Diagnostic accuracy of ultrasonography, breast MRI, and 18F-FDG PET/CT for axillary lymph node metastasis in breast cancer patients.

e12023 Background: The aim of this study was to evaluate the diagnostic performance of ultrasonography (US), breast magnetic resonance imaging (MRI), and 18F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) for axillary lymph node metastasis in breast cancer patients. Methods: Between 2012 and 2013, 180 breast cancer patients received curative breast cancer surgery. Preoperative ultrasonography, breast MRI, and 18F-FDG PET/CT were performed. Complete axillary lymph node dissection was performed when a metastatic lymph node was detected preoperatively or intra-operatively. Results: Among 180 patients, 53 patients (31.1%) had axillary metastatic lymph node. The sensitivity of US, MRI, and PET/CT for the detection of axillary lymph node metastasis was 55% for US, 72% for MRI, and 44%, respectively. The specificity of US, MRI, and PET/CT for the detection of axillary lymph node metastasis was 88%, 78%, and 92%, respectively. The positive predictive value and negative predictive value of US, MRI, and PET/CT for the detection of axillary lymph node metastasis were 67%, 60%, and 71%, respectively, and 81%, 86%, and 79%, respectively. In multivariate logistic regression analysis, pathologic lymph node metastasis was associated with positive finding for axillary metastasis of US, MRI, and PET/CT, independently (P = 0.028, P = 0.03, and P = 0.017, respectively). With the combining three modalities, scoring system for detecting axillary lymph node metastasis was developed and the AUC for positive lymph node metastasis was 0.811 (95% CI, 0.737 - 0.886). Conclusions: The diagnostic accuracy of US, MRI, and 18F-FDG PET/CT were similar. However, combining three imaging modalities could increase the diagnostic accuracy of axillary lymph node metastasis, preoperatively.