Exploitation of a shared genetic signature between obesity and endometrioid endometrial cancer

Abstract

Aims: The findings in epidemiological studies suggest that endometrioid endometrial cancer (EEC) is associated with obesity. However, evidence from gene expression data for the relationship between the two is still lacking. The purpose of this study was to explore the merits of establishing an obesity-related genes (ORGs) signature in the treatment and the prognostic assessment of EEC.

Methods: Microarray data from GSE112307 were utilized to identify ORGs by using weighted gene co-expression network analysis. Based on the sequencing data from TCGA, we established the prognostic ORGs signature, confirmed its value as an independent risk factor, and constructed a nomogram. We further investigated the association between grouping based on ORGs signature and clinicopathological characteristics, immune infiltration, tumor mutation burden and drug sensitivity.

Results: A total of 10 ORGs were identified as key genes for the construction of the signature. According to the ORGs score computed from the signature, EEC patients were divided into high and low-scoring groups. Overall survival (OS) was shorter in EEC patients in the high-scoring group compared with the low-scoring group (P < 0.001). The results of the Cox regression analysis showed that ORGs score was an independent risk factor for OS in EEC patients (HR = 1.017, 95% confidence interval = 1.011-1.023; P < 0.001). We further revealed significant disparities between scoring groups in terms of clinical characteristics, tumor immune cell infiltration, and tumor mutation burden. Patients in the low-scoring group may be potential beneficiaries of immunotherapy and targeted therapies.

Conclusions: The ORGs signature established in this study has promising prognostic predictive power and may be a useful tool for the selection of EEC patients who benefit from immunotherapy and targeted therapies.

Keywords: endometrioid endometrial cancer; immune correlation analyses; obesity-related genes; targeted treatment; weighted gene coexpression network analysis.

Figure 1

The flow plot of this study.

Figure 2

Identification and functional annotation of ORGs. (A) The cluster dendrogram of ORGs in obesity. (B) Correlation of WGCNA modules and obesity. (C) GO analysis of ORGs. (D) KEGG analysis of ORGs. ORGs, obesity-related genes. WGCNA, weighted gene co-expression network analysis. GO, Gene Ontology. KEGG, Kyoto Encyclopedia of Genes and Genomes.

Figure 3

Establishment of the ORGs signature in the train cohort. (A) Univariate Cox regression analysis for screening prognostic ORGs. (B,C) LASSO regression analysis for variable selection and avoid overfitting. (D) PCA plot for different ORGs scoring groups. (E) Scatter diagram for the ORGs score and survival status of EEC patients. (F) Heat map for the key 10 ORGs expression with ORGs grouping. (G) Kaplan–Meier curves of survival difference between two groups. ORGs, obesity-related genes. LASSO, least absolute shrinkage and selection operator. PCA, principal component analysis. EEC, endometrioid endometrial cancer.

Figure 4

Validation of the ORGs signature in the test cohort and the entire cohort. PCA plot (A), scatter plot (B), heat map (C), and Kaplan–Meier curves (D) for the test cohort. PCA plot (E), scatter plot (F), heat map (G), and Kaplan–Meier curves (H) for the entire cohort. ORGs, obesity-related genes. PCA, principal component analysis.

Figure 5

Correlation analyses of clinical features. Univariate (A) and multivariate (B) Cox regressions of ORGs score, age, tumor grade, and FIGO stage. The ROC (C) and C-index (D) of ORGs score, age, tumor grade, and FIGO stage. (E) The nomogram for predicting prognosis of EEC patients. (F) The calibration curves of the nomogram. ORGs, obesity-related genes. FIGO, the International Federation of Gynecology and Obstetrics. ROC, receiver operating characteristic curve. EEC, endometrioid endometrial cancer.

Figure 6

Subgroup analysis of EEC patients. Kaplan–Meier curves of patients with white race (A), age < 65 (B), age ≥ 65 (C), FIGO stage I–II (D), FIGO stage III–IV (E), tumor grade 1–2 (F), tumor grade 3 (G), BMI ≥ 30 (H), BMI < 30 (I). EEC, endometrioid endometrial cancer. FIGO, the International Federation of Gynecology and Obstetrics. BMI, body mass index.

Figure 7

Analysis of immune activity. Comparison of the discrepancy of immune cell infiltration (A) and immune function (B) between two groups based on ssGSEA. (C) Differences in the expression of LAG-3 and PD-1 between the two groups. (D) TME analysis based on the ESTIMATE algorithm. ssGSEA, single-sample gene set enrichment analysis. TME, tumor microenvironment.

Figure 8

Analyses of mutation data. (A) Visualization of somatic mutations in different ORGs groups. (B) Differential analysis of TMB in ORGs scoring groups. (C) Kaplan–Meier curves of survival differences between the high- and low-TMB groups. (D) Kaplan–Meier curves of survival differences between ORGs scoring groups in low-TMB patients. (E) Distribution of MSI status in the different scoring groups. (F) Differential analysis of the ORGs score in patients with different MSI status. Kaplan–Meier curves of survival differences between ORGs scoring groups in patients with MSI-H (G) and MSS (H). ORGs, obesity-related genes. TMB, tumor mutation burden. MSI, microsatellite instability. MSI-H, high microsatellite instability. MSS, microsatellite stable.

Figure 9

Drug sensitivity analyses. Analysis of drug sensitivity differences in olaparib (A), talazoparib (B), niraparib (C), 5-fluorouracil (D), oxaliplatin (E), and cyclophosphamide (F).