M I N I - R E V I E W

Contemporary Thyroid Nodule Evaluation and

Management

Giorgio Grani, Marialuisa Sponziello, Valeria Pecce, Valeria Ramundo, and

Cosimo Durante
Department of Translational and Precision Medicine, Sapienza University of Rome, 00161 Rome, Italy

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

ORCiD numbers: 0000-0002-0388-1283 (G. Grani); 0000-0003-0736-1047 (M. Sponziello);
0000-0002-5693-2988 (V. Pecce); 0000-0003-2709-533X (V. Ramundo); 0000-0002-1791-5915 (C. Durante).

Context:  Approximately 60% of adults harbor 1 or more thyroid nodules. The possibility of
cancer is the overriding concern, but only about 5% prove to be malignant. The widespread
use of diagnostic imaging and improved access to health care favor the discovery of small,
subclinical nodules and small papillary cancers. Overdiagnosis and overtreatment is associated
with potentially excessive costs and nonnegligible morbidity for patients.

Evidence Acquisition:  We conducted a PubMed search for the recent English-language articles
dealing with thyroid nodule management.

Evidence Synthesis:  The initial assessment includes an evaluation of clinical risk factors and
sonographic examination of the neck. Sonographic risk-stratification systems (e.g., Thyroid
Imaging Reporting and Data Systems) can be used to estimate the risk of malignancy and the
need for biopsy based on nodule features and size. When cytology findings are indeterminate,
molecular analysis of the aspirate may obviate the need for diagnostic surgery. Many nodules
will not require biopsy. These nodules and those that are cytologically benign can be managed
with long-term follow-up alone. If malignancy is suspected, options include surgery (increasingly
less extensive), active surveillance or, in selected cases, minimally invasive techniques.

Conclusion:  Thyroid nodule evaluation is no longer a 1-size-fits-all proposition. For most

nodules, the likelihood of malignancy can be confidently estimated without resorting to
cytology or molecular testing, and low-frequency surveillance is sufficient for most patients.
When there are multiple options for diagnosis and/or treatment, they should be discussed
with patients as frankly as possible to identify an approach that best meets their needs. (J Clin
Endocrinol Metab 105: 1–15, 2020)

Key Words:  biopsy, risk assessment, risk factors, TIRADS, ultrasonography, watchful waiting

T
he prevalence of thyroid nodules in the general Although epidemiological studies suggest a small but
population is high—up to 60% as documented by real increase in the incidence of thyroid cancer, likely
high-resolution ultrasonography—but very few of these stemming from exposure to environmental risk factors
lesions ultimately prove to be malignant (about 5%) (1). (2), the growing number of thyroid cancer diagnoses
is largely attributable to the increasingly widespread
ISSN Print 0021-972X  ISSN Online 1945-7197 use of diagnostic imaging technology and medical sur-
Printed in USA veillance, together with improved access to health care
© Endocrine Society 2020.
This is an Open Access article distributed under the terms of the Creative Commons
in general, all of which favor the discovery of small,
Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits un-
restricted reuse, distribution, and reproduction in any medium, provided the original
work is properly cited. Abbreviations: ACR, American College of Radiology; BSRTC, Bethesda System for Re-
Received 22 March 2020. Accepted 27 May 2020. porting Thyroid Cytopathology; FNAB, fine-needle aspiration biopsy; GSC, Genomic
First Published Online 3 June 2020. Sequencing Classifier; PTC, papillary thyroid cancer; TIRADS, Thyroid Imaging Reporting
Corrected and Typeset 16 July 2020. and Data System; US, ultrasonography.

doi:10.1210/clinem/dgaa322 J Clin Endocrinol Metab, September 2020, 105(9):1–15   https://academic.oup.com/jcem   1

2  Grani et al   Contemporary Thyroid Nodule Evaluation J Clin Endocrinol Metab, September 2020, 105(9):1–15

subclinical thyroid nodules and small papillary thyroid benign conditions; inherited syndromes associated with
cancers. These considerations have raised concern over thyroid cancer; a family history of thyroid cancer), but
the costs and potential morbidity associated with the even in these cases, the benefits of early detection have
short- and long-term management of patients with thy- yet to be demonstrated (6). So, the implementation of
roid nodules, which includes periodic outpatient visits sonographic screening programs for thyroid nodules is
and cervical ultrasound examinations, fine-needle aspir- discouraged.
ation biopsy (FNAB), genomic testing, and, in some in-
determinate cases, diagnostic thyroid lobectomy. On the Initial assessment
whole, there is a clearly perceived need for a more re- The initial assessment of a clinically evident or in-
fined, tailored, and careful approach to the management cidentally discovered thyroid nodule includes cervical
of these highly prevalent lesions. Similar considerations sonography and evaluation of clinical risk factors.

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

can be extended to that amount of nodules that are Depending on the results that emerge, the use of other
proven to be malignant, but have a low-risk phenotype, diagnostic tools, such as FNAB for cytology and mo-
and can be safely managed through more conservative lecular testing, will be indicated for a small subset of
surgery or even active surveillance programs (3, 4). The the lesions.
aim of this review is to provide an overview of currently
recommended practices for the initial workup and sub- Thyroid ultrasound and sonographic risk-
sequent management of patients with thyroid nodules. stratification systems
Thyroid ultrasonography (US) is the first-line tool
Search strategy for thyroid imaging. The thyroid gland is superficial in
We conducted a search of PubMed using the fol- the neck, with the posterior border usually located less
lowing terms: “thyroid nodule”[tiab] AND (“disease than 4 cm below the skin surface. High-resolution linear
management”[MeSH Terms] OR (“disease”[All probes provide excellent image definition of the gland.
Fields] AND “management”[All Fields]) OR “disease The examination is safe and painless, requires no prep-
management”[All Fields]) AND ((“2017/01/01”[PDAT]: aration, and can be performed rapidly in different care
“2020/12/31”[PDAT]) AND English[lang]). From the settings. To characterize thyroid nodules and obtain an
215 records returned, we selected the articles that were initial estimate of their risk for malignancy, the exam-
most relevant, with a preference for more recent pub- iner should focus on the echogenicity of the nodule; its
lications. We scanned the reference lists of the papers composition (solid, cystic, mixed), shape, and margins;
identified to find other relevant articles. the presence within the nodule of calcifications or other
hyperechoic foci; and the characteristics of all cervical
Primum non nocere lymph nodes (7). Findings consistently associated with ma-
Clinicians encountering patients with thyroid nodules lignancy include hypoechogenicity; infiltrative, irregular,
today are faced with the task of avoiding the overdiagnosis or lobulated margins; intranodular microcalcifications;
of low-risk cancers without jeopardizing the chances of and a taller-than-wide shape. In addition to the nodule
identifying those rare advanced or higher risk tumors that itself, all US studies must include a thorough exploration
will require prompt treatment. Overdiagnosis implies the of all cervical lymph node compartments, and the pres-
discovery of conditions that will never cause morbidity ence of any suspicious lymph nodes should be noted (8).
or death. As such, their identification can trigger a cas- The diagnostic sensitivity and specificity of these features
cade of deleterious events: people are transformed into vary, and no single feature has proved capable of reliably
patients, with inevitable emotional consequences and po- distinguishing malignant lesions from those that are be-
tential exposure to risks related to overmedicalization and nign (9). In addition, recognition and reporting of these
overtreatment. At the public health level, overdiagnosis features are characterized by substantial interobserver
overstretches the capacities of health systems, increases and even intraobserver variability (10).
spending, and subtracts resources from patients with the To address these shortcomings, several national
greatest health care needs. It is little wonder that the U.S. and international professional organizations have de-
Preventive Services Task Force now recommends against veloped US-based risk-stratification systems (often
thyroid cancer screening in asymptomatic adults, because referred to as Thyroid Imaging Reporting and Data
its harmful effects outweigh its potential benefits (5). The System or TIRADS, terms derived from those used for
recommendation against screening does not apply to pa- breast cancer imaging) that assign thyroid nodules to
tients with known risk factors for thyroid cancer (e.g., categories characterized by increasing risks (or risk
childhood radiation exposure in the form of radioactive ranges) for cancer, based on the presence or absence of
fallout or radiotherapy, including low-dose forms for the above-mentioned nodule features (Table 1) (11-15).
Table 1.  An Overview of the Standardized Thyroid Nodule US Risk Stratification Systems Proposed or Endorsed by National or International
Practice Guidelinesa 
Risk Score AACE/AME/ACE (12) ATA (13) EU-TIRADS (14) K-TIRADS (15)
Suspicious US ▪  Marked hypoechogenicity ▪  Irregular margins ▪  Non-oval shape ▪  Microcalcification
features ▪  Spiculated or lobulated margins (infiltrative, ▪  Irregular margins ▪  Taller-than-wide shape
▪  Microcalcifications microlobulated) ▪  Microcalcifications ▪  Spiculated/microlobulated margins
▪  Taller-than-wide shape ▪  Microcalcifications ▪  Marked hypoechogenicity
▪  Extrathyroidal growth ▪  Taller-than-wide shape
doi:10.1210/clinem/dgaa322

