Out of this nettle, danger, we must pluck this flower, safety

In this issue of Blood, Greenberg and colleagues report the final results of the ECOG E1996 study, a randomized prospective clinical trial designed to assess both efficacy and safety of recombinant erythropoietin in patients with MDS.¹ 

Erythropoiesis stimulating agents (ESAs) epoetin and darbepoetin are by far the most commonly prescribed drugs for patients with myelodysplastic syndrome (MDS).²  In part, this is because ESAs are recommended for treatment of anemia in lower-risk MDS patients by widely used clinical practice guidelines, such as those of the National Comprehensive Cancer Network (NCCN; www.nccn.org). Yet ESAs are neither approved for this indication by the US Food and Drug Administration (FDA), nor is long-term ESA safety well established in patients with MDS.³ 

ESA safety has received considerable attention during the past 4 years, as a number of concerns have arisen with respect to the risks of ESA exposure in patients with various neoplasms—concerns that emerged only after tens of thousands of patients with solid tumors were treated with ESAs.⁴  Worrisome safety data in patients with cancer-associated anemia have prompted the FDA to revise ESA labels multiple times, including addition of a series of sobering “black box” warnings to prescribing information. The FDA has also mandated that ESA sponsors must rigorously examine ESA risks in various patient populations.

The lack of a specific ESA label for MDS has had practical consequences for patients. In 2007, the Centers for Medicare and Medicaid Services (CMS) proposed a National Coverage Decision (NCD) that would have ended reimbursement for ESAs when used to treat MDS-associated anemia. CMS only backed down from this stance and revised the restrictive NCD to exclude MDS after a broad-based campaign advocating preservation of ESA access for MDS patients, supported by the American Society of Hematology (ASH), the American Society of Clinical Oncology, the MDS Foundation, the Aplastic Anemia & MDS Foundation, and other organizations. CMS ultimately left ESA coverage decisions for MDS up to regional carriers, which unfortunately means that ESA reimbursement policies now vary from region to region—the sort of “postcode medicine” more commonly associated with nationalized health services in the United Kingdom and elsewhere than with health care delivery in the United States.

Various investigators have published results from open-label studies of ESAs in MDS, used either as single agents or in combination with colony stimulating factors (CSFs, for which there is in vitro evidence of synergistic effects on erythropoiesis), with major hemoglobin response rates ranging from 10% to 40%.^5,6  But there are almost no prospective, controlled, randomized trials.

In this issue of Blood, Greenberg and colleagues help fill that data gap by presenting the final results of the Eastern Cooperative Oncology Group (ECOG) E1996 prospective clinical trial of ESAs for patients with MDS, which was first reported in abstract form at the ASH annual meeting in 2004 and is now published with a lengthy median follow-up time of 5.8 years. The investigators enrolled 110 patients with lower-risk MDS, who were randomized to receive either epoetin alfa, with or without filgrastim, or supportive care. The erythroid response rate was 34% for the active treatment arm using 2006 International Working Group criteria, compared with 5.8% for the supportive care arm.

ESA responders lived longer than nonresponders in the E1996 study, presumably because ESA responses mark patients who have more hematopoietic reserve and less severe disease. Although overall survival and leukemic transformation rates were equivalent between E1996 randomization groups, crossover of the supportive care arm to active treatment after 4 months limits our ability to draw conclusions about longer-term safety of ESAs in MDS, and the study was also powered so that only quite dramatic survival differences could have been detected (ie, 80% power to detect a 46% difference in hazard rate).

Several recent retrospective studies suggested that ESA treatment may be associated with a survival benefit in MDS, but these encouraging retrospective analyses are potentially confounded by patient selection bias.^7,8  It is clear that larger prospective studies with primary safety endpoints, including both overall survival and progression-free survival, are needed. However, as for Hotspur's metaphorical dangerous nettle in Henry IV, Part I, actually conducting such studies to pluck the flower of safety is a prickly business, because of logistical hurdles, patient preferences, and entrenched practice patterns. An ambitious industry-sponsored controlled study of epoetin alfa in MDS, EPO-ANE-3018, began enrolling patients a few months ago (the accrual goal is 450 patients), but results will not be available for at least a few years. Therefore, E1996 represents an important contribution—not because it definitively answers all important ESA safety questions, but because it suggests that ESA exposure is relatively safe in MDS at least in the very short term—and because for several years to come, the E1996 results will remain virtually the only controlled data available in this area.