The n e w e ng l a n d j o u r na l of m e dic i n e

original article

Ferric Carboxymaltose in Patients with Heart

Failure and Iron Deficiency
Stefan D. Anker, M.D., Ph.D., Josep Comin Colet, M.D.,
Gerasimos Filippatos, M.D., Ronnie Willenheimer, M.D.,
Kenneth Dickstein, M.D., Ph.D., Helmut Drexler, M.D.,*
Thomas F. Lscher, M.D., Boris Bart, M.D., Waldemar Banasiak, M.D., Ph.D.,
Joanna Niegowska, M.D., Bridget-Anne Kirwan, Ph.D., Claudio Mori, M.D.,
Barbara von Eisenhart Rothe, M.D., Stuart J. Pocock, Ph.D.,
Philip A. Poole-Wilson, M.D.,* and Piotr Ponikowski, M.D., Ph.D.,
for the FAIR-HF Trial Investigators

A BS T R AC T

From the Department of Cardiology, BACKGROUND

Campus Virchow-Klinikum, Charit Uni-
versittsmedizin, Berlin, and the Center
Iron deficiency may impair aerobic performance. This study aimed to determine
for Clinical and Basic Research, San Raf- whether treatment with intravenous iron (ferric carboxymaltose) would improve
faele, Rome (S.D.A.); Hospital del Mar symptoms in patients who had heart failure, reduced left ventricular ejection frac-
and Universitat Autonoma de Barcelona,
Barcelona (J.C.C.); Athens University Hos-
tion, and iron deficiency, either with or without anemia.
pital Attikon, Athens (G.F.); Heart Health
Group and Lund University, Malm, Swe- METHODS
den (R.W.); Stavanger University Hospital, We enrolled 459 patients with chronic heart failure of New York Heart Association
Stavanger, and the University of Bergen,
Bergen both in Norway (K.D.); Medi
(NYHA) functional class II or III, a left ventricular ejection fraction of 40% or less (for
zinische Hochschule Hannover, Hannover, patients with NYHA class II) or 45% or less (for NYHA class III), iron deficiency (ferritin
Germany (H.D.); University Hospital Z level <100 g per liter or between 100 and 299 g per liter, if the transferrin saturation
rich, Zurich (T.F.L.), SOCAR Research, Nyon
(B.-A.K.), and Vifor Pharma, Glattbrugg
was <20%), and a hemoglobin level of 95 to 135 g per liter. Patients were randomly as-
(C.M., B.E.R.) all in Switzerland; Rus- signed, in a 2:1 ratio, to receive 200 mg of intravenous iron (ferric carboxymaltose) or
sian State Medical University, Outpatient saline (placebo). The primary end points were the self-reported Patient Global Assess-
Diagnostic Consulting Center 1, Moscow
(B.B.); Military Hospital, Wroclaw (W.B., ment and NYHA functional class, both at week 24. Secondary end points included the
P.P.); Centrum Medyczne, Telmont Cen- distance walked in 6 minutes and the health-related quality of life.
trum Medyczne, Sp. z O.O., Warsaw (J.N.),
and Medical University, Wroclaw (P.P.) RESULTS
all in Poland; and London School of Hy-
giene and Tropical Medicine (S.J.P.) and Among the patients receiving ferric carboxymaltose, 50% reported being much or
the National Heart and Lung Institute, Im- moderately improved, as compared with 28% of patients receiving placebo, according
perial College London (P.A.P.-W.) both to the Patient Global Assessment (odds ratio for improvement, 2.51; 95% confidence
in London. Address reprint requests to Dr.
Anker at Applied Cachexia Research, De- interval [CI], 1.75 to 3.61). Among the patients assigned to ferric carboxymaltose,
partment of Cardiology, Charit Universi 47% had an NYHA functional class I or II at week 24, as compared with 30% of pa-
ttsmedizin Berlin, Campus Virchow-Klini- tients assigned to placebo (odds ratio for improvement by one class, 2.40; 95% CI,
kum. Augustenburger Pl. 1, D-13353 Berlin,
Germany, or at [email protected]. 1.55 to 3.71). Results were similar in patients with anemia and those without anemia.
*Deceased. Significant improvements were seen with ferric carboxymaltose in the distance on
Members of the Ferinject Assessment the 6-minute walk test and quality-of-life assessments. The rates of death, adverse
in Patients with Iron Deficiency and events, and serious adverse events were similar in the two study groups.
Chronic Heart Failure (FAIR-HF) study
group are listed in the Appendix. CONCLUSIONS
This article (10.1056/NEJMoa0908355) was Treatment with intravenous ferric carboxymaltose in patients with chronic heart
published on November 17, 2009, at NEJM.
org. failure and iron deficiency, with or without anemia, improves symptoms, functional
N Engl J Med 2009;361:2436-48.
capacity, and quality of life; the side-effect profile is acceptable. (ClinicalTrials.gov
Copyright 2009 Massachusetts Medical Society. number, NCT00520780.)
2436 n engl j med 361;25 nejm.org december 17, 2009

Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts Medical Society. All rights reserved.
 Ferric Carboxymaltose in Heart Failure and Iron Deficiency