▪  Pathologic adenopathy ▪  Rim calcifications

with small extrusive
soft-tissue component
▪  Evidence of
extrathyroidal
extension
Category Low-risk: Benign: Benign (EU-TIRADS 2): Benign:
Risk of malignancy: 1% Risk of malignancy: <1% Risk of malignancy: ≈ 0% Risk of malignancy: <1-3
FNAB > 20 mm FNAB is not indicated FNAB is not indicated FNAB ≥ 20 mm
(selective)b Purely cystic nodules pure/anechoic cysts; Spongiform
Cysts (fluid component (no solid component) entirely spongiform nodules Partially cystic nodule with comet-tail
>80%). artifact
Mostly cystic Pure cyst
nodules with Very low suspicion: Low-risk (EU-TIRADS 3): Low suspicion:
reverberating artifacts Risk of malignancy: Risk of malignancy: 2%- Risk of malignancy: 3%-15%FNAB
and not associated with < 3%FNAB ≥ 20 mm or 4%FNAB > 20 mm ≥15 mm
suspicious US signs. observation Oval shape, smooth margins, Partially cystic or isohyperechoic
Isoechoic spongiform Spongiform or partially cystic isoechoic or hyperechoic, without nodule without any of 3 suspicious
nodules, either nodules without any of the any feature of high risk US features
confluent or with US features defining low-,
regular halo. intermediate-, or high-suspicion
patterns
Low suspicion:
Risk of malignancy: 5%-10%FNAB ≥
15 mm
Isoechoic or hyperechoic solid
nodule, or partially cystic nodule
with eccentric solid area without:
microcalcifications, irregular margin,
extrathyroidal extension, taller
than wide shape
https://academic.oup.com/jcem  3

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

Table 1.  Continued

Risk Score AACE/AME/ACE (12) ATA (13) EU-TIRADS (14) K-TIRADS (15)
Intermediate-risk: Intermediate suspicion: Intermediate-Risk Intermediate suspicion:
Risk of malignancy: 5–15% Risk of malignancy: 10–20%FNAB (EU-TIRADS 4): Risk of malignancy: 15%-
FNAB >20 mm ≥10 mm Risk of malignancy: 6%-17% 50%
Slightly hypoechoic (vs. thyroid tissue) Hypoechoic solid nodule with FNAB >15 mm FNAB ≥10 mm
or isoechoic nodules, with ovoid-to- smooth margins without: Oval shape, smooth margins, Solid hypoechoic nodule without any s
round shape, smooth or ill-defined microcalcifications, extrathyroidal mildly hypoechoic, without uspicious US
margins extension, or any feature of high risk feature or partially cystic or
May be present: taller-than-wide shape isohyperechoic
Intranodular vascularization nodule with any of the following:
Elevated stiffness at elastography, microcalcification, nonparallel
Macro or continuous rim calcifications orientation (taller-than-
Indeterminate wide), spiculated/
hyperechoic microlobulated margin
spots
High-risk: High suspicion: High-risk (EU-TIRADS 5): High suspicion:
Risk of malignancy: 50%-90%c Risk of malignancy: Risk of malignancy: 26%-87% Risk of malignancy: > 60
4  Grani et al   Contemporary Thyroid Nodule Evaluation

FNAB ≥10 mm (5 mm, selective)d >70%-90% FNAB > 10 mm FNAB ≥ 10 mm (>5 mm selective)
Nodules with ≥1 of the following: FNAB ≥10 mm Nodules with ≥ 1 of the following: Solid hypoechoic nodule with any of
Marked hypoechogenicity Solid hypoechoic nodule or solid Non-oval shape the following:
(vs. prethyroid muscles) hypoechoic component Irregular margins Microcalcification
Spiculated or lobulated margins of partially cystic nodule Microcalcifications Nonparallel orientation (taller-than-
Microcalcifications with ≥1 of the following: Marked hypoechogenicity wide)
Taller-than-wide shape Irregular margins (infiltrative, Spiculated/
(AP > TR) microlobulated) microlobulated margin
Extrathyroidal growth Microcalcifications
Pathologic adenopathy Taller-than-wide shape
Rim calcifications with small
extrusive soft tissue
Extrathyroidal extension
From Tumino D, Grani G, Di Stefano M, et al. Nodular thyroid disease in the era of precision medicine. Front Endocrinol (Lausanne). 2020;10:907.
Abbreviations: AACE/ACE/AME, American Association of Clinical Endocrinologists/American College of Endocrinology/Associazione Medici Endocrinologi; ATA, American Thyroid Association; EU-TIRADS,
European Thyroid Association (ETA) Thyroid Imaging Reporting and Data System; K-TIRADS, Korean Society of Thyroid Radiology Thyroid Imaging Reporting and Data System.
a
The TIRADS developed and endorsed by the American College of Radiology (ACR) is also widely used. Unlike the systems shown in the table, in which risk is defined by the association of 2 or more
nodule features, the ACR system individually assesses 5 key aspects of the nodule (composition, echogenicity, shape, margins, and echogenic foci) and expresses the result in terms of a numerical score.
The nodule is then assigned to a risk class based on the sum of the 5 scores.
b
Growing nodule, high-risk history, before surgery or local therapies.
c
In accordance with the presence of 1 or more suspicious findings.
d
FNAB is recommended for subcapsular or paratracheal nodules and those associated with suspicious clinical findings (e.g., dysphonia); suspicious lymph nodes or extrathyroidal spread; a positive personal
or family history of thyroid cancer; or a personal history of head and neck irradiation.
J Clin Endocrinol Metab, September 2020, 105(9):1–15

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

doi:10.1210/clinem/dgaa322 https://academic.oup.com/jcem  5

Within each risk class, size cutoffs are used to identify effectiveness of these systems in the real-world clinical
lesions whose FNAB can safely be deferred. Avoiding un- practice also depends on the adoption of uniform ter-
necessary FNABs is an important goal. Aside from cost minology and accurate, nonambiguous definitions of the
considerations, these procedures can be associated with features being assessed. For these reasons, the European
complications, albeit minor and transient (e.g., mild Thyroid Association (14), the Korean Society of Thyroid
bruising, soreness, swelling, neck discomfort) (16). More Radiology (15), and the ACR (32) have developed spe-
important, inconclusive cytology results are by no means cific lexicons to be used with their risk-assessment
rare, and they often lead to additional testing (frequently systems. The definitions of certain critical nodule descrip-
quite expensive) and/or diagnostic surgery, undertaken tors, such as echogenicity (33), shape (34), hyperechoic
for the purpose of confirming that the nodule is indeed foci (35), and extrathyroidal extension (36), can signifi-
benign. The accuracy of the risk estimates generated by cantly impact the diagnostic performance of thyroid

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

the systems shown in Table  1 has in some cases been US. As a result, interobserver agreement for US-based
validated in retrospective (17-20) and/or prospective risk-stratification systems remains only fair to moderate
studies (21-23), and their performance has also been (kappa, 0.34-0.44) (10). Specific training involving joint
confirmed by a recent meta-analysis (24). These classi- evaluation of images can be useful to increase operators’
fications also guide the timing of subsequent long-term ability to recognize these features, thereby improving
follow-up evaluations and the eligibility of suspicious the reproducibility for all classifications, even among
nodules for management limited to active surveillance. trained clinicians with similar levels of experience (37,
As shown in Table 1, the nodule aspects considered 38). To address this problem, the International Thyroid
in in the risk-estimation process are fundamentally the Nodule Ultrasound Working Group, a multidisciplinary
same with all these systems—structural composition, alliance of physicians with expertise in thyroid nodule
echogenicity, shape, margins, and echogenic foci—and sonography, is attempting to devise a unified, inter-
risk classes are defined by sets or clusters of 2 or more national set of guidelines that is based on a validated
nodule features (11). The exception to the latter rule lexicon and incorporates state-of-the-art techniques and
is the American College of Radiology (ACR) TIRADS. research data (39). Standard B-mode US examinations
With this system, the same key aspects of the nodule can also be expanded to include elastographic (strain or
are assessed individually. Each is rated with a numerical shear-wave techniques) analysis (40) of the stiffness of
score, and the sum of the 5 scores determines the risk the nodular tissue and contrast-enhanced assessments of
class to which the nodule is assigned. its perfusion and vascularity (41). Both approaches have
There is some degree of heterogeneity across sys- produced promising results and their use for US-based
tems in terms of the definitions of certain nodule risk-stratification has been proposed (40, 41). Thus far,
features, the relative weight assigned to individual however, adoption of these proposals has been limited
features, and the size criteria used for FNAB re- owing to availability and reproducibility issues. Other
commendations. Substantial variability has also novel approaches include the use of software applica-
been observed in the different systems’ ability to tions capable of performing automated analysis for ex-
decrease the number of unnecessary FNABs. In 1 tracting quantitative parameters from US images. These
recent prospective analysis (21), the ACR TIRADS tools may be the basis for computer-aided diagnosis sys-
outperformed the 4 other widely used systems tested tems that can generate an automated “second opinion.”
in reducing the number of biopsies performed on Some findings suggest that machine learning approaches
nodules ultimately diagnosed as benign: more than are as accurate (42) or even more accurate (43) than
one-half of the biopsies would have been classified by expert radiologists in discriminating between malignant
the ACR system as deferrable, with a false-negative and benign thyroid nodules.
rate of only 2.2%. Although developed mainly for
detection of papillary thyroid cancers (PTCs), the Clinical risk factors
sonographic risk-stratification systems also seem to The prevalence of thyroid nodules increases with age,
provide reliable recommendations for FNAB of fol- and most are detected in individuals older than 40 years
licular thyroid cancers (25), medullary thyroid can- of age. In addition to the sonographic appearance of
cers (26), and metastases to the thyroid gland (27). the nodule, other factors have to be considered when
Most physicians in the United States (28), Spain (29), deciding whether or not FNAB should be performed.
and Italy (30) report using TIRADS or similar classi- Some are thought to be predictive of nodule develop-
fications when performing thyroid ultrasonography. ment or malignancy (e.g., serum levels of TSH, autoanti-
However, most US reports in routine practice provide bodies, obesity), but the evidence for these associations
insufficient data for risk stratification (31). The optimal is currently inconclusive. TSH should nonetheless be
6  Grani et al   Contemporary Thyroid Nodule Evaluation J Clin Endocrinol Metab, September 2020, 105(9):1–15

