Breaking through the transarterial chemoembolization resistance barrier: Reshaping the treatment path for advanced liver cancer with triple therapy

Abstract

In this article we commented on an article published recently by Jiao et al. This retrospective study confirmed that the triple therapy of transarterial chemoembolization (TACE) combined with programmed death protein ligand 1 inhibitors and molecular targeted therapy can significantly reverse TACE resistance in advanced hepatocellular carcinoma. Compared with TACE alone, the triple therapy reduced the resistance rate from 38.8% to 9.7% and increased the median progression-free survival and median overall survival by 92.3% and 26.8%, respectively. TACE induces tumor antigen release and upregulates programmed death protein ligand 1, activating the effect of immune checkpoint inhibitors while molecular targeted therapy inhibits postembolization vascular regeneration, forming a dynamic synergistic network of embolization-immune activation-vascular inhibition. The maximum tumor diameter, capsule loss, and bilateral distribution were identified as independent predictors. This study provided level I evidence and promoted the transformation of advanced hepatocellular carcinoma treatment from single local intervention to an integrated model of local control-systemic treatment. In the future it will be necessary to analyze the dynamic evolution rules of the tumor microenvironment through cross-omics strategies, further explore biomarkers, optimize treatment sequences, and conduct multicenter prospective trials to verify long-term survival benefits and guide the optimization of individualized sequential treatment.

Key Words: Transarterial chemoembolization resistance; Immune checkpoint inhibitors; Molecular targeted therapy; Hepatocellular carcinoma; Comment

Core Tip: This article discussed a research article published recently focusing on the synergistic value of triple therapy of transarterial chemoembolization combined with programmed death protein ligand 1 inhibitors and molecular targeted therapy to overcome transarterial chemoembolization resistance in advanced liver cancer. The article covered relevant topics and provided insights into the application of the triple therapy in clinical practice.

INTRODUCTION

Hepatocellular carcinoma (HCC), the sixth most common malignant tumor and the third-leading cause of cancer-related deaths worldwide, poses significant therapeutic challenges: 70% of patients are diagnosed at an advanced stage, losing the opportunity for surgical cure[1]. The 5-year survival rate for advanced HCC is < 12%, indicating a substantial unmet clinical need[2,3].

Transarterial chemoembolization (TACE), a cornerstone local treatment for unresectable HCC, achieves disease control by embolizing the tumor-feeding arteries and releasing local chemotherapy drugs[4]. However, its efficacy is consistently limited by a high rate of drug resistance. The objective response rate (ORR) is only 40%-50%[5]. Drug resistance is common with 30%-50% of patients developing TACE resistance (progression of intrahepatic target lesions after three consecutive TACE sessions within 6 months)[6]. The mechanism of resistance is complex, including hypoxia-induced activation of the hypoxia-inducible factor (HIF)-1α/vascular endothelial growth factor (VEGF) pathway[7] after embolization, promoting angiogenesis, upregulation of programmed death protein ligand 1 (PD-L1) expression to form an immunosuppressive microenvironment, and incomplete embolization due to tumor heterogeneity (especially for lesions ≥ 5 cm or multifocal). The indications are limited with significantly reduced efficacy for patients with Child-Pugh C grade, main portal vein tumor thrombus (VP4), and tumor diameter > 10 cm (ORR < 10%)[6]. The treatment of advanced HCC has evolved from single local therapy (TACE) to a combined local-systemic approach.

Sorafenib was approved in 2007, marking the beginning of targeted therapy for advanced HCC, but the median overall survival (OS) was only 10-12 months[8]. The CheckMate 040 study in 2017 confirmed the value of nivolumab as a second-line treatment[9]. The IMbrave150 study established the first-line treatment position of atezolizumab + bevacizumab (median OS 19.2 months)[10]. Recent studies have revealed that the combination of TACE and systemic therapy can overcome drug resistance through synergistic mechanisms[7,8,11,12] including: (1) Immune microenvironment remodeling: Local release of chemotherapeutic drugs by TACE induces tumor antigen exposure, and local hypoxia upregulates HIF-1α, promoting PD-L1 expression[7] while PD-L1 inhibitors block immune checkpoints and restore CD8⁺ T cell cytotoxic function[11]; and (2) Angiogenesis regulation: VEGF expression increases 2.1 times after TACE, and targeted drugs (such as lenvatinib) inhibit the VEGF receptor signaling pathway[12], blocking abnormal angiogenesis and promoting CD8⁺ T cell infiltration. The three form a closed-loop regulation of embolization-immune activation-vascular normalization. However, there are still some challenges, such as a lack of large-sample evidence for the preventive effect of the triple therapy on TACE resistance, controversy over the optimal sequence of administration (TACE and systemic therapy), and the lack of biomarkers to predict drug resistance or efficacy.

