Key Takeaways
Immunotherapy has transformed liver cancer treatment, with immune checkpoint inhibitor combinations improving survival and expanding first-line treatment options for advanced hepatocellular carcinoma (HCC).
HCC remains difficult to treat because tumor heterogeneity, immune resistance, underlying liver disease, and late diagnosis complicate therapeutic development and treatment selection.
New liver cancer treatment strategies are showing mixed results. Combining immunotherapy with liver-directed therapies can improve progression-free survival, but overall survival benefits remain uncertain or unproven for some regimens.
Emerging immunotherapies offer new approaches to overcoming resistance, including bispecific antibodies, alternative immune checkpoint targets, and engineered CAR T-cell therapies directed against GPC3.
Predictive biomarkers remain a major unmet need in HCC. Research involving tertiary lymphoid structures, circulating tumor DNA, and tumor immune signatures could eventually improve patient selection and treatment monitoring.
Optimizing treatment sequencing is a growing priority as more patients receive immunotherapy first-line, creating demand for evidence-based strategies to manage progression and sustain clinical benefit.
Immunotherapy Has Changed Liver Cancer. What Comes Next?
Immune checkpoint-based combinations have transformed the treatment of advanced hepatocellular carcinoma (HCC), producing survival improvements after years of limited progress in systemic therapy. Their emergence has expanded treatment options and introduced new opportunities to intervene against a malignancy that accounts for approximately 90% of primary liver cancers and contributes substantially to global cancer mortality. However, the therapeutic transformation remains incomplete. Many patients do not experience durable responses, and treatment decisions remain complicated by the biological diversity of HCC and the chronic liver disease that frequently accompanies it.1
These challenges have become more consequential as the treatment landscape has expanded. Clinicians now have several effective immune-based regimens but limited tools for predicting which patients will benefit most from each. Meanwhile, developers are investigating new immune targets, engineered cellular therapies, and combinations that integrate systemic treatment with liver-directed interventions. Recent clinical findings have shown both the potential of these strategies and the difficulty of translating improved disease control into longer survival.
Further progress will depend on a deeper understanding of immune resistance, better methods for selecting and monitoring treatment, and clinical development strategies that account for the distinctive biological and functional characteristics of HCC.
The New Foundation of HCC Treatment
For much of the modern era of systemic HCC therapy, sorafenib represented the principal treatment option for patients with advanced disease. Its introduction established that systemic treatment could extend survival, but therapeutic progress remained limited for nearly a decade. Additional tyrosine kinase inhibitors (TKIs) eventually expanded treatment choices, while the emergence of immune checkpoint inhibitor (ICI)-based combinations produced more substantial changes in first-line treatment.2
The phase III IMbrave150 trial demonstrated the potential of combining checkpoint inhibition with antiangiogenic therapy.3 Atezolizumab, which targets programmed death-ligand 1 (PD-L1), was paired with the vascular endothelial growth factor (VEGF) inhibitor bevacizumab. An updated analysis showed median overall survival of 19.2 months with the combination compared with 13.4 months for sorafenib, establishing a significant improvement over the previous standard.
Other pivotal studies demonstrated the effectiveness of different immunotherapeutic approaches. In the HIMALAYA trial, tremelimumab, targeting cytotoxic T lymphocyte-associated protein 4 (CTLA-4), was combined with the PD-L1 inhibitor durvalumab, producing median overall survival of 16.43 months compared with 13.77 months for sorafenib.4 The CheckMate 9DW trial subsequently demonstrated a survival benefit with nivolumab and ipilimumab, which inhibit programmed cell death protein 1 (PD-1) and CTLA-4, respectively, compared with investigator-selected lenvatinib or sorafenib.5 The U.S. Food and Drug Administration (FDA) approved the combination for first-line treatment of unresectable or metastatic HCC in April 2025.6
Although these studies expanded the therapeutic landscape, their results do not establish the relative superiority of one leading immunotherapy regimen over another. Clinical decisions must account for differences in toxicity, underlying hepatic function, comorbidities, and patient preferences. Bevacizumab-containing regimens, for example, require particular attention to bleeding risk and esophageal varices, while combinations incorporating multiple checkpoint inhibitors introduce considerations related to immune-mediated adverse events. Treatment selection consequently remains heavily dependent on clinical characteristics and the distinct safety profiles of available regimens.
