Title
Author
DOI
Article Type
Special Issue
Volume
Issue
1School of Clinical Medicine, Dali University, 671000 Dali, Yunnan, China
2Department of Infectious Diseases, Yantai Qishan Hospital, 264000 Yantai, Shandong, China
3Department of Obstetrics, The Affiliated Yantai Yuhuangding Hospital of Qingdao University, 264000 Yantai, Shandong, China
4Department of Neurology, The Affiliated Yantai Yuhuangding Hospital of Qingdao University, 264000 Yantai, Shandong, China
5Genetic Testing Center, The First Affiliated Hospital of Dali University, 671000 Dali, Yunnan, China
*Corresponding Author(s):zhyy@dali.edu.cn (Yuanyuan Zhang); wgm1991@dali.edu.cn (Guangming Wang)
† These authors contributed equally.
| History | Submitted: 21 July 2023 | Accepted: 22 August 2023 | Published: 15 March 2025 |
| Copyright: | ©2025 The Author(s). Published by MRE Press. |
Background: The protein encoded by the erythroblast transformation-specific-related gene (ERG) plays an important role in the occurrence and development of various human cancers. However, its biological importance in endometrial carcinoma (EC) remains unclear. Here, we aimed to assess ERG expression in EC and determine its prognostic value via bioinformatic analyses. Methods: Data from public databases were used to evaluate ERG expression in patients with pan-cancer or EC and elucidate its influence on tumour immune cell infiltration. ERG mutation types and clinical characteristics of patients were analysed. The Kaplan-Meier method was employed to evaluate the EC overall survival rate and its correlation with ERG expression levels. Moreover, receiver operating characteristic curves were generated to assess the specificity and accuracy of ERG expression in diagnosing EC. Finally, pathway enrichment analyses were performed to elucidate gene function. Results: ERG is differentially expressed in various cancers and is downregulated in EC. Moreover, copy number variation is the most common ERG mutation type in EC. Low ERG expression correlates with the clinical stage of cancer (p = 0.011), histological type (p = 1.6734 × 10−5), histological grade (p = 2.2165 × 10−8), and degree of invasion (p = 0.019). It is also associated with a relatively poor prognosis. Low ERG expression is also associated with immune cell infiltration and several pathways relevant to the development and progression of EC. In particular, ERG contributes to EC occurrence and development by regulating various transcriptional processes and carcinogenic signalling pathways. Furthermore, ERG expression is related to immune invasion and ET diagnosis and prognosis. Hence, ERG expression can be used as a prognostic biomarker for EC. Conclusions: Our findings can serve as a reference for identifying targets of molecular biological therapy and immunotherapy for EC.
Cite this article
Jianglei Ma, Huijie Zhang, Chenfang Zhang, Xueyan Luo, Fuhui Duan, Yuanyuan Zhang, et al.ERG expression as a potential biomarker for endometrial carcinoma prognosis.European Journal of Gynaecological Oncology,2025,46(3):46-55 DOI:10.22514/ejgo.2025.035
1. Introduction
Endometrial carcinoma (EC) is the sixth most common cancer among women and the most common malignant tumour of the female reproductive tract, with 417,000 new cases reported worldwide in 2020 [1]. The lifetime risk of EC in women is approximately 3%, and the median age at diagnosis is 61 years [2]. Although current treatments have improved clinical outcomes for patients with EC, the associated prognosis remains poor as evidenced by the high incidence rate (21.9–86.6/100,000) and low patient survival rate. In other words, patients diagnosed with International Federation of Gynecology and Obstetrics stage III or IV EC have 5-year overall survival rates of 57%–65% and 20%–26%, respectively [3, 4]. Hence, more effective biomarkers and new targets for the treatment of EC are needed.
The protein encoded by the erythroblast transformation-specific-related gene (ERG) has important roles in cell proliferation, angiogenesis, apoptosis and vascular cell remodelling [5]. Meanwhile, genome rearrangements and chromosome translocations cause overexpression and increased ERG activity in several cancers, including sarcomas, acute myeloid leukaemia, and prostate cancer [6]. In fact, genomic rearrangement in ERG occurs in approximately 50% of patients with primary prostate cancer [7]. Upregulated ERG expression is associated with low survival rates among patients with prostate [8] and central nervous system tumours, including glioblastoma and haemangioblastoma [9]. In contrast, downregulated ERG expression is associated with a short survival period in renal cancer [10]. However, its role in EC progression and prognosis remains unclear.