Against this background, the study by Jiao et al[13] made a breakthrough contribution. Through a large-sample retrospective analysis (n = 604), it was confirmed that using 1:1 propensity score matching to balance baseline differences (such as tumor burden, proportion of portal VP4), ensuring the reliability of the conclusion. The triple therapy reduced the TACE resistance rate from 38.8% to 9.7% (P < 0.001) with a median OS exceeding 20.8 months (vs 16.4 months for single-agent TACE), shifting from extending survival after resistance to preventing resistance occurrence and moving the treatment threshold forward. Using the progressed intrahepatic target lesions after ≥ 3 TACE sessions conducted within 6 months (based on the Expert Consensus on TACE Resistance in China)[14], it avoids the interference of new lesions or extrahepatic metastasis in the determination of resistance, more accurately reflecting the local efficacy of TACE.

A risk scoring model based on three major drug resistance factors (tumor size, capsule integrity, and biliary distribution) can identify patients who are high risk with TACE resistance at an early stage and guide upgraded combined treatment [such as combination of programmed death 1 (PD-1) inhibitors or enhanced embolization]. The maximum diameter of the tumor (especially when it is ≥ 5 cm) is significantly associated with the formation of collateral circulation[15], reducing the complete embolization rate [odds ratio (OR) = 1.01, P < 0.001]. Capsule loss facilitates cancer cell infiltration along portal vein branches (perivascular invasion), a key mechanism for HCC progression[16], increasing the risk of portal VP4 (OR = 4.52, P = 0.03). Biliary distribution (caused by multifocal lesions, making treatment coverage difficult) reflects the multicentric origin or intrahepatic metastasis tendency of liver cancer[17], leading to incomplete embolization (OR = 2.13, P = 0.003).

It is confirmed that the toxicity of combined treatment is controllable (the rate of grade ≥ 3 adverse events is 19.4% with no treatment-related deaths), and a precise management strategy of reducing the dose of PD-1 inhibitors and adjusting the dose of targeted drugs is proposed. The rationale for combining anti-PD-1 with TACE rests on two sequential biological events: TACE-induced antigen release peaks immediately after the procedure (within 24-48 h), and subsequent immune checkpoint upregulation, including PD-L1, occurs predominantly within 72 h after TACE. Thus, immune intervention should be scheduled to bridge antigen availability with adaptive immune activation to determine an operationally optimal window-balancing biological processes and clinical safety based on data from EMERALD-1 and TALENTACE study. Early intervention with multi-target receptor tyrosine kinase inhibitors (MTT, such as lenvatinib) can block VEGF-mediated immune escape[18]. This study addressed the core clinical problem of TACE resistance through mechanism-driven triple therapy design, converting basic research findings (the regulation of the hypoxia-immune-angiogenesis axis) into clinical solutions, providing a new treatment paradigm for patients with Barcelona Clinic Liver Cancer stage B/C stage disease that combines the depth of evidence-based medicine and clinical operability.

TACE RESISTANCE DILEMMA

TACE is confronted with indication limitations and patient selection dilemmas. TACE requires patients to have liver function at Child-Pugh A or B grade (with limited evidence for B grade), but patients with advanced HCC with liver cirrhosis often cannot tolerate embolization due to liver decompensation (such as ascites and coagulation disorders). In China the proportion of liver cirrhosis caused by hepatitis is high, and the risk of liver function deterioration after TACE is significant. Patients with Child-Pugh C grade are completely excluded from the standard treatment plan. Small tumors (≤ 3 cm) are more suitable for radical treatment (surgery/ablation), and TACE may be overtreatment; for large tumors (> 10 cm) the incomplete embolization rate is as high as 70%, and the ORR is only 5%-10%. For multifocal/diffuse liver cancer embolization is difficult to cover all lesions and is prone to activate the extrahepatic collateral circulation. Main portal VP4 (VP4 grade) is an absolute contraindication for TACE as embolization may aggravate liver failure. Although branch tumor thrombus (VP1-3 grade) can be cautiously attempted, the efficacy is significantly reduced (median OS is only 9-12 months).

