Efficacy and Safety of TACE-HAIC Combined with Molecular Targeted Therapy and Immune Checkpoint Inhibitors for Unresectable Hepatocellular Carcinoma

Introduction

As the third leading cause of cancer-related mortality worldwide, hepatocellular carcinoma (HCC) is often diagnosed at intermediate or advanced stages, rendering most patients ineligible for curative treatment.¹ Recent guideline and research on the management of unresectable hepatocellular carcinoma (uHCC) show that locoregional therapies, such as transarterial chemoembolization (TACE), hepatic arterial infusion chemotherapy (HAIC), ablation, and radiotherapy, play a key role.^2,3 At the same time, current guideline and clinical studies have demonstrated significant progress in systemic therapy for uHCC, especially with the combination of molecular targeted therapy (MTT) and immune checkpoint inhibitors (ICIs).^4–6 TACE achieves local tumor control through selective arterial embolization, which induces tumor necrosis. However, this approach can trigger compensatory responses, including the development of collateral tumor vessels and up-regulation of hypoxia-inducible factor 1-alpha (HIF-1α). This promotes abnormal angiogenesis and fosters an immunosuppressive tumor microenvironment, ultimately contributing to disease progression and resistance to TACE.^7,8

In recent years, the integration of TACE with systemic therapies has become increasingly recognized and utilized in the management of intermediate and advanced uHCC.⁹ A meta-analysis has demonstrated that the combination of TACE/HAIC with targeted therapy and immunotherapy significantly improves survival outcomes and tumor response compared with targeted therapy combined with immunotherapy alone, with a manageable safety profile.¹⁰ Several studies have shown that TACE combined with molecular targeted therapy and immune checkpoint inhibitors is superior to TACE alone in terms of PFS. However, the LEAP-012 trial did not show a statistically significant OS benefit, and EMERALD-1 study has not published mature OS results yet.^11–13 Although the triple combination therapy has achieved significant breakthroughs over TACE alone for intermediate and advanced hepatocellular carcinoma, it still has limitations. Its efficacy is often diminished in cases involving massive tumors, portal vein tumor thrombus (PVTT), or multifocal intrahepatic lesions.

HAIC is a key locoregional therapy for HCC, providing long-lasting, high-concentration chemotherapy through intra-arterial infusion to improve antitumor effects.^14,15 And in high burden HCC, combination therapy including HAIC leads to better therapeutic effects.^16,17 Whether adding HAIC to the current triple regimen therapy of TACE combined with MTT and ICIs (TTI) confers additional clinical benefit in uHCC remains unclear. To address this, we performed a two-center retrospective cohort study comparing the efficacy and safety of a quadruple regimen therapy of TACE-HAIC combined with MTT and ICIs (THTI) versus the TTI triple regimen, aiming to provide evidence for optimizing treatment strategies in this population.

Materials and Methods

Study Design and Patient Selection

We conducted a retrospective, two-center cohort study of consecutive patients with unresectable hepatocellular carcinoma treated at the School of Clinical Oncology of Fujian Medical University and Nanchong Central Hospital from September 2019 to August 2024. Patients were assigned to either the THTI group, defined as those who received TACE-HAIC plus MTT and ICIs, or the TTI group, defined as those who received TACE-HAIC plus MTT and ICIs. Propensity score matching (PSM) with 1:1 ratio was done to balance baseline factors and form comparable cohorts. The Independent Ethics Committee of the Clinical Oncology School of Fujian Medical University, Fujian Cancer Hospital (No. K2026-168-01) approved this retrospective study, as did the Ethics Committee of Beijing Anzhen Nanchong Hospital, Capital Medical University & Nanchong Central Hospital (No. 2025–207). Informed consent was waived because the study is retrospective. Inclusion criteria were: (1) conclusion criteria were: (1) histopathologically confirmed HCC or diagnosis according to clinical guidelines²˒⁴; (2) determined unresectable by multidisciplinary team (MDT) due to high tumor burden, insufficient future liver remnant, or distant metastasis; (3) Barcelona Clinic Liver Cancer (BCLC) stages B or C; (4) Child-Pugh class A or B; (5) Eastern Cooperative Oncology Group (ECOG) performance status 0 or 1; and (6) with at least one measurable intrahepatic lesion as defined by the modified Response Evaluation Criteria in Solid Tumors (mRECIST). Exclusion criteria were: (1) contraindication to TACE or HAIC procedure; (2) a history of other active malignancy; (3) severe cardiac, pulmonary, renal, or hepatic dysfunction; (4) Child-Pugh class C or ECOG PS >1; (5) incomplete follow-up data.

