Do Surrogate Endpoints Fail the Litmus Test?

A letter published in JAMA Internal Medicine questions the FDA's acceptance of surrogate endpoints for what the authors describe as "costly, toxic drugs that do not improve overall survival."

Objective response rate, progression-free survival (PFS), disease-free survival (DFS)—these surrogate end points are now commonplace in oncology drug development. The FDA regularly accepts these surrogate endpoints as clinical trial outcomes in place of overall survival (OS) data, with the objective of accelerating drug development and improving patient access to much-needed—and sometime life-saving—therapeutics. The expectation though, is that the developing company would be responsible and obligated to conduct post-marketing studies evaluating OS.

Now, a letter published in JAMA Internal Medicine, authored by Chul Kim, MD, MPH, from the National Cancer Institute, and Vinay Prasad, MD, MPH, from the Knight Cancer Center at Oregon Health and Sciences University, questions the FDA’s acceptance of these surrogates for what the authors describe as “costly, toxic drugs that do not improve overall survival.”

The authors evaluated drugs approved by the agency over a 5-year period between January 1, 2008, and December 31, 2012. The analysis included the pathway for approval (accelerated vs traditional) and the surrogate endpoint that may have been used in the study. For drugs that had received FDA approval based on a surrogate, the authors conducted a literature search to identify publications that’d report the drug’s impact on overall survival.

Of the 54 approvals during the study period, the authors recognized 36 (67%) drugs that were approved on the basis of surrogate endpoints. All drugs that received accelerated approval did so on the basis of surrogate endpoints, while only 21 of 39 (54%) traditional approvals used a surrogate endpoint. Of the commonly used endpoints, rate of response was employed 53% of the time, while PFS or DFS was used 47% of the times. Over a median follow-up period of 4.4 years, 5 drugs improved OS, 18 failed to improve OS, while 13 had drugs were either untested or had no reported survival results.

This study begs the question: if expensive cancer drugs are approved without evidence of a clinically-relevant benefit, the importance of value tools such as NCCN’s Evidence Blocks and DrugAbacus developed by Peter Bach, MD, at the Center for Health Policy and Outcomes at Memorial Sloan Kettering Cancer Center, hold all the more relevance.

While study author Prasad called on the FDA to Prasad called on FDA to “set a reasonable period of time — and we can debate what that might be – by which drugs must show they either improve how long patients live or how well they live,” the FDA remained confident of their approval procedure. Said an agency spokesperson, “We have had multiple discussions over a number of years within the global scientific and patient community, including with the Oncologic Drugs Advisory Committee, to gain consensus of the use of progression-free survival and response rate as endpoints to support approval of drugs that treat cancer. It has been widely accepted that the benefit of a drug can be demonstrated by a variety of endpoints, not just overall survival.”