NEW YORK—Although the most appropriate uses of molecular testing in clinical management of thyroid cancer are still being assessed, there is consensus that such testing is a valuable tool—if and when it is used correctly.
That was the takeaway message from a panel here at the World Congress of the International Federation of Head and Neck Oncologic Societies and the Annual Meeting of the American Head and Neck Societies.
“Molecular testing is the most important advance in diagnostic thyroid testing in the last 30 years,” said Robert Witt, MD, FACS, Professor of Otolaryngology-Head & Neck Surgery at Thomas Jefferson University and Director of the Head & Neck Multidisciplinary Clinic at the Helen F. Graham Cancer Center, speaking in an interview after the meeting.
For the 20 to 25 percent of thyroid nodules that generally in the past have gone on for surgery because cytology has deemed them indeterminate, molecular testing can give a more precise answer to whether or not the lesions are malignant, he explained. “I think you're going to see a tidal wave change in terms of management of indeterminate thyroid nodules in the immediate years.”
Still, the session's moderator, Maisie L. Shindo, MD, Professor of Otolaryngology-Head and Neck Surgery and Director of the Thyroid and Parathyroid Center Oregon Health & Science University, cautioned that while the development of such tests gives patients more options, “we have to be very cautious about using these tests properly.
“They are not inexpensive, and the proper tests need be ordered to get the right information. And you really have to factor in clinical assessment—examining the patient, getting a history, and reviewing the patients' symptoms—that's one of the most important points.”
Shindo, Witt, and the other panelists discussed appropriate uses of molecular testing in thyroid cancer in the following case studies.
A Case for a Gene Expression Classifier
The first case study was of a 36-year-old man who had a thyroid nodule discovered by his primary care provider. The patient did not have compressive symptoms; there was no history of radiation exposure; and he had no family or personal history of thyroid malignancy. His clinical exam found a firm, three-centimeter palpable left thyroid nodule, but no palpable lymphadenopathy. And the right nodule was normal.
The primary care physician ordered a fine needle aspiration (FNA), and the report was follicular neoplasm (Bethesda category 4) with significant vascularity. Thyroid-stimulating hormone (TSH) level was 2.4 mIU. Based on those biopsy results, the patient was sent to the head and neck surgeon.
Panelist Julie Ann Sosa, MD, Professor of Surgery and Medicine and Chief of the Section of Endocrine Surgery at Duke Cancer Institute and Duke Clinical Research Institute, said that with the patient having no contralateral nodularity, she would perform ipsilateral thyroidectomy and not resection.
But, Witt said he would consider this case a “perfect” occasion to consider using a gene expression classifier test. “Research has shown that for the gold standard—histology—compared with a gene expression classifier, you're going to have a 94 percent negative predictive value for a follicular neoplasm,” he explained during the session, citing a paper previously published in the New England Journal of Medicine (2012;367:705-715).
The FNA cytology would also yield a 94 to 95 percent negative predictive value if the results of that test were benign, he noted. And according to the 2013 National Comprehensive Cancer Network guidelines: “If you have a genetic test that is equivalent to a benign FNA cytology, it is very reasonable to do that test and observe that patient, rather than do an operation.”
The point both panelists agreed on was that before either course was taken, discussion of either option—and the risks and benefits associated with each—was needed.
“You need to have a good informed consent with that patient. This [molecular testing data] is relatively new data—but very solid data—and so you need to get a sense of the patient's opinion,” Witt said. Lifestyle factors, such as the patient's profession, are significant, he said: “For example, if the patient is a choir director in a high school and surgery leaves him with one percent vocal cord inability, you've ruined his life.”
Sosa noted: “I absolutely agree that these conversations need to be sophisticated and extensive—discussing the risks and benefits of different approaches.”
And, if molecular testing is going to be pursued, it would need to be repeated, Sosa added. And, age, continued need for surveillance, and the potential increase in size of this nodule are all factors that should be discussed with the patient and weighed against the risks of surgery.
“The take-home message is that in the past the recommendation would have been to take the thyroid out for definitive diagnosis,” Shindo said. “But today, there are additional tests—in this case the gene expression classifier—that can be performed on the biopsy looking at molecular markers to further determine the actual risk of cancer in this nodule.”
Patients like the one in this case study should be evaluated by head and neck endocrine surgeons who focus on thyroid and parathyroid surgery who are knowledgeable about these molecular tests and can properly order or perform them, Shindo added. “Patients should know their options, as well as the limitations of these tests.”
A Case for Molecular Panel Testing
Another case Shindo presented was that of a 31-year-old man who had undergone ultrasound—prompted by his having a very strong family history of thyroid cancer—which showed he had thyroid nodules. The patient was asymptomatic and had had no radiation exposure. He had had an osteochondroma of the humerus, which had not been treated but was being monitored. And his father, mother, sister, and brother had all had papillary thyroid cancer. The patient's thyroid was not enlarged, and he had no lymphadenopathy.
The patient's ultrasound, showed a right-side hyperechoic thyroid nodule: not suspicious appearing. And there was a nine-millimeter nodule on the contralateral lobe, which was suspicious for papillary thyroid cancer. The major discussion with the patient for this case was whether to proceed with surgery and if so, what should the extent of surgery be, Shindo said.
“Clearly this is a patient who appears to meet criteria that justifies suspicion for familial papillary thyroid cancer, which makes the case different from cases of sporadic papillary thyroid cancer. All the lesions seen are relatively benign in appearance. Without the family history, I would not have biopsied this patient. But with the family history—what I suspect is a high index of concern and anxiety—I expect a biopsy may be in his future.”
