Identification of Prognostic miRNAs Associated With Immune Cell Tumor Infiltration Predictive of Clinical Outcomes in Patients With Non-Small Cell Lung Cancer

Abstract

BackgroundA detailed means of prognostic stratification in patients with non-small cell lung cancer (NSCLC) is urgently needed to support individualized treatment plans. Recently, microRNAs (miRNAs) have been used as biomarkers due to their previously reported prognostic roles in cancer. This study aimed to construct an immune-related miRNA signature that effectively predicts NSCLC patient prognosis.MethodsThe miRNAs and mRNA expression and mutation data of NSCLC was obtained from The Cancer Genome Atlas (TCGA). Immune-associated miRNAs were identified using immune scores calculated by the ESTIMATE algorithm. LASSO-penalized multivariate survival models were using for development of a tumor immune-related miRNA signature (TIM-Sig), which was evaluated in several public cohorts from the Gene Expression Omnibus (GEO) and the CellMiner database. The miRTarBase was used for constructing the miRNA-target interactions.ResultsThe TIM-Sig, including 10 immune-related miRNAs, was constructed and successfully predicted overall survival (OS) in the validation cohorts. TIM-Sig score negatively correlated with CD8+ T cell infiltration, IFN-γ expression, CYT activity, and tumor mutation burden. The correlation between TIM-Sig score and genomic mutation and cancer chemotherapeutics was also evaluated. A miRNA-target network of 10 miRNAs in TIM-Sig was constructed. Further analysis revealed that these target genes showed prognostic value in both lung squamous cell carcinoma and adenocarcinoma.ConclusionsWe concluded that the immune-related miRNAs demonstrated a potential value in clinical prognosis.

Identification of miRNA signature for predicting the prognostic biomarker of squamous cell lung carcinoma

“…Then, the minimum absolute contraction and selection operator (LASSO) penalty Cox model was introduced to analyze the variable selection, which has been successfully applied to the creation of multiple biomarker models [ 41 ]. Zhang et al [ 42 ] used LASSO-penalized multivariate survival models to predict the immune-associated miRNAs involved in the development of LUSC. In addition, more in-depth computational models can be used for the identification of miRNA biomarkers in human cancers.…”

Novel prognostic biomarkers, METTL14 and YTHDF2, associated with RNA methylation in Ewing’s sarcoma

“…CellMiner software is based on 60 cancer cells listed by the National Cancer Institute's Center for Cancer Research (NCI). The NCI-60 cell line is currently the most widely used cancer cell sample population for anti-cancer drug testing 26 , 27 . It allows researchers to query data on the 22,379 genes identified in the NCI-60 cell line, as well as the 20,503 compounds analyzed (including 102 drugs approved by the US Food and Drug Administration).…”