R
ecent developments in the manage- deficiency with the use of intravenous iron (ferric
ment of chronic heart failure in patients carboxymaltose) confers symptomatic benefits in
with an impaired left ventricular ejection patients with chronic heart failure.
fraction have changed the natural history of this
clinical syndrome and improved patients out- Me thods
comes.1,2 However, the normal daily activities
of many patients with heart failure remain re- Trial Design and Oversight
stricted; they report symptoms of fatigue and From June 25, 2007, through December 31, 2008,
dyspnea that adversely affect their quality of life, a total of 459 eligible patients were enrolled from
leading to high morbidity.3,4 Therapeutic options 75 sites in 11 countries (Fig. 1 in the Supplemen-
to improve functional capacity in patients with tary Appendix, available with the full text of this
heart failure are limited, and novel therapies are article at NEJM.org). The study design has been
needed. published previously.20 The protocol was approved
Numerous mechanisms unrelated to hemody- by the institutional review board at each partici-
namic dysfunction may underlie impaired exercise pating center, and the trial was conducted in ac-
tolerance in patients with chronic heart failure. cordance with the principles of the Declaration
Among them, inadequate oxygen supply and im- of Helsinki, the International Conference on Har-
paired oxygen use by skeletal muscle during exer- monization Good Clinical Practice guidelines, and
cise contribute to poor clinical status.5,6 In addi- local and national regulations. Written informed
tion, anemia may aggravate symptoms in patients consent was provided by all patients before any
with heart failure.7 Targeting these abnormalities study-related procedures were performed.
may confer functional benefits to such patients. The trial was designed, implemented, and over-
Iron plays a key role in oxygen uptake, trans- seen by the FAIR-HF Executive Committee, to-
port, and storage, as well as oxidative metabolism gether with representatives of the sponsor, Vifor
in the skeletal muscle; it also is involved in eryth- Pharma (Glattbrugg, Switzerland). ClinStar (Mos-
ropoiesis.8,9 Traditionally, iron deficiency has been cow) was responsible for on-site monitoring of
considered to have clinical consequences only in sites in Russia and Ukraine. Kendle (Munich, Ger-
the presence of anemia. Alternatively, a reduced many) was responsible for on-site monitoring in
hemoglobin level can be viewed as the end result other countries, in addition to data collection and
of a process beginning with the gradual depletion data management. SOCAR Research (Nyon, Swit-
of iron stores.9,10 Iron deficiency in patients with zerland) was responsible for data analysis. Anal-
or without anemia attenuates aerobic performance yses were performed independently of the spon-
and is accompanied by reports of fatigue and ex- sor, according to a predefined plan of statistical
ercise intolerance.11 The repletion of iron in pa- analysis. The medical statistics unit at the Lon-
tients who have iron deficiency without heart don School of Hygiene and Tropical Medicine
failure improves cognitive, symptomatic, and ex- performed the same analyses, separately, with
ercise performance.12,13 identical results. The manuscript was prepared and
Recently, it has been recognized that patients submitted for publication by the FAIR-HF Execu-
with heart failure may be prone to the develop- tive Committee. An independent data and safety
ment of iron deficiency as a consequence of a monitoring board reviewed the safety data on an
depletion of iron stores or defective iron absorp- ongoing basis. The authors had access to the study
tion and the reduced availability of iron recycled data and vouch for the accuracy and completeness
in the reticuloendothelial system.14,15 Four small of the reported data and analyses.
studies showed that the correction of iron defi-
ciency with the use of intravenous iron in pa- Recruitment and Follow-up of Study Patients
tients with chronic heart failure may result in Eligible subjects included ambulatory patients who
clinical benefits.16-19 In one of these studies,17 had chronic heart failure of New York Heart As-
the symptomatic benefit was similar in patients sociation (NYHA) class II or III, a left ventricular
with anemia and those without anemia. We de- ejection fraction of 40% or less (for patients in
signed our randomized, double-blind study, called NYHA class II) or 45% or less (for patients in
the Ferinject Assessment in Patients with Iron NYHA class III), a hemoglobin level at the screen-
Deficiency and Chronic Heart Failure (FAIR-HF) ing visit between 95 and 135 g per liter, and iron
trial, to determine whether the correction of iron deficiency. The presence or absence of iron defi-

n engl j med 361;25 nejm.org december 17, 2009 2437

Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

ciency was determined on the basis of results re- tion was less than 20%. Patients were excluded if
ceived from the central laboratory. Iron deficien- they had uncontrolled hypertension, other clini-
cy was diagnosed when the serum ferritin level cally significant heart disease, inflammation, or
was less than 100 g per liter or was between 100 clinically significantly impaired liver or renal
and 299 g per liter when the transferrin satura- function.

Table 1. Baseline Demographic and Clinical Characteristics of the Study Patients in the Intention-to-Treat Population,
According to Study Group.*

Ferric Carboxymaltose Placebo

Variable (N=304) (N=155)
Age yr 67.810.3 67.411.1
Female sex no. (%) 159 (52.3) 85 (54.8)
White race no. (%) 303 (99.7) 155 (100.0)
NYHA class no. (%)
II 53 (17.4) 29 (18.7)
III 251 (82.6) 126 (81.3)
Left ventricular ejection fraction % 31.95.5 33.06.1
Body weight kg 77.014.2 77.616.3
Body-mass index 28.04.8 28.15.1
Blood pressure mm Hg
Systolic 12615 12615
Diastolic 779 7610
Pulse beats/min 7111 7212
6-Minute walk test distance m 274105 269109
Ischemic cause of heart failure no. (%) 245 (80.6) 123 (79.4)
Cardiovascular risk factor no. (%)
Hypertension, treated with drugs 243 (79.9) 128 (82.6)
Dyslipidemia, treated with drugs 144 (47.4) 70 (45.2)
Diabetes mellitus 93 (30.6) 37 (23.9)
Atrial fibrillation 94 (30.9) 44 (28.4)
Medical history no. (%)
Myocardial infarction 168 (55.3) 90 (58.1)
Angina pectoris 171 (56.3) 89 (57.4)
Stroke 24 (7.9) 9 (5.8)
Coronary revascularization 64 (21.1) 31 (20.0)
Laboratory measurements
Hemoglobin g/liter 11913 11914
Mean corpuscular volume m3 91.68.1 91.76.7
Serum ferritin g/liter 52.554.5 60.166.5
Transferrin saturation % 17.712.6 16.78.4
C-reactive protein mg/liter 7.465.34 9.125.48
Sodium mmol/liter 1413 1413
Potassium mmol/liter 4.650.61 4.580.52
Alanine aminotransferase U/liter 20.512.3 18.88.1
Aspartate aminotransferase U/liter 23.110.4 22.47.2
Creatinine mg/dl 1.20.6 1.20.6
Estimated glomerular filtration rate 63.821.2 64.825.3
ml/min/1.73 m2 of body-surface area

2438 n engl j med 361;25 nejm.org december 17, 2009

Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts Medical Society. All rights reserved.
 Ferric Carboxymaltose in Heart Failure and Iron Deficiency

Table 1. (Continued.)

Ferric Carboxymaltose Placebo

(N=304) (N=155)
Concomitant treatment no. (%)
Diuretic 280 (92.1) 140 (90.3)
ACE inhibitor or ARB 281 (92.4) 141 (91.0)
Digitalis glycoside 46 (15.1) 25 (16.1)
Beta-blocker 262 (86.2) 129 (83.2)
Antiplatelet therapy 189 (62.2) 97 (62.6)
Anticoagulant therapy 67 (22.0) 22 (14.2)
Lipid-lowering therapy 142 (46.7) 72 (46.5)
Insulin 27 (8.9) 9 (5.8)
Oral hypoglycemic agent 49 (16.1) 22 (14.2)

* Plusminus values are means SD. To convert the values for creatinine to micromoles per liter, multiply by 88.4. ACE
denotes angiotensin-converting enzyme, ARB angiotensin-receptor blocker, and NYHA New York Heart Association.
Race was self-reported.
The body-mass index is the weight in kilograms divided by the square of the height in meters.
The percent transferrin saturation was calculated as iron (in micromoles per liter)transferrin (in grams per li-
ter)25.1.
The estimated glomerular filtration rate was calculated by the central laboratory according to the Modification of Diet in
Renal Disease formula: 186(serum creatinine [in micromoles per liter]88.4)1.154age (in years)0.2031.21 (if pa-
tient is black)0.742 (if patient is female).

