Cytotoxic CD4+ T cells in cancer: an emerging target for next-generation anticancer immunotherapy? Authors: Imke M B van Wandeloo, Kalijn Bol, Carla van Herpen, I Jolanda M de Vries, Gerty Schreibelt Published: 2025 ABSTRACT AbstractOver the last decades, a growing number of distinct CD4+T helper cells has been identified and our understanding of CD4+T cell differentiation and function in various disease contexts has increased immensely. It has long been thought that the role of CD4+T cells in the tumor microenvironment (TME) was limited to coordinating the immune response by stimulating other immune cells and by secretion of cytokines with antitumor activity, while direct killing of tumor cells has been largely attributed to cytotoxic CD8+T cells. Notably, CD4+T cells with direct cytotoxic activity (CD4+CTLs) have been reported in the context of viral infections, autoimmune disorders, and more recently in patients with various cancer types. These cells have the ability to secrete cytotoxic molecules and kill target cells in a major histocompatibility complex (MHC) class II-dependent manner. In this review, we give an overview of phenotypical characteristics of CD4+CTLs in human cancers and the antitumor mechanisms employed by these cells. Further, we explore their role and clinical relevance in the context of cancer and describe how these cells may be used for the development of novel immunotherapeutic options to benefit patients with cancer with MHC class II-positive tumors. Introduction In recent years, immunotherapies such as immune checkpoint inhibitors (ICI) and adoptive cell therapy (ACT) have become the standard of care for many patients affected by various types of cancer.1 T cell effector functions of both CD4+ and CD8+ T cells play a critical role in antitumor immune responses, but direct killing of tumor cells has been largely attributed to the cytotoxic functions of CD8+T cells. While it is known that CD4+T helper cells also play a key role in mediating the antitumor immune response, it has long been thought that their role was limited to coordinating the immune response by stimulating other immune cells. Naïve CD4+T cells (TH0) can differentiate towards different functional CD4+T cell subtypes, including T regulatory cells (Tregs), follicular T helper cells (TFH), T-helper cell 1 (TH1), TH2, TH9, TH17, and TH22. On stimulation of their T-cell receptor (TCR) and cues from the cytokine environment, naïve CD4+T cells express key transcription factors that drive their differentiation toward these different subtypes. Each subtype secretes different cytokines that allow them to provide aid to the appropriate effector immune cells, or mediate immunosuppression in the case of Tregs.14 Between these different subtypes, substantial phenotypic plasticity exists, meaning that a specific CD4+T cell can adopt characteristics of other subsets during its lifespan, and even express characteristics of multiple subsets simultaneously.5 It is widely recognized that the helper activity displayed by CD4+T cells plays a critical role in the development of an anti-tumor response, as activated CD4+T cells can directly activate cytotoxic CD8+T cells through the production of interleukin (IL)-2. CD4+T cells also support cross-presenting dendritic cells through CD40-CD40 ligand interaction, thereby inducing effector functions of CD8+T cells in an indirect manner.1 6 In addition, CD4+T cells can drive the differentiation of B cells towards affinity-matured, class-switched plasma cells, thereby playing an important role in the induction of the humoral response against tumor antigens.1 7 Aside from coordinating the antitumor response of CD8+T cells and B cells, CD4+T cells can also have a more direct antitumor response. For instance, TH1 cells produce effector cytokines such as interferon-γ (IFNγ) and tumor necrosis factor-α (TNFα), which display direct antitumor activity.1 Besides the abovementioned roles, CD4+T cells can also mediate tumor suppression in a manner that does not require tumor-intrinsic major histocompatibility complex (MHC) class II expression, for instance, by interacting with tumor-associated myeloid cells.8 9 Thus, CD4 T cells play a multifaceted role in anti-tumor immunity. In the last decade, CD4+T cells that display direct cytotoxic activity in a similar fashion to their CD8+T cell counterpart have also been identified in various disease contexts. Initially, it was thought that these cytotoxic CD4+T cells (CD4+CTLs) were an anomaly primarily found in patients with chronic viral infections, such as human cytomegalovirus (hCMV),10 11 human immunodeficiency virus12 13 and viral hepatitis.14 However, the cells have also been found in other viral infections, including dengue virus,15 influenza virus16 and in SARS-CoV-2-infected individuals.17 CD4+CTLs have also been documented in various autoimmune diseases.1821 More recently, the presence of CD4+CTLs has also been established in preclinical murine models of cancer and various types of human cancer.2242 For instance, murine studies have shown that transfer of naïve tumor antigen-specific CD4+T cells into lymphopenic hosts resulted in pronounced expansion and differentiation into a CD4+CTL subset that was able to eliminate established melanoma tumors, independent of CD8+T cells, B cells and natural killer (NK) T cells, without requiring in vitro