The FDA Approved Atebrioz on a Volume Endpoint the FOP Trial Listed as Secondary
The FDA approved Atebrioz for FOP on a change in new bone volume, a measure the trial registry listed as secondary, after the lesion-count primary missed statistical significance at week 24.
When the FDA approved Atebrioz (zilurgisertib) tablets on 25 September 2026, the stated basis was a change in the volume of total new heterotopic ossification in patients 12 and older with fibrodysplasia ossificans progressiva. The pivotal study had asked a narrower question. On the registry, the primary outcome of NCT05090891 is the occurrence of new heterotopic ossification lesions from baseline at week 24, assessed by low-dose whole-body CT excluding the head. That count, 1 of 32 patients (3.1%) on zilurgisertib versus 5 of 31 (16.7%) on placebo, an 81% reduction, produced p=0.0986 and did not reach conventional statistical significance. Months earlier, Mirum's own filing told investors the application was supported by secondary endpoints and that the pivotal study had not reached statistical significance on the primary. The company warned the FDA might find the data insufficient. The agency approved the drug anyway.
An ultra-rare disease and a 63-patient trial
FOP is caused by a mutation in activin A receptor-type 1, which controls new bone growth. Muscle, tendons, and ligaments gradually turn into bone, causing limited movement, deformities, severe disability, and early death. The companies say it affects approximately 300 people in the United States and 900 worldwide. A 63-patient randomized trial in that population cannot be powered the way a common-disease study is powered. The p-value on lesion occurrence did not clear 0.05. PROGRESS Cohort 1 assigned 32 patients to zilurgisertib 100 mg once daily and 31 to placebo for 24 weeks, then a 292-week single-arm open-label extension. Incyte sponsors the study, which is still listed as recruiting because later cohorts remain open. Zilurgisertib is an oral ALK2 inhibitor licensed to Mirum. The starting dose is 100 mg once daily, with or without food.
Volume moved. Lesion count did not clear.
The FDA says efficacy was based on the change from baseline in volume of total new heterotopic ossification versus placebo during the double-blind period, assessed by whole-body CT. At week 24, the Atebrioz group had an average 3.2 cm3 decrease. Placebo had a 24.6 cm3 increase. The company June release reported a mean change in total lesion volume of -3.24 cm3 (SD 19.86) versus +24.64 cm3 (SD 51.94), nominal p=0.004, and a mean new lesion volume of 0.003 cm3 (SD 0.02) versus 6.57 cm3 (SD 20.70), which the company called a greater than 99% reduction, nominal p<0.0001. Those standard deviations matter. The placebo means sit inside a very wide distribution. No median was reported.
The company approval release redescribes the approved measure as total new heterotopic ossification lesion volume, which includes expansion of baseline lesion burden as well as any new discrete bone that developed during the 24-week double-blind period. That is a different question from whether a new lesion appeared. The registry listed volume among secondary outcomes. The documents do not show the FDA secretly rewriting the primary. They show the agency stating a volume basis after the registered occurrence primary did not clear p<0.05. Annualized new flares were 2.34 (SD 6.06) versus 4.55 (SD 7.71). No p-value was given for flares. In the open-label extension to week 48, the company says no new lesions were observed among 61 patients who continued zilurgisertib or who crossed over. That extension is uncontrolled. No adverse events led to discontinuation or dose reduction in the 24-week period. The FDA notes the drug can cause fetal harm based on animal studies. Designations were fast track, priority review, and orphan drug.
A daily pill from age 12 is not the same dataset as a monthly infusion
Ipsen's Sohonos (palovarotene), approved in 2023, was the first FOP drug and works through the retinoic acid receptor rather than ALK2. Pasatru (garetosmab-grts), from Regeneron, was the second. The FDA approved it for adults only, to reduce new heterotopic ossification and clinician-assessed flare-ups. It is an antibody infused at 10 mg/kg over 60 minutes once every four weeks. Its trial enrolled 63 adults for 56 weeks, and the primary endpoint was the number of new lesions by week 56: 2 among 23 patients on 10 mg/kg, 1 among 19 on 3 mg/kg, and 19 among 21 on placebo. Flare-ups over 56 weeks were 9, 53, and 66 in those same groups. Pasatru carried breakthrough therapy among its designations. Atebrioz's FDA page does not.
Atebrioz is a once-daily oral tablet from age 12, approved on a 24-week volume measure after a lesion-count primary missed significance. Pasatru is a monthly IV antibody restricted to adults that hit its lesion-count primary over 56 weeks. Those are different endpoints, different durations, different ages, and different routes. Robert Pignolo, the Mayo Clinic lead investigator, said another option is meaningful in a progressive disease like FOP, particularly for adolescents earlier in its course.
Price, younger cohorts, and a voucher that went to Incyte
No list price has been disclosed. Mirum expects U.S. availability in October through Mirum Access Plus, with eligible patients paying as little as $0 per month, an access figure rather than a list price. PROGRESS Cohort 2, ages 6 to under 12, has completed enrollment. Cohort 3, ages 2 to under 12, is enrolling. An EU application is under review at the European Medicines Agency. The 292-week extension remains uncontrolled. The FDA issued a rare pediatric disease priority review voucher to Incyte, not to the company that will sell the drug. In April 2026 Mirum paid Incyte $16.0 million upfront for commercialization rights, with mid-to-high single digit royalties, regulatory milestones up to $48.0 million, and commercial milestones up to $15.0 million. The filing's pre-approval warning is now a matter of record: secondary endpoints carried the application, the primary missed, and the FDA still signed.