TAMs predominantly exhibit an M2-like phenotype, primarily arising from the phenotypic transformation of resident KCs during tumor progression. Unlike pro-inflammatory, anti-tumor M1 macrophages, M2-polarized TAMs secrete anti-inflammatory factors such as Interleukin-10 (IL-10), transforming growth factor-β (TGF-β), and arginase-1 (Arg1), which suppress immune cell activity and foster an immunosuppressive microenvironment conducive to tumor growth^[21]. Furthermore, M2 TAMs promote tumor progression by secreting angiogenic factors like vascular endothelial growth factor (VEGF) and chemokines such as CCL2, thereby stimulating vasculature formation and enhancing the metastatic potential of cancer cells^[21,22].

In summary, TAMs are extensively infiltrated within HCC, and their polarization state is closely linked to tumor aggressiveness, immune evasion, and patient prognosis. Through the secretion of immunosuppressive factors, metabolic reprogramming, promotion of angiogenesis, and facilitation of metastasis, M2-like TAMs serve as a central driver of immune tolerance and malignant progression in HCC, making them a pivotal target for immunotherapy^[23-26].

MDSCs are a heterogeneous population of immature bone marrow-derived cells with immunosuppressive functions. They originate from myeloid progenitor cells that fail to differentiate normally under pathological conditions such as cancer, expanding as abnormal neutrophils and monocytes^[47]. Initially lacking immunosuppressive activity^[48], these precursor cells are converted into functionally immunosuppressive MDSCs within TME under the influence of inflammatory and immunosuppressive cytokines including sTNF, TGF-β, IL-10, and IL-1β^[49].

Tregs are a subset of T cells characterized by Foxp3, CD25, and CD4 expression, playing a crucial role in immune suppression. They maintain immune tolerance, prevent autoimmune diseases, and regulate the tumor immune environment by suppressing unnecessary immune responses.

The heterogeneity of M2-like TAMs exhibits distinct etiological and spatial specificity. In HBV-related HCC, the bacteria-dominant intratumoral microbial subtype is associated with higher M2 infiltration and specific metabolic reprogramming^[72], while the CK19⁺ subtype specifically enriches secreted phosphoprotein 1-positive (SPP1⁺) TAMs^[23]. During nonalcoholic fatty liver disease (NAFLD) progression, the NAFLD-mSII subtype linked to high cirrhosis risk shows mixed M1/M2 infiltration^[73]. Spatially, M2 macrophages are enriched at the tumor margin, forming a metabolic feedback loop with tumor cells to maintain immunosuppression^[74].

MDSC phenotype and function evolve with underlying liver disease. In HBV-related HCC, CD14⁺HLA-DR^-/low MDSCs are elevated, while during MASH progression, uPAR expression shifts to MDSCs in advanced stages^[61]. Infiltration of MDSCs and SPP1⁺ macrophages correlates with poor prognosis via interactions like SPP1-CD44^[61]. Tregs are integrated in the immunosuppressive network, with their expansion and activation driven by MDSCs in contexts like HBV-related HCC, collaborating with other cells to weaken anti-tumor immunity^[60,75-77].

In summary, M2 macrophages, MDSCs, and Tregs are shaped by HCC etiology and tumor molecular subtypes, forming a heterogeneous immunosuppressive ecosystem [Figure 1]. Deciphering this heterogeneity is essential for understanding immune evasion and developing precision immunotherapies.

CD8⁺ CTLs are key anti-tumor effector cells in HCC TIME, exerting cytotoxic effects by recognizing tumor antigens presented via major histocompatibility complex-I (MHC-I) molecules. Their functional status directly affects tumor progression and immunotherapy efficacy.

Tumor cells reduce MHC-I expression via lysosomal degradation, weakening antigen presentation and inhibiting CTL recognition^[80]. Tumor-associated neutrophils (TANs) highly express PD-L1, inhibiting CD8⁺ T cell activation and cytotoxicity while promoting exhaustion^[81]. High tumor metabolism causes nutrient depletion, leading to CTL functional exhaustion with reduced secretion of cytotoxic molecules and anti-tumor cytokines^[82]. LCFAs accumulation induces mitochondrial dysfunction in CD8⁺ T cells, suppressing fatty acid catabolism and function^[82]. Additionally, tumor cell secretion of immunosuppressive factors, recruitment of inhibitory cells, and high PD-L1 expression contribute to CTL exhaustion via PD-1/PD-L1 pathway activation^[83].

