Molecular pathogenesis of thyroid cancer

Abstract

A number of molecular abnormalities have been described in association with the progression from normal thyroid tissue to benign adenomas to well-differentiated and finally anaplastic epithelial thyroid cancer. These include upregulation of proliferative factors, such as growth hormones and oncogenes, downregulation of apoptotic and cell-cycle inhibitory factors, such as tumor suppressors, disruption of normal cell-to-cell interactions, and cellular immortalization. The progression model for thyroid carcinoma has not been proven, but evidence suggests that an evolutionary molecular process is involved, especially in the development of follicular thyroid cancers for which there are distinct intermediate phenotypes. We present a comprehensive evaluation of factors involved in thyroid tumorigenesis and attempt to describe preliminary attributes of a progression model. The organization of this model should also provide a template for the incorporation of new information as it is derived from large-scale genomic studies.

Full text

Surgical Oncology 12 (2003) 69–90
Molecular pathogenesis of thyroid cancer
Dorry L. Segev
a
, Christopher Umbricht
b
, Martha A. Zeiger
c,
*
a
Department of Surgery, Johns Hopkins Medical Institutions, 600 N. Wolfe Street, Baltimore, MD 21287, USA
b
Departments of Surgery, Oncology, and Pathology, Johns Hopkins Medical Institutions, 720 Rutland Ave, Ross 756, Baltimore, MD 21205, USA
c
Division of Endocrine and Oncologic Surgery, Department of Surgery, Johns Hopkins Medical Institutions, 600 N. Wolfe Street, Carnegie 681,
Baltimore, MD 21287, USA
Abstract
A number of molecular abnormalities have been described in association with the progression from normal thyroid tissue to
benign adenomas to well-differentiated and finally anaplastic epithelial thyroid cancer. These include upregulation of proliferative
factors, such as growth hormones and oncogenes, downregulation of apoptotic and cell-cycle inhibitory factors, such as tumor
suppressors, disruption of normal cell-to-cell interactions, and cellular immortalization. The progression model for thyroid
carcinoma has not been proven, but evidence suggests that an evolutionary molecular process is involved, especially in the
development of follicular thyroid cancers for which there are distinct intermediate phenotypes. We present a comprehensive
evaluation of factors involved in thyroid tumorigenesis and attempt to describe preliminary attributes of a progression model.
The organization of this model should also provide a template for the incorporation of new information as it is derived from
large-scale genomic studies.
r2003 Published by Elsevier Ltd.
Keywords: Thyroid cancer; Oncogenes; Tumor suppressor genes; Growth hormones; Telomerase
1. Introduction
Thyroid cancer is the most common endocrine
malignancy and accounts for the majority of endocrine
cancer deaths each year [1]. Except for medullary
thyroid cancer (MTC) derived from parafollicular C-
cells, thyroid lymphoma, and the very rare case of
intrathyroidal sarcoma, thyroid cancers originate from
follicular cell epithelium and include papillary thyroid
carcinoma (PTC), follicular variant of PTC (FVPTC),
follicular carcinoma (FC), its oncocytic variant, the
H.urthle cell carcinoma (HC), and anaplastic thyroid
carcinoma (ATC).
Several etiological factors generally thought to pre-
dispose to neoplasia have been particularly well
documented in thyroid cancer, including radiation,
nutritional factors, and hereditary predisposition. The
most significant sources of radiation exposure in thyroid
cancer patients are from therapeutic and diagnostic use
as well as environmental disasters. The effects are dose
dependent, and show a strong age dependence, with
exposure in childhood and adolescence showing almost
an order of magnitude higher incidence of cancer [2].
Notably, all forms of radiation exposure appear to
predispose principally to papillary thyroid cancer,
particularly PTCs associated with ret oncogene muta-
tions leading to ret fusion proteins [3]. Although the
relation between radiation exposure and PTC is well
documented, this accounts for less than 10% of thyroid
cancers [4].
A relationship between dietary iodine uptake and
thyroid cancer has been suspected since an early study
showed a higher incidence in areas of endemic goiter [5].
The effects of iodine supplementation on the overall
incidence of thyroid cancer have been ambiguous,
although the decrease seen in a landlocked country such
as Switzerland may suggest a beneficial effect in areas of
preexisting iodine deficiency [6]. The effect of dietary
iodine on histological subtype is more evident, with a
higher incidence of FC in endemic goiter areas. With
the introduction of iodine supplementation, the pro-
portion of PTC to FC is reversed [7], although some
of that trend coincides with the increased recognition
of FVPTC as a distinct pathological entity. Data on
additional dietary factors, particularly goitrogens,
are even less clear cut, but anything leading to
ARTICLE IN PRESS
*Corresponding author.
0960-7404/$ - see front matter r2003 Published by Elsevier Ltd.
doi:10.1016/S0960-7404(03)00037-9

compensatory increases in TSH will increase the risk of
thyroid neoplasia, given the central role of TSH in
thyrocyte proliferation.
Genetic predisposition to non-medullary thyroid
tumors also occurs in a number of syndromes affecting
multiple organ systems, predominantly involving organs
of entodermal origin. This is in contrast to other familial
sydromes predisposing to organ-specific tumors, such as
retinoblastoma or familial forms of breast cancer [3,8].
At the cellular level, several studies have documented
a series of molecular abnormalities associated with the
progression of normal to benign to well-differentiated to
ATC (see Table 1)[9–17]. In this review, we explore the
molecular basis for the development of thyroid cancer
and attempt to describe a comprehensive molecular
progression model (see Fig. 1).
2. Thyroid tumor progression model
A number of factors, including hormones, cellular
proteins, and other genetic changes are likely involved in
the development of thyroid malignancy. In an elegant
series of experiments, Hahn et al. developed an
experimental model to define the genetic elements
necessary for the development of cancer. This model
recapitulates a malignant phenotype based on three
genetic elements: an oncogene (ras), a tumor suppressor
gene (T antigen), and telomerase [18]. In this manu-
script, we have identified several elements implicated
in thyroid carcinogenesis. Conceptually, we attempt
to place these abnormalities into three categories:
(1) factors that promote thyroid and/or tumor prolif-
eration, such as oncogenes and hormones and corre-
sponding signal transduction pathways, (2) factors that
hinder tumor proliferation, such as regulators of
differentiation and cell-cycle progression, and (3) factors
controlling cellular immortalization and death, such as
telomerase and regulators of apoptosis (see Fig. 2a).
Genetically, the distinction can be made between gain of
function mutations, which are generally dominant and
lead to a continuous or abnormal oncogenic signal, and
loss of function mutations, which are recessive. The
latter require the elimination of the other allele by
ARTICLE IN PRESS
Table 1
Summary of molecular factors involved in thyroid oncogenesis
Factors that promote tumor proliferation
TSH Stimulates thyrocyte growth, hormone production
GSP Facilitates binding of TSH receptor ligand binding and downstream increase of cAMP
EGF Inhibits differentiation, upregulates proto-oncogene
VEGF Promote tumor growth through angiogenesis
IGF-I Promote thyroid cell proliferation
MET Important thyroid tyrosine kinase binds HGF and stimulates thyrocyte proliferation
MYC Nuclear transcription factor involved in control of cell growth and differentiation
RAS Membrane linked, conveys signals from tyrosine kinase receptor to MAP kinase cascade
BRAF Cytosolic serine/threonine kinase, binds to RAS and activates MAP kinase cascade
RET/PTC Tyrosine kinase receptor, normally binds TGF-beta related neurotrophic factors
TRK-T Peripheral nervous system tyrosine kinase which normally binds nerve growth factor
cyclin D1 Cell cycle progression mediator, activates cyclin dependent kinase, phosphorylates Rb
TPO Catalyzes iodide oxidation, thyroglobulin iodination, and iodothyronine coupling
CP/LF Bind copper and iron, leading to hydroxyl radical damage and malignant behavior
DAP4 Exopeptidase described in T-cell activation, increased activity in many malignancies
HMGI Nuclear protein family of chromatin structure and formation regulators
Factors that hinder tumor proliferation
TGF-beta Blocks cAMP dependent thyrocyte proliferation through CDK and apoptosis
p21 Cyclin dependent kinase inhibitor, effector of p53-mediated G1 arrest of the cell cycle
p27 Cyclin dependent kinase inhibitor, controls G1 to S phase progression
Rb Allows cell cycle progression in phosphorylated form, controls E2F transcription factors
p53 Transcription factor, essential mediator of cell cycle arrest in response to genetic damage
Pax-8/TTF-1 Paired domain transcription factor, controls thyroid-specific gene expression
E-Cadherin Glycoprotein facilitates cell adhesion and maintains epithelial integrity, blocks invasion
Galectins Carbohydrate binding protein which facilitate cell–cell and cell–matrix interactions
CD44 Membrane glycoprotein associated with cell–matrix adhesion, hyaluronic acid receptor
Factors affecting cellular immortalization and death
Telomerase Reverse transcriptase, extends cell life by maintaining length of chromosomal telomeres
Bcl Family of homologous proteins which are pro- or anti-apoptotic
Fas/Fas-L Ligand/receptor system, promotes apoptosis through activation of caspases
PTEN Lipid phosphatase, through PI-3K or Akt/PKB, causes apoptosis or G1 arrest
MDM2 Oncogene, interacts with p53, E2F1, other gene products, ability to transform cells
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–9070

additional events such as loss of heterozygosity or
silencing in order to lose tumor suppressor function.
While the first two categories of molecular events in
cancer can readily be classified as gain and loss of
function events, the third category can be involved in
either type of genetic alteration. Clearly, this classifica-
tion is somewhat arbitrary, since a number of factors
could be assigned to more than one group. For example,
p53 is a classic tumor suppressor gene, whereas some
p53 mutants are oncogenic and p53 is also a key player
in apoptotic cell death. Similarly, many factors can
affect the apoptotic/anti-apoptotic balance of a cell
directly or indirectly and could therefore be grouped
according to either their proliferative or antiproliferative
influence.
Thyroid neoplasms derived from the thyroid epithe-
lium represent an attractive model for understanding the
molecular pathways of multistage tumorigenesis [19].
The evidence available so far suggests that the early
stages of thyroid cancer development may be the
consequence of the activation of proto-oncogenes or
growth factor receptors, such as gsp, met, NTRK, ras,
ret, and the thyrotropin (TSH) receptor [3]. Inappropri-
ate expression of these genes is associated with the
development of neoplasms ranging from benign toxic
adenomas (gsp and TSH receptor), to differentiated
follicular (ras) and papillary (met, NTRK, ret) carcino-
mas. In contrast, alterations in tumor suppressor genes
such as p53 or Rb are observed in poorly differentiated
forms of thyroid cancer, suggesting that they represent
relatively late genetic events [20,21]. An additional and
unique characteristic of thyroid carcinomas is the
relatively high percentage of gene rearrangements and
chromosomal translocations (ret/PTC, NTRK, Pax-8/
PPAR-gamma), which are rare events in most epithelial
tumors [3,22,23], and may reflect the susceptibility of
thyroid tissue to radioactive iodine uptake [24]. Overall,
most genetic alterations have only been documented in
small subsets of tumors and by themselves, are unreli-
able and insensitive markers of malignancy [14,25]. This
suggests that significant components of the relevant
pathogenic pathways leading to thyroid cancer have yet
to be identified.
3. Factors that promote tumor proliferation
Alterations in the signaling pathways otherwise
normally involved in the growth and differentiation of
specific tissues are believed to initiate the progression
towards tumorigenesis. Specifically, abnormal synthesis
of growth factors that stimulate cellular proliferation, or
constitutive activation of related receptors and signal
transduction pathways, have been implicated in
thyroid tumor formation [12,26]. Interestingly, certain
growth signals not only stimulate proliferation,
but cause the cells to lose normal differentiated function,
i.e. the ability to concentrate iodide and produce
thyroglobulin [9].
3.1. Growth hormones
3.1.1. TSH-R
TSH plays a major role in thyrocyte growth and
differentiated function, including hormone production
[27]. TSH transduces its signal through its receptor
(TSH-R) and results in stimulation of the adenyl cyclase
(AC)/cyclic AMP (cAMP) and phospholipase C (PLC)
pathways. TSH-R is a member of the G-protein
associated 7-transmembrane-segment receptors and is
considered a growth factor receptor [28]. Aberrant
stimulation of TSH-R is associated with benign
ARTICLE IN PRESS
TSH-R, gsp
ras, gsp? PTEN?
RET/PTC, trk, met,
bRAF, p21? ras? Gal3?
cell cycle? TSH-R? hTERT? p27? Gal3? p53, Rb, p21
ras, gsp?
?
Pax-8? TPO? DAP4? HMGI?
Benign
Lesions
Normal
Thyroid Follicular
Adenoma
Follicular
Variant PTC
Hurthle Cell
Adenoma
Follicular
Carcinoma
Papillary
Carcinoma
Anaplastic
Carcinoma
Hurthle Cell
Carcinoma
Fig. 1. Putative progression model for multistage thyroid tumorigenesis. Thyroid neoplasms seem to undergo a number of genetic alterations along
the pathway from normal tissue to adenoma to carcinoma to undifferentiated carcinoma. This figure illustrates the mutations associated with specific
transitions.
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–90 71

overactive thyroid nodules such as multinodular goiter
or Graves’ disease. This can occur through mutations
of TSH-R that cause its constitutive activation (see
Table 2)[29]. Specifically, mutations of the third
intracellular loop signal transduction region of the
TSH receptor have been associated with transformation
of thyrocytes [28].Table 2 summarizes TSH-R muta-
tions described in 8 studies [29–36], reflecting a 25%
mutation rate in hyperfunctioning thyroid adenomas.
Less evidence exists regarding a possible association
between TSH-R mutations and thyroid malignancy.
Although a major factor in thyrocyte malignant
ARTICLE IN PRESS
Fig. 2. (a) Overview of molecular signaling pathways. This schematic demonstrates some of the signal transduction pathways which lead to
regulation of gene expression, progression of the cell cycle, and programmed cell death (apoptosis) in all human cells. Relevant to our thyroid cancer
progression model, we focus on factors that promote tumor proliferation, factors that hinder tumor proliferation, and factors that allow
immortalization. (b) Schematic representation of the cell cycle. Various phase-specific cyclins and their interactions with CDKs regulate the
progression along the cell cycle. At the important G1 phase checkpoint, CDK inhibitors such as p21 and p27 regulate the activity of CDK4, which
controls progression to G1 directly through phosphorylation of Rb and indirectly through regulation of gene expression by release of E2F
transcription factors.
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–9072

