THYROID SPECIAL ARTICLE

Volume 28, Number 10, 2018

ª Mary Ann Liebert, Inc.
DOI: 10.1089/thy.2018.0306

Follicular Thyroid Carcinoma: A Perspective

Gilbert H. Daniels1

Introduction most doubled again since 2002. Most of the increase is due to
micro- and small (1–2 cm) PTC.

P apillary thyroid carcinoma (PTC) and follicular

thyroid carcinoma (FTC) are both well-differentiated
The epidemiology of FTC is more difficult to study because
of changes in diagnostic criteria over time. In the 1960s, any
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

thyroid cancers derived from thyroid follicular epithelium. thyroid tumor with >50% of the tumor mass composed of
Apart from this superficial similarity and a common ability to follicles was considered FTC, whereas now most such tumors
concentrate (radioactive) iodine, they share little in common. would be considered PTC (5). By the late 1980s, the World
PTC and FTC differ markedly in epidemiology, ultrasound Health Organization (WHO) defined FTC as a malignant ep-
examination, cytological findings, genetic landscape, patho- ithelial tumor with evidence of follicular cell differentiation
logical findings, and clinical behavior. The 2015 American but lacking diagnostic features of PTC (5). The definition also
Thyroid Association (ATA) guidelines for thyroid nodules included HCTC and poorly differentiated thyroid cancer
and thyroid cancer (1) focus primarily on PTC. Although (PDTC) (5). Several studies illustrate this changing diagnostic
much is known about FTC (2), many important questions pattern. Three University of Chicago pathologists re-analyzed
remain unanswered (3). This commentary summarizes the 66 cases of FTC diagnosed between 1965 and 2007. Only 19
recent literature on FTC and also reflects my perspective (29%) were still considered FTC, whereas 24 (36%) were re-
from over four decades of clinical experience with FTC. It is classified as PTC (21 classical and 3 follicular variant PTC
designed to highlight the distinctions between PTC and FTC [FVPTC]), 18 (27%) as follicular adenomas (FA), and 5 (8%)
as well as the many uncertainties concerning FTC. as PDTC (6). Pathologists at Kuma Hospital in Japan re-
reviewed 441 FTC diagnosed between 1983 and 2004; 198
Review (45%) were still considered FTC, whereas 44 (10%) were re-
classified as PDTC, 43 (9.8%) as HCTC, and the remainder
The clinical spectrum of PTC ranges from indolent pap-
benign FA, PTC, or FVPTC (7).
illary thyroid microcarcinomas (microPTC) (<1 cm) to large,
LiVolsi and Asa (8) speculated about ‘‘the demise of
highly aggressive tall cell or hobnail variants of PTC. Simi-
FTC.’’ But analysis of the National Cancer Institute’s Sur-
larly, FTC is not a single homogeneous entity. Yet, in my
veillance, Epidemiology, and End Results (SEER) database
experience, many clinical endocrinologists treat FTC as a
from 1980 to 2009 revealed a modest increase in age-adjusted
single disease requiring total (or completion) thyroidectomy
FTC rates in women (31.9%) and men (35.9%) (9) with an
followed by radioactive iodine (RAI) therapy. In the United
incidence rate of 0.88 cases per 100,000 person years. A
States, universal administration of RAI for PTC >1 cm is no
marked regional difference in incidence rates was found,
longer the standard of care; a far more nuanced approach is
raising important questions about the uniformity of diagno-
recommended and practiced (1). The extent of surgery and
sis. At the same time the ratio of PTC to FTC increased from
the use of RAI should also be tailored to the disease risk in
3.98 (1980–1984) to 9.88 (2005–2009); thus over 90% of
FTC. However, FTC outcomes vary considerably between
well-differentiated thyroid cancers are PTC. Over these
studies, suggesting a lack of uniformity in disease criteria.
years, FVPTC supplanted FTC as the most common follic-
Hürthle cell thyroid carcinoma (HCTC) is traditionally
ular patterned thyroid malignancy (9).
considered a variant of FTC. However, HCTC is a unique
The incidence ratio of PTC to FTC is lower in iodine de-
tumor whose pathogenesis and clinical behavior differs from
ficient countries and rises with iodine supplementation (10).
classical FTC (see below). Clinical studies that combine
Although iodine repletion may explain a decline in FTC in
HCTC with FTC, and those which do not comment on the
some countries, the U.S. population is generally iodine re-
inclusion of HCTC, may result in an imprecise picture of FTC.
plete. In some iodine-deficient countries the current inci-
For purposes of this review, the term FTC excludes HCTC.
dence ratio of PTC to FTC remains close to 1 (11).
Epidemiology
Cytology
Between 1973 and 2002, the annual incidence of PTC in
the United States increased from 3.6 to 8.7 cases per 100,000 The hallmark nuclear changes of PTC are often evident on
(4) and, according to the American Cancer Society, has al- cytological examination. Thus the pathological diagnosis of

1
Thyroid Unit, Cancer Center and Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston,
Massachusetts.

1229
 1230 DANIELS

PTC (Bethesda VI) after a fine needle aspiration biopsy of all PTC (24). PTC may occur in many rare hereditary
(FNAB) will be correct in almost 100% of cases, whereas tumor syndromes, but these make up a small percentage of
over 50% of FNAB suspicious for PTC (Bethesda V) will be familial PTC. These hereditary tumor syndromes (with as-
PTC (12). Although a minority of atypia of undetermined sociated mutations in parentheses) include familial adeno-
significance (AUS)/follicular lesion of undetermined signif- matous polyposis (FAP); Cowden’s Syndrome (PTEN and
icance (FLUS) FNAB (Bethesda III) are malignant, nuclear others); LiFraumani ( p53); DICER 1 (DICER 1); Lynch
atypia (AUS) increases the risk of malignancy compared with syndrome (DNA mismatch repair genes MLH1, MSH2,
architectural atypia (FLUS) (13). MSH6, PMS2, or EPCAM); Werner syndrome (WRN); and
In contrast, FTC may be suspected but generally cannot be Carney Complex (PRKAR1A) (25).
diagnosed by FNAB. The distinction between a FTC (ma- Oncogenic drivers in FTC are primarily RAS point muta-
lignant tumor with capsular and/or vascular invasion) and a tions (NRAS, HRAS, and less frequently KRAS) and
benign FA (no invasion) requires pathological examination PAX8PPARc rearrangements; RAS and PAX8PPARc are
of the tumor capsule after tumor excision. The risk of ma- mutually exclusive. The prevalence of RAS and PAX8PPARc
lignancy with a FNAB reading of follicular neoplasm (Be- varies considerably between studies. RAS mutations are
thesda IV) is 10–40% and 6–28% with AUS/FLUS (Bethesda found in up to 40–50% of FTC and 20–40% of FA;
III) (12). Whereas in the past the most common malignancy PAX8PPARc fusions are found in 30–40% of FTC and a
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

diagnosed after FNAB of follicular neoplasm was FTC, it is smaller percentage of FA (2–13%) (26). Finding the same
now FVPTC (14). One recent series represents an extreme oncogenic drivers in FTC and FA means that these drivers are
example of this trend. Of 1379 thyroid nodules undergoing not diagnostic of malignancy and also raises the question of
surgery for FN, 34% had a malignant tumor but none were whether FA with these drivers are premalignant or carcino-
FTC (15). mas in situ. RAS mutations and PAX8PPARc fusions are not
found in classical PTC but are common in FVPTC (see below).
Ultrasonography Additional oncogenic fusions in FTC include DERL-COX6C
and CREB3L2-PPARc (22). As in PTC, TERT promoter mu-
Thyroid nodules with the following characteristics have an
tations may occur in FTC and are independent of the oncogenic
increased risk of malignancy: solid, taller than wide, hy-
drivers. TERT promoter mutations appear to be far less common
poechoic or markedly hypoechoic, and/or irregular or spi-
in FA than FTC but the data are limited (27).
culated borders or calcifications (either micro or macro)
Chromosomal instability occurs in up to 65% of FTC, in-
(1,16). While these criteria help diagnose PTC, they are less
cluding extra copies, unbalanced rearrangements, complex
helpful for the diagnosis of FTC. Compared with PTC, FTC
karyotypes, and chromosomal losses. In one study RAS
are generally larger, tend to be isoechoic, have a hypoechoic
protein activator like 1 (RASAL1) mutations were found
rim and more often lack the suspicious ultrasound features
in almost 5% of FTC (28). Molecular alterations of the
which characterize PTC (17,18). Hoang et al. (17) compared
phosphatidylinositol 3-kinase–PTEN–AKT pathway occur
PTC to FTC and found the following statistically significant
in a variable proportion of FTCs, depending on whether
( p < 0.002) differences: round/tall (74.0% vs. 26.1%), hy-
mutations and/or gene copy number variations are con-
poechoic (72.4% vs. 34.8%), irregular margin (92.9 vs.
sidered (29).
60.9%), fine calcifications (33.9% vs. 0%), no hypoechoic
The prevalence of familial FTC is unknown. FTC may
rim (74.0% vs. 13.0%), and no cystic change (98.4% vs.
occur in rare hereditary cancer syndromes including Cow-
82.6%). Similarly, BRAFV600E mutant tumors (characteristic
den’s, Werner’s, and Carney complex. The most common
of PTC) have an ultrasound appearance that differs from
mutation in Cowden’s syndrome is PTEN (hence the name
those with RAS mutations (characteristic of FTC) (19). Ul-
PTEN-Hamartoma syndrome) but germ-line RASAL1 muta-
trasound cannot reliably distinguish between FA and FTC
tions may also cause a Cowden-like picture (30).
unless a highly vascularized tumor grossly protrudes beyond
its capsule—an uncommon finding (20).
Pathology
Genetic landscape
The pathological diagnosis of classical PTC is relatively
BRAFV600E is the most common oncogenic mutation in straightforward (31). The characteristic nuclear changes of
PTC, occurring in around 50% of classical PTC and in a PTC include nuclear enlargement and overlap, nuclear
higher percentage of the tall cell and hobnail variants. All elongation, irregular nuclear contours, nuclear pseudo-
thyroid tumors which contain BRAFV600E mutations are inclusions or prominent longitudinal grooves, empty ap-
considered malignant. RET/PTC, NTRK, ALK, and BRAF pearance of the nucleoplasm described as optically clear, or
oncogenic fusions occur in a minority of PTC cases and are ground glass nuclei. Papillary architecture is common but
not necessarily specific for malignancy (21,22). BRAFV600E need not be uniformly present. Psammoma bodies (lamellated
and these oncogenic fusions are mutually exclusive, and none calcifications) occur in about one-half of PTC. Multifocal
occur in FTC. Oncogenic fusions are more common in ra- disease is present in 20% or more cases. Tumor spread via
diation induced thyroid cancer (e.g., post-Chernobyl) (23). lymphatics to lymph nodes is common and occurs in at least
TERT (telomerase reverse transcriptase) promoter mutations 25% of PTC cases. When applied to PTC, the term ‘‘lym-
may also occur in PTC and are independent of the driver phovascular invasion’’ generally connotes lymphatic rather
mutations and fusions. PTC is generally a diploid tumor with than true vascular invasion. Lymphatic invasion correlates
little chromosomal instability. with lymph node metastases (LNM); lymphovascular invasion
Familial PTC (familial non-medullary thyroid carcinoma) without LNM may be a surrogate for subsequent LNM (32).
is a heterogeneous group of disorders and occurs in about 5% The prevalence of true (large vessel) vascular invasion in PTC
 FOLLICULAR THYROID CARCINOMA 1231

