Skip to main content
Download PDF

Identification of immune-related cervical cancer prognostic biomarkers and construction of prognostic model based on tumor microenvironment

European Journal of Medical Research volume 30, Article number: 261 (2025) Cite this article

Abstract

Objective

Tumor microenvironment (TME) and the expression of immune-related genes (IRGs) are closely related to the development of cervical cancer (CC). This study aims to explore some IRGs as prognostic biomarkers for CC patients based on TME.

Methods

The abundance of tumor-infiltrating immune cells in CC samples was assessed using single-sample gene set enrichment analysis (ssGSEA). Thus, two immune-related groups are generated according to the immune status. The differentially expressed genes were discovered based on the grouping. Then, univariate Cox and LASSO regression analyses were performed using the R package glmnet. Five IRG prognostic signatures (HLA-DMA, DMBT1, CXCR6, CX3CL1, and SEMA3A) were established after that. The protein expression of some genes was verified by immunohistochemistry (IHC).

Results

The signature of the five IRGs was identified to be an independent prognostic indicator for the overall survival in CC patients. A prognostic risk model was also constructed. CC patients were classified into high- and low-risk groups based on the median risk score. The survival time of patients in the high-risk group was shorter than that of those in the low-risk group. The five genes remarkably related to prognosis were screened, among which HLA-DMA, CXCR6, and CX3CL1 were the protective factors, whereas DMBT1 and SEMA3A were the risk factors. GO and KEGG enrichment analyses showed that the biomarkers of the five IRGs were enriched in the receptor-ligand interaction and chemokine signaling pathway. Moreover, CXCR6 expression was significantly correlated with immune cell infiltration among the five hub genes. IHC results demonstrated that the expression of SEMA3A protein level was increased, and CX3CL1 protein level was decreased in cervical cancer tissue.

Conclusion

Immune-related prognostic biomarkers in CC include HLA-DMA, DMBT1, CXCR6, CX3CL1, and SEMA3A. The risk score for the five genes is more accurate than that for other clinical risk factors in predicting prognosis at 3 and 5 years. The higher the risk score is, the worse the prognosis of CC patients is. Five prognostic biomarkers may participate in regulating TME through chemokine-mediated signaling pathways and receptor-ligand interactions. These findings provide new insights into the immunotherapy of CC.

Introduction

With the second highest incidence rate among global female malignant tumors, cervical cancer (CC) is one of the most common malignant tumors in gynecology. It has become a serious threat to women’s lives as the second greatest cause of cancer death in women between the ages of 20 and 39 [36]. Although improvements in CC prevention, including screening and therapy, have been found, the prognosis has not significantly improved, particularly in patients with metastatic or recurring CC [7]. Therefore, further research on the biomarkers for the prognosis assessment in patients with CC is vital to provide potential targets for CC immunotherapy.

Human papillomavirus (HPV) is a considerable pathogenic element in the complicated pathophysiology of CC, which is also intimately associated with chronic cervicitis and gene modification [40]. The occurrence and development of malignancies cannot be accelerated by a single HPV infection, and the clinical symptoms of the infection depend on the immunological health of the host and the postinfection microenvironment. The interplay among virus-infected cells, immune cells, host matrix, and their derivatives within the postinfection microenvironment is considered a key factor in the persistence, proliferation, and malignancy of viruses [43]. Similarly, the tumor microenvironment (TME) comprises various cell types and extracellular elements, including immune cells, fibroblasts, endothelial cells, and matrix cells, all of which are essential for tumor development [6]. Therefore, TME considerably affects immunotherapy and the prognosis of patients with CC.

According to studies, neutrophils are one type of innate immune cell that actively contributes to the development and growth of tumors through various processes, and patients with high neutrophil counts have a poor prognosis [25]. An increase in tumor-infiltrating CD204 + M2 macrophages suggests a poor prognosis for cervical adenocarcinoma patients [11]. CC patients with high T cell levels in their adaptive immune system have a good prognosis [15]. A high count of cytotoxic CD8 + T lymphocytes is linked to a low recurrence rate in CC radiation patients, whereas a low number of cytotoxic CD8 + T lymphocytes and a large number of regulatory T cells are independent indicators of poor survival in CC [37]. Additionally, increased helper CD4 + T cells around CC tumors reduce the risk of disease recurrence [24]. The targets for CC immunotherapy, programmed death 1 (PD-1) and programmed cell death ligand 1 (PD-L1), can enhance the clinical results of patients with CC [17]. The increased PD-L1 expression in tumor cells creates an immunosuppressive microenvironment that prevents T cell activity and promotes the growth of tumor cells. PD-1/PD-L1 inhibitors eliminate tumor cells by blocking the PD-1/PD-L1 pathway, relieving immunosuppression, and reactivating the immune system [28]. The dysregulation of the gut microbiota may activate inflammatory signaling pathways (such as the NF-κB and IL-17R pathways), leading to the upregulation of pro-inflammatory cytokines, thereby affecting the TME in cervical cancer [29, 42]. These cytokines may further modulate the expression of IRGs, which in turn can influence the function of immune cells and the tumor's ability to evade the immune system. Given the importance of immunotherapy in CC, TME and immune cells are crucial for discovering new treatments and prognostic biomarkers for CC. Therefore, we conducted TME-based cancer typing, screened immune-related genes (IRGs), and provided accurate predictions for disease prognosis and potential targets for CC immunotherapy.

Patients with locally advanced cervical cancer (LACC) typically have a poor prognosis, and traditional treatment modalities such as surgery, radiotherapy, and chemotherapy show limited efficacy. In recent years, immunotherapy has gradually emerged as a new direction for the treatment of cervical cancer. However, its clinical application still faces many challenges. Currently, although immune checkpoint inhibitors (ICIs) and other immunotherapeutic approaches have demonstrated promising antitumor activity in some patients, there is a lack of reliable biomarkers to predict patients' responses to immunotherapy or chemoradiotherapy. For example, the expression level of PD-L1 in cervical cancer has a relatively high positivity rate (34.4%–96%), but its sensitivity and specificity as a predictive biomarker are still not satisfactory. Moreover, indicators such as the level of immune cell infiltration in the tumor microenvironment and T-cell receptor (TCR) diversity show some potential, but they have not yet become reliable prognostic biomarkers. Against this backdrop, the development of new prognostic models based on IRGs has become particularly important [14]. By deeply exploring the relationship between IRGs and the tumor microenvironment as well as the response to immunotherapy, it is expected to provide more accurate prognostic assessment and treatment guidance for patients with LACC, thereby improving their clinical outcomes.

In this study, we classified a novel immune-related group for CC based on immune cell infiltration and IRG expression. We analyzed the transcriptome expression data in The Cancer Genome Atlas (TCGA) dataset and extracted the potential genes associated with tumor immunity. The screening of immunogens related to prognosis was used to construct and validate a prognostic model. As a result, we obtained five immune-related prognostic genes, which may help describe the prognosis of patients with CC.