degeneration seems to be TSH-R pathway activation
independent of ligand binding [37], fewer mutations of
TSH-R have been described in thyroid carcinomas
compared to benign tumors (Table 2). Levels of TSH-
R mRNA are similar for benign thyroid tumors
compared to normal thyroid tissue [38,39], but these
levels decrease with decreasing differentiation [40]. Loss
of expression of the TSH receptor has been documented
in thyroid carcinomas [41,42], and it is speculated that
this is a late event in the progression model [39].
3.1.2. GSP
Given the evidence that the TSH pathway plays a key
role in thyrocyte proliferation and oncogenesis, it
follows that proteins found along the TSH-R dependent
signaling pathway might also be potential thyroid
oncogenes. One such protein is Gs-alpha, or gsp, the
alpha subunit of the heterotrimeric G protein family of
GTP-binding proteins that facilitates TSH-R ligand
binding and activation of AC, resulting in increase of
cAMP (see Table 3)[43]. Like mutations of TSH-R, gsp
mutations are found mostly in hyperfunctioning adeno-
mas, with an average of 26% of hot nodules demon-
strating mutations in 7 studies reviewed (Table 3)
[14,34,44–48]. These results are supported by findings
that transgenic mice overexpressing mutated gsp cDNA
or cholera toxin A1 subunit that mimics the activating
mutation of gsp [49] develop hyperthyroidism and
thyroid adenomas [50]. Mutations of gsp have also been
described by a number of groups in thyroid malignan-
cies [44,47,51,52] and mimicking the mutation with
cholera toxin A1 subunit resulted in malignant trans-
formation of rat thyroid cells [53].
3.1.3. EGF
Epidermal growth factor (EGF) exerts its effects by
initiating a kinase cascade which involves the mitogen
activated protein kinase (MAPK). This chain ultimately
phosphorylates ribosomal S6 kinase (pp90rsk) which
then either stimulates thyrocyte growth or abrogates the
differentiation effects of TSH, such as iodide uptake and
hormone production [54,55]. Interestingly, this interac-
tion with TSH occurs despite the fact that no overlap
was seen between the multiple proteins which TSH is
known to phosphorylate and EGF, to activate [56].
EGF aberrations have been reported in a number of
thyroid lesions. Hyperfunctioning nodules as well as
malignant lesions can express elevated levels of EGF
[57], with the highest expression levels in PTC [58] and
ATC [59]. In addition to overexpression of the ligand,
the EGF receptor (EGF-R), encoded by c-erbB1, can be
overexpressed in a number of thyroid lesions such as
PTC [60], and thereby result in downstream upregula-
tion of the oncogene met (see Section 3.3) [61]. Lesions
which overexpress c-erbB1 or the EGF receptor protein
may be more clinically aggressive [58,60].
3.1.4. VEGF
Angiogenic factors stimulate neovascularization and
appear to play an important role in tumors that grow
beyond 2 or 3 mm in size. An increase in expression of a
subset of these factors including VEGF, VEGF-C, and
angiopoietin-2, as well as downregulation of the angioin-
hibitor thrombospondin-I have been described in thyroid
tumors [62]. One study correlated VEGF and VEGF-R
levels with size of papillary thyroid lesions and recurrence
rate [63]. More studies are needed to better define the role
for this potential marker of tumor aggressiveness.
3.1.5. IGF-1
Insulin-like growth factor (IGF-1), expressed by the
stromal cells of the thyroid, can act in an autocrine
fashion to promote thyrocyte proliferation [64,65].
Increased immunoreactivity in thyroid adenomas and
carcinomas has been described for IGF-1 as well as its
receptor [64,65]. Expression patterns of IGF-1 binding
proteins can vary in thyroid lesions, thereby regulating
IGF-1 pathways at the level of ligand/receptor interac-
tion [66]. The IGF-1 signaling pathway can activate
growth and stimulate differentiation by synergizing with
TSH [11,67], and it can also support growth by inducing
angiogenesis and neovascularization in rat thyroid
cells [68].
3.2. Oncogenes
3.2.1. MET
Hepatocyte growth factor/scatter factor (HGF/SF), a
potent mitogen [69] and stimulus for thyrocyte prolif-
eration, functions by binding to the receptor formed by
the alpha and beta subunits of met, a 190 kDa receptor
tyrosine kinase (RTK) important in thyroid growth [70].
Oncogenic potential of met may occur through gene
ARTICLE IN PRESS
Table 2
Reported frequency of TSH-R mutations in human thyroid tumors
Tumor type TSH-R mutations
Hot nodule 28/113 (25%)
Follicular carcinoma 2/14 (14%)
Papillary carcinoma 1/14 (7%)
Table 3
Reported frequency of gsp mutations in human thyroid tumors
Tumor type gsp Mutations
Hot nodule 25/96 (26%)
Adenoma 2/82 (2%)
Follicular carcinoma 2/30 (7%)
Papillary carcinoma 7/74 (11%)
Anaplastic carcinoma 0/4 (0%)
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–90 73

amplification or overexpression [71], through a defect in
post-translational processing [72], by fusion and aber-
rant dimerization [50,73], or through a variant in which
the receptor is not cleaved into its alpha and beta
subunits [74].
Amplification of met has been described in approxi-
mately 70% of PTC and ATC and in 25% of FC, with
little expression in benign thyroid disease and normal
thyroid tissue [71,75]. Some have postulated that
activation of met may occur through a paracrine
mechanism, as parafollicular C cells secrete HGF/SF
[76]. Inconsistent results have been obtained with
regards to the relationship between met expression and
aggressive and metastatic behavior of thyroid tumors,
with Ruco et al. demonstrating a direct correlation
[77], while Belfiore et al. showed an opposite relation-
ship [78].
3.2.2. MYC
The transcription factor c-myc activates expression of
several cell growth and differentiation regulatory genes.
Synthesis of myc is decreased during progression of the
cell cycle, and shut down completely during differentia-
tion and the associated arrest of proliferation [79].myc
dimerizes with a partner protein Max before it binds to
core DNA sequences, and this relationship is essential
for myc activity [80,81]. Both myc and Max can be
regulated by phosphorylation [82], and alterations of
myc causing constitutive expression have been reported
in a number of human tumors.
TSH induces early expression of myc in the thyroid
[83] and several studies examining immunohistochem-
istry and mRNA expression demonstrate increased
expression during thyroid tumorigenesis [58,84–86].
Furthermore, some authors have even shown a relation-
ship between abundance of myc mRNA and more
poorly differentiated thyroid tumors, implying more
aggressive growth with increased myc expression
[38,40,84].
3.2.3. RAS
The ras oncogenes, which include Ha-ras1, K-ras2,
N-ras, synthesize a group of 21 kDa proteins that play
an important role in tumorigenesis and tumor progres-
sion [87]. Two forms of Ras proteins exist, an inactive
form which is bound to GDP and an active form which
exhibits GTPase activity and thereby conveys signals to
a cascade of mitogen activated protein (MAP) kinases.
Mutations of ras have the effect of constitutive
activation of Ras proteins [87], which in turn lead to
genomic instability and additional mutations and
malignant transformation [88]. Overexpression of ras
seems to be associated with normal human thyroid
growth as well as both benign and malignant neo-
plasia (see Table 4)[89]. Thyroid cells transformed
in vitro with mutant active ras demonstrate increased
proliferation as well as inhibited differentiation, both of
which are markers of the progression towards malig-
nancy [90,91]. Transgenic mice which overexpress ras
develop thyroid hyperplasia as well as follicular and
papillary malignancies [92,93].
In many series, up to 50% of thyroid follicular and
12% of H.urthle cell neoplasms have been reported to
harbor ras mutations (Table 4)[13,14,94–102]. Muta-
tions of ras might also be indicators of more aggressive
behavior, with an increased frequency shown in meta-
static tumors in some series [98,103] but not in others
[94,104].
3.2.4. BRAF
Related to the ras signaling pathway are the RAF
proteins, a family of homologous cytoplasmic serine/
threonine protein kinases which are regulated by
binding to Ras proteins. BRAF is the most efficient at
phosphorylating MAPK family kinases and is important
in apoptotic and proliferative pathways. Recent reports
in a majority of patients with malignant melanomas
demonstrate somatic, single substitution mutations
which result in BRAF signaling independent of binding
to RAS [105]. Similar findings have been described in
colorectal tumors [106] and, recently, in thyroid tumors.
Two separate groups report exciting results with 35–
70% mutations in PTC compared to no mutations in
any other types of thyroid lesions, including FC, HC,
MTC, and benign tumors [107,108]. These findings
highlight the importance of the RAS/RAF/MAPK
cascade in thyroid tumorigenesis, suggest a possible role
for RAF kinase inhibitors in novel treatment strategies
of PTC, and further support the concept of a progres-
sion model separating different subtypes of thyroid
cancer.
3.2.5. RET/PTC
The most frequent genetic abnormality described in
PTC involves ret, a proto-oncogene which encodes a cell
surface RTK. Although somatic mutations of the ret
gene have been described in some cases of MTC
[3,25,109], most abnormalities in ret function occur as
a result of genetic rearrangements at its chromosomal
locus. Fusion of the RTK domain to the 50-terminal
region of heterologous genes that are constitutively
expressed in thyroid follicular cells leads to the
ARTICLE IN PRESS
Table 4
Reported frequency of ras mutations in human thyroid tumors
Tumor type ras Mutations
Hot nodule 3/73 (4%)
Adenoma 48/167 (29%)
Follicular carcinoma 30/88 (34%)
Papillary carcinoma 29/172 (17%)
Anaplastic carcinoma 19/62 (31%)
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–9074

generation of chimeric oncogenes and proteins known as
RET/PTC (ret rearrangements in PTC) [3,25,110].A
number of RET/PTC chimeras have been described and
are listed in Table 5 [97,111–117].
Expression of RET/PTC chimeric proteins is facili-
tated by the heterologous promoters provided by the
fused genes and results in constitutive, ligand-indepen-
dent activation of ret RTK in papillary cancer cells
[3,109,110]. Several studies have proven specificity of the
RET/PTC rearrangement for thyroid malignancies,
demonstrating absence of these rearrangements in 250
non-thyroid tumors [10,22,118,119].
At least 16 studies have examined the frequency of
RET/PTC rearrangements in human thyroid cancers,
and the results are summarized in Table 6
[13,22,31,91,96,113,118–127]. Some reports have sug-
gested that patients with RET/PTC abnormalities
develop tumors with more propensity for aggressive
behavior and distant metastases [119]. There also
appears to be a strong association between exposure to
ionizing radiation, especially in childhood, and RET/
PTC rearrangements in PTCs [122,128]. Interestingly,
while radiation-induced ret rearrangements appear to
lead to PTCs, spontaneous ret mutations are also
associated with familial and sporadic MTC [3].
Clinically, RT-PCR for RET/PTC rearrangements is
a specific molecular method which has been successfully
applied to FNA biopsies and has been shown to
improve the diagnosis of malignancy when used as an
adjunct to FNA cytology [129]. Activation of these
chimeras is quite common in PTC specimens, and may
be useful in differentiating follicular lesions. One study
found these rearrangements in almost all HC with
associated lymph node metastases, a feature usually
more consistent with PTC, and absent in all other HC
examined [130]. A second study correlated RET/PTC
expression with H .urthle cell variant of PTC (HVPTC)
based on clinical, histological, and immunohistochem-
ical features [131].
3.2.6. TRK-T
The oncogene trk (NTRK1, trkA) encodes a cell
surface RTK protein that binds nerve growth factor
(NGF) and plays a cytoskeletal role [132]. Native
expression of this oncogene is restricted to the peripheral
nerve ganglia [133] but, similar to ret activation through
RET/PTC rearrangements, oncogenic activation of trk
can occur through creation of chimeric fusion proteins.
Similar to the RET/PTC paradigm, chromosomal
rearrangements which combine the 30RTK end of the
trk gene with the 50end and promoter region of a
ubiquitously expressed gene result in a constitutively
active RTK. Three types of activating rearrangements
have been described in thyroid tumors, and are
summarized in Table 7 [134–137].
Overexpression of the TRK-T1 chimera in cultured
thyroid cells reduces TG gene expression, but other
genetic events must occur before TRK-T1 can lead to
malignant transformation [138]. Only 3 studies have
examined these alterations in human thyroid tumors,
demonstrating no expression in hot nodules, adenomas,
or FC (0/12, 0/44, and 0/6, respectively), but rearrange-
ments in 11/130 PTC and 1/1 ATC [13,121,139].
3.2.7. Cyclin D1
The cyclin thought to be most important in develop-
ment of thyroid malignancies is cyclin D1, a 36 kDa
protein which facilitates progression of the cell cycle
through the G1-S checkpoint [140]. This occurs through
an interaction between cyclin D1 and its associated
CDKs, with subsequent phosphorylation and deactiva-
tion of the pRb retinoblastoma gene product. Inactiva-
tion of pRb allows progression through the G1 phase of
the cell cycle, or the transition from cell quiescence to
proliferation. The protein is encoded by the CCND1
gene, which is located on chromosome 11q23 and is
expressed at higher levels in a number of human
malignancies.
ARTICLE IN PRESS
Table 5
Summary of RET/PTC chromosomal rearrangements and their
resulting chimeric proteins
RET/PTC chromosomal rearrangements
RET/PTC1 10q paracentric inversion; H4 (ubiquitously expressed
gene of unknown function)
RET/PTC2 t(10;17)(q11.2;q23); regulatory subunit of cAMP
dependent protein kinase A
RET/PTC3 intrachromosomal rearrangements; ELEI, a gene of
unknown function
RET/PTC4 same as ret/PTC3
RET/PTC5 post-Chernobyl patients; RFG-5 (ret-fused gene 5)
RET/PTC6 t(7;10)(q32;q11.2); HTIF-I (human transcription
intermediary factor I)
RET/PTC7 t(1;10)(p13;q11.2); transcription coactivator for
nuclear receptors 6 & 7
RET/PTC8 t(10;14)(q11.2;q22.1); kinetin gene
Table 6
Reported frequency of RET/PTC rearrangements in human thyroid
tumors
Tumor type RET/PTC rearrangements
Adenoma 4/177 (2%)
Follicular carcinoma 0/44 (0%)
Papillary carcinoma 85/532 (16%)
Anaplastic carcinoma 0/20 (0%)
Table 7
Summary of TRK chromosomal rearrangements and their resulting
chimeric proteins
TRK chromosomal rearrangements
TRK-T1 Non-muscular tropomyosine (TPM3) gene
TRK-T2 Translocated promoter region (TPR) gene
TRK-T3 TRK-fused gene (TFG)
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–90 75

