PD−1 is an inhibitory receptor of the B7−CD28 family, expressed on activated T cells, B cells, and a subset of NK cells. Upon engagement by PD−L1, tyrosine residues within PD−1’s ITSM and ITIM motifs are phosphorylated, creating docking sites for SHP2. In its basal state, SHP2’s N−SH2 domain occludes its PTP catalytic pocket; ligand−induced phosphotyrosine binding displaces N−SH2, unleashing phosphatase activity (35). Activated SHP2 preferentially dephosphorylates the co−stimulatory receptor CD28—rather than TCR components such as CD3ζ or ZAP70—thereby attenuating downstream PI3K/Akt signaling, IL−2 production, and T−cell proliferation (36). In myeloid cells, PD−1–SHP2 signaling also impairs phosphorylation of transcription factors IRF8 and HOXA10, inhibiting monocyte differentiation and antigen presentation; intriguingly, SHP2 appears dispensable for PD−1–mediated T−cell exhaustion in chronic viral infection models, underscoring context−dependent pathway regulation (37, 38). Furthermore, numerous studies have established a correlation between PD-L1 upregulation and poor prognosis in CRC (39–41). PD-1/PD-L1 inhibitors counteract this immune suppression by blocking the PD-1/PD-L1 pathway, thereby enhancing T-cell-mediated antitumor immunity. To date, the U.S. Food and Drug Administration (FDA) has approved several PD-1 (Table 2) /PD-L1(Table 3) inhibitors, including pembrolizumab, nivolumab, cemiplimab, and avelumab, for the treatment of various malignancies (42).

The KEYNOTE-177 trial, a landmark international Phase III study, has drawn significant attention for its evaluation of pembrolizumab versus chemotherapy in patients with dMMR/MSI-H mCRC (32). The primary endpoints were OS and PFS. In the final analysis, the median PFS was 16.5 months for pembrolizumab monotherapy compared to 8.2 months for chemotherapy, although no statistically significant difference in OS was observed between the two groups (32). The favorable safety profile of pembrolizumab, as demonstrated in prior trials, further supports its clinical utility (43). These findings underscore the superiority of pembrolizumab as a first-line treatment for dMMR/MSI-H mCRC, offering prolonged PFS and reduced adverse events compared to conventional chemotherapy. Consequently, the FDA approved pembrolizumab for this indication, marking a significant milestone in CRC immunotherapy.

The Phase II CheckMate 142 trial evaluated the efficacy of nivolumab combined with relatlimab, a LAG-3 inhibitor, in patients with dMMR/MSI-H mCRC (44). After a median follow-up of 47.4 months, the ORR, DCR, median PFS, and 2-year PFS rate were 50%, 70%, 27.5 months, and 51%, respectively (44). Compared to pembrolizumab or nivolumab monotherapy, the combination of nivolumab and relatlimab yielded a higher ORR and a longer investigator-assessed PFS, highlighting the durable clinical benefits of this dual checkpoint blockade (43, 45). These results suggest that combining nivolumab with relatlimab represents a promising therapeutic strategy for dMMR/MSI-H mCRC.

Notably, two neoadjuvant trials illustrate the potential for combining immune checkpoint blockade with other modalities. In the randomized phase II PICC trial (NCT03926338), 34 patients with locally advanced dMMR/MSI-H CRC received neoadjuvant toripalimab either alone or alongside the COX-2 inhibitor celecoxib; the combination arm achieved an impressive pCR rate of 88% versus 65% with toripalimab alone, and no relapses were observed at a 12-month follow-up, with only 3% of patients experiencing grade ≥ 3 immune-related adverse events (46). In another ongoing phase II study (NCT04715633), 52 participants with dMMR/MSI-H CRC are being treated preoperatively with camrelizumab plus oral apatinib; interim analyses report an ORR of 42% and a favorable safety profile, with mature pCR and long−term outcomes eagerly awaited. Another ongoing phase II trial NCT06205836 is investigating the safety and efficacy of cemiplimab, both as a monotherapy and in combination with fianlimab, in patients aged 70 years and older with dMMR/MSI-H mCRC. The primary endpoint is the complete response rate (CRR), and the results are highly anticipated.

