The introduction of mandatory risk evaluation and mitigation strategy (REMS) programs, in which patients must participate in order to receive a new therapy, provides an opportunity to examine an issue affecting ethical oversight and publication of scientific study results. Only limited experience exists to indicate how well current REMS programs have performed in improving benefit-risk profiles. The scant information that is available about these programs comes primarily through transcripts of US Food and Drug Administration (FDA) advisory committee meetings.
Little, if any, information from established mandatory REMS programs has been published in the scientific literature. One reason is a seeming paradox: The programs are intended to provide data to inform future decisions about the safety of the product, which presumes public disclosure and discussion within the scientific community of the monitoring results. Nevertheless, data from some REMS programs have not been published as research findings, in part because they are not considered to represent medical research (as distinct from medical practice or quality assurance).
A mandatory REMS program violates one guideline for ethical research—investigators making participation in research a condition of treatment. Many institutional review boards (IRBs) will not make exceptions to this ethical guideline for programs that are mandated for safety reasons, and they may therefore refuse to even evaluate the programs as research.
Assuring Safe Use
The FDA was granted authority to require a REMS for selected newly approved drugs under the Food and Drug Administration Amendments Act of 2007 (FDAAA, US Public Law 100-85). REMS programs may involve (1) required provision of a written medication guide to patients; (2) a formal communication plan to assure adequate education of patients, physicians, pharmacists, and other health-care providers; or (3) “elements to assure safe use.”
The romiplostim (Nplate) REMS, as noted in this issue of ONCOLOGY by Jamali and colleagues at the FDA, includes “elements to assure safe use” in the form of a wholly controlled and monitored release program. Of the 60 approved and deemed REMS,[3,4] only 20 include “elements to assure safe use.” This mechanism affords the FDA and the manufacturer the opportunity to make selected therapies available to patients before sufficient data are available to support a more fully informed benefit-risk assessment. Once those data are collected, it may continue to provide the mechanism to make the drug available while reducing its risks. The program controls access to the drug and collects safety data to inform subsequent policy assessments.
REMS Data Not ‘Research’?
The DHHS Office for Human Research Protections (OHRP) has previously advised investigators that results of mandatory monitoring programs could not be considered research and therefore could not be published in the scientific literature (personal communication, A.A. Mitchell, June 8, 2009). Such an interpretation led to an academic organization agreeing not to publicly disclose the results of a study relating to thalidomide(Drug information on thalidomide) (Thalomid).[5,6] Although the results of some voluntary programs conducted with IRB oversight have been published in the scientific literature,[7-10] results from mandatory programs have been published inconsistently, and the limited public disclosure has generally restricted scientific scrutiny and debate. Some REMS researchers may avoid submitting results for publication, anticipating rejection by journal editors owing to lack of IRB oversight or fearing sanctions from OHRP.
The question of whether collecting and evaluating data from REMS programs constitute research covers old ground. A similar debate arose regarding the distinction between public health practice (activities performed under federal, state, or local government authority, in which the benefits largely accrue to the individual participants) and research (activities intended to create knowledge to benefit a population broader than the participants).[11,12] That discussion has focused primarily on the nature of oversight required to protect the individual participants. In public health practice, governmental agencies (and their employees) are accountable for the well-being of individuals. For research conducted or supported by government agencies, IRBs (and investigators) assure accountability for human subject protection.
In the private sector, IRBs also provide such oversight for research, but activities performed as quality assurance are not subject to IRB review. While institutions and clinicians have an ethical obligation to protect patients, the formal protection afforded by IRB review is denied to participants in such activities. In the case of REMS that are conducted by pharmaceutical companies and their research collaborators, the IRB may be a primary mechanism available to assure human subject rights. Moreover, participating research collaborators are generally obliged to submit their research activities for IRB review and approval, a situation that is confusing at best.
