The Research For Diagnostic Significance Of Plasma LncRNAs For Lung Cancer

Objective: Lung cancer is a common malignant tumor in humans.The early diagnosis rate of lung cancer is very low,and the overall prognosis is poor.The aim of this study was to identify the dysregulated lncRNAs in the plasma of non-small cell lung cancer(NSCLC)and to construct a lncRNA combination model for the diagnosis of lung cancer.By comparing with conventional protein markers,to evaluate the diagnostic efficacy of plasma lncRNA combination model for NSCLC,especially for the NSCLC in early stage.In addition,the diagnostic value of plasma lncRNAs for small cell lung cancer will also be explored.Methods: 1.Based on previous studies and database data,we selected 31 lung cancer-related lncRNAs as preliminary screening targets.Primers were designed and validated by using NCBI Primer-BLAST,UCSC database.RT-qPCR was used to detect the expression of lncRNAs in small plasma samples,and the lncRNAs which can be stablely and reliablely detected were selected as the candidate lncRNAs to construct the diagnostic model for lung cancer.2.Two independent sets of plasma samples from 488 participants were available for the detection of candidate lncRNAs(training set = 265;validation set = 223).The expression results of lncRNAs were analyzed by Logistic regression to construct a lncRNA combination model.ROC curves were plotted to evaluate the diagnostic value of each lncRNA and lncRNA combination model.3.The diagnostic value of conventional protein markers for adenocarcinoma and squamous cell carcinoma were calculated and the optimal protein diagnostic models were constructed respectively.The diagnostic value of lncRNA combination model for lung cancer were evaluated by compared with protein diagnostic models.4.The correlation between the expression of plasma lncRNAs and clinical features of NSCLC patients were evaluated by comparing the differential expression of lncRNAs in NSCLC subgroups with different clinical features.5.The expression levels of candidate lncRNAs and protein markers in the plasma of small cell lung cancer were detected,and the lncRNA combination model and protein combination model for the diagnosisof small cell lung cancer were constructed,respectively,then their diagnostic values were compared.Results: 1.By detecting the expression of 31 lung cancer-related lncRNAs in plasma samples,we screened 5 lncRNAs(RMRP,NEAT1,TUG1,MALAT1 and H19)that were stably expressed in plasma.2.By performing a differential expression screening,we excluded the non-differentially expressed lncRNA H19.RMRP,NEAT1,TUG1 and MALAT1 were included in the diagnostic model by stepwise logistic regression analysis.The lncRNA combination model had a high diagnostic value for distinguishing NSCLC from non-NSCLC controls.The AUC values in the two stages were 0.86 and 0.89,respectively.The sensitivities were85.26% and 87.02%,respectively,and the specificity were 76.23% and 76.05%,respectively.3.We also constructed protein combination models for lung adeno-carcinoma and squamous cell carcinoma(adenocarcinoma,CEA+CA125+CYFR-A21-1;squamous cell carcinoma,SCC+CYFRA21-1)respectively by Logistic regression.Compared with protein combination models,the lncRNA combination model has a higher sensitivity(lncRNA combination model=70.04%-92.02%;protein combination models=18.63%-65.64%)for NSCLC(including the early stage NSCLC and small nodule NSCLC).Especially,the diagnostic sensitivity of lncRNA combination model for the early stage NSCLC was 81.41%-83.34%,and that of the protein combination models was 18.63%-65.64%.For NSCLC with small nodule,the negative likelihood ratio of the lncRNA combination model(0.19-0.43)was also better than that of protein combination models(0.70-0.80).However,protein combination models have higher specificity and positive accuracy.4.NEAT1 was highly expressed in NSCLC with tumor size >3 cm(P = 0.001).TUG1 was associated with lymph node metastasis(P = 0.001),distant metastasis(P = 0.025),and TNM stage(P = 0.001)in NSCLC.5.NEAT1,TUG1,and MALAT1 were differentially expressed in plasma samples of small cell lung cancer.The lncRNA combination model(NEAT1+TUG1+MALAT1)obtained by logistic regression had a diagnostic AUC value of 0.71(95% CI: 0.63-0.78)for small cell lung cancer,which was lower than the AUC(0.95)of protein combination model(pro-GRP+NSE).Conclusions: 1.LncRNA combination model has a good diagnostic efficacy for non-small cell lung cancer,especially for early stage NSCLC and NSCLC with small nodule.2.For the diagnosis of NSCLC,lncRNA combination model has a better sensitivity and protein combination model has a better specificity.These two models can complement eachother to promote the NSCLC diagnosis in clinical practice.3.The diagnostic value of NEAT1+TUG1+MALAT1 for small cell lung cancer is not as good as that of protein marker model(pro-GRP+NSE).