is uncertain. One study from a large referral center found large fewer sections were taken, thus downgrading these FTC to
vessel invasion in 47/698 PTC (6.7%); the median number of the minimally invasive category.
affected blood vessels was 1 (range 1–11) (33). It is unknown Given the difficulty in precisely diagnosing FTC, it is no
whether this experience can be generalized. surprise that important intra- and interobservation variation in
FVPTC is the most common variant of PTC; other variants diagnosis exists. For example, five leading French pathologists
include tall cell, morular cribriform (commonly associated reviewed 41 cases previously diagnosed as FTC (39). The final
with familial adenomatous polyposis), columnar cell, hobnail, consensus diagnosis was cancer in 30 (73%) cases but only
diffuse sclerosing, clear cell, Warthin-like, solid/trabecular, and 24/30 (80%) were diagnosed as FTC (59% overall). There
insular subtypes (31). was unanimous agreement about 13/24 (54%) cases of FTC,
MicroPTCs are very common, comprising about one-half with the highest agreement in those with more extensive VI.
of all PTC reported in the recent Korean PTC ‘‘epidemic’’ Diagnostic discrepancies occurred in 4/7 (57%) cases of
(34) with an even larger reservoir of undiagnosed microPTC. minimally invasive FTC. There was a tendency to defer to the
Conversely, the pathological diagnosis of FTC is more opinion of the group leader at the consensus conference even
difficult and requires great attention to detail (31). FTC does when the individual pathologist did not agree with that
not display cellular or nuclear atypia, hence the difficulty in opinion (known as the ‘‘leadership phenomenon’’).
diagnosing FTC by FNAB. Invasion of the tumor capsule or Given these diagnostic difficulties, many pathologists are
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

capsular blood vessels is the sine qua non of FTC; without reluctant to diagnose FTC by frozen section, although some
invasion, follicular tumors are benign (FA). A thick tumor institutions routinely use frozen section for diagnostic pur-
capsule (the pathological correlate of the hypoechoic rim on poses (40,41).
ultrasound) is generally found in FTC and FA but tends to be If HCTC is excluded, variants of FTC are uncommon to
thicker in FTC. rare and include the clear cell variant and mixed medullary
The WHO (31) defines capsular invasion (CI) as a tumor follicular variant (31,32). Clear cell change also occurs in
that completely transgresses the tumor capsule, often mush- PTC but is more common in FTC (42). Some FTC are de-
rooming beyond the capsule; follicular cells within the capsule scribed as having an insular, solid, or trabecular component,
are considered an artifact, likely related to prior FNAB. but the significance of these findings is uncertain; these are
However, one Armed Forces Institute of Pathology study di- not currently considered specific FTC variants.
agnosed tumors with <50% capsular invasion as FTC (35). The The terms ‘‘minimally invasive’’ and ‘‘extensively inva-
WHO and other pathologists define true vascular invasion (VI) sive’’ FTC have many different meanings, leading to many
as invasion into veins with tumor cells adherent to the vessel different ways of subclassifying FTC. The WHO (31) cur-
walls, either covered by endothelium or within a thrombus or rently divides FTC into three main pathological categories:
fibrin; in other words, large vessel invasion (5,31). (1) minimally invasive (CI alone); (2) encapsulated an-
Heffess’ and Thompson’s (36) comment from 2001 is gioinvasive (without regard to number of blood vessels) and
likely still true today: ‘‘Notwithstanding the wide acceptance (3) widely (i.e., grossly) invasive. In 2014 the Armed Forces
of the diagnostic criteria established by the World Health Institute of Pathology subdivided minimally invasive into
Organization for the classification of follicular carcinomas in those with CI alone, limited VI (<4 blood vessels), and ex-
particular, they have been difficult to apply and have led to a tensive VI (>4 blood vessels) (31). The term ‘‘widely inva-
great deal of confusion. . This confusion is compounded sive’’ was reserved for grossly invasive FTC. Most recent
when applied to ‘low-grade’ or ‘minimally invasive’ follic- publications define minimally invasive FTC as CI alone and/
ular carcinoma because of the poor reproducibility of the or minimal VI. Minimal VI generally refers to <4 blood
classification and the variable results reported in the litera- vessels involved, although some authors consider up to 10
ture.’’ It may also be difficult to be certain whether and how blood vessels as minimally invasive (43). The 2015 ATA
much vascular invasion is present. Common diagnostic errors guidelines recommend distinguishing between FTC with CI
include diagnosis of benign tumors as malignant, diagnosis of alone, with minimal VI (<4 blood vessels) and with extensive
malignant tumors as benign, and classifying tumors with VI (>4 blood vessels) (1). Some studies utilize four distinct
extensive VI as having minimal VI. FTC categories: CI, minimal VI (+ CI), extensive VI and
The diagnosis of CI or VI requires examination of multiple gross extrathyroidal extension (ETE) (44), whereas most
sections through the excised tumor, but this examination studies use the term extensively invasive FTC to include ei-
lacks standardization. Lang et al. (37) recommended at least ther extensive VI or gross ETE. Some studies (45) conform to
10 sections through the nodule, but most pathologists rec- the current WHO categories. One study refers to all FTC
ommend examination of the entire tumor capsule. Since ex- except for those with CI as ‘‘classical FTC’’ (46). Minimally
amination of a larger nodule will require more sections than a invasive FTC is generally far more common than widely
smaller nodule, it may be more appropriate to recommend a invasive FTC, but there are notable exceptions (see below).
certain number of sections per cm of tumor. Lang et al. (38) Given the ambiguity of these terms, it may be time to retire
evaluated 162 patients with minimally invasive FTC (CI the terms minimally invasive and extensively invasive FTC
alone and/or <4 foci of VI). The number of tissue blocks in favor of more specific but admittedly arbitrary criteria
examined per cm tumor was the only variable which signif- (namely: CI alone, minimal VI [<4] + CI, extensive VI [>4]
icantly correlated with distant metastatic disease (DM). None or extensive ETE). Although the validity of the distinction
of 82 patients with examination of >4 blocks/cm developed between <4 and >4 blood vessels is unproven, this analysis
DM compared with 7/80 (8.8%) with examination of <3 seems reasonable until more data are available. Sharing a
blocks/cm. Although open to many interpretations, it is common diagnostic language is a minimum requirement for
plausible to assume that more extensive VI was missed when valid comparisons.
 1232 DANIELS

FTC tends to be larger than PTC. The SEER database FTC at an earlier stage and at a smaller size, which may
(1988–2003) evaluated 30,504 PTC and 2584 FTC and re- influence the pathological features such as LNM. Based on a
ported a median diameter (interquartile range) of 1.5 cm study of 112,128 patients with thyroid cancer, Nguyen et al.
(0.8–2.5) and 3.0 cm (2.0–4.5), respectively; only 8% of PTC (52) (SEER 2004–2014) reported an ‘‘extremely low risk of
were larger than 4 cm compared with 27% of FTC (47). The LNM’’ for FTC primary tumors <5 cm; above 5 cm ‘‘the
SEER database (1988–2009) found 60 times as many mi- risk increases slightly but remains less than 20%.’’ Absolute
croPTC as microFTC (22,174 and 371, respectively) com- numbers cannot be determined from the data provided. It is
pared with an overall PTC:FTC ratio of 10:1 (48). It is also possible that actual disease behavior differs between
possible that follicular tumors <1 cm are not routinely scru- ethnic groups or geographic areas. Examples of the differ-
tinized for invasion or that it is difficult to be certain of ences between studies follow. Song et al. (2017) (53)
capsular or vascular invasion in small follicular tumors. compared FTC (likely including HCTC) diagnosed in three
Single institution–based publications may paint a different time periods: 1973–1995 (n = 68), 1996–2005 (n = 231), and
picture compared with large national database studies. Single 2006–2015 (n = 391). Widely invasive tumors (not other-
institution publications have the advantage of relative diag- wise specified) were found in 10.3%, 16.0%, and 12.3%,
nostic homogeneity, particularly with pathology re-review, respectively; multifocal tumors in 16.2%, 17.7%, and 13.3%,
but the disadvantages of small numbers of patients and in- respectively; ETE in 17.2%, 17.9%, and 11.0%, respectively;
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

stitutional biases. Analyses from national databases have the and LNM in 0%, 3.5%, and 2.6%. The decline in ETE was
advantage of larger numbers and homogenization of results statistically significant ( p = 0.0033).
across many institutions. However, without the possibility of Machens et al. (2005) (54) studied 134 FTC (including
pathological review, diagnostic uncertainty reigns. HCTC) diagnosed between 1994 and 2004; many of these
My mental picture of FTC reflects my experience at the tumors were considered clinically aggressive. Multifocal
Massachusetts General Hospital (MGH) but may not be an tumors were found in 9.0%, LNM in 19.4%, and ETE in
accurate picture for other institutions. FTC is a unifocal 17.2%. Like the WHO, O’Neil et al. (45) divided FTC
tumor, with rare LNM, that uncommonly demonstrates ex- (excluding HCTC, FVPTC, and PDTC) into three groups:
trathyroidal extension (ETE) and is primarily minimally in- CI alone (n = 61), any angioinvasion (n = 52), and widely
vasive. To confirm this impression, I reviewed the MGH invasive disease (gross ETE) (n = 11). LNM were present
pathology reports of all cases of FTC diagnosed from January in 3.2% overall but in 36% of widely invasive tumors.
2003 through March 2018 with institutional review board Podda et al. (55) found no LNM in 42 minimally invasive
approval. The pathology specimens were not re-reviewed; FTC (CI and/or <3 foci of VI). LNM were found in 5 of 29
HCTC, FVPTC, and PDTC were excluded, although diag- (17.2%) widely invasive FTC (>3 foci VI). Overall LNM
nostic uncertainty was apparent in some cases. Only 1/231 were found in 7.0% and multifocal disease in 2.8%. Huang
(0.4%) had LNM; that tumor was noted to have insular and et al. (2011) (40) found no LNM in 89 (0%) minimally
papillary features. Two additional patients had LNM from invasive (CI and/or minimal VI) FTC, whereas 4/145 (2.8%)
concomitant PTC. Only a single patient had equivocal mul- widely invasive FTC (widespread infiltration into blood
tiple foci of FTC (0.4%). ETE occurred in 7 patients (3%), vessels and/or adjacent thyroid tissue) had LNM. Overall
but 5 of these had unusual features such as an insular com- 1.7% had LNM. In this study 62% of FTC were widely
ponent, being solid and insular, being solid and trabecular, invasive.
being a clear cell variant, and having features of FTC, PTC, Based on analysis of SEER data (1988–2003), Zaydfudin
and clear cell tumor, respectively. Between 1990 and 2015, et al. (47) reported LNM metastases in 22% of 30,504 PTC
85% of FTC pathology specimens reviewed at MGH were compared with 2% of 2584 FTC. Goffredo et al. (SEER
minimally invasive; only 15% were widely invasive (49). 2000–2009) (56) reported LNM in 0.9% of 1200 minimally
WHO experts agree that that lymphatic invasion is so rare invasive (not otherwise specified) FTC and 3.6% of 4208
in FTC that when it is present, the diagnosis of FVPTC should widely invasive FTC (not otherwise specified) or 2.8%
be considered instead (31). Lodewijk et al. (50) found con- overall; ETE was found in 3.9% and 12.4%, respectively.
tralateral cancer (a surrogate for multifocality) in 280/794 Overall 78% of FTC were considered widely invasive. HCTC
(35%) PTC and 20/115 (17%) FTC. However, only 1 of the was likely included. Goffredo et al. (57) (National Cancer
20 contralateral tumors in the FTC cases was also FTC (0.8% Database 2010–2011) compared 333 FTC with CI alone with
overall.) FTC series which report multifocal disease gener- 284 FTC with minimal VI (+ CI). Multifocal disease was
ally do not specify the pathology of the additional tumor foci. found in 15.0 and 11.4%, respectively; ETE in 5.4% and
Cowden’s syndrome should be suspected when FTC is truly 9.2%, respectively; and LNM in 2.0% and 0%, respectively.
multifocal, and when FTC is found in conjunction with VI was correlated with larger tumor size. The number of
multiple follicular adenomas (including adenolipomas and involved blood vessels was not specified and HCTC was
microadenomas) (51). likely included.
The prevalence of LNM, multifocal disease, ETE and
the relative distribution of minimally invasive and widely
Outcomes: recurrent/persistent disease, distant
invasive FTC varies considerably between recent publica-
metastases, and mortality
tions, some of which are summarized below. These dis-
parities may be due to the intentional inclusion of HCTC or For many aggressive malignancies, there is a close corre-
the inadvertent inclusion of FVPTC or PDTC (see below), lation between disease recurrence and disease-specific mor-
but differences in diagnostic criteria and extent of tumor tality (DSM). For example, women with recurrent stage I
section are also likely explanations. Locations with a high breast cancer have a 15-year DSM of 32%, and with recurrent
rate of discovery of incidental thyroid nodules may diagnose stage II breast cancer the 15-year DSM is 59% (58).
 FOLLICULAR THYROID CARCINOMA 1233

