Efficacy and safety of sintilimab plus bevacizumab with or without transarterial chemoembolization for unresectable hepatocellular carcinoma: a propensity score-matched retrospective cohort study

Background To compare the efficacy and safety of transarterial chemoembolization (TACE) plus sintilimab and bevacizumab (triple therapy) versus sintilimab and bevacizumab alone (dual therapy) in unresectable hepatocellular carcinoma (uHCC). Methods This single-center retrospective cohort study enrolled consecutive patients with uHCC who received triple or dual therapy between October 2021 and May 2025. Propensity score matching (PSM) was performed using logistic regression (1:1 nearest-neighbor, caliper: 0.02 SD). The primary endpoint was progression-free survival (PFS). The secondary endpoints included objective response rate (ORR), disease control rate (DCR), overall survival (OS), and safety. An exploratory prognostic score for PFS was developed in the matched cohort and used to define risk strata for hypothesis-generating analyses. Results After PSM, 40 patients were included in each group. The triple therapy group achieved a longer median PFS than the dual-therapy group (13.11 vs. 8.57 months; log-rank P = 0.026). The median OS was not reached (log-rank P = 0.20). The ORR (75.0% vs. 60.0%; P = 0.233), DCR (95.0% vs. 82.5%; P = 0.154), and radiographic downstaging rate (17.5% vs. 5.0%; P = 0.154) were numerically higher with triple therapy. The prognostic model demonstrated that, median PFS was 14.46 months, 11.50 months, and 5.62 months in the low-, intermediate-, and high-risk groups, respectively (log-rank P < 0.001). Risk-stratified HR estimates were numerically lowest in the intermediate-risk group; however, the formal treatment-by-risk-group interaction was not statistically significant ( P for interaction = 0.879). Grade 3–4 adverse events were numerically more frequent with triple therapy (37.5% vs. 22.5%; P = 0.222), with no treatment-related deaths. Conclusions In PSM-matched uHCC patients, TACE plus sintilimab-bevacizumab was associated with longer PFS, with generally manageable toxicity but a numerically higher incidence of grade 3–4 adverse events. Exploratory risk-stratified analyses generated hypotheses regarding potential differences in treatment benefit across prognostic strata.