receptor gamma gene rearrangement assay 2.0, both from Invivoscribe, for
clonality assessment of T cells lymphoproliferative disorders.

Results: For the detection of CIN2+ high grade cervical lesions, the sensitivity and
specificity of RT-qPCR assay were 92 % and 98.6 %, respectively. Therefore, HPV
E6/E7 mRNA RT-qPCR assay showed the significantly higher sensitivity (91.1
%) compared to real-time NASBA assay (41.1 %). In normal cytology cases, the
specificity was 98.6 % and 53.7 % by HPV E6/E7 mRNA RT-qPCR assay and HPV
DNA testing. These results revealed that HPV E6/E7 mRNA RT-qPCR assay better
reflects cytological diagnosis.

Methods 22 archived formalin-fixed, paraffin-embedded (FFPE) clinical samples

were extracted with QIAamp DNA FFPE tissue kit, following manufacturers
recommendation. 2 proficiency panel samples were included as well. The extracted
genomic DNA was quantified and amplified using the specimen control size ladder
master mix in the Invivoscribe kits and AmpliTaq Gold DNA polymerase. All
samples were tested in duplicates of 50ng and 100ng of DNA to check for presence
of inhibitors and DNA quality. Next, PCR was performed with TCRG tube A, TCRG
tube B and TCRG 2.0 with samples in duplicate, polyclonal, monoclonal and negative
control. The PCR products were denatured with Hi-Di formamide and GeneScan 600
Liz size standards and anlayzed on ABI 3500 capillary electrophoresis instrument.
Data are automatically displayed as size and color specific peaks.

Conclusion: It is suggested that Real-HPV E6/7 mRNA assay (M&D, Wonju,

Republic of Korea) could overcome the shortcoming of lower specificity in DNA
assay as well as the lower sensitivity of commercialized HPV mRNA real-time
NASBA assay, NucliSENS EasyQ HPV v1.1 (bioMrieux, Marcy, France), with
ThinPrep Pap (Hologic Inc.) samples.

Results Majority of the samples were concordant with the in-house developed test
for TCRG, TCRG clonality assay and TCRG 2.0. However, 2 of the samples showed
discrepant results. TCRG 2.0 requires shorter hands-on time to perform and is easier
to analyze as it is single tube and single color (blue). TCRG clonality assay has 2 tubes
(tube A and B) and each tube is dual-color (green and blue). The dual-color can be
quite confusing to interpret. The clone size detected is different for both assays as the
PCR primers are different. Hence, for different labs using these 2 assays, it is hard to
correlate if the clone detected is the same.

H. C. Macher, G. Suarez-Artacho, J. M. Guerrero, S. Alvarez-Gomez, P.

Molinero, M. A. Gomez-Bravo, P. Jimenez-Arriscado, A. Rubio. Virgen del
Roco University Hospital, SevillE, Spain

Conclusion The TCRG clonality assay and TCRG gene rearrangement assay 2.0
from Invivoscribe generated similar results for clonality assessment of T cells
lymphoproliferative disorders. We had 2 cases with discrepant results and will
emphasize that the results of molecular clonality tests must always be interpreted
in the context of clinical, histological and immunophenotypic data. TCRG clonality
assay adopts the Euroclonalitys primers and is more widely used in laboratories.
Even though TCRG gene rearrangement assay 2.0 is easier to perform and interpret,
more data should be generated for the better comparison of these 2 assays.

B-195
HPV E6/E7 mRNA RT-qPCR Assay for Detecting High Grade of Cervical
Lesion with ThinPrep Pap Samples

G. Kim1, J. Munkhdelger2, H. Wang3, D. Lee4, S. Kim1, Y. Choi5, S.

Park1, Y. Kim1, H. Kim1, S. Ahn1, J. Kim1, H. Jin6, S. Park7, K. Park2, H.
Lee1. 1Department of Biomedical Laboratory Science, College of Health
Sciences, Yonsei University, Wonju, Korea, Republic of, 2Department
of Pathology, Wonju College of Medicine, Yonsei University, Wonju,
Korea, Republic of, 3M&D, Inc., Wonju Eco Environmental Technology
Center, Wonju, Korea, Republic of, 4Department of Clinical Laboratory
Science, Hyejeon College, Hongseong, Korea, Republic of, 5Department
of Biomedical Laboratory Science, Songho College, Hoengseong, Korea,
Republic of, 6Department of Clinical Laboratory Science, College of
Health Sciences, Catholic University of Pusan, Busan, Korea, Republic
of, 7Department of Clinical Laboratory Science, College of Health and
Therapy, Daegu Hanny University, Daegu, Korea, Republic of
Background: Human cervical cancer is the second most common cancer among
women worldwide. Several decades ago, human papillomaviruses (HPV) were found
out to be a major factor of cervical cancer. HPV DNA genotyping assay has been
the method of the choice, since it has shown high analytical sensitivity. The latest
results show oncogenic HPV DNA appeared not only in cancerous tissues, but also
in the normal tissues according to cytological diagnosis. For this reason, HPV test
targeting E6 and E7 mRNA of 5 oncogenic HPVs (HPV genotype 16, 18, 31, 33, and
45) which are known to be responsible for oncogenesis of cervical cancer has been
commercialized using real-time nucleic acid sequence based amplification (NASBA)
assay. The previous data showed that the real-time NASBA assay has higher clinical
specificity than HPV DNA testing (97.1 % vs. 53.7 %). However, the sensitivity of
real-time NASBA assay was lower than that of the HPV DNA testing (41.1 % vs.
100 %).
Methods: In the present study, therefore, HPV E6/E7 mRNA targeting RT-qPCR
assay was designed to detect 16 oncogenic HPV genotypes (HPV genotype 16, 18,
31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 68, and 69), it was performed with RNA
prepared from ThinPrep Pap (Hologic Inc., Bedford, MA, USA) samples, and the
results were compared to real-time NASBA data.

B-196
Donated organ genomic signature in circulating DNA of the liver transplant
recipient to monitorization the transplanted liver health

Background:Health assessment of the transplanted organ is very important due to the

relation of long-term survival of organ transplant recipient and the maintaining of
organ health. It has been described that transplanted organ cells suffering damage
liberate DNA. Thus, during organ rejection, apoptosis cell death deals to the release
of specific transplanted organ DNA to the host plasma. In this context organ cell
free DNA may give us a differential genomic marker of the donated organ health.
In a first approximation the quantification of DNA from Chromosome Y in women
host under male organ transplantation may be a useful tool. The objective of this
work was to validate the usefulness of quantifying specific organ circulating DNA
(cDNA) in serum of transplanted patients as noninvasive diagnostic genomic marker,
in the diagnosis of graft injury. With this purpose we monitorized serum cDNA from
chromosome Y in eight women after male liver transplantation.
Methods: cDNA quantization of the SRY gene was performed by real-time
quantitative PCR before, at the moment of transplantation (day 0) and during the
stay at the intensive care unit. Beta-globine cDNA levels, a general cellular damage
marker, were also quantified. Patients were grouped based on clinical outcome. Group
A, were patients that accepted liver transplantation without any complication (Patients
1-3); group B were patients that accepted the male organ but suffer complication not
related with the transplantation (patients 4-6); patient 7 suffering an autoimmune
hepatitis rejected the first transplanted organ but accepted the second one, and finally
patient 8 underwent a liver transplantation from a male liver donor suffering a sepsis
by colangitis that developed to general organ failure and died.
Results: All patients showed an increase of cDNA levels at the moment of
transplantation that decreased until patient stabilization. Group A, showed the early
peak at day 0 that immediately disappear. Patients from group B showed an increase
of beta-globine gene levels but not of SRY gen ones at the moment of any clinical
complication. Patient 7 showed high levels of of beta-globine gene levels and SRY
gen after the first transplantation which was rejected and decreased after the second
one accepted. Liver transplantation was succeful for patient 8 showing low levels of
SRY gen during most the first weeks after surgery. However the beta-globine levels
persisted elevated due to a colangitis that ended in a sepsis, multiorganic failure and
death. At the moment of multiorganic failure SRY gen levels were also increased.
Conclusion: Our results shows that donor-derived cDNA may be quantified in the
serum of organ transplant recipients, and that high levels of donor DNA might be used
as an indication of graft injury vs other pathologies of the patients

B-197
Genetic variants of glucose-6-phosphate dehydrogenase (G6PD) in Brazilian
children with positive neonatal screening for G6PD deficiency, and correlation
with neonatal jaundice

S. F. FONSECA1, C. S. NOBRE2, V. B. A. S. EIRA1, P. L. M. BELMONTE1,

D. O. ALENCAR3, J. V. THOMAS4, R. H. JACOMO5. 1Universidade de
Brasilia, Brasilia, Brazil, 2UniCEUB, Brasilia, Brazil, 3DLE Medicina
Diagnostica, Rio de Janeiro, Brazil, 4Secretaria de Estado Sade do Distrito
Federal-GDF, Brasilia, Brazil, 5Laboratrio Sabin LTDA, Brasilia, Brazil
Background: G6PD deficiency, the most common human enzymopathies throughout
the world, causes a spectrum of phenotypes including neonatal hyperbilirubinemia
and acute and chronic hemolysis. Genetically, G6PD deficiency is a heterogeneous

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S189

Molecular Pathology/Probes

Wednesday, July 30, 9:30 am 5:00 pm

condition, with approximately 150 mutations and 400 variants identified. Molecular
studies seek to define the origin of the enzymopathy in a determined population
and correlate the G6PD variants with the clinical course of the disease, as well as
identifying the G6PD deficiency in heterozygous females. Previous reports have
shown that the prevalence of G6PD deficiency, in several regions of Brazil, is around
10% among males of African origin and between 1-6% on euro-descendent males. The
data regarding the types of G6PD variants in Brazilian population are fragmented and
scarce. The capital of Brazil, which lies in the Federal District, has a mixed population
representing the different regions of Brazil. The Neonatal Screening Program (NSP)
in Federal District indicates a prevalence of 4.5% for G6PD deficiency. The objective
of the current study was identify the types of variants in the G6PD gene in a group
of children screened through the NSP in the Federal District, and correlate these data
with the presence of neonatal jaundice. Methods: Oral mucosa samples were collected
from eighty boys and four girls diagnosed with G6PD deficiency through the NSP
in January and February of 2014, whose parents signed an informed consent form.
The majority of the newborns presented with residual enzyme activity of around 50%
(moderate deficiency). All representatives of the children filled out a questionnaire
with relevant details regarding family history, history of neonatal jaundice and therapy.
Molecular analysis was carried out using real-time PCR (allelic discrimination). The
G202A and C563T mutations in the G6PD gene were analyzed using specific primers
and probes. Results: Seventy of the 84 families were unable to provide information
regarding ethnic origin of the child, 13 claimed indigenous descent, and one claimed
Portuguese and Spanish descent. 60.7% of the children presented with neonatal
jaundice, 76.5% presented at 48 hours post-natal, and 29% required phototherapy.
Molecular analysis identified a high proportion (98.8%) of neonates positive for the
G202A mutation (variant G6PD A-): 79 boys were hemizygous and 4 girls were
homozygous for this mutation. Only one boy presented the Mediterranean C563T
mutation. Analysis of the correlation between genotype and presence of neonatal
jaundice was compromised by the intense predominance of the G202A mutation in
the sample group. Conclusions: This is the first study carried out in the population
of individuals with G6PD deficiency in the Federal District of Brazil. Although the
sample group studied was relatively small, the high prevalence of a single mutation
suggests that G6PD deficiency in the population of the Federal District is principally
due to the G202A mutation. Neonatal jaundice was frequent among G6PD deficient
children. The absence of cases of heterozygous females in the sample group may
reflect the inability of neonatal enzyme screening to detect G6PD deficiency in these
cases.

B-199
Validation of a quantitative CMV Simplexa kit for clinical use in a private
hospital in Sao Paulo, Brazil

N. H. Muto, L. Oyakawa, J. F. Melo, J. R. Pinho, R. Sitnik. Hospital Albert

Einstein, So Paulo, Brazil
Background: Human Cytomegalovirus (CMV) is a member of the Herpesviridae
family. Primary CMV infection in healthy individuals is asymptomatic or results in
a mild, non-specific illness. After acute infection, CMV establishes latent infection.
Reactivation can occur in immunocompromised patients with an important morbidity
and mortality in this group. Early diagnosis and CMV viral load monitoring in high
risk patients are critical for an efficient infection management. To improve molecular
diagnostic of CMV in our Clinical Laboratory, we validated the quantitative FOCUS
CMV Simplexa kit in plasma samples.
Methods: Nucleic acids from plasma samples were extracted using automated
EasyMag system and submitted to Real Time PCR with Focus Simplexa CMV kit,
using the 3M Integrated Cycler equipment which is a nucleic acid amplification
system based on a centrifugal micro fluidic platform. Validation was conducted
according to CAP guidelines. We evaluated accuracy, linearity, precision (intra and
inter-assay) and sensitivity. Accuracy was tested comparing obtained results with
previous data. Linearity was determined using dilution series of a high viral load
sample, while intra and inter-assay variations were determined using 4 pools of
samples (high, medium, low-medium and low) and a negative plasma obtained from
the blood bank. Sensitivity was established using samples with viral loads close to the
detection limit of the kit.
Results: For accuracy, we compared the results of 53 samples and obtained a
mean log difference of 0.21 Log copies/mL. Analytical linearity was evaluated in
quadruplicates using the calibration CMV standard from FOCUS, ranging from
2,010,000,000 copies/mL (9.3 Log copies/mL) to 750 copies/mL (2.8 Log copies/mL).
The correlation curve obtained had R2=0.9993. Intra and inter-assay variation of all 4
pools of samples were lower than 10%, as expected. Finally, Detection limitwas tested
with samples with 330, 153 and78 copies/mL. The test was able to detect viral load
of 330 copies/mL with 100% confidence and 153 copies/mL with 90% confidence.

S190

Conclusion: The results of validation demonstrated that the kit is reliable and useful
for quantification of CMV in plasma samples, and combined with its fast turnaround
time and decreased hands-on time, make this assay highly suitable for the rapid
diagnostics of CMV infections in the clinical laboratory.

B-200
Sodium citrate at 8% is equivalent to EDTA as anticoagulant of choice for
circulating cell-free DNA analysis: low contamination by blood cells genomic
DNA and inhibition of blood nuclease activity.

G. B. Barra, V. M. Silva, R. H. Jcomo, J. A. R. Vaz, T. H. Santa Rita.

Laboratorio Sabin, Brasilia, Brazil
Background: Despite the intensive research, few circulating cell-free DNA (cfDNA)
analysis have been translated to clinical practice. The lack of preanalytical consensus
is a major obstacle. Traditionally, the EDTA is the anticoagulant of choice for studding
cfDNA. Moreover, because of the lack of cell protection, the cfDNA is susceptible to
blood nucleases, but the impact of these enzymes has long been neglected. Here, we
studied the initial amount of cfDNA, its stability and the blood nucleases activity in
plasmas (EDTA, citrate, heparin) and serum samples.
Methods: Fresh blood from 20 health donors was collected simultaneously in
K3EDTA, sodium citrate 3.2%, sodium heparin, and Z serum clot activator tubes (all
from Greiner-bio-one). The citrate 8% samples were obtained by transferring fresh
blood sequentially to 3 citrate 3.2% tubes. Serum or plasmas were generated within 1015 minutes after the venipuncture. DNA extraction was performed by using Nuclisens
easyMAG (Biomerieux). RNAse P was the target used for cfDNA quantification in a
StepOne qPCR System (Life technologies) by using hydrolysis probe chemistry and
absolute quantification. The results were shown as median in Genomic Equivalents/
mL. Statistical analysis was Friedmans test. The cfDNA stability was evaluated
treating (or not) the samples with 25U of DNAse I for 1h at 37C before RNAse
P assay. To investigate samples nuclease activity a hydrolyze probe and a passive
reference (ROX) were added to the crude samples and the fluorescence increase were
measured for 24h at 37C in the qPCR system. For nucleases inhibition assay a serial
dilution of citrate (0.4 to 14%) was used.
Results: The cfDNA amounts in EDTA (158.7 GE/mL) and in citrate (130 GE/
mL) were similar (p=0.27) and lower than the levels found in heparin (413 GE/mL;
p=0.031-EDTA, p<0.0001-citrate) and in serum (815 GE/mL; p=0.0012-EDTA,
p<0.0001-citrate). The nuclease activity was higher in heparin (arbitrary considered
100%), 90% in serum, 66% in citrate and not detected in EDTA. The nuclease activity
curve in citrate was different from serum and heparin suggesting an inhibitory effect.
The treatment with DNAse I reduced the cfDNA amount in EDTA by 1.1-fold, in
serum by 1300-fold, in heparin by 242-fold and in citrate by 1.3-fold. In the citrate
serial dilution experiment, no nucleases activity was detected from 7%. Increasing the
citrate concentration to 8% did not change the initial cfDNA amount (96.86 GE/mL)
compared to EDTA (129.2 GE/mL; p=0.12) and citrate (89.57 GE/mL; p=0.99). The
nuclease activity was not detected in citrate 8% and treatment with DNAse I did not
alter its cfDNA amount, reduction of 1.01-fold.
Conclusion: The citrate 3.2%, citrate 8% and EDTA have similar initial cfDNA,
although lower when compared to heparin and serum. The nuclease activity was
higher in heparin and serum, partially inhibited in citrate 3.2% and completely
blocked in EDTA and citrate 8%. The divalent ions chelators citrate 8% and EDTA
share a common mechanism of both avoid blood cells genomic DNA contamination
to cfDNA and inhibit blood nucleases. The high-levels of cfDNA in serum and heparin
should be attributed to the coagulation and direct lyses of blood-nucleated cells,
respectively.

B-201
Development of a real-time PCR genotyping assay to detect HLA-B*5701 allele
associated with abacavir hypersensitivity reaction

H. S. Yang, A. Smith, K. L. Lynch, A. H. Wu. University of California, San

Francisco, CA
Background: Abacavir sulfate is a nucleoside reverse-transcriptase inhibitor with
potent antiviral activity against HIV. 5-10% of individuals being treated with abacavir
develop a potentially life-threatening hypersensitivity reaction (ABC-HSR). Human
leukocyte antigen (HLA) B*5701 allele strongly predicts ABC-HSR. Therefore,
Pharmacogenetic screening for the HLA-B5701 allele is recommended prior to
initiation of abacavir therapy. In San Francisco General Hospital, a two-color B57specific immunofluorescence assay with FacsCanto flow cytometry was utilized to
detect all HLA-B57 subgroups. Preliminary positive samples are sent to a reference

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Molecular Pathology/Probes

Wednesday, July 30, 9:30 am 5:00 pm

lab to identify true HLA-B*5701 positive patients. Recent studies have shown that
a HCP5 single-nucleotide polymorphism (SNP), rs2395029, is in perfect linkage
disequilibrium with the HLA-B*5701 allele: the sensitivity of the HCP5 SNP for
carriage of the HLA-B*5701 allele was 100% and specificity was 99%. Objective:
Develop an accurate in-house assay utilizing real-time polymerase chain reaction
(PCR) and fluorescence monitoring. Methods: DNA extraction from blood samples
was performed with a Qiagen DNA mini kit, and DNA concentration was measured
using a NanoDrop ND2000. A rapid-cycle PCR was developed using the RotorGene Q 2plex HRM system. Forward primer: GAGTGCCCATTGAACTACACA,
reverse primer: GCTGGTCTCTGGACACATACTG, wild-type probe: FAM
- AGCTGCCACAGGG - BHQ1 plus, mutant probe: CAL Fluor Orange 560 AGCTGCCCCAGGG - BHQ1 plus. Thermocycling conditions were 20 sec at 95 C,
followed by 40 cycles at 95 C for 3 sec and 60 C for 30 sec. PCR was performed
in a 25-ul volume in the presence of 1X Taqman GTXpress master mix, 900 nmol/L
of each primer, 250 nmol/L of each probe, 2 ul DNA, and DEPC H2O. 2-fold serial
dilutions of a wild-type (T/T) sample and a mutant sample (G/G), with each dilution
amplified in triplicates, were tested to evaluate the linearity and repeatability, as well
as the limit of detection of the genotyping assay. A standard curve was constructed
for each sample on the basis of DNA serial dilution, on which Ct values were plotted
against the log value of the target DNA amount. Blood samples of 49 patients who
were diagnosed of HIV were included in the patient comparison study between the
new RT-PCR assay and PCR-SSOP method of the reference laboratory. Results: Ct
values were obtained from amplification of serial dilutions of a wild-type sample
from100 ug/ul to 3.125 ug/ul and a homozygous mutant samples from 10 ug/ul
to 0.625 ug/ul, respectively. The regression equation of the wild-type sample was
y=-3.4317x + 30.912, with a R2 of 0.9985. The intra-assay coefficient of variation
(CV) for all dilutions ranged from 0.037% to 0.73%. The regression equation of the
homozygous mutant was y=-3.2123x+31.021 with a R2 of 0.9907. CV for all dilutions
ranged from 0.13% to 0.31%. Patient comparison study revealed that this real-time
PCR assay demonstrated 100% sensitivity and 100% specificity when validated
with 10 positive and 39 negative samples previously confirmed by the reference
lab. Conclusion: A real-time genotyping assay was developed to identify positive
and negative HLA-B*5701 alleles. This approach offers a sensitive, rapid and costeffective screening assay prior to abacavir prescription. The genotyping assay has a
wide dynamic range of reliable amplification linearity.

B-202
Development and Validation of a Clinical Sequencing Assay Using RNA-Seq to
Direct Treatment of Relapsed Pediatric Cancers

M. Meyers-Needham, C. Chiang, M. S. Mahadevan, L. A. Silverman, R.

Legallo, I. M. Hall. University of Virginia, Charlottesville, VA
Background: Cancer continues to be the leading cause of death in children in the
US. The aim of this study was to develop a clinically-validated, RNA-sequencing
(RNA-Seq) assay of formalin-fixed, paraffin-embedded (FFPE) tumor tissue to detect
actionable mutations and/or pathway activation in pediatric cancer relapse patients
who have a <20% chance of event-free survival.
Methods: We studied FFPE tissue from 12 tumors previously identified to have gene
amplification and/or over-expression. Purified RNA was quantified and evaluated by
use of Nanodrop (2 - 3,000 ng/uL RNA), Invitrogen Quant-IT Qubit RNA BroadRange Assay Kit (1 ng/uL - 1 ug/uL) and Agilent Bioanalyzer RNA Pico 6000 Kit
(AMR 0.05 - 5ng/uL). RNA Integrity Numbers ranged from 2 - 4.3; 28S/18S values
were 0 for the majority of samples. All purified samples were sonicated to ensure
sufficient fragmentation of RNA. The first 4 samples were sequenced in multiplex on
the Illumina MiSeq platform, and manual analysis of overexpression was performed.
Within-sample normalization of genes of interest was accomplished by selecting
9 housekeeping genes expressed in all samples. When matched normal tissue was
unavailable, data from The Cancer Genome Atlas (TCGA) database was used for
comparison with our results.
Results: At abstract submission, 4/12 specimens have been fully analyzed:
dedifferentiated liposarcoma of kidney, invasive ductal carcinoma of breast, B-cell
lymphoma metastasized to tonsil, and lung adenocarcinoma metastasized to lymph
node. Specimens contained known gene amplification of MDM2, HER2 (ERBB2), or
MYC, or an EGFR mutation, respectively. RNA transcript over-expressions of MDM2
(177-fold), HER2 (20-fold) and EGFR (7-fold) were detected by manual analysis in
tumors with matching gene amplification or mutation. Number of reads ranged from
3.5 - 7.3 million per sample, and thus coverage was insufficient to produce reads in
exons 19-21 of the EGFR gene in order to detect intragenic mutations or deletions.
Therefore only one specimen per flow-cell was sequenced in subsequent runs,
achieving approximately 25 million reads per sample. At time of abstract submission,
the TCGA database had no normal control data of the tissue type for the MYC sample

analysis; MYC data will be analyzed with a bioinformatics pipeline in development

that will define approximately 100 suitable housekeeping genes for normalization.
Conclusions: We describe a proof-of-principle for a clinically validated wholetranscriptome RNA-Seq assay from archival FFPE tumor tissue in order to detect
overexpression of clinically relevant genes in cancer patients.

B-203
Diagnositc yield of chromosomal microarray for individuals with developmental
disabilities or congenital anomalies.

A. Repraz-Andrade, C. Torreira Banzas, M. Pombar Prez, M. Blanco

Prez, O. Blanco Barca, M. A. Andrade-Olivi, J. R. Fernandez Lorenzo.
Complexo Hospitalario Universitario de Vigo, Vigo, Spain
Chromosomal microarray (CMA) is increasingly utilized for genetic testing of
individuals with unexplained developmental delay/intellectual disability (DD/
ID), autism spectrum disorders (ASD), or multiple congenital anomalies (MCA).
Guidelines recommend the use of CMA as first-tier clinical diagnostic test for
individuals with developmental disabilities or congenital anomalies.
OBJECTIVES To implement an algorithm for array comparative genomic
hybridization (aCGH) testing in patients with unexplained unexplained DD/ID, ASD,
or MCA. To assess the diagnostic yield of aCGH for individuals with developmental
disabilities or congenital anomalies and compare it to the literature data.
METHODOLOGY We performed, since 06/2009, aCGH to 161 patients referred
from the Neuropediatrics department of our hospital with unexplained DD/ID, ASD,
or MCA. DNA was extracted with an automated method, QIAamp DNA Blood Mini
Kit in a QIACube instrument (QIAgen). aCGH was performed in the beginning
with Nimblegen CGH ISCA Plus 6x630K arrays (Roche Diagnostics) and since its
exit from the microarray business with the Signature Genomics CGX-HD 4x180K
arrays (Perkin Elmer). Both probes design follow the ISCA consortium guidelines
(International Standards for Cytogenomic Arrays). Results were reported following
ISCN 2013 recommendations. If a CMA variant was observed, parental samples were
analyzed to assess whether it is a de novo or an inherited alteration. Copy number
variations are assigned the following interpretations: Abnormal (well established
syndromes, de novo variants and large changes); VOUS (variants of unknown
significance) and likely benign (not previously reported but inherited from a healthy
parent). Diagnostic yield was defined as the number of patients with abnormal variants
divided by the total number of patients tested.
RESULTS 161 patients and 42 parents were studied. 110 of the 161 patients (68,32%)
had a normal aCGH result. 51 patients (31,68%) showed an abnormal result. After
analyzing every single case and performed parental tests, we classified the alterations
as follows: 21 abnormal (13,04%), 7 likely benign (4,35%), 1 VOUS (0,62%) and
9 are still pending parental aCGH results (5,59%). Abnormal variants deletion or
duplication size varied from 200 Kb to 8 Mb. The diagnostic yield, calculated as 21
abnormal patients divided 161 patients, is 13,04%.
CONCLUSIONS aCGH showed a much higher diagnostic yield than conventional
cytogenetics techniques for the diagnosis of unexplained unexplained DD/ID, ASD,
or MCA. These figures are according to the literature. Our results could be even
higher as there are still 9 cases pending parental aCGH results, so these variants can
be reclassified as abnormal or likely benign. The use of CMA as a first-tier clinical
diagnostic test for individuals with developmental disabilities or congenital anomalies
has proven to surpass the classical approach with conventional cytogenetics. The use of
CMA is cost effective in a child with DD/ID, ASD, or MCA. CMA is not inexpensive,
but the cost is less than the cost of a G-banded karyotype plus subtelomeric FISH plus
other techniques such as MLPA, and the yield is greater.

B-205
A retrospective analysis of Western Blot findings and subsequent Nucleic Acid
Amplification Testing of samples with Immunoassay Screening positive for
Human Immunodeficiency Virus

A. Omar, S. D. S. Cosio, W. B. Y. Lee, M. Wong. Khoo Teck Puat Hospital,

Singapore, Singapore
Background: With the first clinical observation of Acquired Immune Deficiency
Syndrome (AIDS) in 1981, the entire medical landscape pertaining to the diagnosis,
treatment and monitoring of an infection by the Human Immunodeficiency Virus
(HIV) has progressed continuously. The western blot assay is the current method of
choice for confirming HIV results positive by immunoassays. We compared Western
Blot outcomes against viral load quantification on immunoassay-positive samples.
A total of 70 anonymised samples that were screening-positive on the Roche HIV

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S191

Molecular Pathology/Probes

Wednesday, July 30, 9:30 am 5:00 pm

Combi or Combi PT immunoassay (Roche Diagnostics, Switzerland) were sent for
western blot confirmation and also tested for viral load using the improved Roche
Cobas Taqman HIV1 Monitor version 2 (Roche Diagnostics, Switzerland).
Samples are determined to be western blot positive if any 2 bands for p24, gp41,
gp120/160 or 2 of 3 envelope bands with or without Group Antigen and/or polymerase
bands are present, as defined by Centres for Disease Control and Prevention and
World Health Organisation criteria respectively. Indeterminate results are defined as
the presence of bands that do not meet positive criteria while inconclusive findings are
those that do not fit negative, positive or indeterminate criteria.
Methods: A total of 84 anonymised samples were screened for infection by the Human
Immunodeficiency Virus, using the Roche Diagnostics immunoassay HIV Combi or
Combi PT. Seventy samples were determined to be at the grey zone cut-off index of
0.9 or greater, indicating presumptive positivity. These samples were subsequently
sent for western blot analysis at the National Reference HIV Laboratory with split
samples tested using the HIV Monitor v2 on the Roche Taqman 48.
Results: Of 84 samples tested, 14 samples were determined to be screening-negative
with no viral RNA detected by nucleic acid testing. Thirty-four samples were screeningpositive and subsequently confirmed by western with viral loads of log 1.53 to 6.00
copies per milliliter. Two samples were defined as western blot indeterminate, with
no virus detected, while the remaining 7 indeterminate samples and 1 inconclusive
sample had viral loads of log 5.33 to 6.48 copies per milliliter. The mean and standard
deviation of viral loads for western blot non-positive and non-negative samples were
6.17 and 0.44 respectively. Western blot positive samples yielded a lower average
value of log 4.62 copies per milliliter with a standard deviation of 0.73.
Conclusion: While there is a robust concordance between definitive western blot
outcomes and nucleic acid testing, the majority of results classified as inconclusive
and indeterminate were associated with significant viral loads. This is not unexpected
as the western blot is an antibody assay. Re-testing is usually recommended at least
4 weeks after the first result for these patients. Thus, from a clinical perspective,
treatment with anti-retroviral prophylaxis in these patients may potentially be delayed
until such time when the western blot turns positive with full seroconversion, possibly
leading to a sub-optimal outcome for these patients.

B-206
Development and evaluation of a new molecular diagnostic method for HER2
testing

W. Chen1, K. Chiu2, R. Yu2, Y. Liu3, H. Ko4, H. Wu1, W. Hsieh1, Z. Lee5, P.

Zhang5, M. Liu4, K. Chang1, I. Su1, Y. Yeh4. 1National Cheng Kung University
Hospital, Tainan 704, Taiwan, 2Genetics Development Corporation, Lake
Bluff, IL 60044, USA, IL, 3Curiemed Corporation, Hsinchu 30076, Taiwan,
4
General Biologicals Corporation, Hsinchu 30076, Taiwan, 5Shanghai
GenePharma Co. Ltd,, Shanghai 201203, China
Background: HER2 PCR Testing results have been published for decades with
favorable concordance with IHC and FISH testing results. However, on the ground
of insufficient evidence, current guidelines still exclude PCR-based method
from determining HER2 status. To find a remedy, we surveyed and found that the
incompatible and insufficient evidence was primarily due to the lack of consistency in
PCR assay design. Here, we propose a standardized HER2 RT-PCR assay with proven
reproducibility for adoption with the hope that a standardize HER2 RT-PCR testing
will gain the traction in creating more compatible data toward achieving the goal of
having ASCO/CAP to include HER2 RT-PCR assay in HER2 testing guideline.
Methods: One-step RT-PCR with external calibrators was utilized to quantify HER2
RNA copies in samples. To evaluate the performance and concordance between
RT-PCR and FISH test, breast invasive ductal carcinoma (IDC) or adjacent normal
specimen was collected and processed into FFPE or OCT sample immediately after
biopsy or surgery. RNA samples were determined to obtain optimal final concentration
of 25g/mL. HER2 positive cutoff value was established by conducting a statistical
population study on tumor versus adjacent normal samples.
Results: The intra- and inter-run coefficient of variations of the assay were <5% (in
terms of concentration in log10). The dynamic range of the assay was between 5.4x1035.4x1012 copies/mL (R2=0.99). Moreover, around 85% overall percentage agreement
(OPA) was obtained when compared with FISH test with FFPE or OCT samples.
Conclusion: A standardized RT-PCR assay and user procedure presented here can
produce consistent test results regardless of the types of patient samples and has high
potential to be included in the ASCO/CAP HER2 testing guideline.

S192

HER2 RT-PCR versus FISH test in

FISH test
Positive Negative
NCKU hospital
Positive 16
3
FFPE
Negative 4
24
HER2 RT-PCR samples Sum
20
27
Positive
8
2
test
OCT
Negative 2
16
samples Sum
10
18
PPA: Positive Percentage Agreement;
PNA: Negative Percentage Agreement;
OPA: Overall Percentage Agreement

Agreement Score
Sum PPA PNA OPA
19
28
47 80% 89% 85%
10
18
28

B-207
Real-time NASBA Targeting HPV E6/E7 mRNA Overcomes Low Specificity of
HPV DNA Test

G. Kim1, Y. Choi2, J. Munkhdelger3, D. Lee4, S. Kim1, Y. Kim1, H. Kim1,

S. Park1, S. Ahn1, J. Kim1, H. Jin5, S. Park6, K. Park3, H. Lee1. 1Department
of Biomedical Laboratory Science, College of Health Sciences, Yonsei
University, Wonju, Korea, Republic of, 2Department of Biomedical
Laboratory Science, Songho College, Hoengseong, Korea, Republic of,
3
Department of Pathology, Wonju College of Medicine, Yonsei University,
Wonju, Korea, Republic of, 4Department of Clinical Laboratory Science,
Hyejeon College, Hongseong, Korea, Republic of, 5Department of Clinical
Laboratory Science, College of Health Sciences, Catholic University of
Pusan, Busan, Korea, Republic of, 6Department of Clinical Laboratory
Science, College of Health and Therapy, Daegu Hanny University, Daegu,
Korea, Republic of
Background: Cervical cancer is one of the most common cancers in women, and
ranks second by female cancer mortality rate. About 40 of human papillomaviruses
(HPV) can infect the cervix and are divided into high- and low-risk groups based on
their frequent detection in carcinoma or low-grade lesions, respectively. Persistent
infection with these types is the major risk factor for the development of cervical
cancer. So it was recommended HPV genotyping test as complementary diagnosis.
But, HPV DNA was detected both in cancerous tissues and in normal tissues
according to cytological diagnosis. Several studies have suggested that detection of
E6/E7 oncogene transcripts of high-risk HPV types would provide higher specificity
in screening for the risk of development of high-grade of cervical samples. The
synergistic effects of E6 and E7 proteins results in disturbance of cell cycle regulation,
prevention of apoptosis, and, ultimately, transformation and survival of neoplastic and
dysplastic cells.
Methods: This study aims to evaluate the clinical performance of the commercial
kit targeting HPV E6/E7 mRNA and compare it with HPV DNA genotyping for
the detection of high-grade squamous intraepithelial lesions (HSIL) and cancer in a
Korean population. NucliSENS EasyQ HPV kit (bioMrieux, Marcy, France) using a
nucleic acid sequence-based amplification (NASBA) technique is chosen. This assay
utilizes molecular beacon probes for real-time detection and typing of E6/E7 mRNA
from HPV genotype 16, 18, 31, 33 and 45.
Results: HPV DNA tests were positive in 100% and 53% of abnormal (SCC, HSIL,
ASC-H, LSIL, and ASC-US) and normal cytology cases, respectively. Positivity rates
of HPV E6/E7 mRNA assay were 75%, 74%, 60 %, 56%, 29% and 9% for SCC,
HSIL, ASC-H, LSIL, ASC-US, and normal cases, respectively. The data from this
study seems clearly show that the real-time NASBA for diagnosis of cervical cancer
has higher clinical specificity than HPV DNA genotyping test, since the positivity rate
detected by the real-time NASBA in normal samples was significantly lower than that
detected by the HPV DNA genotyping test. However, the positive rate detected by
the real-time NASBA in cancerous samples was much lower than that detected by the
HPV DNA genotyping test. In this respect, the sensitivity of the real-time NASBA for
cervical cancer diagnosis seems to be lower than current HPV genotyping test. The
main reason of this low sensitivity was primarily due to the different prevalence of the
high-risk cervical cancer causing HPV genotypes of Korea.
Conclusion: Data from the current study suggests that the NucliSENS EasyQ HPV
E6/E7 assay (bioMrieux) has higher specificity than DNA assays, and can overcome
DNA assays shortcoming of low specificity in clinical detection of high-grade cervical
lesions and ability to predict the risk of their development in HPV-infected women.
Expanding the types of HPV targeted by mRNA assays may increase sensitivity as
well as specificity of detection of high-grade cervical lesions.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Molecular Pathology/Probes

Wednesday, July 30, 9:30 am 5:00 pm

varies greatly, we determined the genotype frequency of 511 patients in our service
and found a frequency of 51.86% CC, 41.49% CT and 6.65% TT for LCT-13910 and
51.66% GG, 40.70% GA and 7.63% AA for LCT-22018.

B-208
Bioinformatics analysis to determine prognostic mutations of 72 de novo acute
myeloid leukemia cases from The Cancer Genome Atlas (TCGA) with 23 most
common mutations and no abnormal cytogenetics

K. J. Welsh, E. Nedelcu, A. Wahed, Y. Bai, A. Dasgupta, A. N. D. Nguyen.

University of Texas Health Science Center at Houston, Houston, TX
Objective: Acute myeloid leukemia (AML) is a heterogeneous malignancy with
many described karyotypic and molecular abnormalities. Recurrent karyotypic
abnormalities in AML have been well established for treatment. However,
approximately half of AML patients have no karyotype abnormality (CN-AML). An
important issue in the treatment of AML is how gene mutation patterns may help
physicians guide the management of patients in daily practice. The Cancer Genome
Atlas (TCGA) database has available data from 200 cases of de novo AML including
cytogenetics, 260 gene mutations, and survival duration for each case. As previously
reported, in this database a total of 23 genes were significantly mutated and another
237 were mutated in two or more samples. We utilize clustering analysis on this
database to correlate the presence of 23 common mutations with the prognosis of de
novo CN-AML cases.
Methods: Cases with positivity for the most common 23 mutations and no cytogenetic
abnormalities were selected from the TCGA. Unsupervised neural network analysis
with NeuroXL Clusterizer (OLSOFT LLC, Moscow) was performed on these cases to
group them into clusters according to their pattern of mutations and survival duration.
The next step was to find interaction between wild-type genes, gene mutations (with
their associated proteins) and common chemotherapy to gain more insight into
response of CN-AML patients to treatment, which was achieved with Ingenuity IPA
software (Ingenuity Systems, Inc., Redwood City, CA).
Results: 121 cases with positivity for the 23 most common mutations were obtained
from the original set of 200 AML cases. Subsequently, 72 cases with no cytogenetics
abnormalities (CN-AML) were obtained from these 121 cases. Within the 72 CNAML cases, the following mutations were not present: TP53, NRAS, KIT, EZH2,
and HNRNPK, leaving 18 mutations in this subset of patients. Using appropriate
threshold for mutation frequency (75%), clustering was found to be based on only
4 mutations: NPM1, FLT3-ITD, RUNX, and DNMT3A. The following prognostic
groups were found: (a) good: NPM1, CEBPA, or TET2, (b) intermediate: NPM1/
DNMT3A, or other mutations, (c) poor: RUNX1, FLT3-ITD/NPM1, FLT3-ITD/
CEBPA, or FLT3-ITD. Pathway analysis revealed significant causality between the
mutations FLT3-ITD, NPM1, DNMT3A, IDH2, RUNX1, TP53, KIT, and CEBPA and
the chemotherapy agents cytarabine, idarubicin, fludarabine, topotecan, etoposide,
hydroxyurea, dexamethasone, methotrexate, and decitabine.
Conclusions: Combinations of mutations appear to dictate the clinical behavior of
AML in terms of prognosis. This study provides further molecular characterization
and prognostic data for the heterogeneous group of CN-AML patients.

B-209
Validation of a genetic test for lactose tolerance in a Brazilian hospital

N. H. Muto, L. Oyakawa, J. R. Pinho, R. Sitnik. Hospital Albert Einstein,

So Paulo, Brazil
Background: Lactose, the predominant carbohydrate in breast milk, is hydrolyzed
in the intestinal mucosa by lactase-phlorizin hydrolase (LPH). Lactase is produced
during the weaning and undergoes a physiological decline in adult, which leads to
lactose intolerance symptoms. In some populations LPH activity persists throughout
life, a condition known as adult lactase persistence. The conventional diagnostic
test is performed challenging the patient with lactose, which can cause diarrhea and
abdominal pain in intolerant patients. Several studies have shown that two single
nucleotide polymorphisms in the introns of a helicase (MCM6) upstream the lactase
gene [rs4988235 (LCT-13910C/T) and rs182549 (LCT-22018G>A)] correlate with
lactose intolerance through differential transcriptional activation of the lactase
promoter. Intolerance genotype is a recessive trait in both SNPs: CC-13910 and GG22018.
Methods: In order to offer a molecular test that will not need lactose consumption,
we developed two Real Time PCRs reactions to target those SNPs. DNA was isolated
from peripheral blood using EasyMag Extractor (Biomeriux) and submitted to Real
Time PCR using TaqMan SNP Genotyping Assay kits (Applied Biosystems) with
probes specific for each SNP.
Results: During validation, we tested 59 samples with previous conventional test
results and found a 91.5% correlation and 100% reproducibility of all possible
genotypes. Considering that lactose tolerance prevalence in different populations

Conclusion: Molecular test was implemented in our Clinical Laboratory in

November, 2012 and since then has been an excellent option for patients instead of
the conventional test. Approximately, half of tested patients were intolerant to lactose.
Those data are in accordance with previous Brazilian studies, and may reflect the great
admixture in Brazilian population.

B-210
Genotyping of drug-metabolizing enzymes CYP2D6, CYP2C19, CYP2C9,
CYP3A4 and CYP3A5 in patients prescribed pain medications

Z. Su, G. McIntire, S. Holt, F. Wallace, J. Putnam, J. Zimmerman. Ameritox

Ltd., Greensboro, NC
Background: The cytochrome P450 (CYP450) superfamily represents a group of
enzymes which are associated with the metabolism of more than 90% of human
drugs. To date, a large number of allelic variants as well as copy number variations
(CNV) have been reported and the number of alleles is still growing. Many genetic
polymorphisms within the CYP450 superfamily result in altered enzyme expression
or activities, significantly effecting drug metabolism. Based on the genotypes of the
CYP450 enzymes, patients may be categorized as ultra-rapid metabolizer (UM),
extensive metabolizer/normal metabolizer (EM), intermediate metabolizer (IM), and
poor metabolizer (PM). Even though the value of routine pharmacogenetic testing
is still debated, accumulated evidence indicates that genotyping drug-metabolism
enzymes will help physicians to tailor pain treatment to individual patients.
Methods and Materials: DNAs were extracted from buccal swabs collected from 200
patients prescribed pain medications. Five genes in CYP450 superfamily: CYP2D6,
CYP2C19, CYP2C9, CYP3A4, and CYP3A5, were genotyped. Genotypes of the 5
genes were reported using the star (*) allele nomenclature. 16 alleles (*2, *3, *4, *5,
*6, *7, *8, *9, *10, *11, *12, *14, *15, *17, *29, *41) and copy number of CYP2D6;
10 alleles (*2, *3, *4, *5, *6, *7, *8, *9, *10, *17) of CYP2C19, 2 alleles (*2, *3)
of CYP2C9, 6 alleles (*1B, *2, *3, *12, *17, *22) of CYP3A4, and 8 alleles (*1D,
**2, *3, *3B, *6, *7, *8, *9) of CYP3A5 were tested. Allele *1 is assigned by default
when no other listed alleles are detected. Phenotypes were assigned based on the
guidelines from the clinical pharmacogenetics implementation consortium (CPIC)
and/or published general rules: 1) UM: more than 2 copies of functional alleles;
2) EM: 2 copies of functional alleles, or 1 copy of functional allele and 1 copy of
decreased functional allele; 3) IM: 2 copies of decreased functional alleles, or 1 copy
of functional and 1 copy of non-functional alleles, or 1 copy of non-functional and 1
copy of decreased functional alleles; 4) PM: 2 copies of non-functional alleles.
Results: Among the 200 specimens, 11 allelic variants (*2, *3, *4, *5, *6, *9, *10,
*11, *17, *41) in CYP2D6, 3 allelic variants (*2, *9, *17) in CYP2C19, 2 allelic
variants (*2, *3) in CYP2C9, 3 allelic variants (*1B, *3, *22) in CYP3A4, and 4
allelic variants (*1D, *2, *3, *6) in CYP3A5 were detected. The most common
genotypes and phenotypes are: 1) for CYP2D6: *1/*2 (EM, 20%), *1/*4 (EM,
13.5%), and *1/*41 (EM, 10.5%); 2) for CYP2C19: *1/*1 (EM, 38.5%), *1/*17 (UM,
27%), and *1/*2 (IM, 19%); 3) for CYP2C9: *1/*1 (EM, 67.5%), *1/*2 (EM, 22.5%)
, and *1/*3 (IM, 8%); 4) for CYP3A4: *1/*1 (EM, 80.5%), *1/*1B (EM, 10%); and
*1/*22 (EM, 4.5%); and 5) for CYP3A5: *3/*3 (PM, 76.5%), *1/*3 (IM, 16%), and
*3/*6 (PM, 1.5%).
Conclusions: The frequencies of allelic variants and genotypes we found in the five
genes are in accordance with preciously published results. In our targeted population,
the most common allelic variant found was the non-functional *3 allele in CYP3A5,
with a frequency of 86%.

B-211
Development of an LNA-Blocker enhanced, allele-specific, loop-mediated
isothermal amplification (AS-LAMP) method for detection of single base
mutations in beta-thalassemia patients

T. Zhong, R. Hu, J. Ye, L. Jiang, X. Hu. The First Affiliated Hospital of

Gannan Medical University, China, Ganzhou, China
Background: It was found that five mutations, namely 654M, 41/42M, 28M, 17M and
27/28M, are the most commonly occurring single base mutations in beta-thalassemia
patients in China. Being able to detect and distinguish these mutations is important
for patient screening and for diagnosis of the disease. The loop-mediated isothermal
amplification (LAMP) method is known to be a rapid (complete in less than 45 min)
and simple way for detection of gene deletion and insertion events, but is not specific

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S193

Molecular Pathology/Probes

Wednesday, July 30, 9:30 am 5:00 pm

enough to distinguish single base difference. This study was aimed to improve the
sensitivity of LAMP method on single base mutation detection in beta-thalassemia
patients by using Locked nucleic Acid (LNA) modified primers in the reaction.
Methods: The LNA-Blocker enhanced, allele-specific, loop-mediated isothermal
amplification (AS-LAMP) was carried out under isothermal condition. To enhance
the specificity of AS-LAMP, an LNA modified allele-specific primer and an LNA
modified wild type blocking primer were used in the reaction. LNA is a high affinity
DNA analogue that has increased target specificity and high melting temperature. This
method was used to detect and distinguish single base mutations in beta-thalassemia
patients. This method was validated with positive and negative controls and with
145 clinical samples of patient genomic DNA (46 positive samples containing all 5
mutations and 99 negative samples).
Results: The LNA-Blocker enhanced AS-LAMP method showed high specificity
with either plasmid or genomic DNA targets in reactions. The assay had a detection
limit of approximately 100 copies of the target sequence (95.7% in sensitivity, 44/46),
comparable to the traditional AS-LAMP. However, the specificity of LNA-Blocker
enhanced AS-LAMP (96.0%, 95/99) was much higher than the traditional AS-LAMP
(75.8%, 75/99). Multiplex detection of all 5 targets was also achieved with LNABlocker enhanced AS-LAMP.
Conclusion: LNA-blocker enhanced AS-LAMP provides a highly specific isothermal
method for detection of single mutations and for screening of single mutations in
beta-thalassemia patients.

B-212
Detection of fetal aneuploidies by quantitative fluorescent polymerase chain
reaction in the Brazilian population

F. F. Coelho, M. S. Goncalves, G. Lopes, V. O. Almeida, N. Fonseca, A. C.

S. Ferreira. Instituto Hermes Pardini, Vespasiano, Brazil
Background: Aneuploidy occurs from non-disjuntion during gametogenesis and
results in an abnormal number of chromosomal in the gametes. The gamete joined in
fertilization contains an extra copy of one chromosome or one of the chromosomes
is missing. These alterations can results in trisomy or monosomy in the fetus or
embryo which are frequently associated with severe congenital abnormalities,
mental retardation and shortened life expectancy. The majority of first trimester
abortion cases is caused for aneuploidies. Quantitative fluorescent polymerase chain
reaction (QF-PCR) is a method which has been regularly used for the diagnosis of
common chromosomal abnormalities in recent years with low error rates. Objective:
To determine the frequency of the most common chromosomal aneuploidies
causing abortion in Brazilian population. Methods: The study is the retrospective
statistical analyses of data registered on 70 women submitted to molecular studies
of chromosome disorders at the Hermes Pardini Institute, Belo Horizonte, Minas
Gerais state, Brazil in 2013. All patients had spontaneous abortion and the ovular
remains, embryo tissue were analyzed after the DNA extraction. Was compared
the genetic profile of the sample with the maternal blood, a differential that allows
to detect maternal contamination in the sample, improving the reliability of the
results. Contaminated samples were excluded from the study. After that all samples
were tested using quantitative fluorescent polymerase chain reaction (QF-PCR) for
detection or exclusion of aneuploidy in chromosomes 13, 15, 16, 18, 21, 22, X and
Y (Chomo Quant QF-PCR Kits). Results: The age average of the patients was 34.4
4.7 years old. We detected 44 subjects (62.9%) with aneuploidies or euploidies.
Nine cases of trisomy 15 (12.9%), nine cases of trisomy 22 (12.9%), seven cases
of monosomy X (10.0%), four cases of trisomy 13 (5.7%), four cases of trisomy 16
(5.7%), four cases of trisomy 21 (5.7%), five cases of triploidy (7.1%), a case of
trisomy 18 (1.4%) and a case of paternal isodisomy (1.4%) were detected by QF-PCR.
Conclusions: Quantittaive Fluorescent PCR is a robust and accurate method for rapid
aneuploidy detection. One the first suspicions of spontaneous abortions is numerical
abnormalities in 13, 18, 21 and sex chromosomes, representing 22.8% of our cases
of aneuploidy. But the results show the importance of also being analyzed changes in
chromosomes15, 16 and 22 which concentrated 31.5% of the aneuploidies detected.
We are going to analyze a larger number of samples in order to confirm the importance
of study chromossomes 15, 16 and 22 disorders.

B-214
Detection of Von Hippel - Lindau (VHL) Gene Copy Number Variations Using
Digital Droplet PCR

D. Milosevic, S. K. Grebe, A. Algeciras-Schimnich. Mayo Clinic and

Foundation, Rochester, MN
Background: Our current assay for detection of large deletions in the VHL gene uses
a combination of MLPA based probes (MRC-Holland b.v) and Luminex FlexMap
technology. It is a two-day assay that requires 400ng of DNA input, and consists
of overnight probe hybridization, amplification, bead hybridization and Luminex
detection. Due to these complex workflows, assay failures are not rare. Digital
droplet PCR (ddPCR) might represent a faster, more sensitive and more reliable
alternative. DdPCR is based on traditional PCR amplification and fluorescent probebased detection methods, but partitions each reaction into multiple nanodroplets.
Quantitation is based on counting the proportion of droplets that show amplification.
Poisson statistics are then applied to back calculate the copy number in the original
sample. This allows for highly sensitive and reproducible absolute quantification of
nucleic acids without the need for standard curves.
The objective of this study is to develop a method for detection of single or multiple
exon deletions in VHL using digital droplet PCR (ddPCR).
Methods: Prior to PCR cycling, reactions are prepared as for traditional fluorescent
qPCR methods. Fluorescent probes and two primer sets encompassing 80-120bp of
each VHL exon are added to each reaction mixture, which also includes an additional
reference gene. The final reaction mix is then partitioned into thousands of nanodroplets. Ideally, the starting DNA concentration is such that this will result in a
mixture of droplets containing either zero or one template molecule per droplet. The
droplets are pooled back into a PCR tube, and cycling is performed. At the conclusion
of cycling, negative and positive droplets are counted in a flow cytometer like device.
Total assay time is approximately three hours.
Various input concentrations of primers, probes and DNA were tested during
validation. Intra- and inter-assay concordance of copy number assessment was tested
on three different DNA specimens with and without deletions. Patient specimens
containing deletions of one, two, or all three exons of the VHL gene were used for
method comparison between MLPA and ddPCR. Finally, formalin fixed, paraffin
embedded (FFPE) specimens, which habitually fail MPLA, were re-tested by ddPCR
Results: Primer and probe concentrations and DNA input were optimized and
standardized for the method, with 10ng of DNA input required for successful
copy number estimation. A direct comparison of TaqMan probes with 3-MGBlabeled probes showed that MGB-labeling provided better specificity. Copy
number assessment concordance was 100% between runs. The method comparison
showed 100% concordance between MLPA and ddPCR. The FFPE specimens were
successfully assayed by ddPCR while failed testing by MLPA.
Conclusion: DdPCR for copy number estimation of the VHL gene variants is fast,
reliable and accurate. It requires minimal nucleic acid input, with a 40-fold reduction
of input DNA compared to MLPA. Because of this advantage, difficult specimen
types, including FFPE tissue, are now capable of being characterized. Additionally,
same-day results are available with the ddPCR method, reducing total run-time from
48- to 3- hours.

B-215
Absolute Quantification of Graft derived cell-free DNA (GcfDNA) early after
Liver Transplantation (LTx) using Droplet Digital PCR

J. Beck1, J. Schmitz1, P. Kanzow2, O. Kollmar3, M. Oellerich2, E. Schtz1.

1
Chronix Biomedical, Gttingen, Germany, 2Dept. of Clinical Chemistry,
University Medical Center, Gttingen, Germany, 3Dept. of General,
Visceral, and Pediatric Surgery, University Medical Center, Gttingen,
Germany
Background: The diagnostic value of GcfDNA as measure of graft integrity after
LTx has been recently proven [1,2]. The yin and yang of using percentage values
vs. absolute GcfDNA quantification is, nevertheless, under discussion [3]. Where
the ratio of graft to host cfDNA has analytical advantages by eliminating disturbing
variables, such as DNA extraction efficiency, variabilities in host cfDNA may
obfuscate the view on the engrafted organ. The early phase after LTx was used as
model to interrogate whether the percentage or absolute plasma concentration of
GcfDNA is a more valuable graft integrity measure.
Methods: GcfDNA percentage was determined by droplet digital PCR (ddPCR - BioRad) as described [1]. A synthetic sequence of non-human origin (average length

S194

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Molecular Pathology/Probes

Wednesday, July 30, 9:30 am 5:00 pm

of cfDNA) was spiked into 1mL plasma, and quantified with ddPCR after DNA
extraction in one fluorescent channel. The total cfDNA was quantified using two
combined human genomic dPCRs in the second channel as copies/mL (cp/mL). Total
cfDNA was calculated using the spike-in without being extracted to assess the DNA
extraction efficiency in each batch. GcfDNA concentration was defined as of the total
cfDNA(cp/mL) x GcfDNA%. Plasma obtained during the first 10 days after LTx from
15 patients (one split-LTx) was investigated
Results: Ten repeated extractions of the same plasma pool from healthy volunteers
yielded an average of 1069 diploid genomic cp/mL plasma with a CV of 7.5%. Of
185 samples six showed a low (<50%) extraction efficiency; the remainders had
an average of 67%+9%. The total cfDNA was highly variable peaking at 6hr after
reperfusion (3.9x1052.0x105cp/mL) weaning to 1.3x1050.9x105cp/mL at day 10.
The respective GcfDNA was 3.1x1051.8x105cp/mL (6hr) and 1.5x1030.9x103cp/
mL (day10). The correlation between GcfDNA% and GcfDNA(cp/mL) values was
weak (r=0.61;p<0.05). A comparison of the AUC (day1-day5) of AST with GcfDNA
percentage and concentration showed a better association with absolute GcfDNA
(r=0.65;p<0.05) compared to percentages (r=0.31;p=0.27). The initial half life
was 1.30.6days for GcfDNA(cp/mL) and 2.91.6days GcfDNA(%), compared to
2.61.2days for AST.

60,3% (A/A), 33,6% (A/C), 6,1% (C/C); HFE C282Y 96,0%(G/G), 4,0%(G/A), HFE
H63D 78,1%(C/C), 20,3%(C/G), 1,6% (G/G); HFE S65C 98,1% (A/A), 1,9% (A/T).
The reproducibility test showed 100% of concordance among the replicates.
Conclusion: The Open Array genotyping method was used to detect simultaneously
7 different mutations in thrombophilia, folate and hemochromatosis related genes.
Compared to other genotyping methods such as PCR-RFLP and sequencing, this
method is, easy to perform and useful for high-throughput routines. The results found
describe the frequency of SNPs related to diseases not well established by previous
literature for Brazilian population. They are important to highlight the genetic profile
of Brazilian blood donors.

Conclusion: A robust and precise ddPCR method for absolute quantification of

GcfDNA, was developed, combining the analytical advantages of graft/host ratio (e.g.
eliminating possible bias from interferences), with a robust quantification of total
cfDNA. The GcfDNA concentration seems better associated with AST-values early
after LTx and showed more rapid dynamics than GcfDNA percentage. Even though
the initial post Tx phase, with highly variable amounts of total cfDNA, is particularly
complicated, this method may also provide a better view on graft integrity in other
situations, where the host cfDNA is increased due to non-transplantation related
causes. As to whether the clinical utility is improved compared to percentage values
for stable patients as well, is subject to further investigations.
1. Beck J. et al. (2013) Digital droplet PCR for rapid quantification of donor DNA in
the circulation of transplant recipients(...). Clin.Chem. 59:1732-1741.
2. Oellerich M. et al. (2014) Use of Graft-Derived Cell-Free DNA as an Organ
Integrity Biomarker to Reexamine Effective Tacrolimus Trough Concentration(...).
Ther.Drug.Monit. Epub
3. Lo YM (2011) Transplantation monitoring by plasma DNA sequencing. Clin.Chem.
57:941-942.

B-220
Frequency of G1691A (FV), G20210A (FII), C677T, A1298C (MTHFR),
C282Y, H63D E S65C (HFE) among Brazilian blood donors and evaluation of
OpenaArray method for SNP detection.

V. D. T. Niewiadonski1, J. Vieira2, C. Almeida3, N. Gaburo Jr.1, E. C.

Sabino2. 1DASA, Sao Paulo, Brazil, 2Instituto de Medicina Tropical- USP,
Sao Paulo, Brazil, 3Fundao Pr-Sangue Hemocentro de So Paulo, Sao
Paulo, Brazil
Background: OpenArray is a new PCR technology that uses a microscope plates
and allow the detection of many SNPs at the same time.
Objective: We have used this technology to describe the genotypic frequencies of some
SNPs related to venous thrombosis (G1691A and G20210A), to hyperhemocistinemia
conditions (C677T, A1298C), and to hereditary hemochromatosis (C282Y, H63D and
S65C), among Brazilian blood donors.
Methods: We tested 400 blood samples from Fundao Pr-Sangue Hemocentro of So
Paulo, Brazil. The DNA was isolated using QIAamp DNA Mini Kit (Qiagen, Hilden,
Germany). The DNA concentration and quality was determined by spectrophotometry
method using NanoDrop 2000 (Thermo Scientific, Wilmington, DE). The OpenArray
platform (Life Technologies, Foster City, CA) was used to genotype the samples
through TaqMan Genotyping detection system. The reproducibility of this method
was evaluated running nine samples in triplicate in three different assays. The DNA
samples were also tested using TibMol (TibMol, Berlim, Germany) reagents at Light
Cycler 2.0 (Roche) plataform, with FRET as detection system, to compare the results
and determine the accuracy of Open Array method. The genotypic frequency for each
SNP was performed using by Taqman Genotyper v1.3 software (Life Technologies)
Results: Of 400 samples, 392 showed valid results. The results showed agreement
of 100% for all assays tested, exception to HFE C282 (95.5% agreement). The only
discordant sample result for HFE C282Y was confirmed by direct sequencing which
showed that OpenArray result were correct. The calculated frequencies of each SNP
found were FV G1691A 98,8% (G/G), 1,2% (G/A); FII G2021A 99,5% (G/G), 0,5%
(G/A); MTHF C677T 45,5% (C/C), 44,8% (C/T), 9,8% (T/T); MTHF A1298C-

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S195

Nutrition/Trace Metals/Vitamins

Wednesday, July 30, 9:30 am 5:00 pm

Wednesday, July 30, 2014

Poster Session: 9:30 AM - 5:00 PM
Nutrition/Trace Metals/Vitamins

B-221
Serum IL-4 concentration in patients suffering from head and neck cancer

M. F. Godoy1, E. Fabios1, O. Casbarien2, M. S. Feliu1, A. Navigante2,

N. Slobodianik1. 1School of Pharmacy and Biochemistry, Buenos Aires,
Argentina, 2Angel Roffo Institute ., Buenos Aires, Argentina
Interleukin 4 ( IL-4) , is a cytokine that induces differentiation of naive helper T cells
(Th0 cells) to Th2 cells. It has many biological roles, including the stimulation of
activated B-cell and T-cell proliferation, and the differentiation of B cells into plasma
cells. It is a key regulator in humoral and adaptive immunity. IL-4 induces B-cell
class switching to IgE, and up-regulates MHC class II production. IL-4 decreases
the production of Th1 cells, macrophages, IFN-gamma, and dendritic cell IL-12.It
presents an anti-inflammatory role (1). Previous results in a group of adult patients
suffering from head and neck cancer(H&N) showed an altered nutritional and
inflammatory status (2). The aim is to analyze the serum IL-4 concentration in adult
patients(n=17) suffering from H&N ,at the beginning of the specific treatment. The
study was approved by the Ethics Committee of the University of Buenos Aires and
met the recommendations stated in the Helsinki Declaration. All participants gave
written consent before recruitment. Reference values were obtained from a healthy
adult group (n=41, R )(3). Blood samples were collected from fasting patients. Serum
IL-4 concentration was determined by Elisa method (Human IL-4 ELISA set, BD
OptEIATM).Results (pg/mL) expressed as Mean SD were compared to reference
values; H&N showed statistical decreased concentration compared with R at a level
of p <0.01: 5.52.1 vs 8.23.9 with a range between 2.5-9.5 pg/mL This finding point
out to and confirm an inflammatory status in the studied group.
References 1) Margni R. La respuesta inmune. En: Inmunologa e Inmunoqumica.
Fundamentos, 5 Edicin, Buenos Aires: Ed. Mdica Panamericana; 1996, p. 14-32.
2)M S Feliu , C Silva, P Cresta, O Casbarien, A Ross, A Navigante, N Slobodianik.
Biochemical parameters in patients suffering from head and neck cancer. . 5th
International Immunonutrition Workshop .Puerto Vallarta, Mxico 6-8 abril 2011.
Oral Presentation
3)M Godoy , I Fernandez , M S Feliu , E M Insani , N H Slobodianik. Valores sricos
de interleuquina-4 en un grupo de adultos sanos. Acta Bioqum Cln Latinoam 2013;
47 (2): 419-20
Supported by University of Buenos Aires ( Code:20020100100044)

B-223
New HPLC method for determination of vitamin K1 and K2 in human serum

J. Cepova, E. Klapkova, R. Prusa. Faculty Hospital Motol, Prague 5, Czech

Republic
Objective: Vitamin K1 is an essential cofactor in the synthesis of activ blood-clothing
factors, vitamin K2 prevents bone loss and fractures. Objective of this study was to
evaluate new HPLC metod for determination of vitamin K1 and vitamin K2 in human
serum.
Methods: We developed a new HPLC method for the determination of vitamin K1
and vitamin K2 in human serum with fluorescence detection after post-column zinc
reduction. Vitamin K1 and K2 were purchased from Sigma-Aldrich, The internal
standard was obtained from Immundiagnostik AG, Germany. 20 l of internal standard
were added to 500 l of serum and two mL of ethanol were added to precipitate
the proteins. The mixture was extracted with 4.0 mL of hexane for 10 min and than
centrifuged at 3000 rpm for 5 min. The lower layer was additionally re-extracted with
further 4 mL of hexane. The organic layers were than evaporated at 50 C under
a stream of nitrogen. The dry residue was reconstituted with 2 mL of hexane and
solid phase extraction was than used (Sep-Pak, 500 mg, Waters). The cartridges were
preconditioned with hexane, and vitamin K was eluted with diethylether in hexane
(3:97, v/v). Eluates were evaporated under a stream nitrogen at 50 C. The separation
was accomplished on a BDS Hypersil C18, 3 m column at 22 C. The detection was
performed at 246 nm (excitation) and 430 nm (emission). The mobile phase consisted

S196

of 880 mL methanol, 100 mL acetonitrile, 1.1 g zinc acetate, 10 mL acetic acid and 10
mL water, flow rate 1.0 mL/min. The retention times were 5.7 min, 9.2 min, 10.8 min
for vitamin K2, internal standard, vitamin K1, respectively.
Results: A linear relationship between serum concentration and peak area was
obtained for both substances with correlation coefficient r2=0.9959 for vitamin K1 and
r2=0.998 for vitamin K2. The intra and interday accuracy and precision were evaluated
on two QC samples by multiple analysis and coefficients of variation were less than
8%. Mean recoveries of the corresponding compounds were 94.5% and 104%. No
interference has been found between vitamin K1 and vitamin K2 or IS.
Conclusion: The analytical method developed to quantitate vitamin K1 and vitamin
K2 in serum has been succesfully validated.
Supported by the project (Ministry of Health, Czech Republic) for conceptual
development of research organization 00064203 (University Hospital Motol, Prague,
Czech Republic)

B-225
Performance Evaluation of LOCI Vitamin B12 and Folate Assays on the
Dimension EXL integrated chemistry system with LOCI module

C. Briggs, T. Johnson, S. A. Lewisch, L. Schiavoni, J. Thomas, C. Tyler.

Siemens Healthcare Diagnositics, Newark, DE
Introduction: The objective of this study was to evaluate the performance of the
fully automated LOCI vitamin B12 (VB12) and folate (FOLA) assays for the
Dimension EXL integrated chemistry system with LOCI module newly
developed by Siemens Healthcare Diagnostics Inc. The vitamin B12 and folate
assays are homogeneous, competitive immunoassays based on Luminescent Oxygen
Channeling Immunoassay (LOCI) technology. LOCI reagents include two synthetic
bead reagents, chemibead and sensibead, and a biotinylated analyte receptor.
The sensibead is coated with streptavidin and contains a photosensitive dye. The
chemibead is coated with an analyte analog and contains a chemiluminescent dye as
the signal generation component. Before the immunological portion of the reaction is
initiated, the patient sample is pretreated with sodium hydroxide and dithioerythritol
to release analyte from endogenous binding proteins. The assays are calibrated with
the five level multi-analyte LOCI Anemia Calibrator. B12 assay time is 32 minutes
and folate is 21 minutes. Sample volume is 12 L and 10 L for the B12 and folate
assays, respectively.
Methods: Precision estimates were obtained per CLSI EP05-A2 protocol (two
replicates twice a day for twenty days) using quality control materials and human
serum pools. Linearity was assessed through dilution of high and low analyte samples
outside of the measuring interval. Specimen equivalence testing was performed
with matched sets of serum and plasma samples. Method comparisons with patient
samples were conducted versus vitamin B12 (VB12) and folate (FOL) assays on the
Dimension Vista system (X-axis). Accuracy was evaluated by recovery of the World
Health Organization (WHO) Vitamin B12 and Folate International Standard 03/178.
Results: Linearity was demonstrated throughout the measuring interval for the B12
assay of 80 to 2000 pg/mL and 0.5 to 20.0 ng/mL for the folate assay (intervals span
from the LoQ for B12 and LoD for folate to the upper calibration standard). With
automated dilution, the measuring interval is extended to 6000 pg/mL for B12 and
100 ng/mL for folate. Equivalent results were obtained among serum, and lithium and
sodium heparin plasma for both B12 and folate assays and EDTA plasma for the B12
assay. B12 levels were tested at 180, 498, and 978 pg/mL and resulted in repeatability
of 5.6%, 2.3% and 2.5%, and within-laboratory precision of 6.5%, 3.7% and 2.8%,
respectively. Folate levels were tested at 2.1, 6.6, and 16.7 ng/mL and resulted in
repeatability of 4.3%, 4.1% and 2.2%, and within-laboratory precision of 7.6%, 5.5%
and 4.0%, respectively. Passing-Bablok regression statistics for Dimension EXL B12
vs. Dimension Vista B12 were: slope: 1.00, intercept: -2.1, r: 0.999, n: 213, range:
62 to 1973 pg/mL. Simple linear regression statistics for Dimension EXL folate vs.
Dimension Vista folate were: slope: 1.01, intercept: 0.05, r: 0.99, n: 138, range: 0.6 to
19.2 ng/mL. Recovery difference from the WHO Standard 03/178 was 2.1% (target
480 pg/mL) for B12 and -7.3% (target 5.33 ng/mL) for folate.
Conclusions: The study results demonstrate good performance of the fully automated
vitamin B12 and folate assays on the Dimension EXL system.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Nutrition/Trace Metals/Vitamins

Wednesday, July 30, 9:30 am 5:00 pm

rice (17.1%), cassava (9.3%), pigmeat (2.1%), bovine meat (0%) and poultry meat
(0%). Excepting the younger, each commodity elicits a constant and specific pattern
of IgG hypersensitivity over age (figure 1B).

B-227
Evaluation of Vitamin D Levels in Routine of a Private Laboratory

N. O. Lima, M. M. Magalhes, S. Sorato, A. G. L. Guimares, J. P.

D. Padilha, W. Pedrosa, A. C. S. Ferreira. Instituto Hermes Pardini,
Vespasiano, Brazil
Background: Humans get vitamin D from exposure to sunlight and from their diet, is
metabolized in the liver to 25-hydroxyvitamin D, wish is used to determine a patients
vitamin D status. Many studies considered major health problems from vitamin D
deficiency resolved. But vitamin D deficiency is common in dark-skinned persons
living in northern countries, or in Asian countries for example. This study tries to
elucidate the prevalence of vitamin D deficiency and the geographic influencing
factors in population of Brazil.

Conclusion: We identify the most common foods that trigger moderate/strong

IgG response in Brazilians and also the hypersensitive frequency for the major
commodities. Some foods (Cola nut, cows milk, egg white and brewers yeast)
showed an age-dependent IgG serum levels that correlates with an expected agedependent consumption. Conversely, the constant and specific patterns found for
commodities suggest that these foods elicit specific immune response, especially
when confronting the extremes (wheat versus meats).

Objective: Compare serum concentrations of vitamin D (25-hydroxyvitamin D) in

samples analysis in a large laboratory, with nationwide coverage in some periods of
the year representing different sun exposes.
Methods: We use a cross-sectional study in a sample of patients who performed the
dosage of 25OHD in a private laboratory in Brazil. We analyzed 19.185 samples
of both genders aged 18 years. Two months were selected for the study: January
and July 2012, representing the months of highest and lowest incident of sunlight,
respectively.
The study was divided empirically into three geographic regions of different latitudes,
and different exposures to sunlight, designated as Region (0): latitudes greater than
-10, Region (1): latitudes between -10 to -20, and Region (2) latitudes lower than
-20.
Results: The samples collected in July showed concentrations of Vitamin D lower than
those collected in January, with a higher proportion of sample with values considered
deficient (20ng/mL). The difference was statistically significant.
The proportion of deficient patients in January and July respectively was 15,51% and
25,50% with statistically significant difference. The chance of been deficient was 1,87
times higher for samples collected in July (odds ratio; IC95%: 1,73-2,01- p<0,0001).
There were difference in the percentage of deficient patients in Region (2) between
moths of January (17,42%) and July (30,25%). Other studies have find low levels of
25OHD in 42%of seniors in So Paulo and mean serum levels of 21,4ng/mL in Rio
Grande do Sul, both states of Region (2).
Conclusion: The role of vitamin D deficiency in increasing the risk of many common
and serious diseases, including some common cancers, type 1 diabetes, cardiovascular
disease, and osteoporosis. The study confirms that the concentrations of Vitamin D
are much lower when derived from regions of lower latitudes and during the month
of July (compared to January) possibly due to a reduction in the incidence of sunlight.
That conclusion should be taken in consideration and included in the clinical analysis
of the patient and not the test result only itself.

B-228
Food-specific IgG antibodies in Brazilians: a descriptive, laboratory
information management system-based study.

M. S. S. B. Caixeta, L. F. R. Velasco, L. F. A. Nery, S. S. S. Costa, G. B.

Barra. Laboratorio Sabin, Braslia, Brazil
Background: The aim of this study was to identify the most common foods that
trigger IgG immune responses in Brazilians of different age groups and describe the
prevalence of this food hypersensitivity for the most common commodities available
for consumption.
Methods: We retrieved the results from our food-specific IgG antibodies test
performed between December 2012 and December 2013, 281 individuals (unique)
were included and 185 (65.8%) were female. The five most common foods that trigger
moderate (24-30 U/mL) or strong (>30 U/mL) IgG antibody responses were analyzed
and presented by age group. The same was done with the commodities. The age
groups were 1-10 (n=34), 11-20 (n=10), 21-30 (n=33), 31-40 (n=73), 41-50 (n=42),
51-60 (n=49), 61-70 (n=29), 71-80 (n=9). The food-specifics IgG assay was Genarrayt
Microarray (Genesis Diagnostics), which detects total IgG against 221 foods in serum
by ELISA.
Results: The five most commons foods that trigger IgG reaction in the patients
were wheat (96.1%), cola nut (84.7%), cows milk (75.4%), egg white (70.1%) and
yeast (brewers) (64.8%) and their age-dependent relative occurrence frequency
can be found in figure 1A. Considering the commodities, their relative occurrence
frequencies were as follows: wheat (described before), bean (57.7%), maize (40%),

B-229
Concentrations of some trace elements in hair of patients with prostate cancer

S. A. K. Saleh, H. M. Adly. Umm AlQura University, Faculty of Medicine,

Biochemistry Department, Makkah, Saudi Arabia
Backgrounds: Deficiency or excess of trace elements can induce metabolic disorders
and cellular growth disturbance, even mutation and tumorigenesis. Many authors
observed direct association between micronutrient deficiencies and the cancer
mortalities. Prostate cancer ranked the sixth most common cancer among males
in Saudi Arabia and there are few studies of the association between trace element
levels and prostate cancer in Saudi Arabia. Objective: This study aimed to explore the
association between concentration of selected hair trace elements including selenium,
zinc, copper, manganese and iron as long-term biological markers with prostate
cancer in Saudi Arabia. Patients and Methods: The study included 58 patients with
prostate cancer, 64 benign prostate hyperplasia (BPH) patients and 52 healthy male
subjects of matched age. Full history and clinical data were recorded for all subjects.
Prostate cancer patients were undergo digital rectal examination (DRE), trans-rectal
ultrasonography (TRUS) guided biopsy of the prostate, computed tomography (CT)
scan of the pelvis, bone scan and histopathological examination, accordingly prostate
cancer stages and metastatic disease were confirmed. Prostate cancer patients were
classified into localized (n = 46) and metastatic prostate cancer (n = 12). Hair samples
collected from the nape section of all subjects and hair trace elements Se, Zn, Cu, Mn
and Fe levels were analysed by ICP-MS (Perkin Elmer 7300). Odd Ratio (OR) of
trace elements levels in hair were adjusted for family history and smoking. Results:
Mean Se and Zn levels in hair of the prostate cancer group were significantly lower
as compared to BPH and healthy groups (p<0.05) whereas the mean levels of hair Cu,
Mn and Fe were significantly higher in prostate cancer than BPH and healthy groups
(p<0.05). Mean hair levels of Se and Zn were significantly differentiated between
localized and metastatic prostate cancer (p<0.05) whereas mean hair levels of Cu, Mn,
Ni and Fe failed to differentiate these groups (p>0.05). Conclusion: Prostate cancer
may be associated with trace element metabolic disorders. Low levels of Se and Zn
and high levels of Cu, Mn and Fe appear to be associated with the risk of prostate
cancer in Saudi Arabia. Additional prospective studies are needed to confirm the
inverse association between Se and Zn levels and prostate cancer.

B-230
A simple, fast, and sensitive high performance liquid chromatographic method
for measuring vitamins A and E in human plasma

C. Yuan, M. Burgyan, D. R. Bunch, E. Reineks, R. Jackson, R. Steinle, S.

Wang. Cleveland Clinic, Cleveland, OH
Background: Sufficient levels of vitamins A and E are important in maintaining
normal physiological functions in human. The objective of this work was to develop

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S197

Nutrition/Trace Metals/Vitamins

Wednesday, July 30, 9:30 am 5:00 pm

a fast and sensitive high performance liquid chromatography (HPLC) method to
measure the major forms of vitamin A (retinol) and vitamin E (-tocopherol and
-tocopherol) in human plasma. Methods: Two hundreds L of plasma samples and
300 L internal standard solution (5.0 mg/L of -tocopherol acetate in ethanol) were
mixed followed by addition of 1.0 mL of hexane. After vortex and centrifugation,
the supernatant was removed, dried, and reconstituted in 200 L of freshly made
3:1 mix of methanol: diethyl ether of which 25 L was injected for HPLC analysis.
Chromatographic separation was achieved using a C18 column (4.6x75 mm, 3.5 m)
with isocratic methanol elution at 1.8 mL/min. The chromatographic time between
injections was 4.0 min. Retinol was detected by UV, whereas tocopherols were
monitored by fluorescence. The six-point calibration was traceable to a NIST standard
material (SRM 968e). Forty leftover patient specimens were used to compare this
method with an independent HPLC-UV method. Plasma samples collected from 51
healthy donors were analyzed to establish the reference interval. Results: Intra-assay
and total coefficient of variation were <6.0% at three levels tested. No interference
was found from lipemic, hemolytic, icteric and uremic samples. Other performance
validation data are listed in the Table. Agreeable results were obtained for retinol and
-tocopherol comparing to the HPLC-UV method. Discrepancy in the -tocopherol
measurement was likely due to difference in the calibration methods as the independent
HPLC-UV method used -tocopherol calibration curve to quantify -tocopherol.
Conclusion: This HPLC method offers rapid and sensitive quantification of vitamins
A and E in human plasma.
Table. Assay Characteristics
Method Comparison
Standard
Reference
Analyte
Intervals Slope Intercept Correlation Error of
(mg/L)
Coeffi
cient
Estimate
(mg/L)
(mg/L)
Retinol
0.03-5.14 88.5-114.6 0.30-1.20 1.040 0.00
0.979
0.04
-tocopherol 0.32-36.02 87.8-96.7 6.0-23.0 1.115 -0.16
0.944
1.10
-tocopherol 0.10-9.99 96.2-100.2 0.30-3.20 1.358 -0.05
0.968
0.14
Linearity
Range
(mg/L)

Accuracy
Range
(%)

Bias
(%)
3.9
9.5
37.9

B-234
Dynamics of 3-epi-25-hydroxyvitamin D3 in Premature Infants During
Neonatal Intensive Care Unit Hospitalization

A. Anderson-Berry1, G. Jones2, E. Lyden1, M. Kaufmann2, C. Hanson1.

1
University of Nebraska Medical Center, Omaha, NE, 2Queens University,
Kingston, ON, Canada
Objective: Evaluate concentrations of 25(OH)D3 and 3-epi-25(OH)D3 in premature
infants over time.
Relevance: Availability of a valid biomarker to assess vitamin D status of infants is of
great clinical importance. The interpretation of 25(OH)D3 in infants is complicated
by the presence of a 3-epi-25(OH)D3 isomer that may falsely elevate the 25(OH)D3
concentration, while functional significance of this metabolite remains unclear.
Methods: 32 infants 32 weeks gestation were randomized to receive 400 or 800
IU/day of vitamin D3 orally. Serum samples were obtained monthly. Vitamin D
metabolites were analyzed in triplicate using LC-MS/MS. Comparisons of categorical
data was done using Fishers exact test; continuous data was compared using the
Wilcoxon Rank Sum test. Spearman correlation coefficients were used to assess the
correlation of vitamin D metabolites. Measurements over time were fit with linear
mixed effect models. P<0.05 was considered statistically significant.
Results Mean gestational age at birth was 30.5 weeks; mean birth weight was 1405
grams. Mean serum 25(OH)D3 concentrations in cord blood were 17.3 ng/mL; mean
concentrations of 3-epi-25(OH)D were 1.3 ng/mL. Both 25(OH)D3 and 3-epi-25(OH)
D3 increased over time (59.0 and 18.6 ng/mL, respectively), however the percent
of the total 25(OH)D concentration that was 3-epi-25(OH)D3 increased significantly
(p<0.0001 for cord blood vs. 8 weeks, figure 1). 25(OH)D3 and 3-epi-25(OH)D3
concentrations were highly correlated at each time point (r=0.86, 0.61, 0.73, and p=
<0.0001, 0.0002, 0.0004 for cord, 4 week, and 8 weeks, respectively).
Conclusion 3-epi-25(OH)D3 concentrations were low in cord blood, but by 4 weeks
gestational age, accounted for a significant proportion of the 25(OH)D3 level in this
population. The high degree of correlation between total 25(OH)D3 levels and 3-epi25(OH)D3 is consistent with 25(OH)D3 serving as the primary substrate for C3epimerization. Vitamin D supplementation was effective in raising 25(OH)D3 levels,
however significant increases in 3-epi-25(OH)D3 also occurred.

S198

B-235
Determination of 25-(OH)-vitamin D2 and D3 in postmenopausal women and
results comparison between immunochemical and chromatographic methods

J. Cepova, R. Prusa, E. Klapkova, M. Pechova. Faculty Hospital Motol,

Prague 5, Czech Republic
Background: Objective of this study was to evaluate and validate HPLC metod
for determination of 25-(OH)-vitamin D3 and 25-(OH)-vitamin D2, and to measure
160 patient samples and compare the results obtained by immunochemical and
chromatographic measurements.
Methods: For determination of 25-(OH)-vitamin D3 and 25-(OH)-vitamin D2 was
used high performance liquid chromatography (HPLC) with UV detection (Agilent
1200). The samples were measured by kit (Recipe, Germany). The separation
was accomplished at 40 C, samples were detected at 264 nm. Chemiluminescent
immunoanalysis was performed on Abbott Architect i4000SR analyzer (Abbott
Laboratories, Germany). We measured 160 patient samples from postmenopausal
women with osteoporosis and postmenopausal women without osteoporosis. For
statistical evaluation we used GraphPad Prism 6.0.
Results: A linear relationship between serum concentration and peak area was obtained
for both substances with correlation coefficient r2=0.9989 for 25-(OH)-vitamin D3 and
r2=0.9986 for 25-(OH)-vitamin D2. The limit of detection for 25-(OH)-vitamin D3 was
4.9 nmol/L and for 25-(OH)-vitamin D2 13.8 nmol/L. The intra and interday accuracy
and precision were evaluated on two QC samples by multiple analysis. Intraday CV
for 25-(OH)-vitamin D3 and 25-(OH)-vitamin D2 were 5.7% and 2.8%. Intraday CV
were 3.7% and 6.4%. Within-day accuracy expressed by the calculated bias between
observed and theoretical concentrations for 25-(OH)-vitamin D3 and 25-(OH)-vitamin
D2 were 4.6% and 6.7%. Mean recoveries of the corresponding compounds were
94.5% and 104%. No interference has been found between 25-(OH)-vitamin D3 and
25-(OH)-vitamin D2 or internal standard. We also compared the levels of 25-(OH)vitamin D3 measured by HPLC with the levels of 25-(OH)-vitamin D measured by
immunochemical method. Furthemore, we compared the sum of 25-(OH)-vitamin
D3 and 25-(OH)-vitamin D2 with 25-(OH)-vitamin D. The first data were tested by
Mann Whitney test, the second part by unpared t test. The data showed significantly
differences, p = 0.0061 for comparison of 25-(OH) vit D3 with 25-(OH)-vit D, and
p < 0.0001 for the sum of 25-(OH)-vit D3 and 25-(OH)-vit D2 with 25-(OH)-vit D.
Conclusion: We assumed that the data measured by immunochemical method will be
higher than the data measured by HPLC due to the large number of cross-reactions,
but the results (expressed as median SEM) were contrary: 75.3 38.8 nmol/L
measured by HPLC vs. 64.7 23.3 nmol/L measured by immunochemical method.
Supported by the project (Ministry of Health, Czech Republic) for conceptual
development of research organization 00064203 (University Hospital Motol, Prague,
Czech Republic)

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Nutrition/Trace Metals/Vitamins

Wednesday, July 30, 9:30 am 5:00 pm

method. The concentration of each analyte was divided into low and high levels
for data analysis. Comparison study and between-assay agreement were examined
using Bland-Altman analysis, Passing-Bablok regression and inter-rather agreement
analysis (Kappa).

B-236
Comparison of the Roche Cobas Electrochemiluminescent Vitamin D Assay to
the DiaSorin Chemiluminescent Method

R. Rosecrans, J. Greene. NorthShore University HealthSystem, Highland

Park, IL
Background: Low vitamin D has been implicated in cancer, cardiovascular
disease, diabetes, multiple sclerosis, autoimmune diseases, even autism, and
headaches resulting in an exponential increase in vitamin D testing. Clinical
diagnostic manufacturers have responded in meeting the vitamin D testing demand.
Immunoassay manufacturers of vitamin D assays claim their methods correlate to LC/
MS/MS, the reference method, but the literature has been varied. This study compares
the Roche Cobas (Roche Diagnostics, Indianapolis, IN) electrochemiluminescence
binding assay to the DiaSorin (Stillwater, MN) chemiluminescent assay for vitamin D.
Methods: The primary method used by the laboratory is the DiaSorin Total Vitamin
D assay. 200 patient samples were chosen for the comparison study. To insure an
adequate concentration distribution, patient samples were grouped as follows: less
than 10.0ng/mL, 10.1 to 20.0ng/mL, 20.1 to 40.0ng/mL, 40.1 to 60.0ng/mL, and
greater than 60ng/mL. After determination on the DiaSorin Liaison, the samples were
frozen and held until transport to a laboratory performing the Roche Cobas assay.

Results: A negative bias was found for all total 25OH-D, 25OH-D2 and 25OH-D3
(ranged from -4.47 to -0.04 ng/mL). The Bland-Altman analysis demonstrated slight
concentration-dependent bias. Overall, Passing-Bablok fit showed that the in-house
LC-MS/MS method was statistically equivalent to Chromsystems (p value > 0.05)
with a high to very high correlation coefficient (r = 0.876 to 0.986), Table 1. The
ability to properly classify patients according to their vitamin D status was very
satisfactory for the tested method which the Kappa values were 0.864, 1.00 and 0.944
for total 25OH-D, 25OH-D2 and 25OH-D3 (concordance >90%), respectively.
Conclusion: The in-house LC-MS/MS method for total 25OH-D, 25OH-D2 and
25OH-D3 determination correlated very well with the NIST traceable method.
Strength of agreement for classifying patients into the low or optimal vitamin D
levels with the reference assay was very good. The observed bias had little impact
on clinical decision therefor is clinical acceptable. We conclude that our LC-MS/MS
method met the minimum requirements for the assessment of vitamin D status in
clinical laboratories.

Results: Linear regression analysis between the Roche Cobas and the DiaSorin
Liaison demonstrated a correlation coefficient of 0.7272. Bland-Altman plots
demonstrate value differences ranging from (-)44 ng/mL to (+)54 ng/mL. Patient
subgroups were divided and analyzed separately. One patient group (n =91) had values
between 5 to 20ng/mL as determined by the DiaSorin method. Linear regression
analysis of this subgroup demonstrates a correlation coefficient of 0.6749. The second
population subgroup (n=172) had DiaSorin vitamin D results between 20.1 to 40.0ng/
mL and a correlation coefficient of 0.1742 when compared to the Roche Cobas assay.
Roche Cobas had an overall positive bias compared to the DiaSorin that resulted
in 48%,15.8%, and 14.8% fewer patients being classified as deficient (<10ng/mL),
insufficient (10.1-31ng/mL) and sufficient (> 32ng/mL), respectively.
Conclusion: The Roche Cobas vitamin D assay has a positive bias compared to the
DiaSorin Liaison for patients in the range of 4.0 to 40.0 ng/mL. 48%fewer patients
were classified as deficient by the Roche assay, 15.6% fewer classified as insufficient,
and 14.8% fewer as optimal. After vitamin D dissociation from its binding protein,
the DiaSorin Total vitamin D assay adds an isoluminol labeled vitamin D that
competes with an anti-vitamin D Ab bound to magnetic particles in a competitive
binding assay. The Roche assay is different, after separation of vitamin D from its
bind protein the vitamin D is incubated with ruthenium labeled vitamin D binding
protein. Next, biotinylated labeled vitamin D is added which binds to the unoccupied
vitamin D binding protein sites. The biotinylated vitamin D is bound to a solid phase
which interacts with the streptavidin. Potential sources of error may be heterophile
antibody interference for each assay or in the Roche assay the presence of exogenous
biotin. Biotin is present in some foods, cosmetics, hair and nail products, and OTC
supplements. This study further demonstrates the need for vitamin D standardization
and a more vigorous approach by all manufacturing companies on possible assay
interferences.

B-237
Measurement of serum total 25-hydroxy vitamin D and its metabolites by
liquid chromatography - tandem mass spectrometry: Agreement with the NIST
traceable Chromsystems method

S. Vanavanan1, A. Chittamma1, P. Meemaew1, S. Promnuch1, B. Intachak1,

M. Rochanawutanon2. 1Clinical Chemistry Division, Department
of Pathology, Faculty of Medicine, Ramathibodi Hospital, Mahidol
University, Bangkok, Thailand, 2Department of Pathology, Faculty of
Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand
Background: Serum 25-hydroxy vitamin D (25OH-D) is considered to be a reliable
indicator of vitamin D status. Liquid chromatography - tandem mass spectrometry
(LC-MS/MS) was recently proposed as a reference method. We compared our inhouse LC-MS/MS method for total 25OH-D and its metabolites (25OH-D2 and D3)
measurement with the NIST traceable Chromsystems. Agreement of vitamin D status
between both LC-MS/MS systems was assessed based on clinical cut-off value for
total 25OH-D.
Methods: Seventy eight serum patient samples were randomly selected from
Ramathibodi Hospital. Determination of total 25OH-D, 25OH-D2 and 25OH-D3
were performed by two LC-MS/MS methods using the Chromsystems as a reference

B-238
Evaluation of a Random Access Total 25-Hydroxy Vitamin D (THVD)
Immunoassay (IA): Patient Correlation with HPLC-Mass Spectrometry (MS)

N. Yadak1, S. Freeman2, C. Hoang3, C. Finch Cruz3, E. S. Pearlman3.

1
University of Tennessee Health Sciences Center, Memphis, TN, 2Veterans
Administration Medical Center, Memphis, TN, 3Veterans Affairs Medical
Center, Memphis, TN
Background: The VAMC evaluated a THVD assay to be employed on our Vitros-5600
platform [OCD; Raritan, NJ]. Specimen results with compared to HPLC-mass
spectrometry (MS) taken to be the reference method.
Methods: Specimens (n=42) received frozen from a reference lab [(RL)/ARUP; Salt
Lake City, UT] had been assayed for THVD using MS. These were thawed shortly
before IA and run in random order according to manufacturers directions. Two of the
specimens had IA values below the analytical measurement range and were excluded
from the regression analysis. There were 13 samples with adequate residual volume
to be returned as blinded liquid samples and re-assayed with MS. In a separate study
40 in-house specimens were run twice using IA with the specimens being stored at 4
degrees overnight between assays. Regression analysis used Table Curve-2D software
(Systat; San Jose, CA).
Results: The relationship between IA (Y) and MS (X) data was fit reasonably well by
the linear equation [with 95% CIs]:
Y = 1.015 [0.88, 1.15] X + 1.25 [-8.30, 5.80] (n=40, r-sq = 0.856).
Using a THVD concentration of 30 mcg/L as the threshold for optimality there
were six specimens with MS values >30 mcg/L that were < 30 mcg/L by IA and 2
specimens with the reverse situation. The median (range) within-pair precision for
IA (n=40) was 2.56% (0-11.83%) with the CV being <10% in 39/40 instances. A plot
of CV (Y) vs. Mean (X) suggests a non-linear relationship with a CV that decreases
with increasing mean concentration but the 95% CI on the slope includes zero. The
median (range) within pair precision (n=13) using MS was 10.5% (2-30.9%) with 4
specimens having a CV >20%. All MS specimens with initial THVD i concentrations
of >30 mcg/L (n=8)] however remained optimal and likewise for those initially with
sub-optimal THVD.
Conclusions: The overall performance of the IA from OCD appears satisfactory but
there may be some increase in the number of patients with suboptimal concentrations
with IA compared to MS. IA reproducibility after 24 hours of refrigeration was very
good. Although the number of specimens retested by MS was small there is the
suggestion that a freeze thaw cycle and transportation time may result in significant
imprecision although clinical reclassification was not observed.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S199

Nutrition/Trace Metals/Vitamins

Wednesday, July 30, 9:30 am 5:00 pm

B-240
High prevalence of Vitamin D Deficiency in Korean pregnant women and
association between maternal 25-hydroxyvitamin D level in pregnancy and
neonatal outcomes

R. Choi1, S. Oh2, H. Yoo3, Y. Cho4, S. Kim4, J. Chung4, S. Lee1. 1Department

of Laboratory Medicine and Genetics, Samsung Medical Center,
Sungkyunkwan University School of Medicine, Seoul, Korea, Republic
of, 2Department of Obstetrics and Gynecology, Samsung Medical Center,
Sungkyunkwan University School of Medicine, Seoul, Korea, Republic of,
3
Biostatistics Team, Samsung Biomedical Research Institute, Seoul, Korea,
Republic of, 4Division of Endocrinology and Metabolism, Department
of Medicine, Thyroid Center, Samsung Medical Center, Sungkyunkwan
University School of Medicine, Seoul, Korea, Republic of
Background: There is growing concern about functional impacts of maternal vitamin
D status on multiple adverse health outcomes in mothers and on their offspring
and low maternal levels of 25-hydroxyvitamin D [25(OH)D] has been suggested
to be associated with some adverse obstetrical and neonatal outcomes. However,
there were no reliable data based estimation of vitamin D status using LC-MS/MS
in Korean pregnant women. Objective: The aim of this study was to carry out the
first population-based survey on vitamin D status in Korean pregnant women to
assess vitamin D status during pregnancy and the effect of vitamin D deficiency on
pregnancy outcomes; premature rupture of membrane, preterm birth, and child born
small for gestational age. Method: Korean pregnant women (n=220) were recruited
prospectively and tested for 25(OH)D levels in serum using liquid chromatographytandem mass spectrometry with assessment of maternal characteristics. Their 25(OH)
D levels were compared with those of 500 healthy nonpregnant women. We analyzed
vitamin D status according to demographics, seasons, and obstetrical characteristics
together with the assessment of obstetrical and neonatal outcomes. Results: The
median concentrations of 25(OH)D in Korean pregnant women (n=220) and healthy
nonpregnant women (n=500) were 12.6 ng/mL and 15.4 ng/mL, respectively. The
overall prevalence of vitamin D deficiency [25(OH)D < 20 ng/mL] in pregnant women
and healthy nonpregnant women were 77.3% (170/220) and 79.2% (396/500), and
the prevalence of severe vitamin D deficiency [25(OH)D < 10 ng/mL] were 28.6%
(63/220) and 7.2% (36/500), respectively. The prevalence of vitamin D deficiency was
higher in winter (100%) than in summer (45.5%) in Korean pregnant women. The 1st
trimester had a higher risk of vitamin D deficiency than the 3rd trimester (adjusted
OR 4.3; 95% CI 1.2-15.2; P < 0.05). No associations were observed between vitamin
D deficiency and pregnancy or birth outcomes including premature rupture of
membrane, preterm birth, and child born small for gestational age. Conclusions: The
prevalence of vitamin D deficiency was high in pregnant women in Korea and showed
the highest during the 1st trimester of pregnancy. Although there was no association
between vitamin D deficiency and pregnancy outcome, further research about the long
term consequences of vitamin D deficiency during pregnancy on the mother and the
offspring is warranted.