Randomization preparation and administration of the study drug

Before the iron-correction phase was begun, a clin- (including at least one physician) were aware of
ical history, physical examination, 6-minute walk the group assignments and therefore were not in-
test results, and 12-lead electrocardiogram were volved in any study assessments. To ensure that
obtained for each patient. We also performed as- patients were unaware of the study drug they were
sessments of the health-related quality of life in all receiving, black syringes were used to adminis-
patients. Using a central interactive voice-response ter the study treatment and a curtain (or some-
system, we randomly assigned eligible patients, thing similar) was used to shield the injection site
in a 2:1 ratio, to receive either ferric carboxymalt- from the patients view.
ose (provided by Vifor Pharma) or placebo (normal The central laboratory sent results regarding
saline). measures of iron metabolism and hemoglobin
only to the study personnel who were aware of
Study Therapy and Blinding the group assignments. These persons were re-
The total iron dose required for iron repletion was sponsible for evaluating for the presence of ele-
calculated at baseline, according to Ganzonis for- vated iron-metabolism measurements or severe
mula21 and the mean of the two hemoglobin val- anemia and for implementing the following pro-
ues obtained during the screening period. The fer- cedures, as defined in the protocol. If the ferritin
ric carboxymaltose or saline was administered as level exceeded 800 g per liter or was between
an intravenous bolus injection of 4 ml (which is 500 and 800 g per liter with a transferrin satura-
the amount of ferric carboxymaltose in a water tion of more than 50%, or if the hemoglobin level
solution for injection that is equivalent to 200 mg was higher than 160 g per liter, ferric carboxymalt-
of iron). The dosing frequency was weekly until ose was discontinued and placebo was given in-
iron repletion was achieved (the correction phase) stead. In this case, the ferritin, transferrin satu-
and then every 4 weeks during the maintenance ration, and hemoglobin levels were reassessed.
phase, which started at week 8 or week 12, de- After the ferritin level had dropped to less than
pending on the required iron-repletion dose. 400 g per liter, the transferrin saturation to un-
Because ferric carboxymaltose is a dark-brown der 45%, and the hemoglobin level to less than
solution that is easily distinguishable from the 160 g per liter, treatment with ferric carboxymalt-
saline placebo, study personnel responsible for the ose could be restarted. If severe anemia (hemo-

n engl j med 361;25 nejm.org december 17, 2009 2439

Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

globin level, 90 g per liter) developed, the study

Figure 1 (facing page). Self-Reported Patient Global
treatment was permanently discontinued. The fol- Assessment and New York Heart Association (NYHA)
low-up of such patients continued, and further Functional Class at Week 24, According to Assigned
management of anemia was performed at the in- Study Treatment.
vestigators discretion. Panel A shows the data regarding the self-reported Pa-
In addition to the dosing visits, at weeks 4, 12, tient Global Assessment, and Panel B, data regarding
the NYHA functional class. The odds ratio for an im-
24, and 26, patients were assessed for efficacy and
provement in the self-reported Patient Global Assess-
safety. ment with ferric carboxymaltose as compared with pla-
cebo was 2.51 (95% confidence interval [CI], 1.75 to 3.61;
Primary and Secondary End Points P<0.001). Data on the self-reported Patient Global Assess-
The primary end points were the self-reported Pa- ment were missing for 30 patients known to be alive and
not in the hospital at week 24 (18 receiving ferric carboxy
tient Global Assessment (Fig. 2 in the Supplemen-
maltose and 12 receiving placebo); for these patients,
tary Appendix) and NYHA functional class (adjust- the last available assessment was used as the assess-
ed for the class at baseline) at week 24. Secondary ment for week 24. The two patients who were hospital-
end points included the self-reported Patient Global ized at 24 weeks (both receiving ferric carboxymaltose)
Assessment and NYHA functional class at week were assigned an assessment of much worse. The 9 pa-
tients who died before week 24 (5 who had been receiv-
4 and week 12, as well as the distance on the 6-min-
ing ferric carboxymaltose and 4 who had been receiving
ute walk test and the overall score on the Kansas placebo) were categorized as dead. Panel A does not in-
City Cardiomyopathy questionnaire22 (on which clude 12 patients in the ferric carboxymaltose group and
the overall score ranges from 0 to 100, with a high- 6 in the placebo group who were known to be alive at
er score indicating a better quality of life) and the week 24 but who had no data on the self-reported Pa-
tient Global Assessment at any time during the study.
European Quality of Life5 Dimensions (EQ-5D)
The odds ratio for an improved NYHA class with ferric
Visual Analog Scale23 (on which the score ranges carboxymaltose as compared with placebo was 2.40
from 0 to 100, with a higher score indicating bet- (95% CI, 1.55 to 3.71; P<0.001). Data on the NYHA class
ter health) at weeks 4, 12, and 24 (all adjusted for were missing for 28 patients known to be alive and not
the baseline data). in the hospital at week 24 (16 receiving ferric carboxy-
maltose and 12 receiving placebo); for these patients,
Safety end points were serious and nonseri-
the last available NYHA classification was used as the
ous adverse events, hospitalization, and death, classification for week 24. The 2 patients who were hos-
as assessed up to week 26. Standardized defini- pitalized at 24 weeks (both receiving ferric carboxymal
tions for the causes of death and hospitalization tose) were considered to have NYHA class IV. The 9 pa-
were developed by members of the FAIR-HF Ex- tients who died before week 24 (5 who had been receiving
ferric carboxymaltose and 4 who had been receiving pla-
ecutive Committee and were supplied to all in-
cebo) were categorized as being dead (having NYHA
vestigators.20 class V). The data in Panel B do not include data for 10
patients in the ferric carboxymaltose group and 5 in the
Statistical Analysis placebo group who were known to be alive at week 24
Data analysis for efficacy was performed accord- but who had no data on the NYHA functional class at
any time during the study. In both panels, percentages
ing to the intention-to-treat principle for each as-
may not sum to 100 because of rounding.
signed study group. The planned sample size of
402 patients 268 in the ferric carboxymaltose
group and 134 in the placebo group was cal- line value. The alpha value was adjusted accord-
culated on the basis of assumptions that the study ing to the method of Benjamini and Hochberg.25
would have a statistical power of 90% (with a two- Missing data on the NYHA class and the self-
sided alpha of 0.025) to detect a mean difference reported Patient Global Assessment were imputed
in the NYHA class of 0.50 and in the self-reported from data collected during the previous follow-up
Patient Global Assessment ranking of 0.90 (the visit, according to the lastobservation-carried-
primary end points) between the two study groups. forward method, for patients who were known
Assuming that 10% of patients would not com- to be alive and not hospitalized at the time of the
plete the week 24 visit, the planned sample size assessment. For hospitalized patients, a missing
was increased to 442. Both primary end points NYHA class was assigned as class IV and a miss-
were compared between the two study groups by ing self-reported Patient Global Assessment was
means of polytomous regression.24 For data on the assigned as much worse. For patients who had
NYHA class, the model was adjusted for the base- died, a missing NYHA class was assigned as

2440 n engl j med 361;25 nejm.org december 17, 2009

Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts Medical Society. All rights reserved.
 Ferric Carboxymaltose in Heart Failure and Iron Deficiency

A Self-Reported Patient Global Assessment at Wk 24

50
Ferric carboxymaltose Placebo

40
34 35
Patients (%)

30 28
26

20 17 18
16
10
10

2 3 3 2 2 3
1 1
0

d
pr ly

pr tle

wo ly
ed

ed

ed

e
pr h

rs

rs
ge

ea
rs
im rate

e
im uc

im lit

at
ov

ov

ov

wo

wo

D
an
M

er
A
e

ch

le

h
od

od

uc
t
Un

lit
M

M
A
No. of Patients
Ferric carboxymaltose 47 100 77 54 5 2 2 5
Placebo 15 26 41 52 4 4 3 4

B NYHA Functional Class at Wk 24

80

70
65
60

50
50 Ferric carboxymaltose
Patients (%)

