DOI:10.2214/AJR.07.2155
AJR 2008; 190:481-488
© American Roentgen Ray Society
Diffusion-Weighted Single-Shot Echo-Planar Imaging with Parallel Technique in Assessment of Endometrial Cancer
Shu-Huei Shen1,
Yi-You Chiou1,
Jia-Hwia Wang1,
Ming-Shyen Yen2,
Rheun-Chuan Lee1,
Chiung-Ru Lai3 and
Cheng-Yen Chang1
1 Department of Radiology, Taipei Veterans General Hospital, National Yang-Ming
University School of Medicine, 201, Sect II, Shih-Pai Rd., Taipei, 112,
Taiwan, Province of China.
2 Department of Obstetric and Gynecology, Taipei Veterans General Hospital,
National Yang-Ming University School of Medicine, Taipei, Taiwan.
3 Department of Laboratory and Pathology, Taipei Veterans General Hospital,
National Yang-Ming University School of Medicine, Taipei, Taiwan.
Received March 1, 2007;
accepted after revision July 28, 2007.
The study is supported in part by research grant VGH95A-045.
Address correspondence to J. H. Wang
(wangjh{at}vghtpe.gov.tw).
Abstract
OBJECTIVE. The purposes of this study were to determine the
feasibility of diffusion-weighted imaging (DWI) with a single-shot echo-planar
sequence and parallel technique for depicting endometrial cancer and to
examine the role of this technique in preoperative assessment.
SUBJECTS AND METHODS. A total of 31 patients were recruited for MRI
evaluation of suspicious endometrial lesions found on transvaginal sonography.
Twenty-four of the patients were proved to have endometrial cancer (patient
group), and seven to have benign diseases (control group). The MRI
examinations included diffusion-weighted single-shot echoplanar sequences and
contrast-enhanced T1-weighted 3D fat-suppressed spoiled gradient-echo
sequences. The apparent diffusion coefficient of endometrial cancer in the
patient group and of normal endometrium in the control group were measured on
the apparent diffusion coefficient map of each diffusion-weighted image and
compared for the two groups. In the patient group, myometrial invasion was
evaluated with the two sequences. The diagnostic accuracy rates of each pulse
sequence were compared.
RESULTS. The mean apparent diffusion coefficient of endometrial
cancer was 0.864 x 10–3 mm2/s and that of
benign endometrial lesions was 1.277 x 10–3
mm2/s. The difference between the two groups was significant
(p = 0.0058). The diagnostic accuracy for myometrial invasion was
61.9% for DWI and 71.4% for gadolinium-enhanced T1-weighted 3D fat-suppressed
spoiled gradient-recalled echo images. In five cases, DWI provided information
about tumor extent and depicted the tumor focus, findings that changed
preoperative staging.
CONCLUSION. DWI performed with parallel imaging technique has
potential as a method for differentiating benign from malignant endometrial
lesions. It also provides valuable information for preoperative evaluation and
should be considered part of routine preoperative MRI evaluation for
endometrial cancer.
Keywords: diffusion-weighted imaging • echo-planar imaging • endometrium • MRI • uterine neoplasm • uterus
Introduction
Adenocarcinoma of the endometrium is the third most frequent gynecologic
cancer in women. More than one half of the patients are older than 50 years,
and 15% are younger than 40 years. There is no reliable imaging method for
differentiating benign from malignant endometrial lesions
[1]. Because of its low cost,
relatively noninvasive nature, and high sensitivity in the detection of
abnormality, transvaginal sonography has been proposed as the first-line
diagnostic tool for evaluating postmenopausal bleeding
[2,
3]. Several sonographic signs
based on heterogeneity and irregular endometrial thickening have been proposed
for diagnosing endometrial carcinoma. These signs, however, are not reliable
for differentiating benign proliferation, hyperplasia, polyps, and cancer
[3–5].
In patients with postmenopausal bleeding who are found to have endometrial
thickening on transvaginal sonography, endometrial biopsy or dilation and
curettage (D&C) is necessary for a definitive diagnosis. Office-based
endometrial biopsy, however, has a high rate of inadequate sampling, and the
sensitivity is less than 50%
[3,
6]. Although the false-negative
rate of D&C is relatively low (< 10%)
[6,
7], the procedure is relatively
invasive. Both biopsy and D&C are difficult to perform on patients with
vaginal or cervical stenosis. In patients without sexual experience,
noninvasive imaging would be more suitable than a transvaginal procedure.
Because of its excellent soft-tissue contrast resolution, MRI is considered
the most accurate imaging technique for preoperative assessment of endometrial
cancer. Several MRI signs have been proposed for differentiating endometrial
cancer from other pathologic conditions. Grasel et al.
