What the research found

The FDA has formally alleged that avacopan's original approval relied on manipulated efficacy data. The ADVOCATE trial tested avacopan against standard care (rituximab or cyclophosphamide plus a prednisone taper) in 331 patients with severe ANCA-associated vasculitis, with two primary endpoints: remission at week 26 and sustained remission at week 52. The week 52 endpoint was critical—the FDA had explicitly told the manufacturer it would not approve the drug without demonstrating superiority on that measure.

According to the agency's formal notice, when the trial's locked database initially showed the week 52 analysis failed (p=0.1025, failing to reach statistical significance), company biostatisticians conducted an unblinded search of the data looking specifically for patients they believed were misclassified. They identified 9 patients whose reclassification would have flipped the result to a positive outcome—predominantly favoring the avacopan arm. The FDA argues this was a targeted search conducted only because the primary analysis had failed, and that the three control-arm patients included appear to have been selected in a way that could not meaningfully affect the superiority claim.

Why it matters for you

This episode illustrates a structural vulnerability in how drug approvals work—one that directly affects the treatments available to you. If you're considering avacopan for vasculitis or similar conditions, you're making a choice based partly on trial data you cannot independently verify. The allegation here isn't that avacopan is ineffective (real-world evidence may tell a different story), but that the specific statistical evidence used to approve it may have been selectively manipulated after the blinded analysis was conducted.

For MyKine users tracking autoimmune markers or considering immunosuppressive therapies, this raises a practical question: how confident should you be in a drug's efficacy when the pivotal trial underwent post-hoc readjudication? If you're currently on avacopan or considering it, this doesn't mean you should immediately stop—real-world safety and tolerability data, and your own biomarker response, matter. But it does mean you should be especially vigilant about tracking your own outcomes (remission markers, vasculitis activity scores, organ function) rather than relying solely on the label's claims.

Caveats

  • Allegation, not adjudication: The FDA has proposed withdrawal and made formal accusations, but ChemoCentryx (now Amgen) has not yet had a hearing or submitted their full defence; this remains a contested matter.
  • Data access: We are relying on the FDA's characterization of internal emails and documents; the full documentary record is not yet public.
  • Endpoint readjudication itself is not inherently fraudulent: Correcting misclassified patients is sometimes legitimate. The FDA's concern is about the timing and selectivity of the search—it occurred only after unblinded failure, and the choice of which patients to reclassify appears asymmetrical.
  • No comment on real-world efficacy: This concerns the pivotal trial data only, not whether the drug works in practice.