Identification and validation of an immunological microenvironment signature and prediction model for epstein-barr virus positive lymphoma: Implications for immunotherapy

BackgroundEpstein-Barr virus (EBV) is considered a carcinogenic virus, which is associated with high risk for poor prognosis in lymphoma patients, and there has been especially no satisfying and effective treatment for EBV+ lymphoma. We aimed to identify the immunological microenvironment molecular signatures which lead to the poor prognosis of EBV+ lymphoma patients. MethodsDifferential genes were screened with microarray data from the GEO database (GSE38885, GSE34143 and GSE13996). The data of lymphoid neoplasm diffuse large B-cell lymphoma (DLBC) from the TCGA database and GSE4475 were used to identify the prognostic genes. The data of GSE38885, GSE34143, GSE132929, GSE58445 and GSE13996 were used to eluate the immune cell infiltration. Formalin-fixed, paraffin-embedded tissue was collected for Real Time Quantitative PCR from 30 clinical samples, including 15 EBV+ and 15 EBV- lymphoma patients. ResultsFour differential genes between EBV+ and EBV- lymphoma patients were screened out with the significance of the survival and prognosis of lymphoma, including CHIT1, SIGLEC15, PLA2G2D and TMEM163. Using CIBERSORT to evaluate immune cell infiltration, we found the infiltration level of macrophages was significantly different between EBV+ and EBV- groups and was closely related to different genes. Preliminary clinical specimen verification identified that the expression levels of CHIT1 and TMEM163 were different between EBV+ and EBV- groups. ConclusionsOur data suggest that differences in expression levels of CHIT1 and TMEM163 and macrophage infiltration levels may be important drivers of poor prognosis of EBV+ lymphoma patients. These hub genes may provide new insights into the prognosis and therapeutic target for EBV+ lymphoma.