Molecular Alterations in Thyroid Carcinoma

Introduction

Most thyroid carcinomas can be divided into two broad categories based on their cell of origin: the vast majority (more than 95%) are of follicular cell origin whereas the remaining 3 to 5% is medullary thyroid carcinoma arising from C cells. Follicular cell-derived carcinomas can be further divided into papillary thyroid carcinoma (PTC), follicular thyroid carcinoma (FTC), Hurthle cell carcinoma (HCC), poorly differentiated thyroid carcinoma (PDTC), and anaplastic thyroid carcinoma (ATC)¹. PTC, FTC, and HCC are considered well-differentiated thyroid carcinomas. PTC, the most common type of thyroid carcinoma, is composed of a heterogeneous group with over 10 phenotypes (i.e. variants). Of them, classic variant (CVPTC), follicular variant (FVPTC), and papillary microcarcinoma are the most common variants ¹.

The past decade has witnessed significant progress in the understanding of molecular pathogenesis of thyroid carcinoma, based on studies using next-generation sequencing (NGS) platform ^2–8. In 2014, The Cancer Genome Atlas (TCGA) reported the comprehensive genomic characteristics of PTC ⁵; Ninety-seven percent of PTCs have unique molecular alterations. This includes 74% with single nucleotide variants (for example BRAF or RAS mutation), 15% with fusions, 7% with arm-level copy number alterations, and 1% with deletions. Merely 3% of PTC is characterized as “dark matter”- cases in which the molecular events underpinning the tumorigenesis remains to be discovered. Subsequently, the genomic characteristics of HCC, PDTC, and ATC have also been reported ^2–4. The aim of this review is to provide a concise summary of the molecular pathogenesis of thyroid carcinoma, focusing primarily on recent advances.

Stepwise tumorigenesis in follicular-cell derived thyroid carcinoma

Early molecular events: driver mutations in the mitogen-activated protein kinase (MAPK) pathway

The key molecular alterations in various types of thyroid carcinoma are summarized in Table 1. Constitutive activation of the MAPK signaling pathway plays a central role in the carcinogenesis of thyroid carcinoma. The essential proteins in this pathway are receptor tyrosine kinases (RTKs, which includes VEGFR, RET, ALK, and TRK), RAS, RAF, MEK, and ERK (Figure 1). RAS is a small G-protein that is recruited upon binding of growth factors to RTK and it, in turn, activates downstream pathways, via serine/threonine kinase BRAF. The cascade of downstream events in this pathway ultimately leads to altered cell proliferation, differentiation, and survival, resulting in various forms of thyroid carcinoma. In thyroid cancer, genetic alterations in the MAPK pathway are highly prevalent and mutually exclusive to each other. Based on the cBioportal data from the TCGA PTC cohort, alteration of the MAPK pathway is detected in 83% of all PTCs tested ^5,9. The common alterations are BRAF mutation (predominantly V600E) in 62%, RAS (including HRAS, NRAS, and KRAS) mutation in 13%, RET-PTC rearrangement in 6%, and less frequently NTRK3 fusion in 1.5%, NTRK1 fusion in 1.3%, and ALK fusion in 0.8%.

Table 1.

