Molecular Pathogenesis of Thyroid Neoplasia

Introduction

Over the past decade, there has been an enormous expansion of knowledge regarding the molecular pathogenesis of thyroid tumors and major advances in clinically applying this knowledge to enhance diagnostic approaches to thyroid nodules and informing clinical trials and treatments for patients with progressive disease. This includes not only somatic genomics (i.e., genetic changes only found in tumor cells) but also germline predisposing genomics (i.e., inherited genetic alterations in all cells). In this chapter, we will focus on genomic changes that are currently being used in clinical practice and/or clinical trials to improve preoperative diagnoses, to predict the biological behavior of cancers, and/or to determine treatment options or clinical trial enrollments for patients with established thyroid cancer.

Thyroid Cancer Epidemiology

Thyroid carcinoma rates are increasing worldwide (see Chapter 17 , Differentiated Thyroid Cancer Incidence). At present in the United States, thyroid cancer is the fifth most common cancer in women and the 11th most common cancer in men. It accounts for ~ 95% of malignancies of classical endocrine organs. It is of considerable interest that among all cancer of nonreproductive organs, the 3:1 ratio of female:male cancer incidence is the highest. It is without question that the marked increased in rise in thyroid cancer that began in the early 2000s reflects increased medical surveillance, the use of thyroid ultrasound, and the development of ultrasound-guided fine-needle aspiration (FNA) technologies. The evidence for this includes the much more rapid rise in the diagnosis of small papillary thyroid cancers in the United States and the data from South Korea, where the incidence of diagnosed papillary thyroid carcinoma (PTC) became similar to that of autopsy studies as a result of more universal coverage of thyroid ultrasound in cancer screening. Interestingly, with this emerging data, evidence that many patients with small PTCs can be actively surveilled safely for decades, and a consequent increased emphasis on identifying larger PTCs in clinical guidelines, the incidence of diagnosed thyroid cancer has “flattened,” or even decreased, in the past few years. It is important to note, however, that whereas the percentage of individuals diagnosed with larger thyroid cancers or have died from thyroid cancer has been stable or has decreased, the absolute number of individuals with larger cancer or have died from the disease is increasing. The reasons for this are unclear. It still may be in part due to the increase in detection but also may reflect an increase due to other factors such as environmental exposures or genetic drift in the population. Together, these data highlight the need for additional biomarkers to better identify patients destined to have indolent thyroid cancer versus those likely to harbor more aggressive disease to avoid overtreatment and also to minimize the effects of undertreatment.

Pathology

From a clinical perspective, the thyroid gland is composed of two populations of hormone-producing cells: (1) follicular cells that produce thyroid hormones from the backbone of thyroglobulin and (2) parafollicular cells (C-cells) that represent a minority population of neuroendocrine cells that produce calcitonin and carcinoembryonic antigen (CEA) (see Chapter 2 , Applied Embryology of the Thyroid and Parathyroid Glands). These two cell populations likely derive from the same endoderm progenitor cells, a feature of thyroid development only recently discovered. Although rare mixed tumors can produce both thyroglobulin and calcitonin, the vast majority of thyroid cancer derives distinctly from either cell type (called follicular cell-derived thyroid cancers) or medullary thyroid cancers. The cancers that derive from the follicular cells account for ~ 95% of thyroid cancer cases and can further be divided into subgroups based on histopathological characteristics. The majority of these cancers are well-differentiated and are either PTC or follicular thyroid cancer (FTC) and together are termed as differentiated thyroid cancer (DTC) (see Chapter 19 , Papillary Thyroid Cancer; Chapter 20 , Papillary Thyroid Microcarcinoma; Chapter 21 , Papillary Carcinoma Observation; and Chapter 22 , Follicular Thyroid Cancer). The cancer cells in these tumors typically maintain expression of the thyroid-stimulating hormone (TSH) receptor, the sodium-iodide (Na/I) symporter, and thyroglobulin, features that have important treatment and monitoring implications. Classical forms of PTC develop de novo within the normal thyroid without a clear “premalignant” pathology, with tumors less than 1 cm defined as papillary microcarcinoma. By contrast, FTCs are thought to develop in part from benign follicular adenomas thereby presenting a multistep premalignant-to-malignant model similar to other cancers, including colon carcinoma. Finally, it is important to recognize that the majority of thyroid nodules detected clinically are not neoplastic but are hyperplastic nodules, colloid nodules, or other benign pathologies that are not known to have malignant potential. Autonomous thyroid nodules (“hot” nodules on radioiodine scanning) rarely develop into malignancy and are characterized by specific mutations that activate protein kinase A signaling, thereby increasing thyroid hormone production ( TSHR and GNAS1 mutations).

