$ABVX malignancy “signal,” straight:
2 cancers, 2 different organs, 1 patient each, no clustering … and infections below placebo.
That’s the fingerprint of background ascertainment, not a drug effect.
Why JAK is the wrong comp:
JAK’s malignancy case was convergent … immune-surveillance MoA + preclinical signal + dose-response + a head-to-head loss to anti-TNF, all pointing one way.
Obe has none of those pillars. It only rhymes on the headline.
Safety can kill a product pre-authorization, but it kills on a confirmed, serious, unmanageable risk, not a watched one. Marketing authorisation turns on efficacy and baseline safety; a manageable or merely-possible risk doesn’t block the gate …
it gets routed to post-licensing follow-up, where the data to actually adjudicate it accrues.
Two unclustered cancers in a confounder-heavy population is the textbook “watch it, don’t gate on it” case.
Likely path:
isolated events, watched question deferred to post-marketing pharmacovigilance / a PMR, monitoring language on the label … no box, no Rx deterrent.
Validated if:
the exposure-adjusted Part 1 50mg-vs-placebo malignancy rate lands in IBD background range, and the Type B pre-NDA doesn’t flag malignancy as a major review issue.
My opinion only, no investment advice.
English
