Ethical, Logistical Issues: Experts Outline Criteria to Justify Randomized Studies

It's hard enough to conduct clinical trials for experimental medicines, but it can be even more difficult when patients already have access to the medicine outside the research setting. Thus the Food and Drug Administration should require post-market clinical studies, which expose patients to heightened risk, only if an important public health question is at stake and no other study design can supply needed evidence.

That's the conclusion of a report on "Ethical and Scientific Issues in Studying the Safety of Approved Drugs" published by the Institute of Medicine in May, which raises a number of important questions on this topic as part of its assessment of FDA policies for ensuring the safety of drugs through the product lifecycle.

FDA requested this IOM analysis in 2010 to help devise policies and programs to prevent future Vioxx-type safety crises. The expert panel issued a preliminary assessment in July 2010 to assist deliberations on drug safety issues by FDA advisory committees.

This final report outlines strategies for improving oversight of marketed drugs to better detect safety concerns. The panel advised FDA to create a comprehensive document that maps drug risks and safety concerns throughout the product life cycle, or benefit and risk assessment and management plan.

The proposal drew objections from agency officials, who claimed it would be "challenging" to develop a formal safety document for every approved drug, given the agency's limited resources, and that lengthy rule-making would probably be required to do so.

Risks warrant study

Much of the report discusses how FDA should weigh the need for post-market clinical trials and how such studies should be structured. FDA has required 385 post-market studies since 2008, according to a separate agency report. Post-approval research is most valid, the panel said, where benefits or risks of a drug are "particularly uncertain." That could apply to drugs approved based on surrogate endpoints that provide conflicting evidence about likely health outcomes; to first-in-class drugs with approval based on surrogate endpoints; and to drugs with troubling side effects that could affect a large number of people.

In deciding whether to require sponsors to conduct post-market studies, the experts emphasized that FDA must balance the need to guard against unsafe medicines against its ethical obligation to protect research participants.

Clinical studies should be required only if a regulatory decision cannot be made with existing safety evidence; if the research can reduce uncertainties about the benefit-risk balance of a medicine; if the results will support a decision in a timely fashion; and if the rights and interests of research participants can be protected. Asking individuals to accept additional risk, without any expectation of clinical benefit, is justified only when a pressing public health question is at stake, as when FDA has to decide whether to remove a drug from the market or restrict its use.

Even if post-market research is warranted, the experts acknowledged that such studies are not easy to conduct because patients may refuse to participate in randomized clinical trials (RCTs) once a medicine is readily available, and extensive patient self-selection can skew results.

Adherence also may be a problem, as well as early patient withdrawal from trials. When RCTs are justified, FDA should work with institutional review boards and data monitoring committees to ensure proper protection of participants' rights and interests. Active comparators should be used whenever possible, although placebo controls may be acceptable under certain circumstances.

Post-market studies

Post-market studies should employ informed consent procedures that address the unique aspects of this research, such as the need to assess a specific safety signal.

The difficulties with post-market RCTs may increase reliance on observational studies, particularly to detect rare or delayed risks.

In anticipation of such a shift, the panel advises FDA to establish a new body to advise on the ethical challenges raised in observational studies and surveillance activities.

FDA also should clarify whether its human subject protection regulations apply to post-marketing observational studies and emphasize that it's important for investigators to ensure the security of patient information obtained from health system databases.