GAITHERSBURG, MD-In the fourth FDA session in recent times about erythropoiesis-stimulating agents (ESAs), the Oncologic Drugs Advisory Committee met here on March 13 to discuss the many concerns the agency has about the safety of the agents.
ODAC had also met on May 4, 2004 and May 10, 2007; and on September 11, 2007, the problems were taken up by the Cardiovascular and Renal Drugs Advisory Committee and Risk Management Advisory Committee.
In 1993, ESAs were approved for cancer patients to reduce the number of red blood cell transfusions, but two things have happened since then. First, the risk of transfusion-borne infection has decreased significantly; and second, eight controlled clinical trials now provide evidence or suggest increased risk of mortality and/or tumor promotion in patients who have head and neck cancer, breast cancer, non-small-cell lung cancer (NSCLC), or cervical cancer.
Also, said Vinni Juneja, MD, Medical Officer with the FDA Division of Biologic Oncology Products, some ESA benefits remain unproven, such as improved quality of life, fatigue, and other symptoms of anemia.
Major Safety Concerns
These new studies are either substantially larger or represent a different underlying histology than the ones originally used to establish ESA safety, he said.
Almost from the beginning, safety has been problematic. The 2004 ODAC meeting led to revised product labeling and establishment of pharmacovigilance programs, and the meeting last May resulted in label revision that includes a list of primary tumors with adverse outcomes, a note about the possibility of shortened survival and tumor progression even when ESAs are dosed to a target hemoglobin of less than 12 g/dL, a warning to use only for treatment of anemia due to myelosuppressive chemotherapy, and a recommendation for discontinuation after completion of chemotherapy. Also, a medication guide replaced the patient package insert.
FDA further claims that there are insufficient data to rule out a risk of mortality, shorter time to tumor progression, or lower locoregional control in cancer other than NSCLC. In addition, when ESAs are administered at doses aimed at maintaining hemoglobin levels higher than needed to avoid transfusion (off-label use), the risks are clearly unacceptable.
For these and other reasons, FDA asked ODAC to discuss the problems and take any or all of the following actions:
* Remove the indication to treat anemia caused by cancer chemotherapy.
* Restrict the indication to only patients who will not be cured by treatment intervention, and contraindicate use as an adjunct to adjuvant therapy.
* Restrict use to specific cancer subtypes where safety has been adequately assessed (only NSCLC).
* Contraindicate use where harmful effects have been demonstrated (breast and head and neck cancers).
* Mandate risk management strategies.
ESA Boons & Banes
The FDA approval for epoetin alfa (Procrit, made by Ortho Biotech; and Epogen, made by Amgen) and darbepoetin alfa (Aranesp, Amgen) was based on statistically significant reductions (20%-30%) in patients requiring transfusion.
Paul Eisenberg, MD, of Amgen Global Regulatory Affairs and Safety told the committee that ESAs are the only alternative to transfusion. He acknowledged that the risks of transfusion have been minimized, but they still exist, and that the US blood supply is often at or near crisis.
Moreover, an extensive published literature of placebo-controlled studies supports the benefits of ESAs in improving anemia symptoms such as fatigue. Although these data do not meet FDA's guidance for patient-reported outcome endpoints, they are nevertheless consistent with correction of anemia.
But the risks attached to ESAs continue to mount, as demonstrated by four studies presented last May at ODAC. Since then, two more (PREPARE and GOG191) have added to the evidence.
Dr. Eisenberg acknowledged the safety concerns, but he said that all eight studies were conducted in treatment settings not approved for ESAs. Other controlled trials performed outside the labeled indication (but considered informative with regard to mortality and/or tumor progression) have not suggested increased risk.
He made a final point: Meta-analyses in more than 8,000 subjects do not indicate a clear effect of ESAs on mortality or tumor progression-in contrast to the well-documented risk of thrombovascular events [TVEs].
Tom Lillie, MD, PhD, of Amgen Global Development maintained that the ESA role in tumor progression is unclear: If chronic exposure to ESAs promoted tumor progression, one might expect earlier emergence of cancers and more aggressive disease. Moreover, the erythropoietin receptor (EPO-R) gene is not an oncogene and is not amplified in tumors. Nor is EPO-R mRNA increased in tumors.
He did, however, acknowledge that the role of EPO-R in tumor biology is unclear.
What Might be Happening
Neither FDA nor members of ODAC said they doubted that ESAs have a valuable clinical benefit in chemotherapy-induced anemia, and everyone at the meeting agreed that ESAs should be used only to avoid transfusion. However, the very nature of the mechanisms by which they work might contribute to a clinical detriment.
First, TVEs are common in cancer patients as well as in certain chemotherapy regimens, and they are a recognized complication of ESAs. Thus, said Dr. Eisenberg, they could, in and of themselves, cause increased mortality.
Second, the increased risk of tumor promotion and/or angiogenesis may be a function of EPO-R. ESAs bind to the EPO-R on red blood cells, with resultant phosphorylation of Jak2 kinase and activation of downstream signaling. EPO-R is found on RBC precursors, the only cells in which functional EPO-R has been demonstrated. But it might also appear on tumor cells, which, if true, could cause increased tumor progression and shortened survival.
Dr. Eisenberg said that a recent NCI workshop concluded that despite extensive investigation, there are no data to support this scenario. There is no evidence that EPO-R functions as an oncogene.
Third-and paradoxically-tumor hypoxia, a known predictor of poor response to radiation, may be worsened at high hemoglobin concentrations.
