Construction And Verification Of A Prognostic Model For Clear Cell Renal Cell Carcinoma Integrating Clinical Features And Long Non-coding RNA

ObjectiveConstruct and validate the potential composite model integrating clinical factors and long non-coding RNAs(lnc RNAs),and evaluate the predictive value of this model in clear cell renal cell carcinoma(cc RCC).This model will provide a theoretical basis for predict overall survival of cc RCC in white patients.MethodsThe transcriptome,mi RNA,and clinical data of 459 white patients with cc RCC were downloaded from the Cancer Genome Atlas database.383 subjects who met the inclusion and exclusion criteria were randomized into training cohort(n = 255)and test cohort(n = 128)at a ratio of 2:1.The model was constructed using the training cohort,and validated in the test cohort.In the training cohort,the prognostic lnc RNA model was constructed through the univariate and multivariate Cox regression analysis,and least absolute shrinkage and selection operator regression analysis.Subsequently,the composite model was constructed.The concordance index(95%CI),area under timedependent receiver operator characteristic curve(95%CI),nomogram,and corresponding calibration curve were used to evaluate the prognostic power of the models.Whether the model can benefit patients was evaluated by the decision curve.Subsequently,the prognostic model constructed by the training cohort was verified in the test and entire cohorts.In order to better explore the function of lnc RNAs in the prognostic model,a competitive endogenous RNA network was drawn.Gene ontology and pathway enrichment analysis were performed according to the target genes in the network,and a protein-protein interaction network was constructed.According to protein-protein interaction network,the top ten hub genes were screened through 12 algorithms,and the expression values of these hub genes were analyzed using the UALCAN online website.Results1.In the training cohort,12 lnc RNAs were screened as prognostic related biomarkers for white cc RCC patients,of which five(AC008556.1,AC012404.1,AC092296.1,AC099684.2,and SPINT1-AS1)were protective factors(hazard ratio <1),and the other seven(AC108752.1,AC131097.1,AL606519.1,FOXP4-AS1,LINC00261,LINC02446,and LINC02475)were risk factors(hazard ratio>1).AC012404.1,AC092296.1,AC099684.2,AC108752.1,AC131097.1,AL606519.1,and LINC02475 were seven novel candidate prognostic biomarkers.2.Risk score,age,and tumor node metastasis stage were all independent prognostic factors for patients with cc RCC.Based on these factors,a composite prognostic model of risk score,age,and tumor node metastasis stage was constructed.The performance evaluation of composite model showed that concordance index(95%CI)was 0.863(0.830-0.896),0.863(0.814-0.912),and 0.841(0.812-0.870)in the training,test,and entire cohorts respectively;the 5-year area under time-dependent receiver operator characteristic curve(95%CI)was 0.923(0.879-0.967),0.861(0.777-0.945),and 0.879(0.834-0.924)in the training,test,and entire cohorts respectively.The prognostic nomogram has good predictive ability.The 3-year and 5-year decision curves showed that application of this model can benefit patients in clinic.3.Competitive endogenous RNA network contained 50 target genes;gene ontology function annotation yielded 59 functional items;pathway enrichment analysis yielded four pathways;29 nodes in the protein-protein interaction network had connections.4.The top ten hub genes(VEGFA,GATA4,DDIT4,NOG,CAV1,CCND1,LOX,PRDM1,NTRK2,and BMP6)were screened out.Conclusions1.This novel prognostic signature incorporating risk score,age,and the tumor node metastasis stage can be applied as an accurate potential tool for cc RCC prognostic evaluation in white patients.2.AC012404.1,AC092296.1,AC099684.2,AC108752.1,AC131097.1,AL606519.1,and LINC02475 are seven novel candidate prognostic biomarkers,which need to be further studied.