Key molecular alterations in thyroid carcinoma

Tumor types	Common molecular alterations	Uncommon molecular alterations
PTC	BRAF (62%), predominantly BRAFV600E RAS (13%) RET-PTC (7%) TERT promoter mutation (9%)	E1F1AX ALK fusion NTRK1 or NTRK3 fusion
Encapsulated FVPTC and NIFTP	RAS (30–52%) PAX8-PPARγ(0–38%) THADA fusion (0–22%)	BRAFK601E (3–7%) Absence of BRAFV600E
FTC	RAS (49%) PAX8-PPARγ(30–58%) TERT promoter mutation 17%	TSHR mutations BRAFK601E E1F1AX
HCC	Widespread chromosomal losses Alteration of mitochondrial genome RAS (9–15%) TERT promoter mutation (22–27%) TP53 mutation (7–12%)	CHCHD10-VPREB3 HEPHL1-PANX1 TMEM233-PRKAB1
PDTC	BRAF 33% RAS 45% TERT promoter mutation (40%) TP53 mutation (10%)
ATC	BRAF 29% RAS 23% TERT promoter mutation (73%) TP53 mutation (59%)	Tumor suppressors: ATM, RB1, MEN1, NF1, and NF2 Mutation affecting PIK3CA-AKT-mTOR pathway, mismatch repair genes, SWI-SNF complex, and histone methyltransferase pathway
MTC	RET (40–60%) RAS (up to 20%)	MET ALK fusion

FVPTC: follicular variant of papillary thyroid carcinoma, NIFTP: noninvasive follicular thyroid neoplasm with papillary-like nuclear features, PTC: papillary thyroid carcinoma, FTC: follicular thyroid carcinoma, HCC: Hürthle cell carcinoma, PDTC: poorly differentiated thyroid carcinoma, ATC: anaplastic thyroid carcinoma, MTC: medullary thyroid carcinoma.

Figure 1. Molecular pathogenesis of thyroid carcinoma.

(A) Constitutive activation of MAPK pathway through mutations or fusions of its key components, including receptor tyrosine kinase (RTK, e.g. RET, NTRK1, NTRK3, and ALK), RAS, and BRAF, plays a central role in tumorigenesis of thyroid carcinoma. Various tyrosine kinase inhibitors, e.g. multiple kinase inhibitors sorafenib and lenvatinib and BRAF inhibitors vemurafenib and dabrafenib, have been investigated in treating advanced thyroid carcinoma. (B) There is a strong genotype-phenotype correlation, in which papillary thyroid carcinoma (PTC) classic variant and tall cell variant typically harbor BRAF ^V600E hotspot mutation, whereas follicular-patterned carcinomas, i.e. follicular carcinoma and encapsulated follicular variant of papillary thyroid carcinoma, are usually RAS-mutated. Inserts: Nuclear features of follicular carcinoma and follicular variant of papillary thyroid carcinoma. Arrows: capsular invasion.

Key points:

1. Thyroid carcinoma is characterized by molecular alterations in MAPK pathway.

2. BRAF ^V600E mutation is the most common mutation in papillary thyroid carcinoma.

Genotype-phenotype correlation

Interestingly, the TCGA study has shown that there is a strong correlation between histologic phenotypes and underlying genotypes. A review of cBioPortal TCGA data shows that the BRAF ^V600E mutation is highly prevalent in tall cell variant (TCV) and CVPTC with a frequency of 89% and 67% respectively. Meanwhile, a RAS mutation was detected in only 6% of CVPTCs and in none of the TCVs tested. In contrast, FVPTC has a high frequency of RAS mutations (38%), and infrequent BRAF ^V600E mutation (13%) ^5,9. Two distinct molecular groups have emerged based on the mutation status of BRAF ^V600E and RAS: the BRAF ^V600E-like and the RAS-like groups. The BRAF ^V600E-like group characterized by robust MAPK pathway activation, less differentiation, and dampened response to radioactive iodine (RAI), is enriched with tall cell and classical variants; whereas the RAS-like group shows activation of both MAPK and PI3K/AKT signaling pathways, is highly differentiated and contains mostly FVPTC ⁵. This genotype-phenotype correlation has been confirmed by multiple additional studies ^10–15. PAX8-PPARγ rearrangement, a common molecular event in thyroid carcinoma, is detected predominantly in follicular-patterned carcinoma, including 30–58% of FTC and 38% of FVPTC, whereas it is quite rare in CVPTC and TCV, accounting for less than 1% of all CVPTC and none of the TCV tested in the TCGA PTC cohort ^16–18. Therefore, it is now clear that TCV and CVPTCs are enriched with BRAF ^V600E mutation and/or RET-PTC fusion, whereas the FVPTC is associated with a high frequency of RAS mutations and the PAX8-PPARγ rearrangement, a low frequency of BRAF ^V600E mutation, and a pattern of chromosomal gains/losses as well as a protein expression profile akin to FTC and follicular adenoma.

