Distinct Molecular Subtypes of Diffuse Large B Cell Lymphoma Patients Treated with Rituximab-CHOP Are Associated with Different Clinical Outcomes and Molecular Mechanisms.

Haifeng Yu,Shuiyun Han,Xi Chen,Tao Lei,Qinghua Lyu,Shuailing Peng

doi:10.1155/2021/5514726

Abstract

Objective Our purpose was to characterize distinct molecular subtypes of diffuse large B cell lymphoma (DLBCL) patients treated with rituximab-CHOP (R-CHOP). Methods Two gene expression datasets of R-CHOP-treated DLBCL patients were downloaded from GSE10846 (n = 233, training set) and GSE31312 (n = 470, validation set) datasets. Cluster analysis was presented via the ConsensusClusterPlus package in R. Using the limma package, differential expression analysis was utilized to identify feature genes. Kaplan-Meier survival analysis was presented to compare the differences in the prognosis between distinct molecular subtypes. Correlation between molecular subtypes and clinical features was analyzed. Based on the sets of highly expressed genes, biological functions were explored by gene set enrichment analysis (GSEA). Several feature genes were validated in the molecular subtypes via qRT-PCR and western blot. Results DLBCL samples were clustered into two molecular subtypes. Samples in subtype I displayed poorer overall survival time in the training set (p < 0.0001). Consistently, patients in subtype I had shorter overall survival (p = 0.0041) and progression-free survival time (p < 0.0001) than those in subtype II. Older age, higher stage, and higher international prognostic index (IPI) were found in subtype I. In subtype I, T cell activation, lymphocyte activation, and immune response were distinctly enriched, while cell adhesion, migration, and motility were significantly enriched in subtype II. T cell exhaustion-related genes including TIM3 (p < 0.001), PD-L1 (p < 0.0001), LAG3 (p < 0.0001), CD160 (p < 0.001), and CD244 (p < 0.001) were significantly highly expressed in subtype I than subtype II. Conclusion Two molecular subtypes were constructed in DLBCL, which were characterized by different clinical outcomes and molecular mechanisms. Our findings may offer a novel insight into risk stratification and prognosis prediction for DLBCL patients.

Highlights

Diffuse large B cell lymphoma (DLBCL) is the most common aggressive non-Hodgkin’s lymphoma globally [1]
According to different cell origins, DLBCL is divided into three subtypes including germinal center B cell-like (GCB; 41%) and activated B cell-like (ABC; 35%) subtypes and others based on gene expression profile, which has become the standard method of prognosis in clinical practice [2]
The GSE10846 dataset including 233 DLBCL patients treated with R-CHOP based on the GPL570 [HG-U133_Plus_2] Affymetrix Human Genome U133 Plus 2.0 Array platform was used as the training set [17, 18]

Summary

Introduction

Diffuse large B cell lymphoma (DLBCL) is the most common aggressive non-Hodgkin’s lymphoma globally [1]. According to different cell origins, DLBCL is divided into three subtypes including germinal center B cell-like (GCB; 41%) and activated B cell-like (ABC; 35%) subtypes and others based on gene expression profile, which has become the standard method of prognosis in clinical practice [2]. Its prognostic effectiveness of this classification has not been uniformly proven due to the heterogeneity of classification structures [3]. The international prognostic index (IPI) is an effective clinical tool for predicting risk stratification and prognosis. It cannot guide personalized therapy [4]. The rituximab, cyclophosphamide, doxorubicin, and prednisone (R-CHOP) therapy is currently proven to be one of the most effective treatment regimen for most DLBCL subtypes. The 5-year overall survival rate of patients receiving first-line treatment is 60-

Methods

Results

Discussion

Conclusion