| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Toride Kyodo General Hospital, 2-1-1 Hongo,
Toride-shi, Ibaraki 302-0022, Japan.
2 Department of Radiology, School of Medicine, Tokyo Medical and Dental
University, Tokyo 113-8549, Japan.
3 Department of Human Pathology, Graduate School, Tokyo Medical and Dental
University, Tokyo 113-8549, Japan.
4 Department of Pathology, Tokyo Metropolitan Geriatric Medical Center, Tokyo
173-0015, Japan.
5 Department of Obstetrics and Gynecology, School of Medicine, Tokyo Medical and
Dental University, Tokyo 113-8549, Japan.
Received December 11, 2002; accepted after revision September 25, 2003.
Address correspondence to K. Shimada
(ken-shimada{at}mse.biglobe.ne.jp).
OBJECTIVE. Our purpose was to clarify the relationship between the tissue vascularity shown on triple-phase dynamic MRI and the number of intratumoral vessels and degree of hyalinization, which are two histopathologic changes in leiomyoma.
SUBJECTS AND METHODS. The subjects were 10 premenopausal patients with 20 leiomyomas who had undergone surgery without preoperative gonadotropin-releasing hormone analogue treatment. Intratumoral vessel density was determined by the mean number of intratumoral vessels with at least one smooth-muscle layer in the optic fields magnified 100 times. Hyalinization grade was determined by the severity of hyalinization, histopathologically classified in three grades. The enhancement index (EI) of the leiomyoma was calculated using the formula EI(t) = [S(t) – S(0)] / S(0), where S(0) is the signal intensity on enhanced T1-weighted images and S(t) is the signal intensity on each dynamic phase image (t = 20, 60, and 180 sec). The histopathologic parameters of intratumoral vessel density and hyalinization grade were compared with the enhancement indexes obtained from the triple-phase dynamic MRI.
RESULTS. We found positive correlations between intratumoral vessel density and EI60 and between intratumoral vessel density and EI180 (in both cases, p = 0.0028 and r = 0.69). We found significant differences among the mean enhancement indexes for each hyalinization grade at all dynamic phases (p < 0.01). The leiomyomas with lower intratumoral vessel densities tended to show greater hyalinization.
CONCLUSION. Our results showed that leiomyomas with only slight hyalinization or with abundant vessels were well enhanced, but the leiomyomas with severe hyalinization enhanced poorly.
Uterine leiomyoma is a common pelvic tumor, with an incidence of 20–50% in women 35 years old or older [1]. Most leiomyomas are asymptomatic [2]. However, some patients may present with abnormal uterine bleeding, pressure on adjacent organs, pain, infertility, or a palpable abdominal pelvic mass [2]. The standard treatments for such symptomatic uterine leiomyomas are hysterectomy and myomectomy [1, 2], but transcatheter uterine artery embolization is now also widely used [1–8].
Jha et al. [1] reported that hypervascular leiomyomas observed on dynamic MRI responded well to treatment by uterine artery embolization. Leiomyomas successively treated by uterine artery embolization showed loss of enhancement on contrast-enhanced MRI [1–5]. However, the detailed histopathologic changes responsible for leiomyoma vascularity shown on dynamic MRI remain unknown.
Uterine artery embolization treatment directly embolizes bilateral uterine arteries and causes hyaline degeneration in the leiomyoma [6, 7]. Therefore, we focused on two kinds of histopathologic changes in the leiomyoma: the number of vessels and the degree of hyalinization. We theorized that these histopathologic changes were closely related to the leiomyoma vascularity shown on dynamic MRI.
The purpose of this retrospective study was to clarify the relationship between these two histopathologic changes in the leiomyoma— the number of vessels and the degree of hyalinization—with patterns of enhancement shown on triple-phase dynamic MRI.
Subjects and Methods
We histopathologically evaluated the intratumoral vessel density and hyalinization grade of resected leiomyomas. Leiomyoma vascularity was measured by the relative tissue enhancement shown on triple-phase dynamic MRI. We examined the relationship between intratumoral vessel density and hyalinization grade and the tissue enhancement shown on triple-phase dynamic MRI.