FVPTC can be further classified into infiltrative and encapsulated forms based on the absence or presence of a complete tumor capsule or well-circumscribed tumor border. The molecular profile and clinical behavior of infiltrative FVPTC resemble those of CVPTC, characterized by a 36% rate of BRAF ^V600E mutation or RET-PTC fusion and 65% risk of nodal metastasis. Meanwhile, the encapsulated FVPTCs, especially those that show no evidence of invasion, are enriched in RAS mutations, lack BRAF ^V600E mutations, and have a negligible risk of lymph node metastasis and recurrence ^19,20.

Inspired by the observation that noninvasive encapsulated FVPTC is highly indolent, with an extremely low risk of lymph node metastasis and/or recurrence, and has a molecular profile distinct from CVPTC, in 2016 Nikiforov et al. revised the terminology to noninvasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP). Molecular profiling performed in a subset of the consensus cohort as well as subsequent studies on NIFTP consistently show that NIFTP lack the BRAF ^V600E mutation and are instead associated with RAS mutations in 30–52%, the BRAF^K601E mutation in 3–7%, and PPARγ and THADA fusions in 0–22% of cases^21–23.

It is now clear that encapsulated follicular-patterned neoplasms, a group of encapsulated thyroid neoplasms with exclusive follicular growth pattern and absence of true papillae, are molecularly similar, enriched in RAS mutations and PAX8-PPARγ fusion. This group includes FA, FTC, encapsulated FVPTC with capsular and/or vascular invasion, and NIFTP. Without invasion (i.e. FA and NIFTP), they follow a highly indolent course with negligible risk of regional spread and recurrence. With invasion (i.e. FTC and encapsulated FVPTC), they have a propensity to spread distantly rather than to regional lymph nodes.

Prior to the birth of NIFTP, multiple studies, including a large meta-analysis of 5655 PTC patients, showed that BRAF mutation is an adverse molecular signature in PTC; associated with advanced stage, high frequency of nodal metastasis, and increased risk of extrathyroidal extension and recurrence ^24,25. As the implementation of NIFTP nomenclature exempts a proportion of RAS-related highly-indolent tumors from a frank diagnosis of carcinoma, the prognostic significance of BRAF mutation in PTC needs to be re-evaluated in the post-NIFTP era.

Follicular thyroid carcinoma

Given the preceding discussion on follicular-patterned carcinoma, it is no surprise that FTC is characterized by RAS point mutations and PAX8-PPARγ fusion, detected in 49% and 30–58% of cases, respectively ^18,26–29. TERT (telomerase reverse transcriptase) promoter mutation has been detected in 17% of FTCs, a frequency that is higher than that seen in PTC (9%) ^5,30. TERT promoter mutation is generally considered an aggressive molecular signature in thyroid carcinoma, and its significance is further elaborated in a later section.

Hürthle cell carcinoma: a unique thyroid cancer characterized by widespread chromosomal losses and mitochondrial DNA mutations

HCC is a thyroid carcinoma that demonstrates unequivocal vascular or capsular invasion and is composed of at least 75% Hürthle cells; characterized by their abundant eosinophilic granular cytoplasm, hyperchromatic/vesicular nuclei, and prominent round central nucleoli ¹. In the fourth edition of the World Health Organization (WHO) classification, HCC is no longer considered a variant of FTC, but rather an independent entity ¹.

