Author + information
- Received June 14, 2006
- Revision received August 30, 2006
- Accepted September 11, 2006
- Published online February 20, 2007.
- Piotr Ponikowski, MD, PhD⁎,3,6,
- Stefan D. Anker, MD, PhD†,‡,1,3,4,6,
- Joanna Szachniewicz, MD⁎,
- Darlington Okonko, BSc, MRCP†,2,
- Mark Ledwidge, PhD§,
- Robert Zymlinski, MD⁎,
- Enda Ryan, MRCP§,
- Scott M. Wasserman, MD∥,
- Nigel Baker, MSc∥,5,
- Dylan Rosser, BSc∥,5,
- Stuart D. Rosen, MD†,
- Philip A. Poole-Wilson, MD†,
- Waldemar Banasiak, MD, PhD⁎,
- Andrew J.S. Coats, DM¶,3,4 and
- Ken McDonald, MD§,3,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Ken McDonald, St. Vincent’s University Hospital, Elm Park, Dublin 4, Ireland.
Objectives This study sought to investigate whether darbepoetin alfa, an erythropoiesis-stimulating protein (ESP), improves exercise capacity in patients with symptomatic chronic heart failure (CHF) and anemia.
Background Anemia is common in patients with CHF.
Methods In a multicenter, randomized, double-blind, placebo-controlled study, CHF patients with anemia (hemoglobin ≥9.0 to ≤12.0 g/dl) received subcutaneous placebo (n = 22) or darbepoetin alfa (n = 19) at a starting dose of 0.75 μg/kg every 2 weeks for 26 weeks. The primary end point was change in exercise tolerance from baseline to week 27 as measured by peak oxygen uptake (ml/min/kg body weight). Other end points included changes in absolute peak Vo2(ml/min), exercise duration, and health-related quality of life.
Results Differences (95% confidence interval) in mean changes from baseline to week 27 between treatment groups were 1.5 g/dl (0.5 to 2.4) for hemoglobin concentration (p = 0.005), 0.5 ml/kg/min (−0.7 to 1.7) for peak Vo2(p = 0.40), 45 ml/min (−35 to 127) for absolute peak Vo2(p = 0.27), and 108 s (−11 to 228) for exercise duration (p = 0.075). Patients receiving darbepoetin alfa compared with placebo had an improvement in self-reported Patient’s Global Assessment of Change (79% vs. 41%, p = 0.01) but no significant differences in the Kansas City Cardiomyopathy and Minnesota Living with Heart Failure Questionnaire scores. Darbepoetin alfa was well tolerated.
Conclusions In patients with symptomatic CHF and anemia, darbepoetin alfa increased and maintained hemoglobin concentrations and improved health-related quality of life. A trend toward increased exercise time but not peak Vo2was observed. (Impact of Darbepoetin Alfa on Exercise Tolerance and Left Ventricular Structure in Subjects With Symptomatic Congestive Heart Failure (CHF) and Anemia; http://clinicaltrials.gov/ct/show/NCT00117234?order = 1; NCT00117234).
Anemia is recognized as a common comorbidity in patients with chronic heart failure (CHF) (1–3). Initial data from a recent, large observational study of patients with CHF attending cardiology clinics showed that 33% had a hemoglobin concentration <12.0 g/dl (4). Several studies have shown that anemia in patients with CHF is associated with an increase in left ventricular mass (2), a greater incidence of rehospitalization (5), and higher mortality (2,5–8). Recently, a small number of preliminary studies have reported that correction of low hemoglobin concentrations using erythropoiesis-stimulating proteins (ESPs) may improve cardiac and renal function and reduce the need for hospitalization and diuretics in patients with severe HF (9,10).
Another consequence of anemia in CHF is reduced exercise tolerance and poor health-related quality of life (1,11,12). One small, single-center, randomized, single-blinded study has suggested that treatment of anemic CHF patients with an ESP may improve measures of exercise tolerance, such as peak Vo2and exercise duration, as well as health-related quality of life (13). To date, no study has investigated these end points in a larger randomized, double-blind, placebo-controlled trial.
