Peripheral Calcification in Thyroid Nodules

CME Article

Peripheral Calcification in
Thyroid Nodules

Ultrasonographic Features and
Prediction of Malignancy

Dae Young Yoon, MD, PhD, Joon Won Lee, MD, Suk Ki Chang, MD,
Chul Soon Choi, MD, PhD, Eun Joo Yun, MD, PhD, Young Lan Seo, MD, PhD,
Keon Ha Kim, MD, Hee Sung Hwang, MD

Objective. The purpose of this study was to investigate the association between peripheral calcifica-
tion in thyroid nodules detected on ultrasonography and thyroid malignancy. Methods. We retro-
spectively analyzed the ultrasonographic features of 65 pathologically proven thyroid lesions showing
peripheral calcification for their correlation with histopathologic results. The following ultrasonographic
parameters were assessed for each nodule: size (maximal dimension), shape (anteroposterior dimen-
sion/transverse dimension ratio), internal echogenicity (hypoechoic, isoechoic, hyperechoic, or invisi-
ble), halo sign (present or absent), type of calcification (stippled, curvilinear/smooth margin, or
curvilinear/irregular margin), and extent of calcification (arc or rim). Results. Twelve (18.5%) of 65 thy-
roid nodules with peripheral calcification were malignant, and 53 (81.5%) were benign. Patient demo-
graphics (age and sex) and ultrasonographic features of the nodules (size, shape, internal echogenicity,
halo sign, and type and extent of calcification) did not show any significant differences between benign
and malignant groups. Conclusions. The relatively high prevalence of malignancy and no reliable cri-
terion for malignancy in thyroid nodules with peripheral calcification indicate that fine-needle aspira-
tion or careful ultrasonographic follow-up may be warranted in these cases. Key words: calcification;
thyroid carcinoma; thyroid nodule; ultrasonography.

Abbreviations T hyroid nodules are very common, found by palpa-
AP/T, anteroposterior dimension/transverse dimension; tion in 4% to 7% of the asymptomatic population,1
FNA, fine-needle aspiration in 17% to 27% of cases on ultrasonography,2–4 and
in 50% of cases at autopsy.5 Although most thyroid
Received April 25, 2007, from the Department of nodules are benign, approximately 4% to 14% of such
Radiology, Kangdong Seong-Sim Hospital, nodules are malignant.1,6–8 Thus, it is important to identify
Hallym University College of Medicine, Seoul, which nodules are more likely to be malignant. Several
Korea (D.Y.Y., J.W.L., S.K.C., C.S.C., E.J.Y., Y.L.S.); ultrasonographic characteristics that have been studied as
Department of Radiology, Samsung Medical Center, potential predictors of thyroid malignancy include irregu-
Sungkyunkwan University College of Medicine, lar margins, hypoechogenicity, the absence of a halo, a pre-
Seoul, Korea (K.H.K.); and Department of Radiology, dominantly solid composition, intranodular vascularity,
Hallym University Sacred Heart Hospital, Hallym and the presence of calcification.2,7,9,10
University College of Medicine, Anyang, Korea Calcification is a common finding on thyroid imaging,
(H.S.H.). Revision requested May 15, 2007. Revised and various patterns of calcification may be seen on thy-
manuscript accepted for publication May 21, 2007. roid ultrasonography, including microcalcification,
coarse and dense calcification, and rimlike peripheral
Address correspondence to Dae Young Yoon, calcification.11–13 Among these, both microcalcification
MD, PhD, Department of Radiology, Kangdong and coarse calcification within the nodule are known to
Seong-Sim Hospital, Hallym University College of be associated with an increased likelihood of malignan-
Medicine, 445 Gil-dong, Kangdong-gu, Seoul 134- cy.7–16 To our knowledge, however, there are insufficient
701, Korea.