Shindo added that she would follow the American Association of Thyroid Cancer Guidelines (disclosing that she was a member of that guideline's writing committee), which suggest that for a patient suspected to have papillary thyroid cancer based on the Bethesda classification showing a malignancy rate of 60 to 75 percent—the management of that suspected cancer should be treated as if it were papillary thyroid cancer.
“I would approach this patient feeling pretty certain he has papillary thyroid cancer—and I'm not sure that additional testing would necessarily change my management,” Shindo said. “For micropapillary thyroid cancers, I absolutely think that lobectomy is adequate but this is potentially familial papillary [thyroid cancer], where I'd be concerned about bilateral disease.”
Witt, however, said he would instead call for a molecular alteration testing panel for the patient that included BRAF, RAS, RET, and Gamma. Based on the patient's family history and the Bethesda 5 classification, surgery will be needed, so the genetic testing results could help inform the extent of surgery needed, he explained.
“If the nodule is BRAF positive, you know the patient definitely has thyroid cancer. Your decision then is not am I going to do lobectomy or not, but rather am I going to do lobectomy or am I going to do total thyroidectomy?”
He added that currently there are still no good prospective studies to suggest that molecular alteration testing should change management or change prognosis for a particular patient—but, recent research showing that one out of five BRAF tumors will act aggressively, along with the patient's family history would influence his opinion on whether or not he would do total thyroidectomy—a surgery that he said under other circumstances he would not consider for a one-centimeter papillary thyroid carcinoma.
The point of this case, Shindo explained, is that given the patient's very strong family history of thyroid cancer—that makes it very likely the patient could carry various mutations that could put him at a high risk of thyroid cancer—the molecular panel test is one that may be very helpful in confirming whether or not the patient has cancer and determining the extent of surgery.
The downside of using molecular tests, though, she noted, is the potential for false negatives. Long-term monitoring in serial studies is needed, and given the current evidence, it would not be “wrong” to proceed with surgery for either of these particular patients, she said. “But the molecular marker tests could potentially allow the patient to avoid surgery.”
Ongoing Molecular Testing Research
Data from two new studies investigating genetic testing in thyroid cancers were also presented during the meeting:
Computer-based Protein Modeling
Computer-based protein modeling was found to help predict clinical outcomes for patients with multiple endocrine neoplasia (MEN) type 2, a rare genetic disease known to be a strong predictor for developing medullary thyroid cancer, based on mutation alone, according to data presented during an oral session (Abstract S450).
Previous research identified a relationship between a distinct mutation in the RET gene and MEN2 disease. This study used software that measured the variation in the amino acid sequence of the RET protein, as well as the biochemical difference between mutant and wild-type proteins and the resultant alteration structure by biophysical modeling. The combination of those measures predicts the amount of structural difference observed with a given mutation.
The data showed that RET mutations associated with a more aggressive clinical phenotype (according to groups defined by the American Thyroid Association guidelines for prophylactic thyroidectomy in MEN2) generally had higher scores as measured by the software.
“We were able to determine whether the mutation predicted phenotype,” the study's senior author, David I. Kutler, MD, Associate Professor at Weill Cornell Medical Center and New York Presbyterian Hospital, explained in an interview. “By looking at how badly the protein was disrupted in the mutated gene, we could determine how aggressive the cancer would be—which is kind of novel.”
The implications in terms of managing these patients is being able to better predict which patients' tumors will act more aggressively and therefore should be treated more aggressively, Kutler said. He added that in the longer term, the hope is that the same software might also be useful for looking at other types of cancer and other diseases.
New Gene Signature for Indeterminate Thyroid Nodules
Another group of researchers presented a study in which they developed a tool—a combined gene expression and mutation assay—to determine the presence or absence of malignancy in thyroid fine needle aspiration (FNA) with indeterminate cytology (Abstract P1161).
The tool uses the same strategy as the gene expression classifiers that are currently used, but tests only nine genes (compared with the Afirma gene expression classifier, which tests 142) and so is estimated to be much more cost-effective.
“Patients with indeterminate or inconclusive cytological thyroid biopsies are almost always led to surgery based on inconclusive criteria,” the study's lead author, Sotiris Nikolopoulos, PhD, Founder and Scientific Director of the Center for Molecular Analysis and Research SA in Agrinio, Greece, explained via email. “A molecular test based on our findings would reveal the real picture of a suspicious specimen (either malignant or not malignant), thus avoiding unnecessary surgeries for some patients or confirming malignancy (i.e., the need for surgery) for others.”
The study included 121 patients with thyroid nodules who had FNA biopsies under ultrasound guidance. The researchers examined FNA cytology for the nodules and transferred an extra aspirate of the nodule into a tube containing a nucleic acid stabilization buffer, which was then processed for gene expression analysis.
The team evaluated nine genes encoding receptor tyrosine kinases and other oncogenes, as well as the presence of the BRAF V600E mutation.
There was agreement between the gene signature and the cytology assessment for 117 of the 121 cases.
The findings, Nikolopoulos noted, could eventually change how indeterminate thyroid nodules are diagnosed. The next step needed to put the method to clinical use is multicenter trial validation, which is expected to be completed in the next six to 12 months, he said.
“I think the time has come for molecular testing to be introduced into the clinic, not as a substitute but as an extension of the traditional cytological examination of a specimen.”
Other Experts Said...
For more information about the role of genetic testing in thyroid cancer, here's what experts told OT in an article published in the 5/25/14 FOCUS: Thyroid Cancer supplement: