| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Vancouver Coastal Health Authority, University
Hospital, 2211 Wesbrook Mall, Vancouver, BC V6T 2B5, Canada.
2 British Columbia Cancer Agency, Vancouver, BC, Canada.
Received August 22, 2005;
accepted after revision October 31, 2005.
Address correspondence to A. L. Spielmann
(Audrey.Spielmann{at}vch.ca).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Forty-nine women who presented for consultation for uterine artery embolization were retrospectively reviewed. The MRI and sonography scans were independently evaluated and compared for uterine size, fibroid size and location (categorized as paraendometrial, intramural, subserosal, or pedunculated) of the four largest fibroids in each patient, and the total number of fibroids present.
RESULTS. One hundred twenty-two fibroids were measured. The uterine volume was significantly smaller as measured on MRI compared with sonography (p = 0.01). We found good MRI and sonography correlation of the volume of the single largest fibroid in each patient (R = 0.87) but poor correlation of fibroid location (R = 0.17). MRI detected 31 paraendometrial fibroids and three pedunculated fibroids that were thought to be intramural fibroids on sonography. Five fibroids thought to be paraendometrial on sonography were confirmed to be subserosal or intramural on MRI. Discrepancy in the total number of fibroids was noted, with additional fibroids found on MRI in 31 of 49 patients and erroneously suspected on sonography in five of 49 patients. Pelvic MRI affected management in 11 of 49 patients, leading to cancellation of uterine artery embolization in four patients. In another seven patients who were originally thought to be poor candidates on the basis of sonographic findings, uterine artery embolization was performed. MRI did not alter the management plan in 38 patients.
CONCLUSION. MRI provided considerable additional information compared with sonography and affected clinical decision making in a substantial number of patients. MRI should be considered in all patients being evaluated for uterine artery embolization.
Keywords: embolization • fibroids • genitourinary imaging • pelvic imaging • uterus • women's imaging
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
At our institution, all potential candidates for uterine artery embolization undergo evaluation using both sonography and MRI. To determine if pelvic MRI is necessary in this application, we compared the sonography and MRI findings in a group of potential candidates to evaluate whether the information provided by MRI affected treatment decisions with regard to the suitability of uterine artery embolization in individual patients.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Unenhanced MRI was performed with a 1.5-T Signa Horizon Echospeed scanner (GE Healthcare) using a torso phased-array coil. Buscopan (butylscopolamine bromide, Boehringer Ingelheim), 20 mg, was administered intramuscularly before the examination to reduce bowel peristalsis, and a tampon was inserted. The patient was restricted to clear fluids 4 hours before the examination and voided immediately before the study. The protocol included a sagittal T1-weighted sequence (TR range/TE, 450-800/20, 28-cm field of view, 256 x 256 matrix, and 2 excitations) and oblique coronal, axial, and sagittal fast spin-echo T2-weighted sequences with reference to the long axis of the uterus (TR/effective TE, 4,000/100, 20- to 24-cm field of view, 512 x 256 matrix, echo-train length of 8-16, and 2 excitations). All scans were obtained at 5-mm thickness with a 2-mm gap. Fat saturation using selective presaturation of lipid resonance frequency was used on the oblique axial and sagittal fast spinecho T2-weighted sequences.
Sonography was performed on an HDI 3000 or 5000 ATL machine (Advanced Technology Laboratory) with a curved array 2-4-MHz transducer (Advanced Technology Laboratory) by experienced technologists using a dedicated fibroid worksheet specifically designed for preembolization evaluation. Transabdominal scanning with a full bladder was performed on all patients with the addition of transvaginal scanning in two patients. Transabdominal scanning was thought to be more useful than transvaginal scanning in most patients because of the large bulk of the uterus. The addition of transvaginal scanning was at the discretion of the attending sonography radiologist at the time of scanning. All studies were reviewed on a PACS workstation.
