OBJECTIVES: To determine whether significant differences in functional and quality of life (QoL) outcomes can be expected in a trial that showed substantial reductions in heterotopic ossification (HO) volume in patients with FOP, an ultra-rare genetic disease.

METHODS: A patient-level model was used to simulate patient histories up to the end of a 2-year, 2-arm trial, varying the effect of the intervention on extra bone growth as measured by HO volume. Outcomes of interest were the Cumulative Analogue Joint Involvement Scale, CAJIS; FOP-Physical Function Questionnaire, FOP-PFQ; and PROMIS. Baseline estimates with treatment (n=99) were based on the MOVE trial (NCT03312634). Baseline estimates without treatment (n=111) were based on the international FOP Natural History Study (NCT02322255). Sets of 1000 trial simulations were run, with varying levels of HO reduction and varying levels of random noise, while assuming linear relationships between HO volume and the outcome measures. Calculations assessed the likelihood of finding a relationship between HO volume and outcomes, and the likelihood of finding an intervention effect on outcomes.

RESULTS: There was a low probability of finding a significant effect of treatment on CAJIS, FOP-PFQ or PROMIS in the simulated 2-year trials, although the implicit relationship could be found cross-sectionally (pooling data from both groups). The low probability of finding a treatment effect was present even when treatment completely halted all new HO and no random noise was included.

CONCLUSIONS: A significant effect on HO was unlikely to result in significant effects on functional and QoL outcomes in the simulated trials, although linear relationships between HO volume and CAJIS, FOP-PFQ, and PROMIS were postulated. Decreased statistical power was a result of substantial HO volume heterogeneity, insufficient patient numbers, and large amounts of random noise. The lack of statistically significant results in the context of this simulation does not imply that there is no treatment effect.