Today it is possible to specifically evaluate the risk of malignancy of indeterminate thyroid nodules (Thy3 or Bethesda III or Bethesda IV) thanks to molecular genetics. There are several methods: some already validated, internationally recognized, patented and marketed and others experimental, under study and not yet validated. This article will discuss the latest version of the commercial Thyroseq method (Thyroseq V3), the most advanced molecular genetics method in the world.
EndocrinologiaOggi Center, in Rome (Italy), with consolidated experience in the field of thyroid molecular testing, is an Italian/European reference center for the execution of the most advanced molecular testing (Afirma GSX X Atlas, Mir-ThyPe, ThyGeNeXT + ThyraMIR and, most of all, Thyroseq V3).
INTRODUCTION
It is known that thyroid needle aspiration allows, in most cases, a certain diagnosis of the nature of a nodule (benign or malignant). However, in approximately 20% of cases, this distinction is not possible. This is the case of indeterminate thyroid nodules (Thy 3a or Thy3f, Bethesda III or Bethesda IV) in which the cells present in the benign nodule are indistinguishable from those found in the malignant one. In these “indeterminate” nodules, only the surgical removal of the nodule and the surrounding tissue (therefore through a real surgical intervention) and its analysis (histological examination) allows the definitive diagnosis of the nature of the nodule.
For these nodules, surgery is the only way to ensure the benignity of an indeterminate nodule.
It is also known that, after the operation, only 20% of the operated indeterminate nodules will be malignant and, instead, a about 80% of these are benign. In essence, 80% of patients undergo a surgical “unnecessary” thyroidectomy, with an undoubted over-treatment, performed with the sole purpose of obtaining a certain diagnosis.
The most difficult challenge for the endocrinologist, therefore, is to try to reduce the number of “unnecessary” diagnostic surgeries as much as possible. To this end, various molecular genetic methods have been developed, including the most reliable: Thyroseq V3.
This article aims to describe the Thyroseq V3 method, explaining its rationale and illustrating its evolution, available scientific data and potential.
PRELIMINARY CONSIDERATIONS
Some preliminary considerations:
– Absolute pre-operative certainty (meaning 100%) of benignity or malignity of an indeterminate Thy3 thyroid nodule cannot be given by anyone (with a few exceptions), not even by the most advanced molecular techniques. Therefore, one must be wary of those who guarantee certainties in this regard. In medicine, 100% does not exist. It does not exist for Thy2 nodules (benign), which are 98% benign, and even more so it does not exist for Thy3 indetermniate nodules. If 100% certainty is not achievable, it is desirable to get as close as possible by stratifying the risk of malignancy of the nodule using molecular genetics. The final decision on what to do is always the result of a careful evaluation of multiple anamnestic, clinical, instrumental, laboratory and genetic aspects, which must always be done in open sharing with the patient. This choice can be suggested, but it is always agreed with the patient who has the final say.
RATIONALE
Why is molecular genetics useful in indeterminate thyroid nodules?
To answer this question, you need to know that thyroid cancer is, like all tumors, a disease of the genome. The initiation and progression of thyroid cancer occur due to the accumulation of genetic and epigenetic alterations such as: somatic mutations, chromosomal rearrangements, alterations in gene expression, microRNA dysregulation, etc. These genetic alterations of thyroid cancer are now almost all known. Those discovered first concerned the MAP kinase (MAPK) and PI3K/AKT pathways. Among these, we recall the point mutations of BRAF (BRAFV600E) and RAS. These, together with the rearrangement of some genes (RET/PTC and PAX8/PPARγ), in addition to being the best known, are also the most frequent, representing, overall, approximately 70% of the mutations found in differentiated thyroid cancer. However, scientific research in recent years has made great progress, and the publication of the Cancer Genome Atlas has greatly reduced (from 25% to 3.5%) the percentage of thyroid cancer cases for which the genetic alterations are unknown.
Therefore, in indeterminate thyroid nodules it is possible to look for the presence of these genomic alterations to understand whether a TIR3 nodule is truly benign or malignant. This is the rationale for the use of molecular diagnostics.
