Objective: To evaluate the efficacy of selected ultrasound (US) features of thyroid focal lesions useful for establishing indications for fine-needle aspiration biopsy (FNAB) with regard to the lesion's size.
Methods: US imaging features of 1141 thyroid nodules (shape, echogenicity, pattern of blood flow, presence of microcalcifications and the presence of other nodules in the thyroid) and their palpability were compared with the post-operative histopathological outcomes. The efficacy of the selected sets of the features was assessed for small nodules (SN)< or =15 mm and large nodules (LN)>15 mm, as well as separately for nodules< or =10 mm.
Results: Logistic regression analysis showed that in SN hypoechogenicity (odds ratios, OR: 3.18), microcalcifications (OR: 19.12), solitary occurrence (OR: 3.29) and height-to-width ratio> or =1 (OR: 8.57) were independent risk factors for malignancy. The optimal set of small lesions that should be biopsied includes all lesions presenting at least one of the above-mentioned features (sensitivity 98%, specificity 44%). In the LN group, the selection criteria based on the shape of lesions and hypoechogenicity were less sensitive than in the SN group, but they allowed further reduction in the number of performed FNABs. Large nodules primarily selected for FNAB should be hypoechoic, more tall then wide or contain microcalcifications (sensitivity 84%, specificity 72%).
Conclusions: The obtained results provide rationale for using features from the US examination in selecting both small and large nodules for FNAB. In the case of LN, the usefulness of sonographic features is less sensitive, but more specific than in the case of SN.