| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Genitourinary Imaging |
1 From the Departments of Radiology (S.D., S.M.H., G.P.K.) and Epidemiology and Biostatistics (W.C.J.H.), Erasmus Medical Center, Dr Molewaterplein 40, 3015 GD Rotterdam, the Netherlands. From the 2001 RSNA scientific assembly. Received September 16, 2002; revision requested November 26; final revision received July 14, 2003; accepted August 22. Address correspondence to S.M.H. (e-mail: smhussain62@hotmail.com).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: In a retrospective descriptive study, two radiologists systematically reviewed MR findings in 20 patients with a clinically proved anovaginal fistula and looked for the main fistula tract, the internal opening in the anal canal and/or vagina, secondary fistula tracts, abscesses within the rectovaginal septum, and sphincter damage. Interobserver variability was calculated, and clinical records were searched for possible underlying causes that could explain the complexity of anovaginal fistulas. The 
value was calculated. Patients with or without a complex anovaginal fistula were compared in regard to the presence of any underlying disease or condition. Statistical significance was calculated with the Fisher exact test.
RESULTS: In all 20 patients, anovaginal fistulas were identified on T2-weighted MR images as predominantly high-signal-intensity linear abnormalities extending between the anal canal and the vagina. In all patients, the fistulas were typically located in the sagittal plane, and the mean distance from the anal verge to the fistula was 25.0 mm (range, 13–32 mm). The internal opening in the anal canal was detected in all patients. The internal opening in the vagina was detected in 19 (95%) patients. In seven (35%) patients, an anovaginal fistula with an additional abnormality was found and included an abscess within the rectovaginal septum (n = 1), a perianal fistula (n = 3), and a perianal fistula in combination with an abscess (n = 3). Defects of the external anal sphincter were present in three (15%) patients. There was complete agreement between observers for all items on endoanal MR images, except for the presence of secondary fistula extensions (agreement, 90%; , 0.74). History of obstetric trauma, pelvic floor surgery, or Crohn disease was present in 10 (50%) patients. Of these patients, six (60%) had a complex anovaginal fistula and four (40%) had a simple anovaginal fistula. In the remaining 10 patients without relevant medical history, one (10%) had a complex anovaginal fistula. This difference tended toward statistical significance (P = .057).
CONCLUSION: Endoanal MR imaging allows evaluation of anovaginal fistulas and additional abnormalities, such as abscesses within the rectovaginal septum, secondary perianal fistula tracts, and sphincter damage.
© RSNA, 2004
Index terms: Anus, abnormalities, 757.411, 757.458 • Anus, MR, 757.121411 • Fistula, genitourinary system, 855.459 • Magnetic resonance (MR), endoscopic, 757.121411, 855.121411 • Vagina, abnormalities, 855.458, 855.459 • Vagina, MR, 855.121411
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Currently, an anovaginal fistula is diagnosed on the basis of typical clinical symptoms, such as recurrent vaginal infection with vaginal flatus and vaginal defecation, and a gynecologic or surgical examination performed with general anesthesia. In many institutions, the clinical diagnosis or suspicion of an anovaginal fistula is confirmed with imaging studies (5). These studies may include conventional fistulography and anal endosonography (6,7). Conventional fistulography may not be possible in many patients because the internal openings of the fistulas may not be visible at physical examination. Mainly because of this problem, fistulographic findings are reported to be correct in only 16% of patients with anal fistulas (6). Endoanal ultrasonography (US) is often inaccurate in the depiction of anovaginal fistulas, mainly because endoanal US has inherent low soft-tissue contrast (7,8).
At our institution, endoanal MR imaging is routinely performed in the assessment of perianal fistulas. The role of endoanal MR imaging in the detection and classification of perianal fistulas is well established (7). Endoanal MR imaging has proved to be an excellent imaging modality in the assessment of anal sphincter anatomy, anal sphincter damage, and perianal fistulas (7,9). This modality provides multiplanar images with high inherent contrast resolution and high spatial contrast resolution of the anal canal, rectum, rectovaginal septum, and vagina (5,9). In particular, T2-weighted MR imaging sequences can depict lesions, such as fistulas and fluid collections, with high signal intensity (10). For surgical planning, the exact location of the main fistula and possible secondary extensions or abscesses must be established, and sphincter damage must be assessed. Thus, the purpose of this study was to evaluate endoanal MR imaging in the assessment of anovaginal fistulas and associated findings.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
We found 36 consecutive patients who were referred because of suspected or proved anovaginal fistula. Patients were included in this study if they had either (a) evidence of anovaginal fistula at exploratory surgery or at examination by the surgeon or gynecologist with anorectoscopy and/or colposcopy or (b) persistent clinical symptoms of recurrent vaginal infections combined with the passage of gas, feces, or both through the vagina. Symptoms were present for more than 3 months.
