DOI:10.2214/AJR.07.3974
AJR 2008; 191:1366-1370
© American Roentgen Ray Society
Susceptibility-Weighted MRI of Endometrioma: Preliminary Results
Mayumi Takeuchi1,
Kenji Matsuzaki and
Hiromu Nishitani
1 All authors: Department of Radiology, University of Tokushima, 3-18-15,
Kuramoto-cho, Tokushima, 7708503, Japan.
Received March 9, 2008;
accepted after revision May 10, 2008.
Address correspondence to M. Takeuchi
(mayumi{at}clin.med.tokushima-u.ac.jp).
Abstract
OBJECTIVE. Susceptibility-weighted MRI combines magnitude and phase
information from fully velocity-compensated gradient-echo sequences and
depicts as signal voids the susceptibility effects caused by local
inhomogeneity of the magnetic field. Our objective, based on MRI of 60
pathologically proven ovarian cystic lesions, including 42 endometriomas and
18 nonendometrial cysts, was to evaluate hemosiderin deposition within the
walls of endometriomas on susceptibility-weighted MR images. Two radiologists
blinded to the final diagnosis retrospectively reviewed the images in
consensus.
CONCLUSION. On susceptibility-weighted MR images, punctate or curved
linear signal voids along the cyst wall were observed in 39 endometriomas
(92.9%) and in no nonendometrial cysts. The signal voids were more prominent
on 3-T than on 1.5-T images, reflecting the higher sensitivity of 3-T MRI to
magnetic susceptibility effects. Thirty-two endometriomas (76.2%) met
definitive MRI criteria, that is, hyperintensity on T1-weighted images and
hypointensity on T2-weighted images, and 41 endometriomas (97.6%) were
correctly diagnosed with susceptibility-weighted MRI.
Keywords: endometrioma • MRI • susceptibility-weighted imaging
Introduction
Endometriosis is a common chronic disease affecting women of reproductive
age [1]. Repeated intracystic
bleeding and rupture of endometriomas can cause severe pelvic adhesions and
infertility, so early diagnosis is important for adequate treatment
[1]. Although sonography is the
first choice of imaging examination, MRI is more specific than sonography in
the diagnosis of endometrioma and can be used as a secondary imaging
examination
[2–4].
Multiple hyperintense cysts on T1-weighted images (multiplicity) and a
hyperintense cyst on T1-weighted images that is hypointense on T2-weighted
images (shading) are definite MRI signs of endometrioma, having high
sensitivity (90%), specificity (98%), and accuracy (96%)
[3]. Some endometriomas,
however, do not meet these criteria, and diagnosis based on conventional MRI
findings can be difficult
[3].
Most endometriomas contain chocolate-colored hemorrhagic material formed by
recurrent cyclical bleeding from ectopic endometrial tissue and can cause
diagnostic shading on T2-weighted images
[2–4].
Some endometriomas, however, are filled with watery fluid on gross cut
sections and may not exhibit typical MRI findings
[3]. Togashi et al.
[3] evaluated MRI images of 86
endometriomas and found that six lesions (7%) with predominantly low signal
intensity on T1-weighted images were misdiagnosed as other adnexal masses.
Shading on T2-weighted images was found in 55 lesions (64%)
[3]. The cyst walls of
endometriomas usually are thick and fibrotic. Within the wall, a cluster of
hemosiderin-laden macrophages are commonly found at histopathologic
examination [1,
5,
6]. To diagnose endometrioma
with MRI, visualization of hemosiderin deposition in the cyst wall may be
helpful, especially in the diagnosis of endometriomas that do not exhibit
typical MRI findings.
Susceptibility-weighted MRI is a relatively new MRI technique that
maximizes sensitivity to susceptibility effects and has exquisite sensitivity
to blood products such as hemosiderin and deoxyhemoglobin
[7–9].
Susceptibility-weighted imaging combines magnitude and phase information from
fully velocity-compensated gradient-echo sequences. The magnetic
susceptibility effects generated by local inhomogeneity of the magnetic field
caused by hemosiderin or deoxyhemoglobin are visualized as signal voids
[7–9].
