Thyroid gland
General
Molecular testing in FNA

Author: Rachel Jug, M.D., Xiaoyin "Sara" Jiang, M.D. (see Authors page)

Revised: 15 February 2017, last major update May 2016

Copyright: (c) 2016-2017, PathologyOutlines.com, Inc.

PubMed search: "molecular testing" thyroid "FNA"
Cite this page: Molecular testing in FNA. PathologyOutlines.com website. http://pathologyoutlines.com/topic/thyroidglandmolectestingfna.html. Accessed April 28th, 2017.
Definition / general
Why do we need molecular testing in FNA cytopathology?
  • There are many factors that can lead to a diagnosis of an indeterminate thyroid nodule, based on The Bethesda System for Reporting Thyroid Cytopathology, classified as Bethesda III ("atypia of undetermined significance (AUS) / follicular lesion of undetermined significance (FLUS)") or IV ("suspicious for follicular neoplasm (SFN) / follicular neoplasm (FN)")
  • Ancillary molecular testing is needed to further stratify the risk of these lesions to aid in clinical decision making (e.g. close followup vs repeat FNA vs surgical lobectomy vs thyroidectomy)

Molecular / cytogenetics description
Purpose of molecular tests:
  • Further characterize lesions diagnosed as indeterminate on FNA to aid clinical decision making
  • Note: with the reclassification of the encapsulated follicular variant of papillary thyroid carcinoma (EFVPT) to noninvasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP), the number of malignant diagnoses will significantly decrease, necessitating reevaluation of the predictive value of each of these molecular tests, since the rate of malignancy and specificity of these panels will likely decrease, especially category III lesions (AUS / FLUS) (Journal of the American Society of Cytopathology 2016;LIII:I)

Background - Genetic alterations in thyroid lesions:
  • Papillary thyroid carcinoma most commonly contains the following genetic alterations: RET (13% - 43%), BRAF mutation (29% - 69%), NTRK1 rearrangement (5% - 13%), Ras mutation (0% - 21%) (Nat Rev Cancer 2006 Apr;6:292)
  • Follicular thyroid carcinoma most commonly contains the following genetic alterations: Ras mutation (40% - 53%), PPARG rearrangement (25% - 63%) (Nat Rev Cancer 2006 Apr;6:292)
  • About 25% of cases lack the common driver mutations
  • The noninvasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP) and invasive encapsulated follicular variant of papillary thyroid carcinoma (EFVPT) possess molecular profiles similar to follicular adenomas / carcinomas (higher rates of Ras than BRAF mutations) (Mod Pathol 2010 Sep;23:1191)
  • Conversely, the infiltrative follicular variant of papillary carcinoma has a molecular profile more similar to that of classic papillary thyroid carcinoma (higher rates of BRAF than Ras mutations)
  • The molecular profile of encapsulated and infiltrative follicular variant parallels their biological behavior (Mod Pathol 2010 Sep;23:1191)
  • Hence, FNA-indeterminate diagnoses may suggest the need for molecular testing to further prognosticate thyroid lesions into these molecular / behavioral subsets of lesions

Overview of Molecular tests / platforms:
  • ThyroSeq2 (UPMC)
    • Sample Collection: Residual aspirated material collected from routine FNA placed in nucleic acid preservative solution (J Clin Endocrinol Metab 2011 Nov;96:3390)
    • Test method: Next generation sequencing assay and RT-PCR detects hundreds of point mutations and gene fusions in more than 60 thyroid cancer genes
    • Performance: Reports specific gene mutations and their corresponding risk of malignancy; Bethesda III - AUS / FLUS on cytology classified as benign or malignant with a positive predictive value (PPV) of 88% and negative predictive value (NPV) of 97%
      Bethesda IV - SFN / FN on cytology classified as benign or malignant with a positive predictive value (PPV) of 83% and negative predictive value (NPV) of 96%
    • Caveats: Interpretation requires integrating results (found gene mutations and corresponding risk of malignancy) into clinical context as opposed to a binary result (benign / suspicious) to dictate need for surgery
    • Information from UPMC with list of tested mutations: ThyroSeq®
    • References: Journal of the American Society of Cytopathology 2016;LIII:I, Thyroid 2015 Nov;25:1217