B-242
Serum biomarkers that predict clinical outcomes in an immobilized population:
Predictors of lean body mass loss

G. J. Davis1, S. H. Gawel2, M. Luo2, N. K. Edens2, N. E. Deutz3, R. R.

Wolfe4, S. L. Pereira2. 1Companion Diagnostics R&D, Abbott Laboratories,
Abbott Park, IL, 2Abbott Nutrition, Columbus, OH, 3Texas A&M Univ,
College Station, TX, 4Univ. Arkansas Medical Sciences, Little Rock, AR
Background:Loss of LBM during extended bed rest (BR) (i.e. hospitalization) is a
major contributor to functional decline and loss of mobility, especially in older adults.
This problem is generally under-recognized due to lack of practical diagnostic tools
to measure lean body mass (LBM) over hospitalization. Identifying blood biomarkers
that predict a hospitalized individuals risk of losing LBM could provide a practical
alternative to expensive and tedious existing methods for LBM measurement (MRI,
DXA, CT). Ease of identification of susceptible populations will increase awareness
of the problem, allow for timely intervention, and have a huge impact on the health
economics of hospitalization.
Methods: Eighteen healthy subjects (age 60-76 y, 3 male, 15 female) were confined
to 10 days of complete BR and received either placebo (n=8) or treatment (Ca-hydroxy--methylbutyrate-HMB) (n=10) over BR. Fasting serum samples were
obtained prior to the start of BR (D1) and analyzed using multiplexed immunoassay
array Human DiscoveryMap 1.0 (RBM-Myriad). LBM was assessed by Dual
energy X-ray absorptiometry (DXA) before and at the end of BR (D10). Baseline

S200

biomarker data from both groups were merged, and multiple-hypotheses testing and
partition analysis (with 5-fold cross validation) were used to identify baseline markers
that predict LBM loss over BR.
Results: Over the 10 day BR period, change in total LBM varied between individuals
(-4.47 kg loss to 0.82 kg gain) indicating some subjects were more predisposed to
LBM loss over others. Of the 187 markers analyzed at baseline, 63 were excluded
due to low detection levels in 30% subjects. One pair of markers was found to
correlate with percent change in LBM over BR: Tissue inhibitor of metalloprotease-1
(TIMP1) and Tenascin C (TNC) [R2=0.71, all subjects; R2= 0.76, females]. Subjects
with TIMP1 141 ng/ml at D1 had larger losses of total LBM at D10, whereas subjects
with TIMP1< 141 and TNC 461 ng/ml at D1 did not lose total LBM over BR. Two
additional markers were found to correlate with percent change in leg lean mass over
BR: Matrix metalloprotease-3 (MMP3) and Apolipoprotein A2 (APOA2) [R2=0.59,
females]. Females with MMP3< 6.93 ng/ml at D1 were more likely to lose leg lean
mass at D10 compared with females with MMP3 6.93 and ApoA2< 276 ng/ml at D1
who did not lose muscle at D10.
Conclusion: Panels of blood biomarkers may be useful in predicting key clinical
outcomes such as LBM loss over immobilization (e.g. hospitalization). Validation of
these markers in large clinical studies is needed. Sponsored by Abbott Laboratories

B-244
Performance Characteristics of the ARCHITECT Active-B12
(Holotranscobalamin) Assay: A Marker of Vitamin-B12 Deficiency

S. D. Merrigan, W. E. Owen, J. A. Straseski. ARUP Laboratories, Salt Lake

City, UT
Background Vitamin B12 (cobalamin) is a necessary cofactor in methionine and
succinyl-CoA metabolism. Some studies estimate the prevalence of deficiency
may be as high as 30% in the elderly population. Cobalamin deficiency is a serious
health risk due to its role in carbon metabolism, cell division, DNA synthesis and
the clinically important outcomes of anemia and progressive and irreversible
neurologic dysfunction. Ten to thirty percent of circulating cobalamin is complexed
with transcobalamin and is called holotranscobalamin (holoTC). HoloTC can readily
enter cells and is therefore considered the bioactive form. The objective of our study
was to evaluate the analytical performance of the ARCHITECT i2000SR Active-B12
(Holotranscobalamin) assay and compare results to manual and automated
immunoassays.
Methods Manufacturer-specified limits of blank (LoB), detection (LoD), and
quantitation (LoQ), imprecision, interference and linearity were evaluated for the
ARCHITECT i2000SR Active-B12 (Holotranscobalamin) assay (Abbott Diagnostics,
Abbott Park, IL) per CLSI guidelines. Residual de-identified serum samples were
used to compare results from the ARCHITECT HoloTC assay with the results
from the automated Abbott AxSYM Active-B12 (Holotranscobalamin) assay
(Abbott Diagnostics) and the manual Active-B12 (Holotranscobalamin) Enzyme
Immunoassay (EIA) (Axis-Shield Diagnostics, Dundee, Scotland, United Kingdom).
Results Manufacturers claims of <0.4, <1.9 and <5.0 pmol/L for LoB, LoD, and
LoQ, respectively, were verified for the ARCHITECT HoloTC assay. Total withinassay imprecision was 5.7% at a mean concentration of 16.2 pmol/L (SD 0.9 pmol/L),
and 4.9% at a mean concentration of 46.7 pmol/L (SD 2.3 pmol/L), verifying the
manufacturers imprecision claims. Interference studies demonstrated <10%
deviation in recovery for hemolysis, icterus, and lipemia up to concentrations of 200
mg/dL, 20 mg/dL, and 850 mg/dL, respectively. The ARCHITECT HoloTC assay
was linear up to the highest concentration measured (113.4 pmol/L). The largest mean
deviation from the calculated recovery was 8.5% at an expected holoTC concentration
of 26.2 pmol/L. Method comparison of the ARCHITECT HoloTC assay to the
AxSYM HoloTC assay for samples with results within the AMR gave the following
Deming regression statistics with 95% confidence intervals: (ARCHITECTHoloTC)
= 0.941 0.062(AxSYMHoloTC) + 1.2 2.5 pmol/L, Sy/x = 6.4, r = 0.947 (n=98).
Average bias between the methods was -0.9 pmol/L (-3%). Method comparison of
the ARCHITECT HoloTC assay to the Active-B12 EIA gave the following Deming
regression statistics with 95% confidence intervals: (ARCHITECTHoloTC) = 1.105
0.046(EIAActive-B12) 6.8 2.7 pmol/L, Sy/x = 11.0, r = 0.950 (n=221). Average bias
between methods was -1.7 pmol/L (-3%). A medical decision point analysis at 35.0
pmol/L was calculated using Deming statistics. The AxSYM HoloTC assay agreed
within the 95% confidence intervals of the ARCHITECT assay (33.7 35.7 pmol/L),
while the Active-B12 EIA comparison was slightly below (31.7 34.2 pmol/L).
Concordance between the assays at 35 pmol/L was 93% (AxSYM) and 94% (EIA).
Conclusions This assay performed acceptably for LoB, LoD, LoQ, imprecision,
interference, linearity and method comparison to the predicate device (AxSYM).

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Nutrition/Trace Metals/Vitamins

Wednesday, July 30, 9:30 am 5:00 pm

An additional comparison to a manual Active-B12 EIA method performed similarly,

with minor exceptions. This study determined that the ARCHITECT HoloTC assay is
suitable for routine clinical use.

B-245
Achieving 25(OH)vitamin D2 and 25(OH)vitamin D3 Equimolarity for the
Dimension EXL Vitamin D Total Assay1,2

E. Garcia, W. Bedzyk, J. Li, S. Manoj, T. Q. Wei. Siemens Healthcare

Diagnostics Inc., Newark, DE
Background:The Dimension EXL Vitamin D Total (VitD) assay is a homogenous,
competitive immunoassay based on LOCI technology. The VitD assay utilizes 4
reagents: a releasing reagent, two latex bead reagents and a biotinylated monoclonal
antibody reagent. The releasing reagent releases 25(OH)vitamin D2 and 25(OH)
vitamin D3 from the endogenous vitamin D binding proteins (DBP). The released
25(OH)vitamin D2 and 25(OH)vitamin D3 are then bound by the biotinylated assay
antibody to produce the assay signal. The assay is intended for the equimolar
determination of 25(OH)vitamin D2 and 25(OH)vitamin D3.
Methods: Recovery of 25(OH)vitamin D2 or 25(OH)vitamin D3 was assessed using
the 25(OH)vitamin D2 or 25(OH)vitamin D3 spiked human serum samples. Available
anti-25(OH)vitamin D monoclonal antibodies were screened for anti-25(OH)vitamin
D2 specific binding using a prototype LOCI VitD assay. One anti-25(OH)vitamin
D2 antibody was selected and added to the biotinylated antibody reagent to bind the
excess 25(OH)vitamin D2, which led to 25(OH)vitamin D2 /25(OH)vitamin D3 3
equimolarity.
Results: A dose response curve showed the recovery ratio of 25(OH)vitamin D2
/25(OH)vitamin D3 changed from 118% to 84 % with the increase of anti-25(OH)
vitamin D2 antibody from 0 to 80 g/mL. Adding 20 g/mL of the antibody in the
assay reagent achieved 100% recovery ratio or 25(OH)vitamin D2 /25(OH)vitamin D3.
Conclusion: By achieving 25(OH)vitamin D2 /25(OH)vitamin D3 equimolarity
and employing highly sensitive LOCI technology, the fully automated Dimension
EXL Vitamin D Total assay provides accurate and precise total 25(OH)vitamin D
measurement on the Dimension EXL system.
1. Under development. Not available for sale. Due to local regulations, not all products
will become available in all countries.
2.Patent pending.

B-246
Development of a Vitamin D Total Assay* with LOCI Technology on the
Dimension EXL System

J. Li, Z. Teng, M. Drinan, R. Janzen, L. Larson, M. Stranz, D. Clark, B.

Wessel, P. Singh, S. Manoj, T. Q. Wei. Siemens Healthcare Diagnostics
Inc., Newark, DE
Background: The Siemens Dimension EXL System incorporates multiple
detection technologies, including LOCI technology, which enables high sensitivity
immunoassay formats. Siemens is currently developing a Vitamin D Total assay
utilizing LOCI technology on the Dimension EXL System.
Methods: The Dimension EXL Vitamin D Total assay (VitD) is a homogeneous
competitive chemiluminescent immunoassay based on LOCI technology. It measures
the total 25(OH)vitamin D concentration (comprised of 25(OH)vitamin D2 and
25(OH)vitamin D3) in both serum and plasma. The VitD LOCI components include
a releasing reagent, two synthetic bead reagents and a biotinylated monoclonal
antibody. The first bead reagent (Sensibeads) is coated with streptavidin and contains
photosensitive dye. The second bead reagent (Chemibeads) is coated with a 25(OH)
vitamin D3 analog and contains chemiluminescent dye. The sample is incubated with
the releasing reagent to release 25(OH)vitamin D molecules from the vitamin D
binding proteins. The reaction mixture is then incubated with biotinylated antibody
to form a 25(OH)vitamin D/biotinylated antibody complex. Chemibeads are added
to remove the excess free biotinylated antibody, and then Sensibeads are added to
bind to the biotinylated antibody. Aggregates of the Chemibead-analog/antibodybiotin/streptavidin-Sensibeads are formed as a result. Illumination of the reaction
mixture by light at 680 nm generates singlet oxygen from Sensibeads, which
diffuses into the Chemibeads and triggers a chemiluminescent reaction. The resulting
chemiluminescent signal is measured at 612 nm and is inversely proportional to the
concentration of total 25(OH)vitamin D in the sample.

Repeatability and within lab CVs were less than or equal to 2.4% and 3.2% CVs
respectively between 10-100 ng/mL. A patient sample correlation (n=215) showed a
Passing-Bablok regression: Dimension EXL Vitamin D Total assay = 1.02 ADVIA
Centaur Vitamin D Total assay, Reference Measurement Procedure (RMP).** + 1.84
ng/mL, r=0.94, range = 3 - 131 ng/mL. Less than 10% cross-reactivity was observed
at 500 pg/mL for 1,25(OH)2vitamin D2 and 1,25(OH)2vitamin D3, at 100 ng/mL for
3-epi-25(OH)vitamin D3, and at 1000 ng/mL for Vitamin D2 and Vitamin D3. This
assay is equimolar and aligned to the ID-LC/MS/MS 25(OH)vitamin D Reference
Measurement Procedure (RMP).
Conclusion: The Dimension EXL Vitamin D Total assay demonstrates acceptable
precision, accuracy and turnaround time for the total 25(OH)vitamin D measurement
on the Dimension EXL System.
*Under development. Not available for sale. Due to local regulations, not all products
will become available in all countries.
**Under FDA review. Not available for sale in U.S. Due to local regulations, not all
products will become available in all countries.

B-247
Development of Candidate Standard Reference Material 3949 Folate
Vitamers in Frozen Human Serum

J. E. Camara, M. S. Lowenthal, J. S. Pritchett, K. W. Phinney, L. C. Sander.

NIST, Gaithersburg, MD
The National Institute of Standards and Technology (NIST) provides a variety of
Standard Reference Materials (SRMs) for the analysis of nutrient levels in clinical
matrices. NIST currently offers SRM 1955 Homocysteine and Folate in Human Serum,
which possesses certified values for homocysteine and 5-methyltetrahydrofolate (5mTHF) and reference values for folic acid, also known as pteroyl-glutamic acid
(PGA), over three concentration levels of material. SRM 1955 Level 1 required
dilution and SRM 1955 Level 3 required spiking of a serum pool to achieve the target
concentrations. Once out of stock, this SRM will be replaced with a new candidate
material, SRM 3949 Folate Vitamers in Frozen Human Serum. Input from The
National Institutes of Health Office of Dietary Supplements (NIH-ODS) and the
Centers for Disease Control and Prevention (CDC) indicates interest in an updated
material with folate levels reflecting those currently observed in the population.
This new SRM will have three concentration levels with low, medium, and
high certified values for both 5-mTHF and PGA. NIST also intends to assign
reference values for the additional minor folate metabolites tetrahydrofolate
(THF),
5-formyltetrahydrofolate
(5-fTHF),
5,10-methenyltetrahydrofolate
(5,10-methenylTHF), and the oxidation product of methyl folinate (MeFox). The goal
levels for 5-mTHF and PGA, respectively, are: Level 1, 10 nmol/L and 1 nmol/L;
Level 2, 50 nmol/L and 10 nmol/L; Level 3, 30 nmol/L and 5 nmol/L. In addition,
Level 3 has goal levels of 5 nmol/L, 5 nmol/L, 5 nmol/L, and 3 nmol/L for THF,
5-fTHF, 5,10-methenylTHF, and MeFox, respectively.
To produce SRM 3949, pilot sera were collected from 15 individual donors, five of
which were given a 400 g folic acid supplement one hour prior to blood draw in
an attempt to increase serum levels of 5-mTHF and PGA for the high level material
without the requirement for additional spiking. To stabilize the folates, 0.5 % (w/v)
ascorbic acid was added as soon as possible after collection of serum. These pilot
sera were screened for five folates plus the oxidation product, MeFox, at the CDC by
ID-LC-MS/MS. Screening results ranged from 5 nmol/L-72 nmol/L for 5-mTHF, 0.4
nmol/L-32 nmol/L for PGA, 0.25 nmol/L-2.2 nmol/L for THF, and 0.13 nmol/L-3.2
nmol/L for MeFox. Both 5-fTHF and 5,10-methenylTHF were below the limits of
detection for all sera. Four pilot sera from donors administered a folic acid supplement
displayed significantly elevated levels of both 5-mTHF and PGA. Based on these
results, a blending protocol was specified to obtain the desired folate concentrations in
each of the three SRM 3949 concentration levels. The endogenous levels of 5-mTHF
and PGA in all three concentration levels and enhanced folate stability via ascorbic
acid addition are improvements over the original SRM 1955 that should better serve
the end users.
The candidate material has been blended and packaged and will undergo additional
analyses by ID-LC-MS/MS at both NIST and the CDC. NIST is also investigating
updates to its current ID-LC-MS/MS method for folates in serum based on the current
CDC method, which may be applied to the certification measurements of SRM 3949.

Results: The method requires 8 L of serum or plasma and is linear from 4 to

150 ng/mL. Time to first result is 32 minutes with a stable calibration for 7 days.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S201

Nutrition/Trace Metals/Vitamins

Wednesday, July 30, 9:30 am 5:00 pm

B-248
The Frequency of Vitamin B12 Deficiency in Metformin-treated Brazilian Type
2 Diabetes Patients.

C. B. Damio1, G. F. Taboada1, A. O. Rodrigues1, M. F. M. C. Pinheiro2, M.

C. Freire2, R. A. Cruz Filho1, G. P. Cardoso1, G. A. B. Lima1. 1Divison of
Endocrinology, Universidade Federal Fluminense, Rio de Janeiro, Brazil,
2
DASA, Rio de Janeiro, Brazil
Background:Vitamin B12 is an essential micronutrient required for optimal
hematopoietic, neurocognitive and cardiovascular function. Metformin is a biguanide
recommended as initial medical therapy for type 2 diabetes mellitus (T2DM).
Despite the known effectiveness, there are disadvantages in its use. Studies indicate
a prevalence of 14% to 30% of vitamin B12 deficiency among patients undergoing
long-term treatment with metformin, with a still controversial mechanism. The
objective of this study was to evaluate the frequency of vitamin B12 deficiency and
the factors associated with serum vitamin B12 levels in metformin-treated Brazilian
type 2 diabetes patients.
Methods:Cross-sectional study that included 231 T2DM patients in metformin
therapy. All the patients were followed at the Endocrinology Division of the
University Hospital. The serum B12 levels were measured by chemiluminescence
method (Access, Beckman Coulter, CA, USA). Vitamin B12 deficiency was defined
by a serum level below 180 pg/mL. SPSS 13.0 was used for statistical analysis. The
Mann Whitney test was used to compare numerical variables between groups, the
McNemar test to assess the association between binary variables, p value <0.05 was
considered to be statistically significant.
Results:Median age was 61 (34 to 79) years and 72% were women. The median time
of T2DM was 12 (3 to 41) years, duration of metformin use was 8 (3 to 30) years
and dose of metformin was 1,700 (500 to 2,550) mg/day. The vitamin B12 mean
levels were 272 (68 to 1,000) pg/mL. The frequency of cobalamin deficiency was
26.8%. The patients with vitamin B12 deficiency had longer disease duration (14 vs
10 years; p= 0.042) and longer metformin use (10 vs 8 years; p= 0.016). Vitamin B12
levels were significantly lower in patients that were using H2 antagonists or proton
pump inhibitors (210 vs 292 pg/mL; p= 0.002). After multiple regression analysis,
only duration of metformin treatment and the use of H2 antagonists / proton pump
inhibitors were significantly correlated to cobalamin levels.
Conclusion:Our study confirmed that the frequency of vitamin B12 deficiency is high
in metformin-treated T2DM patients. The patients with vitamin B12 deficiency were
using metformin for a longer time, suggesting a cumulative effect of the drug. The use
of H2 antagonists or proton pump inhibitors negatively influenced the serum levels
of vitamin B12, demonstrating the role of acid gastric reduction as a predisposing
factor to vitamin B12 deficiency. Finally, as vitamin B12 deficiency is a cause of
peripheral neuropathy, it should be considered in the differential diagnosis of diabetic
neuropathy.

B-249
Performance of Roche Elecsys Vitamin D Assay in Different Patient Populations
and in Patients with Vitamin D2 supplement.

X. Yi, N. Babic, M. H. Varnamkhasti, E. K. Y. Leung, K. T. J. Yeo. The

University of Chicago, Chicago, IL
Background: Demand for vitamin D (vit D) testing has increased worldwide,
partially due to mounting evidence linking vit D status to overall health
and well-being. Currently available methodologies include immunoassays
and liquid chromatography tandem mass spectrometry (LC-MS/MS). It has
been reported that the accuracy of some immunoassays is dependent on the
concentration of vitamin D binding protein (VDBP); excess VDBP amounts
may interfere with antibody binding, leading to vit D underestimation
(e.g. in pregnant subjects). In addition some immunoassays are optimized
for vitamin D3 (D3) and may not detect vitamin D2 (D2). We studied
the performance of Elecsys vit D assay (Roche, IN) in different patient
populations with varying VDBP concentration, and in patients taking D2
supplement. Method: A total of 211 patient specimens from 4 clinical areas:
intensive care unit (ICU), obstetrics (OB), gastroenterology (GAST) and
primary care (PCG) were collected and analyzed by in-house developed
LC-MS/MS assay, Roche Elecsys assay and DiaSorin (Stillwater,
MN) radioimmunoassay (RIA), respectively. Unlike immunoassays,
our validated LC-MS/MS method can quantitate 25-OH vit D2 and D3
concentrations individually. The results were analyzed by Passing-Bablok
regressions and Bland-Altman plots. Results: Comparison studies showed

S202

the following for the entire patient cohort (n=211): [RIA vit D] = 0.93 [LCMS/MS vit D] +1.80, mean bias =1.2 ng/ml; [Elecsys vit D]= 0.63 [LCMS/MS vit D] + 3.60, mean bias = -6.7 ng/ml. For patients with only D3
detected (n=153), the correlation showed [Elecsys vit D] = 0.79 [LC-MS/
MS vit D] + 3.04, mean bias = -2.4 ng/ml. In 58 patients found to have
detectable D2: [Elecsys vit D] = 0.47 [LC-MS/MS vit D] + 2.72, mean bias =
-15.2 ng/ml; when D2 concentration is >50% of total vit D (n=27), [Elecsys
vit D] = 0.30 [LC-MS/MS vit D] + 5.18, mean bias = -17.8 ng/ml. There
was lesser underestimation of D2 by the RIA method when compared to
LC-MS/MS method. Regression analyses revealed significant differences
between the various patient populations (patients with detectable D2 were
excluded): PCG, [Elecsys vit D] = 0.90 [LC-MS/MS vit D] + 2.54, mean
bias = 0.45 (n=33); ICU, [Elecsys vit D] = 0.74 [LC-MS/MS vit D] + 3.92,
mean bias = -1.17 (n=31); OB, [Elecsys vit D] = 0.71 [LC-MS/MS vit D]
+4.64, mean bias = -2.62 (n=31); GAST, [Elecsys vit D] = 0.73 [LC-MS/
MS vit D] - 1.54, mean bias = -8.15 (n=21), respectively. Conclusion: The
Roche Elecsys vitamin D assay underestimates measurement of vitamin
D concentrations in patients who have higher concentrations of D2 and in
OB and GAST groups. There was good agreement between Roche Elecsys
vitamin D assay with LC-MS/MS assay for the PGC and ICU groups when
patients with D2 were excluded.

B-250
An Improved Cleanup Strategy for Patient Samples using Anion Exchange
Solid Phase Extraction for the Analysis of Vitamin B6 Status

M. Boraski1, S. Baek2, N. Bhavsar1, M. Pikulski1. 1Clinical Pathology

Laboratories, Austin, TX, 2Biotage, Charlotte, NC
Background: Low levels of vitamin B6 are implicated in health problems involving
the nerves, skin, mucous membranes, and circulatory system. In children, vitamin B6
deficiency is related to cases of anemia and seizures. Vitamin B6 dependent seizure
disorders are an important and treatable cause of childhood epilepsy. Pyridoxal
5-phosphate (PLP), the primary biologically active form of vitamin B6, is the
preferred method for assessing vitamin B6 status. Most of the published procedures
for PLP involve extraction via protein precipitation from plasma followed by
derivatization to enhance the fluorescence signal. In this work, we present a more
selective sample extraction by using anion exchange solid phase extraction (SPE).
This procedure is also followed up with derivatization. Our objectives were to
automate the procedure by using SPE, decrease the noise and increase the resolution
in the chromatography, improve the HPLC column lifetime and create a more robust
method with cleaner sample extracts.
Method: To 200L of plasma, 30L semicarbazide/glycine derivatizing reagent
was added. The samples were then vortexed and incubated. The samples were
then processed using strong anion exchange SPE cartridges (EVOLUTE AX from
Biotage). The eluent was then evaporated and reconstituted in HPLC grade water.
After vortexing, 60L of the sample was injected onto a HPLC-FLD equipped with
a pre-column (KrudKatcher, Phenomenex) and a 4.6x100mm, 3, Gemini-NX C18
HPLC column (Phenomenex). HPLC mobile phase A was 10mM sodium phosphate
dibasic in 0.3% acetic acid and mobile phase B was acetonitrile/methanol (70:30)
in a 5 minute gradient (1mL/min). The excitation and emission wavelengths were
optimized at 370 and 470nm, respectively.
Results: The improvements in our method using the SPE cleanup included increased
column lifetime (3X), improved resolution and chromatographic robustness and
improvements in the run failure rate. The recovery was increased from 73% to 91%.
Three levels of plasma quality controls (QCs) were tested at 33, 92, and 360nmol/L.
The first two QCs were commercially available (Chromsystems) and the third was
prepared in-house from pooled patient samples. The coefficient of variation percent
(CV%) for intra-assay precision for the QCs were 5.1%, 4.6%, and 6.9% and interassay precision were 10.3%, 5.3%, and 7.3%. The linear regression analysis data for
six calibration curve points spanning 2-400nmol/L yielded an R2=0.9994. The lower
limit of quantitation (LLOQ) is 2nmol/L. A method comparison was performed with
an outside laboratory utilizing 50 patient samples. The Deming regression analysis
resulted in correlation R=0.96 for PLP using HPLC-FLD.
Conclusion: An improved method for the analysis of pyridoxal 5-phosphate has been
developed and has proven to be accurate, precise and robust.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Nutrition/Trace Metals/Vitamins

Wednesday, July 30, 9:30 am 5:00 pm

B-251
Reference intervals for intestinal disaccharidase activity determined from a
non-reference population

S. A. Hackenmueller, D. G. Grenache. University of Utah, Salt Lake City,

UT
Background: Dietary disaccharides are important exogenous sources of glucose.
Because the small intestine is normally impermeable to disaccharides, the
activities of intestinal disaccharidases are required for hydrolysis into component
monosaccharides that are subsequently absorbed. Decreased or absent activities
of one or more disaccharidases can result in carbohydrate maldigestion. The
measurement of disaccharidase activities in small intestine mucosa is considered
the gold standard test for diagnosis of disaccharidase deficiency. Due to the inability
to obtain intestinal biopsies from a healthy reference population, laboratories that
perform disaccharidase testing have adopted historical cutoff activities that are used to
identify a disaccharidase deficiency (lactase, <15; maltase, <100; palatinase, <5; and
sucrase, <25 U/g protein). The objectives of this study were to validate these historical
activity cutoffs using the Hoffman method for reference interval determination and
to evaluate disaccharidase activities and demographics of individuals for whom
intestinal disaccharidase testing was performed.
Methods: 14,827 samples for which all four disaccharidase test results were available
were extracted from the laboratory information system. For cutoff validation, results
of 0 U/g protein were excluded. Enzyme activities were log transformed and the
Hoffman method was used to calculate a reference interval. This method involved
determining the cumulative frequency distribution for each enzyme and performing
linear regression over the linear portion of the distribution. The reference interval
(RI) was calculated as RIlower=2.5(m)+b and RIupper=97.5m+b (m=slope; b=intercept).
The frequencies and patient demographics of all possible disaccharidase activity
phenotypes were determined from the entire population.
Results: The reference intervals for each enzyme were calculated to be 5-55 for
lactase, 105-380 for maltase, 9-32 for palatinase, and 26-110 U/g protein for sucrase.
The ratios of historical cutoffs to the calculated lower reference limits were 0.95, 0.56
and 0.96 for maltase, palatinase and sucrase, respectively, indicating the historical
cutoffs were less than the lower reference limit for each enzyme. The ratio for lactase
activity was 3, indicating the historical cutoff was within the calculated reference
interval. Examination of the frequency distribution of the activities of each enzyme
revealed that maltase, palatinase, and sucrase were unimodal while lactase showed a
bimodal distribution. The intersection of the two lactase populations corresponded
to a lactase activity of 10 U/g protein, which produced a ratio of 1.5 if taken to be
the lower reference limit. The median patient age of the entire data set was 13 years
(range, <1-88 years) and 45% were male. Using the historical cutoffs, 52% of samples
had no enzyme deficiencies. Deficiencies of lactase, maltase, palatinase, and sucrase
were present in 47, 11, 3, and 13% of the samples, respectively. 35% of samples were
deficient only in lactase. 3% of samples were deficient for all four enzymes.
Conclusion: The historical cutoffs for maltase and sucrase were validated. To align
with calculated reference intervals, the palatinase cutoff should increase to 9 U/g
protein, and the lactase cutoff should decrease to 10 U/g protein. Disaccharidase
testing is most commonly performed in patients <18 years. Lactase deficiency is
the most frequently observed single-disaccharidase deficiency. Pandisaccharidase
deficiency is rare.

B-252
Validation of a 25-OH Vitamin D (total) ELISA on the DRG:HYBRiD-XL, a
Fully Automated Analyzer for Immunoassays and Clinical Chemistry

M. Herkert1, C. Lauf1, B. Uelker1, T. Dudek1, T. Zeller1, C. E. Geacintov2.

1
DRG Instruments, Marburg, Germany, 2DRG International, Springfield,
NJ
Background: Vitamin D plays an important role in regulating body levels of calcium
and phosphorus and bone mineralization. Physiological Vitamin D levels result from
dietary intake and production in the skin after sun exposure. 25-hydroxyvitamin
D (25-OH-D), the major metabolite of Vitamin D circulates bound to Vitamin D
binding protein (VDBP). Determination of 25-OH D in serum or plasma supports the
diagnosis and therapy control of postmenopausal osteoporosis, rickets in children,
osteomalacia, renal osteodystrophy, neonatal hypocalcemia and hyperparathyroidism.
Objective: To validate the 25-OH D Elisa on the DRG:HYBRiD-XL.
Methods: 25-OH-D was validated on the DRG:HYBRiD-XL, a fully automated
analyzer for immunoassays and clinical chemistry parameters. Assay Procedure: 25
l of human serum are incubated for 30 min with 50 l denaturation buffer to release

the Vitamin D bound to VDBP. Thereafter, 200 l neutralization buffer, 50 l Enzyme

Conjugate (biotinylated Vitamin D) and 50 l Enzyme Complex (Streptavidin-HRP)
are added. The reaction volume is mixed and transferred to a well coated with VDBP.
After incubation for 60 min, the well is washed with wash buffer. Then 200 l of TMB
substrate are added to the well. After incubation for 30 min, 150 l of the blue TMB
substrate is transferred to a cuvette and measured at 645 nm (450 nm reference wave
length). Quantification is done based on a master standard curve that is barcoded on
the kit box.
Results: 25-OH-D can be quantified from serum and plasma (EDTA, heparin, citrate)
on the DRG:HYBRiD-XL. The dynamic range of the assay is between 2.3-130 ng/
mL. The limit of detection (according to CLSI guideline EP-17A) is 5.6 ng/mL,
the limit of quantification is 11.4 ng/mL. The mean within-run precision (EP-5A)
is 8.9%, the mean between-run precision is 12.9%. Recoveries of 6 samples were
found from 85.0-114% (mean 96.5%). The linearity (EP6-A) of 6 samples ranges
from 85.4-114.5% (mean 93.7%). Cross-reactivity (EP7-A2) was evaluated for 25OH Vitamin D3 (96.3%), 25-OH Vitamin D2 (74.7%), Vitamin D3 (3.8%), Vitamin
D2 (3.2%), and 1,25 (OH)2 Vitamin D3 (0.9%). Bilirubin and Hemoglobin (up to 0.5
mg/mL) and Triglycerides (up to 30 mg/mL) have no influence on the assay results.
Method comparison (EP9-A2-IR) with the 25-OH-D manual Elisa (EIA-5396) from
DRG Instruments gave a correlation coefficient of 0.958 (n=147; y=1.032x+0.679).
A method comparison with the 25-OH-D Liaison (Diasorin) gave a correlation
coefficient of 0.931 (n=147; y=1.028x+1.54). Median 25-OH-D values in the USA
largely depend on collection month (winter: 19.1 ng/mL; summer 23.6 ng/mL), race
(Afro-Americans 17.9 ng/mL; Hispanics 21.2 ng/mL; Caucasian 29.1 ng/mL), and
latitude of collection (north 19.8 ng/ml; mid 19.1 ng/mL; south 28.8 ng/mL).
Conclusions: 25-OH-D can be determined on the DRG:HYBRiD-XL with good
precision, and results show good correlation to DRGs manual Elisa and to the Liaison
from Diasorin. The concentration of 25-OH-D in serum decreases during winter time,
with dark skin colour and with higher latitude.

B-253
Vitamin D Sufficiency Thresholds: Are Age-Specific Values Needed?

L. M. Soares1, W. Pedrosa1, S. M. E. Santos2, L. S. Vasconcellos2. 1Hermes

Pardini, Belo Horizonte, Brazil, 2Universidade Federal de Minas Gerais,
Belo Horizonte, Brazil
Background: Vitamin D insufficiency has become a global health problem that
has been associated with metabolic bone disease and a great variety of chronic
illnesses. Thresholds for vitamin D sufficiency have been based mainly on total
25-hydroxyvitamin D (25OHD) concentrations at which serum parathyroid hormone
(PTH) increases, but there is great controversy surrounding the precise level at
which these changes occur. Current guidelines mainly suggest values of 20 or 30
ng/mL. Although vitamin D requirements are thought to vary with age, there are no
differences on the reported thresholds. The aims of this study were to analyze the
relationship between 25OHD and PTH in different age groups and to evaluate the
need of specific reference values for each one of them.
Methods: This was a cross-sectional analysis of 23,276 paired serum PTH and 25OHD
levels measured from January 2012 to December 2012 in a large Brazilian reference
laboratory. Data on laboratory tests and demographic variables were available from a
computerized database. Serum 25OHD was measured through Chemiluminescence,
using the Architect 25-OH Vitamin D Assay (Abbott, Illinois, USA). Serum PTH
was measured through Chemiluminescence, using the Access Intact PTH Assay
(Beckman Coulter, California, USA).
Results: Laboratory tests were equally distributed throughout the different seasons of
the year. Eighty-one percent of the studied population was female and the mean age
was 54.718.3 years. Serum 25OHD ranged between 8 and 160 ng/mL, with a mean
of 27.711.4. Serum PTH levels were inversely correlated with serum 25OHD. The
23,276 patients were split by their 25OHD values into 50 groups and the median PTH
of each was calculated, as well as the rate of results exceeding the upper limit of PTH.
Next, the data were broken down into five age classes (0 to 20, 20 to 40, 40 to 60,
60 to 80, and older than 80 years-old). The median PTH and the rate of high results
were plotted against the mean 25OHD and the graphs demonstrated that the PTH
concentrations were consistently higher in the oldest adults at different 25OHD levels.
The median PTH associated with 25OHD concentration of 20 ng/mL in each age
group was: 23 (0 to 20 years-old), 38 (20 to 40 years-old), 40 (40 to 60 years-old), 48
(60 to 80 years-old), and 57 pg/mL (over 80 years-old). When 25OHD concentration
of 30 ng/mL was considered, the median PTH of each group was, respectively, 23, 32,
34, 43 and 46 pg/mL.
Conclusion: We concluded that the serum concentrations of 25OHD required to
overcome hyperparathyroidism are different depending on the age and that the cut-

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S203

Wednesday, July 30, 9:30 am 5:00 pm

off values obtained in older adults shouldnt be applied to other groups. We assume
that the development of specific reference values according to age will attenuate the
overdiagnosis of vitamin D insufficiency and overtreatment of otherwise healthy
young patients.

B-254
C-ing is Believing: Enhanced Specificity for Vitamin C using HPLC with
Electrochemical Detection and Automatic Alternating Column Regeneration

Z. D. Clark1, E. L. Frank2. 1ARUP Laboratories, Salt Lake City, UT,

2
Department of Pathology, University of Utah Health Sciences Center, Salt
Lake City, UT
BACKGROUND: Vitamin C (L-ascorbic acid) is a water-soluble
micronutrient that is essential for human health. Deficiency of vitamin
C causes the fatal disease scurvy. Since the 1940s, vitamin C has been
measured spectrophotometrically after derivatization to form colored
products. These methods are subject to lack of specificity, limited
sensitivity, and interference from other compounds. Newer, more
specific techniques utilize chromatographic separation. While a highspecificity mass spectrometer may be an appealing detector choice, we
found electrochemical detection (ECD) provided adequate selectivity and
sensitivity at a fraction of the cost.
OBJECTIVE: The goal of this study was to develop a high-throughput HPLC-ECD
method for the measurement of vitamin C in plasma.
METHOD: Protein is precipitated from plasma using 10% meta-phosphoric acid. An
internal standard, 3,4-dihydroxybenzylamine, is added and the solution is incubated
with dithiothreitol in phosphate buffer to reduce dehydroascorbic to ascorbic acid.
Following re-acidification, a 5 L aliquot is injected onto an Agilent HPLC system.
The analytes exiting the analytical column undergo an electrochemical reaction in
a coulometric cell. The current generated is proportional to analyte concentration
and is measured by the CoulochemIII detector. The new assay was developed
and validated using a single pump and a single analytical column. The injectionto-injection time was 13 min. Subsequently, to increase the method throughput and
shorten turnaround time, a dual LC pump system with a 2-position/10-port switching
valve capable of performing automatic alternating column regeneration was validated
and implemented. The injection-to-injection time was reduced 2-fold.
RESULTS:
Parameter
LOQ
AMR
Linearity
Imprecision
Mean value
Within run CV
Btwn run CV
Total CV

Results
1.9 mol/L
5 - 5,000 mol/L
y=0.977x-0.040; observed error 1.9%
Low Control
High Control
23.62 mol/L
117.62 mol/L
5.2 %
2.1 %
3.6 %
3.0 %
6.3 %
3.7 %
Discarded specimens (n=44):
Accuracy: Method
y=0.983x-8.93; Sy/x=7.42; R=0.9901
Comparison
Fresh ref. interval specimens (n=41):
y=0.834x-1.08; Sy/x=6.77; R=0.9167
Not detected after injection of sample with 14,200 mol/L
Carryover
of ascorbic acid (AA).
33 common drugs and endogenous compounds tested.
Only isoascorbic acid (erythorbic acid), a non-endogenous
Analytical Specificity:
stereoisomer of AA, co-eluted.
Interference
Gross hemolysis (166 mol/L) reduced measured AA
concentration by 12%.
CONCLUSION: We have successfully developed and validated an HPLC-ECD
method for the measurement of vitamin C in plasma. Advantages of this method
include higher analytical specificity and significantly simplified sample preparation
compared to the previously used spectrophotometric method. Additionally, much
shorter injection-to-injection time compared to HPLC methods utilizing a single LC
column was achieved by employing an automatic alternating column regeneration
system.

S204

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Nutrition/Trace Metals/Vitamins

Pediatric/Fetal Clinical Chemistry

Wednesday, July 30, 9:30 am 5:00 pm

(median, range 0-1420). The average value of eosinophils in the group with PCT <2
ng/mL was 103/mm3, versus 28/mm3 in the group with PCT> 2 ng/mL (p <0.01).
At a cut-off value of <=0/mm3, the eosinophil count yielded a sensitivity of 51.4%
and a specificity of 83.6%, a positive predicted value (PPV) of 17.2% and a negative
predicted value (NPV) of 96.3%, with an area under the curve of 0.711 (95% CI:
0674-0747). At a cutoff value of 120/mm3, the NPV of the absolute eosinophil count
was 98.6%, thus allowing to rule out a PCT>2 ng/mL value in 21.4% (141/568) of
children.

Wednesday, July 30, 2014

Poster Session: 9:30 AM - 5:00 PM
Pediatric/Fetal Clinical Chemistry

CONCLUSION: The absence of eosinopenia can be considered as an inexpensive

warning test for predicting PCT <2 ng/mL in pediatric patients. The presence of
an absolute eosinophil count> 120/mm3 allows to rule out a PCT>2 ng/mL, thus
avoiding PCT determination in 21.4% of children, which is of interest especially in
the context of countries with limited resources

B-255
Increased expression of aquaporin 9 in placenta from pregnant women with
gestational diabetes

T. Vilario-Garca1, A. Prez-Prez1, P. Guadix1, V. Dietrich2, A. Damiano2,

J. L. Dueas1, C. Gonzlez-Rodrguz1, V. SANCHEZ-MARGALET1.
1
VIRGEN MACARENA UNIVERSITY HOSPITAL, SEVILLE, Spain,
2
University of Buenos Aires, Buenos Aires, Argentina
Background: Leptin is expressed in human placenta acting as an autocrine signal
for the trophoblast. In fact, leptin is now considered an important regulatory signal
in fetoplacental physiology. Placenta leptin expression is increased in pathological
pregnancies, such as gestational diabetes, and it may play a role in the overgrowth
of placenta, which supplies a growing fetus with nutrients and water. This function
of placenta is partly based on the expression of aquaporins/glyceroporins, such as
aquaporin 9 (AQP9).
The aim of this study was to analyse the expression of AQP9 in placenta from
gestational diabetes compared with that from normal pregnancy. In addition, we
studied the in vitro effect of leptin on the AQP9 expression by trophoblast explants
from control pregnancies, in order to check the possible leptin regulation of AQP9
expression in human placenta.
Methods: We collected 30 placentas (20 from control pregnancy and 20 from
gestational diabetes), after caesarean delivery at term. The expression of AQP9
was determined by quantitative RT-PCR, immunoblot and immune histochemistry.
In addition, explants from control placenta were incubated in vitro with increasing
concentrations of leptin (0-100 nM) during 6 h and AQP9 expression was also
measured. Mean values were analysed by ANOVA followed by Bonferronis post test.
Results: We have found that AQP9 expression was significantly increased by
30% in placenta from gestational diabetic pregnancy. Western blot confirmed
the increased protein level of AQP9 in placenta from gestational diabetic women.
Immune histochemistry demonstrated the localization of the overexpressed AQP9
in the sincitiotrophoblast. In vitro incubation of trophoblast explants with increasing
concentrations of leptin demonstrated the dose-dependent effect of leptin on the
expression of AQP9, with a maximal effect at 10 nM increasing three-fold the basal
leptin expression.
Conclusions: Data show that AQP9 expression is increased in placenta from pregnant
women with gestational diabetes and this may contribute to supply more nutrients to
the fetus. The increased expression of AQP9 in gestational diabetes may be mediated
at least in part by leptin, since leptin levels are known to be increased in gestational
diabetes and we have found that leptin stimulates the expression of AQP9 in vitro. In
conclusion, we have shown that leptin positively regulates the expression of AQP9 in
the trophoblast and this effect may mediate the observed increase in AQP9 expression
in trophoblast from pregnant women with gestational diabetes.

B-257

B-258
Neonatal umbilical cord blood cardiac troponin as reflecting fetal growth, age
and well-being.

J. M. Potter1, Z. Kesckes1, F. Nouri-Girones1, G. Koerbin2, B. Jayasinghe3,

P. E. Hickman1. 1Australian National University, Canberra, Australia,
2
University of Canberra, Canberra, Australia, 3Canberra Hospital,
Canberra, Australia
Background: The advent of highly sensitive assays for cardiac troponin have made
possible investigations to better understand the biology and pathophysiology of these
cardiac markers. The aim of the current study was to document the umbilical cord
blood concentrations of troponins I (cTnI) and T (cTnT) using high sensitivity assays
and correlate these with maternal and fetal clinical history.
Methods: Umbilical cord blood was collected immediately following delivery from
416 babies, including 12 sets of twins. Cardiac troponins were assayed using hs-cTnI
on Abbott Architect and hs-cTnT on Roche E4111. Clinical history was obtained from
clinical notes. Ethics permission was obtained from ACT Health Human Research
and Ethics Committee for the study and consent was obtained from mothers for their
participation.
Results: All results were above LoD for both assays (1.0 ng/L cTnI; 5.0 ng/L cTnT).
Minimum cTnI concentration was 1.2 ng/L (median 6.9n/L; Q1 4.5, Q3 12.1), cTnT
7.0ng/L (Q1 27.1, Q3 51.0). Troponins were statistically significantly correlated
[cTnT=21.45 ln(cTnI)-0.82; R2 = 0.3418, P<0.001]. Babies who were born before 32
weeks gestation (n=17) had higher median cTnI of 14.7ng/L (Q1 6.7, Q3 22.7) and
cTnT of 58.0 ng/L (Q1 56, Q3 102) compared with those born after 41 weeks gestation
[n=24, median cTnI 6.0ng/L (Q1 4.5, Q3 10.43); cTnT of 31.2 ng/L (Q1 23.0, Q3
48.6). Babies with the highest cTnI (>48 ng/L, >95%ile) had markedly elevated cTnT
(n=6, median 197 ng/L, range 151-297).
Conclusion:
The relationship between cTn and current objective measures of fetal well-being are
assessed to determine whether cTn measurement in this clinical setting is of value.

B-259
Dynamic changes in circulating amino acids and acylcarnitines in children
and adolescents: A CALIPER study of healthy community children and new
pediatric reference intervals

T. Teodoro-Morrison1, L. Kyriakopoulou2, J. Raizman1, V. Bevilacqua1, A.

Cohen2, M. Chan2, B. Wan2, M. Yazdanpanah2, Y. Chen2, D. Colantonio2,
K. Adeli2. 1University of Toronto, Toronto, ON, Canada, 2Hospital for Sick
Children, Toronto, ON, Canada

Absence of eosinopenia as predictor of procalcitonin < 2 ng/mL in pediatric

patients

e. fernandez rodriguez, c. perez mendez, s. garcia alonso, e. fernandez

garcia. Hospital Cabuenes, Gijon, Spain
BACKGROUND: The role of eosinopenia as a marker of sepsis has recently been
evaluated. The aim of our study was to assess whether eosinophilia was a reliable
criteria to predict procalcitonin (PCT) values under 2 ng/mL in pediatric patients
presenting to the emergency department to rule out sepsis.
METHODS: From January 2012 to December 2013, a total of 605 consecutive
children aged 3 to 36 months (median: 16.5 m; 46.6% male) were enrolled in the study.
Procalcitonin levels and eosinophil counts were measured on admission. PCT was
measured with a two-step two-site sandwich electrochemiluminescence immunoassay
(ECLIA) on the ROCHE cobas e 411 system. An undetectable eosinophil count
(<0.01 x 10(9) or <10/mm3) was considered as eosinopenia.
RESULTS: The absolute eosinophil count in the study population was 30/mm3

Objective: An up-to-date and comprehensive database of pediatric reference intervals

is essential for the accurate management of children with metabolic disease. The
Canadian Laboratory Initiative for Pediatric Reference Intervals (CALIPER) program
has established pediatric reference intervals from healthy community children for a
comprehensive list of common biomarkers relevant to the diagnosis and management
of inborn errors of metabolism (IEM).
Methods: Healthy children and adolescents from birth to 19 years were recruited
based on informed parental consent. A cohort of over 500 healthy individual samples
was used to calculate pediatric reference intervals for 36 acylcarnitines (LC-MS/
MS) and 37 amino acids (Waters MassTrak amino acid analyzer). Over 100 healthy
individual samples were focused in the 0 - 2 week age range to ensure reliable
reference intervals were established in the newborn period. Reference intervals were
calculated using non-parametric statistics according to CLSI-C28-A3 and partitioned
based on age and sex.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S205

Pediatric/Fetal Clinical Chemistry

Wednesday, July 30, 9:30 am 5:00 pm

Results: The majority of analytes demonstrated reference intervals requiring 2 - 4 age
dependent partitions, with the most common being within the newborn period of 0 - 2
weeks. Also, several analytes displayed a unique reference interval during puberty,
some of which demonstrated sex-based partitions. These were often related to energy
metabolism and growth in the muscle. Finally, the minority of analytes demonstrated
one reference interval for all ages and sexes.
Conclusions: The reference intervals established here for acylcarnitine and amino
acid profiles will aid in the accurate diagnosis and monitoring of children suspected
of IEM. Furthermore, reference intervals extending to 19 years of age will aid in the
management of children and adolescents treated for metabolic disease. Importantly,
these reference intervals are established for the indicated instrumentation and should
be validated on other platforms and for local populations as recommended by CLSI.

Analyte

Age

CA19-9

0 - < 1 year

Age

Number of samples

0-<2wk
2wk-<13yr
13yr-<19yr
BHB
0-<1yr
1yr-<19yr
(mmol/L)
0-<1yr
Total carnitine 1yr-<12yr
(umol/L)
12yr-<19yr

Valine
(umol/L)

145
253
: 46
: 51
91
219
67
136
: 53
: 56

Lower limit
(95% CI)
92 (73-96)
126 (123-139)
166 (157-179)
155 (145-162)
0 (0-0)
0 (0-0)
21 (17-26)
28 (27-31)
32 (30-33)
24 (21-27)

Upper limit
(95% CI)
326 (299-349)
356 (339-364)
301 (289-309)
259 (246-269)
1.73 (1.16-2.13)
0.12 (0.11-0.13)
78 (74-81)
61 (55-74)
59 (57-61)
53 (50-54)

B-260
Closing the gaps in pediatric population reference values for cancer
biomarkers: A CALIPER study of healthy community children

V. Bevilacqua , M. Chan , D. Armbruster , B. Shodin , K. Adeli . University

of Toronto, Toronto, ON, Canada, 2The Hospital for Sick Children, Toronto,
ON, Canada, 3Abbott Diagnostics, Abbott Park, IL
1

1 1

Background: The CALIPER (Canadian Laboratory Initiative in Pediatric Reference

Intervals) program, a national research initiative aimed at closing the gaps in pediatric
reference intervals, sought to develop a database of covariate-stratified reference value
distributions for 11 key circulating tumor markers including those used in assessment
of patients with childhood or adult cancers.
Methods: Healthy community children from birth to 18 years of age were recruited
to participate in the CALIPER project with informed parental consent. Participants
completed questionnaires and were assessed according to established inclusion
and exclusion criteria. Serum samples from approximately 400-700 children were
analyzed on the Abbott Architect ci4100 and reference intervals were established for
Alpha-Fetoprotein (AFP), Ant-Thyroglobulin (Anti-Tg), Human Epididymis Protein
4 (HE4), Cancer Antigen 125 (CA125), Cancer Antigen 15-3 (CA15-3), Cancer
Antigen 19-9 (CA19-9), Pro Gastrin-Releasing Peptide (ProGRP), Carcinoembryonic
Antigen (CEA), Squamous Cell Carcinoma Antigen (SCC), as well as Total and Free
Prostate Specific Antigen (PSA) according to CLSI C28-A3 statistical guidelines.
Results: Significant fluctuations of biomarker concentrations by age and/or gender
were observed in 10 of 11 biomarkers investigated. Results for three of the markers
examined (CA19-9, CA125 and SCC) are shown in Table 1. Age partitioning was
required for CA15-3, CA125, CA19-9, CEA, SCC, ProGRP, Total & Free PSA, HE4
and AFP while gender partitioning was also required for CA125, CA19-9, Total &
Free PSA.
Conclusion: The establishment of pediatric reference intervals for tumor biomarkers
will not only aid in harnessing the full potential of tumor markers in a pediatric
population but also in research aimed at determining the value of tumor marker use
in various cancers.

< 2.00, 54.40

1 - < 14 years < 2.00, 27.40

14 - < 19
years
CA125

< 2.00, 12.02

SCC

59

< 2.00, (46.39,

58.90)

< 2.00, 37.6

56

172

< 2.00, (22.80,

32.60)

< 2.00, 27.40

172

66

< 2.00, (10.32,

13.13)

< 2.00, 12.02

66

0 - < 4 months 2.4, 22.2

56

4 months - < 5
7.0, 32.6
years

191

5 - < 11 years 4.7, 28.5

Table 1. Pediatric reference intervals for biomarkers used in children with suspected
metabolic disease. (BHB-beta-hydroxybutyrate).
Analyte

Age- and Sex-Specific Pediatric Reference Intervals for 3 Tumor Biomarkers

Female
Male Reference
Reference
Interval
Interval
Reference
No. Of Confidence
Reference
No. Of
Interval
Samples Interval
Interval
Samples

11 - < 19
years
0 - < 1 week
1 week - <
1 year
1 - < 8 years

174

5.5, 27.7

122

> 70.0

44

0.4, 6.7

114

0.4, 1.7

8 - < 19 years 0.5, 2.1

81
142

(2.0, 3.0), (20.2,

24.4)
(4.9, 8.0), (31.6,
35.3)
(4.4, 5.8), (24.4,
29.4)
(5.4, 6.5), (23.5,
30.2)
> 70.0
(0.4, 0.5), (5.6,
7.9)
(0.3, 0.5), (1.6,
1.8)
(0.4, 0.5), (2.0,
2.2)

2.4, 22.2

56

7.0, 32.6

191

4.7, 28.5

174

5.5, 39.1

128

> 70.0

44

0.4, 6.7

114

0.4, 1.7

81

0.5, 2.1

142

Confidence
Interval
< 2.00,
(31.04,
42.63)
< 2.00,
(22.80,
32.60)
< 2.00,
(10.32,
13.13)
(2.0, 3.0),
(20.2, 24.4)
(4.9, 8.0),
(31.6, 35.3)
(4.4, 5.8),
(24.4, 29.4)
(3.9, 6.7),
(30.4, 41.0)
> 70.0
(0.4, 0.5),
(5.6, 7.9)
(0.3, 0.5),
(1.6, 1.8
(0.4, 0.5),
(2.0, 2.2)

B-261
Sex-based differences in gestational age at lung maturity as determined by
lamellar body counts

C. Bookhout, C. Beamon, R. Strauss, F. Lin, M. Gearhart, C. HammettStabler. UNC Hospitals, Chapel Hill, NC
Introduction: In late gestation, as fetal lungs prepare to transition to an air environment,
alveolar epithelial cells called type II pneumocytes synthesize and store a mixture
of phospholipids and proteins known as pulmonary surfactant. Around week 32 of
gestation, surfactant release into the amniotic fluid begins in the form of structures
called lamellar bodies (LBs). Adequate surfactant is an important predictor of fetal
lung maturity, especially in cases of prematurity.
Lamellar bodies can be quantified in amniotic fluid using the platelet channel of
automated hematology analyzers due to similarity in size. Inadequate surfactant
production is associated with respiratory distress syndrome, which is more likely to
occur in patients with lower lamellar body counts (LBCs).
Respiratory distress is more common in male late preterm infants than in females,
possibly related to sex-based differences in lung maturity. Female fetuses have
demonstrated earlier development of lung maturity than males in previous studies
using other markers of lung maturity, such as the lecithin/sphingomyelin ratio,
presence of phosphatidylglycerol, and loss of phosphatidylinositol; however, sex
differences in lamellar body counts have not been assessed in the literature. This study
aimed to determine if female fetuses demonstrate evidence of lung maturity using
LBCs at an earlier gestational age than males.
Methods and Analysis: The population for this retrospective cohort study included all
pregnant women who had amniocentesis with LBC analysis on the Advia Hematology
System at our institution from 2003-2012 and subsequently delivered within 72
hours. Data were collected from our laboratory database of amniotic fluid LB counts.
Gestational age at the time of the amniocentesis, fetal sex, and additional demographic
and outcome data were collected from medical records. Lung maturity was defined as
a LB count >35,000. Linear regression analysis of the data was done using Microsoft
Excel, with statistical analysis performed by analysis of covariance (ANCOVA) to
evaluate the relationship between gestational age, lamellar body count, and sex.
Results: 263 deliveries were included for analysis, with 128 female (49%) and 135
male (51%) infants. Gestational ages ranged from 30-41 weeks and lamellar body
counts from 4332 to >200,000. Lung maturity with LBC>35,000 was demonstrated in
105/128 female (82%) and 107/135 male (79%) infants. The regression line of LBC
versus gestational age for females was y=10188x-311669, with r=0.47, and for males
it was y=10391x-317262, with r=0.47. The two lines did not differ significantly in
either slope (p=0.93) or intercept (p=0.95).
Conclusion: Our study did not demonstrate a significant difference in gestational
age at lung maturity by fetal sex as measured with LBCs. These results differ from
prior data showing a higher degree of lung maturity in female than male infants using
other indices, despite larger sample sizes in our study. Broad inclusion criteria and
relatively wide scatter of data may have contributed to the non-significant results.
Sex differences in LBCs were not found to explain increased prevalence of neonatal
respiratory distress in males.

S206

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Pediatric/Fetal Clinical Chemistry

Wednesday, July 30, 9:30 am 5:00 pm

Pediatric reference intervals for 29 endocrine and special chemistry biomarkers

on the Beckman Coulter DxI Immunoassay System: A CALIPER sudy of
healthy community Children

equipment, in a private laboratory, between September 2013 and February 2014. The
statistics included Box-Cox transformation, Tukeys test, Kolmogorov-Smirnov test
and estimation of RI by non-parametric method of the percentiles. The confidence
interval was determined as 90% for the reference limits (percentile 2.5 and 97.5). The
statistics were calculated by MedCalc software.

M. Chan, V. Bevilacqua, Y. Chen, S. Bustos, C. ODwyer, K. Adeli.

CALIPER Program, Pediatric Laboratory Medicine, The Hospital for Sick
Children, Toronto, ON, Canada

Results: Tukeys test was used in the transformed data and it identified 2.39% of
individuals as outliers. All results considered as outliers were excluded from the
original data. The RIs were determined by non-parametric method of the percentiles
of residual data (Table 1).

B-262

BACKGROUND: Appropriate interpretation of laboratory test results requires

carefully established reference intervals based on a healthy population. Growth and
development can markedly influence circulating concentrations of biomarkers, so
accurate reference intervals established on the basis of a healthy pediatric population
are essential for test result interpretation. The CALIPER (Canadian Laboratory
Initiative on Pediatric Reference Intervals) program, a national research initiative
aimed at closing the gaps in pediatric reference intervals, sought to develop a database
of covariate-stratified reference intervals for endocrine and special chemistry markers
on the Beckman Coulter DxI Immunoassay System.
DESIGN AND METHODS: Healthy children and adolescents were recruited as part
of the CALIPER study. After informed parental consent was obtained, participants
filled out a questionnaire including demographic information and provided a blood
sample. We measured 29 proteins using the Beckman Coulter DxI Immunoassay
System utilizing 443 - 636 samples per assay. The variance, age- and sex-specific
concentrations of analytes were visually inspected from scatterplots of protein
concentration as a function of age for both genders. Pediatric reference intervals
were calculated according to Clinical Laboratory Standards Institute (CSLI) C28-A3
guidelines. Partitions based on age and/or sex were determined and statistically
evaluated using the Harris and Boyd method. After removal of outliers, reference
intervals were calculated using the non-parametric rank method, with values ranked
and the 2.5th and 97.5th percentiles calculated.
RESULTS: We observed a complex pattern of change in most protein concentrations
from the neonatal period to adolescence. The changes in concentration observed for
each of the examined proteins were classified into 1 of 5 categories: (a) high variance
and high concentration within the neonatal period that decreases abruptly shortly
after birth: AFP, ferritin and prolactin; (b) high variance at birth that is significantly
reduced around 1 year of age: free T4 and thyroglobulin (c) high variance and high
concentration within the neonatal period that decrease gradually with age: SHBG and
vitamin B12; (d) high variance at birth that decreases abruptly around 1 year of age
and increases again in adolescence: cortisol, DHEAS, folate, testosterone (males),
FSH (especially females) and LH (especially females); and (e) constant variance
throughout life but variable concentration according to age: free T3 and total T4.
Estradiol (and progesterone to a lesser extent) concentrations and variance were low
from birth, then increased in females during adolescence. Insulin increased slightly
with age for both genders. Ostase (bone-specific alkaline phosphatase) displayed
constant variance and concentration, with a sharp decrease in adolescence, especially
in females.
CONCLUSIONS: This study shows the complex expression profiles of several
endocrine and special chemistry biomarkers as a function of age and gender. This
allowed the calculation of age- and sex-specific reference intervals for 29 endocrine
and special chemistry markers, specific to the Beckman Coulter DxI Immunoassay
System. The completion of this study will aid in accurate diagnosis and laboratory
assessment of children being monitored by Beckman Coulter immunoassays in
healthcare institutions worldwide. It is however important that these reference
intervals be validated for the local pediatric population as recommended by CLSI.

B-263
Using the clinical laboratorys database for indirect estimation of the reference
intervals of serum creatinine.

A. C. Dias, A. L. Barbosa, S. F. Fonseca, S. S. S. Costa, L. F. A. Nery. Sabin

Laboratory of Clinical Analysis, Braslia, Brazil
Background: Reference intervals (RI) age-specified of serum creatinine are important
for triage, diagnostics and monitoring of chronic kidney disease. Ideally, RIs must be
determined by sampling a healthy population. There are two methods of sampling: a
direct one and an indirect one. The direct method, that follows the recommendations
of CLSI and IFCC, involves the selection of reference individuals. The aim of this
study is to estimate the RIs of serum creatinine through the indirect sampling method,
suggested by Horn.
Methods: Transversal study conducted in 44.592 individuals, ages between 1 and 74,
both genders, which performed serum creatinine (enzymatic method) in Advia 2400

Table 1. Reference intervals for creatinine concentratios in serum (mg/dL)

calculated with a nonparametric method.
Lower limit
Upper limit
Males
2.5 percentil (90% 97.5 percentil
Age group
n
outliers
(%)
CI)
(90% CI)
1 to < 3 years
347
48.86 5
0.19 (0.18 to 0.20) 0.38 (0.36 to 0.40)
3 to < 5 years
417
51.86 13
0.26 (0.25 to 0.26) 0.46 (0.44 to 0.50)
5 to < 7 years
434
44.92 9
0.30 (0.29 to 0.31) 0.53 (0.51 to 0.55)
7 to < 9 years
496
40.76 7
0.33 (0.32 to 0.34) 0.58 (0.56 to 0.60)
9 to < 11 years 501
36.40 10
0.37 (0.35 to 0.37) 0.62 (0.61 to 0.62)
11 to < 13 years 482
37.37 13
0.41 (0.41 to 0.42) 0.71 (0.69 to 0.72)
13 to < 15 years 613
37.04 8
0.46 (0.44 to 0.47) 0.86 (0.83 to 0.89)
Women (18 to <
25536 n.a.
600
0.50 (0.50 to 0.50) 0.94 (0.94 to 0.95)
75 years)
Men (18 to < 75
14700 n.a.
401
0.70 (0.69 to 0.70) 1.25 (1.24 to 1.25)
years)
Conclusion: The direct method of sampling is preferred. However, the difficulty to
obtain a representative number of reference individuals, especially for pediatrics
population, can be overcome by using the indirect method. This method demonstrated
to be an excellent alternative method to determine the RIs, which for serum creatinine
are compatible with the main description in the literature.

B-264
A Urine-based Immunoassay for Urocortin 3 and Diagnosis of Sleep Apnea

T. Pitcher1, K. Goudy2, M. Linder3, R. Valdes3. 1University of Louisville,

Louisville, KY, 2PGxL Technologies, Louisville, KY, 3University of
Louisville and PGxL Technologies, Louisville, KY
Introduction: Obstructive sleep apnea (OSA) affects 2-3% of children in the US and
is characterized by obstruction of the upper airway during sleep resulting in disruption
in ventilation, hypoxemia, and sleep fragmentation. Children suffering from OSA are
more likely to develop behavior difficulties, learning disabilities, pulmonary/systemic
hypertension and decreased growth. Currently, the gold standard for diagnosing
OSA is an overnight sleep study which is labor intensive, limited by availability
and expensive. Development of a non-invasive test using a urine biomarker would
provide an inexpensive and more accessible test to diagnose OSA. In a previous
study, proteomic analysis by mass-spectrometry revealed a 4 kilo Dalton peptide in
urine which was significantly increased in children with OSA compared to controls
(Pub Med #). The peptide was identified as the stress-coping peptide Urocortin 3
(UCN3); 38 amino acids in length and is expressed in kidney tubules, heart and brain.
UCN3 is involved in modulating stress responses, osmoregulation, and in regulating
the hypothalamus-pituitary-adrenal axis. The objective of this present study was to
develop a urocortin 3 (UCN3) two-site immunometric assay for eventual use as a
rapid non-invasive diagnostic screening tool for OSA in children
Methods: To develop monoclonal and polyclonal antibodies targeting the mature
human peptide, we immunized mice and rabbits with synthetic UCN3 constructs
conjugated with KLH. Following conventional prime and boost immunization
strategies, the mice were sacrificed, the spleens were harvested and immortalized.
ELISA was used to identify the clones with reactivity for unconjugated UCN3. Full
length UCN3 and UCN3 fragments were coated in 96-well plates and assayed with
hybridoma supernatants (neat and 1:10 dilution) to determine reactivity. The ideal
clones were defined as those that yielded the highest signal to full length UCN3 and
to either the N-terminal or C-terminal half of the peptide. All of the antibodies were
affinity purified and used to develop a two-site immunometric assay. Antibody pairs
were identified using checkerboard ELISA and affinity characterization using Biacore
(GE life science). The highest affinity mAb was used for antigen capture and the
polyclonal was used for detection.
Results: Analysis of the clones yielded twenty hybridomas reactive towards UCN3,
with five clones exhibiting absorbance >1.0 to full length and either the N- or
C-terminus of UCN3. Checkerboard and Biacore analysis using these five clones
and an anti-UCN3 polyclonal antibody (pAb) identified that mAb 4D3 was the ideal
capture antibody and the anti-UCN3 pAb as the detection antibody. These antibodies
generated an assay with a linear range from 100 to 3100 ng/mL, an intra-assay CV
of 2.7 and 3.7% (N=2) at 100ng/ml and 3160ng/ml, respectively. Recovery of UCN3
spiked into urine was 127% at 100 ng/mL and 100% at 3100 ng/mL (N=3).

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S207

Pediatric/Fetal Clinical Chemistry

Wednesday, July 30, 9:30 am 5:00 pm

Conclusions: The set of monoclonal and polyclonal antibodies developed provides a
Urocortin 3 ELISA-based assay with analytical characteristics suitable for subsequent
studies to test the clinical predictive value measuring UCN3 in urine for the diagnosis
of OSA in pediatric subjects.

B-265
CLSI-based transference of the CALIPER database of pediatric reference
intervals to Beckman Coulter Clinical Chemistry Assays

V. Bevilacqua1, M. Chan1, Y. Chan1, S. Bustos1, C. ODwyer1, E. Randell2, K.

Adeli1. 1CALIPER Program, Pediatric Laboratory Medicine, The Hospital
for Sick Children, Toronto, ON, Canada, 2Eastern Health Authority &
Faculty of Medicine, Memorial University, St. Johns, NL, Canada
OBJECTIVES: Reference intervals represent the range of results that are commonly
observed in a population of healthy individuals. These intervals are defined as the range
that encompasses the central 95% of the distribution of test results from reference
individuals sampled from a healthy reference population. Comparison of a given test
result to an appropriate reference interval enables proper clinical assessment. Accurate
pediatric reference intervals obtained in healthy community children are essential
for the accurate diagnosis and management of diseases in children. The CALIPER
program has established a comprehensive database of age- and sex-stratified pediatric
reference intervals for over 70 common biochemical markers, proteins, lipids, and
enzymes, as well as endocrine markers and fertility hormones. However, this database
was only directly applicable for assays performed on the Abbott ARCHITECT ci4100
system. We therefore sought to expand the scope of this database to biochemical
assays performed on the Beckman Coulter Syndchron DxC800 chemistry platform,
allowing for a much wider application of the CALIPER database.
DESIGN AND METHODS: Based on CLSI C28-A3 and EP9-A2 guidelines,
CALIPER pediatric reference intervals were transferred to Beckman Coulter
chemistry assays performed on the Synchron DxC800 instrument, using specific
statistical criteria. First, 200 pediatric pooled patient serum specimens were analyzed
on the Abbott ARCHITECT ci4100 and the Beckman Coulter DxC800, and the
data was subjected to regression analysis, standardized residual, Bland-Altman, and
quantile-quantile (Q-Q) plots. A total of 34 chemistry, enzyme, lipid/lipoprotein and
protein markers were assessed. Regression analysis was performed to determine
correlation of values between analytical systems, resulting in determination of R2
for each analyte. Analytes with R2 values equal to or greater than 0.95 were deemed
transferable to the Beckman Coulter DxC800 platform. To assess the validity of the
transferred reference intervals, 100 serum samples from the CALIPER cohort (healthy
community children) were assayed on the Beckman Coulter DxC800 and validation
was assessed based on CLSI C28-A3 criteria.
RESULTS: Most (22) of the analytes showed good correlation between the Abbot and
Beckman systems with R2 values equal to or greater than 0.95. Ten analytes showed
strong correlation, with R2 values between 0.77 and 0.94. Two analytes (carbon
dioxide and calcium) showed poor correlation, and were not transferable to Beckman
Coulter DxC800. Most transferred reference intervals determined using the Beckman
system were validated through the analysis of CALIPER reference samples.
CONCLUSIONS: The current study allows successful transference of a large number
of routine and special chemistry markers from the CALIPER database to the assays
on the Beckman Coulter DxC800 analytical systems. This will greatly extend the
utility of validated CALIPER reference intervals which will be directly applicable for
assays performed on the Beckman Coulter DxC800 chemistry platform. Validation of
the reference intervals should facilitate the broad application of CALIPER reference
intervals at pediatric centers worldwide.

B-266
Development of a Pregnant Subject Biospecimen Bank

A. M. Gronowski, R. Colvin, C. Kramer, K. H. Moley. Washington

University, St. Louis, MO
Background: A need exists in the research community for serum, urine, cord blood,
placenta and other tissues from pregnant women in order to study disorders that affect
both the mother and fetus and to establish gestational age-specific reference intervals.
Objective: To create an infrastructure to recruit pregnant subjects, collect biospecimens
throughout pregnancy and make them available to researchers along with clinical
information and outcomes.
Methods: In 2008, a system was put into place to consent subjects and track them
across a major medical center throughout their pregnancy. Biospecimens, including

S208

serum/plasma, urine, vaginal swabs, follicular fluid, placenta, cord blood, semen, and
infant heel stick blood are collected. Clinical data is gathered and stored in a database.
A business plan was developed and a cost-recovery fee structure was implemented.
Specimens are provided to researchers immediately or frozen and stored for short
or long periods of time, at the researchers discretion. Long term storage utilizes an
already established university specimen repository.
Results: Funding has been obtained to support this structure for six years. The cost
to launch a similar Biobank is ~$200,000 per year and to maintain this infrastructure
has been ~ $250,000 per year and includes 3.5-5.0 FTE. Average enrollment is 12
women/week from 4 recruitment sites. Over 4,700 samples have been distributed to
11 researchers from 5 different university departments, and over 40,000 samples have
been banked for future research. Major clinical outcomes are shown in the Table.
The Biobank has helped university researchers receive grant funding including R01,
SCOR, ICTS, and March of Dimes.
Conclusions: Here we describe the successful formation of a biorepository for
specimens collected from subjects longitudinally throughout pregnancy. Over a six
year period, we demonstrate that: funding can be obtained, enrollment is sufficient,
accumulation of clinically significant outcomes can be achieved, and the Biobank
allows researchers to fund and conduct research.
Table. Outcome data from subjects included in biobank
Number
%
1669
84.9
297
15.1

Pregnant
Not pregnant
Status of Pregnant patients (if known)
Delivered
Loss
Unknown
Complications
Preeclampsia
Pregnancy induced hypertension
Intrauterine growth restriction
Macrosomia
Fetal anomalies
Chorioamnionitis
Gestational diabetes

1353
37
279

81.1
2.1
16.7

165
139
101
87
84
102
92

9.9
8.3
7.5
6.4
6.2
6.1
5.5

B-267
Evaluation of a Discriminatory Zone for Serum Beta-human chorionic
gonadotropin (hCG) in Early Pregnancy

M. A. V. Willrich, N. A. Baumann, N. V. Tolan, G. G. Klee, D. Brown, C.

C. Coddington. Mayo Clinic, Rochester, MN
Background: The beta-human chorionic gonadotropin (hCG) discriminatory zone
is the concentration of serum hCG at which a gestational sac should be visible on
sonography in a normal intrauterine pregnancy (IUP) and has been used to aid in
management of women presenting with pain and/or vaginal bleeding in early
pregnancy. The reliability of an hCG discriminatory zone has been debated in the
literature. In addition, large inter-individual variation of serum hCG concentrations
and the lack of standardization between assays prohibits the use of a universal hCG
cut-off. The aim of this project was to perform a retrospective study to determine a
discriminatory zone for the serum hCG assay used in our institution.
Methods: Inclusion criteria included females with clinician-ordered ultrasound (US)
performed and serum hCG measured within 48h between June 2010 and December
2012 (n=554). The Roche Cobas intact hCG+ assay was used to measure hCG on a
Cobas e immunoassay analyzer (Roche Diagnostics). Chart review was performed on
a subset of unique patients with serum hCG concentrations between 1,000-5,000 mIU/
mL to determine pregnancy outcomes (n=106). Ultrasound reports were reviewed
and presence of embryonic structures recorded. Serum hCG concentrations were
correlated with presence or absence of gestational sac (GS) and pregnancies were
categorized as ectopic or IUP. The IUP group was further categorized into viable or
non-viable pregnancies. ROC curve analysis was performed.
Results: In the subset of chart-reviewed women, the median age was 31 years old
(range:18-51) and serum hCG concentration was 2320973 mIU/mL (meanSD).
Serum hCG concentrations were independent of estimated weeks of gestation
(p=0.9611). GS was present in 66 cases (hCG 24121030 mIU/mL) and not visible by
US in 39 cases (hCG 2180870 mIU/mL). Of those cases with no GS present, chart
review revealed that 15 were ectopic (hCG 2372902 mIU/mL) and 24 were IUP
(hCG 2060846 mIU/mL). Serum hCG concentrations were not different between
ectopic or IUP (24461128 mIU/mL versus 2287933 mIU/mL, respectively,
p=0.7056). Serum hCG concentrations between viable and non-viable IUPs were
similar (2397943 mIU/mL versus 2220929 mIU/mL, respectively, p=0.3198). ROC
curve analysis identified that in the absence of a GS, a serum hCG concentration of
1700 mIU/mL would be 87% sensitive and 63% specific in differentiating an ectopic
pregnancy from an IUP (AUC=0.63, p=0.2783), and a cut-off of 3900 mIU/mL would
provide a sensitivity of 13% and a specificity of 95%.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Pediatric/Fetal Clinical Chemistry

Wednesday, July 30, 9:30 am 5:00 pm

Conclusions: While defining an hCG discriminatory zone would prove valuable to

clinicians, especially in the absence of visible gestational sac on sonography, we
found a large overlap of serum hCG concentrations in females with ectopic and IUP.
The use of a single measurement of serum hCG below or above a given threshold does
not provide enough sensitivity or specificity to definitely diagnose ectopic pregnancy.
However, when used in conjunction with clinical judgment and sonography, a higher
hCG cut-off of 3900 mIU/mL provides increased specificity and may aid in the
diagnosis and management of patients.

B-269
Quantitative amino acid analysis using liquid chromatography tandem mass
spectrometry and aTRAQ reagents. Do we have a new gold standard?

A. Liu1, J. Hobert2, M. Erali1, J. Pickering1, N. Longo3, M. Pasquali2, I.

De Biase2. 1ARUP Instute for Clinical and Experimental Pathology, Salt
Lake City, UT, 2Department of Pathology, University of Utah School of
Medicine, Salt Lake City, UT, 3Department of Pediatrics, University of
Utah School of Medicine, Salt Lake City, UT
Background: Defects in the metabolism or the transport of a specific amino acid or
a group of amino acids leads to disorders generally referred to as inherited disorders
of amino acid metabolism. Accurately quantifying amino acids in biological fluids
(plasma, urine, or cerebrospinal fluid) is essential for the diagnosis and follow up of
aminoacidopathies, as well as useful for the nutritional assessment of patients with
non-metabolic conditions. Amino acid analysis is conventionally performed on an
ion-exchange chromatography (IEC) based amino acid analyzer, which provides
excellent separation and reproducibility with minimal sample preparation. The IEC
method has several disadvantages: long run time, large sample volume, and lack of
analyte specificity due to interfering substances. To address our large clinical load
and improve specificity, we have optimized the aTRAQ method (AB SCIEX) with
ion-pairing reverse-phase liquid chromatography and tandem mass spectrometry, and
transferred the assay in our clinical lab.
Methods: Samples were labeled with aTRAQ reagents prior to instrument analysis by
API 4000 in conjunction with SHIMADZU HPLC. The chromatographic separation
was performed using the ABSciex amino acid analysis column or the Phenomenex
Gemini-NX column, C18, 4.6x150mm, 5um. Amino acid quantitation was obtained
using 6-point external calibration with stable isotope dilution. MultiQuant software
was used for data analysis.
Results: Our method allowed accurate quantification of 47 physiological amino acids
and related compounds, including the isomers alloisoleucine and isoleucine, and four
additional analytes: sulfocysteine, agininosuccinic anhydrides, formiminoglutamic
acid, and glycylproline. The assay analytical performance was evaluated using
standard solutions, spiked samples of varying concentrations, and de-identified
clinical specimens. The assay was linear from 1 umol/L to 2500 umol/L. The total
imprecision was less than 10% and the recoveries were between 90 and 110% for most
amino acids. The assay was compared with IEC method with 115 plasma samples and
38 urine samples, run in parallel. The comparison yielded good correlation for most
amino acids [slopes between 0.93 and 1.04, y-intercepts between -7.0 and +7.0, R2
greater than 0.96]. In urine, as expected, correlation was poor for several amino acids
due to interfering substances that cannot be separated by IEC. Reference intervals
were established with a focus on the pediatric population. De-identified clinical
samples from patients with known disorders of amino acids metabolism or transport
were also analyzed with this method and were 100% concordant with IEC.
Conclusion: We have optimized the aTRAQ procedure for amino acid analysis to
achieve lower imprecision and better batch to batch reproducibility. Compared with
the IEC method, this assay has shorter instrument run time and increased specificity.
We have made the assay robust and ready for implementation in the clinical lab.
We are able to report 47 amino acids and related compounds from a single sample
analysis. This represents a broad, all-inclusive panel for amino acids analysis that, in
our opinion, could replace ion-exchange chromatography in the clinical lab.

B-270
Fetal male lineage determination by analysis of Y-chromosome STR haplotype
in maternal plasma

T. H. Santa Rita, C. F. Chianca, L. F. R. Velasco, C. F. Sousa, L. F. A. Nery,

S. S. S. Costa, G. B. Barra. Laboratorio Sabin, Brasilia, Brazil
Background: The paternity testing is increasingly becoming a clinical laboratory
test and there are some situations that are necessary to perform fetal kinship analysis

before the delivery. The prenatal paternity testing is invasive and most frequently done
by testing of chorionic villi or amniotic fluid, which are associated with a stressful
sampling that bears a small but existing risk for both mother and child. Actually, with
the availability of SNP microarrays and next-generation sequencing, the prenatal
paternity testing is been performed non-invasively by analysis of the cell-free fetal
DNA in maternal plasma. However, such new methodologies are associated with
extras know-how, equipment, and reactions costs. In this way, it is highly desirable
to perform non-invasive fetal kinship analysis by using the legacy paternity testing
technology. However, the higher maternal DNA background limits the detection
of fetal DNA markers (STRs). So, it is necessary to explore the fetus and mother
genetics differences (e.g. the Y chromosome in case of male fetus). Thus, the aim
of this study is to determine the fetus Y-STR haplotype in maternal plasma during
pregnancy and estimate, non-invasively, if the alleged father and fetus belong to the
same male lineage.
Methods: The study enrolled couples with singleton pregnancies and known
paternity. All participants signed informed consent and the local ethics committee
approved the study. Fetal gender was determined by qPCR targeting DYS-14 in
maternal plasma and it was confirmed after the delivery. The first consecutive 20
and 10 mothers bearing male and female fetuses, respectively, were selected for the
Y-STR analysis. The median gestational age was 12 weeks (range 12-36). Peripheral
blood was collected in EDTA tubes (mother) and in FTA paper (father). Maternal
plasma DNA was extracted by NucliSens EasyMAG (Biomeriuex). All DNA samples
were subjected to PCR amplification by ampFLSTR Yfiler (Life Technologies),
PowerPlex Y23 (Promega) and an in-house multiplex, which together accounts for 27
different Y-STR. The PCR products were detected with 3500 Genetic Analyzer (Life
Technologies) and they were analyzed using GeneMapper-IDX (Life Technologies).
Fetuses haplotypes (in Y-Chromosome Haplotype Reference Database standard
format that consider only 15 of the 27 tested Y-STR) were compared to other 5328
Brazilian haplotypes available on www.yhrd.org.
Results: Between 22 and 27 loci were successfully amplified from maternal plasma in
all 20 cases of male fetuses. None of the women bearing female fetuses had a falsely
amplified Y-STR. The haplotype detected in maternal plasma matched the alleged
father haplotype in all cases. One case showed a mutation in the DYS438 locus, which
was confirmed after the delivery. Seventeen fetuses haplotypes were not found in
YHRD and three of them occurred twice, which corresponded to paternity probability
of 99.981% and 99.944%, respectively.
Conclusion: High discriminatory fetal Y-STR haplotype could be determined from
maternal plasma during pregnancy starting at 12 weeks of gestation. Moreover, all
male fetuses could be attributed to the alleged father male lineage early in pregnancy.
This strategy is an alternative for fetal kinship analysis before the delivery. The main
limitation is that its only applied for mothers bearing a male fetus.

B-271
Pediatric reference value distributions for vitamins A and E in healthy
community children: Establishment of new age-stratified reference intervals
from a CALIPER cohort

J. Raizman, A. H. Cohen, T. Theodoro-Morrison, B. Wan, M. Chen, C.

Wilkenson, V. Bevilaqua, K. Adeli. University of Toronto, Toronto, ON,
Canada
Objective: Vitamin A (retinol) and vitamin E (alpha tocopherol) are fat soluble
micronutrients measured in the pediatric population to monitor deficiencies due to
malabsorption secondary to gastrointestinal (GI) disorders. A major challenge of
vitamin A and E testing is lack of reliable pediatric reference intervals (RI) which
limits accurate interpretation of patient results. We report new pediatric RI for both
vitamins as part of the Canadian Laboratory Initiative for Pediatric Reference Intervals
(CALIPER). Methods: Healthy community children were recruited with parental
consent and whole blood samples collected from 342 healthy children 1 day to 19
years of age. Retinol and alpha tocopherol were extracted from serum using hexane
before concentrations were measured with high performance liquid chromatography.
Age and sex-specific RI were calculated using non-parametric and robust methods
based on CLSI C28-A2 guidelines. Results: Comparison of vitamin A and E levels in
males and females demonstrated a tight correlation and did not reveal any significant
differences requiring no sex partitioning. Further analysis by age demonstrated
distinct partitioning patterns. Both vitamin A and E showed age partitioning at 0 to <1
years with levels determined as early as the first day of life. Interestingly, vitamin A
exhibited a complex pattern necessitating 4 distinct age partitions trending toward a
general rise in levels over time. Vitamin E required 2 age partitions. Levels rose within
the first year of life but were reduced slightly after this period requiring only one broad
partition between 1 to <19 years. Ratios of vitamin E to cholesterol and triglyceride

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S209

Pediatric/Fetal Clinical Chemistry

Wednesday, July 30, 9:30 am 5:00 pm

were also calculated, correlating well to vitamin E levels. Conclusions: This study
establishes pediatric RI for vitamin A and E in a healthy population from neonates to
early adulthood. These values will be beneficial in assessing accurate vitamin status
when monitoring children with GI disorders or malnutrition.

1.58
2
2.53
51.36
33

Higher
Samples
confidence
(n)
interval
100
0.92,1.01
71
0.82,1.02
50
0.65,1.01
85
1.86,9.67
245
14,3.78

Lower
confidence
interval
1.54,1.61
1.90,2.09
2.33,2.70
48.26,54.58
6.45,6.92

3.7

6.7

82

3.53,3.78

6.45,6.92

8.53

44.48

83

7.51,9.52

40.37,47.76

Analyte (micro Age

mole/L
(years)

Lower Upper
Limit limit

Vitamin A

1 - < 11
11 - < 16
16 - < 19
0-<1
1 - <19

0.97
0.93
0.82
5.92
14.5

1 - <19
1 - <19

Vitamin E
Vitamin E/
cholesterol
Vitamin E/
triglyceride

B-272
Diagnosis of Primary Hyperoxaluria Type III, A Novel Hereditary Disorder of
Hydroxyproline Metabolism, By Gas Chromatography-Mass Spectrometry

L. Hasadsri, P. Loken, D. S. Milliner, J. C. Lieske, P. Rinaldo, S. Tortorelli,

D. Oglesbee. Mayo Clinic, Rochester, MN
Background and Objectives: Primary hyperoxaluria type III (PH3) is a newly
discovered disorder caused by deficiency of mitochondrial 4-hydroxy-2-oxoglutarate
(HOG) aldolase, which catalyzes the final step in the metabolism of hydroxyproline.
The condition is characterized by the infant to childhood onset of recurrent
nephrolithiasis and progressive nephrocalcinosis. Timely detection of primary
hyperoxalurias, in particular PH3, remains a significant challenge. Patients have often
reached end-stage renal disease by the time they are diagnosed. Here we describe
a novel method for the detection of urinary metabolites in primary hyperoxaluria
types I, II, and III by gas chromatography-mass spectrometry. We also summarize
the clinical and laboratory features of 9 known (i.e., mutation confirmed by sequence
analysis) and 2 novel cases of PH3 uncovered by our assay.
Methods: Patient samples and de-identified clinical information were obtained in
collaboration with the Mayo Clinic Hyperoxaluria Center and Rare Kidney Stone
Consortium. Samples analyzed included urine from unaffected controls, patients with
PH1-3, and patients with hyperoxaluria of unknown etiology. Urine specimens were
methoximated and extracted with 4:1 (v/v) ethyl acetate/propan-2-ol then evaporated
to dryness and derivatized with BSTFA+TMCS in pyridine. Samples were re-extracted
in isooctane and analyzed on an Agilent 5975 series, using hydrogen as the carrier
gas. The following commercially available internal standards were used: glycolate-D2,
oxalate-13C2, and glycerate-D3, in addition to custom-synthesized HOG-13C2.
Validation: The linear range of detection for HOG was 0.024 - 600 g/mg creatinine
(Cr). Pooled normal, intermediate, and elevated HOG urine samples (n=30 for each)
were used to determine intra-assay precision (%CV = 2.83, 0.73, and 4.73, respectively)
as well as inter-assay precision (%CV = 0.89, 0.48, and 2.60, respectively). In the
absence of a comparison method for the analysis of HOG, the accuracy of this test
was validating by correlation of test results with clinical findings. Blinded results
from analysis of one hundred normal and 34 abnormal specimens were submitted
for interpretation; 100% interpretive concordance was achieved for the 134 samples.
Results: The reference range of HOG concentrations in controls was 0.01 - 2.21 g/
mg Cr. In patients with PH1, the range was 0.01 - 0.96 g/mg Cr. Patients with PH2:
0.01 - 3.10 g/mg Cr. Testing of urine samples from patients with hyperoxaluria of
unknown etiology (i.e., negative for PH1 and/or PH2 by molecular testing) uncovered
two male infants with elevated levels of HOG, which were 116.86 and 261.78 g/
mg Cr, respectively. Mutation analysis of the HOGA gene in one of these patients
revealed homozygosity for a known pathogenic mutation.

B-273
Pediatric reference intervals for specialty endocrine and chemistry biomarkers
on the Abbott Architect ci4100 System: A CALIPER study of healthy
community children

V. Bevilacqua1, M. Chan1, Y. Chen1, S. Bustos1, C. ODwyer1, F. Quinn2,

B. Shodin2, D. Armbruster2, K. Adeli1. 1CALIPER Program, Pediatric
Laboratory Medicine, The Hospital for Sick Children, Toronto, ON,
Canada, 2Abbott Diagnostics, Abbott Park, IL
BACKGROUND: Appropriate interpretation of laboratory test results requires
carefully established reference intervals based on a healthy population. Growth
and development can markedly influence circulating concentrations of biomarkers,
thus accurate reference intervals established from a healthy pediatric population are
essential for test result interpretation. The CALIPER (Canadian Laboratory Initiative
on Pediatric Reference Intervals) program, a national research initiative aimed
at closing the gaps in pediatric reference intervals, sought to develop a database
of covariate-stratified reference intervals for a number of endocrine and special
chemistry markers.
METHODS: Healthy children and adolescents were recruited as part of the CALIPER
study. After informed parental consent was obtained, participants filled out a
questionnaire including demographic information and provided a blood sample. We
measured a number of specialty and endocrine and biochemical markers (Alpha-1
antitrypsin, AGP, Amylase P, Anti-CCP Anti-TPO, Beta 2 microglobulin, C Peptide,
Ceruloplasmin, Cholinestarase E, hs-CRP, Cystatin C, DHEA-Sulfate, Glucose,
IgE, Insulin, SHBG, Testosterone (2nd GEN, and Bioavailable/Free Testosterone
indexes) using the Abbott ARCHITECT ci4100 system and reference intervals were
established utilizing 367 - 763 samples per assay. The variance, age- and sex-specific
concentrations of analytes were visually inspected from scatterplots of analyte
concentration as a function of age for both genders. Pediatric reference intervals
were calculated according to Clinical Laboratory Standards Institute (CSLI) C28-A3
guidelines. Partitions based on age and/or sex were determined and statistically
evaluated using the Harris and Boyd method. After removal of outliers, reference
intervals were calculated using the non-parametric rank method if the sample size was
larger than 120, with values ranked and the 2.5th and 97.5th percentiles calculated.
If the sample size was between 40 and 120, the robust method was used in reference
interval calculation.
RESULTS: We observed a complex pattern of change in most analyte concentrations
examined from the neonatal period to adolescence. The changes in concentration
observed for each of the examined proteins were classified into 1 of 4 categories:
(a) high variance and high concentration within the neonatal period that decreases
abruptly shortly after birth: Beta-2-microglobulin, high-sensitivity CRP and Cystatin
C; (b) gradual concentration increase with age: albumin BCG, albumin BCP and
pancreatic amylase; (c) high variance and high concentration within the neonatal
period that decrease gradually with age: Sex hormone-binding globulin (SHBG) and
cholinesterase (less pronounced); and (d) high variance at birth that decreases abruptly
around 1 year of age and increases again in adolescence: C-peptide, DHEA-S and 2nd
GEN testosterone (males). Ceruloplasmin showed a unique expression pattern with
a sharp increase in concentration during the neonatal period followed by a gradual
decrease over time.
CONCLUSIONS: This study shows the complex expression profiles of several
endocrine and chemistry biomarkers as a function of age and gender. This allowed for
establishment of age- and sex-specific reference intervals for these biomarkers, which
will aid in accurate diagnosis of pediatric patients monitored by immunoassays on
the Abbott ARCHITECT ci4100 in healthcare institutions worldwide. It is however
important that these reference intervals be validated by each laboratory for the local
pediatric population as recommended by CLSI.

Conclusions: In contrast to previously published methods, our assay can rapidly detect
all analytes of clinical utility for the diagnosis of primary hyperoxaluria types I-III.
Patients with PH1 and PH2 do not have significantly increased excretion of HOG
as compared to unaffected controls. Analysis of urine specimens from patients with
hyperoxaluria of unknown etiology can potentially lead to a diagnosis of PH3.

S210

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Point-of-Care Testing

Wednesday, July 30, 9:30 am 5:00 pm

Diagnostics, Indianapolis, IN) ACCU-CHEK Inform I (Inform-I) blood glucose
meter uses GDH-PQQ methodology. A mutant quinone GDH is used in ACCUCHEK Inform II (Inform-II) which can distinguish glucose from maltose. In this
study, we want to determine if maltose interference has been eliminated with InformII using samples containing maltose and samples from patients treated with icodextrin
peritoneal dialysis.

Wednesday, July 30, 2014

Poster Session: 9:30 AM - 5:00 PM
Point-of-Care Testing

B-276
Assessment of Harmonization among Siemens Point-of-care and Central
Laboratory Blood Gas Platforms

S. Fennell, C. Bethoney, K. LaRock, D. Poirier. Siemens Healthcare DX,

Norwood, MA
Background: Determine correlations between Siemens point-of-care (POC) and
central lab blood gas platforms in order to demonstrate harmonization across the
product portfolio.
Relevance: AACCs International Consortium for Harmonization of Clinical
Laboratory Results has been working with a variety of stakeholders regarding
harmonization among results from different methods and labs for the same measurand.
[1] Malone states, Harmonization means achieving comparable results among
different measurement procedures. Further, When lab measurement procedures
give different results for the same specimen, patients may get the wrong treatment,
because decision criteria are not appropriate for the procedure in use. In order to do
this effectively, results need to be harmonized.
Methods: Method comparison studies were performed with whole blood among POC
(RAPIDPoint) and central lab (RAPIDLab) blood-gas platforms in accordance with
the CLSI EP-09 guideline. Correlation statistics including Deming slopes, intercepts,
and coefficients of determination (r2) were generated for the following comparisons:
RAPIDLab 1265 Blood Gas System vs. RAPIDPoint 500 Blood Gas System
RAPIDLab 348EX Blood Gas System vs. RAPIDPoint 500 Blood Gas System

Methods: Samples containing 240 mg/dL, 360 mg/dL, and 720 mg/dL of maltose
were prepared by spiking a whole blood sample with maltose. At each level of maltose,
a control sample was prepared by adding equal amount of water to an aliquot of the
blood sample. Glucose results in these samples were measured in triplicates with both
Inform-I and Inform-II. Previously frozen plasma samples from three patients who
underwent icodextrin peritoneal dialysis were also tested with Inform-I, Inform-II,
and results compared with that obtained with Beckman Olympus AU5400 (AU5400)
which is free from maltose interference.
Results: Glucose results in samples containing different amounts of maltose and
samples from the three patients obtained with Inform-I, Inform-II, and AU5400 are
shown in the table below:
Falsely
Increased
Falsely Increased
Glucose
Glucose with
with
Inform-I (mg/dL)
InformII(mg/dL)
240 (mg/dL)
114
241
151
10
360 (mg/dL)
313
121
225
16
720 (mg/dL)
563
128
491
34
Patient 1
217
327
238
110
21
Patient 2
196
245
218
49
22
Patient 3
276
286
289
10
13
Conclusion: Both the Inform-I and Inform-II exhibited maltose interference which
increases with maltose concentration. However, significant reduction of maltose
interference was observed with Inform-II. The increases in the glucose results
obtained with Inform-II in samples of patients who underwent icodextrin peritoneal
dialysis were minimal, therefore, may not change the clinical decisions to manage
blood glucose levels in these patients.
Sample
Containing
Maltose

Glucose
with
AU5400
(mg/dL)

Glucose
with
Inform-I
(mg/dL)

Glucose
with
Inform-II
(mg/dL)

RAPIDPoint 405 Blood Gas System vs. RAPIDPoint 500 Blood Gas System
Results: Deming regression statistics for each comparison across intervals for each
measurand are shown in Table 1. The slopes for each measurand fell between 0.96
and 1.25, with r2 0.9829.
Conclusion: Harmonization at medical decision levels and average concentrations
was demonstrated among Siemens POC and central lab blood gas platforms for the
measurands evaluated.
[1] Malone B. AACCs Thought Leadership Series: Why Harmonization Matters.

B-279
Magnetic Immunoassay for Quantitative Point-of-Care Tests and Rapid
Measuring of Protein Concentration

P. I. Nikitin1, A. V. Orlov1, V. A. Bragina1, M. P. Nikitin2, T. I. Ksenevich1,

B. G. Gorshkov1. 1General Physics Institute, Russian Academy of Sciences,
Moscow, Russian Federation, 2General Physics Institute and Institute of
Bioorganic Chemistry, Russian Academy of Sciences, Moscow, Russian
Federation
Background: Currently, protein markers of diseases are widely detected by rapid
Lateral Flow (LF) strips based on color or fluorescent labels. The measurements are
carried out by recording such labels from the thin surface layer of the LF membranes
only. The advantages of employment of magnetic beads (MB) as labels in bioassays
are well established as the beads are not affected by color of the samples, reagent
chemistry or photobleaching, are highly stable and could be counted over the whole
volume of solid phase. MB can be used in quantitative immunochromatographic
assays to facilitate rapid diagnostics of diseases and monitor therapy efficiency.
In the present work, a rapid quantitative MB-based assay has been developed and
demonstrated by measuring in human blood of concentration of the tumor marker of
prostate specific antigen (PSA) used as a model. Such highly sensitive registration in
wide dynamic range of concentration of PSA and other tumor markers is attractive for
disease diagnostics and relapse monitoring after surgical removal of tumors.

B-277
Investigation of Maltose Interference on the Roche ACCU-CHEK Inform II
Blood Glucose Meter

L. Song, D. Stene, L. Boyd. David Geffen School of Medicine at UCLA,

Los Angeles, CA
Background: Maltose can be present in the blood of patients who were treated with
peritoneal dialysis using icodextrin or maltose-containing immune globulin up to 2
weeks after the treatment. Maltose can interfere with glucose measurement using
glucose dehydrogenase pyrroloquinolinequinone (GDH-PQQ) and cause falsely
elevated glucose results. Patients can develop severe hypoglycemia if treated with
insulin in response to these falsely elevated glucose results. The Roche (Roche

Methods: The magnetic nanolabels were recorded over the whole volume of test zone
on LF strips using original method of non-linear MB remagnetization and frequency
miing (P.Nikitin & P.Vetoshko, EP 1262766, 2001). Recently, this method was
successfully used for toxin detection in complex biological media by MB counting
at 3D-filter solid phase (A.Orlov et al. Anal. Chem. 2013, 85, 1154-1163). Direct
comparison showed that the sensitivity of the electronic detection method is on the
level of the gamma-radioactive technique for counting of MP based on the 59-Fe
isotope (M.Nikitin et. al. J. Appl. Phys. 2008, 103, 07A30). Thus, MBs combined with
highly sensitive detectors allow realizing many advantages of old radioimmunoassays,
but in much more safe and affordable ways. In present work such advantages have
been demonstrated by magnetic LF strips based on dry chemistry.
Results: It has been shown with blood samples of 25 patients that the limit of
quantitative PSA detection computed in compliance with IFCC/CLSI guidelines
for quantitative methods is 25 pg/mL over the wide dynamic range exceeding 3

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S211

Point-of-Care Testing

Wednesday, July 30, 9:30 am 5:00 pm

orders of concentration magnitude. CV was less than 10% at low concentrations.
Importantly, the developed dry chemistry assay features simplicity and 4 times
better LOD at duration of 20 min as compared with several-hour-long commercially
available ELISA kits. It has been shown that because of high sensitivity, quantitative
registration, simplicity and short duration, the developed POC method combines the
advantages of laboratory methods and rapid tests based on dry chemistry.
Conclusion: The replacement of traditional optical (gold, colored latex, etc.) labels in
immunochromatographic assay formats by the magnetic beads combined with highly
sensitive MB detection over the whole volume of the LF strips has allowed to develop
a quantitative and highly sensitive in wide dynamic range of more than 3 orders of
concentration magnitude immunoassay. The LOD of 25 pg/mL demonstrated by the
magnetic immunochromatographic assay while detection of tumor marker PSA in
human blood allows us to consider it as an attractive diagnostic POC platform for
highly sensitive quantitative detection of proteins in biological fluids.