In the thyroid, cyclin D1 overexpression has been
shown by immunohistochemistry as well as Northern
blot analysis to correlate with prognosis and prolifera-
tive activity of thyroid neoplastic cells as compared with
adenomas [141]. One third of thyroid carcinomas
studied by Jiang et al. were found to overexpress cyclin
D1, with an inverse correlation between mutations of
Rb (a cell-cycle inhibitor discussed below) and expres-
sion of cyclin D1 [142]. Overexpression of cyclin D1 was
demonstrated to be an independent predictor of lymph
node metastases in a large series of PTC patients [143].
Furthermore, statistically significant increases in cyclin
D1 expression were seen in ATC compared to their less
invasive counterparts [144]. Muller et al. also demon-
strated significantly high levels of cyclin D1 in clinically
aggressive malignancies of the thyroid [145]. In one
study of 128 lesions, nuclear staining was seen in only
1.7% of HA compared with 18% of HC [146].
3.3. Functional proteins
Several proteins have been found to be overexpressed
in thyroid neoplasms, whose role in the pathogenetic
process is unclear, since they do not readily fit into
currently known signal transduction pathways or cell-
cycle regulation mechanisms. One has to assume that
their overexpression or abnormal expression provides
some growth or survival advantage. This abnormal
expression pattern may also provide useful diagnostic
and therapeutic targets.
3.3.1. Thyroid peroxidase
The enzyme thyroid peroxidase (TPO) is present in all
thyroid follicular cells and catalyzes iodide oxidation,
thyroglobulin iodination, and iodothyronine coupling.
Loss of expression is associated with reduction in iodide
trapping in thyroid follicles, impairment of thyroid
hormone synthesis, and a resultant decrease in auto-
regulation of most thyroid functions, including growth.
This enzyme is immunologically altered in malignancy,
preventing recognition by monoclonal antibodies raised
against normal TPO (MoAb-47) [147,148].
TPO expression and mutation seem important in the
progression and evaluation of thyroid lesions. Krohn
et al. demonstrated, by LOH and direct sequencing,
mutations in the TPO gene in cold thyroid nodules of
patients who otherwise lacked germline or somatic
mutations of the gene [147]. Two large FNA studies
have shown significantly increased accuracy of FNA
detection of FVPTC [149] and malignant thyroid tumors
in general [150], when cytological analysis is combined
with TPO immunohistochemistry. Furthermore, very
promising clinical results have been demonstrated for
differentiation of follicular lesions by TPO immunode-
tection. Of 5 studies performed on a total of almost 500
follicular thyroid tumors, 400 of 457 FA (88%) and 31
of 32 FC (97%) were accurately diagnosed by anti-TPO
antibody staining in both FNA and surgical specimens
[148,151–154]. Diagnostic criteria for FA included more
then 80% antibody staining. Others have suggested that
focality of staining is even more important than percent
area stained, demonstrating that most FC stain only
focally [155]. Furthermore, the false positive results seen
corresponded with atypical FA, consistent with the
progression model of adenoma to differentiated carci-
noma [152,154].
3.3.2. Ceruloplasmin and lactoferrin
Ceruloplasmin (CP) and the lactoferrin (LF) are
glycoproteins, which share a remarkable amino acid
sequence homology. CP is a 160 kDa serum enzyme
which binds most of the circulating blood copper. The
biological behavior of CP is similar to that of LF, an
iron-binding glycoprotein observed in salivary gland
secretions and milk. High levels of serum CP have been
observed in a number of neoplasms, including gastric,
colon, and pulmonary, and reflect the disturbance of CP
catabolism and the high concentrations of copper in the
cytoplasm of some neoplastic cells [156]. Increased levels
of LF have also been observed in various malignancies.
Immunohistochemical studies of resected surgical
tissue have compared expression of CP and LF in
follicular thyroid tumors, with very encouraging results.
Two groups have demonstrated 100% staining of 26 FC
for both CP and LF [157,158], with little or no staining
in 1 of 30 FA. Two other studies looked only at CP and
obtained similar results, with only one positive adenoma
and no negative carcinomas from 90 thyroid follicular
samples [156,159]. These results were confirmed in 36
additional tumors [160].
3.3.3. DAP4 (dipeptidyl-aminopeptiodase)
Dipeptidyl-aminopeptidase IV (DAP IV) is a highly
specific membrane-linked serine protease known to be
involved in T-cell activation. The activity of this
exopeptidase is increased in many cancers, including
liver and breast [161–163]. DAP IV is absent from
normal thyroid tissue but seems to be expressed at high
levels in malignant thyroid cells [164].
Most groups utilize a specialized stain described by
Lodja to assess enzymatic activity of DAP IV [165].A
grading system was chosen by each group, the most
common using a 12-point score based on the percentage
of positively stained epithelial cells and their staining
intensity, with a set cutoff point for malignancy [165]. In
two papers utilizing this technique, all 8 surgical
specimens of FC and only one of 37 FA met criteria
for malignancy [165]. Tang et al. prospectively evaluated
the utility of this system for FNA and frozen section
samples, demonstrating 4/6 FC and 6/7 FA accurately
diagnosed by FNA, while 7/11 FC and 7/7 FA were
correctly assessed by DAP IV staining of frozen section
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–9076

material [162]. A different FNA study showed 100%
accuracy in diagnosing 4 FC and 15 FA, with an overall
sensitivity and specificity of 94% and 92% for all types
of thyroid malignancies [164].
Using the same staining technique but a different
grading system, Gonzalez-Campora et al. showed
accurate diagnosis of 14 FA, 3 FC, 1 HA, and 1 HC,
with the highest staining in the HC. However, less
promising results were seen using a third grading system,
demonstrating 9 of 10 FC accurately diagnosed but 19
of 32 FA falsely classified as malignant [163]. Finally,
DAP IV expression has been studied with monoclonal
antibody immunohistochemistry with similar results,
demonstrating malignancy in all 33 FC and only 5 of 52
FA. Of interest, a third category of FA with incomplete
capsular invasion was examined in this study and 50%
of these tumors showed malignant-type evidence of
DAP IV staining [166].
3.3.4. HMGI proteins
The high mobility group I (HMGI) family is a class of
nuclear proteins which regulate chromatin structure and
formation [17,167]. Although they have no inherent
transcriptional activity, they seem to interact with a
number of transcription factors, including the HIPK2
serine threonine kinase, to regulate gene transcription
[168,169]. Normal physiologic expression of HMGI is
generally limited to embryogenesis, but activation has
been reported to occur in experimental cancer models as
well as in a number of human malignancies [170].
HMGI protein expression correlated with the appear-
ance of a malignant phenotype in rat thyroid trans-
formed cells [169]. Transfection with antisense rescued
thyroid cell lines from exhibiting a malignant pheno-
type [171] and induced apoptosis in 2 thyroid ATC cell
lines [172].
Moretti et al. demonstrated expression of HMGI in
thyroid malignancies but not in FA, and have proposed
the use of these proteins as molecular markers of
malignant thyroid transformation [3]. Immunohisto-
chemical studies for HMGI were positive in 18 of 19 FC
compared to only 44 of 200 FA and 0 of 12 normal
tissues. This correlated with immunohistochemistry
and RT-PCR of FNA samples from follicular lesions
[167]. Another study of frozen thyroid tissues
using semiquantitative RT-PCR demonstrated low
levels of HMGI in 8 FA and HA, with high levels in
5FC[173].
4. Factors that hinder tumor proliferation
Several general mechanisms are involved in suppres-
sion of cell proliferation and include cell-cycle regulation
(see Fig. 2b), cellular differentiation, and cell adhesion.
The inhibition of one or more of these mechanisms is a
common theme in tumor progression, and a number of
factors have been identified which hinder thyrocyte
proliferation. Cyclins are a class of cell-cycle modulators
that play a central role in controlling neoplastic growth
through interactions with cyclin-dependent kinases
(CDK). CDKs are activated after forming complexes
with the cyclin proteins, and each activated complex can
phosphorylate specific targets, thus facilitating progres-
sion through different phases of the cell cycle. Loss of
regulatory control of the cell cycle leads to abnormal
proliferation, a hallmark of oncogenesis. CDK inhibi-
tors (CDKI) hinder cell-cycle progression by impairing
the activity of cyclin/CDK complexes. Neoplastic, or
uncontrolled, growth has been associated with over-
expression of the cyclins D and E, which are positive
regulators of the cell cycles, or underexpression of the
CDKIs which are inhibitors of cell-cycle progression
[174].
4.1. Cell-cycle regulators
4.1.1. TGF-beta
The cytokine TGF-beta, a known epithelial cell
growth inhibitor, exerts its effect by binding to a type
II receptor which then recruits, phosphorylates, and
forms a heterodimeric complex with a type I receptor.
Activity of the TGF-beta type I receptor results in
induction of apoptosis and alteration of cell-cycle
regulatory pathways [175]. In the thyroid, TGF-beta
can selectively abrogate cyclic AMP pathways, inhibit
growth [176,177], and regulate differentiation
[10,11,178]. Alterations can occur at a ligand level, with
altered TGF-beta expression seen in malignant thyroid
lesions when compared with normal epithelium [179],or
at the level of the type II receptor, with an indirect
correlation noted between size of PTC and expression of
TGF-beta type II receptor [180].
4.1.2. p21
Another inhibitor of the cell cycle is the CDKI p21,
an effector of G1 cycle arrest mediated by p53. Tumors
with documented mutations of p53, a growth inhibitor
and tumor suppressor, are associated with lower levels
of p21 [181]. This CDKI is rarely seen in normal thyroid
tissue or benign proliferative lesions such as goiter, but
immunohistochemical expression has been documented
in one third of thyroid cancers [182].
4.1.3. p27
KIP1 (kinase inhibitor protein), also known as p27,
is a tumor suppressor gene which encodes a nuclear
protein member of the CDKIs. These proteins nega-
tively regulate cyclin activity and control G1 to S phase
progression. In resting cells, as compared with neoplas-
tic ones, p27 expression is high [183]. Levels of p27 fall
when the cell enters the cell cycle and have been shown
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–90 77

to correlate with loss of differentiation by histological
evaluation [184]. Furthermore, expression of p27 may be
an important prognostic factor in predicting behavior of
many types of carcinomas [185].
Levels of p27 were shown to be decreased in several
studies of thyroid tumors when compared with normal
thyroid tissue, with lowest p27 staining occurring in
poorly differentiated carcinomas [183,186]. A study of
p27 immunohistochemistry in FC showed a strong
inverse correlation between p27 expression and tumor
prognosis [184]. Underexpression of p27 in PTC was an
independent predictor of lymph node metastases [187].
Of note, measurement of p27 levels is possible in FNA
[188] and has been predicted to be a useful diagnostic
and prognostic tool for thyroid malignancies. Immu-
nostaining of 39 follicular lesions and logistic regression
analysis of p27 label showed that this method was
effective in distinguishing FA from FC, with labeling
indices of 48 for FA compared with 16 for FC [185]. Of
59 FVPTC, varying levels of p27 staining were seen,
with strong positive staining exhibited in only 4 cases,
compared to 53 of 57 FA [186].
4.1.4. Rb
The retinoblastoma gene (Rb) encodes a nuclear
phosphoprotein which plays a major inhibitory role in
cell-cycle regulation [189,190]. In its underphosphory-
lated state, Rb binds and thereby inactivates both cyclin
and E2F proteins [191,192]. This usually occurs in
resting cells, in phase G0 or G1 of the cell cycle [193].
Various factors influence Rb phosphorylation, including
interaction with viral oncoproteins (SV40 large T
antigen [194], human adenovirus EA1 [195], and
papillomavirus E7 [196]) or the cyclin family [192,197].
In a cell-cycle specific manner, these factors can bind to
the underphosphorylated form [198] and, in phosphor-
ylating Rb, can cause release of the bound cyclin or
E2F, freeing these proteins to facilitate progression
through the G1/S phase checkpoint of the cell cycle
[192,199–201] Transgenic mice overexpressing the SV40
large T antigen in a thyroid tissue specific manner were
found to develop thyroid tumors [202]. Similarly,
overexpression of HPV type 16E7, which inactivates
Rb in a specific p53-independent manner, led to the
development of toxic thyroid lesions and carcinomas
[203].
In human thyroid tumors, some groups have found
no mutations in benign lesions, but in 55% of thyroid
carcinomas, with a higher propensity for mutations in
more clinically aggressive tumors [11,21]. However,
other groups were unable to demonstrate any genetic
abnormalities of Rb or loss of Rb protein expression by
immunohistochemistry [31,204]. Although evidence
suggests that this protein likely plays an important role
in regulation of thyrocyte cell-cycle progression, its
association with human thyroid oncogenesis remains to
be better elucidated.
4.1.5. p53
The policeman of the genome is p53, a tetrameric
nuclear phosphoprotein transcription factor [11]. Cells
with genetic injury as a result of radiation [205],
ultraviolet light or mimetic drugs [206], or other stresses
[207] appear to rely on an upregulation of p53
expression to facilitate DNA repair. These include
transcriptional activation of cell-cycle inhibitors, an
arrest of the G1-S transition point to allow for effective
stimulation of native DNA repair mechanisms, and even
apoptosis in the case of unrecoverable DNA alterations
[207–209]. It would follow that cells lacking p53 activity
(thereby lacking repair mechanisms) would be more
prone to accumulation of genetic injury and subsequent
malignant behavior [210]. In fact, p53 is the most
frequently altered gene in human tumors, with somatic
mutations described in over half of a wide variety of
human lesions [211]. Genomic mutations of p53 have
been described in Li-Fraumeni syndrome, a familial
syndrome predisposing patients to breast carcinomas,
brain lesions, sarcomas, and other diverse tumors
[212,213]. Mutations of p53 are considered late events
in the sequence of human carcinogenesis [214].
Immortalized thyroid carcinoma cell lines demon-
strated homozygous p53 abrogation in three of four
lines studied [20,95]. In a separate study, cells grown
from poorly differentiated human thyroid lesions were
successfully rescued by p53 overexpression, resulting in
less proliferation and malignant behavior [215]. In
addition to its important role in DNA injury repair,
p53 has also been implicated in the cell differentiation
process, as shown in may cell types including hemato-
poietic and muscular cell lines [216]. This is corrobo-
rated by reports in which knockout mice lacking
expression of the p53 gene develop aggressive, poorly
differentiated tumors [217]. Re-expression of this gene in
a cell line derived from an ATC can rescue cells and
restore expression of differentiated genes such as TPO
and Pax-8 [218].
4.2. Cellular differentiation and adhesion
Cellular differentiation, and terminal differentiation
in particular, is usually associated with decreasing
proliferative potential. Similarly, the establishment of
cellular adhesion and interactions requires the exit from
active cycling. Epithelial cells show anchorage depen-
dent cell survival, which is mediated by integrin
receptor-mediated interactions with the extracellular
matrix. Their disruption leads to anoikis, a specialized
version of apoptotic cell death [219]. It is therefore
necessary for neoplastic cells to overcome the growth
inhibition of these factors in carcinogenesis.
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–9078