We report here the results of a randomized, double-blind, placebo-controlled study that determined the effect of treatment with subcutaneous darbepoetin alfa, a long-acting ESP that stimulates erythropoiesis in the same manner as endogenous erythropoietin (14), on exercise tolerance as measured by changes in peak Vo2, absolute peak Vo2, and exercise duration in patients with symptomatic CHF and anemia. Other end points included safety, health-related quality of life, functional status (New York Heart Association [NYHA] functional class), brain natriuretic peptide (BNP) concentration, body weight, and on-study hospitalization.
Individual institutional review committees approved the study, and all patients gave written informed consent before any study-specific procedures were performed. The main inclusion criteria were: age ≥21 years; symptomatic CHF with exercise limitation on the treadmill; reproducible peak Vo2≤16 ml/kg/min during screening obtained from at least 2 and up to 5 exercise tests at separate visits, with results from the last completed test within 15% of the previous one; hemoglobin concentration ≥9.0 to ≤12.0 g/dl; left ventricular ejection fraction (LVEF) ≤40%; receiving stable, recommended therapy for CHF (angiotensin-converting enzyme inhibitor or angiotensin receptor blocker, and a beta-blocker if tolerated); not likely to receive cardiac transplantation within 6 months after randomization; being iron replete (transferrin saturation ≥15%); resting blood pressure <160/100 mm Hg; and serum creatinine ≤3.0 mg/dl. Patients were excluded if they received blood transfusions or ESP within 12 weeks of randomization or were receiving any other investigational drugs.
This was a phase 2, randomized, double-blind, placebo-controlled study conducted in 3 centers (Fig. 1).Patients were randomly assigned in a 1:1 ratio to receive either subcutaneous placebo or darbepoetin alfa at a starting dose of 0.75 μg/kg every 2 weeks for 26 weeks. Darbepoetin alfa (Aranesp, Amgen, Inc., Thousand Oaks, California) was provided in single-dose vials (200 μg/ml). Placebo was supplied in identical single-dose vials. Randomization was within the study centers in blocks of 2 and 4, and without further stratification. The randomization sequence, which was computer-generated and held by an independent statistician, was implemented via a central randomization coordinator and was concealed from investigators throughout the study. Patients and personnel at the study centers were blinded to the identity of the study drug, hematology results (except at baseline), and hormone markers. The central randomization coordinator calculated the dose in μg/kg and the volume of drug to be administered, providing only information on the volume.
Hemoglobin concentration was monitored weekly, with central data assessment at the core laboratory. During the titration phase, darbepoetin alfa doses were titrated by ±25% of the current dose to maintain a gradual hemoglobin concentration increase of ≥0.5 g/dl but ≤1.5 g/dl every 3 weeks. When hemoglobin was ≥13.0 g/dl and had increased by at least 2.0 g/dl from baseline, the dose was reduced by 25% from the current dose. During the maintenance phase, the dose was adjusted by ±25% to maintain the hemoglobin concentration at the target of 13.0 to 15.0 g/dl. If hemoglobin was ≥15.5 g/dl, study drug was withheld; dosing resumed at the previous schedule and at 75% of the most recent dose after at least 2 weeks had passed and once the hemoglobin concentration was ≤14.0 g/dl. To maintain the study blind, volume changes of placebo were implemented to mimic the dose changes in the darbepoetin alfa group. All patients were prescribed 200 to 300 mg/day of elemental oral iron, unless baseline serum ferritin was >800 ng/ml. The treatment phase lasted for 26 weeks, and patients underwent a safety follow-up examination 4 weeks after receiving the final dose of study drug.
The primary end point was the change from baseline to week 27 in exercise tolerance, as measured by peak Vo2adjusted for body weight (ml/kg/min). The study was designed to have approximately 80% power to detect a difference of 1.1 ml/kg/min between treatment groups in mean change in peak Vo2from baseline to week 27, assuming SD of 1.3 ml/kg/min and a significance level of 0.05. This effect was estimated using data from a previous clinical study (13), and was chosen on the basis of clinical relevance.
Other end points included changes from baseline to week 27 in absolute peak Vo2(ml/min), exercise duration (in seconds), hemoglobin, NYHA functional classification, health-related quality of life, BNP, body weight, and on-study hospitalizations. Health-related quality of life was assessed using 2 different disease-specific instruments, the Kansas City Cardiomyopathy Questionnaire (KCCQ) (15) and the Minnesota Living With Heart Failure Questionnaire (MLHFQ) (16), as well as a brief generic instrument, the Patient’s Global Assessment of Change (PGA) (17). Safety end points were the incidence of adverse events, changes in laboratory measurements and vital signs, and seroreactivity to ESPs.
