Fig. 5

Time-to-event analysis of period from diagnosis to death in patients treated with miglustat for > 2 years (n = 17) versus untreated patients and those who received miglustat for < 2 years (n = 26). Patients who discontinued miglustat after < 2 years of treatment because of neurological worsening were excluded (n = 4). The Kaplan Meier curves for this analysis were truncated when approximately 10% of patients were still under observation in each group, due to low relevance of graphical representation based on limited patient numbers beyond this time point. For patient 6 who was diagnosed in early infancy, time-to-event analysis began from start of miglustat treatment. Mean clinical scores at diagnosis were not different between the two groups (9.4 in patients treated for > 2 years versus 9.1 in untreated patients and those receiving miglustat for < 2 years). A statistically significant delay to death was noted in patients treated with miglustat for > 2 years versus untreated patients and those receiving miglustat for < 2 years (p = 0.029; log-rank test)