Structurally, CTLA-4 has a high homology with CD28, but it has a greater affinity for B7 homologs than CD28: CTLA-4 has a monomeric affinity of about 0.2 μM for CD80 and a monomeric affinity of about 2 μM for CD86 (47). During transendocytosis, CD80 continues to bind to CTLA-4 and is endocytosed with CTLA-4 into advanced endosomes and lysosomes, inducing ubiquitination of lysine residues of CTLA-4, which ultimately leads to the degradation of CTLA-4, thereby reducing the amount of CTLA-4 on the cell surface, inhibiting the recycling of CTLA-4 and weakening its negative regulatory effect on the immune response (47). However, CD86 will rapidly dissociate and degrade after being captured by CTLA-4 and transendocytosis, while CTLA-4 will not be modified and can be recycled to the cell surface for further ligand capture, thus maintaining the negative regulatory function of CTLA-4 on the immune response to a certain extent (47). PD-L1 and CD80 can undergo cis-interaction within the cell, i.e., on the same cell surface, PD-L1 and CD80 can form heterodimers. This cis-interaction is able to inhibit the binding of PD-L1 to PD-1 and the interaction of CD80 with CTLA-4, thereby enhancing the immune response to some extent (48). In summary, CTLA-4 inhibitors may be used to treat CRC with a combination of drugs that modulate CD80 and CD86 expression, or in combination with PD-L1 inhibitors, may achieve better therapeutic outcomes (49). At the same time, according to the individual differences of patients (such as the expression levels of CD80, CD86 and PD-L1 on the surface of tumor cells), personalized combination treatment plans are formulated to improve the pertinence and effectiveness of treatment.

Leveraging the molecular choreography of CTLA-4, CD28, and PD-L1 offers a blueprint for precision therapy in CRC. Tumors with dominant CD80 signaling—readily stripped by CTLA-4 and routed to lysosomal oblivion—are primed for CTLA-4 blockade, whereas CD86-centric lesions, whose ligand dissociates and spares CTLA-4 recycling, may demand dual checkpoint inhibition to extinguish residual negative feedback. Concomitant quantification of PD-L1-CD80 cis-heterodimers on tumor cells stratifies patients further: where this inhibitory pair is abundant, adding a PD-L1 antagonist liberates both CD80 for CD28 co-stimulation and PD-L1 from PD-1 restraint, amplifying anti-tumor immunity. Thus, an integrative map of surface CD80, CD86, and PD-L1 densities—together with germline CTLA-4 variants that tune ligand avidity—can guide calibrated sequencing or combination of CTLA-4 and PD-L1 inhibitors, converting mechanistic insight into bespoke regimens that maximize efficacy while minimizing collateral toxicity. These clinical studies will provide key insights into the treatment of CRC (Table 4).

The TME harbors a significant population of regulatory T cells (Tregs), which express elevated levels of CTLA-4 and contribute to immunosuppression. In this context, CTLA-4 inhibitors, including ipilimumab and tremelimumab, have been developed and widely utilized in clinical practice. These agents exert their antitumor effects by alleviating T-cell suppression and depleting Tregs (50, 51). Building on this, Guo et al. (52) conducted a preclinical study on 2MW4691, a bispecific antibody designed to balance Treg depletion and T-cell activation. The experimental results suggest that 2MW4691 holds promise as a candidate for cancer therapy, warranting further evaluation of its tolerability in clinical trials.

In the CheckMate 142 trial, the combination of nivolumab with low-dose ipilimumab demonstrated a high response rate and reliable safety in 119 patients with dMMR/MSI-H CRC (53). After a median follow-up of 4 years, the ORR for the combination therapy was 65%, with a remarkable 48-month OS rate of 71%, underscoring the durable efficacy of this regimen (54). Another large-scale Phase III trial, CheckMate 8HW, enrolled 707 patients to evaluate the PFS of nivolumab plus ipilimumab versus nivolumab monotherapy in dMMR/MSI-H CRC. Across all treatment lines, the combination of nivolumab and ipilimumab outperformed nivolumab alone, offering a novel therapeutic option for patients with mCRC (55).

Tremelimumab, another well-characterized CTLA-4 inhibitor, differs from ipilimumab as an IgG2 monoclonal antibody. It has demonstrated promising therapeutic efficacy across various malignancies, including lung cancer (56), biliary tract cancer (57), and urothelial carcinoma (58). In a Phase Ib/II trial, the combination of tremelimumab and durvalumab with chemotherapy exhibited robust clinical activity in RAS-mutated mCRC (59). No safety concerns were observed in the Phase Ib portion, and the subsequent Phase II trial primarily assessed PFS, reporting a 3-month PFS rate of 90.7% and a secondary ORR of 64.5% (59).