Cardiopulmonary exercise testing
Exercise testing was carried out on a treadmill using a modified Naughton protocol (18). Measurements of oxygen uptake, carbon dioxide production, and minute ventilation were taken breath-by-breath and averaged over 15-s intervals. Peak O2uptake was measured as an average of the last 15 s of exercise. Maximal exercise capacity was attained if the respiratory exchange ratio was >1.00 and had increased by at least 0.15 from the resting value.
All data were analyzed according to the intent-to-treat principle. Imputations were made using the last-observation-carried-forward method. Pre-specified sensitivity analyses to evaluate the impact of missing data were also performed for the primary end point, including a repeated measures analysis of covariance model, and an analysis based on the completer set (subjects who completed 27 weeks of treatment). These analyses included baseline peak Vo2and study center as covariates. The results from these sensitivity analyses were similar to the results from the primary analysis with last value carried forward imputation. For the primary analysis, an analysis of covariance model was used, with treatment as the main effect and baseline peak Vo2and study center as covariates. Other statistical analyses included summary statistics (point estimates of week-27 peak Vo2and change from baseline to week 27 for each treatment group); estimates and 2-sided 95% confidence intervals (95% CI) for the difference between least-squares means of the 2 treatment groups, adjusted for the effects of center and baseline values of the variable being analyzed. Results were considered to be statistically significant if p < 0.05. Health-related quality of life domain scores were computed using published scoring guidelines (15,19). All analyses were carried out using SAS software version 8.2 (SAS Institute Inc., Cary, North Carolina).
Of 93 CHF patients screened, 52 (56%) were excluded from the study. The main reasons were LVEF >40% (n = 16), hemoglobin concentration not within specified range (n = 16), and peak Vo2>16 ml/kg/min (n = 10). Forty-one patients were randomly assigned to receive either darbepoetin alfa (n = 19) or placebo (n = 22). No patients withdrew before receiving the study drug, but 3 patients (16%) in the darbepoetin alfa group (adverse event, n = 2; withdrawn consent, n = 1) and 3 (14%) in the placebo group (death, n = 1; withdrawn consent, n = 1; protocol-specified criteria, n = 1) discontinued the study early.
Treatment groups were generally balanced with respect to demographics and baseline characteristics (Tables 1 and 2).⇓⇓The characteristics of patients enrolled at separate centers did not affect the study results. There were some differences in hemoglobin values determined by the core laboratory and by the investigative sites, respectively. Consequently, 17 patients (41%) were recruited with hemoglobin concentrations >12.0 g/dl (n = 10 in the darbepoetin alfa and n = 6 in the placebo group, with hemoglobin >12.0 to 13.0 g/dl; n = 0 in the darbepoetin alfa and n = 1 in the placebo group, with hemoglobin >13.0 g/dl). All 17 patients met eligibility criteria specified in the protocol because screening hemoglobin concentrations obtained at local laboratories were <12.0 g/dl.
All patients randomly assigned to treatment were assessed for the primary end point (imputed peak Vo2values were used for 8 patients, 4 in each treatment group). Exercise tolerance, as measured by the change in peak Vo2from baseline to week 27, did not change significantly in either treatment group (Table 3,Fig. 2A).The mean treatment difference was 0.5 ml/kg/min (95% CI −0.7 to 1.7, p = 0.40). Sensitivity analyses of pre-specified baseline factors (e.g., gender, age, race, smoking history, beta-blocker use, duration of CHF, and NYHA functional class) showed no significant effect on change in exercise tolerance.
From baseline to week 27, absolute peak Vo2(mean ± SE) in the darbepoetin alfa arm was 43 ± 30 ml/min compared with placebo −3 ± 28 ml/min (Table 3, Fig. 2B). The mean treatment difference in absolute peak Vo2changes was 45 ml/min (95% CI −35 to 127, p = 0.27).
