Abstract
Neuroinflammation occurs in response to acute ischemic stroke, and has been speculated to underlie secondary poststroke pathologies, such as depression, that often develop over time poststroke. However, no study has examined whether neuroinflammation is present in chronic stroke patients (e.g., ≥ 1 year poststroke). This study tested whether neuroinflammation is present in chronic stroke patients, and is associated with neurodegeneration, using [(11)C]PBR28 PET and diffusion MRI. Eight patients with middle cerebral artery (MCA) ischemic stroke incurred 1-3 years prior and 16 healthy controls underwent [(11)C]PBR28 PET to measure glial activation and diffusion MRI to measure microstructural integrity by mean diffusivity (MD) and fractional anisotropy (FA) using an integrated PET/MRI scanner. Group differences in [(11)C]PBR28 binding, MD and FA were analyzed voxelwise across the whole brain excluding the infarct zone defined as voxels containing the infarct in any patient. Compared to controls, patients showed elevations in [(11)C]PBR28 binding in several brain regions outside the infarct zone, including regions with presumed direct neuroanatomical connections to the infarct (e.g., ipsilesional internal capsule and thalamus) and those without known direct connections (e.g., contralesional thalamus and cingulate gyrus). Patients also showed widespread elevations in MD, with a subset of these regions having reduced FA. In patients, MD was more elevated in regions with co-localized elevations in [(11)C]PBR28 binding than in contralateral regions without elevations in [(11)C]PBR28 binding. This pilot study supports the presence of extensive glial activation along with widespread loss in microstructural integrity in non-infarcted tissue in a cohort of patients with chronic MCA stroke. The loss in microstructural integrity was greater in regions with co-localized glial activation. It is possible that stroke risk factors (e.g., hypertension) contributed to these tissue changes in patients.