Stem cell-like CD8⁺ T cells (TCF1⁺ TIM3^- CD28⁺) exhibit strong proliferative potential, while terminally differentiated cells (TCF1^- TIM3⁺ PD-1⁺) lack proliferation ability and show typical exhaustion characteristics^[84]. Functional reprogramming is coordinated by the JAK-STAT pathway, with IL-2/IL-15 activating STAT5, IL-21 activating STAT3, and IL-6 activating STAT3/STAT1 heterodimers. In chronic hepatitis B patients, STAT3 phosphorylation correlates with CTL functional recovery^[85,86].

Different CD4⁺ helper T cell subsets affect tumor immunity via specific cytokines, with Th1 and Th17 playing complex roles in HCC. Th1 cells secrete Interferon-γ (IFN-γ), IL-2, and TNF-α to promote CD8⁺ T and NK cell activation and enhance anti-tumor immunity. Their differentiation depends on IFN-γ-STAT1-T-bet signaling, with T-bet regulating chemokine expression and migration^[87]. IFN-γ activates macrophages and enhances DC antigen presentation. Probiotic BLE may enhance liver Th1 function by upregulating T-bet and promoting cytokine secretion^[88]. Th17 cells secrete IL-17, activating NF-κB and STAT3 pathways to promote tumor development^[87]. Differentiated from naïve CD4⁺ T cells under IL-6 and TGF-β1 induction, Th17 cells are regulated by retinoic acid-related orphan nuclear receptor γt (RORγt), with IL-23 enhancing RORγt expression via STAT3 to promote expansion and stabilization^[89-91].

B cells play a dual role in TIME, both promoting anti-tumor immunity and mediating suppression. Tumor-infiltrating B cells (TIL-Bs) and tumor-associated B cells (TABs) exhibit significant heterogeneity. B cell presence in liver cancer correlates with improved patient survival, with activated subsets enhancing CD8⁺ T cell activation^[92]. μMT mice lacking B cells show aggravated HCC progression and impaired CD8⁺ T cell activation^[92]. B cells participate in tertiary lymphoid structures (TLS) formation, promoting local immune memory and effector cell generation^[93,94].

B cells also mediate immune suppression by secreting IL-10 or expressing PD-L1. In HCC, TABs are recruited by TAMs via the CXCL12/CXCR4 axis, expressing PD-L1 to inhibit immune effector cells^[95]. B cell-derived GABA reprograms macrophages to induce an immunosuppressive environment^[96]. IgG⁺ plasma cells promote tumor-promoting macrophage formation in HCC, while IgA⁺ plasma cells induce MDSC activation in colorectal cancer liver metastasis^[97]. B cell abundance and activation status predict survival and immunotherapy response, with high infiltration indicating better prognosis and stronger anti-tumor immunity^[98,99].

NK cell activity is regulated by activating and inhibitory receptors, with natural killer group 2A (NKG2A) and T Cell Immunoreceptor with Ig and ITIM Domains (TIGIT) pathways crucial for function inhibition. In HCC patients, NKG2A upregulation reduces NK cell anti-tumor effects, while blocking NKG2A restores activity and enhances antibody-dependent cellular cytotoxicity (ADCC)^[100]. In HBV-related HCC, TIGIT⁺ TIM-3⁺ NK cells exhibit exhaustion with weakened killing function, reduced cytokine secretion, and impaired proliferation^[101]. Fibrinogen-like protein 1 (FGL1) inhibits CD8⁺ T and NK cells via LAG-3 binding, promoting HCC progression^[102]. LCACs accumulation induces iNKT cell senescence, weakening immune surveillance^[103]. Enhanced STAT3 activity and PRDM10 downregulation in liver NK cells contribute to dysfunction and exhaustion^[104,105]. Soluble MHC Class I Chain-Related A (sMICA) released by tumor cells disrupts NKG2D signaling, impairing NK cell activation^[106]. These mechanisms collectively lead to NK cell exhaustion and immune evasion.

