Abstract TP377: Predictors of Diffusion-weighted Imaging Lesions and One-year Case-fatality in Intracerebral Hemorrhage

Abstract

Introduction: Diffusion-weighted imaging (DWI) lesions in intracerebral hemorrhage (ICH) have been associated with vasculopathy, blood pressure (BP) lowering, and poor functional outcomes. We sought to identify predictors of DWI lesion formation and one-year case-fatality in ICH. Methods: In this retrospective study of primary ICH at our institution from 2009-2011, cases were identified by ICD-9 code and verified by physician review. Only subjects undergoing brain MRI within 72 hours of presentation were included. In addition to standard chart review, we abstracted all BPs available prior to MRI. Vasculopathy severity was assessed by white matter disease score and microbleed burden. Univariate analysis was stratified by DWI lesion presence; logistic regression analyses assessed predictors of DWI lesions and all-cause one-year case-fatality. Results: Of 500 subjects in the overall ICH cohort 98 met inclusion criteria for this analysis, of which 20 (20.4%) had DWI lesions. Selected univariate analysis is shown in the Table. In the best clinical model fit, intraventricular extension (OR 5.63 (1.60, 19.85), p=0.007), microbleeds (OR 3.43 (0.95, 12.34), p=0.06), and delta systolic BP lowering (((max systolic BP-min systolic BP)/max systolic BP)*100) (OR 0.97 (0.95, 0.99), p=0.0005) were associated with DWI lesion presence. Overall one-year case-fatality was 16.3%; age (OR 1.04 (1.00, 1.07), p=0.05), GCS (OR 1.30 (1.09, 1.56), p=0.004), and DWI lesions (OR 3.80 (1.14, 12.61), p=0.03) were independently associated. Conclusions: Predictors of DWI lesions in ICH include underlying vasculopathy but the impact of BP lowering on DWI lesion formation remains unclear. Further investigation of BP effects on DWI lesion formation is warranted with larger cohorts given the association with DWI lesions and one-year case-fatality. Future trials of BP control in ICH should consider stratifying by imaging biomarkers for underlying vasculopathy.