Thyroid Nodule Risk
Stratification and FNA
Guidelines
Mark A. Lupo, MD, FACE, ECNU
Thyroid & Endocrine Center of Florida
Assistant Clinical Professor of Medicine
Florida State University, College of Medicine
Sarasota, Florida
Objectives
Review features of benign and malignant
nodules
Highlight importance of pattern recognition for
risk of malignancy assessment
Introduce ACR TIRADs System
Discuss current guidelines for nodule FNA
US Appearance of Nodules
Composition (solid/cystic)
Echogenicity (darkness on US)
Margins
Presence of calcifications
Shape
Vascularity
High
Suspicion
70-90%
Intermediate
Suspicion
10-20%
Low
Suspicion
5-10%
Very low
Suspicion
<3%
Benign
<1%
ATA 2015: Nodule Sonographic Pattern Risk of Malignancy
Haugen et al. Thyroid; October 2015 (epub)
AACE 2016 Nodule Guidelines
Gharib et al. Endocrine Practice. May 2016
Characteristics Suggesting Low Risk Nodules
Hyperechoic nodule (especially in autoimmune thyroiditis)
Pseudonodules in Hashimoto’s
Pure cyst without a solid component
Colloid within nodule “comet tails”
Spongiform Echotexture
Hyperechoic with background of Hashi’s
Nodule
Hyperechoic with background of Hashi’s
Nodule?
CLEFT SIGN – Due to Fibrosis
Cleft Sign
Right Transverse Right Sagittal
Thick hyperechoic fibrotic band separates the posterior and
anterior components of the lobe in transverse view creating the appearance
of a hypoechoic nodule (arrow) with hyperechoic halo
(note: true halos are only hypoechoic)
In Sagittal view, it is evident that there is no discrete nodule.
Colloid within Nodule with “Comet Tails”
Colloid Within Nodule “Cat’s Eye” (Comet
Tail, Stepladder or Ringdown artifact)
aggregation of multiple
microcystic components
in more than 50% of the
volume of the nodule
“honeycomb of internal
cystic spaces”
Only 1 in 360 spongiform
nodules malignant
99.7% Specificity (Moon)
“Spongiform” nodules
Moon Radiology 2008; 247: 762-70
Bonavita AJR 2009; 193:207-13
Spongiform echotexture
Spongiform Echotexture
Note: Bright Linear Reflectors all
posterior to the microcystic areas
Thin-walled Cyst
Enhancement
Thin-walled Cyst
Complex cyst
Halos & Margins Halo: sonolucent rim around an iso/hyperechoic nodule
representing capsule
Thin/regular – lower risk
Thick/irregular – higher risk
Hypoechoic Nodules ---- Margins:
Smooth and regular
Poorly Defined: interface between nodule and surrounding
parenchyma is difficult to delineate
Lower risk, seen in hyperplastic nodules
Irregular: the demarcation between the nodule and parenchyma
is clearly visible but demonstrates and irregular, infiltrative or
spiculated course.
Higher risk
Halo
Edge Artifact
Halo
No Halo, but a Smooth Margin
Halo
Thin Halo
Benign Follicular AdenomaThick, Irregular Halo
Follicular CA
Subtle differences in halo
Irregular halo
Thinner halo
Isoechoic
Nodules
Margins
Poorly defined, but not
infiltrative (spongiform)Irregular
Infiltrative/Irregular Borders
Suspicious Sonographic Features
Hypoechoic
Microcalcifications
Infiltrative margins
Taller than wide on transverse view
Abnormal cervical lymph nodes
NOT DOPPLER?
Hypoechoic
Compare to strap muscles and SCM
Hypoechoic
Hypoechoic
Markedly Hypoechoic
Compare to strap muscles and SCM
Macrocalcifications: hyperechoic spots with
acoustic shadowing (“dense calcifications”)
Low-intermediate risk
Microcalcifications: hyperechoic spots
without acoustic shadowing (thought to
represent psammoma bodies) (<1mm)
High Risk
Rim (eggshell) calcifications: curvi-linear
hyperechoic calcification along periphery
• Higher risk if interrupted
Calcifications
Macrocalcification Patterns
Kim et al. J Ultrasound Med 2008, 27:1179
Coarse Macrocalcification in PTC
Bonavita AJR 2009; 193:207-13
Eggshell Calcifications with Shadowing
Smooth Eggshell
Maybe Reassuring
Interrupted Eggshell
Not reassuring
PTC in Graves’
Left lobe showing background heterogeneity and a calcified nodule
which proved to be papillary thyroid cancer on FNA biopsy in this patient with
Graves’ Disease.
