Quantitative evaluation of axillary lymph nodes in breast cancer on dual-phase dual-energy CT by precise match with pathology.

Abstract

BackgroundIdentification of axillary lymph node (aLN) metastasis in breast cancer (BC) is important. Dual-energy computed tomography (DECT) is a promising innovation in the field of CT. However, its role in evaluating aLNs remains unclear.PurposeTo investigate the diagnostic performance of DECT in evaluating aLN metastasis in BC patients.Material and MethodsData were prospectively collected from treatment-naïve BC patients who underwent DECT for staging, ultrasound-guided biopsy for suspicious aLNs, and placement of tissue marker in the pathology-positive aLNs. Further cross-sectional imaging was performed preoperatively to locate the marker-labeled LN and help to identify the pathologically proven LN on DECT. Maximal short diameter (MSD) and 13 DECT parameters were measured on metastatic aLNs and contralateral normal aLNs. The univariate, least absolute shrinkage and selection operator and multivariable logistic regression were performed to find independent parameters for predicting metastasis. The diagnostic performance was assessed using receiver operating characteristics (ROC) analysis.ResultsA total of 76 axillary LNs (38 metastasized, 38 normal) from 38 patients were finally included. All DECT parameters showed significant difference between metastatic and normal LNs (all P < 0.001). Arterial enhancement fraction (AEF) and MSD were independent predictors of metastasis (P = 0.010 and 0.014, respectively). The area under the ROC curve (AUC) of AEF was the highest (0.967). The combined AUC of AEF and MSD was significantly higher than that of MSD alone (0.994 vs. 0.943; P = 0.025).ConclusionDECT is a promising tool for preoperative evaluation of aLNs in BC patients, with MSD and AEF having the best diagnostic performance.