| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, Academic Medical Center, University of Amsterdam, P.
O. Box 22700, 1100 DE Amsterdam, The Netherlands.
2
Department of Radiology, University Hospital Rotterdam Dijkzigt, Erasmus
Medical Center Rotterdam, Dr. Molewaterplein 40, 3015 GD Rotterdam, The
Netherlands.
3
Department of Surgery, University Hospital Rotterdam Dijkzigt, Erasmus Medical
Center Rotterdam, 3015 GD Rotterdam, The Netherlands.
Received July 12, 2001;
accepted after revision September 6, 2001.
Address correspondence to J. Stoker.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Nineteen consecutive patients (age range, 28-56 years; median age, 39 years) with clinical indications of an anovaginal or rectovaginal fistula were included in our retrospective study. Endoluminal sonography was performed using a 7.5-MHz transducer. Endoluminal MR imaging was performed at 0.5 T for 10 patients and 1.5 T for nine patients; axial T2-weighted gradient-echo, coronal and sagittal T2-weighted turbo spin-echo (0.5 T), or axial and radial T2-weighted turbo spin-echo and axial T2-weighted fatsaturated turbo spin-echo (1.5 T) images were obtained. For a variety of reasons, surgery of the fistula was not attempted in six of these 19 patients. The imaging findings were compared with the findings obtained during surgery in the remaining 13 patients.
RESULTS. In 12 of the 13 patients, the fistula was found during surgery: seven of the fistulas were anovaginal, and five were rectovaginal. Findings of endoluminal sonography were true-positive in 11 patients, true-negative in one, and false-negative in one. Findings of endoluminal MR imaging were true-positive in 11 patients, false-negative in one, and false-positive in one. Positive predictive value for endoluminal sonography and endoluminal MR imaging were 100% and 92%, respectively. Imaging findings for anal sphincter defects were comparable.
CONCLUSION. Endoluminal sonography and endoluminal MR imaging have comparable positive predictive values in revealing the location of anovaginal and rectovaginal fistulas.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Findings based on a patient's clinical history and physical examination are often inconclusive, and imaging is performed to establish the location and course of the fistula track [3]. Several imaging techniques have been introduced that are useful in the workup of patients with anovaginal and rectovaginal fistulas, but reports in the literature concerning these techniques are sparse. Initially, only proctography and vaginography were available. The sensitivity of proctography has been reported to be only 34% [4]. In a relatively large study of 27 patients, the results for vaginography were better than those for proctography, with 19 of the 24 fistulas detected (sensitivity, 79%) [5]. This superiority of vaginography has been attributed to the use of an occluding vaginal balloon [6, 7]. During proctography, the preferential flow will often be toward the proximal colon, whereas during vaginography, the flow is toward the vagina. The disadvantage of balloon vaginography is that the balloon occludes the opening of low fistula tracks [5, 6]. The drawbacks of both barium enemas and vaginography are that the relation of the fistula track to the anal sphincter is difficult to appreciate and the extent of anal sphincter defects cannot be identified. Vaginal fistulas can be identified on CT. However, the low contrast resolution is a limiting factor, and therefore, the use of intraluminal contrast may produce better results. To our knowledge, no series on the use of CT in depicting anovaginal and rectovaginal fistulas have been published.
Attempts to improve accuracy of endoluminal sonography and MR imaging have been made. Endoluminal sonography is an established technique for identifying anal sphincter defects and can also be used for locating vaginal fistulas [8, 9]. MR imaging can be used, but the results depend on the MR imaging coils chosen. The spatial resolution of body coil MR imaging or external surface coils is too limited to visualize the often small-diameter tracks and accompanying anal sphincter defects [10,11,12]. Recent studies have proven the value of high-resolution MR imaging using endoluminal coils for detailed depictions of anal sphincter anatomy and diseases (e.g., sphincter defects or perianal fistulas) [12,13,14]. An evaluation of the role of high-resolution endoluminal MR imaging and a comparison of this modality with endoluminal sonography in locating anovaginal or rectovaginal fistulas have not been performed. The objective of this study was to compare the accuracy of endoluminal sonography with that of endoluminal MR imaging in revealing the location of anovaginal and rectovaginal fistulas confirmed by surgery.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Nine patients had a history of obstetric trauma (seven had experienced anal sphincter rupture, and two had undergone episiotomy). Seven patients had a history of previous anorectal surgery (three for perianal fistulas, two for rectal tumors, one for fecal incontinence, and one for anovaginal fistula after surgery for bartholinitis). At later diagnostic workup, three patients were found to have Crohn's disease.