measured in all patients to rule out the possibility of a neoplasm or suspicious for a follicular [or Hürthle cell]
hyperfunctioning nodule. The latter lesions do not re- neoplasm”). For most papillary thyroid cancers, as well
quire biopsy because they are virtually always benign. as medullary, poorly differentiated, and undifferentiated
Recognized risk factors for thyroid malignancy are carcinomas, the cytology report will usually be unam-
medical irradiation during childhood (44), accidental biguously diagnostic (Bethesda class  VI, malignant),
exposure to ionizing radiation from fallout in child- whereas some degree of uncertainty persists for nodules
hood or adolescence (45, 46), a family history of thyroid assigned to Bethesda class V (suspicious for malignancy)
cancer, or hereditary syndromes that include a predispos- nodules, which is associated with a very broad range of
ition to thyroid cancer (e.g., PTEN hamartoma tumor malignancy risks (53).
syndrome, Carney complex, Werner syndrome) (13).
Nodules that are firm, fixed, or rapidly growing require What to do with indeterminate results?

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

prompt evaluation (47). Recently, the intraglandular lo- As noted, the term “indeterminate cytology” re-
cation of the nodule has also been confirmed to be an fers to Bethesda class  III or class  IV findings, which
independent risk factor for malignancy. Nodules arising are associated with expected malignancy rates of 10%
in the isthmus are the most likely to be diagnosed as to 30% and 25% to 40%, respectively. The options
cancer, whereas those found in the lower third of a lobe suggested for identifying these nodules includes re-
carry the lowest risk (48), as compared with those in the peat FNAB for cytology and/or molecular testing and
middle (49) or upper pole of the lobe (50). diagnostic lobectomy. Some data suggest that a repeat
These factors are typically not considered in risk US risk stratification can be useful in predicting ma-
stratification algorithms, but may influence the course lignancy and in planning further steps for the man-
of action in shared decision-making with patients (39). agement of indeterminate nodules (55-58), or at least
those in Bethesda class III (59, 60). However, if nodules
Cytology and molecular testing are properly selected beforehand, and the pretest risk
Fine-needle aspiration cytology is the next step in of malignancy is high, the utility of this approach may
the triage of a thyroid nodule. It should be reserved for be reduced (61, 62).
lesions found to be sufficiently suspicious on the basis Cytological assessment of a second fine needle as-
of US and clinical findings. The results play a key role pirate is commonly used, but it provides a definitive
in optimizing subsequent management. The Bethesda diagnosis for only 40% of class I (nondiagnostic) and III
System for Reporting Thyroid Cytopathology (BSRTC) nodules (63). If the second cytological study is still in-
was discussed in 2007 by a panel of experts at the U.S. determinate, diagnostic surgery (usually lobectomy) has
National Institutes of Health in Bethesda, MD. The first traditionally been the only route to a definitive patho-
edition of the system was published in 2010, and an logical diagnosis. It is obviously expensive and associ-
updated version followed in 2018 (51). The BSRTC is ated with some risks. And if the nodule proves to be
widely used in the United States, and it has served as malignant, reoperation (completion thyroidectomy) is
a model for similar tiered classification schemes devel- often indicated, with added risks and costs. Up to 60%
oped more recently in other parts of the world (52). of patients undergoing lobectomy for an indeterminate
The robust diagnostic framework provided by the nodule are likely to be over- or undertreated at initial
BSRTC offers valuable guidance in developing man- surgery (64).
agement strategies for patients with thyroid nodules Molecular testing of the FNAB samples is a newer
(53). Nonetheless, several potential diagnostic pitfalls approach that can reduce the need for diagnostic sur-
exist that can lead to false-positive, false-negative, gery. The tests developed for this purpose over the past
nondiagnostic, or indeterminate results (54). Cytology 10  years are based on 3 main molecular approaches:
itself has limitations: it cannot, for example, distinguish testing for somatic mutations, gene expression evalu-
between follicular-patterned hyperplastic/adenomatoid ation, and microRNA (miRNA)-based classifiers
nodules, follicular adenomas, follicular carcinoma, and (65-71). The current version of the ThyroSeq test (ver-
some cases of follicular variants of papillary thyroid sion 3)  involves targeted next-generation sequencing
carcinoma. Thyroid cytology can be considered only analysis of 112 cancer-related genes for point mutations,
a screening test for these follicular-patterned lesions, gene fusions, copy number alterations, or abnormal gene
the results of which will almost invariably reported as expression. When validated on 257 cytologically inde-
“indeterminate,” that is, assigned to Bethesda class  III terminate nodules with surgical pathology reports, it
(“atypia of undetermined significance” or “follicular le- displayed good sensitivity (94%) and specificity (82%),
sions of undetermined significance”) or IV (“follicular with a negative predictive value of 97% and a positive
doi:10.1210/clinem/dgaa322 https://academic.oup.com/jcem  7

predictive value of 66%. The authors concluded the test sequencing mutation analysis (ThyGeNEXT) (positive
might eliminate the need for diagnostic surgery in up to predictive value 74%, negative predictive value 94%
61% of patients with indeterminate nodules (66). The (68)), but the assay has yet to be subjected to inde-
Afirma test was originally based on microarray ana- pendent validation. Molecular tests require a dedicated
lysis of mRNA expression profiles. The current version, needle pass, the collection of residual material in the
the Afirma Genomic Sequencing Classifier (GSC), in- needle hub, liquid cytology remnants, or the recovery
cludes 12 classifiers composed of 10,196 genes (RNA of cells from routinely prepared cytology slides. In some
sequencing approach). Compared with the previous cases, a repeat dedicated biopsy is needed; in others (i.e.,
version, the new test correctly classifies more indeter- ThyroSeq, ThyGeNEXT/ThyraMIR, and ThyroPrint)
minate nodules as benign and displays improved spe- can also be performed using the original cytology slide.
cificity (68.3%) and positive predictive value (47.1%), Meaningful comparison of these tests in terms of

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

with a sensitivity of 91.1% and a negative predictive their diagnostic performance is extremely difficult for
value of 96.1% (67). It is expected to further reduce several reasons. The currently available data come from
the frequency of diagnostic surgery, based on the results studies that differ significantly from one another in co-
of independent studies (72-74). In short, the ThyroSeq hort selection criteria, sample sizes, malignancy rates,
and Afirma assays currently have positive and nega- study designs, and applied reference standards (e.g.,
tive predictive values that make them suitable for use not all lesions are surgically resected, the issue of inter-
in both rule-in and rule-out testing. ThyroPrint, a gene- pathologist variability is not always addressed). In add-
expression classifier based on interrogation of only 10 ition, no attempts have been made to conduct direct
genes has also displayed good performance (sensitivity head-to-head comparisons, using more than 1 test on
96%, specificity 87%, and positive and negative pre- the same samples. Furthermore, the high costs of these
dictive values of 78% and 98%, respectively) in an in- tests limit their use in many countries. However, when
ternal, multicenter validation study (70). hypothetical modeling was used to compare surgery
A well-known limitation of mutation-based ap- versus molecular testing for the management of inde-
proaches is related to the occurrence of RAS mutations terminate nodules, both of the major molecular ap-
in a wide variety of thyroid tumors, including follicular proaches discussed previously proved to be considerably
adenoma, noninvasive follicular thyroid neoplasm more cost-effective than diagnostic lobectomy, and the
with papillary-like nuclear features, encapsulated and Thyroseq v. 3 was more cost-effective than the Afirma
unencapsulated follicular-variant PTC, classic PTC, GSC (80). If, on the basis of all clinical, imaging, and
medullary thyroid cancer, poorly differentiated thyroid cytologic findings, the sole aim of surgery is diagnostic,
cancer, and anaplastic thyroid cancer. When clonally molecular testing should definitely be considered.
present, mutant RAS is an oncogene, and nodules har-
boring these mutations should be considered neoplastic. Follow-up examinations: what to look for
However, recent findings show that the presence of a Fig.  1 shows the simulated management strat-
RAS mutation alone is not a helpful marker of malig- egies and outcomes of 1000 newly discovered thyroid
nancy. The few cancers with this mutation prove to be nodules (21, 72-74, 81-83). The overall management
low-risk tumors with fairly indolent behavior (75, 76). pathway is based on the US-risk stratification of the
Other available approaches are based on expression target lesions and the cytology assessment (if any).
levels of miRNAs, small, highly conserved, noncoding These scenarios do not include symptomatic thyroid
RNA molecules capable of influencing the expression nodules that are already candidates for resection re-
of messenger RNAs and impacting multiple pathways gardless of their sonographic features. In these cases,
(77). Differential miRNA expression has been described a biopsy might be performed to clarify the best sur-
in distinct thyroid cancer subtypes and is also linked gical approach, but the results would not change the
to the differentiation or progression status of these tu- indication for surgery itself (84). The distributions of
mors (78). MiRNAs have also been proposed as circu- US-defined risk classes, US-defined FNAB indications,
lating biomarkers of thyroid cancer in peripheral blood and Bethesda cytology class were derived from pub-
(79). As a result, miRNA gene expression classification lished findings (21). For illustration purposes, all inde-
has been proposed as a complementary molecular test terminate nodules are shown as undergoing molecular
that can further improve predictive values and refine testing, although other options are offered. However, if
surgical management (69, 71). Promising results have these alternative approaches, guided by clinical and US
been reported for a combination assay that includes data, had been used, it is unlikely that the final number
miRNA classification (ThyraMIR) and next-generation of resected nodules and their malignancy rate would be
8  Grani et al   Contemporary Thyroid Nodule Evaluation J Clin Endocrinol Metab, September 2020, 105(9):1–15

