Introduction

Urethrorectal fistula is an uncommonly reported condition in dogs and is typically congenital in nature.¹ Urethrorectal fistulas may lead to leakage of urine from the rectum, persistent urinary tract infections, dysuria, diarrhea, or tenesmus.^1,2 Diagnosis is often confirmed via a positive contrast urethrocystogram. This report describes a rectovesicular fistula secondary to trauma, associated with a similar clinical presentation to congenital urethrorectal fistulas. Rectovesicular fistulas have not been previously reported in dogs, and this report highlights the unique imaging and surgical techniques used to diagnose and surgically address a rectovesicular fistula. Given the excellent prognosis for dogs who undergo surgical correction of congenital urethrorectal fistulas, the rectovesicular fistulous tract in this case was surgically repaired and treated similarly to previously described urethrorectal fistulas.^1,2 Additionally, surgical correction of the fistulous tract allowed for repositioning of a traumatically retroflexed urinary bladder to a more normal position within the abdomen.

Case Report

A 7 mo old 16.2 kg intact male Australian cattle dog was referred to the Veterinary Medical Teaching Hospital for further evaluation of a suspected urethrorectal fistula. The owners reported persistent leaking of urine from the dog’s rectum at rest and when posturing to urinate, as well as intermittent diarrhea since finding the dog as a stray 2 mo before. The primary veterinarian performed a fecal floatation, and no parasites were detected. Metronidazole (15.4 mg/kg per os [PO] q 12 hr for 7 days) was commenced without resolution of the dog’s clinical signs.

A radiographic positive contrast retrograde urethrocystogram (Figure 1) was performed consisting of serial right lateral radiographs. A 10 French red rubber catheter was inserted retrograde into the penile urethra. During injection of diluted contrast medium (10 mL iohexal diluted with 50 mL lactated Ringer’s solution) through the urinary catheter, clear fluid was seen exiting the rectum, presumed to be contrast medium. As the red rubber catheter was advanced, the tip was visualized exiting the rectum. Review of the images revealed contrast medium filling the penile and membranous urethra, accumulating in a small amorphous viscus (Figure 1, black arrow heads) in the expected region of the prostatic urethra, and ultimately traveling dorsally to enter the rectum (Figure 1, white arrow heads) at the level of the pelvic canal. The urinary bladder was not identified in the expected location in the images provided; however, multiple malunion pelvic fractures were observed including nonunion fractures of the caudal ischia and pubic bone. The dog was then referred to the Texas A&M University Veterinary Medical Teaching Hospital for advanced imaging and treatment.

View Figure

• Download as Powerpoint
• Download Figure

On general physical examination, the urinary bladder was not identified on abdominal palpation. The dog was ambulatory on all limbs, but lameness was appreciated in the left pelvic limb. Pain was only elicited on hip extension of the pelvic limbs bilaterally. No neurologic deficits were appreciated. An abbreviated serum chemistry panel was performed that showed a packed cell volume of 42% (reference range 35%–55%), total solids of 6.0 g/dL (reference range 5.7–7.8 g/dL), metabolic acidosis (pH 7.274; reference range 7.38–7.49; HCO₃⁻ 11.5 mmol/L; reference range 21.2–27.8 mmol/L; pCO₂ 24.5 mm Hg; reference range 29.8–40.8 mm Hg) with hyperchloremia 122 mmol/L (reference range 110.6–115.5 mmol/L), hypercalcemia 1.46 mmol/L (reference range 1.23–1.35 mmol/L), hypermagnesemia 0.57 mmol/L (reference range 0.38–0.52 mmol/L), and a mild azotemia (blood urea nitrogen 37 mg/dL; reference range 7–32 mg/dL; creatinine 1.6 mg/dL; reference range 0.2–2.5 mg/dL).

