LAUSR

Autosomal dominant cerebellar ataxias have a marked heterogeneous genetic background, with mutations in 34 genes identified so far. This large amount of implicated genes accounts for heterogeneous clinical presentations, making genotype-phenotype correlations a major challenge in the field. While polyglutamine ataxias, linked to CAG repeat expansions in genes such as ATXN1, ATXN2, ATXN3, ATXN7, CACNA1A and TBP, have been extensively characterized in large cohorts, there is a need for comprehensive assessment of frequency and phenotype of more 'conventional' ataxias. After exclusion of CAG/polyglutamine expansions in spinocerebellar ataxia genes in 412 index cases with dominantly inherited cerebellar ataxias, we aimed to establish the relative frequencies of mutations in other genes, with an approach combining panel sequencing and TaqMan® polymerase chain reaction assay. We found relevant genetic variants in 59 patients (14.3%). The most frequently mutated were channel genes [CACNA1A (n = 16), KCND3 (n = 4), KCNC3 (n = 2) and KCNA1 (n = 2)]. Deletions in ITPR1 (n = 11) were followed by biallelic variants in SPG7 (n = 9). Variants in AFG3L2 (n = 7) came next in frequency, and variants were rarely found in STBN2 (n = 2), ELOVL5, FGF14, STUB1 and TTBK2 (n = 1 each). Interestingly, possible risk factor variants were detected in SPG7 and POLG. Clinical comparisons showed that ataxias due to channelopathies had a significantly earlier age at onset with an average of 24.6 years, versus 40.9 years for polyglutamine expansion spinocerebellar ataxias and 37.8 years for SPG7-related forms (P = 0.001). In contrast, disease duration was significantly longer in the former (20.5 years versus 9.3 and 13.7, P=0.001), though for similar functional stages, indicating slower progression of the disease. Of interest, intellectual deficiency was more frequent in channel spinocerebellar ataxias, while cognitive impairment in adulthood was similar among the three groups. Similar differences were found among a single gene group, comparing 23 patients with CACNA1A expansions (spinocerebellar ataxia 6) to 22 patients with CACNA1A point mutations, which had lower average age at onset (25.2 versus 47.3 years) with longer disease duration (18.7 versus 10.9), but lower severity indexes (0.39 versus 0.44), indicating slower progression of the disease. In conclusion, we identified relevant genetic variations in up to 15% of cases after exclusion of polyglutamine expansion spinocerebellar ataxias, and confirmed CACNA1A and SPG7 as major ataxia genes. We could delineate firm genotype-phenotype correlations that are important for genetic counselling and of possible prognostic value.