Sixteen patients were excluded from the study because (a) no information was available concerning any of the inclusion criteria because patients were referred from other institutions (n = 9), (b) MR imaging revealed a perianal fistula instead of an anovaginal fistula that was subsequently confirmed at surgery (n = 5), or (c) anovaginal fistulas could not be confirmed with a gynecologic examination and anosigmoidoscopy (n = 1) or with a barium enema examination and general anesthesia (n = 1). Thus, the remaining 20 patients formed the study group.
In all 407 patients with endoanal MR imaging reports that were negative for anovaginal fistula, a careful search of clinical findings was performed by two authors (S.D., S.M.H.). The follow-up period was 2–5 years. In this survey, we were unable to detect any instances of anovaginal fistulas that were not mentioned initially on the MR imaging report but were subsequently mentioned in the clinical records or proved to exist, either with endoscopic or gynecologic examinations or at surgery or follow-up MR imaging.
The mean age of patients in the study group (n = 20) was 36 years (range, 20–55 years). All patients had anovaginal fistulas that were proved with exploratory surgery (n = 12), persistent clinical symptoms of the passage of gas and feces through the vagina with recurrent vaginal infections (n = 5), or surgery and/or gynecologic consultation and examination (n = 3). Two authors (S.D., S.M.H.) compared MR findings with the detailed descriptions of surgical findings in patient records, and patient follow-up status was assessed with clinical records and follow-up data available in the hospital information system.
MR Imaging
All MR imaging examinations were performed by using an endoanal coil with a 1.5-T MR imager (Gyroscan NT Intera 1.5; Philips Medical Systems, Best, the Netherlands). The endoanal coil is commercially available (Philips Medical Systems) and consisted of a fixed, rectangular, 60-mm-long rigid receiver coil with a width of 16 mm. The coil is contained within an 80-mm-long cylindrical coil holder with a diameter of 19 mm. Before the introduction of the coil into the anal canal, a condom was placed over the coil and US gel was used as a lubricant. The coil was introduced while the patient was lying in the left lateral position. After the coil was introduced, each patient carefully turned onto her back, and the position of the coil was rechecked.
In each patient, the following three sequences were used. Transverse T2-weighted contrast-enhanced fast field-echo imaging was performed (repetition time msec/echo time msec, 23/14; acquisition time, 5 minutes 39 seconds; matrix, 205 x 256; flip angle, 60°; field of view, 140 mm; section thickness, 2 mm with no gaps; and two signals acquired). Contrast-enhanced fast field-echo is a term used by Philips Medical Systems and does not indicate contrast material administration. Transverse T2-weighted fast spin-echo [SE] MR imaging was performed with and without fat saturation (5,086/100; acquisition time, 2 minutes 23 seconds; matrix, 186 x 256; flip angle, 90°; field of view, 120 mm; section thickness, 4 mm with a 0.4-mm gap; and three signals acquired). Coronal and sagittal T2-weighted SE MR imaging was performed without fat saturation (2,454/100; acquisition time, 2 minutes 34 seconds; matrix, 186 x 256; flip angle, 90°; field of view, 120 mm; section thickness, 4 mm with a 0.4-mm gap; and four signals acquired).
Image Review
All MR images were transported from local digital media to a viewing station and were systematically and independently reviewed by two radiologists. Interobserver variability was calculated on the basis of the statistics. After calculating the value, images with disagreement were reevaluated with both observers working together, and consensus was reached. One radiologist (S.D.) had 1 year of experience in abdominal imaging, which included MR imaging of the pelvic floor, and the other (S.M.H.) had practiced abdominal imaging for 6 years, with scientific interest in MR imaging of abdominal and anorectal diseases. All MR images were assessed for the following items: (a) the presence of anovaginal fistulas, (b) the localization of anovaginal fistulas, (c) the distance of anovaginal fistulas from the anal verge in the sagittal plane, (d) the internal opening of anovaginal fistulas in the anal canal, (e) the internal opening of anovaginal fistulas in the vagina, (f) the presence of edema or abscesses in the rectovaginal septum, (g) the presence of secondary perianal or other extensions, and (h) internal and external anal sphincter damage. An abscess within the rectovaginal septum was defined as a localized and well-circumscribed dilatation of the main fistula or secondary extension with high signal intensity on the T2-weighted images, while edema was indicated by an ill-defined area of intermediate to high signal intensity around the area of the anovaginal fistula. Sphincter damage was identified as a local area of low signal intensity on the T2-weighted images, which indicated scarring or discontinuity of the internal or external anal sphincters.