Why HCC Remains a Difficult Therapeutic Target
The biological diversity of HCC complicates efforts to develop treatments that work consistently across patients. Tumors can differ considerably in their genetic alterations, molecular signaling, cellular composition, and interactions with surrounding tissue. This heterogeneity exists between patients, among lesions within the same patient, and across different regions of individual tumors. Distinct tumor-cell populations may therefore exhibit different sensitivities to treatment, while resistant populations can persist or expand during therapy.7
The underlying causes of liver disease add another dimension to this complexity. Chronic hepatitis B or C infection, alcohol-associated liver disease, and metabolic dysfunction can produce different inflammatory and molecular environments in which HCC develops. These influences interact with the genetic and cellular changes that accompany tumor evolution, contributing to substantial variation in the mechanisms that sustain malignant growth and shape the surrounding immune environment.7,8
Immune evasion presents a related challenge. HCC tumors can establish microenvironments that suppress antitumor immunity through interactions among malignant cells, regulatory immune populations, stromal components, and inhibitory signaling pathways. Persistent antigen exposure can contribute to dysfunctional T cell states characterized by diminished effector activity, while suppressive immune-cell populations and signaling molecules can independently interfere with the activation and function of antitumor lymphocytes. Checkpoint blockade can restore aspects of immune activity, but its effectiveness may remain constrained by other features of the tumor microenvironment.8,9
Transforming growth factor-beta (TGF-β) signaling is one mechanism under investigation for its contribution to immune suppression in HCC. Its relevance extends beyond explaining the behavior of the tumor microenvironment: developers are also exploring ways to engineer immune therapies that remain functional despite inhibitory signals. Such strategies reflect the recognition that targeting PD-1, PD-L1, or CTLA-4 may not adequately address every mechanism responsible for ineffective antitumor immunity.8,10
The condition of the liver itself further distinguishes HCC from many other malignancies. Most cases develop in patients with chronic liver disease, requiring clinicians to evaluate both cancer progression and the functional capacity of an organ that may already be compromised. Cirrhosis and portal hypertension can increase treatment-related risks, while deteriorating hepatic function may restrict the ability to tolerate additional therapy. Treatment selection therefore depends on the severity of liver dysfunction as well as tumor burden, vascular invasion, and other characteristics of the malignancy.
These considerations also complicate clinical development. Pivotal trials of systemic HCC therapies have generally concentrated on patients with relatively preserved liver function, particularly those classified as Child-Pugh A, leaving limited prospective evidence for patients with more advanced hepatic impairment. A systematic review and meta-analysis of checkpoint inhibitor treatment in Child-Pugh B patients found antitumor activity without a significant increase in treatment-related adverse events compared with Child-Pugh A patients, but survival outcomes remained inferior. Although some patients with impaired liver function may benefit from immunotherapy, identifying those most likely to tolerate and benefit from treatment remains difficult.11,12
Disease stage imposes additional constraints. Resection, transplantation, and ablation can offer potentially curative treatment for appropriately selected patients with early HCC, whereas more advanced tumors often require liver-directed or systemic approaches. Difficulties with early detection can narrow therapeutic opportunities, while differences in disease stage and underlying hepatic function complicate efforts to extend successful regimens into new clinical settings.
