Elastosonography (ES) is a newly developed method that is used for the differential diagnosis of benign and malignant thyroid nodules. In different studies, ES scoring has been compared with histopathological findings, and sensitivity and specificity of the scoring were calculated. In this study, it determines the strain index (SI) as well as the ES to score thyroid nodules, and establishes the role for these parameters in the differential diagnosis of thyroid nodules using histopathological analysis as a reference standard. Real-time ES in transverse axis (TA) and longitudinal axis (LA) was performed in 391 nodules of 292 patients. ES scoring was made for all the nodules. SI in TA and LA was calculated for four times in each nodule and mean values were determined. The results were compared with final histopathological diagnoses. In histopathological examinations, 125 (31.97%) of 391 nodules were malignant and 266 (68.03%) were benign. Of these histopathologically benign nodules, 189 (%71.05) were also probably benign according to elastosonographic scoring (scores of 1, 2, or 3), while 77 (28.95%) were probably malignant (scores of 4 or 5). Among 125 histopathologically malignant nodules, 52 (41.60%) were probably benign and 73 (58.40%) were probably malignant according to elastosonographic scoring. There was a significant relation between scoring and histopathological findings (χ(2)=36.513; P<0.001). Accordingly, sensitivity and specificity of ES scoring were 58.4 and 71.0%, respectively. ROC analysis value obtained for strain ratios in LA (AUC: 75.5%; P<0.001) had a higher significance compared to ROC analysis value obtained for strain ratios in TA (AUC: 66.0%). Thus, ROC analysis evaluation was applied only for SI in LA. The optimal SI cut-off value in LA for all the nodules was found to be 16.709 (sensitivity: 73.4%, specificity: 70.0%) (AUC: 75.4±0.03%; 70.2-80.5%). SI cut-off value corresponding to 90% sensitivity in this axis was 4.516 (specificity: 35.7%). Sensitivity and specificity of SI values that were determined according to morphological features of nodules in gray-scale ultrasonography were higher. For hypoechoic nodules with microcalcifications and without a halo, SI cut-off value, sensitivity, and specificity were 17.020, 84.3, and 81.1%, respectively. Our study is the first clinical-wide series study that measured, used, and compared the ES scoring and SI cut-off values for the differential diagnosis of benign and malignant thyroid nodules. This study indicates that measurement of SI with ES as a noninvasive procedure may be used as an adjunctive method to the conventional methods for the differential diagnosis of thyroid nodules.
Read full abstract