The Effect of Erythropoietin on Exercise Capacity, Left Ventricular Remodeling, Pressure-Volume Relationships, and Quality of Life in Older Patients With Anemia and Heart Failure With Preserved Ejection Fraction
Rose S. Cohen MD,
From the Division of Cardiology, Department of Medicine, Columbia University Medical Center, New York, NY
Mathew S. Maurer, Clinical Cardiovascular Research Laboratory for the Elderly, 5141 Broadway, 3FW Room 035, Allen Pavilion of New York Presbyterian Hospital, Columbia University Medical Center, New York, NY 10032 E-mail: firstname.lastname@example.org
Anemia is a significant comorbidity among patients with heart failure both in the setting of a reduced ejection fraction and with preserved ejection fraction (HFPEF).1–3 In persons with heart failure, there is a relationship among anemia, clinical symptoms, left ventricular (LV) structure, hemodynamics, morbidity, and renal function.4–12 Previous studies in patients with systolic heart failure have demonstrated a significant effect of increasing hemoglobin with either erythropoietin13–18 or iron17,19,20 on functional capacity, ventricular structure, and quality of life, although a recent randomized trial failed to show clinical benefit in such patients.21 Such studies have not specifically focused on the population with HFPEF, who account for more than half of persons with chronic heart failure.
Chronic anemia is known to result in compensatory LV hypertrophy, higher myocardial chamber volumes, and a high–cardiac output state. These structural and hemodynamic changes could be detrimental in the population with HFPEF, as the clinical syndrome of heart failure may be exacerbated. Thus, correction of anemia may provide significant benefits in terms of preload reduction, less effort intolerance, and less dyspnea, leading to improved quality of life. Accordingly, the primary hypothesis of this prospective open-label study is that the administration of subcutaneous erythropoietin to a cohort of anemic patients with HFPEF would be associated with significant changes in LV structure (reduced end-diastolic volume [EDV], regressed LV mass) and function (LV chamber contractility, ventricular-vascular coupling, stroke volume, and cardiac output) as well as improvements in exercise capacity and quality of life.
Study Design and Participants
This was a prospective, open-label, 12-week cohort study among community-dwelling, independently living, older patients with anemia and HFPEF. Participants were recruited from internal medicine clinics as well as specialty cardiology and renal clinics at an urban medical center setting (New York Presbyterian Hospital, New York City, NY) The diagnosis of heart failure was based on the National Health and Nutrition Examination Survey congestive heart failure criteria with a score ≥322 and were considered to have a preserved ejection fraction if 3-dimensional echocardiographically determined ejection fraction was >45%. Anemia was defined as hemoglobin <12 g/dL.23 Informed consent was obtained from all participants. The Columbia University Medical Center institutional review board approved the study.
Patients were excluded from the study if they had uncontrolled hypertension (systolic blood pressure >160 mm Hg and/or diastolic blood pressure >90 mm Hg); resting heart rate >120 beats per minute; baseline 6-minute walk >450 m; valvular heart disease greater than mild by transthoracic echocardiography; infiltrative cardiac disease, such as hemochromatosis and amyloidosis; hypertrophic cardiomyopathy; chronic pulmonary disease (forced expiratory volume in 1 second <60% of predicted); renal failure (glomerular filtration rate <15 mL/min); hemoglobin level <9 g/dL; exercise limited by angina; claudication or neurologic diseases; severe liver dysfunction; cardiac surgery <3 months prior; known iron deficiency anemia from chronic gastrointestinal blood loss, uterine bleeding, or other chronic bleeding; significant alcohol or illicit drug use; known hypercoagulable state; or active hematologic disease. Patients were also excluded if they had a history of deep venous thrombosis or pulmonary embolus within 12 months before study entry; had a history of cerebral vascular accident or transient ischemic attack within 6 months or an acute coronary syndrome within 6 months of study entry; had an allergy or sensitivity to human serum albumin; or had a known hypersensitivity to mammalian cell-derived products.
