Stage-specific riluzole effect in amyotrophic lateral sclerosis: a retrospective study.

Objectives: To estimate the effect of riluzole on the stage-specific risk of progression of ALS. Methods: Patients from the PRO-ACT dataset were staged employing two methods (King's and FT9). Hazard ratios associated with riluzole treatment were estimated for forward transition between stages, using unadjusted and adjusted Markov multistate models. Results: Of 1903 patients, 1587 had received riluzole. Riluzole-treated patients survived non-significantly longer than those who did not (median 22.9 months vs. 18.3 months from time of initial observation, log rank p = 0.16). After adjusting for age and ALSFRS-R slope at first visit, riluzole significantly reduced risk of the following transitions: (1) King's stages: 1->2 (hazard ratio (HR) = 0.81), and 2->3 (HR = 0.82), 4->death (HR = 0.57), and (2) FT9 stages: 1->2 (HR = 0.84), 3->4 (HR = 0.71), and 4->death (HR = 0.67). In contrast, the beneficial effect of riluzole in bulbar-onset patients was in early rather than late King's stages. Conclusions: This examination of cohorts closely followed in clinical trials finds a beneficial effect of riluzole that is predominantly but not exclusively in later stages of ALS. This analytic framework has utility to discern stage-specific treatment effects, and for refined health economic analyses.