Materials and methods

Data collection

From the website of TCGA (https://cancergenome.nih.gov), we obtained the data, including transcriptome data, mutation data, and related clinical information, for 306 CC patients. Additionally, we obtained 2483 and 3713 IRGs from ImmPort portal (https://www.immport.org/home) and InnateDB (https://www.innatedb.ca), respectively. The prognostic IRG risk model was validated using the GSE44001 (n = 300) dataset, which was retrieved from the GEO database (http://www.ncbi.nlm.nih.gov.ge-o). We used Strawberry Perl software (version 5.30.1, 64 bit) to integrate the transcriptome data and the corresponding clinical information of CC patients obtained from TCGA and GSE44001 datasets. The transcriptome data included gene names, sample numbers, and expression levels. The clinical information included the patient’s age, TNM staging, pathological staging, histological grading, survival time, and survival status. We obtained the IRGs from InnateDB and ImmPort in TCGA transcriptome expression data and removed duplicate genes, obtaining 2499 IRGs for subsequent analysis.

Clusters of CC samples by ssGSEA

ssGSEA was used to calculate the immune scores of 29 immune-related pathways in different CC samples. We used hierarchical clustering (hclust) to group patients. The study included two groups: one with a high immunity (Immunity_H) and the other with a low immunity (Immunity_L). The distribution of samples was shown in a two-dimensional space using the t-distributed Stochastic Neighbor Embedding (tSNE) R program.

Evaluating the consequence of immune clusters

The “estimate” R package is used to calculate the score of the immune microenvironment, such as stromal score, immune scores, and ESTIMATE scores. The expression levels of 19 types of human leukocyte antigen (HLA) genes were used to testify the differences between Immunity_H and Immunity_L groups. Immune cell infiltration was accurately calculated by the CIBERSORT algorithm, and the difference between these two groups was further testified.

Extraction of the differential IRGs for the intersection of CC samples

The differentially expressed genes (DEGs) from the Immunity_H and Immunity_L groups were analyzed by the R package “limma” with the threshold of |log2FC|≥ 1 and FDR of ≤ 0.05. An online website (http://bioinformatics.psb.ugent.be/webtools/Venn/) was used to draw a Venn diagram of DEGs in the two datasets, and the intersection of genes was used for subsequent analysis.

Prognostic model establishment and validation

Univariate Cox regression analysis was applied to screen prognosis-related genes from 25 overlapping genes with the criteria of P ≤ 0.05. Their hazard ratios (HRs) were calculated. LASSO regression analysis was employed to identify genes mostly associated with overall survival (OS). The prognostic prediction model, which included five IRGs, was constructed. The expression level of each gene in the prognostic model was multiplied by its associated coefficient to produce the risk score, as shown below:

$$\begin{aligned} {\text{Risk Score}} = & - 0.00657281741545396 \times {\text{HLA}} - {\text{DMA}} + 0.105876789661672 \times {\text{DMBT}}1\, \\ & \pm \,0.212568449630155 \times {\text{CXCR}}6 \pm 0.145219113732868 \times {\text{CX}}3{\text{CL}}1 \\ & + 0.179849834672435 \times {\text{SEMA}}3{\text{A}}. \\ \end{aligned}$$

With the median value of the risk score as the boundary, the CC samples were divided into high and low risk score groups. TCGA database was used as the training group, whereas the GSE44001 dataset was used as the test group for model validation. Moreover, we drew a Kaplan–Meier (K–M) curves analysis diagram. The area under the receiver operating characteristic (ROC) curve (AUC) was used to predict the model’s accuracy at 1, 3, and 5 years. The calibration curves were used to evaluate the predictive performance of the model. Univariate and multivariate Cox analyses were employed to assess the prognostic values of the risk score and other clinical–pathological parameters in the entire set.

Differential analysis of TME between high and low risk score groups

The differences in TME, including ESTIMATE scores, immune scores, and matrix scores, between the high and low risk score groups in TCGA database were compared.

Functional prediction of prognostic IRGs

The functions of five prognostic IRGs were predicted through Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, with P < 0.05 being considered statistically significant.

Tumor immune estimation resource version 2 (TIMER2) database analysis

The correlations between five hub genes expression and tumor purity, as well as immune cell infiltration, including B cells, CD8 + T-cells, CD4 + T-cells, macrophages, neutrophils, and dendritic cells, were analyzed using the TIMER2 database (https://cistrome.shinyapps.io/timer/). The correlation between gene expression and immune infiltration was estimated using the Pearson correlation test.

Immunohistochemistry (IHC) staining analysis

The cervical tissues were collected from 15 patients with high-grade cervical squamous intraepithelial lesions (HSIL), 15 patients with cervical cancer and 10 patients with cervical cancer of lymph node metastasis (CC-LNM) who had surgery from January 2018 to December 2019. All patients were received surgical treatment in the Department of Gynecology and Obstetrics at the First Affiliated Hospital of Fujian Medical University in Fuzhou, People’s Republic of China. None of the patients had received therapeutic medications or previous surgical interventions before sample collection. IHC staining was used to measure the protein expression of SEMA3A and CX3CL1 in cervical tissues. Slides were incubated with anti-SEMA3A (bs-121R, Bioss, Beijing, China, diluted 1:200), and anti-CX3CL1 (10,108–2-AP Proteintech, Wuhan, China, diluted 1:400). The IHC scores of SEMA3A and CX3CL1 protein were evaluated by two independent pathologists. This study was approved by the Ethics Committee of the First Affiliated Hospital of Fujian Medical University in accordance with the Declaration of Helsinki (Approval Number: [2021435]. All the patients signed an informed consent form.

Statistical analysis

R program (version 4.2.0) was used to compile all statistical data. The “pheatmap” package was used to draw heatmaps, the “ggpubr” package was used to generate box plots and violin plots, and the “glmnet” package was used for LASSO regression analysis. The “Survival” package was used to calculate the univariate Cox regression analysis and construct ROC curves. The “Survival” package and “SurvMiner” package were used to draw the K–M curves of the OS between the high and low risk score groups. The “ClusterProfiler” package was selected for the functional GO and KEGG enrichment analyses.

Results

Grouping based on immune-related cells and pathways

According to the activity and infiltration level of 29 immune cells and pathways in each CC sample in TCGA data, the immune score was calculated by the ssGSEA method. Then, the CC samples were divided into Immunity_H (n = 111) and Immunity_L (n = 195) groups on the basis of the immunological score (Fig. 1A). In addition, the tSNE analysis proves that the immune score could distinguish between Immunity_H and Immunity_L groups (Fig. 1B). The heatmap shows that the Immunity_H group had high immune, matrix, and ESTIMATE scores (Fig. 1C). Moreover, the Immunity_L group exhibited a high tumor purity.

