Riluzole in patients with hereditary cerebellar ataxia. Authors' reply

Overall, we endorse their suggestions, especially the usefulness of a delayed start design and the need for open-label phases in which all patients would receive the active drug for longer periods after a positive shortduration randomised trial. Because of regulatory and economic hurdles, open-label phases were difficult to plan in our populations of patients: only three patients remained on riluzole therapy from the first trial3 (one died from multisystem atrophy, but two patients are doing well, without adverse effects or progression, after 5 years of follow-up).