Treatment Protocols

TACE procedure was performed using the Seldinger technique to puncture and catheterize the femoral artery. Superselective catheterisation of the tumor feeding artery (ies) and chemotherapeutic agents (epirubicin 30mg/oxaliplatin 50mg) infusion. Embolization with 10–20mg epirubicin and 5–20mL of lipiodol. If required, it will be supplemented with gelatin sponge particles and/or embolic microsphere.

HAIC procedure: A microcatheter was inserted into the primary tumor feeding artery after TACE to administer FOLFOX (oxaliplatin 85 mg/m² over 2–3 hours, calcium folinate 400 mg/m² over 2 hours, fluorouracil 400 mg/m² bolus and 2400 mg/m² continuous infusion over 46 hours) or RALOX (oxaliplatin 100 mg/m² + raltitrexed 3 mg/m² over 5–6 hours). Repeated TACE-HAIC or TACE were performed at intervals of 3–4 weeks.

For molecular targeted therapy, continuous oral medication was administered at the following doses: Continuous oral medication is used, the corresponding drug doses are: lenvatinib (8mg qd, when body weight≤60kg or 12mg qd, when body weight >60kg), apatinib (0.25g qd), donafenib (0.2g bid), sorafenib (0.4g bid), bevacizumab or bevacizumab analogs (15 mg/kg, once every 3 weeks intravenously). ICIs were intravenously administered as follows: 200 mg of tislelizumab, 200 mg of camrelizumab, 200 mg of sintilimab,1200mg of atezolizumab, or 200mg of pembrolizumab, every three to four weeks.

Outcomes Assessment

The primary endpoints of this study were progression-free survival (PFS) and overall survival (OS). Secondary endpoints consisted of objective response rate (ORR), disease control rate (DCR), and treatment-related adverse events (TRAEs). PFS and OS were defined as the time from treatment initiation to disease progression or death, and to death from any cause, respectively. Adverse events were documented and assessed according to the Common Terminology Criteria for Adverse Events. Radiological examinations were performed after every 1–2 treatment cycles. Two experienced abdominal radiologists read the tumour response on contrast enhanced CT or MRI scans using the mRECIST criteria.

Statistical Analysis

All statistical analyses were conducted with R (version 4.5.0, R Foundation) and SPSS (version 27.0). Categorical data were expressed as counts and percentages. To balance baseline covariates, propensity scores were derived from a logistic regression model, followed by 1:1 nearest neighbor matching.Categorical variables were compared using the chi-squared test or Fisher’s exact test as appropriate. Survival curves were generated using the Kaplan-Meier (K-M) method, and differences between groups were assessed by the Log rank test. Independent prognostic factors were identified via univariate and multivariate Cox proportional hazards regression analyses, with results presented as hazard ratios and 95% confidence intervals. A two-sided P-value < 0.05 was considered statistically significant.

Results

Baseline Characteristics and Propensity Score Matching

Between September 2019 and August 2024, a total of 274 uHCC patients who met the TACE-HAIC criteria were first included. Of these, 156 patients were included in THTI group and 118 patients were included in TTI group for the main analysis (Figure 1). Before PSM, there were significant imbalances in baseline characteristics between the two groups. In the THTI group, there were a higher percentage of patients with Child–Pugh B liver function (32.7% vs 20.3%, P=0.023), lymph node metastasis (65.4% vs 50.8%, P=0.015), tumor diameter >10 cm (44.9% vs 31.4%, P=0.023), and portal vein tumor thrombosis (43.6% vs 30.5%, P=0.027). Therefore, the THTI group had a more advanced disease stage and higher tumor burden at the baseline. In order to minimize selection bias, PSM was conducted in a 1:1 ratio, and ultimately, there were 107 patients that were well-matched in both groups. After PSM, baseline characteristics are well balanced between groups (all P>0.05), including age, sex, ECOG PS, BCLC stage, Child-Pugh grade, hepatitis B virus (HBV) infection, albumin (ALB), alpha-fetoprotein (AFP), tumor number, tumor size, lymph node metastasis, PVTT, and extrahepatic metastasis (Table 1), making it more reliable to compare the following efficacies and safety.

Table 1 Baseline Characteristics of the Study Patients Before and After Propensity Score Matching

Figure 1 Flowchart of patient selection for this study.

