OBJECTIVES: Two different cost-effectiveness models for voretigene neparvovec (Luxturna^®) for RPE65 – Mediated Inherited Retinal Dystrophy have been considered in independent evaluations by the Institute for Clinical and Economic Review (ICER) and by National Institute for Health and Care Excellence (NICE). The models differed in structure, parameters, assumptions, and results. We sought to compare these models, analyze the use of alternative models for each evaluation and check the impact of key inputs and assumptions of the model results.

METHODS: We replicated both models with the use of MS Excel and VBA for applications. The replicated models were validated by comparing their results with published information. The ICER model was further adapted to the UK and the model assessed by NICE was run from the US perspective. Next, several model inputs and assumptions were varied to test their impact on the model results.

RESULTS: The outcomes of replicated models were well aligned with published data. It was proved that the use of an alternative model from perspective of a given country could change the results of economic evaluation, however, the conclusions regarding the cost-effectiveness of voretigene neparvovec remind the same. The assumptions regarding discounting, methods of survival extrapolation, duration of treatment effect, estimation of visual acuity/visual field, utility sources and methods as well as rate of residual treatment effect had significant impact on the results of both models.

CONCLUSIONS: The way of modelling can differ between countries and impact the economic evaluation. This is of particular importance for gene therapies.