Routine second-opinion cytopathology review of thyroid fine needle aspiration biopsies reduces diagnostic thyroidectomy

Abstract

Follicular thyroid carcinoma cannot be distinguished reliably from benign follicular neoplasia by fine needle aspiration (FNA) biopsy. Given an estimated 20% risk of malignancy, many patients with indeterminate FNA biopsies require thyroidectomy for diagnosis. Some centers have shown significant discordance when a second pathologist evaluates the same FNA biopsy. We sought to determine whether routine second-opinion cytopathology reduces the need for diagnostic thyroidectomy, especially in patients with indeterminate FNA biopsies. In all, 331 thyroid FNA biopsy specimens obtained from outside centers from 2004 to 2009 were reviewed at our institution. The FNA biopsy results were categorized into nondiagnostic (Bethesda I), benign (Bethesda II), indeterminate (follicular/Hurthle cell neoplasm, follicular/Hurthle cell lesion; Bethesda III & IV), and malignant (papillary or suspicious for papillary or other malignancy; Bethesda V and VI). Second-opinion cytology was compared with the initial opinion in 331 cases and with final operative pathology in the 250 patients who progressed to thyroidectomy. The average patient age was 51 with a predominant number of female (79%) participants. The overall cytology concordance for all 331 FNA biopsies was 66% (218/331). Concordance was highest at 86% (74/86) with malignant FNA biopsies. Concordance in the 129 patients with indeterminate FNA biopsies was only 37% (48/129). Indeterminate FNA biopsies were reread as nondiagnostic in 21% (27/129) of patients and as benign in 42% (54/129) of patients. Twenty-two patients with an indeterminate FNA biopsy reread as benign progressed to operative therapy for reasons other than cytology (eg, symptomatic nodule and radiation exposure/high risk) and were found to be benign in 95% (21/22) of patients on operative pathology for a 95% negative predictive value. An additional 11 patients with an indeterminate FNA reread as benign had follow-up FNA biopsies, each of which was benign. Indeterminate FNA biopsies on initial cytology had a malignancy rate of 13% (17/129) on operative pathology compared with 29% (14/48) for indeterminate FNA biopsies from second opinion. A second opinion improved FNA biopsy accuracy from 60% to 74%. Overall, second-opinion cytology of indeterminate FNA biopsies avoided diagnostic operation in 25% (32/129). Routine second opinion review of indeterminate thyroid FNA biopsies can potentially obviate the need for diagnostic thyroidectomy in 25% of patients without increases in false negatives.