The paper: While agnostic literally means “nothing can be known,” last month’s JAMA editorial questions if we really should be comfortable not knowing. For starters, we do not know that there is, in fact, no mutation-tissue interaction influencing the efficacy of pembro. Merck’s supporting data includes 14 pooled cancer types with half of those represented by one(!) studied patient each (we're talking one patient with breast cancer, for instance). Consider for a moment vemurafenib, which has a response rate of 98% among BRAF-mutant hairy-cell leukemia, 48% among BRAF-mutant melanoma, and 0% among BRAF-mutant myeloma. The authors also question the cost of casting wide PCR and IHC nets to test every cancer patient for qualifying mutations, not to mention the costs of a growing number of tests as tissue-agnostic approvals expand across a wide array of mutations. They close with a call for expansive, continually-enrolling post-marketing trials spanning multiple institutions and cancer sites.
Bottom line: The FDA has rightly put the ball in our hands, and it is time for oncologists to take a concerted shot at answering these known unknowns.