A prognostic gene model of immune cell infiltration in diffuse large B-cell lymphoma

PeerJ

article

Hao Zhou¹  Chang Zheng¹  De-Sheng Huang¹ 

[1] Department of Epidemiology, School of Public Health, China Medical University;Department of Impression Evidence Examination Technology, Criminal Investigation Police University of China;Key Laboratory of Cancer Etiology and Prevention ,(China Medical University), Liaoning Provincial Department of Education

关键词: Diffuse large B-cell lymphoma;    Immune cells;    Immunoscore;    Tumor microenvironment;    Natural killer cell;    CIBERSORT algorithm;    Lasso regression;   

DOI  :  10.7717/peerj.9658

学科分类：社会科学、人文和艺术（综合）

来源: Inra

【 摘 要 】

BackgroundImmune cells in the tumor microenvironment are an important prognostic indicator in diffuse large B-cell lymphoma (DLBCL). However, information on the heterogeneity and risk stratification of these cells is limited. We sought to develop a novel immune model to evaluate the prognostic intra-tumoral immune landscape of patients with DLBCL.MethodsThe ESTIMATE and CIBERSORT algorithms were used to estimate the numbers of 22 infiltrating immune cells based on the gene expression profiles of 229 patients with DLBCL who were recruited from a public database. The least absolute shrinkage and selection operator (Lasso) penalized regression analyses and nomogram model were used to construct and evaluate the prognostic immunoscore (PIS) model for overall survival prediction. An immune gene prognostic score (IGPS) was generated by Gene Set Enrichment Analysis (GSEA) and Cox regression analysis was and validated in an independent NCBI GEO dataset (GSE10846).ResultsA higher proportion of activated natural killer cells was associated with a poor outcome. A total of five immune cells were selected in the Lasso model and DLBCL patients with high PIS showed a poor prognosis (hazard ratio (HR) 2.16; 95% CI [1.33–3.50]; P = 0.002). Differences in immunoscores and their related outcomes were attributed to eight specific immune genes involved in the cytokine–cytokine receptor interaction and chemokine signaling pathways. The IGPS based on a weighted formula of eight genes is an independent prognostic factor (HR: 2.14, 95% CI [1.40–3.28]), with high specificity and sensitivity in the validation dataset.ConclusionsOur findings showed that a PIS model based on immune cells is associated with the prognosis of DLBCL. We developed a novel immune-related gene-signature model associated with the PIS model and enhanced the prognostic functionality for the prediction of overall survival in patients with DLBCL.

【 授权许可】

CC BY