Study Drug Administration and Dosing
Epoetin alpha was administered weekly by subcutaneous injection using a prespecified dosing algorithm (See Supplementary Appendix). The dosing algorithm was designed to make adjustments based on the rate of rise of the hemoglobin over a 1-week period, as well as the absolute hemoglobin value. Participants initially received active treatment with 10,000 units of erythropoietin given weekly by subcutaneous injection. They were carefully monitored (eg, every week) when beginning therapy to avoid rapid increases in hemoglobin/hematocrit and/or increasing blood pressure. No dose adjustments were made for the first 3 doses of erythropoietin (10,000 units/wk) unless the hemoglobin rose too rapidly (>0.3 g/dL) in any given weekly interval.
Blood Volume Analysis
Blood volume was determined after intravenous administration of iodine131–labeled albumin (Volumex; Daxor Corp, New York, NY) as previously described.24,25 Plasma volume was determined as the zero-time volume of distribution of the radiolabeled albumin obtained by semilogarithmic extrapolation of values measured from at least 3 samples drawn 12 minutes after injection at 6-minute intervals. Spun hematocrit was determined from each sample, and plasma radioactivity of each sample was measured in a semi-automated counter (BVA-100 Blood Volume Analyzer; Daxor Corp, New York, NY). Blood volume and red blood cell volumes were calculated from the plasma volume measurement and then compared with normal values for age, sex, height, and weight based on the patient’s ideal weight.26
Two- and Three-Dimensional Echocardiography
Standard 2-dimensional transthoracic echocardiography was performed in each participant. End-diastolic measurements of LV internal dimension, septal thickness, and posterior wall thickness were acquired according to the standards of the American Society of Echocardiography.27 Doppler indices of the mitral inflow pattern including peak E wave velocity, peak A wave velocity, and isovolumetric relaxation time as well as lateral mitral annual velocities were recorded for 3 beats and averaged. LV filling pressures were estimated by the formula .28
The equipment and procedures of freehand 3-dimensional transthoracic echocardiography (3-DE) have been previously described in detail.29,30 3-DE was performed using conventional real-time echocardiography, 3-dimensional acoustic spatial locater, personal computer, and custom software. The data derived include LV chamber (EDV, myocardial volume, stroke volume (SV), and ejection fraction (EF=SV/EDV). Myocardial volume was multiplied by 1.05 g/dL to determine ventricular mass. Echocardiography was performed by study personnel blinded to clinical information.
Both a 6-minute walk test31,32 and cardiopulmonary exercise test were performed at baseline and after 3 months of study drug treatment by study personnel blinded to clinical information. The total distance walked after 6 minutes to the nearest meter was recorded. Patients performed an upright bicycle exercise test where the workload was increased every 3 minutes by 25 W according to a standard protocol. After 3 minutes of data at rest, exercise began at a workload of 0 W and increased every 3 minutes by 25 W until symptom-limited peak exercise was reached. Expired gas analysis was performed continuously throughout the test with the Innocor system (Innovision A/S, Odense, Denmark).33 Peak oxygen consumption (VO2) was defined as the highest value of VO2 achieved in the final 30 seconds of exercise.
Quality of Life
The Kansas City Cardiomyopathy Questionnaire34—a valid, reliable, self-administered, 23-item questionnaire that quantifies physical limitations, symptoms, self-efficacy, social interference, and quality of life—was employed. The Kansas City Cardiomyopathy Questionnaire summary score and subscores were calculated (range, 0–100; higher scores indicate better health status) at baseline and after 3 months of follow-up.
Echocardiographic Estimates of Ventricular Chamber Properties
The end-systolic pressure-volume relationship (ESPVR) and end-diastolic pressure-volume relationship (EDPVR) were estimated using validated single-beat techniques. The slope, defining end-systolic elastance (Ees), and a volume axis intercept (Vo) of the ESPVR were estimated noninvasively by the single-beat method [Ees(sb)] described by Chen and associates.35 To account for covariance in Ees and Vo, both of which determine the position of the ESPVR, the values of these parameters derived from each participant were used to predict the V120 (the volume to achieve an end-systolic pressure of 120 mm Hg), which is calculated from the Ees and Vo of each patient: V120 = Vo + 120 / Ees. Effective arterial elastance (Ea), a lumped index of vascular hemodynamic load primarily related to total peripheral resistance and heart rate, was estimated by Ea ≈ Pes / SV,36 where Pes is the LV end-systolic pressure estimated by 0.9 × systolic blood pressure.37
To characterize the EDPVR (where EDP = αEDVß; α is a scaling constant and ß is a diastolic stiffness constant), a recently developed and validated single-beat approach was used.38 This approach relies on the empiric observation that volume-normalized EDPVRs share a common shape, thereby allowing estimation of α and ß to define the entire EDPVR from a single measured pressure-volume point. Measured end-diastolic pressure and EDV were used to derive α and ß in each participant. To account for covariance in α and ß,39 both of which impact on the shape and position of the EDPVR, the values of these parameters derived from each participant were used to predict the EDV at a common end-diastolic pressure of 30 mm Hg to yield a pressure-independent index of heart size or ventricular capacitance (EDV30).
