| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 All authors: Department of Radiology, Niigata University School of Medicine, 757, 1-Bancho, Asahimachi-dori, Niigata 951-8510, Japan.
Received October 1, 1999;
accepted after revision December 8, 1999.
Address correspondence to H. Seki.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. In 42 patients with endometrial carcinoma, T2-weighted MR imaging using rapid acquisition with relaxation enhancement, dynamic MR imaging using gradient-echo sequences, and contrast-enhanced T1-weighted MR imaging using spin-echo sequences were performed before treatment. We evaluated patterns of enhancement in the cervix and tumor. In 39 of the 42 patients who underwent surgical treatment, we compared MR imaging findings with histologic results concerning cervical involvement.
RESULTS. Enhancement of the cervical epithelium was greater than that of the tumor and cervical stroma on dynamic MR imaging in most patients. In assessing cervical involvement, the accuracy of T2-weighted, dynamic, and contrast-enhanced T1-weighted MR imaging was 85%, 95%, and 90%, respectively; no statistically significant difference was observed. False-positive cases on T2-weighted MR imaging were correctly identified as having no cervical involvement on dynamic MR imaging using the finding of continuous enhancement of the cervical epithelium. We found this finding to be reliable in assessing tumor involvement of the cervix.
CONCLUSION. We believe that, in combination with T2-weighted MR imaging sequences, dynamic MR imaging is useful in assessing endometrial carcinoma involvement of the cervix.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Assessment of cervical involvement has been made mainly by fractional curettage, which has been shown to be highly inaccurate [7, 8]. Several authors have mentioned that T2-weighted MR imaging using conventional spin-echo sequences is useful in evaluating endometrial carcinoma involvement of the cervix [9,10,11,12]. In addition, it has recently been shown that T2-weighted MR imaging using rapid acquisition with relaxation enhancement (RARE) sequences provides excellent image quality and anatomic detail [13]. Several previous studies have reported that dynamic MR imaging may improve accuracy in assessing preoperative staging of endometrial carcinoma [12, 14,15,16,17,18]. One report suggested that contrast-enhanced MR imaging might be helpful in evaluating endometrial carcinoma involvement of the cervix [12]. However, to our knowledge, few studies have compared dynamic MR imaging with T2-weighted MR imaging using RARE sequences in the assessment of tumor involvement of the cervix [14].
In this study, we evaluated enhancement patterns of the cervical epithelium, stroma, and tumor on dynamic MR imaging and correlated MR imaging with histologic findings. Then we tried to determine the value of dynamic MR imaging in evaluating endometrial carcinoma involvement of the cervix.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
MR imaging was performed with a 1.5-T superconducting magnet (Vision; Siemens, Erlangen, Germany). A phased array coil was used, and patients were imaged in the supine position. All patients were given 20 mg of butylscopolamine bromide (Buscopan; Tanabe Seiyaku, Osaka, Japan) intramuscularly to reduce the peristaltic movement of the intestine. T2-weighted MR imaging using RARE sequences was performed with 3500/96 msec (TR/TE), two signals acquired, an echo train length of seven, and a 196 x 512 matrix. The section thickness was 6 mm with a 1-mm intersection gap, and the rectangular field of view was 188 x 250 mm. T2-weighted MR images were obtained in axial planes and parasagittal planes that were parallel to the longitudinal axis of the uterus. T1-weighted MR imaging using spin-echo pulse sequences was also performed in axial and parasagittal planes with 550/12 (TR/TE), two signals acquired, a 192 x 512 matrix, a 6-mm section thickness, a 1-mm intersection gap, and a rectangular field of view of 188 x 250 mm. The acquisition time of T1- and T2-weighted MR imaging was 3 min 34 sec and 3 min 24 sec, respectively. Subsequently, multislice (five slices) dynamic MR imaging was performed using a spoiled gradient-echo sequence, which was a fast low-angle shot technique, with 110/4 or 120/4 (TR/TE), a 75° or 80° flip angle, a 160 x 256 or 144 x 256 matrix, a 6-mm section thickness, a 1-mm intersection gap, and a rectangular field of view of 188 x 250 mm. These images were obtained parallel to the parasagittal T2-weighted MR imaging and were centered at the imaging plane in which the endometrium and cervical epithelium were observed to define the position of the internal cervical os. Data acquisition began simultaneously with initiation of an IV rapid bolus injection (approximately 2 ml/sec) of 0.1 mmol/kg of gadopentetate dimeglumine (Magnevist; Nihon Schering, Osaka, Japan). Dynamic MR imaging was performed every 20 sec for 4 min without breath-holding; the first 80 sec after administration of the contrast material, the subsequent 80 sec, and the last 80 sec were defined as phase 1, phase 2, and phase 3, respectively. After dynamic MR imaging, contrast-enhanced T1-weighted spin-echo MR imaging was performed 5-6 min after administration of the contrast material in axial and parasagittal planes with the same parameters as those of unenhanced T1-weighted spin-echo imaging. T2-weighted MR imaging and unenhanced and contrast-enhanced T1-weighted spin-echo MR imaging were performed with a superior spatial presaturation pulse and with a saturation band placed anteriorly to eliminate ghost or blurring artifacts caused by movement of subcutaneous fat. Total imaging time was approximately 25 min. Total examination time was about 35 min including the positioning time.
