Search for a command to run...
Background: It is unknown whether the association between prior anticoagulation use and outcome in patients with intracerebral hemorrhage (ICH) is dependent on the location of the hemorrhage. We aimed to determine if the association between anticoagulation and outcome differs in patients with deep vs lobar ICH. Methods: We conducted a retrospective cohort analysis of patients with spontaneous ICH admitted to a tertiary hospital from 2010-2024 and presenting within 24 hours of symptom onset. Deep hemorrhages were those that affected the basal ganglia, thalamus, or brainstem. Lobar hemorrhages were those involving the cerebral cortex with or without extension to the subcortical white matter. Multivariable logistic regression, stratified by hematoma location, was used to assess the association between oral anticoagulation use and the outcomes of hematoma expansion ([HE], absolute increase >6mL or relative increase >33%) and hospital mortality in both deep and lobar hemorrhages. Results: A total of 798 patients were included (mean [SD] age, 70 [15]; 352 females [44.1%]), of whom 432 (54.1%) had deep and 366 (45.9%) had lobar hemorrhages. There were 163 (20.4%) cases associated with oral anticoagulation; 88 (11.1%) with vitamin K antagonists (VKA) and 75 (9.5%) with direct oral anticoagulants (DOAC). Among patients with deep hemorrhage, anticoagulation was associated with both HE (42/100 [42.0%] vs 87/327 [26.2%]; adjusted odds ratio (OR) 1.73 [95% CI, 1.04-2.88]; P=0.036) and hospital mortality (35/100 [35.0%] vs 61/332 [18.4%]; OR 3.55 [1.73-7.27]; P=0.001). Among patients with lobar hemorrhage, anticoagulation was not associated with either HE (19/63 [30.2%] vs 126/303 [41.6%]; OR 0.61 [0.33-1.14]; P=0.122) or hospital mortality (17/63 [27.0%] vs 76/303 [25.1%]; OR 1.06 [0.50-2.27]; P=0.875). Compared to no anticoagulation exposure, VKAs were associated with HE in deep ICH (OR 2.51 [1.33-4.73]; P=0.005) while DOACs were not (OR 1.06 [0.51-2.20]; P=0.873]. Both VKA (OR 4.06 [1.71-9.64]; P=0.002) and DOAC (OR 2.95 [1.11-7.80]; P=0.029) were associated with mortality in deep ICH. Neither VKA or DOAC predicted HE or mortality in lobar ICH. Conclusions: Anticoagulation is associated with hematoma expansion and mortality in deep but not lobar ICH. Our findings underscore the importance of blood pressure control in anticoagulant users. Whether hematoma location modifies the effect of anticoagulation reversal on outcomes in ICH should be clarified in future studies.