Abstract TMP30: High Prevalence Of Unrecognized Cerebral Amyloid Angiopathy Related Inflammation Among Patients With Cerebral Amyloid Angiopathy And White Matter Disease

Abstract

Introduction: Cerebral amyloid angiopathy related inflammation (CAARI) is an autoimmune inflammatory condition that occurs in patients with cerebral amyloid angiopathy (CAA) and can lead to rapidly progressive cognitive decline. Little is known about the prevalence of subclinical and clinical CAARI. We sought to determine the prevalence of subclinical and clinical CAARI among patients with CAA and white matter hyperintensities (WMH). Methods: In this multicenter, single healthcare system, retrospective cohort study, we reviewed electronic medical records and MRI brain scans of patients presenting with a diagnosis code for amyloidosis or CAA that were evaluated between 1/2010 - 6/2020. We included patients meeting modified Boston criteria for CAA who had WMH on T2 FLAIR sequences. Patients without available hemosiderin sensitive sequences (SWI or GRE) and T2 FLAIR sequences and those without WMH on T2 FLAIR imaging were excluded. Two independent vascular neurologists blinded to background clinical information reviewed each MRI brain scan for the presence of CAARI. The clinical course and outcomes were reviewed and reported. Results: Out of 1100 patients reviewed, 511 met modified Boston criteria for CAA and 193 met the final study inclusion criteria. 55 (28.5% of those with CAA and WMH, and 10.8% of all CAA) patients had MRI brain imaging suggestive of CAA-RI. 21 (38.2%) were male, 38 (69.1%) were Caucasian, and the mean (SD) age was 72.9 (8.9) years. CAA-RI was recognized in only 10 (18.2%) patients initially while 20 (36.4%) were diagnosed up to 9 months later (median 0, IQR 0-9 months). At time of earliest detection of CAARI on imaging, common concurrent findings were cognitive impairment (74.5%), macro-hemorrhages (52.7%), headache (30.9%), seizures (14.5%), and ischemic infarcts (14.5%). Only 18 (32.7%) patients were started on immunosuppression. Nineteen (34.5%) patients expired during the observation period of which only 8 (42.1%) were ever diagnosed with CAA-RI by their treating clinician. Conclusion: The prevalence of subclinical CAA-RI in our study was high. Most cases of radiographic CAARI went unrecognized and untreated. Further studies are needed to assess if treatment of subclinical CAARI may prevent cognitive decline in these patients.