LAG-3, a next-generation immune checkpoint, is highly expressed on exhausted T cells and has emerged as a promising therapeutic target. Initially identified for its ability to bind MHC class II molecules and inhibit T-cell activation (60). However, even in the absence of MHC II, LAG-3 binds to the TCR-CD3 complex and traces immune synapses, subsequently disrupting the binding of the tyrosine kinase p56lck (Lck) to CD4 and CD8 co-receptors, thereby limiting TCR signaling and downstream T cell activation (61). This effect is mediated by the EP motif, suggesting that TCR or CD3 is the primary cis-ligand for LAG-3 (61). LAG-3 regulates T-cell activation through steric modulation of antigen presentation: LAG-3 dimers compress the lateral spacing of MHC-II molecules on dendritic cells, occluding the CD4 co-receptor docking site and hindering TCR triggering (62). Epitope mapping of clinical-stage LAG-3 antagonists has revealed that blockade can be achieved without directly targeting the MHC-binding interface, suggesting alternative mechanisms to disrupt this inhibitory axis (62). These mechanistic insights not only deepen our understanding of checkpoint biology but also inform the design of next−generation immunotherapeutics. Furthermore, dual deficiency of PD-1 and LAG-3 in CD8⁺ T cells enhances tumor clearance, accompanied by increased interferon-γ (IFN-γ) release and upregulation of interferon-responsive genes (63).

The importance of alternative ligands such as galectin-3 (Gal-3), fibrinogen-like protein 1 (FGL1), LSECtin, and α-syn PFF needs to be further confirmed (64). Despite ongoing debates regarding LAG-3 signaling pathways and ligand interactions, the clinical synergy between LAG-3 inhibitors and other immune checkpoint blockers has been well-documented (65) (Table 5). Relatlimab (a LAG-3 monoclonal antibody) and nivolumab has received FDA approval for the treatment of unresectable or metastatic melanoma, marking a significant milestone in cancer immunotherapy (66).

In the context of CRC, the combination of nivolumab and relatlimab has demonstrated clinical benefits in patients with dMMR/MSI-H CRC, as evidenced by the CheckMate 142 trial (44). A Phase III trial evaluating favezelimab (a LAG-3 inhibitor) plus pembrolizumab in CRC reported no treatment-related deaths and notable antitumor activity in PD-L1-positive patients (67). Additionally, the global KEYFORM-007 trial is investigating the safety and efficacy of the co-formulated favezelimab/pembrolizumab (MK-4280A) in PD-L1-positive CRC patients, with an enrollment of 505 participants. This trial also compares the combination to standard-of-care treatments such as TAS-102 (trifluridine and tipiracil) and regorafenib, with results eagerly anticipated. Tebotelimab, a bispecific antibody targeting both PD-1 and LAG-3, has shown encouraging therapeutic responses in solid tumors and hematologic malignancies, with an ORR of 19% and a favorable safety profile (68). These findings highlight the potential of dual checkpoint blockade in overcoming immune resistance and enhancing antitumor immunity.

Currently, ICIs show more efficacy in dMMR/MSI-H CRC patients. However, the heterogeneity of CRC can restrict ICIs’ effectiveness. Moreover, ICIs’ use often leads to drug resistance and immune-related side effects. Thus, a comprehensive patient evaluation before treatment is necessary. Since ICIs have limited efficacy in pMMR/MSS CRC patients, they usually need to be used with other therapies.

Chemotherapeutic agents, such as oxaliplatin and 5-fluorouracil, can induce immunogenic cell death, releasing tumor antigens and neoantigens that activate dendritic cells and enhance antigen presentation (70, 71). Additionally, chemotherapy modulates the immunosuppressive TME by reducing the population of Tregs and myeloid-derived suppressor cells (MDSCs) while increasing CD8⁺ T-cell infiltration, thereby converting “cold” tumors into “hot” tumors (72, 73). Furthermore, chemotherapy may upregulate PD-L1 expression on tumor cells, enhancing the targeting efficacy of ICIs (74, 75). This synergistic interplay between chemotherapy and ICIs creates a dual mechanism of “immune priming and immune maintenance,” offering a promising strategy for pMMR/MSS CRC.

