Materials and Methods

Medical records from the Animal Medical Center, New York, and the Matthew J. Ryan Veterinary Hospital of the University of Pennsylvania were reviewed. Records for dogs with obstructive pyonephrosis treated with a ureteral stent between August 2008 and November 2012 by 2 authors (ACB and CWW) were included in the study. Pertinent information including signalment, history, initial clinical signs, previous treatments including antimicrobial treatment, imaging findings, clinicopathologic data (CBC, serum biochemical analyses, urinalysis, and coagulation profile), results of bacteriologic culture of urine and urolith analysis, when available, was recorded. Surgical and anesthetic reports were reviewed for any alterations to the standard procedure or intraoperative complications that occurred. Perioperative (< 14 days), short-term (14 to 30 days after the procedure), and long-term (> 30 days after the procedure) outcome was recorded for each patient.

Patients were classified as having sepsis on initial examination if they met 2 of the following criteria: high temperature (> 39.17°C [102.5°F]), high heart rate (> 120 beats/min), high respiratory rate (> 50 breaths/min), and leukocytosis (> 16,000 leukocytes/mL) in the presence of an infectious source.

All dogs were prescribed an appropriate antimicrobial, chosen on the basis of results of microbial culture and antimicrobial susceptibility testing, for 6 to 8 weeks following the procedure. Dogs with documented nephroliths or ureteroliths were transitioned to an appropriate stone prevention or neutralizing diet unless they remained azotemic, in which case a diet suitable for their stage of renal disease was prescribed. For animals with calcium oxalate uroliths, potassium citrate (75 mg/kg [34 mg/lb], PO, q 12 h) was recommended long term, and urine pH and serum potassium concentrations were monitored. For dogs with suspected struvite stones, antimicrobial treatment and dissolution dietary therapy was continued for 1 month beyond radiographic and ultrasonographic evidence of dissolution of the urinary tract stones.

Follow-up evaluation consisted of CBC and serum biochemical analysis, urinalysis, bacteriologic culture of urine, and serial imaging (urinary tract ultrasonography and abdominal radiography). This was recommended at 1 to 2 weeks and at 1 month after stent placement, then every 3 months for 2 years, and then every 6 months. Owners and referring veterinarians were contacted by phone at the end of the study to obtain survival data and evidence of long-term complications and to assess client satisfaction with treatment and clinical status of pets.

All dogs were treated with antimicrobials IV prior to anesthesia. Antimicrobials were chosen on the basis of results of microbial culture and antimicrobial susceptibility testing; if these were not available, then ampicillin or sulbactam (22 mg/kg [10 mg/lb], IV, q 8 h) and enrofloxacin (15 mg/kg [6.8 mg/lb], IV, q 24 h) were administered. All procedures were performed with animals under general anesthesia. Before the procedure, patients were typically premedicated with methadone (0.1 mg/kg [0.045 mg/lb]) and given IV fluid therapy (5 mL/kg/h [2.27 mL/lb/h]). General anesthesia was induced with propofol to effect (typically 5 to 10 mg/kg [2.27 to 4.5 mg/lb]) and maintained with isoflurane (typically 1.5 to 2%). If a patient became hypotensive during the procedure and did not respond to an IV fluid bolus of 10 mg/kg given over 20 minutes, a constant rate infusion of dopamine (1 μg/kg/min [0.45 μg/lb/min], IV) was initiated and generally continued as needed.

Patients were positioned in dorsal recumbency, and the entire abdomen, perineal region, vulva, or prepuce was clipped of hair and aseptically prepared. After preparing the external genital opening with dilute chlorhexidine solution, a cystoscope was advanced through the urethra into the bladder. In female dogs, this was achieved with an appropriately sized rigid 30° cystoscope.^a–c In male dogs, a flexible ureteroscope was used.^d Once the endoscope was within the bladder, an appropriately sized angle-tipped hydrophilic guide wire^e–g was advanced through the working channel of the endoscope into the affected ureteral orifice at the UVJ (Figure 1). The guide wire was advanced a short distance into the ureteral lumen to maintain access through the ureteral papilla while an open-ended ureteral catheter^h–j was advanced over the guide wire into the ureter. The wire and catheter combination were advanced up the ureter, around any obstructive lesion when possible, to the UPJ. The guide wire was removed, and a urine sample was obtained via the ureteral catheter for urinalysis and aerobic bacterial culture and antimicrobial susceptibility testing. If the guide wire would not easily pass around the obstruction, then retrograde ureteropyelography was performed by infusing a mixture^k (50:50) of iohexol (240 mg/mL) and saline (0.9% NaCl) solution into the ureter through the catheter under fluoroscopic guidance (Figure 2). The location of the obstruction was noted, as was any contrast extravasation which would indicate a ureteral perforation or tear. Following ureteropyelography, the guide wire was readvanced through the catheter and into the renal pelvis. Next, the catheter was advanced over the guide wire into the renal pelvis. The wire was then removed, and debris within the pelvis was serially aspirated and lavaged with a mixture (50:50) of sterile saline solution and dilute contrast material until the pelvis appeared clear of purulent debris. This was monitored under fluoroscopic guidance to avoid overfilling the renal pelvis. After the renal pelvis lavage was complete, the guide wire was reintroduced into the ureteral catheter and curled within the renal pelvis under fluoroscopic guidance. The ureteral catheter was then removed over the guide wire. Markers on the side of the catheter were used to measure the length of the ureter from the UPJ to the UVJ as it was withdrawn. An appropriately sized double pigtail ureteral stent was then advanced over the guide wire, through the cystoscope, and up the ureter to the level of the renal pelvis.^l–p The proximal end of the stent was then advanced over the wire until it coiled within the renal pelvis. Once the stent was in proper position within the renal pelvis, the guide wire was withdrawn from the lumen of the stent and the distal end of the stent was pushed into the urinary bladder through the cystoscope, allowing it to coil. Appropriate placement of the stent was confirmed with fluoroscopy and cystoscopy. Lastly, a stone basket retrieval device was used to collect any small cystoliths present for analysis.

