Abstract
Atezolizumab plus bevacizumab (Ate/Bev) and lenvatinib (Len) are first-line therapies for unresectable hepatocellular carcinoma (uHCC). However, Ate/Bev's high cost limits its common use in real-life practice, while Len is usually covered by national health insurance (NHI). We conducted this study to compare their effectiveness and safety in real-world settings. We retrospectively evaluated 346 uHCC patients treated with first-line Ate/Bev (n=80) or Len (n=266) from December 2019 to December 2022, using 1:2 ratio propensity score matching (PSM) analyses. Compared to the Len group, the Ate/Bev group exhibited higher incidences of Child-Pugh class B (14.1% vs. 5.7%, P=0.014), larger main tumors (58.8% vs. 40.2%, P=0.003), and more main portal vein invasion (25% vs. 12.8%, P=0.008). Treatment-related adverse events were notably lower in the Ate/Bev group (56.3% vs. 72.3%, P=0.007). After PSM, no significant differences were observed in the objective response rate (21.9% vs. 21.6%, P=0.983), progression-free survival (5.1 vs. 6 months, P=0.783), and overall survival (13.3 vs. 14.1 months, P=0.945) between the Ate/Bev (n=73) and Len (n=142) groups. Patients in the Ate/Bev group received more sequential post-treatments compared to the Len group (45.2% vs. 24.6%, P=0.009). Len-based therapies (n=28, 84.8%) and mono- or combined-immunotherapy (n=19, 54.3%) were the most frequently administered sequential therapies following Ate/Bev and Len, respectively. Patients with uHCC who received first-line self-paid Ate/Bev seemed to have lower liver function reserve and more advanced tumor characteristics compared to those who underwent NHI-reimbursed Len. However, the treatment outcomes and safety profiles were similar between these two groups.