GM1 Ganglioside Therapy in Acute Ischemic Stroke

The statistically significant difference between GMl-treated and placebo-treated patients that we found did not refer to the mean values of neurologic scores, which were similar among the groups, as clearly depicted in our Tables 4 and 5, but to the degree of neurologic improvement, as obtained by subtracting the Canadian Neurological Scale score at entry from the scores on days 1-15, 21, and 120, respectively. However, it is not clear what these differences in degree of neurologic improvement could mean in terms of functional recovery. We completely agree with Hoffbrand and colleagues that the success of a stroke treatment depends more on whether the patient is able to return home to useful function than on the activities of daily living scales are demanding and could not be used for stroke patients in the acute phase because performance is often limited by factors that are not dependent on their neurologic deficit. Furthermore, taking into account functional scales for long-term follow-up, we cannot know how much of their recovery is due to rehabilitation. In this context, it is interesting, rather than surprising, that we found a good correlation between mean neurologic global scores and the disability index, neither of which was statistically significant, confirming the functional value of the Canadian Neurological Scale score. Nevertheless, we think that no available parameters for evaluation of functional outcome in stroke patients, whether based on neurologic or disability scales, are sufficient. At present, one way to overcome this handicap could be to define more homogeneous and comparable subgroups of patients by stratifying them according to prognostic parameters such as age, severity of neurologic status at admission, and interval time between onset of symptoms and treatment. We think that only multicenter studies, capable of larger enrollment of a continuous series of patients, could help to reach this goal. Our study only partially fulfilled these criteria; for instance, the time interval of 12 hours between stroke onset and treatment, although shorter than that of previous studies, was well over the socalled "therapeutic window." Moreover, our data are hardly comparable with those of previous studies due to significant differences in the study design, such as treatment not started in the acute phase, too small a sample size, and different outcome measures. We therefore believe that the results we obtained need to be further tested in a larger population of 800 patients, stratified according to sex, age, and Canadian Neurological Scale score at entry, and who are treated within 5 hours of stroke and for a longer period (21 days). Our Early Stroke Treatment trial will have the enrollment of the established number of patients, and we think that at the end of 1990 we will be able to provide our definitive results.