EC has been broadly divided into two groups based on histomorphology since 1983. Type I is endometrioid tumors, which are the most common subtype, and most of the tumors express the estrogen receptor (ER) and have a favorable prognosis. In contrast, type II is estrogen-independent and mainly represents a serous carcinoma with a poor outcome (9, 10). However, the classification based on tumor histopathology is subjective, has reproducibility challenges, and high-level EC cannot be reliably based on histological criteria. Misdiagnosis is possible due to the presence of mixed high-grade histological components. Biological information to enhance diagnosis is urgently required in this situation.

To overcome these limitations, the Cancer Genome Atlas (TCGA) Research Network, based on mutational burden and somatic copy-number variations, revealed four molecular subtypes of EC with distinct prognoses: DNA polymerase epsilon (POLE) ultramutated (POLEmut) with an excellent prognosis, microsatellite instability hypermutated (MSI-H) with an intermediate prognosis, copy number high (CNH) with the worst prognosis, and copy number low (CNL) with an intermediate prognosis (11). However, the TCGA studies only include endometrioid and serous EC, and the additional cost of entire genome sequencing greatly limits its practical application. Subsequent studies have found cheaper and easier surrogates. To broaden the utility of TCGA classification, Talhouk proposed a practical molecular POLE mutation-based classification model, the Proactive Molecular Risk Classififier for Endometrial Cance (ProMisE) (12). In this study, immunohistochemistry (IHC) for mismatch repair (MMR) proteins and immunohistochemical staining for p53 were used instead of molecular assessments of mutational burden and copy number variations. Therefore, the ProMisE model classified four subgroups: mismatch repair deficient (MMRd), POLE-ultramutated (POLEmut), p53-wild type (p53wt), and p53-abnormal (p53abn) (13, 14). In 2016, Stelloo et al. validated a more pragmatic, cost-effective, and clinically applicable molecular classification system called the Translational Research in Post-Operative Radiation Therapy in EC (TransPORTEC) system. Therefore, EC can be subdivided into four subgroups: the POLE-mutated, MMRd, p53abn, and no specific molecular profile (NSMP) (15, 16). POLE-mutant ECs have a highly favorable prognosis and do not require adjuvant treatment. MMRd ECs have an intermediate prognosis. Numerous recent studies have focused on the immunotherapy for EC (17, 18). And reports have demonstrated these two groups may benefit from immunotherapy (19, 20). P53-mutant ECs are typically associated with more advanced stages, higher rates of lymphatic vascular space infiltration (LVSI), and diverse pathological types. Most P53-mutant ECs are serous adenocarcinomas and have a poor prognosis. The NSMP group was the most frequent, and its prognosis was uncertain (16), as shown in Figure 1 .

In view of the important prognostic and therapeutic implications of the European Society of Gynecological Oncology (ESGO), European Society of Pathology (ESP), and European Society for Radiotherapy and Oncology (ESTRO) guidelines, new prognostic risk groups were defined of adopting the molecular classification for EC treatment in 2020 (21). ECs with POLE-mutant confined to the uterus are regarded as low-risk, whereas p53-abnormal ECs are regarded as “high-risk” in the presence of invasion (22).

The tumor microenvironment plays a crucial role in the occurrence and development of malignant tumors, and has a significant impact on the clinical outcomes of EC (23). Immune cells in a normal endometrium are vital for protection against external pathogens. The immune cells, mesenchymal cells, endothelial cells, inflammatory mediators, and extracellular matrix (ECM) molecules are components of the ITME. Infiltrating stromal and immune cells are the major components of the ITME and play a significant role in the biological behavior of cancer (24). Related studies have shown that many immune cells and cytokines are found in EC tissues and can stimulate an endogenous anti-tumor immune response. Therefore, these patients may benefit from immunotherapy (25–27).

