Toward individualized prognosis in primary breast diffuse large B-cell lymphoma: incidence trends and a predictive...

Primary breast diffuse large B-cell lymphoma (PB-DLBCL) is a rare malignancy with limited population-level prognostic data. We aimed to characterize incidence trends, estimate dynamic conditional survival (CS), and develop a prognostic model for overall survival (OS) using Surveillance, Epidemiology, and End Results (SEER) data.

We identified 792 patients with PB-DLBCL from the SEER database (2000-2021). Age-adjusted incidence rates were calculated to assess trends over time. 10-year CS probabilities were estimated to evaluate dynamic survival patterns. The cohort was randomly divided into training (n=554) and validation (n=238) sets. Best subset regression (BSR) and stepwise backward Cox regression were applied to identify the most informative prognostic variables, which were incorporated into a CS-nomogram. Model performance was assessed using calibration plots, time-dependent receiver operating characteristic curves, concordance index (C-index), and decision curve analysis (DCA).

The age-adjusted incidence of PB-DLBCL remained stable over two decades, with no significant trend [annual percent change (APC) =-0.16%]. CS analysis revealed that patients who survived the first post-diagnosis year had progressively improved long-term survival probabilities (10-year CS: 49% at diagnosis vs. 57% after 1-year survival; 73% for 5-year survivors). BSR with stepwise backward Cox regression identified age, Ann Arbor stage, and chemotherapy as the optimal prognostic factors. The CS-nomogram incorporating these variables effectively stratified patients into high- and low-risk groups, demonstrating robust discrimination and calibration in both training and validation cohorts. DCA confirmed the model's clinical utility in guiding individualized patient management.

PB-DLBCL showed stable incidence and dynamic improvement in survival among patients who surpassed early post-diagnosis milestones. Age, stage, and chemotherapy were key prognostic determinants. The CS-nomogram provided a practical, individualized tool for dynamic survival prediction and risk stratification, supporting personalized clinical decision-making in this rare lymphoma subtype.