Medullary Carcinoma



Medullary Carcinoma


Paul W. Biddinger





HISTORICAL COMMENTS

The term medullary carcinoma of the thyroid was first used in reports published in 1955 and 1959 by Hazard and colleagues.1,2 They distinguished this carcinoma by its solid nonfollicular pattern, amyloid stroma, and biologic behavior intermediate that of papillary and anaplastic thyroid carcinomas. They chose the term medullary because of the discrete gross appearance and sheetlike histologic pattern that seemed similar to medullary carcinoma of the breast.3 Although Hazard et al. coined the term medullary thyroid carcinoma and were the first to define it as a distinct clinicopathologic entity, earlier reports appear to have described the same tumor (Fig. 14.1).4,5,6

In 1966, Williams7 suggested that medullary carcinomas derived from the C (parafollicular) cell, noting that this origin would resolve the discrepancy between its undifferentiated appearance and relatively good prognosis. Meyer and Abdel-Bari8 in 1968 reported calcitonin-like activity in an extract of medullary carcinoma. In the following year, Bussolati et al.9 demonstrated calcitonin in histologic sections of medullary carcinoma by means of immunofluorescent staining. Immunohistochemical stains for calcitonin and other components of medullary carcinoma have become part of the standard clinical repertory over the last four decades, expanding our appreciation for medullary thyroid carcinoma’s range of morphologic and clinical expression.

In 1961, Sipple10 reported a case of pheochromocytoma associated with thyroid carcinoma and reviewed five others from the literature. None of the cases carried the specific diagnosis of medullary thyroid carcinoma, but this is not surprising given that they were reported before or only shortly after the establishment of this diagnostic term. Sipple syndrome has since become the eponymous name for multiple endocrine neoplasia type 2A (MEN2A).

Schimke and Hartmann11 documented the association of medullary thyroid carcinoma and pheochromocytoma and its autosomal dominant inheritance in 1965. In 1968, Steiner et al.12 described a large kinship with medullary carcinoma, pheochromocytoma, and hyperparathyroidism, affirming the autosomal dominant inheritance with high penetrance and variable expressivity. They proposed the term multiple endocrine neoplasia type 2 to distinguish it from the established entity of multiple endocrine adenomatosis involving the pituitary, parathyroids, and pancreas, which is now designated multiple endocrine neoplasia type 1.

Williams and Pollock13 reported the association of multiple mucosal neuromata with medullary thyroid carcinoma and pheochromocytoma in 1966. Subsequent reports by Schimke et al.14 and Gorlin et al.15 in 1968 confirmed this association as a distinct clinicopathologic entity, which we now recognize as multiple endocrine neoplasia type 2B (MEN2B). MEN2B is also known as Wagenmann-Froboese syndrome, an infrequently used eponym that recognizes case reports from the 1920s of individuals who had facial characteristics of MEN2B.16,17 In 1986, Farndon et al.18 recognized a familial form of medullary thyroid carcinoma without extrathyroidal manifestations. Their observation of hereditary medullary carcinoma distinct from MEN2A and MEN2B has been verified by subsequent studies and is now designated familial medullary thyroid carcinoma (FMTC).


INCIDENCE AND EPIDEMIOLOGY

In the United States, medullary carcinoma accounts for approximately 2% to 4% of thyroid malignancies.19,20,21 The incidence of medullary carcinoma in the United States was 0.21 per 100,000 population in 2005 to 2006.22 It occurs on both a sporadic and a hereditary basis. Most cases are sporadic and have a peak incidence in the fifth or sixth decade.23 The pattern of inheritance is autosomal dominant. In most studies, the male to female ratio is about equal, although a slight female predominance has been reported for sporadic disease.23,24

Internationally, the reported rates for medullary carcinoma vary approximately from 0.02 to 0.35 per 100,000 person-years, with medullary carcinoma being the third most frequent type of thyroid malignancy after papillary and follicular carcinomas.25 The international rates of medullary carcinoma among males tend to be higher compared with that for females.