[5] reported that they could
not differentiate most endometrial polyps from carcinoma on the basis of
morphologic features on T2-weighted MR images. Endometrial polyps have a
higher incidence of having a central core and an intratumoral cyst, and
myometrial invasion and necrosis are highly predictive of carcinoma. Park et
al. [7] found that the
enhancement patterns of endometrial carcinoma were different from those of
other endometrial pathologic conditions. There is, however, no consensus about
the utility of MRI in differentiation of benign from malignant endometrial
lesions.
Findings on diffusion-weighted MRI (DWI) can provide insight into the water
composition of tumors and normal tissue. Pathologic processes such as
inflammation and neoplasia tend to alter structural organization by
destruction or regeneration of membranous elements or by a change in
cellularity [8]. Thus changes
in permeability, osmolarity, and active transportation can occur concurrently.
All of these changes can affect proton mobility and diffusivity, which can be
observed with DWI. Tumors, such those of as brain
[9], liver
[10], and colorectal
[11] cancer, have a lower
absolute diffusion value than does normal tissue. To our knowledge, the
appearance of endometrial cancer on DWI performed with parallel imaging
technique has not been described, and its usefulness for diagnosing
endometrial cancer has not been studied. The primary goal of this study was to
determine the feasibility of using single-shot echo-planar DWI performed with
parallel technique to differentiate endometrial cancer from benign endometrial
lesions. Another goal was to assess the role of the technique in preoperative
assessment of endometrial cancer.
Subjects and Methods
Patient Recruitment
During the period January–October 2006, a total of 31 patients were
referred for MRI by the gynecologic clinic because of abnormal vaginal
bleeding and the transvaginal sonographic finding of an endometrial lesion.
Patients with a pacemaker, intracranial coil, or other contraindication to MRI
were excluded for safety reasons. The study was approved by our institutional
review board, and an institutional review board–approved consent form
was used to obtain written informed consent from all patients. All patients
underwent D&C before the MRI examination, the interval between procedures
being 3–20 days (average, 9.5 days). According to the results of
pathologic examination of the D&C specimen, 24 patients (average age, 55.8
years; range, 33–82 years) had endometrial cancer (patient group), and
seven (average age, 45.2 years; range, 36–68 years) had benign disease
(control group). Four of the patients with benign lesions had endometrial
hyperplasia, and three had an endometrial polyp. All of the patients in the
patient group subsequently underwent hysterectomy. Those in the control group
did not undergo further surgery and participated in follow-up for at least 6
months without clinical or sonographic evidence of tumor growth.
Imaging Protocol
The MRI examination was performed with a 1.5-T unit (Twin Excite, GE
Healthcare) with an eight-channel phased-array body coil. The imaging
sequences included T2-weighted turbo spin echo in the sagittal and oblique
axial planes (TR/TE, 4,650/90; echo-train length, 17; matrix size, 512 x
256; field of view, 20 x 20 cm; number of signals averaged, 4; slice
thickness, 5 mm; gap, 1 mm; number of sagittal slices, 20) and T1-weighted
spin echo in the oblique axial plane (400/9; matrix size, 320 x 224;
field of view, 20 x 20 cm; number of signals averaged, 2; slice
thickness, 5 mm; gap, 1 mm; number of slices, 20). The oblique axial plane was
determined in relation to the sagittal plane and perpendicular to the uterine
body to best show the relation between the endometrium and the myometrium. DWI
with single-shot echo-planar imaging (EPI) and array spatial sensitivity
encoding technique (8,000/83; matrix size, 128 x 256; field of view, 24
x 24 cm; number of signals averaged, 4; slice thickness, 5 mm; gap, 1
mm; phase-encoding direction, anteroposterior) was performed in the sagittal
and oblique axial planes.
In the first five cases, DWI with a b value of 0 (T2-weighted EPI) and DWI
with b values of 500, 800, and 1,000 were performed. T2-weighted EPI provided
the best reference images of the anatomic relations and excluded the
T2-weighted shine-through effect. With the increase in b value, the signal
contrast between the tumor and adjacent normal tissue background became
stronger. In all cases, a b value of 1,000 was chosen to best depict tumor
extent. The scanning time of DWI for each plane was less than 2 minutes. The
apparent diffusion coefficient (ADC) map of each diffusion-weighted image was
produced with a workstation (Advantage Windows version 4.2.3, GE Healthcare).
Contrast-enhanced T1-weighted 3D fat-suppressed spoiled gradient-recalled echo
(FSPGR) images (matrix size, 256 x 224; slice thickness, 5 mm; zero-fill
interpolation, 2; receiver bandwidth, 62.5 kHz; flip angle, 12°) in the
early arterial (30 seconds), venous (60 seconds), and late phases (180
seconds) were acquired in the sagittal plane.