Microcalcification vs Comet Tail
Microcalcifications
Microcalcifications
Dense and
Microcalcifications
Microcalcifications in Papillary Cancer
May be confused with comet tails
When in doubt - assume the worse
sagittal
NOT microcalcifications!
sagittal
Small hyperechoic linear streaks just
posterior to small cystic area
posterior acoustic enhancement!
Microcalcifications
Calcifications in PTC
LongitudinalTransverse
PTC
Cystic PTC
Cystic PTC - UGFNA
4cm Left Cystic Nodule FNA - PTC
Invasion of Strap Muscle
Invasion of Carotid Sheath
Invasion of Larynx
Cystic Papillary
carcinoma
invading the
Larynx. Note
the classic
intra-cystic
Cauliflower
appearance.
Taller than wide
Kim AJR 2002; Cappelli Clin Endocrinol 2005; Moon Radiology 2008
Nodule is taller than wide on
the transverse view
What About Intranodular Flow?
B-Mode Power Doppler
NO LONGER CONSIDERED AN INDEPENDENT
RISK FACTOR WHEN DECIDING ON FNA
PTC – lack of Doppler
Don’t be re-assured by lack of
Doppler flow !
US Prediction of Thyroid Cancer
Sensitivity Specificity
Microcalcifications 45% 85%Absence of halo 66% 46%Poorly defined margins 64% 77%Hypoechoic 80% 45%Increased Doppler flow 67% 81%Taller than Wide 48% 92%MicroCa2+ + irreg margin 30% 95%MicroCa2+ + hypoechoic 28% 95%Solid + hypoechoic 73% 69%
Brkljacic J Clin Ultrasound 1994; Takashima J Clin Ultrasound 1994; Rago Euro J Endorinol 1998; Cerbone Horm Res 1999; Leenhardt J Clin Endocrinol Metab 1999; Kim AJR 2002; Papini J Clin Endocrinol Metab 2002; Cappelli Clin Endocrinol 2006
Interobserver variability of US features
Kappa values
Agreement
<0.2 slight
0.21-0.4 fair
0.41-0.6moderate
0.61-0.8 substantial
>0.8 ~perfect!
Moon Radiology 2008 247:762; Choi Thyroid 2010 20:167
Factors Contributing to
Variability of US Reporting
US machine properties
Probe frequency
Gain, power output
Use of harmonics or post image acquisition software e.g.
smoothing
Image interpretation
Static vs. real time
Use of video
Patient factors
Depth of nodule
US Pattern Recognition
While individual
features are suggestive,
combining features and
pattern recognition are
the keys to US risk
stratification of thyroid
nodules
TIRADS-Horvath 2009
Category US PATTERN MALIGNANCY
RISK %
TIRADS 1 Normal thyroid gland 0
TIRADS 2 Cyst with or without comet tail
Spongiform
Mixed cystic/solid with solid portion mixed
0
TIRADS 3 Hashi’s pseudonodule <5
TIRADS 4 5-80
4A Solid or mixed hyper/isoechoic with thin
capsule
5-10
4B Hypoechoic with irreg margins
Hyper/iso/hypo with thick capsule
10-80
TIRADS 5 Hypo/Iso nonencapsulated with microcalcs >80
Horvath J Clin Endocrinol Metab 2009;90:1748
TIRADS-Kwak 2011Category Number of suspicious US
features
CANCER
RISK %
TIRADS 3 NONE 1.7
TIRADS 4A 1 3
TIRADS 4B 2 10
TIRADS 4C 3-4 44-73
TIRADS 5 5 88
Suspicious US features: solid
marked hypoechogenicity, hypoechogenicity
microlobulated or irregular margins
microcalcifications
taller than wide shape
Kwak Radiology 2011;260:892
Kwak 2011 features – A – hypoechoic, B – microca, C – Markedely hypoechoic D – Irregular Margins
Adapted TI-RADS based on Russ 2011.
Russ, Leboulleux, Leenhardt, Hegedus Eur Thyroid J. 2014 Sep;3(3):154-63
Russ 2013
Adapted TI-RADS based on Russ 2011.
Russ, Leboulleux, Leenhardt, Hegedus Eur Thyroid J. 2014 Sep;3(3):154-63
TIRADS - Interobserver
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
TIRADS 2-5 TIRADS 4-5
Ka
pp
a v
alu
es
Cheng et al. Russ et al.
