Development and validation of nomograms to predict prognosis of Burkitt's lymphoma: a SEER-based study

Abstract

Abstract Background: The purpose of this study was to establish two practical and valid nomograms to predict overall survival (OS) and cancer-specific survival (CSS) in patients with Burkitt's lymphoma. Methods: A total of 3972 patients with Burkitt's lymphoma diagnosed in 2000-2015 were screened from the SEER database and randomized into training cohorts (N=2780) and validation cohorts (N=1192). Univariate and multivariate Cox regression analyses were performed to select independent risk factors affecting prognosis, followed by the construction of nomograms for OS and CSS. The reliability of the nomogram was validated with C-index and calibration curve. DCA plots were used to compare the clinical value of the nomogram with Ann Arbor Stage staging. In addition, patients were divided into high-risk and low-risk groups according to the mean of their scores as a cut-off, and survival was compared using the Kaplan-Meier method. Results: According to the results of multivariate Cox regression analysis, the common independent prognostic factors affecting OS and CSS were age, race, marital status, year of diagnosis, primary site, stage, and chemotherapy. Based on these variables, two prediction models were constructed. In the training cohort, the C-index of the nomogram for OS was 0.741. DCA plots indicated that our nomogram had more clinical net benefits than the Ann Arbor staging system. Conclusion: A comprehensive assessment of the incidence and survival prognosis of Burkitt's lymphoma was conducted using a large database, and two nomograms were created to predict patient risk and prognostic factors, thereby guiding clinicians in individualized clinical practice.