FDA Approves Atebrioz (Zilurgisertib) — Third-Ever FOP Drug, Approved on a Secondary Endpoint After the Prespecified Primary Missed
On September 25, 2026, the FDA approved zilurgisertib (Atebrioz) — a once-daily oral ALK2 inhibitor developed by Mirum Pharmaceuticals under license from Incyte — "to reduce the volume of total new heterotopic ossification (HO) in adults and pediatric patients aged 12 years and older with fibrodysplasia ossificans progressiva (FOP)." It is the third FDA-approved therapy for FOP, after palovarotene (Sohonos, 2023) and garetosmab (Pasatru, August 2026) — and unlike garetosmab's adult-only label, it is indicated down to age 12.
The regulatory nuance is the notable part. In the phase 2 PROGRESS trial (Cohort 1, n=63), the prespecified primary endpoint missed: "One patient (3.1%) in the zilurgisertib arm developed new lesions vs five patients (16.7%) in the placebo arm, an 81% relative reduction (P = .0986). This difference did not reach statistical significance." The FDA instead anchored approval on the ranked secondary endpoint — change in total new HO volume on whole-body CT1: "At week 24, patients receiving zilurgisertib had a mean decrease of 3.2 cm³, compared with a mean increase of 24.6 cm³ in the placebo group (nominal P = .004)." Supporting measures: new-lesion volume down >99% (0.003 vs 6.57 cm³, P<.0001), annualized flares 2.34 vs 4.55, and through week 48 "no new HO lesions were observed among 61 patients with available whole-body CT data."
Safety was clean enough that "no AEs led to discontinuation or dose reduction"; the label carries an embryo-fetal toxicity warning and contraception requirements. The application carried fast track, priority review, and orphan drug designations; pediatric cohorts (6–<12 and 2–<12 years) are still enrolling.
Why it matters: FOP — the progressive "stone man syndrome" — went from zero approved drugs in 2023 to three in three years2, and this is the first oral ALK2 inhibitor (targeting the ACVR1 pathway upstream of the others). But the approval also shows the FDA accepting a volume-based imaging endpoint after the lesion-count primary missed — a data point for how it weighs hierarchical endpoints in ultra-rare diseases.
-
An instance of Narrowing a drug's clinical claims is the ultimate escape hatch for applications doomed by trial design disputes. — With the lesion-count primary dead at P=.0986, the application survived by anchoring on the ranked secondary imaging endpoint — endpoint flexibility salvaging a review whose headline measure failed. ↩︎
-
An instance of First-in-history drug approvals now arrive in waves, not once per generation. — A disease emptying its no-therapy list in three years is the approval wave operating at full speed. ↩︎