Construction And Validation Of A Novel Prognostic Signature Of MicroRNAs In Lung Adenocarcinoma

Backgrounds and aims:Lung cancer(LC)is the leading cause of malignancy-related mortality,with approximately one million deaths each year,which is a big concern for public health worldwide.Up to 50% of LC cases are diagnosed as lung adenocarcinoma(LUAD),which is the most common pathological type occurred in LC patients.Despite of the remarkable progress in image scan,bronchoscopy,and novel treatment options,tumor progression is the major obstacles for favorable survival rate.5-year survival rate for lung adenocarcinoma patients is 55.6%(local),compared with 4.5% for metastatic patients.Additionally,40% of stage IB and 66% of stage II patients have to face a high relapse rate associating with a poor prognosis within 5 years.Recent reports have showed that adjuvant chemotherapy confers a survival benefit of 4-15% for the patients with resected stage II–III rather than the patient with stage I 7,8.The limited survival benefits indicate that the current staging system might lack the deficiency of the current staging system and the presence of more unknown tumor characteristics and proved less effective for LUAD patients.Therefore,it is very important to establish a model for diagnosis,classification and prognosis of LUAD,which is conducive to more accurate judgment of the patient's prognosis and facilitates the development of personalized and precise treatment methods.MicroRNA(miRNA)is a type of non-coding RNA with a length of 17-25 bp,which can affect protein-coding genes at the post-transcriptional level by binding to the 3 'untranslated region(3'-UTR)of m RNA.Mi RNAs regulate about half of m RNA expression through this mechanism.A large number of articles have been reported that miRNAs can participate in physiological processes such as cell differentiation,apoptosis,proliferation and cell cycle.In addition,miRNAs are also involved in the development of various cancers,such as pancreatic cancer,colorectal cancer,esophageal squamous cell carcinoma,ovarian cancer,glioma,cervical cancer and prostate cancer.With the application of gene sequencing in tumors,miRNAs are considered as new biomarkers for predicting prognosis and drug resistance.Herein,integrating multiple miRNAs may be more effective and predictive than single miRNAs.The purpose of this study was to screen miRNAs closely related to the prognosis of LUAD,construct a survival prediction model for LUAD patients,and provide a reference for evaluating the prognosis of patients.Methods:The corresponding LUAD miRNA gene expression data and patient clinical profiles were obtained from The Cancer Genome Atlas(TCGA)database,and single factor survival analysis was used to screen out miRNAs related to LUAD's overall survival(OS).The LUAD patients were divided into training group and validation group according to the principle of random allocation.Afterward,1,000 resample LUAD training matrices based on the training set was applied to identify the potential prognostic miRNAs.Through least absolute shrinkage and selection operator Lasso(LASSO)regression analysis,the miRNAs with ?coefficient(?-co-ef)more than 0 was obtained,thus establishing a prognostic model based on miRNAs.The median value of the prognostic model in the training group is defined as the cut-off value,and we assigned the LUAD patients as the high-risk or low-risk patients.Different survivals between the predicted two groups in both training and validation groups mean that the prediction model is meaningful.Using the online prediction tool to accumulate the weighted context ++ score to select the most reliable target genes,we study the functional role of the target genes at the molecular level through Hallmark gene sets and Kyoto Encyclopedia of Genes and Genomes(KEGG).In addition,fresh cancer tissues and adjacent tissue samples from 20 patients with LUAD were collected to verify the expression of miRNAs and target genes in the model,further revealing the potential relationship between the miRNA and its target genes,and validating that the model has certain predictive value.Results:This article identified a 6-miRNA-based LUAD survival prediction model.The formula of individual risk scoring model for patients with LUAD is:-0.365 *mi R-1468 + 0.394 * mi R-299-0.322 * mi R-4709-0.570 * mi R-5571 + 0.166 *mi R-582-0.143 * mi R-3653.In the training and validation groups,Kaplan Meier curve analysis showed that the OS of the high-risk group was worse than that of the low-risk group,which indicated that the prediction model had favorable accuracy for LUAD patients.Hallmark gene set and KEGG functional analyses showed that the target genes of these six miRNAs have potential connections with various cancer biomolecular pathways,such as the signature UV response,Wnt and m TOR signaling pathways.The further detection of mi R-582 and its target genes showed that the expressions of mi R-582-3p and mi R-582-5p in LUAD were lower than those in adjacent tissues,which have tissue differences;the expression of CEP55,RASA1,CHODL and DIXDC1 in LUAD tissue were increased significantly,which is in line with the reverse regulation trend of miRNAs,verifying that the model has certain predictive value.Conclusion:Taken together,the prognostic model of LUAD has certain value for the evaluation of the prognosis of patients with LUAD.