Presentation (CTI)

OBJECTIVES: To assess the feasibility and value of using a patient-level simulation to support cost-effectiveness modelling of B-VEC (topical gene therapy) in Dystrophic Epidermolysis Bullosa (DEB), in the absence of patient-level trial data.
METHODS: The cost-effectiveness model for B-VEC is structured around health states defined by the percentage of body surface area (BSA) with recurrent or chronic open wounds. However, the pivotal Phase 1/2 (GEM-1) and Phase 3 (GEM-3) trials for B-VEC undertook randomization at wound-level, not patient-level, and did not report the number or size of wounds per patient or percentage of BSA covered with wounds. Therefore, data on the effect of B-VEC on percentage of BSA with recurrent or chronic wounds are not available. To overcome this limitation, a patient-level simulation was developed using wound-level data from GEM-1, GEM-3, their open-label extension, and published literature. Simulated patients were tracked in parallel B-VEC and placebo arms over a 3-year time horizon. The simulation captured wound closure dynamics, duration of closure, reopening size, and risk of chronicity to estimate BSA involvement over time. These outcomes were used to estimate transition probabilities between health states, as well as B-VEC vial consumption.
RESULTS: The simulation successfully generated plausible patient trajectories consistent with clinical expert expectations, although there is no long-term patient-level evidence to validate outcomes against. All simulated patients receiving B-VEC reached a very mild disease state (<2.1% BSA with open wounds) within three years, with declining treatment use over time. Placebo patients remained predominantly in their initial health states.
CONCLUSIONS: Patient-level simulation offers a viable solution for modelling treatment impact and resource use in the absence of patient-level trial data. This approach may be particularly valuable in rare diseases like DEB, where traditional data sources are limited.