41 Placebo
40

30 29

20

10
6
1 1 3 2 3
0
sI

II

d
sI

sI

ea
sI
as

as

as

D
as
Cl

Cl

Cl

Cl

No. of Patients
Ferric carboxymaltose 17 121 148 3 5
Placebo 2 43 97 4 4

class V and a missing self-reported Patient Global for interaction were performed as part of the sub-
Assessment was assigned as died.Anker
AUTHOR: The analysis group analysis RETAKE: regarding
1st the primary end points
2nd
regarding the primary end points was
FIGURE: 1 of 3 restricted by adding an interaction
3rd
term to the ordered poly-
to data for patients who had at least one valid tomous regression Revised model. Cox proportional-haz-
ARTIST: MRL
self-reported Patient Global Assessment or NYHA ards regressionSIZE models were used to estimate the
6 col
value, respectively, during theTYPE: Line period.
follow-up Combo 4-C
hazard H/Tratios for
33p9safety outcomes on the basis of
For the continuous variables, changes fromAUTHOR,thePLEASE
theNOTE:
treatment received. Event rates are reported
Figure has been redrawn and type has been reset.
baseline value and values at weeks 4, 12, and 24 per person-year at risk. All analyses were con-
Please check carefully.
were compared between the ferric carboxymaltose ducted with the use of SAS software, version 9.1
JOB: 36125 ISSUE: 12-17-09
group and the placebo group by comparing the (SAS Institute).
means within each of the two study groups at
each visit with the use of a model for repeated R e sult s
measures adjusted for baseline values. For cate-
gorical end points, differences in the distribution Characteristics of the Study Patients
for each of the two study groups were tested by The clinical characteristics of 459 patients are pre-
means of ordered polytomous regression. Tests sented in Table 1. A total of 304 patients were ran-

n engl j med 361;25 nejm.org december 17, 2009 2441

Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

domly assigned to receive ferric carboxymaltose

Figure 2 (facing page). Main Secondary Outcomes
and 155 to receive placebo. The two groups were during the Study, According to Assigned Study
similar with respect to the baseline clinical and Treatment.
laboratory characteristics and the use of cardio- Shown are data across the study period for the self-
vascular medications at the time of enrollment. reported Patient Global Assessment (Panel A), the
New York Heart Association (NYHA) functional class
Follow-up (Panel B), change in distance on the 6-minute walk test
(Panel C), the change in the score on the European
Of the 304 patients assigned to receive ferric car- Quality of Life5 Dimensions (EQ-5D) Visual Analog
boxymaltose, 26 (8.6%) did not complete the 24 Scale (on which the score ranges from 0 to 100, with
weeks of follow-up; of these 26, 5 had died and 21 higher scores indicating better health) (Panel D), and
had withdrawn (Fig. 1 in the Supplementary Ap- the change in the overall score on the Kansas City Car-
diomyopathy questionnaire (on which the overall score
pendix). Of the 155 patients assigned to receive
ranges from 0 to 100, with a higher score indicating a
placebo, 20 (12.9%) did not complete the 24 weeks better quality of life) (Panel E). The data in Panels A
of follow-up; of these 20, 4 had died and 16 had and B are odds ratios, for the ferric carboxymaltose
withdrawn. (FCM) group as compared with the placebo group and
shown on a log2 scale, of being in a better assessment
Primary End Points category (Panel A) or NYHA functional class (Panel B).
In both panels, for data on the self-reported Patient
The self-reported Patient Global Assessment at Global Assessment that were missing for patients
week 24 was improved in the ferric carboxymaltose known to be alive and not in the hospital at each time
group, with 50% of patients reporting that they point, the last available assessment was used. Patients
were much or moderately improved, as compared who were hospitalized at each time point were given
an assessment of much worse (in Panel A) or an
with 28% of patients in the placebo group (odds
NYHA class of IV (in Panel B). Patients who died be-
ratio for being in a better rank, 2.51; 95% confi- fore week 24 were categorized as dead (in Panel B, cor-
dence interval [CI], 1.75 to 3.61; P<0.001) (Fig. 1A). responding to NYHA class V). Data were not included
Similarly, the NYHA functional class at week 24, for patients who were known to be alive at the time
after adjustment for the baseline value was im- point but had no previous data. For Panel A, the odds
ratios at week 4 and week 12 are 3.44 (95% confidence
proved in the ferric carboxymaltose group, with
interval [CI], 2.34 to 5.07) and 3.19 (95% CI, 2.20 to
47% having an NYHA functional class I or II, as 4.63). In Panel B, the odds ratios at week 4 and week
compared with 30% in the placebo group (odds 12 are 3.96 (95% CI, 1.98 to 7.93) and 3.42 (95% CI,
ratio for improvement by one class, 2.40; 95% CI, 2.04 to 5.72). Panels C, D, and E show the mean (SE)
1.55 to 3.71; P<0.001) (Fig. 1B). changes in the variable at weeks 4, 12, and 24. In those
panels, the P values are for the comparison between
the two study groups, and the I bars denote the stan-
Secondary End Points dard error.
The use of ferric carboxymaltose, as compared with
placebo, significantly improved the self-reported
Patient Global Assessment and NYHA class at Safety and Biochemical Analyses
weeks 4 and 12 (P<0.001 for all comparisons) (Fig. Survival status was available for all patients through
2A and 2B, and Fig. 3 and 4 in the Supplemen- at least week 24. The rates of death, hospitalization,
tary Appendix). Significant improvements were and serious and nonserious adverse events reported
also seen in the distance on the 6-minute walk by the investigators were similar in the two study
test and in the quality of life, as evaluated by the groups (Table 2). There was a trend toward a lower
EQ-5D visual assessment score and the overall rate of first hospitalization for any cardiovascular
Kansas City Cardiomyopathy score, at weeks 4, reason among patients receiving ferric carboxy-
12, and 24 (P<0.001 for all comparisons) (Fig. 2C, maltose as compared with those receiving placebo
2D, and 2E). (hazard ratio, 0.53; 95% CI, 0.25 to 1.09; P=0.08).
The hazard ratio for death or first hospitalization
Subgroup Analyses for any cardiovascular reason among patients who
A consistent benefit regarding the two primary end received ferric carboxymaltose as compared with
points was observed in all prespecified subgroups those who received placebo was 0.61 (95% CI, 0.32
(Fig. 3). The treatment effect was similar in pa- to 1.18; P=0.14).
tients with anemia and in those without anemia Data on serious and nonserious adverse events
(prospectively defined as a hemoglobin level 120 g are also shown in Table 2, and in Table 1 in the
per liter at baseline). Supplementary Appendix. The study treatment

2442 n engl j med 361;25 nejm.org december 17, 2009

Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts Medical Society. All rights reserved.
 Ferric Carboxymaltose in Heart Failure and Iron Deficiency

A Self-Reported Patient Global Assessment B NYHA Functional Class

P<0.001 P<0.001 P<0.001
8.0 P<0.001 P<0.001 P<0.001 8.0

FCM Better

FCM Better
Odds Ratio (95% CI)

Odds Ratio (95% CI)

4.0 4.0

2.0 2.0

1.0 1.0
Placebo

Placebo
Better

Better
0.5 0.5

0.0 0.0
0 4 8 12 16 20 24 0 4 8 12 16 20 24
Weeks since Randomization Weeks since Randomization
No. of Patients No. of Patients
FCM 282 291 292 FCM 304 287 294 294
Placebo 146 149 149 Placebo 155 147 150 150