PTC has a low risk of death (classical 2.7%, tall cell variant have a higher mortality (72), a finding possibly related to
6.7%, FVPTC 0.6%) (59) but a high risk of recurrence. The lead-time bias.
precise prevalence of recurrent/persistent PTC is uncertain, Some FTC studies report a decline in DM, DSM or tumor
but likely ranges from at least 25% for structural recurrences size and an increase in disease-free survival over time. Song
(60) to as high as 45% (61) when biochemical persistent et al. (53) compared FTC outcomes in three periods (1973–
disease (elevated thyroglobulin (Tg) without structural dis- 1995, 1996–2005, and 2006–2015) and noted a decline in the
ease) is included. LNM are the most common site of recur- rate of DM (14.7% of 68, 11.3% of 231 and 5.2% of 391), a
rent/persistent PTC; non-nodal soft tissue recurrences and change that was apparently not significant based on a survival
distant metastatic disease (DM) are far less common. Risk analysis. Yu et al. (73) compared 1997–2001, 2002–2006 and
factors for recurrent PTC (1,62) include LNM, extra- 2007–2011, with 3-year disease-free survival rates of 77.8%,
thyroidal extension, multifocal disease, age, and tumor size. 93.7%, and 100%, respectively, (p = 0.008) and a decline in
The mortality risk for PTC increases with age, extent of the median tumor diameter (4.2 cm to 2.8 cm and 2.9 cm,
extra-thyroidal extension, DM, and to some extent, LNM. By respectively), possibly due to earlier diagnosis.
multivariate analysis vascular invasion in classical PTC is not Many clinicians assume that the risk of DM and DSM in
considered an independent risk factor (33). For the same FTC increases with pathological progression from CI alone to
disease stage, DM that are 18F-fluoro-deoxyglucose (FDG)/ minimal VI (<4 blood vessels) to widely invasive with ex-
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

positron emission tomography (PET) positive have a worse tensive VI (>4 blood vessels) or gross ETE. The ATA low risk
prognosis than those which are FDG/PET negative (63). of recurrence group includes FTC with CI alone or minimal VI
It is uncertain whether BRAFV600E mutations indepen- (<4 blood vessels) + CI. The high risk of recurrence group
dently predict mortality in PTC (64) when comparable includes extensive VI (>4 foci) (1). However, the WHO (31)
pathological stages are considered and less aggressive vari- currently does not subclassify the extent of VI. Increased tu-
ants (e.g., FVPTC) are excluded, but associated TERT pro- mor size, older age and, possibly, the molecular profile also
moter mutations likely increase the risk of recurrences, DM correlate with DM and mortality. Multivariate analyses of risk
and DSM (65). BRAFV600E mutations also make tumors less for DM (summarized by Grani et al. (2)) from many different
likely to be RAI-avid (66). institutions report increased risk associated with older age,
Some PTC variants (tall cell, columnar cell, diffuse scle- larger tumor size, stage, angioinvasion, and ETE.
rosing, hobnail) appear to predict more aggressive disease One single institution study is instructive, particularly
and higher DSM, but it is uncertain whether that risk is in- concerning the timing of DM appearance. Sugrino et al.
dependent of tumor stage (67). (72) evaluated 134 FTC patients diagnosed between 1989
In contrast to PTC, when FTC recurs DM are usually and 1998. All pathology specimens were re-reviewed to
present; recurrent LNM are rare in FTC (68). As with PTC, exclude HCTC, clear cell FTC, FVPTC, and PDTC.
DM are primarily found in the lung and bone, but metastases Follow-up was sufficiently long to discover DM. DM were
to the liver, brain and other sites also occur. Although FTC present in 13 (9.7%) at presentation or within 1 year of
DM are not uniformly fatal, DSM in FTC is rare without DM. diagnosis; 9 of these 13 died of their disease, with disease
Radioactive iodine (RAI) therapy has great efficacy, but cure specific survival (DSS) of 41.0 and 30.8% at 5 and 10 years,
of bulky thyroid cancer metastases is rare (69,70). respectively. An additional 23 (19.1%) patients subse-
It is generally accepted that DM and DSM rates are higher quently developed DM (overall DM: 26.8%) at a mean
with FTC than PTC. For example, in a single institution study interval of 70 months after surgery (range 12–189 months),
(71) of 1503 well differentiated thyroid cancers, the 19 deaths and 9 of these 23 (39%) died of their disease, with overall
from FTC (of 221 FTC, 8.6%) was almost identical to the 18 10-, 15-, and 20-year DSS in this group of 89.9%, 84.6%,
deaths from PTC (of 1282 PTC, 1.4%) despite a PTC:FTC and 79.2%, respectively. Late DM were found in 3/32
ratio of 5.8:1.0. Similarly, based on the SEER database (1988– (10.3%) without VI (it is uncertain if these were CI alone),
2009) PTC accounted for 52.1% of DM (56) compared with 20/111 (18%) with minimally invasive FTC (CI and/or
40.4% from FTC and 7.4% from HCTC, despite the annual limited VI), 3/10 (30%) with widely invasive FTC (exten-
incidence ratio of PTC:FTC of 10:1. However, another study sive VI or gross ETE), and 3/7 (42.9%) with LNM. Overall,
using SEER (1988–2004) reported DM in 1.0% of PTC (305/ the majority of patients developing DM had minimally
30,504) compared with 3.0% in FTC (78/2584; possibly in- invasive FTC. Comparing those with primary tumors <4 or
cluding HCTC); overall 80% of DM were from PTC (47). >4 cm, DM occurred in 4/45 (8.9%) and 19/76 (25%), re-
The risk of DM in FTC ranges from a low of 3.0% to spectively ( p = 0.03). For age <45 years or >45 years, DM
almost 30% (see below). Possible explanations for these occurred in 4/58 (6.9%) and 19/76 (30.2%), respectively
variations include inclusion of HCTC (or less commonly ( p = 0.001). For those without DM, presentation age >45
PDTC), early diagnosis of FTC due to incidental detection of and size >4cm were independent predictors of DM, whereas
thyroid nodules, insufficient duration of follow-up, less for the entire group widely invasive disease was an inde-
stringent criteria for FTC (i.e., inclusion of FA), and reports pendent predictor of DM.
from centers of excellence where patients with DM may be The reported DM and DSM risk of each FTC invasive
referred prior to surgery, thus increasing the risk of DM. Once category varies considerably from study to study. Several
DM are present, either from PTC or FTC, the outcome is very small single institution studies report either no DM or no
similar, with a high correlation between macroscopic disease DSM in FTC with CI alone (46,49,74). Some studies report
progression and mortality (70). However, RAI avidity is no DM or DSM in minimally invasive FTC (CI and/or min-
more common in RAS than BRAFV600E mutant tumors (69). imal VI) (40,55). However, most studies report at least some
Up to one-half of the DM in FTC are present at diagnosis (or DM in each category of invasion as well as all size and age
within one year of diagnosis). Early onset DM is reported to categories.
 1234 DANIELS

In a concerning but fortunately unique study, D’Avanzo PTC (42), with reports of more aggressive and equally ag-
et al. (75) reported deaths in 5/45 (11%) FTC (likely in- gressive behavior compared with FTC (31). Little is known
cluding HCTC) patients with CI alone. One patient presented about the behavior of mixed medullary/follicular carcinoma.
with distant metastatic disease and 5 developed local recur- The impact of insular, trabecular, or solid changes in FTC is
rence, 4 of whom died. The pathology slides were not unknown. The behavior of FTC in genetic syndromes such as
available to confirm the diagnosis or extent of ETE (personal Cowden’s or Werner’s is also unknown.
communication). The impact of molecular abnormalities on FTC outcome,
O’Neill et al. (45) found DM in 2/61 (3.3%) FTC with CI beyond conventional pathological staging, is also uncertain,
alone, 6/52 (11.5%) FTC with vascular invasion and 5/11 although some studies suggest they may be helpful (76). Song
(45.0%) FTC with wide local invasion; disease free survival et al. (53) reported an association between RAS mutant FTC
at 40 months in these three groups was 97.0%, 81.0%, and and persistent disease and DM. Fukahori et al. (77) reported a
49.0%, respectively ( p < 0.01). DM were found in 1/55 positive association between NRAS codon 61 mutations and
(1.8%) patients under age 45 compared with 12/57 (17.5%) DM and RAS mutations in general with poor overall survival.
>age 45 ( p = 0.02) and in 5/38 (13.1%) with tumors <20 mm, Some studies suggest that tumors with PAX8PPARc are
4/56 (7.1%) between 21 and 40 mm, and 4/30 (13.3%) smaller and present at a younger age but are more likely to
>40 mm (p = .055; not significant, p < 0.05). have vascular invasion (26). Melo et al. reported a positive
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