Imaging Analysis
All of the images were reviewed by two radiologists specializing in
genitourinary radiology. The ADCs of endometrial cancer in the patient group
and normal endometrium in the control group were measured. At the workstation,
three 15-mm2 regions of interest were selected, and the average
ADCs were acquired. In the patient group, myometrial invasion was determined
by correlation of T2-weighted turbo spin-echo (TSE) images with
diffusion-weighted images and gadolinium-enhanced T1-weighted 3D FSPGR images.
On DWI, tumor extent was determined by careful comparison of T2-weighted
echo-planar and diffusion-weighted images with a b value of 1,000. Any
abnormally high signal intensity on high-b-value images was depicted and
correlated with T2-weighted EPI findings. Myometrial invasion was classified
into three categories according to the International Federation of Gynecology
and Obstetrics classification: IA, no myometrial invasion; IB, invasion to
less than one half of the thickness of the myometrium; IC invasion to one half
or more of the thickness of the myometrium. The results were compared with the
operative findings and the pathologic finding on the hysterectomy specimen to
determine the diagnostic accuracy of each pulse sequence.
Statistical Analysis
The ADCs of endometrial cancer and benign endometrial tissue were compared
by use of Wilcoxon's rank sum test. In the patient group, the correlation
between tumor grade and ADC also was analyzed by use of Wilcoxon's rank sum
test. The diagnostic accuracy of DWI and gadolinium-enhanced T1-weighted 3D
FSPGR imaging for myometrial invasion was calculated with the pathologic
finding as the reference standard.
Results
In three patients with proven endometrial cancer, ADCs were not measured
because no definite tumor focus could be identified and the regions of
interest could not be properly selected. In these cases, no tumor was
identified in the endometrial cavity at gross examination, probably because of
previous curettage; only a small tumor focus was identified at microscopic
examination. In the other cases of endometrial cancer, the mean ADC was 0.864
± 0.311 (SD) x 10–3 mm2/s (n
= 21). The mean ADC of the control group was 1.277 ± 0.219 x
10–3 mm2/s (n = 7). The difference
between the two groups was significant (p = 0.0058). In the patient
group, all of the malignant tumors were adenocarcinoma. Most of them were of
the endometrioid type (grade 1, n = 8; grade 2, n = 9; grade
3, n = 2). One tumor was of the mixed endometrioid and mucinous type
(grade 2), and one was of the clear cell type (grade 1). There was no
significant correlation between ADC and tumor grade (p = 0.1915).
On DWI, susceptibility artifact was mainly caused by gas in bowel loops,
but the artifacts generally did not interfere with image interpretation. In
patients with abundant bowel gas in the pelvic cavity, however, distortion and
susceptibility artifacts can be prominent, and precise measurement and mapping
can be difficult. For accurate delineation of tumor extent, careful comparison
and correlation between high-resolution T2-weighted TSE and T2-weighted
echo-planar and diffusion-weighted images with a high b factor was mandatory.
In one patient with a metallic prosthesis in the hip joint, the lower part of
the cervix and part of the lesion were obscured (Fig.
1A,
1B,
1C,
1D,
1E,
1F). No obvious hematoma,
edema, or other possible postcurettage change that might have interfered with
image interpretation was identified, except for visible tissue defect that
might have been caused by curettage. In four cases, spots of slightly high
signal intensity were identified within tumor. This effect might have been due
to intratumoral hemorrhage rather than postcurettage change. All of these
patients had undergone curettage more than 14 days before MRI. No obvious
hemorrhage was identified in the seven patients who underwent MRI less than 7
days after D&C.
View larger version (154K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —83-year-old woman with benign endometrial polyp. T2-weighted
fast spin-echo sagittal (A) and axial (B) images show polypoid
tumor (star, A; arrows, B) protruding from
endometrial cavity.
|
|
View larger version (148K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B —83-year-old woman with benign endometrial polyp. T2-weighted
fast spin-echo sagittal (A) and axial (B) images show polypoid
tumor (star, A; arrows, B) protruding from
endometrial cavity.
|
|
View larger version (117K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C —83-year-old woman with benign endometrial polyp. Sagittal
(C and D) and axial (E and F) diffusion-weighted
images at same level as A and B show tumor (arrows)
barely visible on dark background with high b value (b = 1,000) (D and
F). Metallic prosthesis in right hip joint causes severe susceptibility
artifact and obscures caudal part of lesion. C and E, b = 0.