This sequence is more sensitive to the susceptibility difference between
tissues than is conventional T2*-weighted imaging
[7–9].
The technique has been used in imaging of the CNS and in body imaging
[7–9].
To our knowledge, there has been no report of susceptibility-weighted imaging
of endometriosis of the female pelvis. In this study we retrospectively
evaluated hemosiderin deposition in the walls of endometriomas by using
susceptibility-weighted imaging. We also compared the imaging findings of
susceptibility-weighted imaging at both 1.5 T and 3 T.
Materials and Methods
We started using susceptibility-weighted imaging for MRI of the female
pelvis in July 2006. A retrospective review of the gynecologic surgical
database at our institution from July 2006 to December 2007 yielded the
records of 61 patients who had undergone MRI examinations that included
susceptibility-weighted imaging before resection of ovarian cystic masses.
Sixty-one surgically proven ovarian cystic lesions, including 43 endometriomas
in 37 women (mean age, 37 years; range, 23–67 years) and 18
nonendometrial benign cystic masses (six simple cysts, two serous
cystadenomas, seven mucinous cystadenomas, two cystadenofibromas, and one
degenerated fibroma) in 18 women (mean age, 50 years; range 27–80 years)
were evaluated. Our institutional review board does not require approval for
this type of retrospective study.
Among the 61 lesions, 10 lesions (nine endometriomas, one nonendometrial
benign cystic mass) in 10 patients were evaluated at both 3 T and 1.5 T
continuously on the same day in a prospective comparison study. Institutional
review board approval was obtained for the prospective portion of the study,
and informed consent was obtained from all 10 patients. Twenty-four lesions
(13 endometriomas, 11 nonendometrial benign cystic masses) in 22 patients were
evaluated at 1.5 T, and 27 lesions (21 endo metriomas, six nonendometrial
benign cystic masses) in 22 patients were evaluated at 3 T.
Fast spin-echo T2-weighted images (TR/TE,
4,000–9,000/98.1–105.3) and fat-saturated spinecho T1-weighted
images (466.7–700/7.6–9.6) were obtained for all patients with
both a 1.5-T superconducting MRI system (Signa Excite HD, GE Healthcare) and a
3-T superconducting MRI system (Signa 3T HD, GE Healthcare) with eight-channel
body-array torso coils. Susceptibility-weighted images consisting of both
magnitude and phase images from 2D fast spoiled gradient-recalled echo (FSPGR)
acquisition in the steady state (650–700/30; flip angle,
15–20°; matrix size, 288 x 192; field of view, 28 x 28
cm; section thickness, 8 mm; section gap, 1 mm; total imaging time, 4 minutes
34 seconds) were obtained for all patients. To enhance the visibility of
signal voids caused by the magnetic susceptibility effects, postprocessing was
applied to the magnitude images multiplied with a phase mask generated from
the filtered phase data with software provided by the manufacturer. Sequence
parameters at 3 T and 1.5 T were almost identical except for flip angle
(15° at 3 T, 20° at 1.5 T).
Two radiologists with 18 and 9 years of experience in body MRI
qualitatively evaluated the images for the presence of signal voids along the
cyst wall on susceptibility-weighted images. They were blinded to the
histopathologic and clinical diagnoses of the lesions and looked at all images
together. The signal intensity of the lesions on T1- and T2-weighted images
and the presence of shading on T2-weighted images also were assessed.
Agreement between the two radiologists was reached in consensus after careful
individual evaluation. We applied the definite criteria for endometrioma as
"a hyperintense cyst on T1-weighted images exhibiting hypointensity on
T2-weighted images (shading)"
[3]. We compared the imaging
findings on susceptibility-weighted images at both 1.5 T and 3 T for the 10
lesions. The sizes of signal voids along the walls of endometriomas and
degrees of susceptibility artifacts on susceptibility-weighted images were
visually compared at both 1.5 T and 3 T. The radiologists were blinded to
whether images were obtained at 3 T or 1.5 T.
