How "suspicious" is that nodule? Review of "suspicious" Afirma gene expression classifier in high risk thyroid nodules.

BACKGROUND: The Afirma gene expression classifier (GEC) is being increasingly utilized to confirm the benign nature of indeterminate FNA cytology results thus avoiding unnecessary surgical procedures. However the "suspicious" result of the Afirma GEC does not classify these indeterminate nodules further in determining appropriate management. This study investigated the outcome of the thyroid nodules deemed to be "suspicious" by the Afirma GEC in a high risk population.

METHODS: The pathology database was searched for all thyroid nodules with Afirma test results over a three year period, 2013-2015. All thyroid nodules with a "suspicious" Afirma GEC result were investigated. Patient medical records were retrospectively reviewed for clinical history, FNA results, radiologic findings, management and follow-up.

RESULTS: Afirma result was suspicious in 69 cases. On cytologic evaluation 3.0% of the cases were non diagnostic (ND), 9% benign, 62% AUS, and 26% suspicious for neoplasm (SN). There was no follow up in 13% of cases and 87% were resected (50% lobectomies and 50% total thyroidectomies). On surgical resection 82% were benign, with 45% follicular adenoma (FA), and 37% nodular goiter (NG). The remaining 18% were malignant.

CONCLUSION: The rate of malignancy in nodules suspicious by Afirma was 18.3% (11/60). The rate of malignancy in nodules suspicious for neoplasm (SN) on cytology interpretation was 31.2% (5/16). False positive rate of Afirma was 56% (32/57). We conclude that cytology interpretation has a higher rate of predicting malignancy, in nodules interpreted as SN, when compared with the Afirma test, by almost twofold Diagn. Cytopathol. 2017;45:308-311. © 2016 Wiley Periodicals, Inc.