|
|
View larger version (121K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D —83-year-old woman with benign endometrial polyp. Sagittal
(C and D) and axial (E and F) diffusion-weighted
images at same level as A and B show tumor (arrows)
barely visible on dark background with high b value (b = 1,000) (D and
F). Metallic prosthesis in right hip joint causes severe susceptibility
artifact and obscures caudal part of lesion. C and E, b = 0.
|
|
View larger version (120K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1E —83-year-old woman with benign endometrial polyp. Sagittal
(C and D) and axial (E and F) diffusion-weighted
images at same level as A and B show tumor (arrows)
barely visible on dark background with high b value (b = 1,000) (D and
F). Metallic prosthesis in right hip joint causes severe susceptibility
artifact and obscures caudal part of lesion. C and E, b = 0.
|
|
View larger version (119K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1F —83-year-old woman with benign endometrial polyp. Sagittal
(C and D) and axial (E and F) diffusion-weighted
images at same level as A and B show tumor (arrows)
barely visible on dark background with high b value (b = 1,000) (D and
F). Metallic prosthesis in right hip joint causes severe susceptibility
artifact and obscures caudal part of lesion. C and E, b = 0.
|
|
Endometrial cancer had intermediate signal intensity on T2-weighted TSE
images. The intensity was lower than that of normal endometrium and higher
than that of adjacent myometrium. Because the junctional zone was apparent on
T2-weighted EPI, myometrial invasion was readily detected. On DWI with a b
value of 1,000, endometrial cancer appeared bright and obvious in contrast to
the dark background. On gadolinium-enhanced T1-weighted 3D FSPGR images, both
endometrial cancer and normal endometrium had less contrast enhancement than
myometrium. Strong enhancement of the subendometrial myometrium in the
arterial phase allowed easy detection of myometrial invasion.
In our study, the following three patterns of endometrial cancer were
observed: polypoid tumor in the endometrial cavity with a stalk-like structure
(n = 9) (Fig. 2A,
2B,
2C,
2D), endometrial thickening
with or without myometrial invasion (n = 4), and predominantly
myometrial infiltration (n = 8). The diagnostic accuracy of
myometrial invasion was calculated for both DWI
(Table 1) and
gadolinium-enhanced T1-weighted 3D FSPGR imaging
(Table 2). Tumors that
infiltrate the myometrium or the stroma of the cervix can show marked or
heterogeneous enhancement after gadolinium administration, which makes it
difficult to differentiate them from adjacent normal myometrium. Under such
circumstances, DWI with a b value of 1,000 clearly showed the boundary between
normal myometrium and tumor (Fig.
3A,
3B,
3C,
3D). With higher imaging
resolution and strong subendometrial enhancement, however, gadolinium-enhanced
T1-weighted 3D FSPGR images had higher sensitivity for superficial layer
myometrial invasion than did DWI. In two patients with category IB tumors, DWI
did not depict superficial layer invasion, but T1-weighted 3D FSPGR images
did. The overall diagnostic accuracy for myometrial invasion was 71.4% for
gadolinium-enhanced T1-weighted 3D FSPGR imaging and 61.9% for DWI.
View larger version (155K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —41-year-old woman with endometrial adenocarcinoma.
Contrast-enhanced T1-weighted 3D fat-suppressed spoiled gradient-recalled echo
sagittal image shows lesion has minimal enhancement compared with normal
myometrium. Star indicates polypoid lesion with stalk.
|
|
View larger version (124K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C —41-year-old woman with endometrial adenocarcinoma. Sagittal
T2-weighted echo-planar image (b = 0) shows tissue contrast enhancement is
strong and zonal anatomy of uterus is enhanced. Junctional zone
(arrows) is clear.
|
|
View larger version (80K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2D —41-year-old woman with endometrial adenocarcinoma. Sagittal
diffusion-weighted image (b = 1,000) shows tumor as area of high signal
intensity with dark background. Pathologic examination disclosed endometrial
cancer with superficial myometrial invasion.
|
|
View this table:
[in this window]
[in a new window]
|
TABLE 1: Diagnostic Accuracy of Myometrial Invasion on Diffusion-Weighted Imaging
Correlated with Pathologic Finding
|
|
View this table:
[in this window]
[in a new window]
|
TABLE 2: Diagnostic Accuracy of Myometrial Invasion on Gadolinium-Enhanced
T1-Weighted 3D Fat-Suppressed Spoiled Gradient-Recalled Echo Images Correlated
with Pathologic Finding
|
|
View larger version (112K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3C —71-year-old woman with endometrial adenocarcinoma. Sagittal
diffusion-weighted images with b values of 0 (C) and 1,000 (D)
show tumor has high signal intensity (star, D). Boundary is
evident.
|
|
View larger version (89K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3D —71-year-old woman with endometrial adenocarcinoma. Sagittal
diffusion-weighted images with b values of 0 (C) and 1,000 (D)
show tumor has high signal intensity (star, D). Boundary is
evident.
|
|
In five cases, DWI provided information about tumor extent and depicted the
tumor foci. This information led to changes in preoperative staging and was
not obtained with the other imaging sequences. In one case of tumor growth at
the uterine fundus, focal endometrial thickening was found at the endocervical
canal on T2-weighted images; thus another tumor focus was suspected (Fig.
4A,
4B,
4C,
4D,
4E,
4F,
4G,
4H,
4I). Findings on
gadolinium-enhanced T1-weighted 3D FSPGR images were inconclusive. On DWI, no
focus of high signal intensity was identified at the corresponding location in
the endocervical canal. At pathologic examination, no tumor growth was found
at the endocervical canal. Another patient, in addition to having a main tumor
in the endometrial cavity, had a tumor focus in the left fallopian tube that
was difficult to appreciate on T2-weighted images and gadolinium-enhanced
T1-weighted 3D FSPGR images but was obvious on DWI (Fig.
5A,
5B,
5C,
5D,
5E,
5F,
5G,
5H,
5I).
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A —82-year-old woman with endometrial adenocarcinoma.
T2-weighted turbo spin-echo sagittal image shows endometrial thickening at
uterine fundus and endocervical canal (arrows). Both are suspected
tumor focuses.
|
|
View larger version (133K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B —82-year-old woman with endometrial adenocarcinoma.
Contrast-enhanced T1-weighted 3D fat-suppressed spoiled gradient-recalled echo
sagittal images do not exclude possibility of tumor growth for both
lesions.
|
|
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C —82-year-old woman with endometrial adenocarcinoma.
Contrast-enhanced T1-weighted 3D fat-suppressed spoiled gradient-recalled echo
sagittal images do not exclude possibility of tumor growth for both
lesions.
|
|
View larger version (133K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4D —82-year-old woman with endometrial adenocarcinoma. Sagittal
diffusion-weighted image with b values of 0 (D–F) and 1,000
(G–I) show only lesion (arrow, G) at fundus has
high signal intensity. Lesion was pathologically confirmed only tumor
focus.
|
|
View larger version (132K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4E —82-year-old woman with endometrial adenocarcinoma. Sagittal
diffusion-weighted image with b values of 0 (D–F) and 1,000
(G–I) show only lesion (arrow, G) at fundus has
high signal intensity. Lesion was pathologically confirmed only tumor
focus.
|
|
View larger version (126K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4F —82-year-old woman with endometrial adenocarcinoma. Sagittal
diffusion-weighted image with b values of 0 (D–F) and 1,000
(G–I) show only lesion (arrow, G) at fundus has
high signal intensity. Lesion was pathologically confirmed only tumor
focus.
|
|
View larger version (120K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4G —82-year-old woman with endometrial adenocarcinoma. Sagittal
diffusion-weighted image with b values of 0 (D–F) and 1,000
(G–I) show only lesion (arrow, G) at fundus has
high signal intensity. Lesion was pathologically confirmed only tumor
focus.
|
|
View larger version (118K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4H —82-year-old woman with endometrial adenocarcinoma. Sagittal
diffusion-weighted image with b values of 0 (D–F) and 1,000
(G–I) show only lesion (arrow, G) at fundus has
high signal intensity. Lesion was pathologically confirmed only tumor
focus.
|
|
View larger version (116K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4I —82-year-old woman with endometrial adenocarcinoma. Sagittal
diffusion-weighted image with b values of 0 (D–F) and 1,000
(G–I) show only lesion (arrow, G) at fundus has
high signal intensity. Lesion was pathologically confirmed only tumor
focus.
|
|
View larger version (128K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A —58-year-old woman with endometrial adenocarcinoma. Several
tumor foci were present at endocervical canal (not shown) and fundus with
hydrometra. Axial diffusion-weighted images with b values of 0
(A–D) and 1,000 (E–H) show images with high b value
depict lesion (arrow, E) with high signal intensity in left
pelvic cavity.
|
|
View larger version (133K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B —58-year-old woman with endometrial adenocarcinoma. Several
tumor foci were present at endocervical canal (not shown) and fundus with
hydrometra. Axial diffusion-weighted images with b values of 0
(A–D) and 1,000 (E–H) show images with high b value
depict lesion (arrow, E) with high signal intensity in left
pelvic cavity.
|
|
View larger version (135K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5C —58-year-old woman with endometrial adenocarcinoma. Several
tumor foci were present at endocervical canal (not shown) and fundus with
hydrometra. Axial diffusion-weighted images with b values of 0
(A–D) and 1,000 (E–H) show images with high b value
depict lesion (arrow, E) with high signal intensity in left
pelvic cavity.
