Revolutionizing tumor immunotherapy: unleashing the power of progenitor exhausted T cells

Surface molecular characterization of Tpex cells

TCF-1 and PD-1 are the quintessential signature of molecules adorning the surfaces of Tpex cells. Additional notable surface molecules include C-X-C chemokine receptor 5 (CXCR5), SLAMF6, and CD62L (Figure 2)^15,20–23. TCF-1, encoded by the Tcf7 gene, is a critical transcription factor downstream of the Notch signaling pathway. Its role extends beyond the thymic development of T cells and encompasses the formation of the memory phenotype²⁴. TCF-1 primarily maintains the memory phenotype of Tpex cells through upregulation of key transcription factors, including inhibitor of DNA binding 3 (ID3), Eomes, c-Myb, and B cell lymphoma-2 protein (Bcl-2)^25,26. The absence of TCF-1 mirrors the dysfunction observed in tumor-specific CD8⁺ T cells, which is similar to the unresponsiveness observed in patients undergoing ICI treatment in clinical scenarios²⁷. SLAMF6, another crucial surface marker associated with precursor exhaustion, is co-expressed with TCF-1 in most Tpex cells^17,28. Furthermore, within the TME in patients with non-small cell lung cancer (NSCLC), a distinct subset of CXCR5⁺ PD-1⁺ CD8⁺ T cells displays heightened TCF-1 expression²⁹. Intriguingly, this phenomenon does not manifest uniformly across various tumor types, e.g., melanoma¹⁷. This divergence suggests a potential relationship between CXCR5 expression on Tpex cells and tumor types; therefore, further research is necessary to reveal these specific relationships.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Surface molecules characterized during the differentiation of CD8⁺ T cell subpopulations. Differences in the expression of surface molecules, transcription factors, and effector molecules during CD8⁺ T cell exhaustion in Tpex cells, Tex^eff cells, and Tex^term cells. Int, intermediate.

The hallmark of CD8⁺ T cell exhaustion, C-X3-C motif chemokine receptor 1 (CX3CR1), has a critical role in the intricate dynamics of the TME. CX3CR1⁺ CD8⁺ T cells, characterized by their expression of CX3CR1, have emerged as key players in the anti-tumor defense within the TME (Figure 2). The downregulation of CX3CR1 is closely linked to the terminal exhaustion state of CD8⁺ T cells, and marks a shift in their functional capabilities. Single-cell RNA sequencing (scRNA-seq) has revealed that Tpex cells are precursors to SLAMF6⁻ CX3CR1⁺ CD8⁺ T cells. Investigations of chronic viral infections have elucidated the journey of TIM-3⁻ CX3CR1⁺ CD8⁺ T cells and traced their origin to Tpex cells. Notably, these cells can further differentiate into a subset characterized by TIM-3⁺ CX3CR1⁺ CD8⁺ T cells endowed with proliferative capabilities³⁰. In the context of melanoma and breast cancer, a specific population of TCF-1^Low PD-1⁺ TIM-3^int CD8⁺ T cells are the primary responders to anti-Galectin-9 therapy³¹. Although Tpex cells might not have inherent anti-tumor capabilities, continuous differentiation of Tpex cells into functional Tex^eff cells can mediate sustained and more robust anti-tumor response. Notably, in PD-1 blockade therapy, Tpex cells engage in asymmetric division, thereby maintaining the stability of the Tpex cell population while simultaneously generating a cohort of highly effective CD8⁺ T cells³². This intricate balance of differentiation highlights the nuanced yet critical role of Tpex cells in shaping anti-tumor immunity after immunotherapeutic interventions.

Ecological localization of Tpex cells in the TME

In the context of chronic viral infections, Tpex cells are found predominantly in peripheral lymphoid tissues, including the lymph nodes, bone marrow, and spleen^22,23,33,34. However, the spatial distribution of Tpex cells within tumor tissues is highly complex. Connolly and colleagues³⁵ have identified a significant population of tumor-specific Tpex cells in the tumor-draining lymph nodes (TdLNs) of the mouse KP-NINJA lung cancer (LC) in situ tumor model; intriguingly, their abundance does not correlate with tumor progression. Similarly, both the mouse B16 melanoma tumor model and patients with head and neck squamous cell carcinoma (HNSCC) exhibit a higher proportion of Tpex cells within TdLNs compared to TME^17,36. In patients with NSCLC and HNSCC, Tpex cells are localized primarily within the tertiary lymphoid structures (TLS) of the tumor and show minimal infiltration into the tumor parenchyma^37,38. These observations suggest a potential key role of TLS in sustaining the functionality of Tpex cells. Furthermore, within the TME in renal cancer, these stem-like T cells, marked by positive TCF-1 expression, tend to concentrate in regions densely populated with APCs. Remarkably, these areas functionally mirror TLS, thus emphasizing the significance of such structures in shaping the dynamics of Tpex cells in the intricate landscape of the TME¹⁶.