B-280
Development of a new rapid point-of-care assay for quantitative measurement
of D-Dimer in whole blood and plasma

S. Yokokawa, M. Morita, K. Kohno, M. Yamamoto, H. Yago. SEKISUI

MEDICAL CO., LTD., Ryugasaki, Ibaraki Pref., Japan
D-dimer is a fibrin degradation product (FDP) composed of cross-linked fibrin
degraded by plasmin and well known as one of the markers for thrombotic disorders
such as deep vein thrombosis (DVT), pulmonary embolism (PE), disseminated
intravascular coagulation (DIC) and for coronary artery diseases. We have developed
a new rapid and quantitative assay for D-dimer in whole blood and plasma. This assay
is based on lateral flow immunochromatography with colloidal gold and employs
two different anti-human D-dimer mouse monoclonal antibodies. The test cartridge
is inserted into the immunochromato-reader RAPID PIA (Sekisui Medical Co.,
Ltd.), and a sample (120L) of whole blood or plasma is added to the well of the
cartridge. After 10 minutes, the reader automatically measures the density of colloidal
gold captured at the test line by the antigen-antibody sandwich reaction. The lower
detection limit for D-dimer was 0.2g/mL, and the upper quantitation limit was 15 g/
mL. No prozone effect was observed in D-dimer samples of concentrations from 15
through 244 g/mL. The within-run C.V. (n=5) at 0.55 g/mL, 3.5 g/mL, and 7.1 g/
mL was 1.8%, 5.1% and 4.9%, respectively. The between-run C.V. (n=5) at 0.55 g/
mL, 3.5 g/mL and 7.1 g/mL was 6.3%, 4.5% and 4.7%, respectively. The method
comparison with the approved IVD reagent, the principle of which is latex-enhanced
immunoturbidimetry, yielded a correlation coefficient of 0.992 and an equation of Y
(present method) = 1.01X - 0.17 (n = 50 citrated plasma specimens). Furthermore,
a high level of correlation was observed between citrated plasma and whole blood
(R:0.997 ; slope:1.03;intercept:0.00). We concluded that this newly developed
assay is accurate, precise and simple for the measurement of D-dimer in whole blood
or plasma at the bedside. We believe that this assay will be a useful tool for rapid
screening patients with thrombotic disorders and coronary artery diseases.

B-281
Performance of the Nova StatStrip Glucometer in a Pediatric Hypoglycemic
Population

C. Tan, K. Walker, B. L. Woolsey, P. S. Thornton, N. Farrell, V. LeungPineda. Cook Childrens Medical Center, Fort Worth, TX
Background: In May of 2013, our hospital replaced the legacy glucometer with the
Nova StatStrip. Our Department of Endocrinology has a fasting study protocol where
patients are fasted until their blood glucose is below 50 mg/dL. At this point a variety
of critical samples are drawn. However, due to the accuracy limitations of our legacy
glucometer, the current protocol needs a confirmation of blood glucose concentrations
from the core laboratory. Therefore, we sought to determine the performance of the
Nova StatStrip in this pediatric hypoglycemic population. If the accuracy of the
glucometer proved acceptable it could lead to a modification in the fasting study
protocols that could lead to a decrease in the length of the study.
Objective: Our goal was to determine the accuracy of the Nova StatStrip glucometer
in measuring samples close to the 50 mg/dL range in real clinical setting conditions.
Methodology: Precision of the Nova StatStrip glucometer at low glucose values was
evaluated at four different values. Furthermore, quality control material and patient
samples were evaluated in the Nova StatStrip, and compared to readings from our core
laboratory analyzer without sample delay.

statistical difference between POC and core lab methods (n=23). Patient samples
(n=30) co-relations indicated an average negative bias of 2 mg/dL when the Nova
glucometer was compared to our central lab method. In addition, we saw a decrease
in the rate of patient ID errors due to the use of a 2D barscan system present in the
Nova StatStrip but absent from our legacy glucometer. The Nova StatStrip proved to
have acceptable accuracy in measuring hypoglycemic samples when compared to our
central laboratory method. These results support the use of POC glucose, in lieu of
waiting for glucose results from the core laboratory, for activating our fasting study
draws, shortening the length of the procedure for our patients.

B-282
Analytical and technical aspects of POCT-Troponin in the Emergency
Department: Comparison with central laboratory hsTnT

J. von Recum1, D. Meyer zum Bschenfelde2, A. Slagman1, J. Searle1, F.

Holert1, C. Mller1, M. Mckel1. 1Charit - University Medicine Berlin,
Berlin, Germany, 2Labor Berlin Charit Vivantes GmbH, Berlin, Germany
Introduction The use of point-of-care-testing (POCT) for cardiac troponin involves
discrepancies in comparison to central laboratory based measurements. In this
analysis, troponin-discrepancies are documented and analyzed to evaluate possible
reasons and generate data that is likely to help eliminating factors of insecurity in
association with technical and handling issues of different platforms and assays used.
Methods Cardiac troponin-T on AQT-90 (EDTA whole blood, Radiometer) POCT
is the standard troponin test in our Emergency Department. We set up parallel
measurements of hsTnT (heparin-plasma, Cobas602, Roche) in the central laboratory
from the same blood draw and Troponin-I (EDTA whole blood, AQT90, Radiometer)
in two separate timeframes of 3 and 4 months within one year (winter and summer
period). Troponin-discrepancies were defined as outlined below (Table 1). In the
first timeframe immediate re-measurement of the POCT-sample was performed.
Measurements of admission-samples resulting from 4946 patients were analyzed
regarding discrepancies between hsTnT/TnT and TnT/TnI.
Results 183 discrepancies were detected resulting from 164 patients. 37 between
TnT and hsTnT and 146 between TnT and TnI (Table 1). 19 patients showed two
discrepancies (hsTnT/TnT and TnT/TnI). Characterizing the discrepancies of
timeframe 1 (n=20) we found 10 discrepancies in more than one measurement and 10
non-reproducible analytical errors. In three patients AQT TnT was constantly elevated
while hsTnT and AQT TnI were negative. In two patients hsTnT was elevated while
AQT TnT was negative. The latter were a 94years old female with a fall and a 81years
old male with acute kidney failure.
Conclusion In our population the percentage of discrepancies between hsTnT and
AQT TnT did not exceed 0.75% of nearly 5000 analyzed individuals. The use of AQT
TnT at the POC is technically reliable under real life conditions. Further studies need
to clarify the diagnostic accuracy, specifically when lower cutoffs are used for hsTnT.
Definition of discrepancies
TnT (AQT) vs. hsTnT (Cobas)
TnT (AQT) vs. TnI (AQT)
hsTnT < 45
hsTnT > 55
TnI < 23 ng/L TnI > 46 ng/L
ng/L
ng/L
TnT < 27
TnT <
ok
discrepancy
ok
discrepancy
ng/L
17 ng/L
TnT > 33
TnT >
discrepancy ok
discrepancy
ok
ng/L
33 ng/L
Discrepancy Details in % (n=4946)
hsTnT < TnT hsTnT > TnT
TnT > TnI
TnT < TnI
Timeframe 1
0.73% (n=17) 0.13% (n=3)
3.5% (n=82) 0.17% (n=4)
(n=2344)
Timeframe 2
0.38% (n=10) 0.27% (n=7)
1.88% (n=49) 0.42% (n=11)
(n=2602)
total
0.55% (n=27) 0.2% (n=10)
2.64% (n=131) 0.3% (n=15)
(n=4946)
Characterization of the TnT discrepancies of timeframe 1 (n=2344)
hsTnT <
hsTnT < TnT hsTnT > TnT
TnT
persistent analytical
discrepancies, to be
0.13%
persistent
0.34% (n=8) 0.09% (n=2)
investigated for possible (n=3)
interferences
non0.38% (n=9) 0.04% (n=1)
persistent
Table Definitions and details of Troponin discrepancies.

Results & Conclusions: Coefficient of Variations of the Nova StatStrip at low

glucose levels was less than 4%. Comparing low glucose QC material showed no

S212

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Point-of-Care Testing

Wednesday, July 30, 9:30 am 5:00 pm

Greater physician and patient satisfaction

B-283

Improved relations between the lab and the ED, as they now work in unison

The modified method for microbilirubin determinations at low source settinghospitals

K. Thongsukkaeng1, P. Pheunsongkram1, U. Doungwao1, W. B.

Treebuphachatsakul2. 1Phukhieo Hospital, Chaiyaphum Province,
Thailand, 2Department of Medical Technology, Faculty of Allied Health
Sciences, Naresuan University, Phitsanulok, Thailand
Background: Bilirubin determinations are often required in the routine management
of newborns. Bilirubinometer requires a micro volume of blood sample and is
convenient used at pediatric units as point of care testing. There is no bilirubinometer
available at low source setting hospitals in rural area of Thailand. The objective of
this study were to evaluated the performances of bilirubinometer and validate the
performances of the modify method using a micro volume of plasma sample for
bilirubin measurement in newborns.
Methods: Precision, accuracy, and linearity of bilirubinometer were evaluated.
One hundred plasma samples of newborns at Phukhieo Hospital were determined
for bilirubin. Sixty microliters of each blood sample was collected into heparincontaining capillary tube. All samples were immediately centrifuged, then measured
for bilirubin by bilirubinometer and a modified method by the automated clinical
chemistry analyzer. Paired data of bilirubin results were analyzed by using paired
different T-test.
Results: Bilirubinometer has revealed good precision, accuracy, and linearity
for bilirubin determination in newborns. Microbilirubin results obtained from
bilirubinometer and a modified method were correlated (r=0.978) and paired difference
data between two methods were not statistical significant differences (p>0.05).
Conclusion: Bilirubinometer was convenient used for microbilirubin in newborns
on site and a modified method was the alternative way for low source setting in
Thai Hospitals at clinical laboratory. However, laboratory should manage to provide
shorten laboratory turnaround time for microbilirubin measurement.

B-284
Implementing a point of care (POC) laboratory in order to reduce patients
length of stay in the ED as well as meet ED critical care standards.

T. Nolen1, H. Hasebe2. 1Bert Fish Medical Center, New Smyrna Beach, FL,
2
Mitsubishi Chemical USA Inc., Chesapeake, VA
Background: Hospitals, providers and patients are all eager to reduce the length
of emergency department (ED) visits. For hospitals, faster patient turnover means
more patients can be seen, generating higher revenue. For physicians, faster turnover
means releasing patients who dont need further care so they can spend more time
with those who do. For patients, less time in the ED means they can return to daily life
sooner and, potentially, reduce their medical costs. Speeding up the rate of turnover
is dependent on reducing turnaround time (TAT) for lab results. We were interested
to determine whether point of care (POC) laboratory testing, based in the ED itself,
could lower TAT when compared to the performance of a satellite laboratory. To test
this theory we set up a POC lab in the Bert Fish Medical Centers ED.
Methods: Although our POC lab operated within the ED, it was controlled by the
main laboratory. Four emergency medical technicians, employed by the ED, managed
all the testing, under the supervision of one medical technologist, a lab employee.
A key difference about our approach to POC testing is that we did not use nurses to
perform any testing procedures. Based on our research, nurses are already too busy to
take on the additional burden of managing POC testing.
Results: In our POC, we used a more sensitive troponin assay with chemiluminecent
POC testing helped improve TAT. With this test, the POC lab was 42% faster than the
main lab when troponin test results were negative_54 minutes in the POC lab versus
97 minutes in the main lab. The impact was significant, since negative results made up
91% of the total. The remaining 9% positive results were retested in the main lab to
exclude other cardiac conditions, so there was no time savings for this batch. In effect,
testing at the POC level becomes a lean process because it cuts out the extra steps
required for testing in the main lab.
Conclusion: Our study demonstrated that establishing a POC lab can yield a number
of benefits. These include:
Faster TAT for troponin testing, as well as testing for myoglobin and CK-MB
Shorter ED stays, and as a result, shorter wait times for new patients in need
Earlier detection of cardiac risk, due to the faster detection of troponin, leading to
increased survivability

We also noted that the cost of running a POC lab is about $300,000 per year, versus
$1-1.5 million for a satellite lab. In part, this is because the salaries of the medical
assistants running the POC lab are lower than those of medical technologists in
satellite labs. Even with the hiring of additional assistants, overall operating costs of
the POC labs are lower.
Our conclusion is that given how the POC can perform more quickly and at lower cost
than traditional satellite labs, hospitals should consider moving to the POC model.

B-286
Development of a fully-quantitative lateral flow assay system for the detection
of a novel combination of sepsis markers

E. E. Oldridge1, E. D. Carrol2, T. Myers3, N. Tort4, J. Flint1, C. Danks1.

1
Forsite Diagnostics Ltd, York, United Kingdom, 2University of Liverpool,
Liverpool, United Kingdom, 3Microlab Devices Ltd, Leeds, United
Kingdom, 4Biokit, Barcelona, Spain
Sepsis, or severe bacterial infection (SBI), is the leading cause of death in intensive
care units in high-income countries, and its incidence is on the rise. Sepsis is
estimated to affect 18 million people worldwide and in 2003 $14.6 billion were spent
on hospitalizations for the disease in the US. Sepsis diagnosis is often delayed due
to inadequate diagnostic tools. Consequently, prompt diagnosis and treatment is of
paramount importance to reduce morbidity and mortality associated with the disease.
We have developed an innovative point-of-care (POC) diagnostics system
incorporating lateral flow assay technology to measure a novel combination of biomarkers: procalcitonin (PCT), neutrophil gelatinase-associated lipocalin (NGAL) and
resistin. In combination, these markers provide a superior indication of SBI in febrile
children presenting to the emergency department.
PCT, a precursor of calcitonin, is a 116 amino acid protein that has been proposed as
a marker of disease severity in conditions such as septicemia, meningitis, pneumonia,
urinary tract infection (UTI) and fungal and parasitic infection. NGAL, or lipocalin 2,
is a 25 kD lipocalin that has a role in innate immunity and is highly up-regulated by
inflammatory stimuli. Resistin has been shown to play an important regulatory role in
adipogenesis, glucose homeostasis and insulin sensitivity. PCT is a well recognized
marker of sepsis, and Resistin and NGAL have recently been shown to be significantly
elevated in intensive care patients with sepsis.
The POC system incorporates 3 simplex lateral flow assays utilizing gold nanoparticle technology to detect PCT, NGAL and Resistin. It requires no sample pretreatment and gives fully-quantitative results within 10 minutes on a hand-held reader.
We show that the system is able to measure the bio-marker combination from human
plasma, producing fully-quantitative values which are incorporated into a clinical
diagnostic algorithm to assess response to therapy. The results show that the linear
ranges of the individual assays span a large concentration range: PCT: 0.1-1000 ng/
ml, Resistin: 5-1000 ng/ml and NGAL: 5-1000 ng/ml. We also discuss the ability
of the assays to successfully differentiate between 5 clinical categories on which to
base the clinical diagnosis and treatment: SBI rule out, SBI rule in, uncertain (further
clinical assessment and investigations necessary), likely severe infection and likely
life threatening.
This test has the potential to advance the effective use of bio-markers in the
management of both child and adult sepsis and impact across the whole POC market
by providing a better diagnosis, rapid treatment and reduced hospital admissions.

B-287
Impact of improved glucose monitoring in the neonatal intensive care unit:
an evaluation of analytical and clinical performance of the point of care Nova
Statstrip

J. Raizman1, T. Henderson1, J. Shea2, S. Silverman1, S. Redmond1, A.

Moore1, J. Dubois3, K. Adeli1. 1The Hospital for Sick Children, Toronto,
ON, Canada, 2Saint John Regional Hospital, Saint John, NB, Canada,
3
Nova Biomedical, Waltham, MA
Objective:To evaluate the analytical and clinical performance of new glucose point of
care testing (POCT) in the intensive care unit (NICU).
Methods:Within-run imprecision, correlation with a plasma hexokinase assay, and
interferences with hematocrit were studied on the Nova-StatStrip and SureStep-Flexx
meters. Meter and lab results were analyzed over 2 years in the NICU. Outcomes

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S213

Point-of-Care Testing

Wednesday, July 30, 9:30 am 5:00 pm

measured were rate of hypoglycemia, frequency of critical results, average length of
stay (LOS), clinical sensitivity/specificity for detecting critical results, and clinical
accuracy.
Results:Imprecision (CV) using control material (2.8-15.7 mmol/L, n=20) and whole
blood patient pool specimens (5.1-23.7 mmol/L, n=20) ranged from 3.19-4.97% and
1.39-7.21%, respectively, for SureStep, and 1.05-4.98% and 1.51-3.71%, respectively,
for Nova. Method comparison to the Ortho Vitro950 hexokinase method (n=120)
revealed correlations of y= 0.867x + 0.82 (R=0.990) with a mean bias of -0.64 mmol/L
for SureStep, and y=1.016x+0.04 (R=0.997) with a mean bias of -0.21 mmol/L for
Nova. Hematocrit interfered with SureStep by reducing glucose measurements at
increasing hematocrit (23-67%) compared to Ortho, an effect absent on Nova. Studies
to assess clinical performance of the meters revealed fewer readings per NICU visit
(24%, p=0.001) in patients monitored with Nova compared to SureStep. This was
associated with a reduction in frequency of hypoglycemia results (53%, p=0.053) as
well as a trend towards critically low 3.0 mmol/L (35%, p=0.112) and high results
9.0 mmol/L (40%, p=0.009). The sensitivity/specificity for detecting critically low
results was 70.2/98.7% and 80/99.5% for SureStep and Nova, respectively. Patients
monitored with Nova trended towards increased LOS in the NICU (16%, p=0.181).
Clinical comparisons demonstrated correlations of y=0.983x+0.24 (R=0.931) for
Nova (n=607) and y=1.1x-0.036 (R=0.869) for SureStep (n=977) compared to lab
values.
Conclusion:Nova demonstrated superior precision and accuracy compared to
SureStep. Improved analytical performance translated into improved detection of
critical glucose results, demonstrating the importance of implementation of accurate
POCT in the NICU.
Clinical performance of the Nova StaStrip and SureStep Flexx in the NICU
Nova StatStrip SureStep Flexx % Change p-value
240
250
24
10
13.1
24
0.001

Total number of admissions

Total results per admission
Hypoglycemia results per
admission
Critical low results per
admission
Mean length of stay in NICU
(days/patient)
Sensitivity for critical low
results (%)
Specificity for critical low
results (%)

0.16

0.34

53

0.053

0.68

1.04

35

0.112

25.1

21.1

16

0.181

80

70.2

99.5

98.7

B-288
Evaluation of new glucometers ( Easy Touch GC) for bedside use

M. E. Adekiitan, G. E. Imana, O. O. Adedeji. Lagos State University

Teaching Hospital, Lagos, Nigeria
BACKGROUND Glucometers have greatly improved the clinical care of diabetics.
It shortens time for making critical decisions. It is portable, inexpensive and easy
to use. Newly acquired glucometers (Easy Touch) in our hospital were evaluated
by calibration and precision profile determination. The results were compared with
values obtained using routine laboratory method.
METHOD Standard glucose solution 500mg/ dl was serially diluted in de-ionized
water to 250mg /dl, 125mg/dl, and 100mg/dl. The solutions were analyzed with three
glucometers to determine their linearity. .
Blood samples were taken in duplicate from 39 patients into fluoride oxalate
containers for the determination of glucose concentrations. A set was analyzed in the
Hospital Laboratory by spectrophotometric glucose oxidase method while the other
was analyzed with the glucometers. Duplicate measurements of the blood samples
by 3 individuals using the glucometers were performed to determine their precision
profile.
RESULTS The glucometers showed good linearity using the standard glucose
solutions. Their correlation coefficients were R1 = 0.883, R2 =0.983, R3 =0.949; R1,
R2 and R3 represent the 3 glucometers respectively. However, the readings from the
glucometers were much lower than the actual glucose concentration at 500mg/dl.
This may be due to the use of de-ionized water as solvent for the standard solutions
used in the calibration as against the matrix of blood. This could imply that the
glucometer readings may not be accurate at critically high glucose levels.
Good correlations were obtained between the readings of the glucometers and
laboratory results at blood glucose concentration <500mg/dl. Thus, the correlation
coefficients were 0.824, 0.900 and 0.845 for glucometers R1, R2 and R3, respectively.
The coefficient of variations (CVs) obtained with the glucometers were 10.43%,
3.75% and 6.50% for R1, R2 and R3, respectively. With the routine laboratory

S214

method, the CVs were 10.94%, 9.84% and 11.69%. When the two methods were
compared, the CVs obtained with the glucometers varied widely, whereas, the CVs
of the laboratory method were fairly constant. This could imply that the performance
of the glucometers was operator dependent because they were operated by 3 different
individuals, which was not the case with the laboratory method that was performed
by only one person.
The imprecision profile, representing, the mean differences between the readings of
the glucometers and laboratory results, were 14.8 4.67, 23.7 6.08 and 15.5 3.83
for glucometers R1, R2 and R3 respectively. These values were high and could have
serious implications in the interpretation of glucose values for the management of
diabetes mellitus.
CONCLUSION The readings from the glucometers showed inaccuracy at very high
glucose concentrations and their performance could be operator dependent. In view
of the good correlations with the laboratory method (at <500mg/dl), the glucometers
should be standardized against the laboratory method regularly when used.
Keywords: Glucometer, calibration, precision.

B-289
Comparison of the UriScan 2 ACR regent dipsticks for Albumin and
Creatinine in Urine with Quantitative Methods

S. Lee, S. Park, M. Kang. YD Diagnostice, YonginSi, KyungiDo, Korea,

Republic of
Background: Microalbuminuria is a predictive maker for renal disease and the
identification of patients at high risk of developing complications of diabetes or
hypertension. The UriScan 2ACR(YD Diagnostics, Korea) is a urine chemistry
point-of-care test for the semi-quantitative measurement of albumin and creatinine and
calculation of albumin:creatinine ratio(ACR). The aim of this study is to comparing
quantitative method and UriScan 2ACR strip for measurement of albumin and
creatinine ratio in urine.
Methods: The samples for this study used to random urine which were collected from
a total 641 patients at three sites. The concentration of albumin and creatinine are
measured to UriScan 2ACR after quantifying by gold standard quantitative method.
Results: The clinical performance results of the UriScan 2ACR strip s for detecting
microalbuminuria (ACR) were showed to accuracy 86.9%. Also, the sensitivity,
specificity, PPV (positive predictive value), and NPV (negative predictive value)
is 89.2%, 85.6%, 77.4%, and 93.5%. The result of microalbumin measurements in
the spot urine samples of UriScan 2ACR strip were showed to accuracy 89.9%.
Also, the sensitivity, specificity, PPV (positive predictive value), and NPV (negative
predictive value) is 91.2%, 89.0%, 84.5%, and 93.9%, respectively. The results of
creatinine were accuracy 70.0%.
Conclusion: UriScan 2ACR strip tests had good agreement with gold standard
quantitative methods. Therefore, measurement of UriScan 2ACR strip in the
spot urine sample is an efficient method for screening the general population for
microalbuminuria. The dipsticks tests were easy to use, simple, and useful screening
tool for point-of-care test.

B-290
Validation of the Abbott i-STAT total -hCG cartridge for use in rural Alberta
hospitals

A. K. Fzry1, L. Redman2, S. M. H. Sadrzadeh2, J. C. Wesenberg3, A. A.

Venner3. 1Alberta Health Services, Edmonton, AB, Canada, 2Calgary Lab
Services, Calgary, AB, Canada, 3Alberta Health Services, Red Deer, AB,
Canada
Background: Human chorionic gonadotropin (hCG) has significant clinical utility,
yet many rural hospitals lack testing volume and/or instrumentation for quantitative
measurement. However, many of these hospitals already have i-STAT analyzers, and
Abbots new -hCG cartridge offers a feasible option for onsite quantitative testing.
This study aimed to evaluate the i-STAT -hCG cartridge.
Methods: Linearity, imprecision and accuracy were evaluated. For linearity, a plasma
sample (hCG=1799 IU/L) was diluted from 1/2 to 1/256 and measured in duplicate.
For imprecision, two levels of each of Clinica (24 IU/L and 1455 IU/L) and Bio-Rad
Immunoassay Plus (6 IU/L and 21 IU/L) quality controls were measured daily over
20 days. Accuracy was assessed by measuring hCG in plasma samples on both the
i-STAT and the Siemens Dimension Vista (n=37), Beckman Coulter DxI 800 (n=39),
or Roche Cobas 6000 (n=42) analyzers. In addition, whole blood samples were

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Point-of-Care Testing

Wednesday, July 30, 9:30 am 5:00 pm

measured with the i-STAT; samples were then centrifuged and the plasma measured
with central lab analyzers (n=43, n=34, n=52, respectively). The -hCG ranged from
<1 IU/L to 161,207 IU/L.
Results: Linearity was demonstrated from 9-1799 IU/L. Total imprecision was
acceptable (Bio-Rad: mean=24 IU/L, CV=7.7%; mean=21 IU/L, CV=4.4%; Clinica:
mean=24 IU/L, CV=5.1%; mean=1455 IU/L, CV=3.0%). Comparison of plasma on
the i-STAT yielded acceptable correlations (Vista: regression line with slope=1.1105,
y-intercept=9 IU/L, R2=0.9849; DxI slope=0.9727, y-intercept=3 IU/L, R2=0.9943;
Cobas slope=1.0043, y-intercept=4 IU/L, R2=0.9997). Comparison of whole blood on
the i-STAT gave similar results (figure 1).
Conclusion: Linearity, imprecision and accuracy of the i-STAT -hCG cartridge were
acceptable for both whole blood and plasma samples. It can be utilized in clinical
settings without access to a large chemistry analyzer with quantitative hCG. It is
specifically useful for patients requiring a stat quantitative hCG result.

B-292
Novel POC analysis for CBC, using PixCell Medical device, HemoScreen

A. Bransky1, Y. Ben-Yosef1, B. Marom1, H. Ben-Asher1, C. DSouza2.

1
PixCell Medical Technologies Ltd., Yokneam Ilit, Israel, 2University of
Westminster, Faculty of Science & Technology, Department of Biomedical
Sciences, London, United Kingdom
Background: Point of care (POC) in the community and hospital out patients is
increasing to enable rapid sample testing for anaemia, monitoring oncology treatment
and reduce patient waiting times. Current devices provide single parameters within
a Complete Blood Count (CBC). However, the HemoScreen developed by PixCell
Medical will provide a CBC and five-part differential within minutes.
Objective: To present a novel device, HemoScreen, for performing CBC using
viscoelastic-focusing and microfluidic technology linked with image-based analysis
and to demonstrate its ease of use and reliability for use in POC settings.
Methodology: HemoScreen, uses a disposable cartridge with 20ul of blood, finger
prick or venous sample, placed into an analytical device. Employing viscoelasticfocusing and flow-based optical imaging, the cells are focused into a single plane,
analysed by image-processing and classification algorithms to calculate the cell counts
and related red cell indicies. The HemoScreen will offer a safe, reliable maintenance
free method for performing CBC including possible flagging for morphological
abnormalities. Internal quality control includes a built-in self-test for electronics,
mechanics and software, carried out before each test, and at regular intervals during
device operation. EQA is undertaken using a commercial control set at three different
levels, low, normal and high.
Currently, 3 samples with 10 replicates each have been processed for precision.
Approximately 60 samples were tested for accuracy, following the CLSI standard
criteria. The Sysmex XE-2100 served as a reference method.
Results and validation: HemoScreen precision and accuracy results are summarized
in the table below:

B-291
The Effect of Maltose on the Radiometer 837 Blood gas Analyzers

D. Semmel, G. Haddad, S. Fan. Boston Medical Center, Boston, MA

Background: Glucose measurement can be performed on commercially available
blood gas analyzers (BGAs) for faster turnaround time, which are based traditionally
on either the glucose dehydrogenase reaction or the glucose oxidase reaction.
Other sugars, including maltose, may interfere with these reactions, causing the
glucose measurement to be falsely elevated. Octagam 5% liquid is a commercially
available intravenous immune globulin. Its preparation contains maltose, which has
the potential to interfere with glucose measurement on the BGAs. Patients treated
with this drug who also necessitate blood glucose levels may demonstrate falsely
elevated measurements if their testing is performed on the BGAs. As it is not feasible
to screen every patient for interfering drug preparations before performing glucose
measurements, we felt compelled to demonstrate whether the maltoses effect on
glucose measurement was clinically significant when using the BGAs.
Methods: To simulate patients being treated with Octagam, two levels of clinically
relevant maltose concentrations were prepared using maltose solution admixed with
whole blood from patients known not to be on any medication produced with maltose
or other sugar based preparations. There were a total of eleven specimens, whose
glucose levels ranged from 25 to 550 mg/dL. Simulated high and low drug levels
were prepared by spiking concentrated maltose into whole blood specimens, resulting
in a high level of 6.4 g/L to simulate maximum dosage and low level of 1.6 g/L
to simulate low dosage. The spiked samples were tested for glucose in duplicates on
the Radiometer 837 BGAs. Results were then corrected for the spiked volume, and
averages of the duplicates were plotted against the original glucose concentrations.
Results: The correlation coefficient (R^2) of the best fitting line for both high and
low maltose levels were good. Additionally, the predicted glucose measurements
at various levels can be extrapolated from the best fitting line. As seen by this best
fitting line, there is no clinically significant difference between the original and spiked
measurements, either in the high level or low level specimens.
Conclusions: There is no clinically significant difference between the original and
spiked results. Therefore, there is no need to eliminate the use of the Radiometer
BGAs as a laboratory instrument for fast and accurate glucose measurements in
urgent specimens for patients undergoing treatment with maltose-containing drug
preparations or to screen patients for use of such drug preparations.

Precision
Accuracy
N=30
Parameter
CV (%) Correlation coefficient (r)
WBC (x109/L) 8.0
0.988
RBC (x1012/L) 3.6
0.958
HGB (g/dL)
6.1
0.949
MCH (pg)
7.2
0.809
HCT (%)
4.0
0.963
MCV (fl)
1.0
0.934
RDW (%)
1.3
0.957
9
PLT (x10 /L) 5.5
0.960
Additional studies are being performed as it has been
further improved by optimizing the current method.

Slope Intercept
0.979 0.26
0.966 0.167
1.013 -0.125
0.916 1.91
1.003 -0.048
0.976 1.691
0.894 1.552
0.948 6.7
shown that precision can be

Conclusion: The HemoScreen is an innovative device that has potential to deliver

a CBC within minutes, for use in POC settings, pharmacies, physician offices,
oncology, neonatal and adult ICU, or even in the home.

B-293
Novel POC analysis for Leukocytes and five-part differential, using PixCell
Medical device, HemoScreen

C. DSouza1, A. Bransky2, B. Marom2, H. Ben-Asher2, Y. Ben-Yosef2.

1
University of Westminster, Faculty of Science & Technology, Department
of Biomedical Sciences, London, United Kingdom, 2PixCell Medical
Technologies Ltd., Yokneam Ilit, Israel
Background: Point of care (POC) is increasing to enable rapid testing while reducing
patient waiting times. Current POC devices provide single parameters within a
Complete Blood Count (CBC). However, the HemoScreen developed by PixCell
Medical will provide a CBC and five-part differential within minutes. Haemoglobin
and neutrophil counts in particular are essential for the monitoring of cytotoxic
chemotherapy and sepsis whereas a 5-part differential is useful in all patients. The
HemoScreen will offer all the parameters at the patient bedside or doctors office that
are currently only provided by the hospital laboratory.
Objective: To introduce the HemoScreen device for Leukocytes with five-part
differential technology and demonstrate ease of use, safety and reliability of the
device for use within POC settings.
Methodology: HemoScreen uses a disposable self-contained reagent cartridge, which
employs micro-fluidic technology. The cartridge uses 20ul of capillary blood collected
directly from the finger or venous blood. Once inserted into the analyser, digital
imaging, and advanced algorithms are employed to calculate CBC results. Following

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S215

Point-of-Care Testing

Wednesday, July 30, 9:30 am 5:00 pm

RBC lysis the differential is obtained by chemical staining of leukocytes. Leukocytes
are classified based on multiple attributes such as cell size, nuclei size, nuclei lobes
and cellular content. Contamination and device maintenance are eliminated as the
flow of liquids occurs within the cartridge.
Intra-assay precision and accuracy were conducted using venous whole blood samples
analysed in accordance with CLSI standards on both the Sysmex XE-2100 (reference
method) and the HemoScreen. Three samples with 10 replicates were tested for
precision. For accuracy 32 samples have been processed. Further samples will be
used for evaluation of accuracy, including highly abnormal patient samples to validate
this device.
Results and Validation: Results of comparability study are summarized below:
Precision
N =30
CV Acceptance
(%) Criteria

Accuracy
N=32
Correlation
Parameter
Acceptance
coefficient Slope Intercept
(Units)
Criteria
(r)
WBC (x106/L) 8.0 CV <10% 0.988
0.979 0.26
r>0.95
NEUT (x106/L) 8.0 CV <10% 0.989
1.081 -0.107 r>0.95
LYMP (x106/L) 10.9 CV <15% 0.965
1.041 0.135
r>0.9
6
MONO (x10 /L) 15.5 CV <20% 0.931
0.976 0.135
r>0.8
6
EOS (x10 /L)
22.8 CV<40%
0.987
1.032 0.012
r>0.9
Calculations
were
not
done
due
to a low
BASO (x106/L) 8.8 CV<40%
count
Conclusion: A five-part differential is essential when monitoring oncology and septic
patients in a POC setting and the HemoScreen has the potential to deliver this rapidly.

B-294
Extensive Evaluation of Sample Interferences on Point-of-Care Glucose Meters

A. N. Steele, Z. Godwin, M. Howes, N. K. Tran. University of California

Davis, Davis, CA
Background: Intensive insulin therapy (IIT) guided by tight glycemic control
(TGC) reduces morbidity and mortality in critically ill patients. Accurate glucose
measurements are necessary for safe TGC. However, endogenous and exogenous
interferences, such as ascorbic acid (AA), beta-hydroxybutyrate (BHB), galactose
(GAL), glutathione (GLUT), lactose (LAC), and N-acetylcysteine (NAC) may impact
glucose monitoring systems (GMS) accuracy. These compounds may be the result
of medical interventions and/or critical illness. The objective of this study was to
determine the effect of these five interferences on current generation POC GMS
performance and the impact of autocorrecting biosensors in improving glucose
measurement accuracy.
Methods: We investigated the effects of AA, BHB, GAL, GLUT, LAC, and NAC on
the Nova Biomedical (Waltham, MA) StatStrip hospital (GMS 1-5) and Xpress meters
(GMS 6-10), Roche Diagnostics (Indianapolis, IN) Inform II meter (GMS 11-13),
and Abbott Laboratories (Abbott Park, IL) Precision Xceed Pro (GMS 14). All POC
GMSs incorporated autocorrecting features within their biosensors. Whole blood
from 12 healthy adult (age18 years) volunteers was used for testing. Specimens were
adjusted to two clinically relevant levels (moderate, high) for each interferent and
at different five glucose levels (range: 50-500 mg/dL). A negative control (without
interfering compound) was also included for each sample series. Samples were tested
on each GMS five times for each individual glucose and interference level. Results
were compared to a plasma reference. Two-way ANOVA followed by pairwise
analyses compared each GMS versus the reference method at each interference level.
Results: AA significantly affected GMS 11-13 at both levels (mean [SD] bias: -56.1
[38.9] mg/dL, P<0.001), while GMS 14 was only significantly affected at high AA
levels (-18.0 [31.2] mg/dL, P<0.001) only. BHB significantly affected GMS 3 (-30.8
[20.8] mg/dL, P<0.001) and 7 (-28.6 [19.5] mg/dL, P<0.001) at high interference
levels, GMS 4 (-28.6 [13.6] mg/dL, and P<0.001) and 8 (-30.0 [18.3] mg/dL, P<0.001)
at moderate levels. For GMS 14, BHB affected the device at both interference levels
(42.4 [37.7] mg/dL, P<0.001). GAL significantly affected GMS 1 (-29.0 [15.5] mg/
dL, P<0.001), 3 (-38.6 [24.3] mg/dL, P<0.001), 9 (-23.4 [18.8] mg/dL, P<0.001),
11-13 (81.6 [13.1] mg/dL, P<0.001), and 14 (-47.1 [35.4] mg/dL, P<0.001) at both
interference levels. GLUT significantly impacted GMS 11-13 (43.5 [27.2] mg/dL,
P<0.001). LAC significantly affected GMS 11-14 at moderate and high interference
levels. NAC significantly affected GMS 11 (20.0 [34.0] mg/dL, P<0.001) and 12 (19.0
[35.6] mg/dL, P<0.001) at high levels, and significantly affected GMS 14 at both
interference levels (-114.5 [44.4] mg/dL, P<0.001).
Conclusions: Accurate glucose monitoring improves glycemic control and outcomes
in critically ill patients. Sample interferences results in erroneous GMS measurements
and increases the risk for dangerous hypoglycemic events during IIT. GMS 1-10 was
observed to be the most robust against the evaluated interfering substances. We advise
caution for facilities using GMS 11-14 due to significant AA, BHB, GAL, GLUT,
LAC, and NAC inferences observed by our study.

S216

B-295
Extensive Evaluation of Hematocrit Interference on Point-of-Care Glucose
Meters

A. N. Steele, Z. Godwin, M. Howes, N. K. Tran. University of California

Davis, Davis, CA
Background: Intensive insulin therapy (IIT) guided by tight glycemic control
(TGC) reduces morbidity and mortality in critically ill patients. Accurate glucose
measurements are necessary for safe TGC. However, confounding factors, such
as abnormal hematocrit (HCT), results in inaccurate measurements on point-ofcare (POC) glucose monitoring systems (GMS). Abnormal HCT is common in
intensive care unit patients as a result of pathologic and iatrogenic mechanisms.
Therefore, accurate POC glucose measurements are necessary for safe IIT for TGC.
The objective of this study was to determine the effect of altered HCT on current
generation POC GMSs and the impact of autocorrecting biosensors in improving
glucose measurement accuracy.
Methods: We investigated the effects of abnormal HCT on the Nova Biomedical
(Waltham, MA) StatStrip hospital (GMS 1-5) and Xpress meters (GMS 6-10), Roche
(Indianapolis, IN) Inform II meter (GMS 11-13), and Abbott (Abbott Park, IL)
Precision Xceed Pro (GMS 14). All POC GMSs incorporated autocorrecting features
within their biosensors. Whole blood from 12 healthy adult (age18 years) volunteers
was used for testing. Specimens were adjusted to five different HCT levels for five
glucose levels (range: 50-500 mg/dL). Samples were tested on each GMS five times
for each individual glucose and interference level. Results were compared to a plasma
reference. Two-way ANOVA followed by pairwise analyses compared each GMS
versus the reference method at each HCT level.
Results: Study results are summarized in Table 1. Asterisks indicate HCT significance
GMS bias against the reference method.
Conclusions: Abnormal HCT is common in critically ill patients. Automatic
hematocrit correction is instrumental for accurate glucose monitoring and TGC.
GMS 1-10 exhibited acceptable performance at all HCT levels with the exception of
GMS 3. Performance was acceptable for GMS 11-13. However, we advise caution
for facilities using GMS 14 due to significant bias over a broad range of HCT levels.
Table 1. Comparison of Mean (SD) GMS Results
GMS GMS GMS
GMS GMS GMS
HCT
1
2
3
4
5
6
-13.8 -12.4 -13.8
-16.6 -8.0 -15.4
20
(8.9) (12.3) (8.7)
(12.5) (9.2) (12.9)

GMS
7
-14.0
(14.9)

30

-12.6 -12.6 -16.8 *** -14.0 -12.6 -6.0

(5.1) (11.5) (12.3)
(5.4) (6.4) (3.7)

-12.4
(9.2)

40

-15.8 -18.2 -22.2 *** -19.2 -12.2 -15.0 -13.0

(8.0) (16.9) (15.2)
(11.1) (8.2) (5.8) (14.5)

50

-19.6 -22.0 -23.2 *** -18.2 -18.0 -18.8 -16.8

(6.7) (15.7) (8.2)
(10.7) (12.8) (8.2) (9.4)

65

-24.8 -27.2 -25.4

(12.3) (18.6) (20.6)

-25.4 -25.6 -24.6 -18.0

(9.2) (16.1) (12.2) (16.8)

GMS
8
-8.6
(10.0)

GMS
9
-24.4
(24.6)

GMS
10
-9.0
(9.9)

GMS
11
13.0
(9.1)

GMS
12
14.0
(9.2)

GMS
13
14.0
(7.4)

GMS
14
-8.8
(13.2)
-18.4
-11.2 -13.2 -9.0 12.2 13.8 8.8
***
(6.1) (8.4) (11.4) (10.6) (11.8) (8.6)
(12.7)
-38.8
-17.6 -13.2 -8.2 4.6
8.6
2.6
***
(12.0) (17.7) (5.3) (12.5) (9.2) (10.2)
(31.7)
-52.6
-20.0 -17.0 -11.6 0.4
3.2
-4.6
***
(9.8) (9.6) (6.8) (22.7) (9.4) (16.5)
(30.8)
-72.8
-25.8 -21.2 -14.6 -7.2 -7.8 -16.0
***
(15.8) (12.4) (7.5) (17.6) (6.1) (18.7)
(45.0)

Note: Glucose levels were measured in mg/dL

Abbreviations: Glu, Glucose; SD, standard deviation
***P<0.001 as determined by Two-Way ANOVA and Tukeys HSD

B-296
Optimization of the turn-around-time of CRP measurement in the emergency
setting by using the Microsemi analyzer

H. Baum, H. Sander, A. Leni, K. Mages. Regionale Kliniken Holding RKH

GmbH, Ludwigsburg, Germany
Introduction: C-reactive protein (CRP) is an established marker in the diagnosis and
follow-up of patients with infectious diseases. In the acute setting it can help to assess
and prioritize these patients and therefore improve diagnosis and further treatment.
The Microsemi CRP analyzer is a small analyzer, which determines with the use
of only 18 l EDTA-blood a complete blood cell count and CRP within 4 minutes.
Furthermore, the Microsemi can be linked to a LAN for rapid result upload into a
hospital information system (HIS). The aim of this study was to determine the turnaround-time (TAT) of CRP measured with the Microsemi analyzer and compared it
with a fully automated method in a routine labor setting.
Material and Methods: Serum and EDTA samples from 66 patients with an urgent
test-request were selected for analysis out of the daily routine. Serum-CRP was
measured on a Vitros 5600 analyser whereas the Microsemi was used to test CRP
with EDTA samples . Time of blood collection was determined using the time of
ordering in the order-entry system of the hospital information system (HIS). Using
the laboratory information system (LIS) the time of arriving in the lab, uploading the
result from the routine analyzer into the LIS, and reporting the result into the HIS

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Point-of-Care Testing

Wednesday, July 30, 9:30 am 5:00 pm

were collected. For the Microsemi the measuring time for CRP is 4 minutes. To these
4 minutes we added a mean time for other work of 5 minutes.

B-298

Results: The method comparison showed a good correlation between both assays
with r = 0.9888 and CRP (Microsemi) = 1.057 x CRP (Vitros) - 0.235.

Comparison of Six Point-of-Care Glucose Monitoring (POCGM) Devices in

Diabetic and Hemodialysis Patients.

Mean time from blood collection to arrival in the lab was 32 (2-235) minutes. Mean
time from arrival of the blood in the lab to reporting of the results into the HIS was
37 (21-103) minutes for the Vitros and 9 minutes of the Microsemi. Taking together
it results in a mean total TAT of 69 (23 - 338) minutes for the Vitros and 41 (13244) minutes for the Microsemi. Therefore, using the Microsemi the results could
be reported in mean 28 (10-94) minutes earlier compared to the routine processing
procedure.

W. Tamimi1, N. Alotaibi2, A. Alsadhan2, S. Aljasser1, R. Dafterdar3, A.

Khan2, A. Alajlan1. 1King Fahad National Guard Hospital, King Saud Bin
Abdulaziz University for Health Sciences, Riyadh, Saudi Arabia, 2King
Fahad National Guard Hospital, Riyadh, Saudi Arabia, 3Medical Services
Department, Ministry of Defense, Riyadh, Saudi Arabia

Conclusion: Using the Microsemi for CRP measurement in the emergency situation
there is the possibility to report much faster this critical parameter to the clinician
without loss of analytical accuracy.

B-297
Evaluation of Clinitest hCG device susceptibility to high-dose hook effect by
intact human chorionic gonadotropin (hCG) and hCG beta-core fragment at
concentrations observed in early natural pregnancy

D. M. Milhorn, N. L. Korpi-Steiner. UNC, Chapel Hill, NC

Background: Qualitative detection of urinary human chorionic gonadotropin (hCG)
using point-of-care testing devices is common practice in the evaluation of suspected
pregnancy. False-negative findings due to high-dose hook effect are known to occur
at elevated urinary concentrations of intvact hCG and/or hCG beta-core fragment
(hCGcf) with select POC hCG devices. In early pregnancy, hCG and hCGcf
concentrations vary depending on days relative to ovulation/conception. The aims
of this study were to evaluate hook effect of intact hCG alone, hCGcf alone, and
combinations of hCG and hCGcf exhibited in early pregnancy using the Clinitest
hCG device with accompanying Clinitek Status (Siemens Healthcare Diagnostics)
readout. The distribution of days relative to ovulation at the time of Clinitest hCG
testing for patients presenting at 1 institution was further evaluated.
Methods: Hook effect by intact hCG and hCGcf using Clinitest hCG devices with
Clinitek Status readout was evaluated using hCG-negative urine matrix containing 7
levels of purified intact hCG or hCGcf (0 - 2x106 pmol/L) alone, and combinations
of intact hCG and hCGcf (n=20) corresponding to physiological concentrations
( 50 pmol/L and 2% for concentrations 10,000 pmol/L and > 10,000 pmol/L,
respectively) detected every other day in early natural pregnancy between 13 and
49 days relative to ovulation (n=37, intact hCG range 10 5.9x104 pmol/L; hCGcf
range 10 2.3x105 pmol/L; data courtesy of McChesney et al. Human Reprod 2005.
20(4):928-35). Prepared samples were tested in duplicate. Estimated days relative to
ovulation at the time of Clinitest hCG testing for female patients 18 years presenting
to UNC Hospitals with borderline or positive Clinitest hCG results (n=182) were
calculated using retrospective analyses of estimated date of delivery (EDD), date of
Clinitest hCG testing, and a gestation slide chart (Perrygraf) with assumed 40 weeks
gestational age at EDD and 14 day relationship between gestational age and days
relative to ovulation.
Results: Clinitest hCG results per Clinitek Status readout for urine samples containing
intact hCG or hCGcf were as follows: positive for intact hCG at all concentrations
(500-2x106 pmol/L) tested, positive between 500 to 5x104 pmol/L hCGcf, borderline
at 5x105 pmol/L hCGcf, and negative at 0, 1x106 and 2x106 pmol/L hCGcf. Positive
Clinitest hCG results were detected for combinations of intact hCG and hCGcf in
urine corresponding to concentrations detected in early pregnancy between 16 to 49
days relative to ovulation. The median estimated days relative to ovulation at the time
of Clinitest hCG testing among 182 pregnant females presenting at 1 institution was
38 days, with an inter-quartile range of 27.5 to 58.5 days.
Conclusion: The Clinitest hCG device with Clinitek Status readout demonstrated no
detectable hook effect by intact hCG. Hook effect at high concentrations of hCGcf
was observed though did not interfere with Clinitest hCG detection of combined
intact hCG and hCGcf concentrations known to occur in early natural pregnancy.
The majority of pregnant females presenting at 1 institution have Clinitest hCG
testing performed during early pregnancy and are unlikely to exhibit false-negative
Clinitest hCG results due to hook effect by hCGcf.

Background: Many patients with diabetes control their blood sugar level on a daily
basis with Point-of-Care Glucose Monitoring (POCGM) devices. Accurate readings
are of high importance to successfully self-manage their diabetes. There is variety of
POCGM home-use devices available to measure glucose for diabetic patients. Each
has a different degree of accuracy and imprecision. Therefore, in this study, we have
evaluated six glucose meters from different manufacturers for patients home-use.
Patients & Methods: A total of 80 blood samples were collected from venous blood
obtained from 20 healthy adults, 40 diabetic and 20 hemodialysis patients during July
2012 in our hospital. For each hemodialysis patient two blood samples were collected
before and after dialysis. Blood glucose level was measured in these samples by
different six POCGM devices from different manufacturers (denoted as A, B, C, D, E
& F) and was compared to the reference glucose- hexokinase method Architect 16000
(Abbott). Two manufacturers (A & B) utilize strip that used the enzyme PQQ-GDH
(pyrroloquinolinequinone dependent glucose dehydrogenase). The test strips utilize
flavin adenine dinucleotide (FAD) with glucose-dehydrogenase (GDH) enzyme by
manufacturers C, D and E. The strip from manufacturer F utilizes glucose oxygenase
enzyme. Results: When the ISO 15197 standards were applied, most of POCGM
devices have shown a good agreement and accuracy with the reference glucosehexokinase method in the range of 82-98% with the exception of one device (F)
which has shown only 39% agreement. All of the six devices have shown an average
negative bias with the reference glucose-hexokinase method in the range of -6.2% up
to -24% (p-value <0.0001). Conclusion: The home-use POCGM devices produced
comparable results in relation to the glucose-hexokinase reference method. Devices
that use FAD-glucose-dehydrogenase method have shown better accuracy in these
studied populations.

B-299
Evaluation of Glucose Meter Accuracy Using Locally-Smoothed Median
Absolute Difference (LSMAD) Analysis

M. Howes, A. N. Steele, Z. Godwin, N. K. Tran. UC Davis, Davis, CA

Background: Glucose monitoring system (GMS) accuracy is crucial for the
management of critically ill patients. Intensive insulin therapy (IIT) is used for
tight glycemic control (TGC) and relies on accurate GMS measurements. Poor
GMS performance results in inappropriate insulin dosing and increases risk for
dangerous glycemic excursions. Performance assessment of GMSs is instrumental for
determining appropriate devices for critical care TGC. We hypothesize that traditional
measures of performance may be inadequate for evaluating current generation GMSs
when compared to the new locally smoothed median absolute difference (LSMAD)
method.
Methods: We evaluated the performance of the Nova Biomedical (Waltham, MA)
StatStrip hospital (GMS 1-5) and Xpress meters (GMS 6-10), Roche Diagnostics
(Indianapolis, IN) Inform II meter (GMS 11-13), and Abbott Laboratories (Abbott
Park, IL) Precision Xceed Pro (GMS 14) against a plasma reference. We collected 202
unique remnant arterial blood gas samples for paired testing. Traditional parametric
statistics including ANOVA, Bland-Altman, and least square linear regression (LSLR)
analyses were compared to the non-parametric LSMAD method. A 5 mg/dL tolerance
threshold served as a LSMAD performance benchmark.
Results: We found no statistically significant differences via ANOVA, Bland-Altman,
or LSLR analyses. With LSMAD analysis (Figure 1), GMS 1-10 exceeded the 5 mg/
dL tolerance threshold for mean (SD) values greater than 86.9 (17.67) mg/dL. GMS
11-13 exhibited breakout points at two mean points: 11.0 and 17.0 mg/dL. For GMS
14, we observed a breakout point of 54 mg/dL.
Conclusions: Accurate glucose monitoring is instrumental for appropriate IIT
in critically ill patients. Current generation GMSs have improved performance.
Traditional methods used to analyze device performance proved inadequate in
identifying significant differences between GMSs. In contrast, LSMAD illustrated
the median performance of GMS 11-14 to be inadequate at clinically significantly
hypoglycemic ranges_suggesting that these devices may be inappropriate for critical
care.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S217

Point-of-Care Testing

Wednesday, July 30, 9:30 am 5:00 pm

B-301

An analytical evaluation of the Abbott i-STAT hCG test cartridge

A. M. Sowder1, M. L. Yarbrough2, R. Nerenz2, R. Mortensen3, A. M.

Gronowski2, D. G. Grenache1. 1University of Utah School of Medicine, Salt
Lake City, UT, 2Department of Pathology and Immunology, Washington
University School of Medicine, St. Louis, MO, 3ARUP Laboratories, Salt
Lake City, UT
Background: The qualitative detection of human chorionic gonadotropin (hCG) in
urine is commonly used as a rapid test to determine pregnancy status. Urine hCG
tests are less analytically sensitive than quantitative serum hCG tests and are prone
to false-negative results. Despite these drawbacks, urine hCG tests are often favored
over serum hCG tests because they can be performed at the point-of-care. The ability
to perform quantitative hCG testing in whole blood at the point-of-care would be
advantageous. The i-STAT total hCG cartridge (Abbott Diagnostics, Abbott Park, IL)
is a quantitative hCG test to be used with whole blood or plasma with an intended
use for the detection of early pregnancy. The purpose of this study was to perform an
analytical validation of the i-STAT hCG test.

B-300
Evaluation of point-of-care (POC) glucose test performance of Patient Care
Technicians (PCT) and Registered Nurses (RN)

S. N. Narla, J. Epps, K. Hermayer, P. Arnold, Y. Zhu. Medical University of

South Carolina, Charleston, SC
Background: The glucose results attained by POC glucose meters are highly
dependent on critical thinking and performance of the personnel performing the test.
In the majority of clinical settings, either PCTs or RNs perform the POC glucose tests.
The objective of this study is to evaluate the POC glucose test performance of PCTs
and RNs by comparing the accuracy of their POC glucose results with the clinical
chemistry (CC) lab analyzer results.
Method: Abbotts Precision XceedPro glucose meter based on the glucose
dehydrogenase method was used for POC test. In the CC lab, glucose is analyzed on
Beckman Coulter UniCelDxC800 analyzers by an oxygen rate method employing a
Beckman Coulter Oxygen electrode. Glucose results acquired with both POC glucose
and CC lab analyzer with a difference in blood collection time of 5 mins were
chosen and the data was collected retrospectively for 15 days, i.e., from 01.16.2014
to 01.30.2014. The average % variance between POC and CC lab tests was analyzed
for the tests performed by PCTs and RNs. The POC and the CC lab test results were
analyzed based on the personnel performing the POC tests by paired two-tailed t test.
Results: 402 tests were analyzed with glucose ranges of 35 mg/dL to 463 mg/dL in
which 155 were performed by PCTs, 221 by RNs and 26 by other personnel. The
average %variance of tests performed by PCTs compared to the CC lab was 8.40,
and the results of POC and the CC lab were statistically significant (p=0.0121).
Also, 51(33%), 47(30 %) and 57(37%) of these tests showed >10%, 5%-9.9% and
<5 % variance respectively compared to the CC lab. For tests performed by RNs,
the average % variance was 7.31, and the results by POC and the CC lab were not
statistically significant (p=0.7910). Also, 52(23%), 66(30%) and 103(47%) of these
tests showed >10%, 5%-9.9% and <5% variance respectively. According to CLSI, for
glucose <100 mg/dL and >100 mg/dL, the difference from the POCT to CC lab test
should not exceed 12 mg/dL and 12.5% respectively. 24% of tests performed by PCTs
and 17% tests performed by RNs did not meet these criteria. Also in both the groups,
there were 15 individuals with at least 2 tests showing >10% variance, out of which
12 were throughout the 15-day period and 5 with at least 2 tests >10% variance on the
same day, there were 2 individuals who fell in both categories.
Conclusion: Overall performance by both RNs and PCTs is within the acceptable
(20%) allowable error. However, there is no statistical significance between POC
and CC lab results for tests performed by RNs, whereas the tests performed by PCTs
were statistically significant. Also, there are some individuals whose performance is
not up to the required standards, therefore, it is important to identify these people and
educate them about the potential sources of errors caused by the operators and critical
thinking. It is also important to perform timely education and competency assessment
on the personnel performing the POC testing.

S218

Methods: hCG-free whole blood was obtained from volunteers. Residual serum and/
or plasma samples sent to the laboratory for physician-ordered hCG tests were used
as a source of hCG. Aliquots of the hCG-free whole blood were used to prepare
samples with specific target hCG concentrations. Whole blood and plasma were used
to evaluate the precision, linearity, analytical sensitivity, and accuracy of the i-STAT
hCG test. Institutional Review Board approval was obtained for this study.
Results: Precision was determined from two samples analyzed in two replicates,
twice per day for 10 days. Whole blood repeatability and within-laboratory CVs were
14.4 and 15.6% at 10.1 IU/L and 6.5 and 6.5% at 1176.6 IU/L, respectively. Plasma
repeatability and within-laboratory CVs were 5.6 and 9.9% at 11.2 IU/L and 4.2 and
4.8% at 1273.7 IU/L, respectively. Linearity was evaluated from six samples prepared
to span the claimed analytical measuring range of 5-2,000 IU/L. For whole blood,
linear regression produced a slope of 0.99, a y-intercept of 6.1, and an r2 of 0.999.
For plasma, linear regression produced a slope of 1.0, a y-intercept of 1.6, and an r2
of 0.999. Analytical sensitivity was determined from a set of five samples prepared to
contain 0, 5, 10, 15, and 20 IU/L of hCG and each analyzed in 10 replicates. The limitof-blank, defined as the mean+3 SD of the 0 IU/L sample, was 0 IU/L for whole blood
and plasma. The limit-of-detection, defined as LOB+3 SD of the 5 IU/L sample, was
2.9 and 1.7 IU/L for whole blood and plasma, respectively. The limit-of-quantitation,
defined as the hCG concentration that yielded a CV of 20% was 8.0 and <5 IU/L
for whole blood and plasma, respectively. Accuracy was evaluated using 20 samples
tested in one replicate on the i-STAT and compared to corresponding plasma hCG
concentrations measured on the Architect Total -hCG (Abbott Diagnostics) assay.
For whole blood (i-STAT) vs. plasma (Architect), Deming regression produced a
slope of 0.96, a y-intercept of -33, and r of 0.997. For plasma (i-STAT) vs. plasma
(Architect), Deming regression produced a slope of 1.09, a y-intercept of -22.6, and
r of 0.994.
Conclusions: The Abbott i-STAT total hCG cartridge demonstrates acceptable
performance for quantifying hCG in whole blood or plasma.

B-303
Evaluation of Accu-Chek Inform II performance with cobas c501 and
Modular P800 Using Specimens from Patients in Emergency Department,
Medical-Surgical and Intensive Care Units.

V. M. Genta1, L. Wyer2, M. Ferguson1, S. S. Church2, D. Graubner3, Y.

Shen1. 1Sentara Virginia Beach General Hospital, Virginia Beach, VA,
2
Sentara Healthcare, Norfolk, VA, 3Sentara Helthcare, Norfolk, VA
Background: Glucose meters, offering a rapid evaluation of patient blood glucose values
at the point-of-care, promote prompt medical intervention. The interchangeability of
the glucose meter results with those obtained with the laboratory method is essential
for their seamless interpretation. The performance of the glucose meter method
may be affected by the blood matrix of patients treated in different hospital units.
We report the results of a limited study comparing Accu-Chek Inform II with the
laboratory method using blood specimens from patients treated in Medical-Surgical
Unit (MSU), Intensive Care unit (ICU) and Emergency Department (ED).Methods:
Patient specimens: 1230 (MSU 520, ICU 290, ED 420). Glucose meters: AccuChek Inform II (Roche Diagnostics), Strip lot #471353; Laboratory Methods: cobas
c501 and Modular P800 (Roche Diagnostics).The patient specimens were collected
by venipuncture in green-top tubes and assayed in parallel and within 30 minutes
with both methods. For 324 patients the HCT value was available.The observations

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Point-of-Care Testing

Wednesday, July 30, 9:30 am 5:00 pm

were collected in Minitab (Version 16, Minitab) statistical software and were
analyzed with multivariable weighted least squares regression analysis (MWLSR),
locally weighted scatterplot smoother (lowess), regression diagnostics and graphic
representations.Results: The scatterplot of the glucose meters values (y axis) by the
laboratory methods values (x axis) by the hospital unit, showed a linear relationship
between methods. This was confirmed by the lowess. This plot showed increased
variability for increasing glucose values; this prompted the use of a weighted least
squares regression model. The absolute (for values 30-100 mg/dL) and relative (for
values 101-600 mg/dL) difference plots showed that for the grand majority (99%) of
specimens the differences were within the total error (CLIAs criterion target value +/6mg/dl, or +/-10%, greater).The MWLSR model (y=3+1.0x+2Unit+0.7HCT) showed
that while there were no statistically significant differences between regression
lines for HCT (P=0.42), there were statistically significant differences for Units
(P<0.0001). The slope for ED (beta1=1.01) was statistically significantly different
from those for MSU (beta1=0.95) and ICU (beta1=0.95). However, these differences
were not clinically significant. The pure error test by data subsetting showed possible
lack of linearity for high values.The lowess for the plot of the standardized deleted
residuals by the fitted value showed a slight curvature for values >400mg/dL and 9
potential outliers (3<|value|<4).The leverage (Hi<0.1), Cooks distance (<0.05) and
DIFTs(<|0.06|) did not show any influential observations. Conclusion: This study
showed that there was a linear relationship between Accu-Chek Inform II method
and the laboratory methods (cobas c501 and Modular P800). For the grand majority
of the specimens, the absolute and relative differences were within the CLIAs
criterion for total error. Consequently, these results suggest that the two methods can
be used interchangeably with the laboratory methods for evaluating patient glucose
blood levels. However, due to the limited number of specimens some matrix effects
secondary to either disease or treatment may not have been manifest. Further studies
should be performed to corroborate these findings.

B-304
Analytical validation of a blood glucose meter device in the emergency
department of a university hospital

M. S. Feitosa, N. C. C. S. Nogueira, R. L. Furlan, L. S. Vasconcellos.

Universidade Federal de Minas Gerais, Belo Horizonte, Brazil
Background: The university hospital in Belo Horizonte is a 500 bed tertiary hospital.
Despite federal regulation that demands the use of hospital equipments, the glucose
meters (and its strips) in the emergency department are appropriate for domestic use
only, and they are not monitored by laboratory staff. Furthermore, their analytical
performance has not been validated. As part of a project of implementation of glucose
meters suitable for hospital use, we conducted analytical validation of the Precision
XCEED PRO (PXP) Abbott, quality control and calibration in the emergency
department, which is reference for several clinical conditions in the public health
system.
Methods: Imprecision, accuracy and linearity were evaluated as part of the validation
plan. Within-run imprecision was first evaluated using two level QC solutions tested
10 times each. Between-run imprecision was conducted running each control four
times for five days. Accuracy was performed comparing 20 results of PXP with
Vitros 5600 in the core laboratory. A capillary sample was collected to perform
glucose testing in PXP. Immediately after, one fluoride tube and one arterial
heparin tube were collected and sent to the core lab to be tested in Vitros 5600. The
experiment was conducted for five days, so at least five patients were tested each day.
Linearity was performed using a five level linearity kit from RNA Medical. Each level
was tested in quadruplicate.
Results: Within-run imprecision was 3.45% for control 1 (mean = 88.7mg/dL) and
3.26% for control 2 (mean=273.0 mg/dL). Between-run imprecision within 5 days
were 6.47% (mean = 88.9mg/dL) and 4.35% (mean = 287.2mg/dL). All values were
considered acceptable. Comparison between capillary samples and fluoride plasma
were performed with correlation coefficient (r) = 0.99, slope = 1.087, intercept =
-4.53 and 95% of samples with results in the acceptable range < 20% of total error,
according to ISO 15197 specifications. Comparison between heparin samples and
plasma fluoride were performed also with correlation coefficient (r) = 0.992, slope =
0.969, intercept = -4.53mg/dL and all samples with results in the acceptable range <
20% of total error. Linearity was evaluated ranging from 24mg/dL to 427 mg/dL with
software QCM3.0.
Conclusions: Precision EXCEED PRO (PXP) is suitable for hospital environment.
Validation experiment performed with this particular device showed that its
performance meets the quality requirements established by the laboratory and the
literature.

B-306
Evaluation of the Performance of a Commonly-Used Glucometer in a Tertiary
Hospital in Nigeria.

C. P. Onyenekwu, E. U. Egbuagha, E. C. Azinge. Department of Clinical

Pathology, Lagos University Teaching Hospital, Surulere, Lagos, Nigeria
Background:Point-of-care-testing (POCT) for glucose is the most common type of
near patient testing performed at various sites within the Lagos University Teaching
Hospital (LUTH). A previous study at our hospital among POCT operators found
among others, a lack of awareness concerning evaluation of POCT devices and the use
of quality control materials for POCT; absence of liaison with the central laboratory
for the verification of suscpicious results, and absence of external quality assurance
programs for POCT. This finding prompted us to evaluate the accuracy of the AccuChek Active (Roche) glucometer, which is the glucometer in use at 90% of the sites
performing POCT for glucose, including the emergency units and the diabetic clinic.
Methods:The study was approved by the Health Research and Ethics Committee
of LUTH and was conducted over a four-week period. Glucose levels in capillary
blood samples from 31 diabetics and 18 non-diabetics attending the Lagos University
Teaching Hospital (LUTH) were measured with an Accu-Chek Active glucometer
calibrated to plasma samples. All glucometer readings were conducted by the same
individual, in order to reduce operator variability. Venous plasma was collected from
the patients into fluoride oxalate vacutainers within five minutes of finger-prick
tests. The central laboratory measured the plasma glucose on Siemens Dimension
Xpand Plus analyser. The laboratory analysis was performed within one hour after
sample collection. Precision studies were not conducted as control solutions for AccuChek Active are apparently not easily available in Nigeria. Accuracy was assessed
with Pearsons correlation, % bias and Bland-Altman plot of the results from the
two methods. The ISO 15197:2013 and the American Diabetes Association (ADA)
requirements for glucose testing were employed in assessing quality. Data was
extracted on Microsoft Excel and analysed using Analyse-It for Excel software
version 2.30, Leeds, United Kingdom.
Results:The glucometer results were highly correlated with those of the central
laboratory (r=0.99), bias ranged from 0.0 to 20.0%. Higher bias levels were observed
with high glucose results. A Bland-Altman plot of the difference between each pair
of results (glucometer and central laboratory), against the central laboratory result,
showed increasing variability of glucometer results at high glucose levels. The
glucometer met the ISO recommendation with 95.9% of the results having a bias
<15%, however, it did not meet the ADA requirements of <5% bias for glucose testing.
Conclusion:The Accu-Chek Active glucometer had a high correlation with the central
laboratory method but showed increased variability at high glucose levels. This may
have implications for patient care particularly at the diabetic outpatient and emergency
care units where insulin-dose adjustments are made according to glucometer readings,
without verification of results from the central laboratory.

B-307
Utilization of a Superior Monoclonal Antibody Pair against Procalcitonin in
Development of Fluorescence-based Lateral Flow Immunoassay

L. Q. Wang, S. He, Y. Z. Guo, D. F. Li, Y. Liu, J. Ge. Abzymo Bioscience

Ltd., Beijing, China
Background: After the first report in 1993 on Procalcitonin (PCT) in patients with
bacterial infection elevated significantly, the use of PCT in identifying the bacterial or
non-bacterial origin of systemic inflammation has been gaining widespread support.
Because the level of PCT in the blood stream of healthy individuals is below 0.5 ng/
ml, we aimed to generate a set of specific monoclonal antibodies with higher affinity
and develop a rapid and sensitive POCT test for use in the emergency rooms and
clinical laboratories.
Methods: Balb/c mice (6-8 weeks) were immunized with recombinant PCT emulsfied
with Freunds or Titermax adjuvant. Four times of intraveneous injections were given
at 50g per injection in 3 week intervals. ELISA were conducted for monitoring PCTspecific sera titer, four mice with the highest titer were selected for the cell fusion.
Three days before the fusion, the last shot was done intraveneously. The splenocytes
of the four mice were fused with the mouse myeloma cell line SP2/0 using PEG-1500.
The microplate wells exhibiting hybridoma growth were screened for the production
of anti-PCT antibodies with both direct and indirect methods. Positive hybridoma
cultures with higher titer and specificity were selected and subcloned by three
round of limited dilutions. All the 98 MAbs were purified by protein A-sepharose
affinity chromatography and were evaluated by colloidal gold-based LFT. The LFT
test containing 4763 kinds of combination in which each MAb was coated as the

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S219

Point-of-Care Testing

Wednesday, July 30, 9:30 am 5:00 pm

capture antibody and the others were detected as the detecting antibody. After that,
18 matched MAb pairs were selected for further development of fluorescence-based
lateral flow immunoassay (FLFIA) and the antibody pair (PCT79/PCT83) was shown
to have the best sensitivity, specificity, stability and coincidence rate. Furthermore, the
results from clinical samples tested by FLFIA with this antibody pair (PCT79/PCT83)
were compared and evaluated with those of electrochemiluminescence immunoassay
(ECLIA) method.
Results: Ninety-eight hybrid clones were screened eventually using the abovementioned methods. After the evaluation in colloidal gold tests,18 matched MAb pairs
were screened from 4763 combination. In FLFIA, the best antibody pair (PCT79/
PCT83) was selected with the sensitivity of 0.1 ng/ml, detection range of 0.1~100
ng/ml; inter-assay CVs <15%. The results from clinical samples revealed that our
strip using the antibody pair (PCT79/PCT83) correlates well with the ECLIA method
(R=0.988).
Conclusion: We described here that we have generated a superior MAbs pair against
PCT which is proved to have promising potential applications in development of
FLFIA. They seem to be the ideal candidate antibodies suitable for development of a
quantitative POCT.

B-308
Bioelectronic Platform for Sensitive and Versatile Diagnostic Applications*

D. Georganopoulou1, A. Gaustad1, E. Van Groll1, R. Hoo1, T. J. Meade2,

P. Bao1. 1Ohmx Corporation, Evanston, IL, 2Northwestern University,
Evanston, IL
Background A new bioelectronic platform is described that is designed for multiple
Point-of-Care (POC) diagnostic applications, including protein, DNA, and small
molecule diagnostics. A self-assembled monolayer (SAM) technology is presented
that demonstrates quantitative, ultra-sensitive, precise, and accurate measurement of
a number of clinical analytes in biological samples (e.g. whole blood, urine, semen,
prostatic fluid, saliva, etc). Cyclic voltammetry techniques measuring a ratiometric
signal allow for a rapid, self-calibrating, fully quantitative dose response with broad
dynamic range (over 4 logs of analyte concentration). This capability allows any
applicable clinical assays to be executed on the Ohmx platform, using a minimal
sample volume (1-50 uL), with performance levels similar to reference lab tests. The
analytical performance of the sensor, and the clinical validation for multiple analytes
(Hemoglobin A1c, hs Troponin I, TSH, hs CRP, DNA, and lactate) are discussed.
Methods Following standard solution bioassays (immunoassays, hybridization,
or enzymatic reactions) an electrophore substrate specifically reacts with the selfassembled monolayers on the gold micro-electrodes. For all assays, a dose response
spanning the analytes relevant clinical range was obtained using commercially
available calibrators. Further clinical validation is presented with 50 clinical samples
tested for A1c, 72 samples tested for Troponin I and 50 samples for TSH. The
clinical samples were provided by hospital collaborators and pre-tested with clinical
immunoanalyzers. A correlation statistical study is shown between the Ohmx test and
reference methods.
Results The Ohmx HbA1c is a single measurement test in 2 uL of whole blood, a
TAT of 3 minutes with a linear response ranging from 2.7% (6.0 mmol/mol) to 19.8%
(192.9 mmol/mol), an intraday precision of CV less than 3% and with R2= 0.93 linear
correlation with BioRad clinical immunoanalyzer. The Ohmx troponin I assay is a
high sensitivity assay that has an LOD of 1 pg/mL with a TAT of 15 min, with CVs
from 3-10% and a linear correlation with R2= 0.95 with Siemens immunoanalyzers.
TheTSH Ohmx test has an LOD of 0.0024 uIU/mL, a TAT of 15 min and R2= 0.9
correlation with the Beckman Coulter DXI. SNP differentiation is shown for DNA
factors II, IV, and MHTFR, and a lactate test is shown with a linear range from 0.2
to 28 mM.
Conclusions There is significant correlation between the Ohmx tests and clinical
immunonalyzers. Analytical and clinical performance for all POC tests offered rivals
performance of reference labs. The platform versatility is shown with tests that include
proteins, DNA, and small molecules with a cartridge that has a COGS of $0.40.
*Assays currently under development and not for clinical use

B-309
Analytical and Diagnostic Characteristics of High-Sensitivity Troponin Assays:
Examination of PATHFAST cTnI

E. Spanuth1, B. Ivandic2, R. Thomae3, E. Giannitsis4. 1DIAneering GmbH,

Heidelberg, Germany, 2DIAneering - Diagnostics Engineering & Research,
Heidelberg, Germany, 3Mitsubishi Chemical Europe, Dsseldorf, Germany,
4
Department of Internal Medicine III, University Hospital Heidelberg,
Heidelberg, Germany
Background The analytical characteristics of high-sensitivity cardiac troponin
(cTn) assays comprise an imprecision (CV) at the 99th percentile value 10%
and measurable concentrations above the limit of detection (LoD) and below the
99th percentile in at least 50% of healthy individuals. The PATHFAST cTnI assay
(Mitsubishi Medience Corporation, Japan) has already shown promising analytical
validity and is usable for point-of-care testing. We thought to evaluate it`s analytical
and diagnostic characteristics and to examine whether the assay could be classified
as high-sensitive.
Methods To establish the analytical criteria cTnI was determined using PATHFAST
in 120 healthy individuals (60 men and 59 women, 21-69 years old, median 42
years) in whom cardiac disorders were excluded by extensive evaluation including
cardiac magnetic resonance imaging and NT-proBNP < 125 ng/L. The diagnostic
characteristics were investigated by comparison of cTnI and cTnT (Roche`s highsensitivity assay cobas hs-cTnT) in 181 patients admitted to the chest pain unit at
presentation, 3 and 6 hours later. The results were related to the discharge diagnoses.
Results The cTnI concentrations measured in the healthy individuals ranged from
0.4 to 17.2, mean 2.1 (95% CI:1.6-2.6) ng/L, without age dependency. Men revealed
higher levels than women, means (IQR) were 2.8 (1.2-2.6) and 1.1 (0.7-1.3) ng/L.
The CLSI nonparametric method revealed a 99th percentile value of 16 ng/L below the
manufacturer recommended value of 20 ng/L. The quantification of cTnI above the
LoD (1.0 ng/L ) and below the 99th percentile was possible in 79 of 120 individuals.
The imprecision profile according to NCCLS revealed 20%, 10% and 5% CVs at
cTnI concentrations of 2, 3 and at 20 ng/L, respectively. In the patient group the
discharge diagnosis was non-ST-segment elevated myocardial infarction (NSTEMI)
in 72 patients. The cTnI median values at 0, 3 and 6 hours were 46, 166 and 399
ng/L, respectively. To evaluate the diagnostic validity for detection of NSTEMI the
results of PATHFAST cTnI and cobas hs-cTnT were compared by ROC analysis.
AUC values at 0, 3 and 6 hours were 0.919, 0.962 and 0.958 for cTnI and 0.923,
0.964 and 0.969 for hs-cTnT, respectively. cTnI revealed AUC values for absolute
changes from admission to 3 hours and from admission to 6 hours of 0.920 and 0.931
in NSTEMI patients.
Conclusions PATHFAST cTnI demonstrated complete fulfillment of the analytical
criteria for high-sensitive cTn assays: The imprecision (CV) at the manufacturer
recommended 99th percentile value was 5%. Quantification of cTnI was possible
in 65.8% of healthy individuals. The examination of the diagnostic characteristics
revealed complete concordance with cobas hs-cTnT detection of NSTEMI as well as
for assessment of absolute changes of cTn values (rise and/or fall) during the course
over time in NSTEMI patients. PATHFAST cTnI showed highly sensitive detection
of NSTEMI with increasing sensitivity at admission and after 3 to 6 hours, not going
along with decreased specificity. The PATHFAST cTnI assay allows high-sensitivity
determination of cTnI within 16 min from whole blood samples and might be useful at
the point-of-care setting for early rule- in and rule-out diagnosis of NSTEMI.

B-311
Evaluation of a point-of-care HbA1C method in an underserved community
clinic and measurement of the impact of implementing a high-quality assay
with immediate results.

A. V. Jean-Noel1, B. Williams1, M. M. Perez2, M. Spinola2, F. Azzara2, H.

Bidwell2, G. L. Horowitz1, N. V. Tolan1. 1Beth Israel Deaconess Medical
Center and Harvard Medical School, Department of Pathology and
Laboratory Medicine, Boston, MA, 2Bowdoin Street Health Center, Beth
Israel Deaconess Medical Center, Boston, MA
Background:Measurement of hemoglobin A1c (HbA1C) assesses glycemic control
and provides an estimate of the average glucose concentration (eAG) over the
previous 2-3 months. The American Diabetes Association recommends that %HbA1C
be measured twice annually for diabetic patients meeting glycemic goals, in an effort
to reduce the risk of microvascular disease. In this study, we evaluated a NGSPcertified CLIA-waived point of care (POC) method for HbA1C by finger-stick and
measured the impact on workflow and compliance with pay-for-performance goals in
an underserved community clinic.

S220

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Point-of-Care Testing

Wednesday, July 30, 9:30 am 5:00 pm

Methods:The boronate affinity Afinion AS100 HbA1C POC assay (Alere Analytics)
was evaluated for performance by the clinic nursing staff, in addition to the laboratory
technologists. Precision studies, across various cartridge and QC lots, included the
calculation of intra- (n=10) and inter-assay (n=30, 150 days) coefficients of variation
(CV) for two levels of manufacturer QC (6% and 8% HbA1C). Method correlation
against the Cobas Integra 800 Tina-quant HbA1C immunoassay method (Roche
Diagnostics) was assessed through linear regression, mean bias, Bland-Altman plots
and evaluation of clinical concordance. This included EDTA samples (n=27) and
concurrent venipuncture/capillary samples (n=20). EDTA samples collected from
patients with HbC and HbS were also included. Compliance with our institutions
pay-for-performance goal of measuring HbA1C twice annually (63mo) for diabetic
patients, was compared before and after implementing POCT.

CONCLUSIONS: The biosensor based creatinine assay under development

demonstrated analytical performance in whole blood with a TE meeting the NKDEP
recommendations for reporting eGFR.

Results:The intra- and inter-assay precision was <2% and 3% CV, respectively.
For HbA1C ranging from 3.9-14.9%, the method correlation for EDTA samples
(y=0.94x+0.37, r2=0.99) and paired venipuncture/capillary samples (y=0.93x+0.28,
r2=0.97) showed excellent agreement. The mean (SD) of bias of %HbA1C was
-0.1%(0.3) and -0.2%(0.2) for EDTA and paired venipuncture/capillary samples,
respectively. There was no significant bias observed for samples with the hemoglobin
variants included. Overall, clinical concordance was 91%(43/47) for %HbA1C within
the ranges of 10%(5/6), where the absolute bias was less than 0.5%. Representing
32% of all HbA1Cs ordered at this clinic, 189 tests were performed by POCT in the
first two months, allowing for real-time therapy assessment and feedback to diabetic
patients. Of these samples, 146(77%) were collected without any additional laboratory
testing ordered for the patient that would have required venipuncture. Perhaps more
important, this means 43(23%) patients, who had venipuncture for other laboratory
testing, also had POCT performed for the sole benefit of the real-time assessment.
Introduction of POCT was associated with an initial 6% increase in the clinics
compliance with a pay-for-performance goal of serial testing performed within nine
months. In addition, POCT was provided for 57(32%) patients whose prior HbA1C
determination exceeded these performance goals.
Conclusion:Accurate representation of long-term method performance is possible
when the evaluation is conducted by the end-user. Compared to the immunoassay
method, the Afinion HbA1C POC method demonstrated excellent precision, accuracy,
and clinical concordance. With on-going cross-checks, this study demonstrates the
potential for these methods to be used interchangeably; including eAG calculation,
screening for, and diagnosis of diabetes. Replacing venipuncture collection with
POCT was shown to initially, increase compliance with pay-for-performance goals.

B-312
A Whole Blood POC Enzymatic Creatinine Assay that Meets the eGFR
Reporting Requirements by NKDEP

C. Xu, M. Curley, L. Boone, H. Chuang, O. Beeks, H. Visnick.

Instrumentation Laboratory, Bedford, MA
BACKGROUND: Estimated Glomerular Filtration Rate (eGFR) from serum
creatinine is considered a better assessment of renal function than serum creatinine
alone. To assure the reported eGFR is within clinically acceptable error limits, the
NKDEP Laboratory Working Group has recommended the total error (TE) limits for
creatinine measurement in the critical creatinine range (1-1.5 mg/dL). A rapid whole
blood creatinine assay with eGFR in the point of care test (POCT) environment is
gaining attention. This publication assessed the TE of our biosensor based whole
blood creatinine assay per NKDEP recommendations for eGFR capability.
METHODS: Performance of the creatinine assay was assessed based on the
NKDEP Whole Blood Protocol, and was tested on six GEM Premier analyzers after
appropriate modifications to accommodate the creatinine sensors. Heparinized whole
blood samples with creatinine concentrations from 1.0-1.5 mg/dL were collected. The
hematocrit levels were adjusted to form four sample test groups: plasma, low Hct
(20-30%), normal Hct (no adjustment), and high Hct (50-60%). Whole blood samples
were assayed in random order with 10 replicates per sample type per cartridge,
followed by plasma samples (N=10). Three level IDMS traceable reference serum
pools were assayed before and after the test samples. The samples were tested in
parallel on an ABL 800 Flex analyzer (Radiometer) as reference.
RESULTS: The sample results were normalized to the IDMS traceable reference
serum. The mean creatinine was 0.993 mg/dL, the 3.5% bias (95% CI: 3.03%, 4.01%)
vs. ABL, and the 3.5% imprecision (95% CI: 2.8%, 4.2%) were both within the 5%
and 8% limits of NKDEP recommendations. The bias and within-run precision per
cartridge were plotted against the acceptable TE budget for reporting eGFR (Fig.1).

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S221

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

Wednesday, July 30, 2014

Poster Session: 9:30 AM - 5:00 PM
Cardiac Markers

B-314
One year retrospective cTroponin T observations : Women present themselves
at a higher age with ACS.

J. P. van Straalen, J. C. Fischer, M. M. Vis, A. Sturk, R. J. de Winter.

Academic Medical Center, Amsterdam, Netherlands
Background: Generally presumed women present themselves with Acute Coronary
Syndrome (ACS) at higher age than men.
Method: We studied the use of cardiac Troponin T (cTnT) in our patient population
with respect to results, age and gender. cTnT was measured with hs cTnT assay
(Roche Diagnostics) using a limit of detection of 0.005 g/l and an upper limit of
reference value of >0.015 ug/L. Data on cTnt results from a serially sampled patient
population (results, gender and age) were extracted from the laboratory information
system over a one year period for those patients with cardiologists involved in their
medical treatment. In this patient group only the result of the first blood drawing was
incorporated in the dataset investigated.
Results: 1665 patients (703 females (42%), 962 males (58%)) were included. 696 of
these patients had a cTnT< 0.015 g/l and 969 patients a cTnT >0.015 ug/L.
Table1
Conclusion: Our data indicate that women do present themselves at higher age with
respect to men with increased cTnT level as indicator of ACS.
cTroponin T
>0.015 g/l
Female
Male

Age (years)
NMean 25 th percentile 50 th percentile 75 th percentile
patients
364
74
67
76
83
605
67
57
67
79

B-315
Association of Lipid, Inflammatory, Cardiac, and Renal Biomarkers with
C-Reactive Protein in Cardiovascular Risk Categorization - A Factor Analysis
Approach

S. Jovicic1, S. Ignjatovic1, R. Kangrga2, M. Dajak2, N. Majkic-Singh1.

1
Center for Medical Biochemistry, Clinical Center of Serbia, and Faculty of
Pharmacy, University of Belgrade, Belgrade, Serbia, 2Center for Medical
Biochemistry, Clinical Center of Serbia, Belgrade, Serbia
Background. C-reactive protein (CRP) strongly and independently predicts
cardiovascular complications, and its use is recommended for risk assessment in
primary prevention by several institutions. Also, there is evidence of other factors,
contributing to and maintaining the intensity of atherosclerotic processes, which
might identify cardiovascular risk contribution not originated from traditional risk
factors. The aim of this study was to examine, using factor analysis, the nature of
influence of biomarkers of inflammation, lipid metabolism, renal, and cardiac function
on cardiovascular risk, and their possible connection and relations to CRP values.
Methods. Principal component analysis was used to investigate clustering of
inflammatory markers [serum amyloid A (SAA), fibrinogen, 1-acid glycoprotein
(A1AGP), haptoglobin, C3 and C4 complement components], lipid metabolism
[total, HDL, non-HDL and LDL cholesterol, triglycerides, apolipoprotein A-I (apo
A-I), apolipoprotein B (apo B), lipoprotein (a) (Lp(a))], renal [creatinine, cystatin C
(Cys-C), estimated glomerular filtration rate (eGFR)], cardiac function [N-terminal
pro-natriuretic peptide type B (NT-proBNP), high sensitivity cardiac troponin T
(hs-cTnT)], and traditional cardiovascular risk factors [age, gender, body mass
index (BMI), systolic blood pressure (SBP)], obtained from 242 apparently healthy
individuals.

and non-HDL cholesterol); 4) ,,hemodynamic factor (age and NT-proBNP); and 5)

,,metabolic factor (triglycerides and HDL cholesterol). The Kaiser-Meyer-Olkin
measure of sampling adequacy was 0.75. In multiple regression analysis, five factor
model had the best predictive value for CRP concentrations >1 mg/L (OR 6.53, 95%
CI 4.06-10.50, P<0.0001 for systemic inflammation; OR 1.44, 95% CI 1.04-2.00,
P=0.028 for ,,cardiorenal factor; OR 1.76, 95% CI 1.23-2.5, P=0.002 for ,,atherogenic
cholesterol; OR 1.91, 95% CI 1.33 - 2.73, P<0.0001 for ,,hemodynamic factor;OR
1.90, 95% CI 1.33 - 2.73, P<0.0001 for ,,metabolic factor), while ,,cardiorenal
factor and ,,atherogenic cholesterol completely lost their significance (P>0.05) for
predicting CRP concentrations >2 mg/L and >3 mg/L. The ability of the factor-based
logistic regression model was compared with multivariable logistic model containing
all 25 variables in predicting the presence of CRP concentrations >1 mg/L, >2 mg/L,
and >3 mg/L. The area under the receiver operator characteristics curve (AUC) of the
five factor model was 0.889 and was not statistically significantly different from the
25 variable model (AUC=0.922) (P=0.2113). However, the differences between the
two models examined were statistically significant in predicting the values of CRP>2
mg/L and CRP>3 mg/L.
Conclusion. Systemic inflammation, cardiorenal function, atherogenic lipid
profile, hemodynamic and metabolic status might independently contribute to the
pathophysiology of chronic, subclinical inflammation in atherosclerosis. They might
represent underlying dimensions accompanying the elevation of CRP concentration
and increased cardiovascular risk.

B-316
Verifying A Cut-Off Value for the Beckman TnI+3 Assay on the DxI 800 and
Access-2 Analyzers.

V. PATEL, J. Qiu, J. Wang, A. Khan, J. Ekpe, M. Sydlowska, B. Goldsmith.

TEMPLE UNIVERSITY HOSPITAL, PHILADELPHIA, PA
Background: Cardiac troponin (cTn) assays have been available in clinical laboratories
for nearly two decades and considered a highly sensitive marker for myocardial
damage. An elevation of cTn is used, together with other diagnostic criteria, to rule in/
out a myocardial infarction (MI). Laboratories measure either cTnI or cTnT isoforms
of troponin. Following a recall of cTnI reagents from the DxI Immunoassay analyzer
in October 2010, Beckman Coulter recently re-introduced a Troponin-I (AccuTnI +3)
assay for the DxI 800 and Access-2 analyzers. We evaluated whether the stated cutoff
determined by the manufacturer was appropriate for our patient population.
Design: We measured plasma cTnI concentrations in 94 patients presenting to our
Emergency Department (ED) in whom no history or evidence of cardiac disease was
found following a review of each patients chart. Our population consisted of 30 Males
(ages 16-63 years) and 64 females (ages 5-54 years). Specimens were spun, aliquoted,
frozen within 24 hours at -20o C, and analyzed within 30 days of collection. Where a
history could not be determined results were not included. cTnI was measured using
the AccuTnI+3 assay (Beckman Coulter, Brea, California) on both the DxI 800 and
Access-2 Immunoassay Systems. EP Evaluator (Data Innovations, Burlington, VT)
and Excel spreadsheet calculations were used to evaluate the data.
Results: cTnI results for the DxI 800 showed (ng/mL): Males range=0.00-0.76, mean
0.03; Females range=0.00-0.36, mean 0.02. For the Access-2: Males range=0.00-0.86,
mean 0.04; Females 0.00-0.50, mean 0.02. We determined the cutoff for plasma
specimens assayed on the DxI 800 of 0.03 ng/mL and for the Access-2 of 0.04 ng/mL.
The ranges (males and females) demonstrate that there were results on non-cardiac
patients above the cutoff.
Conclusions: The 99th percentile at a coefficient of variation (CV) of <10% is used
to define the cutoff concentration between a non-MI and a MI event. Though the
manufacturer determined a cutoff of 0.2 ng/mL, 0.04 ng/mL was selected to provide
one consistent cutoff for our hospital. We suggest using non-cardiac specimens that
reflect the population to determine an appropriate cutoff. Caution must be used to
ensure, as best as possible, exclusion of patients with a history or presenting symptoms
of cardiac disease.

Results. Factor analysis identified five clusters, i.e. principal components (factors),
which explained 65.3% of the total variance (29.0% factor 1, 13.2% factor 2,
9.0% factor 3, 8.5% factor 4, and 5.6% factor 5). Based on factor loading of
0.5 clasters were interpreted as 1) ,,systemic inflammation (fibrinogen, SAA,
A1AGP, haptoglobin, C3 and C4 complement components); 2) ,,cardiorenal factor
(creatinine, uric acid, Cys-C, hs-cTnT and gender); 3) ,,atherogenic cholesterol (LDL

S222

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

B-317
Short Term Variation Important In Evaluating Goodness of Troponin Assays
For Diagnosing Myocardial Damage

G. S. Cembrowski1, A. R. Cembrowski2, A. N. Kunst1, D. T. Holmes3, F.

S. Apple4. 1Alberta Health Services, Edmonton, AB, Canada, 2Department
of Ecology and Evolutionary Biology, University of Toronto, Toronto, ON,
Canada, 3Department of Pathology and Laboratory Medicine, University
of British Columbia, Vancouver, BC, Canada, 4Department of Laboratory
Medicine and Pathology, Hennepin County Medical Center, Minneapolis,
MN
Background: While total imprecision is usually used to evaluate the quality of troponin
assays, most diagnoses of myocardial necrosis are based on serial troponin changes
within 8 to 12 hours. We have used a novel method to isolate the short term variation
in sequential, intra-patient results, a mixture of intra-individual biologic variation (sb)
and co-existing analytic variation (sa). We have used this method to derive the short
term variation of two troponin assays, the Ortho VITROS Troponin I ES and the
Beckman Coulter AccuTnI.
Methods: Two different data sets were extracted: all of the patient troponins that
were measured by the main Beckman DxI analyzer at University of Alberta Hospital
over a 20 month period and serial VITROS troponins of 1271 patients in whom the
diagnosis of myocardial infarction (MI) was being considered (measured on one of
two VITROS systems over 6 months at Hennepin County Medical Center). For both
clinical environments, patient bloods were collected into lithium heparin. Two subsets
of patient results were studied and compared, those patients whose serial troponins
were all less than 41 ng/L and patients with serial troponins all less than 61 ng/L. We
tabulated the pairs of intra-patient troponins that were separated by 2 to 3h, 3 to 4h,
4 to 5h, 5 to 6h, 6 to 7h, 8 to 9h, and 9 to 10h. The standard deviations of duplicates
(SDD) of the paired troponins were calculated for each time interval. The graphs of
SDD vs. time interval were approximately linear; the y intercept (y0) provided by the
weighted linear regression equation represents the sum of sa and sb with y0 = (sa2
+sb2)1/2.
Results: For patients with series of troponins under 41 ng/L (average of 14 and 13 ng/L
for Beckman and Ortho, respectively), the average short term variation (including
biologic variation) was 3.0 ng/L (SEM=0.8 ng/L) for the Beckman and 1.1 ng/L
(SEM=0.3 ng/L) for the Ortho. For patients with series of troponins under 61 ng/L
(average of 18 and 15 ng/L for Beckman and Ortho), the short term variation was 3.0
ng/L (SEM=1.3 ng/L) for the Beckman and 2.2 ng/L (SEM=0.6 ng/L) for the Ortho.
Conclusion: Perhaps not surprisingly, the total short term variation (including biologic
variation) is approximately equal to the short term variation derived from running
low level quality control materials. The biologic variation of troponins for patients
with healthy myocardiums is thus very low. For the Beckman DxI, the short term
variation (3 ng/L) at a level of 25 or 35 ng/L is considerable. If +/-3SD limits are used
(99.7% confidence limits), then variation around these levels is +/- 9 ng/L, variation
that can cause both false positive and false negative interpretations. For the Ortho
VITROS, there will be fewer misinterpretations. We recommend that serial patient
troponin values be collected and mined for total variation. The analytical systems
that minimize the short term total variation are most useful diagnostically.

comparison studies between the PON1 method commonly available for PON1 assay
and a similar non-ELISA commercially available PON1 kit method have resulted
in a weak Spearman correlation displaying a coefficient R = 0.40 for the range
104.9- 245.6 U/L. Conclusion: The current study will bridge some of the gaps on our
understanding to the enzyme performances. The outcome should encourage additional
studies in clinical setting to investigate other missing aspects of the factors known to
affect PON1 enzyme function and performance.

B-319
Assessing the incremental value of additional biomarkers versus a highsensitivity cardiac troponin I assay for predicting a short-term serious cardiac
outcome in an early chest pain population

C. Shortt1, G. Pond1, K. Phan2, K. Sohn3, A. Worster1, S. Hill1, P. A. Kavsak1.

1
McMaster University, Hamilton, ON, Canada, 2McGill University,
Montreal, QC, Canada, 3University of Toronto, Toronto, ON, Canada
BACKGROUND: Patients presenting with chest-pain to the emergency department
(ED) who are at risk for a cardiovascular event in the short-term, must be identified.
With the advent of high-sensitivity cardiac troponin (hsTn) assays, most ED patients
will have measurable cardiac troponin concentrations. It is unclear whether additional
biomarkers of myocardial stress, inflammation, vascular or other endocrine functions
can improve prognostic ability when compared to hsTnI alone.
METHODS: After ethics approval, the presentation EDTA plasma sample (stored
-800C) from patients (onset of chest-pain within 6h) in the RING study (Clin Chem Lab
Med 2011;49:19158) were measured for these analytes (platforms):hsTnI, insulin,
myeloperoxidase, 25-OH vitaminD (Abbott:ARCHITECT); IL-6 (Beckman:Access);
PlGF, sFlt-1, proBNP (Roche:Elecsys); EGF, NTproBNP, E-selectin, P-selectin,
sICAM-3, Thrombomodulin, IL-6, IL-10, MCP-1, bFGF, PlGF, sFlt-1, VEGF (MSD:
Multi-Array:); IL-2, IL-4, IL-6, IL-8, IL-10, VEGF, IFNgamma, TNFalpha, IL-1a,
IL-1b, MCP-1, EGF (Randox:Evidence). Using MedCalc, Statsdirect, and R program,
we performed ROC curve analyses to select the top three biomarkers with the highest
AUC for the composite outcome (percutaneous intervention, coronary artery bypass
surgery, significant arrhythmia, refractory ischemic pain, heart failure, myocardial
infarction, stroke, cardiac arrest, or death) at 72h and combined these three biomarkers
with hsTnI to assess if inclusion of these biomarkers increased the AUC.
RESULTS: Of 140 patients (median age (interquartile range)=60y (49-68); 65%
male), 23 had a composite outcome. Abbott hs-cTnI by itself had an AUC of 0.88
(95%CI:0.82-0.93) for prediction of the composite outcome. The top three biomarkers
were Roche proBNP (AUC=0.73; 95%CI:0.65-0.81); Roche PlGF (AUC= 0.71;
95%CI:0.63-0.78); and Abbott insulin assay (AUC=0.70; 95%CI:0.61-0.77). The
addition of insulin, PlGF and proBNP were not significant (p-values of 0.94, 0.36 and
0.12 respectively) and lowered the AUC (0.86) for the Abbott hsTnI assay.
CONCLUSIONS: The Abbott hsTnI assay alone, provides superior short-term
prognostic utility in patients presenting early after chest-pain.

B-318
Paraoxonase-1 enzyme activity assay for clinical samples: Validation and
correlation studies

M. Garelnabi1, A. Younis2. 1University of Massachusetts Lowell, Lowell,

MA, 2Department of Obstetrics and Gynecology, Mercer University School
of Medicine, & Central Georgia Fertility Institute, Macon, GA
Introduction: Paraoxonase-1 (PON1) enzyme is reported in various types of tissues
and linked to numerous pathophysiological disorders; representing a potential
biomarker in many pathological conditions such as cardiovascular diseases. Methods:
We conducted several small studies to evaluate PON1 performances affected by
sample types, storage, and interferences. In addition; we carried out some limited
studies to compare the performance of the clinical research widely used PON1
assay to similar commercially available PON1 kit assay method. Results: Type of
anticoagulant studies have shown that samples collected in NaF, citrate, ETDA
clot activator, and sodium heparin have increased PON1 levels 49%, 24.5% and
19.8%; 11.4% and 8% respectively compared to the serum. Whereas samples in
lithium heparin have 10.4% decreases in its PON1 levels compared to the serum.
Biological interference such as hemolysis has little effect on PON1 levels; however
samples spiked with lipids have shown 13% reduction on PON 1 levels. The method

B-320
A high sensitive Homocysteine (hsHCY) assay improves this predictive
biomarkers clinical value

G. Liang1, A. Jaffe1, J. Larkin2, R. Jaffe1. 1Health Studies Collegium,

Ashburn, VA, 2ELISA/ACT Biotechnologies, Sterling, VA
Plasma Homocysteine (HCY) levels predict accelerated development of heart and
blood vessel disease and HCY is one of the few tests that predict all cause morbidity

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S223

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

and mortality. A significant practical limitation to HCY testing has been the leakage
into plasma of red cell HCY starting within minutes after phlebotomy severely
limiting community clinical use of HCY assays. A more stable specimen with reduced
variability is needed, especially at lower, more predictive levels. Strategic, clinical,
and lab results for an hsHCY assay that addresses this need is reported here.
Specimens (N=30) were collected in tubes containing EDTA (current preferred
specimen) and a novel cell preservative EAB. Specimens were stored at 41C until
analysis. Plasma was isolated and analyzed at 0.5, 4, 8, 24, 48, 72, and 96 hours
after phlebotomy using a Diazyme Analyzer 700. The data of HCY values were
statistically analyzed by one-way ANOVA.
HCY levels started to increase within 30 minutes in EDTA specimens while the EAB
cell preservative sample values were stable for at least 48 hours. Indeed initial HCY
levels in EDTA samples averaged 32% higher than ones in the EAB preservative
(p<0.0001). The initial HCY variance in EDTA is 10.23 2.23, while the variance in
EAB is 7.63 1.60. This suggests that HCY is released into plasma from red blood
cells in EDTA samples starting minutes after phlebotomy and does not occur in EAB
cell preserved specimens over 48 hours.
Initial and more precise HCY can be measured and maintained for at least 48 hours
when the EAB cell preservative is used. HCY increase appears to starts within minutes
after blood draw in EDTA but not EAB preservative. The EAB cell preservative
allows a more stable and accurate HCY assay for this important predictive biomarker.
More studies of high sensitivity HCY (hsHomocysteine) are needed to further validate
this predictive biomarker clinically and globally.

