Companies designing cancer trials should take one lesson from last week's Oncologic Drugs Advisory Committee meeting: if overall survival is the primary endpoint, sponsors should not expect that showing progression-free survival as the secondary endpoint will get them an approvable drug if the primary endpoint fails.
This is so even though FDA has unambiguously stated that progression-free survival (PFS) is an appropriate endpoint that can support approval for some forms of cancer. But the chances of persuading the agency that clinical efficacy has been documented are much higher if PFS is selected as the primary endpoint and trials are powered to show overall survival as a secondary endpoint.
PFS is an attractive measure of efficacy for a variety of reasons. The contribution of the investigational agent is clearer than in trials that look at overall survival (OS), and a modest effect that might be missed in an OS design can be detected.
But failure to plan for the complexities required to obtain a robust result, or to tightly control trial conduct, can doom an application, according to Richard Pazdur, director of FDA's Division of Oncology Drug Products. "From a practical perspective, progression-free survival, as a primary endpoint for drug approval, takes meticulous, prospective