Abstract
Objective: The Afirma gene expression classifier (GEC) is a molecular test to further classify indeterminate fine-needle aspiration (FNA) as benign or suspicious for malignancy. Study Design: A total of 158 FNAs with Bethesda category III/IV cytology were sent for an Afirma GEC test. We correlated the Afirma GEC results with surgical outcome and also compared the data after Afirma's implementation with the data before. Results: Among the 158 FNAs, the Afirma result was benign in 63 (40%), suspicious in 85 (54%) and unsatisfactory in 10 (6%). In total, 73 (86%) suspicious Afirma cases had surgery and 28 (38%) showed carcinoma. In contrast, only 8 (13%) benign Afirma cases had surgery and all of them were benign. The sensitivity, specificity, negative predictive value and positive predictive value (PPV) of Afirma were 100, 15, 100 and 38%, respectively. The PPV was 20% in cases with follicular lesion of undetermined significance, but was 50% in cases suspicious for follicular neoplasm (SFN). The surgical excisional rate was significantly decreased in SFN cases after the Afirma test. Conclusions: The Afirma GEC is useful for further risk stratifying SFN cases.