Beyond Checkpoint Inhibition: New Combinations and New Targets
One strategy for extending immunotherapy involves integrating systemic treatment with established liver-directed interventions. Transarterial chemoembolization (TACE), which delivers chemotherapy into the hepatic arterial circulation while restricting blood flow to the tumor, has long been used in selected patients with unresectable HCC. Combining TACE with checkpoint inhibition and antiangiogenic agents offers a potential means of improving disease control, but recent trials have demonstrated the difficulty of translating such combinations into longer survival.1,13,14
The phase III EMERALD-1 trial evaluated durvalumab with or without bevacizumab alongside TACE. Adding both agents improved median progression-free survival from 8.2 to 15.0 months compared with TACE alone, whereas durvalumab without bevacizumab did not significantly improve progression-free survival. However, the final overall survival analysis reported in 2026 found no survival benefit for the three-component regimen, with median overall survival of 29.9 months versus 33.3 months in the control group. Grade 3 or 4 adverse events possibly related to treatment were also more frequent with the combination, occurring in 27.3% of patients compared with 7.5% of controls. The final survival findings were reported in a conference abstract.12,15
The phase III LEAP-012 trial investigated pembrolizumab and lenvatinib alongside TACE in patients with unresectable, nonmetastatic HCC. Median progression-free survival improved to 14.6 months compared with 10.0 months for TACE plus placebo, but the reported interim analysis did not establish a statistically significant overall survival benefit. Grade 3 or higher treatment-related adverse events occurred in 71% of patients receiving the combination compared with 32% of controls.14
Progression-free survival can capture improvements in disease control, but overall survival may also be affected by subsequent therapies and underlying liver disease. A statistically nonsignificant overall survival result does not exclude other clinical benefits, particularly when follow-up remains incomplete, as in the reported LEAP-012 analysis. The final EMERALD-1 analysis raises a more substantial question about the survival value of adding durvalumab and bevacizumab to TACE, particularly given the additional treatment-related toxicity.14,15
Efforts to move immunotherapy into earlier disease settings have also encountered difficulties. The phase III IMbrave050 trial initially suggested a recurrence-free survival benefit with adjuvant atezolizumab plus bevacizumab after curative-intent treatment in patients at high risk of recurrence, but the advantage was not sustained during longer follow-up.16 Similarly, the phase III KEYNOTE-937 study did not demonstrate improved recurrence-free survival with adjuvant pembrolizumab following resection or local ablation, according to results reported in a 2026 conference abstract.17 These outcomes indicate that the benefits demonstrated in advanced disease cannot be assumed to extend to patients who have undergone potentially curative treatment.
Other development programs are moving beyond established checkpoint combinations to investigate additional immune targets and therapeutic formats. Bispecific antibodies offer one approach by engaging two immune targets within a single molecule. In a phase II study, KN046, which targets PD-L1 and CTLA-4, was combined with lenvatinib in patients with advanced HCC. Among 55 participants, the objective response rate was 45.5%, with median progression-free survival of 11.0 months. The findings demonstrated preliminary antitumor activity, although the single-arm design did not establish whether the combination offers an advantage over existing first-line regimens.18
Targeting alternative immune checkpoints may offer opportunities to address resistance after prior immunotherapy. The phase II LIVERTI trial investigated domvanalimab, targeting T cell immunoreceptor with immunoglobulin and ITIM domains (TIGIT), together with the PD-1 inhibitor zimberelimab in patients whose HCC had progressed on previous checkpoint inhibition.19 The confirmed objective response rate was 17.2% among 29 patients, and the trial failed to meet its prespecified primary efficacy threshold. Its results illustrate the difficulty of restoring antitumor immune activity after progression on checkpoint-based treatment, even when an additional inhibitory pathway is targeted.
KN046 and LIVERTI illustrate different attempts to extend immune-targeted treatment, but their results cannot be directly compared. The regimens, study designs, and patient populations differ, particularly with respect to prior checkpoint inhibitor exposure. Determining which immune mechanisms remain therapeutically actionable after progression will require studies specifically designed to investigate resistance and subsequent treatment responses.
Engineered cellular therapies offer a different therapeutic approach. Glypican-3 (GPC3), a protein expressed at elevated levels in many HCC tumors, has emerged as a target for chimeric antigen receptor (CAR)-T cell therapy. However, the suppressive tumor microenvironment creates obstacles to the function and persistence of transferred immune cells. The investigational product C-CAR031 seeks to address one of these barriers through a dominant-negative TGF-β receptor II designed to protect GPC3-directed CAR T cells from TGF-β-mediated inhibition.10
A first-in-human study published in 2026 reported an objective response rate of 44.4% among 36 patients with heavily pretreated advanced HCC.10 Median response duration was 4.4 months, and cytokine release syndrome occurred in 34 patients, including two grade 3 events. Investigators also identified changes in tumor samples suggesting that GPC3 antigen loss and increased TGF-β levels may contribute to resistance. These exploratory findings indicate that overcoming one immunosuppressive mechanism may leave other obstacles unresolved, including changes in target-antigen expression. The study demonstrated preliminary clinical activity for a cellular therapy engineered to address the tumor microenvironment, while identifying potential resistance mechanisms that warrant further investigation.
