Early versus late anticoagulation on outcomes in large hemispheric infarction: a propensity score-matched analysis

The optimal timing of anticoagulation initiation in patients with large hemispheric infarction (LHI) remains uncertain, particularly the trade-off between thromboembolism and hemorrhagic transformation risk. This study aimed to compare clinical outcomes between early and late anticoagulation (AC) initiation in patients with LHI.

We performed an analysis in a retrospective cohort of 264 consecutive patients with LHI (early AC ≤ 7 days; late AC > 7 days). Propensity score matching (PSM) was performed to balance baseline characteristics between early and late groups (43 patients per group). We assessed the efficacy and safety outcomes of anticoagulation at 1 year and 3 months after stroke onset. Primary efficacy outcome was composite of ischemic stroke recurrence or systemic embolism. Cox proportional hazards regression was performed to obtain adjusted hazard ratios (HRs). Secondary efficacy outcomes included favorable functional outcome (modified Rankin Scale [mRS] 0-2) and all-cause mortality. Safety outcomes included symptomatic intracranial hemorrhage, major hemorrhage. Subgroup analyses assessed primary efficacy outcomes across prespecified variables.

Before PSM, the early AC group demonstrated a significantly lower rate of stroke recurrence or systemic embolism compared to the late AC group (8.3% vs. 27.7%; HR = 0.55, 95% CI 0.33-0.92; log-rank P < 0.001), which persisted after PSM (9.3% vs. 27.9%; log-rank P = 0.024). On multivariate Cox regression, early AC was independently associated with a markedly reduced risk of stroke recurrence before PSM (adjusted HR = 0.307, 95% CI 0.144-0.655; P = 0.004), however with the reduced sample size and limited outcome events, there was no statistical significance after PSM (HR = 0.443, 95% CI 0.179-1.099; P = 0.158). Good functional outcome (mRS 0-2) favored early AC in the overall cohort (45.3% vs. 18.1%; OR = 3.75, 95% CI 2.00-7.06; P < 0.001), though this difference was attenuated after matching (39.5% vs. 27.9%; P = 0.362). Rates of symptomatic hemorrhage (6.6% vs. 9.6%; P = 0.389), major hemorrhage (8.3% vs. 13.3%; P = 0.204), and all-cause mortality (8.8% vs. 10.8%; P = 0.615) did not differ significantly between groups, a pattern that was maintained in the matched cohort. Subgroup analyses showed statistically significant differences in stroke recurrence or systemic embolism among patients with moderate stroke severity (HR = 0.21, 95% CI 0.03-1.50, P = 0.007), absence of malignant MCA infarction (HR = 0.43, 95% CI 0.14-1.34, P = 0.028), no prior stroke or TIA (HR = 0.42, 95% CI 0.14-1.31, P = 0.022), and below 75 years old (HR = 0.28, 95% CI 0.04-1.97, P = 0.048).

Early anticoagulation initiation after large hemispheric infarction was independently associated with a substantially reduced risk of stroke recurrence and systemic embolism without a significant increase in hemorrhagic complications or mortality. These findings support the clinical benefit of early anticoagulation and provide further evidence for LHI patients.