B-321
Development of a New Latex-Enhanced Immunoturbidimetric Assay for the
Determination of Type IIA Secretory Phospholipase A2 (sPLA2), a Biomarker
of Increased Cardiovascular Risk

G. Shanbhag, P. Lowry, M. Benchikh, A. Chacko, R. McConnell, S.

FitzGerald. Randox Laboratories Limited, Crumlin, United Kingdom
Background: Secretory phospholipase A2 (sPLA2) enzymes are biomarkers of
increased cardiovascular risk and are targets of emerging therapeutic agents. They are
associated with incident coronary artherosclerosis in healthy men and women and with
recurrent adverse cardiovascular events in patients with acute coronary syndromes.
sPLA2-IIA is a member of the phospholipase A2 family and is a potent biomarker for
cardiovascular risk assessment. It is widely expressed in hepatocytes, macrophages,
platelets and vascular smooth muscle cells and is up regulated in response to proinflammatory compounds such as interleukin-1, interleukin-6, tumor necrosis
factor-, interferon- and oxidized low-density lipoprotein (LDL). The objective of
this study was to develop a new quantitative, latex-enhanced immunoturbidimetric
immunoassay to determine levels of sPLA2-IIA in human serum / plasma which have
prognostic value in patients with Coronary Heart Disease (CHD) and can be employed
to assess risk of future cardiovascular events. The assay is applicable to a variety of
automated, clinical analyser systems, which ensures the reliability and accuracy of the
measurements and facilitates the testing procedure.
Methods: The assay is a latex-enhanced immunoturbidimetric assay based on the
principle of measuring changes in scattered light. The latex particles are coated with
anti-sPLA2-IIA antibodies, which in the presence of sPLA2-IIA rapidly agglutinate.
When a sample containing sPLA2-IIA is introduced, the agglutination reaction is
initiated and the change in scattered light is measured as a change in absorbance that
is directly proportional to the concentration of sPLA2-IIA in the sample. The assay
is applicable to a variety of automated systems. A correlation study was conducted
using a commercially available enzyme immunoassay. Results:The assay presented
a limit of detection of 15 ng/mL sPLA2-IIA (analytical range: 0 to 500 ng/mL
sPLA2-IIA). The within-run precision and the total precision expressed as %CV, were
typically <5% and <8% respectively. In the correlation study, citrate plasma samples,
spanning the analytical range, were tested with this immunoturbidimetric assay and a
commercially available enzyme immunoassay. Linear regression on the resulting data
generated an r-value >0.9.
Conclusion: The data generated indicates that this latex-enhanced immunoturbidimetric
assay is applicable to the detection of sPLA2-IIA in human plasma and serum. The
assay is of value as a new analytical tool for assessment of cardiovascular risk in
clinical settings. Its applicability to a variety of automated clinical analysers ensures
the reliability and accuracy of the measurements and facilitates the testing procedure.

S224

B-324
Comparison of hs-cTnT and a conventional cTnI assay for the detection of
ischemia induced myocardial injury and type 4 myocardial infarction post PCI.

V. Van Hoof, A. Vorlat, D. Coenen, R. Hammami, S. van Kerckhoven, J.

Bosmans, S. Haine, T. Vandendriessche, C. Vrints, M. Claeys. University
Hospital Antwerp, Edegem, Belgium
Background: comparison of a high sensitive troponin T (hs-cTnT) assay with a
conventional troponin I (cTnI) assay, for the follow-up of patients post-percutaneous
coronary intervention (PCI), more particular for the detection of ischemia induced
myocardial injury (IIMI) and type 4 myocardial infarction (type 4 MI).
Methods: PCI and stenting was performed in 103 stable cardiac patients with
significant coronary artery stenosis. hs-cTnT (Modular, Roche) and conventional
cTnI (Dimension Vista, Siemens) were measured at 4 time points (0, 90, 180 and 360
min post-PCI). IIMI during stenting was defined as at least one cTn value > upper
reference limit (URL = 99th percentile of the reference population) and an absolute
rise of >50% of the 99th pctl URL in the 360 min blood sample. Type 4 PCI MI
was defined in the 360 min blood sample according to the Third Universal Definition
of Myocardial Infarction: elevation of cTn values (>5 x 99th pctl URL) in patients
with normal baseline values (=<99th pctl URL) or a rise of cTn values >20% if the
baseline values were elevated and were stable or falling (Thygesen et al Circulation
2012;126:2020-2035).
Results: both assays correlated fairly well (r= 0,9081, 95%CI 0,8898 to 0,9235). IIMI
during stenting was detected in 68 patients (66%) with the hsTnT assay and in 69
patients (67%) with the TnI assay. The number of patients who answered the criteria
for IIMI at 90 min and at 180 min was significantly less (18.2% for TnI and 20.9% for
hsTnT at 90 min; 27.3% for TnI and 38.2% for hsTnT at 180 min). A frequency table
for IIMI showed agreement between both methods for 31 patients without IIMI and
for 65 patients with IIMI (93% concordance). Discrepancies were found for 7 patients
(3 positive for hsTnT but negative for TnI and 4 positive for TnI but negative for
hsTnT). Chi square test and Fishers exact test were very significant (p<0.0001). The
contingency coefficient between both methods was 0.637. Type 4 MI was detected
in 33 patients (32%) with the hsTnT assay and in 31 patients (30.1%) with the TnI
assay. A frequency table for type 4 MI showed agreement between both methods for
61 patients without type 4 MI and for 22 patients with type 4 MI (81% concordance).
Discrepancies were found for 20 patients (11 positive for hsTnT but negative for TnI
and 9 positive for TnI but negative for hsTnT). Chi square test and Fishers exact test
were very significant (p<0.0001). The contingency coefficient between both methods
was 0.465.
Conclusion: myocardial injury caused by a short period of myocardial ischemia
during PCI was detected with comparable sensitivity by the hsTnT assay and
by the conventional TnI assay. For both assays, values at 90 min and at 180 min
underestimated the presence of IIMI as established at the 360 min time point. There
were more discrepancies between both assays for the detection of type 4 MI post PCI.

B-325
Quality Assessment and Reagent Lot-to-Lot Consistency in High-Sensitivity
and Contemporary Troponin T Assays

A. K. Saenger1, U. Klause1, A. Ziegler2, S. Menassanch-Volker2, K.

Hallermayer3. 1Roche Diagnostics, Indianapolis, IN, 2Roche Diagnostics,
Rotkreuz, Switzerland, 3Roche Diagnostics, Penzberg, Germany
Background: Introduction of high-sensitivity cardiac troponin T (hs-cTnT, Roche
Diagnostics) in 2009 outside the US has facilitated and expedited earlier diagnosis
of NSTEMI and demonstrated prognostic value in a variety of patient populations.
Accompanying this trend are new analytical challenges with hs-cTn assays due to
the significant impact small analytical confounds may have in the diagnosis of AMI
and interpretation of serial troponin results. Lot-to-lot reagent stability for hs-cTn,
particularly at low concentrations, is of utmost importance.
Objective: Assess lot-to-lot performance of the hs-cTnT and contemporary cTnT
assays across multiple reagent lots, instruments and targeted at various cTnT
concentrations.
Methods: For the period between 2008-2014 (cTnT Gen. 4) and 2013-2014 (hscTnT/hs-cTnT-STAT), we reviewed all cTnT reagent lot data (Roche Diagnostics,
Penzberg). Acceptability of reagent performance utilized protocols designed to assess
variability and bias between reagent lots and instruments. Residual pooled patient
specimens were used in all assessments of reagent lot-to-lot performance. Data were
analyzed collectively and validated between multiple platforms (Cobas e411/e601/
e602; Elecsys 2010; Modular E170).

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

Results: Unique reagent lot data were available for the hs-cTnT, hs-cTnT-STAT and
cTnT Gen.4 assays (n = 3, 5 and 42, respectively). Lot-to-lot reagent performance
demonstrated minimal bias across the reportable range for hs-cTnT (0-1%), hs-cTnTSTAT (3-17%) and cTnT (5-7%). Very low bias at the 99th percentile of the hs-cTnT
assay (14 ng/L) with the STAT (range: 13.00-13.09 ng/L) and routine assays (13.9113.98 ng/L) was observed. The largest bias for hs-cTnT was noted at the LoD (1317% at 5 ng/L).
Conclusion: Excellent lot-to-lot comparability was achieved with the hs-cTnT
and cTnT assays, allowing laboratories to confidently integrate the assays into
their clinical AMI decision making protocols. Assurance of reagent processes and
transparency regarding performance criteria is critical for implementation of troponin
and interpretation in the context of serial sampling and biological variability.

Conclusion: Cystatin-C is one of the promising early risk marker for the Coronary
Artery Disease patients. Importance of Cystatin-C as one of the biomarker for
Coronary Artery Disease patients, positive correlation with cholesterol and body mass
index will be discussed.
Mean and standard deviation values of Cystatin-C and Lipid Profile
Mean
Mean SD
t-value
p-value
(Controls)
Cases)
Cholesterol
162.5 25.9830 201.9 56.8112 5.37
< 0.0001
HDL-C
43.1818 .7746 39.0138 .6248 -5.22
< 0.0001
LDL
119.3 24.9667 161.9 54.0915 6.10
< 0.0001
Cholesterol
Cystatin-C
0.9738 0.2067 1.3883 0.3822 8.27
< 0.0001
23.3668
24.3015
BMI
2.00
0.0467
3.0605
3.1302
87.8939
93.1778
WC
4.15
< 0.0001
7.7819
8.8080

Parameter

B-327
Diagnostic Accuracy of the Trinity Biotech Meritas Cardiac Troponin I Point
of Care Assay

M. M. Murakami, Y. Sandoval, S. W. Smith, K. M. Schulz, K. M. Smith,

P. R. Stack, B. J. Kilburn, F. S. Apple. Hennepin County Medical Center,
Minneapolis, MN
Background: The diagnosis of acute myocardial infarction (MI) is based on clinical
factors and an increased cardiac troponin (cTn), with a rising and /or falling cTn
pattern required. In addition, utilizing point-of-care (POC) technology to measure cTn
may assist in a more rapid management of patients presenting to rule in or rule out MI.
The goal of this study was to validate the diagnostic accuracy of the Trinity Biotech
Meritas POC cTnI assay based on the 99th percentile value (19 ng/L).

B-326
Cystatin c as biomarker for Coronary Artery Disease patients in southern
India.

P. Srilakshmi1, R. Kondreddy1, V. Madrol1, A. R. Saeid2, S. J. Dhoipode2,

S. Shakila2, H. A. Beloch2, A. M. Jarari3, L. T. Peela4, J. R. Peela2.
1
Department of Biochemistry, Mamata Medical College, Khammam,
India, 2Faculty of Medicine,Quest International University Perak, Ipoh,
Malaysia, 3Department of Biochemistry, Faculty of Medicine,University
of Benghazi, Benghazi, Libyan Arab Jamahiriya, 4Great Eastern Medical
School, Srikakulam, India
BACKGROUND: Coronary Artery Disease is the leading cause of mortality and
morbidity across the globe. Prevalence of Coronary Artery Disease is alarmingly
increasing in developing countries. India is also experiencing the same with the
increasing urbanization, changing lifestyles and obesity. Present study is focused on
estimation of Cystatin-C in Coronary Artery Disease patients in correlation with lipid
profile, obesity and other risk factors.
Materials and methods: The study was conducted in the department of biochemistry,
Mamata Medical College and General Hospital, Khammam, Andhra Pradesh, India.
The patients attending outpatient and wards of cardiology and general medicine
departments of hospital and local cardiac centers were included in this study. Study
group comprised of 145 patients diagnosed as Coronary Artery Disease with an age
group of 30-50 patients with associated risk factors i.e., Diabetes, hypertension,
smoking. And alcoholism. 66 sex and age matched subjects were recruited as control
group (non Coronary Artery Disease cases) using the same criteria.
All patients and controls were measured cystatin-C and lipid profile by using authentic
methods available. These patients were also divided as per their Body Mass Index
with associated risk factors.
Results: In present study, there is significant increase in the values of serum
cholesterol (p<0.0001), Low density lipoprotein cholesterol (0.0001) and significant
decrease in High density lipoprotein cholesterol (p<0.0001) level was observed in
Coronary Artery Disease cases when compared with controls. Cystatin-C (p<0.0001)
was significantly elevated in people with Coronary Artery Disease when compared
with controls. The levels of cystatin-C is correlating positively with total cholesterol,
low density lipoprotein cholesterol, Body Mass Index and waist circumference.

Methods: 293 Patients presenting with symptoms suggestive of myocardial ischemia

presenting to Hennepin County Medical Centers emergency department with cTnI
orders based on clinical indications were evaluated in this study. Plasma (EDTA) was
obtained at 0h (baseline), and 6h. cTnI was measured by the Meritas cTnI assay (LoD,
12 ng/L; 10% CV at 24 ng/L). All charts were adjudicated for MI, predicated on
the Third Universal Definition of Myocardial Infarction guidelines based on the 99th
percentile of the Abbott ARCHITECT cTnI assay used routinely in the hospital.
Results: MI was diagnosed in 8.6% (n=25) patients. MIs were comprised of 15 type
1 and 10 type 2. Clinical sensitivity improved over the time of serial testing, from
64.0% at 0h to 88.0% over 6h. Specificity was 88.8% at 0h and 86.8% over 6h;
with ROC AUC values of 0.84 and 0.94 at 0h and 6h, respectively. In a subset of 92
patients in which EDTA whole blood was measured in parallel to plasma, similar ROC
AUCs were found, 0.96 vs. 0.97, for maximum cTnI over 6h (p=0.43). Conclusion:
Our findings demonstrate that the Trinity Biotech Meritas POC cTnI assay is a
diagnostically useful aid in ruling in and ruling out acute MI in an emergency room
setting.

B-328
Improved Diagnostic Accuracy for Myocardial Infarction of the Abbott
ARCHITECT High Sensitivity Assay Compared to the Contemporary cTnI
assay in an unselected US Population

Y. Sandoval, M. M. Murakami, S. A. Love, K. M. Schulz, R. Ler, J.

Nicholson, S. W. Smith, F. S. Apple. Hennepin County Medical Center,
Minneapolis, MN
Background: High-sensitivity (hs)-cTn assays have not yet been cleared for clinical
use in the US. This study compared the diagnostic accuracy for myocardial infraction
(MI) of the Abbott ARCHITECT hs-cTnI assay to the Abbott ARCHITECT
contemporary cTnI assay.
Methods: Retrospective analysis of 310 unselected patients with symptoms suggestive
of acute coronary syndrome (ACS), in which serial cTnI measurements were
obtained on clinical indication. Fresh EDTA plasma specimens were simultaneously
measured with both the contemporary and hs-cTnI assays. Unique to this study was
adjudication of MI [using the 3rd Universal Definition of MI subtype (type1 and type
2) classification] independently predicated on the 99th percentiles of a normal
population for a) the contemporary cTnI assay (99th percentile, 30 ng/L), b) the hscTnI assay (overall 99th percentile, 26 ng/L), and c) the gender-specific hs-cTnI (99th
percentiles: male 34 ng/L and female 16 ng/L). Sensitivity, specificity, and positive
and negative predictive values were calculated using baseline and maximum values.
Results: 24% Fewer MIs were identified based on the hs-cTnI assay: 33 (10.7%) using
the overall hs-cTnI assay cutoff and 32 (10.3%) using the gender-specific cutoffs

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S225

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

compared to 43 (13.9%) using the contemporary cTnI assay. Using the hs-cTnI assay,
type 1 MIs decreased from 14 to 11 and type 2 MIs decreased from 29 to 22. The table
demonstrates our diagnostic accuracy findings. ROC areas under the curve showed
that baseline diagnostic accuracy was improved using the hs-cTnI assay: overall
cutoff 0.734; female cutoff 0.763, male cutoff 0.705; compared to the contemporary
cTnI assay: 0.691.
Conclusions: Our data demonstrate that the hs-cTnI assay would not result in
an over diagnosis of MI, and with careful adjudication, appears to result in fewer
misclassifications. The hs-cTnI assay demonstrated superior specificity, with the best
diagnostic accuracy in females. Future outcomes studies will be important.
hs-TnI and cTnI Sensitivity & Specificity
Troponin
Sensitivity %
Specificity %
NPV
Assay (99th
PPV % (95%CI)
(95%CI)
(95%CI)
%(95%CI)
percentile)
Baseline cTnI
90.2 (85.1,
55.8 (39.9, 70.9)
65.5 (59.5, 71.2) 20.7 (13.7, 29.2)
(>30 ng/L)
94)
Baseline hs93.1 (88.9,
54.6 (36.4, 71.9)
73.2 (67.6, 78.3) 19.6 (12.0, 29.2)
TnI (>26 ng/L)
96.1)
Baseline
Gender93.4 (89.2,
Specific hs-TnI 56.3 (37.7, 73.6)
71.2 (65.5, 76.5) 18.4 (11.2, 27.5)
96.3)
(M>34 and
F>16 ng/L)
Maximum
100.0
cTnI (>30
100.0 (91.8, 100.0) 58.1 (51.9, 64.0) 27.7 (20.9, 35.5) (97.7,
ng/L)
100.0)
Maximum
100.0
hs-TnI (>26
100.0 (89.4, 100.0) 68.8 (63.0, 74.3) 27.7 (19.9, 36.7) (98.1,
ng/L)
100.0)
Maximum
Gender100.0
Specific hs-TnI 100.0 (89.1, 100.0) 65.8 (59.9, 71.4) 25.2 (17.9, 33.7) (98.0,
(M>34 and
100.0)
F>16 ng/L)

B-329
Zonulin as a potential biomarker of metabolic inflammation and pulmonary
endothelial permeability

T. B. Dschietzig1, R. Kluesener2, F. P. Armbruster1, C. Melzer2.

1
Immundiagnostik AG, Bensheim, Germany, 2Dept. of Cardiology, Charit
Berlin (Campus Mitte), Berlin, Germany
Background: In obesity and metabolic syndrome, disturbed intestinal permeability
and low-grade chronic systemic inflammation appear to act in a vicious circle called
metabolic inflammation. Zonulin, a tight junctions modulator and key regulator
of intestinal permeability, has been shown to be up-regulated in individuals with
type-1 diabetes and to play a role in gut-related dysfunctional auto-immunity. In
addition, there are some preliminary reports indicating a possible association between
zonulin and metabolic inflammation or type-2 diabetes as well. Moreover, zonulin
is implicated in the regulation of general endothelial/epithelial permeability, and its
association with increased pulmonary permeability has been demonstrated in animal
experiments.
Methods: This study aimed at investigating plasma zonulin and its dependence
on various clinical and biochemical factors in 225 patients carrying automatic
implantable cardioverters/defibrillators (AICD), with 75% of them suffering from
systolic heart failure, 69% from coronary artery disease (CAD), and 27% from type-2
diabetes (T2D).
Results: Univariate linear regression analysis showed that zonulin levels were
associated with plasma creatinine, plasma nitrotyrosine, severity of CAD, left
ventricular ejection fraction, and NYHA functional class, but not with high-sensitivity
C-reactive protein (hsCRP), body mass index, weight, height, sex, or age. After
multiple linear regression analysis, the negative association with creatinine (p =
0.006) and the positive one with NYHA class (p = 0,013) remained significant. In the
subgroup of individuals with T2D, multiple regression revealed a significant positive
affection of zonulin by hsCRP only (p = 0.025).
Conclusion: These findings may support reports on zonulins involvement in the
phenomenon of metabolic inflammation in T2D patients. The association of zonulin
with NYHA may reflect its newly established role in altering endothelial/pulmonary
permeability in heart failure. The robust
negative correlation with creatinine is unexpected and needs further clarification in
experimental and clinical studies.

S226

B-330
A Multi-Center Analytical Evaluation of the ARCHITECT STAT High Sensitive
Troponin-I Assay

M. Krintus1, P. Boudry2, N. Capell3, U. Koller4, G. Lefevre5, L. Lennartz6,

J. Lotz7, M. Nybo7, M. Plebani8, J. Shih9, O. Skadberg10. 1Nicolaus
Copernicus University, Collegium Medicum, Bydgoszcz, Poland, 2CHR
Mons-Hainaut, Mons, Belgium, 3Dr. Peset University Hospital, Valencia,
Spain, 4Vienna Hospital Association, Hospital Hietzing, Vienna, Austria,
5
AP-HP, Hopital Tenon, Paris, France, 6Abbott Laboratories, Wiesbaden,
Germany, 7Odense University Hospital, Odense, Denmark, 8University
Hospital, Padova, Italy, 9Abbott Laboratories, Abbott Park, IL, 10Stavanger
University Hospital, Stavanger, Norway
Introduction: Troponin is the preferred biomarker for the diagnosis of acute myocardial
infarction in the presence of symptoms of ischemia, with the recommended cutoff
at the upper reference limit (URL) or 99th percentile value with precision of <10%.
Troponin assays have improved in sensitivity to meet these guidelines.
Objective: This study evaluated the analytical performance of a new high sensitivity
troponin-I assay on the ARCHITECT instrument to confirm the data provided by the
manufacturer.
Methods: The ARCHITECT STAT high sensitive Troponin-I (hsTnI) assay is a double
monoclonal, sandwich assay with chemiluminescent detection. Nine laboratories
in Europe participated in this study using either ARCHITECT i2000SR or i1000SR
instruments in their routine laboratories. Total precision, limit of blank (LoB), Limit
of detection (LoD), limit of quantitation (LoQ), linearity and interference were
determined following guidance from CLSI documents EP5-A2, EP17-A, EP6-A,
and EP7-A,respectively. Method comparison was performed using the ARCHITECT
STAT Troponin-I assay(contemporary TnI) as the referent method and 2598 samples
that spanned the dynamic range of the assays. The 99th percentile URL was determined
using 1769 samples from healthy populations (screened with a questionnaire) or blood
donors from seven countries. Ethics approval or waiver was received for specimens
collected for the reference interval and method comparison studies.
Results: Precision ranged from 1.5 to 8.9% using the manufacturers controls. The
LoB, LoD and LoQ ranged between 0.1-1.4, 0.5-2.1 and 4.6-8.5 ng/L, respectively,
confirming the information in the package insert. The lowest concentrations
corresponding to a total CV of 10% was 5.6 ng/L. Common interferences did not
affect the hsTnI results. The overall 99th percentile URL was determined to be 19.3
ng/L and was higher in men (27.0 ng/L) than in women (11.4 ng/L). Troponin was
detectable in between 52.1 and 87.8% of the apparently healthy population depending
on the LoD value used. Concordance between the investigated hsTnI assay and the
contemporary TnI assay at the 99th percentile cutoff was found to be 95%.
Conclusions: These results demonstrate that the ARCHITECT STAT high sensitive
troponin-I assay is a precise and highly sensitive method for measuring troponin I
on a high throughput analyzer. This new assay meets the criteria of a high sensitivity
troponin test with the 10% CV concentration below the 99th percentile URL.:

B-331
Comparison of sCD14-ST (Presepsin) with Eight Biomarkers for Mortality
Prediction in Patients Admitted with Acute Heart Failure

E. Spanuth1, G. Hess2, E. Giannitsis3. 1DIAneering - Diagnostics

Engineering & Research, Heidelberg, Germany, 2Prof Hess Medical
Consulting GmbH, Mainz, Germany, 3Department of Internal Medicine III,
University Hospital Heidelberg, Heidelberg, Germany
Background sCD14-ST represents a 13 kDa fragment of sCD14 which is released
after conversion of CD14++ monocytes into CD14+/16+ monocytes after M-CSF
activation. sCD14-ST may play a role in acute heart failure (AHF) as monocyte TLR4
expression has been shown to be increased in this condition and as it was related to
disease severity.
Objective To evaluate the diagnostic and prognostic value of sCD14-ST in AHF in
comparison with cardiovascular markers (NT-proBNP, hscTnT, GDF-15, sFlt-1,),
inflammatory markers (C-reactive protein (CRP) and procalcitonin (PCT)), kidney
markers (neutrophil gelatinase-associated lipocalin (NGAL)), and soluble ST2.
Methods The marker concentrations were measured in base-line plasma samples
obtained from 60 patients (50 to 90 years old, median 77 years; 26 females, 34 males)
with AHF attending the emergency room (ER). Patients with myocardial infarction
or sepsis were excluded. Outcome measure was mortality at 2 years. sCD14-ST was
determined by using the PATHFAST Presepsin assay (Mitsubishi Chemical), NTproBNP, hscTnT, GDF-15, sFlt-1, PCT, and CRP by using the ELECSYS tests (Roche

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

Diagnostics), and ST2 and NGAL using the Presage assay (Critical Diagnostics, San
Diego, CA, USA) and the NGAL Rapid ELISA Kit (BioPorto Diagnostics, Denmark),
respectively.
Results Baseline NT-proBNP ranged from 361 - 27287 ng/L, median (IQR) = 5773
(2207 - 8488) ng/L confirming the diagnosis of AHF. During the 2years follow up 25
patients (41.7%) died. The results of the biomarker determination are summarized in
the table.
Tab. 1: Prognostic validity criteria of 9 markers for mortality prediction in
emergency patients with acute heart failure
Non-survivors,
Survivors, n=35
n=25
p-value ROC analysis AUC (95% CI)
Median (IQR)
Median (IQR)
sCD14-ST,
1414 (1069763 (601-1144)
0.0001 0.789 (0.662-0.885)
ng/L
1712)
GDF-15,
2885 (17663979 (27250.0392 0.681 (0.546-0.798)
ng/L
4582)
7717)
ST2, g/L 57 (34-83)
79 (50-120)
0.0453 0.675 (0.540-0.792)
0.022 (0.020.044 (0.02PCT, g/L
0.0314 0.667 (0.531-0.785)
0.037)
0.13)
hscTnT,
21 (12-33)
31 (20-47)
0.0280 0.646 (0.509-0.767)
g/L
25.1 (7.28CRP, g/L
13.1 (3.8-25.3)
0.1227 0.640 (0.504-0.762)
62.0)
NT-proBNP, 5453 (19016161 (35180.1609 0.607 (0.470-0.732)
ng/L
6919)
9664)
sFlt-1, ng/L 113 (90-179) 145 (106-210) 0.3410 0.605 (0.468-0.731)
1.69 (0.740.7472 0.505 (0.370-0.639)
NGAL, ng/L 1.60 (0.85-3.0)
2.69)
Conclusion sCD14-ST, hscTnT, PCT, GDF-15 and ST2 differed significantly between
survivors and non-survivors.
Surprisingly, sCD14-ST was found to be the best prognostic marker for mortality
prediction in patients admitted with AHF to the ER. The data may provide new
information on the pathogenesis of heart failure and may improve therapeutic
approaches in the future.

B-332
Validation of a Novel Equation for Estimating Low-Density Lipoprotein
Cholesterol

J. W. Meeusen, A. Lueke, A. S. Jaffe, A. K. Saenger. Mayo Clinic,

Rochester, MN
Background: Aggressive low density lipoprotein cholesterol (LDL-C) lowering
strategies are recommended for primary and secondary prevention of cardiovascular
events. A newly derived equation for LDL-C estimation was recently reported which
addressed limitations in the commonly used Friedewald calculation method (LDLCF). The novel method (LDL-CN) adjusts for the large inter-individual variability in
triglyceride (TG) to very-low-density lipoprotein (VLDL-C) ratio using an adjustable
factor empirically determined based on patient TG and non-high-density lipoprotein
cholesterol (non-HDL-C). LCL-CN reportedly classified patients with superior
concordance to measured LDL-C compared to the Friedewald method, particularly in
patients with LDL-C <70mg/dL. We evaluated the performance of the novel method
in a data set from patients with LDL-C directly measured by -quantification (BQLDL-C), the gold-standard reference method.
Methods: Retrospective analysis identified 23,055 subjects, excluding those with
TG >400mg/dL. Serum total cholesterol (TC) and TG were measured on a Roche
Cobas c501. High-density lipoprotein cholesterol (HDL-C), LDL-C (BQ-LDL-C)
and VLDL-C concentrations were determined by -quantification. Our analytical
performance of these lipid measurements is directly certified by the Center for Disease
Control and Prevention Lipid Standardization Program. The Friedewald LDL-C
was calculated as (LDL-CF =[TC] - [HDL-C] - [TG/5]); and the novel method was
calculated as (LDL-CN =[TC] - [HDL-C] - [TG/X]), where X is an adjustable factor
based on the 180-cell method described by Martin, et al (JAMA 2013;310(19):20612068).
Results: Overall, LDL-CF underestimated BQ-LDL-C, while LDL-CN tended to
overestimate BQ-LDL-C. The median LDL-CN was 112mg/dL (IQR 88-141),
significantly higher than the BQ-LDL-C median of 108mg/dL (IQR 84-136;
P<0.0001). The median difference for [LDL-CF] - [BQ-LDL-C] was -1mg/dL (95%CI
-1-1) and the median difference for LDL-CN was 3mg/dL (95%CI 1-4).
Both estimation methods significantly deviated from the reference LDL-C method
when BQ-LDL-C was <70mg/dL (P<0.0001). The median difference for LDL-CF
was -2mg/dL (95% CI -3 - -1) compared to +2mg/dL (95%CI 1-3) for LDL-CN.
Consequently, the LDL-CN method calculated fewer <0mg/dL compared to LDL-CF
(1 vs. 9, respectively).

Overall, LDL-CF correctly classified 16,593 (72%) patients and LDL-CN correctly
classified 16,749 (73%) patients according to guideline cutoffs. LDL-CF was more
sensitive at identifying patients with BQ-LDL-C <70 mg/dL at 85% compared to 76%
for LDL-CN. However, LDL-CF was less specific at 86% compared to 91% for LDLCN. The largest discrepancy in classification was observed in subjects with a BQLDL-C <70 mg/dL and triglycerides between 200-399 mg/dL, where sensitivity and
specificity for LDL-CF were 88% and 86%, compared to 53% and 99% for LDL-CN.
Conclusions: We compared both novel and Friedewald estimated LDL-C against the
gold standard LDL-C reference method, in contrast to the prior study which relied on
validation of a subset of samples by -quantification to allow the use of the vertical
auto profile method for direct LDL-C measurement. In our patient cohort, the novel
method significantly overestimated LDL-C. Conversely, the Friedewald method
tended to underestimate LDL-C; however the bias was not statistically significant.
We conclude that the novel method has some benefits but whether the improvements
are significant enough over the Friedewald calculation to justify making the change in
routine clinical practice is unclear.

B-333
Elevated Serum Levels of Von Willebrand Factor Antigen (vWF:ag) Predict
for Early Death and Shorter Survival in Patients with Primary Systemic Light
Chain (AL) Amyloidosis Independently of Cardiac Biomarkers

I. Papassotiriou1, E. Terpos2, N. Simos1, A. Gougoutsi2, N. Kanellias2,

E. Eleutherakis-Papaiakovou2, D. Chiotis2, G. Amentas2, M. Kavasi2,
C. Pamboucas2, E. Psimenou2, S. Toumanidis2, M. Roussou2, M. A.
Dimopoulos2, E. Kastritis2. 1Department of Clinical Biochemistry, Aghia
Sophia Childrens Hospital, Athens, Greece, 2Department of Clinical
Therapeutics, University of Athens School of Medicine, Athens, Greece
Background: Cardiac involvement is the main determinant of prognosis in AL
amyloidosis, but the role of endothelium in this disease has not been extensively
studied and no marker of endothelial dysfunction has been evaluated, as yet. We
aimed to study the prognostic role of vWF:Ag in patients with AL amyloidosis who
were treated with novel agents.
Patients and Methods: The analysis included 81 consecutive patients with newly
diagnosed AL amyloidosis, with median number of involved organs was 2; heart was
involved in 62%, kidneys in 74%, peripheral nerve in 24%, liver in 9% and soft tissue
in 21% of patients. Median NTproBNP level was 2,318pg/mL (range 33-75,000pg/
mL); 36% had NTproBNP levels 4,000pg/mL (Roche Diagnostics, CH) and 28%,
38% and 34% of patients had Mayo stage -1, -2 and -3, respectively. vWF antigen
(vWF:ag) levels were measured using a latex particle-enhanced immunoturbidimetric
assay (ACL Top 3G, Instrumentation Laboratory, USA).
Results: The median serum level of vWF:Ag in patients with AL amyloidosis was
181U/dL (range 20-557U/dL), significantly higher compared to healthy controls
(median: 84U/dL, range 48-124U/dL; p <0.001). There was no significant association
of vWF:Ag levels with renal, cardiac, nerve or liver involvement, as well as with
the levels of NTproBNP or Mayo stage and there was no significant correlation
with the degree of renal dysfunction, serum albumin levels or proteinuria. Levels of
involved free light chains did not have any correlation with vWF:Ag levels either.
The prognostic significance of vWF:Ag revealed that vWF:Ag levels within the top
quartile (230U/dL) were associated with a very poor outcome (median survival 4
months vs. 47 months, p=0.001). Because the most important predictor of early death
in patients with AL amyloidosis is cardiac involvement, we performed a multivariate
analysis, which included NTproBNP levels: vWF:Ag levels 230U/dL were
independently associated with survival (HR:2.64, 95%CI1.2-5.8, p=0.01), along with
NTproBNP levels 4,000pg/ml (HR:4.17, 95%CI1.98- 8.8, p<0.001). As survival
curves indicated that patients with vWF:Ag 230U/dL had a significant probability
of early death, we performed an analysis to identify whether vWF:Ag is a significant
predictor of early death, adjusting for the levels of NTproBNP 4,000pg/mL. Thus, in
multivariate analysis, vWF:Ag 230 U/dL was independent predictor of death within
6 months from initiation of therapy (HR:15, 95%CI2.6-84, p=0.002) together with
NTproBNP 4,000pg/mL (HR:16.8, 95%CI3-94, p=0.001). A combination of the two
risk factors was able to identify patients at a very high risk of early death: 75% of
patients with both risk factors present died within 3 months from initiation of therapy.
Conclusion: vWF:Ag levels are elevated in patients with AL amyloidosis but are not
correlated with other features of the disease, such as pattern of organ involvement and
cardiac biomarkers. For first time, we found that high levels of vWF:Ag are associated
with a high risk of early death and shorter survival in patients with AL amyloidosis,
independently of cardiac biomarkers. In addition, vWF:ag levels improve the
prognostic ability of cardiac biomarkers in patients with AL amyloidosis. Our data
also justify the investigation of the role of endothelial dysfunction in AL amyloidosis.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S227

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

B-335
Cardiac troponin T and cardiac troponin I in patients with Chronic Renal
Disease: A Meta-Analysis.

D. C. Gaze, P. O. Collinson. St Georges Hospital & Medical School,

London, United Kingdom
BACKGROUND: There are considerable methodological differences amongst
studies evaluating the prognostic significance of an elevated cardiac Troponin T and I
(cTnT, cTnI) in patients with Chronic Kidney Disease (CKD). This makes it difficult
to study the factors involved or the mechanisms responsible for cTn elevations in
CKD. The heterogeneity of patients included and the small sample size in many
of the studies as well as the use of multiple assays and cut off values has resulted
in imprecise estimation of prognostic significance. This systematic meta-analysis
reviews the clinical value of measuring cardiac troponin in CKD.
METHODS: Suitable papers were identified from Medline/EMBASE searches. Each
paper was abstracted for demographics, clinical outcome, assay used and cut off value.
Primary end-point was all-cause mortality. Meta-analysis was performed reporting
diagnostic odds ratio (dOR) by the DerSimonian-Laird random effects model.
Publication bias was assessed by Funnel plots and Eggers regression asymmetry test;
heterogeneity by Q-test. Summary areas under the receiver operator characteristic
curve(sROC) were calculated. Meta-regression was used to assess the independent
effects of sample size and increasing renal dysfunction on the dOR.
RESULTS: 183 datasets were identified. 107(58%) reported using cTnT and 96(52%)
using cTnI. 33% of studies did not include follow up. In total, 22,820 patients were
included. Median age=61(38-73)years. 58% were male. Median length of follow up
was 24mts. The overall dOR for all-cause mortality was 2.88 (95%CI=2.23-3.71),
(Sensitivity=53%,Specificity=65%,sROC=0.749) irrespective of assay used and data
were heterogeneic without evidence of publication bias. For cTnT, 18,058(79%)
subjects were studied. The dOR was 2.79 (95%CI=2.12-3.70, (Sensitivity=64%,Spec
ificity=60%,sROC=0.943). The 2ndgen and 3rdgen assays had similar dOR. For cTnI,
16,747(73%) subjects were studied. The dOR was 2.38 (95%CI=1.78-4.51, (Sensit
ivity=22%,Specificity=88%,sROC=0.905). Sample size had a positive influence on
the overall dOR(p=<0.0001) which was negated for cTnT(p=0.829) but significant
for cTnI(p=<0.0001). 30(15%) reported pre-dialysis creatinine concentrations
(median=682.31,95%CI=639-763mol/L) which was negatively correlated to the
dOR irrespective of assay used but of no significance(p=0.09), which was negated for
cTnT(p=0.687)but significant for cTnI(p=0.039).
DISCUSSION: cTnT and cTnI are predictors of all-cause mortality in patents with
CKD but no evidence of acute coronary syndrome. The evidence base is larger for
CKD patients who have measurable cTnT compared to cTnI. The ability to predict allcause mortality is higher with cTnI compared to cTnT, but the values are similar. cTnI
exhibits greater specificity in CKD compared to cTnT for the determination of allcause mortality, suggesting the overall diagnostic performance of cTnT and cTnI in
CKD patients may reflect different associated risk of death possibly due to differences
in pathophysiology. These data suggest that the presence of renal dysfunction alone is
not related to the ability of cTn to predict all-cause mortality.

B-336
A simple fluorescence total homocysteine microassay for very low sample
volume

Y. Tan, S. Li, L. Tang-Hall, N. Zhang, Q. Han, R. M. Hoffman. AntiCancer,

Inc, San Diego, CA
Background: Total plasma/serum homocysteine (tHCY) is an independent risk
factor for cardiovascular and other diseases. However, the tHCY assays on the
current market are limited due to the sample volume which is not suitable for routine
screening for newborn or for small animal studies. Therefore, we developed a low
sample-size novel portable HCY fluorescence assay to meet those needs.
Methods: The novel portable tHCY fluorescence assay is a single-enzyme two-step
assay. Reagent I (RI) is a combination of a reducing reagent, homocysteine -lyase,
and a compound consisting of a Schiff-based of N,N-dibutyl-p-phenylenediamine
(DBPDA) and pyridoxal 5-phosphate (PLP). A sample (5 l) of either plasma or serum
is added to RI where reduction of tHCY takes place and enzymatic reaction occurs,
producing H2S, which binds to the DBPDA-PLP, producing a pre-chromophore.
The first-step takes 5 min. In the second-step, Reagent II (RII), an oxidant [Ferric
Chloride (FeCl3) in acid] is added. The oxidation reaction is complete in 10 minutes,
and produces a chromophore, which can be measured by fluorescence at Ex 660/Em
710 nm. The total assay time is 15 min. The assay is carried at room temperature. A
fluorescence reader, including a small portable device, is the only equipment needed.

S228

Results: This novel portable tHCY fluorescence assay was compared to the A/C
Portable tHCY Assay [FDA 510(k) 080851] for 40 plasma samples. The correlation
coefficient is 0.95 and the slope is 0.96. The precisions of within and between assay
were below10%. The linearity of tHCY in the assay is 4.2-44.8 mol/L, and the
detection limit is 4.2 mol/L. The interferences of L-CYS, L-MET, lipid and protein
were all below 10%.
Conclusion: This new portable tHCY fluorescence assay is a highly-sensitive and
simple single enzyme two-step assay that can be used with a very small sample
volume. A simple portable fluorescence reader is the only equipment required. This
assay only needs a 5 l sample, which meets the needs for newborn routine tHCY
screening and for small animal studies.

B-337
Development of a New Specific and Highly Sensitive Enzyme-Linked
Immunosorbent Assay to Detect Heart Type Fatty Acid Binding Protein in
Human Serum

M. Summers1, C. Richardson1, T. McFadden1, R. McConnell1, K. Makris2,

J. Lamont1, S. FitzGerald1. 1Randox Laboratories Limited, Crumlin,
United Kingdom, 2Laboratory of Metabolic Bone Diseases, KAT General
Hospital, Athens, Greece
Background: The human genome comprises nine putative fatty acid-binding
proteins (FABPs) with an approximate molecular weight of 15 kilodaltons. FABPs
are structurally well conserved but interestingly exhibit only moderate sequence
homology. FABPs bind and transport intracellular hydrophobic ligands through
cellular compartments. Heart-type FABP or FABP 3 (H-FABP) is a member of the
FABP family that exhibits high expression in both cardiac and skeletal muscle. Due
to its low molecular weight and cytoplasmic location, H-FABP is released quickly
into the circulation following myocardial injury. Consequently, this protein has been
demonstrated to be an early indicator of acute myocardial infarction. Furthermore,
elevated levels of H-FABP have also been reported following acute stroke. The
availability of analytical methods allowing the determination of this protein in serum
is relevant in clinical settings. This study reports the development of a new specific
and highly sensitive enzyme-linked immunosorbent assay (ELISA) for the detection
of H-FABP, which represents a useful analytical tool to facilitate clinical research
applications.
Methods: A colorimetric sandwich immunoassay was employed. The in-house
made monoclonal capture antibody was immobilised and stabilised on the 96-well
microtitre plate surface. The analyte, if present in the sample, is bound to the capture
antibody and then a second in-house made monoclonal antibody labelled with
horseradish peroxidase is bound to the analyte. Absorbances were read at 450nm. The
signal is proportional to the concentration of the analyte in the sample. Recognition of
native H-FABP was confirmed with analysis of 15 selected stroke samples (6 ischemic
stroke and 9 haemorrhagic stroke) compared with 15 healthy controls. Difference was
assessed with Kruskal Wallis Test(Medcalc version 12.7.8.0).
Results:The ELISA was specific for H-FABP showing %cross-reactivity
values <0.05% with the other FABP family members. The assay presented a functional
sensitivity of 0.2 ng/ml (assay range of 0 to 50ng/ml). Native recognition of H-FABP
in serum samples was demonstrated by comparing H-FABP concentrations in 15
selected stroke samples (Median 5.016ng/ml) versus 15 controls (Median 1.041ng/
ml).The stroke samples were significantly higher than the healthy controls (P<0.0001).
Conclusion: The results show that the developed ELISA for measurement of H-FABP
in human serum is both highly specific and sensitive. Furthermore, the median
concentration values were significantly higher in the selected stroke patient samples
when compared with controls. This indicates clinical utility of the developed ELISA
in the context of stroke. This represents a new analytical tool for clinical research
applications related to this important biomarker.

B-338
Troponin concentrations in young women measured by high-sensitivity
troponins T and I

T. C. Aw, S. K. Phua. Changi General Hospital, Singapore, Singapore

Introduction: The high-sensitivity troponins (hs-cTn) have a reported limit of
detection (LOD) of 5 ng/L (Apple. Clin Chem 2012;58:59) for hs-TnT (Roche) and
1.5 ng/L for hs-TnI (Abbott) (Aw. Clin Chim Acta 2013;422:26-8). Both assays have
lower 99th percentile upper reference limits (P99URL) in women than men. The
hs-TnI URL in females aged 40-65 is 17.9 ng/L with 11.1% of the hs-TnI results
undetectable (below the LOD) (Aw. Clin Chim Acta 2013;422:26-8). Such low hs-cTn
levels are more noticeable in younger subjects. Disproportionate representation of

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

low troponin levels will confound the determination of the P99URL for women. We
decided to investigate hs-cTn levels in a large cohort of young women below 40 years.
Methods: We measured serum troponins (hs-TnT and hs-TnI) in 260 apparently
healthy (via questionnaire) ambulant female subjects aged 20-39 years. An additional
94 women (aged 30-39) were recruited to determine a robust P99URL for hs-TnI in
the entire cohort of 354 women. Pregnant subjects and those with a personal or family
history of hypertension, diabetes, renal, heart, or muscle disease were excluded.
Statistical analyses were performed using MedCalc 12.0 (Mariakerke, Belgium).
Results: Overall 98.5% (256/260) of the subjects studied had hs-TnT values below
the assay LOD while only 27.7% (72/260) of the hs-TnI values were undetectable as
shown in the Table below. Undetectable hs-cTn values are most pronounced in the
20-29 age group. The female (age 20-39) P99URL for hs-TnI (n=354) is 7.7 ng/L
(90% CI 5.1-12.6), median hs-TnI and interquartile range were1.7 and 0.7-2.4 ng/L
respectively.
Table. Proportion of hs-cTn values < LOD
hs-TnT
hs-TnI
< LOD (%)
< LOD (%)
20-29 (72)
100
72.2
30-34 (91)
97.8
7.7
35-39 (97)
97.9
11.3
Overall (260)
98.5
27.7
Conclusion: hs-cTn values are much lower in healthy younger females (age 20-39)
than older subjects. These women may not be suitable subjects for inclusion in the
reference population for the determination of hs-TnT P99URL while women 30-39
can be included for hs-TnI reference range studies.
Age Group (n)

B-339
Turnaround time for a sensitive cardiac troponin I assay at Point of Care

R. H. Christenson1, S. Duh1, S. Carden2. 1Univ of Maryland Sch of Med,

Baltimore, MD, 2Univ of Maryland Medical Center, Baltimore, MD
Objective: We compared turnaround times for cardiac troponin I (cTnI) measurement
in the adult Emergency Department (ED) from order to analytical system (preanalytical), system to result (analytical/post-analytical) and Brain-to-Brain (Order to
Result) for the Pathfast point-of-care system (Misubushi) and for a Vitros 5600 central
lab analyzer and automated track system (Ortho Clinical Diagnostics).

B-341
Reducing CK-MB Utilization: The Calgary Laboratory Services (CLS)
Experience

E. Flynn, I. Seiden-Long, L. deKoning. Calgary Laboratory Services,

Calgary, AB, Canada
Background: Creatine kinase-MB isoenzyme (CK-MB) was the main cardiac
biomarker in Calgary until Troponin T (TnT) testing was implemented in 2001,.
However, CK-MB continued to be used not only for acute myocardial infarction
(AMI), acute coronary syndrome (ACS) evaluation but also for other indications.
High-sensitivity cardiac Troponin T (hs-TnT) testing became available in early
2012. Hs-TnT could be used for virtually every clinical condition where CK-MB
had previously been the test of choice. At the time of hs-TnT implementation, CLS
was performing >4000 CK-MB tests annually. A laboratory utilization initiative was
directed towards significantly reducing CK-MB testing.
Methods: (1) Stakeholder consultation with Calgary Zone Departments of Cardiology,
Cardiovascular Surgery, and Emergency Medicine; (2) Stakeholder and lab agreement
to restrict CK-MB ordering to Cardiologists and Cardiovascular Surgeons only; (3)
Educational article in the CLS newsletter (LabLink) to all medical providers; (4)
On an agreed-upon date, all computerized physician order entry (CPOE) order sets
containing CK-MB used by non-Cardiology specialists were removed from the CPOE
database (SunriseTM Clinical Manager [SCM]). Monthly CK-MB test workloads were
obtained from the laboratory information system (Millennium, Cerner) pre- and
post-intervention; (5) Monthly lists were compiled via SCM of non-Cardiology/nonCardiovascular Surgery providers who continued to order CK-MB; e-mails were sent
to the individual providers as to the superiority of hs-TnT over CK-MB.
Results: Monthly CK-MB workload decreased from a pre-intervention baseline of
646 tests/month (at the time of hs-TnT implementation in January 2012) to a low of
34 tests/month (September 2013) post-intervention (see Figure 1).
Conclusion: The above interventions resulted in a 95% reduction in CK-MB testing
in the Calgary Zone. Multiple utilization strategies including engagement of clinical
stakeholders, education and limiting CPOE resulted in sustained workload reduction
for this test.

Relevance: Point-of-Care availability has been shown to improve the disposition time
for patients with suspected myocardial ischemia in the ED. Both cTnI assays have
10% total CV at the 99th percentile of normal subjects. We compared the difference
in turnaround times in a busy ED associated with a 750-bed tertiary care medical
center. This study helps define turnaround time expectations for in an ED population
presenting to a large metropolitan hospital.
Methodology: ED orders for both the Pathfast point-of-care and the Vitros automated
systems are placed by clinicians in the hospital information system and barcode labels
print; this is Order time. When the specimen is presented to the point-of-care system
the label is scanned; after the sample is tubed to the lab, the specimen is placed on the
automated track system and scanned; the time scanned is Instrument time. Both the
point-of-care and automated lab systems are interfaced and auto-verify results if no
issues impacting analytic testing quality are detected; this is Report time. The total
turnaround time, is termed Brain-to-Brain. Results for 73 patients were audited to
determine if there were differences in diagnostic results between the Pathfast point-ofcare and automated Vitros system. The study was conducted for 100 consecutive days.
Results: Data are displayed in the table. There was no diagnostic difference between
the Pathfast and Vitros systems (p<0.001).
Conclusions: Use of the Pathfast system decreased Brain-to-Brain turnaround time
by 30 minutes or more (24% to 60% faster) compared to the automated OCD system.

B-342
Prognostic Values of Combination of High-Sensitivity Cardiac Troponin I and
B-Type Natriuretic Peptide in Outpatients with Chronic Kidney Disease

F. Kitagawa1, M. Hasegawa2, A. Kuno1, J. Takeda1, T. Ishikawa1, H. Naruse1,

Y. Yuzawa2, J. Ishii1. 1Department of Joint Research Laboratory of Clinical
Medicine, Fujita Health University Hospital, Toyoake, Japan, 2Department
of Nephrology, Fujita Health University Hospital, Toyoake, Japan
Background: The risk stratification for cardiovascular events is clinically important in
patients with chronic kidney disease (CKD: estimated GFR<60mL/min/1.73m2). We
prospectively investigated whether combining cardiac troponin I (hsTnI), measured
with a new high-sensitivity assay, and B-type natriuretic peptide (BNP) would be
effective for the risk stratification in 371 outpatients (median age of 69 years) with
CKD not on dialysis. They were also divided into tertiles of serum hsTnI levels; T1:
<7.8 pg/mL, T2: 7.8-13.9 pg/mL and T3: >13.9 pg/mL. They were divided into tertiles
according to plasma BNP levels; tertile 1 (T1): <24.7 pg/mL, T2: 24.7-81.9 pg/mL

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S229

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

and T3: >81.9 pg/mL. Results: Median estimated GFR was 24mL/min/1.73m2. A
medical history of cardiovascular disease was present in 32.9%. Log hsTnI levels
were positively correlated with log BNP levels (r = 0.43, p < 0.0001). Cardiovascular
event rate were 3.2%, 10.4% and 32.0% in T1, T2 and T3 of hsTnI, and were 2.4%,
13.4% and 30.5% in T1, T2 and T3 of BNP, respectively (p < 0.0001 in both). On a
multivariate Cox regression analysis, both hsTnI (p = 0.002) and BNP (p = 0.0002)
were independently associated with cardiovascular events. Cardiovascular event
rates according to combined setting of hsTnI and BNP tertiles were shown in Figure.
Conclusions: The combination of hsTnI and BNP may be useful for predicting adverse
outcome in this population.

Fig. 1: Presepsin plasma concentrations (Medians, IQR) in healthy controls and

patients with cardiac diseases
Conclusion Presepsin concentration showed a strong relationship with the severity
of cardiac diseases reflecting inflammatory processes in the pathogenesis of
cardiovascular conditions. Our data show that presepsin could provide incremental
diagnostic information to distinguish cardiac and non-cardiac chest pain. Further
investigation is needed to confirm the possible rule-out diagnosis of myocardial
infarction as well as discrimination between cardiac and non-cardiac chest pain at
admission in the ER.

B-343

B-344

Presepsin (sCD14-ST) in Acute Coronary Syndromes and Heart Failure

E. Spanuth1, R. Thomae2, E. Giannitsis3. 1DIAneering - Diagnostics

Engineering & Research, Heidelberg, Germany, 2Mitsubishi Chemical
Europe, Dsseldorf, Germany, 3Department of Internal Medicine III,
University Hospital Heidelberg, Heidelberg, Germany

Troponine T results in a patient population: Should women have their own

upper reference limit for cardiac Troponin T?

J. P. van Straalen, J. C. Fischer, M. Vis, A. Sturk, R. J. de Winter. Academic

Medical Center, Amsterdam, Netherlands

Background Recently, research activities revealed that TLR4 expression on

circulating peripheral CD14+ monocytes may play a pivotal role in the pathogenesis
of cardiac diseases. Monocyte activation is triggered by different factors like LPS,
ischemia, tissue hypoxia, left ventricular distension, or increasing filling pressures
of the heart. It was shown that CD14++ monocytes could be induced by M-CSF
(macrophage colony stimulating factor) and changed into CD14+CD16+ monocytes
and shed off CD14 into blood circulation as soluble sCD14. Simultaneously, a 13 kDa
fragment of sCD14 is formed, named sCD14-ST or presepsin.

Background: For a reference population, differences in 99th percentile values for

cardiac Troponin T (cTnT) are reported between men and women (1). This study
investigates these characteristics in the relevant patient population itself.

Objectives To evaluate the diagnostic value of presepsin in cardiovascular conditions.

Methods 112 healthy volunteers, 40 patients with low grade chronic heart failure
(CHF) and 106 patients admitted with chest pain and/or dyspnea were included.
Discharge diagnosis was unstable angina pectoris (UAP) in 17, non-ST-elevation
myocardial infarction (NSTEMI) in 29, and acute heart failure (AHF) in 60 patients,
respectively. Presepsin and cTnI were measured by using the PATHFAST system
(Mitsubishi).
Results The control group revealed 95% and 99% upper reference limits of 258
(90%CI: 237-276) and 304 ng/L (90% CI: 288-320) ng/L, respectively.Presepsin
differed significantly between controls and patients (Fig. 1). The difference between
healthy controls and patients with UAP was more significant with presepsin than with
cTnI (p<0.0001 and p<0.0071). This finding could be confirmed by ROC analysis
yielding the AUC values 0.992 for presepsin and 0.681 for cTnI, p=0.0001. Detection
of NSTEMI revealed AUC values of 0.982 for presepsin and 0.966 for cTnI. Using
a threshold of 304 ng/L for the diagnosis of NSTEMI presepsin reached sensitivity/
specificity values of 100/98,2% compared to 74,1/100% for cTnI.

Results: The dataset comprised 3515 individual patients (1491 female, 2024 male).
The results profile of women and men was as follows:

Method: Data with respect to cTnT results and gender were extracted from the
laboratory information system over a one year period for all patients with cardiologists
involved in their medical treatment. cTnT was measured with the hs cTnT assay
(Roche Diagnostics, limit of detection 0.005 g/l). If cTnT was serially sampled in
a patient, only the cTnT result of the first drawing was incorporated in the dataset.

cTnT<0.005 g/L: 334 female (22%) and 214 male (10%).

0.005<cTnT< 0.015 g/L: 519 female (35%) and 733 male (36%).
Elevated levels of cTnT (>0.014 g/L): 638 female (43%) and 1077 male (53%). So,
using 0.014 g/L as a decision threshold for women as well as for men, elevated levels
of cTnT relative to one reference value, are predominantly found in males.
Assuming that the risk for ACS in this patient population is the same for women as
for men, in our dataset we established the cTnT decision level for which 53% of the
women, i.e. identical to the percentage men, should have an elevated level of cTnT. To
reach this target, we found that the cTnT threshold for women should be adjusted to
0.010 g/L (793 female patients positive instead of 638 female patients).
Conclusion: The results support the use of a lower reference value for women: At
cTnT<0.005 g/L the percentage of females of the whole female patient population is
twice that of men. The cTnT decision value for women to reach the same percentage
of positive patients as for men was found to be 0.010 g/L. These decision thresholds
coincide with 99th reference limits in a reference population (reference value females
<0.010 g/L; males <0.0145 g/L) (1).
1. Mitiwala SR, Sarma A., Januzzi JL, ODonoghue ML. Biomarkers in ACS and
heart failure: should men and women be interpreted differently? Clin Chem 2014;
60: 35-43.

S230

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

B-345
Plasma cotinine is associated with social deprivation and subclinical
atherosclerosis

A. J. Dunlop, H. E. Turner, J. J. Allison, I. Clunie, B. L. Croal, D. W.

S. Stephen, K. A. Deans. Aberdeen Royal Infirmary, Aberdeen, United
Kingdom
Background: Cotinine is the primary metabolite of nicotine and is the preferred
biomarker for verification of self-reported history of cigarette smoking. Cigarette
smoking is a well-recognised classical risk factor for cardiovascular disease. The aims
of this work were: to use plasma cotinine to verify self-reported smoking history; to
study the association between plasma cotinine and social deprivation, and to examine
the association between plasma cotinine and subclinical atherosclerosis (ultrasound
evidence of carotid plaque).
Methods: Plasma cotinine was measured by an in-house liquid chromatographytandem mass spectrometry method. Intra-assay Coefficient of Variation (CV) was
4.7% at a cotinine concentration of 1ng/mL; 0.9% at a concentration of 10ng/mL
and 1.4% at 100ng/mL. Inter-assay CV was 6.7% at a mean cotinine concentration
of 1.8ng/mL and 5.2% at 372.3ng/mL. EDTA plasma samples (n=572) from the
Psychological, Social and Biological Determinants of ill-health (pSoBid) study were
analysed for cotinine.
Results: Current smokers had higher cotinine concentrations (250 (95% CI 171.1 to
385.8)ng/mL) than ex-smokers (0.1 (0.0 to 1.7)ng/mL), p<0.0001 and non-smokers
(0.0 (0.0 to 0.6)ng/mL), p<0.0001. Cotinine was higher in the most deprived group
(4.4 (0.2 to 283.2)ng/mL) compared to the least deprived (0.0 (0.0 to 0.5)ng/mL),
p<0.001. Cotinine was associated with carotid plaque presence (OR of 1.46 (1.19 to
1.79) per tertile increase in cotinine), p<0.001 (see Figure). After adjusting for selfreported smoking history and area-level social deprivation, this association persisted
(adjusted OR 1.40 (1.04 to 1.87)), p=0.025.
Conclusion: Cotinine is independently associated with subclinical atherosclerosis,
even after adjustment for self-reported smoking history and social deprivation.

the Stratus CS (CS) (Siemens Healthcare Diagnostics), range 30-50,000 ng/L 10%
CV 60ng/L 99th percentile 70 ng/L; the Beckman AccuTnI enhanced (B) (Access 2,
Beckman-Coulter) range 1 - 100,000 ng/L, 10% CV 30 ng/L, 99th percentile 40 ng/L,
the Siemens Ultra (S) (ADVIA Centaur, Siemens Healthcare Diagnostics), range.6 50,000 ng/L, 10% CV 30 ng/L 99th percentile 50 ng/L. and cardiac troponin T (cTnT)
by the Roche high sensitivity cardiac troponin T assay hs-cTnT (Elecsys 2010, Roche
diagnostics), range 3 - 10,000ng/L, 10% CV 13ng/L, 99th percentile 14 ng/L.
The universal definition of myocardial infarction utilising laboratory measurements
of cardiac troponin performed at the participating sites together with measurements
performed in a core laboratory was used for diagnosis. All patients were followed
up for 3 months for major adverse cardiac events death, myocardial infarction,
readmission with unstable angina or need for urgent revascularisation (MACE). Delta
troponin was calculated as the difference between the second and first samples. A
significant delta was considered as 50% of the reference interval. Only those with
a troponin with an initially uncertain diagnosis who had sequential sampling were
studied further.
Results: Samples were available from 608/1132 patients enrolled in the study. MACE
occurred in 7 patients. The number of patients with at least one elevated troponin for
each method was as follows, cTnI CS (>70 ng/L) 8 (1.3%) no MACE, cTnI S (>40
ng/L) 12 (2.0%) 1 MACE, cTnI B (>40 ng/L) 6 (1.0%) no MACE and for cTnT (>14
ng/L) 18 (3.0%) no MACE. Troponin elevation did not predict MACE. Addition of
the delta reduced the number of misclassifications as follows cTnI CS to 3/8, cTnI S
to 11/12, cTnI B to 5/6 and for cTnT to 9/18.
Conclusion: In this group troponin elevation occurred in 1.0-3.0% of patients.
Additional of delta troponin provides only modest further exclusion of alternative
elevations of cTnI. Clinicians need to interpret small elevations of cardiac troponin in
the appropriate clinical context but they carry a good short term prognosis.

B-347
Absolute or relative deltas for diagnosis of myocardial infarction and how
should they be calculated.

P. O. Collinson1, D. C. Gaze1, S. Goodacre2. 1St Georges Hospital, London,

United Kingdom, 2University of Sheffield, Sheffield, United Kingdom
Objective: To compare delta values with peak troponin for the diagnosis of myocardial
infarction when more sensitive troponin assays are used for the diagnosis using the
universal definition of myocardial infarction.
Methods: The study was a sub study of the point of care arm of the RATPAC trial
(Randomised Assessment of Treatment using Panel Assay of Cardiac markers), set in
the emergency departments of six hospitals. Prospective admissions with chest pain
and a non-diagnostic electrocardiogram were randomised to point of care assessment
or conventional management. Blood samples were taken on admission and 90 minutes
from admission for measurement of a panel of cardiac markers. An additional blood
sample was taken at admission and 90 minutes from admission, separated and the
serum stored frozen until subsequent analysis. Samples were analysed for cardiac
troponin I (cTnI) by
the Stratus CS (CS) (Siemens Healthcare Diagnostics), range 30-50,000 ng/L 10%
CV 60ng/L 99th percentile 70 ng/L; the Beckman AccuTnI enhanced (B) (Access 2 ,
Beckman-Coulter) range 1 - 100,000 ng/L, 10% CV 30 ng/L, 99th percentile 40 ng/L,
the Siemens Ultra (S) (ADVIA Centaur, Siemens Healthcare Diagnostics), range.6 50,000 ng/L, 10% CV 30 ng/L 99th percentile 50 ng/L. and cardiac troponin T (cTnT)
by the Roche high sensitivity cardiac troponin T assay hs-cTnT (Elecsys 2010, Roche
diagnostics), range 3 - 10,000ng/L, 10% CV 13ng/L, 99th percentile 14 ng/L.

B-346
The value of delta troponin for differential diagnosis of troponin elevation in
non AMI patients in the unselected emergency room population.

P. O. Collinson1, D. C. Gaze1, S. Goodacre2. 1St Georges Hospital, London,

United Kingdom, 2University of Sheffield, Sheffield, United Kingdom
Objective: To determine if delta troponin allows reliable exclusion of AMI in the non
AMI population when more sensitive troponin assays are used for the diagnosis the
universal definition of myocardial infarction.
Methods: The study was a sub study of the point of care arm of the RATPAC trial
(Randomised Assessment of Treatment using Panel Assay of Cardiac markers), set in
the emergency departments of six hospitals. Prospective admissions with chest pain
and a non-diagnostic electrocardiogram were randomised to point of care assessment
or conventional management. Blood samples were taken on admission and 90 minutes
from admission for measurement of a panel of cardiac markers. An additional blood
sample was taken at admission and 90 minutes from admission, separated and the
serum stored frozen until subsequent analysis. Samples were analysed for cardiac
troponin I (cTnI) by

The universal definition of myocardial infarction (MI) utilising laboratory

measurements of cardiac troponin performed at the participating sites together with
measurements performed in a core laboratory was used for diagnosis. Delta troponin
was calculated for all permutations as follows (where a = 0 minute and b = 90 minute
sample). Absolute delta (b - a), absolute positive delta (b - a for b>a and a - b for
a>b), relative delta (b - a)/a, relative positive delta (b-a)/a for b>a and (a-b)/b for
a>b. Diagnostic accuracy was compared for different diagnostic strategies by receiver
operator characteristic (ROC) curve analysis and comparison of area under the curve
(AUC) for diagnosis by each delta calculation and for peak troponin value.
Results: Samples were available from 617/1132 patients enrolled in the study, 357
male age 23.7-92.8 years median 53.8 years. Delta troponin was diagnostically
equivalent to peak troponin for all four troponin methods. Absolute delta was superior
to relative delta for cTnI B ( p 0.0007) and cTnT (p 0.0491) and just failed to reach
significance for cTnI CS (p 0.064). Absolute and absolute positive delta had equivalent
diagnostic performance for all methods. For all methods, expressing relative delta as
a positive percentage made diagnostic performance worse when compared to absolute
delta or peak troponin or both. Absolute positive delta was superior to relative positive
delta for cTnI B and cTnT.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S231

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

Conclusion: Absolute delta is superior to relative delta for the diagnosis of MI. The
calculation should be the subtraction of the two values in temporal sequence.

B-348
Prognostic implications of simultaneous biomarker assessments in patients with
type 2 diabetes mellitus - observations from the SAVOR-TIMI 53 Trial

P. Jarolim, B. M. Scirica, D. L. Bhatt, M. A. Cavender, M. J. Conrad, A. A.

Umez-Eronini, K. A. Im, D. A. Morrow. Brigham and Womens Hospital,
Harvard Medical School, Boston, MA
Background: Cardiac biomarkers improve risk stratification and therefore offer an
attractive strategy for cardiovascular (CV) screening. We evaluated the incremental
prognostic value of multiple biomarkers reflecting different pathophysiologic
processes in stable outpatients with type 2 diabetes mellitus (T2DM) and established
CV disease (CVD) or cardiac risk factors only to ascertain the benefit of biomarker
screening. Methods: Baseline specimens from 12,182 patients enrolled in the
SAVORTIMI 53 (Does Saxagliptin Reduce the Risk of Cardiovascular Events When
Used Alone or Added to Other Diabetes Medications) trial and followed for a median
of 2 years were tested using the high sensitivity troponin T (hsTnT), N-terminal
pro-B-type natriuretic peptide (NT-proBNP), and high sensitivity c-reactive protein
(hsCRP) assays (all Roche Diagnostics). hsTnT was categorized according to the 99th
percentile (14 pg/ml). Results: hsTnT levels greater than 99th percentile were detected
in 25% of patients with risk factors only and 44% of patients with established CVD.
Overall, elevated hsTnT was associated with an increased risk of CVD (adjusted
hazard ratio, HRadj 3.6, p<0.01) and myocardial infarction (MI) (HRadj 2.3, p<0.01),
with similar results in each risk stratum. (Figure, left) There was a stepwise increase in
the rates of CV death and MI with higher quartiles of each biomarker. After adjusting
for clinical risk factors and biomarkers, elevated concentrations of NT-proBNP and
hsTnT remained significantly associated with both CV death and MI, even at lower
level elevations. (Figure, right) Similar results were seen for hospitalization for
heart failure and MI. Conclusion: In this study of over 12,000 patients with T2DM,
regardless of baseline risk, a substantial proportion of stable patients with T2DM
have evidence of structural heart disease or hemodynamic stress, which were strongly
associated with subsequent risk of CV death and MI.

Results: Of the 100 patients studied, 36 were diagnosed with MI (36%). Nine
patients had a type 1 MI (25%) and 27 had a type 2 MI (75%). 5 of 36 MIs underwent
percutaneous coronary intervention (PCI), while 31 did not. The remaining 64
patients without MI were primarily evaluated for their underlying medical conditions
including shortness of breath, chest pain, heart failure, and renal disease symptoms;
none underwent PCI. In the MI group (no PCI), 222 cTnIs were measured, with 107
cTnI values (48%) determined to be excessive (measured after the diagnosis of MI
was determined). There were 52 additional cTnI order sets beyond the initial one (0,
3, 6, 9 hour), with an average of 7.16 cTnI values per MI patient. Furthermore, 23%
of all cTnI measured were from unnecessary 2nd or 3rd order sets. In the no-MI group,
378 cTnIs were measured, with 150 cTnI values (39.6%) determined to be excessive
(measured after the diagnosis of MI was excluded). There were 63 additional cTnI
order sets beyond the initial one (0, 3, 6, 9 hour), with an average of 6.0 cTnI values
per no-MI patient. Furthermore, 18% of all cTnI measured were from unnecessary
2nd or 3rd order sets. Taking into account nursing time for blood draws as well as
laboratory time and supplies (reagents, technical FTE, blood drawing, supplies), we
conservatively estimate excessive expenditures of approximately $88,000, based on
the 32,000 cTnI tests performed per year at our hospital.
Conclusion: Our data show that in a monitored telemetry unit staffed by attending
and resident physicians, a substantially, excessive number of cardiac troponin tests
are ordered after establishing the diagnosis of both MI and no-MI. The excessive cTnI
testing is wasteful. Better education and monitoring of cTn orders in the diagnosis
or exclusion of MI is needed. Laboratorians should take the lead in educating their
clinical colleagues on the use of cTn monitoring from the 2012 Third Universal
Definition of MI guidelines in patients hospitalized with and without ACS.

B-351
Troponin T Identifies Individuals at Higher Risk for Coronary Heart Disease
Within Narrow Blood Pressure Categories in The Atherosclerosis Risk In
Communities (ARIC) Study

Y. Pokharel1, W. Sun1, G. Taffet1, S. Virani1, J. DeLemos2, C. Ndumele3,

T. Mosley4, R. Hoogeveen1, J. Coresh3, S. Solomon5, G. Heiss6, E.
Boerwinkle7, B. Bozkurt1, C. Ballantyne1, V. Nambi1. 1Baylor College
of Medicine, Houston, TX, 2University of Texas Southwestern Medical
Center, Dallas, TX, 3Johns Hopkins University, Baltimore, MD, 4University
of Mississippi Medical Center, Jackson, MS, 5Harvard Medical School,
Boston, MA, 6University of North Carolina at Chapel Hill, Chapel Hill,
NC, 7University of Texas School of Public Health at Houston, Houston, TX
Background: Although systolic blood pressure (SBP) is a recognized risk factor
for coronary heart disease (CHD), there is variable CHD risk for individuals within
narrow SBP ranges, even after accounting for other CHD risk factors. We evaluated
whether cardiac troponin T (cTnT) measured using a high sensitivity assay identifies
individuals at higher CHD risk within narrow SBP categories among 10,754 ARIC
study participants without CHD.

B-349
Cardiac Troponin Testing Is Over Utilized in Patient Care to Rule In and Rule
Out Myocardial Infarction

O. Rodriguez Fraga1, Y. Sandoval2, S. Love2, Z. J. McKinney2, R. Kohler2,

M. M. Murakami2, S. W. Smith2, F. S. Apple2. 1Hospital Universitario La
Paz, Madrid, Spain, 2Hennepin County Medical Center, Minneapolis, MN
Background: The primary goal of this study was to determine whether clinicians
appropriately utilize cardiac troponin I (cTnI) in the assessment of patients at moderate
risk of acute coronary syndrome (ACS) / myocardial infarction (MI). Secondarily, we
assessed the cost of excessive cTnI orders.
Methods: Following IRB approval, we retrospectively reviewed medical records
(EHR) in 100 consecutive patients who had serial cTnI orders in the cardiac renal
(CARE) unit, a unit where patients at moderate risk for MI are evaluated. Patients
were adjudicated as MI or no-MI according to the 2012 Third Universal Definition
of MI guidelines. A cTnI order set consisted of values obtained at 0, 3, 6 and 9 hours
(Ortho-Clinical Diagnostics Vitros ES, 99th percentile 0.034 ug/L). Clinicians were
not limited to any number of order sets. Excessive cTnI measurements were defined
as any cTnI beyond those necessary to diagnose MI or rule out MI, according the 2012
definition. All cTnI results during hospitalization were tabulated.

S232

Methods: We measured cTnT (Elecsys Troponin-T, Roche Diagnostics) on an

automated Cobas e411 analyzer with a limit of measurement of 3 ng/L. Incident
CHD was defined as hospitalization for myocardial infarction (MI), definite coronary
death, MI confirmed by electrocardiogram and coronary revascularization. Hard CHD
excluded coronary revascularization procedures. Risks for incident CHD by SBP and
cTnT categories were calculated using fully adjusted Cox proportional hazard model
(table).
Results: The mean age of the participants was 62.8 years (56% women, 22% blacks).
Over a mean follow up of 10.95 years there were 1,377 incident CHD events, which
included 857 hard CHD events. Approximately half of the events (53%) occurred
in individuals with SBP<140 mmHg and cTnT3 ng/L. In fully adjusted models,
increasing cTnT was associated with increasing CHD or hard CHD events across
most SBP categories (table). Of note, individuals with cTnT3 ng/L and SBP<140
mmHg had higher hazards for CHD and hard CHD compared to those with cTnT<3
ng/L and SBP 140-159 mmHg.
Conclusion: Risk for CHD increased significantly with higher cTnT levels within
narrow SBP categories. Furthermore, individuals with well controlled SBP but
elevated cTnT had increased hazards for CHD compared with those with suboptimal SBP but undetectable cTnT, suggesting that cTnT is an important marker of
cardiovascular health with potential value in identifying individuals at higher CHD
risk.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

Table. Risk of incident CHD across SBP and cTnT categories

cTnT (ng/L)
<3
3- 5
6-8
9-13
14
P for trend
SBP (mmHg)
SBP <120 mmHg and cTnT <3 ng/L as the reference
CHD
0.093
2.1 (1.5-3.1)
1.5 (1.0-2.1)
1.2 (0.9-1.7)
1.1 (0.8-1.4)
reference
<120
0.001
2.9 (2.0-4.3)
1.5 (1.0-2.1)
1.4 (1.0-1.9)
1.2 (0.9-1.7)
1.1 (0.8-1.6)
120-129
0.008
2.5 (1.7-3.6)
1.8 (1.2-2.6)
1.5 (1.0-2.1)
1.5 (1.1-2.1)
1.3 (0.9-1.8)
130-139
0.011
2.8 (1.9-4.2)
1.9 (1.3-2.9)
1.7 (1.1-2.5)
1.6 (1.0-2.3)
1.4 (0.9-2.1)
140-149
0.001
3.5 (2.2-5.8)
2.1 (1.4-3.4)
2.2 (1.3-3.5)
1.4 (0.9-2.3)
1.1 (0.6-2.0)
150-159
0.001
3.6 (2.3-5.6)
3.5 (2.4-5.1)
2.4 (1.6-3.7)
1.1 (0.6-2.2)
1.6 (0.9-2.7)
160
Hard CHD
reference
1.0 (0.7-1.5)
1.1 (0.7-1.6)
1.5 (1.0-2.2)
2.6 (1.7-4.0)
0.014
<120
1.1 (0.7-1.6)
1.0 (0.6-1.5)
1.5 (1.0-2.2)
1.4 (0.9-2.2)
3.3 (2.1-5.2)
<0.001
120-129
1.1 (0.7-1.7)
1.2 (0.8-1.9)
1.3 (0.8-2.0)
1.6 (0.9-2.6)
2.4 (1.5-3.9)
0.008
130-139
1.0 (0.6-1.8)
1.2 (0.7-2.0)
1.5 (0.9-2.4)
1.3 (0.7-2.2)
3.0 (1.8-4.8)
0.001
140-149
0.8 (0.3-1.8)
1.0 (0.5-2.1)
2.2 (1.2-4.0)
2.1 (1.2-3.7)
4.8 (2.9-8.2)
<0.001
150-159
1.5 (0.8-2.8)
1.7 (0.9-3.3)
1.7 (1.0-3.1)
3.0 (1.8-4.9)
3.4 (2.0-5.8)
0.079
160
SBP 140-159 mmHg and cTnT <3 ng/L as the reference
CHD
0.8 (0.6-1.2)
1.0 (0.7-1.4)
1.1 (0.8-1.7)
1.6 (1.1-2.5)
0.093
0.8 (0.5-1.1)
<120
0.9 (0.6-1.4)
1.1 (0.7-1.5)
1.1 (0.7-1.7)
2.3 (1.5-3.5)
0.001
0.9 (0.6-1.3)
120-129
1.2 (0.8-1.7)
1.1 (0.8-1.7)
1.4 (0.9-2.1)
1.9 (1.2-2.9)
0.008
1.0 (0.6-1.4)
130-139
1.2 (0.7-2.2)
0.9 (0.4-1.7)
1.9 (1.2-3.0)
2.7 (1.7-4.1)
2.7 (1.7-4.4)
0.001
160
Hard CHD
1.1 (0.6-1.8)
1.2 (0.7-2.0)
1.5 (0.9-2.7)
2.7 (1.6-4.7)
0.014
1.1 (0.6-1.7)
<120
1.0 (0.6-1.8)
1.5 (0.9-2.6)
1.5 (0.8-2.6)
3.5 (2.0-6.1)
<0.001
1.1 (0.7-2.0)
120-129
1.3 (0.8-2.3)
1.3 (0.8-2.4)
1.7 (0.9-3.0)
2.6 (1.5-4.6)
0.008
1.1 (0.6-2.0)
130-139
1.5 (0.7-3.2)
1.8 (0.8-3.8)
1.8 (0.9-3.5)
3.2 (1.8-5.7)
3.6 (2.0-6.7)
0.079
160
Data presented as hazard ratio and 95% confidence interval as calculated using Cox-proportional hazards model
Model adjusted for age, race and gender, anti-hypertensive medication use, estimated glomerular filtration rate,
diabetes, fasting glucose, total/high density lipoprotein cholesterol, body mass index and current cigarette smoking.
P for trend was calculated based on the results of Wald chi-square test on linearity hypothesis of ordered cTnT
categories.
Results were similar when the reference was SBP <120 mmHg and cTnT 5 ng/L
cTnT: high sensitivity cardiac troponin T, SBP: systolic blood pressure, CHD: coronary heart disease

B-353
Myocardial injury in cancer patients - are there differences between women
and men?

C. Shortt, M. Mansour, S. Dhesy-Thind, S. Hotte, D. Leong, A. Worster, P.

A. Kavsak. McMaster University, Hamilton, ON, Canada
BACKGROUND: The prevalence of cardiac injury in a general cancer
population is not well documented; despite the fact that there are cancer
treatments that may elicit cardiotoxicity. Currently, cardiac imaging is the
main approach to assess cardiotoxicity; however, laboratory tests may serve
as a complementary tool in this regard. New commercially available cardiac
troponin assays have obtained regulatory approval (outside of the United
States) with preliminary data from healthy individuals suggesting different
thresholds in women versus men. These new tests are termed highsensitivity cardiac troponin assays. We conducted an observational study
measuring high-sensitivity cardiac troponin I (hs-cTnI) in a convenient
non-select group of ambulatory cancer patients assessing whether hs-cTnI
differs in male vs female cancer patients.

B-352
Analytical Correlation Between Abbott ARCHITECT High Sensitivity
and Contemporary Cardiac Troponin I Assays During Evaluation of Acute
Myocardial Infarction within an Unselected Urban Hospital Population

S. A. Love, Y. Sandoval, M. M. Murakami, K. M. Schulz, R. Ler, J.

Nicholson, S. W. Smith, F. S. Apple. Hennepin County Medical Center,
Minneapolis, MN
Background: The Abbott high-sensitivity (hs) cardiac troponin I (cTnI) assay is
available for clinical use outside the US, while only the contemporary cTnI assay is
used in the US; awaiting FDA clearance. The goal of this study was to examine the
correlation between the hs and contemporary cTnI assay results at 1) baseline and
2) over serial measurements during the course of ruling in/out patients suspected of
myocardial infarction (MI).
Methods: During the fall of 2013, fresh EDTA samples were collected during the
evaluation of patients with symptoms of ACS as clinically indicated. Patients were
included if they had at least a baseline and one additional serial sample. Fresh
specimens were measured simultaneously for both the contemporary and hs-cTnI
Abbott ARCHITECT assays (Abbott ARCHITECT i1000SR/i2000SR).
Results: 1021 Specimens from 310 patients were analyzed. Correlation data for
baseline and all specimens, subdivided by concentration ranges, are shown in the
table. For all samples, there was a negative bias with the hs-cTnI assay at baseline,
mean -6 ng/L. Focusing at concentrations above and below the 99th percentile values,
concordance below 100 ng/L, based on the contemporary cTnI value, was generally
poor, 0.77 at baseline and 0.80 for all samples. Slopes varied substantially across the
dynamic range of the assays.
Conclusion: Our data show that cTnI results are 1) lower using the hs-cTnI assay,
b) show variable correlations depending on specimen concentration range and c)
showed poor concordance. Therefore, results are not interchangeable between the
contemporary and high-sensitive cTnI Abbott ARCHITECT assays.
Correlation Data
Specimen
Range
N
Slope
r
Baseline only all cTnI results 310
0.983
0.98
cTnI < 1000
Baseline only
306
0.756
0.94
ng/L
Baseline only cTnI < 100 ng/L 271
0.863
0.81
All Specimens all cTnI results 1021
1.24
0.99
cTnI < 1000
All Specimens
967
0.868
0.91
ng/L
All Specimens cTnI < 100 ng/L 812
0.829
0.84

METHODS: Following ethics board approval, we measured cardiac troponin levels

(Abbott hs-cTnI assay) in two cohorts (A and B) attending a regional cancer center
with the ordering physicians blinded to the test results. Cohort A consisted of all
cancer center patients with clinical lithium heparin plasma samples available for hscTnI measurement over a period of 40 consecutive days. Cohort B consisted of cancer
center patients whose serum samples collected for measurement of clinical tumor
markers and subsequently were measured for hs-cTnI over a period of 20 consecutive
days. We used the sex-specific upper limits of normal (i.e., the manufacturers reported
99th percentiles) to designate myocardial injury (i.e., women >15.6; men >34.2 ng/L)
in both cohorts. We used Statsdirect and Medcalc software for non-parametric testing
(e.g., Mann-Whitney, spearman correlations) and to compare positivity rates between
men and women and correlate with CEA, CA 125, and total PSA in cohort B.
RESULTS: We measured hs-cTnI levels in 4,757 lithium heparin and 285 serum
samples. Cohort A comprised 2,272 women of median (IQR) age= 67y (57-76) and
2485 men of median (IQR) age= 62 y (52-70); p<0.01 for the between-sex difference).
In cohort A, the prevalence of myocardial injury was higher in females (8.0%;
95%CI:6.9-9.3) compared with males (3.1%; 95%CI:2.5-3.9) (p<0.01). Cohort B
comprised 137 women of median (IQR) age= 64 y (54-73) and 148 men of median
(IQR) age= 68y (59-76); p=0.03 for the between-sex difference). The prevalence
of myocardial injury remained significantly higher in the female population (8.8%;
95%CI:4.5-15) compared with the male population (1.3%; 95%CI:0.2-4.9) (p<0.01)
despite the lower age of women in cohort B. There was no difference in hs-cTnI
positivity rates in females or males between cohort A and B (p>0.20) nor was hs-cTnI
correlated to CEA (rho=0.10 p=0.27), CA 125 (rho=0.12 p=0.33), or total PSA (rho=
-0.10 p=0.40) in cohort B (p>0.25).
CONCLUSIONS: Applying sex-specific ULN the prevalence of myocardial injury
in female cancer patients is higher than male cancer patients. Additional studies are
required to assess what are the contributing causes for this marked difference in this
population and importantly, if these differences portend a worse outcome for women.

B-354
On the 99th percentile reference interval determination for the BeckmanCoulter Cardiac Troponin I assays.

D. C. Gaze1, C. A. Hodges-Savola2, D. R. Holmes2, S. A. Faye2, P. O.

Collinson1. 1St Georges Hospital & Medical School, London, United
Kingdom, 2Beckman Coulter, Chaska, MN
BACKGROUND: The 99th percentile concentration of an apparently healthy
population is used define a positive cardiac troponin to aid in the diagnostic and
prognostic assessment of chest pain patients with suspected acute coronary syndrome.
We sought to determine the 99th percentile upper reference limit of the newly released
AccuTnI+3 and pre-commercial hs-cTnI assay.
METHODS: Serum samples (n=1000) were obtained following routine testing
from apparently healthy donors. Subjects were selected on the basis of the following
criteria: >40 years old with normal urea and electrolytes, liver function tests, glucose
and N-terminal pro-B-type natriuretic peptide. Age and gender was also recorded.
Serum samples were stored frozen at -70oC until batch analysis of cTnI using the
AccuTnI+3 and pre-commercial high-sensitivity cTnI (hs-cTnI) assay for the Access2
and AccuTnI+3 on the UniCel DxI Immunoassay system (Beckman-Coulter, Chaska,
MN). The manufacturers report a total %CV of 5 to 7% in the range 7 to 56,360ng/L

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S233

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

with a European 99th percentile of 40ng/L and a US 99th percentile of 20ng/L for the
AccuTnI+3 assay.
RESULTS: Two subjects were removed from the study due to inadequate sample
volume leaving a population of 998 comprising 433(43%) males and 565(57%)
females. The median age was 59.0 years, interquartile range 57.4 to 60.3 years.
There was no significant difference in age between males and females (p=0.4063).
There was a good correlation between the AccuTnI+3 concentrations obtained on
the two instruments (r=0.90, 95%CI=0.89 to 0.91). However, the 99th percentile
upper reference limit were statistically different at 41 and 34ng/L (P=<0.001) for the
Access II and DxI AccuTnI+3 respectively. Detectable concentrations were observed
in 878 (88%) samples on the Access II but only in 578 (58%) samples on the DxI
instrument. Concentrations in males were significantly higher than females using both
instruments. Using the prototype assay, the hs-cTnI 99th percentile was 27ng/L on the
Access II with detectable concentrations in 98% of subjects. Concentrations of cTnI in
males were significantly higher than females using both assay formats.
CONCLUSIONS: The AccuTnI+3 and prototype hs-cTnI 99th percentile
concentrations are similar to other contemporary and high sensitive commercial cTnI
assays. Differences were seen in the 99th percentile using the AccuTnI+3 on the Access
II and UniCel DxI instruments. As gender differences were observed in the reference
population, confirming findings for other hs-cTn assays; further prospective studies
in chest pain patients are required to assess the clinical utility of gender specific 99th
percentile concentrations.

B-355
A step towards D-dimer assays standardization: Antibodies with Equal
Specificity to D-dimer and High Molecular Weight Fibrin Degradation Products

A. Kogan1, K. Mukharyamova1, A. Kara2, A. Bereznikova1, A. Katrukha1.

1
HyTest LTD, Turku, Finland, 2School of Biology, Moscow State University,
Moscow, Russian Federation
D-dimer is an acknowledged marker of blood clotting. It has been shown that D-dimer
concentration is elevated (higher than 0.5 g/ml) in plasma of patients with pulmonary
thromboembolism, deep vein thrombosis and disseminated intravascular coagulation
of different etiology. Despite the long history of D-dimer use in clinical practice, there
are some problems concerning its quantitative determination.
D-dimer as well as its precursor - fibrin degradation products (FDPs) are present in
blood of patients to varying ratios. Different monoclonal antibodies (Mabs) utilized
in different D-dimer assays recognize D-dimer and FDPs with unequal efficiency and
that causes significant discrepancy in results obtained by such assays. In addition,
differences in Mabs specificities make it impossible to use D-dimer as a standard or a
calibrator in current generation of D-dimer assays.
The aim of this work was to obtain monoclonal antibodies with the same specificity to
D-dimer and FDPs and to design an assay prototype that equally recognizes D-dimer
and FDPs in patient plasma.
Hybridomas producing D-dimer-specific MAbs were generated using standard
techniques. Mabs that did not show cross-reactivity with fibrinogen were tested
with D-dimer and FDPs in different two-site combinations. To obtain preparations
of D-dimer and FDPs with equal concentrations, a fibrin clot was initially partially
and then completely digested by plasmin. The partially digested fibrin was used as
a source of FDPs, whereas the completely digested product was utilized as a source
of D-dimer. Thus, the FDPs and D-dimer preparations contained the same amount
of cross-linked fibrin-derived materials. Two-site antibody combination with
capture antibody DD189 and detection antibody DD255 (labeled with stable Eu3+chelate) gave an equal response with FDPs and D-dimer in the range of the antigen
concentrations from 20 to 1000 ng/ml and was selected for the further analysis.
DD189-DD255 assay was used to analyze plasma protein profiles obtained by gel
filtration on Superdex 200 column. It was shown that plasma samples from different
patient groups contained different ratios of D-dimer and FDPs. D-Dimer levels in
blood of three patients who had gone through a surgical operation were comparable
with the level of FDPs or exceeded it. In contrast, FDPs were the main products of
fibrin degradation in samples from two thrombotic patients. In two septic patients
blood different ratios of D-dimer and FDP were observed.
Our results show that the ratio of D-dimer and FDPs varies between patients with
different diseases. This fact means that a precise determination of fibrin-derived crosslinked products in patients blood requires the use of antibodies with equal specificity
to D-dimer and FDPs. Use of antibodies with the same reactivity to D-Dimer and
FDPs would also allow the use of D-Dimer as a calibrator. This would have a
significant impact on standardization of different D-Dimer assays and would benefit
number of practicing clinicians who currently need to deal with the variability of
different D-Dimer assays.

S234

B-356
An Immunoturbidimetric Assay for the Detection of Thromboxane Metabolites
in Urine

I. Muncy, B. Hurley, A. Bamberg, K. R. Pitts. Corgenix, Inc., Broomfield,

CO
Background: Activated and aggregated platelets play a key role in the pathogenesis of
cardiovascular disease. Activated platelets produce thromboxane A2 (TxA2), a potent
vasoconstrictor and inducer of platelet aggregation. TxA2 is generated by thromboxane
synthase from molecules derived from arachadonic acid by cyclooxygenase-1 (COX1). TxA2 has a short half-life in plasma and is rapidly hydrolyzed to thromboxane B2
(TxB2). TxB2, in turn, is metabolized to 11-dehydro thromboxane B2 (11dhTxB2),
11-dehydro 2,3 dinor thromboxane B2 (11dh2,3dTxB2, a truncated form of
11dhTxB2), and a number of other minor TxB2 metabolites which are excreted by
the kidney. Thus, 11dhTxB2 is a stable metabolite of TxA2 and an in vivo indicator of
platelet activity. Acetylsalicylic acid (ASA) has been known for many years to have
anti-platelet activity. ASA functions by acetylating and irreversibly inhibiting COX1, thus inhibiting the production of TxA2 and its metabolites. Low dose ASA blocks
more than 95% of platelet COX-1 activity. The measurement of stable metabolites
of TxA2, such as urinary 11dhTxB2, is a means of quantitating TxA2 production in
vivo and thus a direct way to analyze ASAs effect post ingestion. Urinary 11dhTxB2
can be quantitated using and ELISA method, but recently an immunoturbidimetric
format has been developed. Here we describe the performance and utility of the
immunoturbidimetric format (TxBCardio) on the Selectra ProM chemistry analyzer.
Methods: An R1 reagent/reaction buffer was developed containing a monoclonal
antibody specific for 11dhTxB2 and the 2,3 dinor form of 11dhTxB2. The monoclonal
antibody facilitates the agglutination reaction with the 11dhTxB2 coated microparticles
in a competitive manner based on the amount of 11dhTxB2 present in the urine sample.
An R2 reagent/coated polystyrene microparticle buffer was developed containing
functionalized polystyrene microparticles coated with 11dhTxB2. These reagents
were tested in several iterations to optimize assay performance characteristics specific
to the Selectra ProM analyzer.
Results: The linear range was determined to be 225-6000 pg/mL with an LOQ of
400 pg/mL. A 1:10 post-dilution allows the linear range to be extended to 60,000
pg/mL. Within run precision is <10% and total precision is less than <20% at all
levels. Control recoveries over 5 days and 2 instruments (n=20) were within labeled
ranges. Correlation to an 11dhTxB2 ELISA resulted in an (r) of 0.987, slope of 0.887
and y-intercept of 291 between the two methods. The 2x2 analysis of creatinine
normalized results from both assays results in a positive predictive value of 94%, a
negative predictive value of 100%, and an overall predictive value of 97%. Of many
common urinary components, interference was observed only with high levels of
hemoglobin and protein. Reagents are stable for at least 35 days once opened.
Conclusions: The data presented here suggest that the TxBCardio reagents can be
used on the Selectra ProM analyzer to adequately detect levels of 11dhTxB2 in urine
samples.

B-357
Evidence of Inappropriate Utilization Of Cardiac Troponin Order Sets After
Initiation Of A Provider Alert Prompt In Electronic Health Record

S. A. Love1, Z. J. McKinney2, Y. Sandoval1, S. W. Smith1, R. Kohler1, M. M.

Murakami1, F. S. Apple1. 1Hennepin County Medical Center, Minneapolis,
MN, 2University of Minnesota, Minneapolis, MN
Background: Two new, hospital-wide practices were implemented. First, a cardiac
troponin I (cTnI) order set was implemented based on serial cTnI testing at 0, 3, 6, and
9 hours, for the diagnosis of myocardial infarction (MI). Second, when any provider
placed additional cTnI orders, a pop-up alert was triggered in the electronic health
record (EHR) alerting for possible excessive ordering. This study determined the
frequency and rationale for providers ordering additional cTnI testing after an initial
order set. Further, we determined by patient diagnosis and provider subspecialty the
scope of additional cTnI testing beyond the initial order set.
Methods: Over 2 months, data was collected on the alert, triggering provider, provider
selected clinical indication, and provider follow up action for each alerts. Providers
were not limited to any number of order sets. Alert initiated inclusion of the patient in
the study. We reviewed the EHR for patients associated alerts, and collected: patient
stay (duration, location), ICD-9 diagnosis, cTnI orders, and timing of cTnI result with
measured values (Abbott Architect; 99th percentile 0.025 g/L).
Results: 1477 alerts were generated by 423 providers. 181 (42%) were resident
physicians, of which 42% were 1st year residents, double that for either 2nd or 3rd year

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

residents (22% each). Alerts were associated with 3045 cTnI results for 702 patients
during 833 encounters. 50% of all the cTnI testing (6044 tests) was associated with
an alert. Alert-triggering providers acknowledged and overrode the alert 1440 times
(97%). For overridden alerts, the provider described the override reason from a predetermined prompted list or by free text. 929 (65%) alerts were from prompted list,
including: 519 (35%) acute coronary syndrome concern (ACS; ST- and non-STelevation MI); 249 (17%) demand ischemia; 50 (3%) non-ACS myocardial necrosis.
71% of all free-text overrides gave no indication. The remaining 149 (29%) reasons
included: chest pain (N = 23); retiming (N = 11); ACS (N = 10); trauma (N =7);
or ED visit (N = 6). In 714 encounters (85%) providers placed a second order set
when there were < 2 cTnI results available at the alert time. The most commonly
associated primary ICD-9 code was 786.5,chest pain (N = 231, 27%). Using all ICD9 designations, 1368 alerts (92%) or 779 encounters (93%) were non-ACS related.
Alerts were predominately (33%) generated by providers treating in-patients within
the cardiac/renal unit (CARE), double the rate for all other units within the hospital.
The 52 ACS patients generated 106 alerts.
Conclusion: Our data show that a) visual alerts did not result in a decrease in orders
by providers, resulting in excessive cTnI testing after a diagnosis was determined, b)
the largest number of ignored alerts was in non-ACS patients, and c) even providers
treating patients diagnosed with ACS practiced excessive cTnI ordering. Our
observations highlight the need for better education regarding the use and ordering of
cTn testing to rule in and rule out the diagnosis of MI.

B-358
Evaluation of standardization capability of current cardiac troponin I (cTnI)
assays by a correlation study: results of an IFCC pilot project

J. R. Tate1, D. M. Bunk2, R. H. Christenson3, J. H. Barth4, A. Katrukha5, J. E.

Noble6, M. Panteghini7, H. Schimmel8, L. Wang9. 1Pathology Queensland,
Royal Brisbane & Womens Hospital, Brisbane, Australia, 2Chemical
Science and Technology Laboratory, National Institute of Standards and
Technology, Gaithersburg,, MD, 3Department of Pathology, University
of Maryland School of Medicine, Baltimore, MD, 4Clinical Biochemistry,
Leeds Teaching Hospitals NHS Trust Leeds General Infirmary, Leeds,
United Kingdom, 5HyTest Ltd, Turku, Finland, 6Analytical Science Group,
National Physical Laboratory, Teddington, United Kingdom, 7Centre for
Metrological Traceability in Laboratory Medicine, University of Milan,
Milan, Italy, 8European Commission, Joint Research Centre, Institute for
Reference Materials and Measurements, Geel, Belgium, 9Biochemical
Science Division, National Institute of Standards and Technology,
Gaithersburg, MD
Background: The International Federation of Clinical Chemistry and Laboratory
Medicine Working Group on Standardization of Cardiac Troponin I (IFCC WG-TNI)
performed a pilot study in collaboration with industry to investigate the feasibility of
preparing a commutable and stable cTnI reference material (RM). The study aimed
to test whether serum pools prepared from patient sera could be used as an RM to
standardize cTnI measurement.
Methods: cTnI-positive serum samples from 90 patients presenting to the emergency
department with suspected acute myocardial infarction were used to prepare
seven pools in the range, 200-10,000 ng/L. Samples and pools were assessed for
method comparison, commutability, and stability, and a normal pool was screened
for interference from cTnI autoantibodies and heterophile antibodies by 16 cTnI
commercial systems according to predefined testing protocols.
Results: Each assay was assessed against median cTnI concentrations measured by
16 systems using Passing-Bablok regression analysis of 79 patient samples with cTnI
values above each assays declared detection limit. An 8- to 9-fold difference in cTnI
concentrations was observed among assays, with Pathfast giving lowest values and
Immulite 1000 TPI highest. After correction by a mathematical recalculation using
slope and y-intercept values, between-assay variation was re-assessed. At 190 ng/L
cTnI concentration, average variation of pools reduced from 49% (range, 43-55%)
to 16% (range, 14-19%), at medium concentrations (814, 1634 and 1845 ng/L) from
35% (range, 34-36%) to 13% (range, 11-15%), and at high concentrations (4155 and
7517 ng/L) from 25% (range, 24-27%) to 7% (range 7.0-7.4%). For patient samples at
low cTnI concentration, average variation reduced from 40% (range, 11-65%) to 22%
(range, 11-38%), at medium concentration from 37% (range, 16-63%) to 20% (range,
7-58%), and at high concentration from 29% (range, 13-63%) to 14% (range, 7-42%).
Overall, the 16 assays demonstrated negligible bias after realignment; however, a few
samples showed substantial positive and/or negative differences for individual cTnI
assays that contributed to larger inter-assay variability than for the serum pools.
Conclusion: cTnI values for pooled samples were equivalent within acceptable limits

after straightforward assay realignment. This evidence indicates that pools are a viable
alternative for providing large volumes of commutable sample for use as a surrogate
matrixed RM for cTnI standardization.