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

Figure 1.  Alluvial flow diagram showing simulated management and outcomes for 1000 newly discovered thyroid nodules. The distributions of
ultrasound (US)-defined risk classes, US-defined fine-needle aspiration biopsy (FNAB) indications, and Bethesda cytology class were derived from
published findings (21). US risk-stratification is that recommended by the American Thyroid Association (ATA) Guidelines. Nodules not classifiable
with the ATA system are included in the intermediate-suspicion category. Nondiagnostic nodules with very-low-suspicion or low-suspicion US
findings can be managed with US surveillance, but repeat FNAB is indicated for those with intermediate- or high-suspicion US findings (81):
in this diagram, all are shown as undergoing repeat biopsy. Bethesda II nodules require repeat biopsy only if the US-based risk class increases
during surveillance (frequency: ~15% over 5 years of follow-up) (82). The false-negative rate is less than 3% (e.g., sampling error; for high-
suspicion nodules with Bethesda II cytology, repeat biopsy is suggested) (83). For illustration purposes, all indeterminate nodules are shown as
undergoing molecular testing (regardless of other possible options). The hypothetical molecular testing approach depicted has a benign call rate
of 65%, a positive predictive value of 50%, and a negative predictive value of 96% (72-74). For high-suspicion nodules classified as benign by
molecular testing, repeat biopsy is indicated. All Bethesda V and VI nodules are referred for surgery. Expected malignancy rates are 80% and 99%,
respectively.

significantly different from those shown in Fig. 1. The negatives (usually the result of sampling errors) are un-
hypothetical molecular testing approach depicted has a common (less than 3%), but repeat biopsy should be
benign call rate of 65%, a positive predictive value of considered for high-suspicion nodules with Bethesda II
50%, and a negative predictive value of 96% (72-74). cytology (83). Most of these nodules (≈85%) will re-
After the initial diagnostic workup described, very few main asymptomatic with no signs of growth and will
of the nodules (10.8%) will be subjected to surgery, but therefore not require any treatment. This estimate is
a high percentage of those that are resected will prove based on findings from a 5-year prospective study of
be malignant (73.1% of the resected nodules; 7.9% of 1567 benign thyroid nodules (85), which have subse-
all nodules). Some nodules (14.5%) will require a repeat quently been confirmed by retrospective studies (86).
biopsy, immediately or during the long-term follow-up, The aim of long-term surveillance should be to detect
to be classified. For Bethesda II nodules, a second cyto- any previously missed malignancy and monitor thyroid
logical assessment is indicated only if the US-based risk nodule growth. The sonographically estimated malig-
class increases during surveillance (frequency: ~15% nancy risk also provides guidance in planning effective
over 5  years of follow-up) (82). Three-quarters of all follow-up (Fig.  2). The algorithm shown in the figure
nodules (74.7%) will be classified as benign and man- is based on literature data, practice guidelines, and the
aged with long-term sonographic surveillance. False authors’ own experience. Thyroid FNAB cytology has a
doi:10.1210/clinem/dgaa322 https://academic.oup.com/jcem  9

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

Figure 2.  Suggested management and follow-up of nodules with no indication for immediate biopsy and those cytologically classified as benign.

very low false-negative rate (<3%); similar findings have nodules (relative risk, 2.5; 95% confidence interval,
been reported for molecular analysis of FNAB fluid 1.6–3.1; P  <  0.001) and therefore warrant repeat bi-
(Table 2). However, nodules with highly suspicious fea- opsy (89). Growth is also a concern because it can cause
tures on the initial US examination—even if they have compressive symptoms. Size increases are more likely in
been cytologically or molecularly classified as benign— younger individuals and patients with multiple or large
warrant a repeat biopsy within 12  months. Repeat nodules (85).
biopsy is also indicated if new suspicious features or sig- On the whole, the recent findings and increasing
nificant growth is observed during follow-up. In a sub- use of comprehensive US-based risk stratification sys-
group analysis of a prospective cohort, the US-estimated tems should reduce the “diagnostic burden” related to
malignancy risk class of 13.2% to 29.4% of the nodules nodules classified as benign: they can be safely followed
increased during the first 5 years of follow-up. However, with visits every 2 or 3 years, and the frequency can be
the risk-class increase was associated with a recommen- even further reduced if no changes have been noted at
dation for FNAB for only 6% to 8% of the nodules previous visits.
whose biopsies had been deferred at baseline (82), and Patients whose nodules are cytologically classified as
none of the increased risk estimates was associated with malignant or suspicious for malignancy (Bethesda classes
discovery of a new malignancy. The development of new V and VI; or suspicious molecular analysis) should gen-
nodules is quite common during surveillance, but only erally be referred for surgery. However, for those with
3% to 7% of these lesions meet the criteria for biopsy. subcentimeter, intrathyroidal cancers with no high-risk
Some authors maintain that a sonographically docu- features, active US surveillance can be proposed in lieu of
mented change in the appearance of a nodule classified immediate surgery (3, 4). The first study published on this
as benign (particularly its margins) is the most reliable issue in 2003 found that more than 70% of microPTCs
indication for repeat FNAB (87). The growth alone remained stable during surveillance, and the frequency
of a such nodules emerged from a meta-analysis as a of spread to the locoregional lymph nodes was quite low
relatively poor predictor of malignancy (88). More re- (~1%) (90). The active surveillance strategy does not
cently, however, a prospective study found that nodules merely delay surgery: the likelihood of disease progres-
displaying significant growth during follow-up (diam- sion diminishes with age, and older patients are therefore
eter increases exceeding 2  mm per year) are signifi- less likely to require surgery during their lifespan (91).
cantly more likely to be malignant than slower growing Similar results have been reported worldwide (92, 93).
Table 2.  Minimally Invasive Techniques for Treatment of Symptomatic, Benign Thyroid Nodules
Method Mechanisms Candidate Nodules Expected Volume Reduction Adverse Effects Cost Estimate
Percutaneous ethanol Dehydration of Cystic or predominantly ~60% • Pain Simplest, least expensive
injection (PEI) cytoplasmic proteins, cystic nodules • Burning sensation option ($50-$100)
coagulation necrosis, • Hematoma
and • Dyspnea
fibrosis • Voice change
Radiofrequency ablation Thermal coagulation Mixed or solid nodules 47.7%-96.9% Overall complication rate 2.11% Equipment $25,000;
(RFA)a necrosis (major complications 1.27%) $800 per session
(101, 103, 104) Major:
• Voice change
• Nodule rupture
• Hypothyroidism
•  Brachial plexus injury
Minor:
• Pain
• Thermal propagation outside
10  Grani et al   Contemporary Thyroid Nodule Evaluation

of the thyroid
• Fever
• Skin burns
• Hematoma
• Transient hyperthyroidism/
transient thyroiditis
Laser ablation (LA)a Thermal coagulation Mixed or solid, 62 ± 22%b • Pain (10%) Equipment with built-in
(101, 102) necrosis functional, or • Fever (8%) laser source: $12,000
nonfunctional nodules • Vasovagal reaction (1%) Nd:YAG laser source:
• Voice change (0.5%) $15,000-$20,000
• Hematoma (0.4%) $500 per session
• Skin burn (0.1%)
Microwave ablation Thermal coagulation Mixed or solid nodules 50%-70 % depending on • Pain (25%) Equipment: $35,000
(MWA) (105-107) necrosis nodule composition • Transient voice change (1%) $400 per session
(solid require more energy • Hematomas
than cystic nodules) • Burns (2/30)
• Horner syndrome (1/30)
High-intensity focused Thermal coagulation Mixed or solid nodules 49%-69% • Hypothyroidism (1.4%-2.3%) Focused thermal tissue
ultrasound (HIFU) necrosis • Hoarseness destruction without
(98-100) • Neck swelling needles;
Equipment: $400,000
$350 per session
a
Laser fibers deliver energy to the target more accurately than radiofrequency electrodes. The efficacies of the 2 techniques are potentially similar in the hands of operators with the same levels of skill
and experience. RFA appears to be superior for benign solid nodules (108); LA seems slightly more effective for nodules > 30 mL (101).
b
The rate of decrease depends on nodule type, vascularity, energy used, operator experience (109).
J Clin Endocrinol Metab, September 2020, 105(9):1–15