Kim et al. (44) evaluated 204 FTC (including 17 HCTC) association between TERT promoter mutations and DM,
diagnosed between 1995 and 2010 after pathological re- ETE, VI, LNM and older age in 12/43 FTC patients with
review. DM occurred in 5/85 (6.0%) with CI alone, 8/80 these mutations, but the number of specimen studied was
(10.0%) with minimal VI, 4/13 (31.0%) with wide invasion small (78). It would be surprising if RAS or PAX8PPARc
without VI, and 12/26 (46.0%) with extensive VI. Overall alone provided important prognostic information beyond
DM occurred in 14.0%. tumor stage, given the presence of these changes in both FA
Glomski et al. (49) reported DM in 22/218 (10.0%) FTC and FTC and common genetic landscape of these tumors
but excluded an additional 25 cases of FTC with DM at when more extensive molecular testing is performed (79).
presentation (19.0%) overall. Of 145 cases with complete
tumor capsule examination 1/123 (0.8%) with CI and/or
Therapeutic considerations and surveillance
minimal VI developed new DM compared with 8/21 (38.0%)
with widely invasive FTC (extensive >4 blood vessels VI or Surgery for PTC is designed to minimize the risk of
ETE). When the tumor capsule was incompletely evaluated, recurrent disease (primarily LNM) and to facilitate RAI
1/45 (2.2%) with minimally invasive disease developed DM, therapy when necessary. The appropriate role of hemi-
compared with 11/17 (65.0%) with widely invasive disease. thyroidectomy in PTC versus total thyroidectomy is evolving
Podda et al. (55) evaluated 234 pathologically confirmed (1), but decisions about the extent of surgery are generally
FTC (possibly including HCTC) diagnosed between 1977 made after a likely diagnosis of PTC based on FNAB. After a
and 2007. Widely invasive FTC (widespread infiltration of total thyroidectomy, the serum Tg concentration 6 weeks
thyroid tissue and/or vascular invasion) was found in 0/4 postsurgery may help inform the decision about whether to
(0%) tumors <1 cm, 2/11 (17.0%) tumors 1 to <2 cm; 15/40 administer RAI; in some patients, a diagnostic radioactive
(37.5%) tumors 2–4 cm and 12/16 (75%) tumors >4 cm. iodine scan may be useful. Measurement of serum Tg may
Huang et al. (40) reported DM in 41/145 (28.3%) with also be helpful after a hemithyroidectomy but has somewhat
widely invasive disease (extensive VI or gross ETE) all of less utility than after a total thyroidectomy (80). After ei-
whom died. No DM were found in 89 minimally invasive ther hemi- or total thyroidectomy, post-operative ultrasound
FTC. It is worth noting that there was an unusual prevalence surveillance uncovers most recurrent PTC (1).
of widely invasive disease in this study. RAI is generally not administered for patients in the low
Goffredo et al. (SEER 1988–2009 (56)) reported DM in 6/ risk ATA category for recurrence, but it is generally admin-
1200 (0.5%) minimally invasive FTC (CI and/or minimal VI, istered for patients at high risk of recurrence and considered
not otherwise specified) compared with 29/4208 (6.9%) for patients at intermediate risk of recurrence with PTC (1).
widely invasive FTC likely including HCTC ( p < 0.001). Which intermediate recurrence risk patients with PTC should
Overall, 1.8% had DM. The unusual 3.5:1.0 ratio of widely to receive RAI is currently the subject of debate.
minimally invasive FTC is noteworthy. In a separate analysis Conversely, FTC may be suspected but is rarely confirmed
based on the National Cancer Database (2010–2011), DM prior to surgery. After obtaining a Bethesda III or IV FNA
were found in 1/333 (0.3%) FTC with CI alone compared cytology result, the final pathology will be benign or low
with 3/284 (1.1%) with minimal VI + CI (number of vessels grade follicular malignancy in the majority of cases, even
not specified) (57). when integrating results of modern molecular testing.
Kuo et al. (48) reported DM in 15/371 (4.1%) micro tu- Therefore, many (if not most) patients diagnosed with FTC
mors (FTC + HCTC), with 10-year DSS of 95.4%. DM were undergo hemi- rather than total thyroidectomy. After hemi-
also found in 110/22,174 (0.5%) microPTC with a 10-year thyroidectomy for FTC, if RAI is deemed necessary, (81)
DSS of 99.3%. Of note, the DM rate for microFTC in this either completion thyroidectomy or RAI lobe ablation is re-
study is higher than for all FTC (47), possibly indicating quired. FTC is primarily a unifocal disease where the primary
selection bias based on overreporting of aggressive mi- tumor is completely removed. Local recurrences are rare.
croFTC or the possible difficulty diagnosing FTC in sub- Therefore, RAI for FTC is generally administered to treat or
centimeter nodules. discover DM, rather than to prevent local recurrences.
If HCTC is excluded, the behavior of FTC variants is un- Compared with PTC, postoperative ultrasound surveillance is
certain. Clear cell changes are more common in FTC than far less useful for FTC because LNM and nodal recurrences
 FOLLICULAR THYROID CARCINOMA 1235

are rare (82). When FTC has extensive local invasion, ul- cerning potential new FTC staging systems primarily focus
trasound may be more useful. on the age cutoff for younger versus older patients (85).
RAI therapy is strongly recommended for FTC patients Although not universally accepted, most FTC studies
who present with DM and for those with extensive (>4 foci) consider the presence or extent of VI as important risk factor
of VI or extensive ETE, all of which fall into the ATA high for DM or DSM, yet VI is not included in the AJCC TNM 8th
recurrence risk category (1). FTC with CI alone and/or edition classification. It is possible that age and tumor size are
minimal VI (<4 blood vessels) fall into the ATA low recur- valid predictors of outcome in FTC, but these are likely to be,
rence risk group; by inference RAI is not recommended for at best, surrogates for VI. It is important to develop new TNM
these patients. However, some FTC patients with CI and/or staging systems for FTC that incorporate VI and validate
minimal VI require RAI, but it is uncertain which of these these by evaluating FTC patients with adequate duration of
patients require RAI and whether and how the risk for DM follow-up from many different institutions. In the future, it is
varies. After hemithyroidectomy, we recommend levothyr- likely that molecular abnormalities will also be incorporated
oxine for all FTC patients regardless of thyroid function into these staging systems. Validation of the current TNM
testing. For FTC with CI alone, we do not recommend classification and ATA recurrence risk categories for FTC is
completion thyroidectomy or radioactive iodine therapy. For also necessary.
FTC with extensive VI or gross ETE, we recommend com- It is likely that many FTC patients will be adequately
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

pletion thyroidectomy (or lobe ablation) followed by radio- staged by the current systems. For patients over age 55 years,
active iodine therapy. For FTC with minimal VI (< 4 blood those with DM at presentation will be stage IV, and those
vessels), we generally do not recommend completion thy- with extensive local invasion will be stage II or III. Most FTC
roidectomy or RAI, except for larger (>5 cm) tumors. It is patients with CI alone or minimal (<4 blood vessel) VI have
important to emphasize the important role of the pathologist an excellent prognosis and will be either stage I or II. How-
at each institution. Practice patterns would prompt more ever, patients with extensive VI will be either stage I or stage
liberal use of completion thyroidectomy and RAI at a given II (if >4 cm), which may not adequately predict their risk for
institution, if a significant minority of FTC patients with CI DM or DSM. These patients are included in the ATA high
alone or minimal VI develop DM. Many important questions risk of recurrence group. Two cases provide a microcosm of
can be posed: Can tumor size, patient age and/or molecular the problem: (1) a 65-year-old man with a 4.1 cm FTC with
profile help decide which FTC patients with CI and or min- CI alone is TNM stage II; and (2) the same patient with a
imal VI need RAI? How effective is serum Tg monitoring at 3.9 cm FTC with extensive VI is stage I. It is likely (but as yet
uncovering DM after total thyroidectomy without RAI? What unproven) that the risk is overestimated in the first patient and
is the utility of serum Tg monitoring for discovering DM after underestimated in the second.
hemithyroidectomy for FTC? Can that utility be improved by
levothyroxine therapy to lower serum thyrotropin (TSH) and
Clinical scenarios
serum Tg? In other words, are smaller changes in Tg more
meaningful if the baseline (postoperative Tg) is lower? What It is important to recognize several rare FTC clinical sce-
is the role of cross-sectional imaging such as CT or FDG narios. FTC may cause hyperthyroidism (or thyrotoxicosis)
PET/CT after hemithyroidectomy when RAI is deemed un- by three different mechanisms. Rarely, activating TSH
necessary? How do we weigh the benefit of early discovery of receptor mutations occur in FTC, clinically mimicking an
DM or the reassurance of a negative scan against the anxiety autonomously functioning adenoma (‘‘hot’’ nodule) with
resulting from a false positive scan (e.g., small indeterminate hyperthyroidism at presentation or when distant metastases
pulmonary nodules)? Does early detection and treatment of appear and grow. This is an exception to the rule that al-
FTC DM improve DSS or disease associated morbidity? most all ‘‘hot’’ nodules on radioiodine scanning are benign
Unfortunately, we do not have complete answers to any of the (86,87). Some patients with concomitant Graves’ disease and
foregoing questions. FTC develop hyperthyroidism as stimulating TSH receptor
antibodies (TRAb) drive the DM to produce thyroid hormone
(88,89). Hyperthyroidism in FTC due to autonomous func-
Staging
tion or TRAb stimulation generally responds to methimazole
Compared with prior editions, the eighth edition of the or RAI therapy. 3,5,3¢-triiodothyronine (T3) thyrotoxicosis
American Joint Committee on Cancer/Tumor-Node- may also occur in FTC with bulky DM due to increased
Metastasis Staging System (AJCC TNM) classification for conversion of administered levothyroxine to T3 by type 2
differentiated thyroid cancer improves our ability to predict iodothyronine deiodinase (90,91). In this scenario, elevated
the risk of death from PTC (83). The ATA risk categories for (or high normal) serum T3 occurs when levothyroxine is
recurrent/persistent PTC have also been validated in many administered, and profound hypothyroidism occurs when
studies (1). However, neither AJCC TNM, 8th edition, nor levothyroxine is discontinued. Because a fully suppressed
ATA recurrence risk categories have been validated specifi- serum TSH is often the therapeutic goal in FTC patients with
cally for FTC. DM, unless serum T3 (or free T3) is measured, deiodinase
In 2007 Lang et al. (84) compared several staging systems induced hyperthyroidism will be missed. If radioactive iodine
for FTC and, at the time, considered TNM AJCC 6th edition is not successful in treating the metastases, the thyrotoxicosis
to be the best, although 13/14 staging systems were consid- may be very difficult to treat. Propylthiouracil inhibits type 1
ered highly predictive of DSS. However, these staging sys- deiodinase but not type 2. Lenvatinib successfully treated one
tems were applied to an atypical population of 181 patients patient with T3 thyrotoxicosis due to increased conversion
diagnosed with FTC between 1961 and 2001, 49% of whom of T4 to T3 (92). Other potential unproven treatments in-
had extensively invasive disease. Current discussions con- clude progressively lowering the levothyroxine dosage or
 1236 DANIELS

