Intracranial hemorrhage in patients with brain metastases treated with therapeutic enoxaparin: a matched cohort study

Abstract

Venous thromboembolism occurs frequently in patients with cancer who have brain metastases, but there is limited evidence supporting the safety of therapeutic anticoagulation. To assess the risk for intracranial hemorrhage associated with the administration of therapeutic doses of low-molecular-weight heparin, we performed a matched, retrospective cohort study of 293 patients with cancer with brain metastases (104 with therapeutic enoxaparin and 189 controls). A blinded review of radiographic imaging was performed, and intracranial hemorrhages were categorized as trace, measurable, and significant. There were no differences observed in the cumulative incidence of intracranial hemorrhage at 1 year in the enoxaparin and control cohorts for measurable (19% vs 21%; Gray test, P = .97; hazard ratio, 1.02; 90% confidence interval [CI], 0.66-1.59), significant (21% vs 22%; P = .87), and total (44% vs 37%; P = .13) intracranial hemorrhages. The risk for intracranial hemorrhage was fourfold higher (adjusted hazard ratio, 3.98; 90% CI, 2.41-6.57; P < .001) in patients with melanoma or renal cell carcinoma (N = 60) than lung cancer (N = 153), but the risk was not influenced by the administration of enoxaparin. Overall survival was similar for the enoxaparin and control cohorts (8.4 vs 9.7 months; Log-rank, P = .65). We conclude that intracranial hemorrhage is frequently observed in patients with brain metastases, but that therapeutic anticoagulation does not increase the risk for intracranial hemorrhage.

Figure 1

Cumulative incidence of intracranial hemorrhage (ICH) in patients with metastatic brain tumors. No differences between enoxaparin and control cohorts were observed in the cumulative incidence of intracranial hemorrhage for any category (Gray test, P > .05) including measurable (A), significant (B), and total (C) hemorrhages. Enoxaparin cohort shown in red and controls in blue.

Figure 2

Volume histograms of intracranial hemorrhages. The distribution of all hemorrhages identified among control (A) and enoxaparin (B) cohorts are shown according to volume (mL). SAH, subarachnoid hemorrhage.

Figure 3

Cumulative incidence of significant intracranial hemorrhage (ICH) in the non-small cell lung cancer and melanoma/renal cell carcinoma subgroups. (A) The cumulative incidence of significant intracranial hemorrhage in patients with non-small lung cancer at 1 year was 15% in the enoxaparin cohort compared with 19% in the control cohort (Gray test, P = .93). (B) In the melanoma plus renal cell carcinoma subgroup, the cumulative incidence of significant intracranial hemorrhage at 1 year was 35% for the enoxaparin cohort vs 34% for the controls (Gray test, P = .88). Enoxaparin cohort shown in solid gray line and controls in hatched black line. NSCLC, nonsmall lung cancer; RCC, renal cell carcinoma.

Figure 4

Kaplan-Meier survival curves comparing enoxaparin and control cohorts. The median survival was 8.4 months in the enoxaparin cohort (gray) and 9.7 months in the control cohorts (black). Log-rank test, P = .65.