The area between the EDPVR and the ESPVR measured as a function of end-diastolic pressure was used to index overall pump function.40,41 This specific area is called the isovolumic pressure-volume area (PVAiso), is independent of afterload, and can be calculated analytically as a function of LV following curve fitting of the EDPVR and the ESPVR: where Pes (V) and Ped (V) are the end-systolic and end-diastolic pressures, respectively, as a function of volume.
Results are expressed as mean ± standard error. Changes in principle measures were compared from baseline to 3-month values by Student t test for paired analyses. Associations between changes in hemoglobin and red cell volume measures and outcome variables (LV volumes, functional parameters, and quality-of-life measures) were determined by use of Pearsons’ correlation coefficient. The primary end point of the study was change in LV EDV after 3 months of study. Preliminary data indicate that the mean LV EDV in patients with HFPEF is 130±34 mL as assessed by freehand 3-dimensional echocardiography. With a total of 10 participants, we had a 90% power to detect a 13-mL (or 10%) difference after 3 months of therapy at an α of .05. SAS for Windows (version 8.0, SAS Institute Inc, Cary, NC) was used for all analyses.
Patients enrolled in this study were older (average age, 68±3 years) and predominantly female (92%), with several comorbidities in addition to HFPEF and anemia (Table I). All participants had true anemia (defined by the assay as red cell volume <95% predicted42) with an average red cell deficit of 503±57 mL (27%±8% deficit). Therapy with erythropoietin was associated with an increase in hemoglobin from 10.8±0.3 to 12.2±0.3 g/dL during the course of the study. The rise in hemoglobin was slow and steady (Figure 1), concordant with the dosing algorithm goals and study target hemoglobin of 12.5–13.5 g/dL. The average weekly dose of erythropoietin was 3926 units. Blood volume measurements performed by the I131 tagged albumin method confirmed the increase in red cell volume with erythropoietin treatment (from 1187±55 to 1333±48 mL). While the final hemoglobin value was below the target hemoglobin range for the protocol of 12.5 to 13.5 g/dL, the target hemoglobin value was reached in 6 of 11 individual patients (55%).
Table I. Demographic and Clinical Characteristics
No. of participants
68±3 (range, 52–86)
33% Black (nonwhite)
58% Hispanic (nonwhite)
Comorbid conditions, No (%)
Coronary artery disease
Chronic renal insufficiency
Blood urea nitrogen, mg/dL
Estimated glomerular filtration rate, mL/min
No significant changes were noted in systolic blood pressure, and a significant decrease was noted in diastolic blood pressure (from 70±3 to 64±2 mm Hg; P<.05) during the study period. This was associated with a modest but not significant change in the number of participants receiving diuretics and calcium channel blockers but no change in the dose of angiotensin-converting enzyme inhibitors and β-blockers. Plasma volume did not differ significantly from baseline to study termination, but it trended toward a decrease in volume (from 3066±169 to 2942±182 mL; P=not significant).
In this cohort of patients, LV EDV (as measured by 3-D echocardiography) decreased with erythropoietin treatment, approximately 8 cc (8%) during the 3-month period (P=.03), while LV mass remained unchanged (Table III). Pressure-volume relationship of overall cohort pre– and post–erythropoietin treatment is shown in Figure 2 along with the PVAiso to end-diastolic pressure relationship. There was a shift in the EDPVR toward smaller volumes, without a significant change in the ESPVR. Overall stroke work, stroke work per grams of ventricular mass, and the PVAiso area trended toward lower values after 3 months of treatment with erythropoietin.