TACE faces technical bottlenecks and efficacy attenuation. The blood supply of liver cancer is complex and has spatial heterogeneity, especially for multinodular tumors with multiple feeding arteries, making embolization difficult and the recurrence rate of residual lesions > 50%. Traditional iodized oil + gelatin sponge cannot achieve sustained drug release, and drug-eluting beads TACE increases local drug concentration but increases the rate of grade ≥ 3 adverse events by 30%. The ORR and survival benefit are insufficient with ORR only 40%-50% (Lencioni criteria), and the depth of response is insufficient with complete necrosis rate < 30%. The median OS of patients with mid-stage HCC is 13-48 months, but for tumors > 10 cm the median OS drops sharply to 9.1 months and the 5-year survival rate to < 10%. After multiple TACE sessions liver function progressively deteriorates (for each increase of 1 point in Child-Pugh score, the risk of death increases by 40%), and about 30% of patients discontinue treatment due to liver failure.

TACE has complications and treatment risks. The incidence of postembolization syndrome is > 80%, presenting as fever, abdominal pain, vomiting, etc. It requires glucocorticoids to relieve symptoms, but it may weaken antitumor immunity. Severe organ damage and the incidence of liver abscess/biliary necrosis is 3%-5%, especially for patients with previous biliary surgery. The incidence of liver failure in patients with Child-Pugh B grade is 8.3%, and the mortality rate is 0.6%. Other complications such as pulmonary embolism (0.9%) and gallbladder infarction (2.1%) are directly related to insufficient super selective techniques. The limitations of TACE are essentially the result of the interaction between tumor biological characteristics, technical bottlenecks, and the host microenvironment. Only by integrating technological innovation and mechanism understanding can TACE be upgraded from a palliative tool to a cure bridge.

SYNERGISTIC EFFECT OF TRIPLE THERAPY

The combination of TACE with PD-L1 inhibitors and molecular targeted therapy can synergistically enhance efficacy. In recent years multiple prospective studies have provided solid evidence for the triple therapy of TACE combined with PD-L1 inhibitors and MTT (Table 1). The EMERALD-1 trial[19] was a global multicenter phase 3 study involving 616 patients. It was the first to demonstrate that durvalumab plus bevacizumab followed by TACE significantly prolonged progression-free-survival (PFS) [hazard ratio (HR) = 0.77, P < 0.001], but the lack of concurrent use of antiangiogenic drugs limited the OS benefit (median OS did not reach significance). The CHANCE001 study[20] was a real-world study in China involving 826 patients. The triple regimen (TACE + PD-L1 inhibitor + MTT) significantly prolonged the median OS from 15.7 to 19.2 months (HR = 0.63, P = 0.001), nearly doubled the ORR (60.1% vs 32.0%, P < 0.001), and the rate of grade ≥ 3 adverse events was only 15.8%, laying a foundation for safety. The LEAP-012 trial[21] was a prospective randomized phase 3 clinical study, originally planned to include 950 patients but actually included 480. It innovatively used lenvatinib plus pembrolizumab before treatment (2-4 weeks before TACE), reducing the PFS risk ratio to 0.66 and confirming that concurrent antiangiogenic drug intervention can block the VEGF surge after embolization (increasing by 2.1 times) and reduce vascular escape.

Table 1 Clinical trials for triple therapy in advanced liver cancer.