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

doi:10.1210/clinem/dgaa322 https://academic.oup.com/jcem  11

Active surveillance protocols initially provide for the benign nature of the lesion has been confirmed with
US neck examinations every 6 months. Once disease 2 serial FNABs and serum calcitonin assessment. For
stability has been documented—in general, with nodules with lower risk features on US or autonomously
2 years of serial US examinations showing no evidence functioning lesions, a single aspirate with benign cy-
of progression—the examination frequency can be re- tology is sufficient (110, 111). The clinical and US-based
duced to every 1 to 2 years or more (94). In contrast, follow-up of benign nodules that undergo treatment re-
if surveillance does reveal evidence of progression quire expert clinical and US evaluation, because the mor-
(i.e., an increase ≥3 mm in the maximum diameter of phologic features may change over time. If regrowth
the nodule, growth toward the dorsal surface of the occurs, a new cytological assessment is indicated.
gland or toward nearby structures, or the appearance When surgery is indicated, decisions on the extent
of lymph node metastases) (95, 96), surgery is indi- of resection will depend on multiple factors, including

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

cated. These data provide a solid background for re- symptoms, the presence of contralateral nodules, thyroid
commendations to avoid immediate biopsy and adopt functional status, comorbidities, family history, surgical
an US-based surveillance strategy for subcentimeter, risk, and patient preferences (84). Common reasons for
intrathyroidal nodules even if sonographically suspi- surgery are large goiters, local compressive symptoms
cious (13). or progressive nodule or thyroid enlargement, or large
toxic multinodular goiters. In most patients with mul-
tiple nodules, both lobes of the gland are involved and
Management total thyroidectomy is necessary. Consensus is lacking
on the procedure of choice for patients with an asym-
Benign nodules metric nodular goiter. In some cases, lobectomy can be
Benign thyroid nodules requiring treatment are rare. considered as a safer alternative to total thyroidectomy.
The most common are hyperfunctioning nodules and However, it requires long-term follow-up, is associ-
nodules whose growth is associated with compres- ated with nodule recurrence risk (114), and may subse-
sion of vital structures like the trachea or esophagus, quently require a second operation (115).
general neck discomfort, and/or cosmetic problems—all
of which can negatively impact quality of life. Surgery Indeterminate and suspicious nodules
is an option in these cases, but there are also several For cytologically indeterminate nodules that cannot
nonsurgical, minimally invasive alternatives. These in- be molecularly classified as benign, lobectomy with
clude US-guided ablation procedures involving percu- isthmusectomy is generally the procedure of choice.
taneous ethanol injection (the traditional method and However, thyroidectomy may be indicated in patients
currently the least expensive) or the application of heat with larger indeterminate nodules (≥3-4  cm), nodules
in the form of laser, radiofrequency, high-intensity fo- displaying progressive growth and/or worrisome fea-
cused US, or microwave energy. Radiofrequency and tures on ultrasound, or a family history of thyroid
laser ablations can significantly reduce nodule volumes cancer or radiation exposure (84). If preoperative mo-
(97). As shown in Table  2 (98-109), these techniques lecular testing is not possible, seeking the opinion of
differ in terms of their indications, adverse effects, and a second pathologist may be worthwhile because thy-
associated costs. Hyperfunctioning nodules can also be roid cytology is characterized by substantial inter- and
treated with radioiodine. intraobserver variability, especially for nodules classi-
High-intensity focused US is a newer needle-free tech- fied as indeterminate (116, 117).
nique that is producing promising results (98-100), but Lobectomy and isthmusectomy (or rarely an
it requires further clinical validation. More evidence and isthmusectomy alone) is usually the least extensive pro-
experience are also needed before microwave ablation is cedure that can be considered when malignancy is sus-
used on a large-scale basis. The use of these techniques pected (84). (Cases eligible for active surveillance, as
for the treatment for symptomatic benign nodules has discussed previously, are the obvious exception.) In pa-
been addressed by several groups of experts (110-113). tients with 1- to 4-cm suspicious nodules, lobectomy or
In general, consensus statements by these groups list total thyroidectomy are both acceptable, whereas patients
US-guided aspiration as the first-line treatment for cystic with large suspicious nodules, suspected extrathyroidal
or predominantly cystic nodules. Ethanol injection can extension, or suspected metastases (locoregional or dis-
be used for recurrences, and thermal ablation techniques tant) should undergo total thyroidectomy.
are reserved for cases in which symptoms persist after Lobectomy offers several advantages over total
ethanol. Thermal ablation can be used for nodules that thyroidectomy. It virtually eliminates the risks of per-
are predominantly solid and/or growing, but only after manent hypoparathyroidism and bilateral recurrent
12  Grani et al   Contemporary Thyroid Nodule Evaluation J Clin Endocrinol Metab, September 2020, 105(9):1–15

laryngeal nerve injury and substantially reduces the Additional Information

rates of permanent unilateral recurrent laryngeal nerve
palsy (0.6% versus 1.3%) (118). Furthermore, 50% to Correspondence and Reprint Requests: Cosimo Durante,
MD, PhD, Dipartimento di Medicina Traslazionale e
80% of the patients who undergo lobectomy do not re-
di Precisione, Università di Roma “Sapienza,” Viale del
quire thyroid hormone replacement therapy (the like-
Policlinico 155, 00161, Roma, Italy. E-mail: cosimo.durante@
lihood varies according to the preoperative TSH level uniroma1.it.
and the presence of thyroid autoimmunity.) (114, 115). Disclosure Summary: The authors have no conflicts of
Minimally invasive US-guided ablation techniques are interest to disclose.
also being proposed by some groups for nonsurgical Data Availability: Data sharing is not applicable to this
treatment of small suspicious nodules (119, 120). article as no datasets were generated or analyzed during the
current study.

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

Conclusions References
The evaluation and management of patients with thy- 1. Durante C, Grani G, Lamartina L, Filetti S, Mandel SJ, Cooper DS.
roid nodules is no longer a 1-size-fits-all proposition. The The Diagnosis and management of thyroid nodules: a review.
JAMA. 2018;319(9):914-924.
tailored approach advocated today requires a careful as-
2. Lim  H, Devesa  SS, Sosa  JA, Check  D, Kitahara  CM. Trends in
sessment of each nodule to determine the likelihood that thyroid cancer incidence and mortality in the United States, 1974-
it is malignant and the chances that it will cause symp- 2013. JAMA. 2017;317(13):1338-1348.
toms. Very few nodules will require an intensive workup 3. Ramundo V, Sponziello M, Falcone R, et al. Low-risk papillary
thyroid microcarcinoma: optimal management toward a more
that includes cytology and molecular testing of FNAB conservative approach. J Surg Oncol. 2020;121(6):958-963.
samples: for the vast majority, a systematic cervical US 4. Filetti S, Durante C, Hartl D, et al.; ESMO Guidelines Committee.
examination with assessment of clinical risk factors will Electronic address: clinicalguidelines@esmo.org. Thyroid cancer:
ESMO Clinical Practice Guidelines for diagnosis, treatment and
provide a reliable foundation for identifying the initial follow-up. Ann Oncol. 2019;30(12):1856-1883.
management strategy. After an appropriate initial as- 5. Bibbins-Domingo  K, Grossman  DC, Curry  SJ, et  al. Screening
sessment, the frequency of subsequent surveillance visits for thyroid cancer: US Preventive Services Task Force
Recommendation Statement. JAMA. 2017;317(18):1882-1887.
for most nodules can be safely reduced compared with
6. Lamartina L, Grani G, Durante C, Filetti S, Cooper DS. Screening
currently used schedules. This is particularly relevant for for differentiated thyroid cancer in selected populations. Lancet
frail, elderly individuals, as they are unlikely to be harmed Diabetes Endocrinol. 2020;8(1):81-88.
by the thyroid tumor itself, and overmedicalization can 7. AIUM practice guideline for the performance of a thyroid
and parathyroid ultrasound examination. J Ultrasound Med.
cause more harm than good. In these populations, sur- 2013;32(7):1319-1329.
veillance can safely be confined to a periodic clinical 8. Kumbhar SS, O’Malley RB, Robinson TJ, et al. Why thyroid sur-
examination. If surgery is needed, resections can often be geons are frustrated with radiologists: lessons learned from pre-
and postoperative US. Radiographics. 2016;36(7):2141-2153.
less extensive. In some cases, minimally invasive, percu- 9. Brito  JP, Gionfriddo  MR, Al  Nofal  A, et  al. The accuracy of
taneous approaches are a viable alternative to surgery. thyroid nodule ultrasound to predict thyroid cancer: sys-
When there are multiple options, they should be dis- tematic review and meta-analysis. J Clin Endocrinol Metab.
2014;99(4):1253-1263.
cussed as frankly as possible with the patient, outlining
10. Persichetti A, Di Stasio E, Coccaro C, et al. Inter- and intraobserver
the advantages, limitations, benefits, and risks of each. agreement in the assessment of thyroid nodule ultrasound fea-
The goal is to identify the best strategy for the individual tures and classification systems: a blinded multicenter study.
patient, in terms of disease outcomes and quality of life, Thyroid. 2020;30(2):237-242.
11. Tumino D, Grani G, Di Stefano M, et al. Nodular thyroid disease
avoiding the pitfalls of overdiagnosis and overtreatment. in the era of precision medicine. Front Endocrinol (Lausanne).
For health professionals, this means learning to work 2019;10:907.
with some degree of clinical uncertainty rather than auto- 12. Gharib  H, Papini  E, Garber  JR, et  al.; AACE/ACE/AME Task
Force on Thyroid Nodules. American Association of Clinical
matically resorting to intensive testing and intervention, Endocrinologists, American College of Endocrinology, and
and by weighing patients’ expectations and demands. Associazione Medici Endocrinologi Medical Guidelines for
Clinical Practice for the Diagnosis and Management of Thyroid
Nodules–2016 Update. Endocr Pract. 2016;22(5):622-639.
Acknowledgments 13. Haugen  BR, Alexander  EK, Bible  KC, et  al. 2015 American
Thyroid Association Management Guidelines for Adult Patients
Financial Support: Writing support was provided by Marian with Thyroid Nodules and Differentiated Thyroid Cancer.
Everett Kent, BSN. V.R.  contributed to this paper as a PhD Thyroid. 2016;26(1):1-133.
candidate enrolled in the Innovative Biomedical Technologies 1 4. Russ  G, Bonnema  SJ, Erdogan  MF, Durante  C, Ngu  R,

