Anticoagulation and Outcomes After Traumatic Cerebral Venous Sinus Thrombosis: A Retrospective Cohort Study.

BACKGROUND

Traumatic cerebral venous sinus thrombosis (tCVST) may complicate traumatic brain injury (TBI), for which anticoagulation practices remain poorly defined. We aimed to investigate anticoagulant practices, early complications, and factors associated with venous recanalization in patients with tCVST.

METHODS

We conducted a retrospective cohort study between 2012 and 2023 in 2 level 1 trauma centers, including TBI patients admitted to intensive care units (ICU). Data on trauma, tCVST features, anticoagulation practices, bleeding complications, venous infarction, and recanalization were extracted from medical records. Predictors of recanalization were identified using multivariable logistic regression.

RESULTS

Among the 91 TBI patients identified with concomitant tCVST (1.7% of ICU TBI admissions), anticoagulation practices were heterogeneous. Fourteen patients (15.4%) were managed conservatively (none or prophylactic anticoagulation), while the remaining patients (77/91, 84.6%) received intermediate-dose or therapeutic anticoagulation, within the first 7 days post-TBI (53/91, 58.2%) or later (24/91, 26.4%). New or worsening intracranial hemorrhage (ICH) under anticoagulation occurred in 9/91 (9.9%) patients, regardless of timing, and tCVST-related venous infarction was documented in 1/91 (1.1%) patient. Among the 79/91 (86.8%) survivors at hospital discharge, 59/79 (74.7%) patients had full or partial recanalization. In multivariable analysis, early anticoagulation was not associated with recanalization (OR=0.64, 95% CI: 0.12-3.07, P =0.6); only tCVST features, including initial partial sinus obstruction (OR=7.12, 95% CI: 1.41-49.25, P =0.03) and presence of depressed skull fracture (OR=0.07, 95% CI: 0.01-0.53, P =0.02) were associated with recanalization.

CONCLUSIONS

Early anticoagulation was not associated with secondary ICH or tCVST recanalization, although limited power and residual confounding may affect these findings. Larger prospective studies are warranted.