Fig. 1
figure 1

Immune-related pathway analysis distinguishes two groups. A Based on the ssGSEA method, hclust divides CC patients in TCGA into Immunity_H and Immunity_L groups. B The scatter plot shows that all patients can be divided into two subgroups through tSNE analysis, verifying the results of hclust. C The heatmap shows the expression and classification of patients in 29 immune cells and pathways

Full size image

Correlation analysis between immunity_h and immunity_L groups

The TME of Immunity_H and Immunity_L groups was analyzed to verify the effectiveness of the grouping. According to the violin plot, the immune, matrix, and ESTIMATE scores in the Immunity_H group were significantly higher than those in the Immunity_L group, and the difference between the two groups was statistically significant (P < 0.001) (Fig. 2A). These results demonstrated that the Immunity_H group had high counts of immune and stromal cells and a high degree of lymphocyte infiltration, whereas the Immunity_L group contained high counts of tumor cells.

Fig. 2
figure 2

Correlation Analysis between Immunity_H and Immunity_L Groups. A The violin plot shows the TME scores of patients in different subgroups. B The distribution of immune cells in the Immunity_H and Immunity_L groups. C The box plot shows the expression spectrum of each HLA gene in different subgroups

Full size image

CIBERSORT was used to calculate the expression of CC immune cell subtypes in TCGA data. The results indicate that some immune cells, such as CD8 + T cells, activated memory CD4 + T cells, T follicular helper cells, M1 macrophages, and resting dendritic cells, were significantly higher in the Immunity_H group than in the Immunity_L group. By contrast, naive B cells, activated dendritic cells, and activated mast cells were higher in the Immunity_L group than in the Immunity_H group. The difference in the expression of these eight immune cell subtypes between the Immunity_H and Immunity_L groups was statistically significant (P < 0.001). These results indicated that the level of immune cell infiltration in the Immunity_H group was significantly higher than that in the Immunity_L group (Fig. 2B).

Further analysis of the differences between 19 types of HLA genes in the Immunity_H and Immunity_L groups showed that the expression levels of different types of HLA genes in the Immunity_H group were higher than those in the Immunity_L group, and the difference between the two groups was statistically significant (P < 0.001) (Fig. 2C).

Acquisition of DEGs

A total of 270 DEGs were obtained, of which 34 were upregulated, and 236 were downregulated, as shown in the form of a volcano plot (Fig. 3A). The differentially expressed IRGs obtained from TCGA database overlapped with the genes in the GSE44001 dataset. Thus, 25 intersecting genes were obtained, as shown in the form of a Venn diagram (Fig. 3B).

Fig. 3
figure 3

Acquisition of differential IRGs. A The volcano plot shows 270 differential IRGs between Immunity_H and the Immunity_L groups. Red dots indicate upregulation, and green dots indicate downregulation. B The Venn diagram shows the overlapping genes of the differential IRGs in the TCGA database and the GSE44001 dataset

Full size image

Construction and validation of the IRG prognostic model

Univariate Cox regression analysis identified 11 genes significantly associated with prognosis. The HR of DMBT1 and SEMA3A is greater than 1, indicating that they belong to prognostic risk factors, whereas the remaining nine differential genes are less than 1, indicating that they belong to prognostic protective factors (Fig. 4A). In LASSO regression analysis, five genes (HLA-DMA, DMBT 1, CXCR 6, CX3CL1, and SEMA3A) were further screened to construct the prognostic model, with λ = 5 as the lowest value in the optimal model (Fig. 4B). LASSO coefficients are shown in Fig. 4C.

Fig. 4
figure 4

Construction and Validation of the IRG Prognostic Model. A Univariate Cox analysis shows that 11 differential IRGs are significantly associated with the prognosis of patients with CC. B The confidence interval for each lambda. C The trajectory of each independent variable. The horizontal axis represents the logarithmic value of the independent variable λ, and the vertical axis represents the coefficient of the independent variable

Full size image

The median risk score of all samples was 150, and this median value was used as the threshold to divide the samples into high- and low-risk groups. In the training set, the distribution of risk score, survival status, and expression of five IRGs was visible (Figs. 5A and 5C). The two risk groups showed a significant difference in the K–M survival analyses, in which the prognosis of patients in the high-risk group was significantly worse than that in the low-risk group (Fig. 5E). In the training set, the 1-, 3-, and 5 year AUC values were 0.776 (95% CI 0.73–0.83), 0.708 (95% CI 0.67–0.79), and 0.698 (95% CI 0.61–0.76), respectively (Fig. 5G). The calibration curve shows that the model prediction results were similar to the actual survival status (Fig. 5I). The testing set was used to validate the prognostic value of this risk model by further proving the prognostic impact of the risk model. The distribution of risk score, survival status, and expression of the five IRGs in the testing set is shown in Figs. 5B and 5D. Similar to the training set, the testing set showed that the prognosis of patients in the high-risk group was significantly worse than that in the low-risk group (Fig. 5F). The AUC values (0.666, 95% CI 0.59–0.73 and 0.684, 95% CI 0.59–0.75) of the 3- and 5 year survival rate ROC curves validated the model’s accuracy (Fig. 5H). The calibration curve also shows that the model prediction results were consistent with the actual prognosis results (Fig. 5J).

Fig. 5
figure 5figure 5

Construction and validation of the CC IRGs prognosis model. A Risk value curve for the test group and (B) training group. C Scatter plot of the survival status of patients in the test group and (D) training group. In the (E) test group and (F) training group, the prognosis of patients in the high-risk group is worse than that of patients in the low-risk group. The prognostic values of the risk scores in the (G) test group and (H) training group for 1-, 3-, and 5 year survival rates are validated. Calibration curves for the (I) test group and (J) training group

Full size image

IRG biomarkers are independent prognostic factor

The univariate and multivariate Cox analyses were performed on the clinical risk factors in TCGA (Figs. 6A and 6B). They demonstrated that the risk score constructed based on the five genes was an independent factor in the prognosis of CC patients. Moreover, the risk score (AUC = 0.785, 95% CI 0.67–0.89) could predict patient mortality more accurately than the other risk factors for CC, such as age (AUC = 0.507, 95% CI 0.43–0.67), stage (AUC = 0.616, 95% CI 0.54–0.81), grade (AUC = 0.56, 95% CI 0.48–0.69), T stage (AUC = 0.503, 95% CI 0.45–0.64)), and N stage (AUC = 0.636, 95% CI 0.56–0.75) (Fig. 6C). This finding indicates that these five IRG biomarkers are relatively reliable for predicting the prognosis of patients with CC.

Fig. 6
figure 6

Differential Analysis of the TME between High- and Low-Risk Groups. A Univariate analysis of the risk score and clinical factors. B Multivariate analysis of risk score and clinical factors. C The ROC curve evaluation of the prognostic value of risk score and clinical factors. D Differential analysis of the TME between the high- and the low-risk groups

Full size image

Differential analysis of the TME between high- and low-risk groups

Differential analysis of TME was performed on the high- and low-risk groups in TCGA database. The box plots show that the immune, stromal, and ESTIMATE scores of the high-risk group were significantly lower than those of the low-risk group, and the difference was statistically significant (P ≤ 2 × 10−16) (Fig. 6D). These results demonstrated that the low-risk group had more immune and stromal cells and a higher degree of lymphocyte infiltration than the high-risk group, whereas the high-risk group contained more tumor cells.