The prevalence of occult medullary carcinoma in thyroidectomy specimens is approximately 0.3%.26 A meta-analysis of autopsy studies revealed a prevalence between 0.1% to 0.2%, with most tumors being <1 cm in diameter.27 The prevalence of occult medullary carcinoma is approximately 0.4% in autopsy series in which the thyroid was completely submitted for histologic examination and immunostained for calcitonin.28,29


Hereditary Medullary Thyroid Carcinoma

Hereditary forms of medullary thyroid carcinoma account for 15% to 30% of total cases.23,24,30,31 The three subtypes are MEN2A, MEN2B, and FMTC. MEN2A is the most common of the three subtypes, accounting for about 75% to 90% of familial cases.23,32,33 FMTC and MEN2B account for about 15% and 5% of cases, respectively.32

Hereditary cases of medullary carcinoma have a frequency of about 1 in 30,000.32 Affected individuals generally present at a younger age, typically in the third decade or earlier depending on the particular mutation. Significant numbers of hereditary cases are now detected and treated at early ages owing to the availability of molecular testing for RET germline mutations. However, medullary thyroid carcinoma is still the most common cause of death of individuals with MEN2A, MEN2B, and FMTC.32







FIGURE 14.1. Drawing from Jaquet’s 1906 publication regarding an apparent case of medullary thyroid carcinoma. The drawing depicts a metastatic lesion in the mediastinum with a uniform population of tumor cells and deposits of amyloid. The letter key to his diagram is as follows: a, amyloid; t, tumor cells; g, blood vessels; th, tumor thrombus in blood vessel; b, connective tissue; ag, blood vessel transformed by amyloid; E, endothelial membrane. (From Jacquet J. Ein Fall von metastasierenden Amyloidtumoren (Lymphosarkom). Virchows Arch Pathol Anat. 1906;185:251-268.)


Multiple Endocrine Neoplasia Type 2A

Multiple endocrine neoplasia type 2A (MEN2A) is characterized by medullary thyroid carcinoma, pheochromocytoma, and primary parathyroid hyperplasia (Fig. 14.2A). Cutaneous lichen amyloidosis or Hirschsprung disease is associated with rare cases of MEN2A.33,34 Medullary carcinoma is usually the first clinical manifestation of MEN2A. The penetrance of medullary carcinoma is 90% to 100% within affected families, and the age at presentation is usually in the third or fourth decade of life.21,33,34 Approximately 50% of individuals affected with MEN2A also develop pheochromocytoma, and about 10% to 30% experience hyperparathyroidism because of parathyroid hyperplasia.32,33,34,35,36 Rare families with features of MEN2A do not have an identifiable RET mutation. In such cases, a clinical diagnosis of MEN2A can be made if at least two of the three classic features are present.21


Multiple Endocrine Neoplasia Type 2B

Individuals with multiple endocrine neoplasia type 2B (MEN2B) have a 100% risk of developing medullary thyroid carcinoma and, in about 50% of cases, pheochromocytoma (Fig. 14.2B).34 MEN2B kindred is also characterized by marfanoid habitus, mucosal neuromas (Fig. 14.3), gastrointestinal ganglioneuromatosis, and abnormal thickening of the corneal nerves.37 Parathyroid hyperplasia rarely occurs in affected individuals. Medullary carcinoma usually presents before the age of 10. MEN2B is the most ominous form of hereditary medullary carcinoma because of its early development and rapid progression to metastatic disease.


Familial Medullary Thyroid Carcinoma

Familial medullary thyroid carcinoma (FMTC) is characterized by medullary thyroid carcinoma alone. FMTC usually manifests during the fifth to sixth decade of life, and these medullary carcinomas tend to have indolent clinical courses.32 Rigorous criteria have been developed for the diagnosis of FMTC to help prevent small MEN2A kindreds from being categorized incorrectly. The criteria are (1) >10 carriers in a kindred, (2) multiple carriers or affected family members older than 50, and (3) no evidence of pheochromocytoma or hyperparathyroidism in family members affected by or at risk for medullary carcinoma.32,33 These criteria may incorrectly place small FMTC kindreds in the MEN2A
category, but this error is offset by the reduced likelihood of overlooking a pheochromocytoma.






FIGURE 14.2. Characteristic abnormalities associated with MEN2A and MEN2B.






FIGURE 14.3. Oral mucosal neuromas in the lips and tongue of a child with MEN2B. A: Nodules in both oral commissures, the upper lip and the tip of the tongue. B and C: Low-power (B) and medium-power (C) photomicrographs of a biopsy showing numerous, irregular nerve bundles.