C 6-Minute-Walk Test D EQ-5D Visual Analog Scale

50 P<0.001 P<0.001 P<0.001 14
P<0.001 P<0.001 P<0.001
12
40
Change in Distance (m)

FCM 10 FCM

Change in Score
30
8
20 6
Placebo 4 Placebo
10
2
0
0
10 2
0 4 8 12 16 20 24 0 4 8 12 16 20 24
Weeks since Randomization Weeks since Randomization
FCM FCM
No. of patients 303 284 280 268 No. of patients 295 274 283 285
Mean 2746 2947 3126 3137 Mean 541 601 621 631
distance (m) score
Placebo Placebo
No. of patients 155 144 141 134 No. of patients 152 140 145 146
Mean 2699 26910 27210 27710 Mean 541 542 562 572
distance (m) score
Mean Study-Treatment 216 377 358 Mean Study-Treatment 61 62 72
Effect Effect
E Kansas City Cardiomyopathy Questionnaire
14 P<0.001 P<0.001 P<0.001

12
Change in Overall Score

FCM
10

6 Placebo

0
0 4 8 12 16 20 24
Weeks since Randomization
FCM
No. of patients 297 277 286 286
Mean 521 621 651 661
score
Placebo
No. of patients 151 140 144 145
Mean 531 562 572 592
score
Mean Study-Treatment 61 82 72
Effect

AUTHOR: Anker RETAKE: 1st

2nd
FIGURE:
n engl j med 2 of nejm.org
361;25 3 december 17, 2009 3rd 2443
Revised
ARTIST: MRL
SIZE
7 col Medical Society. All rights reserved.
Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts
TYPE: Line Combo 4-C H/T 36p6
AUTHOR, PLEASE NOTE:
 The n e w e ng l a n d j o u r na l of m e dic i n e

Self-Reported Patient Global Assessment NYHA Functional Class

Ferric Carboxy- Ferric Carboxy-
maltose Placebo Odds Ratio P Value for maltose Placebo Odds Ratio P Value for
Subgroup no. of patients (95% CI) Interaction no. of patients (95% CI) Interaction
Hemoglobin 0.98 0.51
120 (g/liter) 146 74 148 74
>120 (g/liter) 146 75 146 76
Median ferritin 0.45 0.78
39 (g/liter) 153 72 154 72
>39 (g/liter) 139 77 140 78
Estimated GFR
<60 (ml/min/1.73 m2 119 67 0.22 121 68 0.27
of body-surface area)
60 (ml/min/1.73 m2 173 82 173 82
of body-surface area)
Median age 0.10 0.89
69.7 yr 149 75 149 76
>69.7 yr 143 74 145 74
Sex 0.99 0.29
Male 140 68 142 68
Female 152 81 152 82
NYHA class 0.66 0.17
Class II 52 27 52 27
Class III 240 122 242 123
Median LV ejection 0.86 0.76
fraction
33% 169 70 171 70
>33% 123 79 123 80
Heart failure 0.60 0.55
Nonischemic 56 30 56 30
Ischemic 236 119 238 120
Diabetes 0.87 0.53
No 202 113 204 113
Yes 90 36 90 37
Median BMI 0.94 0.73
27.37 150 71 152 72
>27.37 142 78 142 78
0.5 1 2 4 8 0.5 1 2 4 8

Placebo Better Ferric Carboxymaltose Placebo Better Ferric Carboxymaltose

Better Better

Figure 3. Self-Reported Patient Global Assessment and New York Heart Association (NYHA) Functional Class in Predefined Subgroups,
AUTHOR: Anker RETAKE: 1st
According to Assigned Study Treatment. 2nd
The number of patients in each subgroup FIGURE:
for whom3 ofdata
3 were available is shown. The odds ratios 3rd are given for the ferric carboxymaltose
Revised
group, as compared with the placebo group, on a log
ARTIST: MRL 2 scale. For the results regarding the NYHA functional class in the subgroup of pa-
tients with heart failure of NYHA class II, one patient receiving ferric carboxymaltose had SIZEimprovement to NYHA class I and two pa-
7 col
tients receiving placebo had a worsening to NYHA Line III. Combo
TYPE: class 4-C
The body-mass H/T
index (BMI) is the weight in kilograms divided by the square
36p6
of the height in meters. CI denotes confidence interval, GFR glomerular
AUTHOR, PLEASEfiltration
NOTE: rate, and LV left ventricular.
Figure has been redrawn and type has been reset.
Please check carefully.

was stopped prematurely in 16 (5.3%) of the 304

JOB: 36125 Laboratory
ISSUE: values for ferritin, transferrin sat-
12-17-09

patients assigned to receive ferric carboxymaltose uration, and hemoglobin at week 24 were signifi-
and in 14 (9.0%) of the 155 patients assigned to cantly different between the two study groups
receive placebo. No severe allergic reactions re- (P<0.001 for all comparisons) (Table 3). The mean
lated to the study treatment were reported. Of (SE) difference in the ferritin level (adjusted for
the patients treated with ferric carboxymaltose, baseline) between patients receiving ferric car-
injection-site discoloration was reported for four boxymaltose and those receiving placebo was
patients and injection-site pain for two patients. 24317 g per liter at week 4, 18815 g per liter

2444 n engl j med 361;25 nejm.org december 17, 2009

Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts Medical Society. All rights reserved.
 Ferric Carboxymaltose in Heart Failure and Iron Deficiency

Table 2. Safety End Points and Serious and Nonserious Adverse Events, According to Study Treatment Received.*

Ferric Carboxymaltose Placebo

End Point or Event (N=305) (N=154) P Value
No. of End No. of Patients No. of End No. of Patients
Points or with End Point Points or with End Point
Serious or Event Serious or Event
Adverse/ (incidence/ Adverse/ (incidence/
Any Adverse 100 patient-yr Any Adverse 100 patient-yr
Events at risk) Events at risk)
Safety end point
Death 5 5 (3.4) 4 4 (5.5) 0.47
Death due to cardiovascular causes 4 4 (2.7) 4 4 (5.5) 0.31
Death due to worsening heart failure 0 0 3 3 (4.1)
First hospitalization 28 25 (17.7) 22 17 (24.8) 0.30
Hospitalization for any cardiovascular cause 16 15 (10.4) 18 14 (20.0) 0.08
Hospitalization for worsening heart failure 7 6 (4.1) 9 7 (9.7) 0.11
Any hospitalization or death 33 30 (21.2) 26 19 (27.7) 0.38
Hospitalization for any cardiovascular cause or death 21 20 (13.9) 22 16 (22.9) 0.14
First hospitalization for worsening heart failure or death 12 11 (7.5) 13 10 (13.9) 0.15
Investigator-reported adverse event
Cardiac disorder 12/46 38 (27.6) 23/49 33 (50.2) 0.01
Gastrointestinal disorder 2/29 24 (16.9) 2/7 5 (6.9) 0.06
General disorder or injection-site condition 4/28 23 (16.2) 1/6 6 (8.3) 0.14
Injection-site pain or discoloration 0/6 6 (4.1) 0/0 0
Infection or infestation 2/56 50 (37.0) 0/32 24 (35.8) 0.97
Abnormal laboratory test, vital sign, or physical finding 0/63 32 (23.0) 0/10 10 (14.0) 0.17
Nervous system disorder 3/29 22 (15.6) 3/20 14 (20.3) 0.44
Respiratory, thoracic, or mediastinal disorder 0/9 9 (6.2) 3/13 10 (14.2) 0.06
Vascular disorder 3/24 20 (14.0) 1/13 11 (15.7) 0.80