GO and KEGG functional predictions

Prognostic IRGs were subjected to GO and KEGG enrichment analyses. The GO enrichment analysis showed that BP might be related to the regulation of neuron migration and chemokine-mediated signaling pathways. Cellular component is abundant in the proenzyme granule membrane. Molecular function may be related to receptor ligand and signal receptor activities. The results of KEGG enrichment analysis revealed a high enrichment of immune-related biological processes, such as cytokine-cytokine receptor interaction and chemokine signaling pathway. The results of GO and KEGG enrichment analyses are displayed in bubble and bar charts, respectively (Figs. 7A and 7B).

Fig. 7
figure 7

GO and KEGG Functional Predictions. A GO enrichment analysis predicts the functions involved with the five IRG biomarkers. B KEGG enrichment analysis predicts the pathways involved with the five IRG biomarkers

Full size image

Immune cell infiltration correlation analysis

TIMER2 was used to analyze the correlation between the expression of five prognostic IRGs and tumor purity and immune cell infiltration (Fig. 8A–8E). The results showed that the expression levels of HLA-DMA, CXCR6, and CX3CL1 were negatively correlated with tumor purity (P < 0.05) and positively correlated with the infiltration levels of B cells, CD8 + T cells, CD4 + T cells, neutrophils, and dendritic cells (P < 0.05). The expression level of DMBT1 was negatively correlated with CD4 + T and dendritic cells (P < 0.05). The expression level of SEMA3A was negatively correlated with CD4 + T cells (P < 0.05). In addition, the absolute value of the correlation coefficient between the expression of CXCR6 and tumor purity and immune cells was larger than that between the expression of the four others prognostic IRGs and tumor purity and immune cells. This finding indicates that CXCR6 has a strong correlation with immune cell infiltration.

Fig. 8
figure 8

Correlation of the five genes with tumor purity and immune infiltrating cells. A HLA-DMA. B DMBT1. C CX3CL1. D CXCR6. E SEMA3A

Full size image

Validation of target protein expression

The expression of SEMA3A and CX3CL1 proteins in cervical tissues was verified by immunohistochemistry. As shown in Fig. 9, the expression of SEMA3A protein was higher in cervical cancer tissues (n = 15) than in high-grade squamous intraepithelial lesions (n = 15), and significantly increased in cervical cancer tissues with lymph node metastasis (n = 10). As shown in Fig. 10, the expression of CX3CL1 protein was lower in cervical cancer tissues (n = 15) than in high-grade squamous intraepithelial lesions (n = 15), and significantly decreased in cervical cancer tissues with lymph node metastasis (n = 10). These results confirm that SEMA3A is a prognostic risk factor and CX3CL1 is a prognostic protective factor for cervical cancer, and indicate that SEMA3A and CX3CL1 have good prognostic ability in cervical cancer.

Fig. 9
figure 9

SEMA3A protein was high expression in cervical cancer (200 ×). The representative IHC images of SEMA3A protein expression in high-grade cervical squamous intraepithelial lesions (A), cervical cancer (B), cervical cancer with lymph node metastasis (C), and the expression of SEMA3A protein in different cervical tissues (D)

Full size image
Fig. 10
figure 10

CX3CL1 protein was low expression in cervical cancer (200 ×). The representative IHC images of CX3CL1 protein expression in high-grade cervical squamous intraepithelial lesions A, cervical cancer B, cervical cancer with lymph node metastasis (C), and the expression of SEMA3A protein in different cervical tissues (D)

Full size image

Conclusions

HLA-DMA, DMBT1, CXCR6, CX3CL1, and SEMA3A are immune-related prognostic biomarkers for CC. HLA-DMA, CXCR6, and CX3CL1 are prognostic protective factors, whereas DMBT1 and SEMA3A are prognostic risk factors. The risk scores of HLA-DMA, DMBT1, CXCR6, CX3CL1, and SEMA3A have higher accuracy than other clinical risk factors in predicting the 3- and 5-year prognoses of CC patients. The higher the risk score is, the worse the prognosis of CC patients is. HLA-DMA, DMBT1, CXCR6, CX3CL1, and SEMA3A may regulate TME through chemokine-mediated signaling pathways and receptor–ligand interactions. These findings provide new insights into the immunotherapy of CC.

Discussion

Given that the American Cancer Association estimated 13,960 new cases and 4310 deaths by 2023, CC remains a major public health problem [36]. The most important risk factors affecting the prognosis of patients with CC are stage, lymph node condition, tumor size, depth of tumor invasion into the cervical stroma, and lymph vascular space invasion [27]. The prognosis of patients with metastatic and recurrent CC is extremely poor, with an average survival of 12 months [5]. Treatment methods and prognostic assessment strategies for late and recurrent CC remain lacking. Therefore, we urgently need new immunotherapies and markers for the prognosis assessment in patients with late and recurrent CC.

HLA-DMA is an important component of the antigen processing and presentation pathway of major histocompatibility complex II (MHC-II) [31]. MHC-II molecules can present tumor-associated antigens to CD4 + T cells, thereby activating immune responses. Functional abnormalities in HLA-DMA may lead to reduced antigen presentation efficiency, thereby affecting the recognition and elimination of tumor cells by immune cells. The processing of MHC-II peptides occurs in the endocytic pathway and is heavily dependent on the HLA-DM function [30]. The tyrosine lysosome targeting signal in the cytoplasmic tail of HLA-DM can target HLA-DM to the peptide loading compartment in HeLa cells. Researchers found that this signal plays a critical role in guiding HLA-DM to the processing compartment in antigen-presenting cells [2]. HLA-DMA may indirectly regulate chemokine signaling by influencing the expression and function of MHC-II molecules, thereby affecting the recruitment of immune cells. In cervical cancer, the expression level of HLA-DMA may be associated with tumor immune escape mechanisms. Tumor cells may downregulate the expression of HLA-DMA to reduce the antigen-presenting capacity of MHC-II molecules, thereby evading immune surveillance. Additionally, HLA-DMA may also regulate immune cell functions by affecting receptor-ligand interactions. For example, its interaction with NKG2D ligands may influence the recognition and killing of tumor cells by natural killer (NK) cells. Evidence indicated that HLA-DMA can be a potential biomarker for predicting immune checkpoint therapy and recurrence in ER-negative triple-negative breast cancer [16]. In lung adenocarcinoma, IL-33 induces the maturation and regulation of dendritic cell function by increasing the expression of genes related to dendritic cell function, including antigen presentation genes (HLA-DMA, HLA-DMB, and CD74), as a result, the T cell proliferation–mediated immune regulation of lung adenocarcinoma is induced [44]. In the present study, we found that HLA-DMA is associated with the prognosis of patients with CC. Moreover, the expression of HLA-DMA is negatively correlated with tumor purity and positively correlated with immune cell infiltration, suggesting that HLA-DMA may play a potential role in the immune regulation of cervical malignancies.