manipulation. The transferred cells defaulted to a TH1 phenotype, associated with production of high levels of TNF, IL-2, and IFNγ, with upregulated expression of genes related to CTL effector functions.43 44 Interestingly, while the transfer of the CD4+T cells resulted in initial regression of melanoma tumors, they were unable to eradicate the tumors completely. The addition of cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) blockade on transferred naïve CD4+T cells further enhanced antitumor activity, resulting in complete rejection of established melanomas.43 In addition to the blockade of immune checkpoints such as CTLA-4, antitumor responses can be increased by targeting co-stimulatory molecules such as CD137 (4-1BB). In an FBL-3 tumor model, it was shown that treatment with a CD137 agonist resulted in the induction of tumor-specific granzyme B (GzmB)-producing CD4+CTLs that were able to reject tumor cells. Interestingly, a subset of CD4+forkhead box P3 (FoxP3)+ Tregs in the model acquired TH cell and cytotoxic functions after CD137 agonist treatment, without the loss of FoxP3 expression. These GzmB-producing Tregs were able to contribute to an MHC class II-restricted killing of antigen-loaded target cells, showing that under certain circumstances, Tregs could be reprogrammed and contribute to antitumor activity.45 Both in murine models and in the context of human cancer, CD4+CTLs have been characterized as granzyme-secreting and perforin-secreting cells that have the ability to kill target cells in an MHC class-II restricted manner.323 4345 Response rates of patients to currently existing immunotherapy approaches vary greatly. The discovery of CD4+T cells with cytotoxic activity is therefore interesting, as it provides a new potential therapeutic target. The ontogeny and differentiation of these cells, however, remain incompletely understood and phenotypical characteristics of these cells appear to be partially contradictory and dependent on disease context. Gaining a deeper understanding of the phenotype and function of CD4+CTLs is essential, as it may lead to more effective use of immunotherapy for individual patients. Here, we aim to characterize these human CD4+CTLs in the context of cancer and to explore which antitumor mechanisms these cells employ, based on currently existing literature. Further, we aim to describe how these cells could be used to benefit patients with cancer. CD4+ CTLs hold prognostic value in human cancers Whereas few circulating CD4+CTLs are generally detected in healthy individuals, elevated levels of CD4+CTLs have been observed in a growing number of human cancers.2526 3840 In humans, the presence of CD4+CTLs has been confirmed in patients with melanoma,222438 multiple myeloma,25 diffuse large B cell lymphoma (DLBCL),26 follicular lymphoma,37 bladder cancer,27 breast cancer,22 28 head and neck cancer,22 29 colorectal cancer,30 31 hepatocellular carcinoma (HCC),223234 non-small-cell lung cancer (NSCLC),35 acute myeloid leukemia,36 glioma,39 41 and renal cell carcinoma.42 Frequencies of CD4+CTLs were found to be highly variable between cancer types.22 Tumor-specific CD4+CTLs show expression of genes linked to cytotoxic functions and their ability to lyse autologous tumor cells in an MHC class II-dependent manner has been confirmed in multiple studies as well. CD4+CTLs appear to have a polyfunctional phenotype, as they have the ability to secrete cytokines in addition to their cytotoxic activity.22 In both patients with DLBCL and patients with multiple myeloma, frequencies of circulating CD4+CTLs, were found to be significantly elevated compared with healthy age-matched and gender-matched controls.25 26 In both studies, CD4+CTLs were able to lyse autologous malignant cells ex vivo. Besides elevated circulating levels, CD4+CTLs may also be elevated in the tumor microenvironment (TME) of patients with cancer. Multiple clonally expanded CD4+CTL states were found to be enriched in the TME of patients with transitional cell carcinoma in the urinary tract compared with adjacent non-malignant tissue. These CD4+CTLs made up approximately 15% of tumor-infiltrating CD4+T cells and were able to recognize and lyse autologous tumor cells in an MHC class II-dependent manner ex vivo, but their cytotoxic activity could be suppressed by Tregs.27 The observed elevated levels of CD4+CTLs in the circulation and TME and their confirmed ability to lyse autologous tumor cells raise the question of whether the presence of these cells is clinically relevant. Numerous studies have made observations that indicate that CD4+CTLs may indeed hold prognostic value in human cancers. For instance, patients with stage III multiple myeloma presented with lower levels of CD4+CTLs compared with patients with stage II disease, suggesting that frequencies of CD4+CTLs found in patients with multiple myeloma were negatively associated with disease severity.25 In patients with metastatic urothelial cancer, a signature of genes upregulated specifically in CD4+CTLs was found to predict clinical responses to atezolizumab, a monoclonal antibody (mAb) directed against programmed death-ligand 1 (PD-L1), highlighting that developing enhanced CD4+CTL activity in response to anti-PD-L1 treatment may be clinically beneficial.27 The levels of circulating and tumor-infiltrating CD4+CTLs have also been found to hold