NeuroImage: Clinical (Jan 2022)
The spatio-temporal relationship between white matter lesion volume changes and brain atrophy in clinically isolated syndrome and early multiple sclerosis
Abstract
Background: White matter lesions and brain atrophy are both present early in multiple sclerosis. However, the spatio-temporal relationship between atrophy and lesion processes remains unclear. Methods: Yearly magnetic resonance images were analyzed in 392 patients with clinically isolated syndrome from the 5-year REFLEX/REFLEXION studies. Patients received early treatment (from baseline; N = 262) or delayed treatment (from month-24; N = 130) with subcutaneous interferon beta-1a. Global and central atrophy were assessed using FSL-SIENA to provide yearly percentage volume change of brain and ventricles, respectively. Yearly total lesion volume change was calculated by subtracting the sum of the negative lesion volume change (disappearing + shrinking) from the positive lesion volume change (new + enlarging) for each yearly interval, as determined by an in-house developed semi-automated method. Using linear mixed models, during the period where patients had received ≥1 year of treatment, we investigated whether total lesion volume change was associated with percentage brain volume change or percentage ventricular volume change in the next year, and vice versa. Results: Higher total lesion volume change was related to significantly faster global atrophy (percentage brain volume change) in the next year (B = − 0.113, SE = 0.022, p < 0.001). In patients receiving early treatment only, total lesion volume change was also associated with percentage ventricular volume change in the next year (B = 1.348, SE = 0.181, p < 0.001). Voxel-wise analyses showed that in patients receiving early treatment, higher total lesion volume change in years 2, 3, and 4 was related to faster atrophy in the next year, and in year 4 this relationship was stronger in patients receiving delayed treatment. Interestingly, faster atrophy was related to higher total lesion volume change in the next year (percentage brain volume change: B = − 0.136, SE = 0.062, p = 0.028; percentage ventricular volume change: B = 0.028, SE = 0.008, p < 0.001). Conclusions: Higher lesion volume changes were associated with faster atrophy in the next year. Interestingly, there was also an association between faster atrophy and higher lesion volume changes in the next year.
