Sample Size and Power Calculations Based on Endpoint Responsiveness: Applications to Trials of Disease-Modifying Therapies in Parkinson's Disease

In Parkinson's disease (PD), early phase clinical trials of disease-modifying agents generally use the MDS-UPDRS Part III as the primary outcome measure of motor function, which typically requires a large sample and/or a long (12-18 months) treatment duration, even if a substantial treatment effect is assumed. This is due to the rather slow rate of early PD progression relative to the variability of the MDS-UPDRS Part III assessments. Using placebo-arm data from four recent large Phase 2 trials of early-stage PD (SPARK, PASADENA, PADOVA and LUMA), the sample size needed to detect a range of assumed treatment effects on MDS-UPDRS Part III progression was estimated. The trial sample size estimates obtained for MDS-UPDRS Part III and for smartphone-based sensor-derived motor functional composite (MFC) endpoints were compared using an approach based on relative responsiveness to disease progression. The calculations were also implemented in an interactive, publicly accessible application. Powering a one-year trial using the MFC1 endpoint, derived from three active assessments (6-minute walk, U-turn and Hand Pronation-Supination), or the MFC2, which adds 60-Second Typing, was projected to reduce the required sample size by more than half relative to the MDS-UPDRS Part III, assuming equivalent proportional treatment effects.