The American Thyroid Association (ATA) currently views FMTC as a variant of MEN2A that has decreased penetrance for pheochromocytoma and primary hyperparathyroidism. The ATA criteria for FMTC are (1) absence of pheochromocytoma or primary hyperparathyroidism in two or more generations of a family or (2) a RET mutation that only has been identified in FMTC kindreds.21 The ATA advises caution in diagnosing FMTC in small kindreds or a single affected generation because of the potential that these cases may prove to be MEN2A with a risk of pheochromocytoma.


Sporadic Medullary Thyroid Carcinoma

Sporadic medullary thyroid carcinoma is defined by the lack of a family history of medullary carcinoma and other disorders associated with the multiple endocrine neoplasia syndromes. Sporadic cases are more common than their hereditary counterparts, accounting for 70% to 85% of all cases of medullary carcinoma.24,30,31,38 Although most cases are due to acquired somatic mutations, 1% to 10% or possibly more of apparent sporadic cases show germline RET mutations.32,39,40,41 The presence of a germline mutation changes the diagnosis to a hereditary form of medullary carcinoma. Many of these individuals have de novo germline mutations and thus represent the initial appearance of hereditary medullary thyroid carcinoma in a kindred.


ETIOLOGIC FACTORS

The etiology of familial medullary carcinoma is primarily due to a germline mutation of the RET gene as will be discussed below in the next section. A point mutation is identifiable in almost all individuals with clinical, pathologic, and/or historical findings characteristic of hereditary disease.41 The specific cause of the germline mutations is unknown.

About one- to two-thirds of sporadic medullary carcinomas have somatic RET mutations, and as with hereditary cases, the cause of these RET mutations is unknown. Mounting evidence suggests that other genetic mutations and/or epigenetic events contribute to the development of sporadic and possibly hereditary medullary carcinoma as discussed below. Specific etiologic factors for other mutations or epigenetic changes have not been identified.


PATHOGENESIS AND MOLECULAR GENETICS


Germline RET Mutations

Familial medullary carcinomas are associated with gain-of-function mutations in the RET gene. A 1985 study of a cell line transfected with human lymphoma DNA revealed a novel transforming gene.42 This proved to be a fusion gene containing a portion of a gene that coded for a tyrosine kinase domain. This tyrosine kinase gene became known as RET, its name a contraction acronym of REarranged during Transfection. Genetic linkage analysis localized the MEN2A locus to the centromeric region of chromosome 10 in 1987.43 Subsequent studies revealed that RET mapped to chromosome 10q11.2 and that germline mutations of RET were the primary cause of MEN2A, MEN2B, and FMTC.44,45,46 Most germline mutations are point mutations. RET is the gene associated with both MEN2/FMTC disorders and Hirschsprung disease, but the sets of germline mutations are different. In contrast to activating mutations in hereditary medullary carcinoma, loss-offunction mutations are responsible for Hirschsprung disease.36 Of note, RET is also involved in the pathogenesis of papillary thyroid carcinoma, where a portion of the gene is fused and activated as a result of chromosomal rearrangement known as RET/PTC (see Chapter 11).

RET has 21 exons and approximately 55,000 base pairs.47 It codes for a protein (RET) that is a member of the receptor tyrosine kinase superfamily. The receptor comprises an extracellular domain with calcium-dependent cell adhesion (cadherin) and cysteine-rich regions, a transmembrane domain, and an intracellular domain with regions of tyrosine kinase activity (Fig. 14.4).34

The RET receptor is activated when a complex ligand binds to the extracellular domain. Glial cell line-derived neurotrophic factor (GDNF) binds to GDNF family receptor α1, a glycosylphosphatidylinositol protein located on the cell surface, forming a complex that subsequently binds to RET (Fig. 14.5).36,48,49,50 Binding of this complex results in receptor dimerization and autophosphorylation of tyrosine residues. This in turn initiates downstream signaling through the mitogen-activated protein kinase and other pathways.34 Several other members of the GDNF ligand family have been shown to activate RET after binding to surface proteins. These include neurturin, artemin, and persephin.36

The RET receptor activates signaling pathways responsible for cell proliferation, survival, differentiation, motility, and chemotaxis.34 RET is normally expressed by thyroid C cells and cells of the adrenal medulla, sympathetic ganglia, and kidneys. Populations of neural crest cells express RET during early embryogenesis and as they migrate to various regions of the body.34 The normal development of the autonomic and enteric nervous systems and the excretory system is dependent on RET.51

