Abstract
Background: Significant difference existed in survival outcomes of primary central nervous system lymphoma (PCNSL) patients. The aim of this study was to establish a prognostic nomogram to access personalized survival rate of PCNSL patients.
Methods: 2947 patients were extracted from the Surveillance, Epidemiology, and End Results database as the training cohort. 177 patients diagnosed from 2006-2020 in Sun Yat-Sen University cancer center were identified as the external validation cohort. Cox proportional hazards model was performed to identify independent factors for establishing the nomogram. Concordance index (C-index) and calibration curves were used to test the accuracy of the model.
Results: Age, tumor site, lymphoma subtype, surgery and chemotherapy were significant predictors included in the nomogram. C-index in the training and validation cohort were 0.701 and 0.653, respectively. Calibration curves showed favorable consistency of nomogram-predicted and actual probability of overall survival (OS). The decision curve analysis (DCA) curves exhibited great clinical application. We published a web calculator for convenient clinical practice and established a risk classification system that significantly distinguished patients into four groups. Results of the primary cohort were well confirmed in the validation cohort.
Conclusion: We developed a nomogram and a web calculator of OS prediction for PCNSL, which may assist in medical consultation, treatment decision making and clinical study design.
No relevant conflicts of interest to declare.