Molecular testing is increasingly becoming the new standard of care in evaluating thyroid nodules for the presence of cancer. This methodology is now included in recommendations of leading clinical guidelines, including those from the American Thyroid Association (ATA), National Comprehensive Cancer Network and UpToDate.1-3
As further evidence of molecular testing’s ubiquity, one need look no further than the 2017 ATA annual meeting in British Columbia, where nearly 30 presentations covered new ground on this topic.
The most common application of molecular testing in thyroid nodule assessment is to guide care for patients with indeterminate nodules. Cytopathology examination of ultrasound-guided fine needle aspiration (FNA) biopsies is the standard preoperative tool for evaluating thyroid nodules larger than one cm. However, in 15-30% of nodule aspirates, the results are indeterminate.4
Based on the Bethesda System for Reporting Thyroid Cytopathology, indeterminate thyroid nodules encompass atypia of undetermined significance/follicular lesion of undetermined significance (AUS/FLUS; Bethesda category III); follicular neoplasm or suspicious for a follicular neoplasm (FN/SFN; Bethesda category IV); and suspicious for malignancy (SM; Bethesda category V). Indeterminate nodules have a high enough risk of cancer to warrant additional evaluation.1
Traditionally, most patients with cytologically indeterminate nodules have been referred for diagnostic surgery, even though 70% to 80% of these nodules ultimately prove to be benign by surgical histopathology — meaning the surgery was unnecessary.5 Further, patients with confirmed thyroid cancer may not receive the surgery they need. For example, Esfandiari et al6 reported that over 40% of patients with medullary thyroid cancer (MTC) did not undergo a total thyroidectomy and central neck dissection, which this diagnosis would typically require. These patients may require a second surgery to “complete” the thyroidectomy.
Thyroid surgery can take a significant toll on patients, due to its morbidity, operative risks, and costs. Thus, the decision to operate should not be taken lightly. All cases of thyroidectomies and an estimated 18% of thyroid lobectomies result in hypothyroidism, requiring lifelong thyroid hormone treatment with levothyroxine.7
Patients with hypothyroidism can develop weight gain, fatigue, and sleepiness – conditions that can be difficult to manage even with medication. Furthermore, potential adverse effects of thyroid hormone therapy include cardiac arrhythmias and loss of calcium, raising the risk of osteopenia or osteoporosis.8 Some patients reported simply not feeling well after removal of all or part of their thyroid. Added to this are the psychological toll of poor health and the inconvenience of requiring frequent physician visits and medication refills.
The financial impacts of thyroid surgeries are also significant. In an analysis of private insurance claims data, Singer et al found that the total cost of thyroid surgery and related clinical follow-up for six months following surgery was $21,371.9 The wasted healthcare costs add up quickly when one considers that over 100,000 patients may undergo such surgery unnecessarily each year.5,10
These limitations in cytopathology have led to the emergence of molecular testing to provide cytopathologists and treating physicians with previously undetectable diagnostic information about patients’ thyroid nodules. Testing based on these technologies is enabling physicians to better determine which patients can safely avoid surgery as part of their diagnostic work-up and make better surgical decisions for those patients who require it. Given the risks, harm and costs associated with thyroid surgery, and the strong data demonstrating benefits of molecular testing in thyroid nodule evaluation, I believe its use is simply the right thing to do for patients.
A primary goal of molecular testing of indeterminate FNA samples is to identify patients who can avoid unnecessary surgery – while minimizing the risk of missing cancer. This requires a test to have a high sensitivity and, correspondingly, a high negative predictive value (NPV). Up until a few years ago, such use of molecular testing was primarily limited to academic settings and early tests showed lower-than-desired sensitivity rates.11-13
In 2011, the commercial introduction of a 167-gene classifier (Afirma® Gene Expression Classifier, or GEC) that identified benign nodules among those deemed indeterminate by cytopathology changed the trajectory for how molecular testing is used in thyroid cancer diagnosis. The test is now covered by Medicare and private insurers.
The 167-gene classifier is performed on patient samples that are collected during the initial FNA procedure, removing the need for patients to undergo a follow-up FNA procedure. The test uses mRNA expression data obtained through microarray analysis and machine learning to identify benign nodules among those previously classified as indeterminate. In a blinded, prospective clinical validation study of the locked gene classifier in 49 sites, Alexander et al14 determined that the classifier had a sensitivity of 92% and an NPV of 94% and 95% for AUS/FLUS and FN/SFN, respectively, among patients with a pretest malignancy rate of 32%. These rates are similar in accuracy to those of a benign cytopathology result.