prognostic value in mismatch repair-proficient colorectal cancer (pMMR CRC),30 HCC,34 and glioma.39 In patients with pMMR CRC, infiltration of CD4 CTLs in the central tumor region was identified as an independent protective factor of disease-free survival (DFS). Infiltration of CD4+CTLs in the central tumor region of pMMR CRC increased after neoadjuvant chemotherapy. This increased infiltration of CD4+CTLs in the tumor was associated with better DFS and overall survival (OS).30 Patients with early-stage HCC were found to have increased numbers of CD4+CTLs, but their numbers and cytotoxic activity progressively decreased in patients with HCC with advanced-stage disease.34 The decreased cytotoxic activity of CD4+CTLs in patients with HCC with advanced stage disease was ascribed to increased numbers of Tregs. The study also found that low numbers of circulating CD4+CTLs were associated with poorer survival rates in patients with stage III HCC and poorer DFS rates in patients who underwent surgical tumor resection. In addition, lower levels of intratumoral CD4+CTLs were associated with poorer DFS and OS in comparison to high levels of intratumoral CD4+CTLs.34 In patients with glioma, CD4+CTLs decreased with tumor severity. Patients with grade II glioma had significantly higher frequencies of circulating and tumor-infiltrating CD4+CTLs than patients with a grade III glioma. Moreover, patients with high percentages of circulating GzmB-expressing CD4 T cells had significantly longer DFS after tumor resection compared with patients with low percentages of these cells, regardless of tumor grade.39 In patients with DLBCL, especially those with a germinal center B-cell-like subtype, OS rates were higher when patients presented with MHC class II-positive tumors compared with MHC class II-negative tumors at diagnosis.46 In addition, patients with relapsed or refractory disease were found to have significantly lower MHC class II-expressing CD19+ cells and lower expression intensity of MHC class II on CD19+ cells compared with non-recurrent patients.26 While these studies did not necessarily look at the association between circulating CD4+CTLs and survival rates of patients with DLBCL, the results are interesting as MHC class II-expressing cells are targeted by CD4+CTLs and the frequency and cytotoxic potential of CD4+CTLs was found to be positively correlated with MHC class II expression levels on malignant CD19+B cells.26 In line with these results, expression of the MHC class II cell surface receptor human leucocyte antigen (HLA)-DR on melanoma cells was also found to be associated with progression-free survival (PFS), OS as well as therapeutic response to anti-programmed cell death 1 (PD-1).47 In addition, a correlation with CD4+T cell infiltrate was observed, although further characterization of the infiltrated CD4+T cells was not performed.47 In contrast, one study found that higher densities of tumor-infiltrating CD4+CTLs in patients with oral squamous cell carcinoma (OSCC) were associated with a worse PFS probability.48 Notably, the authors observed that CD4+CTLs may upregulate expression of several inhibitory receptor genes and they may exhibit co-expression modules similar to those observed in Tregs, suggesting that next to their cytotoxic ability, CD4+CTLs may also act as immunosuppressor cells in OSCC.48 Together, these data demonstrate that increased intratumoral frequencies of CD4+CTLs are associated with better survival probabilities, and that a decline in CD4+CTL numbers and activity can be observed with advancing tumor stage, although this generalization is not supported by all studies. The dual role of CD4+CTLs, as observed in the context of OSCC, calls for a need to not overlook potential detrimental roles that CD4+CTLs may have in the context of human cancers. Collectively, findings from most studies support the notion that CD4+CTLs may act as a prognostic biomarker in various human cancers. In addition to this prognostic value, emerging evidence, though still scarce, suggests that CD4+CTLs may also play a pivotal role in mediating immunotherapy responses.49 Phenotypical heterogeneity of CD4+ CTLs exists in different cancers With the advancement of single-cell-resolution technologies such as single-cell RNA sequencing (scRNA-seq), the ability to phenotypically characterize CD4+CTLs increases. In the context of cancer, CD4+CTLs show phenotypical heterogeneity. Table 1 shows an overview of genes expressed by CD4+CTL clusters detected by scRNA-seq in various human cancers. Noticeably, genes encoding cytotoxic effector molecules such as granzymes, perforin, and the antimicrobial peptide granulysin are expressed on human CD4+CTLs. These molecules are encoded by GZMA, GZMB, GZMH, GZMK, GZMM, PRF1, and GNLY, respectively. In addition, expression of the granule-associated protein natural killer cell granule protein 7 (NKG7) was demonstrated on CD4+CTLs. NKG7 is considered to be a valuable marker for the identification of cytotoxic cells, and it is thought to be involved in the exocytosis of cytotoxic granules, thereby playing a critical role in target cell killing by cytotoxic effector cells.50 Even though many clusters described on a single-cell RNA level were found to express genes that encode cytotoxic effector molecules, validation that these clusters are also able to mediate cytotoxicity in vitro