A Food and Drug Administration advisory committee of outside scientists and doctors on Thursday voted 10-3 to recommend approval of Replimune’s melanoma immunotherapy, overriding the agency’s own scientific staff, who raised serious methodological concerns about the trial data. We hope FDA leaders listen to their own reviewers when they decide as soon as this weekend whether to approve the therapy.
Market analysts had predicted the panel would vote against Replimune’s RP1 after FDA staffers this week released thoroughly reasoned briefing documents identifying serious problems with the trial design. The questions were so pointed that the company and its Wall Street backers couldn’t wait to dismiss them. The FDA reviewers cited a carefully documented list of reasons why the outside experts should scrutinize the trial results before rubber-stamping them, as we described in a recent editorial.
RP1 is a re-engineered herpes virus injected into melanoma tumors that bursts cancer cells. The release trains the immune system to track down and kill cancer cells hiding around the body. Many patients with metastatic melanoma don’t respond to, or develop resistance to, other front-line immunotherapies.
Replimune’s therapy turbo-charges the response to other immunotherapies by unmasking cancer cells. In this way it shrinks both superficial and distant tumors — though the trial data fails to establish that the observed shrinkage came from the drug rather than from confounding procedures. A third of patients whose cancer worsened on other immunotherapies experienced remission on RP1, with cancers vanishing in one of six — a number that impresses only if one assumes tumors didn’t merely shrink from the biopsies, the immunotherapy patients had already failed, or the needle itself.
While FDA staffers exposed the weakness of this benefit, one oncologist at the advisory committee hearing pointedly dismissed the methodological concerns the agency’s own scientists had raised. Another oncologist criticized the FDA for demanding rigor rather than accepting preliminary results at face value.
The FDA reviewers said tumors could only have shrunk from run-of-the-mill biopsies performed during the trial, or from the immunotherapy that patients hadn’t previously responded to, or from the mere poking of the tumor with a needle. These are precisely the confounding variables that sound clinical trial design accounts for before a therapy reaches the market. The company and its backers would rather the public marvel at their impatience than engage with the science.
Framing that question as an obstacle to patient care is the real act of deception. Melanoma patients and their family members who testified at the hearing understandably begged the agency to approve the medicine — their desperation is real, and their hope deserves better than a drug approved on data that won’t hold up. The advisory committee chose emotion over methodology. The panelists were moved by compelling patient testimonies — exactly as designed by a company that knows emotional appeals beat statistical rigor in a public hearing. Protecting dying patients by making them experimental subjects in a company-purchased panel verdict: George Orwell, call your office.
At least the FDA’s internal dissenters are doing their jobs. The story here isn’t bureaucratic obstruction — it’s a capture narrative, and the captured have the courage to say no.