Image Analysis
In all 42 patients, a small (1-5 mm2) region of interest was
placed in the tumor, cervical epithelium, and cervical stroma. On dynamic MR
images, the enhancement ratio was obtained for each region of interest as
follows: (Sn-S0) x
100/S0, where Sn indicates signal
intensity at each dynamic phase and S0, signal intensity
on the image obtained before administration of the contrast medium
[17]. Then, we assessed
enhancement patterns of each part.
The contrast-noise ratio between the tumor and cervical epithelium was obtained as follows: (Se-St)/noise SD, where Se indicates signal intensity of the cervical epithelium; St, signal intensity of the tumor; and noise SD, the standard deviation of the intensity of the background noise in the phase-encoding direction. The contrast-noise ratio between the tumor and cervical stroma was obtained as follows: (Ss-St)/noise SD, where Ss indicates signal intensity of the cervical stroma. The Student's t test was used to determine the significant difference of the contrast-noise ratio among the pulse sequences.
Image Interpretation
In 39 patients who underwent surgical treatment, endometrial carcinoma
involvement of the cervix was evaluated using T2-weighted MR imaging, dynamic
MR imaging, and contrast-enhanced T1-weighted spin-echo MR imaging. On
T2-weighted MR imaging, abnormal signal intensity extending into the cervical
canal or stroma or widening of the cervical canal was considered to indicate
cervical involvement
[9,10,11,12,
14]. On dynamic MR imaging and
contrast-enhanced T1-weighted spin-echo MR imaging, cervical involvement was
determined on the basis of the disruption of enhancement of the cervical
epithelium [12,
14]. When the enhanced
cervical epithelium was indistinguishable from enhanced cervical stroma, the
same criteria as those of T2-weighted MR imaging were used.
MR images were interpreted independently by two radiologists who were unaware of the histologic results. Discrepancies in interpretation were resolved by consensus. The degree of interobserver agreement between the two radiologists was assessed using the kappa statistic. Kappa values of 0.41-0.75 were considered to represent a good correlation, and those greater than 0.75 were considered to represent excellent correlation [20].
The data were correlated with histologic findings. Surgical specimens were sectioned along the longitudinal plane of the uterus, and cervical involvement was evaluated grossly and confirmed microscopically. With histologic results as the standard of reference, the accuracy of detection of cervical involvement with MR imaging was determined. The chi-square test or Fisher's exact test was used to compare the accuracy of the three MR imaging techniques. In addition, sensitivity, specificity, positive predictive value, and negative predictive value were calculated for each pulse sequence.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Enhancement Patterns of the Cervix and Endometrial Carcinoma
The enhancement ratio of the cervical epithelium, stroma, and tumor on
dynamic MR imaging is shown in Figure
1. The cervical epithelium revealed significant enhancement in all
patients (Figs.
2A,2B,2C,2D,3A,3B,3C,3D,4A,4B).
Enhancement of the cervical stroma was slight compared with that of the
cervical epithelium in all dynamic phases in 35 patients (Figs.
2A,2B,2C,2D
and
3A,3B,3C,3D)
and in phase 1 and phase 2 in three patients. In the remaining four patients,
the cervical stroma showed significant enhancement similar to that of the
cervical epithelium (Fig.
