Objective: This thesis aims to identify and address the main bottlenecks of the clinical drug development for Amyotrophic Lateral Sclerosis (ALS).
Methods: Part 1 of this thesis aims at exploring and addressing the sources of clinical heterogeneity in trial populations and treatment responses. Part 2 evaluates the performance of classical efficacy outcomes as compared to innovative measures of disease progression. In addition, a comparison of different statistical strategies is given that optimizes the use of information by combining multiple efficacy outcomes. Part 3 evaluates how incorporating interim analyses or data-driven assumptions may improve the use of time and resources. Finally, a data-driven and evidence-based integration of Parts I-III is illustrated for a real-world setting.
Results: Large efficiency gains can be obtained by replacing eligibility criteria for prognostic survival models, by implementing new efficacy outcomes such as plasma biomarkers or remote healthcare technology, by innovating overall design settings with event-based and group-sequential monitoring schemes and by optimizing analytical strategies with the use of a joint modeling framework. Compared to classical trial design, optimizing clinical trial design could lead to considerable reductions in the size and duration of clinical trials, which lowers the burden for patients and optimizes the use of the available resources.
Conclusions: Clinical trials for ALS are complicated by the unpredictable and variable course of the disease, the insensitivity of clinical outcome measures and by suboptimal design assumptions. This thesis expanded the ALS clinical trial toolbox and takes a first step towards data-driven, evidence-based guidance for future settings.