Surgical utility of Afirma: effects of high cancer prevalence and oncocytic cell types in patients with indeterminate thyroid cytology

The Afirma Gene Expression Classifier (GEC) molecular marker assay was developed for the purpose of improving surgical decision-making with indeterminate fine-needle aspiration (FNA) biopsies of thyroid nodules. In this paper, we analyze the performance of the GEC over 27 months in a community hospital-based thyroid surgery practice. We began using GEC and Thyroid Cytopathology Partners (TCP) exclusively for thyroid FNA analysis in January 2011, shortly after the Afirma GEC became commercially available. In this paper, we focus on patients with indeterminate FNA results and the outcomes of GEC analysis, with particular attention paid to the calculation of the negative predictive value (NPV) of the Afirma test. We performed 645 FNAs in 519 patients over 27 months. Overall, 58 FNAs (9%) were read as indeterminate, with 36 of these classified as suspicious by GEC (62%), 20 characterized as GEC benign (34%), and 2 determined to be inadequate due to low mRNA content. Of the 36 suspicious GEC patients, 30 underwent thyroidectomy, and 21 of the 30 had malignant final pathology. Of the 20 benign GEC patients, 5 underwent thyroid surgery, and 2 were discovered to have malignancies. The NPV for the Afirma GEC in our practice environment was 89.6%. In a practice with a high incidence of thyroid cancer in patients with indeterminate FNAs (33% for our practice), the NPV of the Afirma GEC test may not be as robust as suggested in the existing literature.