Liang et al.

BMC Infectious Diseases (2019) 19:548

https://doi.org/10.1186/s12879-019-4166-1

RESEARCH ARTICLE Open Access

An improved algorithm for rapid diagnosis

of pleural tuberculosis from pleural effusion
by combined testing with GeneXpert MTB/
RIF and an anti-LAM antibody-based assay
Qingtao Liang1†, Yu Pang2†, Yang Yang1, Hua Li1, Chao Guo1, Xinting Yang1 and Xiaoyou Chen1*

Abstract
Background: This retrospective study evaluated the performance of a lipoarabinomannan (LAM)-based immunological
method for diagnosing pleural tuberculosis (TB) from pleural effusion samples. Results were compared to those obtained
using conventional culture and molecular testing methods.
Methods: Suspected pleural TB patients who visited Beijing Chest Hospital for medical care between January 2016 and
June 2017 were retrospectively analysed in the study. Pleural effusion samples were tested for Mycobacterium tuberculosis
(MTB) using the BACTEC MGIT 960 System, GeneXpert, and an anti-LAM antibody assay (LAM assay).
Results: Pleural effusion samples were collected from a total of 219 retrospectively recruited participants suspected of
having pleural TB. Thirteen of 155 confirmed pleural TB cases tested positive for MTB via MGIT culture, for a sensitivity of
8.4% [95% confidence interval (CI): 4.0–12.8%]. In addition, GeneXpert and LAM testing identified 22 and 55 pleural TB
cases, for sensitivities of 14.2% (95% CI: 8.7–19.7%) and 35.5% (95% CI: 28.1–43.6%), respectively. The specificities of these
two assays were 100.0% (95% CI: 92.9–100.0%) and 96.9% (95% CI: 88.2–99.5%), respectively. Combined application of
culture and LAM testing identified 60 positive cases, for a sensitivity of 38.7% (95% CI: 31.0–46.4%) that was significantly
higher than that of MGIT culture alone (P < 0.01). Similarly, use of LAM testing in combination with GeneXpert led to
correct diagnosis of 40.0% (95% CI: 32.3–47.7%) of pleural TB cases, a higher rate than obtained using GeneXpert alone
(P < 0.01). In addition, the specificity of the combined assay of GeneXpert and LAM testing was 96.9% (95% CI: 88.2–
99.5%). Patients aged 25 to 44 years were more likely to have positive LAM assay results than those ≥65 years of age
(P = 0.02). Meanwhile, the proportion of diabetic patients with positive LAM assay results was significantly lower than that
of the non-diabetes group (P = 0.03).
Conclusions: An anti-LAM antibody detection assay showed potential for diagnosis of pleural TB from pleural effusion
samples. Combined use of the LAM assay with MGIT culture or GeneXpert methods could improve sensitivity for
improved pleural TB diagnosis compared to results of individual conventional tests alone.
Keywords: Tuberculosis, Pleural, Diagnose, Lipoarabinomannan

* Correspondence: [email protected]
†
Qingtao Liang and Yu Pang contributed equally to this study.
1
Department of Tuberculosis, Beijing Chest Hospital, Capital Medical
University, Beijing Tuberculosis and Thoracic Tumor Research Institute,
Beijing, China
Full list of author information is available at the end of the article

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Liang et al. BMC Infectious Diseases (2019) 19:548 Page 2 of 8