is lacking. IFNG, encoding the cytokine IFNγ, is also often expressed, which is consistent with observations in murine studies where transferred naïve tumor antigen-specific CD4+T cells appear to default to a TH1 phenotype.43 44 The fractalkine receptor CX3CR1 was also found to be expressed by CD4+CTLs in various cancers. In addition, multiple transcription factors, encoded by the genes RUNX3, EOMES, and TBX21, are found to be expressed by tumor-specific CD4+CTLs. The observed high levels of KLRG1, significantly higher levels of class I-restricted T cell-associated molecule (CRTAM) and low levels of both CD27 and CD2812 22 51 provide an argument that CD4+CTLs are likely in a late state of differentiation.22 Cytotoxic CD4+clusters detected by scRNA were often found to be conserved across patients.22 27 36 41 42 Interestingly, some authors observed the presence of more than one distinct CD4+CTL subtype. For instance, Oh et al observed two heterogeneous CD4+CTL states that shared the expression of certain cytotoxic effector molecules, but one population co-expressed high levels of GZMB, PRF1, GNLY, and NKG7, whereas the other population was characterized by the expression of GZMK and lower levels of NKG7 (table 1).27 Moreover, later work again revealed two distinct subpopulations of CD4+CTLs with GZMB and GZMK overexpression that exhibited a TH1 gene signature in patients with muscle-invasive bladder cancer.51 The circulating GZMB-expressing CD4 CTL population consisted of cells with a more naïve-like or effector phenotype, whereas GZMK-expressing CD4 CTLs adopted an effector memory phenotype.51 Other markers that have been suggested as potential identifiers of CD4 T cells with cytotoxic potential in human cancers include CD11b52 and co-expression of B7-H3 and LAG3.53 To summarize, CD4+CTL clusters have been identified in different types of human cancers. These cells are likely in a late state of differentiation and express genes encoding cytotoxic effector molecules. Table 1Expression of genes by cytotoxic CD4+T cell clusters identified by single-cell RNA sequencing in patients with different cancer typesStudyCancer typeMarker expression on cytotoxic CD4+clusters GZMA GZMB GZMH GZMK GZMM PRF1 GNLY NKG7 FASLG IFNG TNF RUNX3 EOMES TBX21 KLRG1 CRTAM CX3CR1 CCL4 CCL4L1/2 CCL5 SLAMF7 CST7 PDCD1 LAG3 Cachot et al, dataset retrieved from Sade-Feldman et al22 24MelanomaCachot et al dataset retrieved from Azizi et al22 28Breast cancerCachot et al dataset retrieved from Puram et al22 29Head and neck squamous cell carcinomaCachot et al dataset retrieved from Zheng et al22 33Hepatocellular carcinomaZheng et al33Hepatocellular carcinomaGuo et al35Non-small-cell lung cancerOh et al† 27Transitional cell carcinoma * * * Oh et al27Transitional cell carcinomaZhang et al† 31Colorectal adenocarcinomaZhang et al31Colorectal adenocarcinomaGuo et al36AMLHan et al37Follicular lymphomaLiu et al41GliomaYang et al† 42Renal cell carcinomaYang et al42Renal cell carcinomaDepending on the publication, scRNA-seq was performed on single cells isolated from tumor and/or matched adjacent normal tissue and/or matched blood sample. Clusters were included in the table when they showed high expression of genes previously linked to effector cytotoxic functions. Markers were included when they were identified in >1 cancer type and/or paper or otherwise of interest.*A minority of IFNG-expressing CD4 cells from this cluster also expressed these markers.†More than one potential cluster identified.scRNA-seq, single-cell RNA sequencing. Expression of genes by cytotoxic CD4+T cell clusters identified by single-cell RNA sequencing in patients with different cancer types Depending on the publication, scRNA-seq was performed on single cells isolated from tumor and/or matched adjacent normal tissue and/or matched blood sample. Clusters were included in the table when they showed high expression of genes previously linked to effector cytotoxic functions. Markers were included when they were identified in >1 cancer type and/or paper or otherwise of interest. A minority of IFNG-expressing CD4 cells from this cluster also expressed these markers. More than one potential cluster identified. scRNA-seq, single-cell RNA sequencing. Ontogeny and differentiation of CD4+ CTLs To date, the origin and differentiation pathway of CD4+CTLs in the context of cancer remains incompletely understood. Early research on the origin and differentiation of CD4+CTLs comes primarily from viral context and autoimmune disease. It has been suggested that CD4+CTLs are composed of a portion of effector/effector memory T cells that have downregulated the expression of CD27 and CD28.4 10 27 51 54 55 These CD4+CD28 cells have been found to be expanded in the circulation of aging individuals and patients with various immune disorders, and they display cytotoxic features, such as the expression of perforin and granzymes.55 Another study has suggested that naïve CD4+T cells that express CRTAM are precursor cells of CD4+CTLs and have the ability to develop a cytotoxic phenotype.56 It has also been proposed that CD4+CTLs may arise from the activation of other CD4+T helper subsets.3 27 However, it appears that the majority of CD4+CTLs observed in patients with cancer display a TH1-like phenotype, suggesting that these CD4+CTLs are derived from or default to a TH1 phenotype. Various transcription