Tumor-infiltrating immune cells can exert an anti-tumor killing effect through a specific cellular immune-mediated immune response. Some subsets can control the growth and progression of tumors, whereas others can promote immune suppression and help tumor cells escape the immune system. Finally, the body reaches a balance between the tumor immune response and tumor immune escape (28, 29). Tumor-infiltrating immune cells include T cells, tumor-associated macrophages (TAMs), natural killer (NK) cells, dendritic cells (DCs), and B cells. ( Table 1 ) T cells can be divided into two subgroups: CD4⁺ and CD8⁺. In terms of immune function, CD4⁺ T cells are mainly regulatory T cells (Tregs), which assist humoral and cellular immunity by secreting various lymphokines. CD8⁺ T cells mainly include inhibitory and cytotoxic T cells. CD8⁺ inhibitory T cells can inhibit humoral immunity. CD8⁺ cytotoxic T lymphocytes (CTLs) can lead to the apoptosis of target cells by releasing perforin, granzyme killer cells, or through the Fas/FasL pathway (30, 31). Cytokines are immune checkpoint molecules and inflammatory factors. The immune checkpoint molecules include members of the B7 family, human leukocyte antigen 1(HLA 1) and lymphocyte activation gene-3 (LAG-3) are shown in. Table 2 (61–64). Inflammatory factors includes interleukin−6 (IL−6), IL-11, tumor necrosis factor a (TNF-α), Cox-2 and interferon.

Inflammatory cells can release IL-6 and TNF-α cytokines to affect cell proliferation signals. Cytokines can promote rapid cell proliferation and differentiation and increase the probability of abnormal mutations. For example, TNF-α at high concentrations can kill EC tumor cells. Meanwhile, EC tumor cells can release TNF-α causing DNA damage and abnormal cell repair (91). TNF-α can enhance tumor invasion by affecting angiogenic factors. Inflammatory cells can also promote the production of cyclooxygenase 2 (COX-2), which participates in tumor generation and invasion by affecting IL-6 and IL-11. Furthermore, inflammatory cells can produce local active nitrogen clusters and reactive oxygen clusters, which can cause cell DNA damage, and easily form abnormal mutations that can promote growth and tumor invasion (92). Che et al. (93) suggested that IL-6 promotes autosynthesis and EC growth via autocrine feedback mediated by the ERK-NF-κB signaling pathway. High expression rates of colony stimulation factor-1 (CSF-1), TNF-α and IL-6 were associated with poor prognosis in EC patients. Lay et al. (92) suggested that IL-11 promotes EC progression by activating STAT. These cytokines may be involved in the immune escape in EC.

ScRNA-seq is a method to measure the expression levels of all genes from individual cells, and reveals heterogeneity at cell level (94, 95). Liu et al. (96–99) applied ScRNA-seq to tumor tissues from patients of cervical cancer to explore the ITME. With the aid of scRNA-seq, the understanding of the ITME in gynecological tumors is improved. Several studies have been studied on EC using scRNA seq. Huang et al. (100) revealed that the small GTPase 3 (RAC3) was specifically distributed in EC tumor cells compared to normal tissues. High levels of RAC3 in EC tissues were reversely associated with CD8⁺ T cell infiltration. Furthermore, RAC3 accelerated tumor cell proliferation and inhibited its apoptosis, without impacting cell cycle stages. Importantly, silencing RAC3 improved the sensitivity of EC cells to chemotherapeutic drugs (100). Yu et al. (101) suggested EC cells can confer malignant phenotype to endothelial cells by Midkine (MDK)-nucleolin (NCL) signal and NCL is associated with suppressed immune activity. EC cells may shape ITME by inhibiting immune cells via MDK-NCL signal. Guo et al. (102) suggested the percentage composition of monocytes, DC and mast cells were higher in paratumor than in tumor. On the other hand, the percentage composition of macrophages was higher in tumor than in paratumor. They found that tumor infiltrating macrophages were associated with increased OS. Wu et al. (103) identified three macrophage subsets, and two of them showed tissue-specific distribution. The tumor-enriched macrophage subset was found to predict immunotherapy responses in EC. Furthermore, six genes were selected from macrophage subset markers that could predict the survival of EC patients, SCL8A1, TXN, ANXA5, CST3, CD74 and NANS, and a prognostic signature was constructed. Further research is needed to study ITME through scRNA-seq.

