Development and validation of a nomogram for false-negative results in fine-needle aspiration of axillary lymph nodes in breast cancer.

[OBJECTIVE] To construct and validate a nomogram for predicting false-negative results of axillary lymph node (ALN) fine needle aspiration (FNA) in breast cancer (BC).

[METHODS] Using the surgical pathological results of axillary lymph nodes (ALNs) in BC patients as the gold standard, we retrospectively analyzed the clinical, pathological, and ultrasonographic characteristics of patients with false-negative lymph FNA results and identified predictive factors. Based on the independent predictors screened, a nomogram prediction model was constructed and validated using the Hosmer-Lemeshow test, receiver operating characteristic (ROC) curve, and decision curve analysis (DCA). The calibration curve was drawn by bootstrap method for internal verification.

[RESULTS] Univariate analysis revealed that the following five factors were statistically significant predictors of false-negative results (P < 0.05): ultrasound features of ALN shape, corticomedullary boundary, hilum status, histological type of the primary tumor from core needle biopsy (CNB), and short-axis diameter of ALNs on ultrasound. Multivariate analysis identified three independent predictors of false-negative FNA results (P < 0.05): ultrasound features of ALN shape, corticomedullary boundary, and histological type of the primary tumor from CNB. A nomogram prediction model was successfully developed based on these independent predictors. The Hosmer-Lemeshow test yielded a P-value of 1, the area under the ROC curve (AUC) was 0.782, and the DCA threshold range for the nomogram was 0.03-0.95. The calibration curve was drawn by bootstrap method for internal verification, and the conclusion was that N = 247, Mean absolute error = 0.012, Mean squared error = 0.00035.

[CONCLUSION] A nomogram model was constructed to predict false-negative FNA results in ALNs of BC patients, demonstrating good predictive performance.