First-line immunotherapy for advanced hepatocellular carcinoma: a network meta-analysis of randomized trials with overall and HBV/HCV-stratified efficacy and safety.

Background: Immune checkpoint inhibitors (ICIs) have revolutionized the front-line treatment of advanced hepatocellular carcinoma (HCC). However, the comparative efficacy and safety of different ICI-based regimens-and their consistency across etiologic subgroups [HBV, HCV, and non-HBV/non-HCV (NBNC)]-remain uncertain. This study conducted a Bayesian network meta-analysis (NMA) of recent randomized controlled trials (RCTs) to compare the first-line immunotherapy strategies both overall and by viral etiology. Methods: A systematic search of PubMed, Embase, the Cochrane Library, and Web of Science was conducted from inception through August 1, 2025. Prespecified primary outcomes were overall survival (OS) and progression-free survival (PFS); secondary outcomes included objective response rate (ORR) and grade ≥3 treatment-related adverse events (TRAEs). Hazard ratios (HRs) and odds ratios (ORs) were pooled in a Bayesian NMA contrasting ICI-based regimens with tyrosine kinase inhibitor (TKI) monotherapy in the overall advanced HCC population and in HBV, HCV, and NBNC strata. Protocol registration: PROSPERO CRD420251131167. Results: Ten RCTs (n=7,301) evaluating 14 first-line regimens were included. Compared to TKI monotherapy, immunotherapy significantly improved OS (HR = 0.79, 95% CI 0.74-0.84) and PFS (HR = 0.70, 95% CI 0.58-0.85), increased ORR (OR = 3.20, 95% CI 2.49-4.12), and did not significantly increase grade ≥3 TRAEs (OR = 1.16, 95% CI 0.75-1.80). Benefits were more pronounced in HBV-positive patients (OS: HR = 0.73, 95% CI 0.67-0.79; PFS: HR = 0.54, 95% CI 0.48-0.61). HCV-positive patients also derived an OS benefit (HR = 0.82, 95% CI 0.72-0.93), whereas PFS improvement in NBNC patients was not statistically significant (HR = 0.74, 95% CI 0.52-1.06). In the NMA, sintilimab plus a bevacizumab biosimilar (Sinti-Bev) showed the greatest OS improvement versus sorafenib (HR = 0.57, 95% CI 0.43-0.75), followed by camrelizumab plus rivoceranib (Camre-Rivo; HR = 0.62, 95% CI 0.49-0.80). For PFS, anlotinib plus penpulimab (Anlo-Penpu) and Camre-Rivo ranked the highest (both HR = 0.52, 95% CI 0.41-0.66). Tislelizumab and nivolumab were associated with lower risks of grade ≥3 TRAEs. In etiology-stratified network analyses, atezolizumab plus bevacizumab (Atezo-Bev) showed the most consistent efficacy profile across virally mediated subgroups, with significant advantages for both OS and PFS in HBV-positive disease and a significant OS advantage in HCV-positive disease. However, no regimen achieved a statistically significant PFS advantage in the HCV-positive subgroup. Conclusions: First-line immunotherapy, compared to TKI monotherapy, provides meaningful survival benefits in advanced HCC, particularly in virally mediated disease, without significantly increasing severe toxicity. Among regimens, Sinti-Bev offers the most substantial OS advantage, while Anlo-Penpu and Camre-Rivo rank highest for PFS. Etiology-stratified analyses highlight Atezo-Bev as the most consistent regimen in virally mediated disease, with significant OS and PFS benefits in HBV-positive patients and a significant OS benefit in HCV-positive patients. Systematic Review Registration: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD420251131167, identifier CRD420251131167.