Ultrasound-guided fine-needle aspiration biopsy of the thyroid

We reviewed the Massachusetts General Hospital experience with ultrasound-guided fine-needle aspiration biopsies (FNABs) of the thyroid to determine the indications, rate of unsatisfactory smears, correlation with excisional biopsy results, and verification of efficient use of personnel time.

All radiologically guided FNABs of the thyroid from January 1993 through June 1997 were reviewed. As a measure of efficient use of technologist time, a sample of times spent by the technologist during the procedure for 20 cases in 1993 and 1997 was compared with that of an equal number of random nonthyroid image guided FNABs.

Two hundred-ninety FNABs were identified in 251 patients, representing 12% of all thyroid FNABs and 11% of all radiologically guided FNABs. Indications in the 251 patients included multiple nodules (78), solitary nodules (61), complex nodules (39), prior failed FNAB (39), thyroid bed abnormalities post-thyroidectomy (21), difficult access (7), and investigation of recurrent tumor in residual thyroid lobe (6). Available records indicated 118 lesions were palpable and 45 were nonpalpable; the physical examination characteristics of the remainder (88) were not stated. Diagnoses included 44 unsatisfactory cases (15%), 103 macrofollicular lesions, 20 microfollicular lesions, 26 mixed macro/microfollicular lesions, 5 oxyphilic lesions, 1 trabecular pattern, 15 nonspecific follicular cell pattern, 9 follicular cell atypia, 30 cysts, 11 thyroiditis, 23 malignant tumors, and 3 other (1 parathyroid, 2 lymph node). Eighty-nine FNABs from 76 patients had subsequent surgical biopsy. Excisional biopsies in 14 unsatisfactory FNABs were benign. In the remaining 75 FNABs from 67 patients, 18 malignancies on FNAB were correctly diagnosed, but 3 other papillary carcinomas were only qualified as atypical follicular cells on cytology. No false-positive cases occurred. Of 15 macrofollicular lesions on cytology, 10 were adenomas on excision, only 2 of which were microfollicular adenomas, and 4 were adenomatous nodules. An aspirate of a parathyroid adenoma was misinterpreted as a macrofollicular lesion of the thyroid. Three microfollicular lesions on FNAB proved to be nodular hyperplasia on excision, and the other 11 were adenomas, 5 of them microfollicular. Average technologist time was significantly longer for thyroid FNABs than nonthyroid FNABs in 1993, but in the 1997 sample no significant difference was identified.

Radiologically guided FNAB of the thyroid is a clinically useful procedure with a high correlation between benign lesions not needing excision (macrofollicular), and lesions that need excision (microfollicular/oxyphilic cell or malignant). Technologist time needed for immediate evaluation tends to decrease with increasing operator experience. Cancer (Cancer Cytopathol)