Results
In only one of the 61 cases of endometrioma were susceptibility-weighted
images suboptimal. In that case, in which images were obtained at both 1.5 T
and 3 T, prominent susceptibility artifacts caused by colonic gas extended to
the lesion. Table 1 summarizes
the signal intensities of the other 60 lesions on T1- and T2-weighted images
and the presence of signal voids along the cyst walls on
susceptibility-weighted images. Punctate or curved linear signal voids along
the cyst wall were visualized on susceptibility-weighted images of 39 of 42
endometriomas (92.9%) (Figs.
1A,
1B,
2,
3A,
3B,
3C,
3D) and on no images of 18
nonendometrial benign cystic masses (Fig.
4A,
4B,
4C). Two of the other three
endometriomas were completely hypointense on susceptibility-weighted images
and appeared as shading on T2-weighted images (Fig.
5A,
5B). One lesion that had high
signal intensity both on T1- and T2-weighted images lacked signal voids along
the wall on susceptibility-weighted images.
View larger version (139K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —46-year-old woman with right-sided ovarian endometrioma.
Axial T2*-weighted 1.5-T MR image (TR/TE, 650/30; flip angle, 20°) shows
combination of punctate and curved linear signal voids (arrowheads)
along cyst wall.
|
|
View larger version (153K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B —46-year-old woman with right-sided ovarian endometrioma.
Axial susceptibility-weighted 1.5-T MR image (650/30; flip angle, 20°)
shows signal voids (arrowheads) more prominent than in A.
|
|
View larger version (119K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2 —40-year-old woman with left-sided ovarian endometrioma.
Lesion had high signal intensity on both T1- and T2-weighted images. Axial
susceptibility-weighted 1.5-T MR image (TR/TE, 700/30; flip angle, 20°)
shows curved linear signal voids (arrowheads) along cyst wall.
|
|
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —47-year-old woman with right-sided ovarian endometrioma.
Axial fat-saturated spin-echo T1-weighted MR image (TR/TE, 600/7.9) shows
predominantly hypointense bilocular cystic mass (arrows). Small
hyperintense areas (arrowheads) suggest presence of hemorrhagic foci
within cyst wall.
|
|
View larger version (124K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3C —47-year-old woman with right-sided ovarian endometrioma.
Axial susceptibility-weighted 3-T MR image (700/30; flip angle, 15°)
(C) shows punctate signal voids (arrows) along cyst wall more
prominently than does axial susceptibility-weighted 1.5-T image (700/30; flip
angle, 20°) (D). Susceptibility artifacts (arrowheads)
caused by rectal gas also are more prominent in C.
|
|
View larger version (133K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3D —47-year-old woman with right-sided ovarian endometrioma.
Axial susceptibility-weighted 3-T MR image (700/30; flip angle, 15°)
(C) shows punctate signal voids (arrows) along cyst wall more
prominently than does axial susceptibility-weighted 1.5-T image (700/30; flip
angle, 20°) (D). Susceptibility artifacts (arrowheads)
caused by rectal gas also are more prominent in C.
|
|
View larger version (137K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A —37-year-old woman with right-sided ovarian mucinous
cystadenoma. Axial fat-saturated spin-echo T1-weighted (TR/TE, 466.7/7.6)
(A) and fast spin-echo T2-weighted (4,000/99.3) (B) MR images
show multilocular cystic mass (arrows) with varying signal
intensities. U = uterus.
|
|
View larger version (144K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B —37-year-old woman with right-sided ovarian mucinous
cystadenoma. Axial fat-saturated spin-echo T1-weighted (TR/TE, 466.7/7.6)
(A) and fast spin-echo T2-weighted (4,000/99.3) (B) MR images
show multilocular cystic mass (arrows) with varying signal
intensities. U = uterus.
|
|
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C —37-year-old woman with right-sided ovarian mucinous
cystadenoma. Axial susceptibility-weighted 1.5-T MR image (675/30; flip angle,
20°) shows no signal voids along cyst wall (arrows). Decrease in
signal intensity (asterisk) due to susceptibility artifact caused by
adjacent intestinal gas is evident.
