Abstract WP306: Determinants of Hematoma Volume Differ by Location in Intracerebral Hemorrhage
Guido J Falcone,Alessandro Biffi,H. B Brouwers,Christopher Anderson,Thomas W Battey,Alison M Ayres,Anastasia Vashkevich,Kristin Schwab,Natalia S Rost,Joshua N Goldstein,Anand Viswanathan,Steven M Greenberg,Jonathan Rosand
DOI: https://doi.org/10.1161/str.44.suppl_1.awp306
IF: 10.17
2013-02-01
Stroke
Abstract:Background and purpose: Admission hematoma volume (AHV) is the strongest predictor of outcome in intracerebral hemorrhage (ICH). Mounting evidence suggests that the biological mechanisms underlying ICH occurrence differ by bleeding location. We aimed to assess if the determinants of AHV vary by bleeding topography, and to estimate location-specific effects for warfarin, a factor known to influence AHV. Methods: Retrospective analysis of 902 ICH cases (388 deep, 356 lobar and 58 cerebellar) enrolled in a single-center, prospective study of ICH. AHV was measured in the first available computed tomography using semi-automated, computer-assisted volumetric methods. Univariate and multivariate linear regression analyses, stratified by ICH location, were implemented to model natural log-transformed AHV as a function of covariates. Admission International Normalized Ratio (INR) values were utilized to establish intensity of anticoagulation. INR was categorized in 3. Results: AHV medians (interquartile range) for deep, lobar and cerebellar ICH were 13 (35), 39 (59) and 13 (24), respectively (p<0.0001). Within these location categories, 18, 21 and 29% of subjects were on warfarin, and 36, 42 and 35% had died at 90 days, respectively. In multivariate analysis, independent predictors of deep AHV were male sex (beta[β]=0.28, standard error[SE]=0.14, p=0.05), history of coronary artery disease (β=0.33, SE=0.17, p=0.05), and degree of anticoagulation (β=0.32, SE=0.08, p=0.00005, tested for trend across INR categories); predictors of lobar AHV were antiplatelet treatment (β=0.27, SE=0.13, p=0.03) and degree of anticoagulation (β=0.14, SE=0.06, p=0.02, comparing INR 3); and predictors of cerebellar AHV were effective anticoagulation (β=0.99, SE=0.41, p=0.02, comparing INR 2) and statin treatment (β=-1.21, SE=0.38, p=0.001). Conclusion: Determinants of AHV differ by ICH location. Anticoagulation is associated with larger AHV in all locations, but the strength and shape of the dose-response curve for this relationship varies with ICH topography. Pre-ICH treatment with antiplatelets and statins was associated with increased mean lobar AHV and decreased mean cerebellar AHV, respectively.
peripheral vascular disease,clinical neurology