Hybrid controlled clinical trials using concurrent registries in Amyotrophic Lateral Sclerosis: A feasibility study.

Hybrid designs with both randomised arms and an external control cohort preserve key features of randomisation and utilise external information to augment clinical trials. In this study, we propose to leverage high-quality, patient-level concurrent registries to enhance clinical trials and illustrate the impact on trial design for amyotrophic lateral sclerosis (ALS). The proposed methodology was evaluated in a randomised, placebo-controlled clinical trial. We used patient-level information from a well-defined, population-based registry, that was running parallel to the randomised clinical trial, to identify concurrently non-participating, eligible patients who could be matched with trial participants, and integrate them into the statistical analysis. We assessed the impact of the addition of the external controls on the treatment effect estimate, precision, and time to reach a conclusion. During the runtime of the trial, a total of 1,141 registry patients were alive; 473 (41.5%) of them fulfilled the eligibility criteria and 133 (11.7%) were enrolled in the study. A matched control population could be identified among the non-participating patients. Augmenting the randomised controls with matched external controls could have avoided unnecessary randomisation of 17 patients (-12.8%) as well as reducing the study duration from 30.1 months to 22.6 months (-25.0%). Matching eligible external controls from a different calendar period led to bias in the treatment effect estimate. Hybrid trial designs utilising a concurrent registry with rigorous matching can minimise bias due to a mismatch in calendar time and differences in standard of care, and may accelerate the development of new treatments.