Leenhardt  L. European Thyroid Association Guidelines
in Clinical Medicine Program of the Sapienza University of
for Ultrasound Malignancy Risk Stratification of Thyroid
Rome. The study was supported by The Umberto Di Mario Nodules in Adults: the EU-TIRADS. Eur Thyroid J.
Foundation ONLUS. 2017;6(5):225-237.
doi:10.1210/clinem/dgaa322 https://academic.oup.com/jcem  13

15. Shin  JH, Baek  JH, Chung  J, et  al.; Korean Society of Thyroid 32. Tessler FN, Middleton WD, Grant EG, et al. ACR Thyroid Imaging,
Radiology (KSThR) and Korean Society of Radiology. Reporting and Data System (TI-RADS): white paper of the ACR
Ultrasonography diagnosis and imaging-based management of TI-RADS Committee. J Am Coll Radiol. 2017;14(5):587-595.
thyroid nodules: revised Korean Society of Thyroid Radiology 33. Grani G, D’Alessandri M, Carbotta G, et al. Grey-scale analysis
Consensus Statement and Recommendations. Korean J Radiol. improves the ultrasonographic evaluation of thyroid nodules.
2016;17(3):370-395. Medicine (Baltimore). 2015;94(27):e1129.
16. Cappelli C, Pirola I, Agosti B, et al. Complications after fine-needle 34. Grani G, Lamartina L, Ramundo V, et al. Taller-than-wide shape:
aspiration cytology: a retrospective study of 7449 consecutive a new definition improves the specificity of TIRADS systems. Eur
thyroid nodules. Br J Oral Maxillofac Surg. 2017;55(3):266-269. Thyroid J. 2020;9(2):85-91.
17. Hoang JK, Middleton WD, Farjat AE, et al. Reduction in thyroid 35. Ha  SM, Chung  YJ, Ahn  HS, Baek  JH, Park  SB. Echogenic

nodule biopsies and improved accuracy with American College foci in thyroid nodules: diagnostic performance with com-
of Radiology Thyroid Imaging Reporting and Data System. bination of TIRADS and echogenic foci. BMC Med Imaging.
Radiology. 2018;287(1):185-193. 2019;19(1):28.
18. Ahmadi  S, Oyekunle  T, Jiang  X’, et  al. A direct comparison of 36. Ramundo V, Di Gioia CRT, Falcone R, Lamartina L, Biffoni M,

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

the Ata and Ti-Rads ultrasound scoring systems. Endocr Pract. Giacomelli L, Filetti S, Durante C, Grani G. Diagnostic perform-
2019;25(5):413-422. ance of neck ultrasonography in the preoperative evaluation for
19. Middleton WD, Teefey SA, Reading CC, et al. Multiinstitutional extrathyroidal extension of suspicious thyroid nodules. World J
analysis of thyroid nodule risk stratification using the American Surg. 2020. doi: 10.1007/s00268-020-05482-6.
College of Radiology Thyroid Imaging Reporting and Data 37. Seifert  P, Görges  R, Zimny  M, Kreissl  MC, Schenke  S.
System. AJR Am J Roentgenol. 2017;208(6):1331-1341. Interobserver agreement and efficacy of consensus reading in
20. Xu T, Wu Y, Wu RX, et al. Validation and comparison of three Kwak-, EU-, and ACR-thyroid imaging recording and data sys-
newly-released Thyroid Imaging Reporting and Data Systems for tems and ATA guidelines for the ultrasound risk stratification of
cancer risk determination. Endocrine. 2019;64(2):299-307. thyroid nodules. Endocrine. 2020;67(1):143-154.
21. Grani  G, Lamartina  L, Ascoli  V, et  al. Reducing the number of 38. Grani  G, Lamartina  L, Cantisani  V, Maranghi  M, Lucia  P,
unnecessary thyroid biopsies while improving diagnostic ac- Durante C. Interobserver agreement of various thyroid imaging
curacy: toward the “right” TIRADS. J Clin Endocrinol Metab. reporting and data systems. Endocr Connect. 2018;7(1):1-7.
2019;104(1):95-102. 39. Tessler  FN. Thyroid nodules and real estate: location matters.
22. Basha MAA, Alnaggar AA, Refaat R, et al. The validity and re- Thyroid. 2020;30(3):349-350.
producibility of the thyroid imaging reporting and data system 40. Cantisani V, David E, Grazhdani H, et al. Prospective evaluation
(TI-RADS) in categorization of thyroid nodules: multicentre pro- of semiquantitative strain ratio and quantitative 2D ultrasound
spective study. Eur J Radiol. 2019;117:184-192. shear wave elastography (SWE) in association with TIRADS clas-
23. Persichetti  A, Di  Stasio  E, Guglielmi  R, et  al. Predictive sification for thyroid nodule characterization. Ultraschall Med.
value of malignancy of thyroid nodule ultrasound classifica- 2019;40(4):495-503.
tion systems: a prospective study. J Clin Endocrinol Metab. 41. Trimboli P, Castellana M, Virili C, et al. Performance of contrast-
2018;103(4):1359-1368. enhanced ultrasound (CEUS) in assessing thyroid nodules: a sys-
24. Castellana M, Castellana C, Treglia G, Giorgino F, Giovanella L, tematic review and meta-analysis using histological standard of
Russ  G, Trimboli  P. Performance of five ultrasound risk strati- reference. Radiol Med. 2020;125(4):406-415.
fication systems in selecting thyroid nodules for FNA. A  meta- 42. Buda M, Wildman-Tobriner B, Hoang JK, et al. Management of
analysis. J Clin Endocrinol Metab. 2020;105(5):1659-1669. thyroid nodules seen on US images: deep learning may match per-
25. Castellana M, Piccardo A, Virili C, et al. Can ultrasound systems formance of radiologists. Radiology. 2019;292(3):695-701.
for risk stratification of thyroid nodules identify follicular car- 43. Zhang B, Tian J, Pei S, et al. Machine learning-assisted system for
cinoma? Cancer Cytopathol. 2020;128(4):250-259. thyroid nodule diagnosis. Thyroid. 2019;29(6):858-867.
26. Valderrabano  P, Klippenstein  DL, Tourtelot  JB, et  al. New 44. Aldrink JH, Adler B, Haines J, et al. Patients exposed to diagnostic
American Thyroid Association sonographic patterns for thyroid head and neck radiation for the management of shunted hydro-
nodules perform well in medullary thyroid carcinoma: institu- cephalus have a significant risk of developing thyroid nodules.
tional experience, systematic review, and meta-analysis. Thyroid. Pediatr Surg Int. 2016;32(6):565-569.
2016;26(8):1093-1100. 45. Cahoon  EK, Nadyrov  EA, Polyanskaya  ON, et  al. Risk of thy-
27. Falcone R, Ramundo V, Lamartina L, et al. Sonographic presen- roid nodules in residents of Belarus exposed to Chernobyl
tation of metastases to the thyroid gland: a case series. J Endocr fallout as children and adolescents. J Clin Endocrinol Metab.
Soc. 2018;2(8):855-859. 2017;102(7):2207-2217.
28. Sawka  AM, Gagliardi  AR, Haymart  MR, et  al. A survey of 46. Land  CE, Kwon  D, Hoffman  FO, et  al. Accounting for shared
American Thyroid Association members regarding the 2015 and unshared dosimetric uncertainties in the dose response for
Adult Thyroid Nodule and Differentiated Thyroid Cancer ultrasound-detected thyroid nodules after exposure to radio-
Clinical Practice Guidelines. Thyroid. 2020;30(1):25-33. active fallout. Radiat Res. 2015;183(2):159-173.
29. Gómez-Pérez  AM, Fernández-García  JC, Iglesias  P, et  al; en 47. Schwetschenau  E, Kelley  DJ. The adult neck mass. Am Fam
representación del Área de Conomiento de Tiroides de al SEEN. Physician. 2002;66(5):831-838.
Diagnosis and treatment of thyroid nodules in Spain. Results of a 48. Jasim S, Baranski TJ, Teefey SA, Middleton WD. Investigating the
national survey. Endocrinol Diabetes Nutr. 2020. doi:10.1016/j. effect of thyroid nodule location on the risk of thyroid cancer.
endinu.2019.10.003. Thyroid. 2020;30(3):401-407.
30. Mauri G, Gitto S, Cantisani V, Vallone G, Schiavone C, Papini E, 49. Ramundo  V, Lamartina  L, Falcone  R, et  al. Is thyroid nodule
Sconfienza  LM. Use of the Thyroid Imaging Reporting and location associated with malignancy risk? Ultrasonography.
Data System (TIRADS) in clinical practice: an Italian survey. 2019;38(3):231-235.
Endocrine. 20202020;68(2):329-335. 50. Zhang  F, Oluwo  O, Castillo  FB, et  al. Thyroid nodule location
31. Symonds CJ, Seal P, Ghaznavi S, Cheung WY, Paschke R. Thyroid on ultrasonography as a predictor of malignancy. Endocr Pract.
nodule ultrasound reports in routine clinical practice provide in- 2019;25(2):131-137.
sufficient information to estimate risk of malignancy. Endocrine. 51. Cibas ES, Ali SZ. The 2017 Bethesda System for reporting thyroid
2018;61(2):303-307. cytopathology. Thyroid. 2017;27(11):1341-1346.
14  Grani et al   Contemporary Thyroid Nodule Evaluation J Clin Endocrinol Metab, September 2020, 105(9):1–15