Current discourse regarding the origin of Tpex cells revolves around 2 primary theories: one positing that Tpex cells undergo self-renewal and proliferation within tumors, and the other positing their continuous migration from TdLNs to tumor tissue. In the mouse KP-NINJA LC in situ tumor model, the differentiation level of tumor-infiltrating CD8⁺ T cells surpasses that in TdLNs. This continuous differentiation pattern from TdLNs to tumors suggests that tumor-infiltrating CD8⁺ T cells are likely to originate from the migration of CD8⁺ T cells in TdLNs³⁵. Examination of patients with hepatocellular carcinoma (HCC) has revealed 2 distinct CD8⁺ T cell subgroups in TdLNs: tumor-specific memory T cells (Ttsm) and Tpex cells. In mouse models of B16 melanoma and Kras^LSL-G12D/+p53^fl/fl lung adenocarcinoma, Ttsm cells in TdLNs have been found to be critical for the response to PD-1/programmed death-ligand-1 (PD-L1) immunotherapy. These cells undergo a linear differentiation trajectory from Ttsm cells to Tpex cells, and subsequently to tumor-infiltrating Tpex and Tex^term cells during tumor control³⁹. Additionally, a noteworthy increase in Tpex cells in TdLNs and their migration to tumors is observed after immunotherapy^28,40. Recent research has further suggested that effector T cells under chronic antigen stimulation transform into Tex^term cells, which exhibit long-term residence within tumors, and the intratumoral anti-tumor response is maintained by sustained recruitment of Tpex cells from peripheral sources. Over time, these Tpex cells migrate to lymph nodes through the lymphatic system²⁸. Consequently, unraveling the intricate ecological positioning and origins of Tpex cells in the TME is imperative for formulating effective immunotherapy strategies.

Regulation of Tpex cells by CD4⁺ T cells

In LC research, the application of multiplex immunohistochemistry staining techniques has revealed intriguing patterns. Tpex cells tend to form clusters near CD4⁺ T cells³⁷. This spatial arrangement hints at potential interactions between Tpex cells and antigen-specific CD4⁺ T cells. The abundant presence of CD4⁺ Th cells within the TME plays a critical role in activating CD8⁺ T cells (Figure 3). This activation is particularly facilitated by the secretion of interleukin (IL)-21, which engages the IL-21R signaling pathway in CD8⁺ T cells and consequently promotes differentiation of Tpex cells into effector cells^6,41. Studies in the mouse B16F10 melanoma model have demonstrated that adoptive therapy with CD4⁺ T cells significantly amplifies the population of CX3CR1⁺ CD8⁺ T cells in the TME, thus effectively controlling tumor growth. In contrast, when the function of CD4⁺ T cells is compromised, the sustained accumulation of effector CX3CR1⁺ CD8⁺ T cells is hampered, as mediated primarily through the IL-21 signaling pathway⁴². Furthermore, in both GL261 (immunoresponsive) and CT2A (immunoresistant) glioblastoma multiforme (GBM) mouse models, effector CD4⁺ T cells have emerged as key contributors maintaining the Tpex cell state within CT2A tumors. This not only enhances the efficacy of αPD-1 therapy but also indicates the critical role of CD4⁺ T cells in orchestrating immunotherapeutic responses. Under CD40 agonist administration to compensate for the deficiency in CD4⁺ T cells in the CT2A TME, the proportion of Tpex cells is sustained. Combined with PD-1 blockade, this dual approach significantly decreases the tumor burden, thus underscoring the potential of this comprehensive strategy⁴³.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Interactions between immune cells and Tpex cells in the TME. (A) DCs in TdLNs create a microenvironment conducive to the survival and development of Tpex cells. After ICI treatment, Tpex cells in TdLNs proliferate and migrate to tumor tissue, and differentiate into Tex^eff cells with functional efficacy. (B) CD4⁺ Th cells maintain the effector functions of Tex cells, and promote the self-renewal and differentiation of Tpex cells into Tex^eff cells through secretion of the cytokine IL-21. (C) IL-10 or TGF-β, secreted by M2 cells, suppresses the function of Tpex cells. In contrast, DCs and M1-type macrophages within the TME synergistically and collaboratively activate Tpex cells, thereby enhancing their anti-tumor effects. (D) Tpex cells in the TME can also enter TdLNs through the lymphatic system.

Regulation of Tpex cells by dendritic cells (DCs)

As guardians of immune activation, DCs not only serve as professional antigen-presenting cells but also are essential in sustaining the function of Tpex cells (Figure 3). Research on chronic viral infections has indicated DC1 cells postpone the differentiation of Tpex cells, thereby extending the window period of CD8⁺ T cell cytotoxic activity. Although DC1 cells might not be the primary drivers of Tpex cell proliferation, they carve out a crucial ecological niche and prevent the overstimulation of Tpex cells⁴⁴. Magen et al.⁴⁵ have investigated the TME in patients with HCC responding favorably to PD-1 therapy and found collaborative effects involving migratory regulatory DCs (mregDCs), C-X-C chemokine ligand 13 (CXCL13)⁺ CD4⁺ Th cells, and Tpex cells orchestrating the activation of T cell clones. This collaboration amplifies the proportion of tumor-specific CD8⁺ T cells, thereby illustrating a dynamic interplay that drives therapeutic success⁴⁵. In the Kras^LSL-G12D/+-p53^fl/fl lung adenocarcinoma model in mice, tumor progression triggers a decline in the migration and activation functions of mature DC1 cells to TdLNs. However, after FLT3L/CD40 agonist therapy, the number and function of DC1 in TdLNs are significantly augmented, thus increasing the numbers of stem cell-like CD8⁺ T cells in TdLNs and enhances their anti-tumor effects⁴⁶. Additionally, in the TME, mregDCs upregulate the expression of major histocompatibility complex-I molecules, and after ICI therapy, CD8⁺ T cells with high CD28 expression on their surfaces further promote the proliferation and differentiation of Tpex cells within the TME^47–49. After αPD-L1 therapy, the number of DCs in TdLNs also increases and consequently enhances the functionality of CD8⁺ T cells, although the detailed mechanism remains to be elucidated⁵⁰.

Regulation of Tpex cells by tumor-associated macrophages

Recent investigations have revealed a distinctive subset of immune cells within esophageal squamous cell carcinoma (ESCC), denoted CD8⁺ Tex-SPRY1. These cells demonstrate anti-tumor functionalities similar to those of Tpex cells and facilitate the polarization of tumor-associated macrophages (TAMs) toward the M1 phenotype, through the secretion of cytokines such as IFN-γ and tumor necrosis factor-α (TNF-α). Notably, M1-type macrophages reciprocally activate CD8⁺ Tex-SPRY1 cells, thereby amplifying their anti-tumor efficacy⁵¹. In the TME of HCC, the status of TAMs is critical in determining the efficacy of αPD-1 therapy. Whereas M2-type macrophages suppress tumor progression, their reprogramming into the M1 phenotype has been shown to bolster the immune response and consequently augment the effectiveness of PD-1 blockade therapy⁵².

ICI therapy

Tpex cells, a distinct subgroup among CD8⁺ T cells, exhibit more robust functionality than Tex^term cells: they inhibit tumor growth more effectively and demonstrate enhanced responsiveness to PD-1 therapy. PD-1 blockade therapy has been shown to augment the proliferative capacity of Tpex cells and to lead to their differentiation into highly efficacious Tex^eff cells^{15,17,33,57–59}. In patients with ESCC, NSCLC, and melanoma, a high proportion of Tpex cells is closely associated with sustained effectiveness and improved clinical outcomes of PD-1 blockade therapy^17,51,57,60. scRNA-seq analysis of tumor samples from patients with NSCLC undergoing PD-1 inhibitor treatment has revealed a significant elevation in Tpex cell levels among those with a positive response to immunotherapy. This increase has been attributed to the expansion of pre-existing Tpex cells in the systemic circulation⁶¹. Furthermore, the CD8⁺ Tex-SPRY1 cell subset, identified by Liu et al.⁵¹ in ESCC, exhibits anti-tumor activity comparable to that of Tpex cells; moreover, the expression level of CD8⁺ Tex-SPRY1 cells positively correlates with the efficacy of PD-1 inhibitor treatment. In a comparison of patients receiving solely radiotherapy vs. a combination of radiotherapy and cytotoxic T lymphocyte antigen-4 (CTLA-4) inhibitor treatment, a significantly elevated proportion of the Tpex cell subset has been observed in the latter group and found to positively correlate with patient survival⁶². For patients with metastatic brain tumors and recurrent GBM undergoing ICI treatment (including αPD-1 or combined αPD-1 and αCTLA-4 therapy), Tpex cells are associated with prolonged survival⁶³. Notably, ICI treatment of recurrent GBM promotes the differentiation of Tpex cells, whereas in metastatic brain tumors, ICI treatment promotes the differentiation of Tex^term cells⁶⁴.

In a mouse model of acute myeloid leukemia, the combination of a bromodomain and extra-terminal domain inhibitor with αPD-1 therapy promotes the proliferation of Tpex cells, thereby exhibiting synergistic effects⁶⁵. Moreover, the heightened expression of lysine-specific demethylase 1 (LSD1) in various cancers contributes to tumor progression and influences the immune response. Targeting LSD1 in mouse models of MC38 colon carcinoma and B16F10 melanoma enhances the durable effects of αPD-1 therapy, primarily by fostering Tpex cell proliferation within the TME and transformation into more efficacious Tex^eff cells. The positive effects on therapeutic outcomes are further underscored by the augmentation of the TCF-1 regulatory network associated with LSD1 depletion⁶⁶. In a separate study, the engineering of particles, specifically R848@M2pep-MPsAFP has been found to successfully reprogram macrophages within the HCC TME, thereby resulting in phenotypic transformation, improved antigen-presenting capabilities, activation of Tpex cells, and significant enhancement of αPD-1 therapy efficacy⁵².

Chen et al.⁴⁷ have demonstrated that concurrent administration of high-mobility group nucleosome binding domain 1 (HMGN1) peptide with αPD-L1 significantly augments the population of Tpex cells within tumors, thereby enhancing the efficacy of αPD-L1 therapy. In investigations of HNSCC, αPD-L1 immunotherapy has been observed to diminish the proportion of Tpex cells in peripheral tissues. Nevertheless, Tpex cells located near DCs tend to show proliferation and differentiation into intermediate-exhausted CD8⁺ T cells (Tex-int). This phenomenon aligns with the increased presence of Tex-int clonotypes circulating in the peripheral blood⁵⁰.

Recent research has demonstrated that the combination of αPD-1 and decitabine increases the proportion of Tpex cells and consequently augments anti-tumor efficacy⁶⁷. However, Mariniello and colleagues⁶⁸ have proposed that the combination of chemotherapy with immunotherapy might influence the effectiveness of these drugs, particularly regarding the Tpex cell subpopulation. Their findings suggest that PD-L1 inhibitors activate Tpex cells and facilitate their differentiation into effector cells, whereas chemotherapy drugs may impede this activation. Notably, the sequence of chemotherapy followed by immunotherapy has been demonstrated to be more efficacious than combination therapy⁶⁸. Importantly, this observation awaits validation in mouse tumor models, and further in-depth mechanistic research is necessary to provide insights for developing clinical treatment strategies.

Cytokine therapy

IL-2, a critical cytokine in regulating T cell maturation and maintaining T cell function, has significant therapeutic promise in cancer immunotherapy^69–73. Given the potential for widespread systemic toxicity and adverse effects associated with IL-2 therapy, current research on engineered IL-2-based drugs focuses primarily on modulating specific aspects of the immune response, through activating CD8⁺ T cells or natural killer cells while inhibiting regulatory T cells^71,74.

In the context of treating chronic viral infections, the combined use of IL-2 and αPD-1 has an effectiveness approximately tenfold greater than that of αPD-1 alone. This synergistic therapy modulates Tpex cells and facilitates their transformation into more efficacious effector T cells⁷⁵. PD1-IL2v, a novel bispecific antibody, simultaneously blocks PD-1 signaling on CD8⁺ T cells and activates IL-2R signaling. This design offers the advantage of selectively activating the IL-2 signal in PD-1⁺ CD8⁺ T cells with tumor antigen specificity. Studies in various mouse models of pancreatic cancer (PK5L1940, FC1242, and Panc02) have indicated that PD1-IL2v treatment significantly increases the population of stem-like Tpex cells in the TME^76,77.

Several studies have explored the therapeutic effects of combining PD1-IL2v with various treatment modalities. In a mouse Panc02 pancreatic adenocarcinoma model, PD1-IL2v treatment leads to a substantially higher proportion of Tpex cells in the TME than single-agent therapy. These cells consequently differentiate into a CD8⁺ TIL subpopulation with enhanced functionality and lower exhaustion (TCF-1⁻ PD-1⁺ TIM-3⁻ GzmB⁺)⁷⁷. When combined with radiotherapy, PD1-IL2v further increases the frequency of Tpex cells and results in more durable anti-tumor responses⁷⁶. Monotherapy with PD1-IL2v in mouse models of pancreatic neuroendocrine tumors and GBM has been found to lead to tumor regression. Combining this therapy with αPD-L1 treatment exhibits stronger anti-tumor effects, with 90% of mice showing no relapse, and a significant increase in the proportion of Tpex cells and their descendant Tex cells in the TME⁷⁸. Current research outcomes on IL-2 drugs underscore the essential role of the Tpex cell subpopulation in activating CD8⁺ T cell anti-tumor responses.

IL-15, another cytokine with substantial potential in cancer therapy, shares the β (CD122) and γ (CD132) chains with IL-2R^79–82. This cytokine plays a crucial role in promoting the proliferation of CD8⁺ T cells, and its absence in the TME has been associated with impaired function of CD8⁺ T cells, tumor relapse, and diminished patient survival^83,84. Previous studies have demonstrated that co-culturing CD8⁺ TILs derived from human renal cell carcinoma with IL-15 enhances the self-renewal capability of these cells⁸⁵.

Combining IL-15 with αPD-1 not only amplifies the proliferation of CD28⁻ PD-1⁺ CD8⁺ TILs exhibiting a Tex^term cell phenotype and CD28⁺ PD-1⁺ CD8⁺ TILs with Tpex cell characteristics, but also intensifies their anti-tumor effects. This action is crucial for the reactivation of CD8⁺ T cells post-PD-1 blockade^48,86. The anti-tumor effect is attributed primarily to the cytotoxic activity of CD28⁻ PD-1⁺ CD8⁺ TILs, which differentiate predominantly from CD28⁺ PD-1⁺ CD8⁺ TILs showing similar characteristics to Tpex cells⁸⁷. The development of pro-IL-15 by Guo⁷⁹ has addressed the toxicity associated with IL-15 while preserving its anti-tumor effects. In the mouse MC38 colon cancer model, pro-IL-15 promotes the proliferation of stem-like TCF-1⁺ TIM-3⁻ CD8⁺ TILs, controlling tumor growth through IFN-γ secretion, particularly when used in combination with PD-L1. This combination exhibits significantly better effects than PD-L1 monotherapy⁷⁹. Moreover, CAR-T cells treated with IL-7 and IL-15 display a notable increase in the expression of memory genes such as Tcf7 and Sell, and show stronger reactivity to PD-1 blockade therapy and sustained anti-tumor effects⁸⁸. This finding opens new avenues for enhancing the efficacy of CAR-T treatment in solid tumors.

Current research on IL-10 has revealed its intricate dual role in immune regulation: it promotes immunosuppressive phenomena and also is crucial in maintaining anti-tumor responses. Previous studies have established a correlation between the concentration of IL-10 in serum and the progression of chronic lymphocytic leukemia (CLL)^89,90. However, recent reports have highlighted the essential role of IL-10R signaling in the maintenance and accumulation of PD-1^intTCF-1⁺ CD8⁺ T cells in CLL. Intriguingly, treatment with αIL-10R may exacerbate the progression of CLL⁹¹.