Since the initial discovery of activating point mutations of RET as the primary cause of the hereditary forms of medullary carcinoma, a wide distribution of mutations have been identified in the extracellular cysteine-rich and the intracellular tyrosine kinase domains, as summarized in Figure 14.4 and Table 14.1.33,34,35,36,41 Mutations of five cysteine-rich domain codons collectively account for about 95% of MEN2A and 85% of FMTC kindred.35 Four are located on exon 10 (609, 611, 618, and 620) and one on exon 11 (634), and almost all mutations involve replacement of cysteine by another amino acid. Codon 634 is the most common site of mutations associated with MEN2A, being involved in 80% to 90% of cases. The most common mutation is substitution of arginine for cysteine (C634A), found in about 50% of MEN2A families.34,35,41 The C634A mutation is also linked to parathyroid hyperplasia.52 Most of the mutations associated with FMTC are found in extracellular cysteine-rich domain codons 618, 620, and 634 in a fairly even distribution.35 FMTC also has been associated with mutations of the intracellular tyrosine kinase domain of RET including codons 768, 790, 791, 804, 848, 883, 891, and 904.35,36,41

About 95% of cases of MEN2B are associated with a point mutation in the tyrosine kinase domain codon 918 of exon 16, resulting in the replacement of a methionine by threonine (M918T).34,41,53,54 A point mutation at codon 883 of exon 15 resulting in alanine replacement by phenylalanine accounts for almost all of the remaining small percentage of MEN2B cases.34,36,55 A small number of MEN2B cases have been found to have double germline mutations involving codons 804+805, 804+806, and 804+904.56,57,58,59,60 De novo mutations leading to MEN2B are more common compared with de novo mutations leading to MEN2A and FMTC.

Mutations in the cysteine-rich domain codons, which account for most cases of MEN2A and FMTC, activate the RET protein by constitutive dimerization (Fig. 14.5). The mutated forms of RET have unpaired cysteine residues that can form disulfide bonds with other RET monomers, and thus, dimerization occurs without the necessity of a ligand.36,61,62 Mutation of the 918 codon, which is associated with MEN2B, affects the catalytic core region of the intracellular tyrosine kinase domain and causes constitutive
activation of the monomeric form of RET.36,54 A conformational change in the binding pocket of the tyrosine kinase domain leads to an altered substrate specificity. A small portion of FMTC cases have mutations involving the intracellular tyrosine kinase domain. The activation mechanisms associated with these mutations are not well defined at this time.






FIGURE 14.4. Schematic diagram of RET gene and associated RET protein. Codons reported to be sites of RET mutations resulting in hereditary medullary thyroid carcinoma are listed below their associated exon. The association between exons and the regions of the RET protein they encode is also shown.


Somatic Mutations in Sporadic Medullary Thyroid Carcinoma


Somatic RET Mutations

Somatic RET mutations have been detected in 30% to 66% of sporadic medullary thyroid carcinomas.34,63,64,65 The replacement of methionine by threonine at codon 918 (M918T) is most common, accounting for about 75% to 95% of somatic RET mutations.34,46,63,64,66 This amino acid substitution at codon 918 is the same as that found in most MEN2B germline mutations. The remainder of somatic RET mutations have been reported at most of the other codons associated with hereditary carcinomas. Medullary carcinomas carrying a somatic RET mutation have been shown to have a higher frequency of metastasis involving regional lymph nodes and distant sites and to have a worse prognosis compared with those without RET mutation.63 Of medullary carcinomas with somatic RET mutations, those with the codon 918 mutation appear to be more aggressive compared with those with other somatic RET mutations.34,63


Somatic RAS Mutations

RAS mutations have been recently found with substantial prevalence in sporadic medullary carcinomas. A study of 25 sporadic medullary carcinomas lacking a RET mutation identified HRAS (codons 12, 13, and 61) mutations in 14 cases (56%) and KRAS mutations (codon 61) in 3 cases (12%), whereas only 1 of 40 RET mutation-positive cases also had a RAS (HRAS) mutation.67 No NRAS mutations were identified in any of the 65 medullary carcinomas. In contrast, the most common RAS mutation found in follicular and papillary thyroid carcinomas affects the NRAS gene (Chapters 10 and 11). Another study of 39 sporadic medullary carcinomas found two cases with an HRAS mutation and one with a KRAS mutation, and these cases were negative for RET mutations.68 Although not all studies of sporadic medullary carcinoma found RAS mutations,69 the available data and our experience Y. Nikiforov personal communication, 2011 suggest that pathogenesis of sporadic medullary carcinomas is associated with nonoverlapping mutations involving the RET and RAS genes.


