An evaluation of perivascular space burden and contrast enhancement on 7T MRI in multiple sclerosis

Changes to the perivascular space (PVS) may be associated with multiple sclerosis (MS) pathology. In this study, we aimed to quantify PVS burden and assess PVS contrast enhancement on 7T MRI in MS.

Participants with MS (N = 100), other inflammatory neurological disorders (OIND, N = 12), and healthy controls (HC, N = 18) underwent 7T MRI. PVS segmentation was performed using an automated algorithm. Post-contrast FLAIR images were reviewed for gadolinium contrast-enhancing PVS (CE-PVS). The count, volume, and fraction of PVS and CE-PVS patterns were quantified.

Overall, 67% of participants were female, with the same proportion across HC, MS, and OIND groups. Median age was 41 years in HC, 46 years in MS, and 35 years in OIND. The segmentation algorithm achieved a median Dice Similarity Coefficient (DSC) of 0.994 compared to manually revised masks. PVS count, volume, and the proportion of PVS volume divided by total white matter (WM) volume did not differ significantly between individuals with and without MS. This finding was no different when restricting the analysis to only PVS in the large size category (> 2.65 mm3). CE- PVS were seen in 29% of MS, 11% of HC, and 8% of OIND, p = 0.431. Participants with progressive MS (PMS, N = 21) had greater PVS count (1341 (841, 1481)) than those with relapsing-remitting MS (RRMS, N = 79) (849 (571, 1283)), adjusted p = 0.037.

7T MRI and an automated algorithm enabled accurate PVS quantification. Although trends towards increased baseline PVS burden and CE-PVS in MS were seen, these differences were not significant. An observed difference in PVS between PMS and RRMS across all time points suggests a potential biological distinction, though, which warrants further investigation.