4A,4B).
Enhancement of the tumor was less than that of the cervical epithelium in 41
patients (Figs.
2A,2B,2C,2D,3A,3B,3C,3D,4A,4B).
The tumor showed significant enhancement similar to the cervical epithelium in
only one patient.
|
|
|
|
|
|
|
|
|
|
|
The cervical epithelium was high in signal intensity compared with the cervical stroma in 38 (90%) of 42 patients on dynamic MR imaging, and in 35 (83%) of 42 patients on contrast-enhanced T1-weighted spin-echo MR imaging (Figs. 2A,2B,2C,2D and 3A,3B,3C,3D). In these patients, enhancement of the cervical epithelium was used as a reference in assessing cervical involvement. In the remaining patients, the cervical epithelium could not be distinguished from the cervical stroma because the cervical stroma had significant enhancement and appeared similar to the cervical epithelium (Fig. 4A,4B).
Contrast Between Tumor and Cervix
The contrast-noise ratio between the tumor and cervical epithelium and that
between the tumor and cervical stroma on T2-weighted imaging, dynamic MR
imaging, and contrast-enhanced T1-weighted spin-echo MR imaging are shown in
Figures 5 and
6, respectively. The contrast
between the tumor and cervical epithelium was greater on dynamic MR imaging
than on T2-weighted MR imaging (p < 0.0001) and contrast-enhanced
T1-weighted spin-echo MR imaging (p < 0.001). In particular, on
dynamic MR imaging, the contrast was greater in phases 2 and 3 than in phase 1
(p < 0.05). On the other hand, contrast between the tumor and
cervical stroma was greater on T2-weighted MR imaging than on dynamic MR
imaging (p < 0.01) and enhanced T1-weighted spin-echo MR imaging
(p < 0.000001). However, in four patients with dynamic MR imaging
and in seven patients with enhanced T1-weighted spin-echo MR imaging, the
tumor-cervical stromal contrast was clear because the cervical stroma showed
significant enhancement (Fig.
4A,4B).
|
|
Assessment of Cervical Involvement
The results of assessment of cervical involvement with T2-weighted MR
imaging, dynamic MR imaging, contrast-enhanced T1-weighted spin-echo MR
imaging, and histologic findings are presented in
Table 1. Cervical involvement
was confirmed histologically in 10 patients. Of these patients, in one patient
with a tumor extending just beyond the internal cervical os microscopically, a
false-negative diagnosis was made on the basis of all three MR imaging
sequence interpretations. In the other nine patients with gross cervical
involvement, a correct diagnosis was made on the basis of the dynamic MR
imaging and contrast-enhanced T1-weighted spin-echo MR imaging. Disruption of
enhancement of the cervical epithelium was observed in eight patients on
dynamic MR imaging and in six on contrast-enhanced T1-weighted spin-echo MR
imaging (Fig.
2A,2B,2C,2D).
A less-enhanced tumor replacing the cervical stroma, which had significant
enhancement, was revealed in one patient on dynamic MR imaging and in three
patients on contrast-enhanced T1-weighted spin-echo MR imaging (Fig.
4A,4B).
Using T2-weighted MR imaging, reviewers made a correct diagnosis in eight
patients (Figs.
2A,2B,2C,2D
and
4A,4B)
and a false-negative diagnosis in one patient because of motion artifacts,
which resulted in poor image quality. Of 29 patients without histologic
findings of cervical involvement, a false-positive diagnosis was made with
T2-weighted MR imaging in four patients. The cervical canal was widened
because of a large tumor extending near the internal cervical os in two
patients (Fig.
3A,3B,3C,3D),
a polypoid tumor descending to the cervical canal in one, and the presence of
an endometrial polyp protruding into the cervical canal in one. Of these four
cases, three were correctly classified as having no cervical involvement using
the finding of continuous enhancement of the cervical epithelium with dynamic
MR imaging (Fig.
3A,3B,3C,3D).