Recent molecular advances, including two comprehensive genomic analyses using NGS platform, have shown that HCC has a unique molecular signature distinct from FTC and PTC ^2,3,31–33. These tumors show widespread chromosomal losses, unique alteration of mitochondrial genomes, especially mutations in the subunits of complex I in the electron transport chain, and novel recurrent rearrangements (e.g. CHCHD10-VPREB3, HEPHL1-PANX1, and TMEM233-PRKAB1) but low frequency of BRAF (0–5%) and RAS mutations (9–15%) ^2,3. These insights have recently been employed in the molecular testing for thyroid fine needle aspiration (FNA) to enhance its performance of these platforms in detecting HCC ^28,34. For example, Thyroseq version 3 has added a molecular group with copy number alterations. Not surprisingly, most nodules harboring copy number alterations are diagnosed as HCCs given the widespread chromosomal loss reported in these lesions ²⁸. Compared to PTC, which has a low frequency of TERT promoter mutation (9%) and TP53 mutation (0.8%), TERT promoter mutation and TP53 mutation are relatively common in HCC with a reported frequency of 22%–27% and 7–12% respectively, suggesting that HCC has a more aggressive molecular signature compared with PTC ^2,3,5. The significance of TERT promoter mutation and TP53 mutation is discussed in more detail in the subsequent section.

It is postulated that mitochondrial DNA mutations lead to an impaired electron transport chain and disruption of the oxidative phosphorylation system. Subsequent aberrant compensatory accumulation of mitochondria in tumor cells results in a Hürthle cell (oncocytic) phenotype histologically. Indeed, it has recently been shown that oncocytic FVPTC, a tumor that shares the oncocytic cytomorphology, also commonly harbors non-silent mitochondrial DNA mutations ²³.

Key points:

1. HCC is enriched with mitochondrial DNA mutations, which lead to aberrant mitochondria accumulation and Hürthle cell phenotype.

2. HCC shows widespread chromosomal losses.

3. HCC has low frequency of BRAF and RAS mutations.

Tumor progression and dedifferentiation: late molecular events occurring in PDTC and ATC

Compared with well-differentiated thyroid carcinoma, PDTC and ATC are associated with a dismal clinic outcome. ATC, in particular, is nearly almost fatal with a median survival of 3 to 6 months after diagnosis ¹. The mortality rate for well-differentiated thyroid carcinoma, PDTC, and ATC is 3–10%, 38–57%, and approximately 100% respectively ^35–37. Considerable efforts have been undertaken in recent years to understand the molecular events that may predict tumor aggressiveness and may serve as treatment targets in advanced thyroid cancer ^6–8,38.

Similar to their well-differentiated counterparts, BRAF and RAS mutations remain the main drivers in PDTC and ATC, occurring in 33% and 45% of PDTC, and 29% and 23% of ATC respectively ^4,6–8,39. The frequency of BRAF and RAS mutations in PDTC varies according to the definition of PDTC. When PDTCs are defined using Turin proposal encompassing the following 3 criteria: 1) solid growth pattern; 2) absence of nuclear features of PTC; and 3) necrosis, mitotic index ≥3/10 high power fields (HPFs), or convoluted nuclei, they contain high frequency (42–64%) of RAS mutations and rare (6–9%) BRAF ^V600E mutations ^39,40. On the other hand, when PDTCs are classified based solely on necrosis or a mitotic index of ≥5/10 HPFs regardless of architectural pattern and nuclear features (MSKCC criteria), they contain both tumors fulfilling Turin proposal and an additional set of tumors that fulfill MSKCC but not Turin proposal. This subset of tumor has high frequency (67–78%) of BRAF mutations and low rate (6–13%) of RAS mutations ^39,40.

TERT promoter mutations, which activate telomerase and attributes to tumorigenesis, are detected at a low frequency in well-differentiated thyroid carcinoma: 10% of PTC, 17% of FTC, and 22–27% HCC ^30,41–45. In contrast, PDTC and ATC are characterized by high rates of TERT promoter mutation, occurring in 40% and 73% respectively ⁴⁶. Furthermore, TERT promoter mutations are subclonal in PTC and clonal in PDTC and ATC, indicating that they are a vital event in tumor progression and evolution. In PTC, TERT promoter mutations appear to be an adverse prognostic factor associated with aggressive histology (TCV), high tumor stage, regional and distant metastases, and increased mortality ^25,41–45.