4.2.1. Pax-8/TTF-1
Recent studies have identified two important tran-
scription factors, Pax-8 and TTF-1, which bind to the
promoter and enhancer sequences of thyroglobulin
(TG), TPO, the Na(+)/I() symporter (NIS), which
are central to thyroid development and function
[220–224], and may be responsible for their regulation
[221–225]. Pax-8 is a paired-domain transcription factor
expressed in the developing and adult thyroid and is
essential for thyroid follicular cell development and
regulation thyroid gene expression [3,226]. Mutations of
Pax-8 have been described in thyroid dysgenesis and
congenital hypothyroidism [227]. Analysis of homozy-
gous Pax-8 knockout mice demonstrates that this
transcription factor is required for the formation of
follicular cells in the thyroid, providing cues for the
differentiation of endoderm primordia [228].
Puglisi et al. investigated over 50 thyroid tumors using
immunoperoxidase staining for Pax-8 and demonstrated
a statistically significant difference in Pax-8 expression
between benign and malignant thyroid disease. Two
other studies confirmed a noticeable decrease in Pax-8
expression with increasing aggressiveness of tumors
[229,230]. Pax-8 may also follow a paradigm in follicular
thyroid neoplasia similar to RET/PTC and TRK-T1
rearrangements in PTC. A chromosomal translocation
forming a fusion protein between Pax-8 and the
peroxisome proliferator-activated receptor (PPAR-
gamma-1, PPARg1) has been described in FC but not
in benign lesions, FA, or PTC [23].
TTF-1 is a homeodomain-containing protein ex-
pressed in embryonic diencephalon, thyroid, and lung
[225]. Mice lacking TTF-1 lack both types of thyroid
cells and fail to develop a thyroid gland [228]. Like
Pax-8 expression, TTF-1 is lower in thyroid carcinomas
than in adenomas [230].
4.2.2. E-Cadherin
The initial step in aggressive tumor behavior is the
dissociation of neoplastic cells from the primary tumor
as a result of broken cell–cell adhesion. The cell
adhesion system includes the cadherin family, a group
of functionally related glycoproteins which are fre-
quently absent or decreased in various epithelial tumor
cells [231]. E-cadherin (uvomorulin) is the dominant
member of this family, expressed by epithelial cells and
essential in maintaining cell polarity and epithelial
integrity.
There is evidence to suggest that decrease in E-
cadherin is important specifically in epithelial thyroid
tumors. Early, less aggressive FC showed greater E-
cadherin expression compared to widely invasive FC
[232]. Normal tissue, with 100% preserved E-cadherin
expression, was compared with follicular lesions,
demonstrating reduction of E-cadherin levels in 100%
of ATC and 50% of follicular lesions [233]. Several
studies have shown that decreased E-cadherin expres-
sion was associated with advanced tumors, higher rates
of synchronous lymph node involvement, and distant
metastases in thyroid cancer [234–236]. However, one
study using immunofluorescence microscopy and Wes-
tern blot failed to demonstrate this [237]. In addition to
studying levels of E-cadherin expression, Graff et al.
showed that, even though tumors may still express
E-cadherin, aberrant methylation of the E-cadherin 50
CpG island in 83% of PTC, 11% of FC, 40% of HC,
and 21% of ATC might cause dysfunction of this
complex and resultant malignant behavior [238].
4.2.3. Galectins
Lectins are carbohydrate-binding proteins that recog-
nize specific oligosaccharide structures on glycoproteins
or glycolipids and facilitate certain cellular functions
such as cell–cell and cell–matrix interactions. Alteration
of glycoconjugates by glycosyltransferases or glycosi-
dases may lead to altered or lost cellular functions and
subsequent malignant behavior [239–241]. The galectins
are a growing family of proteins, which have been
implicated in regulation of cellular growth, differentia-
tion, and malignant transformation in a number of
tissues [239,242,243]. Increased expression of galectins
has been shown in transformed thyroid cells and thyroid
carcinomas, and distribution of these proteins in cancer
samples has been shown to localize to tumor, but not
adjacent regions of normal thyroid [239,241,242]. Most
studies of thyroid tumors have investigated expression
of galectin-3, that has been shown to act as a cell death
suppressor interfering with an apoptotic pathway
involving bcl-2 [244].
A metaanalysis of all immunohistochemistry per-
formed in surgically resected specimens of thyroid
follicular lesions indicates that 22 of 148 FA expressed
galectin-3, while 65 of 75 FC or HC expressed this
protein [245]. These results are promising, with 100%
diagnostic accuracy seen in 17 follicular tumors by Xu
et al. [241], in 25 follicular lesions by Inohara et al. [246],
and in 51 follicular lesions using RT-PCR [239]. Two
studies were not included in this metaanalysis, one of
which showed 78% positive FC and only 11% positive
FA, with similar results in H .urthle cell lesions (59% and
7%, respectively), and another which showed galectin-3
staining in 54 of 57 carcinomas, but only 4 of 125
adenomas [243,247].
4.2.4. CD44
CD44 is a polymorphic family of immunologically
related integral membrane glycoproteins associated with
cell–matrix adhesion, lymphocyte activation and target-
ing, cell migration, and tumor growth and metastasis
[248]. CD44 can be expressed on the cell surface as a
standard receptor (CD44 s, the putative receptor for
hyaluronic acid), as well as multiple isoforms (CD44v),
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–90 79

the expression of which is qualitatively and quantita-
tively altered during tumor growth and progression
[243]. The heterogeneity of CD44 is a consequence of
both alternative splicing as well as posttranslational
modification [249,250]. Aberrant variants of these
products have been postulated to regulate growth
patterns and metastatic potential of thyroid tumors
[250,251].
5. Factors affecting cellular immortalization and death
In addition to acquiring proliferative factors and
overcoming tumor suppressive factors that would
provide a differential growth advantage, neoplastic cells
must acquire the ability to outlive the biologic clock set
for all replicating non-germline cells, and to sidestep
several control checkpoints that can trigger the apopto-
tic cascade. This can be done through several pathways,
including stabilization of the telomere, which is usually
critically shortened after the multiple replication cycles
necessary for acquiring the malignant phenotype.
Instead of repairing DNA damage incurred during
uncontrolled proliferation, cancer cells typically elim-
inate DNA-damage checkpoint controls. Finally,
apoptosis, or programmed cell death, is one of the most
important cellular events in preventing uncontrolled
growth. On a molecular level, this occurs as a result of
genomic DNA cleavage by endonucleases and is closely
regulated in normal cells. Apoptosis can be documented
morphologically by cytoplasmic membrane blebbing,
nuclear condensation, and cellular contraction [252].
Cancer cells have developed numerous strategies to
avoid triggering these events.
5.1. Cellular immortalization
5.1.1. Telomerase
Telomerase is responsible for maintaining the length
of chromosomal telomeres by adding back (TTAGGG)n
hexamer repeats in a dynamic equilibrium with the
sequences lost during DNA replication. It is a specia-
lized reverse transcriptase with an integral RNA moiety
that serves as its own template, and is detectable in germ
cells, stem cell containing tissues with rapid cellular
turnover, immortal cell lines, and cancer cells, but is
undetectable in most normal differentiated somatic
tissues [253–258]. Telomerase activity is essential to cell
survival during neoplastic progression, possibly by
delaying cellular senescence long enough to allow the
accumulation of the multiple genetic alterations
required for the malignant phenotype [259,260]. While
there are alternative, telomerase-independent mechan-
isms that allow the maintenance of viable telomeres and
cellular immortality [261,262], they appear to be used
only occasionally in human cancer.
In well-differentiated thyroid cancer, telomerase en-
zyme activity has been demonstrated in 100% of FCs
and 67% of PTCs [263,264]. More recently, because the
telomerase enzyme assay is both plagued with non-
specific inhibitors and can be falsely positive due to
lymphocytic infiltration, our group examined the
feasibility of hTERT gene detection in thyroid FNA
from suspicious thyroid nodules by RT-PCR and
comparing these results with final histopathology [265].
Overall, the sensitivity and specificity for hTERT gene
expression in the detection of thyroid malignancy was
90% and 60%, respectively (unpublished data). In light
of the rarity of alternative pathways of immortalization
in human cells, the fairly high incidence of telomerase
negative PTCs raises the question whether tumor
progression for at least some PTCs may be unusually
shortened, occurring before critical shortening of
telomeres, and may point to a subset of PTCs with
limited growth potential.
5.2. Regulators of apoptosis
5.2.1. BCL family
Enhanced growth in B-cell follicular lymphomas
through inhibition of programmed cell death was
originally attributed to Bcl-2, the first molecule de-
scribed in the Bcl family [266]. Subsequently, a number
of homologous proteins were identified, some of which
are anti-apoptotic (Bcl-2, Bcl-x, Bcl-w) and some of
which are pro-apoptotic (Bax, Bak, Bok, Bad, Blk)
[267]. Like Fas/Fas-L, these proteins interact with
caspases to regulate apoptosis [267]. The Bcl family
has been associated with thyroid lesions, but the
relationship of Bcl family expression and malignant
behavior remains controversial. Both expression of the
anti-apoptotic Bcl-x as well as the pro-apoptotic Bax
have been associated with FC [268]. Some studies have
claimed a direct correlation between Bcl-2 expression
and differentiation [268], which would support its role in
growth promotion, while others have found an increased
level of Bcl-2 expression in undifferentiated thyroid
tumors [269].
5.2.2. Fas/Fas-L
Fas ligand (Fas-L) binds one of a family of
homotrimeric receptor proteins known as Fas. The
Fas/Fas-L pathway promotes apoptosis by activating
the caspases and their associated pathways [270]. Fas/
Fas-L interactions have been implicated in the develop-
ment of Hashimoto’s thyroiditis, PTC, and FC. It is
speculated that tumor infiltrating immunocytes in the
vicinity of proliferating thyroid lesions may be sensitive
to the actions of Fas, thus allowing growth of these
lesions to evade the immune system [271,272]. Clinically,
Mitsiades et al. found that Fas-L expression correlated
with a more aggressive phenotype in PTC [273].
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–9080

5.2.3. PTEN
PTEN is a dual-function lipid phosphatase acting on
the phosphoinositide 3-kinase (PI3K) and the Akt/PKB
pathways [274]. PTEN dephosphorylates Akt (Akt-P),
causing apoptosis and/or G1 cell-cycle arrest. Conse-
quently, dysfunctional PTEN leads to high levels of
Akt-P, which inhibits apoptosis [275,276]. Both benign
and malignant thyroid disease are well-established
components of the autosomal dominant disorder
Cowden syndrome (CS), characterized by multiple
hamartomas and breast cancer [277]. Although somatic
PTEN mutations are rare in primary epithelial thyroid
tumors, hemizygous deletion occurs in 10–20% of
thyroid adenomas and carcinomas. Expression and
genetic analysis of 139 benign and malignant thyroid
tumors showed that FA, FC, and PTC all had a 20–30%
frequency of hemizygous deletion, while 60% of ATC
had hemizygous PTEN deletions associated with de-
creased PTEN expression. Epigenetic silencing of
PTEN and perhaps inappropriate subcellular compart-
mentalization are two novel mechanisms of PTEN
inactivation that appear pertinent to thyroid carcino-
genesis [278–280].
5.2.4. MDM2
Overexpression of MDM2 is generally linked with an
advantage in cell proliferation and predisposition to
tumorigenesis. The mdm2 oncogene is activated by
overexpression, amplification or enhanced translation
in different types of human tumors, mainly sarcomas
[281]. MDM2 plays a pivotal role in cell proliferation
through functional interactions with the tumor suppres-
sor gene product p53, and possesses transforming
activity [282]. This interaction results in inactivation of
p53 function by a mechanism involving protein degra-
dation [283,284]. MDM2 also interacts with other
proteins, such as retinoblastoma, E2F-1, the TATA
box binding protein and the ribosomal protein L5 [285].
A handful of studies have investigated the role of
MDM2 in thyroid tumors [286–290]. A consistent
finding was that in the majority of cases overexpression
of MDM2 was correlated with p53 overexpression.
MDM2 overexpression only was found in a smaller
subset of cases [286,290]. A study of FA found MDM2
overexpressed in the absence of p53 overexpression in
19% of cases [289]. Interestingly, a recent tissue
microarray analysis of H .urthle cell tumors found a
similar pattern of gene expression [291].
6. Conclusions
As is befitting for a complex process such as thyroid
carcinogenesis, a large number of molecular alterations
have been identified in the various thyroid neo-
plasms. Furthermore, it is highly likely that the advent
of large-scale genomic screening will bring forth an
abundance of additional data that will need to be
considered as well. It is therefore useful to attempt
to organize this molecular information into models of
molecular carcinogenesis as we currently understand
them (see Figs. 1 and 2a)[18,19,292,293].
It is generally accepted that the carcinogenic process
proceeds through multiple discrete steps, since experi-
mental data suggest that a single oncogenic mutation is
not sufficient to induce malignant transformation.
Disruptions including combinations of hyperfunctional
proliferative factors with downregulated growth sup-
pressive factors lead to a proliferative advantage and
clonal expansion in one or multiple foci. As the
proliferating cells approach their replicative limits and
accumulate genetic damage, the neoplastic clones must
achieve immortalization and overcome apoptotic signals
to continue their growth.
The fundamental questions that need to be addressed
are threefold: (1) What are the molecular events that
determine if a hyperproliferative focus of cells will
develop? (2) What changes determine if this focus
will remain benign or will progress to malignancy? and
(3) What determines the histological subtype and the
clinical behavior of the emerging carcinoma?
There is no absolute proof that the transition from
benign to premalignant to malignant and finally invasive
and metastatic occurs in a predictable and orderly
fashion, but the existence of intermediate phenotypes, at
least in the case of follicular neoplasms, is consistent
with this microevolutionary concept. The lack of
recognizable precursors of PTCs, except perhaps for
the occasional papillary microcarcinoma, for which
there are no specific genetic data so far, and follicular
variants of papillary carcinoma (FVPTCs), makes it
more difficult to reconstruct the sequence of the genetic
steps required to develop PTCs.
As discussed in the introduction, the carcinogenesis of
thyroid tumors reflects a unique interaction between
environmental factors such as ionizing radiation, dietary
factors, genetic and epigenetic events. Their interplay
defines a molecular pathway that leads to either FC or
PTC. On the one hand, dietary influences such as iodine
deficiency in combination with activation of ras and
other oncogenes and the inactivation of tumor suppres-
sor genes such as p16 or Pax-8, seem to play a role in the
development of FC. On the other hand, genetic
rearrangements of one of several membrane RTKs
(ret, trk, met), which have been specifically linked to
ionizing radiation, play a role in the development of
PTC. Moreover, the recent finding that the ras/MEK/
MAP Kinase pathway is also implicated in a majority of
PTC, including FVPTC, in the form of BRAF muta-
tions identifies it as a central player in thyroid
carcinogenesis, and raises the possibility that the nature
of the alteration of this signaling pathway may also
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–90 81