The KEYNOTE-651 study evaluated the long-term safety and efficacy of pembrolizumab combined with mFOLFOX7/FOLFIRI in pMMR/MSS CRC patients, demonstrating favorable outcomes in Cohorts A, C, and E (76). In Cohorts B and D, the benefit of ORR appeared more pronounced in the KRAS wild-type subgroup, warranting further investigation (77). Another multicenter Phase II trial, GOIM 2802, showed that bevacizumab combined with the XELOX-2 regimen administered biweekly was effective and well-tolerated in mCRC patients, with an ORR comparable to that of the bevacizumab plus FOLFOX-4 group (78). The CAMILLA trial (NCT03539822), a Phase I/II clinical study, established the safety of cabozantinib plus durvalumab combined with chemotherapy in pMMR/MSS mCRC patients during its Phase I component (79). The Phase II portion, which included a CRC cohort, reported a DCR of 86.2%, highlighting the antitumor activity and manageable toxicity of this combination (80). These encouraging results have prompted the initiation of the Phase III STELLAR-303 trial, further exploring this therapeutic approach.

Radiotherapy, like chemotherapy, can promote the release of tumor antigens and remodel the TME, enhancing systemic immunity through the abscopal effect (81, 82). Radiation-induced upregulation of cell surface molecules on tumor cells further augments the cytotoxic activity of NK cells and T cells (83). The release of type I interferons enhances T-cell priming and dendritic cell activation, contributing to a robust antitumor immune response (84, 85). These immunostimulatory effects make radiotherapy a valuable adjunct to ICIs in pMMR/MSS CRC.

A Phase II randomized trial reported impressive results, with short-course radiotherapy combined with the PD-1 inhibitor toripalimab and CAPOX achieving a complete response (CR) rate of 58.1% in patients with locally advanced rectal cancer (LARC) (86). Another study evaluated the efficacy and safety of tislelizumab (a PD-1 inhibitor) combined with chemoradiotherapy in LARC patients, demonstrating a CR rate of 40.0% (87). Although these results are promising, larger-scale trials are needed to validate these findings. Neoadjuvant chemoradiotherapy (NACRT) with or without sintilimab (a PD-1 inhibitor) achieved a CR rate of 44.8% in pMMR LARC patients, significantly higher than the 26.9% observed in the control group (88). Immunohistochemical analysis suggested that PD-L1-positive patients may derive greater benefit from this treatment, highlighting the importance of biomarker-driven patient selection.

A neoadjuvant therapy study involving 44 LARC patients treated with preoperative chemoradiotherapy (CRT) and nivolumab monotherapy reported 3-year relapse-free survival (RFS) outcomes. In MSS patients receiving CRT, those with high expression of PD-L1, PD-1, CTLA-4, and Ki-67, as well as an elevated CD8/eTreg ratio, exhibited a higher trend of 3-year RFS (89). These findings underscore the potential of combining radiotherapy with ICIs in pMMR/MSS CRC and suggest that neoadjuvant strategies may become a standard treatment option for this patient population.

Anti-angiogenic agents, primarily tyrosine kinase inhibitors (TKIs) and antibodies targeting the vascular endothelial growth factor (VEGF)/VEGF receptor (VEGFR) pathway have demonstrated synergistic effects when combined with ICIs (90, 91). VEGF not only promotes angiogenesis but also exerts direct immunosuppressive effects by recruiting MDSCs and Tregs while inhibiting T-cell function (92, 93). Anti-angiogenic drugs, such as bevacizumab and ramucirumab, normalize tumor vasculature, improve blood perfusion, and alleviate hypoxia within the TME (94, 95). Vascular normalization enhances immune cell infiltration and reduces endothelial cell-induced T-cell apoptosis, creating a favorable environment for ICIs to exert their effects (96).

