Superior cerebellar peduncle hyperintensity: a highly specific imaging marker for SCA27B

Abstract

Background: Spinocerebellar ataxia type 27B (SCA27B), resulting from a GAA-repeat expansion in the FGF14 gene, has recently been identified as a prevalent and likely underdiagnosed cause of late-onset cerebellar ataxia. Clinically, the disorder is characterized by a progressive cerebellar syndrome that may initially manifest with episodic vertigo, diplopia, and gait instability, often triggered by alcohol or physical exertion. Due to significant phenotypic overlap with other disorders such as multiple system atrophy, cerebellar type (MSA-C) and CANVAS, establishing a diagnosis can be challenging. Objective: This study aimed to evaluate the diagnostic specificity of superior cerebellar peduncle (SCP) hyperintensity in SCA27B compared to other genetic and neurodegenerative ataxias. Methods: A retrospective analysis was conducted on 127 subjects, comprising 106 patients with neurodegenerative forms of cerebellar ataxia and 21 age- and sex-matched healthy controls. The ataxia cohort included 14 patients with genetically confirmed SCA27B, 47 with other spinocerebellar ataxias (SCA3, SCA2, SCA6), 12 with MSA-C, 20 with CANVAS, and 13 with late-onset Friedreich ataxia (LOFA). All subjects underwent 3T brain MRI, and three-dimensional FLAIR sequences were independently reviewed by two blinded investigators. The imaging features assessed included bilateral SCP hyperintensity, middle cerebellar peduncle (MCP) hyperintensity, and the "hot cross bun" sign . Statistical comparisons utilized Pearson’s chi-square and Fisher-Freeman-Halton exact tests to confirm significance. Results: The SCA27B cohort had a mean age of 62 years and a mean disease duration of 10.3 years at the time of MRI. Results demonstrated that bilateral SCP hyperintensity was present in 71.4% (10/14) of SCA27B patients. In contrast, this feature was rare or absent in disease controls: it was observed in only 15% of CANVAS patients, 8.3% of SCA3 and MSA-C patients, and was completely absent in SCA2, SCA6, and LOFA cohorts. Consequently, SCP hyperintensity yielded a specificity of 93.8% for the diagnosis of SCA27B. Conversely, classical MSA-C markers, such as MCP hyperintensity and the hot cross bun sign, were common in the MSA-C and SCA2 groups but were entirely absent in all patients with SCA27B. Conclusion: This study demonstrates that bilateral and symmetric hyperintensity of the superior cerebellar peduncles represents a highly specific neuroradiological marker for SCA27B. Anatomically, selective involvement of this principal efferent pathway reflects disruption of cerebellar output circuits in FGF14-related disease. Furthermore, SCP hyperintensity was detected even in patients with mild clinical severity, suggesting it may serve as an early diagnostic indicator. Incorporating careful inspection of the SCPs using 3D FLAIR sequences into routine neuroimaging protocols can facilitate earlier identification of SCA27B and streamline targeted genetic workups for these patients.