52. Nardi F, Basolo F, Crescenzi A, et al. Italian consensus for the clas- 71. Sponziello  M, Brunelli  C, Verrienti  A, et  al. Performance of a
sification and reporting of thyroid cytology. J Endocrinol Invest. dual-component molecular assay in cytologically indeterminate
2014;37(6):593-599. thyroid nodules. Endocrine. 2020;68(2):458-465.
53. Baloch  Z, LiVolsi  VA. The Bethesda System for reporting thy- 72. San  Martin  VT, Lawrence  L, Bena  J, Madhun  NZ, Berber  E,
roid cytology (TBSRTC): from look-backs to look-ahead. Diagn Elsheikh TM, Nasr CE. Real-world comparison of Afirma GEC
Cytopathol. 2020. doi: 10.1002/dc.24385. and GSC for the assessment of cytologically indeterminate thy-
54. Rossi  ED, Adeniran  AJ, Faquin  WC. Pitfalls in thyroid roid nodules. J Clin Endocrinol Metab. 2020;105(3).
cytopathology. Surg Pathol Clin. 2019;12(4):865-881. 73. Wei  S, Veloski  C, Sharda  P, Ehya  H. Performance of the
55. Hong HS, Lee JY. Diagnostic performance of ultrasound patterns Afirma genomic sequencing classifier versus gene expres-
by K-TIRADS and 2015 ATA Guidelines in Risk Stratification sion classifier: an institutional experience. Cancer Cytopathol.
of Thyroid Nodules and Follicular Lesions of Undetermined 2019;127(11):720-724.
Significance. AJR Am J Roentgenol. 2019;213(2):444-450. 74. Angell TE, Heller HT, Cibas ES, et al. Independent comparison
56. Valderrabano P, McGettigan MJ, Lam CA, et al. Thyroid nodules of the Afirma genomic sequencing classifier and gene expres-
with indeterminate cytology: utility of the American Thyroid sion classifier for cytologically indeterminate thyroid nodules.

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

Association sonographic patterns for cancer risk stratification. Thyroid. 2019;29(5):650-656.
Thyroid. 2018;28(8):1004-1012. 75. Guan H, Toraldo G, Cerda S, et al. Utilities of RAS mutations in
57. Grani  G, Lamartina  L, Ascoli  V, et  al. Ultrasonography scoring preoperative fine needle biopsies for decision making for thyroid
systems can rule out malignancy in cytologically indeterminate nodule management: results from a single-center prospective co-
thyroid nodules. Endocrine. 2017;57(2):256-261. hort. Thyroid. 2020;30(4):536-547.
58. Ahmadi S, Herbst R, Oyekunle T, et al. Using the ATA and ACR 76. Medici  M, Kwong  N, Angell  TE, et  al. The variable phenotype
Ti-Rads sonographic classifications as adjunctive predictors of and low-risk nature of RAS-positive thyroid nodules. BMC Med.
malignancy for indeterminate thyroid nodules. Endocr Pract. 2015;13:184.
2019;25(9):908-917. 77. Boufraqech M, Klubo-Gwiezdzinska J, Kebebew E. MicroRNAs
59. Lee  JH, Han  K, Kim  EK, et  al. Risk stratification of thyroid in the thyroid. Best Pract Res Clin Endocrinol Metab.
nodules with atypia of undetermined significance/follicular lesion 2016;30(5):603-619.
of undetermined significance (AUS/FLUS) cytology using ultra- 78. Erler  P, Keutgen  XM, Crowley  MJ, et  al. Dicer expression and
sonography patterns defined by the 2015 ATA Guidelines. Ann microRNA dysregulation associate with aggressive features in
Otol Rhinol Laryngol. 2017;126(9):625-633. thyroid cancer. Surgery. 2014;156(6):1342-50; discussion 1350.
60. Barbosa  TLM, Junior  COM, Graf  H, et  al. ACR TI-RADS 79. Rosignolo F, Sponziello M, Giacomelli L, et al. Identification of
and ATA US scores are helpful for the management of thyroid thyroid-associated serum microRNA profiles and their potential
nodules with indeterminate cytology. BMC Endocr Disord. use in thyroid cancer follow-up. J Endocr Soc. 2017;1(1):3-13.
2019;19(1):112. 80. Nicholson  KJ, Roberts  MS, McCoy  KL, Carty  SE, Yip  L.
61. Chaigneau E, Russ G, Royer B, et al. TIRADS score is of limited Molecular testing versus diagnostic lobectomy in Bethesda
clinical value for risk stratification of indeterminate cytological III/IV thyroid nodules: a cost-effectiveness analysis. Thyroid.
results. Eur J Endocrinol. 2018;179(1):13-20. 2019;29(9):1237-1243.
62. Grani  G, Filetti  S. Sonographically estimated thyroid nodule 81. Park  CJ, Kim  EK, Moon  HJ, Yoon  JH, Park  VY, Kwak  JY.
malignancy risk: strengths and limitations in clinical practice. Thyroid nodules with nondiagnostic cytologic results: follow-up
Endocr Pract. 2019;25(9):966-967. management using ultrasound patterns based on the 2015
63. Allen  L, Al  Afif  A, Rigby  MH, et  al. The role of repeat fine American Thyroid Association Guidelines. AJR Am J Roentgenol.
needle aspiration in managing indeterminate thyroid nodules. J 2018;210(2):412-417.
Otolaryngol Head Neck Surg. 2019;48(1):16. 82. Grani  G, Lamartina  L, Biffoni  M, et  al. Sonographically esti-
64. Stewart  R, Leang  YJ, Bhatt  CR, Grodski  S, Serpell  J, Lee  JC. mated risks of malignancy for thyroid nodules computed with
Quantifying the differences in surgical management of patients five standard classification systems: changes over time and their
with definitive and indeterminate thyroid nodule cytology. Eur J relation to malignancy. Thyroid. 2018;28(9):1190-1197.
Surg Oncol. 2020;46(2):252-257. 83. Ha  SM, Baek  JH, Choi  YJ, et  al. Malignancy risk of ini-
65. Nikiforova MN, Mercurio S, Wald AI, et al. Analytical perform- tially benign thyroid nodules: validation with various Thyroid
ance of the ThyroSeq v3 genomic classifier for cancer diagnosis in Imaging Reporting and Data System guidelines. Eur Radiol.
thyroid nodules. Cancer. 2018;124(8):1682-1690. 2019;29(1):133-140.
66. Steward  DL, Carty  SE, Sippel  RS, et  al. Performance of a 84. Patel  KN, Yip  L, Lubitz  CC, et  al. The American Association
multigene genomic classifier in thyroid nodules with indeter- of Endocrine Surgeons Guidelines for the Definitive Surgical
minate cytology: a prospective blinded multicenter study. JAMA Management of Thyroid Disease in Adults. Ann Surg.
Oncol. 2019;5(2):204-212. 2020;271(3):e21-e93.
67. Patel  KN, Angell  TE, Babiarz  J, et  al. Performance of a gen- 85. Durante C, Costante G, Lucisano G, et al. The natural history of
omic sequencing classifier for the preoperative diagnosis of benign thyroid nodules. Jama. 2015;313(9):926-935.
cytologically indeterminate thyroid nodules. JAMA Surg. 86. Mohammadi  M, Betel  C, Burton  KR, Higgins  KM, Ghorab  Z,
2018;153(9):817-824. Halperin IJ. Follow-up of benign thyroid nodules-can we do less?
68. Labourier E, Shifrin A, Busseniers AE, et al. Molecular testing for Can Assoc Radiol J. 2019;70(1):62-67.
miRNA, mRNA, and DNA on fine-needle aspiration improves 87. Rosário  PW, Calsolari  MR. What is the best criterion for repe-
the preoperative diagnosis of thyroid nodules with indeterminate tition of fine-needle aspiration in thyroid nodules with initially
cytology. J Clin Endocrinol Metab. 2015;100(7):2743-2750. benign cytology? Thyroid. 2015;25(10):1115-1120.
69. Banizs AB, Silverman JF. The utility of combined mutation ana- 88. Singh Ospina N, Maraka S, Espinosa DeYcaza A, et al. Diagnostic
lysis and microRNA classification in reclassifying cancer risk of accuracy of thyroid nodule growth to predict malignancy in thy-
cytologically indeterminate thyroid nodules. Diagn Cytopathol. roid nodules with benign cytology: systematic review and meta-
2019;47(4):268-274. analysis. Clin Endocrinol (Oxf). 2016;85(1):122-131.
70. González  HE, Martínez  JR, Vargas-Salas  S, et  al. A 10-gene 89. Angell TE, Vyas CM, Medici M, et al. Differential growth rates of
classifier for indeterminate thyroid nodules: Development and benign vs. malignant thyroid nodules. J Clin Endocrinol Metab.
Multicenter Accuracy Study. Thyroid. 2017;27(8):1058-1067. 2017;102(12):4642-4647.
doi:10.1210/clinem/dgaa322 https://academic.oup.com/jcem  15