All pelvic MRI and pelvic sonography studies were independently reviewed, without knowledge of the imaging findings from the other technique, by two abdominal and MRI fellowship-trained radiologists. The purpose of this study was to compare sonography with MRI. MRI was effectively used as the gold standard because no objective third gold standard was available in this nonsurgical population. Differences of opinion, which were minor, were resolved by consensus. The MRI and sonography scans were evaluated for the following features: uterine volume measured using a prolate ellipse equation (sagittal x transverse x anteroposterior x 0.5233); the volume and the single longest diameter of the largest fibroids (up to four) in each patient; location of the fibroids, categorized as paraendometrial, intramural, subserosal, or pedunculated; and the total number of fibroids present.
For the purposes of our study, a fibroid was deemed to be paraendometrial if any portion of the tumor contacted the endometrium, even if most of the fibroid extended into the myometrium—that is, an intramural fibroid with endometrial contact. All submucosal fibroids with greater than 50% contact with the endometrium were also included in this group. Although this definition is not of relevance with respect to determining suitability of hysteroscopic resection, it is important when assessing patients before uterine artery embolization. Patients with even a minor endometrial contact are at an increased risk of cavitation of the fibroid with prolonged vaginal passage of tissue, bleeding, or discharge; and thus identification of an endometrial contact allows appropriate and informed counseling of potential candidates for uterine artery embolization in regard to their chances of a satisfactory and complication-free outcome.
In instances of discrepancy of fibroid position between sonography and MRI, MRI was considered correct. The ability to detect the ovaries and the presence of ovarian disorders was compared. Junctional zone thickness and the presence of adenomyosis were also documented on the MRI studies. The junctional zone is the deepest layer of the myometrium adjacent to the endometrium, which is hypointense. Adenomyosis was considered to be present if the junctional zone thickness in any portion measured greater than 12 mm. The sonographic diagnostic criteria of adenomyosis include diffuse uterine enlargement with a normal uterine contour, thickened heterogeneous myometrium in the absence of a discrete mass, streaking, and cysts in the myometrium.
The data were organized matching the MRI and sonography measurements of the fibroids in descending order from largest to smallest volume except when there was an obvious match in location of measured fibroid on the sonography and MRI. This step was done in an attempt to most accurately match the volume data and the fibroid location. The data were evaluated statistically using a two-way analysis of variance. The kappa statistic was used for the comparative measurements of MRI and sonography, and linear regression analysis was performed.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The total number of fibroids seen on MRI and sonography was discrepant in 36 of 49 patients. Additional uterine fibroids (n = 1-26) were detected on MRI in 31 of 49 patients, and additional fibroids were erroneously reported on sonography in five of 49 patients. The maximum length of the four or fewer largest fibroids was measured in each patient on both MRI and sonography, for a total of 122 fibroids. The mean maximum fibroid length was 8.4 cm on MRI (range, 1-16.6 cm). The size discrepancy between the maximum fibroid length on MRI and that on sonography is shown on Figure 1. These discrepancies were not statistically significant (p = 0.16). The volume of the single largest fibroid per patient measured on MRI and sonography correlated well (R = 0.87). However, the correlation was less strong when the average volume of the four or fewer largest fibroids was compared (R = 0.64). More often, sonography tended to underestimate the fibroid volume in comparison with MRI (n = 72/122).
|
|
|
|
|
The average junctional zone thickness of the uteri in this study was 6.4 mm (range, 2-19 mm). Adenomyosis was present in the uterus in five (10%) of 49 patients as diagnosed on MRI. Adenomyosis was not diagnosed on sonography in any patient, probably because of the small number of transvaginal scans obtained and the increased difficulty of diagnosing adenomyosis in combination with uterine fibroids. The average junctional zone thickness in the patients with adenomyosis was 16.1 mm (range, 13.5-19 mm). The ovaries were identified in 82% (80/98) of patients on MRI and 39% (38/98) on sonography. Ten ovarian lesions were detected on MRI, including eight simple or small hemorrhagic cysts, one dermoid, and a probable endometrioma. Only one cyst was detected on sonography but it was shown to be a dermoid on MRI.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although early in the development of this treatment patients underwent uterine artery embolization for fibroids with only sonography for preprocedure evaluation, preprocedure MRI is increasingly being performed. The superior soft-tissue resolution of MRI compared with sonography should allow more accurate determination of uterine and adnexal anatomy. The results of our study show the impact that this further information has on the evaluation of potential candidates for uterine artery embolization and reinforce the belief that preprocedure MRI is necessary before the performance of uterine artery embolization.