Russ, Leboulleux, Leenhardt, Hegedus Eur Thyroid J. 2014 Sep;3(3):154-63
Russ 2013 95.7% Sensitivity
61% Specificity
High
Suspicion
70-90%
Intermediate
Suspicion
10-20%
Low
Suspicion
5-10%
Very low
Suspicion
<3%
Benign
<1%
ATA 2015: Nodule Sonographic Pattern Risk of Malignancy
Haugen et al. Thyroid; October 2015
PTC foll variantBenign Hürthle cell
adenoma
Follicular thyroid cancer
Isoechoic/Hyperechoic and Solid
~20% of all cancers are Iso/hyperechoic:
predominantly follicular/ Hürthle
Hyperplastic noduleHyperplastic nodule
Sonographic Features of Follicular Variant Papillary
Cancer versus Conventional Papillary Cancer
FEATURE FOLLICULAR
VARIANT (n=44)
CONVENTIONAL
PAPILLARY (n=74)
Size (mean) 17 mm 10.5 mm
Ovoid to round shape 95% 73%
Isoechoic 52% 8%
Halo 25% 3%
Taller than Wide 5% 22%
Marked Hypoechogenicity 5% 38%
Microcalcifications 3% 24%
FNAB Dx of Papillary CA 28% 56%
FNAB Indeterminate 46% 19%
Kim DS et al. J Ultrasound Med 2009, 28:1685
R8 US Pattern and suggested FNA cutoffs
Sonographic
Pattern
Estimated
malignancy
risk
FNA size
cutoff
Strength
of rec
Quality of
evidence
High suspicion >70-90% > 1 cm Strong Moderate
Intermediate
suspicion
10-20% > 1 cm Strong Low
Low suspicion 5-10% > 1.5 cm Weak Low
Very low
suspicion
< 3% > 2 cm Weak Moderate
One option is surveillance
Benign < 1% No biopsy Strong Moderate
FNA is not recommended for nodules
that do not meet the above criteria,
including all nodules < 1 cm
Strong Moderate
Haugen et al. Thyroid; January 2016
Micronodules Concerning for PTC
Sonographic appearance of micro PTC and risk
50% 14% 2%
30% <5%
Ito World J Surg 2016;40:523–528; Miyauchi World J Surg 2016 40:516–522
Sonographic
PatternStrength
of rec
Quality of
evidence
High
suspicionRepeat US and US FNA
within 12 months
Strong Moderate
Intermediate/
Low
suspicion
Repeat US at 12-24m
If growth or new suspicious
US feature, repeat FNA OR
continued observation
Weak Low
Very low
suspicion
Utility of surveillance US and
assessment of nodule growth as
an indicator for repeat FNA is not
known. If repeated, US should at
> 24 months
Weak Low
IF 2nd US FNA done with benign cytology, US
surveillance for continued risk of malignancy is
no longer indicated
Strong Moderate
R23 Follow-up of nodules with benign cytology
Haugen et al. Thyroid; January 2016
Stephanie Lee - AACE 2016 AM
Slide from Stephanie Lee - AACE 2016 AM
Other Patterns
hyperechoic, microcalcs
eggshell calc
isoechoic, coarse calc
isoechoic, margin
lobulated smooth
ATA May Have Limitations in
Thyroid Nodule Classification
Compared with TIRADS
Yoon et a. Radiology 2016; 278: 917-924
-Comparison ATA 2015 with TIRADS 2011
-3.4% of 1,293 nodules without ATA Classification
-18.2% of “not specified” ATA nodules malignant
ACR TI-RADS 2017
Tessler et al; J ACR 2017
Point based system
Composition
Echogenicity
Shape
Margins
Echogenic Foci
“Leave no nodule behind”
Tessler et al; J ACR 2017
Cystic or almost completely cystic
almost universally benign
Russ G. Ulttasonography 2016; 35: 25-38
Moon WJ Radiology 2008; 247: 762-770
Lee et al. Thyroid 2009; 19(4): 341-346
Spongiform composition – 0 points
Colloid cyst
Mixed Cystic-Solid -- 2 points
(w/ no suspicious features)
No FNA
ACR TI-RADS System FNA
Indications TIRADS 3-5
Consider FNA
2016 AACE Nodule Guidelines
Benign - <1% ROM
Intermediate – 10-15% ROM
Suspicious – 50-90% ROM
Sonographic
Pattern
Estimated
malignancy
risk
FNA
size
cutoff
AACE
Level of
Evidence
AACE
Strength
of Rec
High
Risk US
50-90% > 1 cm 2 A
Intermediate
Risk US
5-15% > 2 cm 2 A
Low Risk US <1% > 2 cm
or
growing
2 A
US Risk and suggested FNA cutoffs
Conclusions
Ultrasound is the cornerstone of thyroid nodule
evaluation and always includes evaluation of the
regional lymph nodes
Pattern recognition is more valuable than
individual suspicious features
FNA decision making is based on the risk
stratification of the nodule pattern