To determine associations between brain atrophy rate and clinical outcome measures in multiple sclerosis (MS).

The determination and treatment of neurodegenerative aspects of MS remains a challenge. One indirect form of assessment is the whole brain atrophy rate as determined by longitudinal MRI. Integration of brain atrophy rate assessments into clinical practice is hampered by multiple technical and translational factors, including incomplete clinical validation. It remains unclear to what extent clinical assessments are reflected and associated with this measure.

In a large retrospective observational cohort, we used FSL-SIENA to determine annualized percent brain volume change (aPBVC) from longitudinal clinical 3T MRI obtained using a uniform hardware and software acquisition protocol. Clinical outcomes, including confirmed disability progression, the 9-hole peg test (9HPT), the symbol digit modalities test (SDMT), and the timed-25-foot walk (T25w) were determined from retrospective chart review. SDMT, 9HPT, and T25w were averaged between the first and second MRI sessions to produce a single number. Using forward regression models, we assessed clinical predictors of aPBVC, including CDP and secondary clinical outcome measures, as independent variables along with age of onset and type of MS.

123 patients met inclusion criteria (102 RMS, 21 PMS; mean age = 36 ±SD 11) with available clinical and MRI data. In univariable analyses, confirmed disability progression showed the strongest association with concurrent aPBVC (β=0.31, p <0.001), followed by the SDMT (β=0.27, p=0.006) whereas 9HPT and T25 were not significant (p>0.05).  Change in EDSS and cross-sectional SDMT both remained significantly associated in multivariable models (p=0.008 and p=0.04, respectively).