Identification and validation of M2 macrophage-related differential genes in DLBCL

Abstract DLBCL is the most prevalent non-Hodgkin lymphoma, with 30-40% of patients are refractory or relapse after conventional immunotherapy. Growing evidence indicate that alterations in tumor microenvironment (TME) play a pivotal role in DLBCL progression, including polarization of macrophages to M2 phenotype, fostering tumor immunosuppressive status. Therefore, exploring molecular TME biomarkers in DLBCL is essential. In this study, we analyzed association between M2 macrophages with gene expression by CIBERSORT algorithm in 420 samples from the GSE10846 dataset. GO and KEGG enrichment analyses revealed these differential genes involving in cytokine-mediated signaling pathways and cytokine activity. Univariate COX regression analysis determined 28 prognostic-related differential genes, with Hypoxia-inducible factor 1-alpha (HIF1a) being positively correlated. A LASSO logistic regression model was then established with superior sensitivity and specificity. High-risk patients exhibited worse outcomes and GSEA analysis identified several pathways associated with the low-risk group, involving cytokine-cytokine receptor interaction. Moreover, risk scores were significantly correlated with immune checkpoint genes. RT-qPCR and immunostaining analyses on clinical sample and co-culture experiments further confirmed the downregulation of HIF1a in samples with higher content of M2 macrophages and adverse prognosis. These findings highlight the importance of M2 macrophages and MDGs in DLBCL prognosis and suggest the potential clinical utility in predicting patient outcomes.