Objective: The aim was to establish a data-driven model of dynamic biomarkers in SCA3 from the early pre-ataxic to the late ataxic disease stage.

Background: Data-driven models provide a unique picture of disease progression by visualizing the temporal order and evolution of distinct disease biomarkers and symptoms. In particular, they allow a more precise staging, monitoring and prognosis in the clinical practice and have valuable impact on patient stratification and design of interventional trials.

Method: The European Spinocerebellar Type 3/Machado-Joseph Disease Initiative (ESMI) recruited > 300 SCA3 mutation carriers in a longitudinal cohort. The standardized clinical assessment, MRI and biosampling allowed the development and validation of disease markers. The most promising state and progression markers were further studied. We consider the pre-ataxia and the ataxia stage as graded manifestation of one disease process. Consequently the identified parameters were z-transformed and plotted against a consistent time scale of disease duration for ataxic and estimated time from ataxia onset for pre-ataxic SCA3 mutation carriers. We performed a linear regression with backward selection to identify factors that covary with ataxia severity and progression in ataxic SCA3 mutation carriers.

Results: We developed a fine-grained multistage disease model of SCA3 that allowed to establish individual disease profiles from the earliest pre-ataxia to the late ataxic disease stage. NfL, ataxin-3, pons volume and total tau showed a continuous change across the clinical onset of disease, qualifying them in particular for interventional trials in the late pre-ataxic stage. Linear regression revealed that these parameters contribute to a model of ataxia severity with an increased proportion of explained variance in ataxia severity when including NfL, total-tau, ataxin-3 and pons volume in addition to CAG repeat length, sex and age (R² 0.61 vs. 0.47).

Conclusion: The approach of a data-driven model will pave the way towards individualized outcome parameters that are needed for successful trials and will have an immediate impact on current practice, allowing a differentiated diagnosis and more personalized delivery and monitoring of care. In particular in the late pre-ataxic stage of SCA3 the inclusion of pons volume and NfL in addition to CAG repeat length, age and sex increases the accuracy to stratify SCA3 mutation carrier.

« Back to MDS Virtual Congress 2021

MDS Abstracts - https://www.mdsabstracts.org/abstract/data-driven-model-of-dynamic-biomarkers-in-sca3-from-early-pre-ataxic-to-late-ataxic-disease-stages/