FDA Guidance: Overall Survival as a Safety End Point, Not Just Efficacy

Harpreet Singh, MD, former director of the FDA's Division of Oncology 2 and chief medical officer of Precision for Medicine, discusses the FDA's draft guidance on the use of overall survival (OS) as an end point in clinical trials.

There's been a long-term need for guidance on using OS as an end point and how it should be evaluated. This new guidance is particularly interesting to Singh because there have been several advisory committee meetings and public workshops discussing OS: what it means, how it should be interpreted, and why it's being "rebranded" as the ultimate safety end point, even though it was classically viewed as an efficacy end point.

The impetus for this shift was the existence of several drugs—in fact, an entire class of drugs in the hematologic space, notes Singh—that showed very robust benefits in progression-free survival (PFS) but later demonstrated detriments in OS. This was likely due to toxicity, meaning the OS benefit was obscured by the harm, even though patients were getting a PFS benefit.

OS is widely viewed as the gold standard end point in oncology because patients want to live longer. They want to know that taking a drug (which always carries some toxicities and risks) will extend their life, Singh says.

However, powering a trial for an OS end point isn't always possible; in fact, it's frequently impractical.

There are a few key reasons why OS is often impractical, especially in a frontline setting, according to Singh:

1. Long Disease Courses: For diseases like multiple myeloma and prostate cancer, the disease courses are quite long, and it can take 10 or more years to get an OS readout.
2. Confounding by Subsequent Therapies: Patients frequently receive therapies after they progress. If a trial doesn't account for this, the OS data is confounded and isn't a "pure" view of the drug's effect.

Because OS often can't be used as a primary end point, this guidance addresses that exact scenario.

Crucially, this guidance is not about using OS as an endpoint, but rather about situations where trials do not have OS as a primary end point (where the primary is typically PFS.

It clarifies how the FDA will look at OS in these cases, and how they are asking sponsors to plan for an evaluation of OS even when it is not the primary measure.