Conditional survival patterns and individualized prognostic prediction in solitary plasmacytoma of the spine

Background Solitary plasmacytoma of the spine (SPS) is a rare plasma cell neoplasm with limited survival data and prognostic tools. Conditional survival (CS) analysis provides time-dependent survival estimates that may better reflect the evolving prognosis of cancer patients. This study aimed to explore prognostic factors and construct a CS-based nomogram for SPS. Methods We extracted SPS cases from the SEER database. CS analysis was performed to evaluate dynamic survival probabilities. Multivariate Cox regression identified factors significantly associated with overall survival (OS). To optimize variable selection, we integrated least absolute shrinkage and selection operator (LASSO) regression and best subset regression (BSR) with stepwise backward Cox analysis. The final variables were used to construct a CS-based nomogram, which was validated for discrimination, calibration, and risk stratification ability. Results A total of 1660 patients were included. CS estimates showed that survival probabilities improved significantly with each additional year survived after diagnosis. Age, radiotherapy, chemotherapy, household income, and marital status were independent predictors of OS, whereas surgical resection showed no significant association with prognosis. Notably, chemotherapy use was associated with poorer survival. The combined BSR–LASSO–stepwise Cox approach selected age, radiotherapy, and marital status for the final model. The nomogram demonstrated strong predictive performance and effectively stratified patients into distinct risk groups. Conclusion We developed and validated the first CS-based prognostic nomogram for SPS, incorporating age, radiotherapy, and marital status. This model demonstrated promising predictive performance in internal validation and may serve as a foundation for future external validation and clinical application in this rare malignancy.