Pediatric Radiology

, Volume 41, Issue 9, pp 1139–1145

MRI of acquired posterior urethral diverticulum following surgery for anorectal malformations

Authors

    • Department of RadiologyCincinnati Children’s Hospital Medical Center
  • Nicholas C. Weaver
    • Department of RadiologyCincinnati Children’s Hospital Medical Center
    • Vanderbilt University
  • Christopher G. Anton
    • Department of RadiologyCincinnati Children’s Hospital Medical Center
  • Taiwo Lawal
    • Department of SurgeryCincinnati Children’s Hospital Medical Center
    • Department of SurgeryUniversity College Hospital
  • Miller C. Hamrick
    • Department of SurgeryCincinnati Children’s Hospital Medical Center
  • Shumyle Alam
    • Department of UrologyCincinnati Children’s Hospital Medical Center
  • Alberto Peña
    • Department of SurgeryCincinnati Children’s Hospital Medical Center
  • Marc A. Levitt
    • Department of SurgeryCincinnati Children’s Hospital Medical Center
Original Article

DOI: 10.1007/s00247-011-2072-9

Cite this article as:
Podberesky, D.J., Weaver, N.C., Anton, C.G. et al. Pediatr Radiol (2011) 41: 1139. doi:10.1007/s00247-011-2072-9

Abstract

Background

Posterior urethral diverticulum (PUD) is one of the most common postoperative complications associated with anorectal malformation (ARM) correction.

Objective

To describe our MRI protocol for evaluating acquired PUD following ARM surgery, and associated imaging findings.

Materials and methods

Two radiologists retrospectively reviewed 61 pelvic MRI examinations performed for postoperative ARM for PUD identification and characteristics. Associated clinical, operative and cystoscopy reports were also reviewed and compared to MRI.

Results

An abnormal retrourethral focus suspicious for PUD was identified at MRI in 13 patients. Ten of these patients underwent subsequent surgery or cystoscopy, and PUD was confirmed in five. All of the confirmed PUD cases appeared as cystic lesions that were at least 1 cm in diameter in two imaging planes. Four of the false-positive cases were punctate retrourethral foci that were visible only on a single MRI plane. One patient had a seminal vesical cyst mimicking a PUD.

Conclusion

Pelvic MRI can be a useful tool in the postoperative assessment of suspected PUD associated with ARM. Radiologists should have a high clinical suspicion for a postoperative PUD when a cystic lesion posterior to the bladder/posterior urethra is encountered on two imaging planes in these patients.

Keywords

Anorectal malformationPosterior urethral diverticulumChildrenMRI

Introduction

Congenital anorectal malformations (ARM) affect approximately one in 5,000 newborns, and are often associated with other congenital anomalies of the genitourinary system [1]. Surgical management has advanced from abdominoperineal and sacroperineal pull-through to the posterior sagittal anorectoplasty (PSARP). A laparoscopic-assisted repair is used in select types. Technical complications from the original corrective surgery occasionally necessitate reoperation, with posterior urethral diverticulum (PUD) seen as one of the most common urological complications [13]. At initial presentation, a rectourethral fistula is seen in more than 80% of boys with ARM; incomplete fistula dissection and resection may result in a diverticulum [3]. From a technical point of view, a PUD may occur if the surgeon incompletely dissects the most distal aspect of the rectum from the posterior urethra. The lower the rectum lies, the longer the common wall between the rectum and urethra, making the surgical dissection via the abdomen more technically challenging. In the vast majority of male cases, the distal rectum lies below the peritoneal reflection. Utilizing a transperitoneal approach (abdominoperineal or laparoscopy) can lead to this complication. For this reason, PUD is a complication more frequently seen in male ARM patients. A PUD can lead to urinary dribbling, difficulty with catheterization, stone formation, recurrent urinary tract infection (UTI) and adenocarcinoma from long-term exposure of colonic mucosa to urine.