B-359
Preliminary Concentration and Density Analyses of Candidate Reference
Material SRM 2924 C-Reactive Protein Solution

E. L. Kilpatrick, M. S. Lowenthal. National Institute of Standards and

Technology, Gaithersburg, MD
The National Institute of Standards and Technology (NIST) has procured recombinant
C-reactive protein (CRP) to generate SRM 2924, C-Reactive Protein Solution. This
candidate material is intended to serve as a pure substance reference material
traceable to SI units for use as a calibrant in the analysis of future reference materials
containing CRP in biological matrices such as serum. The material was received
from the manufacturer in 1200 vials each containing 1 mL of aqueous buffer with
a target concentration of 0.49 g/L to 0.51 g/L of CRP. Certification of this candidate
material will require the determination of purity, density, molar mass, structure and
accurate concentration with homogeneity across the material space. The analysis
of concentration will be determined by amino acid analysis (AAA) via liquid
chromatography isotope dilution tandem mass spectrometry. A sampling plan was
devised to determine concentration homogeneity by analyzing in four independent
groups selected by stratified random sampling. AAA was performed on the first group
by vapor phase hydrochloric acid hydrolysis (105 C for 28 hr) of dried samples spiked
with isotope labeled free amino acids (isoleucine, leucine, phenylalanine, proline and
valine) and calibrated by similarly processed unlabeled amino acids in a five point
linear regression curve. NMIJ 6201-b No. 2, C-reactive Protein Solution, was analyzed
in triplicate as a control material to assess accuracy. The five calibrants, six samples
and three controls were hydrolyzed within the same hydrolysis vessel. Density
measurements were determined for the same samples by the Lang-Levy method with
a 500 L pipet calibrated using pure water. The regression coefficient for calibrants
of the five individual amino acids ranged from 0.9996 to 0.9999 indicating excellent
linearity within the measurement range and produced similar slope and intercept
coefficients. The results of the first group (n=6) indicated that the concentration of the
material by AAA was 20.0 mol/kg (CV = 1.7 %). The value for the control material
was found to be 39.2 mol/kg (CV = 0.2 %) matching well with the certified value
of 40.0 mol/kg ( 1.6 mol/kg expanded uncertainty ). The higher variation of the
sample results vs control is expected because the sample values include bottle-tobottle variation while the control has only within-bottle variability. The density was
determined to be 1.0050 g/mL (n = 6) which was used to convert the concentration
to a mass/volume basis with a value (0.46 g/L) that is close to that based on UV
absorbance as determined by the manufacturer (0.49 g/L). Although the concentration
is slightly below expectations, the material remains a candidate for certification. These
results do not include the full set of planned sampling and are uncorrected for calibrant
purity and water content. However, they do indicate that preliminary concentrations
and homogeneity are sufficient for further analysis leading to the future certification
of this candidate material as a NIST Standard Reference Material.

B-360
Growth differentiation factor-15 levels predict recurrent cardiovascular events
in patients with an acute coronary syndrome in MERLIN-TIMI 36

P. Jarolim, E. A. Bohula May, M. Bonaca, B. Scirica, J. Kuder, S. Murphy,

M. Sabatine, D. A. Morrow. Brigham and Womens Hospital, Harvard
Medical School, Boston, MA
Background: Growth differentiation factor-15(GDF-15), a marker of myocardial
stress, has demonstrated a clear association with mortality. However, the association
with myocardial infarction (MI) has been less consistent. We assessed the prognostic
performance of GDF-15 for recurrent cardiovascular (CV) events in patients with
non-ST elevation acute coronary syndrome (NSTE-ACS). Methods: GDF-15 (R&D
Systems) was measured at enrollment in 4,035 subjects from MERLIN-TIMI 36, a
randomized, placebo controlled trial of ranolazine in patients with moderate to high
risk NSTE-ACS. Patients were classified according to the previously established
cutpoints for GDF-15 of <1,200 (n=2,525), 1,200-1,800 (n=973) and >1,800 ng/L
(n=537). Endpoints of CV death (CVD), MI and heart failure (HF) were adjudicated
by a central blinded events committee. Results: Patients with higher GDF-15 values
tended to be older and more likely to have hypertension, diabetes and prior CV
disease. The rates of the composite of CVD/MI and CVD/HF, as well as the individual
endpoints through one year were higher in patients with higher baseline levels of
GDF-15 (p-trend<0.0001 for all endpoints; Figure 1, top). This relationship was

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S235

Cardiac Markers

Wednesday, July 30, 9:30 am 5:00 pm

also apparent as early as at 30 days for CVD/MI (3.6% vs 5.5% vs 8.7%, p<0.0001)
and CVD/HF (1.2% vs 3.7% vs 10.1%, p<0.0001). Higher GDF-15 levels remained
significantly associated with CVD and HF, but not MI, after adjustment for multiple
factors (Figure 1, bottom). Conclusions: Following NSTE-ACS, GDF-15 provided
prognostic value for CVD and HF, but not MI.

B-362
Magnitude of short- and long-term intra-patient BNP variation in low BNP
patients: Implications for more rational BNP utilization

G. S. Cembrowski1, A. R. Cembrowski2, A. K. Fzry3, D. T. Holmes4.

1
University of Alberta Hospital, Edmonton, AB, Canada, 2University of
Toronto, Toronto, ON, Canada, 3Alberta Health Services, Edmonton, AB,
Canada, 4University of British Columbia, Vancouver, BC, Canada
Introduction: Edmonton hospitals have offered BNP testing to the emergency
department (ED) for the last 6.5 years and to other wards (nonED) for 5 years. Despite
proactive approaches to BNP ordering, BNP usage is increasing in both settings. We
used a unique data mining approach to evaluate biologic variation in patients with
initial BNPs less than 101 pg/mL (clearly normal). We discovered significant patient
repeats as well as very low biologic variation of BNP.
Methods: We tabulated paired intra-patient BNPs repeated within 365 days of each
other. The initial BNP had to be under 101 pg/mL and the second BNP could be of any
value up to 20,000 pg/mL. We calculated the standard deviation of duplicates (SDD)
of the intra-patient pairs grouped in 5 day intervals : 0-5d, 5-10d, 10-15d, and so forth.
Using weighted linear regression, the SDDs were regressed against the time intervals.
The extrapolation to zero time interval in these regressions represents the sum of the
biologic variation (sb) and the short-term analytic variation (sa): y0=(sa2+sb2)1/2
Results: A total of 2178 patients had initial BNPs that were under 101 pg/mL (1742
ED, 471 nonED). The figure shows SDDs for the intervals of separation between the
first and second test. The y-intercept corresponds to a sb of 14.3 pg/mL (27%) for an
initial average BNP of 53.3 pg/mL.

B-361
The influence of eGFR on high-sensitivity troponins T and I in asymptomatic
non-dialysis-dependent patients

Conclusion: The small biologic variation of patients with initial BNPs under 101 pg/
mL implies that such patients have very little risk of congestive heart failure and that
repeat testing will result in another low BNP. Therefore, a BNP less than 101 pg/mL
should not be followed with a repeat test unless there is a clinical change in the patient.

T. C. Aw, S. K. Phua. Changi General Hospital, Singapore, Singapore

Introduction: Troponins are the preferred cardiac biomarkers. High-sensitivity
troponins (hs-cTn) has enabled earlier diagnosis of acute myocardial infarction and
better definition of cardiac risk. Troponins are frequently elevated in chronic kidney
disease (CKD). CKD is associated with increased risk of cardiovascular morbidity
and mortality. It remains unclear if troponin elevation in CKD is consequent to
increased troponin release from myocardial injury or due to impaired renal clearance.
We decided to investigate the influence of renal dysfunction, as reflected by the
estimated glomerular filtration rate (eGFR), on serum hs-cTn in asymptomatic CKD
patients not on dialysis.
Methods: We measured serum hs-TnT (Roche) and hs-TnI (Abbott) in 530
ambulatory outpatient subjects (265 women) across a range of eGFR (CKD-EPI)
[Stage 1-4 CKD]. Pregnant subjects and those from the emergency, renal and
cardiac departments were excluded as were those recently hospitalized (< 30 days).
The gender-specific 99th percentile upper reference limits (99PURL) for hs-TnT
(Giannitsis. ClinChem2010;56::254) and hs-TnI (Aw. ClinChimActa2013;422:26)
were M: 14.5, 32.7 ng/L and F: 10, 17.9 ng/L respectively. An additional 60 subjects
(31 females) with stage 5 CKD (eGFR < 15 ml/min) who were not on dialysis were
also studied.
Results: Overall hs-TnT was above the 99PURL in 57.5% (305/530) of subjects with
CKD stage 1-4 while hs-TnI was increased in 10.6% (56/530). While hs-TnT and
hs-TnI values were 45.2% concordant in the normal range, only 9.6% of patients had
both hs-TnT and hs-TnI elevation. eGFR impacts hs-TnT more markedly than hs-TnI
(< 60 versus < 45 mL/min respectively) (see Table).
Table. Distribution of Elevated hs-troponin values by CKD stage
CKD stage
1
2
3a
3b
4
5
Sub-total
(eGFR ml/min)
(90 ) (60-89) (45-59) (30-44) (15-29) (<15)
N
104 117
102
100
107
60
590
Hs-TnT elevation - n 23
28
55
80
99
39
344
(%)
(22.1) (23.9) (53.9) (80)
(92.5) (65)
Hs-TnI elevation - n 3
8
8
18
19
22
78
(%)
(2.9) (6.8)
(7.8)
(18)
(17.8) (36.7)
Conclusion: hs-TnT concentrations are more susceptible to renal dysfunction than
hs-TnI. When evaluating patients with chest pain, diagnosis and risk stratification
needs to take into account the influence of eGFR especially in those with stage 3
CKD and beyond.

S236

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Proteins/Enzymes

Wednesday, July 30, 9:30 am 5:00 pm

and Bablok. Inter-and intra-assay imprecision were performed according to the
CLSI protocol (EP5-A2). Determination of hemoglobin subtypes was carried out by
recovery measurements of IFCC Hb-subtype evaluation samples.

Wednesday, July 30, 2014

Poster Session: 9:30 AM - 5:00 PM
Proteins/Enzymes

B-364
Determination of reference value for ALT in our laboratory population

N. Z. Maluf, C. F. d. Pereira, M. L. Garcia, N. Silva, C. C. V. Ricardo, F. S.

Fukuoka. DASA, Barueri, Brazil
Background: As the reference values informed in the instructions of use (IFU)
are just a guidance, we know the ranges can vary from lab to lab and so each lab
may have a different range for whats normal. Thereby, the ranges for ALT (alanine
aminotransferase) numbers may differ slightly depending on the technique and
protocols used by different laboratories worldwide. However, normal reference ranges
are routinely provided by each laboratory in their printed individual patients reports.
This study aims to establish the reference value for ALT in our laboratory population
in order to ensure the best follow-up for each patient. The objetive is to evaluate and
standardize the reference values of the UF1000 measured parameters.
Methods: 100 serum samples from healthy individuals (based on current IFU Siemens
Dimension ALTI method reference values: 12 - 78 U/L) were tested for ALT in two
different instruments: Siemens ADVIA 2400 and Siemens Dimension RxL Max. The
Siemens Dimension ALTI and the Synermed ALT (measured on Siemens ADVIA
2400) are an L-alanine pyridoxal-5-phosphate (P5P) methodology. The reference
value for Synermed ALT is from 7 to 35 U/L. We used DAgostino-Pearson test to
establish the reference values for ALT to our hospitals patients and it represents the
central 95% of results determined from a healthy population.
Results: The results are summarized in the table below.
Conclusion: The obtained ALT reference value (5.5 - 45 U/L) is in agreement with
the values from the literature. Finding our own reference values is crucial to ensure
reliable results and, consequently, improve the patients quality of life.

Results: A direct method comparison of HbA1c net values obtained on r910 against
HPLC (BioRad Variant II) with 90 native samples demonstrated excellent correlation
[r=0.9977; Passing/Bablok: y=1.015 x - 0.23%(DCCT)]. DiaSys HbA1c net FS test
is highly precise with an intra-assay precision of CV<0.7% (for HbA1c values from
5.7 to 13.0%) and an inter-assay precision of CV<2.1% (for HbA1c values from 4.4 to
9.9%). High accuracy of HbA1c net FS was demonstrated by recovery IFCC controls
(with varying Hb and HbA1c levels) within 3% of the target value. Various Hb
variants as HbAA, HbAC, HbAD, HbAE, HbAJ, HbAS, HbCC, HbEE, HbSC, HbSS,
elevated HbF and b-Thalassemia showed no significant interference with HbA1c net
FS.
Conclusion: DiaSys new enzymatic HbA1c assay reveals outstanding specificity
and precision. This test highly correlates to HPLC (NGSP/DCCT) but also to IFCC
reference material and is unaffected by interferences from common Hb variants. By
application of HbA1c net to the fully automated DiaSys system respons910, HbA1c
workflow is optimized, due to the implemented on-board hemolysis eliminating errorprone and time-consuming manual preparation.

B-366
Platelet Hyaluronidase-2 as a Potential Novel Biomarker for Both
Inflammatory Bowel Disease and Platelet Function.

S. Albeiroti, K. Ayasoufi, C. de la Motte. Cleveland Clinic, Cleveland, OH

Background: Inflammatory Bowel Disease (IBD), a group of chronic inflammatory
conditions of the intestine, consists of two main types: Crohns Disease (CD) and
Ulcerative Colitis (UC). Different platelet abnormalities are associated with IBD
including reactive thrombocytosis, decreased mean platelet volume (MVP) and
increased state of activation in the circulation. Our lab has recently reported that
platelets contain the enzyme hyaluronidse-2 (HYAL2), which is a GPI-anchored
protein that digests hyaluronan (HA). HA is a major polysaccharide component of
the extracellular matrix and its deposition is increased during intestinal inflammation.
Importantly, HA fragments resulting from HYAL2 digestion can promote
inflammatory cytokine production. We therefore hypothesize that platelet HYAL2
plays a role in IBD.
Objective: To compare platelet HYAL2 levels between IBD patients and non-IBD
controls and to define normal platelet expression parameters in resting and activated
platelets.
Methods: Platelets from IBD patients and non-IBD controls were isolated, washed and
lysed (n=12). Protein contents of platelet lysates were measured by Bradford assay
and HYAL2 levels were compared using immunoblot assay standardized to protein
content (25 g of total protein). Flow cytometry was used to analyze surface HYAL2
levels of paraformaldehyde fixed platelets from normal donors (n=9). Platelets were
activated using thrombin receptor activating peptide (TRAP) and surface P-selectin
was used to assess platelet activation state.

B-365
Evaluation of a novel enzymatic HbA1c test on the fully automated system
respons910

A. Lein, H. Mueller, J. Rink, S. Rosenthal, K. Hahne, D. Vendt, A.

Grzesista, H. Baethies, T. Maerker, E. Metzmann, M. Grimmler, G. Gorka.
DiaSys Diagnostic Systems, Holzheim, Germany
Background: Glycated Hemoglobin A1c (HbA1c) is a well established parameter
for long-term monitoring and diagnosis of diabetes. Here, we present a novel
enzymatic HbA1c test (HbA1c net FS) for highly specific detection of HbA1c,
excluding putative interferences by common hemoglobin (Hb) variants. HbA1c net
demonstrates excellent precision based on the application type (twin-test). This test
links 2 calibrations and 2 detections for Hb and HbA1c in only one determination.
The test principle is defined by Hb determination after sample hemolysis at 570 nm
and H2O2 release after oxidative cleavage of fructosylated dipeptides in the same
cuvette. H2O2 concentration is determined colorimetrically at 660 nm, whereas delta
absorbance is proportional to the HbA1c concentration. The aim of this study was
to establish a novel enzymatic HbA1c test, with superior performance, convenient
handling and optimized workflow compared to other common methods in the market
(as immunoassays or HPLC).
Methods: Assay adaption and performance verification have been carried out on
respons910. All reagents,calibrators and controls were from DiaSys Diagnostic
Systems GmbH. Method comparisons were performed against HPLC as reference
system. Data have been evaluated by using regression analysis according to Passing

Results: 1) Immunoblot analysis of platelets from IBD patients showed an average

reduction of 41% in total HYAL2 levels compared to their non-IBD counterparts
(p-value < 0.05). 2) Flow cytometric analysis showed that only 6.8 1.3% (SE)
of isolated, normal, non-activated (P-selectin negative) platelets express surface
HYAL2. However, after activation the percentage of platelets expressing surface
HYAL2 increased to 37.6 6.4% (SE) (p-value < 0.001).
Conclusion: Platelet HYAL2 levels are significantly lower in patients with IBD
compared to non-IBD controls, which demonstrates its potential usefulness as a
marker in screening patients with IBD. In addition, we showed that platelet HYAL2
surface expression is significantly higher in activated platelets compared to resting
platelets. This reveals that platelet HYAL2, similar to P-selectin, is a potential marker
for platelet activation, indicating usefulness in platelet function tests.

B-367
Determination of Fecal Pancreatic Elastase using an Automated Immunoassay
Procedure

C. Cruzan, P. Krause, K. Urek, J. A. Maggiore. Doctors Data, Inc., Saint

Charles, IL
Background: The clinical laboratorys role in the assessment of chronic digestive
disorders and bowel diseases continues to expand. Exocrine pancreatic insufficiency
as a contributing source of bowel-related diseases is commonly assessed by measuring

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S237

Proteins/Enzymes

Wednesday, July 30, 9:30 am 5:00 pm

fecal levels of pancreatic Elastase, a serine protease that hydrolyzes dietary proteins.
The absence or insufficiency of pancreatic Elastase contributes to incomplete protein
digestion leading to varying degrees of malnutrition, and to pathologies ranging from
abdominal discomfort to maladies of the large intestine. Manual ELISA techniques
to measure fecal pancreatic Elastase are labor-intensive and prone to high variability
related to sample handling. An automated sample processing and analytical platform
was sought to enhance operating efficiency, reduce variability, and retain analytical
accuracy.

detected. A schematic diagram of the overall urinary protein pattern in IgAN was
established (Figure 1) based on proteins identified in more than 50% of subjects of
each profile pattern.
Conclusion:The type B urinary protein profile pattern indicates increased oxidative
stress compared to type A, whereas type C indicates mild renal impairment compared
to types A and B. Our detailed analysis provides a valuable non-invasive tool for
predicting the degree of renal damage in IgAN.

Methods:The BIOSERV Diagnostics solid-phase enzyme linked immunosorbent

assay (ELISA) is based on a double sandwich technique employing two polyclonal
antibodies engineered to recognize several epitopes on human pancreatic Elastase
peptide sequences; the first is coated to the assay wells of the microplate, and the
second is labeled to biotin which binds to the immobilized Elastase molecule. A
streptavidin-labeled horseradish peroxidase binds to the biotin-Elastase complex,
which oxidizes the tetramethylbenzidine (TMB) chromogenic substrate to a blue
color in the well-characterized reaction. The required sample dilution steps and
the procedural sequence of substrate addition, conjugation, washing, incubation,
and reading were programmed into a Dynex DSX analyzer. Analytical precision,
linearity, recovery, and accuracy of this automated process were assessed. Further,
time studies of technologist direct labor and analytical throughput were undertaken to
assess the benefits of automation.
Results: Using the automated procedure on the Dynex DSX, intra-assay precision
coefficients of variation (CV) (n=20) at pancreatic Elastase levels of 107 and 459
g/g were 8.4% and 4.3%, respectively. Inter-assay CV (n=20) at the same pancreatic
Elastase levels were 12.0% and 6.4%, respectively. The inter-assay CV and intraassay CV using the manual ELISA method at a level of 484 g/g were 4.8% and
8.9%, respectively. Assay linearity was demonstrated between 28 and 500 g/g, with
recovery of expected pancreatic Elastase concentrations ranging from 101.5% to
112.7%. Least-squares regression analysis comparing pancreatic Elastase values from
diluted fecal extracts as determined by manual ELISA and automated DSX ELISA
processing (n=35, range 35 - 485 g/g) yielded a correlation coefficient (R2) of 0.93, y
= 1.10x + 58.6; standard error of estimate = 48.5. Time studies (n = 5) indicate that the
automated procedure provides an average reduction of 147 minutes of direct, handson labor of a technologist processing a 96-well plate of samples, quality controls,
and calibrators. Overall, start-to-finish analytical throughput averages 244 minutes
using manual processing, compared to 303 minutes using the Dynex DSX program
for pancreatic Elastase.
Conclusion: The automated ELISA procedure for fecal pancreatic Elastase using the
Dynex DSX has been shown to maintain accuracy and improve analytical precision
of the BIOSERV Diagnostics ELISA method. Despite increased overall analytical
throughput, the automated procedure provides operating efficiency through the
reduction of direct hands-on time.

B-368
Clinical application of urinary protein profiling using cellulose acetate
membrane electrophoresis for patients with IgA nephropathy

A. Nakayama1, M. Sakatsume2, A. Katayama3, H. Suzuki3, K. Shiba1, S.

Iijima1. 1Bunkyo Gakuin University, Tokyo, Japan, 2Niigata University,
Niigata, Japan, 3Nippon Medical University, Tokyo, Japan
Background:Urinary protein patterns obtained from cellulose acetate membrane
(CAM) electrophoresis are useful for renal damage prediction in patients with
glomerulonephritis such as IgA nephropathy (IgAN). To identify the disease-causing
proteins in each CAM fraction, we established a protein extraction method for
proteomic analysis and developed protein profiles in the patients with IgAN.
Methods:Urine samples from 30 patients with IgAN were each applied onto 10 lanes
of a CAM. After electrophoresis, the first and last lanes were cut, silver-stained, and
used as a guide for fractions in the unstained region. These sections were fragmented,
and proteins were extracted with Laemmles sample buffer, separated by non-reducing
sodium dodecyl sulfate-polyacrylamide gel electrophoresis, silver-stained, and
identified by liquid chromatography-tandem mass spectrometry.
Results:Urinary protein profiles for 30 patients with IgAN showed 3 main patterns
(types A - C): in type A (13 patients), 5 fractions were similar to the normal serum
pattern; in type B (14 patients), mobility of the 1-globulin fraction was relatively
fast, similar to that of albumin; and type C (3 patients) showed a lack of 1-, 2-,
and -globulin fractions. The fast 1-globulin fraction of type B mainly comprised an
albumin dimer, which is formed in response to oxidative stress, and 1-antitrypsin.
Although there were no bands corresponding to 1- and 2-globulin in type C, a
significant amount of Tamm-Horsfall protein, reflecting normal tubular function, was

S238

B-369
Performance Evaluation of ERM-DA471/IFCC standardized Tina-quant
Cystatin C Generation 2 Assay on Roche Clinical Chemistry Analyzers

C. Schmidt1, C. M. Cobbaert2, J. A. Bakker2, H. P. C. Schippers2, W.

Endlich2, M. Flodin3, P. Eliasson3, D. Hermsen1. 1University Hospital of
the Heinrich-Heine University, Central Institute of Clinical Chemistry and
Laboratory Diagnostics, Dsseldorf, Germany, 2Leiden University Medical
Center (LUMC), Department of Clinical Chemistry and Laboratory
Medicine, Leiden, Netherlands, 3University Hospital, Clinical Chemistry
and Pharmacology, Uppsala, Sweden
OBJECTIVE Cystatin C has been proposed as a more sensitive endogenous
biomarker regarding the estimation of glomerular filtration rate when compared to
serum creatinine. But Cystatin C assays of different manufacturers showed a lack
of comparability before the ERM-DA471/IFCC reference material was available. In
our study we evaluated the analytical performance of the new ERM-DA471/IFCC
standardized Tina-quant Cystatin C Gen. 2 immunoassay (Roche Diagnostics,
Mannheim, Germany).
METHODS AND RESULTS The analytical performance of the new assay was
evaluated in three laboratories using Roche/Hitachi MODULAR ANALYTICS,
COBAS INTEGRA 800, cobas c 701 and cobas c 501 analyzers. Cystatin C
concentration was measured turbidimetrically by a particle enhanced immunoassay
standardized against the ERM-DA471/IFCC. The analytical performance of the new
assay was investigated under routine laboratory conditions using samples covering the
entire measuring range of cystatin C (0.4 - 6.8 mg/L).
Within-run imprecision data were collected using three control levels (Roche
Diagnostics) and three self prepared pools of human sera (single run, n = 21 replicates
per sample) covering a concentration range from 0.75 to 5.02 mg/L. Coefficients
of variation (CVs) were determined to be less than 3.3 % for control materials and
less than 2.8 % for pooled samples. Between-day imprecision yielded CVs between
1.7 and 7.3 % (one run/day, up to 17 days) using control sera (three levels, Roche
Diagnostics). The recovery of target values in different control sera was determined
performing triplicate measurements in three independent runs. The recovery of
Cystatin levels in control sera from Roche Diagnostics and other manufacturers
ranged between 86.0 - 107.8 % and 86.4 - 100.5%, respectively.
All method comparison experiments were designed in compliance with CLSI
EP09-A3, performed using > 119 serum samples and analyzed with Passing-Bablok
regression. Statistical analysis of method comparison experiments against the Roche
Cystatin C Gen.1 yielded correlation coefficients > 0.990, slopes between 0.92 and
1.13 (0.89 - 1.16, 95 % confidence interval) and intercepts from -0.13 to 0.07 (-0.16
- 0.10) mg/L. Method comparison studies against other available ERM-DA471/
IFCC standardized Cystatin C immunoassays (Siemens N Latex Cystatin C, Gentian
Cystatin C) performed on different analyzer systems (Siemens BN II Nephelometer,
Abbott Architect, Roche cobas c502) showed excellent correlation coefficients ( r >
0.990). Calculated slopes ranged between 0.93 and 1.04 (0.91-1.06), and intercepts
between -0.10 to 0.04 (-0.13 - 0.07) mg/L, respectively, confirming the successful
standardization of the new Cystatin C Gen. 2 immunoassay to the ERM-DA471/IFCC
reference material.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Proteins/Enzymes

Wednesday, July 30, 9:30 am 5:00 pm

CONCLUSIONS Our study demonstrated an excellent technical performance of the

new Roche Tina-quant Cystatin C Gen. 2 assay. Method comparison results revealed
a high degree of comparability to other available Cystatin C immunoassays that are
traceable to ERM-DA471/IFCC. Due to the excellent study results the new Tinaquant Cystatin C Gen. 2 assay is well-suitable for routine use.
Disclaimer: The study was sponsored by Roche Diagnostics GmbH, Germany. The
Roche Tina-quant Cystatin C Gen.2 assay is not yet cleared for use in the U.S.
COBAS, COBAS C, COBAS INTEGRA, MODULAR and TINA-QUANT are
trademarks of Roche.

B-370
Effects of CH3CN-cosolvent and CuBr-catalyst on the synthesis of
difluorophosphonates as chemical specialty

H. Tsai. Department of Chemistry and Material Engineering, Chung Cheng

Institute of Technology, National Defense University, Ta-hsi, Tao-yuan,
Taiwan
Background: Fluorophosphonates have found applications as hyperlipidaemic
drugs, hormone substitutes, cancer chemotherapy and nervous chemical warfare.
However there is generally a conspicuous lack of methods for the preparation of
difluoromethanephos phonates. Effects of CH3CN-cosolvent and CuBr-catalyst on
the synthesis of difluoro phosphonates (EtO)2P(O)CF2C(O)R were discussed in this
research.
Methods: Synthesis of (EtO)2P(O)CF2Br : 19F NMR: -62.0 ppm (d, J = 90.3); 31P
NMR: -0.65 ppm (t, J = 93.7); 1H NMR: 4.42 ppm (4H, q, J = 7.1), 1.42 ppm (6H,
t, J = 7.1). Procedure for Synthesis of (EtO)2P(O)CF2C(O)CO2Et : (Method A): A
three-necked flask was charged with 15.0 mL of a 2.0 M monoglyme solution of
(EtO)2P(O)CF2ZnBr. 42.0 mmols of freshly distilled ClC(O)CO2Et was added and
stirred at rt for 48 hours. The mixture was filtered, extracted with CH2Cl2, dried over
anhydrous MgSO4, concentrated by rotary evaporation and vacuum distilled at 9495oC/0.15 mmHg to give 54 % of the titled compound. 1H NMR : 1.40 (9H, t, J =
7.1), 4.30 (6H, m, J = 7.1). 13CNMR : 13.9 (s), 16.2 (d, J = 5), 63.5 (s), 65.8 (d, J =
7), 115.5 (t, d, J = 275, J = 199), 158.5 (s),181.8 (t, d, J = 16). 19FNMR : -115 (d, J =
95). 31PNMR : 4.5 (t, J = 95). (Method B): 7.5 mL dry CH3CN, 0.45 mmol CuBr and
42.0 mmols ClC(O)CO2Et were added to a 15.0 mL of 2.0 M monoglyme solution
of (EtO)2P(O)CF2ZnBr. Stirred at rt for 0.5 hour. 58 % of the titled compound was
afforded. (Method C): 0.45 mmols CuBr and 42.0 mmols of ClC(O)CO2Et were added
to a 15.0 mL of 2.0 M monoglyme solution of (EtO)2P(O)CF2 ZnBr, Stirred at rt for 24
hours. The 19F NMR spectrum indicated the presence of (EtO)2P(O)CF2C(O)CO2Et,
(E,Z)-(EtO)2P(O) CF=CFP(O)(OEt)2 and (EtO)2P(O)F.
Results: Reaction of triethylphosphite with dibromodifluoromethane via MichealisArbuzov reaction gave 95 % yield of (EtO)2P(O)CF2Br. Subsequently reacted with
acid washed zinc powder in the presence of monoglyme, (EtO)2P(O)CF2ZnBr was
obtained. It took 48 hours to synthesize (EtO)2P(O)CF2C(O)R from the acylation
of (EtO)2P(O) CF2ZnBr with ClC(O)R (R = CO2Et, OEt, NEt2) at rt However, this
situation can be easily ameliorated upon addition of a catalytic amount of cuprous
bromide and the addition of acetonitrile as cosolvent to the reaction mixture. This
reaction is completed within 0.5 hour to yield (EtO)2P(O)CF2C(O)CO2Et in a 58 %
isolated yield. Furthermore, if the reaction was carried out in the presence of 1.5 %
CuBr-catalyst without the CH3CN as cosolvent, in addition to the formation of 12 %
of (EtO)2P(O)CF2C(O)CO2Et, 18 % yield of (E) and (Z)-(1,2-difluoroethylenediyl)
bisphosphonate (EtO)2P(O)CF=CFP(O) (OEt)2 and 70 % of (EtO)2P(O)F at -82 ppm
(JP-F = 972 Hz) were observed in 19FNMR spectrum.
Conclusion: In the presence of CH3CN-cosolvent, monoglyme solvent and
appropriate CuBr-catalyst, acylation of (EtO)2P(O)CF2ZnBr with ethyl chloroformate,
diethyl carbamoyl chloride or ethyl oxalyl chloride gave good yields of 2-oxo-1,1difluorophos phonates (EtO)2P(O)CF2C(O)R as chemical specialty. However, if these
acylations were carried out without acetonitrile as cosolvent, the mixture products
of (EtO)2P(O)CF2C(O)CO2Et, (E,Z)-(EtO)2P(O)CF=CFP(O)(OEt)2 and (EtO)2P(O)F
were observed.

B-371
Neutrophil Gelatinase-Associated Lipocalin Levels in Patients with Thalassemia
and Sickle Cell Disease: Correlation with Renal Injury

I. Papassotiriou1, E. Voskaridou2, A. Margeli1, A. Kattamis3. 1Department

of Clinical Biochemistry, Aghia Sophia Childrens Hospital, Athens,
Greece, 2Thalassemia Center, Laikon General Hospital, Athens, Greece,
3
Hematology-Oncology Divison, Department of Pediatrics, University of
Athens Medical School, Athens, Greece
Background and Aims: Neutrophil gelatinase-associated lipocalin (NGAL) is a
protein belonging to the lipocalin superfamily initially found in activated neutrophils,
in accordance with its role as an innate antibacterial factor. However, it subsequently
was shown that many other types of cells, including in the kidney tubule, may
produce NGAL in response to various injuries. The increase in NGAL production
and release from tubular cells after harmful stimuli of various kinds may have selfdefensive intent based on the activation of specific iron-dependent pathways, which in
all probability also represent the mechanism through which NGAL promotes kidney
growth and differentiation. NGAL levels clearly correlate with severity of renal
impairment, probably expressing the degree of active damage underlying the chronic
condition. For all these reasons, NGAL may become one of the most promising nextgeneration biomarkers in clinical nephrology and beyond. We aimed to investigate the
clinical significance of NGAL levels and its correlation with renal function in patients
with hemoglobinopathies.
Patients and Methods: 117 adult patients with hemoglobinopathies were included
in the study divided in 3 groups. Group A: 30 patients with transfusion-dependent
thalassemia major (TM); Group B: 29 patients with thalassemia intermedia (TI)
and Group C: 58 patients with HbS/betathalassemia disease, while 20 apparently
healthy individuals served as controls (Group D). In patients and controls along
with standard blood and urine chemistry, measurements of serum Cystatin C and
NGAL were performed. Estimated Glomerular Filtration Rate (eGFR) values were
calculated with an unadjusted for body surface Cystatin C based equation: eGFR (mL/
min)=77.24(Cys C)exp(-1.2623)
Results: The main results of the study showed that: a) NGAL levels were significantly
higher in all the groups of patients compared to controls: Group A 95.045.0g/L,
Group B 139.186.1g/L, Group C 117.837.3g/L vs Group D 50.311.3g/L
(p<0.001), b) Cystatin C levels were significantly higher in patients of Group A
0.940.34mg/L and Group C 1.040.50mg/L compared to controls 0.750.09mg/L
(p0.300), c) NGAL levels and eGFR values (Group A: 96.939.8, Group B:
117.026.0, Group C: 86.227.8 and group D: 109.615.0mL/min, respectively)
correlated significantly in patients of Group A and Group C (r=-0.739, p<0.001 and
r=-0.735, p<0.001, respectively), while NGAL values are independent from eGFR
values in patients of Group B.
Conclusions: These findings illustrate the tubular-glomerular activation feedback
mirrored by NGAL in patients with transfusion-dependent thalassemia major and
HbS/betathalassemia disease, who suffer from renal injuries, indicating that tubular
damage precedes GFR reduction. Upregulation of NGAL in patients with thalassemia
intermedia independently of renal injuries may reflect the compensatory, protective
role of NGAL in response to diverse cellular stresses, including inflammation and
oxidative stress. However, recent reports have implicated NGAL upregulation
as a mechanism that contributes to anemia in the setting of chronic low grade
inflammation. In experimental models, systemic and medullary NGAL has been
demonstrated to induce inhibition of erythropoiesis through induction of apoptosis
and arrest of differentiation of erythroid progenitor cells.

B-372
Evaluation of Dried Blood Spots for Use in Isoelectric Focusing Electrophoresis
in Deficient Alpha-1-Antitrypsin Phenotype Interpretation

M. M. Duran. Geonostics, Lincolnshire, IL

Background: Laboratory diagnostics contribute significantly in the diagnosis
of Alpha-1-Antitrypsin (AAT) deficiency, utilizing AAT serum concentration,
AAT phenotype determination by isoelectric focusing (IEF) electrophoresis, and
genotyping. Dried blood spots (DBS) are a potentially attractive sample type for IEF
phenotype analysis on the Sebia Hydrasys because of the ease of sample collection.
Previous work demonstrated that the common phenotypes of MM, MS, and MZ, as
a DBS sample, were indistinguishable from their companion serum samples. In this
current study, we present the novel methodology of DBS for the 3 most prevalent
deficient phenotypes: SS, SZ, and ZZ.
Methods: Eighteen whole blood samples from known phenotypes SS, SZ, and ZZ

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S239

Proteins/Enzymes

Wednesday, July 30, 9:30 am 5:00 pm

were spotted on filter paper using 50 uL of sample. The blood spots dried overnight
and then the 50 uL dried blood samples were punched and rehydrated with a buffer/
deionized water solution. Ten uL of each extracted DBS sample was applied to the
IEF comb. Eighteen serum samples of the corresponding DBS phenotypes were also
placed on a comb for comparison. IEF was performed on the Sebia Hydrasys using
standard protocol and the resultant gel was stained, washed, and digitally scanned.
Results: In Figure 1, phenotype SS is on the left, phenotype SZ is in the center, and
phenotype ZZ is on the right. For each phenotype, the DBS sample is on the left
and the corresponding serum sample is on the right. All 18 DBS phenotype samples
displayed the unique identifiable banding patterns present in the serum samples.
Conclusion: IEF of DBS samples adequately reveal the S and Z alleles, in the absence
of the M allele, on the Sebia Hydrasys. Work is underway to validate the 100+
additional rare alleles and to establish AAT protein stability on filter paper

Differences between assays were analyzed by Pearson Test and the degree of
agreement between measurements was evaluated using Bland-Altman analysis.
Results: For A40 measurement with the INNOTEST -Amyloid (1-40)
immunoassay, the intra-assay CVs were 1.29% and 2.68% at A40 concentrations
of 5000 and 9800 ng/L respectively (n= 16). The inter-assay CVs were 13.94% and
15.66% at the same A40 concentrations (n = 10). The mean SD recovery of CSF
A40 immunoassay was 85% (slope: 1.0264; intercept: -328.83; R2= 99.8%).
The correlation coefficient between the 2 immunoassays (INNOTEST vs IBL
assay) was R2= 0.735. Interpretation of A42/A40 ratio was concordant between the
2 immunoassays at 92% (68/74) of the cases using a cut-off at 0.05.
Conclusion: This preliminary study showed that INNOTEST -Amyloid (1-40)
immunoassays intra-assay CVs, inter-assay CVs and linearity CVs were less than
20%. INNOTEST -Amyloid (1-40) immunoassay seems to correlate with IBL
Human Amyloid (1-40) immunoassay which is commonly used in the laboratories.

B-375
Evaluation of a new liquid UIBC method on Architect c4000 analyzer

M. Gramegna, M. La Motta, R. Lucini. Sentinel CH. SpA, Milan, Italy

Objective The objective of this study is to evaluate the analytical and clinical
performances of a new liquid UIBC Method on Architect c4000 analyzer.
Relevance: Transferrin is the principal plasma protein for transport of iron. One
molecule of TRF binds two ferric ions and an associated anion, usually bicarbonate in
vivo. Normally 30% of the iron binding sites of transferrin are occupied by Fe3+, the
additional amount of iron that can be bound is the unsaturated iron-binding capacity
(UIBC). The sum of serum iron and UIBC represents the total iron-binding capacity
(TIBC). UIBC measurement is an iron panel parameter used in the diagnosis and
treatment of anemia.
Methodology UIBC is determined directly by saturating the transferrin at an alkaline
pH with a known excess amount of iron. The iron that remains free after transferrin
saturation is reduced to ferrous state and then complexed by ferene-S to form a
stable complex which colour intensity is measured at 580-600 nm. UIBC is therefore
determined by subtracting the quantity of unbound iron from the total added quantity.
The instrument used for this evaluation was an Architect c4000 analyzer, a randomaccess analyzer. To perform this evaluation, modified CLSI protocols were adopted.
Acceptance criteria as total imprecision were 5% for samples 110 g/dL and 2%
for samples 250 g/dL. LOD should be 41 g/dL, The method should be linear up
to 500 g/dL. Claimed goal for on board calibration stability was 7 days and reagent
on board stability was 35 days. Comparison to commercial methods had following
acceptability: slope 0.90 - 1.10, intercept 15, r 0.975.

B-373
Evaluation of INNOTEST -Amyloid (1-40) immunoassay and comparison
with IBL Human Amyloid (1-40) immunoassay

J. Oudart, R. Mahmoudi, G. Fiabane, P. Arthuis-Demoulin, J. Novella, L.

Ramont. CHU de Reims, Reims, France
Background: Combined assay of cerebrospinal fluid (CSF) biomarkers Tau protein,
phosphorylated Tau protein and beta Amyloid peptide 1-42 (A42) is commonly used
in the diagnosis of Alzheimers disease (AD). CSF biomarkers are now included in
diagnosis criteria in addition to clinical evaluation and medical imaging. In discordant
cases where an isolated reduction of A42 peptide is found, the beta Amyloid
peptide 1-40 (A40) measurement associated with A42/A40 ratio allows better
classification of patients. According to the literature, the clinical cut off for this ratio
is less than 0.05 in AD.
The aim of this study was to evaluate the INNOTEST -Amyloid (1-40)
immunoassay and compare it versus IBL Human Amyloid (1-40) immunoassay.
Methods: - Evaluation of Innotest -Amyloid (1-40) immunoassay (Fujirebio
Europe, Belgium): intra-assay precision studies were evaluated on two CSF samples
(5000 and 9800 ng/L respectively), 16 consecutive times. Inter-assay precision studies
were evaluated at the 2 same levels of concentrations (n= 10). Dilutional linearity was
evaluated by successive dilutions of a CSF sample in the sample buffer.
- Comparison between INNOTEST -Amyloid (1-40) immunoassay and
IBL Human Amyloid (1-40) immunoassay (IBL, Japan): seventy four patients
were included in this study. A40 peptide concentration was assessed using the 2
immunoassays for all patients.

S240

Validation: Total imprecision (21 days) gave CV% at 103 g/dL lower than 5%, CV%
at 136 g/dL lower than 4% and at 269 g/dL lower than 2%. LOD was 13.1 g/dL.
The test was linear from 19 g/dL up to 500 g/dL. On board reagent stability was
up to 35 days and on board calibration stability was up to 21 days. Compared vs a
commercial Ferrozine method (n = 82, samples between 6.5 and 486 g/dL) linear
regression gave y = 1.16x - 9.57 and r = 0.998. Compared versus Sentinel UIBC
Liquid REF 17639 (n = 82, samples between 41.5 and 466 g/dL) linear regression
gave y = 1.10x - 10.3 and r = 0.997. Bilirubin (up to 66 mg/dL), hemoglobin (up to
100 mg/dL) and triglycerides (up to 1000 mg/dL) did not interfere.
Conclusions Analytical and clinical performances of new liquid UIBC method
on Architect c4000 analyzer meets the acceptance criteria and it shows all the
requirements for its use as routine clinical chemistry assay.

B-376
Performance characteristics of a cystatin C immunoassay on the Beckman
Coulter AU5800, AU680 and IMMAGE 800 Systems

C. Townsley, T. Nilsen. Gentian AS, Moss, Norway

Background: Cystatin C is a biomarker of kidney function. Its measurement can
be used to estimate glomerular filtration rate. The objective was to evaluate the
performance characteristics of this assay on the Beckman Coulter AU5800, AU680
and IMMAGE 800 systems.
Methods: A cystatin C particle-enhanced turbidimetric immunoassay (Gentian,
Norway) with avian antibodies is standardized against the international calibrator
standard ERM-DA471/IFCC and holds a current FDA 510k for use on other
platforms. Measurement was carried out on the Beckman Coulter AU5800, AU680

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Proteins/Enzymes

Wednesday, July 30, 9:30 am 5:00 pm

and IMMAGE 800 systems at different sites between 2008 and 2013 using a cystatin c
immunoassay from Gentian, Norway. Different reagent lots were used and new serum
samples and dilution series were made for each site study in the case of precision,
linearity, security zone, interference, recovery and limit of quantification. Protocols
based on CLSI guidelines were used.
Results: The measuring range is 0.45 - 8.0 mg/L, linearity was proven with a
minimum range of 0.45-6.9 mg/L on all instruments with prozone of 32 mg/L. The
total C.V. for precision ranged from 1.5 - 6.2% with total C.V. at 1 mg/L of 2.7%.
LoQ for all instruments was observed to be <0.5 mg/L. Interference studies with
potential interferents, Intralipid (10 g/L), hemoglobin (6 g/L) and bilirubin (200
mg/L) showed no interference. Total analysis time was 10 minutes.
Conclusion: The Gentian Cystatin C immunoassay is validated for use on the
Beckman Coulter AU5800, AU680 and IMMAGE 800 systems. The assay shows
acceptable performance characteristics for measuring cystatin C in human serum and
plasma samples on these systems.
Summary of results for studies on AU5800, AU680 and IMMAGE 800
AU5800
AU680
IMMAGE 800
Linearity (mg/L)
0.45 7.07 0.44 - 9.02 0.37 - 6.9
Recovery (%)
96 - 100
98 - 105
100 - 110
Security zone (mg/L)
32
40
40
Precision (total C.V.%)
1.62 - 3.42 1.51 - 2.44 2.7 - 6.2
Bilirubin interference testing
400 mg/L 200 mg/L 800 mg/L
Hemoglobin interference testing 6 g/L
8 g/L
8 g/L
Intralipid interference testing
10 g/L
16 g/L
16 g/L

B-378
New algorithm for alpha-1-antitrypsin (AAT) deficiency investigation including
high resolution capillary zone electrophoresis (CZE-HR) for serum protein
electrophoresis (SPE) for screening

J. P. mond, C. Bergeron, M. Beaulieu. University of Montral Hospital

Center (CHUM), Montral, QC, Canada
Background: AAT inhibits proteases such as neutrophil elastase. AAT deficiency
is an underdiagnosed condition, which predisposes individuals to early onset of
emphysema. We propose an algorithm in which 1st-step is based on AAT nephelometric
measurement in conjunction with high resolution capillary zone electrophoresis
(CZE-HR) SPE. CZE-HR SPE allows clear separation and quantification of AAT in
an isolated fraction along with assessment of clinical conditions affecting the accuracy
of AAT quantification. Genotyping and phenotyping are done in sequence only if
required. We aim to review this algorithm performance after 14 month utilisation.
Methods: Implementation of a new algorithm using i) AAT measurement and CZEHR SPE as screening and interpretation tools, and if required ii) genotyping in 2nd
step, iii) addition of phenotyping in 3rd step. AAT concentration decision cut-off is
1.15g/L in normal CZE-HR SPE patients and 2.0g/L in inflamed ones. Integrated
interpretation report is always provided. This algorithm replaces individual AAT
quantification and/or genotyping requests. CZE-HR SPE was conducted on Sebia
Capillarys 2 automated system with high resolution buffer. Retrospective 14 month
utilisation study is based on 172 consecutive medical requests for AAT deficiency
investigation.

B-377
Evaluation of new Lipase Color Liquid on Architect c4000 analyzer

M. Gramegna, L. Politi, R. Lucini. Sentinel CH. SpA, Milan, Italy

Objective The objective of this study is to evaluate the analytical and clinical
performances of new Lipase Color Liquid reagent on Architect c4000 analyzer.
Relevance Lipase enzymes are produced in the pancreas and also secreted in small
amounts by the salivary glands as well as by gastric, pulmonary and intestinal mucosa.
Determination of lipase is used for diagnosis and treatment of diseases of pancreas
such as acute and chronic pancreatitis and obstruction of the pancreatic duct.
Methodology The method for the determination of lipase is based on the cleavage
of specific chromogenic lipase substrate 1,2-O-dilauryl-rac-glycero-3-glutaric
acid (6-methylresorufin)-ester emulsified in stabilized micro-particles. In the
presence of specific activators of pancreatic lipase as colipase, calcium ions and
bile acids, the substrate is converted in 1,2-O-dilauryl-rac-glycerol and glutaric
acid-6-methylresorufin-ester which decomposes spontaneously in glutaric acid
and methylresorufin. The increase of absorbance at 580 nm, due to methylresorufin
formation, is proportional to the activity of lipase in the sample. For the evaluation
of this reagent, modified CLSI protocols were adopted. Acceptance criteria for total
imprecision were 5% for lipase samples. LOD should be 3.0 U/L. LOQ was
defined as the analyte concentration at which the % CV is less than 20%. The method
should be linear up to 300 U/L. Method comparison was evaluated by comparing
new Lipase Color Liquid reagent on Architect c4000 versus commercial method on
Hitachi Modular P.
Validation Total imprecision (during 40 days on 3 level samples) gave CV% at 24
U/L (L1), 67 U/L (L2) and 157 U/L (L3) lower than 5%. LOD was 1.5 U/L. LOQ was
3.0 U/L. The test was linear from 0 U/L up to 600 U/L. On board calibration stability
and reagent on board stability were up to 40 days. This lipase test (y) was compared
with reference method (x) that uses the same substrate (methylresorufin) and gave the
following results: y = 1.07x + 5.78; r = 0.99; n = 74. The test is not affected by the
presence of hemoglobin up to 150 mg/dL, bilirubin (unconjugated and coniugated) up
to 66 mg/dL. The test is affected by the presence of lipids up to 100 mg/dL.
Conclusions Analytical and clinical performances of new Lipase Color Liquid reagent
on Architect c4000 analyzer meets the requirements for its use as colorimetric test to
measure lipase in serum. Specificity and better precision make this assay very suitable
for the routine measurement of this critical analyte.

Results: Our population consists of 172 individuals (83 men, 89 women; age (years;
mean 2SD) of 51.3 34.6 and 52.5 27.4, respectively). CZE-HR SPE allows
AAT detection of as low as 0.07 g/L (7 mg/dL). In our population, investigation was
conclusive after 1st step in 58% of cases (n=99) where genotyping was not required.
Genotyping lead to be conclusive in an additional 25% of patients (n=43), whereas
phenotyping was required in 17% (n=30). Overall, detection rate of deficiency
was 11%: i) deficiency associated with common variants (SS, SZ, ZZ) was 7.0%
(n=12) and ii) additional 7 patients (4.0%) require SERPINA1 gene sequencing
for confirmation of rare non-S or non-Z deficiency variants. We identified also 23
heterozygotes carriers (13.4%; 17 patients with S allele and 6 with Z allele). One
case previously genotyped as MS was retested with the algorithm approach, and
was correctly phenotyped as Pi*SNull. Despite adding phenotyping testing in our
repertoire, the overall cost of investigation has decreased by 30% when using our
new algorithm strategy as compared to our previous one (AAT measurement and
genotyping for all).
Conclusion: High resolution automated SPE (CZE-HR) is a powerful tool to assess
AAT and to determine many clinical states affecting accuracy of AAT quantification
by common techniques. Those conditions may impact screening effectiveness.
Implementation of our AAT deficiency screening algorithm has allowed us to increase
the efficiency of investigation while eliminating genotyping in majority of cases.
Moreover, we were able to perform expensive phenotyping testing in our laboratory
repertoire while reducing overall testing cost by 30%. This allows us to selectively
require outside sequencing services for only 4.0% of tested patients with unusual
deficiency-associated variants.

B-379
Biochip Array Technology Rapidly Identifies a Platelet-Derived Alzheimers
Disease-Specific Phenotype

M. Veitinger1, R. Oehler2, E. Umlauf1, R. Baumgartner1, C. Gerner3, R.

Babeluk2, J. Attems4, G. Mitulovic5, E. Rappold1, J. Lamont6, M. Zellner1.
1
Institute of Physiology, Medical University of Vienna, Vienna, Austria,
2
Surgical Research Laboratories, Medical University of Vienna, Vienna,
Austria, 3Department of Medicine I , Medical University of Vienna, Vienna,
Austria, 4Institute for Ageing and Health, Newcastle University, Newcaste
upon Tyne, United Kingdom, 5Department of Medical and Clinical
Laboratory Diagnostics, Medical University of Vienna, Vienna, Austria,
6
Randox Laboratories Limited, Crumlin, United Kingdom
Background: Globally >35 million people have Alzheimers disease (AD) or a related
dementia. Current diagnostic tests use neuropsychological review and brain scan
and a number of studies have reported AD-specific cerebrospinal fluid biomarkers
but analysis requires invasive lumbar puncture. There is an urgent need to develop
a new minimally invasive diagnostic procedure. In this study platelets, which share
biochemical features with neurons, were used as a surrogate to characterize ADmodulated proteins. The combination of biochip array technology with a powerful
biomarker algorithm is innovative and generates a reliable ante mortem AD test.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S241

Proteins/Enzymes

Wednesday, July 30, 9:30 am 5:00 pm

Methods: Ethics was granted and 62 patients with clinically suspected AD and 63 age/
sex-matched control subjects were enrolled. 2D-DIGE was performed on gel-filtered
platelets to resolve alkaline platelet proteins using pH6-9 strips and acidic proteins on
pH4-7, subsequently separated in the
second dimension using 11.5%SDS-PAGE. DeCyder software was applied to detect
spots and all sample gels were matched with the master gel. Tryptic digest identified
peptides using Spectrum Mill MS Proteomics Workbench software and UniProt
database. APOE 4 and GSTO1*A140 genotyping was performed. Platelet rich plasma
was derived from whole blood (20min, 120 x g) and stored at -800C. After thawing,
the platelets were isolated and lysed with SDS before addition of 2%BSA/PBS to bind
excess SDS. This lysate was applied to the protein biochip employing immunoassay
sandwich principles. The protein biochip was subsequently imaged on the Evidence
Investigator analyser.
Results: Monoamine-oxidase-B displayed the most significant increase in spot density
in AD patients.The decrease in ApoE spot density was attributed to isoform 3 and
accordingly, this spot exhibited lower detection in 4-positive patients. A second ApoE
spot exhibited increased density and was assigned to isoform 4. The fourth strongest
confirmed AD-related biomarker was identified as tropomyosin1 (Tm1) and adjacent
spots were also identified as Tm1. Glutathione S-transferase omega1 (GSTO1) spots
displayed a strong association with the APOE-genotype in AD patients. APOEgenotyping revealed significantly more AD APOE 4 carriers(66%)than the control
group(11%). Genotyping revealed that exclusively two GSTO1*A140 alleles were
present in non-APOE 4 AD patients (n=20) relative to 38% in controls (30% in nonAPOE 4 controls) and 32% in APOE 4-positive patients. Biochip determination of
the GSTO1*A140 and APOE 4 allele count identified 98% of all samples correctly
for the GSTO1 SNP rs4925 genotype and 100% correct genotyping was achieved for
APOE 4 by normalization with either ERK2 or panApoE concentrations. Biochip
analysis with the cellular loading control replicated also the higher expression of both
quantitative markers Tm1 and MaoB in patients relative to controls.
Conclusion: The most powerful AD biomarker algorithm was identified following
combinatorial review of the proteins/isoforms that exhibited the most significant fold
change. This study has demonstrated the utility of measuring multiple biomarkers from
a single platelet preparation to aid the diagnosis of late-onset AD. The combination
of biochip array technology with a powerful biomarker algorithm is innovative,
generating a reliable ante mortem AD test. It offers the potential to rapidly detect an
AD-specific phenotype using routine blood-based clinical screening.

B-380
The Value of Serum Dipeptidyl Peptidase-IV in Early Detection of
Mucopolysaccharidosis

I. Kurt1, E. Sertoglu2, I. Okur3, S. Tapan1, M. Uyanik1, H. Kayadibi4,

F. Ezgu3, H. I. Aydin5, A. Hasanoglu3. 1Gulhane School of Medicine,
Department of Clinical Biochemistry, Ankara, Turkey, 2Ankara Mevki
Military Hospital, Anittepe Dispensary, Biochemistry Laboratory, Ankara,
Turkey, 3Gazi University Faculty of Medicine, Department of Pediatric
Metabolism and Nutrition, Ankara, Turkey, 4Adana Military Hospital,
Department of Biochemistry, Adana, Turkey, 5Gulhane School of Medicine,
Department of Pediatric Metabolism, Ankara, Turkey
Background/Aim: The mucopolysaccharidosis (MPS) are a group of eleven
lysosomal storage disorders (LSDs) that result from the absence or malfunctioning
of enzymes required for the breakdown of glycosaminoglycans (GAGs). Early
recognition of MPS allows for immediate treatment, before the onset of irreversible
impairment, and, in many cases, prevention of catastrophic health outcomes, including
death. We aimed to investigate the diagnostic utility of serum dipeptidyl peptidase-IV
(DPP-IV) enzyme activity, urinary GAG/Cre ratio, total adenosine deaminase (ADA)
and ADA-1 isoenzyme activity in the diagnosis of MPS.
Material and Method: 31 MPS patients which were previously diagnosed by clinical
and enzymatic analysis were included in the study along with 31 healthy controls
matched with patients for age and gender (8 (4 - 14) years, 61% male for patients and
8 (5 - 11) years, 58% for controls). Serum DPP-IV enzyme activity was measured
according to Beesley et al and reported as nmol/min/mL of plasma. Serum total ADA
and ADA-1 isoenzyme activity was measured according to Guisty G, Galanti B
colorimetric method and reported as U/L. Urinary GAG concentration was measured
by dimethylmethylene blue method, creatinine was measured by the modified Jaffe
method and results were reported as mg GAGs/mmol creatinine. All of the statistical
analyses were carried out by using the SPSS (V15) package for Windows.
Results: Serum DPP-IV enzyme activity, urinary GAG/Cre ratio, total ADA and
ADA-1 isoenzyme activity were significantly higher in patients than in controls
(p< 0.001, p< 0.001, p= 0.038 and p= 0.006, respectively). There were significant
correlations between serum DPP-IV enzyme activity and urinary GAG/Cre ratios,

S242

ADA-1 activity, ADA-1/total ADA (r= 0.498, p< 0.001; r= 0.348, p= 0.006; r= 0.270,
p= 0.034, respectively). Area under ROC curve for DPP-IV enzyme activity was
0.988, p< 0.001 and for urinary GAG/Cre ratio was 0.986, p< 0.001. DPP-IV enzyme
activity and urinary GAG/Cre ratio were the most significant parameters according
to the univariate logistic regression analysis (p= 0.001 and p< 0.001, respectively).
Conclusion: According to our results, measurement of plasma DPP-IV activity can
be used as a biomarker for MPS screening. However, no prominent diagnostic value
emerged (so far) for the soluble plasma DPP-IV, although methods for plasma DPP-IV
activity measurements are available since the 1980s. Although many cases are needed
to decide more precisely, the results of this study indicate that the measurement of
serum DPP-IV activity can be used as first-line screening test complementary to
urinary GAG/Cre ratio for MPS screening.

B-381
Macrophage Inflammatory Protein (MIP-1), an Early Biomarker of Chronic
Kidney Disease

M. Summers, T. McFadden, E. Healy, C. Richardson, J. Lamont, R.

McConnell, S. FitzGerald. Randox Laboratories Limited, Crumlin, United
Kingdom
Background: Chronic kidney disease (CKD) describes abnormal kidney function and/
or structure. It is frequently unrecognised and presents with progressive decline of
renal function leading to end-stage renal failure and death. The Modification of Diet in
Renal Disease (MDRD) classification of renal disease describes the progressive stages
of the disease (stages 1-5) with respect to the estimated glomerular filtration rate
(eGFR). The complexity in diagnosing a patient with CKD at an early stage of disease
has led to most patients not receiving a diagnosis until the disease has progressed to
an advanced stage. Currently there are no reliable methods for assessing early stage
kidney disease. Inflammation plays a key role in the developmental progression of
CKD and identification of inflammatory biomarkers in patients suspected of having
CKD would aid early diagnosis. This study aimed to investigate the potential of MIP1 as an early biomarker of CKD, which would aid the diagnosis of early stage CKD.
Methods: MIP-1 was assessed in serum samples from a total of 202 subjects: 152
CKD patients (51 Stage 1, 50 Stage 2 and 51 Stage 3) and 50 healthy controls. The
analysis was performed with a biochip based immunoassay applied to the Evidence
Investigator analyser. Statistical analysis was performed using MedCalc v12.5, all
data represented as median [95% CI].
Results: Differences in concentration of MIP-1 across the disease groups and
controls were initially assessed using the Kruskal-Wallis test and MIP-1 displayed
significant difference between the diseased subjects and the healthy controls
(significance determined as p<0.05). Post-hoc analysis was performed comparing
CKD groups with controls using Mann-Whitney (with Bonferroni correction). This
showed that MIP-1 displayed significantly higher median concentrations at all CKD
stages (Stage 1-3) compared to control. (10.39 [9.0-13.77], 8.7 [7.29-11.24], 11.9
[9.89-14.23] pg/ml respectively; p<0.0001 for all) compared to control (5.36 [4.895.81] pg/ml). Receiver Operating Characteristic (ROC) curve analysis was conducted
to assess diagnostic performance of diseased versus healthy subjects. Area under the
curve was determined as 0.856(95% CI: 0.800-0.901).
Conclusion: The results indicate increased median concentration of MIP-1 in the
serum of CKD patients (stages1-3) compared to controls. This study highlights the
potential utility of MIP-1 in diagnosing early stage CKD. Early identification will
enable effective clinical management and prevent progression to a later stage.

B-382
FIA-MS/MS multiplex enzyme assay for screening oligosaccharidoses

K. K. Nickander1, S. E. Hofherr2, M. J. Magera1, D. Matern1. 1Mayo Clinic,

Rochester, MN, 2Childrens National Medical Center, Washington, DC
Background: Oligosaccharidoses are autosomal recessive lysosomal storage
disorders of glycoprotein catabolism. Current clinical testing for oligosaccharidoses
is performed by a thin layer chromatography (TLC) screen lacking sensitivity and
specificity. A positive TLC screen is followed by individual fluorometric enzyme
assays in leukocytes or fibroblasts, which although specific, address only one enzyme
at a time. We addressed this widespread deficiency by developing a multiplexed
enzyme assay to screen for several oligosaccharidoses, including -mannosidosis,
-mannosidosis, Schindler disease, sialidosis, galactosialidosis, GM1 gangliosidosis,
Morquio B disease, fucosidosis, and mucolipidoses II /, III /, and III , utilizing
flow injection analysis and tandem mass spectrometry (FIA-MS/MS).
Methods: Leukocytes or fibroblast lysates (100 L) prepared in 0.1 M citrate

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Proteins/Enzymes

Wednesday, July 30, 9:30 am 5:00 pm

phosphate buffer, pH 4.5, are incubated overnight with 3 different substrate mixes (10
L) then extracted using solid phase extraction and FIA-MS/MS (API 3200; AB Sciex,
Framingham, MA). Substrate mix 1 was prepared with 0.2 mM 4-methylumbelliferyl
-D-mannopyranoside, 0.8 mM 4-nitrophenyl -D-mannopyranoside, and 0.8 mM
4-methylumbelliferyl N-acetyl--D-galactosaminide, substrate mix 2 was prepared
with 0.4 mM 4-methylumbelliferyl-N-acetyl--D-neuraminic acid sodium salt
hydrate and substrate mix 3 was prepared with 0.4 mM 4-methylumbelliferyl--Dgalactopyranoside and 0.4 mM 4-methylumbelliferyl -L-fucopyranoside. All mixes
also included 0.2 mM umbelliferone as an internal control. Linearity was assessed
by 1:2 dilutions of substrate starting at 4 times the normal starting concentration and
processed with heat inactivated matrix. Reference intervals were determined from 157
leukocyte and 253 fibroblast specimens. Accuracy was determined from 15 known
deficient specimens.
Results: Precision (inter-assay and intra-assay) performance was within acceptable
limits (20% CV). Method response for the individual substrates demonstrated
acceptable linearity from 6.25-400% the normal starting concentration, with R2
values ranging 0.982-0.999. Acceptable clinical utility was demonstrated by correct
identification of 100% (N=15) of specimens with decreased or minimal residual
enzyme activity.
Conclusions: The FIA-MS/MS method provides a reliable alternative to TLC,
improving sensitivity and specificity without the need for follow-up assays with
individual enzymes. Six different enzymes are assayed screening for 9 different
oligosaccharidoses.

B-383
Assessment of Albuminuria using High Performance Liquid Chromatography
in Diabetic and Nondiabetic Patients

R. Prusa, M. Fortova, E. Klapkova, K. Kotaska. Faculty Hospital Motol,

Prague 5, Czech Republic
Background: Many studies have found higher values of urinary albumin reported
using high performance liquid chromatography (HPLC) in comparison with
immunochemical methods. Two opposite hypotheses explaining the difference
between these methods are as follows: 1) the presence of immunounreactive albumin
in urine that is not detected using immunochemical methods, 2) the presence of coeluting proteins that falsely overestimate the results on HPLC. The aim of our study
was the implementation of HPLC method for albuminuria, testing the hypothesis
about co-eluting proteins, and comparison of albuminuria assessed using HPLC and
immunoturbidimetric methods in diabetic and nondiabetic patient samples.
Methods: We developed the HPLC method for detection of albuminuria under these
chromatographic conditions: multilinear gradient consisting of water (solvent A),
acetonitrile (solvent B), 100 mM NaH2PO4 (solvent C), 1 % trifluoracetic acid (solvent
D), flow rate 2 mL/min, temperature 22 C, UV detection in the wavelength 280 nm,
chromatographic column Zorbax 300 SB-C3, liquid chromatograph Agilent 1200
(Agilent Technologies, USA). We analyzed two mixtures of urine. The first one was
prepared from 30 patient urine samples with immunoturbidimetrically physiological
albuminuria, to which we added albumin standard. The second mixture was prepared
from 30 patient urine samples with mild albuminuria. We compared albuminuria
assessed using HPLC with the immunoturbidimetric method (automatic analyser
Cobas Integra 400, Roche Diagnostics GmbH) in two patient groups: 636 diabetics
[345 males and 291 females, mean age 50.1 years (1-97)] and 456 nondiabetics [250
males and 206 females, mean age 33.9 years (0.1-91)].
Results: Transferrin, -1-acid glycoprotein, -1-antitrypsin, -1-antichymotrypsin
and hemopexin do not interfere with albumin in HPLC method. The elution curve
of prealbumin splits into several peaks, of which a few interfere with albumin. This
interference has no clinical importance. In mixture 1 we did not find a significant
difference between the albuminuria assessed using both methods (79 mg/L vs. 82
mg/L), while in mixture 2 we measured over 26 % higher albuminuria using HPLC
(79 mg/L vs. 100 mg/L). These results suggest the existence of immunounreactive
albumin. We found a statistically significant difference between the methods in both
patient groups (14.6 19.3 mg/L immunoturbidimetrically vs. 25.3 21.1 mg/L HPLC
in diabetics, 30.1 23.6 mg/L immunoturbidimetrically vs. 41.2 27.6 mg/L HPLC in
nondiabetics, median standard error of the mean, p < 0.0001, Mann-Whitney test).
Conclusion: Our results prove that the HPLC method for albumin detection is more
sensitive than immunoturbidimetry. We did not confirm nonspecificity of the HPLC
method. We found statistically significant higher concentrations of urinary albumin
using HPLC in both diabetic and nondiabetic patients.
Acknowledgment: Supported by the project (Ministry of Health, Czech Republic)
for conceptual development of research organization 00064203 (University Hospital
Motol, Prague,Czech Republic).

B-384
Performance Evaluation of the AnshLite Myelin Basic Protein
Chemiluminescent Immunoassay Using Cerebrospinal Fluid

J. Lu1, D. G. Grenache2. 1ARUP Institute for Clinical and Experimental

Pathology, Salt Lake City, UT, 2Department of Pathology, University of
Utah School of Medicine, Salt Lake City, UT
Background: Myelin is the insulating sheath surrounding neurons. Myelin basic
protein (MBP) accounts for about one-third of total central nervous system (CNS)
myelin protein. The cerebrospinal fluid (CSF) concentration of MBP increases in
response to neuronal damage allowing the measurement of CSF MBP to be used
clinically as a nonspecific marker of CNS inflammation. The objective of this study was
to evaluate the performance characteristics of the AnshLite MBP chemiluminescent
immunoassay (CLIA) for the quantitative determination of MBP in CSF.
Methods: MBP was measured using the AnshLite MBP CLIA (Ansh Labs, Webster,
TX, USA). Performance characteristics including precision, linearity, analytical
sensitivity, recovery, method comparison, MBP stability, the effects of freeze/thaw
cycles, and the reference were evaluated using residual CSF specimens sent to ARUP
Laboratories. The University of Utahs Institutional Review Board approved the
project.
Results: Precision was determined using CSF pools with MBP concentrations of 3.51
and 1.54 ng/mL assayed in three replicates once each day for ten days. Within-run
and total CVs were 9.6 and 14.8%, respectively. Linearity was determined by serially
diluting a high-MBP CSF sample with the zero calibrator to create six samples each
tested in two replicates. The assay was linear within the measuring range to 11.25
ng/mL (linear regression y=0.98(x)+0.05, R=1.00). The limit of blank (LOB) and
limit of detection (LOD) were determined by testing the zero calibrator ten times
and 0.45 ng/mL calibrator seven times. The LOB was 0.01 ng/mL calculated as the
mean concentration added to 3 SD. The LOD was 0.08 ng/mL calculated as the LOB
added to 3 SD. Accuracy was determined by recovery studies performed by adding
volumes of one of two calibrators (6 and 14 ng/mL) to two CSF samples with MBP
concentrations of 1.18 and 0.34 ng/mL. Calculated recovery ranged from 86 to 109%.
A method comparison using 42 samples with an MBP concentration range of 0.3058.57 ng/mL was performed using the Beckman MBP ELISA (Beckman Coulter, Inc.,
Brea, CA) as the comparator method. Deming regression yielded y=2.75(x)-2.68,
R=0.95, Sx/y=4.78. MBP stability was determined by storing two CSF specimens
with MBP concentrations of 6.77 and 1.49 ng/mL at room temperature for two days,
4C for two weeks, and -20C for three weeks. MBP decreased <12% relative to
time zero under all conditions. The effects of up to three freeze/thaw cycles were
evaluated by testing two CSF samples with MBP concentrations of 7.02 and 1.59 ng/
mL after each freeze/thaw cycle. The MBP concentration changed by -12.6, +6.94,
and -18.0% relative to time zero after each cycle, respectively. The reference interval
was established using 130 CSF specimens that were negative for oligoclonal bands.
Using non-parametric analysis, this was determined to be 0.0 to 5.5 ng/mL.
Conclusions: The AnshLite MBP CLIA demonstrates acceptable performance
characteristics for quantifying MBP in CSF. MBP is stable for two days at room
temperature, two weeks at 4C and three weeks at -20C. Freeze/thaw cycles of CSF
samples for MBP testing should be avoided.

B-385
Rationalising not rationing: the case of serum protein electrophoresis (SPE)

R. M. Dorizzi, V. Zanardi, P. De Vita, R. Agnoletti, A. Alberelli, S. Babini.

Corelab, Laboratorio AVR, Pievesestina di Cesena, Italy
Background Serum protein electrophoresis (SPE) is highly requested in Italy. The
AVR laboratory, which serves 1,125,000 inhabitants, runs about 300,000 tests/year
while e.g. in Kent 10,557 SPEs are requested in the same period for a catchment
area of 759,000 inhabitants. Since research/monitoring of monoclonal components is
the only appropriate indication for this test, the AVR Laboratory carried since 2010
several actions to increase the appropriateness of the requests: adoption of automated
alerts in the CPOE, benchmarking of the requesting modes of GPs, frequent meetings
with GPs, changes in the report format. Since October 1st 2013 the report contains the
various fractions only when a monoclonal component (MC) is present; in the other
cases only the comment Monoclonal components are not noticeable is reported.
In January 2014 we carried out the first audit of the effect of this action in the fourth
trimester after its implementation.
Materials and methods We extracted the results of the SPEs carried in the fourth
quarter of the years 2009-2013 stored in the LIS (Noemalife, Bologna, Italy). All the
SPEs were performed using Capillarys (Sebia, Marcy lEtoile, France). We assessed

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S243

Proteins/Enzymes

Wednesday, July 30, 9:30 am 5:00 pm

the tests carried out in inpatients and outpatients and the percentage of detected MCs.
Results The results summarized in the table shows a steady decline of the demand for
SPE since 2009 and a constant increase in the percentage of detected MCs consistent
with a progressive increase of the appropriateness of the request. This trend is more
evident in inpatients and it could be suggested that a MC is more frequently detected
in patients with symptoms requiring hospitalization and undergoing more accurate
investigation.

replicates each day for ten days. Repeatability and within-laboratory CVs were 5.4
and 6.5% at 17.3 ng/mL (0.21 mg/g) and 13.8 and 14.5% at 66.9 ng/mL (0.84 mg/g),
respectively. A1A was stable in stool for a minimum of 2 days, 7 days and 3 months
at room temperature, 4 - 8 C and -20 C, respectively. A1A measured in timed stool
obtained from 45 healthy volunteers (21 males, 24 females, ages 21 - 61) ranged
from <0.002 to 0.59 mg/g stool. Using a robust skewed method, the reference limits
for A1A in stool and A1A clearance were 0.47 mg/g and 45 mL/day, respectively.

Conclusions 1) The laboratory can and should promote the appropriateness of

laboratory tests; 2) appropriateness can improve in relatively a short time; 3)
effectiveness of the actions can be measured with simple tools.

Conclusions: The ImmuChrom Human A1A ELISA demonstrates acceptable

performance for quantifying A1A in stool. The assay can also be used in conjunction
with the Roche Diagnostics Tina-quant A1A (ver.2) assay for assessing A1A clearance.

B-387
Lipolysis Suppresses Insulin Signaling and Glucose Uptake

G. R. Mullins, L. Wang, T. E. Harris. University of Virginia, Charlottesville,

VA
Background: Acute hyperglycemia often develops after trauma or major surgery,
particularly surgery within the abdominal cavity such as coronary bypass procedures.
This condition, termed stress-induced hyperglycemia, contributes to mortality and
delays healing in post-surgery and ICU patients and is a result of systemic insulin
resistance (IR) caused by the adaptive stress response. Although intensive insulin
treatment reduces hyperglycemia and mortality in some patients, insulin therapy after
surgery can result in hypoglycemic incidents, emphasizing the need to develop new
treatment strategies. Despite its clinical significance, the pathophysiology of postsurgery acute IR is largely unknown. Understanding the contributing molecular
mechanisms will provide new insights in developing therapeutic approaches to
improve patient outcomes after major surgery. It is known that -adrenergic stimulation
by catecholamines will inhibit insulin stimulated glucose uptake in adipose tissue.
This catabolic signaling leads to the release of stored fat through lipolysis, which may
play a role in the observed stress-induced IR and hyperglycemia. Our objective was to
investigate a possible mechanistic link between lipolysis and catecholamine-induced
IR in adipose tissue, which would likely contribute to stress-induced IR after surgery.

B-386
Performance Evaluation of a Polyclonal Based Fecal Alpha-1-Antitrypsin
ELISA

J. A. Erickson1, R. A. Jensen2, D. G. Grenache3. 1ARUP Institute for Clinical

and Experimental Pathology, ARUP Laboratories, Salt Lake City, UT,
2
Parasitology and Fecal Testing Technical Section, ARUP Laboratories,
Salt Lake City, UT, 3University of Utah School of Medicine, Department of
Pathology, Salt Lake City, UT
Background: Alpha-1-antitrypsin (A1A) is a serum protease inhibitor that inhibits
a wide range of proteases. A1A resists degradation by digestive enzymes and
therefore, can be used to detect the presence of serum proteins in the gastrointestinal
tract. The primary causes of protein-losing enteropathy can be divided into erosive
gastrointestinal disorders, non-erosive gastrointestinal disorders, and disorders
involving increased central venous pressure or mesenteric lymphatic obstruction. The
diagnosis of protein-losing enteropathy is most commonly based on the determination
of fecal A1A clearance. The purpose of this study was to assess the performance
characteristics of a polyclonal antibody based A1A ELISA for the measurement of
A1A in stool and to determine reference limits for A1A in stool and A1A clearance.
Methods: Stool samples included deidentified residual random stool specimens
sent to ARUP Laboratories for routine testing, and timed stool specimens and paired
serum obtained from healthy volunteers. A1A was extracted from stool and measured
using the ImmuChrom Human A1A ELISA (ImmuChrom, GmbH, Heppenheim,
Germany) according to the manufacturers instructions. Serum A1A was measured
using the Tina-quant A1A assay (ver.2) on a cobas 8000 modular analyzer (Roche
Diagnostics, Indianapolis, IN). The performance characteristics evaluated for the stool
ELISA were analytical sensitivity, linearity, accuracy, precision, and A1A stability in
stool. The University of Utahs Internal Review Board approved the project.

Methods/Results: We have used biochemical assays to investigate the effect

of catecholamine-induced lipolysis on insulin signaling and glucose uptake in
adipocytes. Activation of the cyclic AMP/Protein Kinase A pathway acutely
inhibits insulin signaling by disrupting mammalian Target of Rapamycin (mTOR)
complexes 1 and 2 in adipocytes. Activation of lipolysis is required for this PKAinduced inhibition of mTOR. The effect of lipolysis is also transferrable in vitro,
suggesting a lipid product of lipolysis may mediate mTOR complex dissociation. In
addition, the signaling lipid, referred to as the mTOR Dissociative Factor (mDF) is
downstream of Adipose Triglyceride Lipase (ATGL) and depleted by Diacylglycerol
Lipase (DAGL), suggesting it is likely a diacylglycerol released during lipolysis. Here
we summarize our progress in our effort to isolate and identify the mDF through
reverse-phase high-performance liquid chromatography (RP-HPLC). We show that
activation of the PKA pathway inhibits mTOR activity and adipocyte glucose uptake
by 40-50%, as measured by phosphor-specific immunoblots and radiometric glucose
uptake assays, respectively. Although significant, this inhibition is not seen when
lipolysis is genetically or pharmacologically blocked. This suggests that inhibition
of mTOR by lipolysis is a potential mechanism of dampened glucose uptake
during catecholaminergic signaling and provides new insight into stress-induced
hyperglycemia.
Conclusion: This study demonstrates a novel intracellular signaling mechanism
where activation of lipolysis controls the PI3K-Akt-mTOR pathway to inhibit insulin
signaling. Understanding this mechanism provides new insight into insulin resistance
that is concurrent with elevated lipolysis found during obesity or after major surgery.
Understanding the molecular mechanism of stress-induced insulin resistance will
lay the groundwork for future therapeutic treatments and improve patient outcomes
after surgery. In addition, identification of the mDF may provide a new biomarker for
predicting the risk of insulin resistance and hyperglycemia.

Results: The analytical sensitivity was 0.14 ng/mL (0.002 mg/g stool) determined
from 23 replicates of the sample diluent (mean + 3SD). Linearity was evaluated using
diluted stool extracts with elevated A1A concentrations (range, 1.5 - 90.0 ng/mL).
Linear regression produced results of y = 1.02x - 3.15 (r2 = 0.986). Accuracy was
determined from analyte recovery studies by adding sera of known A1A concentration
into previously measured stool extracts. Recovery (measured ng A1A / expected ng
A1A) ranged from 95.2 to 118.4%. Precision was determined from ten stool extracts
obtained from each of two random stool specimens, with one extract analyzed in five

S244

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

EDDP Screening (DRI)

Wednesday, July 30, 2014

TDM/Toxicology/DAU

B-389
Point of Care PK Quantitation Device for Pharmacokinetic Guided Dosing of
Paclitaxel as a Companion Diagnostic Device

C. Lee, C. Park, C. Hsiao, A. Trieu, S. Lee. Autotelic Inc, Fountain Valley,

CA
Background: Paclitaxel chemotherapy is the cornerstone of most anti-cancer
regimens due to its potent cytoxic activity against tumor cells. The variability of
paclitaxel dosing can be as high as 10X across patients, therefore, dosing at a fixed
dose even when adjusted for body weight will leave a significant portion of the patients
underdosed - getting no benefit from the treatment- and another portion overdosedgetting undue toxicity. Therefore, paclitaxel therapy would benefit from TDM guided
dosing. A full pharmacokinetic (PK) study is required to determine whether patients
are getting the appropriate and efficient dosage. However, at the current state, a full
pharmacokinetic requires not only multiple high-volume blood draws and extended
stays in the hospital but also testing method by LC/MS/MS, which are both timeconsuming and expensive. Here we describe the development of Point of Care
TDM (Therapeutic Drug Monitoring) device for PK guided dosing of paclitaxel as a
companion diagnostic device.
Methods: The 8A10 and 3C6 mAbs (mAbs against paclitaxel) were purified from the
antibody-rich harvested medium using MabSelect (GE Healthcare, Pittsburgh, PA).
Commercial antibodies (29B7B3C and 69E4A8E; Santa Cruz Biotechnology Inc.)
were also tested. To synthesize BSA-paclitaxel, we used the method of J-G Leu et al.
as described in Cancer Res. (1993) 53:1388-1391. BSA-paclitaxel and mAbs were
labeled with colloidal gold using the gold labeling kit from BioAssay Works, LLC,
Ijamsville, MD.
Results: Rapid test for paclitaxel based on the lateral flow system was developed.
Of the mAbs tested, only 8A10 and 3C6 were useful. The assay requires the unique
configuration of immobilizing the mAb against paclitaxel onto the membrane followed
by flowing the BSA-paclitaxel-colloidal gold through in presence of test analyte. This
configuration seems more effective than the traditional configuration where BSApaclitaxel is immobilized on the membrane and colloidal-gold labeled anti-paclitaxel
mAb as flow through. This resulted in a competitive assay for paclitaxel where
the signal decreased as the concentration of paclitaxel analyte in blood increased.
Coupled with the current lateral reader technology - especially the one developed by
Qiagen- a rapid quantitative assay for paclitaxel is possible. The assay demonstrated
good linearity and range suitable for paclitaxel TDM. The application of this assay
in a preclinical pharmacokinetic study yield reasonably comparable result to LC/MS
method.
Conclusions: A quantitative lateral flow platform coupled to a reader was developed
to easily detect the paclitaxel concentrations in small amount of blood samples.
Individual pharmacokinetic profiles can be obtained and used to determine the
suitable treatments. In addition, the lateral flow PK quantitative assay can be deployed
at point-of care (in home, doctors office or central lab).

B-390
Comparison of Roche Methadone Screening Assay with DRI Methadone
Metabolite (EDDP) Screening Assay

M. R. Sneidman, H. E. Yu. Geisinger Health System, Danville, PA

Geisinger Medical Laboratories offer immunoassay based screening. Methadone, one
of the drug classes we test, is a synthetic opioid used for narcotic addiction and pain
management. Urine drug testing is used to monitor medication compliance for these
patients. We observed significant number of false positive methadone screens (Roche).
In a five-month period, 88 urine samples were screened positive for methadone, but
27% were negative by GC-MS. Because of the high rate of false positive results, we
evaluated an alternative screening assay DRI Methadone Metabolite (EDDP). The
following results were obtained:

GC-MS Confirmation

POS (n=24)
NEG (n=0)
POS (n=4, false NEG methadone)
NEG (n=5)
NEG (n=1, false POS EDDP)
POS (n=0)
POS (n=10)
NEG (n=10, false POS methadone)
NEG (n=25)
POS (n=0)
NEG (n=15)
NEG (n=15)
The EDDP screening assay (DRI) was more accurate with no false negative and only
1 false positive out of 54 urine samples tested. We also noted that many false positive
methadone screens were from patients who were taking tramadol and suspected that
is the interfering substance causing the false positive results. We therefore replaced
the Roche methadone screening assay with the DRI methadone metabolite (EDDP)
screening assay.
POS (n=29)

Poster Session: 9:30 AM - 5:00 PM

Methadone
Screening (Roche)
POS (n=24)

B-391
Light-Sensitivity of Chlordiazepoxide

T. V. Hartman, M. Dallman, L. J. Langman, P. Jannetto. Mayo Clinic,

Rochester, MN
Objective Due to conflicting information in the literature, this study was conducted to
determine if Chlordiazepoxide (CDP) is light sensitive at a variety of sample storage
conditions.
Relevance If CDP patient specimens are not protected from light for greater than
24 hours, clinical results could be inaccurate leading to erroneous clinical results.
Furthermore, the light-sensitivity of CDP is not sufficiently addressed in the current
literature.
Methodology This study utilized High Performance Liquid Chromatography with
Ultra-Violet Visible Spectroscopy (HPLC/UV-Vis) to compare three identical serum
samples at various concentrations throughout the analytical measurement range (0.5
g/mL - 12.0 g/mL). Briefly, the three samples were split into 4 aliquots and stored
protected and unprotected from light at both ambient and refrigerated temperatures for
analysis on days 1, 3, 7, 16, 21, and 28.
Validation After 28 days, the three aliquots stored protected from light yielded
an average degradation of 20.7% at ambient storage and essentially negligible
degradation at refrigerated storage while the three aliquots not protected from light
yielded an average degradation of 93.5% at ambient storage and 22.8% at refrigerated
storage.
Conclusions Some degradation of CDP found at ambient temperatures was
determined to be due to the storage temperature; however, by comparing the lightprotected samples vs. the non-light protected samples at the ambient storage or
refrigerated storage temperatures, the data clearly shows significant degradation of
CDP due to the exposure to direct light which could significantly affect patient results.

B-392
Modified HPLC-UV Method without Evaporation Step for the Determination
of Serum Ribavarin Level in Patients with Hepatitis C

A. Khan1, I. Altraif1, P. Marquet2, A. Al-Othaim1, K. M. Khan3, N. Al

Hussain1, W. Tamimi1. 1King Fahad National Guard Hospital, King Saud
Bin Abdulaziz Univresity for Health Sciences, Riyadh, Saudi Arabia,
2
Universit de Limoges, faculte de medecine, Limoges, France, 3University
of Karachi, Karachi, Pakistan
Bcakground: The measurement of serum concentration of anti-viral drug ribavarin is
indicated for some patients with hepatitis C, for therapeutic drug monitoring. A simple
and fast high performance liquid chromatographic method for the determination of
ribavarin in serum sample is developed and validated according to FDA method
validation guidelines without the evaporation step. Method: Serum sample (500
L), internal standard (50 L), and 20 mM ammonium acetate buffer of pH =8.5
(500 L) were mixed and allowed to centrifuge for 5 minutes. After centrifugation,
the supernatant was transferred into phenyl boronic acid cartridges (previously
conditioned with 1mL of 3% formic acid and then with 1mL of 20mM ammonium
acetate buffer pH 8.5) for solid phase extraction followed by washing step with
20 mM ammonium acetate buffer (1 mL) under vacuum not exceeding 10 psi and
discarded. ribavarin and internal standard were eluted with 3% formic acid (300 L)
and was injected (90L) into HPLC system. Serum samples from 37 patients were run
simitnutenosly on both methods and compared. Results: A linearity between 0.1 - 6.25
mg/L with 0.06mg/l as limit of detection was obtained. The correlation coefficient
was 0.950 with p value of 0.92 showing a good reproducibility of results. The mean

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S245

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

accuracy was checked at three different concentrations and found to be 108% for each
level. The mean recoveries (extraction efficiency) of ribavarin and internal standard
from serum were found to be 65.5 % and 71.2% respectively at concentrations ranging
from 0.1 to 6.22 mg/L for ribavarin and 50mg/L for internal standard. The intra assay
precision were determined at 0.5, 1.2, & 2.2 mg/L and % CV were found to be 7.1%,
5.1% % 7.4% respectively whereas injection reproducibility were calculated 5.6%,
0.7% and 0.6% at three levels. Maximum 1hour and 30 minutes was consumed to
complete this assay whereas it takes about 3 hours and 30 minutes if the step of
evaporation and gravity elution was used in the older method. Conclusion: The newly
developed HPLC method was faster, accurate and sensitive. It may be applied for the
estimation of ribavarin level in serum samples.

B-393
A HPLC-High Resolution Accurate Mass Spectrometric Method for the
Quantification of Doxorubicin and Doxorubicinol

D. Sartori, A. Breaud, W. Clarke. Johns Hopkins University School of

Medicine, Baltimore, MD
Background: While chemotherapeutic agents have been commonly used in the
treatment in a majority of cancers, these drugs are typically associated with significant
adverse effects. Although chemotherapeutic regimens reduce mortality, morbidity
may be high due to significant drug-induced toxicities. The anthracycline antineoplastic agent doxorubicin is a common treatment modality for hematologic
malignancies, as well as a variety of solid tumors. Doxorubicin, and its active
metabolite, doxorubicinol, is associated with severe cardiotoxicity; therefore, it
is important to generate tools to measure doxorubicin concentrations following
intravenous administration to minimize drug-mediated toxicity. Both invasive (blood)
and non-invasive (saliva) collection schemes have been pursued for other drugs, and
this is largely due to ease of collection, particularly in pediatric populations. Here,
we present a liquid chromatographic-high resolution mass spectrometric (LC-HRMS)
method for the dual quantification of doxorubicin and its metabolite in both serum
and saliva.
Methods: Standard solutions of doxorubicin and its active metabolite, doxorubicinol,
as well isotopically-labeled internal standards for each, were prepared in water and
spiked into drug-free human serum and saliva. Following protein precipitation, a 10
l aliquot was injected onto our LC-HRMS system. Chromatographic separation
was achieved using a Thermo Scientific PFP HPLC column (50 2.1 mm, 5 m
particle size) and eluted under a gradient elution containing acetonitrile with 0.1%
formic acid. Both doxorubicin and doxorubicinol were detected over 5 minutes using
a Q-Exactive hybrid quadrupole-orbitrap mass analyzer (Thermo Scientific). A HESI
(heated electrospray ionization) source was used and the instrument was operated in
full scan (m/z 250-750) mode with a resolution of 70,000 at m/z 200.
Results: The high resolution Q-Exactive mass analyzer was able to detect both
doxorubicin (m/z 544.1813) and doxorubicinol (m/z 546.1970) in serum and saliva
with a mass tolerance of 5 ppm from theoretical mass. Assay development and
subsequent validation was performed following the recommendations of the FDA
for bioanalytical method validation. The analytical measuring range of the assay for
both doxorubicin and doxorubicinol was 5 to 1000 ng/ml. Calibrator solutions were
evaluated at drug concentrations of 5 ng/ml, 15 ng/ml, 50 ng/ml, 75 ng/ml, 125 ng/ml,
250 ng/ml, 500 ng/ml and 1000 ng/ml. Linearity was assessed as the average slope
of 1/x2 weighted linear regression analysis. Across five independent injections from
serum-extracted specimens, the average slope for doxorubicin and doxorubicinol
were 0.953 and 0.951, respectively. Quality control solutions were prepared in serum
with theoretical concentrations of 30 ng/ml and 400 ng/ml for both drugs. Intra-assay
precision across five injections were determined to be 2.9% and 2.5% for the parent
and metabolite drug, respectively.
Conclusion: A LC-HRMS assay has been developed and validated for the
simultaneous quantification of doxorubicin and its active metabolite in both serum
and saliva specimen sources.

B-394
Rapid Enzyme Hydrolysis by a Novel Recombinant Beta-Glucuronidase in
Benzodiazepine Urinalysis

A. A. Morris, S. A. Chester, E. C. Strickland, G. L. McIntire. Ameritox, Ltd,

Greensboro, NC
Background:Benzodiazepines are widely prescribed drugs that are readily abused
for their sedative effects and as such are very frequently targeted in therapeutic drug
monitoring. Only trace amounts of parent drug are present in urine due to extensive

S246

metabolism and conjugation of benzodiazepines and their glucuronides are major

urinary species. Hydrolysis to cleave the glucuronides prior to analysis is necessary
for improved detection. For benzodiazepine analysis, hydrolysis by enzyme is
preferred over acid because the latter produces benzophenones, which convolute result
interpretation. Enzyme hydrolysis can be costly and time-consuming, with reported
incubation times ranging from 0.5 to 20 hours. The assessment and application of
a novel recombinant beta-glucuronidase for rapid hydrolysis in benzodiazepine
urinalysis is presented.
Methods:IMCSzyme recombinant beta-glucuronidase was buffered to
recommended optimum pH and evaluated. Aliquots of drug-free urine fortified
separately with glucuronides of oxazepam, lorazepam, temazepam at 2500 ng/mL
were hydrolyzed with the enzyme in triplicate. The hydrolysis efficiency was assessed
at 55C at incubation times of 0, 15, 30 and 60 mins and at room temperature at
incubation times of 0, 5 and 10 mins. The optimized enzyme hydrolysis was applied to
20 randomly selected positive authentic urine samples for each analyte and compared
to hydrolysis by commonly used beta-glucuronidase from abalone under its validated
optimized conditions. Hydrolysis efficiency for alpha-hydroxyalprazolam glucuronide
was evaluated solely with patient samples positive for that compound. Hydrolyzed
urine samples were analyzed on a TLX-4 Multiplexed HPLC with Agilent 1200 Series
Binary Pumps coupled to a Thermo Scientific TSQ Quantum Ultra Triple-Stage
Quadrupole Mass Spectrometer using a previously validated method. Analytical
column performance was monitored.
Results:The validated benzodiazepines liquid chromatography tandem mass
spectrometry (LC/MS/MS) method had a linear range of 20-5000 ng/mL for oxazepam,
lorazepam, temazepam and alpha-hydroxyalprazolam. At 0 mins (immediately
run after addition), mean analyte recovery >92% from all hydrolyzed glucuronide
controls was observed at both 55C and room temperature. Complete hydrolysis of
the glucuronide controls (mean analyte recovery 100%) was observed at 15 mins
(shortest incubation time tested) at 55C and at 10 mins at room temperature. This
was considerably faster than the optimized incubation time of 30 mins for the abalone
beta-glucuronidase. Hydrolysis at room temperature was also more convenient,
eliminating heat activation. Mean analyte recovery changed by no more than -2%
at longer incubation times at both temperatures. In patient samples, total oxazepam,
lorazepam and temazepam compared well between the two enzyme sources.
Recovery of alpha-hydroxyalprazolam was >30% higher using the recombinant betaglucuronidase versus that from abalone, highlighting the differential hydrolysis of
alphahydroxyalprazolam by the abalone enzyme under conditions optimal for the
other analytes. The IMCSzyme recombinant beta-glucuronidase appeared to be a
cleaner extract and the maximum number of analytical column injections seen with
abalone-treated samples was exceeded with this recombinant source.
Conclusion:The superiority of the IMCSzyme recombinant beta-glucuronidase
was demonstrated with fast benzodiazepine hydrolysis at room temperature. The use
of this enzyme decreases processing time due to reduced incubation time, requires no
heat activation and affords improved column life.

B-395
Comparison of paired immunosuppressant levels in venous and dried blood
spot levels in post-transplant patients

J. A. Dickerson, K. Sadilkova, R. Jack. Seattle Childrens Hospital, Seattle,

WA
Background: Transplant patients are routinely and chronically monitored with
laboratory testing to assess risk of rejection, infection, and to optimize therapeutic drug
doses. Therapeutic drug monitoring of tacrolimus and sirolimus plays a significant role
in the clinical follow-up of transplant patients receiving immunosuppressant (IMS)
therapy. Success of transplant and favorable patient outcome relies on maintaining
adequate therapeutic drug levels. Drug dosing must be individualized due to the
narrow therapeutic range of these drugs, significant inter-individual variability, and
intra-individual variability in absorption and metabolism. We developed, validated,
and implemented a clinical liquid chromatography-mass spectrometry (LC-MS/MS)
assay for simultaneous quantitation of tacrolimus and sirolimus in dried blood spots
(DBS) collected remotely by patients and mailed into the laboratory. The purpose of
this research is to assess the clinical utility of remote collection of dried blood spots
for immunosuppressant monitoring, and compare the IMS level in paired collections
of venous whole blood and dried blood spots.
Methods: The validation of the LC-MS/MS assay in dried blood spots was described
previously, and the assay correlated well with our routine whole blood assay. To
clinically correlate sirolimus and tacrolimus levels in capillary blood collected from a
finger poke with venous whole blood, pediatric, post- transplant patients were asked
to provide up to three paired collections of DBS and venous whole blood (total of
25 collections per drug). To be eligible, Seattle Childrens patients needed to be

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

> 1 yr old, status post a heart, liver, or kidney transplant, and currently monitored
for sirolimus or tacrolimus levels. Thirty-one patients consented and completed at
least one paired collection. The phlebotomist only ordered the clinical whole blood
immunosuppressant level, but collected both the venous and the blood spot. The
participant took the dried blood spot card home with them with a pre-addressed,
postage-paid envelope and mailed it back to the lab. The concentration and the
turnaround times of the dried blood spot were compared with the whole blood sample.
The recorded data include the unique study ID number, age of the patient, organ
transplanted, immunosuppressant therapy, date and time of collection, whole blood
level, date and time DBS was received, date and time DBS was analyzed, and DBS
level.
Results: Tacrolimus in DBS correlated well with venous levels (y = 1.03x + 0.84,
R2= 0.93, n=25). Overall, a small, but statistically significant negative bias was
observed (-0.6 ng/mL, p = 0.0013). A chart review was performed to assess if
clinical management would have changed, and none of the cases revealed a clinically
significant change. Sirolimus in DBS also correlated with venous levels (y = 0.83x +
1.17, R2 = 0.84, n=23). Overall, a small, negative bias was observed which was not
statistically significant (-0.5 ng/mL, p = 0.09). Turnaround time from collection to
receipt of DBS cards was on average 5.8 days, and ranged from 3 to 18 days.
Conclusions: In summary, analysis of IMS levels in DBS is possible, and the difference
noted between capillary and venous blood is within the clinically acceptable limits.

