Prognostic values of a novel multi-mRNA signature for predicting relapse of cholangiocarcinoma

Abstract

Cholangiocarcinoma (CCA) is an epithelial cancer and has high death and recurrence rates, current methods cannot satisfy the need for predicting cancer relapse effectively. So, we aimed at conducting a multi-mRNA signature to improve the relapse prediction of CCA. We analyzed mRNA expression profiling in large CCA cohorts from the Gene Expression Omnibus (GEO) database (GSE76297, GSE32879, GSE26566, GSE31370, and GSE45001) and The Cancer Genome Atlas (TCGA) database. The Least absolute shrinkage and selection operator (LASSO) regression model was used to establish a 7-mRNA-based signature that was significantly related to the recurrence-free survival (RFS) in two test series. Based on the 7-mRNA signature, the cohort TCGA patients could be divided into high-risk or low-risk subgroups with significantly different RFS [p < 0.001, hazard ratio (HR): 48.886, 95% confidence interval (CI): 6.226-3.837E+02]. Simultaneously, the prognostic value of the 7-mRNA signature was confirmed in clinical samples of Ren Ji hospital (p < 0.001, HR: 4.558, 95% CI: 1.829-11.357). Further analysis including multivariable and sub-group analyses revealed that the 7-mRNA signature was an independent prognostic value for recurrence of patients with CCA. In conclusion, our results might provide an efficient tool for relapse prediction and were beneficial to individualized management for CCA patients.

Keywords: Gene Expression Omnibus database; cholangiocarcinoma; least absolute shrinkage and selection operator model; mRNA signature; recurrence-free survival..

Figure 1

Identification of differentially expressed genes in cholangiocarcinoma from public CCA datasets. (A-F) Volcano plots of DEGs in the 5 indicated datasets. (X-axis: log2(FC); Y-axis: -log10(FDR) for each gene. Genes with FDR < 0.01 and FC >1.5 or <-1.5 were considered as DEGs in each series. Blue: down-regulated genes; Grey: non-differential genes; Red: up-regulated genes). (G-H) Overlapping analyses of DEGs in TvsP (G) and TvsN (H) groups, DEGs shared within 2 datasets or more were regarded as credible DEGs in each Venn diagram. (I) Overlapping analysis of GEO and TCGA datasets.

Figure 2

Biological processes (BP) enrichment analysis and KEGG pathway analysis. (A-B) GO biological processes (BP) enrichment analysis and KEGG pathway analysis of upregulated DEGs. (C-D) GO biological processes (BP) enrichment analysis and KEGG pathway analysis of downregulated DEGs.

Figure 3

Construction of a 7-mRNA signature from the TCGA cohort. (A) 10-fold cross-validation for tuning parameter selection in the LASSO model. The dotted vertical lines are drawn at the optimal values by minimum criteria (lambda.min, left vertical dotted line) and 1-SE criteria (lambda.1se, right vertical dotted line). (B) LASSO model at optimal lambda value, 7 mRNAs with nonzero coefficients were selected.

Figure 4

Evaluation of the risk score formula for relapse in the TCGA cohort. (A) Waterfall plots for distribution of risk score and relapse status of individual patients. (B) Recurrence rate between the high- and low-risk at the indicated time. (C) The Kaplan-Meier survival curve of recurrence-free for patients between two different groups. (D) Time-dependent ROC curve at 1 year, 3 years, 5 years and more than 5 years. (E) Comparison of prognostic accuracy between the signature and single mRNAs. (F) Comparison of prognostic accuracy between the signature and clinical characteristics. P-values were calculated using the log-rank test. HR, hazard ratio; AUC, area under ROC curve; RFS, recurrence-free survival. ****, p <0.001.

Figure 5

Validation of the 7-mRNA signature for relapse prediction in the Ren Ji cohort. (A) Waterfall plots for distribution of risk score and relapse status of individual patients. (B) Recurrence rate between the high- and low-risk at the indicated time. (C) The Kaplan-Meier survival curve of recurrence-free for patients between two different groups. (D) Time-dependent ROC curve at 1 year, 3 years, 5 years and more than 5 years. (E) Comparison of prognostic accuracy between the signature and single mRNAs. (F) Comparison of prognostic accuracy between the signature and clinical characteristics. P-values were calculated using the log-rank test. HR, hazard ratio; AUC, area under ROC curve; RFS, recurrence-free survival. ****, p <0.001.

Figure 6

Kaplan-Meier survival analyses of the TCGA cohort, according to the 7-mRNA-based classifier stratified by clinicopathological characteristics. (A, B) Gender, (C, D) Age, (E, F) CA 199 levels, (G, H) Tumor size, (I, J) Pathologic stage, and (K, L) AJCC stage.

Figure 7

Kaplan-Meier survival analyses of the Ren Ji cohort, according to the 7-mRNA-based classifier stratified by clinicopathological characteristics. (A, B) Age, (C, D) Tumor thrombus, and (E, F) AJCC stage.