factors, including T-bet, Eomes,23 57 58 Runt-related transcription factor 3 (RUNX3), Th Inducing POZ-Kruppel Factor (ThPOK),5961 Hobit,62 B lymphocyte-induced maturation protein-1 (Blimp-1),63 and B-cell lymphoma 6 (Bcl-6)64 have been suggested to regulate the induction of a cytotoxic program in CD4+T cells. Which of these transcription factors function as main upstream inducers appears to be dependent on disease context and whether it concerns human or murine CD4+CTLs. The antagonistic transcription factors ThPOK and RUNX3 play a role in the development of thymocytes into mature T cells, where they are regarded as the master regulators of commitment to the CD4 or CD8 T cell lineages. T cells commit to the CD4 helper lineage when ThPOK is induced in thymocytes expressing an MHC class II-reactive TCR, whereas RUNX3 induction in thymocytes with MHC class I specificity is necessary to induce the cytotoxic program in thymocytes committed to the CD8 T cell lineage.54 In humans, increased expression of RUNX3 and T-bet were found to play a role in the acquisition of cytotoxic function in CD4+T cells. In hCMV-seropositive healthy adults, terminally differentiated CD4+T cells displayed cytotoxic activity.61 The transcriptional program of these CD4+CTLs was enriched in RUNX3, T-bet, and Eomes. Contradictory to studies in murine models, in which reprogramming of CD4+T cells towards MHC class II-restricted CD4+CTL is associated with the upregulation of RUNX3 and subsequent downregulation of ThPOK,54 59 60 these human cells did not downregulate expression of ThPOK.61 Further in vitro analysis revealed that naïve CD4+T cells acquired cytotoxic activity when cultured with TH1 polarized cytokines. In addition, cytotoxic program acquisition by naïve CD4+T cells was found to be dependent on RUNX3 and T-bet, and was negatively regulated by the sustained expression of ThPOK.61 In contrast to murine models, in which Eomes is critical for cytotoxic programming of CD4+T cells,58 65 66 the role of Eomes in the cytotoxic activity acquisition by human CD4 T cells was found to be redundant.61 In murine models, Blimp-1 has recently gained attention as a transcription factor for CD4+CTLs. Depletion of Tregs in tumor-bearing mice resulted in excess availability of IL-2 in the TME, which was found to drive production of GzmB and support gain of cytotoxic activity by CD4+T cells in a Blimp-1 dependent manner.63 In humans, however, the transcription factor Hobit, a Blimp-1 homolog, was identified to be highly expressed by human CD4+CD28 T cells, and linked to cytotoxic potential of CD4+T cells after infection with hCMV.62 It has been suggested that the development of TFH cells and CD4+CTLs occurs in an opposing manner, as the development of CD4+CTLs is suppressed by the TFH master transcription factor Bcl-6 and transcription factor 7,64 although TFH-like CD4+CTLs have also been described in literature.67 SLAM family member 7 has also been proposed as a candidate to increase cytotoxic function in CD4+CTLs in human cancers.22 CD4+ CTLs possess granule-dependent cytotoxicity and lyse tumor cells in an MHC class II-dependent manner There seems to be a general consensus that CD4+CTLs are able to recognize and lyse tumor cells in an MHC class II-dependent manner by degranulation of lytic granules containing granzymes and perforin (figure 1), which can be hampered by tumor-infiltrating Tregs.27 63 CD4+CTLs can recognize cognate antigens that are processed and presented by antigen-presenting cells (APCs) in the context of MHC class II molecules.3 The acquisition of cytotoxic functions is expected to be dependent on the strength of the interaction between the TCR and antigen presented on MHC class II, the antigen dose, and cues from the cytokine environment.4 68 The expression of MHC class II on several different cell types, including tumor cells, can be upregulated by IFNγ or TNFα.6971 IFNγ is able to induce the expression of the MHC class II transactivator CIITA, via the JAK-STAT pathway, leading to enhanced MHC class II gene expression.7274 Cachot et al demonstrated that treatment of melanoma cell lines with IFNγ and/or TNFα enhanced the expression of MHC class II.22 Various cells in the TME can be a source of IFNγ, including CD4+CTLs that have defaulted to a TH1 phenotype.43 44 Therefore, tumor cells may become susceptible to CD4+CTL activity in the course of an already ongoing immune response. Interestingly, it was shown that CD4+CTLs possess direct, contact-dependent, and granzyme-dependent cytotoxic activity against melanoma cell lines, but with delayed killing kinetics compared with CD8+T cells.22 CD4+CTLs recognize and lyse tumor cells in an MHC class II-dependent manner by degranulation of lytic granules containing granzymes, perforin, and granulysin. Naïve CD4+T cells expressing CRTAM have the ability to develop a cytotoxic phenotype. Transcription factors implicated in the acquisition of the cytotoxic program include ThPOK, RUNX3, Eomes, T-bet and Blimp-1. CD4+CTLs express SLAMF7, FasL, NKG7, and CX3CR1, have downregulated expression of CD27 and CD28 and can be activated by targeting the costimulatory molecules CD137 and CD134. CD4+CTLs that have defaulted to a TH1-like phenotype secrete IFNγ, which in turn can induce MHC class II expression on target cells. The figure is created in BioRender. Blimp-1, B lymphocyte-induced