mr findings of struma ovarii
TRANSCRIPT
MR findings of struma ovarii
Jong Chul Kim*, Sung Soo Kim, Jin Yong Park
Department of Diagnostic Radiology, Chungnam National University Hospital, 640 Dacesa-dong, Jung-ku, Taejeon 301-721, South Korea
Received 20 September 1999; accepted 1 March 2000
Abstract
This study was performed to characterize MR findings of struma ovarii. In 10 patients, T1- and fast spin echo T2-weighted MR images
were obtained in the axial, coronal, and sagittal planes using 1.5 T MR units, and they were retrospectively evaluated for the site, size,
components, signal intensity, and contrast enhancement. MR images showed a unilateral complex mass with a multilobulated surface and
thickened septa, corresponding pathologically to thyroid follicles and the stroma. Cystic portions had variable signal intensities on T1- and
T2-weighted images. The contents of cystic components showing low signal intensities both on T1- and T2-weighted images were viscid
gelatinous materials (n = 4). Solid portions were relatively well-enhanced. In conclusion, struma ovarii has some characteristic MR
appearance of a multilobulated complex mass with thickened septa, multiple cysts of variable signal intensities, and enhancing solid
components. D 2000 Elsevier Science Inc. All rights reserved.
Keywords: Struma ovarii; MR images; Multilobulated; Ovary neoplasms; Ovary; Cysts; Ovary; MR
The mature cystic teratomas constitute the majority of
ovarian germ cell tumors [1,2], and thyroid tissues are
found in 5±15% of these tumors on pathologic examina-
tions [2±4]. Struma ovarii is a rare ovarian tumor com-
posed solely or predominantly of thyroid tissue or the
tumor in which hyperthyroidism results from the ovarian
thyroid tissue [1,2,5±8]. Pure struma ovarii with micro-
scopic or macroscopic thyroid tissue constitutes 2% of all
germ cell tumors of the ovary [4] and 2.7% of all ovarian
teratomas [3]. Preoperative clinical or radiologic diagnosis
of struma ovarii, however, is very difficult.
There are limited numbers of radiologic reports of
struma ovarii [2,4,7,9±16], and the majority of these are
case reports [2,7,11,13±15]. Even though magnetic reso-
nance (MR) imaging is known to be useful in the detection
and characterization of adnexal masses, to our knowledge,
there are only several reports concerning the MR findings
of this rare tumor [4,7,11,16]
The purpose of this study was to evaluate whether there
were any characteristic imaging findings in struma ovarii
through the retrospective analysis of MR findings.
1. Materials and methods
During the recent 11 years, 10 cases of pathologically
proven struma ovarii after salpingo-oophorectomy were
retrospectively identified in our hospital and its affiliated
hospitals by reviewing medical records.
The age of the 10 patients ranged from 37 to 73 years
(average, 51). Of the 10 patients, nine women complained
of palpable lower abdominal masses, and one woman was
admitted due to symptoms and signs of hyperthyroidism
such as facial flushing, palpitation, etc. Thyroid function
test after admission was positive in the latter patient.
GE Signa 1.5 T MR unit (General Electric Medical
Systems, Milwaukee, WI, USA) was used to obtain
T1-weighted images (TR/TEeff 410±660/11±17 ms), fast
spin echo T2-weighted images (TR/TE 3200±6000/85±
132 ms; echo train length = 8, 12) and gadolinium-
enhanced T1-weighted images in the axial, coronal, and
sagittal planes. Contrast-enhanced images were obtained
after intravenous injection of gadopentetate dimeglumine
(Magnevist, Schering, Berlin, Germany, 0.1 mmol/kg). In
3 of 10 patients, fat saturation techniques were performed
to differentiate whether hyperintense areas in the tumors
on T1-weighted images were due to hemorrhage, fat, and/
or fluid of high protein content. The sequence parameters
were as follows: section thickness = 8±10 mm; field of* Corresponding author. Tel.: +82-42-220-7835; fax: +82-42-253-0061.
E-mail address: [email protected] (J.C. Kim).
0899-7071/00/$ ± see front matter D 2000 Elsevier Science Inc. All rights reserved.
PII: S0 8 9 9 - 7 0 7 1 ( 0 0 ) 0 0 1 58 - 3
Journal of Clinical Imaging 24 (2000) 28± 33
view = 26 � 26 cm; matrix = 512 � 224; numbers of
excitation = 1.5±4.
MR findings of struma ovarii were retrospectively
evaluated for the site, size, components (cystic, solid,
mixed, or complex), signal intensity, and absence or degree
of contrast enhancement, by three experienced radiologists
who reached a consensus on the analysis. These radiologic
findings were compared with pathologic ones.
2. Results
MR findings of our 10 patients were summarized in
Table 1.
All tumors in 10 patients were unilateral (right, five; left,
five), and the maximal tumor size ranged from 5 to 16 cm
(mean, 9.9 cm). In one patient, the symptoms and signs of
hyperthyroidism such as facial flushing, palpitation, etc.,
subsided after the ipsilateral salpingo-oophorectomy.
MR images showed a unilateral complex mass of
multiple cysts and some solid components with a multi-
lobulated surface and thickened septa in all patients. In 2
of 10 patients, most of struma ovarii appeared to be
consisted of only multiseptated cysts. The maximal thick-
ness of the septa was 3±10 mm in each patient on
contrast-enhanced T1-weighted image, and that of the
peripheral cyst wall was 7±15 mm. The cystic portions
of the tumors were variable in size (range of maximal
diameter, 0.5±6.5 cm), shape (round, oval, semilunar,
irregular, triangular, lobulated, etc.), numbers (range 3±
40), and location (peripheral, central, eccentric, even dis-
tribution). Their cystic portions showed variable signal
intensities without contrast enhancement on T1-weighted
images, and were mainly high with different signal in-
tensities of fast spin echo T2-weighted images (Figs. 1±3).
In 4 of 10 tumors, some cystic portions of the tumors,
either scattered (n = 3) (Fig. 2) or peripherally located (n =
1) (Fig. 3), were hyperintense both on T1- and T2-
weighted images. Fat saturated MR images in three of
these four patients revealed the non-fatty nature of the
cystic contents of the masses, because the hyperintense
areas on T1-weighted images were not suppressed. These
hyperintense cystic areas on T1-weighted images were
pathologically proved to be due to hemorrhagic products
(n = 1) and viscous colloid components (n = 3) (Figs. 2
and 3). There are four cases with some cystic components
showing low signal intensities both on T1- and fast spin
echo T2-weighted images (Figs. 1±3). The contents of
these cysts were also viscid gelatinous materials. Both the
viscid fluid and viscid gelatinous materials were histolo-
gically eosinophilic, similar to the colloid in the normal
thyroid follicles.
Table 1
MR findings of 10 cases of pathologically proven struma ovarii
T1WI T2WI Gd-T1WI
No. Age Site Size (cm) Cystic Solid Cystic Solid Cystic Solid
(1) 73 R 10 � 9 � 7 iso iso hyper sl. hyper UE ME
hypo hypo
(2) 42 R 7 � 6 � 5 iso iso hyper sl. hyper UE IE
hypo hyper
(3) 65 L 10 � 8 � 5 iso iso hyper sl. hyper UE IE
hyper hyper iso
hypo hypo
(4) 39 L 8 � 10 � 7 hyper iso UE ME
iso hypo
hyper hyper
(5) 57 R 12 � 11 � 9 hypo iso hyper iso UE IE
iso hypo
hyper hyper
(6) 37 L 7 � 10 � 10 hypo iso hypo iso UE ME
iso hyper
(7) 53 L 16 � 13 � 9 hypo iso hyper iso UE IE
iso sl. hyper hyper
(8) 40 R 5 � 4 � 2 hypo iso hyper iso UE IE
iso hyper
hyper hyper
(9) 63 L 8 � 5 � 3 hyper iso hyper sl. hyper UE IE
hypo hyper
(10) 45 R 6 � 11 � 9 hyper iso hyper iso UE IE
hypo hypo
Note: MR, magnetic resonance; T1WI, T1-weighted image; T2WI, T2-weighted image; Gd-T1WI, gadolinium-DTPA-enhanced T1-weighted image with
or without fat saturation; hyper, hyperintense; hypo, hypointense; iso, isointense with surrounding pelvic muscles; sl., slightly; UE, unenhanced; IE, intense
contrast enhancement; ME, moderate enhancement.
