Evaluation of the Afirma Gene Expression Classifier in Repeat Indeterminate Thyroid Nodules.

Published

Journal Article

- Molecular testing in indeterminate thyroid nodules is a rapidly evolving field with variable reported outcomes.- To report our experience at a tertiary thyroid referral center with the Afirma Gene Expression Classifier (Veracyte, San Francisco, California) in repeat fine-needle aspirations of thyroid nodules with a previous indeterminate cytologic result.- Results of cytopathology and the Afirma test were collected from August 2013 to March 2015, as were diagnoses from surgical resection when performed.- One hundred and fifteen thyroid nodules were evaluated by Afirma. The fine-needle aspiration diagnostic categories for these nodules were 100 (87%) Bethesda III, 10 (9%) Bethesda IV, 3 (2%) Bethesda II, 1 (1%) Bethesda V, and 1 (1%) Bethesda I. Afirma results for 52 of the nodules (45%) were benign, 57 (50%) were suspicious, and 6 (5%) specimens yielded no result because of low messenger RNA content. Three of the benign nodules (6%) were treated surgically, and all were benign on final surgical pathology. Forty-six (81%) of the suspicious nodules were treated surgically; final surgical pathology revealed 30 (65%) were benign and 16 (35%) malignant, yielding a positive predictive value of 35%.- In our experience, 50% of the indeterminate nodules were classified as suspicious by Afirma, with a 35% rate of malignancy in these nodules at surgical resection, in comparison with a historical rate of malignancy at our institution of 11% for Bethesda III nodules and 23% for Bethesda IV. Our use of Afirma is consistent with prior reports in that it has a low positive predictive value in indeterminate thyroid nodules.

Full Text

Duke Authors

Cited Authors

  • Harrison, G; Sosa, JA; Jiang, X

Published Date

  • July 2017

Published In

Volume / Issue

  • 141 / 7

Start / End Page

  • 985 - 989

PubMed ID

  • 28467214

Pubmed Central ID

  • 28467214

Electronic International Standard Serial Number (EISSN)

  • 1543-2165

International Standard Serial Number (ISSN)

  • 0003-9985

Digital Object Identifier (DOI)

  • 10.5858/arpa.2016-0328-oa

Language

  • eng