Dr. Lillie talked about targeted vs achieved hemoglobin. Randomized studies in cancer and chronic renal failure demonstrate the risk of targeting high hemoglobin [above 12 g/dL]. High targets require increased drug exposure, especially in poor responders, who then have an increased risk of mortality as well as limited benefit.
He said that Amgen and Johnson & Johnson will recommend additional label modifications to reduce inappropriate exposure in chemotherapy-induced anemia: hemoglobin initiation at 10 or less, limitations on dose escalation, and discontinuation for nonresponders.
PREPARE & GOG191 Studies
The open-label, randomized PREPARE study of darbepoetin alfa, with a hemoglobin target of 12.5 to 13 g/dL, was conducted in 733 patients receiving neoadjuvant breast cancer treatment. Darbepoetin was given over the duration of the chemotherapy, which consisted of epirubicin, cyclophosphamide, and paclitaxel.
The primary endpoints were relapse-free and overall survival. Secondary endpoints were comparison of the two chemotherapy arms, and comparison of relapse-free and overall survival between the darbepoetin and placebo arms.
Patients receiving ESAs had shorter three-year, relapse-free, and overall survival than those on placebo.
The randomized controlled GOG191 study was conducted in 114 patients (out of a planned 160) with cervical cancer receiving concurrent cisplatin and radiotherapy. Epoetin alfa was given at 40,000 units per week for seven weeks, with a target of 12 to 14 g/dL. The dose was increased to 60,000 units per week if hemoglobin could not be maintained at 12 g/dL.
The primary endpoint was progression-free survival, and the secondary endpoints were overall survival and local control.
The study was terminated early because of an increase in TVEs in patients treated with epoetin alfa compared with controls. Both local and distant recurrence were more frequent in the epoetin patients, and three-year progression-free and overall survival were lower in the epoetin group.
Risk Management
Adrian Thomas, MD, of Global Safety and Benefit Risk Management for Johnson & Johnson's Pharmaceutical Group, told the committee that his company and Amgen have developed a risk-minimization plan that includes:
* Design of a new large placebo-controlled study (6,186 patients) to evaluate overall survival and tumor progression in patients with solid tumors.
* Safety updates to health care professionals.
* Development of a medication guide to keep patients informed and enhance physician-patient discussions about risk.
* Physician education.
* Communication with cancer patients and advocacy groups.
* Labeling changes-for example, initiation hemoglobin level.
In addition to these proposals, FDA said the sponsors should consider two other points. First, an informed-consent process could facilitate communication between patients and physicians. Patients would be required to consent to ESA treatment-after physicians had discussed its risks and benefits, as well as alternative treatments.
Second, advertising and promotion could be limited by no direct-to-consumer ads, restrictions on physician incentives, and promotion limited to defined specialties and journals.
Still despite the assurances of Amgen and Johnson & Johnson, there was still concern about safety. Richard Pazdur, MD, Director of the FDA Office of Oncology Drug Products, said, Eight controlled clinical studies provide evidence of increased mortality and/or tumor promotion when the agents are given to patients with certain cancers. These data suggest a tightly linked association.
The agency has placed less weight on other randomized studies because many were small, had limited data collection, limited follow-up, or inconsistent use of ESAs. But they do not negate the evidence of harmful effects.
FDA Questions
In view of these strong and significant reservations, FDA asked ODAC the following questions:
First, and the one on which all subsequent questions were based: Considering all the available data on the benefits and risks of ESAs in the treatment of anemia in chemotherapy-induced anemia, do you recommend that these products remain on the market for this indication? With remarkably little discussion or argument, the members voted 13-1 in favor.
The next four questions were about labeling:
* To date, only clinical trials in NSCLC have excluded an increased risk of death in patients receiving ESAs. Should the current indication be restricted only to patients with lung cancer? The vote was 8-6 against such a restriction.
* The PREPARE trial demonstrated decreased relapse-free and overall survival in breast cancer patients on neoadjuvant chemotherapy. The risks and benefits are different for patients with metastatic or incurable disease. Should the current indications be modified to state that ESAs are not indicated for patients receiving potentially curative treatments? Eleven members voted yes, two said no, and one abstained.
* For the most part, adverse findings have been found in two malignancies: breast, and head and neck. Should ESAs not be used in those two cancers? The vote was 9 to 5 in the affirmative.
* Not all patients with anemia require a transfusion, and the label does not specify the hemoglobin level at which ESA treatment should be initiated. Assuming a patient is asymptomatic and has no comorbid conditions, specify the hemoglobin level at which to initiate treatment. Members of the committee agreed that too many variables are at play to determine a specific hemoglobin level. They said that physician judgment should be the operant factor here but that a hemoglobin of 10 g/dL was reasonable.
The final two questions concerned minimizing risk:
* Informed consent would require a patient's authorization to undergo treatment with an ESA, and both patient and physician signatures would be required. In the process, the physician would discuss risks, benefits, and alternative treatments. Should FDA require informed consent for such treatment? The vote was 8 yes, 5 no, and 1 abstention.
* Restricted distribution systems can require identification and enrollment of health care providers who agree to prescribe only in accordance with product labeling and who commit to patient education regarding safe use. Patient registration can also be required. Should FDA mandate a restricted distribution system for oncology patients receiving ESAs? Ten members voted that it should not, one voted for such a system, and two abstained.
© 2008 Lippincott Williams & Wilkins, Inc.