|Home | About | Journals | Submit | Contact Us | Français|
The erythropoiesis-stimulating agents (ESAs) erythropoietin and darbepoetin prevent transfusions among chemotherapy-associated anemia patients. Clinical trials, meta-analyses, and guidelines identify mortality, tumor progression, and venous thromboembolism (VTE) risks with ESA administration in this setting. Product labels advise against administering ESAs with potentially curative chemotherapy (United States) or to conduct risk–benefit assessments (Europe/Canada). Since 2007, fewer chemotherapy-associated anemia patients in the United States and Europe receive ESAs. ESAs and the erythropoietin receptor agonist peginesatide prevent transfusions among chronic kidney disease (CKD) patients; clinical trials, guidelines, and meta-analyses demonstrate myocardial infarction, stroke, VTE, or mortality risks with ESAs targeting high hemoglobin levels. U.S. labels recommend administering ESAs or peginesatide at doses sufficient to prevent transfusions among dialysis CKD patients. For dialysis CKD patients, Canadian and European labels recommend targeting hemoglobin levels of 10 to 12 g/dL and 11 to 12 g/dL, respectively, with ESAs. ESA utilization for dialysis CKD patients has decreased in the United States.
Erythropoiesis-stimulating agents (ESAs) treat chemotherapy- or chronic kidney disease (CKD)-associated anemia. Beginning in 1989 in the CKD setting, regulatory agencies approved erythropoietin and darbepoetin (in Canada) and erythropoietin, darbepoetin, and peginesatide (United States) to prevent transfusions, and erythropoietin and darbepoetin to treat symptomatic anemia (Europe). In 1993 and 2002, erythropoietin and darbepoetin, respectively, received regulatory approvals for prevention of transfusions or treatment of symptomatic anemia among persons with cancer and chemotherapy-associated anemia. For CKD patients receiving hemodialysis, although partial anemia correction is beneficial, one trial reported mortality/myocardial event risks with epoetin administration targeted to complete anemia correction.1 For CKD patients with diabetes not receiving hemodialysis, one trial identified cerebrovascular accident risks with darbepoetin targeting high hemoglobin levels versus rescue therapy for moderately severe anemia.2 For CKD patients not receiving hemodialysis, two trials identified myocardial infarction risks with peginesatide versus darbepoetin.3 In oncology, two trials identified poorer survival with epoetin.4,5 These trials necessitate reassessment of ESAs (Table 1).
Data sources included reports from regulatory agencies, product manufacturers, and meta-analyses identified by Ovid and Cochrane searches (search terms: ESAs, meta-analysis, guidelines [2007 to May 2012]). Clinical trials and database studies cited in regulatory documents were reviewed.
Among 75 abstracts identifying guidelines or meta-analyses, 18 were included. Studies without primary data were excluded. Sixteen regulatory documents, 16 phase III clinical trials, 13 regulatory agency advisory committee materials, and 1 database study cited in regulatory documents were reviewed (Fig. 1).
Meta-analyses and guidelines identified decreased transfusions when cancer patients with chemotherapy-associated anemia received ESAs.6-9 Quality of life benefit was reported in the Canadian review but not in other overviews.10 Seven phase III trials identified mortality and/or tumor progression risks when chemotherapy- or radiation therapy-treated cancer patients received ESAs; one trial identified mortality risks when cancer patients not receiving chemotherapy received ESA darbepoetin.6 One meta-analysis identified a nonsignificant 1.03-fold higher relative survival benefit when cancer patients received ESAs (p > 0.05).7 Three subsequent meta-analyses identified statistically significant 1.10-fold, 1.15-fold, and 1.17-fold mortality risks, and three meta-analyses identified statistically significant 1.57-fold, 1.67-fold, and 1.69-fold VTE risks when cancer patients received ESAs (p < 0.05 for each study).6,8-10
In 2007, FDA statisticians concluded that quality of life claims for ESA-treated cancer patients were invalid.11 FDA’s Oncology Drug Advisory Committee (ODAC) warned against administering ESAs to breast, head, and neck, and non–small cell lung cancer patients because of tumor promotion concerns.12 American Society of Clinical Oncology/American Society of Hematology (ASCO/ASH) guidelines recommended administering ESAs to patients with chemotherapy-associated anemia and hemoglobin levels < 10 g/dL, targeting 12 g/dL hemoglobin levels, highlighting mortality concerns at higher hemoglobin levels.13 Black Box warnings described VTE, cardiovascular, and mortality risks with ESA treatment of chemotherapy-associated anemia with target hemoglobin levels > 12 g/dL.11 Revised labels identified cancer progression/mortality risks among lymphoid and non–small cell lung cancer patients receiving ESAs targeted to hemoglobin levels more than 12 g/dL.
