We aimed to study the validity of six published ultrasound criteria for risk stratification of thyroid nodules in the former severely iodine deficient population of Austria.
Retrospective, single centre, observer blinded study design. All patients with a history of thyroidectomy due to nodules seen in the centre between 2004 and 2014 with preoperative in-house sonography and documented postoperative histology were analyzed (n = 195). A board of five experienced thyroidologists evaluated the images of 45 papillary carcinomas, 8 follicular carcinomas, and 142 benign nodules regarding the following criteria: mild hypoechogenicity, marked hypoechogenicity, microlobulated or irregular margins, microcalcifications, taller than wide shape, missing thin halo.
All criteria but mild hypoechogenicity were significantly more frequent in thyroid cancer than in benign nodules. The number of positive criteria was significantly higher in cancer (2.79 ± 1.35) than in benign nodules (1.73 ± 1.18; p < 0.001). Thus, with a cut-off of two or more positive criteria, a sensitivity of 85% and a specificity of 45% were reached to predict malignancy in this sample of thyroid nodules. As expected, the findings were even more pronounced in papillary cancer only (2.98 ± 1.32 vs. 1.73 ± 1.18, p < 0.001). The six ultrasound criteria could not identify follicular cancer.
Our findings support the recently published EU-TIRADS score. Apart from mild hypoechogenicity, the analyzed ultrasound criteria can be applied for risk stratification of thyroid nodules in the previously severely iodine deficient population of Austria.