Table III. Echocardiographic and Physiologic Variables
Abbreviations: Pes, pressure at end systole; V30, ventricular capacitance at an end-diastolic pressure of 30 mm Hg; V120, volume to achieve an end-systolic pressure of 120 mm Hg. aP<.05. bP=.054. Values are mean ± SE.
2-Dimensional echocardiographic parameters
Mitral valve E velocity
Mitral valve A velocity
E’ (tissue Doppler)
3-Dimensional echocardiographic parameters
Stroke volume, mL
End-diastolic volume, mL
Ejection fraction, %
Left ventricular mass, g
Estimated left-ventricular end-diastolic pressure, mm Hg
Single-beat end-systolic elastance, mm Hg/mL
Single-beat volume axis intercept, mL
Pes/end-systolic volume, mm Hg/mL
Stroke work, mL mm Hg
Stroke work/mass ratio, mL mm Hg/g
PVAIso 20 mm Hg, mL* mm Hg
All measures of functional capacity improved significantly (P<.05) with erythropoietin, including peak VO2 (15%), exercise time (32%), and 6-minute walk (15%) (Table II). Finally, 8 of 9 scales or subscales on the Kansas City Cardiomyopathy Questionnaire (Table IV) improved. However, there were no significant associations between changes in hemoglobin or red cell volume and any functional, structural, or quality-of-life measures.
Table II. Preliminary Clinical Findings
Abbreviations: N/A, not available; RER, respiratory exchange ratio. aOriginal n=12, one patient discontinued for unrelated medical reasons. bP<.05. Values are mean±SE.
Blood pressure parameters, mm Hg
Systolic blood pressure
Diastolic blood pressure
Mean arterial pressure
Medication use, No. of patients (%)
Diuretic (any type)
Angiotensin-converting enzyme inhibitor
Calcium channel blocker
Blood volume parameters
Total blood volume, mL
Plasma volume, mL
Red cell volume, mL
6-Minute walk test, m
Bicycle ergometer exercise time, s
Peak exercise heart rate, beats/min
Table IV. Quality of Life
aP<.01. Values are mean ± SE.
Quality of life
This is the first study to describe the effect of erythropoietin treatment in elderly patients with HFPEF and anemia. We explored many potential effects of short-term therapy including ventricular remodeling, cardiac output, pressure-volume relationships, and ventricular-vascular coupling, as well as other clinical parameters of functional capacity and quality of life. These data demonstrate that correction of mild to moderate anemia with short-term erythropoietin therapy is associated with improvements in functional parameters and reductions in LV volumes and ventricular capacitance without significant effects on systolic properties.
Safety and Tolerability
Widespread enthusiasm for the benefits of erythropoietin therapy was dampened by several studies in patients with chronic kidney disease, trauma, and cancer who experienced significant adverse effects from erythropoietin therapy including thrombotic events43 and cardiovascular events in patients targeted to receive a higher hemoglobin.44,45 Meta-analysis has reaffirmed concerns about an increased risk of death and poor blood pressure control46 in patients targeted for a higher hemoglobin, resulting in the issuance of black box warning by the US Food and Drug Administration (FDA).47 The FDA recommended careful monitoring of hemoglobin values after dose adjustments, careful monitoring and control of blood pressure, and limiting the rate of rise of hemoglobin to <1 g/dL in any 2-week period, all of which were taken into account in the design and execution of this pilot/feasibility trial. The dosing algorithm employed called for weekly monitoring of hemoglobin levels, with dose adjustments that were based on the current hemoglobin level as well as rate of rise in the preceding week. This algorithm affected a rise in hemoglobin that was slow and steady and did not result in any significant adverse events (eg, thrombotic episodes or decompensated heart failure or related hospitalization) or any significant increase in blood pressure. Additionally, the dose of erythropoietin employed was lower than anticipated and differed significantly from current treatment guidelines.48 However, such an approach was associated with significant patient burden and alterations in blood pressure medications in a large percentage of participants (33%). Whether such an approach will be safe and executable in a larger number of participants is unknown but is currently being studied.49
Ventricular Structure and Function
The vast majority of patients with HFPEF have concomitant hypertension, and a significant percentage have chronic renal dysfunction and LV hypertrophy, all risk factors for the development of HFPEF in large population-based studies.50 Anemia contributes to LV hypertrophy, which is an important underlying substrate among patients with HFPEF. Accordingly, improving anemia in patients with both chronic renal failure and heart failure in meta-analyses has been shown to reduce LV mass and LV volumes. In these studies, average baseline LV mass was significantly higher (289 g) than in the current cohort, and LV EDV was greater (148 mL). Meta-analysis demonstrated that erythropoietin resulted in significant declines of 15% in LV mass index and 16% in EDV.51 The discrepancies between these results and ours could be attributable to several factors, including the limited sample size in our study, less severe ventricular remodeling in our participants, and the short duration (3 months compared with >6 months in most trials) of therapy and limited increase in hemoglobin achieved with therapy in the current study. The use of noninvasive pressure-volume analysis, demonstrates that there was a trend toward a reduction in ventricular capacitance concordant with ventricular remodeling.
