We read with much interest the article by Schulte et al1 because we reported the same type of trial in 8 patients with Machado-Joseph disease (MJD) in 1995.2 Their results were negative, whereas ours and those of 3 other studies3- 5 were positive. The authors point out 3 major problems that may explain this discrepancy; the first is the small subject number (at most, 8 patients) in the previous studies. Second, the treatment period with trimethoprim-sulfamethoxazole was short, with a maximal duration of 4 weeks. Finally, quantitative methods were missing in the previous studies except for ours2 and the one by Mello and Abbott.3 It goes without saying that the first 2 issues are important in clinical trials. However, we believe that the third problem, the lack of quantitative methods to detect the changes in ataxia, is the most important. In our previous study,2 we adopted timed tests to detect minor changes in neurologic deficits because we thought that mild but significant improvements would be missed with semiquantitative methods such as grades 1 (mildly impaired), 2 (moderately impaired), and 3 (markedly impaired). In this sense, it was excellent that Schulte and colleagues selected static posturography as an assessment of stance and used the motor performance test of Schoppe6 as an assessment of dexterity. If we had used the ataxia score of Klockgether et al7 as our clinical rating score, our results might have been negative rather than positive. It is our opinion that a variety of quantitative methods should be used in clinical trials for spinocerebellar ataxias until a highly reproducible and reliable computer-assisted device appears in the near future.
Sakai T, Miyoshi K. The Use of Quantitative Methods in Clinical Trials for Spinocerebellar Ataxia. Arch Neurol. 2002;59(6):1044-1045. doi: