Germ cell tumors can arise not only in the gonads but also in other midline regions of the body, including the central nervous system. Primary central nervous system germ cell tumors (CNS GCTs) are a rare form of intracranial malignancy, typically classified into two groups based on histopathological characteristics: germinomas and non-germinomatous germ cell tumors (NGGCTs). Germinomas comprise approximately 64-75% of CNS GCTs, while more than half of NGGCTs are mixed germ cell tumors containing various histological components (1). Germinomas are highly responsive to both radiotherapy and chemotherapy, and current treatment strategies rely on a combination of these modalities. However, significant clinical challenges persist. These tumors often occur in young patients, and the adverse effects of radiotherapy and chemotherapy, such as hormonal imbalances and long-term neurological and cognitive dysfunction, are a serious concern (2–4). Furthermore, some patients exhibit treatment resistance and disease recurrence, leading to poor disease control and limited improvements in prognosis following recurrence (5). This underscores the urgent need for new risk stratification methods and innovative therapeutic approaches to improve patient outcomes.

In recent years, the tumor microenvironment has gained significant attention in cancer research due to its pivotal role in tumorigenesis and malignant progression (6). Numerous studies have documented the presence of abundant immune cell infiltration and immune checkpoint expression in gonadal germ cell tumors (7–12). Similarly, research on CNS GCTs has identified comparable patterns of immune infiltration (13–17). Immune checkpoint molecules, such as the PD-1/PD-L1 axis and CTLA-4, are crucial in modulating T cell responses to tumor antigens (18, 19). The interaction between these molecules and their ligands can suppress anti-tumor immune responses, facilitating tumor immune evasion (18–20). As a result, immune checkpoint inhibition therapy has become a promising strategy for treating malignant tumors. However, research specifically focused on CNS GCTs remains limited, and the relationship between immune cell infiltration within the tumor microenvironment and tumor development, progression, and patient prognosis is not yet fully understood. Moreover, data on CTLA-4 expression and its influence on the prognosis of CNS GCT patients are sparse. Further investigation and clinical studies are needed to evaluate the potential of immune checkpoint inhibition therapy for CNS GCTs.

To address these gaps, we conducted a comprehensive analysis of T cell infiltration characteristics and immune checkpoint expression in our CNS GCTs cohort. We explored the correlation between these factors and their association with clinical characteristics and patient prognosis. Our findings unveiled the distinct tumor microenvironment of CNS GCTs, providing insights into the potential application of immune checkpoint blockade (ICB) therapy for this disease.

In recent years, significant progress has been made in understanding the pathogenic mechanisms of CNS GCTs, particularly in genetics, including chromosomal variations and genomic alterations (21–27). As a unique and complex tumor type, increasing attention has been directed toward the tumor microenvironment (TME) of CNS GCTs. A deeper understanding of the TME is essential for advancing therapeutic approaches.

In this study, we observed significant T cell infiltration in the majority of both germinoma and NGGCT cases, with notable heterogeneity in the extent of T cell infiltration across individual cases. Several studies have confirmed that a prominent feature of the germinoma tumor microenvironment is the infiltration of large numbers of immune cells. Saito et al. reported that 70-80% of infiltrating lymphocytes in germinomas are T cells (16). Another study found that T cells play a major role in the tumor microenvironment, accounting for about 30.9% of all lymphocytes, with considerable variation in immune cell infiltration across cases (13). Additionally, CD3+ T cells were found to dominate the immune microenvironment of germinomas, with helper T cells (Th cells) outnumbering cytotoxic T cells (Tc cells) (17). Similarly, abundant immune cell infiltration, particularly a high proportion of T cells, has been observed in germ cell tumors of other systems, such as testicular seminomas (7, 10, 11). This extensive immune cell infiltration is considered one of the “two cell patterns” in the histopathology of germ cell tumors. Regarding immune checkpoints, our study revealed that CTLA-4 was positively expressed in over 60% of germinoma cases and more than 40% of NGGCT cases. Li B et al. also reported significant CTLA-4 expression in the immune-hot subtype of CNS GCTs, 94.4% of which were germinomas (27). A study on testicular germ cell tumors (TGCTs) also showed high CTLA-4 expression in both seminoma and non-seminoma cases (seminoma: 87.2%; non-seminoma: 79.5%) (10), a higher proportion than observed in our cohort. Furthermore, we found that PD-1 and PD-L1 were highly expressed in both germinoma and NGGCT cases, consistent with previous studies on intracranial and extracranial germ cell tumors (10, 13–15).

