Clinical Relevance of Intracranial Hemorrhage Volume in Primary and Metastatic Brain Tumors

Background

Venous thromboembolism occurs in approximately 15-30% of individuals with primary or metastatic brain tumors. Spontaneous intracranial hemorrhage (ICH) is also a frequent complication of brain tumors thus complicating the decision to administer therapeutic anticoagulation. We previously demonstrated that anticoagulation does not increase the risk of ICH in patients with brain metastases; however, in the setting of glioma, the risk of major ICH is 13-fold higher during exposure to anticoagulation. Absent a standardized definition of major intracranial hemorrhage, major was based on bleed volume greater than 10ml, symptoms, or requiring surgical intervention. The aim of this study was to evaluate the 10 mL bleed volume cutoff with clinical presentation and outcomes for ICH with brain tumors.

Methods

In a retrospective matched-cohort study, we evaluated patients who had primary brain tumors (n=133) and those with metastatic tumors (n=293). Matched controls were identified using a round-robin algorithm that identified best-match according to baseline demographics. A blinded review of radiographic imaging was performed and the volume of hemorrhage calculated using the ½ ABC method. Patient charts were reviewed for all surgical procedures associated with hemorrhages as well as for the presence of symptoms such as focal neurologic deficit, headache, nausea, or change in cognitive function. Statistical comparisons were performed by one sided Fisher's test and log rank test on Kaplan-Meier estimates of survival.

Results

A total of 61 out of 133 patients with primary brain tumors (45.8%), and 138 patients out of 293 patients with metastatic tumors (47%) were identified with an intracranial hemorrhage. In metastatic tumors, 87.5% (21 of 24) patients with a bleed volume of 10 mL or higher were symptomatic compared with 36% with smaller volumes (P<0.001). Similarly, glioma patients with a hemorrhage volume of 10 mL or higher was associated with the development of clinical symptoms (83% vs 17%, P<0.001). Enoxaparin administration did not influence the likelihood of symptomatology for hemorrhages > 10 mL either in glioma (P=0.3) or brain metastases (P=0.65). A threshold of 10 mL ICH also discriminated between the need for neurosurgical procedure (odds ratio 8.50, 95% CI 3.29-21.70, P<0.001). The median survival following any hemorrhage was poor (3.2 months) and was not significantly influenced by the size of hemorrhage (P=0.42).

Conclusions

Intracranial bleed volume of 10 mL or more in patients with primary or metastatic brain tumors consistently correlated with clinical symptoms and need for surgical intervention. Volume assessment is an important element in assessing the clinical significance of intracranial hemorrhage and this data supports the implementation of a 10 mL volume threshold to characterize intracranial hemorrhages in cancer patients.