Therefore, in this study, we performed bioinformatic analyses of ERG expression in EC to evaluate its influence on tumour immune cell infiltration, and its prognostic and diagnostic value. Furthermore, we aimed to characterise the most common ERG mutation types associated with EC and identify the signalling pathways related to tumour development.
2. Materials and methods
2.1 Patient dataset and ERG expression analysis
ERG expression levels in different cancer types were determined based on mRNA expression data from The Cancer Genome Atlas (TCGA) database, using TIMER2.0 (http://timer.cistrome.org/) [11]. Gene Expression Profiling Interactive Analysis (GEPIA) (http://gepia.cancer-pku.cn/) was used to analyse ERG expression in EC, RNA-sequencing data for 174 EC tissues and 91 normal adjacent tissues from TCGA were analysed [12]. The Human Protein Atlas (http://www.proteinatlas.org/) was employed to verify immunohistochemical staining results for ERG protein abundance in EC based on average optical density (AOD) values [13]. AOD refers to the concentration of reactants per unit area and is the sum of the optical density values for each point in the image divided by the target distribution area. The data and information used in this study were obtained from public databases.
2.2 ERG mutation type and frequency analysis
Data obtained from the cBioPortal database (http://cbioportal.org) were used to analyse different types and frequencies of ERG alterations in EC, including mutations, deletions and copy number variations [14].
2.3 Analysis of clinical characteristics
After arranging the TCGA data, including those on clinical and histological indicators, patients were divided into high- and low-expression groups according to the median ERG expression level. Correlations between ERG expression and clinical and histological indicators were analysed using chi-square tests.
2.4 Survival and receiver operating characteristic curve analysis
The Kaplan-Meier (KM) plotter (http://kmplot.com/analysis/) was used to predict the correlations between ERG expression and overall survival (OS) in patients with EC [15]. The KM plotter software was used to determine the survival period of patients stratified according to clinical characteristics and to calculate the hazard ratio (HR), log-rank p-value, and associated 95% confidence interval. We used the pROC package in R (v4.1.0; The R Project for Statistical Computing, Vienna, Austria) to plot a receiver operating characteristic (ROC) curve from RNA-sequencing data of patients with EC and healthy individuals. The area under the curve (AUC) was calculated to assess the sensitivity and specificity of ERG expression in diagnosing EC.
2.5 Analysis of immune cell infiltration
We used Gene Set Cancer Analysis (GSCA; http://bioinfo.life.hust.edu.cn/GSCA/#/), an online tool for immune genomics analysis based on the ImmuCellAI algorithm [16], to analyse the correlation between ERG expression and tumour infiltration by 24 types of immune cells in EC [17]. It was also used to determine the correlation coefficient and false discovery rate (FDR)-adjusted p-value (pFDR).
2.6 Gene ontology term, biological process, and Kyoto Encyclopedia of genes and genomes pathway enrichment analyses
The LinkedOmics portal (http://www.linkedomics.org/) was used to identify ERG-related genes in EC [18]. The Pearson method was used for correlation analysis, with pFDR of < 0.01 as the selection criterion. Gene Ontology (GO) term, biological process (BP), and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed using the DAVID v6.8 bioinformatics tool (https://david.ncifcrf.gov) [19]. The top 10 enriched pathways (p < 0.01) were selected and visualised in dot plots.
2.7 Prediction of sensitivity to therapeutic drugs
Half-maximal inhibitory concentration (IC50) of 138 therapeutic drugs [20] (including chemotherapeutic drugs and targeted inhibitors, with efficacy information) in EC was calculated using the pRRophetic package in R.
2.8 Statistical analysis
Data are presented as mean ± standard deviation and were statistically evaluated using SPSS (v26.0; SPSS Inc., Chicago, IL, USA). Statistical analyses were performed using R software v4.1.0 (64-bit). The chi-square test was used to analyse intergroup differences. The KM plotter and a Cox regression model were used for survival analysis. Pearson’s correlation analysis assessed correlations between ERG expression and tumour-infiltrating immune cells. Results with a two-sided p-value of < 0.05 were considered statistically significant.
