Diagnostic and Prognostic Values of a Branched-chain Amino Acid-associated Gene Signature in Lung Cancer

Abstract

BackgroundLung cancer is the malignancy with the highest morbidity and mortality worldwide. Several studies have indicated that branched-chain amino acids (BCAAs) can regulate cancer progression and metabolism. In this study, a BCAA-associated prognostic signature was established and assessed to verify whether it can be used in clinical treatment of patients with non-small-cell lung carcinoma (NSCLC).MethodsThe candidate BCAA-associated genes were screened from TCGA cohort by differential expression analysis. Two datasets, the training and test sets, were divided from the lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC), and LUSC-only datasets in TCGA cohort and the novel model were constructed based on the training dataset. Moreover, all prognosis-related genes were screened via univariate Cox analysis. The LASSO Cox method was used to avoid collinearity. To further probe the clinical effect, a large number of clinical characteristics were included in the prognostic risk model. The outcomes for the calibration curves and nomogram proved the capability of the prognostic signature.ResultsBased on the 366 samples in the training set, six survival-related genes (CAV1, GRIA1, LDHA, AOX1, GRIN2D, and AKR1C1) were screened using the LASSO Cox regression analysis. Patients were divided into high- and low-risk groups. The signature demonstrated good diagnostic and prognostic capabilities.ConclusionWe developed the first-ever BCAA-related signature with both prognostic predictive power and diagnostic efficacy. The prognostic signature and diagnostic model suggested that the expression of BCAAs could serve as a powerful potential marker in cancer clinical treatment in the future.