The diagnosis of no cervical involvement was correctly made in only one
patient with contrast-enhanced T1-weighted spin-echo MR imaging. The remaining
25 patients were correctly classified as having no cervical involvement with
all three MR imaging sequences.
|
The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of the three MR imaging techniques in assessing cervical involvement are given in Table 2. Although the diagnostic accuracy in assessing cervical involvement with dynamic MR imaging was greater than that with T2-weighted MR imaging and contrast-enhanced T1-weighted spin-echo MR imaging, there was no statistically significant difference. On the other hand, the specificity and positive predictive value with dynamic MR imaging were higher than those with T2-weighted MR imaging and contrast-enhanced T1-weighted spin-echo MR imaging. This was attributed mainly to a decrease in the number of false-positive cases with dynamic MR imaging.
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Several authors have reported that T2-weighted MR imaging using conventional spin-echo sequences was useful in assessing cervical involvement, with diagnostic accuracy reported to be 85-90% [9,10,11,12]. In addition, recent technical improvement in MR imaging using RARE sequences with a phased array coil has significantly improved signal-noise ratios and allowed a high-resolution matrix to be obtained [13, 14]. In this study, the accuracy in assessing cervical involvement using T2-weighted RARE sequences with a phased array coil was 85%, which was not an improvement over the results reported in the literature using T2-weighted conventional spin-echo MR imaging sequences with a body coil. In addition, the tumor-cervical epithelial contrast was poor on T2-weighted MR imaging using RARE sequences, even though the tumor-cervical stromal contrast was excellent, which led to false-positive diagnoses in patients with a large tumor near the internal cervical os or a polypoid tumor protruding into the cervical canal. These findings suggest that it is sometimes difficult to distinguish tumors with no cervical involvement from those with invasion to the cervical epithelium or superficial cervical stroma using T2-weighted MR imaging, even when RARE sequences with a phased-array coil are used.
In our study, the specificity and positive predictive value for assessing cervical involvement with dynamic MR imaging were higher than those with T2-weighted MR imaging and contrast-enhanced T1-weighted spin-echo MR imaging but were not statistically different. On dynamic MR imaging, enhancement of the cervical epithelium was useful in evaluating cervical involvement. In particular, in false-positive cases with T2-weighted MR imaging caused by a dilated cervical canal without cervical involvement, the correct diagnosis of no cervical involvement could be made with dynamic MR imaging using the finding of continuous enhancement of the cervical epithelium. In most cases, enhancement of the cervical epithelium was greater than that of the tumor and cervical stroma especially, in phase 2 or 3 of dynamic MR imaging. The tumor-cervical epithelial contrast on dynamic MR imaging was superior to that on enhanced T1-weighted spin-echo MR imaging. Therefore, we consider that, in combination with T2-weighted MR imaging, dynamic MR imaging is a reliable method for the assessment of cervical involvement.
When the tumor directly invades the cervical stroma without invasion of the cervical epithelium, continuous enhancement of the cervical epithelium may be observed with dynamic MR imaging, which may lead to a false-negative diagnosis. However, Bigelow et al. [21] reported invasion of the cervical epithelium in most patients with endometrial carcinoma involvement of the cervical stroma. This fact indicates that enhancement of the cervical epithelium on dynamic MR imaging can be used as a reference for the assessment of cervical involvement in most cases.
In conclusion, we believe that dynamic MR imaging is helpful in improving the assessment of endometrial carcinoma involvement of the cervix. In particular, we consider that enhancement of the cervical epithelium on dynamic MR imaging is a reliable reference in assessing cervical involvement.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  E. Sala, S. Wakely, E. Senior, and D. Lomas MRI of Malignant Neoplasms of the Uterine Corpus and Cervix Am. J. Roentgenol., June 1, 2007; 188(6): 1577 - 1587. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. E. J. Pijl, J. Doornbos, M. N. J. M. Wasser, H. C. van Houwelingen, R. A. E. M. Tollenaar, and J. L. Bloem Quantitative Analysis of Focal Masses at MR Imaging: A Plea for Standardization Radiology, June 1, 2004; 231(3): 737 - 744. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Manfredi, P. Mirk, G. Maresca, P. A. Margariti, A. Testa, G. F. Zannoni, D. Giordano, G. Scambia, and P. Marano Local-Regional Staging of Endometrial Carcinoma: Role of MR Imaging in Surgical Planning Radiology, May 1, 2004; 231(2): 372 - 378. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) exceeded that of cervical stroma ([UNK]) and endometrial
carcinoma (
), which was especially significant in phase 2 (80-160 sec)
or phase 3 (160-240 sec).
