Construction and validation of lncRNA prognostic model for bladder cancer / 中华检验医学杂志

ABSTRACT

Objective:This study aims to construct a prognostic model of bladder cancer (BLCA) based on lncRNA.Methods:BLCA lncRNA expression data and clinical information were downloaded from TCGA. Univariate Cox regression was used to evaluate the correlation between the expression level of each lncRNA and overall survival (OS), and the lncRNAs with a corrected P-value<0.01 were selected as candidate predictors. In the training queue, the prediction model is constructed by methods such as least absolute shrinkage and selection operator, and multi-factor stepwise Cox regression, and verified in the verification queue at the same time.. Evaluation the area under the curve of time-dependent receiver operating characteristic (tROC) and Harrel C index. According to the median risk score of the prediction model, patients were divided into high-risk group and low-risk group and the differences in clinicopathological characteristics between the two groups were compared by t-test or chi-square test. Results:Establish a BLCA prognostic model based on 13 lncRNAs, of which LINC01465, ARHGAP5-AS1, ZFHX4-AS1, MAFG-AS1 are prognostic risk factors (β regression coefficients are 0.32, 0.16, 0.06, 0.20, respectively, all>0), and the rest are protection factors (β regression coefficients are all<0); the prediction model of the overall survival in the first year, the third year, and the fifth year in the complete cohort has an area under the tROC curve of 0.79, 0.82, and 0.80 respectively, and the Harrell C index is 0.74. Its predictive ability is better than the previously published BLCA prognostic model based on lncRNA. Adjusting for confounding factors including age and tumor stage found that the risk score of this model was an independent poor prognostic factor for overall survival in BLCA patients (hazard ratio 4.05; P<0.001). Comparison of clinicopathological characteristics of patients in the high-risk and low-risk groups showed that in the high-risk group, there were more old patints (70.0 vs. 66.1, P<0.001), more non-papillary patients (74.2% vs. 61.2, P=0.005), more high-stage patients (37.6% vs. 28.0%, P<0.001 for stage Ⅳ patients), and more high-grade tumors (98.0% vs. 92.0%, P=0.005). Conclusion:In this study, a prognostic model of bladder cancer based on 13 lncRNAs was constructed. This model has good predictive ability and can provide value for clinical decision-making and patient consultation.