Background and aims: The aim of the present study was to evaluate real-time elastography (RTE) and contrast-enhanced ultrasound with Sonovue (CEUS) for the differentiation of benign and malignant thyroid nodules using cytologic/histologic analysis as reference method. While first studies have reported excellent results of RTE for the differentiation of thyroid nodules, data on CEUS of the thyroid are rare.
Methods: RTE was classified as: score 1=soft, 2=predominantly soft, 3=predominantly hard, 4=hard nodule. Power Doppler perfusion (PD) was classified as: pattern 1=no/minimal perfusion; 2=perinodular; 3=little intranodular; 4=hyperperfusion. Fourty-six nodules of 45 patients were included in the present study. All examined nodules were ≧1 cm in size and non-functioning or hypo-functioning on radionuclide scanning. FNA was performed in 41 nodules, 13 of which received an operation due to a suspicious or non-diagnostic result. Five patients received primary operation.
Results: Five patients had to be excluded due to non-diagnostic FNA. Therefore, 41 nodules in 40 patients were available for analysis. FNA revealed cancerous tissue in 4 and suspicious tissue in one patient. All 5 patients were operated and papillary carcinoma was found in 4, and follicular carcinoma in one patient, respectively. In all other patients cytology/histology revealed benign tissue. A significant correlation between cytology/histology and ultrasound measurement was found only for RTE (r=0.32; P<0.05). When using PD pattern 3&4 for the diagnosis of malignant nodules sensitivity, specificity, PPV, and NPV were 80, 47, 17, and 94%, respectively. When using elastography score 3&4 for the diagnosis of malignant nodules sensitivity, specificity, PPV, and NPV were 80, 75, 31, and 96%, respectively. And when using PD and RTE criteria together sensitivity, specificity, PPV, and NPV increased to 80, 83, 29, and 97%, respectively. The only malignant nodules which both methods missed, was the follicular carcinoma. Sensitivity and NPV for the diagnosis of papillary carcinoma was 100%. No specific CEUS pattern could be identified to differentiate between benign and malignant nodules.
Discussion: RTE can be used with high NPV in the work-up of thyroid nodules to exclude papillary thyroid cancer. However, follicular carcinoma remains a challenging problem. CEUS does not improve the characterization of thyroid nodules.