Other Genetic Abnormalities

Other genetic mutations and/or epigenetic events are likely to play a role in the development of sporadic and hereditary medullary carcinoma. The disease phenotype of hereditary tumors generally correlates with RET mutations of specific codons, but individuals with the same germline mutation still show some
variation in regard to clinical manifestations of disease. Several potential oncoproteins and tumor suppressors are present in the signaling pathways utilized by the RET receptor, and mutations in genes encoding these proteins may play significant roles in the development of medullary carcinoma. Major signaling proteins activated downstream of RET are RAS and phosphatidylinositol 3-kinase. As discussed above, RAS mutations have been found in a significant number of sporadic medullary carcinomas. Loss
of function of negative regulators of RET signaling such as the tyrosine phosphatases LAR, PTPRJ, or SHP-1 is another potential mechanism of tumorigenesis. Murine studies suggest that mutations of the tumor suppressor genes RB1 and TP53 may play a role in medullary carcinoma.






FIGURE 14.5. A: Drawing showing the normal binding of GDNF to GDNF family receptor α1 and subsequent binding of the complex with RET receptor and activation through dimerization. B: Drawings showing constitutive dimerization of RET in MEN2A and FTMC and monomeric activation in MEN2B due to change in tyrosine kinase binding pocket.








Table 14.1 Sites of RET Germline Mutations and Associated Forms of Hereditary Medullary Thyroid Carcinoma











































































































































Sites of Germline Mutations




Relative Risk of Aggressive Medullary Carcinoma (2001)a


ATA Risk Level (2009)b


Associated Form of Hereditary Medullary Carcinoma


Domain


Exon


Codon


Extracellular cysteine-rich


8


533



A


FMTC


10


609


1


B


MEN2A and FMTC



611


2


B


MEN2A and FMTC



618


2


B


MEN2A and FMTC



620


2


B


MEN2A and FMTC


11


630


B


MEN2A and FMTC



634


2


C


MEN2A and FMTC



649



A


FMTC


Intracellular tyrosine kinase


13


768


1


A


MEN2A and FMTC



790


1


A


MEN2A and FMTC



791


1


A


MEN2A and FMTC


13/14


804+778



B


MEN2A and FMTC


14


804


1


A


MEN2A and FMTC



804+805



D


MEN2B



804+806



D


MEN2B


14/15


804+904



D


MEN2B


15


883


3


D


MEN2B



891


1


A


MEN2A and FMTC


16


912



A


FMTC



918


3


D


MEN2B


Note: The list of mutations in this table accounts for most cases of hereditary medullary thyroid carcinoma but is not all-inclusive. Additional, but relatively rare, sites of RET germline mutations include codons 321, 515,600, 603, 606, 631, 666, 777, 819, 833, 844, 866, 904, and 922. Most of these are associated with FMTC.


a Relative Risk Classification: 3, highest; 2, high; 1, least high. From 2001 Consensus Document of Seventh International Workshop on MEN; Brandi ML, Gagel RF, Angeli A, et al. Guidelines for diagnosis and therapy of MEN type 1 and type 2. J Clin Endocrinol Metab. 2001;86:5658-5671.

b ATA Risk Level: D, highest, C & B, intermediate, A, lowest. From Kloos RT, Eng C, Evans DB, et al. Medullary thyroid cancer: management guidelines of the American Thyroid Association. Thyroid. 2009;19:565-612.



Loss of Heterozygosity

Analyses of both hereditary and sporadic cases of medullary carcinoma have shown loss of heterozygosity at one or more chromosomes other than 10q including 1p, 3p, 3q, 11p, 11q, 13q, 17p, 18p, 18q, 19p, 19q, and 22q.70,71,72 About 25% of hereditary and 75% of sporadic cases show imbalances when analyzed by comparative genomic hybridization (CGH).70 Medullary thyroid carcinomas with the somatic RET M918T mutation tend to have the highest number of chromosomal imbalances detected by CGH.71 These chromosomal imbalances are due to both gains (amplification) and losses (deletions). The affected regions are known to be the location of tumor suppressor genes as well as the genes that code for proteins of the GDNF family of receptors and their ligands.70 The developing picture of pathogenesis suggests that RET mutations cause C-cell hyperplasia, at least in familial cases, creating a favorable environment for the development of medullary carcinoma and that additional somatic mutations lead to expansion of clones with more aggressive biologic properties.70


DNA Ploidy

Multiple studies have examined DNA ploidy of medullary carcinomas by flow cytometry.73,74,75,76,77,78 Overall, about 30% of cases are aneuploid and the remaining 70% of cases diploid. Most studies found that DNA aneuploidy was an adverse prognostic indicator with univariate analysis but was not an independent prognostic factor when subjected to multivariate analysis.