Anemia results in reductions in systemic vascular resistance, increases in sympathetic nervous system activation, and expansion of plasma volume, which can augment preload volume, contributing to the effort intolerance and dyspnea experienced by persons with HFPEF. Accordingly, correction of anemia would ameliorate these hemodynamic changes and result in declines in EDV, SV, and cardiac output. Accordingly, treatment of anemia through an amelioration of altered loading conditions would be anticipated to reduce ventricular work, as shown in the reduction in the pressure-volume area (Figure 1, bottom panel).
Functional Capacity and Quality of Life
Effects of erythropoietin on quality of life have been limited predominantly to the cancer population, with a Cochrane Review suggesting that therapy may improve quality of life and functional class52 in this population. Also, in pre-dialysis patients and dialysis patients, erythropoietin corrects anemia and also improves quality of life and submaximal exercise performance.53–55 However, for patients with heart failure, limited data are available. Early unblinded studies and phase II results using erythropoietin in patients with systolic heart failure have found overall significant improvements in exercise capacity and quality of life.15,56 However, more recent randomized trials did not demonstrate a significant benefit on exercise duration, New York Heart Association class, or quality-of-life score compared with placebo.21 Since a majority of heart failure cases involve patients with a preserved ejection fraction who have a similar decrement in quality of life as those with systolic heart failure,57 characterizing the role of erythropoietin in this population is warranted. While the data from this open-label trial are encouraging, with almost all scales and subscales on the Kansas City Cardiomyopathy Questionnaire demonstrating statistically significant increases in scores (concordant with improved quality of life), compatible with a moderate to large clinical benefit,58 given the unblinded and open-label nature of this study, causality cannot be concluded.
This study is limited by its open-label design, its limited sample size, and the limited duration of treatment. While the sample size is limited, our study demonstrated the feasibility of performing clinical trials in a patient population previously unstudied and not typically included in clinical trials. Such preliminary data are difficult to obtain for reasons of patient accessibility to study facilities, study participant burden, and patient comorbidities. Our findings can be considered a prerequisite in order to determine relative safety and feasibility of the approach prior to the undertaking of larger randomized phase II and III clinical trials, which are currently under way.49 Notably, in order to safely raise hemoglobin values with weekly injections, the study participants were at their target hemoglobin level only at the very end of the study period (generally week 11 and 12) and they remained anemic for the first 2 months of the study, which may have blunted the true effect of erythropoietin (or raised hemoglobin values) on ventricular remodeling. While we found that the increase in hemoglobin was accompanied by significant increases in functional capacity, improvements in quality of life, and reductions in ventricular volumes, these associations cannot be directly attributed to the intervention given the open-label study design and absence of statistical correlation and may have occurred by chance or be attributable to some other confounding factor (eg, medication adjustment).
Disclosures: This research was supported by the NIH/NIA (R01AG027518-01A1). Dr Rose Cohen was supported by the ACCF/Merck Fellowship 2005–2006 and a grant from the New York Academy of Medicine Glorney Raisbeck Fellowship 2006.
In a population previously unstudied (older adults with HFPEF and concomitant anemia), erythropoietin was well-tolerated and effected a significant increase in hemoglobin and red cell volume without significant increases in blood pressure or other adverse effects during the 3-month study period. Additionally, there were improvements in exercise capacity, improvements in quality of life, and reduced ventricular capacitance during the course of the trial. These data suggest that ongoing evaluation of erythropoietin therapy in patients with HFPEF and anemia is warranted.