Background: Diffuse large B cell lymphoma (DLBCL) is a heterogeneous disease with poor outcome. Based on the cell of origin (COO), it is classified into germinal center B (GCB) and non-GCB subtypes. The IL17A and tumor associated macrophages (TAM) in the tumor microenvironment have been involved in modulation of tumor immunity in colon cancer and non Hodgkin lymphoma. EZH2 has been associated with poor prognosis of several tumors. The aim of the current work is evaluation of the impact of IL 17A, CD68, EZH2 proteins expression and cell of origin classification in predicting response to R-CHOP as well as patient outcome in DLBCL. Materials and methods: IL17A, CD68, EZH2 proteins were evaluated by immunohistochemistry in 60 cases of diffuse large B cell lymphoma. Also Tumors were classified into GCB & non-GCB types based on Hans algorithm using CD10, BCL6 & MUM1 immunohistochemistry. The pattern of proteins expression was correlated to patients' characteristics, response to R-CHOP and Patient survival. Results: Overexpression of IL17A protein was significantly associated with involvement of more than one extranodal site, high LDH level, PS ≥ 2 as well as high IPI (p = < 0.001, 0.002, < 0.001, 0.004, < 0.001 respectively). Both EZH2 protein overexpression and non-GCB type were significantly associated with advanced tumor stage, higher LDH level, PS as well as IPI (p < 0.001). However, CD68 immunopositivity was not related to any of the studied parameters (p = > 0.05). We found that Cases with IL17A or EZH2 overexpression or non-GCB phenotype were highly resistant to R-CHOP (p < 0.001). Moreover they were associated with short-term DFS, PFS, and OS (p < 0.05). CD86 immunoreactivity was significantly associated with longer DFS (p = 0.016), but not PFS, OS or response to therapy (p ˃ 0.05). Conclusion: IL17A & EZH2 proteins and COO classification could be used for risk stratification of newly diagnosed DLBCL patients and predicting their prognosis and response to therapy. Also, a panel of markers should be used for examining TAM instead of CD68 alone.