Biomarkers: From Understanding Resistance to Selecting Treatment
The expanding range of therapeutic mechanisms has increased the importance of biomarkers that can inform treatment decisions. Three applications are particularly relevant: identifying patients likely to respond before treatment begins, monitoring response or resistance during therapy, and characterizing the biological changes responsible for progression. Although these objectives overlap, each requires different evidence before it can guide clinical practice. No validated biomarker currently supports routine selection among the available checkpoint-based regimens for HCC.
PD-L1 expression, which has predictive value in certain other malignancies, has shown limited utility as a standalone indicator of immunotherapy response in HCC. Researchers have therefore examined broader molecular and immune signatures, including the distribution and functional characteristics of immune cells within tumors. Such approaches may capture aspects of tumor biology that measurements of individual proteins cannot adequately represent, although intratumoral heterogeneity complicates their interpretation.9,20
Tertiary lymphoid structures (TLS), organized aggregates of immune cells that develop within or near tumors, represent one area of investigation. A 2024 study published in Nature Immunology found that higher intratumoral TLS density at the time of surgery following neoadjuvant immunotherapy was associated with pathological response and improved relapse-free survival.21 Investigators also identified differences in TLS organization within areas of tumor regression, including T-cell-rich regions and markers associated with T-cell memory. The findings suggest that the spatial organization of immune activity may provide information about therapeutic response beyond the presence or abundance of individual immune-cell populations.
Circulating tumor DNA (ctDNA) offers a different means of characterizing treatment response. Analysis of tumor-associated genetic material in blood could potentially allow researchers to monitor molecular changes over time without repeated tissue sampling. In the KN046 phase II study, ctDNA status before the third treatment cycle was associated with prognosis, while exploratory analyses in LIVERTI found that ctDNA dynamics may provide pharmacodynamic information about response to treatment. These findings support continued investigation of liquid biopsies as tools for monitoring treatment activity, although they do not establish that ctDNA-guided decisions improve patient outcomes.18,19
For drug developers, a crucial next step is determining whether biomarker findings can inform prospective clinical trial design. Molecular and immune characteristics associated with treatment response could potentially support patient stratification or the investigation of more biologically defined populations. However, a biomarker that correlates with favorable outcomes is not necessarily predictive of greater benefit from a particular treatment. Establishing that distinction requires studies capable of relating the biomarker to differential therapeutic effects, followed by evidence that using the information can improve clinical decision-making.
Biomarker-informed development may also require strategies for addressing sampling limitations. Measurements derived from one region of a heterogeneous tumor may not adequately represent other lesions or the changes that occur during treatment. Prospective evaluation of tissue and circulating biomarkers could help establish which measurements offer reproducible information and whether incorporating them into trial design or clinical decisions improves outcomes.
The Unresolved Challenge of Treatment Sequencing
The adoption of checkpoint-based combinations in first-line HCC treatment has created a growing need for evidence on subsequent therapy. Although several TKIs and other systemic agents remain available after disease progression, much of the research supporting their use predates the widespread adoption of immunotherapy combinations. The effectiveness of a drug after sorafenib cannot necessarily establish its value following treatment with atezolizumab plus bevacizumab or dual checkpoint inhibition.
The 2024 American Society of Clinical Oncology guideline identifies several potential second-line options following atezolizumab plus bevacizumab, including TKIs and ramucirumab for patients with alpha-fetoprotein levels of at least 400 ng/mL.2 Selected immunotherapy combinations may also be considered in particular circumstances, although supporting evidence for some approaches remains limited to case series or expert consensus. Recommendations following durvalumab plus tremelimumab similarly emphasize TKIs, with the guideline acknowledging the limited quality of the available evidence.
