Purpose: To evaluate safety, yield, and accuracy of ultrasonography (US)-guided core-needle biopsy of the thyroid gland.
Materials and methods: Findings at 209 consecutive core-needle biopsies of lesions of the thyroid gland in 198 patients (median age, 48 years; age range, 14-80 years) were retrospectively reviewed. In 138 (66%), findings at previous fine-needle aspiration cytologic (FNAC) analysis were nondiagnostic on one to five occasions. Biopsy was performed as an outpatient procedure with direct US guidance by using nonadvancing 16-18-gauge core needles. Hospital records were reviewed 6 months to 5 years following biopsy to determine final diagnosis, delayed complications, and influence of biopsy findings on subsequent patient treatment. Final diagnoses were determined on the basis of findings at excisional histologic analysis, clinical course, or other laboratory values. The sensitivity, specificity, and accuracy of US-guided core-needle biopsy were calculated.
Results: One hundred ninety-nine (95%) specimens were adequate for histologic diagnosis. The sensitivity, specificity, and accuracy of core biopsy in differentiating neoplastic (malignant and benign) from nonneoplastic lesions of the thyroid gland were 96% (74 of 77), 89% (109 of 122), and 92% (183 of 199), respectively. The sensitivity, specificity, and accuracy of core-needle biopsy in the detection of malignant neoplasms were 61% (11 of 18), 100% (181 of 181), and 96% (192 of 199), respectively. After US-guided core-needle biopsy, 115 (58%) of 198 patients were treated conservatively, and no evidence of missed tumor manifested during the follow-up period. In the 83 patients who underwent surgical resection, biopsy was performed for therapeutic reasons in 76 (92%) and for diagnostic reasons in seven (8%). There were three cases of small postbiopsy hematomas and one of minor hemoptysis, but none required hospital admission. There were no major complications.
Conclusion: US-guided core-needle biopsy of the thyroid gland is a safe outpatient procedure with a high diagnostic yield and accuracy, and frequently it obviates surgery in patients in whom findings at FNAC analysis are recurrently nondiagnostic.