Background immunological tests based on detection of antibodies

Tuberculosis (TB), caused by Mycobacterium tubercu- present in pleural effusion specimens offer promise for
losis complex (MTBC or MTB), is a serious global public improving pleural TB diagnosis [17]. Performance of im-
health concern [1]. Despite great progress made in re- munologically based testing had been previously evalu-
cent decades toward reducing TB disease burden, 10.1 ated as a tool for diagnosing pleural TB from blood
million incident cases and 1.6 million deaths are cur- samples [17], with sensitivity estimates ranging from 26
rently observed each year worldwide [1]. In addition to to 59% and specificity estimates ranging from 81 to
damaging the lungs as the most commonly affected tis- 100% [2, 18, 19]. Despite the fact that the WHO has ad-
sue, tuberculosis can involve any other organ or tissue to vised against the use of commercial serological tests for
cause so-called extrapulmonary tuberculosis (EPTB) [2]. diagnosing active TB [20], a serological test for diagnosis
Globally, reported proportions of extrapulmonary cases of pleural TB may be of value, since current tools are
range from 15 to 25% across countries [2, 3], with a not diagnostically effective for this disease. Such a test
worsening burden of EPTB disease resulting from pa- would be based on detection of anti-MTB antibodies
tient co-infection with human immunodeficiency virus produced by the underlying inflammatory reaction that
(HIV) [2]. Unfortunately, this TB/HIV co-infection sce- leads to pleural TB. This concept is based on limited
nario has been relatively neglected by TB control pro- data from previous studies, which had demonstrated that
grams, mainly due to its low overall contribution to TB lipoarabinomannan (LAM), an integral component of
transmission within the community [4]. Moreover, al- the MTB cell wall, can stimulate production of anti-
though pulmonary TB cases are often easily recognizable LAM antibodies by the human host [21]. We therefore
due to typical radiological features and positive bacterio- conducted a retrospective study to assess the perform-
logical evidence, EPTB is frequently more difficult to ance of an immunological method (the LAM assay) to
diagnose due to non-specific clinical and radiological detect anti-LAM antibodies in TB pleural effusion speci-
features of EPTB that are often subject to variable inter- mens. We then compared LAM assay sensitivity and
pretation [5]. specificity to corresponding results obtained using con-
Tuberculous pleurisy, one of the most common mani- ventional culture and molecular testing methods.
festations of EPTB, is the most common cause of pleural
effusion in patients in many countries [6, 7]. As for other Methods
forms of EPTB, pleural TB poses a great diagnostic chal- Study subjects
lenge, since conventional acid-fast bacilli (AFB) and cul- We retrospectively analyzed patients suspected of having
ture methods have poor sensitivity as tools for pleural TB who visited Beijing Chest Hospital to seek
diagnosing this disease [8]. A promising approach, PCR, medical care between January 2016 and June 2017. In-
has been applied to the detection of mycobacterial DNA clusion criteria were: i) suspected pleural TB based on
in pleural fluid, with molecular diagnostic sensitivity standard clinical and radiological criteria, including a
ranging from 29 to 75% depending on target sequence persistent cough of 2 weeks or more, unexplained fever
amplified and DNA extraction procedure [9–11]. Re- for 2 weeks or more, weight loss, and radiological evi-
cently, the use of GeneXpert, a fully automatic molecu- dence of pleural effusion; and ii) results of patient
lar diagnostic system, provides rapid and accurate pleural effusion specimen testing performed via myco-
detection of EPTB. However, this assay can be difficult bacterial culture, GeneXpert and anti-LAM antibody test
to perform in resource-limited countries, due to high (LAM assay) methods. Patients who were receiving anti-
cartridge costs and infrastructural requirements [12]. TB treatment were excluded from the final analysis. Par-
Additionally, given the poor sensitivity of GeneXpert in ticipant demographic profiles and clinical information
pleural TB, WHO has not recommended the use of were obtained from medical records.
GeneXpert for the diagnosis of pleural TB [13–15]. A total of 226 participants suspected of having pleural
Therefore, the lack of a reliable test for detecting MTB TB were retrospectively reviewed. Of the 226 cases, 7
in pleural effusion specimens undoubtedly leads to mis- were excluded due to invalid GeneXpert results (n = 2)
diagnosis or missed diagnosis of pleural TB. These chal- and contaminated culture results (n = 5). Ultimately, re-
lenges thus highlight the urgent need for development sults from 219 patients were used in the final analysis.
of more effective diagnostics for timely diagnosis of On the basis of laboratory examination results and clin-
pleural TB [12]. ical symptoms, 72 cases (32.9%) belonged to confirmed
Pleural TB results from entry of MTB antigens into pleural TB, 83 (37.9%) to clinically diagnosed pleural TB,
the pleural space after the rupture of a subpleural focus. and 64 (29.2%) to non-pleural TB. Forty-seven (73.4%,
This entry of antigens leads to generation of a host anti- 47/64) non-pleural TB cases were afflicted with malig-
body response and accumulation of pleural effusion that nancies and the other 17 cases (26.6%, 17/64) with
result from a hypersensitivity reaction [16]. Therefore, pneumonia (Fig. 1).
Liang et al. BMC Infectious Diseases (2019) 19:548 Page 3 of 8