maturation protein-1; CD4+ CTLs, Cytotoxic CD4+ T lymphocytes; CRTAM, class I-restricted T cell-associated molecule; Eomes, Eomesodermin; FasL, Fas ligand; GNLY, granulysin; GzmA, granzyme A; GzmB, granzyme B; GzmH, granzyme H; GzmK, granzyme K; GzmM, granzyme M; IFNγ, interferon gamma; IL, interleukin; MHC, major histocompatibility complex; NKG7, natural killer cell granule protein 7; RUNX3, Runt-related transcription factor 3; SLAMF7, SLAM family member 7; TCR, T cell receptor; ThPOK, Th inducing POZ-Kruppel factor. Using single-cell-resolution technologies, studies have shown that genes encoding cytotoxic effector molecules are expressed on human CD4+CTLs in patients with cancer. On antigen recognition in the context of MHC class II molecules, CD4+CTLs are able to release granules that contain granzymes, perforins, and the antimicrobial peptide granulysin. Perforin is a cytolytic protein, encoded by the PRF1 gene, that forms pores in the target cells, allowing for diffusion of granzymes into the target cells. Granzymes are serine proteases that, in turn, initiate apoptosis of the target cells.75 The notion that the cytolytic activity acquired by tumor-specific CD4 T cells depends on recognition of tumor cells through MHC class II and degranulation of lytic granules was further confirmed as simultaneous in vitro blockade of GzmB and perforin activity was found to prevent lysis of tumor cells.43 While the abilities of GzmA and GzmB to induce cell death have been confirmed, it remains unclear whether GzmH, GzmK, and GzmM also possess generic cytotoxic activity in vivo.75 Besides the release of granzyme-containing and perforin-containing cytotoxic granules, additional mechanisms might be involved in CD4+CTL cytotoxicity. For instance, cytotoxic effector cells are also able to kill tumor cells through FasL-related or Tumor Necrosis Factor-Related Apoptosis Inducing Ligand (TRAIL)-dependent mechanisms.76 Transcriptomic analysis revealed a modest increase in the expression of FasL on CD4+CTLs in patients with melanoma,22 which suggests that the Fas-FasL pathway may play a role in inducing apoptosis of tumor cells by CD4+CTLs. In contrast with these results, blockade of FasL or TRAIL in melanoma-bearing murine models failed to prevent tumor rejection.43 In addition, no clear increase of the TRAIL death receptor/ligand pathway was observed in CD4+CTLs in patients with melanoma.22 Only in OSCC patients, expression of TNFSF10, encoding the protein TRAIL, was observed on CD4+CTLs.48 Together, these data, summarized also in figure 1, suggest that the major mechanism in which CD4+CTLs kill tumor cells is dependent on degranulation of granules containing cytotoxic effector molecules. The extent to which FasL-mediated and TRAIL-mediated pathways contribute to tumor rejection in cancer patients has to be elucidated in future studies. Application of CD4+ CTLs as antitumor therapy In response to chronic antigen stimulation, CD8+T cells present in the TME may become exhausted.77 In addition, downregulation of MHC class I molecules presents one of the major mechanisms used by various types of solid tumors to escape the immune system.78 This poses a problem for immunotherapies using the cytotoxic potential of CD8+T cells to their advantage. Even after an initial clinical response, acquired resistance to immunotherapy can occur due to clonal selection and outgrowth of tumors bearing specific mutations. Previous studies have associated therapy resistance with mutations resulting in disrupted IFNγ signaling and genetic alterations in components of the tumor cell-intrinsic MHC class I-antigen presentation machinery, such as loss of β2 microglobulin (B2M), an essential component in the folding, transport and expression of MHC class I on the cell surface.79 80 MHC class II expression is present in various human tumors and can be upregulated in response to IFNγ.6971 Since CD4+CTLs have the ability to secrete cytokines and directly kill tumor cells in an MHC class II-dependent manner, these cells may contribute to antitumor responses and may be able to (partially) compensate for the loss of CD8-mediated killing in patients with disrupted MHC class I expression, and the subsequent lack of antigen presentation in the context of MHC class I.4 It has even been observed that CD4+T cells may play a critical role in modulating responsiveness of mismatch repair-deficient tumors harboring B2M alterations to ICI, highlighting that CD4+T cells may indeed be able to reject B2M-deficient tumors.81 In vivo enhancement of CD4+ cytotoxic activity Current literature primarily presents options for the potential enhancement of CD4+cytotoxic activity in vivo. For instance, neoadjuvant chemotherapy in patients with pMMR CRC led to increased infiltration of CD4+CTLs into the tumor.30 Immunostimulatory mAbs also hold potential to enhance cytotoxic activity of CD4+T cells. These mAbs can antagonize receptors that suppress immune responses, such as CTLA-4 and PD-1, or act on receptors that amplify immune responses, such as CD137 and CD134.82 In murine models, both anti-PD-1 and anti-lymphocyte activation gene-3 (LAG-3) mAbs exhibited antitumor effects on MHC class II-expressing tumors in a cytotoxic CD4+T cell-dependent manner. Combination of anti-PD-1 and anti-LAG-3 mAbs exhibited even stronger antitumor efficacy.83 Another study found that dual