Predicting sentinel lymph node metastasis in breast cancer: a study based on the SEER database.

BACKGROUND

Sentinel lymph node biopsy (SLNB), a standard surgical procedure for clinically axillary-negative breast cancer patients, significantly reduces complications compared with axillary lymph node dissection, but it is still a relatively invasive procedure with some complications, affecting patient's quality of life. To identify patients who might benefit from avoiding SLNB, this study aimed to develop a nomogram for predicting sentinel lymph node metastasis (SLNM) in breast cancer patients using the SEER database.

METHODS

We identified breast cancer patients whose 1-5 lymph nodes were examined in the SEER database as those who underwent SLNB. Patients were randomly assigned to the training and validation cohorts at a 3:1 ratio. Univariate and multivariate logistic regression were used to evaluate the relationships between SLNM and patients' clinicopathological characteristics. A nomogram was constructed, and its performance was validated via ROC curves, calibration curves, and decision curve analysis.

RESULTS

Age, race, primary site, T stage, M stage, histological grade, pathological type, estrogen receptor status, and progesterone receptor status are independent predictive factors for SLNM in patients with breast cancer. We successfully developed a predictive nomogram for sentinel lymph node status, with AUC values of 0.711 and 0.700 for the training and validation cohorts, respectively.

CONCLUSION

Our study successfully established an SLNM nomogram that provides richer predictive information. The model exhibits good clinical efficacy and serves as a reference value for populations potentially exempt from SLNB.