Diagnostic criteria at room temperature. Bound antibody was detected using

A combination of clinical, microbiological, histological, goat anti-human IgG antibody conjugated to horseradish
and radiological findings was used for pleural TB diag- peroxidase. Antigen-antibody complexes were demon-
nostic confirmation according to Chinese national guide- strated by the presence of a pink line formed by antibody-
lines [22]. Briefly, pleural effusion specimens obtained conjugated horseradish peroxidase (within immunocom-
from patients suspected of having pleural TB were sub- plexes) that reacted with the substrate solution to produce
jected to routine laboratory analysis, including AFB a colored line. A line with intensity equal or greater than
smear, mycobacterial culture, GeneXpert, and other that of the positive control was recorded as a positive
tests. Pleural TB cases were classified as confirmed result.
pleural TB cases and clinically diagnosed pleural TB
cases based on laboratory examination results and clin- Statistical analysis
ical symptoms, respectively. Patients with at least one In view of the low overall rate of positive results ob-
positive MTB culture result via conventional culture tained for pleural TB patients, a composite reference
method or GeneXpert and/or granulomatous inflamma- standard (CRS) was created from clinically diagnosed
tion suggestive of TB from histological examination of pleural TB case samples and from cases with laboratory
pleural biopsy tissue samples were defined as confirmed confirmation for use as the gold standard. Results ob-
pleural TB cases. Patients without any experimental tained from testing the CRS using different methods and
diagnostic evidence, but who had exhibited clinical- combinations of methods were compared. Samples from
radiological manifestations based on clinical symptoms patients with malignancies and other respiratory diseases
prior to receiving treatment for TB, were defined as clin- served as negative controls. Sensitivity, specificity, posi-
ically diagnosed pleural TB cases [11]. tive predictive value (PPV), and negative predictive value
were calculated to evaluate the performance of diagnos-
Laboratory examination tic tests for detection of pleural TB. The chi-square test
Pleural effusion samples were collected in sterile 4-ml was used to compare categorical variables and differ-
tubes and transported to the laboratory for examination. ences were declared significant for P < 0.05. All statistical
A volume of 2.0 mL of each pleural effusion sample was analyses were performed using SPSS version 17.0 (Chi-
centrifugated at 3000×g for 15 min then each pellet was cago, IL, USA).
resuspended in phosphate buffered saline (PBS). A vol-
ume of 500 mL of each suspension was inoculated into a Results
separate 7-mL MGIT tube supplemented with 0.8 mL of Participants
oleic acid-albumin-dextrose-catalase (OADC) along with We first compared the distribution of demographic char-
PANTA™. MGIT tubes were placed into the MGIT 960 acteristics between pleural and non-pleural TB cases. As
instrument and cultures with growth were automatically summarized in Table 1, we observed that the percentage
reported by the instrument. Species identification testing of male patients in the pleural TB group was signifi-
was performed for all positive cultures using the Tibilia cantly higher than that of female patients (odds ratio
Rapid Test (Chuangxin, Hangzhou, China) [23]. (OR) [95% confidence interval (CI)]: 4.66[1.81–11.97],
For GeneXpert MTB/RIF testing, 1.0 mL of pleural ef- P < 0.01). In addition, percentages of pleural patients
fusion was mixed with 2.0 mL of sample reagent aged < 25 years (OR [95% CI]: 4.00[1.03–15.60], P = 0.04)
followed by incubation at room temperature for 15 min. and 25–44 years (OR [95% CI]: 2.82[1.14–6.98], P = 0.02)
2.0 mL of each inactivated sample mixture was trans- were significantly higher than in the non-pleural TB
ferred to a separate Xpert MTB/RIF cartridge then car- group. By contrast, no significant difference was ob-
tridges were inserted into the GeneXpert instrument. served with regard to residence or diabetes status be-
Results confirming the presence of MTB were automat- tween pleural TB and non-pleural TB groups (P > 0.05).
ically reported by the instrument within 90 min [24].
Detection of anti-LAM antibodies was performed ac- Performance of laboratory diagnostics
cording to the manufacturer’s instructions (Biovision, Thirteen of 155 pleural TB cases were detected by
Beijing, China). Briefly, a volume of 10 μL of pleural effu- MGIT culture testing, for a sensitivity of 8.4% (95% CI:
sion was tested for the presence of antibodies specific for 4.0–12.8%). In addition, GeneXpert and LAM testing
highly purified LAM antigen immobilized to test strips. identified 22 and 55 pleural TB cases, for a sensitivity of
LAM antigen had been secreted by MTB during active in- 14.2% (95% CI: 8.7–19.7%) and 35.5% (95% CI: 28.1–
fection prior to purification. When a pleural effusion sam- 43.6%), respectively. LAM assay sensitivity was signifi-
ple was applied to a test strip, it contacted the antigen line cantly higher than that of MGIT culture and GeneXpert
on the strip to permit antibodies in the specimen to bind (P < 0.01). No significant difference was observed in sen-
to antigen on the strip. Strips were incubated for 10 min sitivity between MGIT culture and GeneXpert methods
Liang et al. BMC Infectious Diseases (2019) 19:548 Page 4 of 8

Table 1 Demographic characteristics of participants suspected of having pleura TB