DMBT1 is a gene encoding a variable splicing protein involved in mucosal innate immunity and widely distributed in various parts of the human body; it is mainly derived from epithelium and highly expressed in the respiratory system (i.e., trachea and lungs), digestive system (particularly small intestine, salivary glands, and stomach), brain, and reproductive system [10, 20]. Studies showed that DMBT1 is related to mucosal defense and epithelial differentiation, and the abnormal expression of DMBT1 is a common cause of epithelial cell carcinoma. Multiple studies revealed the complex relationship between DMBT1 and various cancers. Upregulation of DMBT1 is found in certain glioblastomas, gastric adenocarcinomas, and lung cancer tissues [21,22,23, 33]. These findings suggest that DMBT1 may be a local regulatory factor in body balance, possibly linked through the regulation of mucosal inflammation and epithelial regeneration [10]. In terms of immune cell recruitment, DMBT1 is involved in the chemotaxis and activation of immune cells by binding to various cell surface receptors. In cervical cancer, the role of DMBT1 may be more complex. On one hand, it may influence the infiltration of immune cells by modulating the chemokine network. On the other hand, its abnormal expression may be associated with immune suppression in the TME. DMBT1 may promote immune evasion of the tumor by affecting the balance of cytokines in the TME. Additionally, the expression level of DMBT1 in the TME may affect the balance of receptor-ligand interactions, thereby regulating the functions of immune cells. In the present study, DMBT1 is considered one of the immune-related prognostic markers for CC and negatively correlated with CD4 + T cells and dendritic cells, suggesting that DMBT1 may have a potential carcinogenic role as one of the risk factors for CC.

CX3CL1 is a unique CX3C chemokine that acts as an adhesion molecule in membrane form and a chemokine in soluble form. CX3CR1 is a highly specific receptor for CX3CL1. The receptor CX3CR1 can inhibit tumor growth and cause specific antitumor immunity by transferring the CX3CL1 gene into tumor cells, thereby leading to increased survival of the tumor-bearing host [41]. Regarding immune cell recruitment, CX3CL1 can attract immune cells such as monocytes, dendritic cells, and natural killer cells, thereby regulating the immune response within the TME. However, the role of CX3CL1 in tumors is dualistic. On one hand, it can enhance anti-tumor immune responses by promoting the infiltration of immune cells. On the other hand, CX3CL1 may also promote tumor progression by modulating the function of immunosuppressive cells, such as regulatory T cells (Tregs). Abnormal expression of CX3CL1 has been found in various types of cancer. Studies showed that the imbalance of CX3CL1/CX3CR1 expression is related to the malignant tendency of high-grade meningiomas [13]. In patients with hepatocellular carcinoma, the high expression of CX3CL1/CX3CR1 is associated with a good prognosis and low local and distant recurrence rate [38]. CX3CL1 is an important chemokine for recruiting tumor-infiltrating lymphocytes and a positive prognostic factor in colorectal cancer, breast cancer, and lung cancer. In addition, HPV oncogenes E5, E6, and E7 are vital for evading host innate immunity (including dendritic cells, Langerhans cells, and natural killer cells) and downregulating adaptive immunity, whereas helper CD4 + T cells and specific cytotoxic CD8 + T lymphocytes are crucial for eliminating HPV-infected keratinocytes [3]. CX3CL1 plays a key role as a regulator of cytotoxic T cell–mediated immunity. CX3CL1 is expressed on mature dendritic cells, thereby promoting natural killer cell activation, IFN-γ production, cytotoxic T cell response promotion, and memory T cell development in adaptive immunity [12]. In cervical cancer, the expression of CX3CL1 is associated with tumor aggressiveness. Studies have shown that high expression of CX3CL1 may be related to tumor progression, invasion, and metastasis. CX3CL1 influences the function of immune cells through its receptor CX3CR1, thereby regulating chemokine signaling in the tumor microenvironment. For example, high expression of CX3CL1 may increase the infiltration of Tregs, thereby suppressing the activity of cytotoxic T cells and promoting tumor immune evasion. The expression of CX3CL1 is reduced in high-grade cervical intraepithelial neoplasia. Therefore, the level of CX3CL1 may predict the prognosis of CC patients, and its continuous reduction as a protective factor in the microenvironment promotes the progression of cervical malignancies. In some studies, CX3CR1 is considered a novel marker for effector memory CD8 + T cells and a potential marker for PD-1 treatment-responsive and chemotherapy-resistant CD8 + T cell subsets [41]. This conclusion suggests the importance of this chemokine pathway in successful antitumor immune responses and immune therapy responses.

The CXCR6-CXCL16 chemotactic axis regulates the homing, activation, proliferation, and cytotoxicity of immune cells. CXCR6 is a receptor for the chemokine CXCL16, which exists in membrane or soluble form. Its main function may be to locate T cells and interact with cells expressing CXCL16 on the membrane, thereby promoting T cell differentiation. In tumor immunology, CXCR6 enhances the interaction between cytotoxic T lymphocytes and dendritic cells (DCs), particularly the contact with the CCR7 + DC3 subset, thereby promoting the trans-presentation of IL-15 and maintaining the effector functions of T cells. Some studies showed that the expression of CXCR6 plays an important role in the antitumor function and immunity of CD8 + T cells [4]. In mice with CXCR6 deficiency or blocked interaction between CXCR6 and its ligand, CD8 + T cells have poor control over tumor proliferation, particularly in the liver natural killer T cells [19]. In HPV-positive head and neck squamous cell carcinoma, the expression of CXCR6 is associated with the infiltration of memory CD8 + T cells, activated natural killer cells, dendritic cells, and M1 macrophages [32]. The expression of CXCR6 in CC is significantly associated with lymph node metastasis. On the one hand, the high expression of CXCR6 may recruit more CD8 + T cells into the tumor tissue, forming an inflamed TME, thereby inhibiting tumor progression. On the other hand, the interaction between CXCR6 and its ligand CXCL16 may also affect chemokine signaling in the tumor microenvironment, thereby regulating tumor immune evasion and progression. Adding CXCR6 to specific T cells can therapeutically enhance their intratumoral accumulation in the animal models of pancreatic, ovarian, and lung cancer, and the survival can be prolonged [19]. CXCR6 is part of the immunological characteristics that can predict the response to anti-PD-L1-based immunotherapy in various cancers. Therefore, CXCR6 is not only a potential target for immunotherapy but also a potential biomarker for assessing the tumor immune microenvironment and prognosis.