DCs are key antigen-presenting cells, cross-presenting tumor antigens to activate naïve T cells and induce anti-tumor immunity. After capturing antigens, DCs present them to CD8⁺ T cells via MHC-I molecules, promoting CTL differentiation and tumor cell killing^[81,107]. In TME, DCs often exhibit antigen-presenting defects. In HCC, DC dysfunction is characterized by reduced maturity, decreased CD80/CD86 expression, and reduced IL-12 secretion, preventing effective T cell activation and exacerbating immune tolerance^[108,109]. The JAK2-STAT3 pathway activation negatively impacts DC maturation and function^[109]. Tumor-associated exosomes (TAEs) alter DC cytokine secretion, reducing IL-12 and increasing IL-10 and TGF-β to impair immune activation^[110,111].

CD8⁺ T cell infiltration and function are pivotal in HCC. HBV-related HCC shows higher infiltration and tumor reactivity, while metabolic dysfunction-associated steatotic liver disease (MASLD)-related HCC exhibits T cell dysfunction due to tumor glucose consumption via NSUN2-mediated metabolic reprogramming^[112,113]. HBV/MASLD co-existing HCC has elevated precursor-exhausted CD8⁺ T cells, associated with better immunotherapy responses^[112], while Wnt-non-mutant metastatic sites show terminal CD8⁺ T cell exhaustion mediated by immunosuppressive B cells^[114]. In alcohol-associated HCC, reduced CD69^- CD4⁺ T cell frequency with elevated PD-L1 suggests impaired immune surveillance^[115]. In MASH-HCC, naïve CD4⁺ T cell crosstalk with SPP1⁺ macrophages via macrophage migration inhibitory factor (MIF) signaling promotes Treg infiltration^[116]. In MASLD-driven HCC, activated B cell subsets correlate with favorable prognosis, co-localizing with CD45RO⁺ CD8⁺ T cells in TLS^[92,117]. However, in Wnt-non-mutant metastatic microenvironments, B cells adopt an immunosuppressive phenotype to drive CD8⁺ T cell exhaustion^[114]. HBV-related HCC shows increased “adaptive NK cells” with limited anti-tumor activity^[118,119]. In MASH-HCC, tumor cell metabolism-associated genes (e.g., ACSL4) modulate NK cell-mediated clearance via ferroptosis regulation^[120]. DC heterogeneity in HCC is still insufficient. In summary, immune cell abundance, function, and spatial distribution are regulated by etiology, tumor genetics, and local signaling networks, shaping TIME characteristics and treatment responses.

The heterogeneity of the immune microenvironment across individual patients is likely a key determinant of variable responses to immunotherapy. Patients exhibit inherent differences in their immune systems, reflected in baseline variations in immune responsiveness, immune evasion mechanisms, and immune tolerance.

Beyond these intrinsic factors, the immunological profile of liver cancer patients is shaped by a combination of genetic, environmental, and lifestyle influences. Underlying liver disease etiologies - such as hepatitis virus (HBV or HCV) infection, alcoholic liver disease, and nonalcoholic fatty liver disease - further remodel the immune landscape in which the tumor develops^[131]. The main heterogeneity of immune features in HCC caused by different etiologies is summarized in Table 1.

In addition to these contextual factors, differential response rates to immunotherapy in HCC are closely linked to the composition and functional states of specific immune cell populations within the TME. The activity, abundance, and immunosuppressive factor secretion of various cell types - including Tregs, NK cells, DCs, MDSCs, TAMs and CAFs - collectively influence the capacity for immune evasion and ultimately shape the efficacy of immunotherapeutic interventions^{[7,60,76,136]}.

HCC is characterized by pronounced local progression. Different tumor regions - such as the core, invasive margin, and peritumoral area - display marked variations in immune cell infiltration, metabolic profiles, and mechanisms of immune evasion.