C-CELL HYPERPLASIA AND RELATIONSHIP TO MEDULLARY CARCINOMA


Historical Comments

C cells, also known as parafollicular cells, are neuroendocrine cells of the thyroid that produce calcitonin and can give rise to medullary carcinoma. The first published descriptions of the
C cells are usually credited to E. Cresswell Baber79,80 who in 1876 and 1878 described them within the dog thyroid. He noted their parafollicular location and different appearance from follicular cells, calling them parenchymatous cells. In 1894, Hürthle81 provided additional description of what he thought were identical cells and applied the term protoplasm-rich cells (protoplasmareichen Zellen) (Fig. 14.6). Although it appears that Hürthle described C cells, his name is often linked to metaplastic oncocytic follicular cells instead.






FIGURE 14.6. Illustrations from Hürthle’s 1894 article describing protoplasm-rich cells in the dog thyroid gland, which are now thought to represent C cells. The upper illustration shows the cells wedged between follicular cells, whereas the lower illustration shows them in a parafollicular location. (From Hürthle K. Beiträge zur Kenntniss des Secretionsvorgangs in der Schilddrüse. Arch Gesammte Physiol. 1894;56:1-44.)

Nonidez82 demonstrated in 1932 that the thyroid contained a population of cells with argyrophilic cytoplasmic granules and based on their location called them parafollicular cells (Fig. 14.7). Pearse83 coined the term C cell in 1966 for cells he thought were comparable to those described by Baber, Hürthle, and Nonidez and that he postulated contained calcitonin. Bussolati and Pearse84 confirmed the presence of calcitonin in C cells the following year.


Origin and Distribution of C Cells

The ultimobranchial bodies populate the thyroid with C cells after they fuse with the medial anlage during embryogenesis (see Chapter 2). After the ultimobranchial bodies are incorporated into the larger medial thyroid anlage, they begin a dissolution phase. Cells of the ultimobranchial bodies disperse into the lateral lobes and give rise to C cells. The ultimobranchial bodies usually fuse with the middle to upper-middle region of the lateral lobes, and this is where the highest concentration of C cells is found in the normal thyroid gland.85,86,87 Few, if any, C cells are found in the upper or lower poles or isthmus.






FIGURE 14.7. Illustration from Nonidez’s 1932 article demonstrating argyrophilic granules within the cytoplasm of parafollicular cells of the dog thyroid gland. Parafollicular cells are designated “d” and elongated follicular cells “e.” (From Nonidez JF. The origin of the “parafollicular” cell, a second epithelial component of the thyroid gland of the dog. Am J Anat. 1932;49:479-505.)

C cells are difficult, if not impossible, to recognize in routine histologic sections stained with hematoxylin and eosin (H&E). The prime exception is certain cases of C-cell hyperplasia as discussed below. Histologic clues for identifying C cells include cytoplasm that is clear or lightly staining and nuclei that are larger and more granular compared with adjacent follicular cells. Fortunately, immunohistochemical staining for calcitonin is readily available to most pathologists for the definitive identification of C cells.


C-Cell Hyperplasia

C-cell hyperplasia is recognized as the precursor of hereditary forms of medullary carcinoma.88 The earliest reports of C-cell hyperplasia derived from studies of thyroids resected from individuals with MEN2A and MEN2B in which C-cell hyperplasia was a frequent finding. Thyroids removed prophylactically because of RET germline mutations may only show C-cell hyperplasia.89



Definition of C-Cell Hyperplasia

The criterion for C-cell hyperplasia used most commonly is >50 cells per low-power (100×) field (LPF).30,90,91,92,93 In general, this is a reasonable and practical guideline for diagnostic histopathologists. There are, however, caveats for this criterion, and these are discussed later in this section.