prescribing multiple daily doses of liothyronine to suppress (FA). After careful review of the original pathology—often
serum TSH without causing symptomatic thyrotoxicosis reviewing additional tumor sections if possible—the original
(93). Cowden’s syndrome should be suspected in patients diagnosis may be changed to FTC (or FVPTC) (49,103), but
with rapid growth of thyroid nodules after a less than total in some cases the original tumor cannot be diagnosed as
thyroidectomy or when FTC is found in conjunction with malignant even after careful pathological re-review. Con-
multiple microadenomas, FA, and adenolipomas (51). Clas- sidering the rarity of these events, it is difficult to know
sic Cowden’s syndrome includes macrocephaly, trichi- whether there are any pathological or molecular character-
lemmomas, and papillomatous papule as well as an increased istics of FA that would warrant heightened surveillance, or in
risk for breast cancer, thyroid cancer, endometrial cancer, rare cases treatment, as if the tumor was a FTC. Without
colonic polyps, and possibly kidney cancer (30). Although of supporting data, we have adopted this approach. We often
unproven benefit, we measure head circumference in patients recommend long-term surveillance (with serum Tg moni-
with FTC, particularly young patients. If the head circum- toring) for patients with very large FA (arbitrarily over 5 cm)
ference is above the 95th percentile for height and sex (94) or particularly when the pathologists use terminology that ac-
above 58.6 cm in a male or 57.8 cm in a female, we consider knowledges either uncertainty or features which are atypical
germline genetic testing for PTEN mutations or PTEN im- but insufficient to diagnose FTC. In some particularly con-
munostaining on the pathology specimen. cerning cases after discussion with the patient, we recom-
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

mend radioactive iodine therapy/scanning. If the initial

Follicular variant of papillary thyroid cancer operation was a hemithyroidectomy, this approach requires
completion thyroidectomy or RAI lobe ablation followed by
The Cancer Genome Atlas divided PTC into two broad
RAI therapy. There are insufficient data to know if Tg
groups based on their molecular landscapes: (1) with
monitoring with or without levothyroxine therapy, appro-
BRAFV600E-like changes and (2) with RAS-like changes (95).
priate cross-sectional imaging, or a specific molecular profile
BRAFV600E mutations occur in up to one-third of FVPTC
would be helpful in providing guidance for atypical FA. It is
(96). These tumors are locally infiltrative, are often associ-
also reasonable to express similar concerns about very large
ated with LNM, and in general behave similarly to conven-
or atypical NIFTP tumors.
tional PTC, albeit with a follicular architecture. Some
investigators suggest diagnosing and treating these tumors as
classical PTC (97,98). PTC with RAS-like changes are all Hürthle cell thyroid cancer
FVPTC and behave differently from those with BRAFV600E
HCTC has traditionally been classified as a variant of
mutations. These RAS mutant tumors are encapsulated; the
FTC. However, it is a unique tumor requiring separate
diagnosis of malignancy depends on the presence of capsular
analysis and classification, a position supported by the
and/or vascular invasion. In other words, they share a mo-
ATA guidelines (1) and the WHO (31). The Hürthle cell,
lecular profile and pathological criteria with follicular, not
also known known as the oncocytic or oxyphilic cell, is large
papillary, tumors. Encapsulated, noninvasive FVPTC (often
with increased granular cytoplasm, large nuclei, and an in-
with RAS mutations) have recently been renamed as ‘‘non-
creased cytoplasmic-to-nuclear ratio. The cell is packed with
invasive follicular tumors with papillary like nuclear fea-
an increased number of abnormally enlarged mitochondria.
tures’’ (NIFTP) (99). Despite the nuclear atypia, these tumors
Hürthle cell tumors may infarct after a FNA. Like FTC, HCTC
behave in an indolent fashion. On the other hand, encapsu-
tends to be encapsulated, and capsular and/or vascular invasion
lated invasive FVPTC (CI and/or VI) are essentially identical
are necessary for the diagnosis of malignancy. Like FTC,
to FTC in their behavior, molecular profile, and pathological
HCTC with vascular invasion spreads to distant sites, but un-
criteria, except for the nuclear changes noted in FVPTC.
like FTC, LNM and soft tissue recurrences are more common.
Although it may be premature to expect pathologists to re-
HCTC demonstrates intense FDG uptake on PET scans even
name these tumors (e.g., FTC with nuclear atypia), it is not
with relatively small tumors. In contrast to FTC, HCTC is
premature to recommend that these tumors be analyzed in
much less likely to be RAI avid. Given the propensity for local
parallel with FTC with assessment of outcome based on CI,
recurrence, ultrasound surveillance may be helpful for HCTC.
minimal VI, extensive VI, ETE, size, age, and molecular
The inclusion of HCTC in many studies of FTC may prejudice
profile. If these encapsulated invasive FVPTC are indeed
the results.
identical to FTC (in all respects except nuclear atypia),
Like FTC, HCTC is currently categorized in different
changing the name to FTC would result in an effective
ways. Although it seems logical to categorize HCTC on the
doubling of the annual incidence of FTC (adding 1 case per
basis of CI, minimal VI, extensive VI, LNM, and gross ETE,
100,000 per year). This calculation is based on the annual
this categorization is not specifically mentioned by the WHO
incidence of PTC (around 15/100,000), the fraction of PTC
(31). Classification of HCTC by these criteria, as well as sex,
which are FVPTC (30%) (100) and the fraction of FVPTC
age and molecular profile, requires additional studies to as-
that are considered encapsulated invasive (23–31%) (101,102).
sess the DSM and recurrence risk for these categories. The
Subtracting NIFTP from the annual incidence of FVPTC will
genetic landscape of HCTC is different from the one in FTC
not change this result; the ratio of FVPTC/PTC will decline
and PTC. The HCTC genome is marked by recurrent mito-
but the percentage of encapsulated invasive FVPTC will rise
chondrial DNA mutations in complex 1 of the electron
proportionally.
transport chain (104), and widespread loss of heterozygosity
across many chromosomes often with genome-wide haploi-
Metastasizing follicular ‘‘adenomas’’
dization (105). These tumors also exhibit a distinct gene
In rare cases, DM compatible with FTC occurs many years expression profile (106); RAS point mutations and PAX8P-
after a thyroid tumor was originally diagnosed as benign PARc fusions are much less common than in FTC (106).
 FOLLICULAR THYROID CARCINOMA 1237

Although HCTC has a reputation for being a particularly provide appropriate guidance about the utility of RAI in FTC
aggressive form of well differentiated thyroid cancer, it is patients with CI and/or minimal VI.
uncertain whether the prognosis differs from FTC when To provide better answers, we need to standardize the di-
comparable disease stage is compared (107). Some data agnosis, nomenclature, and reporting of FTC. In private,
suggest that the prognosis for HCTC may be improving over pathologists admit that review of multiple tumor sections is,
time and that survival is comparable or superior to FTC; at a minimum, tedious, raising the question of a potential role
however, additional studies are necessary. It is also uncertain for artificial intelligence/machine learning for surveying
whether the diagnostic criteria for HCTC have changed over these sections. Given the relatively small numbers of cases at
time (108,109). any single institution, multicenter collaboration with central
pathology review in conjunction with advanced molecular
diagnostics is likely necessary to advance knowledge and
Poorly differentiated thyroid cancer
treatment of FTC. Given the length in time to appearance of
PDTC (31) has a prognosis intermediate between well- DM, these studies will need to utilize samples from earlier
differentiated thyroid cancer and anaplastic (undifferentiated) dates or extend the studies over many years. In the absence of
thyroid cancer, with a high prevalence of LNM, gross ETE, a clearly superior diagnostic schema, it seems reasonable to
DM, and DSM. The Turin diagnostic criteria include con- analyze four diagnostic categories (CI, minimal VI, extensive
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

ventional criteria for follicular cell–derived carcinoma, solid VI, ETE) and correlate these with size, age, molecular
trabecular or insular growth pattern, absence of the conven- analysis, and outcomes. Encapsulated invasive FVPTC cases
tional nuclear features of PTC, and at least one of the following should be studied in parallel to decide whether these are in-
three features: (1) convoluted nuclei, (2) >3 mitoses per 10 deed FTC or distinct tumors. Using these same criteria, a less
high power fields, and/or (3) tumor necrosis. Geographic dif- ambitious but possibly highly informative project would
ferences in diagnosis are apparently common (110). analyze data from many institutions on the ‘‘outlier’’ cases—
We include PDTC in this commentary for several reasons. those patients with ‘‘metastasizing FA’’ or FTC with CI
First, FTC series with a high prevalence of gross ETE might and/or minimal VI who develop DM. Preliminary studies of
include cases with PDTC. Second, PDTC often develops this population are beginning to appear (76). The terms
from well-differentiated FTC. Third, when only a small ‘‘minimally invasive FTC’’ and ‘‘widely invasive FTC’’ are
percentage of the tumor has poorly differentiated features, it ambiguous and might profitably be abandoned. Additional
might be categorized as FTC but behave in a more aggressive molecular testing comparing FA and FTC and indolent versus
fashion. Although it is generally assumed that PDTC is not aggressive FTC may provide additional diagnostic or thera-
RAI-avid, we have seen patients with PDTC with RAI-avid peutic information. Only with appropriate collaborative
DM, suggesting that the DM are from the FTC rather than the studies can we begin to understand whether the different
PDTC component. It is uncertain whether insular or solid/ outcomes reported by different institutions are due to dif-
trabecular features in an FTC are signs of a more aggressive ferences in pathology assessment or the nature of the tumor.
tumor and possible transition to PDTC. Clinicians also need more practical information. What is
the role of levothyroxine therapy after hemithyroidectomy
for FTC? Does TSH suppressive therapy prevent recur-
Guidelines
rences? Does lowering serum TSH with levothyroxine im-
Guidelines serve many purposes, such as providing guid- prove the prognostic value of serum Tg measurements by
ance for clinicians, and they are important educational and starting at a lower baseline? How well does serum Tg con-
reference resources for experts and nonexperts alike. An centration predict DM after total thyroidectomy with or
additional benefit is the necessary focus on what is known, without RAI? Does Tg surveillance facilitate earlier diag-
what is unknown, and what is only suspected or assumed. nosis of DM? Does early diagnosis of DM improve the DSM
Codifying the unknown and the assumed invariably raises or decrease morbidity of DM? What is the role of cross-
important questions and leads to research studies to answer sectional imaging to discover DM in patients not treated with
those questions. RAI? Is there a role for long-term surveillance or treatment of
The ATA 2015 guidelines (1) focus almost exclusively on large or atypical FA or NIFTP?
PTC with little discussion of the controversies concerning New staging systems incorporating VI as an important
FTC. Although anaplastic thyroid carcinoma and medullary variable should be explored for FTC and compared with the
thyroid carcinoma in aggregate have a lower annual inci- current TNM and risk of recurrence classifications.
dence compared with FTC, each has its own set of ATA Despite major knowledge gaps, publication of specific
guidelines. It is clearly the right time to provide new separate guidelines for FTC, HCTC and PDTC will undoubtedly ad-
guidelines for FTC, HCTC and PDTC to shine the spotlight vance knowledge and treatment for these tumors.
on these important and problematic tumors. Although beyond the scope of this perspective, new ther-
apies for radioactive iodine refractory FTC are needed. These
include further exploration/improvement of redifferentiation
Directions for future study
therapy as well as development of novel targeted therapies,
We do not know how well the WHO, AJCC TNM, eighth immunotherapies, and combination therapies.
edition, or ATA risk of recurrence categories apply to FTC
outcomes, specifically FTC with CI, minimal VI, or extensive
Conclusion
VI. We do not know how well size, age and molecular profile
predict outcomes independent of ETE or extent of vascular After his totally inaccurate obituary was published in 1897,
invasion. Until we have better information, it is difficult to Samuel Clemens (Mark Twain) commented, ‘‘The report of
 1238 DANIELS

