Value of MRI Outcomes for Preventive and Early-Stage Trials in Spinocerebellar Ataxias 1 and 3

Objective: To examine the value of MRI outcomes as endpoints for preventive and early-stage trials of two polyglutamine spinocerebellar ataxias (SCAs). Methods: A cohort of 100 participants (23 SCA1, 63 SCA3, median Scale for the Assessment and Rating of Ataxia (SARA) score = 5, 42% preataxic, and 14 gene-negative controls) was scanned at 3T up to 6 times (median follow-up 3 years) in the READISCA study. Structural, microstructural, and neurochemical outcomes were assessed for sensitivity to change. Associations between change in MR and clinical and patient-reported outcomes were evaluated. Sample sizes were estimated for preventive trials (at preataxic stage) and early-stage trials using the most sensitive outcomes. Results: Infratentorial volumes, middle cerebellar peduncle (MCP) diffusivities and select neurochemical outcomes were more sensitive to change than SARA in both SCAs with moderate-to-high effect sizes (Cohen's d ≥ 0.5). The pons volume was the most sensitive outcome at both preataxic (d = 1.09) and ataxic (d = 1.48) stages. The most sensitive MR measures overlapped between genotypes, except that SCA1 showed faster progression in the cerebellum and SCA3 in the pons. Changes in MCP diffusivities and pontine neurochemical measures were associated with changes in SARA and FARS-ADL (|r| = 0.28–0.47). Interpretation: The estimated sample sizes to detect a 50% reduction in progression with 80% power using the pons volume as primary outcome measure indicate the feasibility of preventive trials (n = 73 per arm in a 1-year trial) in common SCAs and predict a 6-fold reduction in required sample sizes relative to SARA in early interventional trials. These findings support the use of MRI endpoints in early-stage SCA trials.