3. Results
3.1 ERG expression in EC
TIMER2.0 analysis of pan-cancer ERG expression indicated that ERG is differentially expressed in various human tumours, including EC (Fig. 1A). GEPIA showed that ERG expression was lower in EC tissues than in normal adjacent tissues (Fig. 1B). Similarly, immunohistochemical staining data from the Human Protein Atlas showed that ERG protein expression was lower in EC (AOD = 0.150) than in normal adjacent tissues (AOD = 0.282; Fig. 1C). These results suggest that low ERG expression is associated with EC.
Fig. 1.ERG expression in various cancers and endometrial carcinoma (EC). (A) ERG expression in various carcinomas determined using The Cancer Genome Atlas data and TIMER2.0. (B) ERG expression in EC determined using GEPIA. Tumour and normal tissue are indicated by red and grey, respectively. (C) Immunohistochemical staining results, retrieved from the Human Protein Atlas, for erythroblast transformation-specific-related gene (ERG) protein expression in EC and normal adjacent tissues. The rectangles are zoomed in portions of normal and endometrial cancer tissue. *p < 0.05, **p < 0.01, ***p < 0.001. AOD: average optical density.
3.2 Identification of ERG mutations in EC
We used the cBioPortal database to determine the frequency of ERG–TMPRSS2 fusions and other ERG alterations in EC. The analysis showed that 3% of all EC samples (n = 1954) had genetic alterations, and copy number variations (amplifications) as well as splicing, truncating, and missense mutations were the most common mutation types (Fig. 2); gene fusions were not detected in the EC samples.
Fig. 2.Frequency of ERG mutations in patients with endometrial carcinoma determined using the cBioPortal database. The different colours of the modules represent the types of mutations and the research related to ERG in endometrial cancer. ERG: erythroblast transformation-specific-related gene; TCGA: The Cancer Genome Atlas.
3.3 Correlations between ERG expression and clinical features in EC
The chi-square test was performed to determine the clinical features of patients with EC associated with ERG expression. There was a significant difference between low and high ERG expression in terms of clinical stage (p = 0.011), histological type (p < 0.001), histological grade (p < 0.001), tumour invasion degree (p = 0.019), and menopause status (p = 0.008). The endometrioid type was identified as the main histological type (Table 1). No significant differences were observed between the tumour ERG expression level with age and body mass index (BMI). These results indicate that low ERG expression may be related to EC invasion and poor patient prognosis.
| Characteristic | Low ERG expression | High ERG expression | p-value | χ2 | |
| n | 277 | 277 | |||
| Clinical stage, n (%) | |||||
| Stage I | 154 (27.8%) | 189 (34.1%) | 0.011 | 11.128 | |
| Stage II | 28 (5.1%) | 24 (4.3%) | |||
| Stage III | 75 (13.5%) | 55 (9.9%) | |||
| Stage IV | 20 (3.6%) | 9 (1.6%) | |||
| Age, n (%) | |||||
| ≤60 yr | 101 (18.3%) | 106 (19.2%) | 0.684 | 0.165 | |
| >60 yr | 174 (31.6%) | 170 (30.9%) | |||
| BMI, n (%) | |||||
| ≤30 kg/m2 | 109 (20.9%) | 103 (19.8%) | 0.280 | 1.166 | |
| >30 kg/m2 | 144 (27.6%) | 165 (31.7%) | |||
| Histological type, n (%) | |||||
| Endometrioid | 182 (32.9%) | 230 (41.5%) | <0.001 | 21.996 | |
| Mixed | 17 (3.1%) | 7 (1.3%) | |||
| Serous | 78 (14.1%) | 40 (7.2%) | |||
| Histologic grade, n (%) | |||||
| G1 | 29 (5.3%) | 70 (12.9%) | <0.001 | 35.249 | |
| G2 | 47 (8.7%) | 74 (13.6%) | |||
| G3 | 193 (35.5%) | 130 (23.9%) | |||
| Tumour invasion, n (%) | |||||
| <50 | 114 (23.9%) | 147 (30.9%) | 0.019 | 5.444 | |
| ≥50 | 117 (24.6%) | 98 (20.6%) | |||
| Menopause status, n (%) | |||||
| Pre | 18 (3.6%) | 17 (3.4%) | 0.008 | 9.644 | |
| Peri | 2 (0.4%) | 15 (3.0%) | |||
| Post | 227 (44.8%) | 228 (45.0%) | |||
G: histopathological cancer grading (G1: well differentiated; G2: moderately differentiated; G3: poorly differentiated); ERG: erythroblast transformation-specific related gene; BMI: body mass index. |
3.4 Prognostic and diagnostic significance of ERG in EC
Patients were divided into high- and low-expression groups according to the median tumour ERG expression level, and the KM plotter was used to assess survival rate. The OS period was shorter (indicative of poor prognosis) in the low-expression group than in the high-expression group (HR = 0.60, p = 0.015; Fig. 3A). Low ERG expression correlated with poor survival metrics in patients with BMI >30 kg/m2 (HR = 0.51, p = 0.023; Fig. 3B), and tumour invasion <50% (HR = 0.41, p = 0.043; Fig. 3C), or age ≤60 years (HR = 0.40, p = 0.043; Fig. 3D). The ROC curve had an AUC value of 0.964 (0.938–0.990; Fig. 3E), which could be used to distinguish healthy individuals from patients with EC. There was no significant difference between ERG expression and prognosis in clinical staging, histological type, histological grade, and menopause status (Supplementary Figs. 1–12). The ROC curve analysis results also indicated that ERG expression has diagnostic value in EC.
Fig. 3.Prognostic and diagnostic values of ERG expression in patients with endometrial carcinoma (EC). (A–D) Kaplan-Meier plotter analysis showing the prognostic relationship between ERG expression in patients with EC and (A) overall survival, (B) body mass index >30 kg/m2, (C) tumour invasion <50%, or (D) age ≤60 years. (E) Receiver operating characteristic curve showing the diagnostic value of ERG expression. HR: hazard ratio; ERG: erythroblast transformation-specific related gene; BMI: body mass index.
3.5 Correlation between ERG expression and tumour-infiltrating immune Cells in EC
We used the GSCA database to explore whether ERG expression correlates with tumour immune invasion and immunotherapy response in EC. Tumour ERG expression correlated positively with the presence of cluster of differentiation 4 (CD4)+ T cells (r = 0.067, pFDR < 0.001), central memory T cells (r = 0.38, pFDR < 0.001), T follicular helper (Tfh) cells (r = 0.33, pFDR < 0.001), mucosal-associated invariant T (MAIT) cells (r = 0.27, pFDR = 0.003), natural killer (NK) cells (r = 0.21, pFDR = 0.011), CD8+ naive cells (r = 0.20, pFDR = 0.024), and T helper 2 (Th2) cells (r = 0.19, pFDR = 0.028). Meanwhile, negative correlations were observed with monocyte (r = −0.38, pFDR < 0.001), dendritic cell (DC) (r = −0.36, pFDR < 0.001), macrophage (r = −0.35, pFDR < 0.001), effector memory T cell (r = −0.32, pFDR < 0.001), exhausted T cell (r = −0.28, pFDR < 0.001), neutrophil (r = −0.23, pFDR = 0.007), and regulatory T (Treg) cell (r = −0.20, pFDR = 0.016) counts and infiltration score (r = −0.17, pFDR = 0.048) (Fig. 4, Supplementary Figs. 13–27). These results indicate that ERG expression may be associated with the immune invasion observed in EC.
Fig. 4.Correlation between tumour ERG expression and tumour immune cell infiltration. Solid red circles, significant positive correlation; solid blue circles, significant negative correlation; the circle size corresponds to the correlation strength. ERG: erythroblast transformation-specific related gene; FDR: false discovery rate; CD: cluster of differentiation; MAIT: mucosal-associated invariant T; NK: natural killer; DC: dendritic cell.
3.6 Identification of potential pathways of ERG involvement in EC
To analyse the potential pathways related to ERG in EC, genes related to ERG were identified using the LinkedOmics portal. Using pFDR < 0.01 as a selection criterion, we identified 4962 ERG-related genes (Fig. 5A), which were imported into the DAVID bioinformatics tool, and GO term and KEGG pathway enrichment analyses were conducted. The main enriched BPs included regulation of transcription (DNA-templated), regulation of transcription from RNA polymerase II promoters, cilium assembly, heart development, regulation of catalytic activity, cell division, protein photoluminescence, and positive regulation of endothelial cell proliferation (Fig. 5B). The main enriched KEGG pathways included Huntington’s disease, proteasome, spinocerebellar ataxia, and calcium signalling pathways as well as pathways associated with cancer, neurodegeneration-multiple diseases, cortisol synthesis and secretion, Parkinson’s disease, renin secretion, and gap junctions (Fig. 5C).
Fig. 5.Gene Ontology (GO) term, biological process (BP), and Kyoto Encyclopedia of Genes and Genome (KEGG) pathway enrichment analyses. (A) Genes related to ERG. Positively associated genes are indicated in red and negatively associated genes in green. (B) Dot plot of the top 10 most significantly enriched GO BPs obtained using the DAVID v6.8 bioinformatics tool. Dot size indicates the number of genes enriched in the BP and colour indicates the p-value. (C) Dot plot of the top 10 most significantly enriched KEGG pathways. Dot size indicates the number of genes enriched in the BP and dot colour indicates the p-value.
3.7 Prediction of drug sensitivity
The IC50 value is a measure of drug efficacy; a low IC50 value indicates that the drug is effective at low concentrations. Additionally, the IC50 value can be used to describe the drug resistance of cancer cells. The pRRophetic analysis of ERG expression in EC and the IC50 values of various therapeutic drugs showed that 24 drugs, including a C-Jun N-terminal kinase inhibitor (AS601245) and checkpoint kinase inhibitor (AZD7762), are expected to be significantly more effective in patients with low ERG expression than in patients with a high expression (Fig. 6). The prediction of drug sensitivity can provide clinical guidance for EC treatment.
Fig. 6.IC50 values of 24 drugs in the high- and low-ERG expression groups. High expression and low expression are indicated by red and blue, respectively (p < 0.001). ERG: erythroblast transformation-specific related gene.
4. Discussion
ERG is essential for postnatal vascular development and is expressed throughout the life cycle of endothelial cells. Moreover, it is a regulator that maintains vascular stability, growth, and participates in tumour angiogenesis and expansion [21].
The results of this study indicate that ERG is aberrantly expressed in numerous cancer types, suggesting it plays a central role in various human cancers. The ROC curve analysis demonstrated that ERG expression had a diagnostic value for EC. Additionally, ERG expression was related to tumour immune invasion and participates in processes and pathways associated with EC development and progression. The potential prognostic value of ERG expression was assessed in EC patients, and it was found that ERG expression was downregulated in those patients; the GEPIA database was used to verified and confirmed its universal downregulation in EC. Therefore, low ERG tumour expression was associated with poor prognosis. A previous study showed that ectopic ERG expression may upregulate hexokinase 2 and phosphoglycerate kinase 1, induce glycolysis, and promote the progression of cervical cancer [22]. However, the precise role of ERG in terms of the occurrence and development of EC had not been previously elucidated. In prostate cancer, ERG–TMPRSS2 fusion is important in tumour development [23]. Meanwhile, although we identified copy number variations as the most common ERG alterations in EC, gene fusion was absent. Moreover, while the risk of EC reportedly increases with age, BMI, and menopausal status (higher risk in postmenopausal women) [3], our results revealed that ERG expression in EC had no significant correlation with age or BMI.
A study on the significance of ERG expression in renal cell carcinoma prognosis showed that the downregulation or deletion of ERG positively correlates with a poor prognosis [10]. Similarly, our results indicated that downregulated ERG expression is associated with a worse prognosis in EC. Moreover, in a previous study where EC prognosis was determined by constructing ERG signals, and low OS and progression-free survival rates were observed in the high-risk patient group [24]. Several factors, including tumour immune cell infiltration, affect EC prognosis, and tumour-infiltrating immune cells can form an ecosystem within the tumour microenvironment that regulates tumour progression and has potential prognostic value [25]. We found that ERG expression in EC is associated with immune cell infiltration. According to the specificity of major histocompatibility complex class I and II molecules, T cells are classified into CD4+ and CD8+ T cells, with CD4+ T cells comprising Th1, Th2, Tfh and Treg cells [26]. Upon encountering antigens, naive CD8+ and CD4+ T cells proliferate and differentiate into various types of effector cells to fight infections or tumours. A portion of these cells survive long-term and become memory T cells. Furthermore, MAIT cells, a subset of T cells that have regulatory properties similar to NK cells in autoimmune models and diseases, can potentially kill cancer cells [27]. Our tumour immune cell infiltration analysis revealed that tumour ERG expression correlated positively with CD4+ T, naive CD8+, MAIT and NK cell infiltration and negatively with DC, monocyte, macrophage, and Treg cell infiltration. DCs are derived from macrophages or DC progenitor cells found in bone marrow and are part of the heterogeneous family of myeloid antigen-presenting cells that comprise various subsets [28, 29]. Myeloid cells have immunosuppressive characteristics and can promote the development of tumours [30]. Hence, the presence of monocytes, tumour-associated macrophages, and Treg cells indicates invasive progression and poor prognosis [31, 32, 33]. Our results suggest that tumour ERG expression regulates immune infiltration, which may affect EC prognosis, suggesting that immunotherapy may be effective for EC depending on the ERG expression status.