90. Ito  Y, Uruno  T, Nakano  K, et  al. An observation trial without 107. Heck  K, Happel  C, Grünwald  F, Korkusuz  H. Percutaneous
surgical treatment in patients with papillary microcarcinoma of microwave ablation of thyroid nodules: effects on thyroid func-
the thyroid. Thyroid. 2003;13(4):381-387. tion and antibodies. Int J Hyperthermia. 2015;31(5):560-567.
91. Ito  Y, Miyauchi  A, Kihara  M, Higashiyama  T, Kobayashi  K, 108. Ha  EJ, Baek  JH, Kim  KW, et  al. Comparative efficacy of
Miya  A. Patient age is significantly related to the progression radiofrequency and laser ablation for the treatment of benign
of papillary microcarcinoma of the thyroid under observation. thyroid nodules: systematic review including traditional pooling
Thyroid. 2014;24(1):27-34. and bayesian network meta-analysis. J Clin Endocrinol Metab.
92. Cho  SJ, Suh  CH, Baek  JH, et  al. Active surveillance for small 2015;100(5):1903-1911.
papillary thyroid cancer: a systematic review and meta-analysis. 109. Sui WF, Li JY, Fu JH. Percutaneous laser ablation for benign thyroid
Thyroid. 2019;29(10):1399-1408. nodules: a meta-analysis. Oncotarget. 2017;8(47):83225-83236.
93. Saravana-Bawan  B, Bajwa  A, Paterson  J, McMullen  T. Active 110. Papini  E, Pacella  CM, Solbiati  LA, et  al. Minimally-invasive
surveillance of low-risk papillary thyroid cancer: a meta-analysis. treatments for benign thyroid nodules: a Delphi-based consensus
Surgery. 2020;167(1):46-55. statement from the Italian minimally-invasive treatments of the
94. Brito JP, Ito Y, Miyauchi A, Tuttle RM. A clinical framework to thyroid (MITT) group. Int J Hyperthermia. 2019;36(1):376-382.

Downloaded from https://academic.oup.com/jcem/article/105/9/dgaa322/5850848 by guest on 08 August 2020

facilitate risk stratification when considering an active surveil- 111. Kim JH, Baek JH, Lim HK, et al.; Guideline Committee for the
lance alternative to immediate biopsy and surgery in papillary Korean Society of Thyroid Radiology (KSThR) and Korean
microcarcinoma. Thyroid. 2016;26(1):144-149. Society of Radiology. 2017 Thyroid Radiofrequency Ablation
95. Ito  Y, Miyauchi  A, Inoue  H, et  al. An observational trial for Guideline: Korean Society of Thyroid Radiology. Korean J
papillary thyroid microcarcinoma in Japanese patients. World J Radiol. 2018;19(4):632-655.
Surg. 2010;34(1):28-35. 112. Hahn  SY, Shin  JH, Na  DG, et  al.; Korean Society of Thyroid
96. Haser GC, Tuttle RM, Su HK, et al. Active surveillance for papil- Radiology (KSThR); Korean Society of Radiology. Ethanol
lary thyroid microcarcinoma: new challenges and opportunities Ablation of the Thyroid Nodules: 2018 Consensus Statement
for the health care system. Endocr Pract. 2016;22(5):602-611. by the Korean Society of Thyroid Radiology. Korean J Radiol.
97. Trimboli P, Castellana M, Sconfienza LM, et al. Efficacy of thermal 2019;20(4):609-620.
ablation in benign non-functioning solid thyroid nodule: a sys- 113. Dobnig  H, Zechmann  W, Hermann  M, et  al. Radiofrequency
tematic review and meta-analysis. Endocrine. 2020;67(1):35-43. ablation of thyroid nodules: “Good Clinical Practice
98. Lang BHH, Woo YC, Chiu KW. High-intensity focused ablation Recommendations” for Austria: an interdisciplinary state-
(HIFU) of single benign thyroid nodule rarely alters underlying ment from the following professional associations: Austrian
thyroid function. Int J Hyperthermia. 2017;33(8):875-881. Thyroid Association (ÖSDG), Austrian Society for Nuclear
99. Lang BHH, Wong CKH, Ma EPM. Single-session high intensity fo- Medicine and Molecular Imaging (OGNMB), Austrian
cussed ablation (HIFU) versus open cervical hemithyroidectomy Society for Endocrinology and Metabolism (ÖGES), Surgical
for benign thyroid nodule: analysis on early efficacy, safety and Endocrinology Working Group (ACE) of the Austrian Surgical
voice quality. Int J Hyperthermia. 2017;33(8):868-874. Society (OEGCH). Wien Med Wochenschr. 2020;170(1-2):6-14.
100. Lang BH, Wu ALH. High intensity focused ultrasound (HIFU) 114. Zatelli MC, Lamartina L, Meringolo D, et al. Thyroid nodule recur-
ablation of benign thyroid nodules - a systematic review. J Ther rence following lobo-isthmectomy: incidence, patient’s characteris-
Ultrasound. 2017;5:11. tics, and risk factors. J Endocrinol Invest. 2018;41(12):1469-1475.
101. Pacella CM, Mauri G, Cesareo R, et al. A comparison of laser 115. Barczyński M, Stopa-Barczyńska M. Hemithyroidectomy for be-
with radiofrequency ablation for the treatment of benign thy- nign euthyroid asymmetric nodular goitre. Best Pract Res Clin
roid nodules: a propensity score matching analysis. Int J Endocrinol Metab. 2019;33(4):101288.
Hyperthermia. 2017;33(8):911-919. 116. Olson  MT, Boonyaarunnate  T, Aragon  Han  P, Umbricht  CB,
102. Pacella CM, Mauri G, Achille G, et al. Outcomes and risk fac- Ali SZ, Zeiger MA. A tertiary center’s experience with second re-
tors for complications of laser ablation for thyroid nodules: a view of 3885 thyroid cytopathology specimens. J Clin Endocrinol
multicenter study on 1531 patients. J Clin Endocrinol Metab. Metab. 2013;98(4):1450-1457.
2015;100(10):3903-3910. 1 17. Gerhard R, Boerner SL. The value of second opinion in thy-
103. Chung SR, Suh CH, Baek JH, Park HS, Choi YJ, Lee JH. Safety roid cytology: a review. Cancer Cytopathol. 2014;122(8):
of radiofrequency ablation of benign thyroid nodules and recur- 611-619.
rent thyroid cancers: a systematic review and meta-analysis. Int 118. Rosato L, Avenia N, Bernante P, et al. Complications of thyroid
J Hyperthermia. 2017;33(8):920-930. surgery: analysis of a multicentric study on 14 934 patients oper-
104. Wang  JF, Wu  T, Hu  KP, et  al. Complications following ated on in Italy over 5 years. World J Surg. 2004;28(3):271-276.
radiofrequency ablation of benign thyroid nodules: a systematic 119. Zhang  M, Tufano  RP, Russell  JO, et  al. Ultrasound-guided
review. Chin Med J (Engl). 2017;130(11):1361-1370. radiofrequency ablation versus surgery for low-risk papillary
105. Korkusuz Y, Kohlhase K, Gröner D, et al. Microwave ablation of thyroid microcarcinoma: results of over 5  years’ follow-up.
symptomatic benign thyroid nodules: energy requirement per ml Thyroid. 2020;30(3):408-417.
volume reduction. Rofo. 2016;188(11):1054-1060. 120. Choi  Y, Jung  SL. Efficacy and safety of thermal ablation
106. Wang  B, Han  ZY, Yu  J, et  al. Factors related to recurrence of techniques for the treatment of primary papillary thyroid
the benign non-functioning thyroid nodules after percutaneous microcarcinoma: a systematic review and meta-analysis.
microwave ablation. Int J Hyperthermia. 2017;33(4):459-464. Thyroid. 2020;30(5):720-731.