Trial name	Therapeutic regimen	Sample size	Endpoint results	Limitation
EMERALD-1[17]	TACE + durvalumab ± bevacizumab	616	PFS benefit (HR = 0.77, triple therapy: 15.0 months, single-agent TACE: 8.2 months)	No OS difference, and excluded Child-Pugh B/C and VP4 grades
CHANCE001[18]	TACE + PD-L1 + MTT	826	OS benefit (HR = 0.63, triple therapy: 19.2 months, single-agent TACE: 15.7 months)	Retrospective design; no control of the dosage differences of targeted drugs
LEAP-012[19]	Lenvatinib + pembrolizumab → TACE	480	PFS benefit (HR = 0.66, triple therapy: 14.6 months, single-agent TACE: 10.0 months)	No OS benefit from reduction in sample size and insufficient follow-up time, deficiency in statistical power, and regional differences
TALENTACE[20]	TACE + atezolizumab + bevacizumab	342	PFS benefit (HR = 0.71, triple therapy: 11.3 months, single-agent TACE: 7.03 months)	The on-demand TACE design leads to heterogeneity in treatment frequency; no evolution of drug-resistant clones
Jiao et al[13]	TACE + PD-(L)1 + MTT	144	TACE resistance reduction, OS benefit (HR = 0.76, triple therapy: 20.8 months, single-agent TACE: 16.4 months)	Retrospective study with a small sample size

TACE: Transcatheter arterial chemoembolization; PD-L1: Programmed cell death-ligand 1; MTT: Multi-target receptor tyrosine kinase inhibitors; HR: Hazard ratio; VP4: Main portal vein tumor thrombus; OS: Overall survival; PFS: Progression-free-survival; PD-1: Programmed death 1.

Similarly, some studies have explored breakthroughs in response rates and conversion potential. The TALENTACE study[22] used atezolizumab plus bevacizumab followed by TACE, achieving an ORR of 81.3% (mRECIST) and a median PFS of 11.3 months, providing an efficient conversion possibility for Barcelona Clinic Liver Cancer stage B/C stage patients. The TALENTACE study included patients with liver cancer who met the TACE indication and had not received systemic treatment with a requirement that the sum of the maximum tumor diameter and the number of lesions should be ≥ 6, Eastern Cooperative Oncology Group Performance Status 0-1, and Child-Pugh A grade. Patients were randomly assigned at a 1:1 ratio to two groups: Group A: T + A combination therapy (atezolizumab 1200 mg intravenous every 3 weeks + bevacizumab 15mg/kg intravenous every 3 weeks) was initiated 14 days to 8 weeks after TACE treatment; and Group B: Patients only received TACE treatment as needed. From March 2021 to August 2023, a total of 342 patients were included in the study (171 in group A and 171 in group B). TACE-PFS was significantly prolonged. The median TACE-PFS was 11.30 months in group A, and 7.03 months in group B (HR = 0.71, 95% confidence interval: 0.55-0.92, P = 0.009), reaching the primary endpoint of the study. The ORR was 81.3% in group A and 66.7% in group B. The PROLONG study[23] was a real-world study involving 197 patients. TACE plus lenvatinib ± PD-1 inhibitor achieved an ORR of 48.3%, a disease control rate of 86.2%, and the proportion of patients with tumor necrosis volume > 50% increased to 58%, significantly better than historical TACE monotherapy data (necrosis rate < 30%).

The TRIPLET study[24] was a perioperative phase 2 study. Neoadjuvant TACE plus camrelizumab plus apatinib achieved a conversion resection rate of 17.1% and a pathological complete response rate of 5.9%, creating surgical opportunities for unresectable HCC. Other ongoing explorations of novel treatment strategies for intermediate HCC include: The EMERALD-3 trial evaluating a single dose of tremelimumab combined with regular intervals of durvalumab (STRIDE regimen) ± lenvatinib plus TACE (NCT05301842); the ABC-HCC trial comparing atezolizumab/bevacizumab with TACE (NCT04803994); and the EMERALD-Y90 trial evaluating transarterial radioembolization combined with durvalumab/bevacizumab (NCT06040099).