Over the course of the study, exercise duration (adjusted mean ± SE) decreased by 67 ± 63 s in the placebo group, compared with an increase of 42 ± 59 s in the darbepoetin alfa group (Table 3, Fig. 2C). The adjusted mean treatment difference between the 2 groups was 108 s (95% CI −11 to 228, p = 0.075).
In an exploratory analysis, no significant correlation between change in hemoglobin concentration and change in exercise tolerance (peak Vo2and absolute peak Vo2) was evident in either treatment group (all r <0.37, p > 0.12). A significant association was found between changes in hemoglobin concentration and changes in exercise duration in patients receiving darbepoetin alfa, but not in the placebo group (r = 0.64, p = 0.003) (Fig. 3).
Reproducibility of cardiopulmonary exercise testing and power calculation on the basis of the actual results
The within-subject coefficients of variance for peak Vo2that were derived from the final 2 valid exercise tests (last screening and baseline) ranged from 0.01% to 28% with a mean ± SE coefficient of variance of 5.2% ± 1.0%. When the results were split by the 3 participating centers, they did not significantly vary, with the following mean (range) coefficients of variance: 2.1% (0.04% to 3.6%), 8.7% (0.5% to 28%) and 4.3% (0.01% to 18%). The SD of the changes from baseline to week 27 in peak Vo2for patients randomly assigned to the placebo group was 2.23 ml/kg/min (range −7.9 to 3.4 ml/kg/min), with the following SDs (ranges) at the center: Dublin 0.69 ml/kg/min (0.1 to 1.4 ml/kg/min), London 0.78 ml/kg/min (−1.6 to 3.4 ml/kg/min), Wroclaw 2.23 ml/kg/min (−7.9 to 0.33 ml/kg/min). The observed SD for repeated assessments was larger than the planned SD (1.3 ml/kg/min difference), indicating high variability.
Secondary end point assessments
Change in hemoglobin concentration
During the study, an increase in hemoglobin concentrations was observed in the darbepoetin alfa group, with little change occurring in the placebo group (Table 3, Fig. 4).At the end of the study, the change from baseline in hemoglobin was 2.4 ± 0.4 g/dl vs. 0.9 ± 0.5 g/dl for the darbepoetin alfa and placebo groups, respectively. The difference in the mean change between the treatment groups was 1.5 g/dl (range 0.5 to 2.4 g/dl) (p = 0.005). Mean values for iron parameters were similar and either in the normal range for serum iron and ferritin, or at the low end of the normal ranges for total iron binding capacity and transferrin saturation, for both treatment groups at baseline and throughout the study (data not shown).
NYHA functional class
In a continuous analysis of changes in NYHA functional class, the mean changes from baseline to week 27 were −0.11 (range −0.30 to 0.08) and −0.09 (range −0.31 to 0.13) for the darbepoetin alfa and placebo groups, respectively. Only 2 patients (11%) in the darbepoetin alfa arm experienced an improvement from baseline to week 27 of 1 NYHA functional class or more, compared with 6 patients (27%) in the placebo group (p = 0.25).
Health-related quality of life
A significantly larger proportion of patients receiving darbepoetin alfa reported an improvement at week 27 in PGA, compared with patients receiving placebo: 15 of 19 (79%) vs. 9 of 22 (41%) patients, respectively (p = 0.01) (Fig. 5).When health-related quality of life was measured using the KCCQ, there were no statistically significant differences (Table 4)in the mean change from baseline to week 27 between darbepoetin alfa and placebo groups in either of the KCCQ individual domain scores or the 3 summary scores (data not shown). Clinically significant changes of ≥5 points (20,21) were recorded for the individual domain, and summary scores for both the placebo and darbepoetin alfa groups (Table 4). Assessment of health-related quality of life using the MLHFQ instrument showed no significant difference between the darbepoetin alfa and placebo groups in the mean changes from baseline to week 27 for all 3 scores. Clinically significant changes of ≥5 points (22,23) were recorded for the total MLHFQ score for both the placebo and darbepoetin alfa groups (Table 4). The magnitude of clinically significant changes for the emotional and physical MLHFQ domains is not well defined.
Occurrences of adverse events for both treatment groups are listed in Table 5.Fifteen (79%) and 13 (59%) patients receiving darbepoetin alfa and placebo, respectively, experienced at least 1 adverse event while being treated. Two deaths occurred during the study, 1 in each treatment group. The causes of death were recorded as circulatory insufficiency and sudden death in the darbepoetin alfa and placebo groups, respectively. Neither death was considered by the investigators to be related to treatment.
