Intracerebral haemorrhage in patients with brain metastases receiving therapeutic anticoagulation

Background Venous thromboembolism is common in patients with solid malignancies and brain metastases. Whether to anticoagulate such patients is controversial given the possibility of intracerebral haemorrhage (ICH). We evaluated the added risk of ICH in patients with brain metastases receiving therapeutic anticoagulation. Methods We performed a matched, retrospective cohort study of 291 patients (100 receiving therapeutic anticoagulation vs 191 controls) with brain metastases managed at Brigham and Women's Hospital/Dana-Farber Cancer Institute between 1998 and 2015. For each patient, all MRI studies of the brain were reviewed to identify ICH. Propensity score matching and multivariable Cox regression were used to mitigate confounding. Results The risk of ICH was comparable in patients receiving anticoagulation versus controls preanticoagulation. Postanticoagulation, we observed significant or borderline-significant associations between anticoagulation and development of any ICH (HR 1.31, 95% CI 0.96 to 1.79, p=0.09), ICH as identified by gradient echo/susceptibility-weighted imaging (HR 1.46, 95% CI 1.06 to 2.01, p=0.02), symptomatic ICH (HR 1.80, 95% CI 1.01 to 3.22, p=0.05), extralesional ICH (HR 5.82, 95% CI 1.56 to 21.7, p=0.009) and fatal ICH (HR 5.68, 95% CI 0.60 to 54.2, p=0.13). Anticoagulation was associated with differentially higher ICH risk in patients with prior ICH versus no prior ICH (HR 2.20 vs 0.68, respectively, p interaction <0.001) and symptomatic ICH risk in melanoma versus other primary malignancies (HR 6.46 vs 1.36, respectively, p interaction=0.02). Conclusions Anticoagulation is associated with clinically significant ICH in patients with brain metastases, especially those with melanoma or prior ICH. The indication for anticoagulation and risk of intracerebral bleeding should be considered on an individual basis among such patients.