Survival Outcomes

Median follow-up was 34.0 (95% confidence interval [CI]: 31.3–36.9) and 33.8 months (95% CI: 31.9–38.8) in the THTI and TTI group, respectively.

Survival analysis showed a significant PFS and OS benefit for the THTI group compared with the TTI group before and after PSM. Before PSM, median PFS was 13.4 months (95% CI: 11.7–15.2) in THTI group compared with 10.9 months (95% CI: 9.5–12.8) in TTI group (hazard ratio [HR]: 0.63, 95% CI: 0.49–0.82, P<0.001) (Figure 2A). After PSM, median PFS benefit was maintained, with 14.0 months (95% CI: 12.8–17.0) in the THTI group compared with 10.6 months (95% CI: 9.4–12.6) in the TTI group (HR: 0.59, 95% CI: 0.45–0.79, P<0.001) (Figure 2B). Before matching, the median OS for THTI and TTI group were 29.7 (95% CI: 27.1–34.3) and 25.4 months (95% CI: 23.5–27.7), respectively (HR: 0.62, 95% CI: 0.45–0.84, P = 0.002) (Figure 2C). After matching, the median OS was 30.8 months (95% CI: 28.3–34.9) in the THTI group versus 24.8 months (95% CI: 23.2–27.5) in the TTI group (HR: 0.56, 95% CI: 0.39–0.81, P=0.001) (Figure 2D).

Figure 2 Kaplan-Meier curves of PFS and OS for THTI and TTI groups. The K-M curves of PFS (A) and OS (B) for the two groups before PSM,and of PFS (C) and OS (D) after PSM.

Among patients in the TTI group, 38 received HAIC-based treatment after disease progression (a group), while 71 did not (b group). Although a group had a significantly shorter median PFS than b group (8.8 vs. 11.4 months, P = 0.043)(Figure 3A), median OS did not differ significantly between the groups (25.5 vs. 24.3 months, P = 0.941)(Figure 3B).The PFS and OS rates analysis is shown in Table S1. Both before and after PSM, patients receiving THTI had significantly higher 1-year PFS rates (55.8% vs 42.4%, P=0.028; 59.8% vs 41.1%, P=0.006) and 2-year PFS rates (18.7% vs 6.8%, P=0.007; 15.0% vs 6.5%, P=0.047) versus TTI group. Both before and after PSM, the 1-year OS rates (89.1% vs 87.3%, P=0.643; 90.7% vs 86.0%, P=0.287) and 2-year OS rates (49.4% vs 46.6%, P=0.652; 47.7% vs 44.9%, P=0.681) were similar between groups, with the THTI group having marginally higher rates than the TTI group.

Figure 3 K-M curves of PFS (A) and OS (B) for a and b groups.

Tumor Response

The tumor responses assessed by mRECIST are summarized in Table 2. Before matching, the THTI group demonstrated superior antitumour activity, with a significantly greater ORR (64.7% vs 48.3%, P=0.006) and DCR (93.6% vs 84.7%, P=0.017). After matching, ORR was 64.5% for THTI group and 44.9% for TTI group (P=0.004), and the DCR was 94.4% for THTI group and 84.1% for TTI group (P=0.015). These data demonstrate that HAIC added to triple regimen therapy markedly enhances the intensity of tumor responses and the ability to control diseases.

Table 2 Best Tumor Response Evaluated by mRECIST before and After PSM

Survival Analysis of Responsive and Non Responsive Subgroups

Patients who achieved CR or PR according to tumour response were categorized as responses (R) and those with SD or PD as non-responses (NR). Among the total patients, 101 were responders to THTI (THTI-R) and 55 were non-responders to THTI (THTI-NR), while 57 were responders to TTI (TTI-R) and 61 were non-responders to TTI (TTI-NR). Importantly, THTI-R group had significantly better median PFS (17.1 vs 13.8 months, P = 0.007) and OS (38.1 vs 27.5 months, P = 0.002) than TTI-R (Figure 4A and B). On the other hand, no survival difference was seen between THTI-NR and TTI-NR patients (median PFS: 8.6 vs 6.9 months, P = 0.619; OS: 21.2 vs 20.1 months, P = 0.781) (Figure 4C and D).

Figure 4 K-M curves of PFS and OS in R and NR groups. The K-M curves of PFS (A) and OS (B) for THTI-R group versus TTI-R group, and of PFS (C) and OS (D) for THTI-NR group versus TTI-NR group.