J.C. Kim et al. / Journal of Clinical Imaging 24 (2000) 28±33 29
Solid portions or septa of the tumors were isointense
relative to the adjacent pelvic muscles on T1-weighted
images, iso- or hyperintense on T2-weighted images, and
intensely (n = 7) or moderately (n = 3) enhanced on
postcontrast T1-weighted images (Figs. 1±3). Most of
these solid portions were scattered throughout the tumors
or abutted on the cystic areas with variable size (0.7±5
cm) and irregular shapes.
The surgical specimens revealed that the 10 tumors
were complex multilobulated masses with multiple cysts,
thickened septa, and solid portions. Adult thyroid tissues
were found in all patients on microscopy. Thyroid follicles
were filled with high proteinaceous gelatinous fluid of
eosinophilic colloid, and the solid portions of the tumors
consisted of the thyroid tissues composed of multiple
follicles and stroma containing abundant blood vessels
and fibrous tissue (Fig. 1d). There were no ascites and
no evidence of malignancy in the pelvic lymphadenopathy.
3. Discussion
Ovarian teratomas are classified as the mature, imma-
ture, and monodermal or highly specialized ones. Struma
ovarii is one of the monodermal or highly specialized type,
together with a carcinoid [17]. This tumor is uncommon,
constituting only 2.7% of all ovarian teratomas [3,18,19]
and only 0.3% of all ovarian tumors [7,18]. It has been
reported that about 85% of patients with struma ovarii has
been present before menopause [2,20,21]. In our study, the
age range of 10 patients including four postmenopausal
women were 37±73 years (average, 51). This age range in
struma ovarii patients was slightly higher than the reported
prevalent age range of 20±44 years in cystic teratomas
[22,23]. In our study, all tumors occurred unilaterally,
similar to the other reports [24].
Clinically, most cases of struma ovarii are silent or
present with nonspecific symptoms similar to the other
Fig. 1. Sagittal pelvic MR images of right struma ovarii in a 73-year-old woman (No. 1). (a) T1-weighted image reveals a large complex pelvic mass (black
arrows). The signal intensity of the mass is similar to that of the uterus (u), and the mass also contains some hypointense areas (white arrowheads). (b) Fast spin
echo T2-weighted image demonstrates a multiloculated tumor of hyperintense cystic areas with some isointense solid portions and septa (black solid arrows),
and hypointense uterine myomas (black open arrows). (c) Gadolinium-enhanced sagittal T1-weighted image reveals non-enhancing cystic contents of variable
signal intensities with moderately enhancing solid portions and septa (white arrowheads) of the tumor. (d) Pathologic microscopy revealed that the mass was
composed of thyroid tissue containing multiple follicles filled with eosinophilic viscous colloid.
J.C. Kim et al. / Journal of Clinical Imaging 24 (2000) 28±3330
benign ovarian tumors. However, if the mature thyroid
tissue in this tumor is hyperfunctioning, symptoms and
signs of hyperthyroidism may be present [2,7,25]. In one
of our 10 patients, the symptoms and signs of hyperthyr-
oidism due to struma ovarii subsided dramatically after
surgical removal of struma ovarii. In a small number of
cases, ascites and pleural effusion may be found (so-
called pseudo-Meigs' syndrome) [3,4], but our patients
had no evidence of ascites either on MR imaging or
on pathology.