ICIs have emerged as a pivotal tool in HCC immunotherapy, achieving substantial clinical progress and becoming an indispensable treatment strategy. ICIs function primarily by blocking negative regulatory signals between tumor cells and immune cells, thereby restoring T cell-mediated anti-tumor activity and reversing tumor immune evasion. The most extensively utilized ICI targets in clinical practice include PD-1 and its ligand PD-L1, as well as CTLA-4. Additionally, emerging targets such as LAG-3, TIGIT, TIM-3, V-Domain Ig Suppressor of T Cell Activation (VISTA), and Siglec-9 are garnering increasing research interest.

Monoclonal antibodies targeting the PD-1/PD-L1 pathway, such as the anti-PD-1 agents nivolumab and pembrolizumab and the anti-PD-L1 agent atezolizumab, have demonstrated therapeutic efficacy in advanced HCC across multiple clinical trials. The IMbrave150 trial established atezolizumab combined with the anti-angiogenic bevacizumab as a first-line standard, significantly improving overall survival (OS) and progression-free survival (PFS)^[150]. Single-cell sequencing data suggest differential benefits across molecular subgroups of advanced HCC with this regimen. Real-world evidence, such as from the AB-real study (n = 296), reported a median OS of 15.7 months, a median PFS of 6.9 months, and an objective response rate (ORR) of 30.8%^[151].

Additionally, the combination of nivolumab (anti-PD-1) and ipilimumab (anti-CTLA-4) has shown significant clinical activity with durable responses, particularly in sorafenib-pretreated patients. In the second-line setting, this combination achieves an ORR of approximately 30% and a median OS of 22 months, demonstrating superior anti-tumor efficacy^[151].

CTLA-4 is another key immune checkpoint, which primarily mediates immunosuppression by regulating the activation of naïve T cells. Anti-CTLA-4 monoclonal antibodies, such as tremelimumab and ipilimumab, have demonstrated efficacy in HCC treatment. Their combination with PD-1/PD-L1 inhibitors can yield synergistic effects, enhancing the overall response rate. Emerging immune checkpoint targets, including LAG-3, TIGIT, and VISTA, have attracted growing interest^[151,152]. These molecules are involved in modulating T cell activity and maintaining immune tolerance. Preclinical and early clinical studies suggest that inhibitors targeting these pathways may help overcome resistance to PD-1/PD-L1 blockade, whether used alone or combined with anti-CTLA-4 therapy, offering potential for new advances in HCC immunotherapy.

Although ICIs have shown considerable promise in HCC, only a subset of patients achieve a meaningful response, while challenges like therapeutic resistance and immune-related adverse events (irAEs) remain common. The complex TIME of HCC, rich in immunosuppressive cells and signaling pathways, facilitates immune evasion and contributes to low ICI response rates. Additional factors, including tumor heterogeneity, loss of tumor antigen expression, and impaired antigen presentation, further limit efficacy. Consequently, significant research efforts are focused on developing ICI-based combination strategies to overcome resistance and improve outcomes. These include combinations with anti-angiogenic tyrosine kinase inhibitors (TKIs), other targeted molecular therapies, local-regional treatments [e.g., Transarterial Chemoembolization (TACE)], and multi-checkpoint inhibition^[153-156].

The advancement of precision immunotherapy in HCC will increasingly depend on biomarker screening to identify responsive patients. Studies suggest that potential predictive markers for ICI efficacy include the density of tumor-infiltrating CD8⁺ T cells, expression of immune-related genes, serum levels of inflammatory cytokines (e.g., IL-8), and liver function-related indicators [e.g., alpha-fetoprotein (AFP), neutrophil-to-lymphocyte ratio (NLR)]^[157-159]. Furthermore, a deeper understanding of tumor genomics and the TIME will help explain the differential ICI responses between immunologically “hot” and “cold” tumors, thereby providing a foundation for individualized treatment strategies^[160,161].

Tregs are highly enriched in HCC. Notably, PD-1-expressing Treg subsets demonstrate potent immunosuppressive activity within both the tumor and peripheral blood, effectively hindering effector T cell activation and anti-tumor immunity^[162]. Key pathways governing Treg function include immune checkpoints such as PD-1/PD-L1 and LAG-3, whose blockade can attenuate Treg-mediated suppression. Furthermore, Treg recruitment into the TME is critically regulated by chemokine signaling. For example, the CXCL12/CXCR4 axis promotes Treg infiltration in HCC, and its inhibition can reduce Treg accumulation and alleviate local immunosuppression^[38,55].

