TACE plus systemic therapy in advanced hepatocellular carcinoma: a comparison of lenvatinib and donafenib

Background To evaluate transcatheter arterial chemoembolization (TACE) combined with systemic therapy versus TACE monotherapy in advanced hepatocellular carcinoma (HCC), and to compare the efficacies of lenvatinib and donafenib when combined with TACE and immune checkpoint inhibitors (ICIs). Methods This retrospective study included 176 patients with advanced HCC receiving initial TACE, categorized into combination therapy (n = 128) or TACE monotherapy (n = 48) groups. A subgroup analysis compared TACE plus ICIs combined with lenvatinib (n = 78) or donafenib (n = 33). Propensity score matching (PSM) was utilized to minimize baseline confounding variables. Progression-free survival (PFS) and overall survival (OS) were analyzed. Results Combination therapy achieved a significantly higher objective treatment effective rate than monotherapy (75.0% vs. 41.7%, P < 0.001), with significantly prolonged PFS and OS, a benefit maintained after PSM. Among the specific regimens, the lenvatinib-based combination yielded a higher objective effective rate than donafenib (83.3% vs. 63.6%, P = 0.023). Post-PSM, the lenvatinib group maintained a significant PFS advantage ( P = 0.007), while OS showed a marginal trend toward improvement ( P = 0.057). Multivariate analysis identified treatment without combined systemic therapy and ineffective treatment response as independent predictors of inferior OS. Severe TACE-related adverse events (Grade 3/4) were comparable between combination and monotherapy cohorts (18.8% vs. 14.6%, P = 0.518), as well as between the lenvatinib and donafenib subgroups (21.8% vs. 15.2%, P = 0.422). Conclusion TACE combined with systemic therapy significantly improves clinical outcomes in advanced HCC without increasing severe toxicity. Furthermore, the triple therapy of TACE, ICIs, and lenvatinib confers superior PFS and tumor response rates compared to the donafenib-based regimen, along with a trend toward OS improvement that did not reach statistical significance.