determine which histological subtype ultimately evolves.
This also raises the possibility of a thyroid tumor
progression model that includes genetic events leading
from normal thyroid to FA to FVPTC to PTC and,
eventually, ATC. Finally, additional ‘‘hits’’ in other
pathways, such as immortalization (telomerase activa-
tion) and apoptosis (bcl2, fas-ligand), tumor suppressor
genes (p53), and genes involved in tissue invasion
(E-cadherin, CD44), determine clinical behavior and
dedifferentiation of thyroid cancers, such as the evolu-
tion to metastatic disease or the progression to ATC.
Not all clinically useful tumor markers will fall into
one or several of these theoretical tumor progression
categories, since one can anticipate that some of these
markers, which fundamentally would include anything
that distinguishes tumors from normal or benign cells,
may be altered as ‘‘collateral damage’’, without playing
any prominent role in the carcinogenic process. Never-
theless, we have attempted to organize thyroid-specific
molecular markers into these categories, with the
understanding that this will need to be revised as our
understanding of the role these factors play in thyroid
carcinogenesis is refined.
7. Abbreviations
AC Adenyl cyclase
ATC Anaplastic thyroid carcinoma
cAMP Cyclic adenosine monophosphate
CDK Cyclin dependent kinase
CDKI Cyclin dependent kinase inhibitor
cDNA Complementary deoxyribonucleic acid
CP Ceruloplasmin
DAP4 Dipeptidyl aminopeptiodase 4
DNA Deoxyribonucleic acid
E2F Adenovirus E1A inducible factor
EGF Epidermal growth factor
FA Follicular adenoma
FC Follicular carcinoma
FVPTC Follicular variant of papillary thyroid carci-
noma
GDP Guanosine diphosphate
gsp G protein subunit alpha
GTP Guanosine triphosphate
HC H.urthle cell carcinoma
HGF/SF Hepatocyte growth factor/scatter factor
HMGI High mobility group I
HPV Human papilloma virus
HVPTC H.urthle cell variant of papillary thyroid
carcinoma
IGF-1 Insulin-like growth factor 1
KIP Kinase inhibitor protein
LF Lactoferrin
MAP Mitogen activating protein
MAPK Mitogen activating protein kinase
MTC Medullary thyroid carcinoma
mRNA Messenger ribonucleic acid
NIS Human Na(+)/I() symporter
PCR Polymerase chain reaction
PI3K Phosphoinositol 3-kinase
PKB Protein kinase B
PLC Phospholipase C
PPAR Peroxisome proliferator activated receptor
PTC Papillary thyroid carcinoma
Rb Retinoblastoma
RNA Ribonucleic acid
RTK Receptor tyrosine kinase
RT-PCR Reverse-transcriptase polymerase chain re-
action
SV40 Simian virus 40
TG Thyroglobulin
TGF-beta Transforming growth factor beta
TPO Thyroid peroxidase
TSH Thyroid stimulating hormone (Thyrotropin)
TSH-R Thyroid stimulating hormone receptor
TTF-1 Thyroid transcription factor 1
VEGF Vascular endothelial growth factor
References
[1] Sarlis NJ. Expression patterns of cellular growth-controlling
genes in non-medullary thyroid cancer: basic aspects. Reviews in
Endocrine Metabolic Disorders 2000;1(3):183–96.
[2] Figge JT, Jennings T, Gerasimov G. Radiation and thyroid
cancer. In: Wartofsky L, editor. Thyroid cancer. Human Press:
Totowa, NJ; 2000. p. 515.
[3] Moretti F, Nanni S, Pontecorvi A. Molecular pathogenesis of
thyroid nodules and cancer. Baillieres Best Practical Research
Clinical Endocrinology and Metabolism 2000;14(4):517–39.
[4] Ron E, Kleinerman RA, Boyce JD. A population-based case-
controlled study of thyroid cancer. Journal of the National
Cancer Institute 1987;79:1–4.
[5] Wegelin C. Malignant disease of the thyroid and its relation to
goiter in men and animals. Cancer Reviews 1929;3:297–301.
[6] Braverman LE, Utiger RD, editors. Werner & Ingbar’s The
Thyroid, 8th ed. Philadelphia, PA: Lippincott; 2000. p. 1081.
[7] Bakiri F, et al. The relative roles of endemic goiter and
socioeconomic development status in the prognosis of thyroid
carcinoma. Cancer 1998;82:1146–52.
[8] Fagin JA. Molecular genetics of tumors of thyroid follicular
cells. In: Braverman LE, Utiger RD, editors. Werner & Ingbar’s
the thyroid. Philadelphia, PA: Lippincott, Williams & Wilkins;
2000. p. 886–98.
[9] Dumont JE, et al. Growth factors controlling the thyroid gland.
Baillieres Clinical Endocrinology and Metabolism
1991;5(4):727–54.
[10] Fagin JA. Molecular genetics of human thyroid neoplasms.
Annual Review of Medicine 1994;45(1):45–52.
[11] Farid NR, Shi Y, Zou M. Molecular basis of thyroid cancer.
Endocrine Reviews 1994;15(2):202–32.
[12] Frauman AG, Moses AC. Oncogenes and growth factors in
thyroid carcinogenesis. Endocrinology and Metabolism Clinics
of North America 1990;19(3):479–93.
[13] Said S, Schlumberger M, Suarez HG. Oncogenes and anti-
oncogenes in human epithelial thyroid tumors. Journal of
Endocrinological Investigation 1994;17(5):371–9.
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–9082

[14] Su!
arez HG. Genetic alterations in human epithelial thyroid
tumours. Clinical Endocrinology (Oxford) 1998;48(5):531–46.
[15] Wright PA, et al. Mutation of the p53 gene in a differentiated
human thyroid carcinoma cell line, but not in primary thyroid
tumours. Oncogene 1991;6(9):1693–7.
[16] Wynford-Thomas D. Origin and progression of thyroid epithe-
lial tumours: cellular and molecular mechanisms. Hormone
Research 1997;47(4–6):145–57.
[17] Ringel MD. Molecular diagnostic tests in the diagnosis and
management of thyroid carcinoma. Reviews of Endocrine
Metabolic Disorders 2000;1(3):173–81.
[18] Hahn WC, et al. Creation of human tumour cells with defined
genetic elements. Nature 1999;400(6743):464–8.
[19] Weinberg RA. Oncogenes, antioncogenes, and the molecular
bases of multistep carcinogenesis. Cancer Research 1989;49:
3713–21.
[20] Fagin JA, et al. High prevalence of mutations of the p53 gene in
poorly differentiated human thyroid carcinomas. Journal of
Clinical Investigation 1993;91(1):179–84.
[21] Zou M, et al. Inverse association between cyclin D1 over-
expression and retinoblastoma gene mutation in thyroid
carcinomas. Endocrine 1998;8(1):61–4.
[22] Santoro M, et al. Ret oncogene activation in human thyroid
neoplasms is restricted to the papillary cancer subtype. Journal
of Clinical Investigation 1992;89(5):1517–22.
[23] Kroll TG, et al. PAX8-PPARgamma1 fusion oncogene in
human thyroid carcinoma. Science 2000;289(5483):1357–60.
[24] Rabes HM. Gene rearrangements in radiation-induced thyroid
carcinogenesis. Medical and Pediatric Oncology 2001;36(5):
574–82.
[25] Learoyd DL, et al. Molecular genetics of thyroid tumors and
surgical decision-making. World 2000;24(8):923–33.
[26] Aaronson SA. Growth factors and cancer. Science 1991;
254(5035):1146–53.
[27] Roger PP, Dumont JE. Factors controlling proliferation and
differentiation of canine thyroid cells cultured in reduced
serum conditions: effects of thyrotropin, cyclic amp and growth
factors. Molecular and Cellular Endocrinology 1984;36(1–2):
79–93.
[28] Vassart G, Dumont JE. The thyrotropin receptor and the
regulation of thyrocyte function and growth. Endocrine Reviews
1992;13(3):596–611.
[29] Parma J, et al. Somatic mutations in the thyrotropin receptor
gene cause hyperfunctioning thyroid adenomas. Nature
1993;365(6447):649–51.
[30] Parma J, et al. Somatic mutations causing constitutive activity of
the thyrotropin receptor are the major cause of hyperfunctioning
thyroid adenomas: identification of additional mutations acti-
vating both the cyclic adenosine 30,50-monophosphate and
inositol phosphate-Ca
2+
cascades. Molecular Endocrinology
1995;9(6):725–33.
[31] Arturi F, et al. Iodide symporter gene expression in human
thyroid tumors. Journal of Clinical Endocrinology and Meta-
bolism 1998;83(7):2493–6.
[32] Paschke R, et al. Identification and functional characterization
of two new somatic mutations causing constitutive activation of
the thyrotropin receptor in hyperfunctioning autonomous
adenomas of the thyroid. Journal of Clinical Endocrinology
and Metabolism 1994;79(6):1785–9.
[33] Porcellini A, et al. Novel mutations of thyrotropin receptor gene
in thyroid hyperfunctioning adenomas. Rapid identification by
fine needle aspiration biopsy. Journal of Clinical Endocrinology
and Metabolism 1994;79(2):657–61.
[34] Russo D, et al. Genetic alterations in thyroid hyperfunctioning
adenomas. Journal of Clinical Endocrinology and Metabolism
1995;80(4):1347–51.
[35] Russo D, et al. Activating mutations of the TSH receptor in
differentiated thyroid carcinomas. Oncogene 1995;11(9):1907–11.
[36] Takeshita A, et al. Rarity of oncogenic mutations in the
thyrotropin receptor of autonomously functioning thyroid
nodules in Japan. Journal of Clinical Endocrinology and
Metabolism 1995;80(9):2607–11.
[37] Spambalg D, et al. Structural studies of the thyrotropin receptor
and Gs alpha in human thyroid cancers: low prevalence of
mutations predicts infrequent involvement in malignant trans-
formation. Journal of Clinical Endocrinology and Metabolism
1996;81(11):3898–901.
[38] Brabant G, et al. Human thyrotropin receptor gene: expression
in thyroid tumors and correlation to markers of thyroid
differentiation and dedifferentiation. Molecular and Cellular
Endocrinology 1991;82(1):R7–12.
[39] Shi Y, Zou M, Farid NR. Expression of thyrotrophin receptor
gene in thyroid carcinoma is associated with a good prognosis.
Clinical Endocrinology (Oxford) 1993;39(3):269–74.
[40] Shi Y, et al. Studies of the human TSH receptor gene in
physiology and pathology. Hormone and Metabolic Research
1993;25(12):632–6.
[41] Bronnegard M, et al. Expression of thyrotropin receptor and
thyroid hormone receptor messenger ribonucleic acid in normal,
hyperplastic, and neoplastic human thyroid tissue. Journal of
Clinical Endocrinology and Metabolism 1994;79(2):384–9.
[42] Cai WY, et al. Analysis of human TSH receptor gene and RNA
transcripts in patients with thyroid disorders. Autoimmunity
1992;13(1):43–50.
[43] Bourne HR, Sanders DA, McCormick F. The GTPase super-
family: conserved structure and molecular mechanism. Nature
1991;349(6305):117–27.
[44] Suarez HG, et al. Gsp mutations in human thyroid tumours.
Oncogene 1991;6(4):677–9.
[45] Hamacher C, et al. Expression of functional stimulatory guanine
nucleotide binding protein in nonfunctioning thyroid adenomas
is not correlated to adenylate cyclase activity and growth of these
tumors. Journal of Clinical Endocrinology and Metabolism
1995;80(5):1724–32.
[46] Yoshimoto K, et al. Rare mutations of the Gs alpha subunit
gene in human endocrine tumors. Mutation detection by
polymerase chain reaction-primer-introduced restriction analy-
sis. Cancer 1993;72(4):1386–93.
[47] O’Sullivan C, et al. Activating point mutations of the gsp
oncogene in human thyroid adenomas. Molecular Carcinogen-
esis 1991;4(5):345–9.
[48] Lyons J, et al. Two G protein oncogenes in human endocrine
tumors. Science 1990;249(4969):655–9.
[49] Zeiger MA, et al. Thyroid-specific expression of cholera toxin A1
subunit causes thyroid hyperplasia and hyperthyroidism in
transgenic mice. Endocrinology 1997;138(8):3133–40.
[50] Michelin S, et al. Characterization of a c-met proto-oncogene
activated in human xeroderma pigmentosum cells after treat-
ment with N-methyl-N0-nitro-N-nitrosoguanidine (MNNG).
Oncogene 1993;8(7):1983–91.
[51] Goretzki PE, et al. Mutational activation of RAS and GSP
oncogenes in differentiated thyroid cancer and their biological
implications. World Journal of Surgery 1992;16:576–82.
[52] Clementi E, et al. A new constitutively activating mutation of the
Gs protein alpha subunit-gsp oncogene is found in human
pituitary tumours. Oncogene 1990;5(7):1059–61.
[53] Zeiger MA, et al. Transformation of rat thyroid follicular cells
stably transfected with cholera toxin A1 fragment. Endocrinol-
ogy 1996;137(12):5392–9.
[54] Duh QY, et al. Epidermal growth factor receptors and adenylate
cyclase activity in human thyroid tissues. World Journal of
Surgery 1990;14(3):410–7 (discussion 418).
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–90 83