The size of the largest fibroid has been shown in a number of studies to be inversely related to the success of uterine artery embolization, which is defined as symptomatic relief, fibroid volume reduction, and patient satisfaction [6-9]. In a study of 200 patients, Spies et al. [6] reported that a smaller baseline size and a submucosal location are most likely to result in a positive outcome. Larger baseline dominant fibroid volume predicted less volume reduction at both the 3- and the 12-month follow-up examinations [6]. Pelage et al. [10] suggested that uterine artery embolization should be declined if one or more fibroids are larger than 10 cm.
The literature on this subject is not unanimous, however. Other earlier, smaller studies have reported poor correlation between initial uterine and fibroid volumes and therapeutic outcome [11-13]. More recently, a study of 47 women treated with uterine artery embolization for fibroids measuring 10 cm or larger (some as large as 19 cm) showed no difference in complications (both minor and major) when those women were compared with a group of women with smaller fibroids (n = 105) [14]. The authors reported no statistical difference in the postprocedure recovery, symptom control, and tumor and uterine reduction rates. However, the improvement of menorrhagia at 4 and 12 months was significantly inferior in the group with the larger fibroids. In our study, it did not appear that the use of MRI significantly affected the evaluation of the size of dominant fibroids, with one exception: One large (16.6 cm) fibroid exceeded the field of view of the sonography transducer, and its size was consequently underestimated (12.5 cm) on sonography.
The position of uterine fibroids in the uterine wall is clearly a predictor of complication after uterine artery embolization. Patients with submucosal fibroids treated with uterine artery embolization are thought to be at an increased risk for the vaginal passage of necrotic tissue, which can result in significant pain, bleeding, infection, and vaginal discharge for prolonged periods of time [15-17] (Figs. 3A and 3B). In a large series, vaginal expulsion of fibroids was noted in 18 (4.4%) of 408 patients who underwent uterine artery embolization for fibroids [17]. In the largest series to date reviewing overall complication rates of uterine fibroid embolization, Spies et al. [6] found that the most common complication requiring hospitalization was the passage of fibroid tissue, often accompanied by pain, infection, or bleeding. This complication appears to occur exclusively in patients with submucosal fibroids or those intramural fibroids that have a large submucosal component. The suitability of uterine artery embolization in patients with submucosal fibroids is debatable. However, patients can be better counseled regarding the risks of this procedure if the position of the fibroid in the uterine wall is assessed beforehand. Our study has shown that accurate assessment of fibroid position requires pelvic MRI.
Pedunculated subserosal fibroids pose a risk of infarction of the stalk of the fibroid and subsequent migration of the fibroid into the peritoneal cavity, which could possibly result in parasitization of the blood supply from the peritoneum or elsewhere, necrosis, or sepsis. Pedunculated subserosal fibroids are considered to be a relative contraindication to uterine artery embolization, depending on the fibroid size and the size of the stalk, although the inclusion criteria for uterine artery embolization vary considerably among institutions. Identification of pedunculated fibroids, particularly those with a narrow stalk, before uterine artery embolization is critical and in our experience can only be confidently achieved using MRI.
The diagnosis of adenomyosis can be difficult on sonography, particularly in the setting of uterine fibroids. The use of transvaginal sonography has been shown to improve the sonographic diagnosis of adenomyosis, but authors caution that predicting the presence and severity of adenomyosis is limited, even with transvaginal sonography, when fibroids are present [18-20]. Adenomyosis was present in 10% of the women in our study as diagnosed on MRI. The sensitivity and specificity of MRI in the diagnosis of adenomyosis have been reported to be as high as 88-93% and 67-91%, respectively, in the absence of fibroids [21-24].
Uterine artery embolization in the setting of isolated or dominant adenomyosis or adenomyosis combined with fibroids has been shown to result in a more variable response to the therapy [7, 11, 25, 26]. In a study by Goodwin et al. [11], adenomyosis was diagnosed at postsurgical histopathology in three of six patients who required hysterectomies after uterine artery embolization because of inadequate response to the procedure. Their patients were imaged with sonography before the embolization. The adenomyosis tissue in the hysterectomy specimens was viable and showed no evidence of infarction.