Endoluminal Sonography
Before sonography, a limited rectal cleaning was performed. Endoluminal
sonography was performed by a radiologist using a sonography scanner
(Brüel and Kjaer, Naerum, Denmark) with a
rotating probe providing a 360° image. A 7.5-MHz transducer with a minimum
beam width of 1.1 mm and a focal length of 3.0 cm was used. A hard plastic
sonolucent cone with a diameter of 18 mm covered the transducer. For acoustic
coupling, the cone was filled with degassed water, and after sonography gel
was applied to the cone, it was covered by a condom, which then also was
coated with sonography gel. With the patient in the left lateral position, the
sonography probe was introduced into the rectum and was rotated so that 12
o'clock was anterior. The probe was then withdrawn until the landmark of the
hyperechoic puborectalis sling was identified. The complete region of the
anovaginal and rectovaginal septum was imaged to locate the fistula. Hard
copies of axial images of the anovaginal and rectovaginal septum were made as
well as those of the puborectal muscle and the internal and external
sphincters at four levels in the anal canal
(Fig. 1).
|
An anovaginal or anorectal fistula was defined as a continuous hypoechoic linear structure with possible hyperechoic reflections (air) between the anus or rectum and the vagina (Fig. 2). Graded compression and decompression were performed when necessary to differentiate between a vessel and a track. An external sphincter defect at the level of the anal sphincter was defined as a disturbance of the normal hyperechoic region of the external sphincter. An internal sphincter defect was defined as a discontinuity of the hypoechoic internal sphincter ring.
|
Endoluminal MR Imaging
Endoluminal MR imaging was performed at 0.5 T (Gyroscan T5-II; Philips
Medical Systems, Best, The Netherlands) in 10 patients and at 1.5 T (Gyroscan
ACS-NT; Philips Medical Systems) in nine patients. No bowel preparation was
necessary, but all patients had already had limited rectal cleaning for the
endoluminal sonography performed earlier the same day.
Before imaging, 1 mL of butylscopalamine bromide (Buscopan, 20 mg/mL; Boehringer Ingelheim, Ingelheim, Germany) was injected intramuscularly to reduce bowel motion. An endoluminal coil (Philips Medical Systems) with a diameter of 17 mm and a length of 8 cm was covered by a coil holder with a diameter of 19 mm and a length of 10 cm. A condom was placed over the coil holder, and lubricant was applied to the condom surface.
An axial three-dimensional proton density—weighted gradient-echo sequence (TR/TE, 30/13; flip angle, 60°; field of view, 112 x 140 mm; imaging matrix, 205 x 256; slice thickness, 2 mm; slice gap, 0; and excitations, 2) and coronal and sagittal T2-weighted turbo spinecho sequences (2800/120; echo-train length, 13; field of view, 90 x 120 mm; imaging matrix, 201 x 256; slice thickness, 4 mm; interslice gap, 0.4 mm; and excitations, 8) were performed at 0.5 T. At 1.5 T, an axial T2-weighted turbo spin-echo sequence (2500/100; echo-train length, 10; field of view, 90 x 120 mm; imaging matrix, 227 x 256; slice thickness, 4 mm; interslice gap, 0.4 mm; and excitations, 4), a fat-saturated T2-weighted turbo spin-echo sequence (2500/100; echo-train length, 10; field of view, 90 x 120 mm; imaging matrix, 227 x 256; slice thickness, 4 mm; interslice gap, 0.4 mm; and excitations, 4), and a radial T2-weighted turbo spin-echo sequence (2500/100; echo-train length, 6; field of view, 90 x 120 mm; imaging matrix, 228 x 256; slice thickness, 3 mm; interslice gap, 0.3 mm; excitations, 4) were performed. The axial sequences were performed perpendicular to the endoluminal coil, and the coronal, sagittal, and radial sequences were performed parallel to the endoluminal coil. The complete anovaginal and rectovaginal septum was evaluated (Fig. 3).
|
An anovaginal or anorectal track was defined as a continuous linear structure connecting the anus or rectum to the vagina. The lumen of the track could be collapsed (hypointense), fluid-filled (hyperintense), airfilled (hypointense), or a combination of two or three of these conditions (Figs. 4 and 5). The anal sphincter was evaluated for the presence and extent of accompanying sphincter defects. An external sphincter defect was defined as a discontinuity of the relatively hypointense external sphincter, disturbed architecture, or altered signal intensity. An internal sphincter defect was defined as a discontinuity of the hyperintense internal sphincter ring.