my death was an exaggeration’’ (111). Similarly, the ‘‘de- 13. Valderrabano P, Khazai L, Thompson ZJ, Sharpe SC,
mise’’ of FTC is an exaggeration. Unfortunately, FTC contin- Tarasova VD, Otto KJ, Hallanger-Johnson JE, Wadsworth
ues to be a serious, sometimes lethal disease which requires JT, Wenig BM, Chung CH, Centeno BA, McIver B 2018
additional thought and study. Cancer risk associated with nuclear atypia in cytologically
indeterminate thyroid nodules: a systematic review and
Acknowledgments meta-analysis. Thyroid 28:210–219.
14. Daniels GH 2011 What if many follicular variant papillary
I would like to thank Josh Murphy for his excellent edi- thyroid carcinomas are not malignant? A review of fol-
torial assistance. licular variant papillary thyroid carcinoma and a proposal
for a new classification. Endocr Pract 17:768–787.
Author Disclosure Statement 15. Conzo G, Avenia N, Ansaldo GL, Calo P, De Palma M,
Dobrinja C, Docimo G, Gambardella C, Grasso M,
There are no disclosures or conflicts of interest to disclose.
Lombardi CP, Pelizzo MR, Pezzolla A, Pezzullo L, Pic-
coli M, Rosato L, Siciliano G, Spiezia S, Tartaglia E,
References Tartaglia F, Testini M, Troncone G, Signoriello G 2017
1. Haugen BR, Alexander EK, Bible KC, Doherty GM, Surgical treatment of thyroid follicular neoplasms: results
of a retrospective analysis of a large clinical series. En-
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

Mandel SJ, Nikiforov YE, Pacini F, Randolph GW, Sawka

AM, Schlumberger M, Schuff KG, Sherman SI, Sosa JA, docrine 55:530–538.
Steward DL, Tuttle RM, Wartofsky L 2016 2015 Amer- 16. Tessler FN, Middleton WD, Grant EG, Hoang JK, Berland
ican Thyroid Association Management Guidelines for LL, Teefey SA, Cronan JJ, Beland MD, Desser TS, Frates
Adult Patients with Thyroid Nodules and Differentiated MC, Hammers LW, Hamper UM, Langer JE, Reading
Thyroid Cancer: The American Thyroid Association CC, Scoutt LM, Stavros AT 2017 ACR Thyroid Imaging,
Guidelines Task Force on Thyroid Nodules and Differ- Reporting and Data System (TI-RADS): White paper of
entiated Thyroid Cancer. Thyroid 26:1–133. the ACR TI-RADS Committee. J Am Coll Radiol 14:587–
2. Grani G, Lamartina L, Durante C, Filetti S, Cooper DS 595.
2017 Follicular thyroid cancer and Hürthle cell carcino- 17. Hoang JK, Lee WK, Lee M, Johnson D, Farrell S 2007 US
ma: challenges in diagnosis, treatment, and clinical man- Features of thyroid malignancy: pearls and pitfalls.
agement. Lancet Diabetes Endocrinol 6:500–514. Radiographics 27:847–860; discussion 861–845.
3. Dionigi G, Kraimps JL, Schmid KW, Hermann M, Sheu- 18. Jeh SK, Jung SL, Kim BS, Lee YS 2007 Evaluating the
Grabellus SY, De Wailly P, Beaulieu A, Tanda ML, Sessa degree of conformity of papillary carcinoma and follicular
F 2014 Minimally invasive follicular thyroid cancer carcinoma to the reported ultrasonographic findings of
(MIFTC)—a consensus report of the European Society of malignant thyroid tumor. Korean J Radiol 8:192–197.
Endocrine Surgeons (ESES). Langenbecks Arch Surg 19. Kakarmath S, Heller HT, Alexander CA, Cibas ES, Krane
399:165–184. JF, Barletta JA, Lindeman NI, Frates MC, Benson CB,
4. Davies L, Welch HG 2006 Increasing incidence of thy- Gawande AA, Cho NL, Nehs M, Moore FD, Marqusee E,
roid cancer in the United States, 1973–2002. JAMA 295: Kim MI, Larsen PR, Kwong N, Angell TE, Alexander EK
2164–2167. 2016 Clinical, sonographic, and pathological characteris-
5. Asa SL 2017 The evolution of differentiated thyroid tics of RAS-Positive versus BRAF-positive thyroid car-
cancer. Pathology 49:229–237. cinoma. J Clin Endocrinol Metab 101:4938–4944.
6. Cipriani NA, Nagar S, Kaplan SP, White MG, Antic T, 20. Kobayashi K, Hirokawa M, Yabuta T, Masuoka H, Fu-
Sadow PM, Aschebrook-Kilfoy B, Angelos P, Kaplan kushima M, Kihara M, Higashiyama T, Ito Y, Miya A,
EL, Grogan RH 2015 Follicular thyroid carcinoma: how Amino N, Miyauchi A 2016 Tumor protrusion with in-
have histologic diagnoses changed in the last half-century tensive blood signals on ultrasonography is a strongly
and what are the prognostic implications? Thyroid 25: suggestive finding of follicular thyroid carcinoma. Med
1209–1216. Ultrason 18:25–29.
7. Hirokawa M, Ito Y, Kuma S, Takamura Y, Miya A, Ko- 21. Acquaviva G, Visani M, Repaci A, Rhoden KJ, de Biase
bayashi K, Miyauchi A 2010 Nodal metastasis in well- D, Pession A, Giovanni T 2018 Molecular pathology of
differentiated follicular carcinoma of the thyroid: its in- thyroid tumours of follicular cells: a review of genetic
cidence and clinical significance. Oncol Lett 1:873–876. alterations and their clinicopathological relevance. Histo-
8. LiVolsi VA, Asa SL 1994 The demise of follicular car- pathology 72:6–31.
cinoma of the thyroid gland. Thyroid 4:233–236. 22. Yakushina VD, Lerner LV, Lavrov AV 2018 Gene fusions
9. Aschebrook-Kilfoy B, Grogan RH, Ward MH, Kaplan E, in thyroid cancer. Thyroid 28:158–167.
Devesa SS 2013 Follicular thyroid cancer incidence pat- 23. Efanov AA, Brenner AV, Bogdanova TI, Kelly LM, Liu
terns in the United States, 1980–2009. Thyroid 23:1015– P, Little MP, Wald AI, Hatch M, Zurnadzy LY, Nikifor-
1021. ova MN, Drozdovitch V, Leeman-Neill R, Mabuchi K,
10. Feldt-Rasmussen U 2001 Iodine and cancer. Thyroid 11: Tronko MD, Chanock SJ, Nikiforov YE 2018 Investiga-
483–486. tion of the relationship between radiation dose and gene
11. Santos JE 2017 Thyroid cancer consequences of iodine mutations and fusions in post-Chernobyl thyroid cancer. J
deficiency — a persisting worldwide problem: character- Natl Cancer Inst 110:371–378.
istics in different populations, Portugal and South Africa. 24. Peiling Yang S, Ngeow J 2016 Familial non-medullary
S Afr J Surg 55:69. thyroid cancer: unraveling the genetic maze. Endocr Relat
12. Cibas ES, Ali SZ 2017 The 2017 Bethesda System for Cancer 23:R577-R595.
Reporting Thyroid Cytopathology. Thyroid 27:1341– 25. Guilmette J, Nose V 2018 Hereditary and familial thyroid
1346. tumours. Histopathology 72:70–81.
 FOLLICULAR THYROID CARCINOMA 1239

26. Nikiforov YE 2011 Molecular diagnostics of thyroid tu- section evaluation worthwhile? Arch Surg 132:674–678;
mors. Arch Pathol Lab Med 135:569–577. discussion 678–680.
27. Proietti A, Sartori C, Macerola E, Borrelli N, Materazzi G, 42. Cipriani NA, Agarwal S, Dias-Santagata D, Faquin WC,
Vitti P, Basolo F 2017 Low frequency of TERT promoter Sadow PM 2017 Clear cell change in thyroid carcinoma: a
mutations in a series of well-differentiated follicular- clinicopathologic and molecular study with identification
patterned thyroid neoplasms. Virchows Arch 471:769– of variable genetic anomalies. Thyroid 27:819–824.
773. 43. Ito Y, Hirokawa M, Masuoka H, Yabuta T, Fukushima M,
28. Liu D, Yang C, Bojdani E, Murugan AK, Xing M 2013 Kihara M, Higashiyama T, Takamura Y, Kobayashi K,
Identification of RASAL1 as a major tumor suppressor Miya A, Miyauchi A 2013 Distant metastasis at diagnosis
gene in thyroid cancer. J Natl Cancer Inst 105:1617–1627. and large tumor size are significant prognostic factors of
29. Paes JE, Ringel MD 2008 Dysregulation of the phospha- widely invasive follicular thyroid carcinoma. Endocr J 60:
tidylinositol 3-kinase pathway in thyroid neoplasia. En- 829–833.
docrinol Metab Clin North Am 37:375–387, viii-ix. 44. Kim HJ, Sung JY, Oh YL, Kim JH, Son YI, Min YK, Kim
30. Ngeow J, Ni Y, Tohme R, Song Chen F, Bebek G, Eng C SW, Chung JH 2014 Association of vascular invasion with
2014 Germline alterations in RASAL1 in Cowden syn- increased mortality in patients with minimally invasive
drome patients presenting with follicular thyroid cancer follicular thyroid carcinoma but not widely invasive fol-
and in individuals with apparently sporadic epithelial licular thyroid carcinoma. Head Neck 36:1695–1700.
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