Amid the significant T cell infiltration and immune checkpoint expression observed, CNS GCTs exhibit clear features of an immunosuppressive tumor microenvironment. First, regulatory T cells (Tregs) typically suppress immune responses by inhibiting the activity of functional T cells within the tumor microenvironment (28, 29). In CNS GCTs, we observed a substantial presence of CD4+Foxp3+ Treg cells, with some Foxp3+ cells also expressing CTLA-4, suggesting that these Tregs actively contribute to the immunosuppressive microenvironment. The expression of CTLA-4 may further amplify this effect, particularly in germinomas. Second, the infiltration levels of different T cell subsets were positively correlated with the expression of immune checkpoints (CTLA-4, PD-1). Immunofluorescence co-staining revealed that a significant number of CD8+ T cells expressed PD-1 on their surface, while PD-L1 expression was closely associated with malignant subtypes (except teratomas). These findings indicate that the high activity of the PD-1/PD-L1 axis allows tumor cells to evade anti-tumor immune responses, likely playing a more prominent role in NGGCTs. Moreover, this heightened activity may lead to effector T cell exhaustion, ultimately resulting in an immune-tolerant microenvironment. Lastly, in germinomas, the proportion of CD4+ Th cells exceeded that of CD8+ Tc cells, possibly indicating that cytotoxic T cells are not fully functional, leading to poor tumor control. However, this pattern was not observed in NGGCTs, likely due to the high heterogeneity of these tumors, which are often mixed and comprised of multiple histological subtypes. Given the complexity of their immune microenvironment, further investigation into the immunological dynamics of NGGCTs is warranted.

The tumor microenvironment, characterized by high T cell infiltration and significant immune checkpoint expression in CNS GCTs, is not only closely linked to the tumor’s biological behavior but may also directly impact patient outcomes and prognosis. In this study, we observed differences in PD-1 expression levels between genders, although the positive expression rate did not show a significant difference. This suggests that gender may play a distinct role in shaping the immune microenvironment of CNS GCTs. Previous studies have highlighted several gender-related factors influencing the development and progression of CNS GCTs, including chromosomal alterations (23, 25), gender-specific differences in mutation rates within the MAPK pathway (21), and mutations in genes associated with sex hormone receptors (26). Additionally, the potential role of sex hormones in modulating the immune system may contribute to these differences. Future research involving larger cohorts could help determine whether gender-specific disparities exist within the tumor immune microenvironment. Such studies would also further elucidate the impact of gender on tumor immunity and the response to immune checkpoint inhibitors, which is critical for developing personalized treatment strategies. We also observed differences in PD-1 expression between primary and recurrent tumors, as well as the impact of preoperative radiotherapy and chemotherapy on CTLA-4 expression, suggesting that these factors may influence the immune microenvironment. PD-1 may play a more prominent role in primary tumors. As a crucial regulator in the immune response process (30), its upregulation in primary tumors could serve as an indicator of immune activity. However, in recurrent cases, its role may be reduced or supplanted by other mechanisms. This implies that immunosuppressive treatments targeting PD-1 may be more effective in primary tumors, while recurrent cases may require alternative therapeutic strategies to address more complex immunosuppressive and immune evasion mechanisms. Moreover, preoperative treatments may alter the immune status of the tumor microenvironment, potentially affecting the response to immunotherapy. Future studies should explore the impact of these factors on immunotherapy, investigating the differences and similarities in treatment strategies for primary versus recurrent diseases, as well as determining the optimal timing for treatment.

