Diagnostic Efficiency of Quantitative Contrast-Enhanced Ultrasound Indicators for Discriminating Benign From Malignant Solid Thyroid Nodules.

Abstract

We aimed to determine the most efficient quantitative parameters to establish a contrast-enhanced ultrasound (US) assessment system for distinguishing between benign and malignant thyroid nodules. A total of 167 patients with thyroid solitary nodules had the diagnosis confirmed by surgery or fine-needle aspiration. Quantitative contrast-enhanced US indicators (time to peak, time from peak to one-half, ascend slope, descend slope, peak intensity, and area under the curve [AUC]) were gathered in nodule and perinodule areas. Univariate and multivariate logistic regression analyses were performed. Receiver operating characteristic curves were generated. Sensitivities, specificities, and positive and negative predictive values were calculated to identify the best cutoff value. The univariate logistic regression model showed that the peak intensity, ascend slope, descend slope, and AUC were significant indicators for discriminating benign from malignant nodules under contrast-enhanced US (P < .0001). For thyroid nodules, low peak intensity, ascend slope, and AUC and high descend slope values were significant indicators of malignancy. However, in perinodule areas, high peak intensity, ascend slope, and AUC and low descend slope values were significantly associated with malignancies. The cutoff values for the nodule peak intensity, ascend slope, descend slope, and AUC were 20.75, 0.91, -0.2, and 1818.23, respectively. The cutoff values for the ratios of the nodule versus perinodule peak intensity, ascend slope, descend slope, and AUC were 0.90, 0.95, 0.96, and 0.96. The nodule-to-perinodule peak intensity ratio showed the best diagnostic efficiency, with 80.41% sensitivity and 80.00% specificity. Quantitative contrast-enhanced US indicators help discriminate benign from malignant thyroid nodules. The nodule-to-perinodule peak intensity ratio showed the best diagnostic efficiency.