Predicting disease progression in progressive cerebellar ataxias: integrating clinical, imaging, fluid, genetic, and digital biomarkers

Abstract

Progressive cerebellar ataxias, including the polyglutamine spinocerebellar ataxias (SCA1, SCA2, SCA3, and SCA6), Friedreich ataxia, and the autosomal recessive and sporadic forms, are rare, clinically heterogeneous neurodegenerative disorders. Disease-modifying therapies, notably antisense oligonucleotides and the approved compound omaveloxolone, are now entering or reaching the clinic, which makes reliable prediction of disease progression a central obstacle to trial success. Rarity, slow and variable decline, and a quantifiable premanifest window together demand sensitive, validated tools to forecast who will progress, how fast, and when symptoms begin. This review synthesizes evidence across five complementary biomarker domains, clinical, imaging, fluid, genetic, and digital, and judges their readiness for forecasting onset and progression. Clinical scales such as SARA capture genotype-specific trajectories but lose sensitivity at the extremes of the disease course. Volumetric, microstructural, and spectroscopic MRI and blood neurofilament light chain change before ataxia onset and predict subsequent decline, whereas repeat length and genetic modifiers set prior risk. Wearable-sensor gait and balance metrics detect change earlier than clinical scales and sharply reduce required sample sizes. We argue that no single modality satisfies every context of use, and that stage-specific, multimodal composites, integrated through harmonized international cohorts and machine learning, offer the most credible path to prognostic enrichment and to shorter, adequately powered, and preventive trials.