Performance of Afirma Gene Sequencing Classifier versus Gene Expression Classifier in thyroid nodules with indeterminate cytology.

Abstract

About 15% to 30% of thyroid fine-needle aspiration (FNA) nodules have indeterminate cytology. The Afirma (Veracyte Inc, South San Francisco, CA) Gene Expression Classifier (GEC)/Gene Sequencing Classifier (GSC) tests were designed to improve risk stratification of the indeterminate thyroid nodules. This study aimed to evaluate and compare the performance of the Afirma GEC and GSC tests in the indeterminate thyroid lesions. Thyroid FNA cases with indeterminate cytology were searched in the pathology database and only those with available Afirma results were selected for this study. Each patient's demographic, sonographic, cytologic, molecular, and subsequent surgical follow-up results were collected and analyzed. There were 100 cases with indeterminate thyroid FNA results, including 49 cases tested by GEC and 51 cases by GSC. In the GEC group, benign call rate (BCR) was 53% (26 of 49) and the calculated negative predictive value (NPV) and positive predictive value (PPV) were 88% and 47% respectively. In the GSC group, the BCR was 63% (32 of 51) and the calculated NPV and PPV were 100% and 64%, respectively. Whereas only 17% (1 of 6) of benign oncocytic lesions were tested benign by the GEC, 60% (3 of 5) of benign oncocytic nodules were tested benign by the GSC. We demonstrated in this study that a little more than half of the indeterminate thyroid nodules had negative Afirma GEC/GSC results and the BCR using the Afirma GSC test was higher than GEC. The Afirma GSC showed higher NPV and PPV than GEC. In addition, the Afirma GSC appeared to be superior for differentiating benign and malignant oncocytic thyroid lesions.