In 2008, ODAC warned that ESAs should not be administered to cancer patients receiving potentially curative chemotherapy.14,15 Medication guides warned cancer patients that ESAs can result in death or other serious toxicities; heart attack, stroke, heart failure, and early death may occur with ESA administration to hemoglobin levels > 12 g/dL; blood clots may occur at any hemoglobin level or ESA dose; ESAs do not improve quality of life; tumors may grow faster and death occur sooner with ESAs targeting high hemoglobin levels.16 National Comprehensive Cancer Network (NCCN) guidelines advised that ESAs were indicated to prevent transfusions among cancer patients receiving palliative, but not curatively intended, chemotherapy.17
In 2010, ESA manufacturers implemented a Risk Evaluation Mitigation System.18 ASCO/ASH guidelines advised that chemotherapy-associated anemia patients with hemoglobin levels < 10 g/dL should be informed of ESA risks and benefits; no target hemoglobin level was recommended.19
In 2007, European Public Assessment Reports (EPARs) warned against achieving hemoglobin levels > 12 g/dL because of toxicity concerns. The European Medicines Agency (EMA) advised that for cancer patients with long life expectancies, risks may outweigh benefits because ESAs may stimulate tumor growth.20 EMA’s Committee for Medicinal Products for Human Use (CHMP) concluded that ESA effects on survival of chemotherapy-treated cancer patients were indeterminant; ESA benefits outweighed risks for chemotherapy-treated cancer patients; symptomatic cancer patients with chemotherapy-associated anemia should receive ESAs targeting hemoglobin levels between 10 and 12 g/dL.20 EMA’s Scientific Advisory Group on Oncology recommended individualized decisions.21
In 2008, after EMAs’ CHMP reviewed a meta-analysis and one clinical trial identifying mortality and VTE risks with ESA administration to cancer patients, they warned of tumor progression, venous thromboembolism, and mortality risks with ESAs; benefits continued to outweigh risks among chemotherapy-associated anemia patients, except for patients with long life expectancy.22 One month later, after reviewing two additional trial reports, CHMP reported tumor progression, progression-free survival, and mortality risks among cancer patients receiving chemotherapy and ESAs.22 A guidance from the National Institute of Health and Clinical Excellence highlighted uncertain effects of ESAs on survival, based on trials reported prior to 2005, and cost-effectiveness estimates greater than generally accepted values.23
In 2009, CHMP reviewed the 2009 Cochrane meta-analysis and warned of risks of mortality with ESA administration to cancer patients receiving chemotherapy. This warning was added to product labels in 2010.
In 2007, Canada Health reported that ESAs were not indicated for cancer patients not receiving chemotherapy; tumor progression occurred among head and neck cancer patients receiving radiation therapy with target hemoglobin levels > 12 g/dL; increased mortality occurred among breast cancer patients receiving chemotherapy with target hemoglobin levels > 12 g/dL. Several trials identified venous thromboembolism, cancer progression, and mortality risks with ESAs.24
In 2009, the Canadian Agency for Drugs and Technologies in Health (CADTH) reported a meta-analysis of chemotherapy-associated anemia patients identifying quality of life and transfusion benefits with ESAs; mortality, VTE, and other toxicity risks; and cost-utility ratios exceeding generally accepted standards.25
Three trials reported prior to 2007 found no survival/cardiovascular benefits and increased toxicity with ESAs targeted to high versus low hemoglobin target levels.1,26,27 One trial included ESA-treated nondialysis CKD patients and identified a statistically significant 1.34-fold higher relative risks (17.5 vs. 13.5%) for death, acute myocardial infarction, congestive heart failure hospitalization, or stroke with ESA treatment targeted to higher target hemoglobin levels.26 Secondary analyses identified ESA nonresponse as a toxicity indicator.28 Another trial of nondialysis CKD patients demonstrated a statistically nonsignificant 1.58-fold increase in cardiovascular risks in high target hemoglobin groups.27 A 2009 trial evaluating nondialysis CKD patients with type 2 diabetes receiving darbepoetin targeted to near normal hemoglobin levels versus rescue therapy for moderately severe anemia found no reduction in death/cardiovascular events and death/end-stage renal disease but 92% higher relative risks of stroke (5.0 vs. 2.6%; p < 0.05).2 Among CKD patients, one meta-analysis found no mortality difference with ESAs targeting higher hemoglobin levels.29 Another found a statistically significant 1.17-fold increased mortality risk with higher target hemoglobin levels.30 A Medicare analysis of hemodialysis patients with diabetes found mortality risks with 30,000 versus 15,000 Units of epoetin/wk of 9 and 13% between 45,000 and 15,000 U/wk (p < 0.05).31