MDSCs, a major class of immunosuppressive cells in the TIME of HCC, promote tumor immune evasion by secreting factors that impair cytokine production and effector T cell function^[121]. The metabolic characteristics and signaling pathways underpinning MDSCs and TAMs are being elucidated. For instance, in liver metastasis, they exhibit upregulated CD36-mediated lipid metabolism, which enhances fatty acid uptake and augments their immunosuppressive activity. Targeting CD36 can thereby suppress the tumor-promoting functions of them^[163,164].

Current strategies to deplete or inhibit Tregs and MDSCs are inherently synergistic. For example, ICIs such as anti-CTLA-4 antibodies can restore effector T cell function while simultaneously suppressing Treg activity^[67]. Furthermore, targeting chemokine receptors like CXCR4 - which is overexpressed in HCC and TME cells - holds promise. In liver cancer, a “dual-hit therapy” using CXCR4 antagonists can block the recruitment of both Tregs and MDSCs to the tumor, thereby alleviating immunosuppression within the TME^[55].

Nanodrug delivery systems enable the targeted delivery of siRNA or small-molecule inhibitors. For example, silencing key regulators such as protein tyrosine phosphatase non-receptor type 2 (PTPN2) can promote the polarization of TAMs toward an M1 phenotype and enhance tumor cell immunogenicity^[165,166]. Moreover, small molecules like vitamin B3 can modulate the GPR109A signaling pathway, inhibiting the immunosuppressive polarization of myeloid cells while boosting the cytotoxicity of CD8⁺ T cells^[167].

The integration of nanotechnology with radiotherapy, chemotherapy, or targeted therapy can further elevate oxidative stress within the TME. This enhances immune cell infiltration and reduces the recruitment of immunosuppressive cells, creating a synergistic effect that ultimately improves the efficacy of immunotherapy^[168-171]. These strategies extend beyond directly targeting immunosuppressive cells; they also remodel the tumor immune landscape by modulating metabolic reprogramming, cellular polarization, and intercellular communication. This multifaceted approach can effectively convert immunologically “cold” tumors into “hot” ones, thereby improving treatment responses in patients with HCC.

Chimeric antigen receptor T cell (CAR-T) cell therapy is under active investigation for HCC, leveraging its high specificity and potent cytotoxicity^[172]. Tumor-associated antigens such as AFP and glypican-3 (GPC3) serve as key targets. AFP-targeted CAR-T cells demonstrate selective cytotoxicity against HCC cells^[173]. Furthermore, GPC3-directed CAR-T therapies have progressed to clinical trials, showing favorable safety and preliminary antitumor activity^[174-176].

Combining AFP- and GPC3-targeted CAR-T cells, or engineering T cells to secrete bispecific T-cell engagers (BiTEs), may enhance anti-tumor immune responses^[177]. However, CAR-T therapy in solid tumors such as HCC remains challenged by the immunosuppressive TME, insufficient T cell infiltration, and tumor antigen heterogeneity, which collectively promote immune evasion and limit therapeutic efficacy^{[174,178,179]}.

TIL therapy involves isolating lymphocytes from a patient’s tumor, expanding them ex vivo, and reinfusing them to enhance anti-tumor immunity. This approach shows particular promise in HCC patients with “hot”, immune-infiltrated tumors, which provide a supportive microenvironment for TIL activity. However, its efficacy in HCC is limited by several factors. The TIME, rich in Tregs and MDSCs, can inhibit reinfused TILs. Furthermore, TIL manufacturing is complex, labor-intensive, and exhibits considerable interpatient variability. These challenges - coupled with the need for personalized protocols - currently restrict the broader clinical application of TIL therapy for HCC^[180-182].

