Profiles of cortical and white matter lesion inflammation in multiple sclerosis by combined 11C-PBR28 and 7 T imaging

Background: Microglia accumulate in a cytotoxic phenotype in active demyelinating white matter lesions (WML), while drop in numbers and change to a homeostatic phenotype in inactive lesions. Whether this activity pattern can be detected in cortical lesions (CL), the relation between inflammation in CL and WML and their clinical correlates are unknown. Goals: To characterize in vivo, using 7 T and magnetic resonance-positron emission tomography (MR-PET) imaging with 11C-PBR28, which binds to activated microglia, profiles of inflammatory activity in CL and WML and their association with clinical burden. Methods: Eleven secondary-progressive, 11 relapsing-remitting multiple sclerosis (SPMS, RRMS) patients, and 14 age- and binding affinity matched healthy controls underwent MR-PET and 7T T2* imaging. 11C-PBR28 binding was measured using normalized standardized uptake values (SUVR) in WML and CL segmented at 7 T. The relative difference in cortical and WM SUVR between patients and controls was used to classify CL and WML as active or inactive. Active WML included lesions active only in the core, lesions active only perilesionally or both. The relationship between profiles of microglia activity in CL and WML within patients was studied using χ2-test. The median of active CL and WML was assessed in patients and used to classify each subject with high or low CL and WML inflammation respectively. Differences in Expanded Disability Status Scale (EDSS) and Symbol Digit Modalities Test (SDMT) scores were investigated in patients with high versus low CL and WML inflammation using Mann-Whitney U-test. Results: In RRMMS, 30/103 (29%) CL and 80/412 (20%) WML, in SPMS 226/410 (55%) and 279/588 (47%) were identified as active. Six out of 22 patients showed different profiles of inflammation in CL and WML and the relation between inflammatory profiles in CL and WML showed only a trend (p = 0.08). Patients with high inflammatory CL activity had lower SDMT scores than patients with low CL inflammation (p=0.02). Patients with high inflammatory WML activity had also lower SDMT scores and higher EDSS than patients with low CL inflammation (p=0.05, p=0.002, respectively). Conclusions: 11C-PBR28 and 7 T imaging reveal distinct inflammatory profiles in CL and WML in MS, which can be used to identify patients with more aggressive disease. Inflammatory activity in CL is not strictly related to WML inflammation, suggesting distinct underlying microglia-mediated events.