Usage trends and performance characteristics of a "gene expression classifier" in the management of thyroid nodules: An institutional experience

Abstract

Background: The gene expression classifier (GEC; Afirma-Veracyte) has proven to be an effective triage modality in the management of thyroid nodules. We evaluate our institutional experience with GEC, specifically examining performance as a first line testing strategy versus in conjunction with repeat fine needle aspiration (FNA), usage trends based on clinical setting, and performance related to diagnostic categories of The Bethesda System for Reporting Thyroid Cytology (TBSRTC).

Methods: All nodules undergoing GEC analysis from 1/2011 to 12/2015 at the Hospital of the University of Pennsylvania were identified using electronic database search methods. Corresponding cytologic diagnoses, GEC results, origin of the sample (in-house vs. satellite site), number and diagnosis of prior FNA's, and clinical and histologic follow-up were collected.

Results: The cohort included 294 nodules. Of these, 145 (49%) were classified as benign, 136 (46%) as suspicious, and 13 (5%) as quantity insufficient by GEC. Surgical resection was performed in 130 (130/294-44%) cases (107, 82% "suspicious" by GEC); final histopathologic diagnosis was benign in 85 (65%) and malignant in 45 (35%) cases. Three false negative diagnoses were identified in the setting of GEC analysis as a first line testing strategy. Most cases with GEC as a first line testing strategy came from satellite clinical sites (112, 66%).

Conclusions: The GEC showed improved performance characteristics when coupled with a repeat FNA. It continues to be of low specificity and positive predictive value in oncocytic follicular lesions. Diagn. Cytopathol. 2016;44:867-873. © 2016 Wiley Periodicals, Inc.

Keywords: Afirma; fine needle aspiration; follicular lesion of undetermined significance; gene expression classifier; indeterminate cytology; thyroid.