* Adverse events are classified on the basis of the system organ classes of the Medical Dictionary for Regulatory Activities. Events that were re-
ported in more than 4% of all the study patients are listed here. One patient who had been randomly assigned to the placebo group re-
ceived ferric carboxymaltose.

at week 12, and 24620 g per liter at week 24 difference in estimated glomerular filtration rate
(P<0.001 for all comparisons). The corresponding (adjusted for baseline) between patients receiving
mean differences in the hemoglobin level were ferric carboxymaltose and those receiving placebo
6.61.1, 10.61.3, and 5.91.5 g per liter, respec- was 3.81.8 ml per minute (P=0.03). There were
tively (P<0.001 for all comparisons), and the cor- no significant differences between the two study
responding mean differences in the mean corpus- groups with respect to adverse-event reporting
cular volume of erythrocytes were 1.50.4, 2.40.5, based on chemical values and other hematologic
and 2.70.7 m3, respectively (all P<0.001). The laboratory test results.
mean difference in the hemoglobin level at week
24 (adjusted for baseline) between the ferric car- Discussion
boxymaltose group and the placebo group was not
significant among patients who did not have ane- Treatment with ferric carboxymaltose for 24 weeks
mia at baseline (2.42.0 g per liter, P=0.21) but in patients who had chronic heart failure and iron
was significant among patients who had anemia deficiency with or without anemia improved symp-
at baseline (9.12.2 g per liter, P<0.001). The mean toms, functional capacity, and the quality of life.

n engl j med 361;25 nejm.org december 17, 2009 2445

Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 3. Levels of Iron-Metabolism Markers and Hemoglobin at Week 24 According to Study Treatment.*

Ferric Carboxymaltose Placebo

Variable (N=305) (N=154) P Value
All patients
Ferritin (g/liter) 31213 748 <0.001
Transferrin saturation (%) 291 191 <0.001
Hemoglobin (g/liter) 1301 1251 <0.001
Mean corpuscular volume (m3) 970 941 <0.001
Patients with anemia (hemoglobin 120 g/liter)
Ferritin (g/liter) 27518 6811 <0.001
Transferrin saturation (%) 291 171 <0.001
Hemoglobin (g/liter) 1271 1182 <0.001
Mean corpuscular volume (m3) 981 931 <0.001
Patients without anemia (hemoglobin >120 g/liter)
Ferritin (g/liter) 34919 8011 <0.001
Transferrin saturation (%) 301 221 <0.001
Hemoglobin (g/liter) 1331 1321 0.21
Mean corpuscular volume (m3) 961 951 0.91

* Plusminus values are means SE. The P value is for the mean treatment effect, adjusted for the baseline value. One
patient who had been randomly assigned to the placebo group received ferric carboxymaltose.
The percent transferrin saturation was calculated as iron (in micromoles per liter)transferrin (in grams per li-
ter)25.1.

Our study also showed that treatment with ferric practice to decide whether symptom management,
carboxymaltose was not associated with an un- by means of treatment with intravenous iron, is
acceptable side-effect or adverse-event profile. indicated.
Our study showed improvement with ferric The dose needed to correct iron deficiency was
carboxymaltose in the two primary end points: calculated according to Ganzonis formula21 and
the self-reported Patient Global Assessment and was provided over a period between 3 and 7 weeks
the NYHA class at 24 weeks. The benefit was evi- (a median of six injections) during the correction
dent after 4 weeks and was maintained through- phase. We required that iron was given in doses
out the study period. These results were consistent of 200 mg per application and that the dosing
across all prespecified subgroups and were con- frequency was weekly during the correction phase
firmed by the observed improvements in distance and monthly during the maintenance phase. The
on the 6-minute walk test distance and in scores results of our study are applicable only to this
on the health-related quality-of-life questionnaires. dosing regimen. Treatment approaches involv-
Our patient population was identified on the ing higher doses and higher ferritin thresholds
basis of laboratory biomarkers, ferritin, transfer- for the interruption of therapy are untested in
rin saturation, and hemoglobin. These variables patients with heart failure.
were also used to calculate the iron-repletion dose The treatment with ferric carboxymaltose was
and to guide decisions about continuation or in- beneficial to both patients with anemia and those
terruption of ferric carboxymaltose. In that sense, without anemia. This suggests that iron deficiency
the trial is a double-blind treatment trial that used is a valid independent therapeutic target. Iron me-
guidance based on blood-sample data for study- tabolism in patients with chronic illness merits a
treatment initiation and monitoring. The study more detailed investigation to unravel the reasons
results suggest that in the assessment of ambu- why the correction of iron deficiency can result
latory patients with symptomatic heart failure and in symptomatic improvements even in the absence
systolic dysfunction, laboratory investigations to of a change in hemoglobin. Our results are con-
detect iron deficiency may be useful in routine sistent with those from four small studies that

2446 n engl j med 361;25 nejm.org december 17, 2009

Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts Medical Society. All rights reserved.
 Ferric Carboxymaltose in Heart Failure and Iron Deficiency

were performed with the use of a different prepa- The benefit was seen in patients with anemia and
ration of intravenous iron.16-19 One of these stud- in those without anemia.
ies15 recruited a subgroup of patients without Sponsored by Vifor Pharma.
Dr. Anker reports receiving lecture fees from Roche Pharma
anemia who had even higher hemoglobin levels and Teva; Drs. Anker, Comin Colet, Filippatos, Willenheimer,
than our patients (i.e., 125 to 145 g per liter). We Dickstein, Lscher, and Ponikowski, fees from Vifor Pharma as
do not know the limit of hemoglobin up to which members of the FAIR-HF Executive Committee; Drs. Anker, Wil-
lenheimer, and Ponikowski, lecture and consulting fees from Vifor
iron deficiency is pathophysiologically important. Pharma and Amgen; and Dr. Willenheimer, lecture fees from Merck
On the basis of our current findings, we cannot and Servier; Dr. Kirwan reports being an employee of SOCAR Re-
recommend treatment with ferric carboxymaltose search, which received fees from Vifor Pharma; Drs. Mori and von
Eisenhart Rothe report being employees of Vifor Pharma and
for patients who have chronic heart failure, iron owning stock in Galenica; and Dr. Pocock reports receiving fees
deficiency, and a hemoglobin level above 135 g per from Vifor Pharma as the FAIR-HF consultant statistician. Finan-
liter, but such therapy is an area of interest for cial and other disclosures provided by the authors are available
with the full text of this article at NEJM.org.
future research. We thank the study coordinators, nurses, and staff at the in-
In conclusion, in stable, symptomatic, ambula- vestigative sites and especially all of the patients involved in the
tory patients with chronic heart failure, an im- FAIR-HF trial; Elena Malyshko (ClinStar) and Miriana Wennekes
(Kendle) and their study teams for study monitoring and manage-
paired left ventricular ejection fraction, and iron ment; Sophie De Brouwer (SOCAR Research) and Tim Clayton
deficiency, treatment with ferric carboxymaltose (London School of Hygiene and Tropical Medicine) for the statisti-
over a 24-week period improves symptoms, phys- cal analysis; and Nicola Waddingham and Giedrius Gaudesius
(Vifor Pharma) for overall study coordination. We dedicate this
ical performance, and the quality of life and has article to the memory of our colleagues Philip A. Poole-Wilson and
acceptable side-effect and adverse-event profiles. Helmut Drexler, who did not live to see the results of this trial.