LIMITATIONS OF TRIPLE THERAPY

Although the triple therapy (TACE + PD-L1 inhibitor + MTT) has shown significant survival benefits, its clinical application still faces challenges. The combination therapy requires patients to have Child-Pugh A liver function (with limited evidence for Child-Pugh B), but in China the proportion of post-hepatitis cirrhosis in patients with HCC is as high as 79.2%, and Child-Pugh B/C patients are excluded due to the risk of liver function deterioration after embolization, accounting for 30%-40% of the potential beneficiary population. The maximum tumor diameter (≥ 5 cm), absence of a capsule, and biliary distribution have been identified as independent predictors of TACE resistance. Even if such patients receive the triple therapy, the ORR is still 40% lower than that of the ideal population. The incidence of grade ≥ 3 adverse events in the triple regimen is as high as 19.4%-38.5%, significantly higher than that of TACE alone (9.7%). Immune-related hepatitis, hypertension, proteinuria, and upper gastrointestinal bleeding require complex dose adjustments[18]. The combination of TACE and PD-1 inhibitors increases the incidence of grade ≥ 3 hepatitis from 4% to 12%[7]. When the interval between PD-L1 inhibitors administration and TACE is < 7 days, the risk of immune-related hepatitis increases while an interval > 21 days may miss the antigen release window. The triple regimen is expensive, increasing medical costs. The definition of drug resistance is not uniform with the Chinese standard (progression after ≥ 3 TACE sessions within 6 months) differing from the Japanese Kudo standard (necrotic volume < 50% after 2 TACE sessions), leading to a 20% deviation in reported resistance rates. Key phase 3 trials had limitations. The EMERALD-1 trial showed no significant improvement in OS with the triple regimen. Although the LEAP-012 trial improved PFS, it excluded patients with portal VP4, limiting the generalization of the results.

Research on the mechanism of drug resistance also faces challenges. HCC has a high degree of intratumoral heterogeneity, and residual tumor cells after TACE show dynamic evolution under clonal selection pressure. Single-cell sequencing shows that the mutation frequency of driver genes such as CTNNB1 and TP53 in different lesions of the same patient can vary by up to 40%, leading to heterogeneous responses to PD-1 inhibitors/targeted drugs[25]. Drug-resistant clones rapidly expand in the hypoxic microenvironment after TACE (VEGF expression increases by 2.1 times within 72 h), but current technologies cannot capture the clonal dynamics in real time. The sensitivity of circulating tumor DNA (ctDNA) detection is only 70%, and it cannot fully reflect clonal evolution. After TACE M2 macrophages promote T regulatory cell proliferation through the interleukin-10/signal transducers and activators of transcription 3 pathway, forming an “immune desert” microenvironment[12]. However, the clinical validation of the co-blocking strategy of co-inhibitory molecules such as T cell immunoglobulin and mucin domain 3/Lymphocyte activation gene-3 is still lacking. Lactate accumulation activates the lactate dehydrogenase A-mediated Warburg effect, inducing CD8⁺ T-cell exhaustion. The lactate dehydrogenase A inhibitor GSK2837808A is effective in preclinical studies, but its liver toxicity limits its clinical application[26]. TACE-induced transforming growth factor (TGF)-β upregulation promotes collagen deposition, reducing drug permeability by 60%. VEGF/TGF-β dual-target inhibitors (such as GFH018) can reverse fibrosis, but the incidence of grade 3 hypertension in phase 3 trials is as high as 28%[27]. Thickening of the tumor vascular basement membrane reduces the tissue concentration of lenvatinib by 50%, and local delivery of drug-loaded microspheres faces a technical bottleneck with an incomplete embolization rate of over 30%[28].

Technical challenges also exist in research. The limitations of dynamic monitoring technology and the difficulty in clinical translation of spatial transcriptomics technology. Stereo-seq can map 3D drug-resistant niches, but the risk of liver biopsy bleeding is 5.7%, limiting its application in advanced patients[29]. The sensitivity of liquid biopsy is insufficient. The detection limit of ctDNA for low-frequency drug-resistant mutations (such as VEGFR2 T790M) is only 0.1%, and the release amount of ctDNA in liver cancer is only one-third of that in colorectal cancer[18].

There is a dilemma in the integration of biomarkers. It is necessary to simultaneously detect multiple-dimensional data such as PD-L1 expression, tumor mutational burden (TMB), and ctDNA mutation spectrum, but the clinical popularization rate is less than 20%[25]. The validation of drug resistance early warning models is insufficient. The area under curve of drug resistance prediction models based on radiomics (such as Radiomics score) drops from 0.92 to 0.75 in external validation[30]. There are adaptability defects in preclinical models. The success rate of patient-derived organoids culture is only 60%, and it cannot simulate the interaction between hypoxia and immune microenvironment after TACE[31]; humanized mice lack a complete liver immune ecosystem, and the response rate to PD-1 inhibitors is overestimated by 30%[12].