B-396
A fast and sensitive high performance liquid chromatographic method for
measuring 6 tricyclic antidepressants in human plasma

C. Yuan, M. Burgyan, S. Wang. Cleveland Clinic, Cleveland, OH

Background: Monitoring certain tricyclic antidepressants (TCA) in blood is needed
due to the large inter- and intra-individual variations. The objective of this work was
to develop a high performance liquid chromatography (HPLC) method to measure
6 commonly prescribed TCA drugs, namely amitriptyline, imipramine, doxepin,
nortriptyline, desipramine, and nordoxepin in plasma. Methods: Five hundreds L
of plasma samples and 50 L internal standard solution (5 g/mL of trimipramine
and protriptyline in 0.1N HCl) were vortex mixed, followed by solid phase extraction
using C18 cartridges. Chromatographic separation was achieved on a cyanopropyl
bonded phase column (4.6x150 mm, 5 m) with an isocratic elution of 22:18:60 mix
of 10mM phosphate buffer (pH 7.0):methanol: acetonitrile at 1.8 mL/min. Detection
was by UV at 214nm and quantification was based on peak height ratios using 5
level calibration. This method was compared with a separate HPLC-UV method using
left-over patient specimens and spiked samples. Results: Baseline separation of all 8
compounds was achieved within 8.0 min. The obtained linearity ranges, mostly 20500 ng/mL, fully encompassed the therapeutical ranges (within 50-300 ng/mL) of all
the analytes. Precision was assessed at three different levels, and intra-assay and total
CVs were <7.8%. No interference was found from lipemic, hemolytic, icteric and
uremic plasma samples. Agreeable results were obtained in the method comparison
study (Table). Conclusion: This newly developed HPLC-UV method offers fast and
sensitive quantification of 6 common TCA drugs in human plasma.
Table 1. Method Comparison.
Standard
Correlation Error of Bias
n Slope (95% CI)*
Coefficient Estimate (%)
(ng/mL)
Amitriptyline 35 1.006 (0.968-1.044) -5.3 (-13.1-2.4) 0.9943
14.1
5.5
Imipramine 32 0.939 (0.885-0.993) 2.1 (-7.9-12.1) 0.9880
15.7
4.9
Doxepin
22 1.046 (0.995-1.097) -5.7 (-153.7-4.3) 0.9945
11.7
-0.4
Nortriptyline 33 1.032 (0.973-1.091) -7.8 (-19.5-3.8) 0.9878
20.2
3.2
Desipramine 32 0.963 (0.930-0.989) 1.1 (-4.0-6.2)
0.9967
8.2
3.3
Nordoxepin 23 0.967 (0.943-0.990) 0.9 (-3.6-5.5)
0.9986
5.7
1.5
*Numbers in parenthesis indicate 95% confidence intervals.
Intercept
(ng/mL)*

B-397
Quantitation of (-)-9-Tetrahydrocannabinol(THC) in Dried Blood Spots Using
LC-MS/MS

the bloodstream. THC is detectable for many hours after consumption and is a good
indicator of recent cannabis consumption. LC-MS/MS is considered a useful tool for
assessing the THC level in blood. The aim of the present study was to develop and
validate a much simplifier and efficient high-throughput LC-MS/MS approach for the
rapid quantification of THC in dry blood spot (DBS) samples.
Methods: The DBS was created by spotting 25uL spiked whole blood onto blank
Whatman 903 paper and allowing a full 24 hours to dry at room temperature. A 3 mm
punch was placed into a 1.5 mL vial with 100 uL exaction solution (acetonitrile with
10 mM ammonium acetate). The vial was vortexed for 10 minutes and soaked for
60 minutes. Then centrifuged and the supernatant transferred to an LC for injection.
No SPE or evaporation step was required. A Shimadzu UFLC XR system was used.
Sample was loaded onto Chromolith-RP18E column (100X3 mm, 3 um) held at 400C.
A gradient LC method was created at a flow rate of 600 ul/min and a total LC run
time of 4 minutes. Mobile phase A is 0.1% formic acid in 100% H2O and B is 0.1%
formic acid in 100% ACN. An IONICS 3Q 220 mass spectrometer system was used
which is equipped with heated coaxial flow ion source and a Hot Source-Induced
Desolvation (HSIDTM) interface was used.
Results: The Calibration curve of the neat THC solution showed a good linearity
over a range of 0.05-100 ng/mL with correlation value of R2=0.994. At LLOQ of 0.05
ng/mL, the accuracy is 96% and CV is less than 10%. For DBS sample, no matrix
interference was observed. Six calibration curves were generated with single injection
over the range of 1-100 ng/mL. Good linearity with 1/x weighting was obtained for
each curve with correlation value R2 >0.99. The average accuracies for the six sets of
THC spiked DBS samples at 1 ng/mL, 10 ng/mL, and 100 ng/mL are 100.9%, 99%,
and 99.4% respectively. And the CVs at 1 ng/mL, 10 ng/mL, and 100 ng/mL are 4.2%,
4.7% and 4%, respectively.
Conclusion: A fast, accurate, and precise LC-MS/MS method with IONICS 3Q 220
mass spectrometer was developed for direct measurement of THC in DBS. Significant
time can be saved in the absence of SPE or LLE sample preparation. In the neat
standard THC solution, an LLOQ of 0.05 ng/mL was achieved with accuracy of 96%
and CV of 3.6%. Six sets of single injection calibration curves for DBS extraction
showed good linearity over a range of 1 to 100 ng/mL. Averaged accuracy was
between 99 and 101%, and the CVs were < 5%. This LC-MS/MS method confirms its
clinical applicability for THC level monitoring in DBS.

B-398
Establishment of Expected Ranges for the Random Serum Concentration of
Lacosamide and Desmethyl Lacosamide

D. A. Payto, N. Foldvary-Schaefer, N. So, M. Bruton, S. Wang. Cleveland

Clinic, Cleveland, OH
Background: Lacosamide (LCM) is an antiepileptic drug approved by the Food and
Drug Administration for adjunctive treatment of partial onset seizures. LCM has a
novel mode of action by selectively enhancing the slow deactivation of the sodium
gated channels. LCM reaches peak concentration within 1-4 hours when taken
orally and has a half-life of 13 hours. In humans the major metabolite is O-desmethy
lacosamide (ODL). Theraputic drug monitoring (TDM) of LCM and ODL is used to
help optimize therapeutic dosing, while limiting adverse effects. TDM is also used for
establishing an individualized therapeutic range and to assess compliance to therapy.
Method: Random serum samples (n=45) were obtained from adult subjects taking
LCM adjunctive therapy for partial onset seizers as part of a clinical trial. The subjects
consisted of 7 males and 16 females. These subjects were on either a 200mg/day
or 400mg/day dose of Lacosamide. The random serum samples were analyzed on a
previously validated LC-MS/MS method for LCM and ODL. Results: A statistical
analysis was performed, in Excel, using the central 95% criteria to establish an
expected concentration range. For subjects taking 200mg/day of LCM the random
serum concentration of LCM and ODL was found to be 2.2 to 9.8 g/mL and up
to 1.6 g/mL respectively. For subjects taking 400mg/day LCM the random serum
concentration of LCM and ODL was found to 3.1 to 19.8 g/mL and 0.5 to 2.5 g/
mL respectively. Histograms, see figure, were constructed using EP Evaluator to show
distribution. Conclusion: The expected random concentration of LCM and ODL for
patients taking 200-400mg/day of LCM is 2.2 to 19.8 g/mL and up to 2.5 g/mL
respectively.

J. Ye, H. Qiao, E. Majdi, L. Cousins. IONICS Mass Spectrometry Group

Inc., Bolton, ON, Canada
Background: Tetrahydrocannabinol (THC) is the major psychoactive ingredient
in the cannabis plant. It has been shown that THC has complex effects on the
central nervous system, correlated with the development of anxiety and psychotic
disorders. After smoke inhaling of cannabis, THC is absorbed and incorporated into

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S247

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

counter or prescription drugs at 1000 ng/g, except dihydrocodeine for Opiates and
Oxycodone assays. ELISA screened positive samples were subsequently quantitated
using previously validated confirmatory methods to detect determinant drugs and
metabolites of each class at the same cut-off. We found that 83.6-98.0% of the ELISA
screened positive samples were eventually confirmed positive.
Conclusion: Fetus exposure to drugs results in adverse effects on newborn health,
and early intervention is crucial. Although GLC-MS is a preferred methodology
to detect minute amounts of harmful substances deposited into the umbilical cord,
solely depending on this technology is operationally difficult to support such critical
care. We validated and determined the 13-panel multiplex ELISA screen method
to be appropriate for implementation and equivalently sensitive as the more timeconsuming drug-class specific GLC-MS methods.

B-401
Evaluation of a liquid chromatography tandem mass spectrometry analytical
method for the quantification of a panel of antiepileptic drugs and their
metabolites in human plasma

C. De Nardi1, A. Morando2, A. Del Plato2. 1Thermo Fisher Scientific

GMBH, Dreieich, Germany, 2Ospedale La Colletta, Arenzano, Italy

B-400
13-Panel toxicological drug screen of umbilical cord tissues with enzyme-linked
immunosorbent assays (ELISAs)

I. Shu, M. Pilkington, C. A. Plate, J. Jones, D. Lewis. United States Drug

Testing Laboratories, Inc, Des Plaines, IL
Relevance: Current best practice for detecting in utero drug exposure is to test
newborns meconium samples. However, a meconium sample is not available
in 8-20% of births in the United States. Umbilical cord is readily available for
collection in every birth. Previous studies have shown that an umbilical cord drug
test performs equivalently well as a meconium drug test. The detection of drugs
requires high sensitivity and fast turn-around for further interventions. Our objective
is to develop and validate a multiplex 13-panel ELISA drug screen that provides
equivalent sensitivities as the chromatography-mass spectrometry (GLC-MS) based
confirmatory methods.
Method: An aliquot of approximately 0.5g of umbilical cord was homogenized in
3.0mL of acetone followed by centrifugation. The supernatant was filtered, 0.2%
succinic acid in acetone was added, the sample taken to dryness, and finally reconstituted
with 700 micro-liter buffer. Laboratory determined volumes of the extracts were
aliquoted to each of thirteen ELISA plates using automated liquid handling devices
for development according to drug assay manufacturer package inserts. The assay
principle is homogeneous-competitive immunoassay, where the intensity of the
developed color is inversely proportional to the samples drug concentration. The
absorbance of each sample well was normalized to that of the negative controls
(B/B0) of the same ELISA plate. Validation was performed according to Scientific
Working Group for Forensic Toxicology (SWGTOX) guidelines. We set the cut-off
levels (decision points) listed below: 0.1ng/g for Tetrahydrocannabinoids; 0.5ng/g
for Cocaine, Opiates, Oxycodone, and Buprenorphine; 1.0ng/g for Barbiturates;
2.0ng/g for Phencyclidine, Benzodiazepines, Methadone, and Meperidine; 4.0ng/g
for Propoxyphene and Tramadol; and 5.0ng/g for Methamphetamine.
Results: Five controls with concentrations ranging between 50-150% of cut-off levels
were prepared for analysis of precision and linearity. Coefficients of variation (CV%)
were 0.9-12.4% within-run (n = 4) and 2.9-19.4% between-run (5 runs, n=20). The
mean 2 standard deviations (SDs) for levels at 50% of cut-off did not overlap
with the decision points. Correlation coefficients (R2) of B/B0 versus concentrations
(expressed in logarithm) were 0.9567-0.9977, demonstrating acceptable linearity. The
mean B0 3.3SDs determined lower limits of detection ranging 3.0-41% of cut-off
levels. The immunoassays did not present hook effect and carry-over at least at 100
times of cut-off levels. The assays did not show interference from common over-the-

S248

Background: the evaluation of an analytical method for the quantification of 17

different antiepileptic drugs and metabolites in human plasma is reported. The
panel includes levetiracetam, theophylline, felbamate, lacosamide, rufinamide,
carbamazepine, oxcarbazepine, carbamazepine diol, carbamazepine epoxide,
10-hydroxycarbamazepine, PEMA, primidone, phenytoin, stiripentol, zonisamide,
phenobarbital and valproic acid. The method is based on liquid chromatography
tandem mass spectrometry performed on a Transcend LC system combined with
a TSQ Quantum Access MAX triple stage quadrupole mass spectrometer, both
from Thermo Scientific. The MassTox TDM Series A kit for antiepileptics from
Chromsystems was used for the scope; the kit included mobile phases, an analytical
column, calibrators, controls and extraction, precipitation and dilution buffers. The
calibration range covered the therapeutic window of concentrations in plasma for each
analyte.
Methods: following the instructions provided by the kit supplier, lyophilized
calibrators (on three levels) and controls (on two levels) were resuspended with
distilled water, protein precipitated using the precipitation buffer (containing 12
different internal standards), vortex-mixed, centrifuged and the supernatant diluted
with the dilution buffer prior to injection. Blank plasma samples from different donors
were also added to the batch. Each calibrator and control sample was prepared in
duplicate. The analytes of interest were divided into three groups and analyzed using
three different gradient elutions, with runtimes of 5.0, 3.8 and 2.2 minutes. A heated
electrospray source was used for sample ionization in both positive and negative mode
and detection was performed by Single Reaction Monitoring (SRM). Data acquisition
and processing was performed using Xcalibur software. Specificity, linearity and
accuracy of the method were evaluated for each analyte. Specificity was tested by
checking the absence of interfering peaks in plasma samples from different donors not
taking any antiepileptic drug. Linearity in the calibration range set by the kit supplier
was evaluated using a linear interpolation with 1/x weighting for all the analytes. The
percentage bias between nominal and experimental concentration for both calibrators
and control samples was used to assess the accuracy of the method; the adopted
acceptance criterion for accuracy was: bias within 15% for calibrators and within
20% for control samples.
Results: the analytical method proved to be specific and accurate, with no interfering
peaks and a maximum percentage bias within the acceptance criteria for both
calibrators and control samples. Linear calibration curves were obtained for all the
analytes of interest, with a correlation coefficient (R2) always above 0.99.
Conclusion: the reported method can be successfully applied to the quantification
of a large panel of antiepileptic drugs and metabolites in human plasma by liquid
chromatography tandem mass spectrometry using a TSQ Quantum Access MAX.

B-402
Quantitation of Pentobarbital in Serum Using Liquid ChromatographyTandem Mass Spectrometry (LC/MS-MS)

V. Ricchiuti, E. Chaffin, F. Lucas. University of Cincinnati Medical Center,

Cincinnati, OH
Background. Pentobarbital is a central nervous system depressant with sedative and
hypnotic properties. It is crucial to monitor levels with an acceptable turn-around
time. Gas Chromatography Flame Ionization Detection (GC/FID) is a commonly used

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

method for the analysis of barbiturates, but can be labor intensive. Our aim was to
validate a quantitative pentobarbital analytical method by LC/MS-MS which will be
used as a routine method in the clinical laboratory with a faster turn around time (TAT)
and will be less labor intensive than GC/FID.
Methods. Pentobarbital present in serum is extracted using a methanol/acetonitrile
protein precipitation, followed by dilution and analysis by Shimadzu Prominence
20A Liquid Chromatograph, followed by the AB SCIEX QTRAP 4500 Mass
Spectrometry System (LC/MS-MS). A 20 uL specimen is mixed with a methanolic,
deuterated internal standard (pentobarbital-D5), which initiates protein precipitation.
The proteins are then further precipitated with the addition of acetonitrile. A portion
of the supernatant is transferred and diluted with methanol:water (50:50) for injection
into the LC/MS-MS. Qualitative identification is based on the presence of the
specific MRM transitions for pentobarbital at the correct retention time. Quantitative
measurement is accomplished by normalization of the peak area with the area of
the internal standard for each specimen, including matrix specific calibrators, and
quality control (QC) materials. Each sample is separately processed by the instrument
software. The program automatically constructs a calibration curve, using the peak
abundance data from the calibrator samples. QC are extracted and analyzed with
patient samples. All data is subjected to analyst and technical QC review, prior to
acceptance for reporting of results.
Results. The relative intra-laboratory reproducibility standard deviations were, in
general, better than 5% at concentrations in the therapeutic range. The estimated
lowest limit of detection (LOD) was zero 0.09 ug/mL and lowest limit of quantitation
(LOQ) was 0.29 ug/mL. LOQ for serum pentobarbital using CV<20% was <0.5 ug/
mL and was used as clinical cut-off. The mean true recoveries for samples in the
analytical measurement range 0.5-100 ug/mL were between 96-106%. Correlation
with GC-FID was excellent (y=0.9x1.1, r=0.99, p<0.05, n=21). Sample recoveries
(n=3) following five freeze/thaw cycles were 98-111%. TAT was approximately 2
hours from receiving of specimen to reporting result to physician.
Conclusion. The quantitative pentobarbital method by LC/MS-MS is an easy and
user-friendly method with an excellent analytical sensitivity. It requires one-step for
extraction of pentobarbital from serum, and then a direct injection into the LC/MSMS system. We were able to achieve excellent TAT with our method, and were able to
meet our institutions goal for TAT.

B-404
Detection of 13 Opioids by LC-MS/MS from dried urine spots

J. M. Boyd1, V. Simon2, J. Bacani3, V. Dias1, S. M. H. Sadrzadeh1.

1
Department of Pathology and Laboratory Medicine, University of
Calgary and Calgary Laboratory Services, Calgary, AB, Canada, 2Calgary
Laboratory Services, Calgary, AB, Canada, 3Faculty of Medicine,
University of Calgary, Calgary, AB, Canada
Introduction: Dried specimens offer a convenient and economical approach for
collection, storage and transport of clinical specimens. Although, dried blood spots
have been used as the specimen of choice in clinical laboratories (Biochemical
Genetics) for years, dried urine spots (DUS) have not been established for routine
use in clinical laboratories. Traditionally, drugs of abuse in clinical laboratories are
detected by immunoassay (IA) and confirmed by GC-MS, which is time consuming
and expensive. LC-MS/MS can offer a one-step detection and confirmation method
for detection and measurement of drugs.
Objectives: To develop a LC-MS/MS method to detect multiple drugs in DUS in onestep.

B-403
Drug monitoring and toxicology: simultaneous determination of total
mycophenolic acid and its glucuronide by HPLC-UV

P. H. Tang. Cincinnati Childrens Hospital Medical Center, University of

Cincinnati College of Medicine, Cincinnati, OH
Background: Mycophenolic acid (MPA) is a potent immunosuppressant. MPA
is primary metabolized to an inactive glucuronide MPAG, which is transported
from liver into bile. Biliary MPAG then enters the GI tract, where it is converted
back to MPA, which is then recycled into the bloodstream via the enterohepatic
circulation pathway. Several studies have documented that variation in MPA plasma
concentrations are unpredictable; variability in plasma concentration of MPA both
within and between individuals is high. Therapeutic drug monitoring (TDM) of
MPA plasma concentrations can help the clinician to develop personalized therapy
strategies to avoid toxicity and maintain efficacy. To support the clinical studies of
mycophenolate mofetil and to provide the clinical service for the TDM of MPA and
MPAG ordered by the researchers and physicians, a simple and rapid HPLC method
using UV detection has been developed.
Methods: The analytical procedures include single dilution step, protein precipitation,
ultracentrifugation and gradient elution chromatography. Separation of MPA, MPAG
and internal standard clonazepam was achieved by using a 5-m Microsorb-MV C18
column (150 x 4.6 mm). These compounds were quantified by UV absorbance at
306 nm.
Results: Linearity was verified over ranges of 0.1-20 mg/L and 1-200 mg/L for MPA
and MPAG, respectively. Recoveries of MPA and MPAG ranged between 93 and
105%. Both within-run (n = 6) and between-run (n = 30) precisions were lower than
10%. Figure 1 illustrates a comparison between the current method and reference
method for MPA. The linear regression statistics indicated an r value of 0.993 (P
< 0.0001). The linear regression equation for correlation was y = 1.018 x + 0.031;
where y, the current method and x, the reference method. No interferences with other
common drugs were observed.
Conclusion: The analytical method is very suitable for both routine clinical use and
pharmacokinetic studies.

Methods: 13 opioids (morphine, hydromorphone, oxymorphone, codeine,

hydrocodone, oxycodone, heroin, 6- mono-acetyl morphine (6-MAM), fentanyl,
norfentanyl, naloxone, tramadol, and meperidine) were analyzed in dried urine spot
extraction. Briefly, 15 L of de-identified urine samples containing the above drugs
was spotted onto Whatman 903 Protein Saver Cards (GE Healthcare Biosciences).
Drugs were extracted using a mixture of methanol, acetonitrile and water (4:4:1).
Extracts, evaporated at RT, reconstituted in 5% acetonitrile before injecting into the
LC-MS/MS. Urine samples were spiked by standards ranging from 100-2000 ng/mL.
HPLC: Agilent 1290 (Palo Alto, CA) with a Restek Ultra Biphenyl column (100mm
x 2.1 mm x 5m). Mobile phase A was 1 mM ammonium formate and 0.1% formic
acid in water; Mobile phase B was 0.1% formic acid in acetonitrile. Separation was
achieved using a gradient elution program starting at 99% A for 1 minute, decreasing
to 95% A for one minute, and then decreasing to 50% A until 13.6 minutes, holding at
50% A until 15.6 minutes, and returning to 99% an over 0.1 minutes. Flow rate was
0.5 mL/min and the injection volume was 10 L.
Mass spectrometric detection was performed using an Agilent 6460 triple quadrupole
mass spectrometer. Source parameters were optimized as follows: Electrospray
voltage +2500V, Sheath gas temperature 380 C, sheath gas flow 11 L/min, Nebulizer
gas was 30 psi, Source gas temperature was 300 C and the gas flow was 9 L/min.
MRM transitions for all analytes were selected and optimized using the Agilent
Optimizer software using 1000 ng/mL of each analyte in methanol. Data analysis was
performed using Agilent MassHunter Quantitative analysis software
Results:
Our preliminary recovery results from DUS by LC/MSMS were: Morphine, 97%;
Oxymorphone, 97%, hydromorphone, 100%; naloxone, 100%; codeine, 100%;
6-MAM, 99%; hydrocodone, 99%; oxycodone, 100%; heroin, 98%; fentanyl, 97%;
norfentanyl, 97%; tramadol, 100%; and meperidine, 97%. These results matched
those generated from identical liquid urine same by traditional immunoassay followed
by GC/MS.
Conclusions: We modified a LC-MS/MS method to detect 13 opioids from DUS,
simultaneously. DUS offers a convenient and economical approach for specimen
collection, transportation and storage. This one-step LC-MS/MS detection
confirmation method eliminates the need for the traditional multi-step approach of
immunoassay and GC-MS and can greatly reduce the operational cost. More drugs
of abuse are being tested.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S249

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

B-405
Effectiveness of Afternoon Dosing Policy on Digoxin Level Monitoring

Y. R. Hsu, G. S. Cembrowski, D. F. LeGatt. University of Alberta,

Edmonton, AB, Canada
Background: Determining serum or plasma digoxin levels is helpful in titrating drug
dosage as well as assessing compliance and toxicity. However, inappropriately timed
sample collection in relation to dose may result in level misinterpretation with the
risk of over or under dosing. To circumvent inappropriately timed collections, the
policy of afternoon dosed digoxin (PM dosing) in all inpatient care facilities in Alberta
Health Services Edmonton Zone was instituted on November 5, 2008. This policy was
also adopted in other but not all Northern Alberta health care facilities. The purpose
of this study was to evaluate the effect of the policy change on digoxin dosing time by
comparing pre- and post-implementation periods.
Method: The evaluation was conducted by comparing the time periods of
approximately four years before and four years after policy implementation
(November 1, 2004 to October 31, 2012). All digoxin levels during this period were
included. Those with available dosing information were grouped into AM (0001h
to 1159h) and PM (1200h to 2400h) dosing groups as well as into inpatients and
outpatients, which also included collection at emergency departments.
Results: A total of 58,236 digoxin levels were performed in Northern Alberta from
November 1, 2004 to October 31, 2012. There was a steady decline in the number of
tests done annually (45.8% decline over eight years), likely reflecting the decreased
prescription of digoxin in recent practice. Of the 58,236 requests, 23,324 (40%)
had dosing information provided on the requisition forms. Within this latter group,
the proportion where dosing occurred PM was significantly higher after policy
implementation (mean 472.1% vs 280.73%, p <0.05). For Edmonton Zone inpatient
facilities, the proportion of requests with PM dosing increased significantly in the
four year period after policy implementation (mean 563.4% vs 271.5%, p <0.05).
The proportion of digoxin levels at increased risk of toxicity (level >2.0 nmol/L)
was significantly higher pre-policy implementation (mean 132.5% vs 5.50.58%,
p <0.05); whereas, the proportion of samples below target digoxin level was higher
post-policy (mean 326.3% vs 162.9%, p <0.05). The proportion of samples within
the target range (0.6 to 1.2 nmol/L for heart failure) was similar pre- and post-policy
(mean 422.6% vs 464.8%, p >0.05).
Conclusion: This study demonstrates the sustained effect of increased compliance to
PM dosing since the institution of this policy in Alberta Health Services Edmonton
Zone in November 2008. However, other intervention strategies are required to
further improve policy adherence, hence minimizing the chance of inappropriately
timed collections. The lack of dosing information on the requisition forms is another
issue that precludes a complete assessment.

B-406
Development of a High-throughput LC-MS/MS Assay for Pain Management
Panel from Urine

H. Qiao, J. Ye, E. Majdi, L. Cousins. IONICS Mass Spectrometry Group

Inc., Bolton, ON, Canada
Background: The widespread use and the potential abuse of opiates, sedatives,
and stimulants drugs have increased the need and in some cases the requirement to
screen patients on a routine basis. Pain panels continue to grow in complexity as
more prescription and non-prescription compounds are added. This has made the
job of toxicological analysis even more challenging. To fulfill these requirements,
a fast, reliable, and accurate LC-MS/MS method has been created for the analysis
of a pain panel comprised of 30 drugs on an IONICS 3Q 120 triple quadrupole mass
spectrometer.
Methods: First, the mixed drug standard solution was spiked into the urine matrix,
then diluted with the mobile phase A (100% H2O, 0.1% formic acid) to make
a series of concentrations ranging from 0.016 to 16 ng/mL. The internal standard
concentration used was 10 ng/mL. The calibrator solutions will be directly injected
without further treatment. IONICS 3Q 120 mass spectrometer equipped with a heated
coaxial flow ion source and Hot Source-Induced Desolvation interface was used.
The time-managed MRM in MolanaTM software was used to optimize the dwell
time for each MRM transition based on the retention times and the number of MRM
transitions within given experiments. The separation was performed on a Shimadzu
Prominence LC system. A 10 uL sample was loaded onto a Restek Ultra II Biphenyl
column (50 x 2.1 mm, 5um) kept at 40 C. A gradient method was created with a flow
rate of 600 uL/min and a total LC cycle time of 7.5 minutes. Solvent B was composed
of 0.1% formic acid in 100% methanol.

S250

Results: A total of 57 MRM transitions were used to monitor 30 drugs including

internal standards. No matrix interferences were observed. LC system carryover was
checked to ensure the validity of the data. An overlay of the extracted chromatograms
of 30 drugs in a 7.5 minute LC run showed that all of the anlaytes were clearly
separated. The calibration curves showed good linearity for all the analytes across the
whole concentration range with a coefficient R2>0.99. All calibration curves used a
linear weighting regression of 1/x. The LLOQs for the 30 drugs were in the range of
0.032 to 2 ng/mL. At LLOQs, the accuracy was between 84-114%, and CVs were <
10% for all analytes.
Conclusion: The results in this study show that in a 7.5-minute LC run, this LC-MS/
MS method can effectively separate the 30 pain panel drugs. The quantitation results
also indicate that this method is accurate, precise, and reproducible. The LLOQs for
all the 30 drugs is in the range of 0.032 to 2 ng/mL, which is 2 to 3 orders lower
than the typical screening cutoff concentration (300 ng/mL), and much lower than the
typical confirmation cutoff concentration (50 ng/mL) for most of the drugs of abuse.
Therefore, this LC-MS/MS method with IONICS 3Q 120 mass spectrometer is an
effective combination for clinical pain management to monitor patient drug use and
program adherence or for drugs of abuse or other workplace drug testing.

B-407
Development of Abbott Phenytoin Assay for the ARCHITECT cSystems
Automated Clinical Chemistry Analyzers

J. Donaldson1, N. Pham1, A. Kong1, S. Hoang1, S. Shaw2, N. Pham2, O.

Ndimbie2, C. Kasal2, L. Ye1. 1Thermo Fisher Scientific, Fremont, CA,
2
Abbott Laboratories, Irving, TX
Introduction: The study objective is to develop a sensitive immunoassay intended
for the quantitative measurement of phenytoin in human serum or plasma on the
ARCHITECT cSystems. The measurements obtained are used in the diagnosis and
treatment of phenytoin overdose and in monitoring levels of phenytoin to help ensure
appropriate therapy.
Methods: The Abbott Phenytoin Assay is a liquid ready-to-use, homogeneous enzyme
immunoassay. The two reagent kit uses specific antibodies to detect phenytoin in
the sample, with minimal cross-reactivity to various over-the counter, structurally
unrelated compounds. The method is based on the competition for a fixed amount of
specific antibody binding sites between enzyme [glucose-6-phosphate dehydrogenase
(G6PDH)]-labeled phenytoin, and phenytoin contained in the sample. In the absence of
phenytoin from the sample, the specific antibody binds the G6PDH-labeled phenytoin
and causes a decrease in enzyme activity. If phenytoin is present in the sample, it
occupies the antibody binding sites, which allows the G6PDH-labeled phenytoin to
interact with the substrate, resulting in enzyme activity. This phenomenon creates a
direct relationship between the phenytoin concentration in sample and enzyme activity.
By measuring the enzymes ability to convert nicotinamide adenine dinucleotide
(NAD) to NADH, its activity is determined spectrophotometrically at 340 nm.
Results: The performance of the Abbott Phenytoin Assay was evaluated on the Abbott
ARCHITECT c8000 analyzer. Based on guidance from Clinical and Laboratory
Standards Institute (CLSI) protocol EP17-A2, the assay demonstrates a Limit of
Quantitation of < 1.8 g/mL using inter-assay precision < 7% CV or < 0.7 g/mL
SD and bias within 10% or 1.0 g/mL over an extended period. The assay is linear
from 1.8 to 40.0 g/mL using guidance from CLSI protocol EP6-A. Assay precision
was evaluated using CLSI guideline EP5-A2. A tri-level commercial control and six
human serum samples containing phenytoin at concentrations ranging from 3.9 g/mL
to 36.1 g/mL were tested. Each sample was assayed in duplicate twice a day for 20
days with at least two hours between runs. The precision ranged from 1.6 %CV to 3.3
%CV for Within-Run and 2.0 %CV to 4.1 %CV for Total-Run. The assay accurately
recovered spiked phenytoin at levels representing sub-therapeutic, therapeutic, and
toxic samples. No significant interference was observed with various endogenous
substances or compounds whose chemical structure or concurrent therapeutic use
would suggest possible cross-reactivity. The assay did not exhibit obvious cross
reactivity with phenytoin derivatives and metabolites at the concentrations tested with
the exception of fosphenytoin. Abbott ARCHITECT Phenytoin patient correlation
studies: new vs. current on-market assay yielded a regression equation of y=1.11x +
0.12 and a correlation coefficient of 0.99. The new reagent has an onboard stability of
40 days and calibration curve stability of 7 days.
Conclusion: The Abbott Phenytoin Assay enables measurement of phenytoin in
human serum or plasma with high precision across the linear range. The ability to
monitor levels of phenytoin with high accuracy can help ensure appropriate therapy.
The assay has applications on the ARCHITECT c16000, c8000, and c4000.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

B-408
Urine drug screening: using GC-MS/MS to augment LC-MS/MS screens

J. M. Colby, A. M. Gordon, A. H. B. Wu, K. L. Lynch. University of

California, San Francisco, San Francisco, CA
Background & Objective Clinical toxicology services, including the detection of
pharmaceuticals and drugs of abuse in biological samples, are routinely offered by
many hospital laboratories and can play an important role in patient management.
In addition to immunoassay screening for routine drugs of abuse, our laboratory
offers a liquid chromatography tandem mass spectrometry (LC-MS/MS) method to
screen patient urine samples for a large number of clinically relevant compounds.
Immunoassays are not available for many of the compounds in our method, which
makes it a challenge to verify results by two methods. Here we report the development
and validation of a gas chromatography tandem mass spectrometry (GC-MS/MS)
method for comprehensive urine drug screening with method performance evaluated
in part by concordance with our LC-MS/MS method.
Methods GC-MS/MS was performed on a Bruker ScionTQ mass spectrometer with
an EI source and a Bruker 436 GC with a 30m BR-5ms column. Chomatographic
runs began at 80C and ramped to 295C over 20 minutes. Retention times (RT)
were established and two multiple reaction monitoring (MRM) transitions were
developed for each of the approximately 150 compounds in the method. Compounds
were identified by RT and the presence/ratio of the two MRM transitions. For method
validation analytes were spiked into 0.5 mL urine, isolated using solid phase extraction
(SPE), eluted in either the acid/neutral or basic fractions, and evaporated to dryness
at 40 under N2 flow. Extracts were reconstituted in 50:50 methanol:acetonitrile and
1 L was injected. Our validation included determining the lower limit of detection
(LLOD), matrix effects, and SPE recovery for each compound. At least 30 patient
samples were analyzed by both GC-MS/MS and LC-MS/MS.
LC-MS/MS was performed on an Agilent HPLC with an ABSciex 3200 QTRAP mass
spectrometer in positive-ESI mode. Compounds were identified by a combination of
RT, one MRM transition and a match between the collected product ion spectrum and
our in-house-built spectral library.
Results Overall the GC-MS/MS method performed quite well. LLODs ranged from
5 ng/mL to 250 ng/mL. A small fraction of compounds were not observed in any
validation samples, likely due to extremely low volatility. Method validation included
determining matrix effects using two donor urine matrices and evaluating the recovery
of our solid phase extraction method for acid/neutral and basic drugs. Comparison of
patient samples between LC-MS/MS and GC-MS/MS showed good concordance. As
expected, many of the compounds that were missed by GC-MS/MS were large and/or
quite polar, for example glucuronide conjugates.
Conclusions GC-MS/MS has many advantages for identification of drugs in
biological matrices. GC-MS/MS instruments are generally less expensive than LCMS/MS instruments and can analyze some compounds that LC cannot and GC-MS/
MS can offer lower limits of detection than GC-MS systems. We have recently had
cases involving pentobarbital and 1,4-butanediol, neither were detected using our LCMS/MS but were easily observed with our GC-MS/MS. Use of both technologies in
our laboratory allows us to screen patient urines for a wider variety of compounds
with a greater assurance of accuracy.

panel consisted of 6-monoacetylmorphine, benzoylecgonine and phencyclidine. This

panel was subjected to dilution (1/10) with an aqueous solution containing the internal
standards prior to online SPE. The benzodiazepine panel consisted of temazepam,
7-amino clonazepam, nordiazepam, alpha hydroxy alprazolam, lorazepam, alprazolam
and oxazepam. Benzodiazepine samples were subjected to enzymatic hydrolysis
followed by centrifugation and then diluted (1/50) prior to online SPE. Online SPE
methods were optimized for each panel. Analysis of all samples was performed at a
rate of <12 seconds per sample using a RapidFire high-throughput system coupled to
a triple quadrupole mass spectrometer.
Results The analytes in the illicit panel had good linearity with R2 values >0.995
within the measurement range of 2.5-500 ng/mL for 6-monoacetylmorphine, 5-1000
ng/mL for phencyclidine and 25-5000 ng/mL for benzoylecgonine. The average
intra- and interday accuracy at LLOQ was 99% for 6-monoacetylmorphine, 105%
for phencyclidine and 102% for benzoylecgonine. The average intra- and interday
imprecision at LLOQ was 17% for 6-monoacethylmorphine, 7% for phencyclidine
and 2% for benzoylecgonine. The standard curve of benzodiazepine analytes had
excellent linearity within the measurement range (50-5000 ng/mL) with R2 values
>0.999. The average intra- and interday accuracy and imprecision at LLOQ for all
compounds were 103% and 8% respectively. No significant carry-over was detected
for any illicit and benzodiazepine analytes. The linearity, accuracy and imprecision
results from SPE/MS/MS analysis were comparable to those from traditional LC/MS/
MS. The SPE/MS/MS analysis was more than 20 times faster than LC/MS/MS.
Conclusions Illicit drugs and benzodiazepines can accurately, precisely and rapidly
be quantified using an ultrafast SPE/MS/MS system. This system is an alternative to
traditional LC/MS/MS, combining efficiency and speed.

B-410
Compliance Rates In Chronic Cancer Pain Patients.

J. A. Hayden, P. Mathias, A. Hoofnagle, G. Baird. University of Washington,

Seattle, WA
Background: A distinction is often made between patients on prescription opioid
therapy for chronic cancer and chronic non-cancer pain. This distinction can have
wide reaching implications, such as impacting which patients are and are not subject
to new legislation. No evidence exists to suggest that these two patient populations
differ in terms of the prevalence of opioid misuse.
Methods: This retrospective cohort study included 1,271 non-cancer and 139 cancer
patients on prescription opioids who submitted urine samples to our laboratories
between May of 2012 and April of 2013. All patients were subject to chart review
as part of normal clinical operations by a single reviewer. Chronic cancer pain was
based on cancer or cancer therapy related pain being the primary diagnosis in the
chart. Urine samples were analyzed for the presence or absence of prescription
opioids (oxycodone, hydrocodone, methadone, morphine, hydromorphone, fentanyl,
buprenorphine, meperidine, propoxyphene) and the heroin-specific metabolite
6-monoacetylmorphine using a clinically validated, laboratory-developed liquid
chromatography tandem mass spectrometry (LC-MS/MS) assay. Cocaine abuse
was determined by an immunoassay (Syvia EMIT II) and methamphetamine was
determined with a separate validated LC-MS/MS assay. Benzodiazepines and other
analgesics (tramadol,carisoprodol, etc) not detected in our assays were not considered.
Results: Cancer patients showed a compliance rate statistically indistinguishable from
that of non-cancer patients (61% and 68%, respectively). The patterns of abused drugs
were not substantially different between these two populations. Both populations were
three times more likely to abuse prescription opioids than illicits (heroin, cocaine and
methamphetamine).

B-409
Ultrafast Quantitative Analysis of Illicit Drugs and Benzodiazepines in Urine
Using High-throughput SPE/MS/MS

F. Mbeunkui, B. Marshall, S. Sullivan, R. Dixon. Physicians Choice

Laboratory Services, Rock Hill, SC
Introduction Pathologists, employers and law enforcement officials use drug testing
extensively today. Drug screening typically involved immunoassay analysis followed
by a confirmatory test by GC/MS or LC/MS detection. Steady increases in the need
for greater analytical capacity and throughput have placed demands on traditional
technologies. The RapidFire (RF) platform provides an automated solid-phase
extraction (SPE) system that gives a throughput of approximately 12 seconds per
sample to the mass spectrometer.

Conclusion: Chronic cancer pain patients are equally likely to misuse prescription
opioids, and in particular to use non-prescribed prescription opioids. This misuse puts
them at risk for adverse outcomes and clinicians should be aware of these dangers
when managing chronic cancer pain.

In the present study, we evaluated the ability of the RF/MS/MS system to quantitatively
measure a panel of illicit drugs or benzodiazepines in urine. Imprecision, accuracy and
linearity results achieved with this ultrafast RF/MS/MS system were comparable to
LC/MS/MS.
Methods Blank matrix containing internal standards was spiked with the analytes of
interest in a range of concentration to prepare the calibration curve. The illicit drug

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S251

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

Methods: The Abbott Theophylline Assay is a liquid ready-to-use, homogeneous

enzyme immunoassay. The two reagent kit uses specific antibodies to detect
theophylline in the sample, with minimal cross-reactivity to various over-the
counter, structurally unrelated compounds. The method is based on the competition
for a fixed amount of specific antibody binding sites between enzyme [glucose6-phosphate dehydrogenase (G6PDH)]-labeled theophylline, and theophylline
contained in the sample. In the absence of theophylline from the sample, the specific
antibody binds the G6PDH-labeled theophylline and causes a decrease in enzyme
activity. If theophylline is present in the sample, it occupies the antibody binding
sites, which allows the G6PDH-labeled theophylline to interact with the substrate,
resulting in enzyme activity. This phenomenon creates a direct relationship between
the theophylline concentration in sample and enzyme activity. By measuring the
enzymes ability to convert nicotinamide adenine dinucleotide (NAD) to NADH, its
activity is determined spectrophotometrically at 340 nm.

B-412
Rapid measurement of tacrolimus, cyclosporin A and sirolimus in blood by
paper spray-tandem mass spectrometry (PS-MS/MS)

R. Shi1, E. M. El Gierari2, N. E. Manicke3, J. D. Faix1. 1Stanford University

School of Medicine, Stanford, CA, 2Stanford Hospital & Clinics, Stanford,
CA, 3Department of Chemistry and Chemical Biology, Indiana UniversityPurdue University Indianapolis, Indianapolis, IN
Background: therapeutic drug monitoring (TDM) of immunosuppressant(s) is critical
in preventing organ rejection after transplant. Automated immunoassays provide
quick turnaround time but are costly and lacking standardization. Conventional
tandem mass spectrometry (MS/MS) requires a liquid chromatography (LC) system
and assay requires pre-analytical manipulation which is not amenable to random
access testing. Paper spray (PS) ionization is a technique that generates gas phase
analyte ions directly from dried blood spots for quantitative MS/MS analysis, without
complex sample preparation, nor a LC system. We planned to evaluate PS-MS/MS for
simultaneous tacrolimus, cyclosporin A and sirolimus TDM in a clinical diagnostic
laboratory, by examining assay precision, accuracy, and analytical measurement range
(AMR), as well as assay specificity and possible analyte ion suppression phenomenon.
Methods: 200 L of whole blood, calibrators, or quality control material were mixed
with 50 L of stable isotope labeled internal standard mixture ( [13C, 2H2]-FK506,
[2H4]-cyclosporin A, and [2H3]-rapamycin). The blood mixture (10 L) was spotted
onto triangular shaped card paper contained in disposable cartridges and dried at
40C. Cartridges were analyzed using an automated paper spray ion source coupled to
a triple quadrupole MS/MS (TSQ VantageTM; Thermo Scientific). Small amount (up to
120 L) of methanol and chloroform mixture with added sodium acetate was delivered
by the ion source to blood containing paper in cartridge before a voltage of 3.5 kV
was applied to generate ion spray. Sodium adduct ions of each immunosuppressant,
along with corresponding internal standard ions were detected in the selected reaction
monitoring (SRM) mode and quantitated by the area under the curve of collected 50
scan within 1.0 min, with an average analysis time of 3 min per sample. Each analyte
result was confirmed by using a second SRM.
Results: the PS-MS/MS method has acceptable AMR and precision, and results
correlate well with those of a FDA approved immunoassay currently used in clinical
labs. Patient samples may be analyzed in batches by PS-MS/MS, or analyzed by
adding on to batch throughout the day, much like in a random access fashion.
Conclusions: PS-MS/MS is much simpler in comparison to a conventional LC-MS/
MS system. It simultaneously provides accurate results for tacrolimus, cyclosporin A
and sirolimus, with fast turnaround time amenable to random access testing protocols.

B-413
Development of Abbott Theophylline Assay for the ARCHITECT cSystems
Automated Clinical Chemistry Analyzers

J. Donaldson1, N. Pham1, A. Kong1, S. Hoang1, S. Shaw2, N. Pham2, O.

Ndimbie2, C. Kasal2, L. Ye1. 1Thermo Fisher Scientific, Fremont, CA,
2
Abbott Laboratories, Irving, TX
Introduction: The study objective is to develop a sensitive immunoassay intended
for the quantitative measurement of theophylline in human serum or plasma on the
ARCHITECT cSystems. The measurements obtained are used in the diagnosis and
treatment of theophylline overdose and in monitoring levels of theophylline to help
ensure appropriate therapy.

S252

Results: The performance of the Abbott Theophylline Assay was evaluated on

the Abbott ARCHITECT c8000 analyzer. Based on guidance from Clinical and
Laboratory Standards Institute (CLSI) protocol EP17-A2, the assay demonstrates a
Limit of Quantitation of < 2.0 g/mL using inter-assay precision < 7% CV or < 0.7 g/
mL SD and bias within 10% or 1.0 g/mL over an extended period. The assay is linear
from 2.0 to 40.0 g/mL using guidance from CLSI protocol EP6-A. Assay precision
was evaluated using CLSI guideline EP5-A2. A tri-level commercial control and six
human serum samples containing theophylline at concentrations ranging from 3.5 g/
mL to 37.0 g/mL were tested. Each sample was assayed in duplicate twice a day for 20
days with at least two hours between runs. The precision ranged from 1.4 %CV to 2.3
%CV for Within-Run and 1.9 %CV to 3.3 %CV for Total-Run. The assay accurately
recovered spiked theophylline at levels representing sub-therapeutic, therapeutic, and
toxic samples. No significant interference was observed with various endogenous
substances or compounds whose chemical structure or concurrent therapeutic use
would suggest possible cross-reactivity. The assay did not exhibit obvious cross
reactivity with theophylline derivatives and metabolites at the concentrations tested
with the exception of caffeine. Abbott ARCHITECT Theophylline patient correlation
studies: new vs. current on-market assay yielded a regression equation of y=1.07x 0.07 and a correlation coefficient of 1.0. The new reagent has an onboard stability of
40 days and calibration curve stability of 7 days.
Conclusion: The Abbott Theophylline Assay enables measurement of theophylline
in human serum or plasma with high precision across the linear range. The ability
to monitor levels of theophylline with high accuracy can help ensure appropriate
therapy. The assay has applications on the ARCHITECT c16000, c8000, and c4000.

B-414
Isotope Dilution Gas Chromatography-Mass Spectrometry (GC/MS) Method
for the Analysis of Hydroxyurea

U. Garg, D. Scott, C. Frazee, G. Kearns, K. Neville. Childrens Mercy

Hospitals and Clinics, Kansas City, MO
Background: Hydroxyurea (HU) is used in the treatment of various malignancies
such as chronic myelogenous leukemia, squamous cell carcinomas, polycythemia
vera and essential thrombocytosis. It is also used in the treatment of sickle cell disease
where it is shown to decrease painful crisis and reduce number of transfusions. There
are limited studies on the pharmacokinetics of hydroxyurea. An accurate, precise
and sensitive method is needed to support such studies and also the monitoring of
therapeutic adherence. Current methods for measurement of HU include colorimetry,
high performance liquid chromatography (HPLC) and gas chromatography-mass
spectrometry (GC-MS). While available GC-MS methods are sensitive and specific,
they involve use of internal standards that are structurally different from HU and
prone to error due to poor (e.g., < 3%) extraction efficiency. Here we describe a novel
GC-MS method for the determination of HU that involves stable labeled HU as an
internal standard.
Materials and Methods: HU was purchased from Sigma-Aldrich and the internal
standard, HU [13C] 15N2 was custom synthesized from Toronto Research, Canada.
The calibrators and controls were prepared in drug-free plasma. To 0.5 mL plasma,
1 mL of phosphate buffer (pH 6.0) and 0.1 mL of internal standard were added. HU
was extracted with ethylacetate. The extract was dried and trimethylsilyl derivatives
of HU were prepared. 1 L of the derivatization mixture was injected into the GC-MS.
Selected ion monitoring was used for identification and quantification of HU. The
monitored m/z ions were: HU (quantification ion 277, qualifier ions - 292 and 249),
HU [13C] 15N2 (quantification ion 280, qualifier ion - 295).
Results and Conclusion: The method was evaluated for reportable range, accuracy,
within and between-run imprecision, and limit of quantification. Reportable range
(linearity) of the method was from 0.1 to 100 g/mL. All results were accurate within
allowable error of <10% or 0.03 g/mL. Within-run and between-run imprecision

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

were <5 and <10% respectively. Lower limit of quantification, at inaccuracy of <10%
or <0.03 g/mL, was 0.1 g/mL. To check specificity and selectivity of the method,
20 negative samples were analyzed. All the drug free samples quantified less than the
lower limit of quantification and failed qualifier ion ratios. Samples were stable for at
least 4 h, 2 months and 6 months at room temperature, -20oC and -70oC respectively.
Samples were also stable after 3 freeze/thaw cycles. Extraction efficiency for 1, 5,
10 and 50 g/mL samples averaged 2.2%, 1.8%, 1.6% and 1.4% respectively. These
data demonstrate that our new, isotope dilution GC/MS method for analysis of HU is
accurate, sensitive, precise and robust.

B-415
Bupropion Exposure in an Infant: A Case Report

S. Delaney1, G. Neuman2, S. Ito2, D. A. Colantonio2. 1University of Toronto,

Toronto, ON, Canada, 2Hospital for Sick Children, Toronto, ON, Canada,
Background: The last two decades has seen a substantial increase in the rates of women
breastfeeding their infants. However, it has been reported that 66 - 80% of nursing
women are on medication. While many drugs are safely taken by nursing mothers,
there is accumulating evidence of toxicity in some breastfed infants. Information
on drug excretion into milk is lacking for most drugs, and early phase drug studies
exclude breastfeeding women. This uncertainty in the risk of drug exposure causes
maternal non-adherence to therapy or avoidance of breastfeeding. This is a clinical
problem in drug safety and is an important womens health issue, uniquely affecting
both mother and infant. The objective of this study is to investigate the risk of drug
exposure in nursing infants. We present a proof-of-principle case of infant bupropion
exposure highlighing the importance and clinical applicability of this study.
Methods: To investigate infant drug exposure through breast milk, we established a
drug safety monitoring program, Drugs in Lactation Analysis Consortium (DLAC),
to measure several drugs commonly used by women breastfeeding. Breast milk is a
complex lipid- and protein- rich matrix, with drugs partitioning to either the aqueous
or lipid phase, thus requiring meticulous sample processing before analysis. We first
investigated whether drugs partition to the aqueous or lipid phase of breast milk. We
then worked to create a simplified drug extraction method using hexane, methanol
and acetonitrile to facilitate efficient drug extraction from breast milk. Extraction
efficiency and stability were determined. Methods were then developed to measure
these drugs using LC-MS/MS. This process was followed in the case presented below.
Results/Case Report: A 6.5 month old previously healthy infant presented to SickKids
Hospital with vomiting and seizures. She was exclusively breastfed; her mother was
taking escitalopram daily for several months, with the recent addition of bupropion.
The usual clinical workup was negative. As we suspected this might be a case of
drug exposure, the infants urine was tested and was positive for bupropion and
escitalopram. Serial breast milk and serum samples were obtained and bupropion and
its major metabolite, hydroxybuproion, were measured using HPLC-MS/MS. Using
a pharmacokinetic model, we determined that the bupropion levels were significant
around the time of the event. Average milk bupropion and hydroxybupropion levels
were 20.3 ng/mL and 68 ng/mL, respectively. Predicted bupropion steady state in
the infant around the time of the event was 0.12 ng/mL. The average observed infant
serum hydroxybupropion level was 11.2 ng/mL. Discharge diagnosis was bupropioninduced seizures.
CONCLUSION: This case highlights the importance of the DLAC study. Given the
increasing use of medication in nursing women, there is urgency to investigate the
potential adverse events associated with infant drug exposure through breast milk.
Future investigations will focus on population pharmacokinetic modeling to simulate
and estimate infant drug exposure and assess potential risks associated with drugs and
breastfeeding. The data generated from this study will help guide decisions for drug
use in nursing mothers.

B-416
Rapid confirm and determination of d and l methamphetamine in human urine
by liquid chromatography tandem mass spectrometry

S. Y. Wu, F. Hassouna. Confirmatrix Clinical Laboratory, Lawrenceville,,

GA,
Background: Prescription methamphetamine (Desoxyn) is composed entirely of the
d isomer. Over-The-Counter cold medications contain l isomer (Vickers inhaler).
Street-methamphetamine is either d-isomer or racemic mixture with d-isomer of
20-100%. In order to narrow down the potential sources of the methamphetamine,
clinical toxicology laboratories need to accurately differentiate between d and the l
isomer forms of methamphetamine. Because of the identical mass spectrum of d and l

isomers, directly identifying specific isomer relies on the chromatographic separation

of isomers in the racemic mixture. We therefore developed an analytical method to
simultaneously confirm and determine d and l methamphetamine isomers in urine
samples by LCMSMS.
Method: 100 ul urine from prescreened positive sample was mixed with 100 ul internal
standard of isotope d and l methamphetamine isomers. Samples were concentrated by
using one step liquid-liquid extraction with diethyl ether and ethyl acetate (1:1). At
least two small amounts of concentrated standards were added to the sample. The
overlay chromatograms of sample and added standards combining with the retention
time of isomer of isotope internal standard were used to confirm the expected isomer.
This separation of isomers was achieved with LC/MS/MS mass spectrometer with
macrocyclic glycopeptide-based 10cm x2.1 mm chiral column and mobile phase of
consisting 100% methanol with1% formic acid and 0.01% NaOH. The x-intercept of
the standard addition calibration curve corresponds to the concentration of analyte. A
query of instrument data processing software was written to perform the calculation
for determination and conversion of d and l isomer concentrations to percentage of
total racemic mixture of methamphetamine.
Results: The analytical curve for d and l methamphetamine was linear over the range
of 40 ng/ml (40% of cutoff 100 ng/ml) to 800 ng/ml with correlation coefficient of
0.9988. The precision of method was evaluated based on three different concentrations
and racemic mixture ratios of d and l methamphetamine over two weeks of intra
runs. CVs (n=25) of 4.2 %, 3.0% and 2.0% were obtained. Limit of quantitation was
determined as 40 ng/ml. Limit of detection was 10 ng/ml with retention time within
2% of isotope internal standard. Recoveries ranged between 95% and 106% for
spiked and pooled samples.
Conclusion : This analytical method was rapid, simple and specific. The accuracy,
precision, repeatability and robustness were found to be within the acceptable limit.

B-417
Development of Abbott Carbamazepine Assay for the ARCHITECT cSystems
Automated Clinical Chemistry Analyzers

J. Donaldson1, N. Pham1, A. Kong1, S. Hoang1, S. Shaw2, N. Pham2, O.

Ndimbie2, C. Kasal2, L. Ye1. 1Thermo Fisher Scientific, Fremont, CA,
2
Abbott Laboratories, Irving, TX,
Introduction: The study objective is to develop a sensitive immunoassay intended
for the quantitative measurement of carbamazepine in human serum or plasma on
the ARCHITECT cSystems. The measurements obtained are used in monitoring
carbamazepine levels to help ensure appropriate therapy. Since carbamazepine
concentrations correlate better with pharmacologic activity than dosage, monitoring
of blood levels can increase efficacy and safety.
Methods: The Abbott Carbamazepine Assay is a liquid stable, particle-enhanced
turbidimetric inhibition immunoassay used for the analysis of carbamazepine in serum
or plasma. The assay consists of two reagents and is based on competition between
carbamazepine in the sample and carbamazepine coated onto a micro-particle for
anti- carbamazepine antibody binding sites. In samples lacking carbamazepine, the
carbamazepine -coated micro-particles rapidly agglutinate in the presence of anticarbamazepine antibodies. The rate of absorbance change is measured photometrically,
and is directly proportional to the rate of particle agglutination. In samples containing
carbamazepine, the agglutination reaction is partially inhibited, slowing the rate of
absorbance change. A concentration-dependent classic agglutination inhibition curve
can be obtained, with maximum rate of agglutination at the lowest carbamazepine
concentration and the lowest agglutination rate at the highest carbamazepine
concentration.
Results: Performance of the Abbott Carbamazepine Assay was evaluated on
the Abbott ARCHITECT c8000 analyzer. Based on guidance from Clinical and
Laboratory Standards Institute (CLSI) protocol EP17-A2, the assay demonstrates a
Limit of Quantitation of < 1.9 g/mL using inter-assay precision < 7% CV or < 0.3
g/mL SD and bias within 10% or 0.4 g/mL over an extended period. The assay
is linear from 1.9 to 20.0 g/mL using guidance from CLSI protocol EP6-A. Assay
precision was evaluated using CLSI guideline EP5-A2. A tri-level commercial control
and five human serum samples containing carbamazepine at concentrations ranging
from 1.9 g/mL to 18.1 g/mL were tested. Each sample was assayed in duplicate
twice a day for 20 days with at least two hours between runs. Precision ranged from
1.0 %CV to 3.0 %CV for Within-Run and 1.9 %CV to 6.3 %CV for Total-Run.
The assay accurately recovered spiked carbamazepine at levels representing subtherapeutic, therapeutic, and toxic samples. No significant interference was observed
with various endogenous substances or compounds whose chemical structure or
concurrent therapeutic use would suggest possible cross-reactivity. The assay did not
exhibit obvious cross reactivity with carbamazepine derivatives and metabolites at

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S253

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

the concentrations tested with the exception of carbamazepine-10, 11-epoxide. Abbott
ARCHITECT Carbamazepine patient correlation studies: new vs. current on-market
assay yielded a regression equation of y=0.91x + 0.56 and a correlation coefficient
of 0.97. New assay vs. HPLC yielded y=1.09x + 0.37 and a correlation coefficient
of 0.96. The new reagent has a 45 day onboard stability and 7 day calibration curve
stability.

tacrolimus may be monitored by 4 immunoassays/analyzers, and by LC-MS-MS

and Mass Spectrometry. More recently, the QMS turbidimetric immunoassay was
developed for tacrolimus.

Conclusion: The Abbott Carbamazepine Assay enables measurement of carbamazepine

in human serum or plasma with high precision across the linear range. The ability
to monitor levels of carbamazepine with high accuracy can help ensure appropriate
therapy. The assay has applications on the ARCHITECT c16000, c8000, and c4000.

Methods: QMS tacrolimus is a turbidimetric immunoassay. Sample preparation

included mixing 200 microliter of samples - patient whole blood with 200 microliter
of the extraction reagent. After vortexing and centrifugation, the supernatant was
transferred to sample cups. Drug in the supernatant and drug coated on microparticle
underwent competitive binding for a limited number of antibody binding sites.
If tacrolimus was absent, tacrolimus-coated microparticle was agglutinated in
the presence of antibody reagent. If tacrolimjus was present, agglutination was
partially inhibited depending on tacrolimus concentration. Thus, agglutination
rate was inversely proportional to everolimus concentrations, and was measured
photometrically. Six calibrators ranged from 0 to 30 ng/mL.

B-418
Chemotherapy with pharmacokinetic (PK) guided exposure optimization: US
based experience with 5-fluorouracil (5-FU) in colorectal cancer (CRC) patients

I. Baburina1, Y. Li1, J. B. Courtney1, M. P. Duda2, S. Diamond2, M. C. Miller1,

F. Braiteh3, S. J. Salamone1. 1Saladax Biomedical, Inc., Bethlehem, PA,
2
Saladax Biomedical Laboratories, Inc., Bethlehem, PA, 3Comprehensive
Cancer Centers of Nevada, Las Vegas, NV,
Background: 5-FU is the backbone of colorectal cancer chemotherapy. Numerous
studies over the last 30 years have demonstrated wide pharmacokinetic variability
of 5-FU, which can lead to undue toxicity and suboptimal treatment. Exposure
optimization of 5-FU based on PK-guided dose adjustment has been tested in multiple
studies, demonstrating improved response rates and reduced toxicity compared
to body surface area (BSA) based dosing. Here we present the experience of a US
based laboratory evaluating 5-FU plasma levels to optimize systemic exposure
in CRC patients of US based oncologists. Methods: Between June 5, 2013 and
January 17, 2014, 5-FU concentrations were determined in 631 patient samples at
Saladax Biomedical Laboratories (a CLIA-certified lab) using a laboratory developed
immunoassay (My5-FU). Systemic 5-FU exposure [the area under the concentration
curve (AUC)] was calculated from the 5-FU level determined in a steady state sample
collected 24+6Hr after the start of 5-FU continuous infusion from 240 CRC patients
(n=250 therapy lines). The calculated AUC and dose adjustment recommendations
to achieve the target range AUC (20 30 mg*hr/L) were provided in the laboratory
report. Differences between cycles and the comparison between dose adjustment
recommendations and actual dose changes were made between evaluable cycle
pairs (defined as two consecutive cycles with AUC results). Actual vs. target AUC,
recommended vs. actual dose adjustment, and ability to adjust exposure to target
range were evaluated. Results: The majority of AUC results (62%) were outside the
target range in the first sample from each therapy line where exposure was determined,
irrespective of the 5-FU dosing or regimen used. For the 250 samples collected during
the first cycle of therapy, 48% had AUCs below the target range, 38% were within the
target range, and 14% were above the target range. In patients dosed at the standard
of 2400 mg/m2 (57% of doses), the majority of AUC results were still out of range.
At this dose, 44% had exposures below the target range, 42% were within the target
range, and 14% were above the target range (n=144). In 288 cycle pairs, 201 (69%)
were outside the target range: doses were decreased in 48% of the 40 above the target
range and doses were increased in 51% of the 161 below the target range. No dose
change was made in 89% of the 87 within the target range. In 101 cycle pairs out of
target range where a dose adjustment consistent with recommendation was made, 42%
of these patients achieved exposure in the target range. Conclusions: 5-FU exposure
optimization is feasible in the US clinical setting. Consistent with other reports, body
surface area based 5-FU dosing resulted in frequent under dosing (48%). The majority
(62%) of dose adjustment recommendations were followed in subsequent cycles.
When AUCs were out of range and dose adjustments were made, 5FU exposure was
optimized in almost half the patients with a single dose adjustment. PK-guided dose
adjustment is a practical approach to personalize and optimize 5-FU exposure.

B-419
Evaluation of Tacrolimus QMS assay by using Indiko and AU680 analyzers and
comparison to Architect

S. Wong1, K. Garrison2, M. Hinsdale2. 1Wake Forest School of Medicine

and Wake Forest Baptist Health, Wisnton-Salem, NC, 2Wake Forest Baptist
Health, Wisnton-Salem, NC,
Background: Tacrolimus has been a widely used calcineurin inhibitor
immunosuppressant for renal, liver and other transplants. Therapeutic ranges are:
10-15 ng/mL for post-transplant and 5 -10 ng/mL for maintenace, with more recent
studies suggesting minimization range to be 3-7 ng/mL. According to CAP surveys,

S254

Objective: This study initially established the clinical efficacy of the QMS tacrolimus
assays by using AU 680 and Indiko, followed by comparison to clinically used assay
by using Architect.

Results: Precision studies of control samples showed the following mean

concentrations and CVs: n= 10, AU 680, 4.14 ng/mL and CV = 2.6%, and Indigo,
3.96 ng/mL and 11.6%., and n =20, AU 680, 9.67 ng/mL and 2.4%., Indiko, 9.79 ng/
mL and 2.3% respectively. Calibration stabilities for both AU680 and Indiko were
shown to be 10 days. Comparison studies of the three analyzers for kidney transplant
samples with concentration ranging from <2.0 to 25.8 ng/mL showed the following
slopes, intercepts and correlation: Arch. vs AU 680 for n = 95, 1.224, 0.62 and 0.976.,
Arch. vs Indiko n = 95, 1.204, 0.59, and 0.965., and AU 680 vs Indiko for n = 97,
0.979, 0.03, and 0.982.
Conclusions: Tacrolimus may be monitored by the QMS assay using two
autoanalyzers with adequate sensitivity and acceptable precision. While both QMS
assays offered comparable tacrolimus determination suitable for monitoring renal
transplant patient, the concentrations were about 20 to 22% higher than those obtained
by the clinically used comparison method using Architect.

B-421
An Immunoassay for Methotrexate in Blood on ARCHITECT i System

R. Smalley1, R. Picard1, B. Burkhart1, R. Frescatore1, C. Glover1, L. Zhu1,

E. Roessner1, K. Majnesj2, A. hrvik2, K. He1, S. Raju1, R. Radwan1, D.
Dickson1, Z. Li1, T. Kettlety1. 1Fujirebio Diagnostics Inc., Malvern, PA,
2
Fujirebio Diagnostics AB, Gothenburg, Sweden,
Background: Methotrexate (MTX) is a cancer therapeutic drug for leukemia,
osteosarcoma, non-Hodgkins lymphoma and others. MTX levels in blood are
monitored in patients to ensure appropriate therapy and determine when to intervene
with the counter-acting rescue therapy. An assay for measurement of MTX in serum
and plasma on the ARCHITECT i System (ARCHITECT Methotrexate) is presented.
Methods: ARCHITECT Methotrexate is a competitive immunoassay. First, the
instrument mixes and incubates a sample (calibrators, controls, sera or plasma) with
anti-MTX antibody-coated magnetic microparticles, biotinylated anti-MTX antibody
and acridinium-conjugated MTX. Then, the instrument washes the microparticles
and adds pre-trigger and trigger solutions to initiate the chemiluminescence reaction.
Signals obtained as relative luminescent units are inversely proportional to the amount
of MTX in the sample.
Results: The limit of quantitation was 0.040 mol/L MTX. The direct measuring
range was from 0.04 to 1.5 mol/L and up to 12000 mol/L with specimen dilution.
The 20-day imprecision study showed a total CV 7.1% within the range of 0.056
to 1705.433 mol/L MTX with 6 controls and 5 panels on 4 instruments using 3 lots
of reagents (n = 80). Deviations from linearity were -4.1% to 7.2% within the range
of 1.881 to 0.012 mol/L MTX. Spike recovery ranged from 92% - 94% for samples
at 0.045, 0.909 and 9.090 mol/L of MTX. In the interference studies, the average
levels of MTX in the individual interfering substance-spiked samples (10 endogenous
substances including human anti-mouse antibody and rheumatoid factor, and 21
therapeutic drugs) were within 90% - 110% of that in the unspiked control samples.
Method comparison of ARCHITECT Methotrexate to TDx/TDxFLx Methotrexate II
(TDx) generated a Passing-Bablok correlation as [ARCHITECT] = 0.00 + 1.01 [TDx]
for samples within the range of 0.040 to 1.415 mol/L (n = 92) and [ARCHITECT] =
0.00 + 1.04 [TDx] for samples within the range of 0.040 to 1624.760 mol/L MTX,
n = 142). Specimen storage study denoted that specimens could be stored at room
temperature for 24 hours or at 2-8oC for 48 hours since the MTX values of the stored
specimens described above were within 10% deviation from that of the baseline
control concentrations. MTX concentrations in specimens collected in K2-EDTA,
sodium heparin, and lithium heparin tubes were within 10% deviation from that
collected in serum tubes across the range of 0.040 to 10.00 mol/L MTX. Reagent
on-board stability showed that the ARCHITECT Methotrexate reagents could remain

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

on the analyzer for a minimum of 30 days with no more than 10% shift from baseline.
Conclusion: The ARCHITECT Methotrexate assay under development was
demonstrated to be an accurate, precise, sensitive and robust assay for the measurement
of methotrexate in human serum and plasma.

B-422
Monitoring Rivaroxaban Anticoagulation Therapy

H. Ketha, E. W. Korman, J. I. Tange, R. K. Pruthi, S. S. Ketha, R. D.

McBane, P. Jannetto, L. J. Langman. Mayo Clinic, Rochester, MN,
Background Currently, laboratory monitoring of direct acting anticoagulants,
dabigatran (direct thrombin inhibitor), rivaroxaban and apixaban (factor Xa inhibitors)
is not being performed based on their predictable pharmacokinetic profile and a lack
of FDA-approved clinical assays to measure drug levels.
Relevance Measurement of drug levels may be useful to ensure drug clearance (e.g.,
in patients with active bleeding or those undergoing invasive procedures with high
bleeding risk), as well as to ensure compliance in patients who develop thrombosis.
Furthermore, the lack of any antidote also supports monitoring of patients.
Objective The first objective was to develop and validate a liquid chromatography
tandem mass spectrometry (LC-MS/MS) assay for measuring serum rivaroxaban
concentration. The second objective was to monitor serum concentrations in patients
receiving rivaroxaban.
Methods An LC-MS/MS assay for measuring serum rivaroxaban concentration was
developed. Briefly, 200 L of serum was added to 700 L methanol containing 15
ng rivaroxaban-d4 (Toronto Research Chemicals, Toronto, Canada). The sample was
vortexed for 30 seconds, centrifuged at 3,000 rpm for 10 minutes and the supernatant
transferred to a test tube. The supernatant was dried under a stream of nitrogen at 50oC
for 45 minutes with the subsequent residue reconstituted in 100 L acetonitrile: water
(30:70). 30 L of the reconstituted residue was then injected onto a C-18 50mm x 3mm
column (Hypersil Gold, Thermo Scientific), and separated using an acetonitrile/0.01%
formic acid and water/10mM ammonium formate/0.01% formic acid gradient over
8.0 minutes at 40oC and analyzed on a mass spectrometer (Agilent 6460, Agilent
Technologies, Santa Clara, CA) using multiple reaction monitoring mode (MRM).
One and two MRM transitions were used as quantifier (m/z 436.1>144.9) and
qualifiers (m/z 436.1>231.1 and 436>73.0) respectively. Rivaroxaban calibrators (2.5,
5, 10, 25, 50, 100, 200 and 500 ng/mL) and quality control samples were prepared
in bovine serum. Thirteen serum samples from a total of twelve patients who were
receiving rivaroxaban were also analyzed.
Results The LC-MS/MS assay was linear across an analytical measurement range of
2.5-500 ng/mL with a slope of 1.004 and a correlation coefficient (r2) 0.998. The intra
assay imprecision was 2.3 % at 2.9 ng/mL (n=10), 2.9 % at 25 ng/mL (n=20), 1.2% at
36 ng/mL (n=20), and 2.0% at 217 ng/mL (n=20). The inter assay imprecision ranged
from 2.7-7.4%. The limit of detection was determined to be 0.5 ng/mL and the limit
of quantitation was set at 2.5 ng/mL. The rivaroxaban concentration in the patient
samples ranged between 3.0-220 ng/mL.
Conclusion We have developed an LC-MS/MS assay for measuring rivaroxaban
concentration in serum. In our initial cross sectional study, rivaroxaban could be
reliably quantitated in samples from patients receiving the drug. Measuring serum
concentration of direct acting anticoagulants is of clinical value, particularly for
managing active bleeding events and in patients undergoing elective procedures. In
the future, developing LC-MS/MS based anticoagulant drug panels are an attractive
option for clinical laboratories to facilitate an improved management of patients
receiving anticoagulation therapy by novel anticoagulants.

B-423
Development of Abbott Phenobarbital Assay for the ARCHITECT cSystems
Automated Clinical Chemistry Analyzers

J. Donaldson1, N. Pham1, A. Kong1, S. Hoang1, S. Shaw2, N. Pham2, O.

Ndimbie2, C. Kasal2, L. Ye1. 1Thermo Fisher Scientific, Fremont, CA,
2
Abbott Laboratories, Irving, TX,
Introduction: The objective of this study is to develop a sensitive immunoassay
intended for the in vitro quantitative measurement of phenobarbital in human serum
or plasma on the ARCHITECT cSystems. The measurements obtained are used in
the diagnosis and treatment of phenobarbital overdose and in monitoring levels of
phenobarbital to help ensure appropriate therapy. Monitoring serum phenobarbital
concentrations has been shown to improve patient therapy by providing physicians
with a tool for adjusting dosage.

Methods: The Abbott Phenobarbital Assay is a liquid stable, homogenous particleenhanced turbidimetric inhibition immunoassay used for the analysis of phenobarbital
in serum or plasma. The assay consists of two reagents and is based on competition
between phenobarbital in the sample and phenobarbital coated onto a micro-particle
for anti- phenobarbital antibody binding sites. In samples lacking phenobarbital, the
phenobarbital -coated micro-particles rapidly agglutinate in the presence of antiphenobarbital antibodies. The rate of absorbance change is measured photometrically,
and is directly proportional to the rate of particle agglutination. In samples containing
phenobarbital, the agglutination reaction is partially inhibited, slowing the rate of
absorbance change. A concentration-dependent classic agglutination inhibition curve
can be obtained, with maximum rate of agglutination at the lowest phenobarbital
concentration and the lowest agglutination rate at the highest phenobarbital
concentration.
Results: The performance of the Abbott Phenobarbital Assay was evaluated on
the Abbott ARCHITECT c8000 analyzer. Based on guidance from Clinical and
Laboratory Standards Institute (CLSI) protocol EP17-A2, the assay demonstrates a
Limit of Quantitation of < 2.0 g/mL using inter-assay precision < 7% CV or < 0.7 g/
mL SD and bias within 10% or 1.0 g/mL over an extended period. The assay is linear
from 2.0 to 80.0 g/mL using guidance from CLSI protocol EP6-A. Assay precision
was evaluated using CLSI guideline EP5-A2. A tri-level commercial control and six
human serum samples containing phenobarbital at concentrations ranging from 2.5
g/mL to 77.2 g/mL were tested. Each sample was assayed in duplicate twice a
day for 20 days with at least two hours between runs. The precision ranged from 1.1
%CV to 3.3 %CV for Within-Run and 1.7 %CV to 6.7 %CV for Total-Run. The assay
accurately recovered spiked phenobarbital at levels representing sub-therapeutic,
therapeutic, and toxic samples. No significant interference was observed with various
endogenous substances or compounds whose chemical structure or concurrent
therapeutic use would suggest possible cross-reactivity. Abbott ARCHITECT
Phenobarbital patient correlation studies: new vs. current on-market assay yielded
a regression equation of y=1.00x + 0.42 and a correlation coefficient of 1.00. New
assay vs. HPLC yielded y= 0.93x + 0.68 and a correlation coefficient of 0.99. The new
reagent has an onboard stability of 40 days and calibration curve stability of 14 days.
Conclusion: The Abbott Phenobarbital Assay enables measurement of phenobarbital
in human serum or plasma with high precision across the linear range. The ability
to monitor levels of phenobarbital with high accuracy can help ensure appropriate
therapy. The assay has applications on the ARCHITECT c16000, c8000, and c4000.

B-424
Development and Validation of a Robust Tandem LC-MS/MS Method for the
Quantification of Antidepressants in Serum

A. K. Petrides, J. Moskowitz, W. Clarke, M. A. Marzinke. The Johns

Hopkins University School of Medicine, Baltimore, MD,
Background: Depression is a rapidly growing issue in the United States, with
more than 11% of Americans taking antidepressant medications, allowing for
antidepressants to become the most commonly prescribed drug in the United States.
There are many drug classes that may be used to treat depression, including the
selective serotonin-reuptake inhibitors (SSRIs) such as citalopram (Celexa) and
sertraline (Zoloft), as well as the dopamine-reuptake inhibitor bupropion (Wellbutrin).
However, treatment success may be variable and can either result in lack of efficacy
due to suboptimal drug administration or non-compliance, as well as potential adverse
side effects. Success is further complicated by environmental and genetic factors, and
the delayed access to an optimal treatment regimen can have deleterious effects. Thus,
methods for drug quantification can become important tools in the assessment of drug
efficacy to optimize treatment regimens, particularly in scenarios where therapeutic
regimens are frequently altered. Here, we present a liquid chromatographic-tandem
mass spectrometric (LC-MS/MS) method for the robust, simultaneous quantification
of the commonly prescribed antidepressants citalopram, sertraline and bupropion, and
its active metabolite, hydroxybupropion, following FDA bioanalytical guidelines.
Methods: Drug-free human serum was spiked with citalopram, sertraline, bupropion
and hydroxybupropion, along with their corresponding isotopically-labeled internal
standards. Following protein precipitation, samples were injected onto the Prelude
turbulent-flow-liquid chromatography system (Thermo Fisher Scientific), consisting
of a Cyclone-P column for on-line solid phase extraction and a Hypersil Gold
C18 column for chromatographic separation. Sample elution was achieved using
acetonitrile (ACN) containing 0.1% formic acid. All four compounds were detected
over 4 minutes using a TSQ Vantage Quadrupole mass spectrometer (Thermo Fisher
Scientific) with a heated electrospray ionization (HESI) source in positive ionization
mode operated in selected reaction monitoring (SRM) mode. Assay validation
followed bioanalytical guidelines, and included precision (intra-assay) and accuracy,
linearity, and functional sensitivity studies. Results: The analytical measuring range

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S255

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

of the assay for each analyte was 5-1000 ng/ml. Linearity was assessed from the slope
of a 1/x2 weighted least squares-fitted linear regression analysis. Three independently
prepared and analyzed calibration curves were used to assess linearity. Average slopes
for citalopram, sertraline, bupropion and hydroxybupropion were 1.033, 1.036, 1.011,
and 1.042, respectively. Further, recoveries across the aforementioned analytical
measuring range for citalopram, sertraline, bupropion, and hydroxybupropion
were 94.7-104.1%, 94.9-104.2%, 92.5-103.3%, and 92-105.2%, respectively. To
assess precision, quality control solutions were prepared at 5 ng/ml (lower limit of
quantification), 25 ng/ml (low), 125 ng/ml (mid) and 850 ng/ml (high) levels. Intraassay precision was determined through analysis of six replicates of each level from
a calibration curve. Intra-assay precision ranged from 2.6-3.6% for citalopram, 3.811.4% for sertraline, 5.4-9.2% for bupropion and 3.0-6.8% for hydroxyburporion. The
functional sensitivity of the assay, which is the computational extrapolation of drug
concentration with a %CV=20%, was 0.05 ng/ml, 0.57 ng/ml, 1.54 ng/ml, and 1.4 ng/
ml using this method for citalopram, sertraline, bupropion and hydroxybupropion,
respectively.
Conclusion: The development and validation of this LC-MS/MS method allows
for the robust and high-throughput quantification of antidepressants commonly
prescribed to our patient population.

B-425
Detection of 55 Drugs and Pain Management Analytes in Urine Using a
Quantitative Liquid Chromatography-Tandem Mass Spectrometry (LC/MSMS). An All-In-One Screening and Confirmatory Method.

V. Ricchiuti, E. Chaffin, E. Eve, F. Lucas. University of Cincinnati Medical

Center, Cincinnati, OH,
Background. Screening drug tests are qualitative, and are conducted to identify
classes of drugs present in the urine using immunoassay-based methods. They rely
on a threshold above which a positive result is produced, and do not detect lower
concentrations of a drug. Confirmatory tests are used to verify a positive screening and
identify a specific drug. They use gas chromatography-mass spectrometry (GC-MS)
or high performance liquid chromatography-tandem MS (LC-MS/MS) methods. Our
goal was to validate an all-in-one screening and confirmatory method by LC-MS/MS
for 55 drugs and metabolites panel with lower limits of detection than immunoassays
and GC-MS.
Methods. 70 urine samples were screened using QuickTox Drug Dipcards (Branan
Medical Corporation, Irvine, CA). Confirmation by GC-MS was a send-out to a
large reference laboratory. LC-MS/MS was performed in-house using the Shimadzu
(Shimadzu Corporation, Kyoto, Japan) Prominence 20A Liquid Chromatograph,
followed by the AB SCIEX QTRAP 4500 Mass Spectrometry System (AB SCIEX,
Framingham, MA). Amphetamines, benzodiazepines, cannabinoids, cocaine, opiates,
opioids, oxycodone, fentanyl and analogues and buprenorphine were analyzed in
positive mode using the TurboIonSpray ion source (Electrospray ionization).
Barbiturates and marijuana metabolite were analyzed using the negative ion mode
for maximum sensitivity. 50 L of urine was mixed with 20 L of combined internal
standard (IS) solution to each tube. Hydrolysis was performed by adding 10 L of
the -glucuronidase solution (100,000 units/mL) to each tube and incubation for
two hours at 55C. 110 L of the curve diluent was added to each tube, mixed and
centrifuged to separate any proteins and the supernatant. Samples were then analyzed
by LC-MS/MS based on the presence of the specific MRM transitions for each
analyte at the correct retention time. Quantitative measurement was accomplished by
normalization of the peak area with the area of the IS for each specimen, including
matrix specific calibrators, and quality control materials. The instrument software
program automatically constructs a calibration curve, using the peak abundance data
from the calibrator samples.
Results. Analytical performance of LC-MS/MS was excellent. CVs were <5% within
the analytical measurement range. Among all the drugs analyzed, fentanyl showed
the lowest estimated limit of detection (LOD) and limit of quantitation (LOQ), 0.003
ng/mL and 0.009 ng/mL, respectively. LOQ for fentanyl at CV<20% was <0.05 ng/
mL and was used as clinical cut-off. LC-MS/MS clinical cut-offs ranged from 0.05
(fentanyl) to 20 ng/ml (barbiturates). Recoveries within the analytical measurement
range were between 95-105%. Qualitative method comparison between GC-MS and
LC-MS/MS using GC-MS cut-offs showed 91% agreement. Quantitative correlation
between GC-MS and LC-MS/MS including all drugs was excellent (slope=1.04,
r=0.99, p<0.05, n=21). 42% of initial drugs screened by immunoassay were confirmed
by GC-MS, however 80% were confirmed by LC-MS/MS using our clinical cut-offs.
Turn-around-time (TAT) was between 24-48 hours from receiving the specimen.
Conclusion. The quantitative drug analysis method by LC/MS-MS is easy and userfriendly with excellent analytical sensitivities. It requires one-step for extraction of
drugs from urine and a direct injection into the LC/MS-MS system. Low volume
of urine needed as well as analytical sensitivity is ideal for the Neonate Abstinence
Syndrome program.

S256

B-426
Formation of 6-monoaetylmorphine in urine specimens with high morphine
concentrations during enzymatic hydrolysis

C. Heideloff1, J. Gabler2, C. Yuan1, L. Zhang3, S. Wang1. 1Cleveland

Clinic, Cleveland, OH, 2Thermo Fisher Scientific, West Palm Beach, FL,
3
Cleveland State University, Cleveland, OH,
Background: 6-monoacetylmorphine (6-MAM), a unique metabolite of heroin, is
known as a definitive indicator of heroin intake. 6-MAM in urine is measured by
GC-MS or LC-MS methods. Due to variable glucuronide conjugation rates between
individuals, some laboratories employ enzymatic hydrolysis using glucuronidase
during sample preparation to improve detection consistency. Acetate buffer is the
primary choice for preparing enzymatic hydrolysis solution. In our routine LCMS/MS analysis, we observed a number of low levels of 6-MAM in patient urine
specimens with morphine at high concentrations. We hypothesized that the acetate
buffer used for enzymatic hydrolysis serves as an acetylating agent leading to the
formation of 6-MAM in urine specimens with high morphine concentrations. Design:
Two sets of studies were performed. In one set, morphine standard and the major
morphine metabolite, morphine-3-glucuronide (M3G), were spiked into morphine
negative urine samples (n=2) to create five levels of morphine specimens (10,00050,000- 100,000- 150,000- 200,000 ng/ml) and M3G specimens (16,250- 81,250162,500- 243,750- 325,000 ng/ml) respectively. Internal standards (6-MAM-D3
and Morphine-D3) were added to the samples and the mixtures were incubated with
1M sodium acetate buffer (pH 4.5) at 60C. Aliquots were taken at timed intervals
(0, 2, 4, 6, 18 hours). In the 2nd set, leftover urine specimens (n=4) with elevated
concentrations of morphine were incubated with 1M sodium acetate buffer (pH
4.5) at 60 C for 0 and 18 hours. An additional study was performed to optimize
the enzymatic hydrolysis procedure using an alternate buffer (1M citrate buffer, pH
4.5) to determine if 6-MAM would still be formed. The urine specimens used in the
2nd study, as well as the 200,000 ng/mL morphine and 325,000 ng/mL M3G spiked
samples from the 1st study were incubated with 1M citrate buffer (pH 4.5) at 0 and 18
hours. Analysis was performed by an LC-MS/MS method. Results: Urine samples
with elevated levels of both free morphine and M3G (>100,000 ng/mL) incubated for
18 hours using acetate buffer formed measurable amounts ( 5 ng/mL) of 6-MAM.
All samples with <100,000 ng/mL and all samples incubated <18 hours did not form
measurable amounts of 6-MAM. In all samples using citrate buffer, no 6-MAM was
at the measurable level. Conclusion: False positive identification of heroin may
be possible in urine specimens with extremely high levels (>100,000 ng/mL) of
morphine when using acetate buffer for enzymatic hydrolysis.

B-427
Development of a Homogenous Enzyme Immunoassay for the Screening of
Synthetic Cannabinoids Applicable to Automated Chemistry Analysers

S. Smyth, V. Anderson, J. Darragh, P. Ratcliffe, S. Smillie, L. Long, M.

Benchikh, R. McConnell, S. FitzGerald. Randox Toxicology Limited,
Crumlin, United Kingdom,
Background. Synthetic cannabinoids are chemical compounds that mimic the effects
of tetrahydrocannabinol, the main active ingredient of cannabis. Originally sold under
the brand name Spice, this brand name has become a generic term to include the
entire class of legal smoking blends sold on the internet.The most common of the
first wave of synthetic cannabinoid compounds available was JWH018. As these
designer drugs continue to be sold there is a need for screening tests, which facilitate
the detection process. Immunoassays are antibody-based tests that provide high
throughput screening. The application of these tests to automated chemistry analysers
is advantageous in clinical laboratory testing settings as it increases the screening
capacity. This study reports the development of a homogeneous enzyme immunoassay
for the screening of JWH018 and related metabolites in urine, applicable to a variety
of automated systems.
Methods. The assay is based on competition between drug in the sample and drug
labelled with glucose-6-phosphate dehydrogenase (G6PDH) for a fixed amount of
antibody in the reagent. In the absence of drug in the sample, JWH018-labelled
G6PDH conjugate is bound to antibody, and the
enzyme activity is inhibited. However, when free drug is present in the sample, the
antibody binds to free drug and the unbound JWH018-labelled G6PDH then exhibits
maximal enzyme activity. Active enzyme converts nicotinamide adenine dinucleotide
(NAD) to NADH, resulting in an absorbance change that can be measured
spectrophotometrically at 340nm. The assay is qualitative utilising a 15ng/ml cut-off
and in this evaluation the RX daytona analyser was used.
Results. The assay was standardised to the major urinary metabolite JWH018 N

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

pentanoic acid, 55 synthetic cannabinoidswere also detected with %cross-reactivity

(%CR) values >5%, including the 5 hydroxypentyl metabolite and 4 hydroxypentyl
metabolite of JWH018 (%CR: 62% and 24% respectively) and AM2201 N- 4hydroxypentyl metabolite (%CR: 114%). The limit of detection in urine (n=20) was
5.2ng/ml (measuring range 0-20ng/ml). Inter-assay precision (n=20) was calculated as
3.07% at 10ng/ml, 1.89% at 15ng/ml and 2.06% at 20ng/ml. Total assay precision
was assessed by running a negative (10ng/ml) and positive (20ng/ml) spiked urine
samples, 2 replicates per run, 2 runs per day over 20 days. All replicates were correctly
reported as negative and positive (n=80). Recovery was assessed at 10, 15 and 20ng/
ml and all samples (n=20) showed recovery of between 115-124%. 101 samples were
assessed against LC/MS and exhibited 92% agreement. When 245 samples were
assessed with this assay and a commercially available biochip based immunoassay an
agreement of 95% was obtained.
Conclusion. The results show that the developed homogeneous enzyme immunoassay
for the screening of synthetic cannabinoids, exhibits optimal analytical. The assay
presents a broad cross-reactivity profile, which increases the screening capacity.
Moreover, the application to automated chemistry analysers improves the screening
for designer drugs in clinical laboratories.

B-428
ARK Voriconazole Assay for the Roche/Hitachi Modular P Automated
Clinical Chemistry Analyzer

S. J. Oh, K. Pham, B. Moon, J. J. Valdez. ARK Diagnostics, inc, Fremont,

CA,
Background: Despite the availability of newer antifungal agents, invasive fungal
diseases remain a leading cause of morbidity and mortality in immunocompromised
patients. Voriconazole (VRZ) is an extended-spectrum triazole indicated for treatment
of invasive fungal diseases. Here an ARK enzyme immunoassay for therapeutic drug
monitoring (TDM) of VRZ is described.
Methods: The ARK Voriconazole Assay is a liquid stable homogeneous enzyme
immunoassay, consisting of two reagents, 6 calibrators (0.00, 1.00, 2.00, 4.00, 8.00
and 16.00 g/mL) and 3 controls (1.50, 5.00 and 10.00 g/mL). The performance
of the ARK assay was evaluated on the Roche/Hitachi Modular P analyzer. Limit of
quantitation, linearity, precision, recovery, specificity and method comparison were
studied.
Results: The assay measured concentrations as low as 0.50 g/mL and was linear
to 16.00 g/mL. Total precision ranged 5.1% to 6.3%CV and within-run precision
ranged 4.3% to 4.8%CV in a 20-day study (CLSI guideline EP15-A2) that evaluated
precision for the 3 quality controls in a synthetic proteinaceous matrix and for
similar concentrations in spiked serum samples. The assay accurately recovered
spiked VRZ samples throughout the assay range. The assay did not crossreact with
the antifungal fluconazole, itraconazole, and posaconazole. Voriconazole N-oxide
(major metabolite) was tested at 10 g/mL and no crossreactivity was observed.
No interference from endogenous substances, anticoagulants and potentially coadministered drugs occurred at the elevated concentrations studied. Seventy-four
specimens at concentrations throughout the range (0.5 to 8.1 g/mL) were assayed
and gave the following Passing Bablock regression results when compared to LC/MS/
MS values: ARK = 1.01 LC/MS/MS + 0.07 (r2=0.98).
Conclusions: The ARK Voriconazole Assay measures VRZ in human serum or plasma
with excellent precision at very low concentrations which is essential for long-term
monitoring of patients. Ability to measure trough levels of VRZ with high accuracy
and fast turn-around time makes this method clinically useful for VRZ TDM.

B-429
Ultrafast, high-throughput quantitative analysis of Carboxy-THC in urine
using Laser Diode Thermal Desorption coupled to tandem mass spectrometry

A. Birsan, S. Auger, P. Picard, J. Lacoursiere. Phytronix, Quebec, QC,

Canada,
Background: The 11-Nor-9-Carboxy-THC (THCC) is a major metabolite of THC
in urine. The confirmation of THCC in patients is often conducted in urine as a
non-invasive procedure to determine the presence of drugs of abuse. The detection
and quantification of THCC are traditionally performed by gas (GC) or liquid (LC)
chromatography coupled to mass spectrometry (MS) or tandem mass spectrometry
(MS/MS) methods.
The laser diode thermal desorption (LDTD) ion source uses an infrared laser diode
to thermally desorb neutral species of THCC molecules from a dried sample. These
neutral species are carried into a corona discharge region, where they undergo

efficient protonation and are directed into the mass spectrometer. The total analysis
time is performed very rapidly (9 seconds).
The objective of this work is to validate this analysis method and test different real
samples using the LDTD-MS/MS. A cross validation study against the LC-MS/MS
approach for the analysis of THCC was done in order to evaluate the performance of
the alternative LDTD-MS/MS developed method.
Methods: A calibration curve and quality control materials were prepared in blank
urine samples. To 500 uL of calibrators, QC and patient specimens, 50 L of internal
standard (THCC-d9, 5ug/mL in MeOH) and 200 uL KOH (3N) were added. The
mixture was vortex-mixed and incubated at 38C for 15 minutes for the glucuronide
hydrolysis. A liquid-liquid extraction was then performed by adding 200 uL of HCl
(6N) and 1000 uL of Hexane:EtAc (95:5). After vortexing and centrifugation at 5000
rpm for 2 min, 6 uL of the organic layer was deposited in the LazWell Plate (precoated with EDTA 200 ug/mL). The LDTD laser power was ramped to 45% in 3
seconds, and shut down after 1 second. Negative ionization mode was used, and API5500 QTrap system was operated in MRM mode with MS transitions of 343->245 and
352->254 for THCC and THCC-d9 respectively.
Results: The calibration curves show excellent linearity with r2 = 0.9991 between the
quantification range of 5 to 5000 ng/mL. Inter-run, intra-run accuracy, and precision
ranges between 95,8 % and 103,1 % with RSD from 8,5 to 13,5%, respectively. No
matrix effect or carryover was observed. The stability of THCC in and out (dry sample)
of solution was evaluated for a period of 3-days. This method was cross validated
with results from a traditional LC-MS/MS method for 49 real patient specimens. A
good correlation between LC-MS/MS and LDTD-MS/MS data is obtained with r2 =
0.9956. All negative samples correlated accordingly.
Conclusion: LDTD technology provides unique advantages in developing an
ultra-fast method for analysis of 11-Nor-9-Carboxy-THC in urine. This method has
demonstrated accurate, precise and stable results with a run time of 9 seconds.

B-430
High-Throughput Quantitative Analysis of 5 Antifungal Drugs in Human
Serum

M. Youssef, V. Miller, W. LaMarr. Agilent Technologies, Wakefield, MA,

Research laboratories traditionally rely on HPLC and more recently LC/MS/MS for
quantitative analysis of antifungal drugs (AFs). The objective of this study was to
develop a rapid and very robust online SPE/MS/MS method for high throughput and
accurate quantitation of 5 antifungal drugs (voriconazole, ketoconazole, fluconazole,
itraconazole, and posaconazole) in human serum. This method employs protein
precipitation followed by dilute and shoot on the SPE/MS/MS system, enabling
analysis of all 5 AFs at 14 seconds per sample producing >20x savings in analysis
time and solvent consumption compared to typical analytical methods. Samples were
prepared by spiking AFs into drug-free human serum followed by adding internal
standard mix, a protein crash with acetonitrile and then diluting samples 10-fold with
a basic dilution. Samples were then analyzed via SPE/MS/MS using a reversed-phase
C18 cartridge at 14 seconds per sample. A simple protein precipitation methodology
followed by (dilutes and shoot) and analysis by SPE/MS/MS allows for the accurate
and precise measurement of these analytes in human serum over a linear range of
(0.2 - 25 mcg/mL). Standard curves consisting of each AF spiked into serum had
excellent linearity within the measured range with an R2 value greater than 0.995.
This methodology is capable of throughputs greater than 240 samples per hour
providing a high-throughput and very efficient mode of analysis. Carryover was
assessed by analyzing the AUC of the blank calculated as % of the mean peak area
of the 0.2 mcg/mL samples. No significant carryover (< 5.0%) was determined for
all of the AFs. QC standards for each AF were run over a series of days to establish
both intra- and interday precision and accuracy values. The accuracies determined
were within 10% and coefficient of variation values were all less than 10% for
concentrations within the measured range. The reproducibility of the method was
evaluated by measuring >2000 sequential injections of all five AFs spiked into serum.
The instrument response was stable for each of the five analytes with coefficient of
variation ranging from 2.3-6.9 % showing the robustness of the system, SPE cartridge
lifetime and consistency of quantitation for the analytes in the panel.A panel of five
antifungal drugs including: voriconazole, ketoconazole, fluconazole, itraconazole,
and posaconazole were quickly, accurately, and precisely measured in serum using
a simple protein precipitation protocol and an SPE/MS/MS system. Samples were
analyzed at 14 seconds per sample, providing a high-throughput method capable of
analyzing more than 240 samples per hour. This methodology provides comparable
results to HPLC, and LC/MS/MS, but at > 20x the speed and efficiency of typical
methods.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S257

TDM/Toxicology/DAU

Wednesday, July 30, 9:30 am 5:00 pm

B-431
Measurement of the antiepileptic drugs, lacosamide and lamotrigine, in human
serum by liquid chromatography-tandem mass spectrometry

D. M. Garby, R. DelRosso, J. A. Edwards, L. A. Cheryk. Mayo Medical

Laboratories, Andover, MA,
Background: Lacosamide and lamotrigine are antiepileptic drugs for whom the
monitoring of drug levels allows for maximization of their seizure suppressing effects
while minimizing the prevalence of adverse side effects. Lacosamide (VIMPAT) is
approved for adjunctive therapy to treat partial-onset seizures in epileptic patients 17
years of age and older. Lamotrigine (Lamictal) is approved for treatment of bipolar
I disorder and a variety of seizure disorders including Lennox-Gastaut syndrome,
primary generalized tonic-clonic seizures and partial-onset seizures. Both lacosamide
and lamotrigine are completely absorbed after oral administration with negligible
first-pass metabolism. Peak plasma concentrations occur within one to five hours after
oral dosing, and the elimination half-life (in adults) is approximately 13 hours for
lacosamide and 25 to 33 hours for lamotrigine. Both are believed to modulate the
activity of voltage-gated sodium channels, in turn stabilizing neuronal activity.
Methods: Stable drug isotopes (lacosamide-13C,D3 and lamotrigine-13C,15N4) are
added to 50uL of serum as internal standards. Protein is precipitated from the mixture
by the addition of acetonitrile. The supernatant is then further diluted with deionized
water. Lacosamide, lamotrigine and their respective internal standards are then
separated from the other serum constituents by liquid chromatography (TLX4, Thermo
Fisher Scientific, Waltham, MA) using a Kinetex 5m C18 50x4.6mm column
(Phenomenex, Torrance, CA) followed by analysis on a tandem mass spectrometer
(QTRAP 6500, AB SCIEX, Framingham, MA) equipped with an electrospray
ionization source in positive mode. Ion transitions monitored in the multiple reaction
monitoring (MRM) mode are m/z 255.9 m/z 211.0 for lamotrigine, m/z 255.9
m/z 144.9 for lamotrigine ion pair, m/z 261.0 m/z 144.9 for lamotrigine-13C,15N4,
m/z 251.0 m/z 108.0 for lacosamide, m/z 251.0 m/z 90.9 for lacosamide ion
pair, and m/z 255.0 m/z 108.0 for lacosamide-13C,D3. Calibrators consist of six
standard solutions ranging from 0 to 50ug/mL.

mobile phase consisting of (A) 0.05% formic acid in 5 mM ammonium formate (aq.)
and (B) 0.05% formic acid in acetonitrile/methanol (50:50 v/v) was ramped linearly
from 2% to 100% B over 10 minutes. To determine the performance of our assay, 100
patient comparison urine samples were diluted 1:5 with 12.5% acetonitrile/methanol
(50:50 v/v) in water and spiked with fentanyl-d5 internal standard. Data were collected
on an ABSciex TripleTOF 5600 System operating in full-scan positive ion mode
with IDA-triggered acquisition of product ion spectra. Data were analyzed using the
MasterView function of PeakView software (AB Sciex) and Analyst TF (AB Sciex,
Version 1.5.1). Criteria for positive identification of a drug included chromatographic
retention time, accurate mass, isotope pattern, and library match. Results from the
HRMS analysis were compared to several other methods (LC-MS/MS, LC-Orbitrap,
gas chromatography-mass spectrometry, immunoassay and patient prescription
history) to determine whether a compound was a true positive or false positive.
Results: Retention times were established for 210 compounds with coefficient of
variations that ranged from 0 to 9%. The lower limit of detection was 25 ng/mL or
less for 83% of the compounds. Using a targeted approach for data analysis and an
in-house spectral library, our HRMS method found 523 positive hits in the patient
comparison samples. Of these candidate hits, 509 (97%) were verified by another
method meaning that only 14 were deemed to be false positives. The HRMS missed
52 compounds that were identified by one or more of the reference methods.
Conclusion: Overall, the HRMS method identified compounds with high confidence;
97% of the compounds found by HRMS were verified by another method. The
majority of the 52 missed compounds could be attributed to differences in lower
limits of detection. In general, the targeted, LC-MS/MS method had slightly better
sensitivity than the HRMS method. However, the ability to detect unknown or
unexpected compounds still makes HRMS very appealing. Therefore, a combination
of LC-MS/MS and HRMS may be the most effective approach for broad spectrum
drug screening.