Appendix
The members of the FAIR-HF study group are as follows: Executive Committee: S.D. Anker (chair), P. Ponikowski (cochair), K. Dickstein,
G.S. Filippatos, T.F. Lscher, R. Willenheimer, J. Comin Colet, H. Drexler (deceased), P.A. Poole-Wilson (deceased); and Data Safety and
Monitoring Board: R.P. Wthrich (chair), J. Lubsen. The FAIR-HF site investigators and institutions were as follows (with numbers of
recruited patients per country in parentheses): Argentina (6): J.H. Altamirano, Instituto de Investigaciones Cardiolgicas Prof. Dr. Al-
berto C. Taquini, Buenos Aires; S.V. Rodriguez, Instituto de Diagnstico Cardiovascular La Plata, Buenos Aires; Czech Republic (17): J.
pinar, Fakultn Nemocnice Brno, Brno; J. Povoln, Oblastn Nemocnice Kladno, Kladno; J. Blohlvek, Vseobecna Fakultn Nemocnice,
Prague; Z. Palch, Fakultn Thomayerova Nemocnice, Prague; J. Hork, Fakultn Nemocnice Kralovske Vinohrady, Prague; Germany (11):
S. Anker, Charit Berlin, Campus Virchow-Klinikum, Berlin; D. Wolf, CardioSec Clinical Research, Erfurt; M. Natour, Praxis Dr. M.
Natour und M. Durak, Heidelberg; Greece (11): A. Manolis, General Hospital of Voula Asklipion, Athens; J. Nanas, Alexandras General
Hospital, Athens; D.T. Kremastinos, University Hospital of Athens Attikon, Athens; D. Alexopoulos, University Hospital of Patras, Pa-
tras; Italy (11): M. Volterrani, Istituto di Ricovero e Cura a Carattere Scientifico San Raffaele Pisana, Rome; G. Vescovo, Unit Locale
Socio Sanitaria 6 Ospedale San Bortolo, Vicenza; Norway (2): K. Dickstein, Stavanger Helseforskning, Stavanger; Poland (60): M. Mysli-
wiec, Wojewodzki Szpital Specjlaistyczny im K. Dluskiego, Biaystok; M. Ogorek, Samodzielny Szpital Wojewodzki im M. Kopernika,
Piotrkow Trybunalski; P. Staneta, Niepubliczny zakad Opieki Zdrowotnej Specjalistyczna Przychodnia Lekarska, Medikard, Plock; J.
Niegowska, Centrum Medyczne, Telmont Centrum Medyczne, Sp. z O.O., Warsaw; M. Duzniewski, Wojewdzki Szpital Brdnowski,
Warsaw; P. Ponikowski, 4 Wojskowy Szpital Kliniczny z Poliklinik, Wroclaw; L. Polonski, Slaskie Centrum Chorob Serca, Zabrze;
Romania (16): M. Radoi, Spitalul Clinic de Urgenta Brasov, Brasov; C.E. Macarie, Institutul de Boli Cardiovasculare Prof. Dr. C.C. Iliescu,
Bucharest; G.A. Dan, Spitalul Clinic Colentina, Bucharest; P.I. Kikeli, Societatea Civila Medicala Procardia Medical, Targu Mures; Russia
(200): E.S. Pasechnik, State Institution of Healthcare Kaluzhskaya Regional Hospital, Kaluga; A.A. Eremenko, Russian Scientific Center
of Surgery of the Russian Academy of Medical Sciences, Moscow; A.A. Gorbachenkov, Central Clinical Hospital of Civil Aviation of the
Ministry of Transport of the Russian Federation, Moscow; A.E. Bragina, City Clinical Hospital 61, Moscow; A.Y. Ivleva, Outpatient
Clinic 3 of President of the Russian Federation Administration, Moscow; B.Y. Bart, Russian State Medical University, Outpatient Diag-
nostic Consulting Center 1, Moscow; G.P. Arutyunov, City Clinical Hospital 4, Moscow; R.A. Khokhlov, Voronezhsky Regional Hospi-
tal 1, Voronezh; S.N. Tereschenko, Moscow State Medicine and Dentistry University of the Federal Agency for Health Care and Social
Development, Moscow; V.A. Lusov, Russian State Medical University, Moscow; V.N. Ardashev, Main Military Clinical Hospital, Moscow;
V.Y. Mareev, Cardiology Research Complex, Moscow; Z.D. Kobalava, Russian State People Friendship University, Moscow; O.P. Alexe-
eva, City Hospital 33, Nizhny Novgorod; V.E. Oleynikov, State Institution of Healthcare Regional Clinical Hospital, Penza; N.A. Kozi-
olova, Permsky Regional Hospital of War Veterans, Perm; D.V. Duplyakov, State Institute of Healthcare Regional Cardiology Dispen-
sary of Samara, Samara; A.A. Petrov, Leningrad Regional Clinical Hospital, St. Petersburg; D.U. Butko, International Clinic and Hospi-
tal Medem, St. Petersburg; M.Y. Sitnikova, Research Institute of Cardiology, St. Petersburg; O.A. Berkovich, St. Petersburg State Medical
University, St. Petersburg; S.A. Boldueva, St. Petersburg State Medical Academy, St. Petersburg; S.R. Minkin, City ConsultiveDiagnos-
tical Center 1, St. Petersburg; V.A. Kostenko, St. Petersburg Scientific Research Institute for Emergency, St. Petersburg; Y.G. Shvarts,
Saratov State Medical University of Roszdrav, Saratov; Y.B. Karpov, Voronezhsky Hospital of Emergency 1, Voronezh; S.V. Nedogoda,
Regional Clinical Hospital of Volgograd, Volgograd; Spain (22): J. Bruguera, Hospital del Mar, Barcelona; J. Gonzlez Costell, Hospital
Universitari de Bellvitge, Barcelona; M. Martnez Sells, Hospital General Universitario Gregorio Maraon, Madrid; J. Quiles Granado,
Hospital Universitario San Juan de Alicante, San Juan de Alicante; Ukraine (103): O.A. Koval, Dnipropetrovsk State Medical Academy,
Hospital Therapy 2, Dnipropetrovsk; G.A. Ignatenko, Donetsk State Medical University, Donetsk; I.G. Kraiz, Central Clinical Hospital
of Ukrzaliznitsya, Kharkiv; O.V. Prohorov, City Clinical Hospital 27, Kharkiv; V.I. Tseluyko, Kharkiv Medical Academy of Postgraduate
Education, Kharkiv; O.M. Parkhomenko, National Scientific Center Institute of Cardiology Academician Strazheska, Kiev; K.M. Amos-
ova, National Medical University, Kiev; L.V. Rudenko, Clinical Hospital of Emergency Medical Service, Kiev; O.M. Gyrina, National

n engl j med 361;25 nejm.org december 17, 2009 2447

Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts Medical Society. All rights reserved.
 Ferric Carboxymaltose in Heart Failure and Iron Deficiency

