A predictive model based on immune related genes for DLBCL

Anti-tumor immunity is the front line of human response to malignancy, which may shad the light to early diagnosis of DLBCL. The introduction of immune related genes will bring new insight in the establishment of a predictive model to facilitate the diagnosis of DLBCL and guide its therapy. Hence, we established an immune related genes based risk model to predict the survival and progression of DLBCL. First, we identify immune related genes in DLBCL via GeneCards. With these genes, we conducted LASSO regression to select those genes with significant contribution to DLBCL and established a risk model to generate risk score. Validation of this risk model in internal test dataset and additional external validation datasets confirm the robust performance of this model. The risk score was also found to correlated with advanced stages and age over 60. Later, a nomogram combining risk score with other common clinical index (age, gender, stage) was established to comprehensively evaluate the survival probability of patients with DLBCL. To guide the treatment, we also found four novel second-line chemotherapies can be used to treat patients with different risk scores. Overall, this novel model can be utilized in clinical practices and guide the treatment.