Patients
The subjects were 10 premenopausal patients 29–51 years old (mean,
39.9 years) with 20 histopathologically confirmed uterine leiomyomas. In this
study, only leiomyomas larger than 3.0 cm in diameter were considered. The
patients underwent triple-phase dynamic enhanced MRI at our institution
between August 1999 and August 2000 and were surgically treated within a mean
of 88.6 days (range, 15–241 days) after their MRI procedure and before
November 2000. The patients were not given preoperative gonadotropin-releasing
hormone analogue treatment. The diameter of leiomyomas measured on MR images
ranged from 3.4 to 21.3 cm (mean, 8.3 cm). Eleven of the 20 leiomyomas were
subserosal, eight were intramural, and one was submucosal.
The MRI examination and treatment were initiated after patients were fully informed and gave their consent. Informed consent was also obtained from all patients to use clinical data, imaging data, and specimens for this study.
MRI Protocols
All MR images were obtained using a 1.5-T superconducting whole-body imager
(Magnetom Vision, Siemens) with a circular polarized phased array coil.
Multisection T2-weighted turbo spin-echo imaging was performed using TR/TE, 4,200/120, and a 270 x 512 matrix. All T2-weighted images were obtained at a section thickness of 7 mm and a 20–60% intersection gap.
Triple-phase dynamic MR images were obtained at times of 20 sec (first phase), 60 sec (second phase), and 180 sec (third phase) using a mechanical power injector after the start of a rapid IV injection at 3 mL/sec of gadodiamide hydrate (Omniscan, Daiichi Pharmaceuticals), 0.1 mmol/kg of body weight, immediately followed by a 20-mL saline flush to ensure complete delivery of the contrast medium for the quantitative measurement. The images were obtained using a multisection 2D fast low-angle shot sequence with a TR/TE, 150/4.1; a flip angle of 80°; a 256 x 128 matrix; and 1 excitation. Unenhanced T1-weighted images were also obtained using the same 2D fast low-angle shot sequence before the dynamic MRI. Both the unenhanced T1-weighted images and each dynamic phase image during the triplephase dynamic study were obtained while the patients breath-held for 20 sec. The section thickness was 5–7 mm, and the intersection gap was 10–20%.
Intratumoral Vessel Density and Hyalinization Grade
Histologic sections of leiomyomas stained with H and E were reviewed by an
experienced pathologist who was unaware of the MRI findings. The number of
histopathologic specimens obtained for each leiomyoma ranged from one to five
(mean, 2.0 specimens).
In each specimen, all intratumoral vessels with at least one smooth-muscle layer were counted while magnified 100 times in 10 different optic fields under a BHSU stereo microscope (Olympus Optical). The mean value of intratumoral vessel density for each leiomyoma was then calculated. The intratumoral vessels that we counted consisted of both arterioles and venules. Capillary vessels were not included. The arterioles and venules could not be distinguished from one another while magnified 100 times.
The severity of hyalinization was graded histopathologically as follows: grade 1, slight, with little or no hyalinization (Fig. 1A); grade 2, intermediate, with limited hyalinization in intercellular space (Fig. 1B); grade 3, severe, with diffuse hyalinization in intercellular space and displacement of myoma cells (Fig. 1C).
|
|
|
Enhancement Indexes of Triple-Phase Dynamic MRI
The enhancement index (EI) of uterine leiomyomas was calculated
using this formula:
 |
Circular regions of interest were placed within the confines of the uterine leiomyoma on several sections, and the signal intensity was measured. The average value for each leiomyoma was then calculated.
Data Analysis
We evaluated the correlations between intratumoral vessel density and each
enhancement index using Spearman's rank correlation test. Hyalinization grade
and enhancement indexes were compared using the Kruskal-Wallis test (overall
test) and Mann-Whitney U test (comparison between two sets of
parameters). Hyalinization grade and intratumoral vessel density were compared
using the same tests. Differences with two-tailed p values of less
than 0.05 were considered significant.
Results
Rapid growth was not clinically observed in the uterus of any patient between the time of MRI examination and surgery. The size of each resected leiomyoma roughly coincided with the size measured on MRI. At histopathologic examination, none of the leiomyomas showed increased mitotic activity, severe cystic degeneration, red degeneration, or myxoid degeneration. The intratumoral vessels in the leiomyomas were approximately 30–200 µm in diameter.
