For those interested in the FDA approval process (and who isn’t!), the Jan. 22/29 issue of JAMA contains a worthwhile set of articles.  Downing et al’s article, “Clinical Trial Evidence Supporting FDA Approval of Novel Therapeutic Agents, 2005-2012,” in particular, has gained attention (from, for example, the Wall Street Journal and Forbes).  The article describes a study of the clinical trial evidence supporting FDA drug approvals, and the authors conclude that the quality of that clinical trial evidence “varie[s] widely.”  As someone who used to work at FDA, this result strikes me as utterly unsurprising (and I imagine many FDA lawyers and scholars had the same reaction).

Here’s a little more background on what the study results actually were (see Downing et al’s paper for more details and results).  Downing et al report that between 2005 and 2012, “FDA approved 188 novel therapeutic agents for 206 indications on the basis of 448 pivotal efficacy trials,” and that the quality of the clinical trial evidence supporting approval of these 188 drugs varied widely across indications.  What the authors mean by “quality,” roughly, is the number of efficacy trials supporting approval, and the design of those trials. Most drugs were approved based on 2 efficacy trials, but for 74 indications (36.8%), drugs were approved based on only one efficacy trial.  Additionally, for 91 indications (45.3%), drugs were approved based solely on trials using surrogate endpoints as their primary outcome.  (Surrogate endpoints are “biomarkers expected to predict clinical benefit.”  Thus, for 91 indications, drugs were approved based on evidence that they affect something that we think, but don’t know with certainty, predicts an actual clinical benefit.)

I say these results are unsurprising because the Federal Food, Drug, and Cosmetic Act (FDCA) specifically authorizes FDA to approve drugs based on clinical trial evidence of such varied “quality.”  In general, section 505(d) of the FDCA requires that, to be approved, the safety and effectiveness of a drug be shown by “substantial evidence.”  Substantial evidence, in turn, is defined as “adequate and well-controlled investigations,” which generally means effectiveness must be shown in at least 2 trials.  Section 505(d) of the FDCA, however, also states that, when FDA determines that data from one adequate and well-controlled trial, along with confirmatory evidence, is sufficient to establish effectiveness, it may approve a drug based on one trial.  In addition to having this leeway for “standard” approvals, when a drug is granted “fast track” status – when it is intended for the treatment of a serious or life-threatening condition and demonstrates the potential to address an unmet medical need – section 506(b) of the FDCA authorizes FDA to approve a drug based on evidence that it has an effect on a surrogate endpoint.

In other words, Congress – for better or worse – has made a policy decision that, in certain circumstances, FDA should not require that drug approval be supported by two trials studying clinical endpoints.  Combine this statutory authority with the pressure that FDA often faces from industry and patients to approve drugs as quickly as possible, and Downing et al’s results seem predictable.  What would have been surprising – and arguably contrary to Congressional intent — would have been a result showing uniformity in the quality of clinical trial evidence supporting drug approvals.

That their results are unsurprising is not to say Downing et al’s study is without value.  This study, hopefully, will educate the public (and medical practitioners who prescribe drugs) about what kinds of evidence actually support many drug approvals.  We want patients and practitioners to understand the quality of evidence that supports drug approvals because the quality of that evidence goes to the amount of certainty we should have about the safety and effectiveness of a drug.  And, to the extent this study increases knowledge about the relative strength of the evidence supporting many drug approvals, it can also enable a more informed debate about whether Congress and FDA have struck the right balance between allowing (or hastening) access to effective drugs for patients who need them, and preventing harms from access to ineffective and unsafe drugs—the incredibly difficult, and ever-present, question about balancing access and safety.