Inactivating TP53 mutations, a genetic hallmark of ATC ⁴⁷, are infrequent in well-differentiated thyroid carcinoma, being detected in merely 0.8% of PTC in the TCGA cohort ⁵. On the other hand, they are highly prevalent in ATC, detected in up to 73% of tested ATC ^4,6–8,39. Similarly, TP53 mutation also distinguishes ATC from PDTC, which has a lower frequency (10%) of TP53 mutation. PDTC and ATC also have mutations in other tumor suppressor genes, such as ATM, RB1, MEN1, NF1, and NF2 at a rate of 0% to 9% ⁴.

Additionally, mutations encoding components of PIK3CA-AKT-mTOR pathway, SWI/SNF nucleosome remodeling complex, mismatch repair genes, and histone methyltransferase are exceedingly rare in PTC but occur in 2–11% of PDTCs and 12–39% of ATCs ^4,39.

A recent study has identified a distinct subgroup of ATC with mutation in mismatch repair genes but with intact BRAF, RAS, and RET oncogenes ²⁹. Such findings indicate that a subset of ATC may arise through BRAF/RAS-independent mechanisms, e.g. microsatellite instability.

In summary, the persistence of BRAF and RAS mutations throughout thyroid cancer development and the acquisition of additional mutations in PDTC and ATC indicate a stepwise tumor progression from well-differentiated to PDTC and ATC in the majority of PDTC and ATC. BRAF and RAS mutations are the driver mutations that occur early in the tumorigenesis and are present at a comparable rate throughout tumor progression, whereas mutations in TERT promoter, TP53, PIK3CA, SWI/SNF complex, mismatch repair genes, and histone methyltransferase pathway accumulate in the process of dedifferentiation and tumor progression, leading to the development of PDTC and ATC ^4,6–8,39.

Genetics of medullary thyroid carcinoma: the RET proto-oncogene

Medullary thyroid carcinoma (MTC) is a neuroendocrine carcinoma originating from C-cells ⁵. Approximately 75% of MTCs are sporadic while the remaining 25% arise in the setting of multiple endocrine neoplasm (MEN) type 2 with germline gain-of-function mutations in RET ^5,48. Three subtypes of MEN2 have been recognized by the WHO: MEN2A, MEN2B, and familial MTC ⁵. The involved RET codons in each subtype are different; in MEN2A, C634R on codon 634 in exon 11 is most commonly altered, affecting approximately 85% of MEN2A patients ^5,48,49 while MEN2B patients have unique germline M918T and A883F mutations ^5,48,49.

RET point mutation is also the most common driver molecular event in sporadic MTCs, reported in 40–60% of cases ^5,48,49. A small percentage of sporadic MTCs may have a RAS mutation ^49,50 or ALK fusion ⁴⁹.