|
|
View larger version (129K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B —39-year-old woman with left-sided ovarian endometrioma. Axial
susceptibility-weighted 3-T MR image (700/30; flip angle, 15°) shows
completely hypointense mass (arrow). Presence of signal voids along
cyst wall cannot be evaluated. Susceptibility artifacts caused by intestinal
gas are prominent.
|
|
Thirty-two of the 42 endometriomas (76.2%) met the definitive MRI criteria,
that is, a cyst hyperintense on T1-weighted images and hypointense on
T2-weighted images. The other 10 lesions (23.8%) did not meet the criteria on
conventional MR images. One of the 10 lesions had low intensity on T1-weighted
images and high intensity on T2-weighted images, and nine lesions had high
intensity both on T1- and T2-weighted images. Signal voids along the cyst wall
on susceptibility-weighted images were visualized in nine of the 10
endometriomas that did not meet the criteria on conventional MR images (Figs.
2 and
3A,
3B,
3C,
3D). With the addition of
susceptibility-weighted imaging to conventional MRI, 41 of 42 endometriomas
(97.6%) were diagnosed correctly.
None of the 18 nonendometrial cystic masses met the criteria for
endometrioma on conventional MR images. Two of these masses, however, had high
signal intensity on T1-weighted images and contained hemorrhagic fluid found
at surgery. Neither lesion had signal voids along the wall on
susceptibility-weighted images. In all eight endometriomas evaluated with both
3- and 1.5-T MRI, the signal voids along the wall on susceptibility-weighted
images were more apparent at 3 T than at 1.5 T (Fig.
3A,
3B,
3C,
3D). On susceptibility-weighted
images, the size of signal voids in all eight endometriomas was larger at 3 T;
however, there was no signal void, which was observed only at 3 T. The
susceptibility artifacts caused by intestinal gas were more prominent on
susceptibility-weighted 3-T images (Fig.
3A,
3B,
3C,
3D). This tendency was
consistently observed in all of these patients. On 3-T images, the artifacts
extended to the lesion of evaluation in two patients.
Discussion
Susceptibility-weighted imaging depicts magnetic susceptibility effects as
signal voids and is more sensitive than conventional T2*-weighted MRI
[7–9].
Because of its exquisite sensitivity to blood products such as hemosiderin and
deoxyhemoglobin, in neuroimaging susceptibility-weighted MRI can yield useful
information about acute stroke by depicting evidence of acute hemorrhage,
intravascular clots, and previous microbleeds
[7–9].
In gynecologic imaging, Yoshigi et al.
[10] found conventional
T2*-weighted MRI (2D FLASH) useful in the diagnosis of ectop ic pregnancy
owing to depiction of extrauterine fresh hematoma in the pelvis. To our
knowledge, our report is the first on evaluation of endometrioma with
susceptibility-weighted imaging.
The ovary is the most common site of endometriosis, which can be converted
into endometriomas as a result of repeated hemorrhage
[6]. The typical endometrioma
is composed of endometrial glands and stroma, which are responsible for the
cyclic bleeding [6]. Deposition
of hemosiderin-laden macrophages in the cyst wall due to repeated hemorrhage
is a pathologic feature of endometrioma
[1,
5,
6]. In our study, signal voids
due to hemosiderin deposition along the cyst wall were well visualized in 39
of 42 endometriomas (92.9%) and in no nonendometrial cysts on
susceptibility-weighted images. Therefore, this imaging sign may be diagnostic
of endometrioma. It was difficult, however, to evaluate signal voids along the
cyst wall in endometriomas that were completely hypointense on
susceptibility-weighted images. This phenomenon was observed in only two of 42
endometriomas (4.8%), and the diagnosis of endometrioma was easily established
on the basis of the presence of shading on T2-weighted images. Thirty-two
endometriomas (76.2%) met definitive criteria on conventional MR images, that
is, hyperintensity on T1-weighted images and hypointensity on T2-weighted
images. With the addition of the presence of signal voids along the cyst wall
on susceptibility-weighted images to the MRI criteria, 41 endometriomas
(97.6%) were correctly diagnosed.