[55] Lamy F, et al. Control of protein synthesis by thyrotropin and
epidermal growth factor in human thyrocytes: role of mor-
phological changes. Molecular and Cellular Endocrinology 1990;
73(2–3):195–209.
[56] Contor L, et al. Differential protein phosphorylation in
induction of thyroid cell proliferation by thyrotropin, epidermal
growth factor, or phorbol ester. Molecular and Cellular Biology
1988;8(6):2494–503.
[57] Lemoine NR, et al. Abnormalities of the EGF receptor system in
human thyroid neoplasia. International Journal of Cancer
1991;49(4):558–61.
[58] Mizukami Y, et al. Immunohistochemical demonstration of
epidermal growth factor and c-myc oncogene product in normal,
benign and malignant thyroid tissues. Histopathology 1991;
18(1):11–8.
[59] Di Carlo A, et al. Epidermal growth factor receptor and
thyrotropin response in human thyroid tissues. Journal of
Endocrinological Investigation 1990;13(4):293–9.
[60] Aasland R, et al. Expression of oncogenes in thyroid tumours:
coexpression of c-erbB2/neu and c-erbB. British Journal of
Cancer 1988;57(4):358–63.
[61] Bergstrom JD, Westermark B, Heldin NE. Epidermal growth
factor receptor signaling activates met in human anaplastic
thyroid carcinoma cells. Experimental Cell Research 2000;
259(1):293–9.
[62] Bunone G, et al. Expression of angiogenesis stimulators and
inhibitors in human thyroid tumors and correlation with clinical
pathological features. American Journal of the Pathology
1999;155(6):1967–76.
[63] Fenton C, et al. The expression of vascular endothelial growth
factor and the type 1 vascular endothelial growth factor receptor
correlate with the size of papillary thyroid carcinoma in children
and young adults. Thyroid 2000;10(4):349–57.
[64] Minuto F, et al. Immunoreactive insulin-like growth factor I
(IGF-I) and IGF-I-binding protein content in human thyroid
tissue. Journal of Clinical Endocrinology and Metabolism 1989;
68(3):621–6.
[65] Williams DW, Williams ED, Wynford-Thomas D. Evidence
for autocrine production of IGF-1 in human thyroid
adenomas. Molecular and Cellular Endocrinology 1989;61(1):
139–43.
[66] Bachrach LK, et al. Insulin-like growth factor binding protein
production in human follicular thyroid carcinoma cells. Growth
Regulation 1995;5(2):109–18.
[67] Williams DW, Williams ED, Wynford-Thomas D. Loss of
dependence on IGF-1 for proliferation of human thyroid
adenoma cells. British Journal of Cancer 1988;57(6):535–9.
[68] Goodman AL, Rone JD. Thyroid angiogenesis: endotheliotropic
chemoattractant activity from rat thyroid cells in culture.
Endocrinology 1987;121(6):2131–40.
[69] Weidner KM, et al. Scatter factor: molecular characteristics and
effect on the invasiveness of epithelial cells. Journal of Cellular
Biology 1990;111(5 Part 1):2097–108.
[70] Eccles N, Ivan M, Wynford-Thomas D. Mitogenic stimulation
of normal and oncogene-transformed human thyroid epithelial
cells by hepatocyte growth factor. Molecular and Cellular
Endocrinology 1996;117(2):247–51.
[71] Di Renzo MF, et al. Overexpression of the c-MET/HGF
receptor gene in human thyroid carcinomas. Oncogene 1992;
7(12):2549–53.
[72] Rodrigues GA, Park M. Dimerization mediated through a
leucine zipper activates the oncogenic potential of the met
receptor tyrosine kinase. Molecular and Cellular Biology 1993;
13(11):6711–22.
[73] Park M, et al. Mechanism of met oncogene activation. Cell
1986;45(6):895–904.
[74] Comoglio PM. Structure, biosynthesis and biochemical proper-
ties of the HGF receptor in normal and malignant cells. EXS
1993;65:131–65.
[75] Di Renzo MF, et al. Expression of the Met/HGF receptor in
normal and neoplastic human tissues. Oncogene 1991;6(11):
1997–2003.
[76] Ivan M, et al. Activated ras and ret oncogenes induce
over-expression of c-met (hepatocyte growth factor receptor)
in human thyroid epithelial cells. Oncogene 1997;14(20):
2417–23.
[77] Ruco LP, et al. Expression of Met protein in thyroid tumours.
Journal of Pathology 1996;180(3):266–70.
[78] Belfiore A, et al. Negative/low expression of the Met/hepatocyte
growth factor receptor identifies papillary thyroid carcinomas
with high risk of distant metastases. Journal of Clinical
Endocrinology and Metabolism 1997;82(7):2322–8.
[79] Eilers M, Schirm S, Bishop JM. The MYC protein activates
transcription of the alpha-prothymosin gene. EMBO Journal
1991;10(1):133–41.
[80] Amati B, et al. Transcriptional activation by the human c-Myc
oncoprotein in yeast requires interaction with Max. Nature
1992;359(6394):423–6.
[81] Amati B, et al. Oncogenic activity of the c-Myc protein requires
dimerization with Max. Cell 1993;72(2):233–45.
[82] Luscher B, Eisenman RN. New light on Myc and Myb. Part I.
Myc. Genes and Development 1990;4(12A):2025–35.
[83] Colletta G, Cirafici AM, Vecchio G. Induction of the c-fos
oncogene by thyrotropic hormone in rat thyroid cells in culture.
Science 1986;233(4762):458–60.
[84] Terrier P, et al. Structure and expression of c-myc and c-fos
proto-oncogenes in thyroid carcinomas. British Journal of
Cancer 1988;57(1):43–7.
[85] Wyllie FS, et al. Structure and expression of nuclear oncogenes
in multi-stage thyroid tumorigenesis. British Journal of Cancer
1989;60(4):561–5.
[86] Yamashita S, et al. Expression of the myc cellular proto-
oncogene in human thyroid tissue. Journal of Clinical Endocri-
nology and Metabolism 1986;63(5):1170–3.
[87] Barbacid M. Ras Genes. Annual Review of Biochemistry
1987;56:779–827.
[88] Finney RE, Bishop JM. Predisposition to neoplastic transfor-
mation caused by gene replacement of H-ras1. Science 1993;
260(5113):1524–7.
[89] Studer H, et al. Histomorphological and immunohisto-
chemical evidence that human nodular goiters grow by
episodic replication of multiple clusters of thyroid follicular
cells. Journal of Clinical Endocrinology and Metabolism 1992;
75(4):1151–8.
[90] Francis-Lang H, et al. Multiple mechanisms of interference
between transformation and differentiation in thyroid cells.
Molecular and Cellular Biology 1992;12(12):5793–800.
[91] Wynford-Thomas D. Molecular basis of epithelial tumorigen-
esis: the thyroid model. Critical Review of Oncogogenesis 1993;
4:1–23.
[92] Rochefort P, et al. Thyroid pathologies in transgenic mice
expressing a human activated Ras gene driven by a thyroglobulin
promoter. Oncogene 1996;12(1):111–8.
[93] Santelli G, et al. Production of transgenic mice expressing the
Ki-ras oncogene under the control of a thyroglobulin promoter.
Cancer Research 1993;53(22):5523–7.
[94] Shi YF, et al. High rates of ras codon 61 mutation in thyroid
tumors in an iodide-deficient area. Cancer Research 1991;
51(10):2690–3.
[95] Wright PA, et al. Radiation-associated and ‘spontaneous’
human thyroid carcinomas show a different pattern of ras
oncogene mutation. Oncogene 1991;6(3):471–3.
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–9084

[96] Bongarzone I, et al. High frequency of activation of tyrosine
kinase oncogenes in human papillary thyroid carcinoma.
Oncogene 1989;4(12):1457–62.
[97] Fusco A, et al. A new oncogene in human thyroid papillary
carcinomas and their lymph-nodal metastases. Nature
1987;328(6126):170–2.
[98] Namba H, et al. H-ras protooncogene mutations in human
thyroid neoplasms. Journal of Clinical Endocrinology and
Metabolism 1990;71(1):223–9.
[99] Karga H, et al. Ras oncogene mutations in benign and malignant
thyroid neoplasms. Journal of Clinical Endocrinology and
Metabolism 1991;73(4):832–6.
[100] Krohn K, et al. Ras mutations are rare in solitary cold and toxic
thyroid nodules. Clinical Endocrinology (Oxford) 2001;55(2):
241–8.
[101] Basolo F, et al. N-ras mutation in poorly differentiated thyroid
carcinomas: correlation with bone metastases and inverse correla-
tion to thyroglobulin expression. Thyroid 2000;10(1):19–23.
[102] Esapa CT, et al. Prevalence of Ras mutations in thyroid
neoplasia. Clinical Endocrinology (Oxford) 1999;50(4):529–35.
[103] Lemoine NR, et al. High frequency of ras oncogene activation in
all stages of human thyroid tumorigenesis. Oncogene 1989;
4(2):159–64.
[104] Namba H, Rubin SA, Fagin JA. Point mutations of ras
oncogenes are an early event in thyroid tumorigenesis. Molecular
Endocrinology 1990;4(10):1474–9.
[105] Davies H, et al. Mutations of the BRAF gene in human cancer.
Nature 2002;417(6892):949–54.
[106] Rajagopalan H, et al. Tumorigenesis: raf/ras oncogenes and
mismatch-repair status. Nature 2002;418(6901):934.
[107] Cohen Y, et al. BRAF mutation in papillary thyroid carcinoma.
Journal of the National Cancer Institute 2003;95(8):625–7.
[108] Kimura ET, et al. High prevalence of BRAF mutations in
thyroid cancer: genetic evidence for constitutive activation of the
RET/PTC-RAS-BRAF signaling pathway in papillary thyroid
carcinoma. Cancer Research 2003;63(7):1454–7.
[109] Learoyd DL, et al. RET/PTC and RET tyrosine kinase
expression in adult papillary thyroid carcinomas. Journal of
Clinical Endocrinology and Metabolism 1998;83(10):3631–5.
[110] De Vita G, et al. Expression of the RET/PTC1 oncogene impairs
the activity of TTF-1 and Pax-8 thyroid transcription factors.
Cell Growth and Differentiation 1998;9(1):97–103.
[111] Pierotti MA, et al. Characterization of an inversion on the long
arm of chromosome 10 juxtaposing D10S170 and RET and
creating the oncogenic sequence RET/PTC. Proceedings of the
National Academy Science of the USA 1992;89(5):1616–20.
[112] Bongarzone I, et al. Molecular characterization of a thyroid
tumor-specific transforming sequence formed by the fusion of ret
tyrosine kinase and the regulatory subunit RI alpha of cyclic
AMP-dependent protein kinase A. Molecular and Cellular
Biology 1993;13(1):358–66.
[113] Santoro M, et al. Molecular characterization of RET/PTC3; a
novel rearranged version of the RETproto-oncogene in a human
thyroid papillary carcinoma. Oncogene 1994;9(2):509–16.
[114] Fugazzola L, et al. Molecular and biochemical analysis of RET/
PTC4, a novel oncogenic rearrangement between RET and
ELE1 genes, in a post-Chernobyl papillary thyroid cancer.
Oncogene 1996;13(5):1093–7.
[115] Klugbauer S, et al. Detection of a novel type of RET
rearrangement (PTC5) in thyroid carcinomas after Chernobyl
and analysis of the involved RET-fused gene RFG5. Cancer
Research 1998;58(2):198–203.
[116] Klugbauer S, Rabes HM. The transcription coactivator HTIF1
and a related protein are fused to the RET receptor tyrosine
kinase in childhood papillary thyroid carcinomas. Oncogene
1999;18(30):4388–93.
[117] Salassidis K, et al. Translocation t(10;14)(q11.2:q22.1) fusing the
kinetin to the RET gene creates a novel rearranged form (PTC8)
of the RET proto-oncogene in radiation-induced childhood
papillary thyroid carcinoma. Cancer Research 2000;60(11):
2786–9.
[118] Ishizaka Y, et al. Detection of retTPC/PTC transcripts in
thyroid adenomas and adenomatous goiter by an RT-PCR
method. Oncogene 1991;6(9):1667–72.
[119] Jhiang SM, et al. Detection of the PTC/retTPC oncogene in
human thyroid cancers. Oncogene 1992;7(7):1331–7.
[120] Sugg SL, et al. Ret/PTC-1, -2, and -3 oncogene rearrangements
in human thyroid carcinomas: implications for metastatic
potential? Journal of Clinical Endocrinology and Metabolism
1996;81(9):3360–5.
[121] Delvincourt C, et al. Ret and trk proto-oncogene activation in
thyroid papillary carcinomas in French patients from the
Champagne-Ardenne region. Clinical Biochemistry 1996;29(3):
267–71.
[122] Bounacer A, et al. High prevalence of activating ret proto-
oncogene rearrangements, in thyroid tumors from patients who
had received external radiation. Oncogene 1997;15(11):1263–73.
[123] Bongarzone I, et al. Frequent activation of ret protooncogene by
fusion with a new activating gene in papillary thyroid
carcinomas. Cancer Research 1994;54(11):2979–85.
[124] Wajjwalku W, et al. Low frequency of rearrangements of the ret
and trk proto-oncogenes in Japanese thyroid papillary carcino-
mas. Japanese Journal of Cancer Research 1992;83(7):671–5.
[125] Zou M, Shi Y, Farid NR. Low rate of ret proto-oncogene
activation (PTC/retTPC) in papillary thyroid carcinomas from
Saudi Arabia. Cancer 1994;73(1):176–80.
[126] Santoro M, et al. Development of thyroid papillary carcinomas
secondary to tissue-specific expression of the RET/PTC1
oncogene in transgenic mice. Oncogene 1996;12(8):1821–6.
[127] Nikiforov YE, et al. Distinct pattern of ret oncogene rearrange-
ments in morphological variants of radiation-induced and
sporadic thyroid papillary carcinomas in children. Cancer
Research 1997;57(9):1690–4.
[128] Mizuno T, et al. Continued expression of a tissue specific
activated oncogene in the early steps of radiation-induced
human thyroid carcinogenesis. Oncogene 1997;15(12):1455–60.
[129] Cheung CC, et al. Analysis of ret/PTC gene rearrangements
refines the fine needle aspiration diagnosis of thyroid cancer.
Journal of Clinical Endocrinology and Metabolism
2001;86(5):2187–90.
[130] Belchetz G, et al. Hurthle cell tumors: using molecular
techniques to define a novel classification system. Archives of
Otolaryngology—Head and Neck Surgery 2002;128(3):237–40.
[131] Cheung CC, et al. Molecular basis off Hurthle cell papillary
thyroid carcinoma. Journal of Clinical Endocrinology and
Metabolism 2000;85(2):878–82.
[132] Barbacid M, et al. The trk family of tyrosine protein kinase
receptors. Biochimica et Biophysica Acta 1991;1072(2–3):115–27.
[133] Martin-Zanca D, Barbacid M, Parada LF. Expression of the trk
proto-oncogene is restricted to the sensory cranial and spinal
ganglia of neural crest origin in mouse development. Genes
Development 1990;4(5):683–94.
[134] Greco A, et al. TRK-T1 is a novel oncogene formed by the
fusion of TPR and TRK genes in human papillary thyroid
carcinomas. Oncogene 1992;7(2):237–42.
[135] Morris CM, et al. Localization of the TRK proto-oncogene to
human chromosome bands 1q23–1q24. Oncogene 1991;6(6):
1093–5.
[136] Greco A, et al. Chromosome 1 rearrangements involving the
genes TPR and NTRK1 produce structurally different thyroid-
specific TRK oncogenes. Genes Chromosomes and Cancer 1997;
19(2):112–23.
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–90 85