Novel targeted therapies in thyroid carcinoma: targeting the MAPK pathway

The growing knowledge of MAPK alterations in thyroid carcinoma has led to multiple clinical trials of tyrosine kinase inhibitors (TKI), with the ability to block the MAPK pathway, for the treatment of advanced thyroid cancer (Figure 1 and Table 2). The presence of unique molecular alterations may drive decisions related to targeted therapy for advanced or metastatic thyroid carcinoma ^51,52. Recently, four multitargeted TKIs have been approved by the Food and Drug Administration (FDA); sorafenib and lenvatinib for RAI-refractory differentiated thyroid carcinoma, and cabozantinib and vandetanib for MTC based on results of phase III clinical trials ^53–55. Table 2 provides a brief summary of selective TKIs in thyroid carcinoma. Lenvatinib targets VEGFR, FGFR, PDGFR, RET, and C-KIT; sorafenib has its effects on VEGFR, RET, and RAF, while cabozantinib inhibits MET, VGEFR, and RET ^53–55. Vemurafenib and dabrafenib, two selective BRAF inhibitors, can restore RAI uptakes and efficacy, demonstrating a partial response in 38% of patients with metastatic or unresectable RAI-resistant BRAF^V600E-mutated PTC ^56–58. Similarly, selumetinib, an inhibitor for MEK1/2, has also shown to reverse the RAI refractoriness in metastatic thyroid carcinoma ⁵⁹. In response to these promising results of targeted therapy, the National Comprehensive Cancer Network (NCCN) guidelines have recently recommended molecular testing in patients with advanced thyroid carcinoma to identify actionable targets and to select patients that are eligible for clinical trials ⁵¹.

Table 2.

Molecular therapeutic targets in advanced thyroid carcinoma: selective tyrosine kinase inhibitors and their effects in thyroid carcinoma.

Drugs	Targets	Cancers treated
Sorafenib 53	VEGFR, PDGFR, RET/PTC, BRAF, C-KIT	RAI-refractory DTC, ATC MTC
Lenvatinib 53	VEGFR, PDGFR, FGFR, RET/PTC, C-KIT	RAI-refractory DTC ATC MTC
Cabozantinib 54	VEGFR, RET, MET	MTC
Vandetanib 55	VEGFR, EGFR, RET	MTC
Vemurafenib 56	BRAF	RAI-refractory BRAFV600E-mutated PTC
Debrafenib 58	BRAF	RAI-refractory BRAFV600E-mutated PTC
Selumetinib 59	MEK1, MEK2	RAI-refractory DTC

DTC: differentiated thyroid carcinoma, RAI: radioactive iodine.

Techniques

In the past decade, several ancillary molecular techniques have been developed to improve the diagnostic accuracy of thyroid FNA and are commercially available in the United States ⁶⁰. A brief summary of these platforms is provided in Table 3. According to the second edition of The Bethesda System for Reporting Thyroid Cytopathology (TBSRTC), the current American Thyroid Association (ATA) guidelines, and the current NCCN guidelines, these molecular platforms may be considered as possible ancillary tests following thyroid FNA to further stratify the risk of malignancy in cases with indeterminate cytology, which in turn comprises of two categories: atypia of undetermined significance ( AUS, Bethesda category III), and follicular neoplasm/suspicious for follicular neoplasm (FN/SFN, Bethesda category IV) ^51,52,60,61.

Table 3.

Common Commercially available molecular testing platforms for thyroid fine needle aspiration

Platforms	Testing methods	PPV a	NPV	Rule in test	Rule out test
Thyroseq genomic classifier version 3 28	Next generation sequencing platform of 112 genes: mutations, fusions, copy number alterations and gene expression	66%	97%	Yes	Yes
Afirma gene expression classifier 34,60,62	Microarray essay of mRNA expression of 142 genes	42%	93–97%	No	Yes
ThyGenX/ThyraMIR 34,62	Combined test of NGS mutation detection and miRNA classifier	74–82%	92–94%	Yes	Yes
RosettaGX 34,62	microRNA classifier	42%	92%	No	Yes

mRNA: messenger RNA, miRNA: microRNA.

^a.The PPV is calculated using a surgical diagnosis of carcinoma or NIFTP.