E-mail: [email protected]

CME Article includes CME test

© 2007 by the American Institute of Ultrasound in Medicine • J Ultrasound Med 2007; 26:1349–1355 • 0278-4297/07/$3.50

Peripheral Calcification in Thyroid Nodules

data to know whether peripheral calcification, as The entire study protocol was approved by
opposed to intranodular calcification, is associ- our Institutional Review Board. Because the
ated with malignancy. patients’ data were evaluated retrospectively
and anonymously, no written informed consent
The purpose of this study was to investigate was necessary.
the association between peripheral calcifica-
tion detected on ultrasonography and thyroid Ultrasonographic Examinations and Image
malignancy. Interpretation
Ultrasonographic examination of the thyroid
Materials and Methods gland was performed with an HDI 5000 system
(Philips Medical Systems, Bothell, WA) or an
Patients Acuson Sequoia 512 system (Siemens Medical
This was a retrospective study analyzing 66 Solutions, Mountain View, CA) equipped with a
patients who had thyroid nodules with peripher- commercially available 8- to 15-MHz linear array
al calcification on ultrasonography. Cases were transducer. All sonograms were interpreted by 2
selected from a radiology report database over a experienced radiologists (D.Y.Y. and S.K.C.), who
3-year period (August 2003–August 2006) at our were unaware of the histopathologic diagnosis.
institution. This subset of reports was then Final decisions regarding the findings were
searched for “peripheral calcifi*,” “ring calcifi*,” reached by consensus. The images were present-
and “rim calcifi*,” where * was a wildcard allow- ed for readers on a picture archiving and commu-
ing matching of any subsequent characters (eg, nication system workstation (Infinitt Technology,
“-cation,” “-cations,” or “-ed”). Peripheral calcifi- Seoul, Korea), in an anonymous random fashion.
cation was defined as bright echoes observed on
the surface of a thyroid nodule with or without The following ultrasonographic parameters
the ability to interpret the tumor by its acoustic were assessed for each nodule: size (maximal
shadows. In contrast, any hyperechoic structure dimension), shape (anteroposterior dimension/
inside a thyroid nodule was considered intran- transverse dimension [AP/T] ratio), internal
odular calcification. echogenicity (hypoechoic, isoechoic, or hypere-
choic in comparison with the background thy-
Thirteen lesions in 10 patients were excluded for roid tissue or invisible because of posterior
the following reasons: ultrasonographically guid- attenuation of the ultrasonic beam), halo sign
ed fine-needle aspiration (FNA) or surgery not (present or absent), and type and extent of
performed (n = 8) and lesions with nondiagnostic peripheral calcification.
FNA results (n = 5). This yielded a final study pop-
ulation of 56 patients (48 women and 8 men; Thirteen nodules were excluded from the analy-
mean age ± SD, 53.5 ± 12.9 years; range, 28–82 sis for shape; in each case, the AP/T ratio could not
years) with 65 thyroid nodules (a single thyroid be assessed because of extensive attenuation of
nodule in 47 patients and 2 nodules in 9 patients). the ultrasonic beam. The types of calcification
were classified into 3 categories as follows: type 1,
The final diagnosis was determined pathologi- stippled (fine or coarse nonlinear particles); type
cally by FNA (n = 44), thyroidectomy (n = 3), or 2, curvilinear, smooth margin; and type 3, curvi-
both (n = 18). Ultrasonographically guided FNA linear, irregular margin (Figure 1). The lesions
was performed by 2 experienced radiologists were also categorized as arc or rim depending on
(D.Y.Y. and S.K.C) with a 23-gauge needle whether the calcification was limited to part of the
according to a previously described technique.17 lesion border or involved the entire lesion border.
Five or more passes were made through the nod-
ule, and specimens were withdrawn by capillary Statistical Analysis
action. Thirty cases with benign FNA results had We compared the ultrasonographic characteris-
ultrasonographic follow-up (>6 months; mean tics of each thyroid nodule with the histopatho-
period, 12.2 ± 4.2 months; range, 6–30 months). logic findings (benign or malignant). Statistical
During the follow-up period, none of these analysis was performed with the Student t test
patients had any changes in nodule volume or and the χ2 test. P < .05 was considered statistical-
imaging features.