To date, studies on ITME in EC have mostly focused on POLE mutations and the MMRd subtype. According to the TCGA data, patients with POLE mutations showed the highest tumor mutation burden (TMB), followed by those with the MSI-H subtype. Reports have illuminated that TMB is highly associated with tumor-infiltrating immune cells, PD-L1 expression and patients’ prognosis in EC (104–106). Meanwhile, studies have also reported that POLEmut and MMRd EC have a high number of CD8+ CTLs expressing PD-1 (29). This is the key mechanism of immune tolerance in POLEmut and MMRd EC, as PD-1 binds to PD-L1 to restricts CD8+ CTLs function. Furthermore, previous studies have shown that CD8⁺ CTLs generally indicate good prognosis (26). These results are consistent with each other, as in clinical practice, POLEmut and MMRd have relatively favorable prognoses and can benefit from immunotherapy. Fusco et al. (107) revealed that a possible hypothesis supporting these results is that tumors with more somatic mutations produce higher levels of immunogenic antigens. Consequently, these tumors can be detected by CD8+ CTLs and are less likely to progress to a late stage. Guo et al. (20) explored 123 MSI EC and found that MSI tumors were enriched with CD8+ CTLs, Treg cells, fewer M2 macrophages, activated DCs, and a higher trend of CD20⁺ B cells infiltration. Patients with MSI EC were identified more often in the early stages, had a lower age, and better survival. Julie et al. (44) found CTLA4, PD-1, PD-L1, TIM-3 and IL-6 expression significantly increased in all EC grades. The number of CD4⁺ T cells was similarly increased in all EC grades, whereas the number of CD8⁺ CTL was only increased in the grade 1 ECs and decreased or remained unchanged in others grades. Daniel (108) studied the POLEmut type and found similar results to those of the prominent CD8⁺ CTL present in POLE-mutant. Mohammad (109) performed a retrospective study and found that a combination of PD-L1 positivity and MMR deficiency may be associated with aggressive features such as LVSI. Dai et al. (7) conducted a retrospective study to comprehensively analyze the ITME of four molecular subtypes in EC comprehensively for the first time. They combined the POLE mutant and MSI-H subtypes into the TMB high (TMB-H) subtype owing to their small sample size. They analyzed the ITME features of 30 EC cases, including TMB, infiltration of anti-tumor-related immune cells and negatively regulatory immune cells, and expression of immune checkpoint molecules. Similar results were found in previous studies, which showed that POLE mutations showed the highest level of TMB, followed by the MSI-H subtype, NSMP, and TP53 mutant subtypes. The TMB-H subtype showed a high degree of infiltration of CD8⁺ T cells and relatively high levels of PD-L1 expression in tumor cells. Although TMB levels were low in the TP53 mutant subtype, the proportions of Treg, M2 macrophages, PD-L1⁺ CD68⁺ macrophages, and CD8⁺ PD-1⁺ T cells were relatively high, indicating a strong immunosuppressive microenvironment in this subtype. In the NSMP subtype, the TMB, proportions of multiple tumor-infiltrating immune cells, and expression levels of immune checkpoint molecules were low, indicating a lack of effective anti-tumor immune responses. Based on the immune microenvironmental features, they summarized the immune phenotype of the three molecular subtypes as normal immune response, absence of immune infiltration, and suppressed immune response. Due to the relatively small sample size, further studies with larger sample sizes are required to confirm these findings. The ITME in different subtypes of EC is shown in Table 3 .