Forms of C-Cell Hyperplasia

C-cell hyperplasia can exhibit focal, diffuse, and nodular growth patterns (Fig. 14.8 and Table 14.2). These patterns are not mutually exclusive, and more than one can be seen in a given case. The number of C cells and degree of aggregation can range in a continuum across the spectrum of hyperplasia. C-cell hyperplasia may be recognizable in H&E-stained sections, particularly in cases associated with hereditary forms of medullary carcinoma.

C-cell hyperplasia also has been observed in a number of other neoplastic and nonneoplastic conditions, although calcitonin immunostaining is generally necessary to identify the C cells. C-cell hyperplasia is one of the findings associated with PTEN-hamartoma tumor syndrome (PHTS).94 PHTS is caused by a germline mutation of the PTEN tumor suppressor gene. Other associated lesions include multiple hyperplastic/adenomatous nodules, chronic lymphocytic thyroiditis, papillary carcinoma, follicular carcinoma, and/or follicular adenomas. C-cell hyperplasia has been observed in chronic lymphocytic (Hashimoto) thyroiditis (Fig. 14.9), multinodular goiter, various forms of hypothyroidism, adjacent to follicular neoplasms and primary thyroid lymphoma, and following prior hemithyroidectomy.95,96,97,98,99 From these observations, the concept has evolved that C-cell hyperplasia may be due to either an intrinsic genetic alteration or an external stimuli such as trophic hormones, hypercalcemia, hypergastrinemia, paracrine factors, or inflammation.38,96 On this basis, C-cell hyperplasia can be separated into “neoplastic” and “reactive (or physiologic)” forms.100






FIGURE 14.8. Growth patterns of C-cell hyperplasia. A: Focal hyperplasia with segmental proliferation. B: Focal and diffuse hyperplasia (encirclement of follicles). C: Nodular hyperplasia with obliteration of follicular lumen. D: Diffuse and nodular hyperplasia with central area suspicious for invasion beyond follicular basement membrane (early medullary carcinoma).

Neoplastic C-cell hyperplasia is characterized by nodular and/or diffuse proliferation of relatively large cells with mild to moderate nuclear atypia that can be appreciated in H&E-stained sections (Fig. 14.10). These cells have more abundant cytoplasm that is clear or lighter staining relative to typical follicular cells in the section. It is of note that some consider C-cell hyperplasia a misnomer when it has “neoplastic” features and/or is associated with hereditary medullary carcinoma and instead prefer the term medullary carcinoma in situ.95,101

Calcitonin immunostaining is indicated whenever features suggestive of C-cell hyperplasia are seen in H&E-stained sections of thyroid. If C-cell hyperplasia is confirmed by calcitonin immunostaining, it is advisable that this finding be reported even if the thyroid was resected for reasons other than medullary carcinoma and/or RET germline mutation. Not every case in which C-cell hyperplasia is identifiable in H&E-stained sections will prove to be associated with a RET germline mutation, but consideration of
genetic testing is advisable because of potential identification of a new hereditary kindred.








Table 14.2 C-Cell Hyperplasia—Definitions of Subtypes and Growth Patterns























Subtypes


“Neoplastic”


“Reactive”


• Detectable on H&E-stained sections


• C cells enlarged


Mild to moderate nuclear pleomorphism distinct from follicular cells


• More abundant cytoplasm, typically clear or lighter staining than follicular epithelium


• Usually bilateral


• NCAM immunostain usually positive


• Associated with RET mutations


• Detectable only with calcitonin immunostain


• C cells not enlarged


• No cytologic atypia


• Usually unilateral


• NCAM immunostain usually negative


• Probably not associated with RET mutations


Growth Patterns


Focal


Diffuse


Nodular


• Segmental proliferation within follicles


• Circumferential intrafollicular collar


• Obliteration of follicular lumen


NCAM, neural cell adhesion molecule.







FIGURE 14.9. C-cell hyperplasia associated with chronic lymphocytic (Hashimoto) thyroiditis. A: Low-power view of H&E-stained section showing features of chronic lymphocytic thyroiditis. B: Calcitonin-stained section of same area as (A) showing numerous C cells. C: High-power view of H&E-stained section showing C cells with clear cytoplasm. D: High-power view of calcitonin-stained section showing encirclement of follicles by C cells.