1350 J Ultrasound Med 2007; 26:1349–1355

Yoon et al

ly significant. All statistical analyses were per- Patient demographics and ultrasonographic
formed with commercially available software features of the benign and malignant nodules
(SPSS version 10.0 for Windows; SPSS Inc, are summarized in Table 1. There were no signif-
Chicago, IL). icant differences in age (P = .37) and sex (P = .98)
Results between the patients with malignant and benign
Of the 65 thyroid lesions included in the study, 53 thyroid tumors.
(81.5%) were benign, and 12 (18.5%) were malig-
nant. The histologic type of all malignant thyroid The mean sizes ± SD as determined by maxi-
tumors was papillary carcinoma. mum dimension were 13 ± 8 mm (range, 3–37
mm) for benign nodules and 14.4 ± 6.8 mm
Figure 1. Diagrammatic representation of the types of thyroid (range, 5–28 mm) for malignant nodules. The
peripheral calcification. A, Type 1: stippled calcification. B, Type maximum dimension did not show statistically
2: smooth curvilinear calcification. C, Type 3: irregular curvilin- significant differences between the benign and
ear calcification. malignant nodules (P = .56) when thyroid nod-
A ules smaller than 10 mm in maximum dimen-
sion, 10 to 15 mm, and larger than 15 mm were
B compared.

Furthermore, we wanted to evaluate whether
the risk of malignancy was altered by the type
and extent of calcification described. The preva-
lence of malignancy ranged from 14.3% for type
2 calcification to 25% for type 1 calcification.
Among nodules with rim-type peripheral calcifi-
cation, 7 (21.9%) of 32 had thyroid cancer,
whereas 5 (15.2%) of 33 nodules with arc-type
peripheral calcification had thyroid cancer. No
significant associations were found between the
type (P = .68) and extent (P = .70) of calcification
and histologic differentiation (Figures 2–6).

In addition, we did not find any significant dif-
ferences between malignant and benign thyroid
nodules for lesion shape (P = .81), internal
echogenicity (P = .50), and the presence of the
halo sign (P = .84).

Discussion

Calcification within the thyroid gland is a com-
mon finding on ultrasonography. It has been
reported to be present in 14% to 55% of thyroid
C sonograms.11,15,17–19 Prior studies also revealed a
strong association between ultrasonographically
detected thyroid calcification and malignancy;
those studies reported malignancy rates of 29%
to 59%.12,16,18–20

To our knowledge, however, only 1 article in the
literature described the association between
peripheral calcification within a thyroid nodule
and thyroid malignancy. Taki et al20 correlated
the types of thyroid calcification with pathologic
results. They found 14 lesions with peripheral

J Ultrasound Med 2007; 26:1349–1355 1351

Peripheral Calcification in Thyroid Nodules

Table 1. Differences Between Benign and Malignant Thyroid Lesions

Parameter Benign (n = 53) Malignant (n = 12) P

Sex 6 (11.3) 2 (16.7) NS
Male 47 (88.7) 10 (83.3) NS
Female 54.3 ± 12.7 50.4 ± 14.0 NS

Age, y* 20 (37.7) 3 (25.0) NS
Maximum diameter, mm 18 (34.0) 6 (50.0) NS
15 (28.3) 3 (25.0)
<10 mm NS
10–15 mm 37 (90.2) 9 (81.8) NS
>15 mm 4 (9.8) 2 (18.2)
AP/T ratio† NS
≤1 19 (35.8) 6 (50.5)
>1 13 (24.5) 4 (33.3)
Internal echogenicity 0 (0)
Hypoechoic 2 (3.8) 2 (16.7)
Isoechoic 19 (35.8)
Hyperechoic 7 (58.3)
Invisible 32 (60.4) 5 (41.7)
Halo sign 21 (39.6)
Absent 4 (33.3)
Present 12 (22.6) 4 (33.3)
Type of peripheral calcifications 24 (45.3) 4 (33.3)
Type 1 (stippled) 17 (32.1)
Type 2 (curvilinear, smooth margin) 5 (41.7)
Type 3 (curvilinear, irregular margin) 28 (52.8) 7 (58.3)
Extent of peripheral calcifications 25 (47.2)
Arc
Rim

NS indicates not significant (P > .05). Data in parentheses are percentages.

