Although studies have shown that PD-L1/PD-1 inhibitors for immunotherapy are effective for numerous cancer types, such as melanoma and lung cancer (20,25,26), the use of PD-L1/PD-1 inhibitors for ovarian cancer treatment, particularly for germ cell tumors, remains controversial (11,17). PD-L1 expression has been studied among different tumor types, including ovarian cancers; however, studies investigating PD-L1 and the tumor microenvironment in germ cell tumors are very rare (27,28).
Our previous study on epithelial ovarian cancer showed that PD-L1 expression was not associated with poor prognosis in patients with epithelial ovarian cancer (11). However, another study noted increased survival in patients with high-grade serous ovarian carcinoma (HGSOC), which is associated with high levels of PD-L1 expression (29).
In this study, we investigated PD-L1 expression and the tumor microenvironment, including tumor-infiltrating lymphocytes, stem cell markers, genetic alterations, and the correlation of PD-L1 with clinical outcomes in patients with ovarian germ cell tumors. Consistent with a previous study on testicular germ cell tumors (19), we demonstrated that PD-L1 was frequently expressed in dysgerminomas, choriocarcinomas, and in dysgerminoma areas of mixed germ cell tumors with an active tumor microenvironment. On the other hand, PD-L1 was not expressed in pure yolk sac tumors in any of the tested samples.
The high expression of PD-L1 in dysgerminoma and pure choriocarcinoma alone or in mixed germ cells suggests the possibility of anti-PD-L1 treatment for patients with chemotherapy-resistant tumors. The response has been described as variable in scattered case reports. Some case reports, for example, described a durable response of choriocarcinoma to anti-PD-L1 treatment (30,31), while others reported no response (32,33). The limited antitumor activity of pembrolizumab in patients with advanced germ cell tumors was observed in clinical trials reported by Tsimberidou (32). However, the generalizability of the results of that study is limited due to the small number of participants and the fact that the majority of the patients had mixed germ cell tumors that were treated with anti-PD-L1 regardless of its expression in the tumor cells. Nevertheless, we showed in our study that PD-L1 expression varies in each component of germ cell tumors with multiple histologic types.
In the tumor microenvironment, tumor-infiltrating lymphocytes, cancer stem cells and other factors all play important roles in the survival of cancer cells and evasion of the immune response (34–36). The presence of CD8+ and CD4+ TILs is a prognostic factor for increased survival in many cancers, including ovarian cancer (34,37,38). These cells aid in controlling the growth of solid tumors by recognizing cancer antigens or overexpressed self-antigens (39)
Our investigations of TILs revealed that OVGCTs that have high expression of PD-L1 also have high expression of CD8, which is expressed by cytotoxic infiltrating lymphocytes; in particular, dysgerminomas have shown enrichment of TILs expressing both CD8 and CD4. Similarly, the T-cell lymphocyte-rich microenvironment was previously documented in testicular seminomas, the counterpart of dysgerminomas (Lobo et al., 2019; Sadigh et al., 2020).
The correlation between CD8 positivity and PD-L1 expression in the current study reflects the role of the microenvironment in regulating the expression of PD-L1. The cytokine-mediated upregulation of PD-L1 on tumor cells in inflammatory-rich tumors plays an important role in suppressing antitumor activity (40). Furthermore, high expression of the TILs in cancer tumors indicates a good prognosis (41).
The other notable cell populations in germ cell tumors are stem cell populations. Stem cells may present different biomarkers in different tissues. Scientists are regularly investigating the role of new stem cell biomarkers that are associated with specific types of cancer (42). We found that PD-L1 expression is associated with the cancer stem cell population in ovarian epithelial tumors (11).Consistent with our findings, a study on breast cancers showed that high PD-L1 expression was associated with stem cells expressing CD44 and OCT4 (43). Although there was no significant difference in the expression of stem cell markers (CD44, LRG5, and ALDH2) or PD-L1 in our small study population of OVGCTs, the statistical analysis of TCGA data for 149 TGCT patients revealed significant differences in the expression of CD44 and LGR5 in PD-L1-positive patients. However, our investigation revealed that OCT3/4 expression was significantly associated with PD-L1 expression in patients with dysgerminoma. OCT 3/4 is known to be expressed in 100% of dysgerminomas and embryonal carcinomas (44), the more undifferentiated types of germ cell tumors, but not in yolk sac tumors (45). OCT3/4 is a marker for normal pluripotency of stem cells but also plays a role in maintaining the stemness of cancer stem cells (46). A recent study revealed the role of PD-L1 in upregulating OCT3/4 in breast cancer through the PI3K/AKT signaling pathway, which indicates the role of PD-L1 in regulating stemness in cancer tissue (43). Similarly, in TGCTs, particularly seminomas, CD44, CD117 and other cancer stem cell markers have been studied for their role in metastasis, tumor progression, and chemotherapy resistance (47). C-KIT (CD117), which regulates germ cell differentiation, is well studied in TGCTs. Studies have shown that some OVGCTs, specifically dysgerminomas, harbor c-KIT even without mutations in the c-KIT gene. Stem cells expressing c-KIT, along with OCT3/4, might increase the survival of underdeveloped oocytes, leading to the formation of these tumors (48). Furthermore, our study revealed that the expression of PD-L1 might be associated with stem cells in dysgerminomas. However, disruption of c-KIT pathways is an area of interest in developing new immunotherapies for germ cell tumors (49).
In addition, genetic studies were performed to link the expression of PD-L1 with genetic alterations in germ cell tumors. We found that patients with dysgerminoma, which is characterized by higher PD-L1 expression, presented a greater number of somatic genetic alterations than patients with yolk sac tumors. The infrequent expression of PD-L1 in yolk sac tumors can reflect the molecular events in yolk sac tumors, which have been reported to have a low tumor mutational burden. A molecular study of 10 ovarian yolk sac tumors revealed that all the tumors were microsatellite stable and exhibited a low mutational burden (50,51). Furthermore, our investigation of TCGA data for TGCTs showed that compared with other germ cell subtypes, dysgerminomas have the highest mutation counts and tumor mutational burden. Based on these findings, we anticipate that genetic alterations influence tumor lymphocyte infiltration and increase PD-L1 expression, which might positively affect treatment response.
Although the statistical analysis of the overall survival of patients with PD-L1 expression in the small group was not definitive our investigation showed that patients with dysgerminoma had the highest survival rate and lower recurrence, which is consistent with the findings of a previous study (52). In contrast, patients with yolk sac tumors have the worst survival rate among all other types.
Like other studies, some limitations were encountered in this study. The main limitation of our study is the low number of tested samples of ovarian germ cell tumors, which is expected in rare diseases (53). Although patients with dysgerminoma are known to have a better prognosis than patients with yolk sac dysgerminoma, our analysis of a small population did not reveal a significant difference in 10-year survival. Therefore, the association between high expression of PD-L1 and patient survival in patients with dysgerminoma was not analyzed. Large populations of patients with dysgerminomas and yolk sac tumors should be studied in the future to investigate survival with and without PD-L1 expression. Furthermore, molecular genetic studies are needed in larger numbers of ovarian germ cell tumors to investigate the associations of genetic biomarkers and tumor mutational burden with the immunotherapy response in ovarian germ cell tumors.