NK cell therapy is increasingly explored for HCC due to its role in anti-tumor immunity. NK cells can recognize and eliminate MHC-I-deficient tumor cells independently of antigen presentation. Specialized subsets such as γδT cells also hold potential for liver cancer immunotherapy, and modulating NK cell function has been shown to enhance anti-tumor responses^[183]. However, NK cell therapy faces challenges within the inhibitory TIME. A key limitation is the lack of sustained proliferative signals after infusion, which curtails long-term efficacy. Strategies such as cytokine engineering or lymphodepletion prior to infusion are being investigated to prolong NK cell persistence and activity, thereby overcoming the suppressive microenvironment and maximizing therapeutic potential^[184,185].

Cellular therapies for HCC remain in early clinical development, facing several key challenges including the inhibitory TIME, antigen heterogeneity and loss, manufacturing complexity, and treatment-related toxicity. Looking forward, a novel approach involves nanoscale lipid vesicles (30-150 nm) derived from CAR-T, CAR-NK, or CAR-M cells, termed CAR-Exosomes. These particles inherit the tumor-targeting specificity of their parental cells while leveraging the intrinsic advantages of exosomes. By combining these strengths, CAR-Exosomes represent a promising strategy to overcome current limitations and may provide a breakthrough in next-generation cancer immunotherapy^[186,187].

In summary, adoptive cell therapies - such as CAR-T, TIL, and NK cell-based treatments - represent promising therapeutic approaches for HCC^[188]. However, their efficacy is limited by the inhibitory TIME and significant tumor heterogeneity. Advances in cellular engineering, rational combination with immunomodulatory agents, and active remodeling of the TIME are therefore critical to enhancing the safety and efficacy of these therapies and improving patient outcomes.

Lactate and its derived lactylation modifications significantly reshape the immune landscape within TIME. Lactate drives the polarization of TAMs toward an immunosuppressive M2 phenotype while impairing the effector functions of CD8⁺ T cells, thereby suppressing immune surveillance^[189-191]. Furthermore, lactate can activate signaling pathways that upregulate immune checkpoint expression, further dampening anti-tumor immunity.

Interventions targeting lactate metabolism - such as inhibiting lactate dehydrogenase (LDH) or modulating lactate transporters (e.g., SLC16A3 knockdown or SLC16A4 upregulation) - have shown promise in alleviating this immunosuppression and enhancing immunotherapy efficacy^[192-194]. Given its pivotal role in the HCC TIME, inhibiting lactate production and transport represents a promising strategy that simultaneously modulates tumor metabolism and remodels the immune landscape. Combining such approaches with immune checkpoint blockade may offer a synergistic strategy to overcome immune evasion and improve treatment responses in HCC patients^[195].

Modulating the gut microbiota represents a promising frontier in liver cancer immunotherapy. As a key regulator of metabolism and immunity, the gut microbiota profoundly shapes the hepatic immune microenvironment and influences HCC development. Through the gut-liver axis, microbial metabolites - such as short-chain fatty acids (SCFAs), bile acids, and lipopolysaccharides - travel via the portal system to modulate hepatic immune homeostasis and inflammatory responses, thereby affecting TIME^[196,197].

Studies indicate that HCC patients exhibit reduced gut microbial diversity, characterized by a decline in beneficial bacteria and an expansion of pro-inflammatory species. This dysbiosis contributes to local immune dysfunction and tumor immune evasion, correlating with altered immune cell infiltration, checkpoint expression, and variable responses to immunotherapy^[198,199]. Interventions aimed at restoring microbial balance - including probiotics, prebiotics, fecal microbiota transplantation (FMT), or antibiotics - can reshape the TIME and enhance therapeutic efficacy^[200,201]. For instance, FMT has shown potential to reverse immune tolerance and improve outcomes in patients resistant to anti-PD-1 therapy^[202,203].

Furthermore, the gut microbiota regulates bile acid metabolism, which in turn influences the activity and metabolic state of hepatic immune cells and modulates immune checkpoint signaling^[204,205]. Thus, targeting the gut microbiome not only helps explain heterogeneity in immunotherapy responses but also offers a novel strategy to overcome resistance. Future combinations of microbiome-based interventions with immune checkpoint blockade hold significant potential to improve survival and therapeutic outcomes in HCC.

In summary, dysregulated lactate metabolism drives immunosuppression in the HCC microenvironment, while the gut microbiota modulates the TIME through metabolic and immunologic crosstalk. Targeting lactate pathways and modulating the microbiota represent promising strategies to overcome current therapeutic limitations, enhance anti-tumor immunity, and improve outcomes in HCC, holding considerable translational potential.