Another noteworthy study has observed that treatment with IL-10-Fc in a mouse B16F10 tumor model stimulates the proliferation of PD-1⁺ TIM-3⁺ CD8⁺ T cells through IL-10R, thereby promoting their sustained cytotoxic function. Importantly, this proliferative process is not contingent on the differentiation of PD-1⁺ TIM-3⁻ progenitor cells. Instead, it is achieved by enhancing the cells’ anti-tumor action through metabolic reprogramming, particularly the upregulation of oxidative phosphorylation in T cells⁹².

Adoptive cell therapy

CAR-T immunotherapy has shown considerable success in the treatment of hematologic malignancies, yet its efficacy in solid tumors remains limited^93–98. This limitation is largely attributable to the tendency of CAR-T cells to enter a functionally diminished “exhausted” state^99,100. Consequently, maintaining the anti-tumor effects of CAR-T therapy has emerged as a critical challenge in therapeutic development. The transcription factor B lymphocyte-induced maturation protein 1 (BLIMP1), encoded by PRDM1, has been identified as a key player in CD8⁺ T cell exhaustion^101,102. Modulating the interaction between BLIMP1 and TCF-1 enhances the persistence and proliferative capacity of CAR-T cells, thereby offering new strategies for improving the treatment of solid tumors^103,104.

In vivo studies have demonstrated that BLIMP1-depleted CD19-CAR-T cells exhibit stronger and more persistent anti-tumor effects in certain cancer models, albeit with a potential decrease in their immediate ability to release cytotoxic molecules, including GzmB and perforin^105,106. In PC3-PSMA and NALM-6 B cell acute lymphoblastic leukemia (B-ALL) models, PRDM1^−/−NR4A3^−/−CAR-T cells have shown greater tumor control and a lower tendency toward the Tex^term cell phenotype compared with traditional CAR-T cells¹⁰⁵. The specific knockout of Regnase-1 in CAR-T cells designed for ALL research promotes the development of Tpex cells, decreases their exhaustion, and consequently enhances sustained anti-tumor action¹⁰⁷. Recent studies have also identified that high expression of ETS1 in CD8⁺ T cells significantly inhibits the transformation of Tpex cells into Tex^term cells—a suppression dependent on decreased BATF activity. Treating the hCD19-B16 melanoma model with ETS1-deficient hCD19-CAR-T cells markedly enhances therapeutic efficacy¹⁰⁸.

Cancer vaccine therapy

Therapeutic cancer vaccines are aimed at activating the immune system to combat tumors. Although most trials have shown modest effects, better therapeutic efficacy has been observed under specific conditions, such as strong T-cell responses or minimal immunosuppression⁹. scRNA-seq analysis of mouse B16 tumors has revealed 2 main CD8⁺ T cell populations in the TME post-vaccination: progenitor-exhausted (TCF-1⁺) and transiently exhausted (CX3CR1⁺ GzmB⁺). These findings suggest that cancer vaccine inoculation is conducive to establishing a persistent CD8⁺ T cell anti-tumor environment^15,109.

Preclinical in vivo experiments in mice have indicated synergistic anti-tumor effects of vaccine therapy combined with ICIs. In melanoma, combining a vaccine with αPD-L1 achieves stronger tumor control than monotherapy with αPD-L1¹⁰⁹. In the mouse MC38 colon cancer model, an adenovirus vector vaccine combined with αPD-1 immunotherapy effectively promotes the accumulation of Tpex cells in TdLNs. These cells then infiltrate tumors and differentiate into effector cells. Moreover, compared with monotherapy with αPD-1, the combination of the vaccine with αPD-1 exhibits stronger clonal capacity of these specific T cells and better anti-tumor effects⁴⁰.

To enhance the precision of vaccine delivery in vivo, Baharom and colleagues^110,111 have developed a nanoparticle vaccine assembled from neoantigen peptides linked to Toll-like receptor 7/8 agonists. Treatment of MC38 tumor-bearing mice with this nanoparticle vaccine significantly increases the proportion of Tpex cells in the spleen, thus demonstrating robust anti-tumor activity^110,111. Crucially, the injection method has been observed to influence the proportion of Tpex cells induced by this vaccine. Intravenous injection yields a greater proportion of Tpex cells (80%) than subcutaneous injection (35%). Notably, the intravenous vaccine injection not only produces a significantly higher percentage of Tpex cells but also demonstrates a more pronounced anti-tumor effect than subcutaneous injection¹¹¹.