Hürthle (oncocytic) cell thyroid cancers had been classified as a variant of FTC; however, the most recent World Health Organization (WHO) classification considers Hürthle cell tumors a distinct type (see Chapter 25 , Hürthle Cell Tumors of the Thyroid). This classification is consistent with recent genomic data that establish Hürthle cells as a distinct form of DTC. Cancers composed of these cells typically do not concentrate radioiodine well but express TSH receptor and thyroglobulin. Hürthle cells are eosinophilic polygonal cells characterized by the presence of large and a greater number of mitochondria with atypical features. These cells can be seen not only in cancers but also in benign adenomas and in association with benign thyroid disorders such as Hashimoto’s thyroiditis.

A highly specific type of thyroid tumor recently classified is termed as noninvasive follicular tumor with papillary-like nuclear features (NIFTP) (see Chapter 23 , Noninvasive Follicular Thyroid Neoplasm with Papillary-like Nuclear Features [NIFTP]). This tumor type previously was termed under one of several names, including encapsulated follicular variant of PTC, follicular adenoma with cellular atypia, and follicular tumor of uncertain malignant potential. These tumors are intrathyroidal; encapsulated; do not display vascular, capsular, or lymphatic invasion; and exhibit nuclear features of PTC. They appear typically to have a benign clinical course and are often characterized by mutations in the RAS genes. Although preoperative diagnosis is elusive due to the requirement for thorough assessment of the tumor capsule for invasion, postoperatively, NIFTPs are thought to be largely cured by surgery alone.

Follicular cell-derived and parafollicular cell-derived thyroid cancers can become poorly differentiated based on histomorphologic and immunohistochemical features. In the case of follicular cell-derived thyroid cancer, these poorly differentiated thyroid carcinomas (PDTCs) sort into two main groups. According to the Turin classification, PDTC includes so-called insular and trabecular variants that are classically FTC-derived. Alternatively, other PDTC are derived from high-grade papillary carcinomas that are enriched for tall and columnar cell variants PTC. Anaplastic thyroid cancer (ATC) appears to develop mostly from well or poorly differentiated PTC or FTC, although in some cases it may occur de novo and often contains higher numbers of cancer stem cells and is characterized by very robust immune cell infiltrates shown to facilitate its progression. These tumors typically do not express TSH receptor, Na/I symporter, or thyroglobulin; are remarkably aggressive; and unless diagnosed and fully removed surgically, are uniformly fatal.

Thyroid Cancer Prognosis

Individuals with DTCs typically have an outstanding prognosis, and cure is common when diagnosed early. Because these tumors are slow-growing and typically indolent in most cases, there are multiple studies demonstrating the safety of active surveillance rather than surgery for individuals with very small PTCs that have no evidence of metastasis. However, it is important to recognize that some patients with DTCs have local metastasis and others may have or develop distant metastasis that can be life threatening. As outlined in detail in other chapters of this text, current treatment strategies rely on risk stratification of patients based on imaging characteristics, tumors, and demographics as well as response to initial therapies (see Chapter 24 , Dynamic Risk Group Analysis and Staging for Differentiated Thyroid Cancer). This wide variety in clinical behavior for DTC tumors highlights the important potential role of molecular markers to optimize and individualize therapy.

