Normalized brain volume predicts cognitive performance in MS: an analysis of a large cohort from fingolimod phase III studies (P7.284)

Objective To evaluate if individualized cut-off values of brain volume (BV) at baseline can predict cognitive-performance over two years. Background BV-loss correlates with cognitive decline in RRMS-patients. The use of BV-values to predict cognitive-performance, on an individual patient basis, requires evaluation. Methods The expected BV, normalized for head-size (NBV), was calculated for individuals, based on the baseline-characteristics of each patient and the study in which they were enrolled, using a multiple-regression model (pooled intent-to-treat populations; FREEDOMS, N=1272; FREEDOMS II, N=1083; TRANSFORMS, N=1280). Patients were stratified into three groups based on the differences between observed and expected NBV-values: low-NBV(L-NBV):>1 standard deviation(SD) below mean, medium-NBV(M-NBV):within±1SD, and high-NBV(H-NBV):>1SD above mean. The value of PASAT-3 (Paced Auditory Serial Addition Test) at baseline, at month 24(M24) and its change over 24-months were analyzed according to the baseline NBV-groups in the pooled FREEDOMS/FREEDOMS II studies after adjusting for treatment effect. Fingolimod effect vs. placebo was calculated for NBV at baseline. Results NBV was available for 3592 patients (mean±SD NBV:1520cm 3 ±82 cm 3 ). Baseline, T2-volume, age, EDSS, disease duration, and sex significantly affected NBV (all, p PASAT-scores differed in the three NBV-groups at baseline (means: L-NBV=44.9, M-NBV=48.1, H-NBV=50.2,p and M24 (adjusted for treatment arm, means: L-NBV=46.7, M-NBV=49.8, H-NBV=51.8,p C hange in PASAT-score over 24 months (adjusted for baseline PASAT and treatment arm) was significantly different among the three groups (means: L-NBV=+0.35, M-NBV=+1.60, H-NBV=+2.03,p=0.004). In the overall FREEDOMS/FREEDOMS II population, fingolimod had a significant effect on the PASAT-score change (adjusted for baseline-PASAT, +1.15, 95[percnt]CI=+0.53, +1.77,p<0.001). Conclusion Individualized cut-offs derived from baseline NBV-values adjusted for baseline-characteristics are predictive of subsequent cognitive-performance. Fingolimod significantly improved PASAT-score in the pooled FREEDOMS/FREEDOMS II population, based on NBV at baseline. Study supported by Novartis Pharma AG Disclosure: Dr. Sormani has received personal compensation for activities with Allozyne, Merck Serono, Teva Neuroscience, Synthon, Actelion, and Biogen Idec as a consultant and/or speaker. Dr. Kappos has received personal compensation for activities with Actelion Pharmaceuticals. Dr. Cohen has received personal compensation for activities with Biogen Idec, Eli Lilly, Novartis, and Vaccinex as a consultant and/or speaker. Dr. Barkhof has received personal compensation for activities with Bayer Schering Pharma, Sanofi, Genzyme, Biogen Idec, Teva, Merck Serono, Novartis, Roche, Synthon BV, and Janssen Research as a consultant. Dr. Sprenger9s institution has received research support from Novartis, ElectroCore, Genzyme, Actelion, Mitsubishi Pharma Europe, and Biogen Idec. Dr. Meier has received personal compensation for activities with Novartis as an employee. Dr. Haering has received personal compensation for activities with Novartis as an employee. Dr. Tomic has received personal compensation for activities with Novartis as an employee. Dr. De Stefano has received personal compensation for activities with Teva Neuroscience, Bayer, Sanofi-Aventis, Biogen Idec, Novartis, and Merck Serono.