Characteristics Diagnostic Class
Pleural TB cases (155) Non pleural TB cases (64) Odds ratios P value Total
n (%) n (%) (95% CI) (219)
n (%)
Sex
Male 121 (78.1) 37 (57.4) 2.60 (1.39–4.85) < 0.01 158 (72.1)
Female 34 (21.9) 27 (42.6) 1.00 Ref. 61 (27.9)
Age group (years)
< 25 24 (15.5) 3 (4.9) 4.00 (1.03–15.60) 0.04 27 (12.3)
25~44 62 (40.0) 11 (18.0) 2.82 (1.14–6.98) 0.02 73 (33.3)
45~64 41 (26.5) 36 (59.0) 0.57 (0.26–1.24) 0.16 77 (35.2)
≥ 65 28 (18.1) 14 (23.0) 1.00 Ref. 42 (19.2)
Residence
Rural 77 (57.4) 29 (63.9) 1.19 (0.66–2.14) 0.56 106 (48.4)
Urban 78 (42.6) 35 (41.0) 1.00 Ref. 113 (51.6)
Diabetes
Yes 17 (11.0) 7 (11.5) 1.00 (0.40–2.55) 0.96 24 (11.0)
No 138 (89.0) 57 (88.5) 1.00 Ref. 195 (89.0)

for pleural TB detection from pleural effusion samples (P = 0.08). In contrast, the combined application of
(P > 0.05). culture-based testing and LAM assay identified 60 positive
We further analyzed the performance of combined diag- cases, resulting in a sensitivity of 38.7% (95% CI: 31.0–
nostic testing for pleural TB diagnosis. When combining 46.4%), which was significantly higher than that of MGIT
MGIT culture and GeneXpert, one additional positive pa- culture alone (P < 0.01). Similarly, by using the LAM assay
tient was detected, yielding a sensitivity of 14.8% (95% CI: in combination with GeneXpert, 40.0% (62/155, 95% CI:
9.2–20.4%) and specificity of 96.9% (95% CI: 92.9– 46.0–62.2%) of pleural TB cases were correctly diagnosed,
100.0%), while the difference between MGIT culture and a higher detection rate than that obtained using GeneX-
MGIT culture+GeneXpert was not statistically significant pert alone (P < 0.01) (Table 2).

Table 2 Performance of diagnostics for the diagnosis of pleural tuberculosis in pleural effusion samples
Method Sensitivity Specificity PPVa NPV
95% CI 95% CI 95% CI 95% CI
Culture 8.4% (13/155) 100.0% (64/64) 100.0% (13/13) 31.1% (64/206)
(4.0–12.8%) (92.9–100.0%) (100.0–100.0%) (24.7–37.4%)
GeneXpert 14.2% (22/155) 100.0% (64/64) 100.0% (22/22) 32.5% (64/197)
(8.7–19.7%) (92.9–100.0%) (100.0–100.0%) (25.9–39.0%)
LAM 35.5% (55/155) 96.9% (62/64) 96.5% (55/57) 38.3% (62/162)
(28.1–43.6%) (88.2–99.5%) (91.7–100.0%) (30.8–45.8%)
Culture+GeneXpert 14.8% (23/155) 100.0% (64/64) 100.0% (23/23) 32.7% (64/196)
(9.2–20.4%) (92.9–100.0%) (100.0–100.0%) (26.1–39.2%)
Culture+LAM 38.7% (60/155) 96.9% (62/64) 96.8% (60/62) 39.5% (62/157)
(31.0–46.4%) (88.2–99.5%) (92.4–100.0%) (31.8–47.1%)
GeneXpert+LAM 40.0% (62/155) 96.9% (62/64) 96.9% (62/64) 40.0% (62/155)
(32.3–47.7%) (88.2–99.5%) (92.6–100.0%) (32.3–47.7%)
Culture+GeneXpert+LAM 40.6% (63/155) 96.9% (62/64) 96.9% (63/65) 40.3% (62/154)
(32.9–48.4%) (88.2–99.5%) (92.7–100.0%) (32.5–48.0%)
a
PPV positive predictive value, NPV negative predictive value
χ = 33.23, P < 0.01(Culture Se. vs. LAM Se.); χ2 = 18.82, P < 0.01(GeneXpert Se.vs. LAM Se.); χ2 = 2.61, P = 0.11(Culture Se.vs. GeneXpert Se.); χ2 = 39.58, P < 0.01(Culture Se.
b 2

vs. Culture+LAM Se.); χ2 = 26.13, P < 0.01(GeneXpert Se.vs. GeneXpert+LAM Se.); χ2 = 3.14, P = 0.08(Culture Se.vs. Culture+GeneXpert Se)
Liang et al. BMC Infectious Diseases (2019) 19:548 Page 5 of 8

