BMC Med Imaging. 2026 Jan 28. doi: 10.1186/s12880-026-02181-x. Online ahead of print.
ABSTRACT
BACKGROUND: We aimed to evaluate the predictive value of preoperative Node-RADS determination at imaging for axillary lymph node (ALN) involvement in cases of breast cancer.
MATERIALS AND METHODS: Node-RADS was determined in all cases using preoperative breast magnetic resonance imaging (MRI) by two radiologists blinded to pathologic results. The ROC curves and AUCs depicted the overall diagnostic performance of the Node-RADS score for lymph node involvement. In addition, the presence of a flare-like perinodal signal was also evaluated as an imaging feature suggestive of extranodal extension. Size and morphological parameters were assessed separately for their association with metastatic involvement.
RESULTS: Both readers demonstrated high diagnostic accuracy in predicting ALN metastasis using the Node-RADS system, with the best diagnostic performance observed at a cutoff value above Node-RADS 2. In logistic regression analysis, heterogeneous internal texture of the lymph node and flare sign were also found to be statistically significantly associated with invasion (p < 0.05). The presence of a fatty hilum was a statistically significant predictor (p < 0.001) associated with a markedly lower likelihood of lymph node metastasis (OR = 0.019, 95% CI = 0.002-0.153).
CONCLUSION: The Node-RADS scoring system demonstrates high diagnostic reliability and reproducibility in the evaluation of axillary lymph nodes in breast cancer. However, the size criterion for ALN assessment may need to be re-evaluated, and the inclusion of size or extranodal extension parameters in the scoring system should be reconsidered.
PMID:41606527 | DOI:10.1186/s12880-026-02181-x