|
|
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5D —58-year-old woman with endometrial adenocarcinoma. Several
tumor foci were present at endocervical canal (not shown) and fundus with
hydrometra. Axial diffusion-weighted images with b values of 0
(A–D) and 1,000 (E–H) show images with high b value
depict lesion (arrow, E) with high signal intensity in left
pelvic cavity.
|
|
View larger version (97K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5E —58-year-old woman with endometrial adenocarcinoma. Several
tumor foci were present at endocervical canal (not shown) and fundus with
hydrometra. Axial diffusion-weighted images with b values of 0
(A–D) and 1,000 (E–H) show images with high b value
depict lesion (arrow, E) with high signal intensity in left
pelvic cavity.
|
|
View larger version (99K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5F —58-year-old woman with endometrial adenocarcinoma. Several
tumor foci were present at endocervical canal (not shown) and fundus with
hydrometra. Axial diffusion-weighted images with b values of 0
(A–D) and 1,000 (E–H) show images with high b value
depict lesion (arrow, E) with high signal intensity in left
pelvic cavity.
|
|
View larger version (83K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5G —58-year-old woman with endometrial adenocarcinoma. Several
tumor foci were present at endocervical canal (not shown) and fundus with
hydrometra. Axial diffusion-weighted images with b values of 0
(A–D) and 1,000 (E–H) show images with high b value
depict lesion (arrow, E) with high signal intensity in left
pelvic cavity.
|
|
View larger version (87K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5H —58-year-old woman with endometrial adenocarcinoma. Several
tumor foci were present at endocervical canal (not shown) and fundus with
hydrometra. Axial diffusion-weighted images with b values of 0
(A–D) and 1,000 (E–H) show images with high b value
depict lesion (arrow, E) with high signal intensity in left
pelvic cavity.
|
|
View larger version (175K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5I —58-year-old woman with endometrial adenocarcinoma. Several
tumor foci were present at endocervical canal (not shown) and fundus with
hydrometra. T2-weighted turbo spin-echo axial image shows enlarged left
fallopian tube (arrows), which has intermediate signal intensity and
might have been easily overlooked.
|
|
Discussion
DWI reflects change in proton mobility caused by alteration of the tissue
environment due to pathologic processes. DWI single-shot EPI is highly
susceptible to off-resonance effects such as main-field inhomogeneity, local
susceptibility gradients, and chemical shift. All of these effects can lead to
severe image degradation. Macroscopic motion is inevitable in the abdomen, and
air and fat are more abundant in the abdomen than in other body parts. Thus,
DWI has been of limited use in abdomen. A new parallel imaging strategy has
reduced the train of gradient echoes in the EPI readout in combination with
faster k-space traversal per unit time. This innovation has resulted in
increased bandwidth per pixel in the phaseencoding direction and a shortened
EPI train, which has improved image quality.
A previous study [10]
showed the usefulness of DWI and ADC in differentiating hepatic lesions. Nasu
et al. [11] found that DWI
performed with parallel imaging technique depicted colorectal tumors as areas
of high signal intensity, whereas normal colonic wall and feces were seen as
areas of low signal intensity. In our preliminary study, the results suggested
that endometrial cancer has a significantly lower ADC than nonmalignant
endometrial lesions. More experience is needed before a definitive conclusion
can be made about ability to differentiate benign from malignant lesions.
Patients in whom transvaginal biopsy or curettage is difficult will benefit
from the noninvasive imaging procedure. Unnecessary surgical intervention for
benign endometrial lesions can be avoided, as can delayed surgical
intervention for endometrial cancer owing to false-negative results of
endometrial biopsy or curettage.
Surgery is the treatment of choice of patients with noninvasive or locally
advanced disease. Preoperative clinical and instrumental staging of local
disease spread and local and distant lymph node involvement is a critical step
in tailoring the extent and radicalness of surgery. It is critical that
preoperative assessment of endometrial carcinoma involve determining the depth
of myometrial invasion, which is a reliable index of lymphatic spread
[12] and is important for
planning surgery. Furthermore, determining the presence and depth of
myometrial invasion is used in most institutions to predict nodal metastasis.
Patients with 50% or greater myometrial invasion have a sixfold to sevenfold
increased prevalence of pelvic and lumboaortic lymph node metastasis compared
with patients in whom myometrial invasion is absent or less than 50%
[12].