ERG is a specific vascular endothelial cell marker involved in regulating Wnt/β-catenin signalling, which affects the expression of vascular endothelial-cadherin—an adhesion molecule involved in endothelial cell–cell connections—and ultimately modulates vascular integrity [21]. In vivo, blood vessels formed in response to vascular endothelial growth factor activity are highly permeable and unstable. ERG promotes the stability of vascular endothelial growth factor-induced blood vessels in vivo and contributes to angiogenesis [34]. In contrast, the absence of ERG leads to angiogenesis defects in the retina and tumours by reducing blood vessel stability [21]. Meanwhile, ERG promotes tumorigenesis when expressed in vitro, in sharp contrast to its effect in vivo [35]. Our GO term enrichment analysis revealed that ERG participates in the transcriptional regulation of endothelial cell proliferation in EC. Similarly, a previous study paired thousands of RNA polymerase II-related enhancers with their target genes in prostate cancer, revealing multiple RNA polymerase II network hubs for receiving, processing, and transmitting regulatory signals. Abnormal regulation of these hubs in cancer can lead to resistance to systemic chemotherapy and molecular-targeted therapy [36]. This can guide the formulation of intervention strategies to prevent or delay cancer onset. We also found a significant correlation between ERG expression and RNA polymerase II transcriptional regulation in EC. We infer that the development and progression of EC and prostate cancer have similar mechanisms. Furthermore, the KEGG pathway enrichment analysis indicated that ERG is involved in calcium signalling, which plays a central role in cancer development [37].
Our study has limitation as well. This is a prospective study, the results are based solely on bioinformatic analyses. Therefore, numerous prospective clinical investigations are needed to validate the predictive value of ERG on the prognostic of EC.
5. Conclusions
ERG expression is downregulated in EC and might interfere with tumour immune cell infiltration and influence patient survival. In particular, low ERG expression in tumours is associated with relatively poor EC prognosis. Thus, ERG expression may serve as a prognostic biomarker and immunotherapy target in EC.
Abbreviations
AOD, average optical density; AUC, area under the curve; BMI, body mass index; BP, biological process; CD4, cluster of differentiation 4; DC, dendritic cell; EC, endometrial carcinoma; ERG, erythroblast transformation-specific-related gene protein; FDR, false discovery rate; GEPIA, Gene Expression Profiling Interactive Analysis; GO, Gene Ontology; GSCA, Gene Set Cancer Analysis; HR, hazard ratio; IC50, half-maximal inhibitory concentration; KEGG, Kyoto Encyclopedia of Genes and Genomes; KM, Kaplan-Meier; MAIT, mucosal-associated invariant T; NK, natural killer; OS, overall survival; ROC, receiver operating characteristic; TCGA, The Cancer Genome Atlas; Tfh, T follicular helper; Th2, T helper 2; Treg, regulatory T.
Availability of data and materials
The datasets generated and analysed during the current study are publicly available, and details are provided in the article.
Author contributions
YYZ and GMW—conceptualised and designed the research study; reviewed and revise the manuscript. CFZ, XYL and FHD—organised the data. JLM—analysed the data. JLM and HJZ—performed software analysis; wrote the manuscript. All authors read and approved the final manuscript.
Ethics approval and consent to participate
The data used in this study were sourced from public databases. The contributors to the database have obtained ethical approval. Our research has no ethical issues and there was no requirement for informed consent.
Acknowledgment
Not applicable.
Funding
This research was funded by the Yunnan Provincial Natural Science Foundation, grant number: 202001BA070001-133; Medical Discipline Leader of the Yunnan Provincial Commission of Health and Family Planning, grant number: D-2017057; Dali Science and Technology Plan Project, grant number: 2022KBG029; and key construction discipline of The First Affiliated Hospital of Dali University.
Conflict of interest
The authors declare no conflict of interest.
Supplementary material
Supplementary material associated with this article can be found, in the online version, at https://oss.ejgo.net/files/article/1900459494759907328/attachment/Supplementary%20material.docx.
