OBJECTIVES: Fibrodysplasia ossificans progressiva (FOP) is an ultra-rare genetic disorder characterised by irreversible heterotopic ossification (HO) in soft and connective tissues. There is no definitive standard of care (SoC) for FOP. Palovarotene has been reported to reduce new HO in patients with FOP (MOVE trial; NCT03312634) versus a natural history study (NHS; NCT02322255). We describe conceptualisation of the first long-term cost-effectiveness model (CEM) comparing palovarotene to SoC.

METHODS: Model conceptualisation relied on FOP literature and MOVE/NHS data, focusing on links between HO and clinically and economically meaningful outcomes such as quality of life (QoL) and costs. The model concept was validated in structured interviews with clinical and health economic experts.

RESULTS: The FOP population eligible for palovarotene is heterogenous, with age influencing HO volume and flare-up rates. To capture the value of palovarotene, multiple concepts to describe the underlying FOP disease process were explored. The most supported structure is a Markov model consisting of states with increasing whole body HO volume. Existing literature can be used to calculate progression through HO states, noting that the transition to higher volume HO states slows down after 25 years of age. There are no studies linking HO volume to QoL and costs. However, increase in HO volume is associated with increases in mobility restrictions as measured by cumulative analogue joint involvement scale (CAJIS) scores, which can be linked to QoL and costs. As the CAJIS-HO relationship is not perfect, explicit quantification of this relationship was avoided, with each HO state associated with a distribution of patients between CAJIS categories.

CONCLUSIONS: By preventing new HO, palovarotene has the potential to positively impact patient outcomes in FOP. We present the methodology to link HO volume to clinically meaningful outcomes for a CEM. Model conceptualisation in ultra-rare diseases with limited data relies heavily on expert validation.