Preoperative diagnostic tests for operable thyroid disease

Although the availability and acceptance of fine‐needle aspiration biopsy (FNAB) of thyroid nodules has increased, many physicians still use thyroid scintigraphy for distinguishing benign from malignant lesions. We evaluated these diagnostic tests in 350 patients who had thyroid surgery in our institution between 1977 and 1990. Histologic confirmation of FNAB was obtained in 265 patients. In the group of patients having surgery, 247 thyroid scintigraphies were performed. Our patients were divided into two groups (1977–1986 and 1986–1990). The first group comprised 173 patients with 173 FNABs and 126 scintigrams. The second group consisted of 177 patients having 92 FNABs and 121 scintigrams. Results of scintigrams were analyzed in the second group only. In 5 out of 120 cases where the FNAB result was “benign or probably benign” the lesion appeared to be malignant postoperatively. If the FNAB result was “malignant or probably malignant” ( n =83) the pathology report confirmed a malignancy in 68 cases (81.9%). In 56 instances of all 265 FNABs the cytology report was not conclusive (“uncertain”); in 21.4% of these cases a malignancy was found postoperatively. An FNAB‐result “(probably) malignant” had a positive predictive value of 0.819 while the negative predictive value of a result “(probably) benign” is 0.950. An “uncertain” result does not take away our concern so this result should have the same consequences as those of a result “(probably) malignant”. In that case, FNAB‐sensitivity is 93.0% and specificity 66.1%. Eighty‐five of the last 116 scintigraphies showed a solitary node. Eleven of these nodes were hot while 74 were cold. In 2 out of 11 hot nodules cancer was diagnosed (18.2%). Twenty‐six of the cold nodules contained a malignancy (35%). Evaluation of 116 99m Tc scintigrams yielded a positive predictive value of finding a malignancy in the thyroid if the scan showed a solitary cold nodule of 0.351. If a cold nodule is supposed to be an indicator of malignancy this test has a sensitivity of 72.2% and a specificity of 43.5%. The accuracy is as low as 52.1%. Our data indicate that the accuracy rate of nuclear imaging techniques in diagnosing thyroid nodules is low. There is no place for routine scintigraphy in preoperative diagnosis of a thyroid lesion. Although observer dependency exists with FNAB, the overall accuracy of this test is high, while this low‐cost test is easy to perform and increasingly available. It enables cancer detection in a cost‐effective manner. We recommend surgery for a clinically suspicious lesion and if the FNAB result is uncertain, probably malignant, or malignant. We advocate direct total thyroidectomy if the FNAB result is malignant, peroperative frozen section after hemithyroidectomy if the FNAB result was probably malignant or in a clinically suspicious lesion regardless of the FNAB result (unless malignant). A diagnostic hemithyroidectomy only, without frozen section, is performed if the FNAB result was uncertain.