Establishment of a prognostic model for diffuse large B-cell lymphoma based on mRNA omics

Objective To screen the prognosis-related genes for diffuse large B-cell lymphoma (DLBCL) based on mRNA omics data and establish a prognostic model.Methods The mRNA-omics data and prognostic data of 234 DLBCL patients in the NCICCR-DLBCL dataset were downloaded from the TCGA database as a training set, while the corresponding data of 223 DLBCL patients in GSE87371 dataset were downloaded from GEO database as a validation set. A prognostic model was established through Cox regression analysis, Kaplan-Meier analysis and LASSO analysis, and its accuracy was validated by Kaplan-Meier survival curve and ROC curve. Finally, the expression of CD70 protein was detected in 117 DLBCL tissues by immunohistochemistry.Results The prognostic model was established, which contained 14 significantly independent mRNA marker (ANO3, ZNF93, RAD9B, PROC, ZSCAN20, RBBP9, SLC18B1, CXorf21, MUC12, CYP2U1, VSIG4, CCDC178, CD70, and GHRH). The high risk group showed remarkably shorter median survival time than the low risk group (P＜0.000 1), and the area under the ROC for 1-year, 3-year and 5-year-survival were 0.92, 0.90 and 0.91, respectively. There were statistical differences between CD70 positive and negative patients in Kaplan-Meier survival curve (P=0.009 4), with shorter survival in the positive group.Conclusions The DLBCL prognosis prediction model based on mRNA expression is successfully established, which provides a feasible solution for the application of mRNA transcriptomics for predicing DLBCL prognosis. It is confirmed in clinical samples that high CD70 expression is closely related to the prognosis of DLBCL, which can be used as one of the indicators for clinical prognosis evaluation.