COMMERCIAL METHODS AVAILABLE
It is essential to understand the difference between commercial and non-commercial tests. Non-commercial tests are molecular genetic tests usually performed at university centers, often within research protocols aimed at publishing scientific studies. These methods vary from center to center, since each center builds its own method based on its own possibilities and the technologies available to it. Therefore, non-commercial methods are very heterogeneous and difficult to compare with each other. Furthermore, non-commercial tests take “inspiration” from commercial tests, which are more studied and already scientifically validated, and propose the search for similar genetic alterations (but usually in smaller numbers) with methods that are usually less cutting-edge. Therefore, these are more obsolete methods, with a certain usefulness but certainly inferior to those on the market and, above all, with a validity that has yet to be demonstrated.
Furthermore, access to non-commercial molecular tests is often difficult for the patient, since it is variable, sometimes discretionary and not standardized, in addition to not having scientific and medico-legal validity. In fact, these, precisely because they are built for research and scientific publication purposes, do not yet possess the scientific strength typical of commercial methods which, instead, can count on scientific studies already published that confirm their applicability in daily clinical practice. A commercial test, in fact, already has the following characteristics:
-It must identify and precisely define the mutational panel or classifier used in a given clinical set (the only aspect in common with non-commercial tests).
– It must have validated the panel/classifier with a pilot study (an aspect that non-commercial tests do not always have).
– It must have been confirmed by subsequent clinical studies, preferably prospective, blinded, multi-center and independent. This last characteristic is exclusive only to molecular tests that have been marketed for some time, such as Afirma and Thyroseq.
Having said this, it follows that commercial molecular tests, also by virtue of their presumed greater scientific strength, also have a medico-legal value that non-commercial experimental tests, by force of circumstances, cannot yet have. Commercial molecular genetic tests currently available worldwide are:
Afirma GSX X Atlas (2012, Veracyte, San Francisco, California)
Thyroseq V3 (2014, CBLPath, New York and University of Pittsburgh, Pennsylvania)
ThyGeNeXT+ThyraMIR (2015, Interpace Diagnostics, Parsippany, New Jersey)
Mir-THYpe (ONKOS Diagnosticos LDTA molecular, Ribeirao Preto, Brazil)
Plus ThyroPrint (GeneproDX, Santiago, Chile)
The first three are patented in the United States. The last two, with less scientific literature, and, at the time of publication of this article, not yet certified by the American College of Pathologists, were developed in South America. Afirma is currently no longer available in Europe.
Finally, it should be remembered that at the moment, the commercial test currently available in Europe for clinical use is Thyroseq V3, which can be performed at the Centro EndocrinologiaOggi in Rome, a reference center in Italy for thyroid molecular genetics.
TYPES OF AVAILABLE TESTS
In detail, molecular genetic tests can be divided into: exclusion test (rule out) or inclusion test (rule in).
An exclusion test is a test that, in the context of an indeterminate Thy3 thyroid nodule, is able to state that a nodule is definitely benign, to exclude malignancy and, therefore, to exclude surgery (hence the term exclusion test). An exclusion test, to identify nodules that are definitely benign, requires a high negative predictive value (NPV) (at least 94% as in the case of standard needle aspiration for Thy2 nodules).
An inclusion test, on the other hand, is a test that, in the context of an indeterminate Thy3 thyroid nodule, is able to identify a nodule that is definitely malignant, to indicate its malignancy, and therefore to strongly suggest surgery (hence inclusion test). An inclusion test, to identify nodules that are definitely malignant, requires a very high positive predictive value (PPV) (at least 98% as in the case of standard needle aspiration for Thy5 nodules).
Among the commercial tests currently available, Thyroseq V3 is the best because it works well as both an exclusion and inclusion test. Obviously, for non-commercial tests, in the absence of sufficient scientific validation, it is not possible to say whether they are exclusion or inclusion tests.
Over time, with the implementation and with the partial fusion between some of these commercial methods, the tests on the market have increasingly better diagnostic performances and increasingly higher and more similar predictive values. In essence, it can be said that, with the exception of a few decimal points, these methods now resemble each other more and more even in terms of accuracy and reliability.
PREDICTIVE VALUE
To fully understand the result of any test, however, it is also necessary to clarify the meaning of some statistical terms, in particular the Predictive Value. This can be: negative predictive value (NPV) or positive predictive value (PPV).
The negative predictive value (NPV) of a molecular test indicates, after this test has classified an indeterminate thyroid nodule as benign, what is the possibility (post-test) that this nodule is actually benign.
In contrast, the positive predictive value (PPN) of a molecular test indicates, after this test has classified an indeterminate thyroid nodule as malignant, what is the possibility (post-test) that this nodule is actually malignant.