FUTURE PERSPECTIVE

This study as a retrospective investigation requires validation through a prospective phase III trial in the future (such as by drawing on the design of the LAUNCH study)[18]. The relatively small sample size (n = 144) may affect the power of the subgroup analysis, and a comparison between TACE + immunotherapy and TACE + targeted therapy should be conducted. In the future, molecular markers such as PD-L1 expression, TMB, and immune cell infiltration should be integrated to achieve more precise patient selection. Single-cell sequencing and spatial transcriptomics should be utilized to analyze the temporal evolution of the tumor microenvironment after TACE (immune cell recruitment within 72 h and the angiogenesis window period from 7 to 14 days) and to determine the optimal time window for intervention with the triple therapy.

Dynamic analysis of epigenetic reprogramming in residual tumor cells after TACE should be conducted to reveal the transcriptional regulatory network of drug-resistant clones; a 3D drug-resistant niche of the tumor microenvironment after TACE should be mapped to locate VEGF/PD-L1 coexpressing drug-resistant nests (spatial niches), guiding local intensified treatment (such as radioembolization combined with CD8⁺ T-cell infusion). ctDNA liquid biopsy should be used to monitor clonal selection under treatment pressure (such as VEGF receptor mutations, PD-L1 copy number amplification), revealing the mechanism of secondary drug resistance. Multiparameter magnetic resonance imaging combined with artificial intelligence algorithms and other radiomics should be used to predict the completeness of embolization, and data from enhanced CT texture features, cfDNA methylation profiles, and peripheral blood T cell receptor diversity should be integrated to build a drug resistance risk warning system.

Further exploration of mechanisms and innovation of therapies can be pursued. For instance, dual immunotherapy (cytotoxic T-lymphocyte-associated antigen-4 + PD-1), T cell immunoglobulin and mucin domain 3/Lymphocyte activation gene-3 bispecific antibodies, and blocking the interleukin-10/signal transducers and activators of transcription 3 signaling pathway may enhance antitumor immunity (such as the HIMALAYA regimen), and individualized vaccines based on TMB combined with the triple therapy can be developed[29]. VEGF/TGF-β dual-target inhibitors (such as GFH018) may overcome the TGF-β-mediated fibrotic barrier after TACE[27], and HIF-2α inhibitors (belzutifan) can block adaptive drug resistance after TACE[25]. Targeting stearoyl-CoA desaturase can inhibit the fluidity of drug-resistant cell membranes and enhance the penetration of lenvatinib. Enhancer of zeste homolog 2 inhibitors (tazemetostat) can eliminate the histone H3, trimethylated lysine 27 epigenetic memory induced by TACE and restore tumor immunogenicity (clinical phase 2 NCT04542850).

Locked nucleic acid drugs targeting circRNA_104797 (VEGF pathway activator) can block the hypoxia adaptive feedback loop[27]. The impact of antibiotic-sensitive microbiota (such as Akkermansia muciniphila) on the efficacy of PD-1 inhibitors should be explored[31], and microbiota transplantation-assisted strategies should be developed. Drug-eluting microspheres (drug-eluting beads TACE) can prolong drug release time[28] and reduce systemic toxicity. The combination of 90Y microspheres and PD-1 inhibitors (NCT04992772 trial) may overcome immune suppression caused by hypoxic microenvironments.

CONCLUSION

The study confirmed that median OS reached 20.8 months in the combination group compared with 16.4 months (95% confidence interval: 11.3-21.5) in the monotherapy group with a statistically significant difference (P = 0.008). It also revealed three major drug resistance drivers from the perspective of tumor biology. The study provided a breakthrough solution to the problem of TACE resistance in advanced HCC. Its value lies in the quantified evidence of survival benefits and in the revelation of the essence of drug resistance mechanisms from the perspective of tumor biology, promoting the transformation of treatment strategies from single local intervention to multimodal coordinated regulation. It promoted three major changes in treatment strategies: (1) From local to systemic: TACE is upgraded from a single local approach to an “ignition point” for immune activation; (2) From empirical to precise: The development of a resistance model based on imaging features enables individualized treatment stratification; and (3) From sequential to coordinated: The optimized sequence (TACE-immunotherapy-targeted therapy) maximizes the synergistic effect. In the future interdisciplinary cooperation (interventional radiology, molecular pathology, artificial intelligence) is needed to explore radiomics prediction models and conduct cost-benefit analyses to accelerate the inclusion of this regimen in international guidelines.