Immunotherapeutic strategies for EC are widely used and can be divided into three categories: anticancer vaccines, ICI, and immunomodulators. In recent years, with the development of tumor cells and the tumor immune microenvironment, immunotherapy, especially ICI, has been applied for the treatment of a variety of malignancies. Immune checkpoints regulate co-stimulation signaling to maintain immune self-tolerance in the body and prevent immune damage caused by the excessive activation of T cells. This is the mechanism by which tumor cells escape immune surveillance and death. By suppressing immune checkpoint activity, ICI activate the tumors recognition and killing functions of immune cells. At present PD-1/PD-L1 and CTLA-4 antibodies were widely used (110). These immunotherapies are most effective for POLEmut and MMRd tumors because of their high TMB and increased immunogenic antigens (111). The higher TMB and PD-L1 expression in MSI-H and MMRd endometrial tumors support the clinical efficacy of PD-1 inhibitors in treating solid tumors with MMRd (112). Moreover, as mentioned earlier, PD-L1 expression increased with TIL abundance in the EC ITME, indicating that PD-1 inhibitors can induce an effective anti-tumor immune response in EC with high PD-L1 expression (104–106). In 2017, the FDA approved pabolizumab for the treatment of solid tumors with unresectable or metastatic MSI-H or MMRd. In 2021, the FDA approved dostarlimab-gxly for adult patients with MMRd solid tumors (113). Furthermore the National Comprehensive Cancer Network (UCCN) and ESGO/ESTRO/ESP guidelines in 2020 recommended anti-PD-1 targeted therapy for patients with advanced MMRd EC (21). Multiple clinical trials are currently underway regarding the treatment of EC with PD-1 antibody. KEYNOTE-016 study, a phase II clinical study in the United States evaluated the efficacy and safety of pabolizumab in 41 cases of metastatic colorectal cancer. The objective response rate (ORR) of tumors with MMRd and proficient MMR (pMMR) were 40% and 0%, respectively (19). KEYNOTE-158 multicenter phase II clinical study further evaluated the efficacy and safety of pabolizumab in patients with MSI-H/MMRd endometrial cancer (114). GARNET Study, a phase I/IIb clinical study conducted in 117 centers in 9 countries, had the largest sample size of monotherapy with PD-1 antibody for advanced or recurrent EC. A total of 271 patients were enrolled in this subgroup, and dostarlimab (TSR-042) was administered to patients with MMRd and pMMR. The ORR for MMRd and pMMR patients were 44.7% and 13.4%, respectively (115). Based on the GARNET study, dostarlimab was approved by the FDA for monotherapy in adult patients with recurrent or advanced EC with MMRd after chemotherapy with previous platinum-containing drugs. ICI therapy can relieve the symptoms of some patients and improve their prognosis; however, many patients show primary or acquired resistance. Therefore, immunocombination therapy is necessary to improve the efficacy and mainly includes combined chemotherapy, immunotherapy, and targeted therapy. Several clinical studies have investigated the efficacy and safety of the combination of PD-1/PD-L1 and CTLA-4 antibodies in patients with advanced/recurrent EC (NCT03015129, NCT03508570, and NCT02982486). With the expansion of clinical research on PD-1/PD-L1 antibodies, the management of their adverse reactions is particularly important. The anti-tumor mechanisms of PD-1/PD-L1 antibody and classical chemotherapy are different, so the adverse reaction profile is also quite different. The molecular mechanism is not yet clear and is mostly considered to affect immune homeostasis.

In 2020, Liu et al. (116) selected a series of immune-related genes between EC and normal endometria, and used Cox regression model analysis to select the genes related to prognosis, and obtained 15 immune-related genes to calculate the risk score. Patients were then divided into high and low-risk groups according to their risk scores. Li and Wen (73) also used immune-related genes in TCGA database to define four immune types: Immunosuppressive (type C1), Interferon γ-dominant (type C2), Inflammatory (type C3), and Immune balance (type C4) respectively. Bagaey et al. (117) performed tumor microenvironment typing based on the expression of immune-related genes in the ITME and identified four immune types: immune-enhanced fibrosis (type IE/F), immune-enhanced non-fibrosis (type IE), fibrosis (type F), and immune depletion (type D). Type E is the most effective immunotherapy and has the best prognosis. Thorsson et al. (118) used immune-related genes to obtain six immune-related types: tissue-based type (type 1), interferon γ-dominant (type 2), inflammatory (type 3), lymphocyte depletion (type 4), immune silent type (type 5), and transforming growth factor (type 6). Types 4 and 6 are mainly giant cells in the tumor microenvironment with less lymphocyte infiltration and the worst prognosis, whereas types 2 and 3 have the best prognosis.

Some investigators have performed immunotyping based on the infiltration of immune cells into the tumor tissue. Cai et al. (119) clustered samples according to immune cell infiltration in tumor tissues and obtained three subsets, CI, CII, and CIII, whose immune cell infiltration was high, medium, and low, respectively. Most specific gene mutations were detected in CI and CII, whereas a higher frequency of TP53 gene mutation and copy number variations were found in CIII. Wang et al. (120) combined the degree of immune cell infiltration with gene expression levels to calculate the tumor microenvironment score and divided EC into high- and low-risk groups. Immune activation and immune checkpoint related genes were frequently expressed in low-risk groups. The high-risk group had a higher frequency of PTEN, CSE 1 L, ITGB3 mutations. These studies indicate that immune features in the EC tumor microenvironment are related to prognosis and can be used as a new basis for EC classification.