MRI has been found accurate in the diagnosis and staging of endometrial
cancer [13,
14] and to have greater
accuracy than other imaging techniques, such as sonography and CT. The
presence and depth of myometrial infiltration can be assessed on T2-weighted
images as interruption of the low-signal-intensity junctional zone by
intermediate to high signal intensity from adenocarcinoma. With postmenopausal
change or myometrial thinning due to endometrial cavity distention, however,
the junctional zone can be poorly visible, making it difficult to assess the
presence and depth of myometrial infiltration. Manfredi et al.
[14] found that the junctional
zone was poorly visible in eight (22%) of their patients. To overcome this
limitation, dynamic MRI should be performed because it can depict the
different enhancement times of the adenocarcinoma and adjacent myometrium.
This capability improves the contrast resolution of the tumor and myometrium.
In previous studies
[13–15],
combining T2-weighted TSE with dynamic MRI improved staging accuracy from 61%
to 90%.
In our study, endometrial cancer was found to have high signal intensity on
DWI; thus the boundary between cancer and normal myometrium was clearly
depicted. When the parallel imaging technique was used, image distortion was
significantly reduced to a minimum in the pelvic cavity, and the correlation
with other reference imaging studies, including T2-weighted EPI and
T2-weighted TSE, was high. However, the overall accuracy for estimating
myometrial invasion was better with gadolinium-enhanced T1-weighted 3D FSPGR
imaging compared with DWI. Although DWI depicted the extent of tumor
infiltration in myometrium better than T1-weighted 3D FSPGR did, DWI is
suboptimal for detection of superficial myometrial invasion owing to its
relatively poor resolution. The main value of DWI as an adjunct sequence
during preoperative assessment is depiction of tumor foci. DWI depicts tumor
foci that are not obvious and are easily missed with the routine sequence. In
patients with endometrial cancer, more than one tumor focus may be present in
the uterus, and tumor seeding to bilateral adnexa and the peritoneal cavity is
not rare.
Correct preoperative depiction of all tumor foci is crucial for staging. On
DWI with a high b value, tumors have high signal intensity on a dark
background; thus all tumor foci within the examination field are obvious, as
shown in our cases. However, the need to correlate DWI findings with reference
imaging findings should be emphasized because the resolution of DWI is
relatively poor and because normal structures and benign lesions, including
lymph nodes, bowel loops, hemorrhagic cysts, and endometriosis, can have high
signal intensity on DWI with a b value of 1,000
[16,
17]. Without careful
correlation of DWI and high-resolution images, normal structures and benign
lesions can cause false-positive results on DWI.
The main limitation of this study was the limited number and range of
benign processes assessed. Because endometrial hyperplasia and polyps were the
only benign processes included in our study, the specificity of signal
intensity on DWI and ADC is unknown. To thoroughly understand the physiologic
importance of uterine DWI findings, more cases with a variety of pathologic
processes, including inflammation, hemorrhage, and different types of benign
and malignant tumors, need to be studied. The other limitation of this study
was the possible sampling error in the benign group. In the patients in the
benign group, only D&C was performed, and none of the patients underwent
hysterectomy. Because endometrial hyperplasia is a histologic precursor of
endometrial adenocarcinoma and can progress to invasive adenocarcinoma through
progressive degrees of cellular and architectural atypia, small tumor foci
embedded within a polyp or hyperplasia can be missed. Although we conducted
follow-up with the patients for more than 6 months to assure no tumor growth,
results of pathologic examination of the hysterectomy specimen would be
stronger support for our conclusion.
DWI with parallel imaging technique has potential as method of
differentiating benign from malignant endometrial lesions. It is not
time-consuming and is easy to perform. As an adjunct sequence for preoperative
evaluation, DWI is sensitive for detecting lesion foci in the examination
field and can help define tumor extent. We believe that DWI with parallel
imaging will have added value as an adjunct to conventional and
contrast-enhanced MRI in routine preoperative evaluation for endometrial
cancer.
Acknowledgments
We thank Sheng Wen-Yung for statistical analysis and consultation on the
study.
References
- Hricak H, Stern JL, Fisher MR, et al. Endometrial carcinoma staging
by MR imaging. Radiology 1987;162
: 297–305[Abstract/Free Full Text]
- Medverd JR, Dubinsky TJ. Cost analysis model: US versus endometrial
biopsy in evaluation of peri- and postmenopausal abnormal vaginal bleeding.
Radiology 2002;222
: 619–627[Abstract/Free Full Text]
- Dubinsky TJ, Parvey HR, Maklad N. The role of transvaginal
sonography and endometrial biopsy in the evaluation of peri- and
postmenopausal bleeding. AJR 1997;169
: 145–149[Abstract]
- Smith-Bindman R, Kerlikowske K, Feldstein VA, et al. Endovaginal
ultrasound to exclude endometrial cancer and other endometrial abnormalities.
