FDA looks to bypass cancer drugs' companion diagnostics with new pilot program: Pazdur

Targeted cancer drugs typically require specific tests to identify patients who could benefit the most from the therapies. But as FDA oncology chief Richard Pazdur, M.D., sees it, these companion diagnostics don’t serve patients that well.

Existing cancer companion diagnostics are bundled with the specific cancer drugs they’re approved for. But the FDA is looking to get around this one-drug-one-test situation by looking at “minimal performance criteria” of tests, Pazdur said.

The concept of a minimal performance criteria, Pazdur said, would allow doctors to use any test that meets those standards, rather than having to stick to specific tests. But the idea doesn’t preclude testmakers from developing and selling drug-specific diagnostics, he added.

The FDA’s medical device regulators are working on a pilot program to implement that proposal, Pazdur said. The director of the Oncology Center of Excellence made the comment during a discussion with FDA Commissioner Robert Califf, M.D., at the Friends of Cancer Research annual meeting on Thursday.

The FDA’s regulation requires that a drug’s labeling includes information about tests necessary to select patients. So sometimes a targeted cancer drug comes with its own diagnostic even if another test is already approved for the same drug target.

Two examples are approved for non-small cell lung cancer with EGFR exon 20 insertion mutations: Takeda’s Exkivity uses Thermo Fisher’s Oncomine Dx Target Test to identify that biomarker, whereas Johnson & Johnson’s Rybrevant requires Guardant Health’s Guardant360 CDx test.

This creates the problem that patients might need to undergo multiple tests—potentially requiring larger samples—to identify the proper treatment even within the same drug class. And it could limit doctors’ choice when planning tests, Pazdur noted. It’s also possible that a doctor may prefer a particular diagnostic because of its cost or accuracy but not prefer the drug tied to it.

Instead of having a specific test, the pilot program the FDA is devising will look to have the minimal performance criteria of a test embedded into a cancer drug’s clinical trial to then select patients based on those criteria. The selection standards will be made publicly available for doctors to follow in clinical practice.

Besides potential waste in companion diagnostics, Califf also pointed out that these tests often have no external validation that their results are accurate and that the medical community have “vastly underestimated the consequences” of potential errors in diagnosis.

“I hope that we come up with more capability of ensuring that the laboratory community is delivering more reliable tests,” the FDA head said.