Thyroseq version 3 is an NGS panel that detects mutations, fusions, copy number alterations, and gene expressions of 112 thyroid cancer-related genes, including commonly altered genes such as BRAF, RAS, and RET-PTC. In a multicenter prospective study of 257 FNA samples with an indeterminate diagnosis, it showed a high NPV of 97% and reasonable PPV of 68% ²⁸. ThyGenX/ThyraMIR is a combined step-wise platform. All samples are initially subjected to ThyGenX, an NGS mutation panel, and those with negative ThyGenX results then undergo ThyraMIR, a microRNA classifier ^34,62. Afirma gene expression classifier is a microarray assay of mRNA expression ^34,60,62, whereas RosettaGX is a microRNA classifier ^34,62. Tests with high negative predictive values-Afirma, ThyGenX/ThyraMIR, and RosettaGX are considered ‘rule-out’ tests, meaning a negative test result can be used to rule out a malignant diagnosis. On the other hand, tests with a high positive predictive value- Thyroseq version 3 and ThyGenX/ThyraMIR are ‘rule-in’ tests, meaning a positive result is associated with a high probability of malignant diagnosis, with the risk of malignancy dependent on the mutation that is detected. These molecular testing platforms have been shown to reduce diagnostic uncertainty and assist clinical decision in managing indeterminate thyroid nodules.

These molecular results should be interpreted in their appropriate context. First, the calculation of sensitivity, specificity, NPV and PPV of the above-mentioned platforms include NIFTP as one of the positive diagnoses. As NIFTP has recently be reclassified as a non-malignant tumor ²¹, the sensitivity and PPV would decrease if NIFTP is removed from the calculation ^63–66. Second, detection of a molecular alteration using thyroseq platform or a suspicious result using Afirma platform does not warrant a total thyroidectomy. The reported malignant rate of a RAS-mutated nodule detected by Thyroseq varies from as low as 10% to 62% ^28,67. Similarly, most thyroid nodules with indeterminate cytology and suspicious Afirma results have a surgical diagnosis of benign nodules or NIFTP ⁶⁸. Hence, a conservative surgical approach, e.g. lobectomy or hemithyroidectomy, may be more appropriate as the initial management for these patients. Lastly, The PPV of these platforms can vary significantly based on characteristics of the population and incidence of thyroid cancer ⁶⁷. Therefore, it may be prudent to determine the PPV of these platform at institutional level.

Conclusions

In this review, we summarized the recent genomic advances, prognostic molecular signatures, promising targeted therapies and commercially available molecular platforms in thyroid carcinoma. Constitutive activation of the MAPK pathway is crucial for the pathogenesis and targeted therapy of thyroid cancer. A strong genotype-phenotype correlation is demonstrated in follicular cell-derived carcinoma. RAS mutations are commonly seen in follicular-patterned carcinoma while the BRAF ^V600E mutation is the most prevalent event in classical and tall cell variants of papillary thyroid carcinoma.

Key points.

• BRAF ^V600E mutation is frequent in classic and tall cell variants of papillary thyroid carcinoma.

• Follicular-patterned carcinoma, i.e. follicular carcinoma and follicular variant of papillary thyroid carcinoma, is characterized by RAS mutation and PAX8-PPARγ fusion.

• Thyroid carcinoma has stepwise molecular tumorigenesis from well-differentiated to poorly differentiated to anaplastic carcinoma in which BRAF and RAS mutations remain the main driver events.

• TERT promoter mutation, TP53 mutation, and mutations in PIK3CA pathway are associated with tumor aggressiveness and are common in poorly differentiated and anaplastic thyroid carcinoma.

• Novel tyrosine kinase inhibitor therapy targeting the MAPK pathway shows promising results in treating advanced thyroid carcinoma.

Synopsis.

Thyroid carcinoma is the most common cancer in the endocrine system. Recent advances, using next-generation sequencing, have shed light on the molecular pathogenesis of thyroid cancer. Constitutional activation of the MAPK pathway through RAS mutation, BRAF mutation, and/or fusions involving receptor tyrosine kinase (e.g. RET-PTC) play a central role in tumorigenesis and open doors to promising tyrosine kinase inhibitor therapy. Several molecular signatures, such as TERT promoter mutation and TP53 mutation, are associated with tumor progression. In this review, we provide a concise and updated summary of the main genetic alterations in thyroid carcinoma.