Re-weighting MDS-UPDRS Part II Items for Optimal Sensitivity to Parkinson’s Disease Progression Using Parkinson’s Progression Markers Initiative Natural History Data
Samuel Dickson1, Basia Rogula2, Jordan Dubow3, Nick Kozauer4, Lauren Powell2, Michele Potashman4, Kim Crimin4, Patrick O'Keefe1, Ellen Korol2, Madeline Crabtree2, Fernanda Nagase2, Vladimir Coric4, Suzanne Hendrix1, Gilbert L'Italien4
1Pentara Corporation, 2Broadstreet HEOR, 3Clintrex, 4Biohaven Pharmaceuticals, Inc.
Objective:
This work examines how composite scores (CS) of re-weighted combinations of the MDS-UPDRS PART II items can better detect meaningful changes and increase success for efficacious disease-modifying treatments.
Background:
Parkinson’s disease (PD) clinical trials require careful planning to progress beyond Phase 2. Many in the PD community recognize the need for improvements in communication, education, funding, recruitment, and compliance, but the need to measure disease with optimized outcome is underappreciated.
Design/Methods:
This study analyzed Parkinson’s Progression Markers Initiative data in subjects with confirmed PD. Three cohorts were defined based on use of levodopa or dopamine agonists (lev/DA) and presence of motor complications: untreated, early-lev/DA (motor complications absent), and later-lev/DA with motor complications (≥25% time of waking day in OFF-state and/or dyskinesia assessed on MDS-UPDRS). Sensitivity of individual items to disease progression was assessed using partial least square regression. Selected items were weighted using model coefficients and summed to create the CS. CS responsiveness to change was assessed using a 2-year mean to standard deviation ratio (MSDR) for treated and 1-year for untreated (due to data limitations).
Results:
CS were generated for untreated (n=428), early-lev/DA (n=424), and later-lev/DA (n=536) cohorts separately. The three most responsive items (with their combined weights) were: turning in bed, getting out of bed/car/chair, tremor (45%); turning in bed, getting out of bed, and speech (56%); and turning in bed, speech, and handwriting (53%) respectively. The MSDRs increased from 0.5431 to 0.5647, 0.4265 to 0.4994, and 0.3128 to 0.3849, respectively. Corresponding sample size decreases were ~7%, 27%, and 34%, reflecting powering improvements of ~3%, 11% and 13%.
Conclusions:
Endpoints derived from the three CS presented can measure clinically meaningful progression with greater sensitivity compared to existing tools.