Pelvic MRI has been shown to be a valuable and comprehensive imaging modality in the assessment of postoperative ARM [1, 4]. MRI in postoperative ARM patients is utilized to assess the presence of an associated presacral mass, to check the trajectory of the pull-through and its relationship to the sphincteric complex, and to inspect the posterior urethra for the presence of a PUD. While fluoroscopic techniques (voiding cystourethrography, enema and colostogram) have proven capable of evaluating associated anomalies of ARM, including PUD, MRI can better demonstrate the 3-D relationships of a PUD to adjacent anatomy, frequently providing a more accurate and confident radiological diagnosis, and assisting in preoperative planning [5]. In some cases, when the fistula to the PUD is small, the PUD may be inconspicuous by fluoroscopic or sonographical imaging but conspicuous on MRI. Evaluation by MRI allows for improved soft-tissue contrast resolution and multiplanar imaging capabilities without the risk of ionizing radiation, but it comes with a relatively high cost and frequent need for sedation [1].

Materials and methods

A retrospective search of the institution’s radiology report database and the Colorectal Center for Children database was performed for all children who underwent a pelvic MRI for postoperative ARM evaluation between January 2007 and June 2010. Any associated operative and cystoscopy reports on children with PUD were also reviewed. Institutional review board approval was obtained and informed consent was waived for this retrospective study, which was compliant with the Health Insurance Portability and Accountability Act.

MR imaging protocol

The anorectal malformation pelvis MRI protocol at our institution has evolved over the past several years. Currently, the MRI exams are performed on either a 1.5T GE Signa HDXT unit (GE Healthcare, Milwaukee, WI) or a 3T Philips Achieva X unit (Philips, Bothell, WA) using an 8-channel or 32-channel cardiac coil respectively. The majority of exams require sedation and appropriate fasting. Except in the rare case in which no anus is present, a 24 French Foley catheter is inserted into the rectum without inflating the balloon. A small amount of water is instilled into the catheter lumen, which aids in identifying the rectum and its relationship to surrounding structures. The urinary bladder is not catheterized.

After obtaining standard three-plane localizer images, our current protocol includes fast spin-echo T2-weighted sequences in the axial and sagittal planes, both with and without fat saturation. The sequences without fat saturation can be particularly useful in highlighting the low signal intensity muscle and bowel wall against the higher signal intensity fat and mucosa, while those sequences with fat saturation better contrast fluid-filled structures, such as a PUD, from surrounding suppressed fat. Further imaging consists of coronal T1 and fast spin-echo T2-weighted images without fat saturation, as well as oblique coronal to rectum fast spin-echo T2, also without fat saturation. When a PUD is identified, a three-dimensional balanced steady-state free precession (SSFP) or heavily T2-weighted 3-D MRCP type sequence are two options that can be reconstructed into 3-D maximum intensity projection (MIP) images and therefore aid in delineating the complex relationship of a PUD relative to surrounding structures.

Image analysis

MRI examinations were reviewed independently in blinded fashion by two board-certified pediatric radiologists (D.J.P., C.G.A.) for identification of a PUD. Discrepancies between the two radiologists were resolved by consensus. Review of clinical, surgical and endoscopic reports was then performed on patients with a PUD identified by MR.

Results

Our database search yielded 61 pelvic MRI examinations performed during the study period that met the criteria of a postoperative ARM examination. An abnormal retrourethral focus suspicious for PUD was identified in 13 of these examinations. These 13 patients included 11 males and two females, ranging in age from 1 to 20 years (mean 8.2 years), all of whom had undergone previous ARM repair including attempted ligation of a rectourethral fistula. Ten of these patients underwent subsequent surgery or cystoscopy after the MRI, and PUD was confirmed in five of these cases. Three of these five patients were initially treated surgically for their ARM via abdominoperineal pull-through and two were laparoscopy assisted. None were initially treated surgically for their ARM via PSARP. Four of these five patients underwent their original surgery at outside institutions and presented to our institution’s colorectal center for evaluation between 47 and 168 (mean 95) months from surgery. Only one of these five patients presented with urinary signs and symptoms (urinary frequency). Only one of these patients was seen at our institution for follow-up after PUD resection, and with the exception of a single urinary tract infection, that patient was urinary symptom free. Only one of these five confirmed cases of PUD was detected by routine renal/pelvis sonography.