BD0801, a humanized anti-VEGF monoclonal antibody, has shown superior antitumor effects compared to bevacizumab in preclinical models, likely due to its potent VEGF/VEGFR blockade and inhibitory effects on human umbilical vein endothelial cells (97). The REGOMUNE trial, a Phase II study, evaluated the safety and efficacy of regorafenib combined with avelumab in MSS CRC patients, demonstrating good tolerability and no unexpected adverse events (98). Biomarker analysis revealed that patients with higher CD8⁺ T-cell infiltration experienced improved median OS and PFS, further supporting the combination of anti-angiogenic agents with ICIs (98). Fruquintinib, a highly selective TKI targeting VEGFR1, 2, and 3, has shown promising results in mCRC patients who have undergone at least two prior lines of chemotherapy. An international Phase III double-blind trial reported a median OS of 9.3 months and a median PFS of 3.7 months in the fruquintinib group, significantly higher than the 6.6 months and 1.8 months observed in the placebo group (99). These findings suggest that fruquintinib represents a valuable therapeutic option for mCRC patients. Other TKIs, including anlotinib (100–102) and sorafenib (103), have also shown potential in CRC treatment. Additionally, the emergence of bispecific antibodies, such as ivonescimab (anti-PD-1/VEGF-A), has expanded the therapeutic arsenal. A Phase Ia clinical trial demonstrated the promising antitumor activity of ivonescimab, with further studies underway to evaluate its efficacy in combination with other therapies (104).

However, prolonged VEGF/VEGFR blockade may lead to excessive vascular pruning, resulting in reduced drug distribution and hypoxia. Therefore, the therapeutic efficacy of anti-angiogenic agents depends on achieving a balance between vascular normalization and immune activation (96). Exploring triple therapy combinations involving anti-angiogenic agents, ICIs, and other modalities may further modulate the TME and enhance treatment outcomes.

The mitogen-activated protein kinase (MAPK) pathway regulates critical cellular processes, including proliferation, differentiation, and apoptosis. Dysregulation of MAPK signaling contributes to uncontrolled cell growth and tumor development, making MEK a clinically relevant target (105, 106). MEKi can reduce the release of immunosuppressive factors and decrease the recruitment of immunosuppressive cells. Preclinical data suggest that MEK inhibition upregulates PD-L1 and MHC-I expression, facilitating the subsequent blockade of the PD-1/PD-L1 pathway (107–109).

A Phase Ib trial evaluated the combination of atezolizumab and cobimetinib (a MEK inhibitor) in patients with advanced solid tumors (110). Among the 84 enrolled mCRC patients (62 pMMR/MSS and 2 dMMR/MSI−H), only seven achieved confirmed responses. Diarrhea was the most frequent adverse event. The regimen demonstrated suboptimal safety, with grade 3–4 treatment−related adverse events occurring in 44% of patients and approximately 70% requiring treatment discontinuation or dose reduction due to intolerability (110). Although initial synergistic activity was observed in mCRC, it was not substantiated in a subsequent Phase III trial. Consequently, despite cobimetinib’s potential to modulate the tumor microenvironment, the atezolizumab–cobimetinib combination proved insufficient to overcome immunotherapy resistance in MSS mCRC patients.

Similarly, the Phase III IMblaze370 trial compared atezolizumab + cobimetinib and atezolizumab monotherapy versus regorafenib in the third−line setting for mCRC2. This study enrolled 363 patients (with dMMR/MSI−H recruitment capped at ≤ 5%) and used OS as the primary endpoint (111). IMblaze370 failed to meet its primary endpoint, demonstrating no significant difference in OS, ORR, or PFS for either atezolizumab−containing arm compared with regorafenib (111). These results indicate that combining MEK inhibitors with immune checkpoint inhibitors offers no substantial clinical benefit in pMMR/MSS CRC and warrants further investigation.

OVs exert their antitumor effects through two distinct yet complementary mechanisms (1): direct cytopathic destruction of malignant cells via selective intracellular replication and subsequent cell lysis, and (2) indirect immunostimulatory effects mediated by the massive release of TAAs and neoantigens following viral-induced oncolysis, which initiates systemic antitumor immune responses (12). Importantly, the presentation of viral antigens additionally triggers robust antiviral immune responses that exhibit cross-reactivity with tumor cells, thereby establishing a bimodal attack on neoplastic tissues (148).

A groundbreaking clinical advance was achieved by Zhao Yongxiang’s team through the development of NDV-GT, a recombinant Newcastle disease virus engineered to express porcine α1,3-galactosyltransferase (α1,3GT). This modification induces tumor-specific hyperacute immune rejection via ectopic α-gal epitope expression (149). In a phase I/II trial involving 23 patients with advanced malignancies, NDV-GT elicited a DCR of 90.00% (18/20) with no severe adverse events. Notably, three CRC patients discontinued treatment due to COVID-19 pandemic-related logistical constraints, underscoring real-world challenges in clinical trial execution (149).

