Endocrine Abstracts

The prevalence of thyroid nodules is increasing in the general population, thus detecting the ones at risk represents the main concern in clinical practice. The aim of the study was to compare four important Thyroid Imaging and Reporting Data Systems (TI-RADS) and assess how elastography and 4D Color Doppler vascularity evaluation improve the estimation of the malignancy risk. 133 nodules were evaluated by means of conventional ultrasound (US), elastography and 4D Doppler. All nodules were classified using the four selected systems and our proposed improved score: The American College of Radiology (ACR), the EU TI-RADS, Horvath TI-RADS and Russ’ French TI-RADS. Outcomes were compared to the pathology exam. The percentage of malignancy was of 26.31%. The ACR and EU TI-RADS showed similar results: good sensitivity of 94.28% and 97.14%, NPV of 93.33% and 95.83%, but fairly poor specificity of 31.81% and 23.46%, a PPV of 35.48% and 31.19% respectively, with an accuracy of 42.8% and 45.8%. Horvath TI-RADS displayed better accuracy of 66.9%, slightly improved specificity (62.24%), but poorer sensitivity (80%). The French TI-RADS, whichadds elastography to the algorithm, showed better outcomes (Se: 91.42%, NPV:96.42%, Sp:82.65%, PPV:65.30%, and accuracy of 84.96%). A mean strain ratio (SR) value greater than 4 was found in 80% of malignant group and in only 12% of the benign nodules. The parameter itself showedgood correlation to the response variable: the pathology report (P < 0.001). Similar accuracy was achieved when considering 4D vascularity as an additional parameter to the French score, butwhen studiedby itself, it proved beneficial for predicting cancer (P < 0.001), thus it may bring important extra information in uncertain cases. Adding elastography and 4D vascularity information to the standard US evaluation of thyroid nodules did improve the estimation of the malignancy risk.