In 2011, draft guidelines from the Kidney Disease: Improving Global Outcomes initiative of the National Kidney Foundation recommended that decisions to initiate ESAs for CKD patients should consider potential benefits of reducing blood transfusions and anemia-related symptoms against risks of harm—including stroke, vascular access loss, and hypertension (grade 1B recommendation). Great caution is urged when administering ESAs to CKD patients with malignancies or stroke (grade 1C). For CKD patients not on dialysis, with hemoglobin levels ≥ 10 g/dL, ESAs are not recommended (grade 2D); if the hemoglobin level is < 10 g/dL, ESA initiation should be based on rate of hemoglobin fall, risk of needing transfusion, risks of ESA therapy, and presence of anemia-related symptoms (grade 2C). For CKD patients on hemodialysis, ESAs should be initiated with hemoglobin levels between 9 and 10 g/dL (grade 2B), although some patients may benefit with improved quality of life if ESAs are begun at higher hemoglobin levels (not graded). ESAs should not be used to maintain hemoglobin levels > 11.5 g/dL (grade 2C) or to intentionally increase hemoglobin levels more than 13 g/dL.
In 2007, the National Kidney Foundation-Kidney Disease Outcomes Quality Improvement (KDOQI) group recommended target hemoglobin levels of 11 to 12 g/dL based on quality of life and transfusion benefits but avoiding hemoglobin levels > 13 g/dL.32 FDA statisticians concluded that based on under-development guidelines, quality of life claims were invalid and recommended that most of these claims be removed from labels.11
In 2008, the FDA’s Cardiovascular and Renal Drugs Advisory and Drug Safety/Risk Management Advisory Committees concluded that CKD patients should not receive large ESA doses and should avoid achieving hemoglobin levels > 13 g/dL.33 Black Box warnings describe death/cardiovascular event risks with ESAs targeting hemoglobin levels > 12 g/dL. Clinicians were advised to administer the lowest ESA doses needed to avoid transfusions.14 Product labels removed quality of life claims with the exception of exercise tolerance and patient-reported physical function, recommended target hemoglobin levels of 10 to 12 g/dL, and warned against administering high ESA doses. Patients were advised that ESAs can lead to death/serious toxicities; heart attack, stroke, heart failure, and early death may occur if ESAs administration results in hemoglobin levels > 12 g/dL; blood clots may develop.34
In 2010, the FDA’s Cardiovascular and Renal Advisory Committee recommended no additional restrictions. In 2011, the FDA warned of cerebrovascular, cardiovascular, or mortality risks associated with ESA administration to non-dialysis CKD patients or when ESAs are administered to CKD patients targeting 11 g/dL or greater hemoglobin levels.11,35,36 Boxed warnings stated that four trials found greater death, serious cardiovascular events, and stroke risks with ESA administration targeting hemoglobin levels > 13 g/dL; dosing should be individualized to achieve and maintain hemoglobin levels between 10 and 12 g/dL.37
In 2011, public health advisories indicated that trials identified death, cardiovascular, and stroke risks with ESAs administered targeting hemoglobin levels > 11 g/dL and that no trial identified target hemoglobin levels, dose, or dosing strategy that did not increase risks.38 Revised labels recommended that for CKD patients on dialysis, ESAs should be initiated when hemoglobin levels are < 10 g/dL and to administer the lowest ESA dosages sufficient to reduce red blood cell transfusions.38 For CKD patients not on dialysis, labels recommend ESA treatment with hemoglobin levels < 10 g/dL if the rate of hemoglobin decline indicates transfusion likelihood and treatment goals are to reduce transfusion-related risks.38 FDA’s Oncologic Drug Advisory Committee reviewed data on peginesatide, a synthetic erythropoietin receptor–activating peptide.3 Structurally, peginesatide is a pegylated dimeric peptide (allowing for sustained activity and once-per-month dosing). Peginesatide bears no homology to erythropoietin, hence any antibodies that form are not cross-reactive. At the FDA Committee meeting, data were presented on two phase III trials for nondialysis CKD patients not on an ESA who received subcutaneous monthly peginesatide (n = 656 patients) versus every 2-week dosing of darbepoetin (n = 327 