The combination of ICIs with molecular targeted agents is now a cornerstone of first-line therapy for HCC. A key regimen, atezolizumab (anti-PD-L1) plus bevacizumab (anti-VEGF), demonstrated superior efficacy over sorafenib monotherapy in the phase III IMbrave150 trial, achieving an ORR of 30% and a complete response rate of 8%. This strategy works synergistically by concurrently inhibiting angiogenesis and remodeling TIME, effectively converting immunologically “cold” tumors into “hot” ones to enhance immune cell infiltration and activity^[206]. The positive correlation between VEGF pathway activity and immunosuppressive gene expression further supports the strong rationale for this combination, highlighting its significant potential to improve treatment outcomes in HCC^[207].

Combining immunotherapy with local treatments such as TACE and radiotherapy represents a promising strategy for HCC^[208]. TACE combined with systemic therapy has become a recommended regimen for advanced HCC, as it enhances the release of tumor-associated antigens and inflammatory cytokines, thereby activating the cancer-immunity cycle and remodeling the TIME to increase immunotherapy sensitivity^[209,210]. Similarly, radiotherapy combined with ICIs can profoundly modulate the TIME, augment T cell-mediated anti-tumor immunity, and has shown encouraging efficacy in clinical settings for refractory HCC^[211,212]. Other local modalities, including radiofrequency ablation, also demonstrate synergistic potential with immunotherapy in preclinical and early clinical studies. This integrated approach offers a promising avenue to improve outcomes for HCC patients^[213,214].

The combination of multi-targeted TKIs with immunotherapy is an increasingly studied strategy in HCC^[215]. TKIs not only directly inhibit tumor growth and angiogenesis but also remodel TIME by enhancing immune cell activity and suppressing immunosuppressive functions, thereby potentially sensitizing tumors to immunotherapy^[210,216]. For instance, the phase III COSMIC-312 trial, evaluating cabozantinib plus atezolizumab in treatment-naïve advanced HCC, shows promising translational potential^[210]. Similarly, the combination of lenvatinib with ICIs has demonstrated robust anti-tumor activity, prompting ongoing research into mechanisms of resistance and optimization of personalized treatment strategies^[217].

Combination therapies targeting key components of the TIME, such as the gut microbiota and TAMs, are an emerging focus. For example, quercetin combined with anti-PD-1 therapy can remodel the TIME by modulating both the gut microbiome and macrophage polarization, potentially enhancing immunotherapy efficacy^[218]. Furthermore, preclinical strategies co-targeting MDSCs and Tregs have shown promise in boosting anti-tumor immunity. These approaches aim to overcome immune evasion by simultaneously disrupting multiple immunosuppressive axes within the TME^[219,220].

Finally, emerging technologies such as nanoparticle-based delivery systems offer innovative platforms for combinatorial treatment. By enabling targeted and synchronized drug release, nanocarriers can enhance immunogenic cell death, promote antigen release, and activate immune cells, thereby improving the precision and efficacy of immunotherapy. For example, an integrated nanoparticle platform combining ultrasound contrast with chemotherapy, photothermal, and immunotherapy has been shown in animal models to potently activate antitumor immunity and suppress metastasis^[221].

Additionally, in clinical practice, the safety and tolerability of combination therapies must be carefully considered. While combining radiotherapy with other treatments has significantly improved efficacy, it also increases the risk of hematologic and liver toxicity, necessitating close monitoring and management^[222]. Furthermore, factors such as sarcopenia, which is associated with treatment prognosis, indicate that combination therapy regimens should be individualized^[223]. This personalized matching approach highly aligns with the etiological heterogeneity of the immune microenvironment in HCC. For instance, in HBV⁺ MASLD⁺ HCC, the presence of high CD8⁺ T cell infiltration and an enriched population of precursor-exhausted T cells suggests that this subgroup of patients may derive the greatest benefit from immunotherapy^[113]. The mechanisms, clinical advantages, and inherent limitations of these diverse combination regimens and other emerging strategies are systematically summarized in Table 2.