thyroid cancer. J Clin Endocrinol Metab 99:E1316–1321. 45. O’Neill CJ, Vaughan L, Learoyd DL, Sidhu SB, Delbridge
31. Lloyd RV, Osamura RY, Klöppel G, Rosai J (eds) 2017 LW, Sywak MS 2011 Management of follicular thyroid
WHO Classification of Tumours of Endocrine Organs, carcinoma should be individualised based on degree of
fourth edition). International Agency for Research on capsular and vascular invasion. Eur J Surg Oncol 37:181–
Cancer [IRAC], Lyon. 185.
32. Prescott JD, Sadow PM, Hodin RA, Le LP, Gaz RD, 46. Robinson A, Schneider D, Sippel R, Chen H 2017 Mini-
Randolph GW, Stephen AE, Parangi S, Daniels GH, Lu- mally invasive follicular thyroid cancer: treat as a benign
bitz CC 2012 BRAF V600E status adds incremental value or malignant lesion? J Surg Res 207:235–240.
to current risk classification systems in predicting papil- 47. Zaydfudim V, Feurer ID, Griffin MR, Phay JE 2008 The
lary thyroid carcinoma recurrence. Surgery 152:984–990. impact of lymph node involvement on survival in patients
33. Wreesmann VB, Nixon IJ, Rivera M, Katabi N, Palmer F, with papillary and follicular thyroid carcinoma. Surgery
Ganly I, Shaha AR, Tuttle RM, Shah JP, Patel SG, 144:1070–1077; discussion 1077–1078.
Ghossein RA 2015 Prognostic value of vascular invasion 48. Kuo EJ, Roman SA, Sosa JA 2013 Patients with follicular
in well-differentiated papillary thyroid carcinoma. Thy- and Hürthle cell microcarcinomas have compromised
roid 25:503–508. survival: a population level study of 22,738 patients.
34. Ahn HS, Kim HJ, Welch HG 2014 Korea’s thyroid-cancer Surgery 154:1246–1253; discussion 1253–1244.
‘‘epidemic’’—-screening and overdiagnosis. N Engl J 49. Glomski K, Nose V, Faquin WC, Sadow PM 2017 Me-
Med 371:1765–1767. tastatic follicular thyroid carcinoma and the primary thy-
35. Thompson LD, Wieneke JA, Paal E, Frommelt RA, Adair roid gross examination: institutional review of cases from
CF, Heffess CS 2001 A clinicopathologic study of mini- 1990 to 2015. Endocr Pathol 28:177–185.
mally invasive follicular carcinoma of the thyroid gland 50. Lodewijk L, Kluijfhout WP, Kist JW, Stegeman I, Plukker
with a review of the English literature. Cancer 91:505– JT, Nieveen van Dijkum EJ, Bonjer HJ, Bouvy ND,
524. Schepers A, de Wilt JH, Netea-Maier RT, van der Hage
36. Heffess CS, Thompson LD 2001 Minimally invasive JA, Burger JW, Ho G, Lee WS, Shen WT, Aronova A,
follicular thyroid carcinoma. Endocr Pathol 12:417–422. Zarnegar R, Benay C, Mitmaker EJ, Sywak MS, Aniss
37. Lang W, Georgii A, Stauch G, Kienzle E 1980 The dif- AM, Kruijff S, James B, Grogan RH, Brunaud L, Hoch G,
ferentiation of atypical adenomas and encapsulated fol- Pandolfi C, Ruan DT, Jones MD, Guerrero MA, Valk GD,
licular carcinomas in the thyroid gland. Virchows Arch A Borel Rinkes IH, Vriens MR 2016 Characteristics of
Pathol Anat Histol 385:125–141. contralateral carcinomas in patients with differentiated
38. Lang BH, Shek TW, Wu AL, Wan KY 2017 The total thyroid cancer larger than 1 cm. Langenbecks Arch Surg
number of tissue blocks per centimetre of tumor signifi- 401:365–373.
cantly correlated with the risk of distant metastasis in 51. Cameselle-Teijeiro J, Fachal C, Cabezas-Agricola JM,
patients with minimally invasive follicular thyroid carci- Alfonsin-Barreiro N, Abdulkader I, Vega-Gliemmo A,
noma. Endocrine 55:496–502. Hermo JA 2015 Thyroid pathology findings in cowden
39. Franc B, de la Salmoniere P, Lange F, Hoang C, Louvel syndrome: a clue for the diagnosis of the PTEN ha-
A, de Roquancourt A, Vilde F, Hejblum G, Chevret S, martoma tumor syndrome. Am J Clin Pathol 144:322–
Chastang C 2003 Interobserver and intraobserver repro- 328.
ducibility in the histopathology of follicular thyroid car- 52. Nguyen XV, Roy Choudhury K, Tessler FN, Hoang JK
cinoma. Hum Pathol 34:1092–1100. 2018 Effect of tumor size on risk of metastatic disease and
40. Huang CC, Hsueh C, Liu FH, Chao TC, Lin JD 2011 survival for thyroid cancer: implications for biopsy guide-
Diagnostic and therapeutic strategies for minimally and lines. Thyroid 28:295–300.
widely invasive follicular thyroid carcinomas. Surg Oncol 53. Song YS, Lim JA, Min HS, Kim MJ, Choi HS, Cho SW,
20:1–6. Moon JH, Yi KH, Park DJ, Cho BY, Park YJ 2017
41. Paphavasit A, Thompson GB, Hay ID, Grant CS, van Changes in the clinicopathological characteristics and
Heerden JA, Ilstrup DM, Schleck C, Goellner JR 1997 genetic alterations of follicular thyroid cancer. Eur J En-
Follicular and Hürthle cell thyroid neoplasms. Is frozen- docrinol 177:465–473.
 1240 DANIELS

54. Machens A, Holzhausen HJ, Dralle H 2005 The prog- Zanzonico P, Larson SM, Refetoff S, Ghossein R, Fagin
nostic value of primary tumor size in papillary and fol- JA 2011 Small-molecule MAPK inhibitors restore radio-
licular thyroid carcinoma. Cancer 103:2269–2273. iodine incorporation in mouse thyroid cancers with con-
55. Podda M, Saba A, Porru F, Reccia I, Pisanu A 2015 ditional BRAF activation. J Clin Invest 121:4700–4711.
Follicular thyroid carcinoma: differences in clinical rele- 67. Silver CE, Owen RP, Rodrigo JP, Rinaldo A, Devaney
vance between minimally invasive and widely invasive KO, Ferlito A 2011 Aggressive variants of papillary thy-
tumors. World J Surg Oncol 13:193. roid carcinoma. Head Neck 33:1052–1059.
56. Goffredo P, Cheung K, Roman SA, Sosa JA 2013 68. Vogrin A, Besic H, Besic N, Music MM 2016 Recurrence
Can minimally invasive follicular thyroid cancer be ap- rate in regional lymph nodes in 737 patients with follicular
proached as a benign lesion?: a population-level analysis or Hürthle cell neoplasms. Radiol Oncol 50:269–273.
of survival among 1,200 patients. Ann Surg Oncol 20: 69. Sabra MM, Dominguez JM, Grewal RK, Larson SM,
767–772. Ghossein RA, Tuttle RM, Fagin JA 2013 Clinical out-
57. Goffredo P, Jillard C, Thomas S, Scheri RP, Sosa JA, comes and molecular profile of differentiated thyroid
Roman S 2016 Minimally invasive follicular carcinoma: cancers with radioiodine-avid distant metastases. J Clin
predictors of vascular invasion and impact on patterns of Endocrinol Metab 98:E829–836.
care. Endocrine 51:123–130. 70. Sabra MM, Ghossein R, Tuttle RM 2016 Time course and
58. Sopik V, Nofech-Mozes S, Sun P, Narod SA 2016 The predictors of structural disease progression in pulmonary
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

relationship between local recurrence and death in early- metastases arising from follicular cell-derived thyroid
stage breast cancer. Breast Cancer Res Treat 155:175–185. cancer. Thyroid 26:518–524.
59. Shi X, Liu R, Basolo F, Giannini R, Shen X, Teng D, 71. Sciuto R, Romano L, Rea S, Marandino F, Sperduti I,
Guan H, Shan Z, Teng W, Musholt TJ, Al-Kuraya K, Maini CL 2009 Natural history and clinical outcome of
Fugazzola L, Colombo C, Kebebew E, Jarzab B, Czar- differentiated thyroid carcinoma: a retrospective analysis
niecka A, Bendlova B, Sykorova V, Sobrinho-Simoes M, of 1503 patients treated at a single institution. Ann Oncol
Soares P, Shong YK, Kim TY, Cheng S, Asa SL, Viola D, 20:1728–1735.
Elisei R, Yip L, Mian C, Vianello F, Wang Y, Zhao S, 72. Sugino K, Ito K, Nagahama M, Kitagawa W, Shibuya H,
Oler G, Cerutti JM, Puxeddu E, Qu S, Wei Q, Xu H, Ohkuwa K, Yano Y, Uruno T, Akaishi J, Kameyama K,
O’Neill CJ, Sywak MS, Clifton-Bligh R, Lam AK, Ito K 2011 Prognosis and prognostic factors for distant
Riesco-Eizaguirre G, Santisteban P, Yu H, Tallini G, Holt metastases and tumor mortality in follicular thyroid car-
EH, Vasko V, Xing M 2016 Differential clinicopatho- cinoma. Thyroid 21:751–757.
logical risk and prognosis of major papillary thyroid 73. Yu XF, Wang WB, Teng XD, Wang HY, Chen X, Wang
cancer variants. J Clin Endocrinol Metab 101:264–274. HH, Ma ZM, Fahey TJ, 3rd, Teng LS 2014 Clin-
60. Mazzaferri EL, Kloos RT 2001 Clinical review 128: icopathological and prognostic analysis of follicular thy-
Current approaches to primary therapy for papillary and roid carcinoma in a single institute over a 15-year period.
follicular thyroid cancer. J Clin Endocrinol Metab 86: Eur J Surg Oncol 40:869–874.
1447–1463. 74. van Heerden JA, Hay ID, Goellner JR, Salomao D,
61. Grant CS 2015 Recurrence of papillary thyroid cancer Ebersold JR, Bergstralh EJ, Grant CS 1992 Follicular
after optimized surgery. Gland Surg 4:52–62. thyroid carcinoma with capsular invasion alone: a non-
62. Krajewska J, Chmielik E, Jarzab B 2017 Dynamic risk threatening malignancy. Surgery 112:1130–1136; discus-
stratification in the follow-up of thyroid cancer: what is sion 1136–1138.
still to be discovered in 2017? Endocr Relat Cancer 24: 75. D’Avanzo A, Treseler P, Ituarte PH, Wong M, Streja L,
R387-R402. Greenspan FS, Siperstein AE, Duh QY, Clark OH 2004
63. Robbins RJ, Wan Q, Grewal RK, Reibke R, Gonen Follicular thyroid carcinoma: histology and prognosis.
M, Strauss HW, Tuttle RM, Drucker W, Larson SM Cancer 100:1123–1129.
2006 Real-time prognosis for metastatic thyroid carci- 76. Xu B, Tuttle RM, Sabra MM, Ganly I, Ghossein R 2017
noma based on 2-[18F]fluoro-2-deoxy-D-glucose-positron Primary thyroid carcinoma with low-risk histology and
emission tomography scanning. J Clin Endocrinol Metab distant metastases: clinicopathologic and molecular char-
91:498–505. acteristics. Thyroid 27:632–640.
64. Xing M, Alzahrani AS, Carson KA, Viola D, Elisei R, 77. Fukahori M, Yoshida A, Hayashi H, Yoshihara M, Mat-
Bendlova B, Yip L, Mian C, Vianello F, Tuttle RM, Ro- sukuma S, Sakuma Y, Koizume S, Okamoto N, Kondo T,
benshtok E, Fagin JA, Puxeddu E, Fugazzola L, Czar- Masuda M, Miyagi Y 2012 The associations between RAS
niecka A, Jarzab B, O’Neill CJ, Sywak MS, Lam AK, mutations and clinical characteristics in follicular thyroid
Riesco-Eizaguirre G, Santisteban P, Nakayama H, Tufano tumors: new insights from a single center and a large
RP, Pai SI, Zeiger MA, Westra WH, Clark DP, Clifton- patient cohort. Thyroid 22:683–689.
Bligh R, Sidransky D, Ladenson PW, Sykorova V 2013 78. Melo M, Gaspar da Rocha A, Batista R, Vinagre J, Mar-
Association between BRAF V600E mutation and mor- tins MJ, Costa G, Ribeiro C, Carrilho F, Leite V, Lobo C,
tality in patients with papillary thyroid cancer. JAMA Cameselle-Teijeiro JM, Cavadas B, Pereira L, Sobrinho-
309:1493–1501. Simoes M, Soares P 2017 TERT, BRAF, and NRAS in
65. Moon S, Song YS, Kim YA, Lim JA, Cho SW, Moon JH, primary thyroid cancer and metastatic disease. J Clin
Hahn S, Park DJ, Park YJ 2017 Effects of coexistent Endocrinol Metab 102:1898–1907.
BRAF(V600E) and TERT promoter mutations on poor 79. Jung SH, Kim MS, Jung CK, Park HC, Kim SY, Liu J,
clinical outcomes in papillary thyroid cancer: a meta- Bae JS, Lee SH, Kim TM, Lee SH, Chung YJ 2016
analysis. Thyroid 27:651–660. Mutational burdens and evolutionary ages of thyroid fol-
66. Chakravarty D, Santos E, Ryder M, Knauf JA, Liao XH, licular adenoma are comparable to those of follicular
West BL, Bollag G, Kolesnick R, Thin TH, Rosen N, carcinoma. Oncotarget 7:69638–69648.
 FOLLICULAR THYROID CARCINOMA 1241