Fig. 1 Classification of patients with suspect pleural tuberculosis

Factors associated with negative LAM results

We compared distributions of demographic and clinical
characteristics between LAM-positive and LAM-
negative groups (Table 3). Compared with the percent-
age of LAM-positive patients in the ≥65 years age group
Table 3 Factors associated with LAM-based results among
(9.1%), the percentage of LAM-positive patients in the pleural TB patients
25–44 years group was higher (OR [95% CI]: 3.55[1.19–
Characteristics LAM positive LAM negative Odds ratios P
10.55], P = 0.02), with no statistical difference observed (n = 55) (n = 100) (95% CI) value
for the group of patients aged < 25 years (OR [95% CI]: No. Col % No. Col %
3.29[0.93–11.61], P = 0.06) or the group of patients aged Sex
45–64 years (OR [95% CI]: 2.22[0.69–7.15], P = 0.18). In
Male 44 80.0 77 77.0 1.20(0.53–2.68) 0.67
addition, the proportion of patients in the diabetes group
with positive LAM assay results (5.5%) was significantly Female 11 20.0 23 23.0 1.00 Ref.
lower than that of the non-diabetes group (14.0%, OR Age group (years)
[95% CI]: 0.22[0.05–0.99], P = 0.03). < 25 10 18.2 14 14.0 3.29 (0.93–11.61) 0.06
25~44 27 49.1 35 35.0 3.55 (1.19–10.55) 0.02
Discussion 45~64 13 23.6 27 27.0 2.22 (0.69–7.15) 0.18
The diagnosis of pleural TB is still an unsolved problem
≥ 65 5 9.1 23 23.0 1.00 Ref.
worldwide due to unreliable laboratory detection test re-
sults [12]. Numerous studies have documented that the Residence
GeneXpert assay is a useful confirmatory (rule in) diag- Rural 26 47.3 51 51.0 0.86 (0.45–1.67) 0.66
nostic test for EPTB from various types of clinical speci- Urban 29 52.7 49 49.0 1.00 Ref.
mens, such as cerebrospinal fluid and tissue samples Diabetes
[25–27]. However, it is not endorsed for use as an initial Yes 2 5.5 15 14.0 0.22(0.05–0.99) 0.03
test for diagnosing patients suspected of having pleural
No 52 94.5 86 86.0 1.00 Ref.
TB, due to its unsatisfactory performance for testing of
Liang et al. BMC Infectious Diseases (2019) 19:548 Page 6 of 8

