The purpose of this study was to clarify ultrasound (US) evaluation of microcalcifications and determine whether the calcifications seen in US scans can reliably predict malignant thyroid tumors. Diagnostic accuracy of microcalcification and other various signs seen in US scans for predicting malignancy was evaluated prospectively in 259 pathologically verified thyroid nodules. Sonographic and pathologic correlation of calcifications was performed on 69 of 99 surgically removed nodules. Pathologic studies revealed that hyperechoic areas with acoustic shadowing represented mostly amorphous dense calcifications and sometimes microcalcifications, but small particles without acoustic shadowing mainly reflected microcalcifications and sometimes large amount of fibrous bands and condensed colloids. Of the various sonographic signs, microcalcification showed the highest accuracy (76%), specificity (93%), and positive predictive value (70%) for malignancy as a single sonographic sign, but its sensitivity (36%) was poor. Although sonographic microcalcification showed relatively high specificity, the accuracy of this finding for malignancy was insufficient.

Ultrasound (US) and US‐guided fine‐needle aspiration biopsy (FNA) were performed in 91 nonpalpable neck nodes of 70 patients, 93% of which had known malignancy. Various sonographic findings were evaluated for predicting malignancy. The accuracy of US‐guided FNA for detecting malignancy was 88%, with 96% sensitivity and 94% specificity. The ratio of minimal to maximal axial diameters of a node was most valid for predicting malignancy with US. A ratio of more than 0.55 yielded the highest accuracy (80%) (92% sensitivity, 63% specificity). Addition of any other factors to this criterion did not improve its accuracy. US and US‐guided FNA are accurate for the assessment of nonpalpable neck nodes. Lymph nodes with a round configuration should be biopsied in patients with known malignancy. © 1997 John Wiley & Sons, Inc. J Clin Ultrasound 25:283–292, 1997

Ultrasound (US) and US-guided fine-needle aspiration biopsy (FNA) were performed in 91 nonpalpable neck nodes of 70 patients, 93% of which had known malignancy. Various sonographic findings were evaluated for predicting malignancy. The accuracy of US-guided FNA for detecting malignancy was 88%, with 96% sensitivity and 94% specificity. The ratio of minimal to maximal axial diameters of a node was most valid for predicting malignancy with US. A ratio of more than 0.55 yielded the highest accuracy (80%) (92% sensitivity, 63% specificity). Addition of any other factors to this criterion did not improve its accuracy. US and US-guided FNA are accurate for the assessment of nonpalpable neck nodes. Lymph nodes with a round configuration should be biopsied in patients with known malignancy.

No abstract