FDA eyes shake-up of swift approval pathway in cancer

The FDA has laid out its thinking on the design of clinical trials used to support filings for accelerated approval of cancer drugs, with clear signs that it wants to tighten up the requirements.

The draft 'Clinical Trial Considerations to Support Accelerated Approval of Oncology Therapeutics' guidance includes the recommendation that cancer drug developers use randomised trials with a control arm, rather than open-label studies in which the experimental therapy is given to all participants.

The accelerated approval pathway was introduced to allow new therapies for serious diseases like cancer to be made available to patients more quickly, on the grounds that early access – often on surrogate measures of efficacy – is followed by a confirmatory study to show it is working as expected.

There have been concerns that those confirmatory trials were not always being carried out for many years, and recent events have not bolstered confidence in the system.

The latest document has emerged after close scrutiny of the early access pathway, which resulted from the FDA’s controversial approval of Biogen’s Alzheimer’s drug Aduhelm (aducanumab) in 2021.

It is used most commonly in oncology, however, and the new guidance comes after a series of cancer therapies were stripped of licenses because confirmatory trials were either not carried out or failed to support earlier efficacy findings, for example Merck & Co’s Keytruda (pembrolizumab) as a third-line gastric cancer therapy and Bristol-Myers Squibb’s Opdivo (nivolumab) in second-line liver cancer.

The preference for randomised controlled trials (RCTs) is the headline change in the document, with the FDA indicating it may still allow single-arm studies in some circumstances – for example, if an RCT is simply not feasible.

One key advantage of going down the RCT route is that longer-term follow-up of patients may serve as the confirmatory evidence needed to convert an accelerated approval to a full license.

The FDA would also like developers to run two RCTs if possible – for example, one with an early endpoint, like response rate, to support accelerated approval and a second with a clinical endpoint, like progression-free survival or overall survival.

The agency says, however, that those should run almost at the same time, as availability of the drug in clinical practice after approval can hold back enrolment of trial subjects. It has also provided guidance on the design of just one RCT if that is more appropriate for a particular development programme.