patients). Primary analyses showed that patients who received peginesatide at one of two doses (0.025 mg/kg/4 weeks or 0.04 mg/kg/4 weeks) had similar hemoglobin levels and those who received darbepoetin had similar hemoglobin levels (11.5,11.7, and 11.4, respectively) and similar mean hemoglobin changes (0.08 with 0.025 mg/kg dose peginesatide and 0.29 with 0.04 mg/kg peginesatide vs. darbepoetin). Safety analyses identified greater rates of adverse events, serious adverse events, deaths, adverse events ≥ grade 3, and adverse events leading to permanent discontinuation of peginesatide (Table 2). For CKD patients who participated in two phase III trials for dialysis CKD patients (intravenous epoetin alfa administered one to three times per week as comparator in the United States and intravenous or subcutaneous epoetin alfa or β one to three times per week as comparator in the United States and Europe),3 primary analyses showed that patients who received peginesatide (n = 1066) versus epoetin (n = 542 patients) had similar hemoglobin levels at weeks 29 to 36 (11.1 vs. 11.3, respectively) and similar mean hemoglobin changes (− 0.24 vs. − 0.09). In these studies, similar proportions of patients experienced adverse events, serious adverse events, deaths, adverse events ≥ grade 3, and adverse events leading to drug discontinuation (Table 2).
In 2012, the FDA granted approval for peginesatide as an ESA for treating anemia due to CKD in adult patients on dialysis.
In 2007, European Public Assessment Reports (EPARs) warned against achieving hemoglobin levels > 12 g/dL when ESAs were administered to CKD patients.
In 2008, Europe’s Pharmacovigilance Working Party recommended administering ESAs to CKD patients with symptomatic anemia, targeting hemoglobin levels between 10 and 12 g/dL. EMA’s CHMP recommended avoiding hemoglobin levels > 12 g/dL.21 The Anemia Working Group of the European Renal Best Practice Work Group and the Kidney Disease: Improving Global Outcomes endorsed KDOQI guidelines.
In 2009, revised EPARs warn of risks of serious problems with heart or blood vessels if high hemoglobin levels are maintained.39
In 2010, the Anemia Working Group of the European Renal Best Practices recommended target hemoglobin values of 11 to 12 g/dL among CKD patients and that ESA doses required to achieve the target hemoglobin should be considered.40 Caution was advised when administering ESAs to patients with type 2 diabetes not undergoing dialysis.
In 2011, the National Institute for Health and Clinical Excellence reported stroke, hypertension, and cardiovascular risks among nondialysis CKD patients and vascular access thrombosis risks among dialysis CKD patients with hemoglobin target levels of 12 g/dL versus lower levels.41 ESA initiation was recommended at hemoglobin levels < 11 g/dL or if symptoms were present, with target hemoglobin levels of 10 to 12 g/dL. Targeting hemoglobin levels > 12 g/dL among nondialysis CKD patients was not cost-effective, whereas cost-effectiveness information on dialysis CKD patients was unavailable.41
In 2007, revised labels warn that maintaining high hemoglobin levels resulted in risks of heart attack, heart failure, stroke, blood clots, and death and recommended maintaining hemoglobin levels between 10 and 12 g/dL.24 Canadian Society of Nephrology guidelines recommend target hemoglobin levels of 10 to 12 g/dL, citing excess mortality with higher hemoglobin targets.42
In 2008, the Canadian Agency for Drugs and Technologies in Health concluded that ESAs improved quality of life and decreased cardiovascular mortality.43 Targeting higher hemoglobin levels increased hospitalization, adverse events, and vascular access thrombosis. A hemoglobin target of 11 g/dL resulted in the largest quality-adjusted life years benefit (incremental costs of Can$25,000) compared with low target hemoglobin levels among CKD patients and was recommended.43
Patterns of ESA use have changed with ESAs. Among cancer patients receiving chemotherapy, ESA utilization increased 3.5-fold from 2000–2001 to 2006–2007 in the United States and Europe, and the frequency of cancer patients on chemotherapy receiving ESAs increased in the United States and Europe but not in Canada. Since 2007, ESA use among cancer patients receiving chemotherapy decreased in the United States (by 60%) compared with Europe (3% decrease) or Canada (3% increase). In 2008, among cancer patients with chemotherapy-induced anemia, rates of ESA use were 61% in Europe, 31% in the United States, and 20% in Canada (Fig. 2). In comparison, in 2007, among cancer patients with chemotherapy-induced anemia, rates of ESA use had been 64% in Europe, 80% in the United States, and 16% in Canada.
