Despite rapid advances in systemic therapy, the outcome of uHCC remains unsatisfactory. The locoregional therapies continue to play a critical role in the comprehensive management of this patient population.14 Our group has previously reported that HAIC with the FOLFOX regimen, as well as comprehensive HAIC-based treatment strategies, significantly improves the prognosis of uHCC.10,15 However, it remains unclear which systemic therapeutic agent would be the optimal partner for HAIC. Lenvatinib has shown superior efficacy among first-line TKIs for uHCC, coupled with a manageable adverse event profile. Separately, the anti-PD-L1 agent durvalumab has also demonstrated promising efficacy and safety in prior studies.16 Therefore, we conducted this prospective, single-armed, phase 2 trial to evaluate the efficacy and safety of FOLFOX-HAIC combined with lenvatinib and durvalumab in uHCC, with the aim of establishing a more effective and better-tolerated treatment strategy.
In a study previously reported by our team, among patients with unresectable hepatocellular carcinoma (largest diameter ≥ 7 cm) without macrovascular invasion or extrahepatic spread, the FOLFOX-HAIC group (n = 159) demonstrated an ORR of 46%, a median PFS of 9.6 months, a median OS of 23.1 months, and an incidence of serious adverse events of 19%.9 In the present study, FOLFOX-HAIC combined with lenvatinib and durvalumab significantly enhanced the efficacy, achieving an ORR of 75.0% and a CR rate of 22.5% according to mRECIST criteria. Even in this cohort with advanced tumors, the conversion surgery rate was 17.5% (7/40). Notably, 42.9% (3/7) of these patients attained pCR. Meanwhile, this combination treatment strategy also provided a substantial survival benefit, with a median PFS of 15.8 months; the 6-month PFS rate was 77.5% (95% CI 63–90%), and the 1- and 2-year OS rates were 97.5% and 94.0%, respectively. In the EMERALD-1 trial, 25.8% of patients were BCLC stage A, 57.3% were stage B, and 16.1% were stage C. The combination of TACE with durvalumab and bevacizumab yielded an ORR of 43.6% and a median PFS of 15.0 months.17 The LEAP-012 trial reported an ORR of approximately 60% and a median PFS of 14.6 months with the combination of TACE, lenvatinib, and pembrolizumab.18 In the CARES-005 trial, TACE combined with camrelizumab and apatinib achieved an ORR of 65.0% and a median PFS of 11.0 months.19 By comparison, our triplet regimen achieved a higher ORR and longer median PFS than the above study. The high rate of intrahepatic tumor control can be attributed to the locoregional efficacy of HAIC, while prolonged survival reflects the durable clinical benefit and long-term effects of systemic agents, particularly immunotherapy.11 Although PFS significantly differed across response categories (CR/PR versus SD/PD and CR/PR/SD versus PD), no statistically significant difference in OS was observed. The potential reasons for the lack of difference in OS may largely be attributed to the impact of subsequent therapies after disease progression, as well as the relatively short follow-up duration, with a median follow-up of 23.1 months and only 3 death events occurring by the end of the study.
Our previous studies have revealed that HAIC has fewer adverse effects than TACE.10 Among tyrosine kinase inhibitors (TKIs) used for HCC, lenvatinib is recognized for a relatively manageable safety profile. Meanwhile, as an anti-PD-L1 agent, durvalumab’s adverse event spectrum is often considered comparable to, although potentially distinct from, that of anti-PD-1 therapies. The results of the present study showed that 34 (85.0%) patients had mild adverse events (AEs) (grade 1-2). Moreover, the most common grade 3-4 AEs were elevated alanine aminotransferase and thrombocytopenia, both of which were manageable with appropriate interventions. No patient developed hepatic decompensation (including ascites, hepatic encephalopathy, or variceal bleeding) during the treatment period. This overall favorable safety profile might be attributed to our active dose adjustment of lenvatinib and durvalumab, as well as the strategy of on-demand HAIC treatment.
As an early-stage trial, this study aimed to evaluate the preliminary efficacy and safety of the combined regimen. The original protocol was planned for 36 participants; however, the rate of treatment completion was slightly lower than expected. To ensure sufficient statistical power and with approval from the ethics committee, the sample size was increased to 40 participants. Post hoc power analysis based on the actual enrollment demonstrated a statistical power of 83.8%, which is considered adequate for exploratory studies. Furthermore, a sensitivity analysis excluding the last four participants still yielded consistent results, supporting the robustness of the findings (Supplementary Table S7, Supplementary Fig. S5).
The present study also has some limitations. First, it was a single-arm, phase 2 trial with a small sample size. Therefore, our findings require validation in a prospective, large-scale, phase 3 randomized controlled trial. Second, all enrolled patients had well-preserved liver function (Child‒Pugh A). The safety and tolerability of the combination regimen in patients with more advanced liver dysfunction remain to be evaluated. Third, the quality of life data were not collected, as this was not included as an endpoint in the protocol. Moreover, since the majority of patients had evidence of HBV infection, external validation in cohorts with other underlying etiologies is necessary. Finally, due to the single-arm design of this study, it remains uncertain whether the observed efficacy of this combination regimen is attributable to an additive or synergistic effect. We will incorporate translational analyses, such as biomarker exploration, to better elucidate the underlying mechanisms in subsequent Phase 3 clinical trials.
In summary, the regimen comprising FOLFOX-HAIC, lenvatinib, and durvalumab demonstrates encouraging therapeutic outcomes and an acceptable safety profile in patients with unresectable hepatocellular carcinoma. Confirmation of these findings through a phase 3 randomized controlled trial is necessary.