More recent studies have begun examining treatment specifically after first-line immunotherapy. A prospective, multicenter phase II trial evaluated regorafenib in 40 patients whose HCC had progressed on atezolizumab plus bevacizumab.22 Median progression-free survival was 3.5 months, and median overall survival was 10.5 months. An exploratory analysis found that patients with later progression during their previous treatment experienced longer overall survival on regorafenib than those who progressed earlier, suggesting that prior treatment dynamics could provide useful information for subsequent therapeutic development. The study's small, single-arm design, however, prevents the finding from establishing a predictive rule.
The multicenter observational LEVIATHAN study compared lenvatinib and sorafenib following atezolizumab plus bevacizumab.23 Among 230 eligible patients, lenvatinib was associated with longer median progression-free survival of 5.5 versus 2.6 months and median overall survival of 11.9 versus 7.4 months. These results reflect the overall, unmatched cohort. Although the investigators also used propensity score matching to address differences in patient characteristics, the nonrandomized design leaves the possibility of residual confounding. Prospective comparative evidence remains necessary to establish whether one treatment sequence consistently improves outcomes.
Treatment sequencing involves more than identifying which therapy retains antitumor activity. Previous toxicities, changes in hepatic function, and the potential for overlapping resistance mechanisms may all affect subsequent choices. These variables can also change during the course of disease, making evidence derived from fixed treatment populations difficult to apply uniformly. The research challenge is to determine whether clinical characteristics, prior response patterns, and emerging biomarkers can help distinguish patients likely to benefit from particular sequences while preserving sufficient liver function to make continued treatment feasible.
Toward More Precisely Directed HCC Treatment
The expanding therapeutic landscape has made treatment decisions in HCC more consequential, even as the evidence needed to guide many of those decisions remains incomplete. Recent studies have shown that improved disease control does not always translate into longer survival, that resistance can persist despite the introduction of additional immune targets, and that promising molecular associations have yet to establish reliable methods for treatment selection.
As therapeutic options expand, clinical development must establish which mechanisms remain effective after resistance develops, which patients can safely receive additional treatment, and which treatment sequences can sustain disease control. Answering these questions will be critical to extending the benefits of immunotherapy across the biologically and clinically diverse population affected by HCC.
References
1. European Association for the Study of the Liver. “EASL Clinical Practice Guidelines on the Management of Hepatocellular Carcinoma.” Journal of Hepatology. 82:315–374 (2025).
2. Gordan, John D, et al. “Systemic Therapy for Advanced Hepatocellular Carcinoma: ASCO Guideline Update.” Journal of Clinical Oncology. 42:1830–1850 (2024).
3. Cheng, Ann-Lii, et al. “Updated Efficacy and Safety Data from IMbrave150: Atezolizumab Plus Bevacizumab vs. Sorafenib for Unresectable Hepatocellular Carcinoma.” Journal of Hepatology. 76:862–873 (2022).
4. Abou-Alfa, Ghassan K, et al. “Tremelimumab plus Durvalumab in Unresectable Hepatocellular Carcinoma.” NEJM Evidence. 1(8) (2022).
5. Yau, Thomas, et al. “Nivolumab Plus Ipilimumab Versus Lenvatinib or Sorafenib as First-Line Treatment for Unresectable Hepatocellular Carcinoma (CheckMate 9DW): An Open-Label, Randomised, Phase 3 Trial.” The Lancet. 405:1851–1864 (2025).
6. “FDA Approves Nivolumab with Ipilimumab for Unresectable or Metastatic Hepatocellular Carcinoma.” U.S. Food and Drug Administration. 11 Apr. 2025.
7. Safri, Fatema, et al. “Heterogeneity of Hepatocellular Carcinoma: From Mechanisms to Clinical Implications.” Cancer Gene Therapy. 31:1105–1112 (2024).
8. Yin, Yue, et al. “Immunosuppressive Tumor Microenvironment in the Progression, Metastasis, and Therapy of Hepatocellular Carcinoma: From Bench to Bedside.” Experimental Hematology & Oncology. 13:72 (2024).