Events that occurred with a >5% difference-in-incidence rate between the treatment groups are detailed in Table 5. Those that were more common in the darbepoetin alfa group vs. placebo were neurologic signs and symptoms not elsewhere classifiable, upper respiratory tract infections, and joint-related signs. The incidences of all other adverse events were similar for the 2 treatment groups. One adverse event (neurologic signs and symptoms not elsewhere classifiable) in the darbepoetin alfa group was considered to be related to treatment by the investigators. Six patients (n = 3 [16%] in the darbepoetin alfa and n = 3 [14%] in the placebo group) experienced at least 1 of the 7 adverse events considered to be of specific interest in patients receiving ESPs (Table 5). No occurrences of deep vein thrombosis, pulmonary embolus, or seizure were reported during the study. With the exception of 1 episode of hypertension recorded in the darbepoetin alfa group, blood pressure remained stable throughout the study. No clinically relevant changes occurred in clinical chemistry, hematology, or vital signs. No anti-darbepoetin alfa antibodies were detected (data not shown).
Mean ± SD baseline serum concentrations for BNP were 418.8 ± 269.4 and 498.3 ± 787.6 pg/ml for the darbepoetin alfa and placebo groups, respectively. The mean changes from baseline to week 27 were not significantly different between the 2 treatment groups: −90.5 pg/ml (range −196.8 to 15.8 pg/ml) and −26.5 pg/ml (range −121.0 to 68.0 pg/ml) for darbepoetin alfa and placebo, respectively. The mean treatment difference between the groups was –64.0 pg/ml (95% CI −192.7 to 64.7 pg/ml, p = 0.32).
The study results suggested a possible preservation of body weight in patients receiving darbepoetin alfa (mean ± SE change from baseline to week 27, +0.1 ± 1.0 kg, compared with −1.2 ± 0.9 kg in the placebo group). The mean difference between the treatment groups was not statistically significant (+1.3 kg, 95% CI −1.3 to 3.9 kg, p = 0.33).
Fewer patients were hospitalized in the darbepoetin alfa arm than in the placebo arm. Specifically, 4 (21%) and 9 (41%) patients in the darbepoetin alfa and placebo group, respectively, required hospitalization for any reason (p = 0.20), whereas the proportion of patients who were hospitalized for any CHF-related reason was 11% and 14%, respectively (p = 1.0).
Serum creatinine, estimated glomerular filtration rate, and transferrin saturation
At baseline and after 27 weeks, the mean serum creatinine concentrations (in mg/dl) were 1.45 (SD 0.53) and 1.43 (SD 0.42) for placebo and 1.32 (SD 0.48) and 1.24 (SD 0.40) for drug. The estimated glomerular filtration rates (ml/min/1.73 m2) were 52 (SD 25) and 49 (SD 22) for placebo and 59 (SD 23) and 61 (SD 20) for drug. The (%) were 34.6 (SD 12.8) and 27.7 (SD 8.9) for placebo and 25.3 (SD 6.6) and 32.1 (SD 13.3) for drug.
This report describes the first randomized, double-blind, placebo-controlled study to evaluate whether therapy with darbepoetin alfa improves exercise tolerance in patients with symptomatic CHF and anemia. The results show that darbepoetin alfa treatment significantly increased hemoglobin concentrations and resulted in improvement in 1 measure of health-related quality of life as determined by PGA. We could not show a statistically significant increase in exercise tolerance as measured by either peak Vo2adjusted for body weight or absolute peak Vo2, but there was a trend for an increase in exercise duration. In patients treated with darbepoetin alfa, a statistically significant association existed between changes in hemoglobin and changes in exercise duration. Therapy with darbepoetin alfa was well tolerated and showed an adverse event profile similar to that of placebo.