Let’s clarify the importance of this data with an example. In a given context, a molecular test (Thyroseq, for example) has a very high negative predictive value (NPV) (97%). This means that if an indeterminate thyroid nodule is classified as benign by Thyroseq, then this nodule has a very high probability of actually being benign (not 100% but 97%, substantially equal to that of the Thy2 nodule which, remember, indicates a benign formation (98%) but never 100%). Obviously, for non-commercial tests it is not possible to provide certain information on the predictive value.
Furthermore, it should be remembered that, unlike sensitivity and specificity which are intrinsic characteristics of a test, the predictive value can vary depending on the prevalence of thyroid cancer in the specific population studied.
Therefore, for a correct interpretation of the molecular test and to provide correct surgical indications, it would be desirable to also know the pre-test risk of neoplasia for each cytological category, which is difficult to know for each individual patient. For this reason, it is advisable to at least stratify the pre-test risk of neoplasia of the patient’s nodule to be subjected to molecular techniques by evaluating:
– family history of thyroid cancer.
– positive history of radiation exposure.
– evaluation of cytological characteristics (for example nuclear atypia, which increases the risk of malignancy).
– ultrasound characteristics (marked hypoechogenicity, microcalcifications, irregular margins, etc.)
With a correct evaluation (and interpretation) of this information, it will therefore be possible to identify the right patient with an indeterminate thyroid nodule to whom the most suitable molecular genomic test can be suggested.
This pre-test analysis, in addition to providing a reference line for a correct indication for the tests, also allows you to make the most of the potential of each molecular method, ensuring greater reliability in the interpretation of the results.
METHODS
Modern molecular methods can evaluate gene expression or DNA. In the first case, molecular genetics evaluates how the various genes in DNA are differently expressed (see specific article). In the second case, DNA is evaluated because, as already mentioned, the DNA alterations of differentiated thyroid cancer are now almost all known (only 3.5% of thyroid tumors have unknown mutations). BRAFv600E, RAS and rearrangements (RET/PTC and PAX8/PPARγ) represent approximately 70% of the detectable mutations. However, each alteration, taken individually, has sensitivity and specificity too low to be clinically relevant in the indeterminate nodule (for example, the prevalence of BRAF mutations in neoplastic indeterminate lesions is only 4.6%).
For this reason, Nikiforov was the first to suggest that by putting them together and analyzing a panel of multiple mutations (BRAF, RAS, RET/PTC, PAX8/PPARg) it was possible to obtain an increase in sensitivity (44% -> 80%) and accuracy (from 93.3% -> 97.4%) of genomic analysis. Starting from this scientific publication, a first commercial molecular test on DNA was created, called “miRInform” (2009), later renamed and replaced by the current ThyGeNext, which evaluated, with PCR, seven gene alterations (BRAF, KRAS, HRAS, NRAS, RET/PTC1, RET/PTC3, PAX8/PPARg). The advent of next-generation sequencing technologies (Next Generation Sequencing, NGS) capable of sequencing and simultaneously discovering genetic alterations in specific areas (for example in thyroid cancer), then allowed the development of the first version of Thyroseq. Unlike miRInform, the DNA study with Thyroseq was performed by NGS (rather than PCR), and in addition to the seven alterations searched for by “miRInform”, new mutational drivers of more recent discovery were also identified (PIK3CA, TP53, TSHR, PTEN, RET, AKT1, CTNNB1, TERT) as well as other gene fusions (BRAF, RET, NTRK1, NTRK3, AKT1, PPARG, THADA). As scientific discoveries on the genetics of thyroid cancer have progressed, the Thyroseq method has undergone further improvements and implementations.
The material to be analyzed must usually be collected from the nodule by needle aspiration and must be collected in specific tubes with preservative solution and shipped in refrigerated boxes or stored at -20 ° C.
Alternatively, it is also possible to collect the material directly from the slides of a previous fine needle aspiration, provided that these contain sufficient material and specific characteristics that must be evaluated on a case-by-case basis (and it must be considered that the extraction process will then cause the definitive destruction of the slides).
The Thyroseq V2 version, also based on NGS, allowed the sequencing of 56 genes involved (analyzed for point mutations, gene fusions and anomalous gene expression).