  • Asuragen miRInform
    • Sample Collection: Performed using samples from residual aspirated material collected from routine FNA that is placed in nucleic acid preservative solution
    • Test method: Limited mutation panel assesses for presence of 17 known thyroid cancer mutations, including BRAF and RAS point mutations and common rearrangements of RET / PTC and PAX8 / PPARγ
    • Test components available as stand alone assay or in varying combinations by other laboratories (for example, Quest Diagnostics)
    • Performance: PPV is 88% for Bethesda III category (AUS / FLUS) and 87% for Bethesda IV category (follicular neoplasm or suspicious for a follicular neoplasm)
    • Caveats: Limited mutational panel, requires interpretation of molecular results in clinical context
    • References: Journal of the American Society of Cytopathology 2016;LIII:I, J Clin Endocrinol Metab 2009 Jun;94:2092

  • Veracyte's Afirma Gene Expression Classifier (GEC)
  • Cleveland Clinic TSH Receptor mRNA test
    • Sample Collection: Two 7 mL lavender EDTA tubes of (peripheral) whole blood (collected prior to FNA) transported on ice; RNA must be isolated within 24 hours of blood draw
    • Test method: Measures serum thyrotropin receptor mRNA in peripheral blood to detect circulating invasive carcinoma cells of follicular origin by a molecular based assay using quantitative RT-PCR (RNA is extracted, reverse transcribed then subject to PCR and measured against a reference preparation of thyroid cancer RNA)
    • TSHR mRNA levels > 1.0 ng/µg total RNA = presence of thyroid cancer (either as a new preoperative diagnosis or may indicate high likelihood of recurrent or persistent disease in patients with known thyroid cancer)
    • Results reported in arbitrary units as reference preparation equivalent TSHR mRNA ng/µg of total RNA (Clin Chem 2002;48:1862)
    • Performance: Overall 81% PPV and 64% NPV but specificity increased when limited to category IV lesions (follicular neoplasm or suspicious for follicular neoplasm - 96% PPV and 77% NPV) (Journal of the American Society of Cytopathology 2016;LIII:I)
    • 72% sensitivity and 83% specificity in differentiation of malignant nodules from benign nodules (J Clin Endocrinol Metab 2004 Aug;89:3705, J Clin Endocrinol Metab 2005 Apr;90:1921)
    • TSHR mRNA combined with FNA had 90% sensitivity and 80% specificity to diagnose thyroid cancer (Surgery 2010 Dec;148:1313)
    • The estimated functional sensitivity of the test is 0.14 ng/µg of total RNA (Clin Chem 2002;48:1862)
    • Caveats: Sample must be shipped overnight to ensure RNA isolation within 24 hours of collection
    • Multiple studies indicate detection and quantification of thyroglobulin and TSH-receptor mRNA in peripheral blood is unlikely to be suitable for follow up of differentiated thyroid cancer, due to the low levels detected in patients without evidence of the disease, suggesting small numbers of thyroid cells may be normally present in the circulation or that there may be ectopic transcription of thyroglobulin and TSH receptor mRNA from blood cells (Rev Esp Med Nucl 2010;29:109, Eur J Endocrinol 2002;147:575)
  • Note: with the reclassification of the encapsulated follicular variant of papillary thyroid carcinoma (EFVPT) to noninvasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP), the number of malignant diagnoses will significantly decrease, necessitating reevaluation of the predictive value of each of these molecular tests, since the rate of malignancy and specificity of these panels will likely decrease, especially category III lesions (AUS / FLUS) (Journal of the American Society of Cytopathology 2016;LIII:I)
  • Molecular / cytogenetics images

    Images hosted on other servers:

    Sample of Asuragen biopsy report

    Veracyte's Afirma
    Gene Expression
    Classifier
    (Download PDF)

    Proposed clinical algorithm

    Predicted NPV and PPV of
    ThyroSeq v2.1 compared to the
    Afirma gene expression classifier
    test in AUS / FLUS nodules