Immune scoring model based on immune cell infiltration to predict prognosis in diffuse large B‐cell lymphoma

Abstract Background Diffuse large B‐cell lymphoma (DLBCL) is genetically heterogeneous in both pathogenesis and clinical symptoms. Most studies on tumor prognosis have not fully considered the role of tumor‐infiltrating immune cells. This study focused on the role of tumor‐infiltrating immune cells in the prognosis of DLBCL. Methods The GSE10846 data set from the National Center for Biotechnology Information’s Gene Expression Omnibus was used as the training set, and the GSE53786 data set was used as the validation set. The proportion of immune cells in each sample was calculated with the CIBERSORT algorithm using R software. After 10 immune cells were screened out (activated memory CD4 positive T cells, follicular helper T cells, regulatory T cells, gamma‐delta T cells, activated natural killer cells, M0 macrophages, M2 macrophages, resting dendritic cells, and eosinophils) by univariate Cox analysis, Lasso regression and random forest sampling analyses were performed, the intersecting immune cells were selected for multifactor Cox analysis, and a predictive model was constructed combined with clinical information. Predictive performance was assessed using survival analysis and time‐dependent receiver operating characteristic curve analysis. Results In total, 539 samples were included in this study, and samples with p < .05 were retained using CIBERSORT. Univariate Cox analysis yielded 10 cell types that were associated with overall survival. Two kinds of immune cells were obtained by Lasso regression combined with the random forest method and were used to construct a prognostic model combined with clinical information. The reliability of the model was validated in two data sets. Conclusions The immune cell‐based prediction model constructed by the authors can effectively predict the prognostic outcome of patients with DLBCL, whereas nomogram plots can help clinicians assess the probability of long‐term survival.