Clinical utility data demonstrated that use of the 167-gene classifier changed patient care as intended. In a multicenter study, Duick et al15 showed that surgeries among patients with “benign” GEC results were reduced by 90% (from 74% to 7.6%), compared to historical surgery rates for patients with indeterminate nodules. Additional studies demonstrated the durability of a benign GEC result, as patients were followed clinically over an extended period of time and remained cancer free.
Subsequent classifiers have been introduced, which also combine genomic data and machine learning to identify benign nodules. One test (RosettaGx Reveal) is performed on a slide smear from the initial FNA biopsy in order to avoid the need for a separate collection tube and FNA pass. In a clinical validation study,16 Lithwick-Yanai et al reported that the test had a sensitivity of 74% for Bethesda III and IV nodules. Further published data for this test are limited.
Gene mutation testing requires examining specific genes for alterations associated with cancer. It often includes testing BRAF, and three RAS genes for specific point mutations and testing for three gene fusions: RET/PTC 1, RET/PTC3, and PAX8/PPARG. This type of molecular testing has been considered primarily as a “rule in” test for cancer due to its typically higher specificity for cancer, yet lower sensitivity. In thyroid nodules with indeterminate cytology, the sensitivity of this seven-gene mutational panel test is variable, with reports ranging from 36% to 100%.17-20
According to the 2015 ATA guidelines,1 the reported variability in sensitivity of mutational analysis with the seven-gene panel in indeterminate nodules suggests that mutation panels of this size may not reliably rule out malignancy with a negative test in this population. At the same time, molecular testing should only be used if the results will change the surgical decision-making.1
Since most patients with indeterminate thyroid nodules are already directed to surgery, use of a mutation panel that rules in cancer would only change management in the few patients that are initially being considered for observation, or if the extent of surgery were altered due to the risk of cancer associated with the specific genomic alteration.
A subsequent version of the mutation panel test (ThyroSeq® 2.1) includes 56 genes and promises a high accuracy in identifying both cancerous and benign nodules, which enables it to rule in and rule out cancer.21,22 These findings were based on single-center studies in which the treating physician and histopathologist were unblinded to some of the molecular results. An evaluation by Valderrabano et al concluded that more study was needed after their analysis demonstrated significantly lower sensitivity and specificity for the test compared to the initial validation findings. For example, in their evaluation, the test’s sensitivity was 70%, meaning that it missed 30% of the cancers in the study group.23
The notion that mutation does not equal malignancy, and vice versa, was demonstrated by Pagan et al.24 In this study, a 524-gene panel encompassing 851 variants and 133 fusions known to be associated with thyroid cancer, was used to evaluate thyroid nodule samples. The researchers found that gene alterations appeared in only 50% of the cancerous samples (ie, 50% sensitivity) and 20% of the benign thyroid samples harbored gene alterations (ie, 80% specificity).These findings suggest that the test lacked accuracy in ruling cancer in or out because gene mutations were found in both benign and malignant samples and similarly were not present in both types of samples.24
While mutational testing alone may not be enough to help the physician determine whether a patient requires surgery, it may help inform what type of surgery to perform. For example, because the BRAF V600E mutation has been shown to have a high positive predictive value for cancer, some physicians choose to perform a full thyroidectomy, rather than a lobectomy, on patients in whom this mutation is found. The same is true in cases of RET/PTC1 and RET/PTC3.
Additionally, several molecular tests have demonstrated the ability to identify medullary thyroid cancer (MTC), an aggressive, but rare form of cancer that is often difficult to detect with cytopathology. Knowing that a patient has MTC before surgery can enable the surgeon to more appropriately plan. This includes ordering the necessary germline genetic counseling and testing, and tumor marker tests that inform the need for additional tests, such as imaging studies to evaluate for metastases.
Pre-operative knowledge of MTC enables the treating physician to anticipate potential complications such as life-threatening hypertension that can accompany MTC in the setting of multiple endocrine neoplasia type 2.
Some molecular tests have also demonstrated the ability to distinguish parathyroid tissue from thyroid nodules. Because parathyroid cells often appear similar to atypical thyroid cells under the microscope, cytopathologists often designate them as indeterminate – typically, as Bethesda III or IV. This can lead to unnecessary thyroid surgery for these patients.