9. Chen, Kaina, Timothy WH Shuen, and Pierce KH Chow. “The Association Between Tumour Heterogeneity and Immune Evasion Mechanisms in Hepatocellular Carcinoma and Its Clinical Implications.” British Journal of Cancer. 131:420–429 (2024).
10. Zhang, Qi, et al. “GPC3-Specific dnTGFβRII-Armoured CAR T Cells for Hepatocellular Carcinoma.” Nature. 657:495–504 (2026).
11. Pinter, Matthias, et al. “Systemic Treatment in Patients with Hepatocellular Carcinoma and Advanced Liver Dysfunction.” Gut. 74:1178–1188 (2025).
12. Xie, Enrui, et al. “Immune Checkpoint Inhibitors for Child-Pugh Class B Advanced Hepatocellular Carcinoma: A Systematic Review and Meta-Analysis.” JAMA Oncology. 9:1423–1431 (2023).
13. Sangro, Bruno, et al. “Durvalumab with or Without Bevacizumab with Transarterial Chemoembolisation in Hepatocellular Carcinoma (EMERALD-1): A Multiregional, Randomised, Double-Blind, Placebo-Controlled, Phase 3 Study.” The Lancet. 405:216–232 (2025).
14. Kudo, Masatoshi, et al. “Transarterial Chemoembolisation Combined with Lenvatinib Plus Pembrolizumab Versus Dual Placebo for Unresectable, Non-Metastatic Hepatocellular Carcinoma (LEAP-012): A Multicentre, Randomised, Double-Blind, Phase 3 Study.” The Lancet. 405:203–215 (2025).
15. Sangro, Bruno, et al. “183O Overall Survival (OS) in EMERALD-1: A Phase III Study of Durvalumab (D) ± Bevacizumab (B) and Transarterial Chemoembolisation (TACE) in Participants (pts) with Unresectable Embolisation-Eligible Hepatocellular Carcinoma (eeHCC).” Annals of Oncology. 37(Suppl. 1):S72 (2026).
16. Yopp, Adam, et al. “Updated Data from IMbrave050: Adjuvant Atezolizumab Plus Bevacizumab for High-Risk Hepatocellular Carcinoma.” Journal of Hepatology. 84:1102–1111 (2026).
17. Chan, Stephen Lam, et al. “Adjuvant Pembrolizumab for Participants with Hepatocellular Carcinoma and Complete Radiologic Response After Surgical Resection or Local Ablation: The Phase 3 KEYNOTE-937 Study.” Journal of Clinical Oncology. 44(Suppl. 2):Abstract 477 (2026).
18. Xu, Da, et al. “The Anti-PD-L1/CTLA-4 Bispecific Antibody KN046 Plus Lenvatinib in Advanced Unresectable or Metastatic Hepatocellular Carcinoma: A Phase II Trial.” Nature Communications. 16:1443 (2025).
19. Hsiehchen, David, et al. “Dual TIGIT and PD-1 Blockade with Domvanalimab Plus Zimberelimab in Hepatocellular Carcinoma Refractory to Anti-PD-1 Therapies: The Phase 2 LIVERTI Trial.” Nature Communications. 16:5819 (2025).
20. Rizzo, Alessandro, Oronzo Brunetti, and Giovanni Brandi. “Hepatocellular Carcinoma Immunotherapy: Predictors of Response, Issues, and Challenges.” International Journal of Molecular Sciences. 25:11091 (2024).
21. Shu, Daniel H, et al. “Immunotherapy Response Induces Divergent Tertiary Lymphoid Structure Morphologies in Hepatocellular Carcinoma.” Nature Immunology. 25:2110–2123 (2024).
22. Cheon, Jaekyung, et al. “Multicenter Phase 2 Trial of Second-Line Regorafenib in Patients with Unresectable Hepatocellular Carcinoma After Progression on Atezolizumab Plus Bevacizumab.” Liver Cancer. 14:446–455 (2025).
23. Lombardi, Pasquale, et al. “Lenvatinib vs. Sorafenib as Second-Line Treatment Post Atezolizumab Plus Bevacizumab for Hepatocellular Carcinoma: The LEVIATHAN Study.” JHEP Reports. 7:101595 (2025).