Notably, our study revealed a significant correlation between tumor markers (β-HCG and AFP) and the proportion of infiltrating CD4+ and Foxp3+ cells, as well as CTLA-4 expression. Specifically, the negative correlation between β-HCG and AFP levels and the infiltration of CD4+ and Foxp3+ cells, as well as CTLA-4 expression, suggests that tumors with lower levels of these markers may be associated with a more pronounced immunosuppressive microenvironment. We hypothesize that lower marker levels could be linked to enhanced immune evasion mechanisms, potentially reflected in increased Treg (regulatory T cell) infiltration and elevated CTLA-4 expression, thereby fostering a stronger immunosuppressive environment. This type of microenvironment aids tumor cells in evading host immune surveillance. In contrast, tumors with higher marker levels may employ alternative mechanisms to evade immune responses, which could involve components of the humoral or innate immune systems (9, 13, 17), or be linked to specific molecular features of the tumor, although these mechanisms remain to be fully elucidated. During pregnancy, HCG exhibits a range of immunoregulatory properties, not only modulating the immune microenvironment within the reproductive system but also affecting the function of peripheral immune cells (31, 32). As for AFP, studies have shown that it can significantly influence the immune microenvironment in liver cancer through multiple mechanisms, including its effects on liver cancer stem cells, various immune cells, fibroblasts, and the endothelial-to-mesenchymal transition process (33). However, in germ cell tumors, only a limited number of prospective cohort studies have confirmed the role of β-HCG and AFP in monitoring treatment response and predicting prognosis (34, 35). Their influence on shaping the immune microenvironment remains largely underexplored. Our findings underscore the heterogeneity and complexity of the tumor microenvironment in germ cell tumors. Therefore, understanding the potential impact of tumor markers on the immune microenvironment could offer valuable insights for developing novel immunotherapies. Future research into how tumor markers influence the immune microenvironment in germ cell tumors is both essential and urgently needed.

Previous studies have shown that different levels of lymphocyte infiltration, their subtypes, and immune checkpoint expression can impact patient prognosis in both intracranial and extracranial germ cell tumors. For instance, high lymphocyte infiltration in germinomas has been associated with better outcomes (13), and in TGCTs, high T cell infiltration, low Treg infiltration, and an active PD-1/PD-L1 axis have been linked to favorable prognosis (7, 8, 10, 12). However, in our study, patients with high CD3+ infiltration had worse PFS and RFS, suggesting that increased non-specific T cell infiltration may not correlate with anti-tumor effects. This could be due to an immunosuppressive tumor microenvironment that impairs T cell function. Conversely, patients with low Foxp3+ Treg infiltration had worse OS, PFS and RFS, indicating that Foxp3+ cells may play a critical role in anti-tumor immunity and are important for controlling CNS GCT progression (28, 36). These findings warrant further investigation, particularly regarding the specific roles of CD3+ and Foxp3+ cells in tumor progression and their interactions with tumor cells and other immune cells.CD4+ and CD8+ T cells are generally regarded as the central players in anti-tumor immunity (37, 38). In this study, we observed that low CD4+ and CD8+ cell infiltration was associated with poorer PFS, with low CD8+ infiltration particularly linked to worse OS. This further suggests that CD4+ and CD8+ T cells contribute to tumor immune defense. Although the infiltration characteristics of various T cell subtypes were correlated with patient prognosis to varying degrees, none demonstrated significant independent prognostic value. This suggests that specific immunosuppressive mechanisms, such as Treg cell infiltration and high immune checkpoint expression, may be present in CNS GCTs, preventing T cells from exerting their full anti-tumor effects. The presence and underlying mechanisms of these suppressive factors warrant further investigation. Patients with negative CTLA-4 expression showed worse PFS and RFS, although CTLA-4 was not identified as an independent prognostic factor. Similarly, PD-L1 expression did not significantly correlate with patient prognosis, a finding that requires further validation.