|
|
Surgery
Thirteen of the 19 patients underwent surgical repair after a median period
of 5 months (range, from 2 weeks to 9 months). An experienced colorectal
surgeon made the decision to perform the repair on the basis of all available
data. Six patients did not undergo surgery: two patients refused (one because
her symptoms decreased), one patient's endoluminal sonography and endoluminal
MR imaging showed that she had no fistula, and three patients were found to
have Crohn's disease and so received antiinflammatory treatment instead of
undergoing surgery. The relatively long interval between imaging and surgery
was related to the limited availability of operating rooms nationwide. No
change of symptoms had occurred, except in the one patient who refused surgery
because of an improvement in her symptoms.
Image Analysis
All sonographic and MR images were evaluated separately by two radiologists
immediately after imaging. They evaluated and scored the images without
knowledge of the findings from the other technique. The quality of the images,
the presence and location of an anovaginal or rectovaginal fistula, and anal
sphincter defects were thoroughly evaluated and recorded. Image quality was
rated as good if no artifacts were present, moderate if some artifacts were
present but the image was not greatly degraded, and poor if the presence of
artifacts was distracting. One surgeon made all treatment decisions and
performed all surgical procedures. The results of the imaging methods were
compared with the detailed description of the surgical findings in the 13
patients who underwent repair. Because the decision to perform fistula surgery
was based mainly on the imaging findings, the positive predictive values of
endoluminal sonography and endoluminal MR imaging were calculated.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Endoluminal Imaging Results
On endoluminal sonography, 16 vaginal fistulas were identified in the 19
patients in our study. In the 13 patients who underwent surgical repair,
endoluminal sonography showed seven anovaginal fistulas and four rectovaginal
fistulas. On endoluminal MR imaging, 17 vaginal fistulas were identified in
our 19 patients. In the 13 patients who underwent surgical repair, endoluminal
MR imaging showed eight anovaginal fistulas and four rectovaginal fistulas
(Table 1). Findings of
endoluminal sonography and endoluminal MR imaging were concordant in nine of
the 13 patients who underwent surgical repair. Endoluminal sonography revealed
seven anal sphincter defects and endoluminal MR imaging delineated eight anal
sphincter defects in the 13 patients treated by surgery
(Table 1).
|
Endoluminal Imaging Versus Surgery
At surgery, seven anovaginal fistulas and five rectovaginal fistulas were
found in 12 of the 13 patients. In one patient, no track was detected.
Findings of endoluminal sonography were true-positive in 11 patients,
true-negative in one patient, and false-negative in one patient. The
false-negative finding was a collapsed track not identified on sonography.
Findings of endoluminal MR imaging were true-positive in 11 patients,
false-negative in one patient, and false-positive in the one patient (the
first in our study) with a very low anovaginal fistula, and so the most
inferior part of the anovaginal septum had not been included in the field of
view. The false-positive finding was a prominent vein in the anovaginal septum
that extended to the intersphincteric space and closely paralleled the anal
canal (Fig. 6). This patient
had a true-negative endoluminal sonographic finding. Positive predictive
values for endoluminal sonography and endoluminal MR imaging were 100% and
92%, respectively. Anal sphincter defects were present in seven patients.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our study, both endoluminal sonography and endoluminal MR imaging proved to have a high positive predictive value for revealing anovaginal and rectovaginal fistulas. The high spatial resolution of both techniques is probably the major contributing factor. A limitation of our study is that accuracy and negative predictive value of the techniques could not be determined because the decision for performing the reference standard surgery (the "gold standard") was based on the imaging results.
Comparison of our results with previously reported experience is hampered by the lack of sufficient data in the literature. To our knowledge, only one study evaluated endoluminal sonography in anovaginal and rectovaginal fistulas [15]. In that retrospective study, 25 individuals with rectovaginal fistulas were studied, and surgery was performed in patients with negative imaging findings. Endoluminal sonography identified only seven (28%) of the 25 tracks, with no false-positive findings. Only fistulas above or at the dentate line (6/15 and 1/7, respectively) were visualized; no track below the dentate line (0/3) was identified. In some studies on endoluminal sonography for perianal fistulas, patients with anovaginal or rectovaginal fistulas were included. However, the results are not sufficiently detailed to evaluate the findings for patients with anovaginal and rectovaginal fistulas separately [16].
