Spinocerebellar ataxia (SCA) has several genetically defined subtypes. A recent multicenter study investigated the clinical symptoms associated with these subtypes, to identify patterns of clinical features that might distinguish between the types of SCA.
Maschke and colleagues recruited 127 patients with SCA types 1–8, and carried out a neurological assessment of their symptoms using a worksheet comprising 33 symptoms and signs. They then calculated the probability of each trait being present in a particular SCA type, and used this to rate the predictive value of individual clinical features for each type of SCA. These were combined to form a Bayesian classifier, which was later simplified to form an algorithm that separated SCA types into three groups—those with upper motor neuron involvement, those with neuropathy, and those with a predominantly cerebellar syndrome.
Overall, the Bayesian classifier was able to correctly predict the type of SCA in 78% of patients. The most accurate predictions were achieved for SCA7 (sensitivity 100%, specificity 98%), SCA8 (sensitivity 91%, specificity 97%), and SCA5 (sensitivity 88%, specificity 95%).
Concluding that many subtypes of SCA do exhibit characteristic clinical features, the authors note that, for practical reasons, an algorithm might be more easily adopted by medical centers than a Bayesian classifier. In addition, they state that algorithms based on detailed clinical data could help prioritize genetic testing for specific types of SCA, thus helping to reduce the costs of such tests.
Rights and permissions
About this article
Cite this article
Kyme, C. Clinical profiling of spinocerebellar ataxia genetic subtypes. Nat Rev Neurol 1, 8 (2005). https://doi.org/10.1038/ncpneuro0007
Issue Date:
DOI: https://doi.org/10.1038/ncpneuro0007