Medical University, Kiev; V.Y. Lishnevska, Institute of Gerontology of the Academy of Medical Science of Ukraine, Kiev; S.S. Pavlyk,
Regional State Clinical Treatment-and-Diagnostic Cardiology Center, Lviv; I.P. Kovalsky, City Hospital 1 of Mykolayiv, Mykolayiv; G.S.
Popik, Odessa State Medical University, Odessa; Z.Y. Telyatnikova, Municipal Institution, City Policlinic 20, Odessa; I.V. Chopey, Uzhgorod
National University, Uzhgorod; V.A. Vizir, Zaporizhzhya State Medical University, Zaporizhzhya.

References
1. Hunt SA, Abraham WT, Chin MH, et new millennium. Trends Cell Biol 2007; ed trial. J Am Coll Cardiol 2008;51:103-
al. 2009 Focused update incorporated 17:93-100. 12.
into the ACC/AHA 2005 Guidelines for 9. Fairbanks V, Beutler E. Iron deficiency. 18. Toblli JE, Lombraa A, Duarte P, Di
the Diagnosis and Management of Heart In: Beutler E, ed. Williams hematology. Gennaro F. Intravenous iron reduces NT-
Failure in Adults: a report of the Ameri- 6th ed. New York: McGraw-Hill, 2001:295- pro-brain natriuretic peptide in anemic
can College of Cardiology Foundation/ 304, 447-70. patients with chronic heart failure and
American Heart Association Task Force 10. Beard JL. Iron biology in immune renal insufficiency. J Am Coll Cardiol
on Practice Guidelines developed in col- function, muscle metabolism and neuronal 2007;50:1657-65.
laboration with the International Society functioning. J Nutr 2001;131:Suppl 2: 19. Usmanov RI, Zueva EB, Silverberg DS,
for Heart and Lung Transplantation. J Am 568S-579S. Shaked M. Intravenous iron without eryth-
Coll Cardiol 2009;53(15):e1-e90. 11. Haas JD, Brownlie T IV. Iron deficien- ropoietin for the treatment of iron defi-
2. Dickstein K, Cohen-Solal A, Filippa- cy and reduced work capacity: a critical ciency anemia in patients with moderate
tos G, et al. ESC guidelines for the diag- review of the research to determine a to severe congestive heart failure and
nosis and treatment of acute and chronic causal relationship. J Nutr 2001;131:Suppl chronic kidney insufficiency. J Nephrol
heart failure 2008: the Task Force for the 2:676S-688S. 2008;21:236-42.
diagnosis and treatment of acute and 12. Davies KJ, Maguire JJ, Brooks GA, 20. Anker SD, Colet JC, Filippatos G, et al.
chronic heart failure 2008 of the Europe- Dallman PR, Packer L. Muscle mitochon- Rationale and design of Ferinject Assess-
an Society of Cardiology: developed in col- drial bioenergetics, oxygen supply, and ment in patient with IRon deficiency and
laboration with the Heart Failure Associa- work capacity during dietary iron defi- chronic Heart Failure (FAIR-HF) study: a
tion of the ESC (HFA) and endorsed by the ciency and repletion. Am J Physiol 1982; randomised, placebo controlled study of
European Society of Intensive Care Medi- 242(6):E418-E427. intravenous iron supplementation in pa-
cine (ESICM). Eur J Heart Fail 2008;10: 13. Satija P, Ondo WG. Restless legs syn- tients with and without anaemia. Eur J
933-89. drome: pathophysiology, diagnosis and Heart Fail 2009;11:1084-91.
3. Jessup M, Brozena S. Heart failure. treatment. CNS Drugs 2008;22:497-518. 21. Ganzoni AM. Intravenous iron-dex-
N Engl J Med 2003;348:2007-18. 14. Opasich C, Cazzola M, Scelsi L, et al. tran: therapeutic and experimental pos-
4. Roger VL, Weston SA, Redfield MM, Blunted erythropoietin production and sibilities. Schweiz Med Wochenschr 1970;
et al. Trends in heart failure incidence and defective iron supply for erythropoiesis as 100:301-3. (In German.)
survival in a community-based population. major causes of anaemia in patients with 22. Green CP, Porter CB, Bresnahan DR,
JAMA 2004;292:344-50. chronic heart failure. Eur Heart J 2005; Spertus JA. Development and evaluation
5. Clark AL, Poole-Wilson PA, Coats AJ. 26:2232-7. of the Kansas City Cardiomyopathy ques-
Exercise limitation in chronic heart fail- 15. Nanas JN, Matsouka C, Karageorgo- tionnaire: a new health status measure for
ure: central role of the periphery. J Am poulos D, et al. Etiology of anemia in pa- heart failure. J Am Coll Cardiol 2000;
Coll Cardiol 1996;28:1092-102. tients with advanced heart failure. J Am 35:1245-55.
6. Massie BM, Conway M, Rajagopalan Coll Cardiol 2006;48:2485-9. 23. Rabin R, de Charro F. EQ-5D: a mea-
B, et al. Skeletal muscle metabolism dur- 16. Bolger AP, Bartlett FR, Penston HS, et sure of health status from the EuroQol
ing exercise under ischemic conditions in al. Intravenous iron alone for the treat- Group. Ann Med 2001;33:337-43.
congestive heart failure: evidence for ab- ment of anemia in patients with chronic 24. McCullagh P. Regression models for
normalities unrelated to blood flow. Cir- heart failure. J Am Coll Cardiol 2006;48: ordinal data (with discussion). J R Stat
culation 1988;78:320-6. 1225-7. Soc [B] 1980;42:109-42.
7. Tang YD, Katz SD. Anemia in chronic 17. Okonko DO, Grzeslo A, Witkowski T, 25. Benjamini Y, Hochberg Y. Controlling
heart failure: prevalence, etiology, clinical et al. Effect of intravenous iron sucrose on the false discovery rate: a practical and
correlates, and treatment options. Circu- exercise tolerance in anemic and nonane- powerful approach to multiple testing. J R
lation 2006;113:2454-61. mic patients with symptomatic chronic Stat Soc [B] 1995;57:289-300.
8. Dunn LL, Rahmanto YS, Richardson heart failure and iron deficiency FERRIC-HF: Copyright 2009 Massachusetts Medical Society.
DR. Iron uptake and metabolism in the a randomized, controlled, observer-blind-

collections of articles on the journals web site

The Journals Web site (NEJM.org) sorts published articles into
more than 50 distinct clinical collections, which can be used as convenient
entry points to clinical content. In each collection, articles are cited in reverse
chronologic order, with the most recent first.

2448 n engl j med 361;25 nejm.org december 17, 2009

Downloaded from www.nejm.org on January 6, 2010 . Copyright 2009 Massachusetts Medical Society. All rights reserved.