Univariate and Multivariate Analysis

Cox proportional hazards regression analysis was performed to identify independent prognostic factors. Univariate analysis showed that the following factors were significant for PFS (all p<0.05): Treatment regimen, Child-Pugh grade, BCLC stage, AFP level, lymph node metastasis and tumour size. Multivariate analysis showed that THTI regimen therapy was an independent protective factor for PFS (HR: 0.62, 95% CI: 0.46–0.83, P=0.002), and AFP > 400 ng/mL (HR: 1.47, 95% CI: 1.09–1.98, P=0.012) is an independent risk factor for PFS (Table S2). For OS,univariate analysis showed that the treatment regimen, Child-Pugh grade, ECOG PS, BCLC stage, AFP, lymph node metastasis, tumor size and extrahepatic metastasis were all significantly associated with OS (P<0.05). Multivariate analysis showed that THTI regimen therapy was an independent protective factor for OS (HR: 0.62, 95% CI: 0.43–0.89, P=0.010), and AFP > 400ng/mL (HR: 1.55, 95% CI: 1.07–2.25, P=0.019),and tumor size > 10cm (HR: 0.64, 95% CI: 0.43–0.95, P=0.025) were independent risk factors for OS (Table S3).

Subgroup Analysis

In most predefined subgroups, the THTI regimen markedly improved PFS and OS, with consistent OS benefits observed across the entire cohort. Subgroup analysis found that the THTI regimen was associated with improved PFS and OS in most predefined subgroups. This is also consistent with an overall survival benefit observed in the overall population. However, no statistical improvement in OS was seen in Child-Pugh B or female subgroups. Similarly, no significant PFS benefit was seen for Child - Pugh B, ECOG - PS 1, or BCLC stage B subgroups. These results imply that the THTI schedule might be more suitably recommended to males and patients having more favorable clinical features, namely those having Child-Pugh A hepatic function, ECOG-PS 0, or BCLC stage C illness. The detailed results are shown in the forest plots for PFS (Figure S1) and OS (Figure S2).

Safety

Incidence of TRAEs are shown in Table 3. Both regimens had tolerable safety profiles, most events were grade 1/2 TRAEs. The frequency of most TRAEs was similar across the two groups. All reported events were managed with supportive care or dose modifications as per protocol, and no treatment-related deaths occurred. Different toxicity patterns could be recognized that correlated to the distinct treatment parts. The THTI regimen because of including HAIC, showed higher rate of chemotherapy side effects like nausea (35.3% vs 27.1%, P=0.152), vomiting (26.9% vs 17.8%, P=0.075). On the contrary, the TTI group, with greater stress on TACE-induced embolization, presented an enhanced occurrence of post-embolization sequela. Also higher fever rate was reported (28.8% vs 17.9%, P = 0.033), higher trend of post procedure liver enzymes transients elevation were observed, with higher aspartate aminotransferase (AST) elevation trend (71.2% vs 60.9%, P = 0.076) and alanine aminotransferase (ALT) (60.2% vs 51.3%, P = 0.143).

Table 3 Treatment-Related Adverse Events

Discussion

This study revealed that the THTI quadruple therapy provides a significant clinical benefit over current TTI triple therapy. After PSM, the significant PFS advantage of the THTI group over the TTI group persisted, with a median PFS of 14.0 months versus 10.6 months (P < 0.001). And the THTI group achieved significantly higher 1-year and 2-year PFS rates, indicating a sustained long-term disease control benefit. Median OS was also significantly prolonged in the THTI group compared with the TTI group (30.8 vs. 24.8 months, P = 0.002).While the 1-year and 2-year OS rates were numerically higher in the THTI group, these differences did not reach statistical significance. Two factors may explain the discordance between the significant OS improvement and the lack of difference in landmark survival rates. First, a proportion of patients in the TTI group received TACE-HAIC as salvage therapy following disease progression, a treatment strategy which may have extended their overall survival.^18,19 This observation also indirectly supports the efficacy of introducing HAIC in cases of TACE resistance. Second, the number of death events in the THTI group remained relatively low at the time of analysis, suggesting that continued follow-up may further elucidate potential OS rate differences. Moreover, the THTI group achieved significantly higher ORR and DCR than the TTI group.

