Background
Lung adenocarcinoma (LUAD) is the major non-small-cell lung cancer pathological subtype with poor prognosis worldwide. Glutamine metabolism (GM) plays an important role in tumor development and progression. Herein, we aimed to construct a GM-associated prognostic gene signature to predict survival in patients with LUAD.
Methods
The gene expression profiles of patients with LUAD were downloaded from the Cancer Genome Atlas (TCGA), and two independent datasets for validation were downloaded from the Gene Expression Omnibus (GEO). 41 GM-related genes were downloaded from the previous literatures and used to establish GM-related risk model (names as Score). The prognostic GM-associated gene signatures were identified using univariate Cox and LASSO analyses. Kaplan-Meier and receiver operating characteristic (ROC) curves were performed to validate the predictive efficacy of the prognostic signatures. The correlation between the GMScore and tumor-infiltrated immune cells was evaluated using CIBERSORTx analysis. A nomogram was used to predict the survival probability of LUAD based on GMScore.
Results
We constructed a prognostic signature comprising 7 GM-related genes (AMDHD1, GAPDH, GCLC, GLS2, HAL, SLC38A2, SLC7A5). The Kaplan-Meier curves validated the reliable predictive ability of the prognostic signature in TCGA and two GEO datasets (p < 0.001, p = 0.001, and p = 0.003, respectively). The area under the curve (AUC) of the ROC curves validated the predictive accuracy of the risk model. We constructed a nomogram to predict the survival probability of LUAD, and the calibration curves showed well predictive accuracy. Finally, functional analysis also revealed distinct immune status between high and low GMScore groups in LUAD.
Conclusion
Our study constructed a GM-related gene signature, and indicated that GMScore could serve as a novel biomarker for predicting patients' survival in LUAD.