Intracystic hemorrhage may be found in adnexal cystic masses other than
endometrioma, such as corpus luteum cysts, follicular cysts, or neoplastic
cysts [4]. In our series, two
nonendometrial cystic masses containing hemorrhagic fluid had no signal voids
along the cyst walls on susceptibility-weighted images. The absence of signal
voids might have occurred because nonendometrial cysts do not usually
repeatedly bleed, and the thick fibrous capsule containing a cluster of
hemosiderin-laden macrophages is considered specific for endometrioma
[1,
5,
6]. The small number of
nonendometrial hemorrhagic cysts (two lesions) was a limitation of our study
because of the possibility that hemorrhagic cysts not related to endometriosis
contain hemosiderin-laden macrophages within their walls and cause signal
voids on susceptibility-weighted images. The retrospective nature of the study
and relatively small population also were limitations of our study.
Prospective studies with larger populations are needed to support our results.
A larger number of cases of hemorrhagic cysts other than endometrioma are
especially required to verify the specificity of signal voids along the
endometrioma wall on susceptibility-weighted imaging.
Because susceptibility-induced loss of signal intensity can increase from
1.5 T to 3 T, signal voids due to hemosiderin deposition on
susceptibility-weighted images were more prominent at 3 T in all eight
endometriomas examined at both 1.5 T and 3 T. Susceptibility artifacts caused
by intestinal gas, however, also were more prominent at 3 T. Further
comparative study at 1.5 T and 3 T may be needed to determine which strength
is suitable for evaluation of endometrioma with susceptibility-weighted
MRI.
We conclude that susceptibility-weighted imaging can contribute to the
diagnosis of endometrioma by depicting hemosiderin deposition in the cyst
wall. Susceptibility-weighted imaging at 3 T may be more sensitive to
hemosiderin deposition than imaging at 1.5 T. However, susceptibility
artifacts caused by intestinal gas also may be more prominent at 3 T.
References
- Olive DL, Schwartz LB. Endometriosis. N Engl J
Med 1993; 328:1759
–1769[Free Full Text]
- Carbognin G, Guarise A, Minelli L, et al. Pelvic endometriosis: US
and MRI features. Abdom Imaging 2004;29
: 609–618[Medline]
- Togashi K, Nishimura K, Kimura I, et al. Endometrial cysts:
diagnosis with MR imaging. Radiology1991; 180:73
–78[Abstract/Free Full Text]
- Bis KG, Vrachliotis TG, Agrawal R, Shetty AN, Maximovich A, Hricak
H. Pelvic endometriosis: MR imaging spectrum with laparoscopic correlation and
diagnostic pitfalls. RadioGraphics 1997;17
: 639–655[Abstract]
- Takeuchi M, Matsuzaki K, Uehara H, Nishitani H. Malignant
transformation of pelvic endometriosis: MR imaging findings and pathologic
correlation. RadioGraphics 2006;26
: 407–417[Abstract/Free Full Text]
- Rosai J. Ovary. In: Ackerman's surgical
pathology, 8th ed. St. Louis, MO: Mosby, 1996:1461
–1539
- Haacke EM, Xu Y, Cheng YC, Reichenbach JR. Susceptibility weighted
imaging (SWI). Magn Reson Med 2004;52
: 612–618[CrossRef][Medline]
- Sehgal V, Delproposto Z, Haacke EM, et al. Clinical applications of
neuroimaging with susceptibility-weighted imaging. J Magn Reson
Imaging 2005; 22:439
–450[CrossRef][Medline]
- Sehgal V, Delproposto Z, Haddar D, et al. Susceptibility-weighted
imaging to visualize blood products and improve tumor contrast in the study of
brain masses. J Magn Reson Imaging 2006;24
: 41–51[CrossRef][Medline]
- Yoshigi J, Yashiro N, Kinoshita T, O'uchi T, Kitagaki H. Diagnosis
of ectopic pregnancy with MRI: efficacy of T2*-weighted imaging.
Magn Reson Med Sci 2006;5
: 25–32[CrossRef][Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