Genomic understanding of thyroid cancer, along with new sequencing technologies and bioinformatics capabilities are driving rapid advances in molecular testing, with significant potential to further improve surgical decision-making and patient care.
In 2017, an enhanced version of the GEC became available (Afirma® Genomic Sequencing Classifier), which data show has a high sensitivity (91%) and significantly increased specificity (from 52% to 68%), significantly improving on the number of surgeries that potentially may be avoided, compared to the previous version.25
The newer test version combines RNA sequencing to detect and measure extensive genomic content (e.g., gene expression, RNA variants and fusions, and other features) along with machine learning to enable better distinction between benign and malignant nodules. Findings presented at the 2017 ATA meeting demonstrated the test’s ability – through additional classifiers – to accurately identify the BRAF V600E mutation, parathyroid tissue, and MTC. 26-28
Also, a newer version of a gene mutation panel test (ThyroSeq® 3) was introduced, which combines more genes (112), as well as a mathematical formula that considers the likelihood of cancer for each gene mutation.29 Results from a multicenter study suggest that the test has a high sensitivity (94%) and slightly reduced specificity (80%).29 The test’s performance was evaluated on indeterminate thyroid nodules in patients that were recommended for surgery. While encouraging, it will be helpful to see how the test performs when used on a wider range of indeterminate nodules – including those that were not recommended for surgery (which can impact test performance) – in a real-world setting.
As the field continues to evolve, molecular testing is poised to provide even more precise and clinically valuable diagnostic information to guide appropriate treatment for patients with thyroid nodules. In addition to refining physicians’ ability to resolve indeterminate thyroid nodules, molecular testing will likely increase in use to inform optimal therapy and measure therapy response, among other uses.
As a physician who has been managing patients with thyroid nodules for over 15 years, I believe the opportunity to provide excellent patient care has never been greater than it is now.
Dr. Shank reported receiving research funding from Veracyle and he is on their speaker's bureau.
1. Haugen BR, et al. 2015 American Thyroid Association Management Guidelines for Adult Patients with Thyroid Nodules and Differentiated Thyroid Cancer. Thyroid. 2015; 26:1-202.
2. NCCN Clinical Practice Guidelines in Oncology: Thyroid Carcinoma. National Comprehensive Cancer Network. Available at http://www.nccn.org/professionals/physician_gls/pdf/thyroid.pdf. Version 2.2017 - May 17, 2017. Accessed 1/3/18.
3. Gharib H et al. American Association of Clinical Endocrinologists, American College of Endocrinology, and Associazione Medici Endocrinologi Medical Guidelines for Clinical Practice for the Diagnosis And Management of Thyroid Nodules – 2016 Update. Endocr Pract. 2016;22(Suppl 1).
4. Hegedus L 2004 Clinical practice. The thyroid nodule. N Engl J Med. 2004;351:1764–1771.
5. Baloch ZW, et al. Diagnostic terminology and morphologic criteria for cytologic diagnosis of thyroid lesions: a synopsis of National Cancer Institute Thyroid Fine-Needle Aspiration State of the Science Conference. Diagn Cytopathol. 2008;36:425-437.
6. Esfandiari NH, Hughes, D. T., Yin, H., et al, The effect of extent of surgery and number of lymph node metastases on overall survival in patients with medullary thyroid cancer. J Clin Endocrinol Metab. 2014;99:448–454,
7. Wormald R, et al. Hemithyroidectomy for benign thyroid disease: who needs follow-up for hypothyroidism? Clin Otoloaryngol. 2001;33(6):587-91.
8. Stoll SJ, et al. Thyroid hormone replacement after thyroid lobectomy. Surgery. 2009;146(4):554-560.
9. Singer J, et al. Impact of a gene expression classifier on the long-term management of patients with cytologically indeterminate thyroid nodules. CMRO. 2016;32(7):1225-1232.
10. Sosa JA, et al. Increases in thyroid nodule fine-needle aspirations, operations, and diagnoses of thyroid cancer in the United States. Surgery. 2013;154(6):1420-6.
11. Lubitz CC, et al, Microarray analysis of thyroid nodule fine-needle aspirates accurately classifies benign and malignant lesions. J Mol Diagn. 2006; 8:490–498; quiz 528.
