Background: Several predictors for treatment failure to interferon-beta (IFN-beta) have been proposed; however, brain atrophy has not been well studied.
Methods: In this prospective and longitudinal study, all consecutive relapsing-remitting multiple sclerosis (RRMS) patients treated with sc IFN-beta-1a were included. Confirmed disability progression or a new relapse between weeks 48 and 144 after beginning with IFN-beta was considered as treatment non-response. EDSS progression, relapses, number of active lesions at 1 year (new or enlarging T2-weighted plus gadolinium-enhancing lesions, categorized in > 2 or ≤ 2), and brain parenchymal fraction (%BVC) volume change within the initial year of treatment were used as predictive factors. Cox regression model was adjusted for age, gender, and disease duration.
Results: Seventy-one patients were included (71·8% female) with a follow-up of 144 weeks. Thirty-four (48%) fulfilled criteria of non-response to IFN-beta treatment. The model showed: (1) relapses+disability progression: HR = 4·6, 95% IC: 3·1-6·7 (P < 0·001); (2) relapses+BVC decrease: HR = 4·1, 95% IC: 3·2-7·3 (P = 0·001); (3) relapses+disability progression+new active lesions: HR = 10·1, 95% IC: 7·1-15·2 (P < 0·001); and (4) relapses+disability progression+new active lesions+BVC decrease: HR = 14·4, 95% IC: 11·4-21·2 (P < 0·001).
Conclusions: Adding BVC measures to previously described predictive failure factors may increase sensitivity to early identify non-responder patients to IFN-beta-1a in the second and third years of therapy.
Keywords: Brain atrophy,; Interferon,; MR,; Multiple sclerosis,; Response.