In addition, clinical records were searched by two authors (S.D., S.M.H.) for possible underlying conditions or diseases, such as obstetric trauma, pelvic floor surgery, or Crohn disease, that could explain the presence of simple or complex anovaginal fistulas. According to the literature (5), anovaginal fistulas were considered to be simple if only one fistulous tract was depicted that extended between the anal canal and the vagina. Anovaginal fistulas associated with any additional fistulous tracts or abscesses were considered to be complex (11,12).
Statistical Analysis
All MR images were systematically and independently reviewed by two radiologists (S.D., S.M.H.). Interobserver variability was calculated on the basis of the statistics. After calculating the value, the same two observers reevaluated the images with disagreement together, and consensus was reached. The number of patients with a complex anovaginal fistula and the number of patients without were compared with regard to the presence of any underlying disease or condition, and the significance was evaluated by using the Fisher exact test. A P value less than or equal to .05 was considered to indicate a statistically significant difference.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
, 0.74) The main anovaginal fistulas could be identified in all patients on T2-weighted images as predominantly high-signal-intensity linear abnormalities extending between the anal canal and the vagina. The fistulas were typically located in the mid- or parasagittal plane, just superior to the external anal sphincter (Fig 1). The mean distance from the anal verge to the fistula was 25.0 mm (range, 13–32 mm). In all 20 patients (100%), the internal opening of the main fistula tract in the anal canal was clearly identified. In 19 patients (95%), the internal opening in the posterior wall of the vagina could be seen. One patient had extensive edema within the rectovaginal septum, which caused difficulties in the identification of the internal opening in the vagina.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Anovaginal fistulas have a relatively thin low-signal-intensity fibrous wall, with a small amount of fluid or air present in the tract (7,14). At MR imaging, anovaginal fistulas appear as areas of high signal intensity on the T2-weighted images and often contain low-signal-intensity bubbles of air (5,7,14,15). A thin rim of low signal intensity surrounding the fistula tract illustrates the fibrous wall (7,14,15). Relatively less active tracts that contain less fluid and more scar tissue may appear as linear abnormalities with predominantly low signal intensity (7,14,15).
Previously, a number of other imaging modalities, including fistulography and transrectal and endoanal US, have been used in the work-up of patients with anovaginal fistula (10). In our experience, fistulography is often an inadequate procedure, because the internal opening of the fistulas in the anal canal and vagina are not always visible at physical examination (6). This makes cannulation for fistulography impossible. Also, this technique does not allow depiction of the anal sphincter, so the fistula cannot be imaged in relation to the anal sphincter complex. It is reported that a correct diagnosis is made with fistulography in only 16% of patients with anal fistulas (6).
Almost 20 years ago, transrectal and endoanal US were introduced and resulted in improved imaging and classification of anal fistulous disease (10,13). Yee et al (3) described a retrospective review of records of 25 female patients who underwent endoanal US before rectovaginal fistula repair. They concluded that transrectal US is not useful for imaging rectovaginal fistulas and cannot be recommended as a diagnostic or screening tool in the identification of rectovaginal fistulas. They did, however, recommend that transrectal US be performed before surgery in all patients with a known rectovaginal fistula to help identify and map occult sphincter defects (3).
A comparative study of endoluminal US and endoluminal MR imaging in 28 patients revealed that classification is better with endoluminal MR imaging (7).
Currently, MR imaging has become a promising modality for imaging of the complex anatomy of the pelvic floor and related structures. MR imaging has an excellent soft-tissue contrast resolution and multiplanar imaging capability, without the need for ionizing radiation (9,10). In our experience, however, the limited spatial resolution of the body coil and other external surface coils prevents identification of small fistula tracts and their relation to the normal anatomic structures (10,14). Developments in coil technology have resulted in phased-array coils, which provide higher signal-to-noise ratios and can be used to improve the spatial resolution (16). Blomqvist et al (17) demonstrated that use of an endorectal coil results in a more detailed depiction of the rectal wall than does use of phased-array coils. Endoluminal MR imaging further improves the signal-to-noise ratio in combination with relatively smaller fields of view (9). This provides exquisite images of the rectum, anal canal, rectovaginal septum, and vagina, with much higher spatial resolution compared with the external surface coils (5,7).
