Accuracy of fine-needle aspiration cytology and frozen-section examination in patients with thyroid cancer

In patients with solitary thyroid nodules (TN) undergoing surgery both fine-needle aspiration (FNA) and frozen-section examination (FSE) are currently performed, but their role is still controversial. We retrospectively analyzed a series of 606 patients with a nontoxic solitary thyroid nodule who underwent both FNAB and FSE prior to thyroidectomy. There were 118 (19.5%) men and 488 (80.5%) women, with a median age of 44 years (range 16-81 years). The results of both FNAB and FSE were compared against the final pathological examination. Definitive histology showed 500 (82.5%) benign nodules, including 239 (39.4%) follicular adenomas, and 106 (17.5%) carcinomas, of which 18 (17.0%) were follicular cancer. In differentiating between benign TN and thyroid tumors the sensitivity, specificity, and accuracy of FNAB were 93.6%, 98.9% and 95.9%, while in the detection of malignancy were 81.1%, 99.4% and 96.2% for FNAB, and 83.0%, 100% and 97.0% for FSE (P = NS), respectively. The combination of both techniques did not improve significantly (P = NS) the results. Both medullary and undifferentiated cancer, and 76 of 80 (95.0%) papillary carcinomas were correctly detected by FNAB. Three (0.5%) false-positive smears suggesting cancer were found. In conclusion, in patients with FNAB revealing thyroid cancer or a benign TN, the extent of thyroidectomy should be established by performing FSE. When FNAB suggests the presence of a follicular tumor, FSE may affect rarely the final intraoperative decision-making, and should be considered unnecessary. However, in patients with FNA revealing thyroid cancer, FSE is useful in confirming malignancy, and may avoid surgical overtreatments. (C) 2003 Elsevier SAS. All rights reserved.