Measurement of brain volume (BV) by MRI-based computational methods has proven clinical relevance as a marker of neurodegeneration. In multiple sclerosis (MS), accelerated BV loss after initiation of anti-inflammatory therapy complicates interpretations of BV changes.

Post-hoc analysis of MRI data from the IMPROVE study, comparing patients treated with subcutaneous IFNβ-1a (44 µg) three times per week (n=120) versus placebo (n=60). T1-weighted MRIs were acquired at weeks 0, 4, 8 12 and 16 using the SIENA-XL method to compare gray matter volume change (PGMVC) and white matter volume change (PWMVC). In the treatment arm, data acquired at week 20, 24, 28, 32, 36 and 40 were used to assess the slope of changes over the treatment period. Slopes of PGMVC and PWMVC change over time were estimated using a mixed-effect linear model. A treatment-by-time interaction term was used to assess whether the slope of changes differed according to treatment arm. In the IFNβ-1a-treated arm, a quadratic term was included to evaluate a plateauing effect over the treatment period.

Up to week 16, PGMVC was -0.14% per month in the placebo group and -0.27% per month in the treated group (p<0.001), and PWMVC was -0.067% per month in the placebo group and -0.116% per month in the treatment group (p=0.27). Over 40 weeks, the decrease in PGMVC in the treatment group showed a significant (p<0.001) quadratic component indicating a plateauing of the PGMVC starting at week 20. Small decreases in PWMVC showed a trend (p=0.06) for a quadratic slope with a plateau starting at week 20.

Monthly-acquired MRI from IFNβ-1a-treated patients showed more pronounced pseudo-atrophy effect in gray matter than in white matter, with a maximum decrease up to week 20.