blockade of PD-1 and CD85j could restore impaired cytotoxic effector functions of intratumoral CD4+GzmB+ T cells in a patient-derived NSCLC explant model.84 In patients with inflamed metastatic urothelial cancer, it has been suggested that anti-PD-L1 mAbs may alter the TME in favor of CD4+CTL activation.27 Furthermore, CTLA-4 blockade has been shown to effectively enhance CD4+CTL responses in patients with melanoma.23 The tumor antigen-specific CD4+CTLs were able to recognize naturally processed antigen presented by APCs, and had the ability to directly lyse autologous melanoma cell lines expressing NY-ESO-1 in an MHC class II-restricted manner.23 mAbs against CD137 and/or CD134 increase costimulation of CD4+T cells,82 and drive cytotoxic function of naïve CD4+T cells, through induction of Eomes.58 66 CD134 and CD137 agonists have undergone clinical testing both as monotherapy and combination therapy and show a tolerable safety profile.828587 In addition, CD4+T cell responses may also be induced in vivo by anticancer vaccines.88 Overall, chemotherapy has been shown to be able to increase CD4+CTL tumor infiltration, but mostly, activation can be increased by treatment with current available and upcoming ICIs and/or CD134/CD137 agonists. CD4 T cell-based adoptive cell therapies In addition to enhancing cytotoxic CD4+T cell responses in vivo, ACT using CD4+CTLs may also be a viable option for the treatment of various malignancies. Preclinical models have shown successful tumor killing after transferring naïve CD4+T cells into melanoma-bearing mice,43 providing early evidence that CD4-based ACTs hold therapeutic potential to establish an antitumor response. Furthermore, co-transfer of CD8+T cells and in vitro generated CD4+CTLs enhanced tumor regression, as these CD4+CTLs exhibited antigen-specific cytotoxicity and supported CD8+-mediated antitumor activity.89 Additionally, durable clinical responses have also been observed in humans.90 91 In a patient with refractory metastatic melanoma, complete tumor regression was observed after administration of an autologous NY-ESO-1-specific CD4+T cell clone.90 Adoptive transfer of tumor-infiltrating lymphocytes (TILs), containing a population of CD4+T cells reactive to an ERBB2IP mutation, in a patient with metastatic cholangiocarcinoma resulted in tumor regression and disease stabilization for over 1 year.91 Moreover, ACT using genetically modified CD4+T cells, transduced with an MHC class II- and HLA-DPB1*0401-restricted TCR specific for melanoma-associated antigen-A3 was demonstrated to be safe in patients with various metastatic solid cancers.92 Out of nine patients treated at the highest dose level, three patients showed an objective partial response.92 Results from these studies indicate that adoptive transfer of autologous antigen-specific CD4+T cells can mediate tumor regression in humans, providing a strong foundation for additional clinical studies focused on the safety and efficacy of harnessing the direct antitumor activity of CD4+T cells in the treatment of advanced stage disease. Further, naturally occurring NY-ESO-1123/137/DRB3*02:02 CD4+T cells displaying high cytotoxicity and secretion of IFNγ and TNFα were detected in melanoma patients.93 Subsequent engineering of human primary CD4+T cells to express this NY-ESO-1123/137/DRB3*02:02 resulted in tumor control after ACT into a murine model,93 indicating that these TCR-engineered CD4+T cells may hold potential for clinical applications aiming to target tumors expressing NY-ESO-1 in HLA-DRB3*02:02 positive patients.93 Chimeric antigen receptor (CAR) T cell ACT is also gaining popularity as immunotherapy. A recent 10-year follow-up report on two patients with chronic lymphocytic leukemia who achieved complete remission on receiving CD19-targeted CAR T cells revealed that therapy response of these patients consisted of an initial phase dominated by CD8+CAR T cells and a long-term remission phase presenting with a persisting activated CD4+CAR T cell population that exhibited cytotoxic characteristics.94 It has been suggested that these cytotoxic CD4+CAR T cells mediate long-term cytotoxicity against CD19+leukemic cells after treatment.94 95 In 2024, lifileucel was US Food and Drug Administration-approved for the treatment of unresectable or metastatic melanoma. Lifileucel uses TILs that are recovered from autologous tumor tissue. The produced TILs are a mix of polyclonal patient-specific CD8+ and CD4+ T cells that are reinfused into the patient and can subsequently recognize and target various tumor-associated neoantigens.96 A recent functional study showed that CD4+TILs, which were isolated and expanded from melanoma biopsies, were also able to kill autologous tumor cells in vitro. As expected, immune recognition by CD4+TILs was dependent on tumor expression of MHC class II. On IFNγ signaling loss, CD4+TIL-mediated cytotoxicity was abrogated toward tumor cells without constitutive expression of MHC class II, but not toward melanoma cells that had constitutive MHC class II expression.80 To date, evidence that CD4+TILs possess direct cytotoxicity against autologous tumor cells and functional validation of CD4+CTLs in humans remains scarce and should be further investigated. Nevertheless, durable clinical responses have been observed in humans after ACT based on autologous CD4+T cells, underscoring that