[137] Greco A, et al. The DNA rearrangement that generates the
TRK-T3 oncogene involves a novel gene on chromosome 3
whose product has a potential coiled-coil domain. Molecular and
Cellular Biology 1995;15(11):6118–27.
[138] Portella G, et al. Human N-ras, TRK-T1, and RET/PTC3
oncogenes, driven by a thyroglobulin promoter, differently affect
the expression of differentiation markers and the proliferation of
thyroid epithelial cells. Oncology Research 1999;11(9):421–7.
[139] Bongarzone I, et al. Age-related activation of the tyrosine kinase
receptor protooncogenes RET and NTRK1 in papillary thyroid
carcinoma. Journal of Clinical Endocrinology and Metabolism
1996;81(5):2006–9.
[140] Doree M, Galas S. The cyclin-dependent protein kinases and the
control of cell division. FASEB Journal 1994;8(14):1114–21.
[141] Basolo F, et al. Cyclin D1 overexpression in thyroid carcinomas:
relation with clinico- pathological parameters, retinoblastoma
gene product, and Ki67 labeling index. Thyroid 2000;10(9):741–6.
[142] Jiang W, et al. Altered expression of the cyclin D1 and
retinoblastoma genes in human esophageal cancer. Proceedings
of the National Academy Science of the USA 1993;90(19):9026–30.
[143] Khoo ML, et al. Overexpression of cyclin D1 and under-
expression of p27 predict lymph node metastases in papillary
thyroid carcinoma. Journal of Clinical Endocrinology and
Metabolism 2002;87(4):1814–8.
[144] Wang S, et al. The role of cell cycle regulatory protein, cyclin D1,
in the progression of thyroid cancer. Modern Pathology 2000;
13(8):882–7.
[145] Muller H, et al. Cyclin D1 expression is regulated by the
retinoblastoma protein. Proceedings of the National Academy
Science of the USA 1994;91(8):2945–9.
[146] Erickson LA, et al. Pathologic features, proliferative activity,
and cyclin D1 expression in Hurthle cell neoplasms of the
thyroid. Modern Pathology 2000;13(2):186–92.
[147] Krohn K, Paschke R. Loss of heterozygocity at the thyroid
peroxidase gene locus in solitary cold thyroid nodules. Thyroid
2001;11(8):741–7.
[148] Christensen L, et al. Thyroperoxidase (TPO) immunostaining of
the solitary cold thyroid nodule. Clinical Endocrinology
(Oxford) 2000;53(2):161–9.
[149] De Micco C, Vassko V, Henry JF. The value of thyroid
peroxidase immunohistochemistry for preoperative fine-needle
aspiration diagnosis of the follicular variant of papillary thyroid
cancer. Surgery 1999;126(6):1200–4.
[150] Faroux MJ, et al. Evaluation of the monoclonal antibody
antithyroperoxidase MoAb47 in the diagnostic decision of cold
thyroid nodules by fine-needle aspiration. Pathology Research
and Practice 1997;193(10):705–12.
[151] Pluot M, et al. Quantitative cytology and thyroperoxidase
immunochemistry: new tools in evaluating thyroid nodules by
fine-needle aspiration. Cancer Detection and Prevention 1996;
20(4):285–93.
[152] De Micco C, et al. Fine-needle aspiration of thyroid follicular
neoplasm: diagnostic use of thyroid peroxidase immunocyto-
chemistry with monoclonal antibody 47. Surgery 1994;116(6):
1031–5.
[153] De Micco C, et al. Thyroid peroxidase immunodetection as a
tool to assist diagnosis of thyroid nodules on fine-needle
aspiration biopsy. European Journal of Endocrinology 1994;
131(5):474–9.
[154] Henry JF, et al. Thyroperoxidase immunodetection for the
diagnosis of malignancy on fine-needle aspiration of thyroid
nodules. World Journal of Surgery 1994;18(4):529–34.
[155] Yamashita H, et al. Immunohistological differentiation of
benign thyroid follicular cell tumors from malignant ones:
usefulness of anti-peroxidase and JT-95 antibodies. Acta
Pathologica Japanica 1993;43(11):670–3.
[156] Song B. Immunohistochemical demonstration of epidermal
growth factor receptor and ceruloplasmin in thyroid diseases.
Acta Pathologica Japanica 1991;41(5):336–43.
[157] Kondi-Pafiti A, et al. Immunohistochemical study of cerulo-
plasmin, lactoferrin and secretory component expression in
neoplastic and non-neoplastic thyroid gland diseases. Acta
Oncologica 2000;39(6):753–6.
[158] Tuccari G, Barresi G. Immunohistochemical demonstration of
ceruloplasmin in follicular adenomas and thyroid carcinomas.
Histopathology 1987;11(7):723–31.
[159] Vaideeswar P, et al. Differentiation of follicular adenoma and
carcinoma of thyroid by immunohistochemical demonstration of
ceruloplasmin. Indian Journal of Pathology and Microbiology
1994;37(2):165–9.
[160] Barresi G, Tuccari G. Iron-binding proteins in thyroid tumours.
An immunocytochemical study. Pathology Research and Prac-
tice 1987;182(3):344–51.
[161] Gonzalez-Campora R, et al. Dipeptidyl aminopeptidase IV in
the cytologic diagnosis of thyroid carcinoma. Diagnostic
Cytopathology 1998;19(1):4–8.
[162] Tang AC, et al. Expression of dipeptidyl aminopeptidase IV
activity in thyroid tumours: a possible marker of thyroid
malignancy. Journal of Otolaryngology 1996;25(1):14–9.
[163] Iwabuchi H, et al. Staining for dipeptidyl aminopeptidase IV
activity in nodular thyroid diseases. Acta Cytologica
1996;40(2):158–63.
[164] Zoro P, et al. Tumor markers in the cytodiagnosis of thyroid
nodules. Detection of dipeptidyl aminopeptidase IV (DAP IV)
activity. Annales de Pathologie 1996;16(4):261–5.
[165] Aratake Y, et al. Dipeptidyl aminopeptidase IV staining of
cytologic preparations to distinguish benign from malignant
thyroid diseases. American Journal of Clinical Pathology 1991;
96(3):306–10.
[166] Kotani T, et al. Diagnostic usefulness of dipeptidyl aminopepti-
dase IV monoclonal antibody in paraffin-embedded thyroid
follicular tumours. Journal of Pathology 1992;168(1):41–5.
[167] Chiappetta G, et al. Detection of high mobility group I
HMGI(Y) protein in the diagnosis of thyroid tumors: HMGI(Y)
expression represents a potential diagnostic indicator of carci-
noma. Cancer Research 1998;58(18):4193–8.
[168] Chiappetta G, et al. High mobility group HMGI(Y) protein
expression in human colorectal hyperplastic and neoplastic
diseases. International Journal of Cancer 2001;91(2):147–51.
[169] Pierantoni GM, et al. High mobility group I (Y) proteins bind
HIPK2, a serine–threonine kinase protein which inhibits cell
growth. Oncogene 2001;20(43):6132–41.
[170] Bandiera A, et al. Expression of HMGI(Y) proteins in squamous
intraepithelial and invasive lesions of the uterine cervix. Cancer
Research 1998;58(3):426–31.
[171] Berlingieri MT, et al. Inhibition of HMGI-C protein synthesis
suppresses retrovirally induced neoplastic transformation of rat
thyroid cells. Molecular and Cellular Biology 1995;15(3):1545–53.
[172] Scala S, et al. Adenovirus-mediated suppression of HMGI(Y)
protein synthesis as potential therapy of human malignant
neoplasias. Proceedings of the National Academy Science of the
USA 2000;97(8):4256–61.
[173] Kim SJ, Ryu JW, Choi DS. The expression of the high mobility
group I(Y) mRNA in thyroid cancers: useful tool of differential
diagnosis of thyroid nodules. Korean Journal of Internal
Medicine 2000;15(1):71–5.
[174] Hunter T, Pines J. Cyclins and cancer. Cell 1991;66(6):1071–4.
[175] Depoortere F, et al. Transforming growth factor beta(1)
selectively inhibits the cyclic AMP-dependent proliferation of
primary thyroid epithelial cells by preventing the association of
cyclin D3-cdk4 with nuclear p27(kip1). Molecular and Biology
of the Cell 2000;11(3):1061–76.
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–9086

[176] Grubeck-Loebenstein B, et al. Transforming growth factor beta
regulates thyroid growth. Role in the pathogenesis of nontoxic
goiter. Journal of Clinical Investigation 1989;83(3):764–70.
[177] Aust G, Scherbaum WA. Expression of cytokines in the thyroid:
thyrocytes as potential cytokine producers. Experimental and
Clinical Endocrinology and Diabetes 1996;104(Suppl. 4):64–7.
[178] Pang XP, Park M, Hershman JM. Transforming growth factor-
beta blocks protein kinase-A-mediated iodide transport and
protein kinase-C-mediated DNA synthesis in FRTL-5 rat
thyroid cells. Endocrinology 1992;131(1):45–50.
[179] Jasani B, et al. Immunocytochemically detectable TGF-beta
associated with malignancy in thyroid epithelial neoplasia.
Growth Factors 1990;2(2–3):149–55.
[180] Matoba H, et al. Expression of transforming growth factor-
beta1 and transforming growth factor-beta Type-II receptor
mRNA in papillary thyroid carcinoma. Hormone and Metabolic
Research 1998;30(10):624–8.
[181] Zedenius J, et al. Alterations of p53 and expression of WAF1/
p21 in human thyroid tumors. Thyroid 1996;6(1):1–9.
[182] Ito Y, et al. Expression of p21 (WAF1/CIP1) protein in clinical
thyroid tissues. British Journal of Cancer 1996;74(8):1269–74.
[183] Resnick MB, et al. Immunohistochemical analysis of p27/kip1
expression in thyroid carcinoma. Modern Pathology 1998;
11(8):735–9.
[184] Tallini G, et al. Downregulation of p27KIP1 and Ki67/Mib1
labeling index support the classification of thyroid carcinoma
into prognostically relevant categories. American Journal of
Surgical Pathology 1999;23(6):678–85.
[185] Erickson LA, et al. Expression of p27kip1 and Ki-67 in benign
and malignant thyroid tumors. Modern Pathology
1998;11(2):169–74.
[186] Wang S, et al. The role of cell cycle regulatory proteins, cyclin
D1, cyclin E, and p27 in thyroid carcinogenesis. Human
Pathology 1998;29(11):1304–9.
[187] Khoo ML, et al. Underexpression of p27/Kip in thyroid
papillary microcarcinomas with gross metastatic disease. Ar-
chives of Otolaryngology—Head and Neck Surgery 2002;128(3):
253–7.
[188] Troncone G, et al. Cyclin-dependent kinase inhibitor p27(Kip1)
expression in thyroid cells obtained by fine-needle aspiration
biopsy: a preliminary report. Diagnostic Cytopathology
2000;23(2):77–81.
[189] Hong FD, et al. Structure of the human retinoblastoma gene.
Proceedings of the National Academy Science of the USA
1989;86(14):5502–6.
[190] Lee WH, et al. The retinoblastoma susceptibility gene encodes a
nuclear phosphoprotein associated with DNA binding activity.
Nature 1987;329(6140):642–5.
[191] Nevins JR. E2F: a link between the Rb tumor suppressor protein
and viral oncoproteins. Science 1992;258(5081):424–9.
[192] Ewen ME, et al. Functional interactions of the retinoblastoma
protein with mammalian D-type cyclins. Cell 1993;73(3):
487–97.
[193] Chellappan SP, et al. The E2F transcription factor is a cellular
target for the RB protein. Cell 1991;65(6):1053–61.
[194] DeCaprio JA, et al. SV40 large tumor antigen forms a specific
complex with the product of the retinoblastoma susceptibility
gene. Cell 1988;54(2):275–83.
[195] Whyte P, et al. Association between an oncogene and an anti-
oncogene: the adenovirus e1a proteins bind to the retinoblasto-
ma gene product. Nature 1988;334(6178):124–9.
[196] Munger K, et al. Complex formation of human papillomavirus
E7 proteins with the retinoblastoma tumor suppressor gene
product. EMBO Journal 1989;8(13):4099–105.
[197] Hatakeyama M, Weinberg RA. The role of RB in cell cycle
control. Progress in Cell Cycle Research 1995;1:9–19.
[198] Ludlow JW, et al. SV40 large T antigen binds preferentially to an
underphosphorylated member of the retinoblastoma suscept-
ibility gene product family. Cell 1989;56(1):57–65.
[199] Mihara K, et al. Cell cycle-dependent regulation of phosphor-
ylation of the human retinoblastoma gene product. Science
1989;246(4935):1300–3.
[200] Shirodkar S, et al. The transcription factor E2F interacts with
the retinoblastoma product and a p107-cyclin A complex in a cell
cycle-regulated manner. Cell 1992;68(1):157–66.
[201] Stiegler P, Kasten M, Giordano A. The RB family of cell cycle
regulatory factors. Journal of Cellular Biochemistry Supplement
1998;30–31:30–6.
[202] Ledent C, et al. Thyroid adenocarcinomas secondary to tissue-
specific expression of simian virus-40 large T-antigen in
transgenic mice. Endocrinology 1991;129(3):1391–401.
[203] Ledent C, et al. Differentiated carcinomas develop as a
consequence of the thyroid specific expression of a thyroglobu-
lin-human papillomavirus type 16 E7 transgene. Oncogene 1995;
10(9):1789–97.
[204] Holm R, Nesland JM. Retinoblastoma and p53 tumour
suppressor gene protein expression in carcinomas of the thyroid
gland. Journal of Pathology 1994;172(3):267–72.
[205] Kastan MB, et al. A mammalian cell cycle checkpoint pathway
utilizing p53 and GADD45 is defective in ataxia-telangiectasia.
Cell 1992;71(4):587–97.
[206] Maltzman W, Czyzyk L. UV irradiation stimulates levels of p53
cellular tumor antigen in nontransformed mouse cells. Molecular
and Cellular Biology 1984;4(9):1689–94.
[207] Vogelstein B, Kinzler KW. p53 function and dysfunction. Cell
1992;70(4):523–6.
[208] Gottlieb TM, Oren M. p53 in growth control and neoplasia.
Biochimica et Biophysica Acta 1996;1287(2–3):77–102.
[209] Finlay CA, Hinds PW, Levine AJ. The p53 proto-oncogene can
act as a suppressor of transformation. Cell 1989;57(7):1083–93.
[210] Hartwell L. Defects in a cell cycle checkpoint may be responsible
for the genomic instability of cancer cells. Cell 1992;71(4):543–6.
[211] Hollstein M, et al. p53 mutations in human cancers. Science
1991;253(5015):49–53.
[212] Malkin D, et al. Germ line p53 mutations in a familial syndrome
of breast cancer, sarcomas, and other neoplasms. Science
1990;250(4985):1233–8.
[213] Li FP, et al. A cancer family syndrome in twenty-four kindreds.
Cancer Research 1988;48(18):5358–62.
[214] Baker SJ, et al. p53 gene mutations occur in combination with
17p allelic deletions as late events in colorectal tumorigenesis.
Cancer Research 1990;50(23):7717–22.
[215] Moretti F, et al. p53 re-expression inhibits proliferation and
restores differentiation of human thyroid anaplastic carcinoma
cells. Oncogene 1997;14(6):729–40.
[216] Feinstein E, et al. Expression of the normal p53 gene induces
differentiation of K562 cells. Oncogene 1992;7(9):1853–7.
[217] Donehower LA, et al. Mice deficient for p53 are developmentally
normal but susceptible to spontaneous tumours. Nature
1992;356(6366):215–21.
[218] Fagin JA, et al. Reexpression of thyroid peroxidase in a
derivative of an undifferentiated thyroid carcinoma cell line
by introduction of wild-type p53. Cancer Research 1996;56(4):
765–71.
[219] Frisch SM, Ruoslahti E. Integrins and anoikis. Current Opinion
in Cell Biology 1997;9(5):701–6.
[220] Chun YS, Saji M, Zeiger MA. Overexpression of TTF-1 and
PAX-8 restores thyroglobulin gene promoter activity in ARO
and WRO cell lines. Surgery 1998;124(6):1100–5.
[221] Komatsu M, et al. Thyroid dysgenesis caused by PAX8
mutation: the hypermutability with CpG dinucleotides at codon
31. Journal of Pediatrics 2001;139(4):597–9.
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–90 87

