ABSTRACT
BACKGROUND: The utility of frozen section (FS) for indeterminate thyroid nodules is controversial. In 2009, the Bethesda System for Reporting Thyroid Cytopathology was established to further subcategorize indeterminate fine-needle aspiration results (follicular lesions, FL) into Bethesda category 3 (BC3) and Bethesda category 4 (BC4). We hypothesize that FS will have less utility in the evaluation of BC3 lesions when compared to BC4. MATERIALS AND METHODS: A total of 479 patients who underwent thyroid lobectomy from January 2008 to July 2014 were retrospectively reviewed. Patients without appropriate Bethesda categorization were excluded. A total of 135 patients (65 FL, 45 BC3, 25 BC4) comprised the study groups. The sensitivity and specificity of FS within these three categories were determined. RESULTS: In the FL group, 6 of 65 patients were found to have thyroid cancer. Three were identified on frozen section (FS) resulting in a sensitivity and specificity of 50% and 100%, respectively. Thus, FS changed the operation in 3 of 65 cases (4.6%). In the BC3 group, 5 of 45 patients were found to have cancer. One was identified on FS resulting in a sensitivity and specificity of 20% and 100%, respectively. Thus, FS changed the operation in 1 of 45 patients (2.2%). In the BC4 group, 4 of 25 patients were found to have cancer. Two were identified on FS resulting in a sensitivity and specificity of 50% and 100% respectively. Thus, FS changed the operation in 2 of 25 patients (8%). CONCLUSIONS: There is improved utility of FS in BC 4 patients as 8% avoided reoperation. However, this benefit hinges on surgeon practice regarding the management of differentiated thyroid cancer >1 cm and <4 cm.