Previous studies have shown that TACE-HAIC combined with targeted therapy and ICIs is effective for uHCC.^20–22 However, these were mostly single center experiences and not compared with the triple combination. Building on this evidence, this multicenter study shows that HAIC addition to triple regimens significantly and further prolonged survival and deepened responses. Although the combination therapies were subject to higher risks of adverse events as compared with monotherapies, the safety profile seen during the research was considered acceptable Grade 3/4 TRAEs were not different from each other and did not differ significantly among THTI and TTI groups. The toxicity profiles differed between the two groups: the THTI group showed a higher tendency for chemotherapy-related toxicities (nausea, vomiting), while the TTI group exhibited a higher incidence of post-embolization syndrome (fever, transient transaminitis). The majority of events were managed with supportive care or dose adjustment which suggests the quadruple regimen is safe to administer at experienced centers.²³

This study also found that HAIC plays a key role in the clinical benefits of the quadruple therapy.Survival analysis stratified by treatment response revealed that achieving an objective tumor response (CR/PR) early in the treatment course served as a crucial predictor for long-term survival benefit. More importantly, this study has shown that the most important advantage of adding the HAIC to the triple therapy was the potential to produce a larger magnitude of response and extend survival even more among treatment responders. Among patients who did not initially have an objective response to treatment, survival outcomes were no significant difference between THTI-NR and TTI-NR. These findings demonstrate that the addition of HAIC to triple combination therapy confers a clinically meaningful survival benefit and enhances tumor response, particularly among patients with favorable characteristics for intensive treatment.

The mechanistic basis for the superiority of the quadruple regimen lies in the synergistic interplay between locoregional and systemic therapies. TACE causes tumor ischemia and necrosis through occlusion of tumor-feeding arteries. However, TACE procedure also upregulates factors such as HIF-1α and vascular endothelial growth factor (VEGF), which promote an immunosuppressive tumor microenvironment and facilitate the formation of collateral circulation, ultimately contributing to tumor recurrence and treatment resistance.^24,25 By delivering continuous, high-concentration and intra-arterial chemotherapy to the liver, HAIC can overcome TACE resistance, mitigate rapid drug washout, and target residual tumor and neovascularization after TACE.^7,14 The combination of TACE and HAIC facilitates rapid tumor debulking while prolonging tumor exposure to chemotherapeutic agents, which may enhance the subsequent efficacy of molecular targeted therapy.^26,27 Molecular targeted therapy acts via VEGFR and FGFR to produce direct anti-angiogenic activity, suppress VEGF-dependent neovascularization following TACE, and interrupt tumor growth signaling pathways.^28,29 ICIs function by blocking inhibitory signals that suppress T-cell activity, thereby enhancing the antitumor immune response primed by prior HAIC and TACE treatments. Collectively, the superior efficacy of the quadruple regimen is derived from the synergistic mechanisms of these four modalities working together.

The critical role of HAIC in enhancing quadruple therapy efficacy may be mechanistically linked to its ability to induce the formation of tertiary lymphoid structures (TLS). Tertiary lymphoid structures are ectopic, organized aggregates of immune cells that develop in non-lymphoid tissues, including tumor.³⁰ The presence of TLS in solid malignancies, including HCC, is usually associated with a good prognosis.^31,32 A variety of therapies, including chemotherapy and immunotherapy, have been shown to trigger TLS neogenesis. Recent studies have provided evidence that HAIC promotes TLS formation in HCC, and TLS-positive tumors patients often show better treatment response and PFS.³³ These regions constitute the spatial niches for CD4+ T cells, dendritic cells, and B cells, whose formation correlates with the proliferation and tumour infiltration of CD8+ T cell subsets following HAIC therapy.³⁴ Consequently, the superior survival benefit observed in the THTI-R group compared with the TTI-R group may be attributed to the dual effect of HAIC in inducing TLS neogenesis. This proposed mechanistic link needs additional supporting validation through foundational and translational study.

This study has several limitations. First, despite the use of PSM, inherent selection bias due to its retrospective design cannot be completely excluded, and the two-center sample limits the generalizability of our findings. Prospective multicenter randomized trials (eg. NCT06904183, NCT06031285) are needed. Second, the short follow-up resulted in immature OS data, with only a minority reaching the endpoint in the THTI group, warranting longer observation. Third, although we hypothesize that HAIC may enhance therapeutic efficacy by promoting TLS formation, this study did not provide direct mechanistic evidence. Future translational research utilizing paired pre- and post-treatment tissue specimens is needed to elucidate the underlying biological mechanisms.

Conclusion

For patients with uHCC, the THTI quadruple therapy provides a superior clinical benefit over current TTI triple therapy, as evidenced by significantly prolonged PFS and OS, along with higher ORR and DCR and a manageable safety profile. Importantly, this therapeutic advantage persists even in patients with high tumor burden.