Genetic modification of adenoviral vectors enables precise modulation of the TME (150). Rongye Jing and colleagues developed rAd.mDCN.mCD40L, a novel oncolytic adenovirus co-expressing CD40 ligand and decorin, a stromal remodeling agent. In CT26 CRC models, this construct demonstrated potent suppression of hepatic metastasis (151). Another studies revealed that adenoviruses encoding glypican-3 core proteins synergized with NK cell adoptive transfer, enhancing NK-mediated tumor cytotoxicity and intratumoral lymphocyte infiltration in CRC xenografts (152). These findings validate OVs as potent adjuvants for cell-based immunotherapies through multidimensional immune activation (152–154).

While FOLFOXIRI chemotherapy remains a cornerstone in mCRC management, its limited efficacy has prompted exploration of OV-chemotherapy combinations. Girod et al. demonstrated enhanced tumoricidal activity in Colo320 CRC cells through co-administration of coxsackievirus B3 PD-H and chemotherapeutic agents, with combination therapy surpassing monotherapy efficacy via virus-mediated chemosensitization (155). To address targeting limitations and therapeutic resistance, a novel approach utilizing mesenchymal stem cell (MSC)-delivered coxsackievirus A21 achieved robust antitumor effects in CRC murine models, highlighting the potential of cellular vehicles for precise viral delivery (156).

Furthermore, the release of TAAs and cytokines following oncolysis is considered a potential mechanism for enhancing ICIs (157). Several clinical studies (158, 159) have evaluated the combination of oncolytic virotherapy with ICIs in CRC, demonstrating favorable clinical activity and manageable toxicity. In a first-in-human signal-seeking study, 34 patients with proficient-mismatch-repair/microsatellite-stable pMMR/MSS CRC received the oncolytic vaccinia virus pexastimogene devacirepvec (Pexa-Vec) in tandem with the PD-L1 blocker durvalumab, with or without the CTLA-4 antagonist tremelimumab (158). Although the combination was well tolerated, its antitumor impact was marginal: median progression-free survival edged from 2.1 to 2.3 months—incremental gains that fall squarely within the efficacy corridor of existing FDA-approved third-line therapies for mCRC (158). Separately, intratumoral delivery of the engineered herpes simplex virus-2 derivative OH2, combined with the PD-L1 inhibitor LP002, yielded an exceptional responder: a single responder experienced a 313-day remission and remains alive at 499 days (159). Such a solitary signal, however, cannot escape the statistical noise inherent to a four-patient cohort; confirmation awaits adequately powered, prospectively designed trials (159).

But, oncolytic virotherapy for CRC faces three main challenges. First, delivery difficulties. Most CRC tumors are internal, making direct injection tough. Second, immune response balance. Pre-existing immunity can block oncolytic virus efficacy. Repeated injections may trigger neutralizing antibodies. Third, lack of predictive biomarkers. Identifying patients most likely to benefit from oncolytic virotherapy is challenging due to the absence of reliable biomarkers. These issues limit oncolytic virotherapy’s effectiveness and broader application in CRC treatment (160).

ACT represents a highly innovative approach to cancer treatment, involving the extraction, modification, and expansion of immune cells ex vivo, followed by re-infusion into patients to enhance anti-tumor immunity. Among various ACT modalities, NK cell therapy, which activates innate immunity, and chimeric antigen receptor T (CAR-T) cell therapy, which offers precise tumor targeting, are the most prominent (161). In CRC, cytokine-induced killer (CIK) cell therapy and CAR-T cell therapy have shown promising results, while newer approaches, such as T-cell receptor-engineered T cells and CAR-macrophages, hold significant potential (161). As technology advances and more clinical trials are conducted, ACT is expected to further improve CRC treatment outcomes, offering hope for better patient survival.