Of the remaining eight unconfirmed cases, three had no endoscopy or surgery performed to date for confirmation, and five failed to demonstrate a PUD on cystoscopy or surgery. Of these five false-positives, one had a seminal vesicle cyst at surgery and the remaining four had no lesion found surgically or endoscopically.

The imaging characteristics of the 13 MRI suspected PUD are presented in Table 1. The typical imaging appearance of PUD encountered on MRI was a variable size, round or oval, T2 hyperintense focus posterior to either the bladder neck or posterior urethra (Fig. 1). We considered a PUD small if it was under 1 cm in size, medium if it was between 1 and 1.9 cm, and large if it was 2 cm or larger. The average size of these 13 foci was 11.6 mm (range 2–35 mm). The PUD was typically located midline or paramidline. Eleven of the 13 lesions demonstrated simple fluid characteristics, while two contained internal complexity. In one patient, an irregularly shaped, T2 hyperintense focus was found, presumably due to incomplete filling and resultant collapse in nonvoiding state, as these cases of PUD were often noted on cystoscopy or surgery to be larger in size than measured on MRI (Fig. 2). In one case, a large, heterogeneous, mixed signal intensity mass was identified posterior to the bladder on MRI, with a corresponding radiograph that demonstrated calcifications and surgical clips in the vicinity of the mass (Fig. 3). This mass was initially interpreted as either a gossypiboma or a teratoma, but ultimately was shown at surgery to represent a PUD filled with calcifications. Four of the five false-positive cases (patient numbers 2, 11, 12 and 13) were punctate (4 mm or less) T2 hyperintense foci, seen in only one imaging plane, and smaller than the fluid-filled structures seen on the true positive cases (smallest true positive PUD measured 11 mm on MRI), which were all seen on more than one plane (Fig. 4). The fifth false-positive case (patient number 4), had a 15 mm cystic lesion on MRI which was seen on multiple imaging planes and interpreted prospectively and retrospectively by both reviewers as a PUD, but on surgery was found to have a seminal vesicle cyst and not a PUD. The average size of the five endoscopy/surgery confirmed PUD was 17.8 mm (range 11–35 mm). Postcontrast imaging was not performed; therefore, the enhancement characteristics of these PUD were not ascertained.
Table 1

Characteristics of 13 postoperative ARM patients with MRI-suspected PUD

Patient number

Age at presentation (months)

Initial ARM repair surgery

Urinary symptoms at presentation

Location of PUD

Shape of PUD

Size of PUD

Seen on multiple imaging planes?

Pathological confirmation

1

123

AP pull-througha

None, but severe hypospadias

Midline

Bilobed

Medium

Yes

Yes. Surgery and cystoscopy

2

87

PSARPb

Dribbling

Paramidline

Round

Punctate

No

No lesion on cystoscopy or surgery

3

168

AP pull-through

None

Paramidline

Oval

Medium

Yes

Yes. Surgery

4

5

PSARP

None

Midline extending to left

Round

Medium

Yes

No. Seminal vesicle cyst on surgery

5

86

Lap asst pull-throughc

Frequency

Midline extending to left

Oval

Medium

Yes

Yes. Cystoscopy

6

27

PSARP

Vesicostomy, urinary tract infection

Paramidline

Round

Medium

Yes

No surgery or cystoscopy to date

7

84

Lap asst pull-through

None

Midline extending to left

Round

Large

Yes

Yes. Surgery

8

50

AP pull-through

None

Midline extending to left

Irregular/multiloculated

Large

Yes

Yes. Surgery

9

66

PSARP

Required intermittent catheterization

Midline

Pedunculated

Medium

Yes

No surgery or cystoscopy to date

10

37

PSARP

None

Paramidline

Oval

Small

Yes

No surgery or cystoscopy to date

11

36

AP pull-through

None

Midline

Round

Punctate

No

No lesion seen on surgery

12

240

PSARP

Required intermittent catheterization

paramidline

Round

Punctate

No

No lesion seen on cystoscopy

13

154

PSARP

None

Midline

Linear

Punctate

No

No lesion seen on surgery

aAbdominoperineal pull-through

bPosterior sagittal anorectoplasty

cLaparoscopically assisted pull-through

https://static-content.springer.com/image/art%3A10.1007%2Fs00247-011-2072-9/MediaObjects/247_2011_2072_Fig1_HTML.gif
Fig. 1

