Identification of a novel prognostic gene signature in pleural mesothelioma: a study based on The Cancer Genome Atlas database and experimental validation.

BACKGROUND

Early detection and prognostic prediction are crucial in improving the survival of patients with pleural mesothelioma (PM). Therefore, this study aimed to develop a gene prognostic risk model for PM patients based on The Cancer Genome Atlas (TCGA) database analysis and experimental validations.

METHODS

Obtaining gene expression data and clinical information of PM from the TCGA database, the dataset was divided into a training set and a testing set. Univariate Cox regression analysis, robust testing, and multivariate Cox regression analysis were performed on the training set to establish a prognostic risk model. Risk scores were calculated for each patient, and the dataset was stratified into high- and low-risk groups. The predictive efficacy and accuracy of the model were evaluated using Kaplan-Meier survival curves and receiver operating characteristic (ROC) curves. The messenger RNA (mRNA) expression levels of genes in the prognostic model in clinical samples and PM cell lines were detected by quantitative reverse transcription polymerase chain reaction (qRT-PCR). Gene expression validation in the prognostic model was conducted using samples from the TCGA and the Genotype-Tissue Expression (GTEx) project databases. The University of ALabama at birmingham CANcer data analysis portal (UALCAN) database was utilized to explore the expression patterns of genes in the prognostic model. Finally, gene set enrichment analysis (GSEA) was performed on genes in the prognostic model to explore their potential biological functions and signaling pathways.

RESULTS

A prognostic risk assessment model consisting of three genes, ubiquitin like with PHD and ring finger domain 1 (), kinesin family member 4A (), and never in mitosis gene A-related kinase 2 () was constructed. The risk score of the prognostic model is calculated as follows: risk score = Expression level of UHRF1 × 1.4525 - Expression level of KIF4A × 1.3270 + Expression level of NEK2 × 1.4167. Patients were further stratified into high- and low-risk groups at this optimal cutoff point. Kaplan-Meier curves demonstrate that, compared to patients in the high-risk group, those in the low-risk group exhibited significantly prolonged overall survival. Visualization of the model through a forest plot revealed a Log-Rank P<0.0001 for the entire model, indicating its potential as an independent prognostic marker for PM. The mRNA expression levels of three genes in the prognostic model significantly elevated in tumor samples and PM cell lines than in non-tumorigenic tissues and cell lines as detected by qRT-PCR. Additionally, these genes exhibited significant differences in expression among PM patients of different stages, tumor subtypes, ages, and metastatic statuses. The overexpressed group of these three genes was significantly enriched in pathways such as DNA replication, mRNA surveillance pathway, nuclear transport, ribosome biogenesis in eukaryotes, and spliceosome pathways.

CONCLUSIONS

Three prognostic marker genes (, , and ) as a gene cluster may serve as prognostic marker genes in PM.