Atezolizumab/bevacizumab or lenvatinib in hepatocellular carcinoma: Multi-center real world study with focus on bleeding and thromboembolic events

Background & Aims Atezolizumab/bevacizumab (atezo/bev) and lenvatinib have demonstrated efficacy as first-line therapies for hepatocellular carcinoma (HCC). However, VEGF-inhibition with these therapies may be associated with the risk of bleeding and thromboembolic events. In this study, we evaluated the efficacy and safety with focus on bleeding and thromboembolic events of atezo/bev versus lenvatinib in a large, multi-center real-world population. Methods This study is based on HCC cohorts from 7 centers in Germany and Austria. Incidences of bleeding or thromboembolic events and efficacy outcomes were assessed and compared. Results In total, 464 patients treated with atezo/bev (n=325) or lenvatinib (n=139) were analyzed. Both groups were balanced with respect to demographics, presence of liver cirrhosis, and variceal status. Duration of therapy did not differ between groups. Within 3 months of therapy, bleeding episodes were described in n=57 (18%) patients receiving atezo/bev compared to n=15 (11%) patients receiving lenvatinib (p=0.07). Variceal hemorrhage occurred in n=11 (3%) patients treated with atezo/bev versus n=4 (3%) patients treated with lenvatinib (p=0.99). Thromboembolic events were reported in n=19 (6%) of patients in the atezo/bev cohort compared to n=5 (4%) patients in the lenvatinib cohort (p=0.37). Incidence of overall bleeding, variceal hemorrhage, and thromboembolic events did also not differ significantly in patients, who received either atezo/bev or lenvantinib for 6 months. Conclusions Safety considerations related to bleeding and thromboembolic events may not be helpful in guiding clinical decision-making when choosing between atezo/bev or lenvatinib. Impact and implications The inhibition of VEGF by current first-line therapies for HCC, such as atezolizumab/bevacizumab or lenvatinib, may be associated with the risk of bleeding and thromboembolic events. Studies comparing the incidence of these side effects between atezolizumab/bevacizumab and lenvatinib, which are preferred treatments over sorafenib for HCC, are needed. Differences in this side effect profile may influence the choice of first-line therapy by treating physicians. Since no significant differences were observed regarding bleeding or thromboembolic events between both therapies in the present study, we conclude that safety considerations related to these events may not be helpful in guiding clinical decision-making when choosing between atezolizumab/bevacizumab or lenvatinib.