A 6-month-old boy with high ARM initially treated by laparoscopically assisted pull-through with large postoperative PUD formed from a remnant of rectourethral fistula. The PUD was subsequently resected surgically. a, b Sagittal and axial T2-weighted MR images without fat saturation demonstrate a round, cystic lesion posterior to the bladder base with dependent debris (arrow). c MIP 3-D reconstruction from an MRCP sequence shows the PUD (arrow)

https://static-content.springer.com/image/art%3A10.1007%2Fs00247-011-2072-9/MediaObjects/247_2011_2072_Fig2_HTML.gif
Fig. 2

A 4-year-old boy with a history of ARM initially treated by abdominoperineal pull-through with irregular, multiloculated PUD formed from a remnant of the rectobulbar urethral fistula. a Axial T2-weighted MRI of the pelvis without fat saturation shows a multiloculated cystic structure posterior to the bladder (arrow). There is also an undescended testicle in the left inguinal canal. b MIP 3-D reconstruction from a balanced SSFP sequence shows the complex shape of the collapsed PUD

https://static-content.springer.com/image/art%3A10.1007%2Fs00247-011-2072-9/MediaObjects/247_2011_2072_Fig3_HTML.gif
Fig. 3

A 7-year-old boy with a history of ARM initially treated with laparoscopic assisted repair with a large, complex PUD posterior to the bladder base. The mass was originally interpreted as gossypiboma versus teratoma. Surgery confirmed a PUD filled with calcification. a, b Sagittal and axial T2-weighted MR images without fat saturation show a large round mass posterior to the bladder base with complex internal material (arrows) and mass effect on the rectum. c Abdomen radiograph shows a calcified mass in the pelvis with adjacent surgical staples. d, e Intraoperative photographs during a posterior sagittal approach, in prone position, demonstrate the PUD before and after removal of the calcifications

https://static-content.springer.com/image/art%3A10.1007%2Fs00247-011-2072-9/MediaObjects/247_2011_2072_Fig4_HTML.jpg
Fig. 4

A 7-year-old-boy with a history of ARM initially treated by PSARP with a punctate T2 hyperintense focus posterior to the posterior urethra to the right of midline (arrow). The lesion was interpreted as a small PUD, but was only seen on axial plane. Cystoscopy and surgery showed there was no PUD

A direct communication between the PUD and the bladder or urethra could not be identified by MRI in the majority of cases, but all cases demonstrated fistulous communication to the urethra by cystoscopy or at surgery. All cases represented the original rectum on pathological inspection.

Discussion

Acquired PUD in postoperative ARM patients is a well-known complication in the pediatric surgical community and, we suspect from our anecdotal experience, a complication that may be increasing in incidence given the popularity of laparoscopy and its utilization for a rectum that lies below the peritoneal reflection [68]. PUD has been reported to occur as a surgical complication in 18% of ARM patients who undergo posterior sagittal anorectoplasty (PSARP) [9]. The PUD is a result of incomplete dissection of the distal rectum at the point in the operation when the surgeon ligates the rectourethral fistula typically present in ARM patients. A considerable length of rectal stump is therefore left attached to the urethra as a blind-ending pouch. It is interesting to note that, in this investigation, none of the five confirmed PUD occurred in patients who underwent PSARP as their initial ARM reconstruction. We believe that this observation supports the notion that a meticulously performed distal rectal dissection via a posterior sagittal approach for a rectum that lies below the peritoneal reflection can avoid this complication [10]. Symptoms of PUD may arise as the residual diverticulum progressively enlarges secondary to infection, parturition, obstruction or trauma [7]. Failure to treat these PUD leads to a risk of recurrent urinary tract infection, urinary dribbling, hematuria, difficult catheterization, stone formation and malignant transformation [3, 11].