*Data are mean ± SD.
†Thirteen nodules (12 benign and 1 malignant) were excluded from the analysis because of marked acoustic shadowing.

calcification in 151 surgically resected thyroid The halo sign, a complete or incomplete hypoe-
nodules, and 43% (6 of 14) of those lesions choic rim surrounding peripheral calcification,
proved to be malignant. We evaluated a relatively was identified in 40% of the lesions in this series.
large series of histopathologically proven cases of The histologic correlate of this halo surrounding
thyroid nodules with peripheral calcification. In
our study, 11 (18.5%) of 65 nodules with this type Figure 2. Longitudinal sonogram of the left thyroid lobe from a
of calcification were associated with cancer, and 58-year-old woman with an adenomatous goiter. A hypoechoic
this prevalence was lower than that previously mass (arrows) with peripheral stippled calcification (type 1;
reported by Taki et al (43%).20 arrowheads) surrounded by a hypoechoic halo is shown. Also
shown is another thyroid mass (calipers).
We observed 3 types of peripheral calcification:
the most frequent type was curvilinear with a
smooth margin (type 2, 43.1%), followed by
curvilinear with an irregular margin (type 3,
32.3%), and stippled calcification (type 1, 24.6%).
Four (25%) of the 16 type 1 lesions were malig-
nant, whereas 4 (19%) of the 21 type 3 lesions and
4 (14.3%) of the 28 type 2 lesions were malignant.
However, this trend did not reach statistical sig-
nificance (P = .68). On the basis of these results,
we suggest that the type of peripheral calcifica-
tion cannot be used as a reliable predictor for
malignancy.

1352 J Ultrasound Med 2007; 26:1349–1355

Yoon et al

Figure 3. Longitudinal sonogram of the right thyroid lobe from matous calcification, which can be detected as
a 52-year-old man with an adenomatous goiter. A well-circum- microcalcification on ultrasonography.11,14,22 By
scribed isoechoic mass (calipers) with rim-type smooth curvilin- contrast, dystrophic calcification consists of
ear calcification (type 2) is shown. nonlaminated amorphous deposits in fibrous
tissue septa rather than the epithelium.11,21 This
peripheral calcification is unclear, but it may cor- type of calcification is thought to correspond to
respond to a fibrous capsule, compressed normal coarse calcification shown on ultrasonography,
thyroid tissue, or a viable tumor. A halo sign has which occurs in both benign and malignant con-
been associated with a greater likelihood of benig- ditions.12 Peripheral calcification is one of the
nity in previous studies.3,7,9 In this study, however, patterns of dystrophic calcification located
a slightly greater proportion of malignant nodules around the nodule. It was generally thought to be
had this characteristic than did benign nodules more frequently associated with benignity23;
(41.7% versus 39.6%, respectively). however, cases of papillary thyroid carcinoma
associated with this type of calcification have
Histologically, thyroid calcification is divided been reported.20,24 Pathologic correlation was
into psammomatous and dystrophic types.11 available for all nodules with peripheral calcifi-
Psammomatous calcification consists of lami- cation in our series. Pathologic examination of
nated round calcium deposits in the epitheli- these nodules showed extensive areas of dys-
um.11,21 It is now well accepted that papillary trophic calcification, which were associated or
thyroid carcinoma frequently forms psammo- not associated with malignant cells.

Peripheral calcification in thyroid nodules
often presents a diagnostic dilemma to radiolo-
gists because of its acoustic shadowing and no
visualization of the internal architecture. In our
series, the AP/T ratio and internal echogenicity
could not be assessed in 16 (20%) and 21 (32.3%)
of the 65 thyroid nodules, respectively; in each
case, the only presenting feature was isolated
posterior shadowing due to complete attenua-
tion of the ultrasonic beam. Furthermore, when
ultrasonographically guided FNA is performed

Figure 4. Transverse sonogram of the left thyroid lobe from a Figure 5. Transverse sonogram of the right thyroid lobe from a
63-year-old man with papillary carcinoma. A hypoechoic mass 48-year-old woman with nodular hyperplasia. A hypoechoic
(calipers) with rim-type smooth curvilinear calcification (type 2) is mass (arrows) with arc-type irregular curvilinear calcification
shown. (type 3) surrounded by an irregular hypoechoic halo is shown.