The issues surrounding a quality assurance program designed to prevent certain hospital-acquired infections in intensive care units provide further evidence of this confusion. In 2006, Pronovost and colleagues published the results of this program; the investigators requested review by the Johns Hopkins University IRB, which determined that it was exempt research and therefore did not require informed consent. Following a complaint, OHRP initiated an investigation of the University’s compliance with federal regulations concerning research and ultimately determined that the IRB’s decision was incorrect and that the program “would likely have been eligible for both expedited IRB review and a waiver of the informed consent requirement.”[14,15]
Subsequently, OHRP issued guidance concerning regulation of quality improvement activities. While these guidelines address the situation in which a quality improvement activity includes a research component, they do not address the use of data collected by a controlled distribution system that is part of a REMS. Guidance from OHRP to acknowledge that these programs can constitute research and that consent or voluntary participation can be waived would facilitate the conduct and reporting of findings from REMS.
Romiplostim and Related Drugs
Could potentially important drug safety information result from research involving the REMS program for romiplostim? Jamali and colleagues present a summary of the information that led to the licensure of this drug, a subcutaneously administered Fc-peptide fusion protein that acts as a thrombopoietin receptor agonist. The approved indication is to treat patients with idiopathic thrombocytopenic purpura (ITP) who have had inadequate responses to corticosteroids, immunoglobulins, or splenectomy. Licensure was granted despite safety concerns that include reactive antibody production, worsening thrombocytopenia after discontinuation, and marrow reticulin deposition (fibrosis). To offset these concerns, Amgen, the manufacturer of romiplostim, agreed to institute a REMS program, which will restrict access to romiplostim, as a condition of licensure. Romiplostim is one of several drugs that stimulate platelet production for which safety issues have emerged over the past decade. The first such drug to be developed is oprelvekin(Drug information on oprelvekin) (Neumega), also known as recombinant human interleukin 11.
Interleukin 11, a multifunctional cytokine, was found to stimulate megakaryocytopoiesis in vitro and to increase platelet production in animals. After a phase I study demonstrated that oprelvekin also increased platelet counts in humans, two randomized placebo-controlled trials supported its licensure in 1997 to ameliorate thrombocytopenia and decrease the need for platelet transfusions in adult patients receiving myelosuppressive chemotherapy for nonmyeloid malignancies who are at risk of severe thrombocytopenia.[18‑20] During prelicensure clinical development, oprelvekin was found to be associated with fluid retention, dilutional anemia, cardiovascular events (including atrial arrhythmias in 15% of patients treated with 50 µg/kg in clinical trials), and papilledema. It has found limited application in oncology practice.
Thrombopoietin is the primary physiologic regulator of platelet production and is more potent than oprelvekin in stimulating platelet production. The gene that encodes it was cloned in 1994, and clinical trials of recombinant forms of human thrombopoietin started quickly. Development of one product derived from thrombopoietin, called megakaryocyte growth and development factor (MGDF), was stopped after it caused antibody responses that inhibited the activity of endogenous thrombopoietin and led to severe and prolonged autoimmune thrombocytopenia in several healthy volunteer platelet donors.
Eltrombopag (Promacta), an orally administered thrombopoietin receptor agonist, was licensed by FDA in 2008. It is approved for the treatment of chronic ITP in patients with an inadequate response to corticosteroids, immunoglobulins, or splenectomy. The product label contains a “black box” warning for hepatotoxicity. Other safety concerns include nausea, vomiting, menorrhagia, myalgia, paresthesia, cataract, dyspepsia, thrombocytopenia, and conjunctival hemorrhage (although some of these are likely related to ITP), as well as potential drug-drug interactions. Like romiplostim, the approval of eltrombopag was conditional on institution of a REMS program including restricted access.
Given the historical precedents of oprelvekin, MGDF, and eltrombopag, the decision to release romiplostim through a mandatory REMS seems prudent. But the question remains: How can we learn about the patient experience with this product and about the success of the program in mitigating risk? Researchers should be encouraged to evaluate the effectiveness of REMS programs, but the current rules discourage them. Researchers and IRBs should be able to contribute to protection of participant welfare without fear of OHRP sanctions. Editors of scientific journals should be free to accept for publication findings from REMS evaluations regardless of whether the program is deemed to be research or quality improvement by the IRB or OHRP.
The information emerging from these programs has value beyond describing a particular drug or REMS program. For example, data characterizing a drug’s safety in special patient populations may be relevant to another medication in the same drug class, and data on patient compliance from one REMS program may inform the design of future REMS programs. When access to information is restricted to regulatory agencies and the sponsoring companies, the public is not well served.
Financial Disclosure: Dr. Kaye is a consultant for Wyeth and Amgen.