While fluoroscopic studies are considered the gold-standard imaging modality for pre- and postoperative evaluation of ARM patients, the role of MRI in both settings has continued to expand [1, 4, 12, 13]. MRI is particularly well-suited for the evaluation of postoperative PUD in this patient population given its superior soft tissue contrast, multiplanar capabilities and lack of ionizing radiation. Ultrasound, while free of ionizing radiation, is user-dependent, requires an adequate acoustic window and may not detect small PUD. In this investigation, only one of the five MRI detected PUD was conspicuous by ultrasound.

In our study, a PUD was most commonly identified as a round or oval, fluid signal intensity mass of variable size posterior to the bladder neck or posterior urethra. Confirmation in more than one imaging plane is critical, as four of our five false-positive cases were punctate T2 bright foci detected in only one imaging plane, and likely simply represented vasculature. All of the confirmed PUD were greater than 1 cm in size. It is important to keep in mind that the actual size of the PUD is in all likelihood larger than appreciated on MR, because the PUD is incompletely distended and collapsed on itself in the nonvoiding state. In addition, identifying communication to the urethra or bladder may not be possible on MRI, and is not critical for confident interpretation of the presence of a PUD. Maximum intensity projections (MIP) and multiplanar reconstructions (MPR) of a 3-dimensionally acquired sequence (e.g., 3-D balanced SSFP or MRCP) can be extremely helpful in preoperative planning.

Differential considerations for PUD should include cystic lesions of the prostate and periprostatic region, such as Mullerian duct cyst, utricle cyst, seminal vesicle cyst, ejaculatory duct cyst, Cowper’s duct cyst and prostatic retention cyst. Mullerian duct cysts, which do not communicate with the urethra, and utricle cysts, which do communicate with the urethra, both occur in the midline and may be difficult or impossible to differentiate from one another by imaging [14]. Mullerian duct cysts typically are found in adults, while utricle cysts are often found in childhood and can be associated with other genitourinary anomalies, particularly hypospadias [14]. Seminal vesicle cysts may be congenital or acquired, and are located off midline in the location of the seminal vesicles. Congenital seminal vesicle cysts are often associated with ipsilateral genitourinary anomalies, such as renal agenesis [15]. Ejaculatory duct cysts are very rare lesions located just off midline adjacent to the central zone of the prostate, can be congenital or acquired and are due to obstruction of the ejaculatory duct [16]. Cowper’s duct cysts are rare lesions located at the level of the urogenital diaphragm, which can be congenital or acquired secondary to duct obstruction [16]. Prostatic retention cysts are acquired lesions typically seen in adulthood, and are most commonly located in the periphery of the prostate gland [16].

While all of these lesions may be considered in the list of differential diagnostic possibilities in a postoperative ARM patient with urinary symptoms, the presence of a cystic mass posterior to the posterior urethra or bladder neck on MRI in this specific patient population should prompt cystoscopic evaluation for PUD. Fluoroscopic voiding cystourethrography may be helpful in equivocal cases. Ultimately, any cystic lesion in this location in children may require surgical intervention to address urinary symptoms and to avoid potential future problems, such as stones or carcinoma. Localizing these lesions by imaging is helpful for preoperative decision-making and planning.

Conclusion

Pelvic MRI can be a useful tool in the postoperative assessment of posterior urethral diverticulum associated with anorectal malformations. Radiologists should have a high clinical suspicion for a postoperative PUD when a cystic lesion posterior to the bladder neck or posterior urethra is encountered on two MR imaging planes in children, especially if the original ARM repair operation was an abdominoperineal pull-through or laparoscopically assisted pull-through and the child is male. This anatomical area must be carefully inspected in all ARM patients undergoing MRI. When a PUD is suspected by MRI, there should be a low threshold for advocating cystoscopic confirmation and surgical intervention, as these lesions pose a malignancy risk.

Copyright information

© Springer-Verlag 2011