MRI Biomarkers Move to the Center of Ataxia Trial Design
Hereditary ataxias have reached a point where clinical trial design is being forced to catch up with biology. In spinocerebellar ataxia and Friedreich ataxia, the disease burden is clear long before conventional trial readouts can reliably capture change, and the field now needs outcome measures that can track degeneration with more precision than bedside scales alone.
The Ataxia Global Initiative MRI Biomarkers Working Group has responded with a consensus statement that places quantitative MRI end-points squarely into the trial conversation for the most mature disease programs: SCA1, SCA2, SCA3 and Friedreich ataxia. The central argument is straightforward. When disorders are rare, progressive and slow-moving, traditional clinical outcomes can be too blunt to detect treatment effects over the time frames that are realistic for drug development.
The pressure comes from both ends of the trial spectrum. In symptomatic disease, investigators need an outcome that changes enough to show whether a candidate therapy is doing anything at all. In presymptomatic hereditary disease, the challenge is even sharper: individuals who carry pathogenic variants may not yet have a measurable clinical deficit, yet they are precisely the population in whom early intervention may offer the greatest biological leverage.
A single sentence captures the stakes. Clinical scales are useful, but they are not built to solve the measurement problem that rare ataxia trials now face.
Why Clinical Scales Alone Are Straining Trial Feasibility
The current gold-standard outcomes remain the Scale for the Assessment and Rating of Ataxia and the Friedreich Ataxia Rating Scale. They are familiar, clinically meaningful and widely used, but their limitations become obvious in a therapeutic area where disease progression is gradual and cohort sizes are small. Natural history data indicate that detecting a 50% slowing of disease over one year would require sample sizes of 71 to 274 participants per arm, depending on design assumptions.
Those numbers are difficult for any rare-disease program, and they are especially punishing in a condition where eligible participants may be geographically dispersed and genotype-specific cohorts are small. Slow progression also compresses signal into a narrow range, increasing the risk that a real treatment effect is drowned out by noise, site variability or ceiling and floor effects.
This is where quantitative imaging enters the frame. Structural MRI, diffusion MRI, magnetic resonance spectroscopy and susceptibility MRI offer objective readouts of the cerebellum, brainstem, spinal cord and white matter pathways that are injured in these diseases. They measure anatomy and tissue integrity directly, rather than relying solely on how those injuries are translated into overt motor disability.
Imaging Reads the Disease Biology Before Disability Fully Declares Itself
In hereditary ataxias, neurodegeneration is not confined to the cerebellum. The disorders reviewed here can also affect the brainstem and spinal cord, and some carry systemic manifestations such as cardiac involvement. MRI biomarkers are attractive because they can follow that distributed pathology in vivo, including structural volume loss, microstructural change and altered tissue composition.
The biologic appeal is obvious. If a therapy is working by slowing neuronal loss, stabilizing white matter tracts or limiting downstream tissue injury, an imaging biomarker may change before a clinical scale does. That makes MRI especially useful in early-phase studies, in enrichment strategies for participant selection, and in presymptomatic trials where symptoms have not yet emerged.
The working group did not treat imaging as a replacement for clinical outcomes. Instead, the statement frames MRI as a candidate complement to clinical scales, one that must still earn its place by showing that it is reproducible, sensitive to change and meaningfully linked to outcomes patients and clinicians care about.
The Consensus Process Focused on Effect Size and Trial Feasibility
The MRI Biomarkers Working Group brought together neurologists, neuroradiologists, physicists, biologists and clinical trial specialists from academia and industry. Its mandate was to identify harmonized procedures for neuroimaging in cerebellar ataxia research and to develop evidence-based recommendations for end-points that can realistically be used in multicenter trials.
The review process was structured around a practical set of criteria: effect size, technical variability, clinical meaningfulness, sensitivity to change over one to two years and feasibility in a multicenter setting. PubMed searches were used to identify quantitative studies in SCA1, SCA2, SCA3 and Friedreich ataxia, focusing on structural MRI, diffusion MRI, magnetic resonance spectroscopy and quantitative susceptibility approaches in brain or spinal cord tissue.
Only studies with inferential statistics and objective quantitative measures were included. Case reports, case series and clinician-rated imaging impressions were excluded. Cross-sectional and longitudinal effect sizes carried the greatest weight in the recommendation process, which is appropriate for a field trying to decide which biomarkers are strong enough to support trial decisions rather than simply describe disease anatomy.
Why the Recommendation Matters for Rare-Disease Development
The implications for trial design are immediate. A biomarker that can detect change with fewer participants and over shorter observation periods can lower the barrier to entry for studies that might otherwise never start. That is not a theoretical gain; it changes whether a rare-disease program is feasible at all.
MRI also helps address a different problem: the possibility of enrolling people before symptoms appear. In genetic ataxias, presymptomatic carriers create a new category of trial participant. Their inclusion demands markers that can identify disease activity before overt functional loss becomes measurable. Imaging is one of the few modalities that can plausibly serve that role across a range of disease stages.
The statement also draws a hard line on what imaging alone cannot yet do. Clinical relevance still has to be demonstrated against patient-reported outcomes, clinician-reported assessments and longitudinal progression on accepted clinical scales. A scan result is not enough by itself; it must map onto deterioration, stability or response in a way that regulators and trialists can trust.
What the Field Still Needs Before MRI Can Serve as a Primary Endpoint
Despite the progress summarized in the consensus statement, several gaps remain. The most important is the need to prove that baseline MRI values, or short-interval changes in those values, predict later clinical change. Correlation with disability measures is useful, but prediction of future progression is the more demanding test.
Multicenter feasibility is another unresolved issue. Imaging biomarkers can only become practical trial end-points if protocols are standardized across sites, acquisition is reproducible and technical variability stays low enough that biological change is not masked. The working group’s call for harmonized procedures reflects that need.
The disease-specific recommendations also matter because one-size-fits-all biomarker thinking does not fit hereditary ataxia. SCA1, SCA2, SCA3 and Friedreich ataxia overlap clinically, but their neuroanatomy and underlying biology are not identical. Trial end-points have to respect those differences if they are going to be credible.
The next step is not to assume MRI is ready for every trial question. It is to test, disease by disease, whether these imaging readouts can do the hardest job in rare-neurology drug development: detect true therapeutic change early enough to matter, with enough precision to guide decisions before time and sample size run out.