12. Prasad NB et al, Identification of genes differentially expressed in benign versus malignant thyroid tumors. Clin Cancer Res. 2008;14:3327–3337.
13. Mazzanti C et al, Using gene expression profiling to differentiate benign versus malignant thyroid tumors. Cancer Res. 2004;64:2898–2903.
14. Alexander EK, et al. Preoperative Diagnosis of Benign Thyroid Nodules with Indeterminate Cytology. N Engl J Med. 2012;367:705-715.
15. Duick DS, et al. The Impact of Benign Gene Expression Classifier Test Results on the Endocrinologist–Patient Decision to Operate on Patients with Thyroid Nodules with Indeterminate Fine-Needle Aspiration Cytopathology. Thyroid. 2012;22(10):996-1001.
16. Lithwick-Yanai G, Dromi N, Shtabsky A, et al. Multicentre validation of a microRNA-based assay for diagnosing indeterminate thyroid nodules utilising fine needle aspirate smears. J Clin Pathol. 2017;70:500–507.
17. Nikiforov YE, et al. Impact of mutational testing on the diagnosis and management of patients with cytologically indeterminate thyroid nodules: a prospective analysis of 1056 FNA samples. J Clin Endocrinol Metab. 2011;96:3390–3397.
18. Nikiforov YE et al, Molecular testing for mutations in improving the fine-needle aspiration diagnosis of thyroid nodules. J Clin Endocrinol Metab. 2009;94:2092–2098.
19. Beaudenon-Huibregtse S et al, Centralized molecular testing for oncogenic gene mutations complements the local cytopathologic diagnosis of thyroid nodules. Thyroid. 2014;24:1479–1487.
20. Eszlinger M et al. Evaluation of a Two-Year Routine Application of Molecular Testing of Thyroid Fine-Needle Aspirations Using a Seven-Gene Panel in a Primary Referral Setting in Germany. Thyroid. 2017;27(3):402-411.
21. Nikiforov YE, et al. Impact of the Multi-Gene ThyroSeq Next-Generation Sequencing Assay on Cancer Diagnosis in Thyroid Nodules with Atypia of Undetermined Significance/Follicular Lesion of Undetermined Significance Cytology. Thyroid. 2015; 25(11):1217-1223.
22. Nikiforov YE, et al. Highly Accurate Diagnosis of Cancer in Thyroid Nodules With Follicular Neoplasm/Suspicious for a Follicular Neoplasm Cytology by ThyroSeq v2 Next-Generation Sequencing Assay. Cancer. 2014;120:3627-3634.
23. Valderrabano P, et al. Evaluation of ThyroSeq v2 performance in thyroid nodules with indeterminate cytology. Endocrine-Related Cancer. 2017;24, 127–136.
24. Pagan M, et al. The diagnostic application of RNA sequencing in patients with thyroid cancer: an analysis of 851 variants and 133 fusions in 524 genes. BMC Bioinformatics. 2015;7(1):61
25. Patel K. Clinical Validation of an Improved Genomic Classifier for Cytologically Indeterminate Thyroid Nodules Using an NGS Platform and Machine Learning Algorithms in an Independent Prospective Multicenter Blinded Cohort Demonstrates Improved Performance. Poster presented at: 3rd World Congress on Thyroid Cancer; July 2017; Boston, MA.
26. Angell TE. Clinical Validation of the Afirma Genomic Sequencing BRAF V600E Classifier. Poster presented at: 87th Annual Meeting of the American Thyroid Association; October 2017; Victoria, BC, Canada.
27. Sosa JA. Clinical Validation of the Afirma Genomic Sequencing Parathyroid Classifier. Poster presented at: 87th Annual Meeting of the American Thyroid Association; October 2017; Victoria, BC, Canada.
28. Randolph G. Clinical Validation of the Afirma Genomic Sequencing Classifier for Medullary Thyroid Cancer. Oral presentation at: 87th Annual Meeting of the American Thyroid Association; October 2017; Victoria, BC, Canada.
29. Steward D. Clinical Validation of ThyroSeq V3 Performance in Thyroid Nodules with Indeterminate Cytology: A Prospective Blinded Multi-Institutional Validation Study. Oral Presentation at: 87th Annual Meeting of the American Thyroid Association; October 2017; Victoria, BC, Canada.