Results: Method performance was assessed using accuracy, precision, linearity

and specimen stability. Accuracy of the method was assessed by comparison to
external reference laboratories performing analysis of these drugs by tandem mass
spectrometry. Comparison to the external laboratories demonstrated excellent
agreement between the methods with overall differences of less than 3%. Precision
studies were performed using commercially available quality control material (AEDII,
UTAK Laboratories Inc., Valencia, CA) and an in-house prepared serum pool fortified
with lacosamide and lamotrigine standard solutions. Intra-run precision coefficients
of variation (CVs) ranged from 1.9% to 2.1% for lacosamide and 1.4% to 2.8% for
lamotrigine. Inter-run precision CVs ranged from 3.3% to 4.7% for lacosamide, and
3.1% to 7.1% for lamotrigine. Linearity studies were performed using serum fortified
with standard drug solutions. Linearity was demonstrated over the assay range (0.2
to 50ug/mL) for each analyte, yielding the following equations: observed lacosamide
value = 1.0331*(expected value) - 0.1118, R2 = 0.9999; observed lamotrigine value =
1.0194*(expected value) - 0.0590, R2 = 1.0000. Specimens were stable when stored at
ambient, refrigerate and frozen temperatures for up to 32 days.
Conclusion: This method provides for the simultaneous and reliable analysis of
lacosamide and lamotrigine in human serum.

B-432
Development of a comprehensive drug screen in urine using liquidchromatography quadrupole time of flight mass spectrometry.

K. L. Thoren, J. M. Colby, S. B. Shugarts, A. H. B. Wu, K. L. Lynch. UCSan Francisco/San Francisco General Hospital, San Francisco, CA,
Background: Recently, there has been much interest in using high resolution mass
spectrometry (HRMS) for drug screening applications as it offers several advantages
over tandem-based methods (LC-MS/MS). Importantly, HRMS instruments are often
used in an untargeted manner. This is especially attractive for broad spectrum drug
screening because it allows for potential identification of unknown or unexpected
compounds. Despite the interest in these techniques, few studies have investigated
the performance of high resolution instruments in comprehensive drug screening. The
objective of this study was to develop a broad spectrum drug screen for urine using
LC-HRMS and to determine its performance in identifying drugs and metabolites in
100 routine clinical samples.
Methods: All chromatographic separations were performed on a Phenomenex
Kinetex 2.6-m C18 column (3 x 50 mm) thermostatted at 30 degrees C. A binary

S258

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Technology/Design Development

Wednesday, July 30, 9:30 am 5:00 pm

care sensors that can provide accurate data without the need for expensive equipment.
Semiconductor manufacturing techniques offer a particularly attractive opportunity
to design a biosensor that is inexpensive, scalable, easy to integrate with portable
electronics, and highly sensitive to target analytes due to device size. Such fabrication
techniques have been honed to perfection with optimization and standardization.
Prototyping of new device designs in a semiconductor manufacturing manner can
critically enable rapid development and commercialization of new in vitro diagnostic
assays.

Wednesday, July 30, 2014

Poster Session: 9:30 AM - 5:00 PM
Technology/Design Development

B-433
Determination of Aminothiol Adsorption Properties of Nano-Sized
Titanium(IV) Oxide by HPLC-FLD

A. Stella1, S. Hsieh2, M. Garelnabi1, J. Horta1, E. J. Rogers1. 1University of

Massachusetts Lowell, Lowell, MA, 2National Taiwan University, Taipei,
Taiwan,
Metal oxide nanoparticles are known for their optical, electrochemical, electrical,
gravimetric, acoustic, and magnetic properties which make them appropriate for
laboratory diagnostics. Among these, titanium(IV) oxide (a.k.a. titanium dioxide,
TiO2) nanoparticles display superior photocatalytic properties. In the present study,
the adsorptive properties of eight TiO2 polymorphs were evaluated when exposed to
the toxic thiol-containing amino acid homocysteine. Homocysteine is an amino acid
known to cause cardiovascular toxicity and neurodegenerative disorders and can be
adsorbed to TiO2 polymorphs to different degrees depending on the physicochemical
characteristics of the polymorph. This adsorption quality may provide an alternative/
point of care biosensing approach to homocysteine measurement.
A homocysteine standard solution was combined with dispersions of each TiO2
polymorph under physiological conditions. After exposure, an HPLC fluorescence
detection (HPLC-FLD) method optimized for quantification of total, unbound
homocysteine was utilized and showed a range of results for the polymorphs studied.
An initial centrifugation (1,280 RCF) step, followed by two microcentrifugation
(11,152 RCF) steps were performed to ensure nanoparticle removal. The supernatants
produced were used for homocysteine analysis. A thiol-specific derivatization
agent, 7-fluorobenzofurazan-4-sulfonic acid (SBD-F) was employed to derivatize
non-adsorbed homocysteine. The excitation wavelength was set to 385 nm and the
emission wavelength was set to 515 nm. The peaks produced by the fluorescence
detector were non-adsorbed homocysteine in the reduced form. It is possible that
the homocysteine may have become oxidized by the nano-sized polymorphs upon
contact, however tris-(2-carboxyethyl)phosphine (TCEP) was used to convert all
homocysteine back to the reduced form for derivatization.
The nano-sized anatase polymorph adsorbed 2.92 g of homocysteine per mg of
polymorph, whereas amorphous TiO2 only adsorbed 3.65 x10-3 g of homocysteine
per mg of polymorph; both values were corrected for primary particle size. Results
were determined by peak area comparison to the 5 mol blank standard. Other
polymorphs produced values between the anatase and amorphous polymorphs.
Surface chemistries are distinguishing characteristics of these polymorphs and
are responsible for their individual physicochemical properties. Variations in the
observations were attributed to the unique combinations of size, surface area and
polarity of each polymorph. The aminothiol adsorptive property of titanium dioxide
polymorphs has potential applications in nanomedicine, biosensing, diagnostics,
and amino acid tagging. This HPLC-FLD evaluation of homocysteine adsorption
effectively determined which polymorphs may be best suited for applications in the
clinical laboratory for diagnostic purposes.

B-434
Versatile Electrical Platform for Accelerated Development and
Commercialization of In Vitro Diagnostic Assays

C. Cheng1, B. Reddy2, F. Lai1, P. Yen3, C. Duarte2, E. Salm2, F. Tsui1,

C. Wen1, T. Chen1, J. Huang1, Y. Hsieh1, C. Mauracher4, G. Bauer4, M.
Yamamoto4, R. Bashir2, Y. Liu1. 1Taiwan Semiconductor Manufacturing
Company, Hsinchu, Taiwan, 2University of Illinois at Urbana-Champaign,
Champaign, IL, 3National Taiwan University, Taipei, Taiwan, 4SONY
DADC Bioscience, Salzburg, Austria,
Background: Diagnostic techniques have become critical to modern health care
not only for patient diagnosis and optimization of treatment, but also for providing
vital information regarding pathways for complex diseases. Ideal biosensors provide
the maximum desirable data with the least amount of complexity. However, to date,
most biosensors either yield very specialized data or are very complex and expensive.
Researchers have attempted to address this need with the development of point-of-

Methods: We present here a versatile electrical biosensor platform consisting of tens

of thousands of devices fabricated with a 0.18 m silicon-on-insulator technology.
The platform consists of a unit cell transistor integrated seamlessly with control
and read-out circuitry that is amenable for immediate commercialization. Each cell
consists of a sub-micron FET Sensor with a near-Nernst pH sensitivity of around
56-59 mV/pH and a resolution of <0.01 pH. The gate oxide is directly exposed to
the target analytes, instead of to the commonly employed floating gate architecture.
This enables many critical advantages, including increased sensitivity due to the
elimination of parasitic coupling capacitances and reduced vulnerability to crippling
factors such as electrostatic discharge. The unit cell can be coupled to a variety of
different surface chemistries for different target analytes and applications.
Results: We verified the device performance and potential applications by detection
of urea level via an enzyme (urease)-catalyzed reaction. A high level of urea in blood
plasma can indicate both potential kidney failure and the onset of kidney diseases.
Compared to reported literature (Lai et al, 2001) and commercially available kits
(SIGMA ALDRICH), our devices exhibited a linear detection response in the
hundreds of picomoles to nanomoles range with a sample volume of 0.1 l, an
improvement of one order in detection limit and three orders in required sample
volume. In addition, we verified the device performance by the detection of DNA
hybridization. As low as picomoles of DNA molecules were easily detectable, a limit
that could be pushed down to femtomoles with integrated polymer microfluidics.
Conclusion: In summary, this work introduces a fully electronic biosensor platform
produced using a semiconductor manufacturing foundry. As proofs of concept of the
functionality of the unit cell, we demonstrated the detection of enzymatic reactions
and DNA hybridization. Foundry fabricated FET sensors with integrated polymer
microfluidics have the potential to enable highly cost effective, mass fabricated POC
devices with better sensitivity and resolution than currently commercially available
solutions. The finely tuned and high throughput nature of a semiconductor foundry
can translate to immediate commercialization of electronic biomedical assays.

B-435
An easy, portable and rapid way for performing PCR reactions

M. Gramegna1, L. J. Turner1, L. Ventura1, M. L. Incandela1, M. Bianchessi2,

M. Cereda2, A. Cocci2, A. Moiana1. 1Sentinel CH, Milano, Italy,
2
STMicroelectronics, Agrate Brianza, Italy,
Background. Limits to the wide diffusion of molecular diagnostics are the elevated
cost of thermalcyclers and the need of trained technicians that perform PCR reaction
and analysis. Another limit is that PCR mixes, enzymes and components are typically
stored at -20 or +2-8C. Repeated freeze-thaw cycles could negatively impact on
the assays performance. We developed innovative technology to produce a freezedried mix for the amplification of nucleic acids, which is ready-to-use, pre-dispensed,
flexible and room-temperature storable. This technology is successfully applicable
to PCR or RT-PCR, End-Point PCR or Real-Time PCR, in singleplex or multiplex
reactions. Objective. The aim of the present work was to couple two technologies,
real-time chip technology and a technology that stabilize PCR reagents at roomtemperature, in order to obtain a rapid, user-friendly and easy-to-store PCR system.
Methods. A PCR mix was prepared as follows: reaction buffer, dNTPs, MgCl2, DNA
polymerase, primers, probe, preservatives and stabilizers. Around 5 L of this mix
was spotted in each chamber of a chip and freeze-dried using an Epsilon 2-12D
freeze-dryer (Martin Christ). Performances were evaluated on a portable PCR system
(STMicroelectronics). Results and conclusions. Real-time PCR performances
evaluated on chip with the portable PCR system matched those obtained with a
classical 7500 Real-time PCR System (Applied Biosystems). This suggests that it is
possible to perform real-time PCR on a small, portable instrument with reduced test
turnaround time, low-volume needs and with a user-friendly software; with freezedried reagents that are easy-to-store. This PCR system can be used at the point of
impact in point-of-care molecular diagnostics.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S259

Technology/Design Development

Wednesday, July 30, 9:30 am 5:00 pm

B-437
Antioxidant profiles of human serum and biological fluids

A. Sharma1, K. Kline2, A. M. Sawaya1, R. L. Leveille1. 1Central Michigan

University, Mt. Pleasant, MI, 2McLaren Central Hospital, Mt. Pleasant,
MI,
Background: Free radicals are strongly associated with numerous human diseases
such as cancer, cardiovascular and neurodegenerative disorders.
Objective: The objective of the present study was to develop a test to assess the
antioxidant profile of human serum and other biological fluids.
Methods: Human serum samples were subjected to serum protein electrophoresis
in agarose gels using conditions similar for separation of lipoproteins or proteins
(migration towards anode). The gels were then stained with an activity stain based
on the ability of antioxidants in serum to reduce ferricyanide to ferrocyanide. In the
presence of ferric ions, strong antioxidant components in serum yielded a dark blue
band.
Results: Evaluation of the activity staining methodology gave positive bands only
with strong antioxidants such as vitamin C, quercetin and Trolox (a water-soluble
derivative of vitamin E). Uric acid and certain amino acids gave very weak bands.
Twenty random serum samples were evaluated for their antioxidant profiles after serum
electrophoresis. Different profiles were observed for various samples suggesting that
this test may be potentially used for diagnosis and management of human diseases that
are strongly associated with free radicals. Some of the antioxidant bands corresponded
to LDL, VLDL and HDL. However, there was no direct correlation between the
intensity of the antioxidant and lipoprotein bands. For example, some samples showed
weak lipoprotein bands but strong antioxidant activities. All samples also showed a
strong reducing activity at the point of origin (gamma globulin region).
Conclusion: This is the first report of visualizing antioxidant activities of serum
components. Profiles of other human biological fluids will also be presented. Studies
are currently underway to assess the serum antioxidant profiles of patients with
various diseases compared to a healthy population.

B-438
Bioenergetic health index: a predictive biomarker of human health that
combines the effects of systemic stress and disease susceptibility

B. K. Chacko, P. Kramer, T. Mitchell, S. Ravi, M. Johnson, G. Benavides,

R. Hardy, V. Darley-Usmar. University of Alabama at Birmingham,
Birmingham, AL,
Background: The increasing incidence of bioenergetics-related diseases and the
personalized nature of these disease progression is one of the major health concerns
worldwide. Emerging literature suggest the existence of substantial diversity in
individual susceptibility to diseases associated with energetic dysfunction such
as diabetes, alcoholoc liver disease and cancer. This opens up a valuable avenue
to develop a personalized predicitive biomarker in the diagnosis and management
of diseases involving bioenergetic alterations. However, no clinical test capable of
determining bioenergetic health exists to stratify these patients. Several studies
studies have demonstrated the influence of genetic, environmental and lifestyle factors
and age in disease susceptibility and progression and also in the individual bioenergetic
function. Our recent findings support an emerging concept that circulating leukocytes
and platelets can act as sensors or biomarkers of bioenergetic dysfunction that occurs
in chronic diseases. It is proposed that chronic disease-induced systemic stress
(inflammation, oxidative stress etc) will cause alterations in leukocyte bioenergetics,
which is the resultant of the intensity of stress and individual health. This suggests
that oxidative stress induced typical changes in leukocyte mitochondrial function can
be used as an indicator of bioenergetic health of individuals. Hence it is hypothesized
that systemic stress alters the bioenergetic capability in leukocytes and platelets and
these alterations can be used to assess individual health and disease susceptibility.
Methods: In this study, by inducing oxidative stress, the impact of stress on
bioenergetic capability of human leukocytes and platelets isolated from different
individuals is demonstrated. The bioenergetic capability of cells is determined using
the extracellular flux analyzer (Seahorse Biosciences). Oxidative stress was induced
in isolated leukocytes (monocytes, lymphocytes and neutrophils) and platelets using
well characterized oxidants, such as dimethylnaphthoquinone and lipid peroxidation
product 4-hydroxynonenal. Using the individual parameters of bioenergetic assay,
the bioenergetic index(BHI) is calculated for each cell type.
Results: We demonstrate that oxidative stress induces different degrees of
mitochondrial dysfunction in different individuals. Our results also show that these
oxidant-induced changes in the profiles of bioenergetic parameters is characterized

S260

by increase in basal respiration, proton leak and non-mitochondrial respiration and

a decrease in ATP-linked and maximal respiration in freshly isolated monocytes,
lymphocytes, neutrophils and platelets. These alterations also show typical profiles for
different individuals. In addition, different degrees of oxidative stress is required for
bioenergetic dysfunction in different individuals suggesting the personalized nature
of this parameter. Using the bioenergetic parameters of oxidant-treated leukocytes
and platelets we also demonstrate the development of the bioenergetic health index, a
functional measure that can define the health of individuals.
Conclusion: These novel findings suggest that peripheral blood leukocytes from
different individuals significantly differ in their response to oxidative stress which is
reflected in the bioenergetic health index in monocytes, lymphocytes, neutrophils and
platelets. Taken together, bioenergetic health index is a novel parameter that is derived
from the mitochondrial function of cells and demonstrates the interplay between
innate health and acquired risk factors. It is also suggested that the bioenergetic
health index has the potential to be used as a clinical test in developing personalized
management/therapeutic strategies in patients.

B-439
Development of an HCV Ag-Ab COMB ELISA Kit and Evaluation in Taiwan
Population

Y. Wu, I. Huang, W. Lu, J. Lin. General Biologicals corporation, Hsinchu,

Taiwan,
Background: Worldwide infection statistics showed 184 million people with
antibodies to Hepatitis C (HCV). HCV infection is important risk factor of
hepatocellular carcinoma, which was ranked in the second place among worldwide
cancer incidences. There is still no vaccine or immune globulin products specific
to HCV infection, so the prevention is relied on screening. To provide a screening
method that reduces window period and testing cost, we developed an enzymelinked immunosorbent assay (ELISA) kit that can detect HCV antigen and antibody
simultaneously.
Methods and results: We used 218 positive and 204 negative serosamples of
Taiwanese population to compare the sensitivity and specificity of our HCV Ag-Ab
COMB kit and Monolisa HCV Ag/Ab kit. The result shows as table; the sensitivity
and specificity of our kit were 100% and 100%, which were better than Monolisa kit.
We also used seroconvertion panels and British Working Standard to evaluate and
compare the performance of our kit and Monolisa kit. The test HCV seroconversion
panels included BBI: PHV901, PHV906, PHV914, PHV917, PHV919, PHV920 and
BCP: 6214, 6215, 6227, total 9 panels. When our kit was compared to the Monolisa
kit, 6 of the 9 seroconversion panels tested gave the same results on both kits. For
2 of the 9 panels tested, our kit detected HCV infection earlier than Monolisa kit.
The analytical sensitivity of our kit (1:32 dilution) was higher than Monolisa kit (1:2
dilution) by using serial dilutions of the NIBSC British Working Standard for AntiHCV.
Conclusion: Although previous data shown the antigen detection limit of our kit was
still higher than compared one, our kit shown a lower antibody detection limit, a better
clinical sensitivity and a better specificity. Besides, the total reaction time only takes
105 minutes. We therefore provide a useful HCV infection screening choice.

Monolisa HCV HCV+

HCVAg/Ab kit
Total

Our HCV Ag-Ab COMB kit

HCV+
HCV215
0
3
204
218
204

Total
215
207
422

B-441
Integration of Microarray and qPCR system for quantitative and multiqualitative analysis of MTB, NTM and MDR

D. Lee, J. Kim, S. Park, S. Ku, M. Peak. K-MAC, Daejeon, Korea, Republic

of,
Background: qPCR and DNA Chip analysis have been widely used for molecular
diagnosis. We developed a novel platform for molecular diagnosis, called Ampli &
Array system, which integrates the qPCR and hybridization in a DNA microarray. In
this system, entire reaction of qPCR and microarray are carried out in same container.
So, quantitative and multi-qualitative analysis can be achieved at the same time. In
this study, the Ampli & Array system was applied to the detection of Mycobacterium
tuberculosis (MTB) to verify the usefulness in diagnosis.
Methods: Probes identifying mutation genes and genotyping of MTB/NTM are
immobilized on the surface of Ampli & Array platform. We can recognize clinically
important target genes, rpoB, katG and inhA that are related with rifampin and

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Technology/Design Development

Wednesday, July 30, 9:30 am 5:00 pm

isoniazid by using these probes. Amplification of the target DNA was carried out in a
normal qPCR machine. After qPCR step, hybridization was performed sequentially by
simply changing temperature at the qPCR machine.
Results: We measured 10 copies of MTB in qPCR process, and detected NTM
genotypes and mutation genes of rpoB, katG and inhA specifically. In hybridization,
we screened 94% of NTM by detecting 5 types of NTM. And, we also classified the
mutation related to MDR. In rifampin related genes, we have screened L511P, D516V,
D516Y, H526Y, H526D and S531L. Also, we have detected 2 types of mutation in
isoniazid related genes.
Conclusion: With the Ampli & Array system, we could quantitatively analyze various
genotypes in samples of large number at once. Also, more accurate results could be
provided with more economical manner. The system encouraged us to overcome the
limitations of current molecular diagnosis/DNA analysis in revolutionary way. For
this reason, we expect that the Ampli &Array system will replace the existing qPCR
and microarray in molecular diagnosis market.

Results: On melting curve analysis, the 100- and 341-bp-amplicons obtained from
crude mouse tissue lysates gave distinct peaks at 81C and 84C, respectively.
Additionally, all of the variants tested (WT/WT, WT/IN and IN/IN) were
distinguishable from each other. In the SNP assay, A/A, A/G, and G/G variants were
detected in the crude specimens. Peaks from 45- and 57-bp-amplicons were detected
by melting curve analysis at 73C and 80C, respectively. However, when tested, the
conventional qPCR system failed to amplify any of the target DNA molecules. We
found that to obtain separation of distinct peaks on melting curve analysis required
melting temperature differences of at least 3C between the amplicons, and amplicon
sizes of 45 bp to 1 kb.
Conclusions: KOD-SYBR is a robust system for polymorphism analysis of crude
clinical samples. The flexibility of being able to use relatively large amplicon sizes
(<2 kb) for melting curve analysis is a distinct advantage of this system. KOD-SYBR
has potential to be a powerful tool for high-throughput genotyping of clinical samples.

B-444
Highly sensitive quantification of oxalate in plasma and dialysate fluid by ion
chromatography.

N. Voskoboev, T. S. Larson, J. C. Lieske. Mayo Clinic, Rochester, MN,

Background: Oxalate is an end product of glyoxalate and glycerate metabolism that is
excreted in the urine. This dicarboxylic acid is a common component of up to 75% of
kidney stones. In addition to kidney stones, more extreme hyperoxaluria due to either
genetic (primary hyperoxaluria) or acquired/secondary causes (enteric hyperoxaluria)
can cause nephrocalcinosis and/or renal failure. If patients with primary (or sometimes
enteric) hyperoxaluria develop renal failure, plasma oxalate levels and removal rates
in dialysate need to be monitored closely to prevent oxalosis and/or rapid loss of
a transplanted kidney. Previously we have validated enzymatic assays of oxalate in
plasma and dialysate. Use of ion chromatography to quantify oxalate has the potential
advantages of improved precision, automation, and integration with the laboratory
information system.
Methods: A Dionex ion chromatography system and IonPac column were used to
measure oxalate. Dionex ion chromatography system was modified to accommodate
a Boric Acid eluent. Waste dialysate and plasma samples from patients with and
without hyperoxaluric diseases were obtained for clinical and analytic validation. All
samples were acidified to a pH 2.5 - 3.0 within 1 hour of collection, previously shown
necessary for oxalate stability. Accuracy was assessed via oxalate spike recovery and
comparison to our laboratorys current enzymatic oxalate oxidase method that is based
on the Trinity Biotech oxalate kit (Trinity Biotech plc, Bray, Co. Wicklow, Ireland).

B-442
A Novel and Robust Real-time PCR System for DNA Polymorphism Analysis
Directly from Crude Clinical Samples

T. Kobayashi, H. Matsumoto, T. Kuroita. Toyobo Co., Ltd., Osaka, Japan,

Background: Real-time quantitative PCR (qPCR), a powerful technology utilized
in many scientific disciplines, is also used for analysis of clinical samples. However,
conventional qPCR systems using Taq DNA polymerase suffer from the following
problems: (1) target size limitations (<200 bp), (2) poor amplification of GC-rich
targets, and (3) PCR inhibition from the biogenic substances present in clinical
samples. To circumvent these problems, we developed a novel qPCR system (KODSYBR) that uses KOD exo(-) DNA polymerase and SYBR Green I dye. Thus far,
KOD-SYBR has been used to amplify large targets (<2 kb), GC-rich targets, and
targets from crude clinical samples (whole blood, tissue lysates, hair roots, Gramnegative and Gram-positive microorganisms).
Objective: To determine the effectiveness of the KOD-SYBR system for DNA
polymorphism analysis of crude clinical samples.
Methods: First, we evaluated the KOD-SYBR system using amplified fragment
length polymorphism (AFLP) PCR with melting curve analysis at the end point.
We used three primers to generate fragments of two different sizes [wild type (WT)
allele: 100 bp; Inserted allele (IN): 341 bp]. Mouse tissue lysates prepared by a rapid
method were used as templates. Next, single nucleotide polymorphisms (SNPs) in the
alcohol dehydrogenase gene were detected using allele-specific primers containing
3 end mismatched bases followed by melting curve analysis. Diluted whole blood
and oral mucosa specimens were used. A G-specific primer-bearing tail sequence at
the 5 end was used to obtain a larger fragment (57 bp) than that of the other primer
set (an A-specific primer and a common primer), which had no tail sequence (45 bp).
The G-specific and common primer set detects an A>G SNP that leads to a missense
mutation (E487K).

Results: During initial development it was demonstrated that oxalate levels increased
in plasma and dialysate samples, even after acidification. This is consistent with
previous observations that oxalate levels can increase in samples exposed to basic
conditions, perhaps due to conversion from ascorbic acid. Thus the Dionex ion
chromatography system and IonPac column were modified to accommodate a Boric
Acid eluent. Using this set-up, plasma and dialysate oxalate measurements were
linear over the range of 1- 50 mcmol/L. Intrassay precision was acceptable (10%) at
1 mcmol/L, and improved to <2 % for values above 10 mcmol/L. Average recovery
with serial dilutions was 102%. Oxalate was stable refrigerated or frozen (-20C or
-80C) when plasma or dialysate was acidified (pH 2.5), but oxalate levels variably
increased in samples that were stored refrigerated or frozen but unacidified. Results
for 48 plasma samples across the normal and abnormal range compared well with the
current assay with a mean difference of only 2% between the two methods.
Conclusion: Ion chromatography using borate buffer can be used to reproducibly
quantitate oxalate in plasma and hemodialysate fluid. Use of common methods based
on aqueous and carbonate eluents are not acceptable for highly sensitive quantification
of oxalate. Results of a new chromatographic method compare well with a previous
enzymatic oxalate oxidase method.

B-445
Evaluation of the Sebia Capillarys 2 Flex Piercing hemoglobin A1c (HbA1c) assay

Z. Zhao1, J. Basilio1, S. Hanson2, A. Tennill2, R. Little2, D. Sacks1. 1National

Institutes of Health, Bethesda, MD, 2University of Missouri, Kansas City,
MO,
Background: Hemoglobin A1c (HbA1c) is used to monitor long-term glycemic
control in patients with diabetes, guide therapy, predict the risk of microvascular
complications, and diagnose diabetes. It is vital that methods to measure HbA1c be
accurate, precise, reliable and subject to minimal interference. Recently, Sebia (Lisse,
France) developed an automated liquid-flow capillary electrophoresis method to

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S261

Technology/Design Development

Wednesday, July 30, 9:30 am 5:00 pm

measure HbA1c on the CAPILLARYS 2 Flex Piercing instrument. We evaluated the
performance of the Capillarys for the measurement of HbA1c.
Methods:Technical evaluation was performed in two clinical centers (center 1, a
NGSP Secondary Reference Laboratory (SRL9) and center 2, a clinical laboratory).
The instruments and reagents were provided by Sebia and used according to the
manufacturers procedures. Linearity, carryover, and interference studies were
performed on fresh or fresh frozen samples collected in each center. Blood samples
were shared between the centers for precision and accuracy studies. Precision
evaluation was performed following CLSI EP-5 using pooled whole blood dailyuse aliquots frozen at -70oC at four HbA1c levels (4.7%, 6.3%, 7.4% and 11.1%) and
compared to National Academy of Clinical Biochemistry (NACB) recommendations.
Accuracy was assessed using 100 single-donor patient samples with < 16 samples
analyzed in duplicate each day. HbA1c values obtained with the Capillarys were
compared to those obtained with two other analyzers used routinely at the two
centers: NGSP SRL9, G8 (Tosoh Bioscience, South San Francisco, CA) in center 1
and D-10 (Bio-Rad Laboratories, Hercules, CA) in center 2. For interference studies,
a difference of 0.2% HbA1c was used as an acceptable limit.
Results: The Capillarys was linear for HbA1c results from 4.2 to 17.6%. There was no
carryover when samples with HbA1c of 4.7 and 14% were analyzed alternately. There
was no interference from labile HbA1c as high as 11%, carbamylated hemoglobin
(blood samples incubated with 0.15 to 1 mmol/L of KCNO at 37C for 3 hours), or
triglycerides (93 to 4666 mg/dL). No additional peaks (i.e. for labile, carbamylated,
etc) were observed on the Capillarys electropherograms. Total CVs were < 2.0% in
both centers. HbA1c values obtained with the Capillarys and comparison methods were
well correlated, with minimal bias. Linear regression analysis yield the following: y
(HbA1c Capillarys center 1)= 1.033 (HbA1c Tosoh G8)-0.34, r=0.997, Syx=0.16; y
(HbA1c Capillarys center 2)= 1.084 (HbA1c Bio-Rad D10)-0.67, r=0.995, Syx=0.22.
Results of the Capillarys were comparable between centers with mean bias < 0.1%
HbA1c (y (HbA1c Capillarys center 1) = 0.982(HbA1c Capillarys center 2) + 0.05,
r=0.996, Syx=0.18). Ninety-five % (center 1) and 97 % (center 2) of single Capillarys
results were within 6% of the SRL9 mean (NGSP manufacturer certification criteria:
> 92.5% of results within 6%).
Conclusion:The analytical performance of Capillarys HbA1c is within NACB and
NGSP recommendations, has minimal interference, and is suitable for clinical
application.

B-446
Homogeneous high-sensitivity CRP assay on MagArray biosensors

H. Yu1, L. Carbonell1, M. Momiyama2, M. Murakami2. 1MagArray Inc,

Sunnyvale, CA, 2IMRA America, Inc, Ann Arbor, MI,
Background: MagArray platform is based on the detection of magnetic particles as
labels in bioassays. In contrast to systems based on optical signals, magnetic signals
are not affected by the common optical interference in complex matrices. In addition,
since the biosensors are designed to detect magnetic particles only when particles
are bound or captured to the sensor surface, this proximity detection mechanism
allows the possibility of homogeneous immunoassays. We report here that we have
developed a one-pot homogeneous assay for CRP with high sensitivity. We wish to
demonstrate MagArray platform is well suited for homogeneous assays that require
both simplicity and sensitivity.
Methods: Antibody pairs for CRPassay were screened and selected on MagArray
platform, and the detection antibody is conjugated to magnetic particles for both
capping CRP and generating signals. The assay consists of two simple steps of mixing
the magnetic particles with serum sample and addition of the mixture to magnetic
sensors. Signals can be read in as short as 2 min for a result, and more accurate results
can be obtained after 5 min. The whole process requires no shaking and rinsing or
other separation steps. Standard curves of CRP in both pure buffer and sera were
established and compared.
Results: The detection sensitivity of CRP in serum on MagArray platform is less than
1 mg/L for a homogeneous assay. The CVs for lower concentrations are less than
10% and less than 5% for medium and high concentrations. No prozone effect was
observed for CRP concentrations of up to 200 mg/L. Our preliminary tests showed no
interference effects (<= 10%) from lipids, HAMA and Rheumatoid Factor.
Conclusion: MagArray platform provides a unique opportunity of detecting proteins
in a simple and homogeneous fashion. Since this assay only involves mixing and

S262

adding reagents to the biosensors, the complexity of the assay format is greatly
reduced. The detection of magnetic signals from magnetic particles in proximity is a
key to applying MagArray platform for a homogeneous immunoassay.

B-447
Modifying tosyl-activated magnetic bead coating protocols to improve
biomolecule binding efficiency

F. Sultan1, N. Gasilova2, H. H. Girault2. 1Merck Chimie SAS, 94126

Fontenay sous Bois Cedex, France, 2Laboratoire dElectrochimie Physique
et Analytique, Ecole Polytechnique Fdrale de Lausanne, Lausanne,
Switzerland,
Background Tosyl-activated magnetic beads are widely used as a solid phase in
immunoassays and biomagnetic separations. With the right conjugation protocol any
biomolecule whether it be antibody, protein, peptide or glycoprotein containing NH2or SH- groups can be covalently coupled to the surface of tosyl-beads. Optimising
the coupling protocol is key to maximizing the binding efficiency of biomolecules
to beads. Using immunoaffinity capillary electrophoresis (IACE) with UV-detection,
bovine -lactoglobulin B levels bound to magnetic beads (MBs) following different
antibody coupling protocols were measured. Results were further verified using a
bicinchoninic acid (BCA) protein test (BCA kit, Pierce, USA) to estimate the amount
of antibody bound to the MBs. A coupling protocol has been developed which
shows higher binding efficiecy for the same concentration of beads when compared
with a published protocol (herein referred to as Protocols E and A respectively).
Moreover, with Protocol E similar levels of antibody loading can be achieved by
incubating at 37oC for 6h rather than 12h at room temperature.
Method Protocol A: 2l of rinsed tosyl-activated MBs were mixed with 8 l of
coating buffer (100mM sodium borate, pH9.5), 8 l of 3M ammonium sulphate
and 8 l of antibodies (5mg/ml). The mixture was incubated for 24h at 37oC under
continuous stirring to avoid sedimentation. After incubation, beads were rinsed with
10mM PBS and stored in PBS containing 0.025% Tween-20 and 0.02% sodium azide.
Protocol E: 4l of rinsed tosyl-activated MBs were mixed with 280l of coating
buffer, 166 l of 3M ammonium sulphate and 54l of antibodies (1.48mg/ml). The
mixture was incubated overnight (~12h) at room temperature under continuous
stirring to avoid sedimentation. Beads were subsequently incubated with a blocking
buffer for 1h at room temperature under continuous stirring. Finally, beads were
rinsed with washing buffer and stored in PBS containing 0.025% Tween-20 and
0.02% sodium azide.
Results Under the same conditions of IACE analysis, electropherograms showed
larger peaks of antigen -lactoglobulin B (5g/ml) in the case of MBs coated using
Protocol E (area of 0.328 (0.009) units) compared with Protocol A (area of
0.155 (0.006) units). Using the BCA test to quantify the amount of bound anti-lactoglobulin B antibody, it was demonstrated that 42g of antibody was bound
per 1mg of MBs using Protocol E compared with 24g for Protocol A. A further
enhancement of Protocol E was made to decrease the antibody incubation time with
the MBs. Comparative peaks were observed from the IACE electropherograms for
-lactoglobulin B levels (50g/ml) following 12h incubation at room temperature and
6h incubation at 37oC (areas of 2.28 (0.06) units and 2.14 (0.05) units respectively).
Conclusion From the IACE detection of -lactoglobulin B levels and BCA analysis of
bound anti--lactoglobulin B antibodies, it can be concluded that Protocol E provides
a higher antibody coating efficiency of MBs than Protocol A. Moreover, the coating
efficiency is retained at 37oC allowing for an overall shortening of experimental time.

B-448
Performance Evaluation of a Protein-Free Reagent to Design Out Common
Interferences in Diagnostic Assays

D. Pauly, T. Jentz, K. Pauly, J. Hein. SurModics, Eden Prairie, MN,

Background: Protein stabilizers, especially those containing bovine serum albumin
(BSA), can lead to increased non-specific binding due to protein interference.
Additional BSA-related issues include: anti-BSA antibodies, heterophilic false
positives, cross reactivity, reliable sourcing, lot-to-lot variability. The ability to
stabilize proteins in-solution, without the use of BSA or alternative proteins, would
be an invaluable way to eliminate these protein-related interferences and issues.
Objectives: To demonstrate the ability to stabilize proteins in solution using a
synthetic protein stabilizing technology versus commercially available proteincontaining stabilizers.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

Technology/Design Development

Wednesday, July 30, 9:30 am 5:00 pm

Methods: We tested 4 stabilizers: A. commercial BSA stabilizer, B. protein-free

stabilizer, C. commercial alternative protein stabilizer (non-BSA), D. 1%BSA in PBS
(negative control). All stabilizers were evaluated in terms of non-specific binding and
retained activity of the target antibody. Data are derived from stability studies utilizing
anti-rabbit antibodies from different host species. Anti-rabbit IgG-HRP conjugated
antibodies (hosts of origin including: guinea pig, sheep, chicken and mouse) were
diluted into the different stabilizing buffers (22 - 120 ng/mL) and equally divided for
storage at 4C and 37C. The test solutions of each stabilizer were evaluated in an
ELISA. At each time point a percent retained activity was determined by comparing
the activity of the aged conjugate (37C) to that of the control conjugate (4C).
Results: After 127 days at 37C the A. commercial BSA stabilizer, B. protein-free
stabilizer, and C. commercial alternative protein stabilizer (non-BSA) demonstrated
averaged retained activities of 79, 77 and 69% respectively across all species. Based
on Arrhenius projections, the stability of the conjugated antibody is equivalent to 3.4
years at 4C. Conclusion:These data represent an effective protein-free approach
to reduce non-specific binding and still perform equal to or better than proteincontaining stabilizing reagents. When protein-related interferences arise, the proteinfree stabilizer provides a valuable alternative for assay developers.
% Retained Activity after 127 days
Stabilizer
Guinea Pig
Sheep
Chicken
Mouse
Mean
A
74.9%
77.4%
72.9%
90.9%
79.0%
B
68.7%
75.9%
73.8%
89.5%
77.0%
C
60.1%
66.3%
66.1%
85.4%
69.5%
D
0
0
0
0
0
% Retained Activity = (Abs OD of 37C sample / Abs OD of 4C Sample) x 100
* Data points were averaged from 4 ELISA wells (n = 4)
** %CVs for each data point were less than 10%

B-452
Direct Single Nucleotide Polymorphism Genotyping from Blood, Plasma or
Serum

C. Yeh, J. Spurgin, J. Athanasuleas, K. Bradley, C. Willmon, J. Chimera.

Atherotech Diagnostics Lab, Birmingham, AL,
Background and objectives: Real-time PCR is a powerful, sensitive and favorable
method for large-scale single nucleotide polymorphism (SNP) genotyping. However,
prior purification of genomic DNA from blood is necessary since PCR inhibitors and
quenching of fluorophores from blood prevent efficient amplification and subsequent
detection of PCR products. We have developed a simplified direct-on-sample SNP
genotyping without prior DNA enrichment/isolation steps to further throughput and
reduce time, cost and labor. This methodology can be applied in pharmacogenomic
analysis on various platforms e.g., fluorescent-, florescence resonance energy
transfer-, or electrochemical voltammetry-based detection.
Methods: Diluents and procedures designed to specifically overcome PCR inhibition
and quenching of fluorescence were evaluated by genotyping on 4 SNPs from Factor
II, Factor V, MTHFR genes, 3 CYP2C9, VKORC1 gene polymorphisms, and 7 SNPs
of PCSK9 gene. Paired DNA and blood, plasma or serum samples are collected and
analyzed for concordance using eSensor and ABI7900HT instruments.
Results: The performances of either DNA purified from blood or the same blood
without DNA purification were analyzed using GenMark eSensor technology on 4
different variants prevalent in Factor II, Factor V and MTHFR genes (Thrombophilia
Risk Panel), and 3 variants from CYP2C9 and VKORC1 genes (Warfarin Sensitivity
Panel). Genotyping of 7 SNPs in the PCSK9 gene was conducted by TaqMan/
ABI7900 platform. Overall, genotyping from purified DNA and the corresponding
blood showed concordance of 84.8% (N=66), 87.5% (N=24) and 100% (N=12) for
PCSK9, Thrombophilia and Warfarin panels, respectively.

B-449
A Novel Method for the Measurement of Serum Viscosity: Evaluation of the
Rheosense microVISC Viscometer

H. Faby, L. C. Rogers. Memorial Healthcare System, Hollywood, FL,

Background: Measurement of serum viscosity is used to evaluate Hyperviscosity
Syndrome (HVS), which is associated with plasma cell dyscrasias, myeloma,
connective tissue diseases, and other inflammatory conditions. Rapid treatment of HVS
is critical to ensure effective reduction of risk from serious complications. Increased
viscosity, caused by the excessive intravascular paraproteins, leads to impaired transit
of blood through the microcirculatory system. The vascular stasis and resultant
hypoperfusion can cause severe complications which include cardiopulmonary
symptoms such as shortness of breath, hypoxemia, acute respiratory failure, and
hypotension; neurological effects such as confusion / mental status changes; ocular
damage including dilation of the retinal veins and retinal hemorrhages; bleeding
from the mucous membranes; and renal failure. Measurement of serum viscosity is
essential for an accurate diagnosis, but traditional methods are labor intensive and
not amenable to STAT analysis. We have evaluated a new instrument, the Rheosense
microVISC, for the rapid analysis of serum viscosity.

To further validate the methodology to perform large-scale high-throughput

genotyping directly from blood, plasma or serum, paired DNA/blood (N=50), DNA/
plasma (N=30) or DNA/serum samples (N=20) were analyzed on 7 SNPs of PCSK9
gene using ABI TaqMan real-time PCR machine. High concordance was achieved and
resulted in sensitivity/specificity of 100%/90.9% for direct blood, 97.9%/66.7% for
direct plasma, and 94.5%/75.0% for direct serum genotyping.
Conclusions: The methodology described is simple and fast that allows accurate gene
polymorphism test directly from a drop of blood, plasma or serum. This method can
be applied to a broad range of clinical genetic tests with the advantages of immediate
sample testing, improving workflow, and lowering workload, cross-contamination,
costs and turnaround time.
Abbreviation: PCR, polymerase chain reaction; PCSK9, proprotein convertase
subtilisin/kexin type 9; MTHFR, methylenetetrahydrofolate reductase; VKORC1,
Vitamin K epoxide reductase complex subunit 1.

Methods: The microVISC (Rheosense, Inc.) is a small portable instrument which

uses VROC (Viscometer/Rheometer-on-a-Chip) technology. The VROC sensor
obtains a viscosity reading by measuring the pressure drop as a sample flows through
a flow channel. Pressure is measured at positions of increasing distance from the
inlet. The slope of the straight line in the plot of the pressure vs. sensor position
is proportional to the viscosity. We evaluated this new instrument for use in the
clinical laboratory. The performance evaluation in this study included within-in run
and between-run precision and linearity, using standards purchased from Rheosense,
Inc. Accuracy was determined by correlation of the microVISC (Rheosense, Inc.)
to a cone and plate viscometer using patient samples. We used serum samples from
patients with a normal comprehensive metabolic panel for verifying the reference
range. Statistical analyses were performed by EP Evaluator
Results: The within-run precision for normal and abnormal controls was < 1%. Mean/
standard deviation/ % CV for between-run precision was 1.54 cP/0.05 cP/3.25% and
4.12 cP/0.10 cP/2.43% for normal and abnormal controls respectively. The linear
range was verified for 0-6.38 cp. The microVISC correlated well with the cone and
plate method: y= 0.836x + 0.064 (r= 0.979). The small negative bias was reflected in
a slightly lower reference range of 1.10-1.60 cP for serum samples.
Conclusion: We validated the microVISC instrument for the rapid, accurate, and
reproducible measurement of serum viscosity in a clinical laboratory setting. The
instrument is small, portable, and easy to use and maintain.

CLINICAL CHEMISTRY, Vol. 60, No. 10, Supplement, 2014

S263

AUTHOR INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
A
Ababio, W.
A-291, A-320
Abbadi, A.
A-257
Abbud-Antaki, R. A.
A-067
Abdalla, E. M.
A-072
Abdella, N.
A-176, A-199,
A-214, A-215
Abdel Moneim, R.
A-098
Abdo, M. M.
A-072
Abou El Hassan, M.
A-270
Abreu, D. Costa
B-185
Abreu, F. Regina Marques A-324
Absah, I.
A-364
Abusoglu, S.
A-107, A-130,
A-135, A-391
Ackermans, M.
A-212
Adams, A. H.
B-184
Adedeji, O. O.
B-288
Adekiitan, M. E.
B-288
Adeli, K. B-259, B-260, B-262,
B-265, B-271, B-273, B-287
Adie, L.
A-024, A-049
Adly, H. M.
B-229
Agarwal, S.
A-129
Aghvanyan, A.
A-035
Agnoletti, R.
B-385
Ahmed, S. A. M.
A-072
Ahn, E.
A-405
Ahn, S.
A-044, A-060,
A-329, B-041, B-046, B-055,
B-080, B-195, B-207
Akbas, H.
A-112, A-361
Akbas, N.
B-020
Akgul, M.
A-040
Akiyama, D.
A-357
Akl, P.
A-263
Aksoy, T.
B-131
Akter, S.
A-230
Akuzawa, M.
A-339
Akyurek, F. Tuncez
A-107
Akyurek, F.
A-107, A-130,
A-135, A-391
Alajlan, A.
B-298
Alstico, V. Teixeira
A-242,
A-330
Alay, E.
A-113
Albahra, S.
A-240
Albeiroti, S.
B-176, B-366
Alberelli, A.
B-385
Albuquerque, R. Lima
A-080
Alcantara Tito, S. R.
A-118
ALENCAR, D. O.
B-197
Alfieri, A.
A-242
Algeciras-Schimnich, A. A-048,
A-056, A-122,
A-226, B-020, B-214
Alhassan Aneyire, S.
A-084
Al Hussain, N.
B-392
Ali, L.
A-230
Al-Inany, H. Gaber
A-119
Aliyeva, O.
B-044
Aljasser, S.
B-298
AlKhaldi, R. M. A-214, A-215
Allison, J. J.
B-345
Almeida, A.
A-357
Almeida, C.
B-220
Almeida, I. Caroline Brito
A-330
Almeida, L. R.
A-244, B-045,
B-048, B-144
Almeida, N. B. C. S.
B-143
Almeida, V. O.
B-212

S264

Al Mulla, F.
A-214, A-215
Alonso, J.
A-293, A-335
Alotaibi, N.
B-298
Al-Othaim, A.
B-392
Alsadhan, A.
B-298
Alsaeoti, S. Omar
A-010
Altawallbeh, G.
A-412
Altinok, D.
A-313
Altraif, I.
B-392
Alvarez, B.
A-293, A-335
Alvarez-Gomez, S.
B-196
Alvi, A. J.
A-343
Amankwah, S.
A-291
Amariko, I.
A-055
Amentas, G.
B-333
Ames, D.
A-384
Amin, P.
A-213
Amin, S.
A-373
Amoah, M.
A-144
Andag, R.
A-360
Anderson, D. A. A-096, A-280
Anderson, V.
B-427
Anderson-Berry, A.
B-234
Ando, Y.
A-339
Andrade, L. E. C. A-115, A-124
Andrade, L. Eduardo Coelho
A-127
Andrade-Olivie, A.
B-134
Andrade-Olivi, M. A.
B-203
Andreguetto, B. D.
B-033
Andres, S. A.
A-042
Anelli, M. Chiara
A-045
Angeles, M.
B-163
ngelo, P. C.
A-289
Ansah, R.
A-291, A-342
Antwi, K.
A-407
Aoki, Y.
A-204, A-207, A-233
Aoyagi, K.
A-187
Apaydin, A. Haydar
A-433
Apple, F. S.
B-317, B-327,
B-328, B-349, B-352, B-357
Arajo, A. L. T.
B-143
Arajo, C. M.
A-166
Araujo, P. B. M. C.
A-177
Arce-Matute, F.
A-017, A-202
Argolo, S. V. L.
A-208, B-023
Ariza-Ariza, R.
A-369
Armbruster, D.
B-160, B-260,
B-273
Armbruster, F. Paul
B-329
Arnold, P.
B-300
Arppe, R.
A-123
Arroyo Huanaco, O. M. A-118
Arthuis-Demoulin, P.
B-373
Arzuhal, A. Ercan A-186, A-216
Asami, M.
A-346
Ascah-Ross, T.
A-377
Asencio, A.
A-211
Asensio, J.
B-134
Aslan, B.
A-134
Assini, A. G.
B-188
Astarita, G.
A-379
Astion, M. L.
B-168
Atalay, C. Resat
A-120
Athanasuleas, J.
B-452
Attems, J.
B-379
Auger, S.
B-429
Augusto Gmez-Ros, G. A-376
Aurand, C.
A-377
Avello, T.
B-087
vila Garca, G.
B-004
Avolio, M.
B-086
Aw, T. C. A-020, B-338, B-361
Ayasoufi, K.
B-366

Aydin, F. A.
Aydin, H. I.
Ayub, S.
Azinge, E. C.
Azzara, F.

A-383
B-380
A-328
B-306
B-311

B
Babeluk, R.
B-379
Babic, N.
B-249
Babini, S.
B-385
Baburina, I.
A-238, B-418
Bacani, J.
B-404
Bach, P.
A-241
Bachman, J.
B-189
Baek, S.
B-250
Baethies, H.
B-365
Bai, Y.
B-208
Bailey, D.
A-221
Baird, G.
B-410
Baker, J.
B-011
Bakker, J. A.
B-369
Baldasano, C. N.
A-238
Baldo, D. C.
A-115
Balherini, R.
B-001
Ballantyne, C.
A-148, B-351
Ballard, K.
A-332
Balsanek, J. G.
A-374
Bamberg, A.
B-356
Ban, M.
B-181
Bandeira, T. Pinheiro Gomes
B-057
Bandeira, V. Pinheiro Gomes
B-057
Banerjee, L.
A-077
Banerjee, S.
A-189
Bannerman- Williams, P. A-320
Banning, P.
B-130
Bao, P.
B-308
Barbera, J. L.
B-134
Barbosa, A. L.
B-263
Barbosa, A. M.
A-430
Barbosa, N.
A-321
Barbosa, T. M. C. C.
A-382
Barclay, S.
B-021
Barden, S.
B-039
Barker, T.
A-261
Barksdale, B.
B-133
Barnett, J.
A-143
Barnidge, D. R. A-066, A-374,
A-404
Barra, G. B.
B-200, B-270
Barra, G. Barcelos
B-228
Barreto, J. A.
A-115
Barth, J. H.
B-358
Bartlett, W. A.
A-259
Bashir, R.
B-434
Basilio, J.
B-445
Bass, D.
A-420
Basso, R. C.
A-152
Basturk, T.
A-134
Bates, P. J.
A-403
Batista, J. Paulo Gervasio A-258
Batista, M. C.
A-205
Bauer, G.
B-434
Bauer, T. W.
A-338
Baum, H.
B-296
Baumann, N.
B-034
Baumann, N. A. A-279, B-020,
B-267
Baumgartner, R.
B-379
Bautista, D.
A-180
Baxter, S.
A-294, A-432
Baykal, T.
A-316

Baykan, O.

A-040,
A-236, A-313
Baz, M. J.
B-087
Beamon, C.
B-261
Beaulieu, M.
B-378
Beck, J.
B-215
Beckham, D.
A-143
Bedoya, A. Maria
B-148
Bedzyk, W.
B-245
Beeks, O.
B-312
Bell, D.
A-377
Belley-Montfort, L.
B-071
Belmonte, P. L. M.
B-197
Beloch, H. Abdul
B-326
Beloch, H.
A-010
Ben-Asher, H.
B-292, B-293
Benavides, G.
B-438
Benchikh, M.
B-321, B-427
Bendet, I.
A-079
Bengtsson, H.-I.
B-026
Ben-Yosef, Y.
B-292, B-293
Berenson, J. R.
A-358
Bereznikova, A.
B-355
Berger, P.
A-064
Bergeron, C.
B-378
Berlanga, O.
A-024, A-049
Bermudez, I.
A-029
Bermudo Guitarte, C.
A-028,
A-047
Bernasconi, R.
A-146
Berntsen, M. S.
B-027
Bertelson, D.
A-427
Bethoney, C.
B-276
Bevilacqua, V.
B-259, B-260,
B-262, B-265, B-273
Bevilaqua, V.
B-271
Bhatt, D. L.
B-348
Bhavsar, N.
B-250
Bhullar, B.
A-325
Bi, C.
A-264
Bianchessi, M.
B-435
Bianco, R.
A-147
Biavatti, M.
A-435
Biavatti, M. W.
A-430
Bidart, J.-.M.
A-064
Bidwell, H.
B-311
Bilello, L.
A-184
Bini, R.
B-061
Birsan, A.
A-394, B-429
Bispo, A. C.
A-177
Bjornson, L. K.
A-184
Blair, H.
A-062
Blanco Barca, O.
B-203
Blanco Prez, M.
B-203
Blankenstein, M.
A-212
Blick, K. E.
A-263
Block, D.
B-034
Block, D. R.
A-279, B-020
Bock, J. H.
A-438
Boelstler, A. M.
B-149
Boerwinkle, E.
B-351
Bohula May, E. A.
B-360
Boisen, M.
B-068
Boisen, M. L.
A-143
Bojko, B.
A-376
Bonaca, M.
B-360
Bonavida, B.
A-358
Bookhout, C.
B-261
Boone, L.
B-312
Boonyod, D.
B-043
Boraski, M.
B-250
Borges, K.
A-007
Born, K.
A-240
Bosi, C. F.
A-435

AUTHOR INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Bosley, S.
Bosmans, J.
Bostom, A. G.
Boston, B.
Botchway, F. A.

A-248
B-324
A-116
B-013
A-291, A-320,
A-342
Botelho, A. C. C. S.
B-183
Botelho, J. C.
A-220
Bottini, P. V.
B-033
Bottrel, R. L. A.
B-188
Boudry, P.
B-330
Bourcy, K.
A-041
Boyd, J. M.
B-031, B-404
Boyd, L.
B-277
Boyle, S.
B-029
Boysen, J.
B-034
Bozkurt, B.
B-351
Bradic, Z.
A-340
Bradley, K.
B-452
Braga, F.
A-259
Bragina, V. A.
B-279
Braiteh, F.
B-418
Bramlett, K.
A-042
Branco, D. Colares Castelo
B-057
Branco, L.
B-068
Bransky, A.
B-292, B-293
Brass, D.
A-103, A-272
Brault, D.
A-260
Breaud, A.
B-393
Briggs, C.
B-225
Brilhante, R. Samia Nogueira
B-057
Brimhall, B.
B-152, B-173
Brinson, A.
B-163
Brocco, G.
A-275
Brochu, V.
B-071
Brockbank, S.
A-222, A-347
Brockmller, J.
A-360
Brooks, Z. C.
B-158
Broome, H.
B-026
Brown, B. L.
A-143
Brown, D.
B-267
Brown, K. B.
B-171
Brunel, V.
A-030, A-031
Bruno, J. J.
A-331
Bruton, M.
B-398
Bcker, D. H.
B-063
Budd, J. R.
B-020
Buffone, G.
A-129
Bulut, E.
A-186
Bunch, D. R.
A-105, A-378,
A-412, A-425, B-230
Bunk, D. M.
B-358
Buno, A.
A-250
Burcham, J. L.
A-117
Burgyan, M.
B-230, B-396
Burkhart, B.
B-421
Burnside, Z.
A-432
Bush, R.
B-009
Bustos, S. B-262, B-265, B-273
Butterfield, J. H.
A-402
Bttler, R.
A-212
Buxeda Figuerola, M.
A-073
Buyukterzi, Z.
A-130, A-135

C
Caball Martn, I.
Cabral, M.
Cabral, W.
Cahalan, S.
Cai, Z.
Caironi, P.

A-073
A-181
A-007
A-425
A-100
A-146

Caixeta, M. S. S. Borges B-228

Calton, L.
A-415
Calton, L. J.
A-419
Calvo, M. D.
B-087
Camara, J. E.
B-247
Camilo, A. L.
A-159, A-173,
A-175, A-196, A-209, B-177
Camilo, A. L. N.
A-163
Camilo, A. Lucia Nascimento
A-185
Camilo, A. Nascimento
B-005
Camoia, J.
B-124
Campagnoli, M. P.
A-155
Campana, P. G.
B-061
Campanholi, L. Bendinelli
A-368
Campbell, J.
A-294, A-432,
A-023, A-059
Campbell, T. G.
B-012
Camporese, A.
B-086
Campos, C. B.
B-191, B-193
Campos, L.
A-321
Caavate-Solano, C.
A-017,
A-202
Cantin, D.
B-071
Cao, J.
B-117
Cao, Z.
A-424
Capell, N.
B-330
Capparelli, A. L.
A-331
Caputo, M.
A-161
Carbonell, L.
B-054, B-446
Carden, S.
B-339
Cardoso, G. P.
B-248
Cardoso, Z. R. M.
A-286
Carlow, D.
A-421
Carmines, P. K. A-204, A-207,
A-233
Carobene, A.
A-259
Carpenter, M. A.
A-116
Carrillo, A.
B-154, B-164
Carrol, E. D.
B-286
Carroll, F.
A-396
Carr-Smith, H.
A-024, A-049
Carvalhares, F. B. F.
A-282
Carvalho, D. Celeste Bezerra de
A-353, A-368
Carvalho, L. G. S.
B-067
Carvalho, T. Nogueira de A-353,
A-368
Casbarien, O.
B-221
Castello, R.
A-161
Castelo, M. G.
B-057
Castro, A. R.
B-038
Castro, L. L.
B-063
Cava Valenciano, F.
A-293,
A-335
Cavender, M. A.
B-348
Celebi, G.
A-101
Cembrowski, A. R.
B-317,
B-362
Cembrowski, G. S.
A-252,
B-317, B-362, B-405
Cepova, J.
B-223, B-235
Cereda, M.
B-435
Cervo, A. M.
B-067
Cetinkaya, R.
A-126
Cevik, O.
A-316
Chacko, A.
A-061,
A-222, B-321
Chacko, B. K.
B-438
Chacon, C. Lopez
B-035
Chaffin, E.
B-402, B-425
Chakraborty, J.
A-077
Chakraborty, S.
A-038

Chaler, E. A.
A-245
Chambers, A. E.
A-189
Chambers, E.
A-423
Chambers, E. E.
A-401
Chan, M.
B-259, B-260,
B-262, B-265, B-273
Chan, N.
A-409
Chan, Y.
B-265
Chanakarnjanachai, S.
B-034
Chang, K.-C.
B-206
Chang, L.
B-012
Chang, W.
B-072
Chang, Y.-C.
B-189
Chapo, J.
A-296
Chapo, J. A.
B-079
Chaudhury, H. S.
A-230
Chebabo, A.
B-049
Chekuri, S.
B-152, B-173
Chen, F.
B-179
Chen, H.
A-358
Chen, M.
B-271, A-246,
A-247, B-175
Chen, Q.
A-006
Chen, T.
B-434
Chen, W.
A-034
Chen, W.-C.
B-206
Chen, X.
A-336
Chen, Y.-C.
A-001
Chen, Y.
B-259,
B-262, B-273
Cheng, C.-W.
B-434
Cheng, D.
B-003, B-016
Cheng, D. L.
A-108, A-137
Cheryk, L. A.
B-431
Chester, S. A.
B-394
Chianca, C. F.
B-270
Chiang, C.
B-202
Chimera, J.
B-452
Chincholkar, P. R.
B-030
Chindarkar, N.
A-392
Chiong, C. S. P.
B-030
Chiotis, D.
B-333
Chittamma, A.
B-237
Chiu, K.-Y.
B-206
Cho, C.
B-115
Cho, J.-E.
A-329, B-041
Cho, S.-N.
A-329, B-041
Cho, S.
A-285
Cho, Y.
B-110
Cho, Y.
B-240
Choi, E.-Y.
B-046,
B-055, B-080
Choi, J.
A-292
Choi, M.
A-224
Choi, R.
B-240
Choi, S.
A-145
Choi, Y.
B-195, B-207
Chollet, D.
A-427
Christensen, S.
A-213
Christenson, R.
A-167
Christenson, R. H.
A-117,
B-339, B-358
Christoulas, D.
A-111, A-151
Chrousos, G.
A-219
Chu, C.
A-390
Chu, D.
A-427
Chuang, C.-K.
A-375
Chuang, H.
B-312
Chun, S.
A-087, A-254
Chung, B.
B-059
Chung, J.
B-240
Church, S. S.
B-303
Cienfuegos, J.
A-042
Cigerli, S.
A-134

Cintra, A.
Citrano, P.
Claeys, M.
Claeyssens, S.
Claggett, B. L.
Clark, D. L.
Clark, D.
Clark, Z. D.
Clarke, W.

B-019
A-125
B-324
A-030, A-031
A-116
A-310
B-246
B-254
A-416,
B-393, B-424
Clements, B.
B-152
Clotilde, L. M.
B-054
Clunie, I.
B-345
Cobacho, A. Francisco
B-185
Cobbaert, C. M.
B-369
Cobbold, M.
A-059
Cocci, A.
B-435
Coddington, C. C.
B-267
Coelho, F. F.
A-007, B-212
Coelho, P. Z.
A-333
Coenen, D.
B-324
Cohen, A.
B-259
Cohen, A. H.
B-271
Cokun, A.
A-259
Colantonio, D.
B-259
Colantonio, D. A.
B-415
Colby, J. M.
B-408, B-432
Cole, J.
A-340
Collinson, P. O. B-335, B-346,
B-347, B-354
Coln-Franco, J. M.
A-096
Colon-Franco, J. M.
A-280
Colvin, R.
B-266
Conejo, L.
A-293, A-335
Connelly, M. A.
A-149
Conrad, M. J.
A-116, B-348
Conta, J. H.
B-168
Conti, N.
A-045
Cook, L.
A-090
Coresh, J. A-138, A-148, B-351
Cormier, J.
B-071
Cornaut, L.
A-180
Corradi, C.
A-007
Corradini, R.
A-045
Cosio, S. D. S.
B-205
Cosma, C.
A-147
Costa, C. Rossi da
A-368
Costa, J. Z.
B-161
Costa, S. S. S.
A-166, B-143,
B-228, B-263, B-270
Cote, L.
A-386, A-390
Counts, D.
A-117
Courjal, F.
A-273
Courtney, J. B.
A-238, B-418
Cousins, L.
A-387, A-397,
B-397, B-406
Cox, S.
B-154, B-164
Cremers, S.
A-203
Croal, B. L.
B-345
Croce, A.
A-260
Cronin, C.
A-309
Crutchfield, C.
A-416
Cruz, G. J.
B-070
Cruzan, C.
B-367
Cruz Filho, R. A.
B-248
Cugini, A.
A-045
Cui, M.
A-366
Cui, W.
A-006, A-036
Curley, M.
B-312

D
Dafterdar, R.
DAgostino, L. E.

B-298
A-331

S265

AUTHOR INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Daher, R. T.
A-367
Dai, J.
B-094
Dajak, M.
B-315
Dallas, J.
A-162
Dallman, M.
B-391
Daly, T. M.
A-338
Damiano, A.
B-255
Damio, C. B.
B-248
Danaceau, J.
A-423
Danese, E.
A-284
Danisman, N.
A-165
Danks, C.
B-286
Darley-Usmar, V.
B-438
da Rosa, J. S.
A-142, A-430
Darragh, J.
B-427
Das, B.
A-168
Dasgupta, A.
B-208
Da Silva, S. Cristina Ferreira
A-281, A-295
Datta, P.
B-094
Dauar, E. H. J. T. A-282, A-306
Daviau, J. S.
A-428
David, A.
B-082
Davis, G. J.
B-242
Davran, F.
A-112, A-361
Dayton, J.
A-183
Dean, S.
A-272
de Andrade, K. R.
A-142
Deans, K. A.
B-345
De Biase, I.
B-269
Debiasi, M.
A-027
de Castro, G. R. W.
A-142
Deegan, S.
A-347
DeFrance, A.
B-017
De Grande, L. A. C.
B-127
De Guire, V.
A-273
Deirmengian, C.
A-125
deKoning, L.
B-341
Del Amo, N.
A-293, A-335
de la Motte, C.
B-366
Delaney, S.
B-415
DeLemos, J.
B-351
Del Guidice, R.
B-175
Dellavance, A.
A-115, A-124,
A-127
Del Plato, A.
B-401
DelRosso, R.
B-431
Deluca, A.
A-067
De Martino, M. Cristina A-281,
A-295
Demerdash, H. M.
A-099
Demir, M.
B-102
Dempster, D.
A-203
Denadin, O. V. Porto
B-057
De Nardi, C.
A-414, B-401
Denardin, O. Porto
A-353,
B-064, B-185
Denardin, O. V. Porto
A-185
Denomme, G.
B-021
De Rosa, R.
B-086
Desemone, J.
A-401
Deustch, J.
B-038
Deutz, N. Ep
B-242
Devaraj, S.
A-129
De Vita, P.
B-385
de Winter, R. J.
B-314, B-344
Dhesy-Thind, S.
B-353
Dhoipode, S. J. A-010, B-101,
B-326
Diamond, S.
B-418
Dias, A. C.
B-263
Dias, V.
B-404
Dias Neto, O. F.
A-306
Dias Neto, O. F. D.
A-282

S266

Diaz, J.
B-134
Diaz-Garzon, J.
A-250
Dickerson, J. A. B-168, B-395
Dickson, D.
A-012, B-421
Dietrich, V.
B-255
Diez, E. A.
A-055
Dilworth, L. L.
A-109
Dimopoulos, M. A.
A-111,
A-151, A-308, B-333
Dinckan, A.
A-112
Diniz, M. E. R.
A-382
Dixon, R.
B-409
Dixon, R. B.
A-389
Dizpenzieri, A.
A-066
Djedovic, N. K.
A-422
Dobrescu, G.
A-035
Dogru, A.
B-131
Dogru, T.
A-101
Dohnal, J.
A-307
Domingues, A. Lucia C. A-127
Donaldson, J.
B-407, B-413,
B-417, B-423
Donato, L. J.
A-402
Dong, Z.
A-063
Donmez, L.
A-126
Doran, T.
A-380
Dorizzi, R. M.
B-385
Doshi, R.
B-047
Dotson, R.
B-163
Doty, A.
B-050
Doungwao, U.
B-283
Doyle, K.
A-241
Drayson, M.
A-023, A-059
Drinan, M.
B-246
Dschietzig, T. Bernd
B-329
DSouza, C.
B-292, B-293
Du, F.
A-358
DU, J.
A-431
Du, L.
A-050, A-063
Du, X.
B-179
Duarte, C.
B-434
Duarte, J.
A-177
Dubois, J.
B-287
Duda, M. P.
B-418
Dudek, T.
B-252
Dueas, J. L.
A-225, B-255
Duffy, D. C.
B-012
Duh, S.-H.
B-339
Duncan, J.
B-081
Duncan, L.
B-034
Duncan, R.
A-222
Dunlop, A. J.
B-345
Dupont, W. D.
A-096
Duque, M.
A-250
Duran, M. M.
B-372
Duretz, B.
A-414
Duro Milln, R.
A-047
Dworakowski, E.
A-203

E
Eckelkamp, L. L.
Edens, N. K.
Edwards, J. A.
Egbuagha, E. U.
Eggert, C.
Eghan, B. Ackon
Eira, V. B. A. S.
Ejilemele, A. A.
Eklund, E. E.
Ekpe, J.
El Ansary, M. M. S.
El Awamy, H.

A-363
B-242
B-431
B-306
A-278
B-103
B-197
A-410
A-171
B-316
A-119
A-010

Elbadawy, A. Hassan Bakr

A-119
Elbeg, S.
A-256
Eleutherakis-Papaiakovou, E.
A-111, A-151, A-308, B-333
El Gierari, E. M.
B-412
Elias, A.
A-174
Eliasson, P.
B-369
El-Khoury, J. M.
A-081,
A-105, A-257, A-378, B-176
Ellis, E.
A-427
El-Maouche, D.
A-200
Eloi-Santos, S. M.
B-183
Elshaari, F.
A-010
mond, J.-P.P.
B-378
Ence, A. T.
A-274
Endlich, W.
B-369
Enzweiler, T.
A-397
Epps, J.
B-300
Erali, M.
B-269
Ercin, C. Nuri
A-101
Erden, A. Ismail
A-216
Erden, G. A-113, A-120, A-165,
A-186, A-216
Erdinc, S.
A-165
Eren, N.
A-134
Erfurth, S.
A-399
Ergil, J.
A-216
Ergul, E.
A-040
Erickson, J. A.
B-386
Ersoy, F.
A-126
Ersoz, N.
A-433
Eskandari, M.
B-039
Esta, N.
B-139
Eve, E.
B-425
Ezgu, F.
B-380

F
Fabios, E.
Faby, H.
Facchin, B. M.
Fahse, M.
Faix, J. D.
Falc-Pegueroles, P.
Falcou-Briatte, R.
Fan, S.-L.
Fandio, J.
Fanizza, C.
Farell, M.
Faro, L. Brito
Farooqi, M. S.
Farrell, N.
Faulhaber, A. C. L.

B-221
B-449
A-435
A-190
B-412
B-004
B-082
B-291
B-159
A-146
B-071
A-324
A-248
B-281
A-205,
A-323
Faye, S. A.
B-354
Fehlberg, L. C.
B-049
Feitosa, M. S.
B-304
Feliu, M. S.
B-091, B-221
Feng, P.
A-172
Feng, W.
A-025
Fennell, S.
B-276
Feres, M. Cristina A-281, A-295
Feres, M. C.
B-061
Ferguson, M.
B-303
Fermer, C.
A-012
Fernandes, O.
A-004, A-258,
A-282, A-306, A-368, B-049
Fernndez, B.
B-181
Fernandez, I.
B-091
Fernandez-Calle, P.
A-250,
A-259
Fernandez Garcia, E.
B-257
Fernandez-Garcia, N.
B-139

Fernndez-Hermida, . B-004
Fernandez Lorenzo, J. R. B-203
Fernandez Rodriguez, E. B-257
Fernndez-Snchez, M. A-225
Ferraz, M. L.
A-124
Ferraz, M. Lucia
A-127
Ferreira, A. C. S.
A-289,
A-382, B-070, B-161, B-188,
B-191, B-193, B-212, B-227
Ferreira, C. E.
A-323
Ferreira, C. E. S.
A-205
Ferreira, P. R. S. A-205, A-323
Ferreira, R. Cuenca
B-128
Ferreira, S. Berlanga
B-001
Festus, O. O.
A-436
Fiabane, G.
B-373
Fiedler, G. Martin
B-008
Filho, J. B. L.
B-161
Finch Cruz, C.
A-332, B-238
Finn, P.
A-116
Fischer, J. C.
B-314, B-344
Fishburn, M.
B-012
Fitzgerald, R.
A-392
FitzGerald, S.
A-061, A-222,
A-294, A-347, A-432, B-321,
B-337, B-381, B-427
Fleisher, M.
A-092, A-421
Flint, J.
B-286
Flodin, M.
B-369
Flores Toledo, S. M.
A-118
Flumian, F. Bull A-353, A-368,
B-019
Flynn, E.
B-341
Foldvary-Schaefer, N.
B-398
Foley, D.
A-415
Fonseca, N.
B-212
Fonseca, S. F.
B-197, B-263
Fontes, R.
A-210
Fortova, M.
B-383
Fortuna, D.
A-231
Fortunato, A.
A-161
Fountain, K.
A-401, A-423
Fournier, D. R.
B-012
Fowler, E.
A-086
Frade, V. Vidotto
A-330
Frahm, J.
A-407
Frana, N. R.
A-115
Frangiamore, S. J.
A-338
Frank, E. L.
B-254
Franquelo, R.
B-139
Frazee, C.
B-414
Freeman, E. L.
A-350, A-351
Freeman, J.
A-167
Freeman, S.
B-238
Freire, M.
A-155, A-177
Freire, M. C.
A-079, B-248
Freire, M. Calafiori
A-210
Freire, M. D. C.
A-197
Freire, M. Di Calafiori
A-080,
A-324, B-019, B-024, B-192
Freitas, R.
A-027
Frescatore, R.
B-421
Frew, E.
B-012
Friedrich, M.
B-178
Frode, T. S.
A-142, A-430,
A-435
Frutos, M.
B-139
Fujimoto, L.
B-163
Fujita, N.
A-069, A-346
Fujita, Y.
B-109
Fukuoka, F. Salvador
B-005,
B-364
Fukuoka, F. S.
B-177
Fukuoka, S.
B-065

AUTHOR INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Fullah, M.
Furlan, R. L.
Furmaga, W. B.
Fusellier, A.
Fzry, A. K.

A-143
B-304
A-240
B-047, B-076
A-252, B-290,
B-362

G
Gabler, J. A-255, A-412, B-426
Gaburo, N.
A-080, A-330,
A-353, B-019
Gaburo Jr, N.
A-004, A-242,
A-258, A-368, B-064, B-185,
B-192, B-220
Gad, A. A.
A-119
Gafary, S.
A-117
Gaino, S.
A-271
Gairloch, E.
A-397
Gales, A. C.
B-049
Galhardo, L. Rocha
A-368
Gallego, C.
B-087
Gandhi, A.
A-095,
B-084, B-124
Garay, R. P.
A-055
Garber, C.
A-264
Garby, D. M.
B-431
Garcia, C. R.
A-229
Garcia, E. Aparecido
A-368
Garcia, E.
B-245
Garcia, M. Lopes
B-364
Garcia Alonso, S.
B-257
Garcia-Chico, P.
B-139
Garcia-Collia, M.
B-139
Garca de la Torre, A.
A-017,
A-202
Garca de Veas Silva, J. Luis
A-028, A-047, A-365, A-370
Garcia-Lario, J. V.
B-134
Garelnabi, M.
B-318, B-433
Garen, K.
A-221
Garg, U.
B-414
Garlipp, C. R.
B-033
Garofalo, P.
A-161
Garrison, K.
B-419
Garry, R.
B-068
Garry, R. F.
A-143
Gasilova, N.
B-447
Gaston, S.
B-020
Gattinoni, L.
A-146
Gaustad, A.
B-308
Gauthey Baraou, M.
A-253
Gavriatopoulou, M.
A-308
Gawel, S. H.
B-242
Gaze, D. C.
B-335, B-346,
B-347, B-354
Ge, J.
B-307
Geacintov, C. E.
B-252
Gearhart, M.
B-261
Geistanger, A.
A-102
Gelati, M.
A-271, A-284
Genc, H.
A-101
Genc, S.
A-383
Genta, V. M.
B-013, B-303
Genzen, J. R.
A-235, A-274
Georganopoulou, D.
B-308
Gerin, F.
A-040, A-236
Gerner, C.
B-379
Gerosa, G.
A-147
Gerz, R.
B-158
Ghadessi, M.
B-181
Giannitsis, E.
A-251, B-069,
B-309, B-331, B-343
Gibson, B.
A-338

Giesen, C. D.
A-276
Gilmore, K.
A-432
Ginis, Z. A-113, A-165, A-186
Giovanelli, R.
A-283
Giovanelli, R. P.
A-438
Girault, H. H.
B-447
Girtman, A. B.
A-404
Gkotzamanidou, M.
A-151
Glezer, E. N.
A-035
Glover, C.
B-421
Goba, A.
A-143
Godin, M.
A-030, A-031
Godoy, M. F.
B-221
Godwin, Z.
B-294,
B-295, B-299
Goepp, J. G.
A-086
Goh, M.
A-020
Gokce, M.
B-131
Gokulu, G.
A-313
Goldman, J.
B-006, B-015
Goldsmith, B.
B-316
Goldstein, D. E.
A-268
Gomes, D. Margareth Valente
A-210
Gomes, E. X.
A-357
Gomez-Bravo, M. A.
B-196
Gomez-Rioja, R.
A-250
Goncalves, M. S. B-188, B-212
Gonzlez, A.
A-293, A-335
Gonzlez, C.
A-335
Gonzalez, E.
A-073
Gonzalez-Redondo, J. M. B-134
Gonzlez Rodriguez, C. A-028,
A-047, A-365, A-370, B-004
Gonzlez-Rodrguez, C. A-052,
A-225
Gonzalez-Rodriguez, C. A-369
Gonzlez-Rodrguz, C.
B-255
Gonzalez-Tamayo, R.
B-139
Goodacre, S.
B-346, B-347
Goodall, M.
A-023, A-059
Gordon, A. M.
B-408
Gorka, G.
B-365
Gorrin, G.
B-040
Gorshkov, B. G.
B-279
Gtze, S.
A-360
Goudy, K.
A-041, B-264
Goudy, K. S.
A-057
Gougoutsi, A.
B-333
Gounden, V.
A-200,
A-277, A-406, A-408, B-104,
B-111, B-182
Goussetis, E.
A-219
Grace, M. E.
A-421
Graells, M.
B-139
Gramegna, M.
A-045, B-375,
B-377, B-435
Grandjean, C.
B-181
Granizo, V.
B-139
Grant, D. S.
A-143
Graubner, D.
B-303
Gravens-Mueller, L.
A-116
Gray, A.
A-402
Grebe, S.
A-190
Grebe, S. K.
A-226, B-214
Greene, D. N.
B-126
Greene, J.
B-236
Grenache, D.
A-269
Grenache, D. G. B-251, B-301,
B-384, B-386
Grenfell, R. F. Q.
A-333
Griesser, H.-W.
A-180
Grimmler, M.
B-365
Gronowski, A. M. B-266, B-301

Grosso, M. J.
Grote, C.
Gruzdys, V.
Grzesista, A.
Gu, J.
Guadix, P.
Guan, W.
Gucel, F.
Gudaitis, D.

A-338
A-399
A-317
B-365
A-408
A-225, B-255
B-117
A-256
A-095, B-084,
B-124
Gudaitis, P.
A-095,
B-084, B-124
Guedes, G. Morgana de Melo
B-057
Guerrero, J. M.
B-196
Guidi, G. C.
A-271, A-275,
A-284
Guilln, R.
A-293, A-335
Guilln Campuzano, E. A-073
Guimares, A. G. L.
B-161,
B-227
Gulati, S.
A-125
Gulbahar, O.
A-256
Gultekin, M.
A-112
Gumusay, O.
B-051
Gundlach, T.
A-180
Guneyk, A.
A-186
Guo, T.
A-232
Guo, X.
A-364
Guo, Y. Zheng
B-307
Gupta, M. K.
A-046
Gurel, H.
A-101
Gutierrez, M.
A-211
Guzman, G.
B-059

H
Hackenmueller, S. A.
B-251
Haddad, G.
B-291
Hafermann, J.
B-007
Hahne, K.
B-365
Haine, S.
B-324
Haklar, G.
A-040,
A-236, A-313
Haliassos, A.
B-010, B-151
Hall, C.
A-012
Hall, I. M.
B-202
Hallermayer, K.
B-325
Hamada, E.
A-359
Hmlinen, E.
A-227
Hammami, R.
B-324
Hammett-Stabler, C.
B-261
Hammett-Stabler, C. A. A-403
Han, Q.
B-336
Han, X.
B-073
Han, Y.
A-358
Hanci, T.
A-113
Handy, B.
A-029
Handy, B. C.
A-246, A-247
Hanson, C.
B-234
Hanson, S.
B-445
Hanson, S. E.
A-267
Hao, J.
A-172
Harbec, H.
A-273
Harding, S.
A-024, A-049
Harding, S. J.
A-328, A-337,
A-343, A-350, A-351, A-354,
A-355, A-356, A-362,
A-371, A-373
Hardy, R.
B-438
Harel, F.
B-071
Harenza, J.
A-090
Hariharan, M.
A-325
Harn, D.
A-333

Harris, T. E.
Harrison, H. H.
Hart, B.
Hartman, T. V.
Hartung, K. J.
Harwick, L.
Hasadsri, L.
Hasanoglu, A.
Hasebe, H.
Hasegawa, M.
Hashad, D. I.
Hashim, I.
Hashim, I. A.
Hashimoto, T.
Hassan, M.
Hassouna, F.
Hayashi, N.
Hayden, J. A.
Hazra, A.
He, K.
He, S.
Healy, E.
Heaney, J.
Hedin, C.
Hedley, B.
Heideloff, C.

B-387
A-141
A-414
B-391
A-266
A-064
B-272
B-380
B-284
B-342
A-098, A-140
A-278
A-248
A-304
A-098
A-302, B-416
A-160
B-410
A-077
B-421
B-307
B-381
A-023, A-059
A-427
A-315
A-081, A-413,
B-426
Heierman, E.
B-009
Heijboer, A.
A-212
Hein, J.
B-448
Heiss, G.
B-351
Helal, A.
A-098
Hemi, R.
A-193
Hemken, P. M.
A-064
Henderson, M.
B-032
Henderson, T.
B-287
Hendrix, B.
B-034
Henemyre-Harris, C.
B-163
Herkert, M.
B-252
Hermayer, K.
B-300
Hermsen, D.
B-369
Hernndez Cruz, B.
A-365,
A-370
Hernando de Larramendi, C.
B-087
Herranz-Puebla, M.
B-087
Hess, G.
B-331
Hewitt, J.
A-325
Hickman, P. E.
B-258
Hidaka, H.
B-118
Hidaka, Y.
A-150
Higgins, J.
A-298
Higgins, S.
B-047, B-076
Higgins, T.
A-183,
A-300, A-318
Hightower, C.
B-173
Higuchi, T.
B-058
Hild, C.
A-239
Hill, S.
B-319
Hinahon, C.
A-042
Hingnekar, C. V.
B-135
Hinsdale, M.
A-417, B-419
Hirano, T.
B-121
Hirayama, A.
A-346
Hoang, C. A-110, B-106, B-238
Hoang, S.
B-407, B-413,
B-417, B-423
Hobert, J.
B-269
Hodges-Savola, C. A.
B-354
Hoff, N.
B-047
Hoffman, B.
A-270
Hoffman, R. M.
B-336
Hofherr, S. E.
B-382
Holert, F.
B-282

S267

AUTHOR INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Holmes, D. T.

A-400, B-317,
B-362
Holmes, D. R.
B-354
Holmes, D.
A-183
Holt, C.
A-224
Holt, S.
B-210
Holweg, C.
A-102
Honda, T.
B-118
Hong, S.
A-329, B-041
Hoo, R.
B-308
Hoofnagle, A.
B-410
Hoogeveen, R.
A-148, B-351
Horn, P.
A-301
Horne, B.
A-103, A-272
Horowitz, G. L.
B-311
Horta, J.
B-433
Hotte, S.
B-353
Howes, M. B-294, B-295, B-299
Hsiao, C.
B-389
Hsieh, S.
B-433
Hsieh, W.-C.
B-206
Hsieh, Y.-J.
B-434
Hsu, Y.-H.R.
B-405
Hu, B.
A-378
Hu, R.
B-211
Hu, T.
A-013
Hu, X.
B-211
Huang, H.
B-040
Huang, I.-P.
B-439
Huang, J.-C.
B-434
Huang, X.
A-034,
A-336, A-341
Huang, Y.
B-179
Hughes, J. P.
A-051
Hui, C.
A-325
Hui, W.
B-031
Hungerford, W.
B-037
Hunsicker, L. G.
A-116
Hurley, B.
B-079, B-356
Hyland, J.
A-081

I
Iannotti, J. P.
Ibraim, I. C.
Ichikawa, T.

A-338
B-070
A-204, A-207,
A-233
Idogun, S. E.
A-071, A-319
Idonije, B. O.
A-436
Ignjatovic, S.
B-315
Ihenachor, O. E.
A-319
Iijima, S.
B-368
Ikeda, S.
B-122
Ikeda, T.
B-122
Ikenaga, H.
A-204, A-207,
A-233
Im, K. A.
B-348
Imai, Y.
B-058
Imana, G. E.
B-288
Imoto, A. A-204, A-207, A-233
Inaba, T.
A-069
Incandela, M. L.
B-435
Ingram, L.
A-327
Inoue, M.
B-065
Inoue, N.
A-150
Intachak, B.
B-237
Iqbal, J.
B-194
Isaza, M.
B-159
Ishii, J.
A-304, B-342
Ishii, N.
A-204, A-207, A-233
Ishii, T.
B-105
Ishii, Y.
A-026, A-187
Ishikawa, T.
A-304, B-342
Ishizuka, K.
A-069

S268

Isik, S.
A-216
Itkonen, O. M.
A-227
Ito, S.
B-415
Ito, Y.
A-345, B-105, B-121
Iturzaeta, J.
A-250
Ivandic, B.
B-309
Ivanova, A.
A-116
Iwanski, A.
A-380
Iwatani, Y.
A-150
Izarra, A. S.
A-055

J
Jaber, F. A.
Jack, R.
Jackson, B. P.
Jackson, R.
Jackson, T. K.
Jcomo, R. H.
Jacomo, R. H.
Jacques, P. F.
Jaffe, A. S.
Jaffe, A.
Jaffe, R.
Jannetto, P.
Jannetto, P.
Jansen, M.
Janzen, R.
Jaques, L. Lara
Jara Aguirre, J. C.
Jaramillo, S.
Jarari, A. M.
Jarari, N. M.
Jardini, D. P.
Jarolim, P.

A-367
B-395
B-027, B-028
B-230
B-166
A-166, B-200
B-197
A-116
B-332
B-320
B-320
B-391
B-422
A-189
B-246
B-001
A-118
B-148
A-010, B-326
A-010
A-282, A-306
A-116, A-122,
B-348, B-360
Jayasinghe, B.
B-258
Jean-Noel, A. V.
B-311
Jensen, R. A.
B-386
Jentz, T.
B-448
Jeon, Y.
A-285
Jeong, T.-D.
A-087, A-254
Jessen, K.
A-248
Jeyaraman, J. Jairaman
B-129
Jha, S. K.
A-106
Ji, L.
A-100, A-431
Jian, J.
B-179
Jiang, L.
B-211
Jianguo, S.
A-025
Jimenez, J.
A-321
Jimenez-Arriscado, P.
B-196
Jin, H.
A-044, A-060,
A-329, B-041, B-046, B-055,
B-080, B-195, B-207
Jin, M.
B-059
Jo, K.
A-145
Johns, C.
B-130
Johnson, H.
A-354, A-356,
A-371
Johnson, M.
B-438
Johnson, S.
B-011
Johnson, T.
B-225
Jones, A.
A-296, B-068
Jones, G.
B-234
Jones, J. B.
A-141, B-009
Jones, J.
B-400
Jones, S.
A-012
Jorge, A. J. L.
A-079
Jovicic, S.
B-315
Juang, A.
B-054
Juan Pereira, L.
A-073
Jung, H.
A-145

K
Kadeer, M. Juma
A-010
Kadosaka, Y.
B-108, B-113
Kakino, A.
B-109
Kale, V.
A-123
Kaleta, E.
A-409, A-424
Kalkan, D.
A-120
Kallmes, D. F.
A-131, A-133
Kalp, K.
A-224
Kalra, B.
A-171, A-229
Kalra, J.
B-146, B-162
Kalra, N.
B-146, B-162
Kamanda-Kosseh, M.
A-203
Kamara, F. K.
A-143
Kameda, T.
B-114, B-119
Kan, C. W.
B-012
Kanamori, K.
A-297, A-303
Kaneko, C.
A-346
Kanellias, N.
A-111, A-151,
A-308, B-333
Kanety, H.
A-193
Kang, C.
A-172
Kang, J.
B-179
Kang, M.
A-145, B-289
Kang, S.
A-153
Kangrga, R.
B-315
Kanneh, L.
A-143
Kano, M.
A-345
Kanzow, P.
B-215
Kao, L.
A-138
Kapoor, R.
B-072
Kappes, J.
A-122
Kara, A.
B-355
Kara, D.
A-216
Kara, M.
A-101
Karaky, N. M.
A-367
Karasawa, H.
A-026
Karatoy Erdem, B.
A-361
Kardos, K.
A-125
Karem, K.
B-038
Karmali, A.
B-047, B-076
Karon, B. S.
A-266
Kasal, C.
B-407, B-413,
B-417, B-423
Kasper, D. C.
A-414
Kassabova, L.
A-418
Kastritis, E.
A-111, A-151,
A-308, B-333
Kasumi, T.
A-339
Katagiri, M.
A-204, A-207,
A-233
Katayama, A.
B-368
Katayama, Y.
B-118
Katrukha, A.
B-355, B-358
Katsumata, Y.
A-150
Kattamis, A.
B-371
Katzmann, J. A. A-066, A-363
Kaufman, J. D.
B-117
Kaufmann, M.
B-234
Kaur, A. A-328, A-337, A-343,
A-350, A-351, A-355, A-362
Kavasi, M.
B-333
Kavira, N.
B-047
Kavsak, P. A.
B-319, B-353
Kawai, Y.
A-297, A-303
Kawano, S.
A-160
Kayadibi, H.
A-101, B-380
Kayahara, N.
B-118
Kearns, G.
B-414
Keeney, M.
A-315
Keevil, B.
A-415, A-419
Kelmendi, A.
A-018

Keno, D.
A-218
Kerdmongkol, J. B-093, B-123
Kero, F. A.
A-397
Kerr, J. R. A-337, A-350, A-355
Kesckes, Z.
B-258
Ketha, H.
A-190, B-422
Ketha, S. S.
B-422
Kettlety, T.
A-012, B-421
Khadka, D.
A-106
Khajuria, A.
A-201, B-007
Khan, A. B-298, B-392, B-316
Khan, K. Muhammed
B-392
Khan, S. H.
A-143
Khoury, R.
A-095,
B-084, B-124
Khupusup, K.
B-132
Kiernan, U. A.
A-407
Kilburn, B. J.
B-327
Kilmartin, P.
A-125
Kilpatrick, E. Leon
B-359
Kim, B.
A-292
Kim, C.
A-401
Kim, G.
A-044, A-060,
A-329, B-041, B-046, B-055,
B-080, B-195, B-207
Kim, H.-S.
A-039
Kim, H. A-039, A-044, A-060,
A-329, B-041, B-046, B-055,
B-080, B-195, B-207
Kim, J.-H.
B-110, B-441
Kim, J.
A-044, A-060, B-041,
B-046, B-055, B-080, B-115,
B-195, B-207
Kim, M.
A-153
Kim, S.
A-044, A-060,
A-329, B-041, B-046, B-055,
B-080, B-195, B-207, B-240
Kim, Y.
A-044, A-060,
A-329, B-041, B-046, B-055,
B-080, B-195, B-207
Kimura, S. C.
B-065
Kimura, T.
B-118
Kincaid, H.
A-102
King, C.
A-213
Kinukawa, H.
A-339
Kisaarslan, T.
B-131
Kitagawa, F.
A-304, B-342
Kitahara, S.-I.
A-346
Kitamura, Y.
A-187
Kitazawa, S.
B-065
Kitpoka, P.
B-132
Klapkova, E.
B-223, B-235,
B-383
Klause, U.
B-325
Klee, G. G.
B-020, B-267
Klein, E.
A-046
Kline, K.
B-437
Kluesener, R.
B-329
Ko, D.-H.
A-254
Ko, H.-Y.
B-206
Ko, K.
B-014, B-062
Koba, S.
B-121
Kobayashi, T.
B-442
Kobayashi, Y.
B-121
Kocabyk, M.
A-256
Kocak, H.
A-112
Koch, C. D.
A-266
Koch, D.
A-154
Koch, D. D.
B-126
Koerbin, G.
B-258
Kogan, A.
B-355
Kohler, R.
B-349, B-357
Kohno, K.
B-280
Kojima, S.
A-187, B-058

AUTHOR INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Kolla, S. D.
Koller, U.
Kollmar, O.
Kondreddy, R.
Kong, A.

B-101
B-330
A-360, B-215
B-326
B-407, B-413,
B-417, B-423
Kong, Q.
A-034
Konukoglu, D.
A-206
Koo, S.
B-115
Korkmaz, A.
A-433
Korman, E. W.
B-422
Koro, A.
B-131
Korpi-Steiner, N.
A-154
Korpi-Steiner, N. L.
B-297
Kse, K.
B-102
Kosovi, M.
A-147
Kotani, K.
B-112
Kotaska, K.
B-383
Kovac, M. Farias
A-338
Kowitski, K.
B-037
Kozak, M.
A-414
Kozo, D. R.
A-238
Kramer, C.
B-266
Kramer, P.
B-438
Krasnozhen-Ratush, O. A-184
Krastins, B.
A-407
Krause, P.
B-367
Krause, R.
B-031
Krempser, K.
B-033
Krintus, M.
B-330
Kritikos, G.
B-040
Kroll, M. H.
A-264, B-093,
B-123
Ksenevich, T. I.
B-279
Ku, S.-J.
B-441
Kubota, R.
A-297, A-303
Kuchipudi, L. S.
B-150
Kuder, J.
B-360
Kulakosky, P.
B-068
Kumakura, J.
B-105
Kumar, A.
A-171, A-229
Kumar, S.
A-035
Kumproa, N.
B-132
Kuno, A.
B-342
Kuno, T.
A-304
Kunst, A. N.
B-317
Kuroita, T.
B-442
Kurosaki, Y.
A-204, A-207,
A-233
Kurt, I.
A-243, A-265, B-380
Kurti, L.
A-018
Kurtipek, E.
A-130, A-135
Kurtoglu, G. Saglam
A-107
Kurtulu, E.
A-206
Kusachi, S.
B-122
Kusano, E. J. U.
B-049
Kusek, J. W.
A-116
Kushnir, M. M.
A-054
Kutasz, E.
A-245
Kutscher, P.
A-278
Kuwata, H.
B-118
Kwon, G.
B-115
Kwon, M.-J.
B-014, B-062
Kwong, T.
A-262
Kyle, P. B.
B-171
Kyriakopoulou, L.
B-259

L
La, M.
Labarbera, J.
Labay, L.
Lacher, D. A.
Lacoursiere, J.

A-252
A-141
B-167
A-051
A-394, B-429

Ladenson, P.
A-268
Ladwig, P. M.
A-374, A-404
Lafuente, A.
A-335
Lai, F.-L.
B-434
Lai, J.
A-414
Lai, S.
A-379
Lai, Z.-C.
B-189
Laing, E. F.
B-103
Lakos, G.
B-019
Lal, A.
B-040
Lam, L.-W.
B-157
LaMarr, W.
B-430
Lambert-Messerlian, G. A-171
Lamont, J.
A-347, B-337,
B-379, B-381
La Motta, M.
B-375
Landeta, E.
A-055
Landis, D.
A-086
Lang, C. D.
A-428
Lang, S.
A-315
Langman, L. J.
B-391, B-422
Langridge, J.
A-379
Larida, B.
A-349
Larkin, J.
B-320
LaRock, K.
B-276
Larramendi, C. Hdo
A-321
Larson, L.
B-246
Larson, T. S.
A-276, B-444
Lasho, M. A.
A-048, A-056
Latini, F.
A-115
Latini, R.
A-146
Lau, P.
B-040
Lauand, J.
B-033
Lauf, C.
B-252
Laulu, S. L.
A-234, A-235,
A-267, A-274
Lawson, S.
B-025
Lazo, M.
A-148
Lazzati, J. Manuel
A-245
Leanse, J.
B-165
Lebedin, Y.
A-005
Leber, A.
A-407
Leblanc, R.
A-273
Lechtenberg, L.
A-229
Lee, C.
B-389
Lee, D.-B.
B-441
Lee, D.
A-044, A-329, B-041,
B-195, B-207
Lee, D. S. L.
B-030
Lee, E.-H.
B-062
Lee, G.
A-405
Lee, H.
A-044, A-060, A-182,
A-285, A-329, B-041, B-046,
B-055, B-080, B-195, B-207
Lee, H. K.
B-166
A-405
Lee, J.-Y.
B-157
Lee, K.-W.
A-001
Lee, K.
B-115
Lee, M.
A-153
Lee, M.-A.
B-170
Lee, S.-G.
B-110
Lee, S.-Y.
B-240
Lee, S.
B-289, B-389
Lee, W. B. Y.
B-205
Lee, W.-I.
A-285
Lee, W.
A-087, A-153, A-254
Lee, Y.
A-145
Lee, Z.-J.
B-206
Lefevre, G.
B-330
Legallicier, B.
A-030, A-031
Legallo, R.
B-202
LeGatt, D. F.
B-405

Le Guillou-Guillemette, H.
B-082
Leguizamon, M. Angelica A-181
Lehman, C. M.
B-184
Lehouillier, C.
B-071
Leichtle, A. B.
B-008
Leilei, L.
A-025
Leimoni, I.
A-259
Lein, A.
B-365
Leite, L. L.
A-079, A-286
Leiva-Salinas, C.
A-211
Lekpor, C. E.
A-291, A-320,
A-342
Lembright, K.
B-025
Leni, A.
B-296
Lennartz, L.
A-260, B-330
Leon, M. Fernanda
B-159
Leonard, S.
B-022
Leong, D.
B-353
Lepage, N.
B-136, B-158
Ler, R.
B-328, B-352
Lessig, M.
A-122
Leung, E. Ki Yun
A-198,
B-165, B-249
Leung-Pineda, V. A-162, B-281
Leveille, R. L.
B-437
Levey, A. S.
A-116
Levine, R.
A-117
Lewinski, M.
B-029, B-074,
B-077, B-081
Lewis, D.
B-400
Lewisch, S. A.
B-225
Li, C.
A-358
Li, D. Feng
B-307
Li, H.
A-221
Li, J.
A-050, A-063, B-245,
B-246
Li, L.
B-073
Li, Q.
A-036
Li, S.
B-336
Li, W.
A-063
Li, Y.
B-418
Li, Z.
A-063
Li, Z.-Q.
A-012, B-421
Liang, G.
B-320
Liang, Y.
A-050
Liao, C.-C.
A-001
Lieske, J. C.
A-276, B-272,
B-444
Lim, J.
B-115
Lim, K. S. L.
B-030
Lim, S.
B-018
Lima, G. A. B.
B-248
Lima, N. O.
B-227
Lima, T. C.
A-430
Lima-Oliveira, G.
A-271,
A-275, A-284
Lin, C.
A-025
Lin, C.-C.
B-170
Lin, C.-N.
B-075
Lin, C.-Y.
A-001
Lin, F.-C.
B-261
Lin, H.-Y.
A-375
Lin, H.-T.
A-001
Lin, J.-C.
A-001
Lin, J.-H.
B-439
Lin, S.-P.
A-375
Lin, S.-J.
B-075
Linder, M.
B-264
Linder, M. W.
A-041, A-057
Lindner, G.
B-008
Lindsay, J.
A-061
Liniger, Z.
B-008
Lino de Souza, A. A.
B-061

Lippi, G. A-271, A-275, A-284

Lird, A.
A-181
Litt, S.
A-096, A-280
Little, R.
B-445
Little, R. R.
A-267, A-268
Liu, A.
B-269
Liu, H.
A-063
Liu, L.
A-062, A-218
Liu, M.
B-206
Liu, P.-M.
B-170
Liu, S.
A-034
Liu, W.-T.
B-170
Liu, X.
A-100, A-358, A-426
Liu, Y.-S.
B-434
Liu, Y.
B-307
Liu, Y.-C.
B-206
Livadara, T.
A-288
Llewellyn-Smith, N.
A-315
Llovet, I.
B-087
Lobreglio, G.
A-287
Lockington, K.
A-363
Lockington, K. S.
A-276
Logan, H. L.
A-274
Loken, P.
B-272
Long, L.
B-427
Longo, N.
B-269
Lopera, S.
B-148
Lopes, E. Pessoa
A-127
Lopes, G.
B-212
Lopes, G. C.
B-188
Lopez, M.
A-407
Lopez-Garrigos, M.
A-211,
B-087, B-134, B-139
Lopez-Hoyos, M.
B-087
Lopez-Yepes, L.
B-087
Lorenzo, C.
B-087
Lotz, J.
B-330
Love, S.
B-349
Love, S. A.
B-328,
B-352, B-357
Lowe, C. F.
B-078
Lowenthal, M. S. B-247, B-359
Lowry, P. A-061, A-222, B-321
Lteif, A.
A-190
Lu, H.
A-034
Lu, J.
A-224, B-384
Lu, W.-S.
B-439
Lu, Z.
A-034
Lucas, F. A-309, B-402, B-425
Lucini, R.
B-375, B-377
Lueke, A.
A-402, B-332
Lugo, J.
A-211
Lundin, M.
A-012
Lunel-Fabiani, F.
B-082
Luo, M.
B-242
Lutsey, P.
A-138
Luz, A. B. G.
A-435
Lyden, E.
B-234
Lynch, K. L.
B-201, B-408,
B-432

M
Ma, L.
Ma, Q.
Mabic, S.
Maceiras, M.
Macher, H. C.
Maciuca, R.
Madrol, V.
Maekawa, M.
Maerker, T.
Magalhes, M. M.
Magera, M. J.

