A digital twin methodology using real patient data for sample size reduction in Alzheimer's disease randomized
Daniel Andrews1,2, Shirin Golchi3, D Louis Collins1,2,4
1Department of Biomedical Engineering, McGill University, Montreal, Canada.
Introduction:
Recruitment for Alzheimer's disease randomized controlled trials (RCTs) is difficult and expensive. To reduce RCT sample sizes, our Digital Twin Trial (DTT) methodology combines an interpretable cognitive decline prediction model with prediction-powered inference.
Methods:
For DTT participants, our model identifies similar individuals ("Digital Twins") from a retrospective database and uses their cognitive scores to predict decline. Predictions adjust observed scores, reducing variance within treatment groups. We simulated 18-month DTTs and standard RCTs using mixed effects models of decline in Alzheimer's Disease Neuroimaging Initiative subjects meeting lecanemab's Phase 3 inclusion criteria.
Results:
Predicted and observed change in Clinical Dementia Rating Sum-of-Boxes correlated at . DTTs required 1,855 subjects versus 2,170 for standard RCTs to detect a simulated 25% decline-slowing drug effect at 0.9 power. DTT Type 1 error was consistent with 0.05.
Discussion:
DTTs could reduce recruitment and cost burdens. Model interpretability could help clinicians trust individualized prognoses.