CIK cells are a heterogeneous population of immune effector cells generated by stimulating and expanding peripheral blood mononuclear cells with cytokines ex vivo (162). This population includes conventional T cells, NK cells, and NKT-like cells, with NKT-like cells being the primary effectors that can recognize and kill tumors in an MHC-unrestricted manner (162). A meta-analysis of 70 clinical trials involving 6,743 CRC patients demonstrated that CIK cell therapy significantly improved clinical outcomes, particularly in terms of quality of life and survival (163). Most studies (66 trials) combined CIK therapy with chemotherapy (e.g., FOLFOX or XELOX regimens) to enhance chemotherapy efficacy and mitigate its side effects. Among these, 45 trials administered CIK cells concurrently with chemotherapy, with infusion timing varying across studies: early (3 trials), mid-chemotherapy (7 trials), late (13 trials), and unspecified (22 trials). Future studies should explore optimal infusion schedules to maximize therapeutic benefits. Additionally, personalized CIK cell therapies tailored to patients’ genetic mutations and immune profiles are under development.

CAR-T cell therapy involves genetically modifying patients’ T cells to express a chimeric antigen receptor (CAR) targeting specific tumor antigens. Upon re-infusion, CAR-T cells recognize and bind to tumor surface antigens, releasing cytokines that induce tumor cell apoptosis (164). In CRC, current trials focus on identifying suitable tumor antigens and improving CAR-T cell efficacy and safety. Common targets include carcinoembryonic antigen (CEA) and NKG2D ligands. Several trials (NCT02349724, NCT02416466, NCT02850536) have assessed the safety of CAR-T cells in CEA-positive patients. Notably, NCT02349724 showed that one patient survived for 23 months with elevated serum IFN-γ levels and no grade 3 or higher adverse events (165). Similarly, no severe adverse events were observed in the other trials, confirming the safety of CAR-T infusion (166, 167). Furthermore, NCT02850536 demonstrated improved CAR-T cell delivery to the liver, resulting in increased OS (167). Despite these promising advances, CAR-T therapy in CRC remains limited and requires further exploration to enhance its efficacy.

ACT for CRC faces several key challenges. The immunosuppressive TME of solid tumors poses a significant barrier to ACT efficacy by inhibiting T cell function (168). Additionally, identifying suitable targets that balance safety and efficacy is challenging. ACT is also associated with substantial toxicities, such as cytokine release syndrome, neurotoxicity, and target-related adverse effects (169). The generation of tumor-specific lymphocytes for each patient is technically demanding and economically costly, which limits the widespread application of ACT. Furthermore, T cell exhaustion, characterized by the loss of effector function and self-renewal capacity, further diminishes treatment effectiveness (169). These complexities collectively restrict the clinical application of ACT in CRC.

Matrix-depletion therapy is an emerging strategy that targets the ECM within the TME. This approach primarily addresses collagen metabolism dysregulation, aberrant activation of matrix metalloproteinases, and the dysregulated expression of factors such as TGF-β (170–172). In CRC, these imbalances are closely linked to TME reprogramming, making ECM components potential therapeutic targets (173, 174). Although matrix-depletion therapy is still in its early stages in CRC, it has shown promise in other stroma-rich malignancies, such as pancreatic cancer.

One key ECM component, hyaluronic acid, plays a critical role in tumor progression. A Phase II trial in metastatic pancreatic cancer patients tested the combination of pegvorhyaluronidase alfa (PEGPH20) with nab-paclitaxel, gemcitabine plus nab-paclitaxel, and gemcitabine alone (175). The results demonstrated that PEGPH20 significantly prolonged PFS and increased ORR, particularly in patients with high hyaluronic acid expression (175). However, a subsequent Phase III trial failed to show benefits in OS or PFS with PEGPH20, leading to its discontinuation in development (176).

Another promising target in matrix-depletion therapy is the Hedgehog signaling pathway, which is implicated in stromal formation. Inhibition of this pathway has been shown to reduce stromal density (177). A Phase II trial combining a Hedgehog pathway inhibitor with cemiplimab in metastatic basal cell carcinoma showed promising anti-tumor activity and safety (178). However, results in sarcoma and myeloma were less encouraging, necessitating further exploration in CRC (179, 180).

Focal adhesion kinase (FAK), a tyrosine kinase involved in integrin signaling, has emerged as another therapeutic target (181). A Phase II trial evaluating a FAK inhibitor in meningioma reported a 33% six-month PFS rate, highlighting its potential as a therapeutic strategy (182).

In summary, while matrix-depletion therapy shows promise, it faces several challenges, including limited specificity and tumor heterogeneity. Moreover, ECM degradation may facilitate circulating tumor cell dissemination, underscoring the need for further safety and efficacy evaluations in CRC.