[222] Schmitt TL, Espinoza CR, Loos U. Transcriptional regulation
of the human sodium/iodide symporter gene by Pax8 and TTF-1.
Experimental and Clinical Endocrinology and Diabetes
2001;109(1):27–31.
[223] Esposito C, et al. PAX 8 activates the enhancer of the human
thyroperoxidase gene. Biochemical Journal 1998;331(Part 1):37–40.
[224] Kogai T, et al. Differential regulation of the human sodium/
iodide symporter gene promoter in papillary thyroid carcinoma
cell lines and normal thyroid cells. Endocrinology 2001;142(8):
3369–79.
[225] Pasca di Magliano M, Di Lauro R, Zannini M. Pax8 has a key
role in thyroid cell differentiation. Proceedings of the National
Academy Science of the USA 2000;97(24):13144–9.
[226] Puglisi F, et al. Expression of Pax-8, p53 and bcl-2 in human
benign and malignant thyroid diseases. Anticancer Research
2000;20(1A):311–6.
[227] Macchia PE, et al. PAX8 mutations associated with congenital
hypothyroidism caused by thyroid dysgenesis. Nature Genetics
1998;19(1):83–6.
[228] Mansouri A, Chowdhury K, Gruss P. Follicular cells of the
thyroid gland require Pax8 gene function. Nature Genetics
1998;19(1):87–90.
[229] Fabbro D, et al. Expression of thyroid-specific transcription
factors TTF-1 and PAX-8 in human thyroid neoplasms. Cancer
Research 1994;54(17):4744–9.
[230] Ros P, et al. Thyroid-specific gene expression in the multi-step
process of thyroid carcinogenesis. Biochimie 1999;81(4):389–96.
[231] Baloch ZW, Pasha T, LiVolsi VA. Cytoplasmic accumulation
of alpha-catenin in thyroid neoplasms. Head and Neck 2001;
23(7):573–8.
[232] Kato N, et al. E-cadherin expression in follicular carcinoma of
the thyroid. Pathology International 2002;52(1):13–8.
[233] Naito A, et al. Clinical significance of E-cadherin expression in
thyroid neoplasms. Journal of Surgical Oncology 2001;76(3):
176–80.
[234] Walgenbach S, et al. Prognostic value of e-cadherin in papillary
thyroid carcinoma. Chirurg 1998;69(2):186–90.
[235] von Wasielewski R, et al. Immunohistochemical detection of E-
cadherin in differentiated thyroid carcinomas correlates with
clinical outcome. Cancer Research 1997;57(12):2501–7.
[236] Scheumman GF, et al. Clinical significance of E-cadherin as a
prognostic marker in thyroid carcinomas. Journal of Clinical
Endocrinology and Metabolism 1995;80(7):2168–72.
[237] Huang SH, et al. Expression of the cadherin–catenin complex in
well-differentiated human thyroid neoplastic tissue. Thyroid
1999;9(11):1095–103.
[238] Graff JR, et al. Distinct patterns of E-cadherin CpG island
methylation in papillary, follicular, Hurthle’s cell, and poorly
differentiated human thyroid carcinoma. Cancer Research
1998;58(10):2063–6.
[239] Gasbarri A, et al. Galectin-3 and CD44v6 isoforms in the
preoperative evaluation of thyroid nodules. Journal of Clinical
Oncology 1999;17(11):3494–502.
[240] Perillo NL, Marcus ME, Baum LG. Galectins: versatile
modulators of cell adhesion, cell proliferation, and cell death.
Journal of Molecular Medicine 1998;76(6):402–12.
[241] Xu XC, el-Naggar AK, Lotan R. Differential expression of
galectin-1 and galectin-3 in thyroid tumors. Potential diagnostic
implications. American Journal of Pathology 1995;147(3):
815–22.
[242] Fernandez PL, et al. Galectin-3 and laminin expression in
neoplastic and non-neoplastic thyroid tissue. Journal of Pathol-
ogy 1997;181(1):80–6.
[243] Bartolazzi A, et al. Application of an immunodiagnostic method
for improving preoperative diagnosis of nodular thyroid lesions.
Lancet 2001;357(9269):1644–50.
[244] Akahani S, et al. Galectin-3: a novel antiapoptotic molecule with
a functional BH1 (NWGR) domain of Bcl-2 family. Cancer
Research 1997;57(23):5272–6.
[245] Coli A, et al. Galectin-3, a marker of well-differentiated thyroid
carcinoma, is expressed in thyroid nodules with cytological
atypia. Histopathology 2002;40(1):80–7.
[246] Inohara H, et al. Expression of galectin-3 in fine-needle aspirates
as a diagnostic marker differentiating benign from malignant
thyroid neoplasms. Cancer 1999;85(11):2475–84.
[247] Nascimento MC, et al. Differential reactivity for galectin-3 in
Hurthle cell adenomas and carcinomas. Endocrine Pathology
2001;12(3):275–9.
[248] Figge J, et al. Preferential expression of the cell adhesion
molecule CD44 in papillary thyroid carcinoma. Experimental
and Molecular Pathology 1994;61(3):203–11.
[249] Ermak G, et al. Deregulated alternative splicing of CD44
messenger RNA transcripts in neoplastic and nonneoplastic
lesions of the human thyroid. Cancer Research 1995;55(20):
4594–8.
[250] Ermak G, et al. Restricted patterns of CD44 variant exon
expression in human papillary thyroid carcinoma. Cancer
Research 1996;56(5):1037–42.
[251] Bohm JP, et al. Reduced CD44 standard expression is associated
with tumour recurrence and unfavourable outcome in differ-
entiated thyroid carcinoma. Journal of Pathology 2000;192(3):
321–7.
[252] Evan G, Littlewood T. A matter of life and cell death. Science
1998;281(5381):1317–22.
[253] Kim NW, et al. Specific association of human telomerase activity
with immortal cells and cancer. Science 1994;266(5193):2011–5.
[254] Shay JW. Aging and cancer—are telomeres and telomerase the
connection? Molecular Medicine Today 1995;1(8):378–84.
[255] Wright WE, Shay JW, Piatyszek MA. Modifications of a
telomeric repeat amplification protocol (TRAP) result in
increased reliability, linearity and sensitivity. Nucleic Acids
Research 1995;23(18):3794–5.
[256] Greider CW, Blackburn EH. Telomeres, telomerase and cancer.
Scientific American 1996;274(2):92–7.
[257] Sommerfeld HJ, et al. Telomerase activity—a prevalent marker
of malignant human prostate tissue. Cancer Research 1996;
56(1):218–22.
[258] Bacchetti S, Counter CM. Telomere and telomerase in human
cancer. International Journal of Oncology 1995;7(3):423–32.
[259] Oshimura M, Barrett JC. Multiple pathways to cellular
senescence: role of telomerase repressors. European Journal of
Cancer 1997;33(5):705–10.
[260] Kiyono T, et al. Both Rb/p16INK4a inactivation and telomerase
activity are required to immortalize human epithelial cells.
Nature 1998;396(6706):84–8.
[261] Tsutsui T, et al. Association of p16(INK4a) and pRb inactiva-
tion with immortalization of human cells. Carcinogenesis
2002;23(12):2111–7.
[262] Stavropoulos DJ, et al. The Bloom syndrome helicase BLM
interacts with TRF2 in ALT cells and promotes telomeric DNA
synthesis. Human Molecular Genetics 2002;11(25):3135–44.
[263] Umbricht CB, et al. Telomerase activity: a marker to distinguish
follicular thyroid adenoma from carcinoma. Cancer Research
1997;57(11):2144–7.
[264] Saji M, et al. Telomerase activity in the differential diagnosis of
papillary carcinoma of the thyroid. Surgery 1997;122(6):1137–40.
[265] Zeiger MA, et al. Human telomerase reverse transcriptase
(hTERT) gene expression in FNA samples from thyroid
neoplasms. Surgery 1999;126(6):1195–8 (discussion 1198-9).
[266] Vaux DL, Cory S, Adams JM. Bcl-2 gene promotes haemo-
poietic cell survival and cooperates with c-myc to immortalize
pre-B cells. Nature 1988;335(6189):440–2.
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–9088

[267] Adams JM, Cory S. The Bcl-2 protein family: arbiters of cell
survival. Science 1998;281(5381):1322–6.
[268] Branet F, et al. Expression of the cell death-inducing gene bax in
carcinomas developed from the follicular cells of the thyroid
gland. Journal of Clinical Endocrinology and Metabolism
1996;81(7):2726–30.
[269] Pollina L, et al. Bcl-2, p53 and proliferating cell nuclear antigen
expression is related to the degree of differentiation in thyroid
carcinomas. British Journal of Cancer 1996;73(2):139–43.
[270] Nagata S, Golstein P. The Fas death factor. Science 1995;
267(5203):1449–56.
[271] Matiba B, Mariani SM, Krammer PH. The CD95 system
and the death of a lymphocyte. Seminars in Immunology
1997;9(1):59–68.
[272] Giordano C, et al. Potential involvement of Fas and its ligand
in the pathogenesis of Hashimoto’s thyroiditis. Science 1997;
275(5302):960–3.
[273] Mitsiades N, et al. Fas ligand expression in thyroid carcinomas:
a potential mechanism of immune evasion. Journal of Clinical
Endocrinology and Metabolism 1999;84(8):2924–32.
[274] Stambolic V, et al. Negative regulation of PKB/Akt-dependent
cell survival by the tumor suppressor PTEN. Cell 1998;95(1):
29–39.
[275] Maehama T, Dixon JE. The tumor suppressor, PTEN/MMAC1,
dephosphorylates the lipid second messenger, phosphatidyl-
inositol 3,4,5-trisphosphate. Journal of Biological Chemistry
1998;273(22):13375–8.
[276] Gustin JA, et al. The PTEN tumor suppressor protein inhibits
tumor necrosis factor-induced nuclear factor kappa B activity.
Journal of Biological Chemistry 2001;276(29):27740–4.
[277] Liaw D, et al. Germline mutations of the PTEN gene in Cowden
disease, an inherited breast and thyroid cancer syndrome. Nature
Genetics 1997;16(1):64–7.
[278] Gimm O, et al. Differential nuclear and cytoplasmic expression
of PTEN in normal thyroid tissue, and benign and malignant
epithelial thyroid tumors. American Journal of the Pathology
2000;156(5):1693–700.
[279] Zysman MA, Chapman WB, Bapat B. Considerations when
analyzing the methylation status of PTEN tumor suppressor
gene. American Journal of the Pathology 2002;160(3):795–800.
[280] Soria JC, et al. Lack of PTEN expression in non-small cell lung
cancer could be related to promoter methylation. Clinical Cancer
Research 2002;8(5):1178–84.
[281] Oliner JD, et al. Oncoprotein MDM2 conceals the activation
domain of tumour suppressor p53. Nature 1993;362(6423):
857–60.
[282] Momand J, Wu HH, Dasgupta G. MDM2-master regulator of
the p53 tumor suppressor protein. Gene 2000;242(1–2):15–29.
[283] Goldberg Z, et al. Tyrosine phosphorylation of Mdm2 by
c-Abl: implications for p53 regulation. EMBO Journal 2002;
21(14):3715–27.
[284] Kubbutat MH, et al. Analysis of the degradation function of
Mdm2. Cell Growth and Differentiation 1999;10(2):87–92.
[285] Lozano G, Montes de Oca Luna R. MDM2 function.
Biochimica et Biophysica Acta 1998;1377(2):M55.
[286] Schmid KW, et al. Possible relation of p53 and mdm-2
oncoprotein expression in thyroid carcinoma: a molecular-
pathological and immunohistochemical study on paraffin-
embedded tissue. Endocrine Pathology 1996;7(2):121–30.
[287] Jennings T, et al. Nuclear accumulation of MDM2 protein in
well-differentiated papillary thyroid carcinomas. Experimental
Molecular Pathology 1995;62(3):199–206.
[288] Zou M, et al. The expression of the MDM2 gene, a p53
binding protein, in thyroid carcinogenesis. Cancer 1995;76(2):
314–8.
[289] Czyz W, et al. p53, MDM2, bcl-2 staining in follicular neoplasms
of the thyroid gland. Folia Histochemica et Cytobiologica
2001;39(Suppl. 2):167–8.
[290] Horie S, et al. Overexpression of p53 protein and MDM2
in papillary carcinomas of the thyroid: correlations with
clinicopathologic features. Pathology International 2001;51(1):
11–5.
[291] Hoos A, et al. Clinical significance of molecular expression
profiles of Hurthle cell tumors of the thyroid gland analyzed via
tissue microarrays. American Journal of the Pathology
2002;160(1):175–83.
[292] Hahn WC, Weinberg RA. Modelling the molecular circuitry of
cancer. Nature Reviews in Cancer 2002;2(5):331–41.
[293] Hanahan D, Weinberg RA. The hallmarks of cancer. Cell
2000;100(1):57–70.
ARTICLE IN PRESS
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–90 89

ARTICLE IN PRESS
Martha A. Zeiger, MD, FACS. She is an endocrine surgeon and Associate Professor of Surgery in the Division of
Endocrine and Oncologic Surgery at Johns Hopkins University Hospital. She has had a longstanding interest in
endocrine surgery as well as the molecular pathogenesis of thyroid neoplasms. In addition to a busy endocrine surgery
practice, she also directs a full time molecular research laboratory dedicated to the molecular aspects of endocrine
tumors.
Christopher B. Umbricht, MD, Ph.D. He is a molecular biologist and internist and an Assistant Professor in the
Departments of Surgery, Oncology and Pathology at the Johns Hopkins University School of Medicine. His research
interests are the molecular pathogenesis of thyroid cancer and breast cancer.
Dorry L. Segev, MD. He is a Chief Resident in Surgery at the Johns Hopkins Medical Institutions. He has had a
longstanding interest in the molecular signaling pathways associated with thyroid and breast cancer. He plans to
pursue a career in Pediatric Transplantation.
D.L. Segev et al. / Surgical Oncology 12 (2003) 69–9090