Axillary Sentinel Lymph Nodes in Breast Cancer: Quantitative Evaluation at Dual-Energy CT

Purpose: To evaluate the diagnostic performance of quantitative parameters derived from dual-energy CT for the preoperative diagnosis of metastatic sentinel lymph nodes (SLNs) in participants with breast cancer. Materials and Methods: For this prospective study, dual-phase contrast agent-enhanced CT was performed in female participants with breast cancer from June 2015 to December 2017. Quantitative dual-energy CT parameters and morphologic parameters were compared between metastatic and nonmetastatic SLNs. The quantitative parameters were fitted to univariable and multivariable logistic regression models. The diagnostic role of morphologic and quantitative parameters was analyzed by receiver operating characteristic curves and compared by using the McNemar test. Results: This study included 193 female participants (mean age, 47.6 years +/- 10.1; age range, 22-79 years). Quantitative dual-energy CT parameters including slope of the spectral Hounsfield unit curve (lambda(Hu)) measured at both arterial and venous phases, normalized iodine concentration at both arterial and venous phase, and normalized effective atomic number at the venous phase were higher in metastatic than in nonmetastatic SLNs (P value range, <= .001 to .031). Univariable and multivariable logistic regression analyses showed that venous phase lambda(Hu) (in Hounsfield units per kiloelectron-volt) was the best single parameter for the detection of metastatic SLNs. The accuracy of the venous phase lambda(Hu) for detecting metastatic SLNs was 90.5% on a per-lymph node basis and 87.0% on a per-patient basis. The accuracy and specificity at venous phase lambda(Hu) was higher than their counterparts in the morphologic parameters (P < .001). Conclusion: Dual-energy CT is a complementary means for the preoperative identification of sentinel lymph nodes metastases in participants with breast cancer. (c) RSNA, 2018