The SEMA3 gene family plays an important role in biological processes. Its functions in regulating the immune system, angiogenesis, local cancer spread, and metastasis are attracting attention increasingly. Recent studies have shown that SEMA3 is secreted by tumor cells, macrophages, and fibroblasts, thereby directly or indirectly affecting cancer cells and the microenvironment to regulate cell-to-cell pathways in the TME or vascular system [39]. SEMA3A induces invasion in pancreatic and colon cancer and is associated with poor survival prognosis. MiR-362 can regulate the expression of SEMA3A in non-small cell lung cancer cells, thereby promoting the formation and metastasis of non-small cell lung cancer cell colonies [18, 45]. SEMA3A is negatively correlated with CD4 + T cells in the relationship between this gene and the CC TME. B7-H4Ig regulates the ability of CD4 + T cells to respond in vivo by binding to SEMA3A. SEMA3A regulates the migration and function of immune cells by interacting with its receptors NRP1 (neuropilin-1) and members of the Plexin-A family. Studies have shown that in the tumor microenvironment, SEMA3A can induce cytoskeletal paralysis of tumor-specific CD8 + T cells, inhibiting their migration and immune synapse assembly, thereby weakening the antitumor immune response [1]. Moreover, high expression of SEMA3A is associated with a reduced number of tumor-infiltrating T cells, suggesting that it may promote tumor progression by inhibiting T cell infiltration. In the tumor microenvironment of cervical cancer, SEMA3A may also influence tumor angiogenesis and immune evasion through interactions with other cytokines and receptors. For example, the synergistic action of SEMA3A with factors such as VEGFA, SEMA3F, and SEMA3C may activate signaling pathways via the NRP2 receptor, promoting angiogenesis and tumor formation. Additionally, the interaction between SEMA3A and NRP1 plays a key role in regulating the chemotaxis and function of immune cells, which may further affect the distribution of immune cells in the tumor microenvironment [9]. However, the role of SEMA3A in tumors is not entirely negative. In some cases, SEMA3A may exert antitumor effects by regulating the activity of immune cells. For example, SEMA3A has been found to inhibit tumor growth in certain tumor models, which may be related to its regulation of immune cell infiltration and activation.

Finally, our study screened five immune genes related to prognosis. As part of the antigen processing and presentation pathway, HLA-DM reaches the processing compartment in antigen-presenting cells through the tyrosine lysosome targeting signal. DMBT1 is considered a local regulatory factor for internal balance, possibly linked to the regulation of mucosal inflammation and epithelial regeneration. In its soluble form, CX3CL1 participates in the immune regulation mediated by the cytotoxic T cell as a chemotactic factor in various cancers. CXCR6 exists in membrane or soluble form as a receptor for the chemotactic factor CXCL16, and its expression plays an important role in the antitumor function of CD8 + T cells and immunity. SEMA3 is secreted by various cells and participates in regulating the cell-to-cell pathways in the TME and vessels. In the final part of the study, we applied IHC to determine the expression of SEMA3A and CX3CL1 protein in patients with high-grade cervical squamous intraepithelial lesions, cervical cancer and cervical cancer with lymph node metastasis, the results confirmed that SEMA3A is a prognostic risk factor and CX3CL1 is a prognostic protective factor.

This article conducted pathway enrichment analysis on five immune genes related to the prognosis of patients with CC. The GO enrichment analysis shows that their biological functions are related to chemokine-mediated signaling pathways, and molecular functions are related to receptor ligand activity and signal receptor activity. Similarly, the KEGG enrichment analysis results revealed that the immune-related biological processes of the five immune genes are enriched in cytokine–cytokine receptor interaction and chemokine signaling pathways. TME includes various cell types, extracellular components (such as fibroblasts, endothelial cells, immune cells and cytokines), hormones, extracellular matrix, and growth factors. Among them, cytokines are divided into four categories: tumor necrosis factor family, chemokine family, interferon family, and hematopoietic factor family. Cytokines are small molecule peptides or proteins secreted by various tissue cells, such as matrix, immune, and tumor cells. They have nutritional functions at the local tissue level or system level. They control the growth, differentiation, or activation of different types of cells; they also function through cytokine receptors on the cell membrane, soluble plasma, or tissue fluid receptors [8]. Cytokines are initially considered messenger molecules in the immune system, guiding white blood cells to the site of inflammation. However, when regulation is disrupted, they may play a role in the malignant transformation, proliferation, survival, angiogenesis, invasion, and metastasis processes in tumor tissues [34, 35]. Studies showed that the dysregulation of some cytokines is closely related to the incidence of precancerous lesions of CC, the progression from precancerous lesions to carcinoma in situ, further invasion, and terminal metastasis [26]. Therefore, HLA-DMA, DMBT1, CXCR6, CX3CL1, and SEMA3A may participate in the regulation of CC TME through chemokine-mediated signaling pathways and receptor–ligand interactions.

Currently, the prognosis prediction of cervical cancer mainly relies on clinical pathological features (such as FIGO stage, histological type, tumor size, etc.) and some molecular biomarkers. For example, the gene CXCL9 has been proven to be associated with the survival prognosis of cervical cancer patients, with its high expression correlating with better prognosis. The five newly identified IRGs (HLA-DMA, DMBT1, CXCR6, CX3CL1, and SEMA3A) in this study serve as independent prognostic biomarkers. They not only enrich the molecular biomarker library for cervical cancer prognosis but also provide potential targets for immunotherapy. Compared with existing prognostic models, the model based on these IRGs can more comprehensively reflect the immune status within the tumor microenvironment, thereby improving the accuracy and individualization of prognostic prediction. Future research can further validate the clinical application value of these IRGs, promoting the optimization of cervical cancer prognosis assessment and treatment strategies.

Although the confirmation of SEMA3A and CX3CL1 expression in cervical cancer tissues by immunohistochemistry (IHC) is an important first step, extending these findings to in vitro or in vivo models would significantly enhance the impact of the research. For example, the roles of SEMA3A and CX3CL1 in immune cell recruitment, proliferation, or cytokine production have not yet been fully elucidated. Further exploration of the specific mechanisms of these biomarkers in the pathophysiology of cervical cancer can be achieved through in vitro cell experiments or in vivo animal models. Functional experiments, such as co-culturing immune cells with tumor cells or using gene knockout/overexpression models, can provide a more comprehensive understanding of the roles of these biomarkers in cervical cancer. In-depth investigation of the functions of these biomarkers will not only offer new targets for the diagnosis and treatment of cervical cancer but also provide a theoretical basis for the development of prognostic models based on IRGs.

In summary, this study identifies the five immune-related prognostic biomarkers, namely, HLA-DMA, DMBT1, CXCR6, CX3CL1, and SEMA3A, thereby providing new ideas for individualized immunotherapy for patients with CC. In future research, we can focus on the relationship between immune-related prognostic biomarkers and the pathogenesis and treatment of CC in the hope that they can be widely used in CC treatment.

Availability of data and materials

No datasets were generated or analysed during the current study.

References

  1. Barnkob MB, Michaels YS, André V, Macklin PS, Gileadi U, Valvo S, et al. Publisher correction: semaphorin 3A causes immune suppression by inducing cytoskeletal paralysis in tumour-specific CD8(+) T cells. Nat Commun. 2024;15(1):3448.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  2. Copier J, Potter P, Sacks SH, Kelly AP. Multiple signals regulate the intracellular trafficking of HLA-DM in B-lymphoblastoid cells. Immunology. 1998;93(4):505–10.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Cosper PF, Bradley S, Luo L, Kimple RJ. Biology of HPV mediated carcinogenesis and tumor progression. Semin Radiat Oncol. 2021;31(4):265–73.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Di Pilato M, Kfuri-Rubens R, Pruessmann JN, Ozga AJ, Messemaker M, Cadilha BL, et al. CXCR6 positions cytotoxic T cells to receive critical survival signals in the tumor microenvironment. Cell. 2021;184(17):4512-4530.e4522.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Elit LM, Hirte H. Management of advanced or recurrent cervical cancer: chemotherapy and beyond. Expert Rev Anticancer Ther. 2014;14(3):319–32.

