Multimodal Characterization of Glymphatic-Related Magnetic Resonance Imaging Markers in Huntington's Disease: A Multi-Cohort Retrospective Study.

Abstract

To characterize magnetic resonance imaging (MRI)-based glymphatic surrogates in Huntington's disease (HD) using MRI measures of perivascular diffusivity and structural perivascular alterations across multiple large cohorts. We analyzed 2,731 MRI sessions from 880 participants across 3 large retrospective HD cohorts. HD gene carriers were grouped by the HD Integrated Staging System into stage 0 to 3. We assessed the diffusion tensor image analysis along the perivascular space (DTI-ALPS) index and MRI-visible enlarged perivascular spaces (EPVS). Linear mixed-effects models examined group differences and associations with motor, cognitive, and functional measures. Likelihood ratio tests evaluated their added explanatory value beyond established imaging, genetic, and demographic variables. Left ALPS indices were lower at more advanced disease stages (stage 1: -3.71%, Cohen's d: -0.25; stage 2: -4.33%, Cohen's d: -0.31; Stage 3: -6.43%, Cohen's d: -0.46; all p < 0.05) compared with controls, independent of demographic and clinical variables. EPVS burden showed region-specific patterns, with subcortical EPVS volume fraction increased in HD (stage 1-3: +49.9% to +63.7%, Cohen's d: 0.47 to 0.55, p < 0.0001), whereas white matter EPVS volume fraction decreased at Stage 3 (-13.8%, Cohen's d: -0.21, p < 0.05). Lower ALPS indices correlated with higher motor impairment (p < 0.0001) and reduced functional capacity (p < 0.01). Both ALPS and EPVS improved prediction of composite HD progression beyond CAG-Age Product Score and striatal volume. Multimodal MRI revealed stage-dependent alterations in MRI-based glymphatic surrogates with low to moderate effect sizes. These findings suggest that DTI-ALPS and EPVS metrics offer complementary, noninvasive indicators of HD pathology derivable from routine imaging protocols. ANN NEUROL 2026.