Clarke, Nathan

Abstract The risk of intracranial hemorrhage (ICH) associated with the use of bevacizumab (BEV) has been a topic of several reviews in the literature, and is considered to be a potential adverse effect of BEV when used in patients with malignant glioma. We performed a single-center retrospective chart-review identifying glioma patients treated with BEV at our institution. 595 unique patients who received BEV with documented follow up were identified. Of these patients, 23 (3.9%) developed ICH within 12 weeks (five half-lives) of a BEV infusion (median 9 days post infusion, 1-24 days IQR). There was a trend towards statistical significance between dose administered and the incidence of hemorrhage (P=0.06), with all 23 hemorrhages occurring in patients treated with 10 mg/kg (n=21/490) per infusion or 15 mg/kg per infusion (n=2/35) (n=525, 88.2% of cohort). No patient on dosing regimens of 7.5 mg/kg (n=72) or 5 mg/kg per infusion (n=16) (n=88, 14.7% of cohort) developed ICH. Therapeutic anticoagulation and concomitant BEV therapy was observed in 103 (17%) patients with 10 patients (10/93, 9.7%) developing an ICH (n=10), at a rate significantly higher than in patients treated with bevacizumab alone (13/479, 2.6%) (p=0.0025). When hemorrhages occurred (with or without concomitant anticoagulation) they were associated with high morbidity and mortality (ECOG 3.54 versus 1.26, p = < 0.0001), with 12 patients (52%) suffering severe debility (n=2) or death (n=10). CONCLUSION: Bevacizumab associated intracerebral hemorrhages remain uncommon events, similar to the baseline rate of spontaneous hemorrhage associated with malignant glioma. We observed a trend towards significance with higher doses per infusion associated with an increased rate of intracerebral hemorrhage. The combination of anticoagulation and BEV appears to confer a greater risk of hemorrhage than BEV alone.