pleural effusion specimens [28, 29]. In this study, detec- exhibit reduced humoral immunity, with more rapid
tion of anti-LAM antibody in pleural effusion samples decay of antibody responses than observed in non-
showed high specificity and moderate sensitivity for diabetic patients [33]. Therefore, poorer humoral im-
diagnosis of pleural TB, as compared with GeneXpert mune function of patients in this population may have
and MGIT culture. led to a lower pleural TB detection rate using the anti-
Here, the LAM assay could identify approximately LAM antibody detection method. However, it should be
40% of pleural TB cases when combined with MGIT cul- noted that the small number of diabetic patients in our
ture or GeneXpert, a success rate significantly higher study population may have undermined the reliability of
than obtained by each test alone. However, the combin- these results. Nevertheless, the low detection rate ob-
ation of MGIT culture and GeneXpert offered limited tained using the LAM assay highlights the need for new
benefit for diagnosing pleural TB from pleural effusion specific biomarkers that are more suitable for detecting
samples, since positive results of these etiological diag- pleural TB from pleural effusion samples in patients
nostic methods are based on the presence of tubercle with diabetes.
bacilli in specimens. Moreover, despite different detec- Compared with other diagnostic tests for pleural TB,
tion limits between culture and GeneXpert methods, a the LAM assay is simple and easy to perform and gener-
high proportion of detection overlap between the two ates results within 1 h after sample collection. Moreover,
methods would undoubtedly weaken justification for the cost of this assay is less than 1.00 USD per test, which
their combined use for pleural TB diagnosis. On the justifies its use as a cost-effective test for detecting pleural
contrary, detection of anti-LAM antibodies yielded bet- TB. Furthermore, since special equipment and skilled op-
ter positivity despite the fact there were few tubercle ba- erators are not always provided to TB laboratories in
cilli but there were adequate LAM molecules in resource-limited settings, another advantage of this assay
circulation in bodily fluids of TB patients and that could is that it can be operated and interpreted without any
elicit significant production of anti-LAM antibodies and complicated procedures or sophisticated instrumentation.
thus gave reasonably good sensitivity. Therefore, the LAM assay is an affordable and promising
Notably, numerous studies have demonstrated that diagnostic test for use in diagnosis of pleural TB, espe-
EPTB has been reported more frequently in females than cially in resource-limited settings.
in males [3, 30]. Conversely, in this report we found that This study had several obvious limitations. First, this
men were more likely to have pleural TB. However, a na- was a retrospective study rather than one based on con-
tionwide epidemiological study from the Netherlands tinuous recruitment of individuals with suspected
could reconcile these conflicting results, since in that pleural TB, which may limit the overall significance of
study EPTB was relatively more prevalent among females, our study conclusions. Second, the positive rate of
while pleural TB was more common in males (8.6% in MGIT culture for detection of MTB in pleural effusion
males versus 6.7% in females, P < 0.01) [5]. Although the specimens obtained here was lower than rates obtained
exact causes for gender biases among TB-related diseases in previous studies, a result possibly due to the small
remain unknown, we hypothesize that immunity, hor- sample volume used for mycobacterial culture. As a con-
mones, and socio-economic factors may be involved [31], sequence, the relatively low sensitivity may have negated
warranting further research. Meanwhile, our analysis the potentially beneficial role of mycobacterial culture
showed that pleural TB was more frequently observed for analysis of various diagnostic combinations. Third,
among patients younger than 44 years of age. One possible we realize that the performance of the LAM assay is far
explanation may be due to the fact that pleural TB is likely from acceptable for applications needed for high-priority
a manifestation of paucibacillary mycobacterial infection target product profiles [34]. The combination use of
that leads to an immunologically-based hypersensitivity multiple biomarkers, such as 38kD, 16kD, Ag85A and
response [16]. Therefore, a decline in immunity with in- MPT64 [35], may improve the sensitivity of this assay for
creasing age may decrease the risk of pleural TB occur- analysis of pleural TB using pleural effusion specimens.
rence in aging TB patients [32]. In line with this Fourth, the performance of the LAM assay was only eval-
hypothesis, further analysis here revealed that younger uated in pleural effusion samples rather than serum sam-
pleural TB patients were more likely to generate an anti- ples in this study. Despite these limitations, our study
LAM antibody response than were elderly patients. provides important insights into the use of an anti-LAM
In addition to advancing age, diabetes is another major antibody-based assay for diagnosis of pleural TB.
comorbidity associated with low rates of pleural TB
diagnosis that may contribute to false-negative LAM Conclusions
assay results. It is now well accepted that both humoral The results of this study demonstrate that the LAM assay,
and cellular immunity are involved in the pathogenesis which detects patient anti-LAM antibodies, shows prom-
of diabetes [33], since patients with diabetes appear to ising performance for achieving pleural TB diagnosis from
Liang et al. BMC Infectious Diseases (2019) 19:548 Page 7 of 8