The data of our study and the study of Yee et al. [15] indicate that endoluminal sonography has a high positive predictive value, but that sensitivity, accuracy, and negative predictive value of the technique may be low. These studies were performed with a 7.5-MHz transducer. Whether the use of a 10-MHz transducer may increase the sensitivity remains to be determined. To our knowledge, no study has evaluated the results of endoluminal MR imaging or compared endoluminal sonography and endoluminal MR imaging of anovaginal and rectovaginal fistulas.
With both endoluminal sonography and endoluminal MR imaging, scrutiny of the complete anovaginal and rectovaginal septum is crucial. Incomplete visualization of the anovaginal septum on MR imaging led to a false-negative finding in our first patient. As in perianal fistulas, the abundant venous vessels in the septum may lead to false-positive findings with either technique, as we discovered [14]. On endoluminal sonography, graded compression and decompression with the transducer or color-flow Doppler sonography will clarify the vascular origin of the structure. The graded compression and decompression technique was used in our patients. In both techniques, revealing the course of the often-tortuous thin-walled structure typically will lead to the correct diagnosis.
On MR imaging, the choice of the type of sequence and the imaging plane are influenced by the presentation of the track, although comparative data are lacking. The track course is through the vaginal septum, which comprises fat and especially abundant veins. On T2-weighted or fat-suppression sequences, the often-hypointense track can be identified against the background of hyperintense veins (Fig. 5). On T1-weighted sequences, the use of IV contrast medium may facilitate identification of the track against the background of enhancing veins. The track often shows little enhancement because trauma, rather than inflammation, is often the cause of the fistula. The images obtained in the off-axis axial imaging plane are very informative. A sagittal sequence can be used to confirm the course of the track. For a complete evaluation of the anal sphincter, an additional coronal sequence is mandatory. Both endoluminal sonography and endoluminal MR imaging can be performed as endovaginal techniques. Our study was not designed to evaluate the role of endovaginal imaging, but one might expect results in identifying fistulas comparable to the endoanal techniques.
Anovaginal and rectovaginal fistulas can be identified on phased array MR imaging, but to our knowledge, no studies with more than a few patients have been performed [17]. Therefore, the role of phased array MR imaging in revealing anovaginal and rectovaginal fistulas has not been determined. In our own anecdotal experience, anovaginal and rectovaginal fistulas often cannot be detected on phased array coil or body coil MR imaging, largely because these tracks typically have a small diameter. In a comparative study of endoluminal MR imaging and phased array MR imaging in the often less subtle perianal fistulas, both techniques had comparable results overall [18]. Phased array MR imaging was somewhat superior in revealing tracks extending outside the sensitive region of the endoluminal coil, whereas the higher resolution images of endoluminal MR imaging resulted in a better identification of a subtle internal opening in some patients. It can be hypothesized that the anovaginal and rectovaginal fistulas, which are often subtle and located close to the coil, are better visualized on endoluminal MR imaging. A comparative study is, however, needed to evaluate this hypothesis.
Apart from revealing a track, imaging can play a role in visualizing concomitant occult anal sphincter defects in patients with anovaginal fistulas caused by obstetric trauma. This capability is important because the presence of concomitant anal sphincter defects influences the functional outcome of treatment of trauma-induced anovaginal fistulas [15, 19]. The study of Yee et al. [15] showed that despite the low accuracy for fistula visualization, endoluminal sonography was an important tool in the treatment of patients with anovaginal fistulas because the modality revealed all the external sphincter defects (23 in 25 patients). Our data on anal sphincter defect identification are concordant with these findings. Studies on using endoluminal imaging to find anal sphincter defects in patients with fecal incontinence have reported similar findings for both techniques [13, 20, 21].
In conclusion, endoluminal sonography and endoluminal MR imaging have comparable positive predictive values in revealing the location of anovaginal and rectovaginal fistulas and concomitant anal sphincter defects.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  F. Berton, G. Gola, and S. R. Wilson Sonography of Benign Conditions of the Anal Canal: An Update Am. J. Roentgenol., October 1, 2007; 189(4): 765 - 773. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Halligan and J. Stoker Imaging of fistula in ano. Radiology, April 1, 2006; 239(1): 18 - 33. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Engin Endosonographic imaging of anorectal diseases. J. Ultrasound Med., January 1, 2006; 25(1): 57 - 73. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Dwarkasing, S. M. Hussain, W. C. J. Hop, and G. P. Krestin Anovaginal Fistulas: Evaluation with Endoanal MR Imaging Radiology, April 1, 2004; 231(1): 123 - 128. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