    Article  CAS  PubMed  Google Scholar 

  6. Frankel T, Lanfranca MP, Zou W. The role of tumor microenvironment in cancer immunotherapy. Adv Exp Med Biol. 2017;1036:51–64.

    Article  CAS  PubMed  Google Scholar 

  7. Green JA, Kirwan JM, Tierney JF, Symonds P, Fresco L, Collingwood M, et al. Survival and recurrence after concomitant chemotherapy and radiotherapy for cancer of the uterine cervix: a systematic review and meta-analysis. Lancet. 2001;358(9284):781–6.

    Article  CAS  PubMed  Google Scholar 

  8. Gururaj A, Kumar R. Polypeptide growth factors and their receptors. Cancer Treat Res. 2005;126:1–14.

    Article  PubMed  Google Scholar 

  9. Hung YH, Hou YC, Hsu SH, Wang LY, Tsai YL, Shan YS, et al. Pancreatic cancer cell-derived semaphorin 3A promotes neuron recruitment to accelerate tumor growth and dissemination. Am J Cancer Res. 2023;13(8):3417–32.

    CAS  PubMed  PubMed Central  Google Scholar 

  10. Kang W, Reid KB. DMBT1, a regulator of mucosal homeostasis through the linking of mucosal defense and regeneration? FEBS Lett. 2003;540(1–3):21–5.

    Article  CAS  PubMed  Google Scholar 

  11. Kawachi A, Yoshida H, Kitano S, Ino Y, Kato T, Hiraoka N. Tumor-associated CD204(+) M2 macrophages are unfavorable prognostic indicators in uterine cervical adenocarcinoma. Cancer Sci. 2018;109(3):863–70.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Kumar S. Natural killer cell cytotoxicity and its regulation by inhibitory receptors. Immunology. 2018;154(3):383–93.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Li G, Hattermann K, Mentlein R, Mehdorn HM, Held-Feindt J. The transmembrane chemokines CXCL16 and CX3CL1 and their receptors are expressed in human meningiomas. Oncol Rep. 2013;29(2):563–70.

    Article  PubMed  Google Scholar 

  14. Li J, Ma Y, Wu Q, Ping P, Li J, Xu X. The potential role of HPV oncoproteins in the PD-L1/PD-1 pathway in cervical cancer: new perspectives on cervical cancer immunotherapy. Front Oncol. 2024;13(14):1488730.

    Article  Google Scholar 

  15. Lippens L, Van Bockstal M, De Jaeghere EA, Tummers P, Makar A, De Geyter S, et al. Immunologic impact of chemoradiation in cervical cancer and how immune cell infiltration could lead toward personalized treatment. Int J Cancer. 2020;147(2):554–64.

    Article  CAS  PubMed  Google Scholar 

  16. Liu D, Hofman P. Expression of NOTCH1, NOTCH4, HLA-DMA and HLA-DRA is synergistically associated with T cell exclusion, immune checkpoint blockade efficacy and recurrence risk in ER-negative breast cancer. Cell Oncol (Dordr). 2022;45(3):463–77.

    Article  CAS  PubMed  Google Scholar 

  17. Liu Y, Wu L, Tong R, Yang F, Yin L, Li M, et al. PD-1/PD-L1 inhibitors in cervical cancer. Front Pharmacol. 2019;10:65.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Luo D, Zhang Z, Zhang Z, Li JY, Cui J, Shi WP, et al. Aberrant expression of miR-362 promotes lung cancer metastasis through downregulation of Sema3A. J Immunol Res. 2018;10:12.

    Google Scholar 

  19. Mabrouk N, Tran T, Sam I, Pourmir I, Gruel N, Granier C, et al. CXCR6 expressing T cells: functions and role in the control of tumors. Front Immunol. 2022;13:1022136.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Madsen J, Mollenhauer J, Holmskov U. Review: Gp-340/DMBT1 in mucosal innate immunity. Innate Immun. 2010;16(3):160–7.

    Article  CAS  PubMed  Google Scholar 

  21. Mollenhauer J, Helmke B, Müller H, Kollender G, Lyer S, Diedrichs L, et al. Sequential changes of the DMBT1 expression and location in normal lung tissue and lung carcinomas. Genes Chromosomes Cancer. 2002;35(2):164–9.

    Article  CAS  PubMed  Google Scholar 

  22. Mollenhauer J, Herbertz S, Holmskov U, Tolnay M, Krebs I, Merlo A, et al. DMBT1 encodes a protein involved in the immune defense and in epithelial differentiation and is highly unstable in cancer. Cancer Res. 2000;60(6):1704–10.

    CAS  PubMed  Google Scholar 

  23. Mueller W, Mollenhauer J, Stockhammer F, Poustka A, von Deimling A. Rare mutations of the DMBT1 gene in human astrocytic gliomas. Oncogene. 2002;21(38):5956–9.

    Article  CAS  PubMed  Google Scholar 

  24. Nedergaard BS, Ladekarl M, Thomsen HF, Nyengaard JR, Nielsen K. Low density of CD3+, CD4+ and CD8+ cells is associated with increased risk of relapse in squamous cell cervical cancer. Br J Cancer. 2007;97(8):1135–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Ocana A, Nieto-Jiménez C, Pandiella A, Templeton AJ. Neutrophils in cancer: prognostic role and therapeutic strategies. Mol Cancer. 2017;16(1):137.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Paradkar PH, Joshi JV, Mertia PN, Agashe SV, Vaidya RA. Role of cytokines in genesis, progression and prognosis of cervical cancer. Asian Pac J Cancer Prev. 2014;15(9):3851–64.

    Article  PubMed  Google Scholar 

  27. Peters WA 3rd, Liu PY, Barrett RJ 2nd, Stock RJ, Monk BJ, Berek JS, et al. Concurrent chemotherapy and pelvic radiation therapy compared with pelvic radiation therapy alone as adjuvant therapy after radical surgery in high-risk early-stage cancer of the cervix. J Clin Oncol. 2000;18(8):1606–13.

    Article  CAS  PubMed  Google Scholar 

  28. Potente M, Carmeliet P. the link between angiogenesis and endothelial metabolism. Annu Rev Physiol. 2017;79:43–66.

    Article  CAS  PubMed  Google Scholar 

  29. Priadko K, Romano L, Olivieri S, Romeo M, Barone B, Sciorio C, et al. Intestinal microbiota, intestinal permeability and the urogenital tract: is there a pathophysiological link? J Physiol Pharmacol. 2022;10:12.