pleural effusion samples. Moreover, the combined use of 4. Norbis L, Alagna R, Tortoli E, Codecasa LR, Migliori GB, Cirillo DM. Challenges
the LAM assay with MGIT culture or GeneXpert could and perspectives in the diagnosis of extrapulmonary tuberculosis. Expert
Rev Anti-Infect Ther. 2014;12(5):633–47.
improve pleural TB diagnostic sensitivity relative to each 5. te Beek LA, van der Werf MJ, Richter C, Borgdorff MW. Extrapulmonary
test alone. In addition, old age and diabetes comorbidities tuberculosis by nationality, the Netherlands, 1993-2001. Emerg Infect Dis.
are main factors associated with false-negative LAM assay 2006;12(9):1375–82.
6. Liam CK, Lim KH, Wong CM. Causes of pleural exudates in a region with a
results. high incidence of tuberculosis. Respirology. 2000;5(1):33–8.
7. Valdes L, Pose A, San Jose E, Martinez Vazquez JM. Tuberculous pleural
Abbreviations effusions. Eur J Intern Med. 2003;14(2):77–88.
AFB: Acid-fast bacilli; EPTB: Extrapulmonary tuberculosis; HIV: Human 8. Trajman A, Pai M, Dheda K, van Zyl Smit R, Zwerling AA, Joshi R, Kalantri S,
immunodeficiency virus; LAM: Lipoarabinomannan; MTBC: Mycobacterium Daley P, Menzies D. Novel tests for diagnosing tuberculous pleural effusion:
tuberculosis complex; OADC: Oleic acid-albumin-dextrose-catalase; what works and what does not? Eur Respir J. 2008;31(5):1098–106.
TB: Tuberculosis 9. Kumar P, Sen MK, Chauhan DS, Katoch VM, Singh S, Prasad HK. Assessment
of the N-PCR assay in diagnosis of pleural tuberculosis: detection of M.
Acknowledgements tuberculosis in pleural fluid and sputum collected in tandem. PLoS One.
We would like to thank the staffs from the Department of Tuberculosis, 2010;5(4):e10220.
Beijing Chest Hospital for their technical support. 10. Trajman A, da Silva Santos Kleiz de Oliveira EF, Bastos ML, Belo Neto E, Silva
EM, da Silva Lourenco MC, Kritski A, Oliveira MM. Accuracy of polimerase
Authors’ contributions chain reaction for the diagnosis of pleural tuberculosis. Respir Med 2014,
Conceived and designed the experiments: QL, YP, XC. Performed the 108(6):918–923.
experiments: QL, YY, HL, CG, XY. Analyzed the data: QL, YY, YP, XC. Wrote the 11. Che N, Yang X, Liu Z, Li K, Chen X. Rapid detection of cell-free
paper: QL, YP, XC. All authors contributed to the final manuscript. All authors mycobacterium tuberculosis DNA in tuberculous pleural effusion. J Clin
read and approved the final manuscript. Microbiol. 2017;55(5):1526–32.
12. Tyagi S, Sharma N, Tyagi JS, Haldar S. Challenges in pleural tuberculosis
Funding diagnosis: existing reference standards and nucleic acid tests. Future
This study was supported by the Beijing Municipal Administration of Microbiol. 2017;12:1201–18.
Hospitals’ Ascent Plan (grant number DFL20151501), and the Key Project of 13. Sehgal IS, Dhooria S, Aggarwal AN, Behera D, Agarwal R. Diagnostic
the Department of Science and Technology, Beijing, China (grant number performance of Xpert MTB/RIF in tuberculous pleural effusion: systematic
D141107005214003). The. review and meta-analysis. J Clin Microbiol. 2016;54(4):1133–6.
funders had no role in study design, data collection and analysis, decision to 14. Denkinger CM, Schumacher SG, Boehme CC, Dendukuri N, Pai M, Steingart
publish, and preparation of the manuscript. KR. Xpert MTB/RIF assay for the diagnosis of extrapulmonary tuberculosis: a
systematic review and meta-analysis. Eur Respir J. 2014;44(2):435–46.
15. World Health Organization. Xpert MTB/RIF assay for the diagnosis of
Availability of data and materials
pulmonary and extrapulmonary TB in adults and children. Geneva,
The datasets used and/or analysed during the current study available from
Switzerland: World Health Organization; 2014.
the corresponding author on reasonable request.
16. Vorster MJ, Allwood BW, Diacon AH, Koegelenberg CF. Tuberculous pleural
effusions: advances and controversies. J Thorac Dis. 2015;7(6):981–91.
Ethics approval and consent to participate 17. Steingart KR, Henry M, Laal S, Hopewell PC, Ramsay A, Menzies D,
This study was approved by the Ethics Committee of Beijing Chest Hospital, Cunningham J, Weldingh K, Pai M. Commercial serological antibody
affiliated to Capital Medical University. The Institutional Review Board detection tests for the diagnosis of pulmonary tuberculosis: a systematic
approved the waiver of informed consent for a retrospective study because review. PLoS Med. 2007;4(6):e202.
this study would bring no more than minimal risk to the subjects. 18. Kunter E, Cerrahoglu K, Ilvan A, Isitmangil T, Turken O, Okutan O, Kartaloglu
Z, Cavuslu S. The value of pleural fluid anti-A60 IgM in BCG-vaccinated
Consent for publication tuberculous pleurisy patients. Clin Microbiol Infect. 2003;9(3):212–20.
Not applicable. 19. Caminero JA, Rodriguez de Castro F, Carrillo T, Lafarga B, Daiz F, Cabrera P.
Value of ELISA using A60 antigen in the serodiagnosis of tuberculosis.
Competing interests Respiration. 1994;61(5):283–6.
The authors declare that they have no competing interests. 20. World Health Organization. Commercial serodiagnostic tests for diagnosis of
tuberculosis: policy statement. Geneva, Switzerland: World Health
Author details Organization; 2011.
1
Department of Tuberculosis, Beijing Chest Hospital, Capital Medical 21. Peter J, Green C, Hoelscher M, Mwaba P, Zumla A, Dheda K. Urine for the
University, Beijing Tuberculosis and Thoracic Tumor Research Institute, diagnosis of tuberculosis: current approaches, clinical applicability, and new
Beijing, China. 2National Clinical Laboratory on Tuberculosis, Beijing Key developments. Curr Opin Pulm Med. 2010;16(3):262–70.
Laboratory on Drug-Resistant Tuberculosis Research, Beijing Chest Hospital, 22. Ministry of Health: Criteria for pulmonary tuberculosis diagnosis. Beijing:
Capital Medical University, Beijing Tuberculosis and Thoracic Tumor Research 2008, WS288–2008. (In Chinese).
Institute, Beijing, China. 23. Pang Y, Dong H, Tan Y, Deng Y, Cai X, Jing H, Xia H, Li Q, Ou X, Su B, et al.
Rapid diagnosis of MDR and XDR tuberculosis with the MeltPro TB assay in
Received: 26 June 2018 Accepted: 5 June 2019 China. Sci Rep. 2016;6:25330.
24. Pang Y, Wang Y, Zhao S, Liu J, Zhao Y, Li H. Evaluation of the Xpert MTB/RIF
assay in gastric lavage aspirates for diagnosis of smear-negative childhood
References pulmonary tuberculosis. Pediatr Infect Dis J. 2014;33(10):1047–51.
1. World Health Organization. Global tuberculosis report 2018. Geneva, 25. Pink F, Brown TJ, Kranzer K, Drobniewski F. Evaluation of Xpert MTB/RIF for
Switzerland: World Health Organization; 2018. detection of mycobacterium tuberculosis in cerebrospinal fluid. J Clin
2. Steingart KR, Henry M, Laal S, Hopewell PC, Ramsay A, Menzies D, Microbiol. 2016;54(3):809–11.
Cunningham J, Weldingh K, Pai M. A systematic review of commercial 26. Nhu NT, Heemskerk D, Thu do DA, Chau TT, Mai NT, Nghia HD, Loc PP, Ha
serological antibody detection tests for the diagnosis of extrapulmonary DT, Merson L, Thinh TT, et al. Evaluation of GeneXpert MTB/RIF for diagnosis
tuberculosis. Thorax. 2007;62(10):911–8. of tuberculous meningitis. J Clin Microbiol. 2014;52(1):226–33.
3. Peto HM, Pratt RH, Harrington TA, LoBue PA, Armstrong LR. Epidemiology of 27. Hillemann D, Rusch-Gerdes S, Boehme C, Richter E. Rapid molecular
extrapulmonary tuberculosis in the United States, 1993-2006. Clin Infect Dis. detection of extrapulmonary tuberculosis by the automated GeneXpert
2009;49(9):1350–7. MTB/RIF system. J Clin Microbiol. 2011;49(4):1202–5.
Liang et al. BMC Infectious Diseases (2019) 19:548 Page 8 of 8

