Answers are grounded in this paper and cite their sources.
Introduction: Atezolizumab plus bevacizumab (Atezo/Bev) improves overall survival (OS) and progression-free survival (PFS) in hepatocellular carcinoma (HCC). Whether adding on-demand transcatheter arterial chemoembolization (TACE) to Atezo/Bev (ABC-TACE) adds benefit, and whether this depends on the tumor response already achieved, remains unclear. Methods: We retrospectively analyzed 567 patients with unresectable HCC given Atezo/Bev at 16 Japanese institutions. Those who underwent on-demand TACE during Atezo/Bev treatment formed the ABC-TACE group (n = 90) and those who did not the Atezo/Bev-alone control group (n = 477). Because TACE was performed at a variable time (median 199 days after initiation), the primary analysis entered TACE as a time-dependent covariate, with landmark analyses at 3, 6, and 9 months. An exploratory analysis stratified the cohort by the response documented before TACE: partial response (PR), stable disease (SD) with or without shrinkage, and progressive disease (PD). Adverse events were graded by CTCAE v5.0, and the albumin-bilirubin (ALBI) score was compared by mixed-effects modeling. Results: Adjusted for ten prespecified covariates, TACE was associated with longer OS (hazard ratio [HR] = 0.48, 95% confidence interval [CI]: 0.32–0.71; p < 0.001) and PFS (HR = 0.62, 95% CI: 0.42–0.92; p = 0.018). Assigning exposure at any time, which ignores immortal time, gave HRs of 0.37 and 0.34. Landmark analyses were concordant in direction. Grade ≥3 adverse events (27.8% versus 35.2%; p = 0.184) and ALBI trajectories (ever- vs never-TACE; p = 0.198) did not differ significantly. Stratified by pre-TACE response (exploratory), OS HRs were 0.12 for PR, 0.42 and 0.53 for the two SD strata, and 1.51 for PD (interaction p = 0.020). The estimates are imprecise (only 2 deaths in the PR stratum followed embolization), the interaction did not persist when the PD stratum was excluded (p = 0.097), and PFS showed no corresponding interaction (p = 0.589). Conclusion: In this real-world cohort, ABC-TACE was associated with longer survival than Atezo/Bev alone. The response-stratified analyses are exploratory and do not establish effect modification. These findings are hypothesis-generating rather than practice-changing, and prospective response-stratified evaluation is required before a response-guided approach can be adopted.
No takes yet. Share an insight, caveat, or question.
Kuroda et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: