Prediction of lung adenocarcinoma prognosis and clinical treatment efficacy by telomere-associated gene risk model.

Background: The most prevalent cause of cancer-related death in China and across the globe is lung adenocarcinoma (LUAD). Telomere shortening (TS) has been found to contribute to the development of LUAD. Therefore, our aim is to investigate the impact of telomere-related genes (TRGs) on immunotherapy and clinical prognosis prediction in LUAD.

Methods: TRGs were obtained from TelNet, while RNA-seq and clinical information were retrieved from the GEO and TCGA databases. TelNet preserves a series of genes known to be engaged in telomere maintenance and also provides information on the type of telomere maintenance mechanism in which the gene is involved. Data pertinent to RNA sequencing and clinical parameters were accessed from two widely-accessed electronic repositories- the GEO and TCGA databases, respectively. We conducted univariate Cox regression analysis in order to recognize prognostic TRGs and employed multivariate Cox regression analysis to develop a risk model for these TRGs. The patients were stratified into high-risk and low-risk groups based on the first quartile of the risk score. The predictive ability and stability of the model were subsequently verified through Kaplan-Meier analysis, ROC curve, and C-index. We investigated the immune landscapes of different risk groups and predicted their responses to immunotherapy. Lastly, we evaluated the sensitivity of different groups to commonly used chemotherapeutic and targeted drugs through drug sensitivity analysis.

Results: Univariate Cox analysis identified 12 prognostic TRGs, while a signature consisting of 4 prognostic TRGs was constructed through multivariate Cox analysis. Survival analysis indicated a significantly shorter survival time in the high-risk group. The predictive immunotherapy analysis suggested that patients in the high-risk group may have a more favorable response to immunotherapy. Finally, we identified 28 appropriate chemotherapeutic and 51 targeted drugs for different patient groups.

Conclusions: The study has successfully developed a prognostic model for LUAD prediction that takes into account TRGs and predicts both prognosis and response to immunotherapy.