An abdominal computed tomography (CT) study with a positive contrast urethrocystogram was performed. The study revealed a caudally displaced urinary bladder within the left perineal region, adjacent to the rectum (Figure 2A and B). The prostate gland was identified immediately cranioventral to the urinary bladder trigone (Figure 2A), and there was rightward deviation and compression of the adjacent rectum (Figure 2B and C). The urinary bladder was rotated caudally 180 degrees over the prostate gland and rotated approximately 45 degrees to the right in the retroflexed position (Figure 2A), which altered the course of both ureters within the caudal abdomen. The right ureter was identified coursing ventral to the colon to insert on the ventral aspect of the urinary bladder. The left ureter was followed over the pelvic brim and seen inserting into the left lateral wall of the urinary bladder. Despite the mispositioning of the urinary bladder, both ureters inserted in the presumed normal location of the ureterovesicular junction of the urinary bladder. Both ureters were normal in diameter. In Figure 2B and C, a small (approximately 2 mm) communication was identified between the urinary bladder apex (Figure 2C) and the adjacent left lateral wall of the rectum (Figure 2B), with the contrast medium identified passing through this communication from the urinary bladder apex to fill the adjacent rectum (Figure 2C), confirming a diagnosis of a rectovesicular fistula.

View Figure

• Download as Powerpoint
• Download Figure

Multiple chronic fractures in variable stages of delayed healing were observed. The dog had a malunion of the sacrum with secondary narrowing and deformation of the sacral vertebral canal. Beginning at the distinguishable third sacral segment and continuing caudally, there was left lateral deviation of the vertebrae. Bilateral, nonunion fractures of the caudal ischia and pubic bones with ventral displacement of the pelvic floor were present. There was subluxation of the left coxofemoral joint and a malunion of the left femoral neck, and the left femoral capital physis was closed.

The dog was anesthetized for repair of the rectovesicular fistula and repositioning of the urinary bladder to a more anatomical location. A Foley catheter was placed within the fistula through the urethra to facilitate identification. Beginning in sternal recumbency, a perineal incision was made approximately 2 cm to the left of the anus. The soft tissues were divided using blunt and sharp dissection until the urinary bladder wall and distal colon could be identified; marked adhesions were present between the urinary bladder and distal colon that extended to the rectum. In the process of attempting to separate the bladder and rectum to reach the fistula, a small hole was inadvertently created in the rectal wall just caudal to the fistula; at this time, the Foley catheter could be seen exiting the rectum (Figure 3). The fistulous tract was found to be very short, with the urinary bladder and rectum essentially flush with one another. The two organs were separated, and the defects in the urinary bladder and rectum were closed separately with 3-0 polydioxanone sutures in a simple interrupted pattern for both organs. The iatrogenic defect in the rectal wall was closed separately with 3-0 polydioxanone sutures in a simple interrupted pattern. The perineal incision was closed routinely. The dog was repositioned into dorsal recumbency for abdominal exploratory. The urinary bladder was identified in a very caudal location with marked adhesions within the pelvic canal. Adhesions were broken down as much as possible to reposition the urinary bladder in a more cranial position, although it was not possible to move it to a completely normal location. A cystopexy to the left caudal abdominal wall was performed using 3-0 polydioxanone sutures. Increasing dilation of the ureters was noted intraoperatively throughout the course of surgery; the urinary bladder was easily accessible and did not appear grossly torsed. Cystoscopy^a (2.8 mm) was performed to further assess the ureteral and urinary bladder anatomy; the ureters appeared to insert in the urinary bladder in a normal in position. The cause for the progressive ureteral dilation was not identified on cystoscopy. The abdomen was closed routinely, with plans to monitor the ureters after the operation through serial focused ultrasound exams.

View Figure

• Download as Powerpoint
• Download Figure

Postoperatively, the dog received a maintenance rate of lactated Ringer’s solution (2.5 mL/kg/hr), and a urinary catheter was placed to measure urinary output. A continuous rate infusion of fentanyl (3 μg/kg/hr) was administered for 24 hr after the operation, then transitioned to carprofen (2.2 mg/kg PO q 12 hr). Cefoxitin sodium (28 mg/kg IV q 8 hr) was administered until discharge. One day after the operation, abdominal ultrasound showed moderate bilateral dilation of the ureters (left: 5.6mm, right: 5.0 mm in diameter) and renal pelves (left: 11.1 mm, right: 7.6mm in diameter). An abbreviated serum chemistry panel was also performed one day after the operation that showed no evidence of azotemia (blood urea nitrogen 20 mg/dL; reference range 7–32 mg/dL; creatinine 0.7 mg/dL; reference range 0.2–2.5 mg/dL).

