To quantify MRI predictors of disability progression and progression independent of relapse activity (PIRA) from patients with multiple sclerosis (MS) in an observational setting.

PIRA is a major contributor of disability worsening in MS and has been extensively studied in clinical trial cohorts.

We obtained demographics, patient-reported measures (patient-determined disease steps [PDDS], relapse, and MS course), and performance measures (walking speed test [WST], manual dexterity test [MDT], and processing speed test [PST]). Inclusion criteria were at least 3 measurements and ≥6 months of follow-up. Confirmed disability worsening (CDW) was defined as a 6-month-sustained increase of ≥1 in PDDS, 20% increase for WST or MDT, or 4-point decrease for PST. Fixed and roving reference methods were applied. CDW was classified as PIRA when it occurred in the absence of relapses between the onset and confirmation of CDW (±30days). MRIs acquired within 6 months of the first visit were analyzed for whole brain fraction (WBF), T2 lesion volume (T2LV), and upper cervical spinal cord area (SCA). The baseline characteristics were compared based on CDW and PIRA status. Logistic regression was performed to predict PIRA.

4909 subjects were included. CDW occurred in 1825 (37%) using fixed baseline and in 2599 (53%) using roving baseline. PIRA represented 37% (679) of all CDW in fixed and 43% (1121) in roving. PIRA patients were older (p<0.05), had better MDT score (p<0.001), and better PST score (p<0.02). PIRA subjects had lower baseline WBF (0.803 vs 0.811, p<0.01) and SCA (74.6 vs 77.1, p<0.01) with no difference in T2LV (p>0.1). Logistic regression showed PIRA was associated with lower baseline MDT, WBF, and SCA (p<0.01).

PIRA represents a major contributor to disability accumulation in MS. Baseline brain volume and spinal cord area were associated with PIRA suggesting a potential neurodegenerative mechanism for this type of disability accrual.