The peak enhancement on the triple-phase dynamic MRI occurred during the third phase (180 sec) in all leiomyomas except for one leiomyoma with severe hyalinization that showed peak enhancement at the second phase (60 sec).
Intratumoral Vessel Density and Enhancement Index
A positive correlation was observed between intratumoral vessel density and
the EIs at 60 and 180 sec (p = 0.0028 and r = 0.69
in both cases). However, the intratumoral vessel density was not correlated
with the enhancement index at 20 sec (p > 0.05, r =
0.44). Figure 2 is a
scatterplot showing the significant positive correlation between intratumoral
vessel density and EI60.
|
Hyalinization Grade and Enhancement Index
At each dynamic phase, the leiomyomas with severe (grade 3) hyalinization
showed the smallest mean enhancement index values, followed by the leiomyomas
with intermediate (grade 2) hyalinization and the leiomyomas with slight
(grade 1) hyalinization (Fig.
3). At all dynamic phases, we found statistically significant
differences among the enhancement indexes for each hyalinization grade
(p < 0.01 or 0.05).
|
When the enhancement indexes of the leiomyomas with severe (grade 3) hyalinization were assumed to be less than 0.1 at the first (20 sec) and the second (60 sec) phase and 0.2 at the third (180 sec) phase, we could differentiate them from the leiomyomas with intermediate (grade 2) and slight (grade 1) hyalinization.
Moreover, for differentiation of the leiomyomas with slight (grade 1) hyalinization from the leiomyomas with more severe (grades 2 and 3) hyalinization, the use of threshold EI20 value of 0.6, EI60 value of 1.0, and EI180 value of 1.2 resulted in an accuracy rate of 85%, 90%, and 90%, respectively.
Hyalinization Grade and Intratumoral Vessel Density
The intratumoral vessel density values overlapped considerably among the
three grades of hyalinization, especially between the leiomyomas with
intermediate hyalinization and the leiomyomas with slight hyalinization
(Fig. 4). However, the mean
intratumoral vessel density of the leiomyomas with severe (grade 3)
hyalinization was the smallest, followed by that of leiomyomas with
intermediate (grade 2) hyalinization and that of leiomyomas with slight (grade
1) hyalinization.
|
Figure 5A, 5B, 5C, 5D, 5E, 5F illustrates a typical case of multiple leiomyomas with a low intratumoral vessel density value and severe (grade 3) hyalinization. Figure 6A, 6B, 6C, 6D, 6E, 6F shows high intratumoral vessel density and slight (grade 1) hyalinization. Leiomyomas with low intratumoral vessel density and severe hyalinization showed little enhancement on any phase of triple-phase dynamic MRI (Figs. 5C, 5D, 5E), but leiomyomas with high intratumoral vessel density and only slight hyalinization showed marked enhancement even on the early dynamic phase images (Figs. 6C, 6D, 6E).
|
|
|
|
|
|
|
|
|
|
|
|
Discussion
Dynamic contrast-enhanced MRI with a gadolinium contrast agent has been widely used for assessment of human tumors [9–12]. Dynamic MRI has been reported to be useful for detecting cellular uterine leiomyomas, which show a good response to gonadotropin-releasing hormone analogue treatment and good enhancement on early dynamic phase images [9]. Therefore, we have been routinely performing triple-phase dynamic MRI for examinations of uterine leiomyomas. In a triplephase dynamic study, we considered that the first-phase (20 sec), the second-phase (60 sec), and the third-phase images (180 sec) corresponded to predominantly early arterial phase, late arterial or venous phase, and equilibrium phase images for the uterus, respectively. We used this imaging technique to examine the relationship between the histopathologic changes of the leiomyomas and tissue vascularity shown on contrast-enhanced MRI.
