A Visual Nomogram Survival Prediction Model in Acquired Immune Deficiency Syndrome (AIDS)-Related Diffuse Large B-Cell Lymphoma

Abstract

Estimating the prognosis of people with newly diagnosed AIDS-related diffuse large B-cell lymphoma (AR-DLBCL) is challenging. We did a prospective, multicenter cohort study using data from 306 consecutive subjects, including training cohort (N = 215) and external validation cohorts (N = 91), to develop and validate a visual nomogram, termed the ARDPI model. Seven co-variates were independently correlated with survival, including age, LMR, CD5, blood EBV-DNA copy number, CD4/CD8, CNS involvement, and anti-HIV therapy (ART), were used to develop the ARDPI model. AUROCs of the model for 1-, 3-, and 5-year survivals were 0.80 (95% CI: 0.72, 0.88), 0.78 (0.69, 0.87), and 0.77 (0.63, 0.91) in the training cohort and 0.85 (0.75, 0.95), 0.80 (0.66, 0.94), and 0.79 (0.61, 0.99) in the external validation cohort. The prediction accuracy of the ARDPI model was better compared with the IPI and NCCN-IPI models. Using the ARDPI model, we identified three risk cohorts with 3-year survivals of 88% (79, 98%), 35% (23, 54%), and 23% (12, 45%) in the training cohort (p < 0.001) and 93% (80, 100%), 46% (27, 78%), and 17% (5, 47%) in the external validation cohort (p < 0.001). The ARDPI accurately predicts the survival of newly diagnosed persons with AR-DLBCL and has clinical benefits. Accuracy is better compared with the IPI and NCCN-IPI prediction models. We also developed a web server to facilitate using our model.