JAMA 1998; 280:1510
–1517[Abstract/Free Full Text]
- Grasel RP, Outwater EK, Siegelman ES, et al. Endometrial polyps: MR
imaging features and distinction from endometrial carcinoma.
Radiology 2000;214
: 47–52[Abstract/Free Full Text]
- Word B, Gravlee LC, Wideman GL. The fallacy of simple uterine
curettage. Obstet Gynecol 1958;12
: 642–648[Medline]
- Park BK, Kim B, Park JM, et al. Differentiation of the various
lesions causing an abnormality of the endometrial cavity using MR imaging:
emphasis on enhancement patterns on dynamic studies and late contrast-enhanced
T1-weighted images. Eur Radiol 2006;16
:1591
–1598[CrossRef][Medline]
- Bammer R. Basic principles of diffusion-weighted imaging.
Eur J Radiol 2003;45
: 169–184[CrossRef][Medline]
- Castillo M, Smith JK, Kwack L, Wiber K. Apparent diffusion
coefficients in the evaluation in high-grade cerebral gliomas. Am J
Neuroradiol 2001; 22:60
–64[Abstract/Free Full Text]
- Taouli B, Vilgrain V, Dumont E, Daire JL, Fan B, Menu Y. Evaluation
of liver diffusion isotropy and characterization of focal hepatic lesions with
two single-shot echo-planar MR imaging sequences: prospective study in 66
patients. Radiology 2003;226
: 71–78[Abstract/Free Full Text]
- Nasu K, Kuroki Y, Kuroki S, Murakami K, Nawano S, Moriyama N.
Diffusion-weighted single shot echo planar imaging of colorectal cancer using
a sensitivity-encoding technique. Jpn J Clin Oncol2004; 34:620
–626[Abstract/Free Full Text]
- Creasman WT, Morrow CP, Bundy BN, et al. Surgical pathologic spread
pattern of endometrial cancer: a Gynecologic Oncology Group study.
Cancer 1987; 60:2035
–2041[CrossRef][Medline]
- Lee EJ, Byun JY, Kim BS, Koong SE, Shinn KS. Staging of early
endometrial carcinoma: assessment with T2-weighted and gadolinium-enhanced
T1-weighted MR imaging. RadioGraphics1999; 19:937
–945[Abstract/Free Full Text]
- Manfredi R, Mirk P, Maresca G, et al. Local-regional staging of
endometrial carcinoma: role of MR imaging in surgical planning.
Radiology 2004;231
: 372–378[Abstract/Free Full Text]
- Akaeda T, Isaka K, Takayama M, Kakizaki D, Abe K. Myometrial
invasion and cervical invasion by endometrial carcinoma: evaluation by
CO2-volumetric interpolated breathhold examination (VIBE). J Magn
Reson Imaging 2005; 21:166
–171[CrossRef][Medline]
- Moteki T, Ishizaka H. Evaluation of cystic ovarian lesions using
apparent diffusion coefficient calculated from recorded turboflash MR images.
Magn Reson Imaging 1999;17
: 955–963[CrossRef][Medline]
- Motoyuki K, Takayuki M, Shigeru K, et al. Diffusion-weighted echo
planar imaging of ovarian tumors: is it useful to measure apparent diffusion
coefficients? J Comput Assist Tomogr2002; 26:250
–256[CrossRef][Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
F. Saremi, A. N. Knoll, O. J. Bendavid, H. Schultze-Haakh, N. Narula, and F. Sarlati
Characterization of Genitourinary Lesions with Diffusion-weighted Imaging
RadioGraphics,
September 1, 2009;
29(5):
1295 - 1317.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. S. Whittaker, A. Coady, L. Culver, G. Rustin, M. Padwick, and A. R. Padhani
Diffusion-weighted MR Imaging of Female Pelvic Tumors: A Pictorial Review1
RadioGraphics,
May 1, 2009;
29(3):
759 - 774.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
R. C. Jha, C. S. Whittaker, A. Coady, L. Culver, G. Rustin, M. Padwick, and A. R. Padhani
Invited Commentary * Authors' Response * Authors' Response
RadioGraphics,
May 1, 2009;
29(3):
774 - 778.
[Full Text]
[PDF]
|
|
|
|
|
|
|
G. Lin, K.-K. Ng, C.-J. Chang, J.-J. Wang, K.-C. Ho, T.-C. Yen, T.-I Wu, C.-C. Wang, Y.-R. Chen, Y.-T. Huang, et al.
Myometrial Invasion in Endometrial Cancer: Diagnostic Accuracy of Diffusion-weighted 3.0-T MR Imaging--Initial Experience
Radiology,
March 1, 2009;
250(3):
784 - 792.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