Finally, we arrived at the latest version (Thyroseq V3) that analyzes an even higher number of genes (112 genes) and was developed with the aim of: a) improving the ThyroSeq v2 panel by including more recently discovered genetic markers; b) analyzing new classes of previously untested genetic alterations, such as copy number alterations (CNA); c) improving its accuracy in Hurthle cell nodules (oxyphils).
Thus was born the latest version, currently also on the market in Italy, called, precisely, Thyroseq V3.
THYROSEQ V3
Like previous versions, Thyroseq V3 was validated first by a single-center study and then by an international multicenter study.
Single-center validation study
The latest version (Thyroseq V3) was validated in a single-center study published in Cancer in 2018, where it was validated on both 238 surgical tissue samples and 175 fine-needle aspirates of indeterminate nodules.
ThyroSeq V3 evaluates 112 genes (including all genes tested by ThyroSeq V2) by NGS sequencing, which with targeted amplification is able to detect 12,135 single nucleotide variations and insertions/deletions, more than 120 types of gene fusions, abnormal gene expression alterations of 90 genes, and copy number alterations in 10 genomic regions in fine-needle aspirate samples. Chromosomal copy number variations are very important, as they are found in 7% of papillary carcinomas without other mutations and in other types of tumors such as Hurthle cell tumors.
It is therefore a large and highly complex tool. The complexity of data analysis made it necessary to create a Genomic Classifier (GC) to classify the test result as negative (probably benign) or positive (probably malignant).
In particular, DNA and mRNA are isolated from the fine needle aspiration sample collected in tubes containing a specific nucleic acid preservation solution (collection tubes).
A score (from 0 to 2) is assigned to each genetic alteration found based on the strength of its association with thyroid malignancy: 0 (no association with cancer); 1 (low probability of cancer); 2 (high probability of cancer). The assigned score derives from: a) a review of the data present in the literature and in the available databases (TCGA, cBioPortal, COSMIC, etc.); b) a review of an internal university database with more than 1000 samples of thyroid tissue or fine needle aspirations with known histological diagnosis; c) RNA analysis in fine needle aspiration samples; d) CytoScan analysis of 17 samples of thyroid tumor tissue.
The final total score for each sample is calculated as the sum of the individual values of the detected genomic alterations (GC score = (xSNV/I)n+xGF+xGEA+xCNA; x = weighted value 0 – 2; n=number of SNVs/Indels; SNV/I, GF, GEA, CNA are indicators of genomic alteration type).
This single-center ThyroSeq v3 study showed a sensitivity of 98.0% and a specificity of 81.8% and above all a high analytical performance, even in blood samples. In fact, it is sufficient for the sample to contain just 2.5 ng of nucleic acids and 12% of thyroid cells for a correct classification of the indeterminate nodule.
This study highlighted a good performance of Thyroseq also in NIFTP (noninvasive follicular thyroid neoplasm with papillary-like nuclear features), follicular adenomas and oxyphilic adenomas (Hurthle cell adenomas) which represent possible precancerous lesions.
Multicenter validation study
After the single-center study, Thyroseq V3 was recently validated by an international multicenter study published in JAMA in 2019.
The genomic evaluation was performed centrally in the United States, at the UPMC Laboratory Molecular and Genomic Pathology, University of Pittsburgh, Pennsylvania.
This study confirmed that Thyroseq V3 has a high sensitivity (94%) and specificity (82%) and above all a negative predictive value (NPV) of 97%. This means that a nodule found to be benign with Thyroseq has a very low residual risk of cancer (3%). This risk is similar to the average risk of cancer (2%) found in nodules diagnosed as benign (Thy2) by fine needle aspiration (NPV value of fine needle aspiration in case of benignity: 98%). The study confirms the validity of Thyroseq V3 also in oxyphilic nodules. (book a thyroid fine needle aspiration)
This study also showed that the few false negatives of Thyroseq V3 (i.e. nodules that, despite being tumors, were classified by Thyroseq as probably benign) were mainly low-risk, non-aggressive tumors.
Although sensitivities are similar, compared to Afirma GSC, Thyroseq actually seems to have a better specificity (82% vs 68%) and in general seems to be able to avoid surgery in 61% of patients with indeterminate nodules.
Finally, Thyroseq V3 has another potential advantage: it provides a molecular profile of the indeterminate nodules that test positive and, therefore, is useful for redefining the risk (and consequently the treatment) of patients with these nodules.
