Riluzole in patients with hereditary cerebellar ataxia

Abstract

We read with interest the Article by Silvia Romano and colleagues, describing a novel and potentially important application of riluzole in patients with spinocerebellar ataxia or Friedreich’s ataxia. The investigators noted that riluzole significantly improved ataxia at 3 months and 12 months. This improvement was assessed as at least a one-point reduction in the Scale for Assessment and Rating of Ataxia (SARA), a rating scale with summed scores of up to a maximum of 40 points in the domains of gait, kinetics, and speech. A posthoc analysis showed improvement in the patients’ quality of life (on the SF-36 scale). In absence of an ideal biomarker for ataxia, Romano and colleagues predefined a one-point reduction in the SARA score as a cutoff for improvement in ataxia. The chosen cutoff suggests that the smallest difference in any of the three SARA domains would represent an improvement in ataxia. We challenge this assumption. First, any selected cutoff should exceed the effects of interobserver and intraobserver variability. Because of the inclusion of heterogeneous patients in Romano and colleagues’ trial, a high variation in SARA scores (2·0–29·5) would induce an increase in the margin of error. Thus, although we noted high interobserver and intraobserver agreements on SARA scores (intraclass correlation coefficient of 0·91–0·98 and 0·96–0·98, respectively), the investigators’ chosen cutoff is highly likely to refer to the margin of error. Second, as indicated by Durr, potentially confounding effects on SARA scores should be taken into account, especially for the comparison of two small and heterogeneous treatment groups. In previous reports, we have shown that SARA scores are not only aff ected by ataxia, but also by age of the child (in terms of observer variability and absolute scores), and by muscle force (in Friedreich’s ataxia). Although we are interested in the potentially positive eff ect of riluzole treatment, the interpretation of the SARA score outcomes are open to question. Until biomarkers for ataxia progression are available, consideration of the SARA construct is important to prevent overinterpretation. The Childhood Ataxia and Cerebellar Group of the European Pediatric Neurology Society (CACG-EPNS), is now doing a SARA validation study in children to provide increased uniformity for the interpretation of longitudinal ataxia trajectories, from childhood to adulthood.