J Ultrasound Med 2007; 26:1349–1355 1353

Peripheral Calcification in Thyroid Nodules

Figure 6. Longitudinal sonogram of the left thyroid lobe from a Therefore, we did not attempt to determine the
58-year-old man with papillary carcinoma. A hypoechoic mass prevalence of peripheral calcification on thyroid
(calipers) with arc-type irregular curvilinear calcification (type 3) ultrasonography in our series. Second, our
surrounded by thin hypoechoic halo (arrows) is shown. The results were limited by the fact that most benign
internal architecture of the mass is not visible because of a conditions were diagnosed on the basis of FNAs.
marked posterior acoustic shadow. Follow-up ultrasonographic evaluation, per-
formed after at least 6 months on patients with
benign cytologic findings, showed no major
changes in all cases. However, the follow-up
period in our study was short for thyroid cancer
(1 year on average).

In conclusion, we found an 18.5% prevalence of
malignancy among thyroid nodules with periph-
eral calcification. In addition, we did not find any
distinct ultrasonographic feature that could dis-
tinguish benign from malignant thyroid nodules.
Further diagnostic evaluation with ultrasono-
graphically guided FNA or ultrasonographic
follow-up should therefore be considered when
assessing any patient with a thyroid mass with
peripheral calcification.

References

on such nodules, poor visualization of the lesion 1. Mazzaferri EL. Management of a solitary thyroid nodule. N
or needle may result in inadequate sampling. Of Engl J Med 1993; 328:553–559.
the 67 ultrasonographically guided FNAs per-
formed on thyroid nodules with peripheral calci- 2. Frates MC, Benson CB, Doubilet PM, Cibas ES, Marqusee
fication in this series, however, 12 procedures E. Can color Doppler sonography aid in the prediction of
yielded cytologic findings that were questionable malignancy of thyroid nodules? J Ultrasound Med 2003;
or malignant, all of which were confirmed at sur- 22:127–131.
gical pathologic examination, whereas 50 proce-
dures yielded benign cytologic findings, and 5 3. Ezzat S, Sarti DA, Cain DR, Braunstein GD. Thyroid inci-
were nondiagnostic. The rate of nondiagnostic dentalomas: prevalence by palpation and ultrasonography.
specimens (7.5% [5/67]) in our study was highly Arch Intern Med 1994; 154:1838–1840.
comparable with those of previous reports,
which revealed nondiagnostic results in 6% to 4. Brander AE, Viikinkoski VP, Nickels JI, Kivisaari LM.
16% of cases.25,26 Importance of thyroid abnormalities detected at US screen-
ing: a 5-year follow-up. Radiology 2000; 215:801–806.
Our study had several limitations based pri-
marily on its retrospective design. First, there 5. Mortenson J, Woolner L, Bennett W. Gross and microscop-
was a possibility of a selection bias because of ic findings in clinically normal thyroid glands. J Clin
the criteria used to select our study population. Endocrinol Metab 1955; 15:1270–1280.
A source of this bias was the dependence on the
original ultrasonography report to detect 6. Cochand-Priollet B, Guillausseau PJ, Chagnon S, et al. The
peripheral calcification of the thyroid gland; the diagnostic value of fine-needle aspiration biopsy under
radiologist reporting on the ultrasonographic ultrasonography in nonfunctional thyroid nodules: a
examination may have described specific pat- prospective study comparing cytologic and histologic find-
terns of calcification or may have only men- ings. Am J Med 1994; 97:152–157.
tioned the presence or absence of calcification.
7. Papini E, Guglielmi R, Bianchini A, et al. Risk of malignancy
in nonpalpable thyroid nodules: predictive value of ultra-
sound and color Doppler features. J Clin Endocrinol Metab
2002; 87:1941–1946.

8. Marqusee E, Benson C, Frates M, et al. Usefulness of ultra-
sonography in the management of nodular thyroid disease.
Ann Intern Med 2000; 133:696–700.

1354 J Ultrasound Med 2007; 26:1349–1355

Yoon et al

9. Kim EK, Park CS, Chung WY, et al. New sonographic crite- 25. Braga M, Cavalcanti TC, Collaco LM. Efficacy of ultra-
ria for recommending fine-needle aspiration biopsy of sound-guided fine-needle aspiration biopsy in the diagno-
nonpalpable solid nodules of the thyroid. AJR Am J sis of complex thyroid nodules. J Clin Endocrinol Metab
Roentgenol 2002; 178:687–691. 2001; 86:4089–4091.

10. Rago T, Vitti P, Chiovato S, et al. Role of conventional ultra- 26. Court-Payen M, Nygaard B, Horn T, et al. US-guided fine-
sonography and color flow Doppler sonography in pre- needle aspiration biopsy of thyroid nodules. Acta Radiol
dicting malignancy in “cold” thyroid nodules. Eur J 2002; 43:131–140.
Endocrinol 1998; 138:41–46.

11. Takashima S, Fukuda H, Nomura N, Kishimoto H, Kim T,
Kobayashi T. Thyroid nodules: re-evaluation with ultra-
sound. J Clin Ultrasound 1995; 23:179–184.

12. Khoo ML, Asa SL, Witterick IJ, Freeman JL. Thyroid calcifi-
cation and its association with thyroid carcinoma. Head
Neck 2002; 24:651–655.

13. Fukatsu H, Makino N, Kodama Y, Ikeda M, Ishigaki T,
Sakuma S. Evaluation of thyroid calcification using com-
puted radiography with image plate. Eur J Radiol 1989;
9:22–28.

14. Komolafe F. Radiological patterns and significance of thy-
roid calcification. Clin Radiol 1981; 32:571–575.

15. Frates MC, Benson CB, Doubilet PM, et al. Likelihood of
thyroid cancer based on sonographic assessment of nod-
ule size and composition [abstract]. In: Radiological Society
of North America Scientific Assembly and Annual Meeting
Program. Oak Brook, IL: Radiological Society of North
America; 2004:395.

16. Kakkos SK, Scopa CD, Chalmoukis AK, et al. Relative risk
of cancer in sonographically detected thyroid nodules with
calcifications. J Clin Ultrasound 2000; 28:347–352.

17. Gharib H. Diagnosis of thyroid nodules by fine needle aspi-
ration biopsy. Curr Opin Endocrinol Diabetes 1996;
3:433–438.

18. Wang N, Xu Y, Ge C, Guo R, Guo K. Association of sono-
graphically detected calcification with thyroid carcinoma.
Head Neck 2006; 28:1077–1083.

19. Seiberling KA, Dutra JC, Grant T, Bajramovic S. Role of
intrathyroidal calcifications detected on ultrasound as a
marker of malignancy. Laryngoscope 2004; 114:1753–
1757.

20. Taki S, Terahata S, Yamashita R, et al. Thyroid calcifications:
sonographic patterns and incidence of cancer. Clin
Imaging 2004; 28:368–371.

21. Klinck GH, Winship T. Psammoma bodies and thyroid can-
cer. Cancer 1959; 12:656–662.

22. Margolin FR, Steinbach HL. Soft tissue roentgenography of
thyroid nodules. AJR Am J Roentgenol 1968; 102:844–
852.

23. Bruneton JN, Livraghi T, Marcy PY, Tramalloni J, Tranquart
F. Thyroid gland. In: Brunetin JN (ed). Medical Radiology:
Radiological Imaging of Endocrine Diseases. Berlin,
Germany: Springer-Verlag; 1999:145–158.

24. Park CH, Rothermel FJ, Judge DM. Unusual calcification in
mixed papillary and follicular carcinoma of the thyroid
gland. Radiology 1976; 119:554.

J Ultrasound Med 2007; 26:1349–1355 1355