Abstract TMP111: Prediction of Cognitive Impairment after Intracerebral Hemorrhage using MRI Small Vessel Disease Score

Introduction: Survivors of Intracerebral Hemorrhage (ICH) are at high risk for cognitive impairment. Previous studies clarified that Cerebral Small Vessel Disease (CSVD) is a primary contributor to post-ICH dementia, but these observations have failed to transform clinical or research practice standards to date. We sought to determine whether a validated MRI CSVD scoring system could readily predict cognitive impairment after ICH. Methods: We analyzed data from ICH survivors with no history of prior cognitive impairment, enrolled in a single-center prospective study. We reviewed MRI scans to compute: 1) a validated 6-point score for CSVD burden based on presence of microbleeds, white matter hyperintensities, lacunes and >20 basal ganglia enlarged perivascular spaces; 2) cortical atrophy. We quantified cognitive performance by: 1) administering the Telephone Interview for Cognitive Status (TICS) test; 2) identifying diagnosis of dementia based on medical records. We utilized linear mixed model analyses to identify predictors of changes in TICS score, and Cox regression to identify predictors of new-onset dementia. We calculated CSVD score cut-offs to maximize sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for diagnosis of dementia. Results: We enrolled 612 primary ICH survivors, and followed them for a median of 46.3 months (Inter-Quartile Range: 35.5-58.7). A total of 214/612 (35%) participants developed dementia. Increasing CSVD scores were associated with faster cognitive decline (Coeff -0.25, Standard Error 0.02) and dementia diagnosis (Hazard Ratio 1.33, 95% CI 1.08-1.64). Age, increasing atrophy scale score, lower TICS score at baseline and systolic blood pressure were also independently associated with faster cognitive decline and new-onset dementia (all p Conclusion: This validated CSVD score is strongly associated with cognitive performance after ICH, with excellent predictive performance for future diagnosis of dementia. Our results support its implementation in clinical care and in future studies of post-ICH dementia.