Most of struma ovarii are benign and can be treated
by surgical resection of the diseased ovary with or with-
out excision of the ipsilateral fallopian tube. Malignant
foci, however, were reported to be found in 5±10% of
struma ovarii [2,24]. In our study, all patients were
treated with salpingo-oophorectomy and total abdominal
hysterectomy. Fortunately, there have been no clinical or
pathological evidences of malignancy or recurrence dur-
ing the postoperative follow-up period (from 7 months to
12 years).
There are not many radiologic reports of struma ovarii
[2,4,7,9 ± 16], and most of these are case reports
[2,7,11,13±15]. In addition, to our knowledge, there are
very few reports describing the MR findings of struma
ovarii [4,7,11,16]. Matsumoto et al. [7] reported one case
of struma ovarii with CT and MR imaging, Dohke et al.
[11] reported two cases, Yamashita et al. [4] described
MR appearances of six tumors, and Joja et al. [16]
reported the appearance on MR images in 12 cases of
struma ovarii.
In our study, similar to other reports [4,7,11,16,25±28],
all tumors were complex multilobulated masses with thick-
ened septa both on MR imaging and on pathology. On
pathology, most of their cystic portions were filled with high
proteinaceous gelatinous fluid of eosinophilic colloid, and
their solid portions consisted of the thyroid tissues and
stroma containing abundant blood vessels and fibrous
tissue. So, multilocularity of struma ovarii may be consid-
ered to be due to multiple large thyroid follicles. The
multiple cysts of struma ovarii may have variable signal
Fig. 2. Axial pelvic MR images of left struma ovarii in a 65-year-old woman (No. 3). (a) T1-weighted image reveals a large complex pelvic mass of variable
signal intensities behind the urinary bladder. (b) Fat saturated T1-weighted image reveals that the hyperintense areas on T1-weighted image are not fatty tissues,
because those areas are not suppressed as subcutaneous or pelvic fat. (c) Fast spin echo T2-weighted image demonstrates a complex multiloculated mass
with hyper- or hypointense cystic areas and isointense solid portions (arrows) relative to the uterine muscle layer (u). (d) Fat saturated gadolinium-enhanced
T1-weighted image shows non-enhancing cystic contents and intensely enhancing solid portions and septa of the tumor.
J.C. Kim et al. / Journal of Clinical Imaging 24 (2000) 28±33 31
intensities depending on the amount, density, or concentra-
tion of viscous proteinaceous fluid in thyroid follicles. In 4
of our 10 cases, hyperintense areas on T1-weighted images
without chemical shift artefact on T2-weighted images were
pathologically confirmed to be due to hemorrhage (n = 1)
and colloid (n = 3) on pathology. The contents of cystic
components showing low signal intensities both on T1- and
fast spin echo T2-weighted images in four cases were also
viscid gelatinous materials. Dillon et al. [29] and Som et al.
[30] reported that viscid fluid showed low signal intensity
on T2-weighted images and high signal intensity on
T1-weighted images. But they also said that as the vicosity
of the fluid increases, the signal intensity on T2-weighted
images decreased and became a signal void and that on T1-
weighted images also decreased. Amr and Hassan [28]
reported that struma ovarii contained thick gelatinous ma-
terial in the tumors. Therefore, the characteristic very low
signal intensities on both T1- and T2-weighted images may
reflect very viscid material such as triiodo thyronine, thyr-
oxine, thyroglobulin, thyroid hormones, etc., within the
tumor [16].
Solid portions or septa of struma ovarii were isointense
relative to the adjacent muscles on T1-weighted images,
iso- or hyperintense on T2-weighted images, and intensely
enhanced on postcontract T1-weighted images. The strong
enhancement of these solid portions can be explained due
to the thyroid tissues and stroma of abundant blood vessels
and fibrous tissue.