28. Lusiba JK, Nakiyingi L, Kirenga BJ, Kiragga A, Lukande R, Nsereko M,

Ssengooba W, Katamba A, Worodria W, Joloba ML, et al. Evaluation of
Cepheid's Xpert MTB/Rif test on pleural fluid in the diagnosis of pleural
tuberculosis in a high prevalence HIV/TB setting. PLoS One. 2014;9(7):
e102702.
29. Rufai SB, Singh A, Kumar P, Singh J, Singh S. Performance of Xpert MTB/RIF
assay in diagnosis of pleural tuberculosis by use of pleural fluid samples. J
Clin Microbiol. 2015;53(11):3636–8.
30. Sreeramareddy CT, Panduru KV, Verma SC, Joshi HS, Bates MN. Comparison
of pulmonary and extrapulmonary tuberculosis in Nepal- a hospital-based
retrospective study. BMC Infect Dis. 2008;8:8.
31. Lin JN, Lai CH, Chen YH, Lee SS, Tsai SS, Huang CK, Chung HC, Liang SH, Lin
HH. Risk factors for extra-pulmonary tuberculosis compared to pulmonary
tuberculosis. Int J Tuberc Lung Dis. 2009;13(5):620–5.
32. Kline KA, Bowdish DM. Infection in an aging population. Curr Opin
Microbiol. 2016;29:63–7.
33. Zaccardi F, Webb DR, Yates T, Davies MJ. Pathophysiology of type 1 and
type 2 diabetes mellitus: a 90-year perspective. Postgrad Med J. 2016;
92(1084):63–9.
34. World Health Organization. High-priority target product profiles for new TB
diagnostics. Report of a consenensus meeting, Geneva. Geneva, Switzerland:
World Health Organization; 2014.
35. Wu X, Yang Y, Zhang J, Li B, Liang Y, Zhang C, Dong M. Comparison of
antibody responses to seventeen antigens from mycobacterium
tuberculosis. Clin Chim Acta. 2010;411(19–20):1520–8.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.