FIGURE 14.10. “Neoplastic” C-cell hyperplasia. Large C cells with clear or lightly stained cytoplasm detectable in H&E-stained sections (A and C). Calcitonin-immunostained sections of same areas (B and D).

Reactive C-cell hyperplasia consists of an increase in C cells that meets the criterion for hyperplasia and requires calcitonin immunostaining for recognition. The C cells are not appreciably enlarged and lack cytologic atypia. Focal, diffuse, and nodular growth patterns are seen in the reactive subtype of C-cell hyperplasia with about the same relative frequency.91,92,100

Although the use of the terms neoplastic and reactive C-cell hyperplasia is common, these designations have not yet gained universal acceptance. Consistent differentiation of these two forms of C-cell hyperplasia on a histologic basis by pathologists may be problematic. Investigators have tried to identify other immunohistochemical markers that distinguish neoplastic from physiologic subtypes. Positive immunostaining for neural cell adhesion molecule (NCAM) favors neoplastic C-cell hyperplasia, but it is not 100% sensitive or specific.93,102 Thus far, no immunostain has emerged that can serve as the criterion standard for distinguishing these putative subtypes. The lack of such a marker, however, has been offset by the development of molecular tests that identify specific RET germline mutations.


Differential Diagnosis of C-Cell Hyperplasia

The differential diagnosis of C-cell hyperplasia includes medullary microcarcinoma, spread of medullary carcinoma, solid cell nests, parathyroid tissue, focal squamous metaplasia, and palpation thyroiditis. Distinction of nodular C-cell hyperplasia from an early medullary microcarcinoma can be challenging. The criterion for medullary microcarcinoma is cells breaching the basement membrane of the follicle and extending into the adjacent stroma. Fibrosis around nests of tumor cells is a clue that invasion has occurred (Fig. 14.11). Immunostains for basement membrane components such as collagen IV may be helpful in some cases.103 Tenascin C, an extracellular matrix glycoprotein, has been found to be expressed consistently in the stroma of both hereditary and sporadic medullary microcarcinomas and also in the stroma of concomitant and isolated foci of C-cell hyperplasia.104 The significant overlapping of stromal expression in medullary microcarcinoma and C-cell hyperplasia limits the diagnostic utility of tenascin C, although the absence of stromal expression supports a diagnosis of C-cell hyperplasia.

Intrathyroidal spread of medullary carcinoma can have small foci similar in size to C-cell hyperplasia, but distinction between the two is generally not an issue with adequate sampling of the thyroid. Focal squamous metaplasia is typically associated with chronic lymphocytic (Hashimoto) thyroiditis. Squamous metaplasia associated with thyroiditis and palpation thyroiditis is discussed in Chapter 4. Solid cell nests are discussed in Chapter 1.


Controversial Aspects of C-Cell Hyperplasia

The definition of C-cell hyperplasia remains problematic. Although >50 cells per LPF (100× magnification: 10× objective
combined with 10× eyepiece) is the most common criterion, even this has its variations as requirements range from this number of cells in just one field, to at least one field in each lobe, or to at least three fields in either or both lobes. Several other criteria have been used in the past.105






FIGURE 14.11. A-D: Two cases of early medullary microcarcinoma arising in an area of “neoplastic” C-cell hyperplasia. A fibrotic response to cells extending beyond the follicles is seen. E and F: Multifocal intrathyroidal spread of medullary carcinoma. Small isolated foci bear resemblance to nodular C-cell hyperplasia (A-C, E, and F, H&E; D, calcitonin immunostain).

Distinct separation of C-cell hyperplasia from the upper limit of normal is difficult. Part of the problem is few studies have documented the concentration of C cells in the normal thyroid. Determining the normal concentration and distribution is difficult. C cells are relatively scant, comprising 0.1% or less of the thyroid cell population, and they are not equally distributed within the thyroid. Studies also indicate that C-cell concentration varies by age with a tendency of neonates and children to have more per unit region compared with adults. However, the ranges of concentration are broad with considerable overlap among the different age groups.105 One autopsy study found >50 C cells in three or more LPFs in 33% of adults, whereas 62% had >50 C cells in at least one LPF.106 Use of the term hyperplasia is
problematic when a high percentage of the “normal” population meets the criterion.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Jul 9, 2016 | Posted by in PATHOLOGY & LABORATORY MEDICINE | Comments Off on Medullary Carcinoma

Full access? Get Clinical Tree

Get Clinical Tree app for offline access