Determinants of long-term survival from atezolizumab plus bevacizumab in unresectable hepatocellular carcinoma

Background: Despite the proven superiority against sorafenib, atezolizumab plus bevacizumab (A+B) lacks long-term efficacy data in unresectable hepatocellular carcinoma (uHCC). This study assessed clinicopathologic factors associated with long-term survival with A+B. Methods: We analyzed patients receiving first-line A+B within AB-Real, a large multicenter registry across Europe, Asia, and the USA. Long-term survivors (LTS) were defined as patients surviving at 24 months. Landmark survival outcomes and associations between baseline clinicopathologic factors and overall survival (OS) were evaluated. Results: Of 1346 patients enrolled, 1085 with Child-Pugh A and ECOG 0-1 received first-line A+B. The median OS was 19.2 months (95% CI: 17.6-20.9), and 2- and 3-year survival rates were 41.4% and 29.3%. Among 695 patients with adequate follow-up, 190 were classified as LTS. The overall response rate was 24.7%, and long-term survivorship was significantly enhanced in radiologic responders (52.2% vs 8.8%, p < 0.001). Compared to non-LTS, LTS more frequently had ALBI grade 1 liver reserve (64.6% vs 38.5%), less PVI (17.9% vs 36.4%), smaller maximum tumor size (3.8 vs 7.0 cm), and lower AFP (median 23.0 vs 240.8 ng/mL) (all p < 0.001). Notably, nearly half (48.8%) of radiologic responders who did not achieve long-term survival had ≥ 2 adverse baseline features, underscoring that tumor burden and liver reserve remain prognostically relevant even among responders. Conclusions: AB-Real provides the first global real-world evidence of long-term efficacy of A+B in uHCC. Long-term survivorship is enhanced in radiological responders and strongly associated with pre-treatment tumor factors and liver function.