A-102
A-426
A-253
A-245
B-196
A-102
B-326
A-359
B-365
B-227
B-382

S269

AUTHOR INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Mages, K.-J.
B-296
Maggiore, J. A.
A-420, B-367
Mahadevan, M. S.
B-202
Mahfouz, R. A.R.
A-367
Mahmoudi, R.
B-373
Maiz, L.
B-087
Majdi, E.
A-387, A-397,
B-397, B-406
Majkic-Singh, N.
B-315
Majnesj, K.
B-421
Makris, K.
B-010,
B-151, B-337
Malaghini, M.
A-258
Malta, F. S. V.
B-188
Maluf, N.
A-357
Maluf, N. Zaidan A-185, A-324,
B-024, B-035
Maluf, N. Z.
B-001, B-005,
B-364
Malvaso, M.
A-273
Mambrin, M. C.
B-091
Mamiya, S.
B-098
Mandel, A.
B-173
Mangueira, C. L. P.
A-205
Manicke, N. E.
B-412
Manoj, S.
B-245, B-246
Mansour, M.
B-353
Mansour, O.
A-099
Mansur, R.
B-019
Mao, H.
A-358
Marani, R. Jesus
B-019
Marca, K.
A-177
Marchetti, C.
A-161
Mardekian, S. K.
B-153
Margeli, A.
A-111,
A-151, B-371
Marom, B.
B-292, B-293
Marques, F. K.
A-289
Marquet, P.
B-392
Marshall, B.
B-409
Martens, F.
A-212
Martin, L.
A-221
Martinez, O.
B-159
Martinez, R.
A-240
Martinez-Ingles, J. R.
B-134
Martinez-Llopis, M. J.
B-087
Martin-Martin, L.
B-134
Martins, J.
B-095
Marzinke, M. A.
B-424
Marzouk, S.
A-140
Mashiba, S.
B-112
Masson, S.
A-146
Mateo, E.
A-007
Mateo, E. C.
B-191
Mateo, E. C. C. A-289, A-382,
B-070, B-188, B-193
Matern, D.
B-382
Mathias, P.
B-410
Mathias, P. C.
B-168
Matsui, S.
A-304
Matsumoto, H.
B-442
Matthews, J.
A-102
Matthias, T.
A-349
Mattsson, L.
A-123
Matyus, S. P.
A-149
Mauracher, C.
B-434
Maym, J.
A-225
Mazarakis, M.
A-308
Mazzotti, D.
B-061
Mazzulli, T.
B-078
Mbeunkui, F.
A-389, B-409
McBane, R. D.
B-422
McBride, T.
A-086
McCann, K.
A-388, A-390

S270

McCash, S. I.
A-092
McCloskey, L. J. A-231, A-428,
B-027, B-028, B-153
McConnell, J. P.
B-117
McConnell, R.
A-061, A-222,
A-347, B-321,
B-337, B-381, B-427
McCudden, C. R.
B-032
McDonald, J. S. A-131, A-133
McDonald, R. J. A-131, A-133
McElhatton, S.
A-294, A-432
McEvoy, J.
A-148
McFadden, T.
A-347, B-337,
B-381
McFall, A.
A-222
McGrowder, D.
A-109
McGuire, C.
B-037
McIntire, G. L.
B-394
McIntire, G.
B-210
McKinney, Z. J. B-349, B-357
McLellan, W.
B-136
McNamara, N.
A-315
Meade, T. J.
B-308
Medeiros, T. Batista
A-127
Medeiros, Y. S.
A-027
Medrano, M.
A-293, A-335
Meemaew, P.
B-237
Meeusen, J. W.
A-402, B-332
Megahed, M.
A-099
Meikle, A. Wayne
A-054
Melguizo, E.
A-369, B-004
Melo, C. P.
B-191
Melo, C. P. S.
B-193
Melo, J. F.
B-199
Melzer, C.
B-329
Menassanch-Volker, S.
B-325
Mendona, S. Aparecida De
Domingues
A-324
Mendu, D. Rao
A-092
Menezes, M. dos Santos B-001
Meng, Q. H.
A-029, B-175
Meng, Q. A-246, A-247, A-426
Menna Barreto, A.
A-155
Merrick, L.
B-078
Merrick, R.
B-130
Merrigan, S. D.
B-244
Mesquita, E. T.
A-079
Metwally, H. Gabr
A-119
Metzmann, E.
B-365
Meyer, R. E.
B-012
Meyers-Needham, M.
B-202
Meyer zum Bschenfelde, D.
B-282
Michelotto, S. S.
B-066
Mika, D.
B-165
Mikailova, P.
A-383
Milhorn, D. M.
B-297
Miller, J.
A-154
Miller, L.
A-092
Miller, M. Craig
B-418
Miller, V.
B-430
Milliner, D. S.
B-272
Millner, L. M.
A-041, A-057
Mills, J. R.
A-066, A-404
Milosevic, D.
B-214
Min, W.-K.
A-087, A-254
Mindicino, H.
A-167
Miranda, M. Pereira
A-281,
A-295
Mitchell, T.
B-438
Mitulovic, G.
B-379
Miwa, T.
A-108
Miyauchi, K.
B-118
Mckel, M.
B-282

Modolo, M. Luisa
B-086
Mody, H. C.
B-038
Mohamed, M.
A-278
Mohammad, H. F.
A-403
Moiana, A.
B-435
Mojiminiyi, O.
A-214, A-215
Mojiminiyi, O. A. A-176, A-199
Moley, K. H.
B-266
Molina, J.
B-134
Molinaro, R.
A-154
Molinaro, R. J.
B-156
Molinero, P.
B-196
Molinos, J. I.
B-134
Molloy, B. J.
A-419
Momiyama, M.
B-446
Mommoh, M. O.
A-188
Momoh, M.
A-143
Monga, D.
B-040
Montagna, R. A.
B-037
Montagnana, M.
A-284
Moon, B.
B-428
Moore, A.
B-287
Moore, D.
A-110, B-106
Mora, C.
A-181
Mora, R.
A-250
Moraes, D. Quintiliano
B-128
Moraes, R. Bueno
A-258
Morando, A.
B-401
Morasse, S.
B-071
Moratori, D.
A-258
Moreira, E. Fernandes
B-005
Moreira, M. L.
A-079, A-197,
B-023
Morgan, B. R.
A-310
Morita, M.
B-280
Moriyama, K.
B-058
Morris, A. A.
B-394
Morris, C.
A-332
Morrison, K.
A-263
Morrow, D. A.
B-348, B-360
Mortensen, R.
B-301
Moseley, S. L.
B-029, B-074,
B-077, B-081
Moses, L. M.
A-143
Moskowitz, J.
B-424
Mosley, T.
B-351
Moulds, J. M.
B-021
Mouro, P. H. O.
B-063
Moyer, V. A.
A-121
Mueller, H.
B-365
Mukadi, P.
B-047
Mukharyamova, K.
B-355
Mller, C.
B-282
Mullins, G. R.
B-387
Muncy, I.
B-079, B-356
Muniyappa, R.
A-200
Muniz, N.
B-099
Munkhdelger, J. B-195, B-207
Murakami, H.
A-026, A-187
Murakami, M.
B-446
Murakami, M. M.
B-327,
B-328, B-349, B-352, B-357
Murata, K.
A-421
Murphy, F.
A-354, A-356,
A-362, A-371, A-373
Murphy, S.
B-360
Murray, D.
A-066
Murray, D. L.
A-066, A-374,
A-404
Murray, J. A.
A-364
Murray, L. M.
B-095
Murray, R.
B-013
Murshed, S.
A-230
Muto, N. H.
B-199, B-209

Muyembe-Tamfum, J.-J.
Myers, T. W.
Myers, T.

B-047
B-029
B-286

N
Na, M.
A-255
Nabhan, C.
B-165
Nader, M.
A-142, A-430
Nadkarni, S.
A-248
Nagano, N.
A-339
Nagashima, K.
A-339
Naides, S. J.
A-218
Nakajima, K.
A-339
Nakamoto, J.
A-218
Nakamura, Y.
A-346
Nakanishi, M.
A-069
Nakano, A.
B-109
Nakayama, A.
B-368
Namasivayam, V.
B-029
Nambi, V.
B-351
Nandkeshwar, R.
B-040
Nang, S.
A-010
Naqvi, H.
B-016
Narayan, H. K. V.
B-135
Narla, S. N.
A-103, A-272,
B-300
Naruse, H.
A-304, B-342
Naseb, N.
B-101
Navarro Alvarez, P.
A-028
Navarro-Casado, L.
B-134
Navarro-Compn, V.
A-369
Navarro Sarabia, F.
A-365,
A-370
Navarro-Sarabia, F.
A-369
Naves, L. A.
A-166
Navigante, A.
B-221
Ndimbie, O.
B-407, B-413,
B-417, B-423
Ndumele, C.
B-351
Nedelcu, E.
B-208
Nelson, A.
B-029
Nelson, P.
A-103
Neren, E. J.
A-002
Nerenz, R.
B-301
Nery, L. F. A.
A-166, B-143,
B-228, B-270
Nery, L. F. Abdalla
B-263
Netzel, B. C.
A-056
Neuman, G.
B-415
Neumark, E.
B-006, B-015
Neville, K.
B-414
Newhouse, C.
B-029
Ng, W. Yoong
B-030
Ngamboli, J.
B-047
Nguyen, A. N. D.
B-208
Nicholson, J.
B-328, B-352
Nickander, K. K.
B-382
Nickolas, T.
A-203
Niederkofler, E. E.
A-407
Niewiadonski, V. Dionisio
Tavares A-004, A-242, B-064,
B-185, B-192, B-220
Nifoci, D.
B-064
Nikitin, M. P.
B-279
Nikitin, P. I.
B-279
Nilsen, T.
B-376
Ning, H.-C.
A-001
Ning, N.
A-006, A-036
Nishimura, P. Yoshie
A-004,
B-064
Niu, J.
A-302
Njoroge, S. W.
A-280

AUTHOR INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Nkrumah, C.

A-084, A-136,
A-144
Noble, J. E.
B-358
Nobre, C. S.
B-197
Noel, S.
B-167
Nogueira, L.
A-007
Nogueira, N. C. C. S.
B-304
Nogues, D.
B-082
Nolen, T.
B-284
Nomura, N.
A-069
Noto, A.
A-146
Nouri-Girones, F.
B-258
Novella, J.-L.
B-373
Nsiah, P.
B-103
Nunez, E.
B-099
Nwaoduah, N. M.
A-428
Nybo, M.
B-330

O
Obata, S.
OCallaghan, L.
Ochmanova, R.
ODwyer, C.

A-204
B-050
A-170
B-262, B-265,
B-273
Oehler, R.
B-379
Oellerich, M.
A-360, B-215
Oglesbee, D.
B-272
Oh, S.-Y.
B-240
Oh, S. J.
B-428
Ohkawa, R.
B-098, B-114,
B-119
hrvik, A.
B-421
Okamura, C.
A-026
Okazaki, M.
B-122
Okitolonda-Wemakoy, E. B-047
Okubo, S.
B-114
Okur, I.
B-380
Okuyama, K.
A-346
Okuyama, R.
A-304
Oladipo, O. O.
A-068
Oldridge, E. E.
B-286
OLeary, M.
A-301
Oliveira, C. L.
A-282, A-306
Oliveira, C. M. M.
A-208
Oliveira, E. Vigliari V de B-001
Oliveira, E. Tavares
B-057
Oliveira, E.
A-124
Oliveira, L. G.
A-323
Oliveira, P. M.
A-124
Oliveira, P. Lima de
B-192
Oliveira, S. G.
B-061
Ollman Saphire, E.
B-068
Olorunju, S. A. S.
B-095
Olsen, K. H.
A-064
Olson, E.
B-009, B-025
Olson, M. T.
A-416
Olugbodi, A. O.
A-068
Olumese, F. E.
A-437
Omar, A.
B-205
Omer, B.
A-383
Omoruyi, F. O.
A-437
Ong, L.
A-169, A-334
Onoagbe, I. O.
A-437
Onyenekwu, C. P.
B-306
Ookura, H.
B-108, B-113
Oon, L.
B-194
Oottamasathien, D.
A-143,
B-068
Oriseseyigbemion-Sakpa, E.
A-071
Orlov, A. V.
B-279
Orluwene, C. G.
A-188
Orth, M.
A-260

Ortuo, M.
Osadolor, H. B.
Osiecki, J.

B-139
A-436
B-029, B-074,
B-077, B-081
Osman, A. Osama Bakr
Seddik
A-119
Osmani, I.
A-018
Osorio, P.
A-152
Oter, S.
A-433
Oto, H.
B-065
Otvos, J. D.
A-149
Oudart, J.-B.
B-373
Ouverson, L. J.
A-279
Owen, W. E.
A-228, B-244
Owiredu, W.
A-136
Oyakawa, L.
B-199, B-209
zahin, A.
B-102
Ozaki, Y.
A-304
Ozcan, N.
A-186
Ozdem, S. S.
A-126
Ozdem, S.
A-126
Ozdemir, O.
A-120
Ozdemir, S.
A-120, A-186,
A-216
Ozel, D.
A-112
Ozenirler, S.
B-051
Ozgurtas, T.
A-243
Ozler, M.
A-433
Ozturk, A.
A-113
Ozturk, G.
A-113,
A-120, A-165, A-216

P
Padilha, J. P. D.
B-161, B-227
Paglia, G.
A-379
Pais, T.
A-321
Paiva, F. Oliveira
A-353,
A-368, B-019
Palavecino, E.
A-417
Palme, S.
A-102
Palmer, G.
B-007
Pamboucas, C.
B-333
Pamidi, P. V. A.
B-180
Panagiotakis, O. B-010, B-151
Pandian, J.
A-174
Pandian, R.
A-174
Panetsos, F.
A-288
Panisa, D.
A-330
Pankow, J.
A-138
Pantani, M.
A-275
Panteghini, M.
B-358
Pantelaki, K.
A-308
Papassotiriou, G. P.
A-111,
A-151
Papassotiriou, I. A-111, A-151,
A-219, A-288, A-308,
B-333, B-371
Paquette, N.
B-071
Parikh, J.
A-327
Park, C. Y.
B-182
Park, C.
B-389
Park, H.
B-014, B-062, B-115
Park, H.-D.
A-392
Park, K.
A-044, B-195, B-207
Park, M.-J.
A-039
Park, S. A-044, A-060, A-329,
A-405, B-041, B-046, B-055,
B-080, B-195, B-207, B-289
Park, S.-D. B-046, B-055, B-080
Park, S.-Y.
B-441
Park, T.
A-285
Park, Y.
B-115
Parker, R. L.
A-274

Parrillas, V.
Parrini, V.
Parvin, C. A.

A-250
A-146
B-150, B-169,
B-174
Pasaoglu, H.
B-051
Pasetti, G.
A-146
Pasquale, D.
B-124
Pasquali, M.
B-269
Pass, H.
A-035
Patel, A. S.
A-171, A-229
Patel, D.
A-410
Patel, H.
A-298
Patel, P.
A-023, B-084, B-012
Patel, R.
A-410
Patel, V.
B-316
Paul, S.
A-364
Paulino, S. V.
B-161
Paulo, B. P.
A-382
Paulsen, R.
A-239
Pauly, D.
B-448
Pauly, K.
B-448
Paus, E.
A-064
Pawliszyn, J.
A-376
Payto, D.
A-255
Payto, D. A.
A-105, A-413,
B-398
Peace-Brewer, A.
A-086
Peak, M.-C.
B-441
Pearlman, E. S. A-110, A-327,
A-332, B-106, B-238
Pechova, M.
B-235
Peck Palmer, O.
A-062
Pedrosa, W.
B-183, B-227,
B-253
Peela, J. R.
A-010,
B-101, B-326
Peela, L. Teja
A-010, B-101,
B-326
Peh, B.
B-194
Pereira, C.
A-357
Pereira, C. F.
A-152
Pereira, C. F. A. A-159, A-163,
A-173, A-175, A-196, A-209,
A-286, B-023, B-067, B-177
Pereira, C. Figueiredo de Araujo
A-185, A-242, A-330,
B-001, B-005, B-035,
B-057, B-128, B-364
Pereira, I. A.
A-142
Pereira, L. Gustavo
B-128
Pereira, R. M. G.
B-049
Pereira, S. L.
B-242
Perez, M. M.
B-311
Perez, R. M.
A-124
Perez Mendez, C.
B-257
Prez-Prez, A. A-052, A-225,
B-255
Prez-Ramos, S. A-017, A-202
Perez-Valero, V.
B-139
Perris, P. D.
B-091
Pessin, M. S.
A-092, A-421
Pesta, M.
A-058
Pesudo, S.
B-139
Peterman, S.
A-407
Petersen, J.
A-410
Petrides, A. K.
B-424
Pfeffer, M. A.
A-116
Pham, K.
B-428
Pham, N.
B-407, B-413,
B-417, B-423
Pham, N.
B-407, B-413,
B-417, B-423
Phan, K.
B-319
Phanthalangsy, C.
A-283

Pheunsongkram, P.
B-283
Phinney, K. W.
B-247
Phipps, R.
B-175
Phua, S. K.
B-338, B-361
Picard, P.
A-394, B-429
Picard, R.
B-421
Pich, M.
A-273
Picheth, G.
A-271,
A-275, A-284
Pickering, J.
B-269
Piech, T.
B-012
Pignatari, A. C. C.
B-049
Pikulski, M.
B-250
Pilkington, M.
B-400
Pillappa, R.
B-106
Pillay, T. S.
B-095
Pinheiro, M. F. M. C.
A-208,
B-248
Pinho, J. R.
B-199, B-209
Piper, K.
A-432
Pirozzi, L.
A-146
Pitcher, T.
B-264
Pitts, K. R.
A-143, A-296,
B-068, B-079, B-356
Planker, A.
B-172
Plant, T.
A-059
Plate, C. A.
B-400
Plaut, D.
B-158
Plaut, D. S.
B-136
Plebani, M.
A-147, B-330
Plouffe, B.
A-117, A-167
Plummer, D.
A-096
Poirier, D.
B-276
Pokharel, D. R.
A-106
Pokharel, Y.
B-351
Poli, G.
A-271, A-284
Politi, L.
B-377
Polivka, J.
A-058, A-058,
A-372, A-372
Pombar Prez, M.
B-203
Poncela, V.
B-087
Poncheri, M.
B-047, B-076
Pond, G.
B-319
Popov, J.
A-218
Porras, A.
B-148, B-159
Porto, L. B.
A-286
Posey, Y.
B-039
Post, W. S.
B-117
Potter, J. M.
B-258
Prabhala, T.
B-182
Prakash, A.
A-407
Prakash, S.
A-162
Preusch, M.
A-251, B-069
Prieto, F.
A-357
Pritchett, J. S.
B-247
Promnuch, S.
B-237
Provuncher, G.
B-012
Prusa, R.
A-259, B-223,
B-235, B-383
Pruthi, R. K.
B-422
Psimenou, E.
B-333
Ptolemy, A. S.
A-221
Purohit, P.
A-164, A-192
Putnam, J.
B-210

Q
Qian, J.
A-358
Qiao, H. A-387, B-397, B-406
Qiu, J.
B-316
Qiu, Y.
A-172, A-252
Quiles, M. G.
B-049
Quilez-Fernandez, J. L. B-134

S271

AUTHOR INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Quillard, M.
Quinn, F.

A-030, A-031
B-273

R
Rabadan, L.
Rabin, B.
Racsa, L.
Racsa, L. D.
Radack, J.
Radwan, R.
Rahman, M. K.
Rainbow, S. J.
Raines, M.
Raizman, J.

B-087
A-349
A-278
A-248
A-162
A-012, B-421
A-230
A-422
A-213
B-259, B-271,
B-287
Raju, S.
A-012, B-421
Ramaiah, S.
A-283
Ramanathan, L. V. A-092, A-421
Ramont, L.
B-373
Rampersaud, D.
A-182
Ramsay, C.
A-064
Randall, C.
B-076
Randell, E.
B-265
Randell, T.
A-221
Raphael, N. A.
B-061
Rappold, E.
B-379
Rashid, G.
B-006, B-015
Rastogi, M.
A-268
Ratanakasetsin, P.
B-043
Ratcliffe, P.
A-061, B-427
Ravago, E.
B-013
Ravi, S.
B-438
Rawlings, A. M.
A-138
Rayamajhi, P.
A-106
Recchi, C.
A-284
Reddy, B.
B-434
Redman, L.
B-290
Redmond, S.
B-287
Reed, K.
B-017
Refetoff, S.
A-198
Rego, F. G.
A-275
Reineks, E.
B-025, B-230
Reineks, E. Z.
A-257
Remaley, A. T.
B-104, B-111
Renis, M.
A-287
Repraz-Andrade, A.
B-203
Repik, T.
A-058
Reyes-Garcs, N.
A-376
Rhea, J. M.
A-154
Ribeiro, M. Rambaldi
B-005
Ribes, J. L.
B-134
Ricardo, C. Cristina Volpe B-364
Ricchetti, E. T.
A-338
Ricchiuti, V.
A-309, B-402,
B-425
Rice, T. W.
A-096, A-280
Richardson, C.
A-222, A-347,
B-337, B-381
Richter-Roche, M.
A-154
Rigone, M. Aparecida
A-353,
A-368
Rimoin, A.
B-047
Rinaldo, P.
B-272
Rink, J.
B-365
RIOS, D.
B-148
Rios Tamayo, R.
A-028
Rissin, D. M.
B-012
Ritchie, G.
B-078
Ritchie, J. C.
B-097
Riviere, A.
A-340
Rivnak, A. J.
B-012
Rizos, D.
B-010, B-151
Roback, J.
B-021

S272

Robeson, B.
B-007
Robinson, J. K.
A-362
Rocchetti, T. T.
B-049
Rocha, M. Fbio Gadelha B-057
Rochanawutanon, M.
B-093,
B-123, B-237
Rockwood, A. L.
A-054
Rodrigues, A. O.
B-248
Rodrigues, C. Seabra
B-185
Rodrigues, R. R.
A-152
Rodrigues, S. S. A-282, A-306
Rodriguez, M.
A-294, A-432
Rodriguez, P.
B-029, B-074,
B-077, B-081
Rodriguez, S.
A-245
Rodriguez Capote, K.
A-183,
A-300, A-318
Rodriguez Fraga, O.
B-349
Rodriguez-Piero, A.
B-139
Rodriguez-Rodriguez, A. B-087
Roehrig, S.
B-029
Roessner, E.
B-421
Roger-Dalbert, C.
B-071
Rogers, E. J.
B-433
Rogers, L. C.
B-449
Rohan, V.
A-058, A-372
Rohlfing, C. L.
A-267, A-268
Rom, W.
A-035
Romani, A.
A-217
Romano, O.
A-007
Romeiro, J. R.
B-063
Romero, M.
B-159
Romro Ospina, A.
A-273
Rraas, T.
A-259
Rosa, D. W.
A-435
Rosa, M. G.
A-079
Rosano, T.
A-401
Rosecrans, R.
B-236
Rosenfeld D.
A-315
Rosenthal, S.
B-365
Rosin, C.
A-185, B-005
Rosmalen, C. F.
A-189
Rossini, L.
B-034
Rousseau, A.
A-180
Roussou, M.
A-308, B-333
Rowe, C.
A-355
Roy, D.
A-035
Rubin, J.
A-148
Rubinstein, M.
A-245
Rubio, A.
B-196
Ruiz, E.
A-335
Ruparelia, F.
A-376
Rupprecht, K.
A-064
Rush, N. I.
B-172
Ryan, E. L.
B-097, B-156

S
Sabatine, M.
B-360
Sabino, E. Cerdeira
B-220
Sachdeva, R.
B-146, B-162
Sacks, D.
B-445
Sadaka, M.
A-140
Sadilkova, K.
B-395
Sadilkova, P.
A-372
Sadrzadeh, S. M. Hossein B-290,
B-404
Saegusa, J.
A-160
Saeid, A. R.
B-326
Saenger, A. K.
A-402, B-325,
B-332
Sagae, N.
B-119
Saggiorno, M.
A-284
Said, A. Rahaman A-010, B-101

Saito, K.
A-069, A-346
Saito, T.
B-058
Sajo Beqaj, S.
B-071
Sakai, N.
A-297, A-303
Sakamaki, K.
A-339
Sakamoto, F. Tadashi Carvalho
B-128
Sakatsume, M.
B-368
Sakiman, Z.
B-129
Salamone, S. J.
A-238, B-418
Salas Garca, .
A-073
Saldana, S.
A-064
Saleh, A.
A-338
Saleh, S. A. K.
A-011, B-229
Sales, J. Alencar
B-057
Salifu, H.
A-320
Salinas, M.
A-211, B-087,
B-134, B-139
Salm, E.
B-434
Salmon, B. P.
A-095, B-084,
B-124
Salvagno, G. L. A-271, A-275,
A-284
Sama, J.
A-364
Sampson, M. L.
B-104, B-111
Samuels, K. A.
A-337
Sanahuja, M. C.
B-091
Sanches, R. C.
A-323
Snchez, C.
B-022
Snchez Jacinto, B. J.
A-118
Snchez-Jimnez, F.
A-052
Snchez Margalet, V.
A-028,
A-047, A-365, A-370
Sanchez-Margalet, V.
A-052,
A-225, B-255
Sandberg, S.
A-259
Sander, H.
B-296
Sander, L. C.
B-247
Sandoval, Y.
B-327, B-328,
B-349, B-352, B-357
Sanso, R. C. A.
A-152
Santa Rita, T. H. B-200, B-270
Santos, D. Cezario dos
B-035
Santos, L.
A-177
Santos, S. Maria Eli
B-253
Santotoribio, J. D. A-017, A-202
Sapkota, L. B.
A-106
Sar, O.
A-134
Sargent, D. D.
A-141
Sartori, D.
B-393
Sastre, J.
B-134
Sato, I.
A-160
Sato, M.
B-098, B-114
Sato, S.
A-187
Satoh, N.
B-105
Savjani, G.
A-171, A-229
Saw, B.
A-334
Saw, S.
A-169, A-334
Sawamura, T.
B-109
Sawaya, A. M.
B-437
Sazci, A.
A-040
Scarpelli, L. Cesar A-242, A-258
Schageman, J.
A-042
Schandl, E. K.
A-016
Scharf, M.
A-155
Scheibe, B.
A-201
Schiavoni, L.
B-225
Schieffelin, J. S.
A-143
Schimmel, H.
B-358
Schippers, H. Pascal C.
B-369
Schmidt, C.
B-369
Schmidt, R. L.
A-235, A-267,
B-184
Schmitz, J.
A-360, B-215

Schmotzer, C.
Schoener, D.
Schofield, R.
Scholes, K. L.
Schrank, Y.
Schranz, D.
Schroeder, T.
Schroff, A.
Schulz, K. M.

A-261
B-029
A-421
A-274
A-210
A-213
A-407
B-012
B-327, B-328,
B-352
Schtz, E.
A-360, B-215
Scirica, B.
B-360
Scirica, B. M.
B-348
Scott, D.
B-414
Searle, J.
B-282
Seiden Long, I.
B-031
Seiden-Long, I.
B-341
Selvin, E.
A-138, A-148
Semmel, D.
B-291
Sen, S.
A-038
Sener, A.
A-316
Serdar, M.
A-243
Seres, Z.
A-174, A-180
Serin, E.
B-131
Sertoglu, E.
A-265, B-380
Sertolu, E.
A-101
Sethi, S.
A-169, A-334
Setterquist, R.
A-042
Sevinc, M.
A-134
Sezer, S.
A-120
Sha, Z.
A-358
Shahangian, S.
A-121
Shahine, E.
A-098
Shakila, S.
A-010,
B-101, B-326
Shalaurova, I.
A-149
Shalom, A.
B-006, B-015
Shamburek, R.
B-104, B-111
Shanbhag, G.
B-321
Sharabi, Y.
A-193
Sharma, A.
B-437
Sharma, P.
A-164, A-192
Sharma, S.
A-410
Sharp, K. L.
A-355
Shaw, S.
B-407, B-413,
B-417, B-423
Shea, J.
B-287
Shen, L.
A-148
Shen, Y.
B-303
Sheng, L.
A-358
Sherlock, C. H.
B-078
Sherman, J.
A-102
Shi, R.
B-412
Shi, S.
A-366
Shi, X.
A-302
Shiba, K. A-297, A-303, B-368
Shieh, M.-J.
B-170
Shih, J.
A-260, B-330
Shim, T.
A-145
Shimakawa, A.
A-345
Shimizu, T.
A-346
Shimomura, Y.
A-339
Shin, K.-S.
A-039
Shinohata, R.
B-122
Shodin, B.
B-260, B-273
Shoji, M.
B-121
Shortt, C.
B-319, B-353
Shou, Z.
A-358
Showell, P. J.
A-328, A-337,
A-343, A-350, A-351,
A-354, A-355, A-356,
A-362, A-371, A-373
Shrestha, L.
B-089
Shu, I.
B-400

AUTHOR INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Shugarts, S. B.
B-432
Shukla, M. R.
B-038
Shukla, P. S.
A-106
Sicuro, F.
A-287
Sidobre, S.
A-102
Sidrim, J. Julio Costa
B-057
Sigdel, M.
A-106
Silva, D. P.
B-067
Silva, J.
B-035
Silva, N.
B-364
Silva, O. Fernandes
A-210
Silva, O. F.
B-023
Silva, V. M.
B-200
Silva Filho, O. F.
A-208
Silva-Moraes, V.
A-333
Silverman, L. A.
B-202
Silverman, S.
B-287
Simkowski, K. W.
B-039
Simon, J.
A-081
Simon, V.
B-404
Simos, N.
A-308, B-333
Singh, P.
B-246
Singh, R. J.
A-190, A-226
Sirikci, O.
A-236, A-313
Sitnik, R.
B-199, B-209
Situ, W.
A-358
Sivakamasundari, R.
A-035
Skadberg, O.
B-330
Skale, D.
A-307
Skenderi, K.
A-219
Skinner, J.
A-064
Slagman, A.
B-282
Slobodianik, N.
B-221
Slobodianik, N. H.
B-091
Smalley, R.
B-421
Smillie, S.
B-427
Smith, A.
B-201
Smith, B.
B-017
Smith, E.
B-029
Smith, K. M.
B-327
Smith, M. B.
A-261
Smith, P.
A-380
Smith, S. W.
B-327, B-328,
B-349, B-352, B-357
Smith, U.
A-327
Smith, Y.
B-017
Smyth, S.
B-427
Sneidman, M.
A-141
Sneidman, M. R.
B-390
Snyder, M.
A-364
Snyder, M. R.
A-374, A-404
So, N.
B-398
Soares, L. M.
B-253
Soga, T.
B-065
Sohn, K.-Y.
B-319
Solanki, M.
A-328
Soldin, S. J.
A-200, A-406,
A-408, B-182
Sol, S.
B-022
Soliera, A. Rachele
A-045
Solomon, S.
A-116, B-351
Somay, G.
A-316
Song, L.
B-277
Song, W.
A-039
Soon, C.-C.
B-157
Sorato, S.
B-227
Soukka, T.
A-123
Sousa, C. F.
B-270
Sousa, L. Bispo
A-080
Southan, L.
A-350
Southan, L. D.
A-337, A-351
Souto, E. Xisto
A-080
Souza, P. R. C.
A-286
Sowder, A. M.
B-301

Spaid, K.
Spanuth, E.

B-182
A-251, B-069,
B-309, B-331, B-343
Spingarn, S.
B-013
Spinola, M.
B-311
Spizz, G.
B-037
Spurgin, J.
B-452
Srilakshmi, P.
B-326
Srisakul, V.
B-043
Srisawasdi, P.
B-093, B-123,
B-132
Stack, P. R.
B-327
Stano, P.
B-086
Steeg, J.
B-162
Steele, A. N.
B-294, B-295,
B-299
Steele, P.
A-301
Steffen, B. T.
B-117
Steffes, M.
A-138
Stein, J. H.
B-117
Steinle, R.
B-230
Stejskal, D.
A-170
Stejskal, J.
A-283
Stella, A.
B-433
Stene, D.
B-277
Stengelin, M.
A-035
Stenman, U. Haken
A-064
Stennett, D.
A-109
Stephen, D. W. S.
B-345
Sternen, D. L.
B-168
Stevenson, L.
A-061
Stewart, W.
A-064
Stickle, D. F.
A-231, A-428,
B-027, B-028, B-153
Stckl, D.
B-127
Stolla, M.
A-262
Stolze, B.
A-408
Stolze, B. R.
A-200
Stone, S. L.
A-343
Stranz, M.
B-246
Straseski, J.
A-224
Straseski, J. A.
A-223, A-228,
A-234, A-235, B-184, B-244
Strathmann, F. G.
A-274
Strauss, R.
B-261
Strickland, E. C.
B-394
St. Romain, S.
A-029
Sturgeon, C.
A-064
Sturk, A.
B-314, B-344
Su, I.-J.
B-206
Su, Z.
B-210
Suarez-Artacho, G.
B-196
Suffin, S.
A-264
Sugimoto, H.
B-058
Suh-Lailam, B.
A-269
Suleymanlar, G. A-112, A-126
Sullivan, S.
A-389, B-409
Sullivan, T.
A-117
Sultan, F.
B-447
Summers, M.
B-337, B-381
Sun, J.
A-302
Sun, W.
B-351
Sun, X.-L.
A-317, B-100
Surtihadi, J.
B-040
Suzuki, H.
A-233, B-368
Suzuki, R.
B-108, B-113
Suzuki, S.
A-162
Svinarov, D. A.
A-418
Swe, S.
A-020
Sydlowska, M.
B-316
Sykes, E.
B-039
Szczesniewski, A.
A-388

T
Tabacof, J.
Taboada, G. F.
Tachibana, R.
Taffet, G.
Taira, E.
Takahashi, H.

A-282, A-306
B-248
A-345
B-351
A-357
A-346, B-108,
B-113
Takeda, J.
B-342
Takenaka, T.
A-204, A-207,
A-233
Tamimi, W.
B-298, B-392
Tan, A.
A-334
Tan, C. H. C.
B-030
Tan, C.
B-281
Tan, E. Bin
B-129
Tan, L.-C.
B-157
Tan, M.
A-169, A-334
Tan, Y.
A-334, B-336
Tanaka, I.
A-339
Tang, P. H.
B-403
Tang, W.
B-013
Tange, J. I.
B-422
Tang-Hall, L.
B-336
Taniguchi, N.
B-112
Tapan, S.
A-265, B-380
Tapper, M.
A-109
Tate, J. R.
B-358
Taylor, T.
B-136
Teerakanjana, N.
B-132
Teixeira, W. G.
A-289
Teng, Z.
B-246
Tennill, A.
B-445
Teodoro-Morrison, T.
B-259
Terpos, E.
A-111, A-151,
A-308, B-333
Ttreault, N.
A-273
Thao, A.
B-003
Theodoro-Morrison, T.
B-271
Thielmann, D.
B-049
Thienpont, L. M.
B-127
Thomae, R.
A-251, B-309,
B-343
Thomas, J. V.
B-197
Thomas, J.
B-225
Thompson, M.
A-280
Thompson, R.
A-109
Thongsukkaeng, K.
B-283
Thoren, K. L.
B-432
Thornton, P. S.
A-162, B-281
Thun, M.
A-349
Tirado, L. Karla B-005, B-024
Tnay, I.
A-040
Tognoni, G.
A-146
Tohami, T.
B-006
Tohmola, N.
A-227
Tokgz, B.
B-102
Tokoro, K.
A-349
Tolan, N. V.
B-267, B-311
Tomi-Olugbodi, A. O.
A-068
Tomita, M.
A-187
Toohey, J. M.
B-027
Topal, T.
A-433
Topcu, I.
B-131
Topolcan, O.
A-058, A-372
Torreira Banzas, C.
B-203
Torricelli, F.
A-045
Tort, N.
B-286
Tortorelli, S.
B-272
Toumanidis, S.
B-333
Tovell, K.
A-183
Townsley, C.
B-376

Tozuka, M. B-098, B-114, B-119

Tran, J.
A-180
Tran, N. K.
B-294,
B-295, B-299
Treebuphachatsakul, W. B.
B-283
Tremblay, M.-H.
B-071
Tremblay, S.
B-071
Trenti, T.
A-045
Trieu, A.
B-389
Trojano, P. Gaya Carvalho B-192
Tsai, C.-C.
A-375
Tsai, H.-J.
B-370
Tsai, M. Y.
B-117
Tsironi, M.
A-219
Tsui, F.
B-434
Tsukushi, T.
A-204, A-207,
A-233
Tubbs, K. A.
A-407
Tufik, S.
A-281,
A-295, B-061
Tuncel, A. F.
B-051
Turkeri, L.
A-040
Turner, H. E.
B-345
Turner, L. J.
B-435
Turpeinen, U.
A-227
Turzyniecka, M. J.
A-070
Tyler, C.
B-225
Tzivaras, A.
A-288

U
Ucar, F.
A-113, A-165, A-186
UCA-Villarrica, S.
A-181
Ueda, M.
B-112
Uelker, B.
B-252
Uh, Y.
B-046, B-055, B-080
Umeda, Y.
B-065
Umez-Eronini, A. A.
B-348
Umlauf, E.
B-379
Unlu, A.
A-107, A-130,
A-135, A-391
Unsal, A.
A-134
Urek, K.
A-420, B-367
Uren, N.
A-040
Usami, Y.
B-119
Usui, S.
B-122
Uta, C
B-102
Uvirova, M.
A-170
Uyanik, M.
A-265, B-380
Uzun, G.
A-126

V
Vabbilisetty, P.
Vaithlingam, S.
Vajapey, U.
Valcour, A. A.
Valdes, R.
Valdes Jr, R.
Valdez, J. J.
Vallone, P. M.
Vanavanan, S.

B-100
A-035
B-040
A-122
A-041, B-264
A-057
B-428
A-090
B-093, B-123,
B-132, B-237
Vandell, V.
A-397
Vandendriessche, T.
B-324
VandenHeuvel, K. A.
A-263
Van Der Gugten, G.
A-400
Van Groll, E.
B-308
Van Hoof, V.
B-324
van Kerckhoven, S.
B-324
Van Natta, K.
A-414
VanSlambrouck, C.
B-165
van Straalen, J. P. B-314, B-344

S273

AUTHOR INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Van Uytfanghe, K.
B-127
Vardanyan, S.
A-358
Vargas, G.
A-129
Varnamkhasti, M. Hosseini
B-249
Varone, C.
A-225
Vasconcellos, L. S.
B-063,
B-183, B-304
Vasconcellos, L. Souza
B-253
Vasishta, A.
B-069
Vats, N.
A-325
Vaz, J. A. R.
B-143, B-200
Veenis, A.
A-417
Veitinger, M.
B-379
Velasco, D.
A-293, A-335
Velasco, L. F. R. B-228, B-270
Vendt, D.
B-365
Venner, A. A.
B-290
Ventimiglia, F. D.
A-331
Ventura, F. Milton Pinto B-057
Ventura, I. J.
A-055
Ventura, L.
B-435
Vergara, R.
A-218
Verna, J. A.
A-331
Vesper, H. W.
A-220
Viana, M. Anesia
A-080
Vicente, G.
A-142, A-430
Vicua, A. G.
A-055
Vidal-Folch, N.
A-226
Vieira, J.
B-220
Vieira, S. Duque Silva
B-001
Vigil de Mello, S. V. G. A-435
Vilanova, B. Costa
B-064
Vilario-Garca, T.
B-255
Vilhena, L. S.
A-382
Villamandos, V.
B-139
Villarrubia, J.
A-293, A-335
Virani, S.
B-351
Virizuela, J. A.
A-052
Vis, M.
B-344
Vis, M. M.
B-314
Visnick, H.
B-312
Vlagea, A.
A-293, A-335
Volanski, W.
A-275
Voma, C.
A-217
von Recum, J.
B-282
Voong, M.
A-334
Vorlat, A.
B-324
Voskaridou, E.
A-288, B-371
Voskoboev, N.
B-444
Vrints, C.
B-324
Vrunski, R.
B-019
Vural, C.
A-216

W
Wadams, H.
A-190
Wagar, E.
A-029, B-175
Wagar, E. A.
A-246, A-247
Wahed, A.
B-208
Waite, G.
A-221, B-031
Walker, K.
B-281
Wallace, F.
B-210
Walters, T.
B-007
Walters, W.
A-182
Waltrick, D.
A-159, A-163,
A-173, A-175, A-196, A-209
Wan, B.
B-259, B-271
Wang, C. A-053, A-053, A-302
Wang, C.-X.
A-050, A-063
Wang, F.
A-006, A-426
Wang, H.
A-034, B-073
Wang, H.-Y.
A-044, A-060,
B-046, B-055, B-080, B-195

S274

Wang, J.
Wang, L. Qi
Wang, L.

B-316
B-307
A-050, A-063,
B-358, B-387
Wang, P.
A-424
Wang, S.
A-081, A-105,
A-255, A-378, A-412, A-413,
A-425, B-025, B-176, B-230,
B-396, B-398, B-426
Wang, T.
B-179
Wang, W.
A-426
Wang, Y. A-050, A-059, A-128
Wang, Z.
A-046
Ward, B.
A-301
Warnick, G.
B-117
Watanabe, M.
A-150
Watanabe, Y.
A-160
Waynick, J.
B-163
Wei, J.
A-358
Wei, T. Q.
B-245, B-246
Weinerman, S.
A-184
Weinstein, D.
B-006
Weiss, E.
B-006, B-015
Welsh, K. J.
B-208
Welsh, M.
B-167
Wen, C.
B-434
Wesenberg, J. C.
B-290
Wessel, B.
B-246
Westgard, S.
B-126, B-154,
B-160
Wheeler, A. P.
A-096, A-280
Wheeler, S. E.
A-062
Whigham, K.
A-239
Wiebe, D.
A-380
Wiencek, J.
B-025
Wildeboer, S. E.
A-438
Wilkenson, C.
B-271
Williams, B.
B-311
Williams, K.
A-238
Williamson, E. E. A-131, A-133
Willmon, C.
B-452
Willrich, M. A. V. A-374, B-267
Wilmik, D.
B-031
Wils, J.
A-030, A-031
Wilson, D. H.
B-012
Wilson, D. P.
A-162
Wilson, R.
B-068
Winn, L. C.
B-022
Wiriyaprasit, R.
B-043
Wisniewski, J.
A-283
Wisniewski, J. L.
A-438
Witt, J. K.
B-133
Wittliff, J. L.
A-042
Wittwer, C.
B-034
Wittwer, C. A.
A-279
Wockenfus, A. M.
A-266
Wohlstadter, J. N.
A-035
Wolak-Dinsmore, J.
A-149
Wolfe, R. R.
B-242
Wolken, J. K.
A-380
Wolsky, R.
B-165
Won, D.
A-254
Wonderling, R.
B-011, B-154,
B-164
Wong, C.
A-108, A-137,
B-003, B-016
Wong, J. S. Y.
B-030
Wong, M.
B-205
Wong, S. L.
A-400
Wong, S.
A-417, B-419
Wongwaisayawan, S.
B-132
Woo, H.-Y.
B-014, B-062
Woods, A.
A-167

Woodworth, A.
Woolsey, B. L.
Worster, A.
Wright, D.
Wu, A. H.B.
Wu, C.
Wu, H.-C.
Wu, L.
Wu, S. Y.
Wu, T.
Wu, Y.-L.
Wu, Z.
Wyer, L.
Wyness, S. P.

A-096, A-154,
A-280
B-281
B-319, B-353
A-301
B-201, B-408,
B-432
A-053
B-206
B-073
A-302, B-416
A-034
B-439
A-420
B-303
A-223

X
Xia, F.
Xiang, Q.
Xie, L.
Xu, C.
Xu, J.
Xu, S.
Xu, W.
Xu, Y.

A-172, B-073
A-092
A-100
B-312
A-366
B-073
A-426
A-358

Y
Yadak, N.
Yadav, A.
Yadav, N. K.
Yadzi, E.
Yago, H.
Yago, M.
Yakubek, D.
Yalcindag, A.
Yamada, T.
Yamada, Y.

A-110, B-238
A-106
A-106
B-031
A-346, B-280
B-139
A-261
A-113
B-112
A-204,
A-207, A-233
Yamaguchi, H.
B-065
Yamaguchi, I.
B-072
Yamaguchi, K.
B-058
Yamamoto, M.
B-280, B-434
Yaman, A.
A-236, A-313
Yamauchi, S.
B-058
Yamout, B. I.
A-367
Yan, W.
A-426
Yang, C.
A-172
Yang, H. S.
B-201
Yang, J.
A-153
Yang, J.
A-358
Yang, Q.
A-053
Yang, Y.
A-050, A-063
Yarbrough, M. L.
B-301
Yasmin, R.
B-037
Yatomi, Y.
B-114
Yavuz Taslipinar, M.
A-113,
A-186
Yazdanpanah, M.
B-259
Yazici, C.
B-102
Ybabez, R.
B-034
Ye, J.
A-387, A-397, B-211,
B-397, B-406
Ye, L.
B-003, B-407, B-413,
B-417, B-423
Yeh, C.-H.
B-452
Yeh, Y.-M.
B-206
Yen, K.
A-102
Yen, P.
B-434
Yeo, C. Pin
B-030
Yeo, C.
B-018

Yeo, K. Teck Jerry

A-198,
B-165, B-249
Yesildal, F.
A-243
Yi, X.
A-198, B-165, B-249
Yildiz, Z.
A-113
Yilmaz, G.
B-051
Yilmaz, V. Taner
A-126
Yilmaz, V. T.
A-361
Yim, H.
B-180
Ylmaz Demirtas, C.
A-256
Yokoba, M.
A-204, A-207,
A-233
Yokobata, K.
B-016
Yokokawa, S.
B-280
Yokota, Y.
B-121
Yoo, H.-J.
B-240
Yoo, J.
A-292
Yoshika, M.
B-108, B-113
Yoshimoto, A.
B-098, B-114
Yoshimura, T.
A-339
Yoshioka, K.
A-069
You, E.
A-285
Younes, S. E.
A-072
Young Jae, K.
A-326
Younis, A.
B-318
Youssef, M.
B-430
Yu, H.
B-054, B-446
Yu, H.-Y.E.
B-390
Yu, R.
B-206
Yu, S.
B-062
Yu, T.
B-189
Yuan, C. B-230, B-396, B-426
Yuan, S.
A-172
Yuan, Z.-X.
A-406
Yuanyuan, S.
A-025
Yuasa, S.
A-069
Yue, B.
A-232
Yueping, W.
A-025
Yundt-Pacheco, J. C.
B-150
Yuzawa, Y.
B-342

Z
Zaghlool, R.
Zaidman, V.
Zanardi, V.
Zaninotto, M.
Zaro, M. J.
Zblewski, A.
Zeller, T.
Zellner, M.
Zeneli, L.
Zeruya, E.
Zgela, M.
Zhang, C.
Zhang, L.
Zhang, N.
Zhang, P.
Zhang, W.
Zhang, X.

A-140
A-245
B-385
A-147
B-087
B-034
B-252
B-379
A-018
A-193
A-167
A-053
A-255, B-426
B-336
B-206
A-046
A-050,
A-063, A-338
Zhang, Y.
A-006, A-036,
A-172, A-262, A-358
Zhao, A.
A-358
Zhao, Z.
A-277, B-445
Zheng, G.
A-050,
A-063, A-063

KEYWORD INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
1,25D
A-180
22q11.2 syndrome
B-191
25-hydroxy vitamin D3
A-107
25-Hydroxy vitamin D A-262, B-237, B-253
25-OH Vitamin D
B-252
25OHD
A-223
3gAllergy
A-359
4-Phenyl butyric acid
A-366
46,XX male
B-185
5-FU
B-418
5-nitro-5-methyl-BAPTA
A-279
6-monoacetylmorphine
B-426

A
A1A clearance
B-386
A1c
A-110
abacavir hypersensitivity reaction
B-201
Absolute Quantification
B-215
Accelerator APS
B-170
ACCU-Chek Inform II
B-277, B-303
Accuracy
B-306
Acetaminophen
A-257
acetaminophen, salicylic acid
A-389
acetate buffer
B-426
aCGH
B-191, B-193, B-203
Active-B12
B-244
Acute coronary Syndrome
B-108, B-343,
B-354, B-360
Acute heart failure
B-331
acute kidney injury
A-131, A-133, A-151
Acute myeloid leukemia
B-208
Acute rejection
A-128
Acylation reaction
B-370
AD modulated proteins
B-379
Add-on
B-156
Addison disease
A-218
Adipocytokines
A-219
Adipokines
A-199
ADMA
A-378, A-413
adsorption
B-433
Adults
B-221
Adverse events
B-162
ADVIA Centaur
A-152
ADVIA Chemistry
B-094
Adiponectin
A-176
age
B-314
Age Groups
B-253
Ageratum conyzoides L.
A-435
ages
A-302
AKI
A-147
AL-amyloidosis
B-333
albumin
A-202, A-276, A-356, A-371
albumin to creatinine ratio
B-289
albuminuria
B-383
Aldolase
B-025
Aldosterone
A-193
Aldosteronism
A-232
allergy
A-324, A-357
Allowable total error (TEa)
B-160
Alpha 2-macroglobulin
A-343
Alpha-1-antichymotrypsin
A-001
Alpha-1-Antitrypsin
B-372, B-386
ALT
B-364
Alternative Calibration
A-416
Alzheimers disease
A-238, B-373, B-379
AMH
A-171, A-224, A-229
Amino acid analysis
B-269, B-359
Amniotic Membrane
A-119
Amphetamines
A-384
Amyloid peptide 1-40
B-373
ANA
A-327
anaemia
A-084
Analyser
A-024

Analytical Evaluation
B-330
Analytical Measurement Range (AMR) A-175,
A-196
analytical performance
A-160, B-148
Analytical quality
B-160
Analytical Quality Control (AQC)
B-129
analyzer
A-108, A-137
Androgens
A-212
Androstenedion
A-212
androstenedione
A-200, A-415
anesthesia personnel
A-216
aneuploidies
B-212
Angiotensin II receptor blocker
A-204
animal CBC
A-428
animal reference ranges
A-428
Ankylosing spondylitis
A-142
Anti -thyroid peroxidase
A-347
anti-CCP antibodies
A-370
anti-convulsant
B-431
anti-inflammatory action
A-430
anti-inflammatory properties
A-435
Anti-Mllerian hormone
A-224
Anti-Streptolysin O
A-328
Anti-thyroglobulin
A-347
Antibodies-to-Infliximab
A-374
antibody
B-355
anticelular antibodies
B-004
antidepressants
B-424
antiepileptic drugs
B-401
Antifungal
B-428
Antifungals
A-419, B-430
Antigen and Activity
A-319
antigen excess
A-273
Antigen/antibody combination
B-073
Antimicrobial Resistance
B-049, B-080
antimicrobial susceptibility
B-057
antimicrobial therapy
B-080
Antinuclear antibodies
A-368
antinuclear factor test
B-019
antioxidant system
A-433
antioxidants
B-437
antitrypsin
B-378
Apo-A1
B-106
Apo-B
B-106
apolipoprotein E
B-122
apolipoprotein E-containing HDL
B-098
Appendicitis
A-313
Appropriateness
A-211, B-087,
B-134, B-139, B-385
Aquaporins
B-255
AR gene
A-100
archetype
A-259
Architect
B-154, B-244
Arginine
A-413
arginine derivatives
A-081
ascites
A-025
Assay
A-180, B-073
Assay Development
A-374, B-245,
B-246, B-448
assay validation
B-384
asthma
A-102
atherosclerosis
B-109, B-345
aTRAQ
B-269
Atypical
A-306
Audit
A-334
Auto verification
B-132
Auto-antibody
A-162
Auto-immune
A-162
Autoantibodies
A-353, A-365
autoantibody
A-358
autoimmune
A-169, A-334
Autoimmune diseases
A-349
Autoimmune Hepatitis
A-124
Automate 1250
B-015

Automated analysers
B-427
Automated Assay
B-094
Automated IFA
A-349
Automated Indices Reporting
B-031
Automated Multiple Immunoassay
B-047
automated system
B-029
Automation
A-161, B-001, B-005, B-011,
B-017, B-021, B-022, B-023, B-024, B-025,
B-030, B-077, B-135, B-367
Autopsy discordance
B-146
Autoverification
B-018

B
B-type natriuretic peptide
A-116
bacteria
A-253
Bacterial culture
A-144
Basic Metabolic Panel
B-028
Beckman
B-262, B-265, B-354
Beckman Coulter systems
B-376
Benchmarking
B-134, B-139
benign
A-010
benzodiazepine
B-409
Benzodiazepines
B-394
Beta glucan
B-072
bias
A-243
bias error
A-264
Bilirubin Oxidase
A-257
bilirubinometer
B-283
bioassay
A-122
Bioavailable 1,25 diOH Vitamin D
A-174
Bioavailable Vitamin D
A-174
biobank
B-266
Biochip
A-222, B-379
Bioelectronic
B-308
Biologic Variability
B-151
biologic variation
B-317, B-362
Biological Variation
A-250
biomarker
A-050, A-058, A-130, A-394,
A-426, B-104, B-112, B-286, B-366, B-381
Biomarker stability
A-012
Biomarkers
A-151, A-280, A-323,
B-242, B-315, B-333
biomolecule coupling
B-447
biosensor
B-446
Biosensors
B-434
Biotin
B-236
BK virus
A-090
bladder
A-040
blaKPC
B-066
Blood
B-026
Blood cardiac troponin I levels
A-339
blood collection
A-271, A-275
blood donation
A-303
Blood Gas
B-276
blood gas analyzers
B-291
Blood Grouping
B-021
blood nuclease
B-200
Blood sample collection
A-244
Blood/Plasma/Serum
B-452
bloody
A-240
BNP
B-362
Body Mass Index
B-326
body surface area
B-418
Bone
A-203
Bone formation markers
A-142
Branched chain amino acids
A-149
Brazil
A-208, B-143, B-228
breast
A-010
Breast Cancer A-013, A-029, A-042, A-052,
A-057, A-060, A-067, B-206
Bupropion
B-415
Burkholderia pseudomallei
B-057

S275

KEYWORD INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
C
C-peptide
A-160
C-reactive protein
B-315
c16000
A-241
Ca 125
A-018
CA125
A-005, A-048
CA15-3
A-029
Cabergoline
A-166
Caeruloplasmin
A-362
CAIS
A-100
Calcium
A-279, B-182
Calcium Oxalate Crystals
B-181
Calculations
B-095
Calibration
B-025, B-050, B-288
calibrations
A-159
Calibrator stability
B-020
Caliper
B-271
Canadian score
A-099
Cancer
A-023, A-059, B-410
Cancer Biomarkers
B-260
cancer genome
B-189
Cancer population
A-092
Cancer treatment
B-353
Candida
B-063
cap color
B-017
Capability indexes
B-013
carbamazepine
B-417
Carbapenemase
B-049
Carboxyhemoglobin
B-180
cardiac biomarkers
A-148
cardiac troponin A-116, B-316, B-327, B-328,
B-348, B-349, B-352, B-357
cardiac troponin I
B-335, B-339,
B-354, B-358
cardiac troponin T
B-335
Cardiometabolic risk factors
A-214
cardiotoxicity
B-353
Cardiovascular
B-089, B-103
cardiovascular disease A-079, B-093, B-095,
B-104, B-111, B-318
cardiovascular event
B-360
cardiovascular events
B-342
Cardiovascular risk
B-321, B-332
Catalyst
B-370
Catecholamine
A-423
Causes
B-175
CBC
A-069
CD95
A-341
celiac disease
A-364
Cell enrichment
A-041
cell-free DNA
B-200
cells
A-020
cellular bioenergetics
B-438
Centaur
A-167
Central laboratory testing
B-306
centrifugation
B-011
Cerebrospinal fluid ferritin
A-099
certified reference material
A-405
cervical cancer
A-036, A-044
CH50
A-355
Changing QC Lots
B-174
Chemiluminescence
B-004, B-236
chemiluminescent immunoassay
B-384
chemistry
A-252
Chemistry Assays
B-154
Chemokines
A-329, B-041
Chemotherapeutic agents
B-393
Children
B-262, B-265
Chlamydia
B-081
Chlordiazepoxide
B-391
Cholesterol
B-101
Chonic pain
B-410

S276

Chromsystems
B-237
chronic kidney disease A-081, B-335, B-342,
B-371, B-381
Chronic renal failure
A-126, B-102
circulating
A-020
Circulating DNA
B-196
circulating peptides
A-013
Circulating tumor cell
A-041, A-060
circulating tumor cells A-006, A-046, A-057
Cirrhosis
A-072
citrate
A-246, B-200
CK-MB
B-341
CKD-EPI equation
A-111
Cleia
A-187, A-359, B-058
Clinical Analytes
B-308
clinical analyzer
A-253
Clinical Chemistry Laboratory
B-175
Clinical Laboratory
A-256
Clinical laboratory test
A-121
clinical outcomes
A-138
clinical performance
B-062
Clinical Research
A-407
Clinical trial
A-146
clinical value
B-320
clinically useful results
B-159
clonidine, glucagon and ITT
A-177
CMA
B-203
CMV
B-199
CO-oximetry
B-180
Coagulation
A-281, A-283, A-295
Collaboration
B-149
colorectal adenocarcinoma
A-050
colorectal cancer
A-001
Colorectal Cancer (CRC) screening
A-045
Column regeneration
B-254
combination of hsTnI and BNP
B-342
commodities
B-228
commutable frozen serum
B-115
companion diagnostics
A-102
Comparability
A-250
Comparison
A-197, A-281
composition
A-109
Computational medicine
B-008
Connectivity
B-018
contrast-induced nephropathy A-131, A-133
copper-free click chemistry
B-100
Coronary
B-326
Coronary heart disease
B-117, B-351
Corrected Calcium
A-184
correlation
A-152, A-223, A-295, B-352
correlations
A-092
cortisol
A-181, A-217
Corynebacterium diphtheriae
B-043
cotinine
B-345
CRE
B-084
creatinine
A-147, A-394, B-177,
B-178, B-184, B-312
CRF
A-134
Critical
B-124
Critical care testing
B-153
Critical ill patients
B-069
Critical laboratory tests values
B-144
critically ill
A-096, B-299
Cross-over Study Design
B-174
Cross-reactivity
A-213
CRP
A-069, B-296
Crude sample
B-442
CSF
A-238, A-371
CSF biomarkers
B-373
cTroponinT
B-314, B-344
Culture
A-125
Cut off value Analyser
B-316
cutoff
A-153
CVD risk
A-164

cyclosporine
A-108
CYP24A1 Mutations
A-190
Cystatin C
A-068, A-354, B-369, B-376
Cystatin-C
B-326
Cytogenetics
A-289
Cytokines
A-329, A-338

D
d methamphetamine
B-416
D-dimer
A-304, B-280, B-355
DAMPA
A-421
Danazol
A-200
Data Analysis
B-010
Data Modeling
B-152
ddPCR
B-215
deamidated gliadin
A-364
decision matrix
B-167
Decoction
A-437
Deficiency
A-141, B-378
dehydroepiandrosterone
A-412
deletion detection
B-214
Delta
B-347
Delta values
B-346
Dengue
B-045, B-070
dengue hemorrhagic fever
B-045
deprivation
B-345
Desmethyl Lacosamide
B-398
Determine Combo Assay
B-039
developmental disabilities
B-203
Diabete mellitus
A-169
diabetes A-136, A-148, A-182, A-211, A-231,
A-437, B-089, B-298, B-365
Diabetes Control
A-176
Diabetes mellitus
A-106,
A-162, A-183,
A-185, A-204, B-348
diabetes type 2
A-170
Diabetes, Hypertension
A-068
Diabetic Cardiomyopathy
A-366
Diabetic mellitus
A-207, A-233
diabetics
B-383
Diagnosis
A-349, B-441
Diagnostic
B-070, B-071
Diagnostic efficacy
A-336
Diagnostic paths
B-008
Diagnostics
B-308
differential diagnosis
A-025
digital
B-012, B-026
digital PCR
B-214
digoxin
B-405
Dihydroxyvitamin D
A-409
Diiodothyropropionic acid (DITPA)
A-198
Dimension EXL 200
B-007
Dimension(r) EXL
B-246
Dimethylarginines
A-378
Dipeptidyl Peptidase-IV
B-380
Direct HbA1c
A-182
Disclosure
B-162
discriminatory zone
B-267
disease susceptibility
B-438
dissacharidase
B-251
Diurnal variation
A-270
dosing policy
B-405
DRG:HYBRiD-XL
B-252
Dried Blood Spot
A-229, A-420
Dried Blood Spots
B-047, B-372, B-395
Dried urine spots
B-404
drug
A-424
Drug Confirmatory by LCMS
B-425
Drug Exposure
B-415
Drug Monitoring
B-402, B-421
Drug resistance
A-145
Drug Screen
B-400

KEYWORD INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Drug Screening
Drug-metabolizing enzymes
drug/gene delivery
Duplex assay
Dynex DSX platform
dysfunction
Dyslipidemia

B-425, B-432
B-210
B-100
A-347
A-309
B-361
B-103

E
early detection
A-013
early pregnancy
B-267
Ebola
B-068
eclampsia
A-285
ED Critical Care
B-284
EDTA
A-246, A-274
Effectiveness
B-158
Efficiency
B-077
eGFR
A-361, B-312, B-376
Elastase
B-367
Electrochemical detection
B-254
electrochemiluminesence
A-226
electrophoresis A-286, B-111, B-378, B-437
Eletrochemiluminescent assay
A-205
ELISA
A-061, A-283, A-296, A-309,
A-327, A-331, B-337, B-400, B-439
Emergency department
B-134, B-319
Emergency Room
B-149
Emergency Service
A-028
emergency testing
B-296
ENAS
B-004
Endemic areas
A-333
Endoplasmic Reticulum Stress
A-366
Endothelia Progenitor Celle
A-219
Endothelial Dysfunction
A-308
Endothelial Nitric Oxide Synthase
A-140
Enhanced Liver Fibrosis
A-127
Enhanced Liver Fibrosis (ELF)
A-124
Enhanced Liver Fibrosis (ELF) Score A-115
Environmental performance
B-143
environmental strains
B-057
Enzimatic creatinine
B-178
enzymatic
B-336
Enzyme
B-394
enzyme immunoassay
B-407, B-413
enzyme resistance
B-066
Enzyme-Linked Immunosorbent Assay B-038
eosin-5-maleimide
A-293
eosinophil count
B-257
EP 23
B-158
Equi-molarity
B-245
error detection
B-032
errors
A-073
Erythrocyte Sedimentation Rate
A-307
erythrocytes
A-298
erythrocytes morphology
B-033
Erytra
B-021, B-022
Erytroleukaemia
A-306
Escherichia coli
B-054
estimated average glucose
A-186
Estradiol
A-123, A-187, A-220, A-228
Ethaol Stability
A-272
ethyl glucuronide
A-386
ethyl sulfate
A-386
EURACHEM/CITAC guide
A-265
Evaluation
A-045, A-183, B-007,
B-020, B-375, B-377
Evaluation Study
B-369
Everolimus
B-016
Evidence based medicine
B-008
Exercise
A-219
Exocrine Pancreatic Insufficiency
A-086
Expected Ranges
B-398
extraction
A-409

Extracts

A-437

F
factor analysis
B-315
Factor XI deficiency
A-292
Factors
A-281
false positive
A-266, A-332, A-384
fasting plasma glucose
A-186
fasting time
A-284
fatigue
A-098
fatty acids
B-091
fatty liver
A-230
FcGRIIIB gene
A-335
Fecal A1A
B-386
Fecal Biomarker Testing
A-086
Fecal Immunoassay Test (FIT)
A-045
female testosterone
A-399
Ferritin
A-345
fertility assays
A-173
Fetal lung maturity
B-261
FFPE tumor tissue
B-202
FGF21
A-170
fibroblast
B-382
fibroblast-like synoviocytes
A-341
fibrosis
B-079
firefighters
A-181
Fixed Means and SDs
B-150
Flow cytometry
A-293, B-366
Flow Cytometry (FACS)
A-360
Fluorescence In Situ Hybridization
A-289
Fluorescence-based Lateral Flow
Immunoassay(FLFIA)
B-307
fluoride oxalate
A-278
Fluorophosphonate
B-370
FMF
A-313
Folate
B-225, B-247
food-specific IgG
B-228
free light chain A-027, A-039, A-273, A-361
Free light chains
A-023, A-059
free PSA
A-026
Free Testosterone
A-234
Free thyroxine
A-406
Free Vitamin D, 25 Hydroxy
A-174
Freelite
A-024, A-321
freeze-dried
B-435
frequency
A-300, A-324
Friedewald Formula
A-118
fructosamine
A-138
fucosylated fraction of alpha-fetoprotein A-326
Fully automated
A-180, B-076
Functional Sensitivity
A-228
functional stimulus tests
A-155
fungal bloodstream infections
B-063
Fungemia
B-063

G
G6PD
A-269
G6PD Deficiency
A-077, B-197
G6PD variants
B-197
GAD
A-169
Gadolinium
A-279
gammopathy
A-363
gas chromatography
B-414
gas chromatography-mass spectrometry B-272
gastric cancer
A-063
GC-MS
A-187
Gel tubes
A-038
gender
B-314
Gene knockout
A-431
gene mutations
A-004
gene polymorphisms
A-112
genetic incompatibilities
A-258

genetic test
B-168
Genomic signature
B-196
genomics
A-067
genotyping
B-064, B-192, B-210, B-442
geriatric
A-252, B-084, B-124
Gestational diabetes
B-255
GFR
B-184
Ghrelin
A-165, A-206
Glicated hemoglobin
A-185, B-005, B-024
Glioblastoma multiforme
A-058
Glomerular filtration rate
A-087, A-105
Glucagon
A-213
Glucocorticoid
A-422
glucometer
B-281, B-288, B-306
glucose
B-298, B-300
Glucose assay
A-432
Glucose hexokinase
A-432
glucose meter
B-277, B-287, B-304
glucose meters
B-299
glucuronide
B-403
Glutaric Aciduria
A-070
glycated albumin
A-138
glycated protein
A-183
glycosaminoglycans
A-375
Glycosylated Hemoglobin
B-101
Graft cfDNA
B-215
graft function
A-112
Granada ID
B-035
Group B Streptoccocus
B-035
Growth deficiency
A-210
Growth differentiation factor-15
B-360
Growth hormone
A-177
Growth hormone stimulation test
A-210

H
H-FABP
Haptoglobin
Harmonization
Hb A1c
Hb Variants
HbA1c

B-337
A-294
B-276
A-154
A-154, A-267
A-182, A-186, A-201,
A-211, A-267, A-268, B-365
HbA2
A-318
HbF
A-201, A-318
HBsAg
B-058
HCG
A-016, A-222, B-290, B-297, B-301
hCG beta core fragment
B-297
HCV
A-072
HCV genotyping
B-040
HDL
B-112
HDL Subfractions
B-105, B-121
HDL-cholesterol
B-122
HDL2
B-105, B-121
HDL3
B-105, B-113, B-121
HDL3 cholesterol
B-108
HE4
A-005, A-048
Head and neck cancer
B-221
Health Care System
B-043
Health screening
A-121
heart failure
A-140, B-343
Heavy chain/ light chain immunoassay A-049
hematocrit
B-295
Hematological parameters
A-280
Hematology
B-292, B-293
hematology flags
A-310
hematuria
A-240, B-033
Hemodialysis
B-298
Hemoglobin
A-110
Hemoglobin A1C
B-311, B-445
Hemoglobinopathies
B-371
hemoglobinopathy
A-300
Hemolysis
A-241
Hemolysis, Icterus, Lipemia
B-031

S277

KEYWORD INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
hemolytic index
A-254
Hemophilia C
A-292
hemorheology
A-307
hemorrhagic
B-068
Heparin Binding Protein
B-069
Heparin Induced Thrombocytopenia
A-309
Hepatitis A
A-077
Hepatitis B
B-051, B-078
Hepatitis C
B-040, B-082, B-392, B-439
hepatitis C active infection
B-059
hepatitis C treatment
B-192
hepatocellular carcinoma
A-326
Hepcidin
A-296, A-427
HER2 Testing
B-206
Hereditary spherocytosis
A-293
hevylite
A-055, A-321
high density lipoprotein
B-114, B-119
High fat diet
B-091
high molecular wight fibrin degradation
products
B-355
high resolution
A-390
high resolution mass spectrometry
B-432
high sensitive
B-320
high sensitive CRP
A-103
High Sensitive Troponin
B-330
High sensitivity
A-345, B-352
high sensitivity cardiac troponin
A-148
high sensitivity troponin
B-325, B-328
High Sensitivity troponin T
B-351
high throughput
A-388, B-429, B-430
high-performance liquid chromatography
(HPLC)
A-383
High-Sensitivity Cardiac Troponin
B-319,
B-353
high-sensitivity troponin
B-341
High-sensitivity troponin I
B-309
highly sensitive assay reagent
A-339
HILIC
A-423
histamine
A-402
HIT
A-332
HIV
B-064, B-073, B-205
HIV-1 p24 Antigen
B-039
HIV-1/2 Antigen/Antibody
B-039
HLA Class II
A-367
HLA-B*5701
B-201
Hodge Test
B-066
Hoffman
B-182
Holotranscobalamin
B-244
HOMA - beta
A-192
Homocysteine
B-320, B-336, B-433
Homogeneous
B-446
hook effect
A-273
Hormone
A-052, A-236
Hormones
A-245
hospitalization
B-242
HPLC
A-417, B-122, B-223, B-230,
B-235, B-383, B-392, B-396, B-403
HPLC_MS/MS
B-269
HPV
B-195, B-207
HPV E6/E7
A-044
HRMS
A-392
hs-troponins
B-338, B-361
hsTnT
B-282, B-324
hTERT
A-044
Human Allergen
A-340
Human cervical cancer
B-195, B-207
human chorionic gonadotropin B-267, B-301
Human papilloma virus (HPV)
B-062
human serum
B-223
hyaluronic acid
B-079
Hydrolysis
B-394
Hydroxocobalamin
B-180
Hydroxyurea
B-414
hypercalcemia
A-184

S278

Hypercholesterolemia
A-248
Hyperemesis gravidarum
A-165
Hypergammaglobulinemia
A-031
Hyperglycemia
B-387
Hyperhomocysteinemia
B-102
hyperoxaluria
B-444
Hypertension
A-136
Hypertriglyceridemia
B-110
Hypervitaminosis D and Hypercalcemia A-190
hypocalcemia
A-246
hypoglycemia
B-287
hypoglycemic
B-281
Hypothyroid
A-164
Hypoxia-inducible factor-1
A-207

I
i-STAT
B-290
ICAM-1
A-063
icodextrin
B-291
identify
B-035
IDMS
A-405
IDS-iSYS
A-221
IFA
A-218
IgD biclonal
A-363
IGF-1
A-011
IGF1
A-197
IgG avidity
A-330
IgG1
A-350
IgG2
A-351
IgG3
A-337
IgM
A-373
Il 6
A-098
IL-18
A-072
illicit drugs
B-409
Imaging
B-026
Immature Platelets
A-315
Immulite 2000
A-221
Immulite 2000 XPi
A-359
Immune Status
B-166
immunoassay
A-056, A-123, A-161,
A-172, A-193, A-213, A-222, A-224, A-228,
A-384, B-012, B-050, B-082, B-225, B-238,
B-367, B-369, B-421, B-427
immunoassay evaluation
B-249
Immunoassay interference
A-198
immunoassays
A-159
immunoassays systems
A-163
ImmunoCAP ISAC
A-357
immunocapture-LC-MS/MS
A-400
immunofixation
A-039, A-286
Immunoglobulin
A-023, A-059, A-438
immunology
B-015
Immunomagnetic
A-041
immunosuppressant
A-108, B-412
Immunosuppressants
B-395, B-419
Immunoturbidimetric
A-294, B-356
Immunoturbidimetric assay
B-321
Impact on pay-for-performance compliance
B-311
imprecision
A-243, A-264
Improve efficiencies
B-170
Improvement
B-133
improving proccess
B-006
inborn error of metabolism
A-070
inborn errors of metabolism
B-259, B-272
Indices
A-129
indirect immunofluorescence assay
A-368
indoors
A-216
Infant
B-234
Infants
A-103
infarction
A-287
Infection prevention
B-086
Infectious Disease
B-074

Infectious diseases
B-069
Inflammation
A-316
inflammatory
A-341
Inflammatory biomarkers
A-372
Inflammatory markers
A-120
Infliximab
A-374
Influenza virus
B-075
Informatics
B-027
Innovative diagnosis
A-333
Instrument interface
B-009
Instrument Manager
B-132
Instrumentation
B-449
Insulin
A-225, A-401, A-407
insulin analogues
A-400
Insulin resistance
A-164
Insulin Signaling
B-387
insulin-growth-factor
A-390
insulin-like growth factor-1
A-221
Insulin-Like Growth Factor-I System A-199
Intact protein mass spectrometry
A-404
Integrated Medical Analytics
B-152, B-173
Interference
A-129, A-201,
A-257, A-267, A-277
Interferences
B-184
Interleukin-10
A-431
interleukin-18
A-112
Interoperability
B-009
Invasive fungal infection
B-072
Inventory Manager
B-164
Invokana
A-231
ion chromatography
B-444
IPF
A-315
Iron
A-296
iron deficiency
A-297, A-298, A-303
Iron deficiency anemia
A-345
Irritable Bowel Syndrome
A-086
ischemic heart disease
A-304
ISO 14000
B-143
Isocitrate-dehydrogenase
A-058
isotope dilution
B-359
ITT
A-155
IVD
B-292, B-293

J
Jaffe
Joint Infection
Jungia sellowii Less

B-178
A-125
A-430

K
kappa
Keto derivation
kidney
kidney diseases
Kidney Function

A-027
A-425
A-203, A-361
B-272
A-105

L
lab test utilization
laboratory
laboratory investigation
Laboratory Value Proposition
Lacosamide
Lactate
Lactation
Lactic Acid, Procalcitonin
lactose tolerance
lambda
Lamellar body counts
lamotrigine
Lassa
Lateral flow
LC MS/MS

B-362
A-073
A-247
B-152
B-398, B-431
A-255, A-278
B-415
A-092
B-209
A-027
B-261
B-431
A-143
A-005, B-286
A-062

KEYWORD INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
LC-HR-MS/MS
B-393
LC-MS/MS
A-054, A-232, A-387, A-396,
A-399, A-410, A-417, A-418, A-420,
A-424, A-427, B-237, B-247, B-397,
B-402, B-404, B-406, B-422
LC-MS/MS vitamin D
B-249
LC/MS
A-377
LCMSMS
B-416
LDL cholesterol
A-118
LDL subpopulation
B-123
LDL-C
B-095
LDTD--MS/MS
A-394, B-429
lean body mass
B-242
lean six sigma
B-163
Leg Ulcers
A-119
leptin
A-063, A-225, A-230
leukocyte
B-382
leukocytes
B-438
LH-receptor
A-189
LHCGR
A-189
Libya
B-101
Light-sensitivity
B-391
line probe assay
B-040
Lineage Leukemia gene
A-289
linearity
A-173, A-209
Linearity study
A-196
Lipase
B-377
Lipemia
A-241, A-284
lipid hydroperoxides
B-098
Lipid metabolism laboratory test
B-087
Lipids
B-091, B-332
lipocalins
A-079
Lipolysis
B-387
Lipoprint
B-099
Lipoprotein (a)
B-097
lipoprotein A
B-089
Lipoprotein subpopulation
B-093
Lipoprotein(a)
B-117
liposome
A-355
liquid chromatography
B-401
liquid chromatography-tandem mass
spectrometry (LC
A-383
liquid enzymatic method
B-179
Liquid stable reagents
A-432
LIS
B-018
liver
B-079
Liver Fibrosis
A-115, A-124, A-127
Liver function
A-431
Liver Transplant
B-196
Liver transplantation
A-128
Locked nucleic Acid
B-211
logistic regression model
A-025
long-term storage
A-245
Loop-mediated isothermal amplification B-211
low sensitivive CRP
A-103
LSMAD analysis
B-299
LTIA
A-346
Luminescence resonance energy
transfer
A-123
lung cancer
A-034, A-035
lutropin
A-226
lysosomal storage disorder
B-382

M
M-spike
machine learning
Magnesium
Magnetic
magnetic beads
Magnetic beads as labels
malaria
Male lineage
Malignant pleural effusions

A-276
B-032
A-217
B-446
B-447
B-279
A-084
B-270
A-017

maltose
B-291
maltose interference
B-277
management
B-023, B-128
marker
A-010
Mass Spectrometry
A-066, A-107, A-200,
A-212, A-262, A-376, A-380, A-382, A-389,
A-390, A-401, A-403, A-407, A-408, A-409,
A-410, A-421, A-425, A-427, B-118, B-238,
B-395, B-408, B-410, B-414, B-424
maternal
B-266
Maternal plasma
B-270
mathematical modeling
A-298
matrix interferences
B-448
Mean platelet volume
A-285
Meaningful use
B-009, B-130
Measuring Reproducibility
B-010
mecA Gene
B-046
Medical Checkup Examinees
A-339
Medical error
B-162
Megalin
A-204
Meritas
B-327
Mesenchymal Stem Cells
A-119
Meta-Calculation
A-414
metabolic syndrome
A-106, A-136, A-188,
B-110, B-123, B-329
Metabolically High Risk
A-199
Metabolomics
A-288, B-075
metanephrine
A-397
metanephrines
A-227
metformin
A-170, B-248
Methadone
B-390
Methadone and EDDP
A-387
methicillin-resistant Staphylococcus aureus
B-086
method comparison
A-234, A-307, A-383,
B-127
Method evaluation
B-311, B-445
method performance
B-126
method validation
B-252, B-318, B-449
Methodology
B-005
methotrexate
A-421, B-421
methylarginine
A-135
Methylated arginines
A-391
MGUS
A-321
microalbuminuria
A-202, B-289
microassay
B-336
microbilirubin
B-283
Microbiology
B-084
microdeletions
B-188
microduplication
B-191
microfluids
B-072
microRNA
A-050
microscopy
A-095
Middleware
B-132
minerals
A-109
Minimal Residual Disease
A-066
MIP-1 alpha
B-381
MiR-28-5p
A-053
miRNA
A-034, A-042
miRNA panel
A-036
MMP-3
A-346
MMRV
B-076
MMRVT
B-047
Modified Control Materials
B-010
Moesin
A-358
Molecular Amplification
B-037
molecular diagnostic
B-029, B-074, B-077,
B-081
molecular marker
A-007
monitoring
A-049, A-055
monitoring patient data
B-127
monoclonal antibodies
A-325, B-307
monoclonal component
B-385
monoclonal gammopathy
A-039

monoclonal gammopathy of undetermined

significance
A-087
morphine
B-426
Mortality
A-126
Mortality prediction
B-331
mouse model of pleurisy
A-430
mRNA
B-195, B-207
mucopolysaccharidosis
A-375, B-380
multi-constituent controls
A-260
Multi-qualitative
B-441
Multiple Instruments
B-150
Multiple Myeloma
A-024, A-028, A-038,
A-047, A-049, A-066, A-151
Multiple Sclerosis
A-367
Multiplex
A-035
Multiplex assay
A-372
Multiplex PCR
B-037
Multiplex Technique
A-353
Multiplexed Immunoassay
B-076
mutation
A-100, A-292
mycophenolate
A-137
mycophenolic
A-137
mycophenolic acid
B-403
Myelin Basic Protein
B-384
Myelodysplasic syndromes
A-306
myelodysplastic disorder
A-335
myeloma
A-055, A-111
myeloperoxidase
B-119
Myocardial infarction B-327, B-328, B-346,
B-347, B-349, B-357

N
N-terminal B-type natriuretic peptide B-348
NADPH
A-217
nano-bio interaction
B-433
national survey
A-051
NDM
B-049
Neonatal Screening
A-208, B-197
Neonate Abstinence Program
B-425
Neopterin
A-130, A-331
nephelometry
A-286
nephropathy
A-071
Nesfatin1
A-165
Neurological disease
A-334
Neutral
A-422
Neutrophile
A-313
newborn management
B-283
newly-developed system
A-297
next generation
A-328, A-337, A-343,
A-350, A-351, A-354,
A-355, A-356, A-362, A-373
Next generation sequencing
B-202
next-generation sequencing
A-090, B-078
NGAL
A-134, B-371
Nitric Oxide
A-288
NMR
B-104, B-111
NMR spectroscopy
A-149
Non esterified fatty acids
A-192, B-179
Non-alcoholic fatty liver disease
A-101
non-HDL cholesterol
B-103
non-small cell lung cancer
A-004
Nontreponemal antibodies
B-038
Normal protein electrophoresis
A-282
normetanephrine
A-397
notification of critical values
B-144
Nova StatStrip
B-287
Novel Anticoagulants
B-422
NSTEMI
B-309
Nucleic Acid Testing
B-205
NYHA class
B-329

S279

KEYWORD INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
O
Obesity
Obesity related factors
Octreotide LAR
OneLab
online SPE
opiates
opioids
osteoporosis
Outbreak
Outcome
ovarian cancer
Ovarian Carcinoma
Ovarian Reserve
Over utilization
oxalate
Oxidative stress
oxidized LDL
Oxidized lipoprotein
ozone therapy

A-214, B-113
A-215
A-166
B-164
B-430
A-403, B-404
A-380
B-235
B-054
A-146
A-012, A-061
A-018
A-171, A-229
B-349
B-444
A-207, A-233,
A-433, B-318
B-109
B-112
A-433

P
Pacients age
Paclitaxel
pain
pain management

A-282
B-389
A-424
A-380, A-396,
B-390, B-406
Pain Management Drugs
A-410
Pain Medications
B-210
Pancreas
B-377
pancreatic cancer
A-006
panel
A-408
Panic values
B-006
paper spray
B-412
Paraoxonase (PON1)
B-102
Paraprotein
A-038
parasites
A-084
parathyroid hormone
A-235, B-253
paroxysmal nocturnal hemoglobinuria A-080
paternity test
A-258
PATHFAST
A-251
PATHFAST cTnI
B-309
Pathogen
B-054
Patient Care
B-171
patient data
B-317
Patient Moving Averages
B-034
Patient Specimens
B-303
Patient-based quality control
B-028
Patients with Syphilis
B-038
patterns
A-006
PCA3 gene
A-007
PCR
B-061, B-071, B-435, B-452
pediatric
A-117, A-129, A-167,
B-259, B-260, B-281
pediatric reference intervals
B-271
Pediatric Reference Range
A-301
percutaneous coronary intervention
B-324
performance
A-172, A-331, B-007, B-300
Performance Evaluation
B-303
periostin
A-102
Personalized medicine
B-189
Pharmacokinetic
B-389, B-418
pharmacokinetics
A-401
phenobarbital
B-423
Phenotype
B-372
phenytoin
B-407
PHI
A-016
phlebotomy
A-271
phospho-lipase A2
A-135
phospholipids, cholesterol lipid
B-100

S280

PIFA
A-332
PINP
A-203
Placenta
B-255
plasma
A-397
Plasma cell morphology
A-282
plasma seperation tube
A-254
Plasma total homocysteine
A-188
plasmatic glucose
A-155
platelet
A-287, A-302, A-358, B-366
Platelets
A-128, A-310, A-316
pleural fluid
A-017
PLP
B-250
pneumatic tube system
B-163
PNH
A-335
POC
B-292, B-293, B-297
POCT
B-153, B-280, B-282, B-300, B-312
point of care
B-068, B-339
Point of Care Testing
B-284
Point-of-care
B-276, B-294, B-295, B-434
point-of-use assay
A-325
Polyclonal Hypergammaglobulinemia A-030
Polycystic Ovary
A-171
Polycystic ovary syndrome
A-120
Polymerase Chain Reaction
B-044
polymorphism
A-040, A-140, A-150
positivity rates
B-167
Post-Analytical
A-414
post-translational modification
A-001
Postnatal
B-193
potassium
B-032
Povidone
A-277
pre eclampsia
A-323
pre-analitycal phase
A-275, A-284
Pre-analytic errors
A-269
pre-analytical
A-238, B-135
pre-analytical error
A-244
Pre-Clinical
A-438
Pre-eclampsia
A-285
preanalytical
B-017
preanalytical validation
A-227
Precision
B-136, B-288
precocious puberty
A-226
Precursor
A-422
Prediabetes
A-206
Predictive
A-126
pregnancy
B-240, B-266, B-301
pregnant women
B-044, B-081
Presepsin (sCD14-ST)
A-251
prevalence
A-087, A-106, B-061
Primary Biliary Cirrhosis
A-115, A-248
Primate
A-438
Private Laboratory
B-227
Pro-inflammatory mediators
A-142, A-435
procalcitonin
A-120, B-257, B-307
Process capability
B-013, B-169
Process optimization
B-031
productivity
B-023
proficiency test
B-115
proficiency testing
A-268, B-043, B-146
Proficiency Testing Schemes
B-151
prognosis
A-067, A-304
Prohibited Substances
A-376
proinsulin
A-160
prolactin
A-152
propoxyphene
B-167
prostaglandin
A-402
prostate cancer
A-011, A-026,
A-046, A-051, B-229
Prostate specific antigen
B-279
prostate-specific antigen
A-051
prostatic cancer
A-007
protease
A-236
Protein C
A-317, A-319
Protein S
A-295

protein-free antibody stability

Proteomics
PSA
pseudohyperkalemia
Pseudohyponatremia
Psoriasis Area Severity Index
PTH
PTX-3
Pulmonary disease
Pulmonary Embolism
pure water
pyridoxal 5-phosphate
Pyrogallol Red
pyrosequencing
Pyruvate

B-448
A-404
A-026
A-247
A-248
A-107
A-184
A-134
B-067
A-130, A-135
A-253
B-250
A-276
A-004
A-255, B-176

Q
Q-Exactive
A-412
QC Rule Design
B-150, B-169
qc rules
B-136
QMS
B-003, B-016, B-419
qPCR assay
A-046
quality
B-126, B-133, B-158
Quality assurance
B-027, B-325
Quality control
B-013, B-050, B-136
Quality control management
A-163
Quality Control Material
A-340
Quality Control Plan (QCP)
B-129
Quality Controls
A-260
Quality Improvement
B-171
Quality Management
A-256, B-127, B-146
QuantiFeron
B-166
Quantimetrix
B-099
Quantitation
A-391, A-392
Quantitative
B-441
Quantitative lateral flow test
B-279
Quantittaive Fluorescent PCR
B-212

R
race/ethnicity
B-117
random
B-182
random error
A-264
RAP1B
A-053
rapid
B-435
rapid assay
A-325
Rapid molecular screening
B-086
rapid serum tube
A-235, A-266
Rapid Test Strip
A-274
RapidFire Mass spectrometry
B-409
rRCV
A-154
Re-engineering Laboratory Process
B-165
ready to use reagent
A-294
real time PCR
B-199, B-209
Real-time PCR
B-046, B-442
Real-time Taqman Probe assay
B-041
REBA MTB-XDR
A-145
recombinant protein
B-359
Reference Interval
A-168, A-252, A-269,
A-318, A-338
reference intervals
A-117, A-378, B-251,
B-259, B-260, B-262, B-265
reference limit
B-344
reference material
A-090, B-247, B-358
Reference range
B-183
reference ranges
A-302
Reference value
B-001, B-179, B-364
regression
A-261
regulatory T cells (Treg)
A-360
relevance
B-128
renal
B-361
renal cell carcinoma
A-053
Renal Impairment
A-030, A-031

KEYWORD INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Renal injury
A-111
Renal tissue renin-angiotensin system A-233
renal transplant recipient
A-116
Renin
A-193, A-232
Reporting Delays
B-156
Reproducibility of Analysis
B-151
reproduction
A-189
resistance
B-078
respiratory tract infection in children
B-048
respiratory viruses
B-048
respons910
B-365
Reticulated platelets
A-301, A-315
Reticulocyte
A-301
Reverse Blot Hybridization Assay
B-080
RFID
B-164
Rhesus monkey
A-428
rheumatoid arthritis
A-098, A-153, A-336,
A-346, A-365, A-370
RIA
A-218
Ribavarin
B-392
Risk Analysis Framework
B-129
risk factors
A-079
Rituximab
A-404
Rivaroxaban
B-422
RNA-binding protein
A-052
RNA-Seq
B-202
Roche Cobas e601
A-029
ROMA
A-048
Routine Biochemical Parameters
A-265
RT-PCR
B-070, B-201, B-206
RT-qPCR
A-060
Running means
B-028

S
safety
A-073
Saliva
B-393
Salmonella typhi
A-144
Sample collection
A-270
Sample matrix
A-251
Sample Preparation
A-377
Sample size requirements
B-174
Sample Volume
A-272
Saudi Arabia
A-011, B-229
sCD14-ST (Presepsin)
B-331, B-343
Schistosomiasis
A-333
Screening
B-067, B-082
SDMA
A-105, A-413
Seasonal Variation
A-141
Sebia Capillarys 2
B-445
Semiconductor
B-434
sensitivity
A-406
Sepsis
A-096, A-146, A-280,
B-046, B-257, B-286
Sequencing
B-064
Serious Cardiac Events
B-319
serum
A-036, A-371, A-426, B-437
serum amyloid A
B-114
Serum biomarker
A-035
Serum Creatinine
A-068, A-131, A-133
serum ferritin
A-297, A-303
Serum Free Light Chain
A-031
serum free light chains
A-028, A-047
Serum HE4
A-012
Serum IL-4
B-221
Serum Pentobarbital
B-402
Serum plasma equivalency
A-261
serum seperation tube
A-254
Serum uric acid
A-101
Severe Hyperbilirubinemia
A-077
Sex differences
B-261
Sexually Transmitted Infections
B-044
Sexually-transmitted
B-037
sfit1/plfg
A-323

SGLT2 inhibitors
A-231
shared epitope
A-365
Shistossomiasis Mansoni
A-127
Short stature
A-210
Sialic acid
A-071
Sickle cell anaemia
A-071, A-319
Sickle Cell Disease
A-288
sigma metric
B-148, B-159
Sigma-Metric
B-169
signal transduction
B-189
Signaling
A-225
Simoa
B-012
single cell analysis
A-057
Single Nucleotide Polymorphisms
A-215,
B-192
single spot urine
B-183
Sirolimus
A-382
SIRS
A-069, A-096
Six sigma
B-154, B-160
Six Sigma metrics
B-126
Six-Sigma Methodology
A-256
Sleep Apnea Biomarker
B-264
slide review
A-310, B-157
small dense LDL
B-110
small dense low-density lipoprotein
B-093,
B-123
Small Molecule Analysis
A-416
small-dense LDL
B-113
small-dense LDL cholesterol
B-108
smoking
A-370
SNOMED CT
B-130
SNP
B-209
SNP detection
B-220
SNP genotyping
B-452
solid phase extraction
A-403
SPE
A-423, B-385
Specific IgE
A-340
specific IgE antibody
A-357
Specimen rejection
B-175
specimen submission
B-163
Specimen Type Verification
A-274
Specimen Types
B-074
Specimens
B-130
Spectrophotometry
B-176
sperm count
A-436
sperm motility
A-436
Sphingomyelin
B-118
sPLA2
B-321
Stability
A-255
Stability samples
A-245
Staff Competency
B-157
standard addition
B-416
standardization
B-358
standards
A-259
STAT testing
B-153
statistical protocols
B-159
Steatohepatitis
A-101
steroids
A-408, A-425
stimulation tests
A-177
sTNFRII
A-061
Stopper
A-272
storage
A-236
stress
A-181
stringent complete response
A-047
stroke
A-099, A-316, A-372, B-337
subclass
A-337, A-350, A-351
Subfractions
B-099
Synovial Fluid
A-125, A-338
Synthetic Cannabinoids
B-427
system
A-418
System stability
B-020
systemic mastocytosis
A-402
Systolic Blood Pressure
B-351

T
T2DM
A-426
Tacrolimus
A-382, B-003, B-419
Taiwan
B-439
tandem mass
A-375
Tandem mass spectrometry
A-391, B-240,
B-401, B-412
Tandem-MS
A-414
TAT
B-284
TAT Complex
A-283
TCRG
B-194
TDM
B-389, B-424
Technology
B-135
Technopath
A-260
Telomere Length
A-215
Telomere Shortening
A-214
temperature
B-128
test strip
A-202
test utilization
B-131, B-357
testing efficiency
B-029
Testosterone
A-161, A-220, A-234, A-399,
A-415, A-416, A-436
tetrahydrocannabinol
A-388, A-392
Text analysis
B-027
thalassemia
A-300, B-211
THC
A-388, B-397
THCC
B-429
The Cancer Genome Atlas
B-208
theophylline
B-413
therapeutic drug monitoring
B-396, B-405,
B-407, B-413, B-417, B-423, B-428
Threshold
A-327
thrombin
A-235
thrombin activatable fibrinolysis inhibitor
A-317
thrombomodulin
A-317
thromboxane
B-356
Thyroglobulin
A-054, A-056,
A-062, A-205
Thyroglobulin antibody
A-062
Thyroglobulin autoantibodies
A-054
thyroid
A-117, A-150
thyroid assays
A-209
Thyroid cancer
A-205
Thyroid function
A-188
Thyroid Function Test
A-168, A-250
thyroid hormone
A-172
Timely Chemotherapy Delivery
B-165
tissue transglutaminase
A-364
TK1
A-016
TnI
B-324
TnI assay
B-316
tosyl activated particles
B-447
Total antioxidant capacity
A-018
total cholesterol
A-405
Total Laboratory Automation
B-170
total protein
A-277
total serum bile acids
B-059
Total Testosterone
A-412
toxicology
A-389, B-408
Toxoplasma gondii
A-330
Trace elements
B-229
Traditional Chinese medicine
A-034
Transferrin
B-375
Transfusion Service
B-022
Treg-specific Demethylation Region
(TSDR)
A-360
tricyclic antidepressant
B-396
triglyceride
B-115
Triglycerides
A-118, B-094
Triiodothyronine (TT3)
A-198
Trinidad & Tobago
A-268

S281

KEYWORD INDEX TO ABSTRACTS OF SCIENTIFIC POSTERS, AACC MEETING, 2014

(Numbers refer to Poster Numbers; see pages S2 to S263)
Triple Quad
A-376
trisomy
B-212
tropical and subtropical regions desease B-045
troponin
A-287, B-149, B-282,
B-317, B-325, B-346, B-347
Troponin T
A-266
TSH
A-208
TST receptor
A-150
Tuberculosis
A-145, A-329,
B-041, B-061
tumor
A-020
tumour market
A-017
turbidimetric
A-328, A-343, A-354, A-362,
A-373
turbidimetric immunoassay
B-003, B-016,
B-417, B-423
turn-around-time
A-278
Turnaround time
A-242, A-261,
B-011, B-030, B-131,
B-157, B-171, B-172, B-339
Turnaround Time Improvement
B-165
type 2 diabetes
A-149
Type 2 Diabetes Mellitus
A-176, A-192,
A-230, B-248

U
U.S. Preventive Services Task Force
A-121
UIBC
B-375
ultrafiltration
A-406
Umbilical Cord
B-400
uncertainty
A-243, A-265
uNGAL
A-147
Unified
A-418
UPLC/MS/MS
A-415, A-419
Urinalysis
A-095, A-113, B-014, B-181
urinary electrolytes
B-183
Urinary Free Light Chains
A-030
urinary glycosaminoglycan/creatinine
ratio
B-380
urinary tract infection
A-113
urine
A-240, A-356, A-386,
A-396, B-001, B-356, B-432
urine culture
A-113
urine dipstick
B-014
urine drug screening
B-408
Urine Immunoassay
B-264
urine sediment analyzer
B-033
Urine Sediment Controls
B-181
URiSCAN
B-014, B-289
Urolithiasis
A-109
UTI
A-095
utilization
B-168, B-341
Utilization management
B-173

V
vacuum tubes
A-275
Vaginal infections
B-071
Validation
A-168, A-185, A-197, B-176,
B-177, B-188, B-193, B-199, B-290, B-304
value stream mapping
B-172
Variability
B-087, B-139
Variants
A-110
variation
A-259
Vascular Endothelial Growth Factor
B-051
Venous stasis
A-271
VHL
B-214
Viscosity
B-449
vitamin A
B-230
Vitamin A and E
B-271
Vitamin B12
B-225, B-248
vitamin b6
B-250
Vitamin C
B-254

S282

vitamin D

A-040, A-081, A-122, A-141,

A-167, A-216, A-220, A-223,
A-262, A-377, A-420, B-051, B-227, B-235,
B-236, B-238, B-240,
B-245, B-246, B-249
vitamin D deficiency
A-153
vitamin D levels
A-122
Vitamin D Metabolic Profiling
A-190
Vitamin D Receptor
A-367
vitamin E
B-230
vitamin K
B-223
Vitamn D
B-234
Vitros-5600
B-106
Von Willebrand Factor Antigen
A-308
Voriconazole
A-417, B-428
vWF-antigen
B-333

W
Waldenstroms Macroglobulinemia
A-308
West Africa
A-143
Western Blot
B-205
Widal test
A-144
women
B-338, B-344
worflow
B-024
Workflow
B-133

Y
Y chromosome
Y-STR
young

B-188
B-270
B-338

Z
Ziehl-Neelsen
Zinc 2 alpha glycoprotein
zonulin

B-067
A-206
B-329