    Google Scholar 

  30. Radley E, Alderton RP, Kelly A, Trowsdale J, Beck S. Genomic organization of HLA-DMA and HLA-DMB comparison of the gene organization of all six class ii families in the human major histocompatibility complex. J Biol Chem. 1994;269(29):18834–8.

    Article  CAS  PubMed  Google Scholar 

  31. Reyes-Vargas E, Barker AP, Zhou Z, He X, Jensen PE. HLA-DM catalytically enhances peptide dissociation by sensing peptide-MHC class II interactions throughout the peptide-binding cleft. J Biol Chem. 2020;295(10):2959–73.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Saiz-Ladera C, Baliu-Piqué M, Cimas FJ, Manzano A, García-Barberán V, Camarero SC, et al. Transcriptomic correlates of immunologic activation in head and neck and cervical cancer. Front Oncol. 2021;11: 714550.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Sasaki H, Betensky RA, Cairncross JG, Louis DN. DMBT1 polymorphisms: relationship to malignant glioma tumorigenesis. Cancer Res. 2002;62(6):1790–6.

    CAS  PubMed  Google Scholar 

  34. Sharma A, Rajappa M, Saxena A, Sharma M. Cytokine profile in Indian women with cervical intraepithelial neoplasia and cancer cervix. Int J Gynecol Cancer. 2007;17(4):879–85.

    Article  CAS  PubMed  Google Scholar 

  35. Sharma M, Satyam A, Abhishek A, Khan R, Rajappa M, Sharma A. Molecular and circulatory expression of insulin growth factors in Indian females with advanced cervical cancer. Asian Pac J Cancer Prev. 2012;13(12):6475–9.

    Article  PubMed  Google Scholar 

  36. Siegel RL, Miller KD, Wagle NS, Jemal A. Cancer statistics, 2023. CA Cancer J Clin. 2023;73(1):17–48.

    Article  PubMed  Google Scholar 

  37. Someya M, Tsuchiya T, Fukushima Y, Hasegawa T, Takada Y, Hori M, et al. Association between cancer immunity and treatment results in uterine cervical cancer patients treated with radiotherapy. Jpn J Clin Oncol. 2020;50(11):1290–7.

    Article  PubMed  Google Scholar 

  38. Tang L, Hu HD, Hu P, Lan YH, Peng ML, Chen M, et al. Gene therapy with CX3CL1/Fractalkine induces antitumor immunity to regress effectively mouse hepatocellular carcinoma. Gene Ther. 2007;14(16):1226–34.

    Article  CAS  PubMed  Google Scholar 

  39. Toledano S, Nir-Zvi I, Engelman R, Kessler O, Neufeld G. Class-3 semaphorins and their receptors: potent multifunctional modulators of tumor progression. Int J Mol Sci. 2019;20: 556.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Xu F, Shen J, Xu S. Integrated bioinformatical analysis identifies GIMAP4 as an immune-related prognostic biomarker associated with remodeling in cervical cancer tumor microenvironment. Front Cell Dev Biol. 2021;9: 637400.

    Article  PubMed  PubMed Central  Google Scholar 

  41. Yan Y, Cao S, Liu X, Harrington SM, Bindeman WE, Adjei AA, et al. CX3CR1 identifies PD-1 therapy-responsive CD8+ T cells that withstand chemotherapy during cancer chemoimmunotherapy. JCI Insight. 2018;3:e97828.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Yang Y, Zhu J, Feng R, Han M, Chen F, Hu Y. Altered vaginal cervical microbiota diversity contributes to HPV-induced cervical cancer via inflammation regulation. PeerJ. 2024;12(12):e17415.

    Article  PubMed  PubMed Central  Google Scholar 

  43. Yuan Y, Cai X, Shen F, Ma F. HPV post-infection microenvironment and cervical cancer. Cancer Lett. 2021;497:243–54.

    Article  CAS  PubMed  Google Scholar 

  44. Zhang J, Chen Y, Chen K, Huang Y, Xu X, Chen Q, et al. IL-33 drives the antitumour effects of dendritic cells via upregulating CYLD expression in pulmonary adenocarcinoma. Artif Cells Nanomed Biotechnol. 2019;47(1):1335–41.

    Article  PubMed  Google Scholar 

  45. Zhang X, Klamer B, Li J, Fernandez S, Li L. A pan-cancer study of class-3 semaphorins as therapeutic targets in cancer. BMC Med Genomics. 2020;13(Suppl 5):45.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We thank TCGA project, GEO, Immport Portal, InnateDB, TIMER2, CIBERSORT, and cBioPortal databases and their contributors for the valuable public datasets used in this study.

Funding

This study was supported by the Natural Science Foundation of Fujian Province (No.2022J01702, No.2022J01218).

Author information

Author notes

  1. Qiufang Bao and Wenlu Tang have contributed equally to this work.

Authors and Affiliations

  1. Department of Obstetrics and Gynecology, the First Affiliated Hospital, Fujian Medical University, Fuzhou, Fujian, People’s Republic of China

    Qiufang Bao & Bin Zhang

  2. Department of Obstetrics and Gynecology, The First Clinical College of Fujian Medical University, Fuzhou, Fujian, People’s Republic of China

    Qiufang Bao, Xiaohong Zhu & Bin Zhang

  3. Department of Obstetrics and Gynecology, National Regional Medical Center, Binhai Campus of the First Affiliated Hospital, Fuzhou, Fujian, People’s Republic of China

    Qiufang Bao & Bin Zhang

  4. Department of Anesthesiology, Longyan First Affiliated Hospital of Fujian Medical University, Longyan, Fujian, People’s Republic of China

    Wenlu Tang

  5. Department of Orthopaedics, Tingzhou Hospital of Fujian Province, Longyan, Fujian, People’s Republic of China

    Wenzhong Tang

  6. Department of Pediatrics, the First Affiliated Hospital, Fujian Medical University, Fuzhou, Fujian, People’s Republic of China

    Linliang Hong

Authors
  1. Qiufang Bao
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Wenlu Tang
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Wenzhong Tang
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Xiaohong Zhu
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Linliang Hong
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Bin Zhang
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

BZ conceived the presented idea. QB developed the theory and performed the computations. WT verified the analytical methods. BZ encouraged QB and WT to investigate and supervise the findings of this work. QB and WT wrote the manuscript with input from all authors. All authors contributed to the design and implementation of the research, analysis of the results, and writing of the manuscript.

Corresponding authors

Correspondence to Linliang Hong or Bin Zhang.

Ethics declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of the First Affiliated Hospital of Fujian Medical University in accordance with the Declaration of Helsinki (approval number: [2021]435). All the patients signed an informed consent form.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

The original online version of this article was revised: The affiliations of Second Author（Wenlu Tang）and third Author（Wenzhong Tang）have been updated.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Bao, Q., Tang, W., Tang, W. et al. Identification of immune-related cervical cancer prognostic biomarkers and construction of prognostic model based on tumor microenvironment. Eur J Med Res 30, 261 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s40001-025-02515-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s40001-025-02515-5

Keywords

European Journal of Medical Research

ISSN: 2047-783X

Contact us