Evaluation of lymph node status and prognostic modeling in elderly breast cancer patients using the SEER database

The incidence of breast cancer remains high among elderly patients, who also exhibit a higher cancer-specific mortality rate than their younger counterparts due to multiple factors. However, prognostic models tailored specifically to elderly breast cancer patients remain scarce. This study aims to evaluate the prognostic impact of various lymph node staging systems on elderly breast cancer patients using a competing risks model. Data for 104,952 elderly breast cancer patients (aged >70 years) diagnosed between 2004 and 2015 were collected from the SEER database. A competing risks model was employed to assess the prognostic impact of several lymph node staging systems, including N stage, positive lymph node number (PLNNs), lymph node ratio (LNR), log odds of positive lymph nodes (LODDS), and the log odds of negative lymph nodes/T stage ratio (LONT). Subgroup analyses were performed according to major clinicopathological variables (age, sex, race, marital status), followed by competing risks analyses of lymph node indicators. Nomograms were then constructed to establish prognostic models for individualized survival prediction. Our findings indicate that PLNNs, LNR, LODDS, and LONT are all independent prognostic factors in both univariate and multivariate competing risks models (P<0.001), with consistent results across various subgroups. The nomograms developed from the training set demonstrated robust predictive performance in the validation set. All four models yielded AUC values greater than 0.8, confirming robust predictive performance across all time points.The C-index for LNR showed the highest performance, with an optimal cutoff value of 0.3 for LNR and 2 for PLNNs. Lymph node status is a critical factor in assessing the prognosis of breast cancer. We conducted a comprehensive comparative analysis of lymph node status specifically in elderly breast cancer patients, thereby enhancing prognostic models tailored for this population. Our findings may contribute to improved prognostic management in elderly breast cancer patients.