A recent study of 30 women that evaluated the performance of uterine artery embolization for adenomyosis with fibroids (n = 27), adenomyosis as the dominant disease (n = 6), and adenomyosis alone (n = 3) showed more favorable results in the setting of adenomyosis [27]. Those authors reported a decrease in the junctional zone thickness with ischemic change in the adenomyosis in some cases, particularly when the disease was focal, asymmetric, and severe. Most of their patients reported improvement in their symptoms. Those authors recommended that concomitant adenomyosis should not be considered a contraindication to uterine artery embolization for fibroids and that uterine artery embolization should be considered for adenomyosis alone. Clearly, further study is required to determine the therapeutic result of uterine artery embolization in the setting of adenomyosis. However, knowledge of the presence of adenomyosis before the procedure is important in order to inform the patient that the response can be more variable in this situation.
The presence of concurrent uterine, ovarian, or pelvic disorders is important in the preprocedure evaluation of women for uterine artery embolization of fibroids. Not only can symptoms of other disorders such as endometriosis, ovarian tumor, or chronic pelvic infection mimic those of uterine fibroids, but also other disease processes may be incidentally found in the initial workup. Clearly, the therapeutic options may be altered if another significant disorder is present. Pelvic MRI is superior to sonography in the diagnosis of ovarian disease, endometriosis, and other pelvic disorders, particularly in the presence of uterine fibroids [19].
Our study has a number of limitations. Although MRI was used as the gold standard in this study, in the absence of histopathologic correlation of our imaging findings our results are not 100% accurate and lead to a bias against sonography. To our knowledge, no prior studies have compared the accuracy of pelvic MRI and sonography with histologic correlation in evaluating uterine fibroids. However, prior histologic studies have shown that MRI is an accurate technique for this application [2]. Sonography is operator-dependent, and the radiologist reviewing the sonograms relies on the technologist to document the visualized fibroids appropriately on the images and the dedicated fibroid study work-sheet. The technologists at our institution are experienced and aware of the importance of accurately assessing fibroids. We believe that the discrepancies described are a limitation of the technique rather than the operator. In addition, it is difficult to know with certainty that the same fibroids were measured and compared on MRI and sonography when more than one fibroid is present. Not only were the fibroids compared by size, largest to smallest, but every effort was also made to correlate the position of the fibroids in the uterus on MRI and sonography when multiple fibroids of similar size were present.
The MRI protocol in this study did not include gadolinium-enhanced sequences, and fibroid perfusion was not a focus of this study. In our institution, gadolinium-enhanced sequences are performed after uterine artery embolization if the response to treatment is less than expected. Routine postprocedure MRI is not performed because of the more restricted access to MRI in Canada. Recent evidence indicates that fibroid perfusion, or lack thereof, after uterine artery embolization is a better predictor of outcome than a change in individual fibroid volume or absolute uterine volume [28]. The ability to assess fibroid perfusion with MRI is an added benefit of that technique.
In conclusion, it is important to obtain detailed information of the presence and nature of uterine fibroids to adequately counsel women as to the risks and benefits of uterine artery embolization. Sonography is widely available, can accurately confirm the presence of uterine fibroids and uterine enlargement, and should be the first line of imaging. However, the preprocedure evaluation of the size, number, and location of uterine fibroids can be made with much greater accuracy using pelvic MRI as compared with sonography. Additional information, such as the presence of adenomyosis and other pelvic disorders, is also revealed using MRI. With the paradigm shift in the clinical approach to the less invasive treatment of uterine fibroids using uterine artery embolization, precise imaging with pelvic MRI is necessary, and it is routinely performed in our practice.
Acknowledgments
We thank David Jung for his help with statistical analysis of the data.
Thanks also to Marcia Lepore for aiding in the typing of the manuscript.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  G. Tropeano, S. Amoroso, and G. Scambia Non-surgical management of uterine fibroids Hum. Reprod. Update, May 1, 2008; 14(3): 259 - 274. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A N Keeling and J F Reidy Imaging and treatment of uterine fibroids, including the role of uterine artery embolization Imaging, December 1, 2007; 19(4): 374 - 384. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