The role of endoanal MR imaging in the detection and classification of perianal fistulas has been reported previously (7). Our results show that endoanal MR imaging can depict anovaginal fistulas, as well as a number of associated abnormalities that are important for surgical management. The disadvantages of endoanal MR imaging include mild discomfort at introduction of the coil and a limited field of view due to signal drop off at some distance from the coil (9). In our study, however, the field-of-view limitations did not pose any problems. In all patients, the anovaginal fistulas were depicted to their full extent. Larger complex fistulas may extend outside the sensitive region of the endoanal coil. In these patients, additional sequences can be performed with body or phased-array coils for complete depiction of the fistulas.
Recently, the value of phased-array coil MR imaging in the identification and classification of anal fistula disease has been reported (16). The use of phased-array coils is advantageous because there is no need to introduce the coil into the anal canal or the rectum, whereas the larger field of view can provide images of the entire lower pelvic region. In our opinion, MR imaging with a phased-array coil may identify most of the larger fistula tracts. The depiction of smaller primary and secondary tracts and anatomy may be problematic with phased-array coils. Endoanal MR imaging proved to be superior to endoanal US and MR imaging with body coils in the detection and classification of anal fistula disease (7).
In our experience, T2-weighted MR imaging sequences are sufficient for depiction of fistulas and the surrounding anatomy. At our institution, we do not routinely use contrast material with MR imaging in the evaluation of patients with fistula disease, as doing so would lengthen the examination time and increase the cost.
Our results indicate that complex anovaginal fistulas are more often seen in patients with a known underlying disease or condition, namely, Crohn disease, previous obstetric trauma, or pelvic floor surgery, which could explain the presence of the fistulas. This difference is not statistically significant, however, and is probably due to the small sample size. Previously, Senatore (18) stated that obstetric injury is the most common cause of anovaginal fistula. Hull and Fazio (11) found that 48 (55%) of all patients who underwent surgery for anovaginal fistula had Crohn disease. Both can be considered important causes of complex anal fistula disease.
The results of our study indicate that it is important to comprehensively evaluate patients before surgery. Complete mapping of the fistula with extensions, abscesses, and sphincter damage is important in the preoperative work-up of the patient and may improve treatment (15,20). Scholefield et al (19) found that preoperative MR imaging was of little use in the surgical treatment of perianal fistulas. Beets-Tan et al (20), however, found that the largest additional value of preoperative MR imaging with the phased-array coil was obtained in patients with complex fistulas associated with Crohn disease and recurrent perianal fistula. Lunniss et al (15) suggested that MR imaging could depict more extensions than surgical exploration.
Preoperative evaluation can help reduce the risk of procedure-dependent complications. This could eventually lead to a reduction in the recurrence of fistulas and the occurrence of postoperative fecal incontinence. In our preliminary review, we also found five patients who were referred for MR imaging in whom an anovaginal fistula was suspected. In these patients, MR imaging showed a perianal fistula instead of an anovaginal fistula, a finding that was subsequently confirmed at surgery. It is likely that the additional information obtained with MR imaging will improve the surgical results in patients with complex fistulas, although this can be reliably established only with a prospective randomized trial of patients who undergo surgery after undergoing preoperative MR imaging and those who undergo surgery without undergoing preoperative MR imaging (20).
Limitations of our study are its retrospective design, the small number of patients, and the lack of a single uniform standard of reference. The observers may have known some of the findings at MR imaging; however, this could not be avoided due to the retrospective nature of our study. We believe the standard of reference should be exploratory surgery, because this is the only clinical technique that may be used to determine the internal structure of the fistula. Not all patients in our study group underwent surgery within the review period. Despite these limitations, we believe our results indicate that endoanal MR imaging has sufficient value in the evaluation of primary anovaginal fistulas and additional abnormalities.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, all authors; study concepts and design, S.D., S.M.H., G.P.K.; literature research, S.D., S.M.H.; clinical studies, S.D., S.M.H.; data acquisition, S.D., S.M.H.; data analysis/interpretation, all authors; statistical analysis, S.M.H., W.C.J.H.; manuscript preparation, S.D., S.M.H., W.C.J.H.; manuscript definition of intellectual content, all authors; manuscript editing, S.D., S.M.H., W.C.J.H.; manuscript revision/review and final version approval, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
) and the vagina containing some air (curved arrow). (b) Sagittal T2-weighted fast SE MR image (2,443/100) confirms the findings of transverse MR imaging and shows the relationship of the anovaginal fistula (thin straight arrow) to the lower edge of the anal sphincter complex (thick straight arrows) that clinically indicates the position of the anal verge. The vagina (curved arrow), puborectal muscle (arrowhead), and endoanal coil (