The dog was discharged 2 days after the operation on carprofen (2.2 mg/kg PO q 12 hr for 12 days) and cefpodoxime proxetil (9 mg/kg PO q 24 hr for 10 days). An abdominal ultrasound was repeated 1 wk after the operation, which demonstrated progressive bilateral dilation of the ureters (left: 6.21 mm, right: 9.44 mm in diameter) and renal pelves (left: 10.26 mm, right: 9.2 mm in diameter). At this time, the owners reported the dog was urinating normally but leaking hemorrhagic diarrhea. Rectal examination was unremarkable, and the diarrhea was treated with a probiotic (Proviable Forte, 1 capsule PO q 24 hr) and discontinuation of the carprofen. Three weeks after the operation, abdominal ultrasound revealed improvement in dilation of the ureters (left: 3.96 mm, right: 4.28 mm) and renal pelves (left: 2.97 mm, right: 4.6 mm). The diarrhea had resolved, and the dog was urinating and defecating normally. A urinalysis was performed with no evidence of a urinary tract infection. A urine culture was not performed. Azotemia was not noted on serial bloodwork at any point during follow-up.

Because of the dog’s improved clinical signs, absence of azotemia, and static nature of ureteral dilation, the owners were instructed to return for regular monitoring of the urinary tract with the primary veterinarian. Given the location and chronicity of the pelvic fractures, orthopedic surgical correction was not required. Against medical advice, the owner declined additional treatment of the left femoral neck malunion and coxofemoral subluxation. The owners were contacted 5 mo after the operation and reported the dog was urinating and defecating normally without evidence of incontinence and no diarrhea.

Discussion

An acquired urethrorectal fistula was initially suspected based on the dog’s clinical signs and radiographic findings. In this case, it was difficult to determine the exact location of the fistula based on radiographic positive contrast urethrocystography alone. Retrograde urethrocystography with CT was necessary to appropriately visualize the anatomical location of the fistula between the urinary bladder and rectum, rather than urethra and rectum, in addition to confirming the retroflexed position of the urinary bladder, which allowed the clinicians to appropriately plan for surgical repair.

Traumatic rectovesicular fistula has not been previously described in the dog. Because pelvic trauma, particularly sacral fractures, are a known risk factor for intra-abdominal trauma in dogs, the rectovesicular fistula most certainly formed secondary to trauma.⁴ This is supported by the concurrent retroflexed urinary bladder, severity of adhesions present, and evidence of multiple healed pelvic fractures involving the pelvic floor and sacrum. A study of experimental colovesicular fistulas was performed in 1975 on canine subjects.³ Colovesicular fistulas were created in subjects that resulted in a variety of clinical signs including pneumaturia, dribbling of urine from the rectum, urinary tract infection, pyelonephritis, sepsis, and death. Although this study did not specifically observe traumatic rectovesicular fistulas, the similarity in location of the pathology provides insight into the expected clinical signs and potential outcomes if surgical intervention is not pursued.^1,2 When successful excision of the fistulous tract is accomplished, long-term prognosis can be excellent depending on the patient’s overall health status and concurrent comorbidities.^1,2 Although the dog had no evidence of pyelonephritis or sepsis, it is likely that urinary tract infection was present because of the nature of his fistula.

A specific cause for the acute ureteral dilation after surgery was not identified. It is possible that severe cystitis developed because of the amount of manipulation the urinary bladder sustained throughout the procedure. Inflammation of the urinary bladder may have resulted in a partial obstruction of the ureterovesicular junction contributing to the bilateral ureteral dilation. Torsion of the urinary bladder resulting in ureteral dilation was also considered, but the urinary bladder was carefully examined throughout the surgery, and the static nature of the ureteral dilation makes this less likely. Because of the acute nature of the ureteral dilation, bacterial infection is thought to be an unlikely cause, and, fortunately, the ureteral dilation improved over time.

Conclusion

Rectovesicular fistulas are a possible sequela to pelvic and sacral trauma in dogs. Similar to urethrorectal fistulas, advanced imaging may be required for an accurate diagnosis and to assist with developing an appropriate treatment plan. Definitive surgical treatment can lead to a good prognosis. If left untreated, the patient’s clinical signs of dribbling urine from the rectum would have persisted and predisposed the patient to developing urinary tract infections, pyelonephritis, sepsis, and premature death. This case highlights a unique and previously undescribed presentation of a traumatic rectovesicular fistula in a dog.