OBJECTIVES: Randomized controlled trials of gene therapies in Duchenne muscular dystrophy (DMD) have studied effects on ambulatory function limited to one year using the North Star Ambulatory Assessment (NSAA). However, the projected lifetime effects of these therapies will need to be considered in health economic evaluations. Drawing on natural history data, we developed a model-based synthesis for projecting plausible effects on longer-term disease milestones, including mortality, based on 1-year outcomes from clinical trials.

METHODS: A quantitative natural history (NH) model of progressive milestones was developed to link 1-year changes in NSAA total score to mortality though intermediate associations with loss of ambulation (LoA) and need for mechanical ventilation. Median time between milestones were estimated from a targeted literature review and a Kaplan-Meier analysis of the Cooperative International Neuromuscular Research Group Duchenne Natural History Study (CINRG) database.

RESULTS: The NH model estimated the median age of LoA at 11.0 years and median time from LoA to first use of mechanical ventilation at 9.4 years, consistent with published estimates. A literature review yielded a median time of 7.0 years from ventilation to death. The model-based median age at death was 27.4 years, consistent with published medians of 27.0-36.2 years. Two scenarios were developed for projecting plausible lifetime treatment effects: 1) a ‘fixed-shift’ effect, where age-at-event for all milestones is delayed by the same number of years, and 2) a ‘persistent slowing’ effect in which milestones are delayed based on a constant deceleration of disease progression.

CONCLUSIONS: Lifetime effects of novel therapies for DMD will ultimately be determined based on real-world evidence over several decades. In the near-term, the NH model presented here, which incorporates NH data, published evidence and plausible assumptions, can complement and supplement the lifetime projections used in health economic evaluations.