80. Van Wyngaarden K, McDougall IR 1997 Is serum thy- cosis in patients with metastatic follicular thyroid cancer
roglobulin a useful marker for thyroid cancer in patients treated with lenvatinib. Thyroid 25:1262–1264.
who have not had ablation of residual thyroid tissue? 93. Celi FS, Zemskova M, Linderman JD, Babar NI, Skarulis
Thyroid 7:343–346. MC, Csako G, Wesley R, Costello R, Penzak SR, Pucino
81. Barbesino G, Goldfarb M, Parangi S, Yang J, Ross DS, F 2010 The pharmacodynamic equivalence of levothyr-
Daniels GH 2012 Thyroid lobe ablation with radioactive oxine and liothyronine: a randomized, double blind, cross-
iodine as an alternative to completion thyroidectomy over study in thyroidectomized patients. Clin Endocrinol
after hemithyroidectomy in patients with follicular (Oxf) 72:709–715.
thyroid carcinoma: long-term follow-up. Thyroid 22: 94. Bushby KM, Cole T, Matthews JN, Goodship JA 1992
369–376. Centiles for adult head circumference. Arch Dis Child 67:
82. Baek HJ, Kim DW, Lee S, Ryoo I, Lee CY, Choi YJ, 1286–1287.
Sung JY 2017 Postoperative ultrasonography surveillance 95. Network CGAR 2014 Integrated genomic characterization
in patients with follicular thyroid carcinoma: a multicenter of papillary thyroid carcinoma. Cell 159:676–690.
study. Radiol Med 122:530–537. 96. McFadden DG, Dias-Santagata D, Sadow PM, Lynch KD,
83. Tuttle RM, Haugen B, Perrier ND 2017 Updated Ameri- Lubitz C, Donovan SE, Zheng Z, Le L, Iafrate AJ, Daniels
can Joint Committee on Cancer/Tumor-Node-Metastasis GH 2014 Identification of oncogenic mutations and gene
Staging system for differentiated and anaplastic thyroid fusions in the follicular variant of papillary thyroid car-
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.

cancer (eighth edition): what changed and why? Thyroid cinoma. J Clin Endocrinol Metab 99:E2457–2462.
27:751–756. 97. Asa SL, Giordano TJ, LiVolsi VA 2015 Implications of
84. Lang BH, Lo CY, Chan WF, Lam KY, Wan KY 2007 the TCGA genomic characterization of papillary thyroid
Staging systems for follicular thyroid carcinoma: appli- carcinoma for thyroid pathology: does follicular variant
cation to 171 consecutive patients treated in a tertiary papillary thyroid carcinoma exist? Thyroid 25:1–2.
referral centre. Endocr Relat Cancer 14:29–42. 98. Daniels GH 2016 Follicular variant of papillary thyroid
85. McLeod DS, Jonklaas J, Brierley JD, Ain KB, Cooper DS, carcinoma: hybrid or mixture? Thyroid 26:872–874.
Fein HG, Haugen BR, Ladenson PW, Magner J, Ross DS, 99. Nikiforov YE, Seethala RR, Tallini G, Baloch ZW, Basolo
Skarulis MC, Steward DL, Xing M, Litofsky DR, Maxon F, Thompson LD, Barletta JA, Wenig BM, Al Ghuzlan A,
HR, Sherman SI 2015 Reassessing the NTCTCS staging Kakudo K, Giordano TJ, Alves VA, Khanafshar E, Asa SL,
systems for differentiated thyroid cancer, including age at El-Naggar AK, Gooding WE, Hodak SP, Lloyd RV, Maytal
diagnosis. Thyroid 25:1097–1105. G, Mete O, Nikiforova MN, Nose V, Papotti M, Poller DN,
86. Niepomniszcze H, Suarez H, Pitoia F, Pignatta A, Dani- Sadow PM, Tischler AS, Tuttle RM, Wall KB, LiVolsi VA,
lowicz K, Manavela M, Elsner B, Bruno OD 2006 Folli- Randolph GW, Ghossein RA 2016 Nomenclature revision
cular carcinoma presenting as autonomous functioning for encapsulated follicular variant of papillary thyroid car-
thyroid nodule and containing an activating mutation of cinoma: a paradigm shift to reduce overtreatment of indo-
the TSH receptor (T620I) and a mutation of the Ki-RAS lent tumors. JAMA Oncol 2:1023–1029.
(G12C) genes. Thyroid 16:497–503. 100. Albores-Saavedra J, Henson DE, Glazer E, Schwartz AM
87. Spambalg D, Sharifi N, Elisei R, Gross JL, Medeiros-Neto 2007 Changing patterns in the incidence and survival of
G, Fagin JA 1996 Structural studies of the thyrotropin thyroid cancer with follicular phenotype–papillary, fol-
receptor and Gs alpha in human thyroid cancers: low licular, and anaplastic: a morphological and epidemio-
prevalence of mutations predicts infrequent involvement logical study. Endocr Pathol 18:1–7.
in malignant transformation. J Clin Endocrinol Metab 81: 101. Kim MJ, Won JK, Jung KC, Kim JH, Cho SW, Park DJ,
3898–3901. Park YJ 2018 Clinical characteristics of subtypes of follicular
88. Ishihara T, Ikekubo K, Shimodahira M, Iwakura T, variant papillary thyroid carcinoma. Thyroid 28:311–318.
Kobayashi M, Hino M, Oobayashi M, Kohno K, Kimura 102. Liu J, Singh B, Tallini G, Carlson DL, Katabi N, Shaha A,
K, Kawamura S, Kurahachi H 2002 A case of TSH re- Tuttle RM, Ghossein RA 2006 Follicular variant of pap-
ceptor antibody-positive hyperthyroidism with func- illary thyroid carcinoma: a clinicopathologic study of a
tioning metastases of thyroid carcinoma. Endocr J 49: problematic entity. Cancer 107:1255–1264.
241–245. 103. Baloch ZW, LiVolsi VA 2000 Encapsulated follicular
89. Stathatos N, Gaz R, Ross DS, Daniels GH 2011 High variant of papillary thyroid carcinoma with bone metas-
radioactive iodine uptake despite a fully suppressed TSH tases. Mod Pathol 13:861–865.
in a patient with thyroid cancer. J Clin Endocrinol Metab 104. Gasparre G, Porcelli AM, Bonora E, Pennisi LF, Toller
96:589. M, Iommarini L, Ghelli A, Moretti M, Betts CM, Mar-
90. Kim BW, Daniels GH, Harrison BJ, Price A, Harney JW, tinelli GN, Ceroni AR, Curcio F, Carelli V, Rugolo M,
Larsen PR, Weetman AP 2003 Overexpression of type 2 Tallini G, Romeo G 2007 Disruptive mitochondrial DNA
iodothyronine deiodinase in follicular carcinoma as a mutations in complex I subunits are markers of oncocytic
cause of low circulating free thyroxine levels. J Clin En- phenotype in thyroid tumors. Proc Natl Acad Sci USA
docrinol Metab 88:594–598. 104:9001–9006.
91. Miyauchi A, Takamura Y, Ito Y, Miya A, Kobayashi K, 105. Corver WE, Ruano D, Weijers K, den Hartog WC, van
Matsuzuka F, Amino N, Toyoda N, Nomura E, Nishikawa Nieuwenhuizen MP, de Miranda N, van Eijk R, Mid-
M 2008 3,5,3¢-Triiodothyronine thyrotoxicosis due to in- deldorp A, Jordanova ES, Oosting J, Kapiteijn E, Hovens
creased conversion of administered levothyroxine in pa- G, Smit J, van Wezel T, Morreau H 2012 Genome hap-
tients with massive metastatic follicular thyroid carcinoma. loidisation with chromosome 7 retention in oncocytic
J Clin Endocrinol Metab 93:2239–2242. follicular thyroid carcinoma. PLoS One 7:e38287.
92. Danilovic DL, de Camargo RY, Castro G, Jr., Papadia C, 106. Ganly I, Ricarte Filho J, Eng S, Ghossein R, Morris LG,
Marui S, Hoff AO 2015 Rapid control of T3 thyrotoxi- Liang Y, Socci N, Kannan K, Mo Q, Fagin JA, Chan TA
 1242 DANIELS

2013 Genomic dissection of Hürthle cell carcinoma re- survival over the last 35 years. Surgery 154:1263–1271;
veals a unique class of thyroid malignancy. J Clin En- discussion 1271.
docrinol Metab 98:E962–972. 110. Volante M, Bussolati G, Papotti M 2016 The story of
107. Chindris AM, Casler JD, Bernet VJ, Rivera M, Thomas C, poorly differentiated thyroid carcinoma: From Langhans’
Kachergus JM, Necela BM, Hay ID, Westphal SA, Grant description to the Turin proposal via Juan Rosai. Semin
CS, Thompson GB, Schlinkert RT, Thompson EA, Diagn Pathol 33:277–283.
Smallridge RC 2015 Clinical and molecular features of 111. Campbell WJ 2006 The Year That Defined American
Hürthle cell carcinoma of the thyroid. J Clin Endocrinol Journalism 1897 and the Clash of Paradigms. Taylor &
Metab 100:55–62. Francis Group, New York.
108. Ernaga Lorea A, Migueliz Bermejo I, Anda Apinaniz E,
Pineda Arribas J, Toni Garcia M, Martinez de Esteban JP, Address correspondence to:
Insausti Serrano AM 2018 Comparison of clinical char- Gilbert H. Daniels, MD
acteristics of patients with follicular thyroid carcinoma Thyroid Unit ACC 730
and Hürthle cell carcinoma. Endocrinol Diabetes Nutr 65: Massachusetts General Hospital
136–142. Boston, MA 02114
109. Nagar S, Aschebrook-Kilfoy B, Kaplan EL, Angelos P,
Grogan RH 2013 Hürthle cell carcinoma: an update on E-mail: [email protected]
Downloaded by KAOHSIUNG MEDICA UNIVERSITY from www.liebertpub.com at 10/19/18. For personal use only.