COVID-19-associated lncRNAs as predictors of survival in uterine corpus endometrial carcinoma: A prognostic model

Abstract

Background: Patients with uterine corpus endometrial carcinoma (UCEC) may be susceptible to the coronavirus disease-2019 (COVID-19). Long non-coding RNAs take on a critical significance in UCEC occurrence, development, and prognosis. Accordingly, this study aimed to develop a novel model related to COVID-19-related lncRNAs for optimizing the prognosis of endometrial carcinoma. Methods: The samples of endometrial carcinoma patients and the relevant clinical data were acquired in the Carcinoma Genome Atlas (TCGA) database. COVID-19-related lncRNAs were analyzed and obtained by coexpression. Univariate, least absolute shrinkage and selection operator (LASSO), and multivariate Cox regression analyses were performed to establish a COVID-19-related lncRNA risk model. Kaplan-Meier analysis, principal component analysis (PCA), and functional enrichment annotation were used to analyze the risk model. Finally, the potential immunotherapeutic signatures and drug sensitivity prediction targeting this model were also discussed. Results: The risk model comprising 10 COVID-19-associated lncRNAs was identified as a predictive ability for overall survival (OS) in UCEC patients. PCA analysis confirmed a reliable clustering ability of the risk model. By regrouping the patients with this model, different clinic-pathological characteristics, immunotherapeutic response, and chemotherapeutics sensitivity were also observed in different groups. Conclusion: This risk model was developed based on COVID-19-associated lncRNAs which would be conducive to the precise treatment of patients with UCEC.

Keywords: COVID-19–associated lncRNA; OS; TCGA; endometrial carcinoma; prognostic model.

FIGURE 1

Identification and construction of a prognostic model for COVID-19–associated lncRNAs in UCEC patients. (A) Univariate Cox regression analysis revealed that the selected lncRNAs are significantly correlated with clinical prognosis. (B–C) Risk score system was constructed using the LASSO Cox regression model. (D) Multivariate Cox regression analysis shows independent prognostic lncRNAs.

FIGURE 2

Prognosis ability of the risk model of the 10 COVID-19–associated lncRNAs in the training set. (A) Distribution of COVID-19–associated lncRNA model–based risk score. (B) Situations of the survival period and survival state between high- and low-risk groups. (C) Clustering analysis heatmap shows the expression levels of the 10 prognostic lncRNAs for the respective patient. (D) Kaplan–Meier survival curves of OS of patients in the high- and low-risk groups. (E) 1-, 2-, and 3-year ROC curves for OS prediction in accordance with COVID-19–associated lncRNAs.

FIGURE 3

Prognosis ability of the risk model of the 10 COVID-19–associated lncRNAs in the testing and whole sets. Distribution of the risk score, OS, gene expression, survival analysis, and ROC curves for forecasting OS in the (A–E) testing set and (F–J) entire set.

FIGURE 4

Kaplan–Meier analysis of overall survival for UCEC patients according to age, event, histological type, and tumor stage.

FIGURE 5

Principal components analysis between low- and high-risk groups based on (A) the entire gene expression, (B) COVID-19 genes, (C) COVID-19–associated lncRNAs, and (D) the risk model.

FIGURE 6

Clinical evaluation by the risk assessment model. (A–D) Scatter diagram shows the (A) age, (B) event, (C) histological type, and (D) clinical stage.

FIGURE 7

Evaluating correlations of the COVID-19–associated lncRNA model with immunotherapeutics and chemotherapeutics in patients with UCEC. (A) Tumor-infiltrating immune cells and (B) TMB differences in high- and low-risk groups. (C–D) Differential response of chemotherapeutics such as (C) gemcitabine and (D) cisplatin based on IC50 for UCEC patients in high- and low-risk groups.