For CKD patients receiving hemodialysis, per-person ESA utilization and hemoglobin levels increased from 2000 to 2007.45 Between 2007 and 2009, per-patient ESA utilization for CKD patients decreased in the CKD setting in the United States (by 20%) but not in Canada (3% increase) or Europe (no change) (Fig. 3). In 2009, mean weekly erythropoietin doses for CKD patients on hemodialysis were 18,000 Units in Canada, 17,500 Units in the United States, and 13,000 Units in Europe. In comparison, in 2007, mean weekly erythropoietin doses for CKD patients on hemodialysis had been 17,500 Units in Canada, 21,000 Units in the United States, and 13,000 Units in Europe. Between 2011 and 2012, ESA use in the CKD setting for patients on hemodialysis decreased by 18%, while rates of transfusions increased by 22%. (Reimbursement for ESA use among CKD patients with hemodialysis in the United States was included as a bundled payment).
Utilization data for CKD patients not on dialysis are available only from the United States. Since 2005, ESA use among nondialysis CKD patients declined from 60 to 46%, and mean monthly dose among nondialysis CKD patients who received ESAs declined 25%, with the largest decline beginning in 2007.46 The mean hemoglobin level in ESA-treated CKD patients not on dialysis declined from 11.5 to 10.6 g/dL. The percent of treated nondialysis CKD patients with hemoglobin levels > 12 g/dL decreased from 27 to 12%, and mean doses declined from 173 to 111 mcg/mo.
Assessments of ESA safety and efficacy has undergone extensive revision. Initial ESA dosing and scheduling recommendations were based on small trials with narrow selection criteria and short follow-up.47 Initial recognition of mortality risks were identified in a phase III trial for CKD patients in 1998 and among two phase III trials of persons with cancer in 2003.1,4,5 Clinical trials with an active comparator for CKD patients and meta-analyses of clinical trials with cancer patients confirmed these safety concerns.2,6,8,10,30,41-43,48
In oncology in the United States, ESA approval is for chemotherapy patients not receiving potentially curative therapy. In Europe and Canada, regulatory approvals are less restrictive for cancer patients with symptomatic chemotherapy-associated anemia. For CKD, safety assessments are restrictive in the United States.2 In the United States, analysis of Medicare claims data for CKD patients on dialysis identifying mortality risks with higher ESA doses supported the FDA’s 2011 advisory to remove target hemoglobin levels.2,31,49 ESA labels in Europe and Canada for CKD patients include trigger and target hemoglobin levels.40,42 In 2011, ESA labels in the United States discouraged ESAs for most CKD patients not receiving hemodialysis. In Europe, this restriction is primarily for CKD patients with type 2 diabetes not receiving hemodialysis; no similar restrictions are noted in Canada.38,50,51
Reassessments of ESAs have occurred internationally, although responses to these reassessments vary. The result is marked decrease in ESA use in the United States but not in Canada or Europe.
This manuscript was supported by funding from a grant from the National Cancer Institute (1R01CA 102713–01), the Centers for Economic Excellence program of the state of South Carolina, and Doris Levkoff Meddin Center for Medication Safety, and a grant from the NIH (1R01CA165609-01A1), SONAR TKI.
Issue Theme Adverse Effects of Drugs on Hemostasis and Thrombosis; Guest Editors, Hau C. Kwaan, MD, FRCP, and Charles L. Bennett, MD, PhD, MPP
Conflict of Interest
Samuel Silver, MD, PhD: Consults for Amgen and 3M; David Goldsmith, MD: Speaking/Consulting Honoraria and Lecture Fees from Amgen, Roche, Sandoz, Takeda.