Anemia is a common comorbidity in CHF, increasing in prevalence in the more advanced stages of the disease (2). The etiology of anemia in patients with CHF is not completely understood, but is associated with many of the hematologic features of anemia of chronic disease (24). In CHF patients, anemia is independently associated with increased rates of hospital admissions and poor survival (1–3,6), but little is known about the impact of treating anemia in these patients. Initial studies using ESPs have shown improved cardiac function, functional capacity, health-related quality of life, and exercise tolerance (10,13). Because of study limitations such as small sample size and lack of double-blind and/or placebo-controlled design, no definitive conclusions could be drawn. It has been reported elsewhere (25) that a certain proportion of CHF patients with anemia adjust their hemoglobin concentrations over time. Over the 26-week treatment period analyzed in the study presented, none of the CHF patients spontaneously had corrected anemia, and there was no significant increase in hemoglobin concentrations in patients treated with placebo.
When interpreting the results of the exercise testing conducted in this study, several issues need to be considered. The reproducibility of the exercise tests performed at baseline was excellent (mean coefficient of variance 5.7%). However, the SD of the change from baseline to week 27 in peak Vo2for subjects randomly assigned to placebo (2.2 ml/kg/min) exceeded the expected value as used in the original power calculation (1.3 ml/kg/min). The preservation and loss of weight seen in the darbepoetin alfa and placebo groups, respectively, may have contributed to this. Thus, the variability of the data may have had an impact on the study’s ability to detect the desired treatment effect of 1.1 ml/min/kg. In a sensitivity analysis, the exercise test traces were reviewed by a single blinded physician (A.J.S.C.), and the differences between the 2 treatment groups were 0.38 ml/kg/min for peak Vo2(p = 0.55) and 44 ml/min for absolute peak Vo2(p = 0.32). Our findings underpin some of the difficulties inherent in conducting a multicenter clinical study on cardiovascular exercise physiology.
These observations may in part explain the apparent discrepancy between our data and those reported by Mancini et al. (13), who conducted a preliminary exercise study on the treatment of anemia with ESPs in symptomatic CHF patients. In their single-center study, the investigators reported a significant increase in mean ± SD peak Vo2from 11.0 ± 1.8 ml/kg/min to 12.7 ± 2.8 ml/kg/min in anemic CHF patients receiving recombinant human erythropoietin. The study was single-blinded and enrolled fewer patients who had more severe CHF (NYHA functional class III and IV) and lower mean baseline peak Vo2values. The change in hemoglobin concentrations after ESP treatment was greater in the study by Mancini et al. (mean ± SD 11.0 ± 0.6 g/dl at baseline and 14.3 ± 1.2 g/dl at the end of treatment), and the results from 3 patients who did not complete the study were not included in the analysis.
In the double-blind, placebo-controlled, MIRACLE (Multicenter InSync Randomized Clinical Evaluation) study, cardiac resynchronization resulted in a median increase in peak Vo2that was 0.9 ml/kg/min greater than in the placebo group (p = 0.009), and in a median increase in exercise duration that was 62 s greater than placebo (p = 0.001) (26). In comparison, we observed a 0.5-ml/kg/min difference in peak Vo2between the placebo and darbepoetin alfa groups (p = 0.40) and an increase in exercise duration from baseline with darbepoetin alfa treatment that was 108 s greater than with placebo (p = 0.075). As such, the improvements in exercise duration, albeit not statistically significant, observed in our study were of similar or greater magnitude than those reported in the MIRACLE study, which recruited 369 patients to show a significant change.
In the present study, more patients in the darbepoetin alfa than in the placebo group experienced improvement in health-related quality of life as assessed by PGA (79% vs. 41%, p = 0.01), a simple measure that has been used previously in cardiovascular trials (13,17). This result confirms data presented by Mancini et al. (13). Their study also described a clinically significant improvement in MLHFQ score from baseline to end of therapy in patients treated with an ESP (compared with a deterioration in MLHFQ score in the placebo group). No formal comparison of the 2 treatment groups was made. We observed clinically significant changes from baseline to week 27 for all KCCQ domain scores and the total MLHFQ score in darbepoetin alfa-treated patients, but there were no significant differences in score changes between treatment groups. This may be attributed, at least in part, to the large placebo effect observed with both health-related quality-of-life instruments. The MLHFQ has been used extensively and has been validated in the CHF population (19,27), but has not shown consistently a clear discrimination between different severities of CHF or between active treatment and placebo (28). The KCCQ is a recently developed and validated CHF-specific instrument that has been shown to be reliable and more sensitive to clinical change than either the MLHFQ or the Short-Form 36 (15). Although the parallel administration of the MLHFQ and the KCCQ in this study could have provided the basis for a sensitivity comparison of the 2 instruments, no conclusions could be drawn because of the notable effects recorded in both placebo groups. Confounding of quality-of-life data by definite improvements in patients receiving placebo has been described in other interventional CHF trials (29).
