A prognostic nomogram and risk classification system of elderly patients with extraosseous plasmacytoma: a SEER database analysis

Background The survival trends and prognostic factors of patients with extraosseous plasmacytoma (EOP) or extramedullary plasmacytoma (EMP) have not been reported in recent years. The objective of this study was to develop a novel nomogram and risk stratification system for predicting the overall survival (OS) of elderly patients with EOP based on the Surveillance, Epidemiology, and End Results (SEER) database. Methods The demographic characteristics of 900 patients aged 60 years and above, diagnosed with EOP between 2000 and 2019, were extracted from the SEER database. The patient population was randomly divided into a training cohort and an internal validation cohort in a ratio of 7:3. Univariate and multivariate Cox regression analyses were conducted to identify independent predictors of prognosis in elderly EOP patients, followed by developing a nomogram for prognostic assessment. The performance of the model was evaluated through receiver-operating characteristic (ROC) curves, C-index, calibration curves for calibration accuracy assessment, and decision curve analysis (DCA) to assess its clinical utility. All elderly EOP patients were stratified into three risk subgroups by cutoff value utilizing X-tile software based on their total OS scores for comparative analysis purposes. Kaplan–Meier (K–M) survival curve analysis was employed to validate any observed differences in OS among these three risk groups. Results Six factors including age, year of diagnosis, marital status, primary site, surgery, and prior tumor history were identified to be independently predictive of the OS of elderly patients with EOP, and these predictors were included in the construction of the nomogram. The 1-, 3-, and 5-year area under the curves (AUCs) for OS were 0.717, 0.754, and 0.734 in the training cohort and 0.740, 0.730, and 0.765 in the validation cohort, respectively. The C-index values in the two cohorts were 0.695 and 0.690. The calibration curves and DCA exhibit commendable consistency and validity, respectively, thereby demonstrating their robust performance. The training set was stratified into low-, medium-, and high-risk subgroups based on the optimal cutoff points (167.8 and 264.8) identified. The K–M curve and cumulative risk curve exhibited statistically significant disparities in survival rates among the groups. Conclusions We developed a nomogram and risk classification system, which can serve as an intuitive and effective tool for clinicians to enhance the prediction of OS in elderly EOP patients, thereby facilitating the formulation of more rational and personalized treatment strategies.