First, we compared intratumoral vessel density with the tissue enhancement shown on triple-phase dynamic MRI. Our results revealed a significant positive correlation between intratumoral vessel density and EI60 and between intratumoral vessel density and EI180. These findings clearly indicate that the quantity of intratumoral vessels in the leiomyoma affects the degree of enhancement at 60 and 180 sec. The correlation coefficient between intratumoral vessel density and EI60 was the same as that between intratumoral vessel density and EI180, although we had theorized that the correlation coefficient for EI60 would be higher because EI180 reflected the permeability of capillary vessels and as the vascularity of the tumor. The reason for this discrepancy is unclear. One possibility is that this may have been caused by the small sample size. Alternatively, no significant correlation was observed between intratumoral vessel density and EI20. The lower correlation may be caused by the assumption of predominantly arterial phase imaging at 20 sec. Real arterial phase imaging using a bolus test or bolus-tracking mechanism might have resulted in a higher correlation coefficient between intratumoral vessel density and enhancement. Furthermore, intratumoral vessel density included both arterioles and venules. This also seemed to contribute to the lower correlation at the first phase (20 sec).
Next, the degree of hyalinization was compared with the tissue enhancement shown on triple-phase dynamic MRI. Our results showed that the leiomyomas with more severe hyalinization were less enhanced. One explanation is that hyalinization involves the presence of homogeneous eosinophilic bands or plaques in the extracellular space (Figs. 1A, 1B, 1C), which represent accumulation of proteinaceous tissue [2]. Thus, hyalinization causes a narrowing of the extracellular space and prevents the gadolinium contrast agent from entering this area (Figs. 1A, 1B, 1C) because gadolinium is an extracellular agent. Moreover, our results indicated that hyalinization and quantity of vessels were likely to interact: fewer intratumoral vessels resulted in more severe hyalinization. This finding may explain why devascularization by uterine artery embolization causes hyalinization.
Accordingly, our results showed that the leiomyomas with only slight hyalinization or abundant vessels were well enhanced and that the leiomyomas with severe hyalinization were poorly enhanced. From these results, we can estimate vessel density and hyalinization grade in the leiomyoma using dynamic MRI— that is, well-enhanced leiomyomas would show only slight hyalinization or abundant vessels and poorly enhanced leiomyomas would show severe hyalinization.
The estimation of vessel density and hyalinization grade using dynamic MRI might also be useful for predicting the effectiveness of uterine artery embolization and selecting proper candidates for uterine artery embolization. Leiomyomas with high vessel density are expected to respond well to embolization, but leiomyomas with severe hyalinization seem to respond poorly. However, at this stage, we do not know whether estimating histopathologic changes by triplephase dynamic MRI is practically useful for predicting the effectiveness of uterine artery embolization because we have not yet applied our results to an appropriate sample of patients. Further study is needed to assess this possibility.
This study has other limitations. First, our study included three patients who waited more than 100 days (144, 158, and 241 days) between the MRI examination and surgery. In these patients, the histopathology of the leiomyoma at the MRI examination might have differed from that at surgery. Parker et al. [13] reported that 28% (371/1,332) of their patients with uterine leiomyoma showed rapid growth, although uterine leiomyomas are benign. We, however, found no rapid growth in the uterus of any patient during the time between MRI examination and surgery. The size of each resected leiomyoma roughly coincided with the size that was encountered on MRI examination. Moreover, none of the leiomyomas showed severe cystic degeneration, red degeneration, or myxoid degeneration at histopathologic examination. Therefore, the histopathology of each leiomyoma at MRI examination did not greatly differ from that at surgery.
A second limitation is that the histopathologic specimens comprised only a part of the leiomyomas. The vascular density and degree of hyalinization might not have accurately represented the leiomyoma because of heterogeneity in the tumors. However, we considered that heterogeneity was to some extent averaged by the measurement of intratumoral vessel density and enhancement indexes in our study.
In summary, the quantity of vessels and the degree of hyalinization were closely related to the leiomyoma vascularity shown on triplephase dynamic MRI. The leiomyomas with only slight hyalinization or abundant vessels were well enhanced, but the leiomyomas with severe hyalinization enhanced poorly.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. D. Peddada, S. K. Laughlin, K. Miner, J.-P. Guyon, K. Haneke, H. L. Vahdat, R. C. Semelka, A. Kowalik, D. Armao, B. Davis, et al. Growth of uterine leiomyomata among premenopausal black and white women PNAS, December 16, 2008; 105(50): 19887 - 19892. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T Tamada, T Sone, D Tanimoto, H Higashi, H Miyoshi, N Egashira, A Yamamoto, and S Imai MRI appearance of primary giant ovarian leiomyoma in a hysterectomised woman. Br. J. Radiol., October 1, 2006; 79(946): e126 - e128. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
