Real-World Outcomes of Atezolizumab with Bevacizumab Treatment in Hepatocellular Carcinoma Patients: Effectiveness, Esophagogastroduodenoscopy Utilization and Bleeding Complications

Our real-world multicenter retrospective analysis suggests that omitting an esophagogastroduodenoscopy (EGD) in specific hepatocellular carcinoma (HCC) patients may be a safe and cost-effective strategy for those undergoing atezolizumab with bevacizumab (A+B), without leading to an elevated risk of bleeding complications. In our study, conducted in Canadian centers during the early access period of A+B, 30% of patients did not undergo pre-treatment EGD. This clinical decision was often based on the absence of cirrhosis, significant thrombocytopenia, or a low likelihood of portal hypertension, as assessed by their physicians. Despite the absence of standardized guidelines and the use of an individualized approach to EGD screening, patients did not experience negative treatment outcomes or worse survival. Our data also indicated that bleeding complications associated with A+B treatment are predominantly non-GI in nature. There may be several reasons not to use EGD routinely in this setting, which include a balance between patient risks and healthcare resources. EGD is an invasive procedure that requires sedation and carries a small risk of complications, along with potential discomfort and anxiety. The limited availability of expert endoscopists' time could also lead to significant delays in initiating effective therapy in the advanced HCC setting where A+B has been shown to prolong life.Abstract The IMbrave150 trial established atezolizumab with bevacizumab (A+B) as standard care for hepatocellular carcinoma (HCC), recommending an esophagogastroduodenoscopy (EGD) within 6 months of treatment initiation to prevent bleeding from esophagogastric varices. The necessity of mandatory EGD for all patients remains unclear. We retrospectively analyzed 112 HCC patients treated with A+B at five Canadian cancer centers from 1 July 2020 to 31 August 2022. A+B was the first-line therapy for 90% of patients, with median overall survival at 20.3 months and progression-free survival at 9.6 months. There was no survival difference between patients with bleeding and those without. Before A+B, 71% (n = 79) of patients underwent an EGD within 6 months, revealing varices in 41% (n = 32) and requiring intervention in 19% (n = 15). The overall bleeding rate was 15% (n = 17), with GI-specific bleeding occurring in 5% (n = 17). In the EGD group, GI-specific bleeding was 6% (n = 5) while in the non-EGD group, it was 3% (n = 1). Non-GI bleeding was observed in 10% (n = 11) of patients. Outcomes for HCC patients treated with A+B in Canada were comparable to IMbrave150. There was no increase in GI bleeding in patients without pre-treatment EGD, possibly supporting a selective EGD approach.