in the postmarketing setting, FDA may be responding to new information about a drug whose benefit–risk profile has been or is being studied by multiple investigators, supported by various funders, using a variety of observational and RCT designs. FDA’s decision about what kinds of studies it should require to assess a drug’s benefit–risk profile after the drug is on the market may depend heavily on its critical assessment of the evolving evidence base and the evidential gaps that may remain. Because in the postmarketing setting patients may have been taking a drug for many years and others will continue to be prescribed the drug, FDA has the option of requiring observational studies that use existing or routinely collected patient information, an alternative that is not available in the premarketing context. Another difference between the premarketing and postmarketing contexts is that after marketing FDA can require both observational studies and RCTs, either simultaneously or sequentially. Although in the premarketing setting FDA can require multiple research designs and studies, its process typically is to pursue the sequence and progression of research from Phase 1 to Phase 3.
In the premarketing setting for a new molecular entity, access to the investigational drug by patients is possible only through participation in research conducted to satisfy FDA premarketing requirements or through a few other selected avenues that are also controlled by FDA, such as expanded access through “compassionate use” (FDA, 2011b). In the postmarketing setting, physicians are free to prescribe a drug to any patient for whom they think it medically appropriate. In most cases, if a doctor and a patient decide that it is in the patient’s medical best interests to take the drug, the patient does not have to enter a research study required by FDA to have access to it.
In the premarketing context, FDA regulators and the manufacturer typically focus on the drug’s benefit–risk relationship compared with a placebo and use active comparators less often, typically when it is either ethically or methodologically necessary, such as to evaluate assay sensitivity. In the postmarketing context, concerns about a drug’s benefit–risk profile are more likely to involve comparisons with other active treatments. For example, although at the time of approval and for some time thereafter a drug that poses serious risks may have a favorable benefit–risk profile, the acceptability of its profile may come into question when a new drug becomes available and appears to offer comparable benefits with less severe risks.
In the case of Avandia® (rosiglitazone), the presence in the marketplace of the clinical alternative pioglitazone, a similar drug with a purportedly more acceptable benefit–risk profile, affected the public controversy and FDA’s response to it. Although the remit of FDA is not to ensure that the drug supply contains only the comparatively “best” drugs for an indication, it has a duty to the public’s health to remove or restrict from the supply drugs that pose unacceptable risks in relation to benefit. In this respect, FDA’s responsibility to ensure that drugs continue to have a favorable benefit–risk profile may on occasion move the focus of required postmarketing research more toward comparative-effectiveness