The study is limited by its relatively small sample size. The multicenter design is likely to have increased the data variability compared with a single-center study. Some patients were recruited with relatively higher than required hemoglobin concentrations, which possibly impacted potential treatment benefits. We speculate that the supportive therapeutic environment created by the frequently scheduled visits with health care providers during the course of the study may have significantly contributed to the health-related quality-of-life improvements observed in patients receiving placebo.
Treatment with darbepoetin alfa in patients with symptomatic CHF and anemia significantly increased hemoglobin concentrations and was well tolerated. A trend toward improved exercise duration was observed, and the increase in exercise time correlated with the increase in hemoglobin. Darbepoetin alfa did not result in significant changes in peak Vo2, absolute peak Vo2, or NYHA functional class. Health-related quality of life as assessed by PGA was improved. The adverse event profile was similar in the darbepoetin alfa and placebo treatment groups. These results support the conduct of larger trials investigating the safety and potential benefit of treatment of anemia with darbepoetin alfa in anemic patients with symptomatic CHF.
Amgen, Inc., participated in discussions regarding study design and protocol development and provided logistical support during the trial. Monitoring of the study was performed by a contract research organization, under contract with Amgen, Inc., who maintained the trial database. The authors had full access to the database. The authors thank Beate D. Quednau, PhD, for expert editorial assistance.
↵1 Dr. Anker is supported by a Vandervell Fellowship (London, England) and by a grant from the Charité Medical School, Berlin, Germany.
↵2 Dr. Okonko is supported by a fellowship from the British Heart Foundation, London, England. The Department of Clinical Cardiology, Imperial College London, is supported by the British Heart Foundation, London, England.
↵3 Drs. Ponikowski, Anker, Coats, and McDonald have received honoraria for consulting from Amgen, Inc. and from Hoffmann-La Roche, Ltd.
↵4 Drs. Anker and Coats have received research grants from Amgen, Inc.
↵5 Drs. Baker and Rosser are employees and shareholders of Amgen, Inc.
↵6 Drs. Ponikowski and Anker contributed equally to this work.
This study was funded by Amgen, Inc.
- Abbreviations and Acronyms
- brain natriuretic peptide concentration
- chronic heart failure
- confidence interval
- erythropoiesis-stimulating protein
- Kansas City Cardiomyopathy Questionnaire
- left ventricular ejection fraction
- Minnesota Living With Heart Failure Questionnaire
- New York Heart Association
- Patient’s Global Assessment of Change
- oxygen uptake per kilogram body weight
- Received June 14, 2006.
- Revision received August 30, 2006.
- Accepted September 11, 2006.
- American College of Cardiology Foundation
- Horwich T.B.,
- Fonarow G.C.,
- Hamilton M.A.,
- MacLellan W.R.,
- Borenstein J.
- Anand I.,
- McMurray J.J.,
- Whitmore J.,
- et al.
- Sharma R.,
- Francis D.P.,
- Pitt B.,
- Poole-Wilson P.A.,
- Coats A.J.,
- Anker S.D.
- Adams K.F.,
- Patterson J.H.,
- Pina I.,
- et al.
- Mozaffarian D.,
- Nye R.,
- Levy W.C.
- Silverberg D.S.,
- Wexler D.,
- Blum M.,
- et al.
- Silverberg D.S.,
- Wexler D.,
- Sheps D.,
- et al.
- Mancini D.M.,
- Katz S.D.,
- Lang C.C.,
- LaManca J.,
- Hudaihed A.,
- Androne A.S.
- Green C.P.,
- Porter C.B.,
- Bresnahan D.R.,
- Spertus J.A.
- Rector T.S.K.S.,
- Cohn J.N.
- Packer M.,
- Fowler M.B.,
- Roecker E.B.,
- et al.
- Rector T.S.,
- Cohn J.N.
- Ryan E.,
- Devlin M.,
- Prendiville T.,
- Ledwidge M.,
- McDonald K.