Anda di halaman 1dari 1

Correlation between the British Thyroid

Association ultrasound grading of thyroid

nodules and histopathology specimens, with
assessment of inter-rater agreement: a one-year
institutional experience
Olivia Francies, Susan Jawad, Simon Morley, Sally Daniels, Sofia Otero

• Total thyroidectomy specimens 142 Solid, hypoechoic
The British Thyroid Association (BTA) guidelines
nodule with an
2014 introduced an ultrasound (US) classification for • Total number with US at UCLH 126 (+1 external)
irregular outline,
thyroid nodules: U1 – U5 • 6 samples excluded for Graves’ disease internal vascularity
• Total number of samples included 121 and containing
echogenic foci
1) Malignancy rate for U grade (as given on the original consistent with
report) calcification
U2 3% Graded U3 - U5
U3 40% = papillary thyroid
U4 79%
U5 100%

2) Sensitivity and specificity for detection of malignant

Sensitivity (%) Specificity (%) Solid, hypoechoic
Original 98.0 44.4 nodule
report Graded U3 & U4
Rater 1 95.9 47.2 = follicular
Rater 2 95.9 43.0
Rater 3 95.8 43.0

U1. Normal thyroid U4. Suspicious: Rater 4 91.8 48.6

(a) solid, hypo-echoic (cf thyroid) High Moderate
U2. Benign: (b) solid, very hypo-echoic (cf
(a) halo, iso-/mildly hyper-echoic strap muscle) 3) Inter-rater agreement: Fleiss kappa score of 0.51 =
(b) cystic +/- ring down sign
(c) disrupted peripheral
calcification, hypo-echoic
(c) micro-cystic/spongiform (d) lobulated outline
(d & e) peripheral egg shell
calcification U5. Malignant: Malignancy rates for each U category are in line with
(f) peripheral vascularity (a) solid, hypo-echoic, lobulated/
irregular outline, micro-calcification
previous scoring systems
U3. Indeterminate/Equivocal: (b) solid, hypo-echoic, • High sensitivity
(a) hyper-echoic, solid, halo lobulated/irregular outline, globular
(b) ?hypo-echoic, equivocal calcification • Only moderate specificity
echogenic foci (c) intra-nodular vascularity Moderate inter-rater agreement overall
(c) mixed/central vascularity (d) shape (taller >wide) (AP>TR)
(e) characteristic associated • Inter-rater agreement for U2 is good
lymphadenopathy • Inter-rater agreement for other U grades is

1) Calculate our institutional rate of
malignancy for each U category
The graph demonstrates good inter-rater agreement for U2 • Use a simplified scoring system for thyroid
2) Calculate the sensitivity and specificity for nodules but greater variability for U3-5 nodules. nodules of benign, indeterminate and
the detection of malignant thyroid nodules suspicious for malignancy
3) Calculate inter-rater agreement for the US Example cases • Calculate the sensitivity and specificity and
classification (U2 – U5) of thyroid nodules inter-rater agreement for this simplified
Isoechoic nodule scoring system
with micro-cystic
• If this is equivalent or better than the current
Uniformly graded U1 – U5 system, consider implementing this
U2 scoring system
• All total and hemi-thyroidectomies for a 12-month period
between 2015–2016 were considered
= benign multi- • Repeat the audit after 6 months
nodular goitre
o Samples excluded: completion thyroidectomies for • Disseminate the results directly to the head
malignancy and diffuse non-nodular disease (Graves’) and neck radiology team
• The U grade in the original US report was correlated with
the final histopathology Solid, hypoechoic
o For multinodular thyroids, the highest U grade given in nodule with
the report was used for analysis lobulated outline
o Incidental micro-papillary carcinomas (where these Graded U3 & U4
nodules were not seen or classified at US) were not
included in the data = follicular variant
• The malignancy rate for each U category was calculated
of papillary cell • Perros P, Colley S, Boelaert K et al. British Thyroid
from the original report carcinoma Association Guidelines for the Management of Thyroid
Cancer. Third edition. Clinical Endocrinology. 2014;
81(Suppl 1): 1-122. Chapter 7 p14-18. John Wiley & Sons
• 4 raters (3 Consultant Radiologists and 1 Sonographer all Ltd. Wiley Blackwell
with an interest in head & neck imaging) retrospectively Solid, hypoechoic
• Lee YH et al. Differentiation between benign and malignant
assessed the stored US images and allocated a U grade nodule with intra- thyroid solid nodules using an ultrasound classification
for the most concerning nodule(s) nodular vascularity system. Korean Journal of Radiology. 2011; 12: 559-567
• Sensitivity and specificity for malignancy for the original Graded U3 - U5 • Kwak JY et al. Thyroid Imaging Reporting and Data System
report and each of the 4 raters was calculated for US Features of Nodules: A Step in Establishing Better
= minimally
• Inter-rater agreement was calculated for the original report invasive follicular Stratification of Cancer Risk. Radiology. 2011; 260(3): 892-
and each of the 4 raters using a Fleiss Kappa score cell carcinoma 9. doi: 10.1148/radiol.11110206