Erythropoietin-Stimulating Agents: Ongoing Concerns With Safety

What Does This Important Study Show? 

Bennett and colleagues performed a rigorous systematic review of studies of ESAs in patients with malignancies with the intent of answering 2 questions: Is the use of ESA for treatment of anemia in patients with cancer associated with (1) increased mortality and (2) increased rates of venous thromboembolic (VTE) events? These questions were answered through review of phase III trials evaluating ESAs for treatment of anemia in cancer published between 1985 and 2005 and obtained through a Cochrane overview published in 2006,3 augmented by trials and additional literature including those published after the Cochrane review was completed. This process identified 8 studies which provided information not available to authors of the Cochrane review; these additional studies encompassed 4,062 patients and all prospectively evaluated survival.

In total, the review evaluating survival rates comprised 51 trials with 13,611 patients. Thirty-eight trials with 8,172 patients were evaluated for information on VTE rates. The anemia was deemed treatment related in 45 trials (11,522 patients) and cancer related in 6 trials (2,089 patients). Trials included in the analysis evaluated patients with both hematologic and solid tumors at various stages. The duration of ESA use varied between 6 weeks and 1 year. Erythropoietin was used in 41 trials enrolling 65% of all patients included in the analysis, with the remainder using darbepoetin. Treatment of the underlying malignancies included chemotherapy, radiotherapy, chemoradiotherapy, palliative radiotherapy, and no treatment. Duration of follow up ranged between 11 months and 5 years.

The meta-analyses found that patients receiving ESAs had higher risks of VTE (334 events in 4,610 patients treated with ESAs compared with 173 events in 3,562 control patients; relative risk, 1.57 [95% confidence interval, 1.31 to 1.87]) and mortality (1,421 deaths in 5,506 patients treated with ESAs compared with 1,211 deaths in 4,519 control patients; hazard ratio, 1.10 [95% CI, 1.01 to 1.20]). Contributing to this analysis, the authors found 8 trials enrolling 4,062 patients that individually demonstrated more rapid tumor progression in patients treated with ESAs. These studies enrolled patients with a diverse group of tumor types including breast cancer, non–small-cell lung cancer, head and neck cancer, lymphomas, and cervical cancer, suggesting that the effect of ESAs on tumor progression was not specific to one form of cancer.

How Does This Study Compare to Previous Data? 

This is the first systematic review to find a statistical increase in the risk of death in patients with malignancy-associated anemia who were receiving ESAs. However, the increased risk of venous thromboembolism with ESA use is well accepted.3, 5 The mechanism by which these agents cause VTE events is unknown and requires additional investigation.

When used to treat kidney disease–associated anemia, ESAs may be associated with adverse outcomes. Thus, increased thrombotic complications were found in the Normal Hematocrit Study (NHS) of hemodialysis patients with concurrent cardiac disease. In this study, ESAs were administered to 618 hemodialysis patients to achieve a target hematocrit of 42% compared with 615 patients who received ESAs to a target hematocrit of 30%. Vascular access thrombosis occurred in 243 patients in the high-hematocrit group, compared with 176 patients in the low-hematocrit group (P = 0.001). After 29 months, 183 deaths and 19 myocardial infarctions occurred in patients in the high-hematocrit group, compared with 150 deaths and 14 nonfatal myocardial infarctions in the normal-hematocrit group. The CHOIR (Correction of Hemoglobin and Outcomes in Renal Insufficiency) study of patients not receiving dialysis and with an estimated GFR from 15 to 50 mL/min found increased mortality and congestive heart failure when ESAs were administered to a target hemoglobin of 13.5 g/dL compared with 11.0 g/dL. This study did not report the number of VTE events although they were included in the composite outcome measure termed “thrombovascular events.”6, 7

There is evidence that ESAs cause thrombosis in other patient groups. Thus, a seminal phase III study in critically ill patients failed to demonstrate reduced transfusion need but did find a significant increase in thrombotic complications with ESA use,8 while use of ESAs in the setting of cardiopulmonary bypass and spinal surgery increased complications (see details in package insert).9 These observations of adverse outcomes led to package insert “black box warnings” that state as of July 30, 2008:

ESAs shortened overall survival and/or increased the risk of tumor progression or recurrence in some clinical studies in patients with breast, non–small-cell lung, head and neck, lymphoid, and cervical cancers.

To decrease these risks, as well as the risk of serious cardio- and thrombovascular events, use the lowest dose needed to avoid red blood cell transfusion.

Use ESAs only for treatment of anemia due to concomitant myelosuppressive chemotherapy.

ESAs are not indicated for patients receiving myelosuppressive therapy when the anticipated outcome is cure.

Discontinue following the completion of a chemotherapy course.

US Food and Drug Administration (FDA), Aranesp (darbepoetin alfa) package insert10

Further discussion of the FDA actions in relation to ESAs can be found on their webpage.9

What Should Clinicians and Researchers Do? 

Current ASCO/ASH (American Society of Clinical Oncology/American Society of Hematology) guidelines11 recommend the use of ESAs as a treatment option for patients with chemotherapy-associated anemia and a hemoglobin concentration that is approaching, or has fallen below, 10 g/dL, with a goal of increasing the hemoglobin and reducing the need for transfusions. However, based on the review published by Bennett and colleagues, clinicians should carefully weigh the risks of this therapy. Available evidence suggests the risk of thromboembolism and death in these already high-risk patients is enhanced by the use of ESAs; therefore, clinicians are implicitly weighing the potential for improved quality of life and reduced need for transfusion against the enhanced risk of venous thromboembolism and death. Although not addressed in the systematic review, particular groups of patients are likely to be at enhanced risk of VTE events; these include multiple myeloma patients treated with thalidomide or lenalidomide, patients with very high risk of VTE events (such as those with some subtypes of metastatic adenocarcinoma and those with primary brain tumors), those with inherited or acquired risk factors for venous thromboembolism (such as prior events), and others. Which subgroup of patients, if any, that might be at high risk of death with ESA use remains unknown.

Most data contributing to the conclusion that ESAs increase the risk of venous thromboembolism and death came from studies which used erythropoietin. Theoretically, darbepoetin might be associated with a different risk profile than erythropoietin; however, prior studies have concluded that these agents have similar biological effects and as such there is no evidence that darbepoetin might be associated with a different risk profile than erythropoietin.5

The biology of the observed increase in the risk of both venous thromboembolism and death is unknown. This is a prime area for further research. Perhaps it is time for oncologists to realize the lessons nephrologists have already learned. As the primary aim of any health care giver remains primum non nocere (first, do no harm), it appears most prudent that practitioners adhere strictly to FDA-approved indications for ESAs. In particular, clinicians should avoid administering this drug to achieve hemoglobin values in excess of 120 g/dL, and they should not administer it to patients outside those recommended in the FDA-approved package insert.