Patients with ATC have a poor prognosis as a group with the only chance of cure being early and complete surgical resection disease eradication before metastasis (see Chapter 28 , Anaplastic Thyroid Cancer and Primary Thyroid Lymphoma). Because of the poor prognosis, most patients with potentially curable localized ATC are treated aggressively, often with surgery and chemoradiation, unless there are major comorbidities. Less aggressive local approaches are often recommended if the cancer cannot be surgically resected and/or it is metastatic to regional or distant sites. Recent data demonstrate that targeting BRAF ^V600E may have noncurative treatment benefits or neoadjuvant treatment benefits for patients with ATC, as discussed in the treatment section later in this chapter. Thus evaluation of somatic genomics of ATC may influence treatment approaches.

Individuals with medullary thyroid carcinoma (MTC) fall into two categories, those with sporadic MTC (~ 75%) and those with inherited MTC (~ 25%) due to an inherited germline mutation in RET (see Chapter 26 , Sporadic Medullary Thyroid Carcinoma, and Chapter 27 , Syndromic Medullary Thyroid Carcinoma: MEN 2A and MEN 2B). For all patients with MTC, primary surgery, including central neck node dissection, can be curative. Such an approach is most effective for patients diagnosed early due to the high frequency of regional nodal spread. Because individuals identified with genetic screening in a family can be diagnosed earlier, even before the cancer develops, they have the best prognosis as a group, whereas individuals with either sporadic MTC or the probands (first in the family) of a family with inherited MTC are typically diagnosed with larger and later stage tumors.

For patients with larger or more aggressive PTCs or PDTCs with incomplete responses after surgery or with known residual metastases, additional treatment with TSH-suppressive doses of thyroid hormone; radioactive iodine; and in selected cases, kinase inhibitors, clinical trials, or external radiation therapy are used. These treatments and their indications are highlighted in the appropriate chapters of this text.

Germline Genetics of Thyroid Cancer

MTC is the form of thyroid cancer most associated with germline predisposition, a feature that influences clinical management. As described earlier approximately 25% of MTC cases are inherited due to germline mutations in RET that cause multiple endocrine neoplasia type 2 syndromes. MTC development is nearly universal in individuals who inherit this mutation, which can vary based on the specific mutation and the specific family. Recommendations for “prophylactic” thyroidectomy based on these data are available and reviewed separately but are universally recommended for individuals with germline inheritance of the more aggressive mutations with the highest penetrance of clinically important MTC. Interestingly, ~ 50% of MTC in patients with sporadic disease harbor a somatic RET mutation at codon 918, which is associated with more aggressive disease. This same mutation, when it occurs in germline, causes MEN2B. Somatic mutations in RAS are also common but not overlapping with RET . Recently, loss of expression of cell cycle regulators RB and p18 have recently been associated with a more aggressive course for MTC.

In contrast to the high frequency of familial disease in MTC, the vast majority of patients with follicular cell-derived forms of thyroid cancer have clinically sporadic disease. Nonetheless, when considering the entire PTC population (about 80% of all DTCs), PTC has among the highest degree of “familiality” (see Chapter 30 , Familial Nonmedullary Thyroid Cancer). In addition, both PTC and FTC can occur as part of rarely diagnosed cancer syndromes such as Cowden’s syndrome ( PTEN mutations), Carney Complex ( PRKAR1A mutations), DICER syndrome ( DICER1 mutations), familial adenomatous polyposis ( APC mutations), Werner syndrome (progeria), and others. In all of these syndromes, benign thyroid nodules can also develop, thus not all nodules in such patients represent thyroid cancer.

For individuals with nonsyndromic forms of PTC, large genome-wide associate studies (GWAS) have been performed by international consortia that have identified a number of single nucleotide polymorphisms (SNPs) that are associated with PTC diagnosis. Several of these have been functionally characterized, and it is possible that they will lead to new approaches to determining who is at risk of having PTC, thereby allowing for population screening recommendations.