Subcortical plaques and inflammation reflect cortical and meningeal pathologies in progressive multiple sclerosis.

It remains elusive whether lesions and inflammation in the sub/juxtacortical white matter reflect cortical and/or meningeal pathologies. Elucidating this could have implications for MRI monitoring as sub/juxtacortical lesions are detectable by routine MRI, while cortical lesions and meningeal inflammation are not. By large-area microscopy, we quantified total and mixed active plaque loads along with densities and sizes of perivascular mononuclear infiltrates (infiltrates) in the sub/juxtacortical white matter ≤2 mm from the cortex, intra-cortically and in the meninges. Data were related to ante-mortem clinical parameters in a false discovery rate-corrected analysis. We compared 12 patients with primary progressive multiple sclerosis (PPMS) and 15 with secondary progressive MS to 22 controls. Fifteen patients and 11 controls contributed with hemispheric sections. Sections were stained with haematoxylin-eosin, for myelin and for microglia/macrophages. B cells and T cells were confirmed in a subset. Immunoglobulin G depositions in selected cortical plaques resembled depositions described before in "slowly expanding" plaques in the white matter. We quantified plaque activity by measuring microglia-dominated and macrophage-dominated areas. Sub/juxtacortical plaques (load and activity) reflected plaque activity in the cerebral cortex. Plaque activity and infiltrates were more pronounced in the sub/juxtacortical white matter than in the cerebral cortex while conversely, the total plaque load was highest in the cortex. Infiltrates correlated trans-cortically and sub/juxtacortical plaque activity reflected cortical and meningeal infiltrates. Sub/juxtacortical infiltrate sizes correlated with shorter survival after progression onset. Two patients with PPMS and putatively fatal brain stem lesions argue against incidental findings. Trans-cortical inflammatory flares and plaque activity may be pathogenic in progressive MS. We suggest emphasis on sub/juxtacortical MRI lesions as plausible surrogates for cortical and meningeal pathologies and, when present, as indicators for cognitive testing.