Efficacy of Lenvatinib versus Bevacizumab in Triple Therapy with Immune Checkpoint Inhibitors and Interventional Therapy for Hepatocellular Carcinoma with Macrovascular Invasion: A Multi-Center Real-World Retrospective Study.

BACKGROUND: Hepatocellular carcinoma (HCC) patients with macrovascular invasion (MaVI) continue to experience a dismal prognosis and the optimal drug combination choices remain undefined. The aim of this study was to compare the clinical efficacy and safety of lenvatinib-based triple therapy versus bevacizumab-based triple therapy, as first-line treatment for patients with MaVI. PATIENTS AND METHODS: This retrospective multicenter study enrolled 184 consecutive patients from three centers in China, all of whom received first-line lenvatinib (n=126) or bevacizumab (n=58) combined with immune checkpoint inhibitors (ICIs) and interventional therapy (TACE and/or HAIC). The primary endpoints were progression-free survival (PFS) and overall survival (OS). Propensity score matching (PSM) was applied to balance baseline clinical characteristics, and inverse probability of treatment weighting (IPTW) was subsequently applied to assess the robustness of the findings. RESULTS: In the overall cohort, median PFS did not differ significantly between the lenvatinib group and the bevacizumab group (hazard ratio [HR]: 0.694; 95% confidence interval [CI]: 0.477-1.008; P = 0.053). However, the PFS difference reached statistical significance after PSM (8.40 months in the lenvatinib group vs 13.00 months in the bevacizumab group; HR: 0.601; 95% CI: 0.389-0.928; P = 0.02) and after IPTW (HR: 0.609; 95% CI: 0.415-0.894; P = 0.011). No significant difference in OS was observed between the two groups in the unadjusted cohort (HR: 0.809; 95% CI: 0.512-1.279; P = 0.362), after PSM (20.23 vs 35.20 months; HR: 0.634; 95% CI: 0.376-1.069; P = 0.085), or after IPTW (HR: 0.686; 95% CI: 0.432-1.089; P = 0.11). Multivariate analysis in the matched cohort showed that bevacizumab-based triple therapy was associated with prolonged PFS compared with lenvatinib-based triple therapy. There was no significant difference between the two groups in the incidence of grade 3-4 adverse events (58.57% vs 66.00%, P = 0.409) and grade 3-4 gastrointestinal hemorrhage (12.86% vs 8.00%, P = 0.399). CONCLUSION: In HCC patients with MaVI, bevacizumab-based triple therapy was associated with prolonged PFS compared to lenvatinib-based triple therapy, with no significant difference in OS.