future research efforts should focus on the potential use and safety of CD4-based ACT in clinical settings. Discussion The aim of this review was to provide an overview of the phenotypical characteristics of CD4+CTLs present in human cancer, and their employed antitumor mechanisms. In addition, we aimed to describe how these cells may be used for the development of novel immunotherapeutic options to benefit cancer patients with MHC class II-positive tumors, based on currently existing literature. Multiple studies show that CD4 T cells with direct cytotoxic activity exist in patients with cancer, where the presence of circulating and tumor-infiltrating CD4+CTLs also appears to hold prognostic value. CD4+CTLs can recognize antigens that are processed and presented by APCs, and they hold the ability to recognize and lyse tumor cells in an MHC class II-dependent manner by degranulation of granules containing cytotoxic effector molecules. Based on the effector mechanism these cells employ, they may be able to compensate for the loss of CD8-mediated killing in patients with disrupted antigen presentation in the context of MHC class I. As such, CD4+CTLs may be an interesting novel therapeutic target. Various transcription factors have been suggested to regulate the acquisition of the cytotoxic program in CD4+T cells, but to date, the ontogeny and differentiation of CD4+CTLs remains incompletely understood. In addition, previously identified CD4+CTL populations appear to be quite heterogeneous, which makes it challenging to identify markers that uniformly define these cells. Resolving knowledge gaps with regard to the phenotype, ontogeny, and differentiation of CD4+CTLs is crucial as this could provide insight into how we could drive isolated naïve CD4+T cells towards a cytotoxic phenotype in vitro. In addition, further assessing changes in MHC class II expression and infiltration of CD4+CTLs between different tumor stages could be interesting. This is especially relevant since recent evidence in melanoma patient samples has revealed that outgrowth of melanoma subclones lacking constitutive and IFNγ-inducible HLA-II can occur, as a result of JAK1/2-STAT1 pathway inactivation, resulting in insensitivity to both IFNγ and CD4+T cell activity.74 Aside from associations between MHC class II expression and/or infiltration of CD4+CTLs and survival outcomes or therapy response, it may also help to identify patient populations that may potentially benefit from targeting these cells specifically. Further, functional evidence in humans that CD4+T cells can possess direct cytotoxicity remains scarce. It would be interesting to culture and expand tumor-derived CD4 T cells from MHC class II-expressing tumors and assess the cytotoxic capacity of these cells towards autologous tumor cell lines ex vivo. In short, future lines of research should focus on establishing protocols for the generation and expansion of CD4+CTLs in vitro and ex vivo or the activation of these cells in vivo, in parallel to identifying patient populations that would benefit from these efforts. Increasing our understanding of CD4+CTLs in the context of cancer and how to target these cells may result in the development of novel immunotherapeutic options that may act synergistically with currently existing immunotherapies, usually aiming to enhance tumor cell killing by CD8+CTLs. Strategies to enhance CD4+CTL activity may benefit patients with tumors with primary or acquired resistance to current available ICI treatment options. Unfortunately, the potential future clinical application of in vivo enhancement of CD4+cytotoxic activity or CD4+CTL-based ACTs is limited to patients with tumor cells that express MHC class II to enable recognition by CD4+CTLs. To overcome this restriction in patient selection, treatment aiming to target CD4+CTLs should be combined with strategies to enhance MHC class II expression and prevent downregulation or loss of MHC class II. IFNγ is the most potent inducer of MHC class II, acting via the JAK/STAT1–CIITA pathway.71 Although intravenous administration of IFNγ gives quite severe toxicity, intratumor injection might be a more feasible route to support CD4+CTL treatment strategies,97 while resistance to IFNγ-signaling could be counteracted by Toll-like receptor 9 agonist.98 Another approach could be to use epigenetic drugs, such as histone deacetylase inhibitors and DNA methyltransferase inhibitors, to restore MHC class II expression or treatments that inhibit oncogenic pathways that suppress MHC class II, for example, with mitogen-activated extracellular kinase (MEK) or epidermal growth factor recepter (EGFR) inhibitors.99 100 Recent studies have also hinted that CD4+CTLs may have immunoregulatory properties,48 101 and little is known about potential CD4+CTL exhaustion.102 While evidence on exhaustion in CD4+CTLs remains scarce, it has been proposed recently that CD4+CTLs may exhibit an immunosuppressive shift and thereby contribute to therapy resistance in patients treated with ICI.49 These challenges must be taken into account when designing strategies for therapeutic MHC class II upregulation and therapeutics targeted towards CD4+CTLs. Nevertheless, here we show that CD4+CTLs may hold prognostic value in human cancers and could therefore represent an emerging target for next-generation immunotherapy. Supplementary material