The nodules that are positive according to Thyroseq V3, based on the type of mutations they express, can be divided into nodules with a BRAF-like profile (usually with more aggressive behavior) or with a RAS-like profile (usually with less aggressive behavior).
The detection of some mutations (BRAF V600E, TERT, TP53), is very specific for papillary thyroid carcinoma, conferring an almost certain probability of tumor (100%). Tumors that express the BRAF V600E mutation are usually classic papillary carcinomas but with a higher risk of locoregional lymph node metastases.
The detection of RAS or RAS-like mutations or PAX8/PPARg fusions, however, does not always indicate with certainty the tumor nature of the lesion. RAS mutations, in fact, can also be found in benign nodules. Therefore, with RAS or RAS-like mutations the risk of tumor is only 60%, since these are present in a varied spectrum of thyroid lesions with a follicular pattern (which varies from benign adenomas, to scarcely aggressive tumors such as NIFTP to invasive follicular carcinomas). As mentioned, generally, the majority of thyroid tumors with RAS or RAS-like mutations are encapsulated, minimally invasive and low risk. If they metastasize, they usually spread hematogenously, skipping the local-regional lymph nodes.
The study highlights that the few false positives to Thyroseq V3 were almost all nodules that, although benign at histological examination, still expressed a RAS or RAS-like mutation. These were, therefore, monoclonal formations, therefore potentially a little more at risk and very different from the classic (and more frequent) hyperplastic nodules which, instead, are polyclonal.
This additional information on the nodules positive to the test, interpreted together with clinical and ultrasound parameters can be of great help in personalising the treatment of the patient with an indeterminate nodule.
Conclusions
Thyroseq V3, the latest and most advanced version of Thyroseq, is probably the most powerful molecular method, which seems to work well both as an exclusion test and as an inclusion test. Thyroseq V3 can be performed at the Centro EndocrinologiaOggi in Rome, Italy. (Book the Thyroideq V3 molecular testing)
Dr. Massimiliano Andrioli
Specialist in Endocrinology
Centro EndocrinologiaOggi, Roma
viale Somalia 33A, Roma
tel/fax 0686391386
cell 3337831426
Studio EndocrinologiaOggi, Lecce
via Ruffano 4, Casarano (Lecce)
tel/fax 0686391386
Bibliography
– Cancer Genome Atlas Research Network. Integrated genomic characterization of papillary thyroid carcinoma. Cell. 2014;159(3):676-690.
– Nikiforov YE, Steward DL, Robinson-Smith TM, et al. Molecular testing for mutations in improving the fine-needle aspiration diagnosis of thyroid nodules. J Clin Endocrinol Metab. 2009;94(6):2092-2098.
– Nikiforov YE, Carty SE, Chiosea SI, 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(23):3627-3634.
– Nikiforov YE, Carty SE, Chiosea SI, 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.
– Nikiforova MN, Mercurio S,Wald AI, et al. Analytical performance of the ThyroSeq v3 genomic classifier for cancer diagnosis in thyroid nodules. Cancer. 2018;124(8):1682-1690.
– Stewart D et al. Performance of a Multigene Genomic Classifier in Thyroid Nodules With Indeterminate Cytology A Prospective Blinded Multicenter Study JAMA Oncol. 2019;5(2):204-212. doi:10.1001/jamaoncol.2018.4616
-Nishino M. Molecular cytopathology for thyroid nodules: A review of methodology and test performance. Cancer Cytopathol. 2016 Jan;124(1):14-27. doi: 10.1002/cncy.21612.
-Nishino M, Nikiforova M. Update on Molecular Testing for Cytologically Indeterminate Thyroid Nodules. Arch Pathol Lab Med. 2018 Apr;142(4):446-457. doi: 10.5858/arpa.2017-0174-RA.
-Ferraz C. Can current molecular tests help in the diagnosis of indeterminate thyroid nodule FNAB? Arch Endocrinol Metab. 2018;62(6):576-584. doi: 10.20945/2359-3997000000081.
– Andrioli M, Carocci S, Alessandrini S, Amini M, Van Doorne D, Pace D, Lauria A, Raffaelli M, Trimboli P. Testing for Afirma in Thyroid Nodules with High-Risk Indeterminate Cytology (TIR3B): First Italian Experience. Endocr Pathol. 2020 Jan 20. doi: 10.1007/s12022-020-09604-7.
APR
2025