MR findings of such complex adnexal masses are found
not only in struma ovarii but also in many other adnexal
masses. These include cystadenoma, cystadenocarcinoma,
endometrioma, mature cystic teratoma, ovarian metastasis,
or tuboovarian abscess [4,10]. Intense or moderate enhance-
ment of solid portions of the tumor may increase the
possibility of struma ovarii rather than the mucinous or
serous cystadenoma, cystic teratoma without fatty tissue, or
ovarian metastasis. Clinical findings of fever, tenderness,
leukocytosis, etc., due to inflammation will favor the
diagnosis of the tuboovarian abscess. Some hyperintense
areas of cystic portions of the tumor which cannot be
suppressed by fat saturation technique on T1-weighted
images and which do not have chemical shift artefact on
Fig. 3. Pelvic MR images of right struma ovarii in a 57-year-old woman (No. 5). (a) Sagittal T1-weighted image demonstrates a large multiloculated mainly
cystic pelvic mass of variable signal intensities. (b) Fast spin echo axial T2-weighted image demonstrates that the mass is composed of hyperintense fluid and
isointense solid portions and septa (arrows) relative to the pelvic muscles. (c) Gadolinium-enhanced axial T1-weighted image shows non-enhancing cystic
contents and intensely enhancing solid portions and septa (white arrowheads) of the tumor.
J.C. Kim et al. / Journal of Clinical Imaging 24 (2000) 28±3332
T2-weighted images may exclude the diagnosis of the pure
teratoma with fatty tissue. Unilaterality may increase the
probability of struma ovarii compared with metastatic tu-
mors to the ovary. Shading artefact due to variable hemor-
rhagic products of different stage may favor the diagnosis of
endometrioma. The complex mass associated with hy-
perthyroidism, ascites, and/or pleural effusion may increase
the possibility of struma ovarii.
Our study has a critical limitation that the sample size is
too small to be statistically significant.
In conclusion, when MR imaging shows a unilateral
complex adnexal mass with a multilobulated surface and
thickened septa and when this mass is composed of multiple
cysts of variable signal intensities and intensely or mode-
rately enhancing solid components, struma ovarii should be
included in the differential diagnosis with high probability.
References
[1] Fox H, Langley FA. Tumors of the Ovary. London: Heinemann, 1976.
[2] March DE, Desai AG, Park CH, Hendricks PJ, Davis PS. Struma
ovarii: hyperthyroidism in a postmenopausal woman. J Nucl Med
1988;29:263± 5.
[3] Talerman A. Germ cell tumors of the ovary. In: Kurman RJ, editor.
Blaustein's Pathology of the Female Genital Tract, 3rd edn. New
York, NY: Springer-Verlag, 1987. pp. 687± 721.
[4] Yamashita Y, Hatanaka Y, Takahashi M, Miyazaki K, Okamura H.
Struma ovarii: MR appearances. Abdom Imaging 1997;22:100±2.
[5] Thomas RD, Batty VB. Metastatic malignant struma ovarii. Clin Nucl
Med 1992;17:577± 8.
[6] Radecki PD, McKrisky PJ. Female pelvis: neoplastic and nonneoplas-
tic masses. In: Friedman AC, Radecki PD, Lev-Toaff AS, Hilpert PL,
editors. Clinical Pelvic Imaging. St. Louis: C.V. Mosby, 1990. pp.
197± 204.
[7] Matsumoto F, Yoshioka H, Hamada T, Ishida O, Noda K. Struma ovarii:
CT and MR findings. J Comput Assist Tomogr 1990;14(2):310 ±12.
[8] Ayhan A, Yanik F, Tuncer R, Tuncer ZS, Ruacan S. Struma ovarii. Int
J Gynaecol Obstet 1993;42:143± 6.
[9] Kozuka Y, Tamaishi Y, Yamawaki T, Sugiyama Y. A case of struma
carcinoid. Obstet Gynecol 1983;50:1991± 4.
[10] Hahn ST, Park SH, Bahk YW, Chung SK. Struma ovarii simulating a
teratodermoid cyst: computed tomographic findings in one case. Radi-
ology 1991;31:89± 91.
[11] Dohke M, Watanabe Y, Takahashi A, et al. Struma ovarii: MR find-
ings. J Comput Assist Tomogr 1997;21(2):265± 7.
[12] Zale Y, Caspi B, Tepper R. Doppler flow characteristics of dermoid
cysts: unique appearance of struma ovarii. J Ultrasound Med
1997;16(5):355±8.
[13] Thomas RD, Batty VB. Metastatic malignant struma ovarii: two case
reports. Clin Nucl Med 1992;17:577± 8.
[14] Lieberman G, Buscombe JR, Hilson AJ, Thakrar D, MacLean A. The
detection of struma ovarii in two patients by radioimmunoscintigra-
phy. Am J Obstet Gynecol 1998;179:262± 3.
[15] Bayot MR, Chopra IJ. Coexistence of struma ovarii and Graves' dis-
ease. Thyroid 1995;5:469±71.
[16] Joja I, Asakawa T, Mitsumori A, et al. Struma ovarii: appearance on
MR images. Abdom Imaging 1998;23:652± 6.
[17] Cotran RS, Kumar V, Robbins SL. Female genital tract. In: Cotran RS,
Kumar V, Robbins SL, editors. Robbins Pathologic Basis of Disease,
4th edn. Philadelphia: Saunders, 1989. pp. 1127±80.
[18] Stevens SK. The adnexa. In: Higgins CB, Hricak H, Helms CA,
editors. Magnetic Resonance Imaging of the Body, 2nd edn. New
York: Raven Press, 1992. pp. 875±7.
[19] Gusberg SB, Danforth DN. Clinical significance of struma ovarii. Am
J Obstet Gynecol 1944;48:532± 42.
[20] Sciarra JJ. Benign neoplasm of the ovary. In: Bird CC, McElin TW,
Victor TA, editors. Gynecology and Obstetrics (Vol. 12). Philadelphia:
JB Lippincott, 1988.
[21] Fox JF, Clement KW. Functioning struma ovarii. Ann Surg 1951;133:
253± 4.
[22] Griffiths CT, Berkowitz RS. The ovary. In: Kistner RW, editor. Gy-
necology: Principles and Practice. St. Louis: Mosby year book, 1986.
pp. 173± 5.
[23] Scoutt LM, McCarthy SM. Female pelvis. In: Stark DD, Bradley WG,
editors. Magnetic Resonance Imaging, 2nd edn. St. Louis: Mosby year
book, 1992. pp. 1973±6.
[24] Marcus CC, Marcus SL. Struma ovarii: a report of 7 cases and a
review of the subject. Am J Obstet Gynecol 1961;81:752± 61.
[25] Szyfelbein WM, Young RH, Scully RE. Cystic struma ovarii. A fre-
quently unrecognized tumor. A report of 20 cases. Am J Surg Pathol
1994;18:785±8.
[26] Smith FG. Pathology and physiology of struma ovarii. Arch Surg
1946;53:603±26.
[27] Szyfelbein WM, Young RH, Scully RE. Struma ovarii simulating
ovarian tumors of other types. A report of 30 cases. Am J Surg
Pathol 1995;19:21±9.
[28] Amr SS, Hassan AA. Struma ovarii with pseudo-Meigs' syndrome:
report of a case and review of the literature. Eur J Obstet Gynecol
Reprod Biol 1994;55:205± 8.
[29] Dillon WP, Som PM, Fullerton GD. Hypointense MR signal in chroni-
cally inspissated sinonasal secretions. Radiology 1990;51:396± 8.
[30] Som PM, Dillon WP, Fullerton GD, Zimmerman RA, Rajagopalan
Z, Maron Z. Chronically obstructed sinonasal secretions on T1 and
T2 shortening. Radiology 1989;172:515± 20.
J.C. Kim et al. / Journal of Clinical Imaging 24 (2000) 28±33 33