This article was adapted from STAT’s latest report, “Failed trial, successful drug: how a negative readout can still lead to FDA approval.”
When a clinical trial readout for what seemed a promising drug comes back negative, investors often contemplate jumping ship and companies wonder if continuing to pursue regulators’ approval is worth the trouble.
But while a failed trial might be bad news, it isn’t always an automatic deal killer for a new drug, as clinical trials analyst and consultant Frank David explains in STAT’s latest report, “Failed trial, successful drug: how a negative readout can still lead to FDA approval.”
There’s a shared sense among industry experts that the agency has grown more permissive about negative data since the mid-2010s, David writes in the report. And while Biogen’s Alzheimer’s drug Aduhelm, which was granted accelerated approval from the U.S. Food and Drug Administration after a failed clinical study, is probably the most well-known example, there are many other drugs that have been approved despite having mixed or wholly negative clinical results.
The report includes David’s exclusive analysis of FDA approvals from 2018 to 2022, and a case history, based on data from STAT Trials Pulse — a machine learning platform developed by AppliedXL that generates real-time insights and risk assessment on studies registered in clinicaltrials.gov — showing how early signs in a drug’s developmental path can predict trouble.
STAT recently talked with David about the report and what mixed or negative trial data really means.
Why isn’t a negative readout from a clinical trial the end of the road for a drug? Doesn’t it mean a drug isn’t effective and so pursuing FDA approval is pointless?
Not necessarily! In some cases, there’s a mix of trials with positive and negative results, and the FDA approves the drug based on the “totality of the evidence,” so to speak. There are also many situations in which a trial shows signs of clinical benefit, but just doesn’t quite hit the statistical threshold for being called positive. Depending on the severity of the disease, what other treatments are available, and how convincing the full data package is, regulators sometimes approve drugs even if there aren’t any frankly positive randomized clinical trials.
How common is it for a drug that has failed a trial to succeed in this way?
It’s hard to give precise odds for how often a drug with a failed trial gets approved, because there aren’t great statistics on FDA rejections. But we can calculate the fraction of FDA approvals that are based at least in part on negative clinical studies. In writing this report, I found that of the 236 drugs approved by the agency from 2018 to 2022, 11 had a mix of positive and negative trial results, and an additional five had no clearly positive clinical studies. So we can say that in a recent five-year period, almost seven percent of approved drugs had at least some negative trial data.
What are some factors that make it more likely that negative data won’t ultimately stand in the way of approval?
Although each of the historical examples profiled in this report is somewhat unique, taken together they illustrate some key themes that are common to drugs that got approved with negative clinical data versus ones that didn’t. The biggest takeaway from these cases is the importance of the full data package. All else being equal, a drug with four successful studies and one failure is probably far more likely to get a favorable reception from regulators than one with one failed trial and no successes. Similarly, robust pharmacologic and translational data that the drug is hitting its target and having important effects in people can be extremely helpful. In many cases, the FDA was able to make a defensible case that the drug was clinically beneficial even though it may have barely missed the mark on a study’s primary endpoint.
In the report, you talk about the importance of a company’s relationship with the FDA in what happens after a trial fails. How does that play a role in decisions?
It’s tempting to think that drug approval is an entirely objective, data-driven exercise, but actually there’s a huge human component. Whenever a company submits an application to the FDA, it needs to make a persuasive case to the agency that the drug should get the green light. That requires expertise in communicating with regulators, which typically comes from years of experience with these sorts of interactions.
When there are negative trial data, the stakes are even higher, and the quality of the relationship is even more important. That’s because the sponsor is asking reviewers to put their reputations on the line and justify why the failed study should be discounted. That’s a pretty high bar that requires not only a plausible, science-driven argument for approval that passes the “red-faced test,” but also a desire by reviewers to stick their necks out and defend it in public. And it’s hard to hit on both of those points without a strong working relationship between the company and the agency.
Admittedly, it can be hard or even impossible from the outside to tell if the relationship between the sponsor and the FDA is extra-tight. But in the opposite scenario, it’s a really bad sign for approval in the face of negative data if you see that the relationship appears fraught. If a sponsor is issuing press releases that make scientifically questionable arguments and seem to go against the FDA precedents and clear guidance, that story is unlikely to end well for the company.
What would you say the take-home message is for investors and drugmakers in all this?
The main point here is that although a clinical failure is a huge disappointment, it’s not an automatic kiss of death for a drug’s chances at the FDA. If you’re a biotech company or investor looking at a study that just went belly-up and you see strong parallels with any of the prior approvals described in this report, then there may still be a path forward. That doesn’t mean you should be an overly optimistic Pollyanna in these cases. But it does mean that you should look objectively at the situation, maybe with the help of unbiased external advisors, to decide if there’s a credible path forward.