However, a key discovery in our study is that PD-1 expression was an independent prognostic factor for PFS and RFS in CNS GCT patients. This not only underscores the potential of PD-1 in risk stratification and subgroup identification of patients, but also emphasizes the promise of anti-PD-1 immunotherapy for CNS GCT, especially in patients with high T-cell infiltration, positive PD-1 expression, and resistance to other treatments. At the same time, considering the current research foundation and treatment status, it is important to interpret our results with caution. Although we observed T cell infiltration and a correlation between immune checkpoints, particularly PD-1, and the clinical features and prognosis of CNS GCT patients, it remains unclear whether positive or elevated expression of immune checkpoints signifies the suitability of immune checkpoint inhibitor therapy. So far, the expression of immune checkpoints, including CTLA-4 and PD-L1, has not been able to predict the response of germ cell tumor patients to immune checkpoint inhibitors (10, 39). A similar lack of predictive value has also been observed in renal-cell carcinoma (40, 41). This is due to our incomplete understanding of the intrinsic relationship between immune checkpoint expression and the response to immunotherapy. Several attempts have been made to treat extracentral germ cell tumors with ICB (42–46), with results that have been both promising and discouraging. While some TGCT patients have not experienced benefits from immunotherapy, other studies have yielded encouraging results. Reports have indicated that monotherapy with PD-1 monoclonal antibodies in germinoma patients, as well as combination therapy with PD-1 and CTLA-4 monoclonal antibodies in CNS NGGCT patients, both demonstrated some therapeutic efficacy (27, 47). In a study involving PD-1 antibody treatment for 7 patients with extracranial GCTs who relapsed after high-dose chemotherapy and stem cell transplantation, 3 patients received at least 6 months of PD-1 therapy, and two of them achieved long-term tumor responses (43). Thus, we are optimistic about the potential future clinical application of immunosuppressive therapy in CNS GCT patients. Therefore, whether immune checkpoint inhibitor therapy can genuinely improve the prognosis of CNS GCT patients still requires validation through future clinical trials. Secondly, the differences in immune microenvironment characteristics between responders and non-responders to ICB therapy need to be elucidated. Further research is required to identify the specific features that can reliably predict which patients are more likely to benefit from immune checkpoint inhibitor treatment. The precise assessment of the applicability of immunosuppressive therapy, along with accurate patient identification methods, will provide the solution to this issue. This study employs IHC to assess the expression and levels of immune checkpoints in CNS GCT patients, a method not yet part of routine clinical practice. However, due to its applicability and cost-effectiveness, coupled with the findings of this study, it could be further validated in the future for its potential role in patient risk stratification. In addition, the optimal use of immunosuppressive agents in the future requires further investigation. Given the presence of the blood-brain barrier (BBB), the ability of immune checkpoint inhibitors to penetrate the BBB must be considered. Some studies have explored the use of ICB drugs to treat recurrent high-grade gliomas through intraoperative intracranial injection and postoperative Ommaya reservoir delivery (48), offering a novel approach for further investigation. Should ICB be used as monotherapy or in combination with other immune checkpoint inhibitors or agents (such as cytokines or immunosuppressive steroids), and how might this impact the therapeutic efficacy of ICB? Additionally, how can we mitigate immune-related adverse events (irAEs) associated with ICB therapy (49)? These are important questions that require further investigation in future research. Moreover, for patients who cannot undergo immunosuppressive therapy or who experience treatment failure, Car-T therapy—demonstrating some efficacy in TGCT treatment (39)—may offer an alternative immunotherapy option.

In addition to the inherent limitations of retrospective studies, another significant limitation of this study is the reliance on FFPE specimens due to the lack of sufficient fresh tissue samples, given current diagnostic and treatment practices. This constrained our ability to conduct detailed analyses of the specific types, functional changes, and spatiotemporal evolution of infiltrating T cells in the CNS GCT tumor microenvironment, as well as the dynamic changes in immune checkpoint expression. Furthermore, FFPE samples may lead to antigen modification, degradation, or denaturation, which could affect the IHC staining results. Additionally, methods like IHC and IF are inherently prone to subjective interpretation. To address these issues, we employed a robust antigen retrieval method, including appropriate control groups, and utilized as objective a quantification approach as possible to minimize potential biases. Moreover, the follow-up period for the cases included in this cohort was relatively short, which may limit the ability to make long-term prognostic interpretations, particularly regarding PFS and RFS. The majority of NGGCT cases in this study were mixed germ cell tumors, which may not fully represent the diversity of single histological subtypes found in NGGCTs. This limitation reduces the generalizability of the findings to all NGGCT subtypes. Since mixed germ cell tumors constitute a large proportion of NGGCT cases, future studies with larger cohorts and longer follow-up periods are needed to better assess the prognostic significance of immune infiltration and checkpoint expression across different histological subtypes and to evaluate their long-term prognostic value. Due to the scarcity of CNS GCT samples, future collaborations across multiple centers and regions will be essential for further advancing research in this area.

In conclusion, our study highlighted the characteristics of T cell infiltration and significant immune checkpoint expression in CNS GCTs, revealing the highly heterogeneous and immunosuppressive nature of the tumor microenvironment. We demonstrated that T cell infiltration and immune checkpoint expression are closely associated with patients’ clinical characteristics and prognosis. Notably, PD-1 expression was identified as an independent prognostic predictor, providing a foundation for improving risk stratification in CNS GCT patients and supporting the potential for future clinical applications of immune checkpoint inhibitors, such as PD-1 monoclonal antibodies.