View Full Size

Endoscopic and fluoroscopic images of a 10-year-old spayed female Pug with left-sided obstructive pyonephrosis positioned in dorsal recumbency. A—Endoscopic image of the left UVJ. B—Angle-tipped hydrophilic guide wire (black arrow) being advanced through the left UVJ. C—An open-ended ureteral catheter (white arrow), guided through the cystoscope, being advanced over the guide wire within the left ureter. D—Fluoroscopic image of the open-ended ureteral catheter (white arrow) within the distal portion of the left ureter during retrograde ureteropyelography documenting the tortuosity of the obstructed ureter. Notice the endoscope (black asterisk) within the urinary bladder guiding the catheter into the UVJ. E—Fluoroscopic image of the guide wire (black arrow) and open-ended ureteral catheter (white arrow) as they are advanced into the renal pelvis under fluoroscopic and endoscopic guidance. F—Endoscopic image of the double-pigtail ureteral stent (black arrow) as it is advanced into the ureter through the working channel of the cystoscope. G—The distal pigtail of the ureteral stent (black arrow) demonstrating urine draining from the renal pelvis through the multiple fenestrations of the stent. H—Fluoroscopic image of the double-pigtail ureteral stent (black arrows) in the left ureter and coiled within the renal pelvis and urinary bladder.

Citation: Journal of the American Veterinary Medical Association 246, 2; 10.2460/javma.246.2.216

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Images acquired during treatment of a 14-year-old spayed female Yorkshire Terrier with right-sided obstructive pyonephrosis. The patient is positioned in dorsal recumbency. A—Fluoroscopic image of the right kidney and proximal portion of the ureter during retrograde ureteropyelography performed under endoscopic and fluoroscopic guidance. The white arrows are the open-ended ureteral catheter that was advanced up to the UPJ around an obstructive stone (black arrow). Notice the multiple filling defects within the renal pelvis (yellow arrows), which represent thick purulent material prior to renal pelvic lavage. B—The renal pelvis and ureter after renal pelvic lavage during a pyeloureterogram. Notice the filling defects are gone and the pigtail end (red arrow) of the ureteral stent within the renal pelvis. The black arrow is marking the obstructive lesion. C—Photograph obtained during renal pelvic lavage demonstrating the endoscope entering the vulva of the female dog and the open-ended catheter draining purulent material from the renal pelvis. D—Syringes containing approximately 60 mL of purulent material drained from the renal pelvis of this patient.

Citation: Journal of the American Veterinary Medical Association 246, 2; 10.2460/javma.246.2.216

• Download Figure
• Download figure as PowerPoint slide

In male dogs, a flexible ureteroscope was used to gain access to the bladder.^d The working channel of this scope was too narrow to accommodate passage of a ureteral catheter, which required an alteration to the described procedure. In these cases, a stiffened, angle-tipped, hydrophilic guide wire^q was advanced though the working channel of the scope, up the ureter, around the obstruction, and into the renal pelvis under fluoroscopic guidance. Once the wire was in place, the scope was removed from the bladder over the guide wire. An angled hydrophilic catheter was then advanced over the guide wire under fluoroscopic guidance.^r The remaining steps in the procedure were identical to those described for female dogs, with fluoroscopic guidance alone.

This procedure was altered in 2 dogs. In one dog, a laparotomy was performed to facilitate debridement of a retroperitoneal abscess and allow placement of a Jackson-Pratt drain on closure of the abdomen. The guide wire was advanced through the ruptured ureter in an antegrade direction to exit at the dog's urethra, at which time a ureteral catheter was advanced over the guide wire in a retrograde manner, allowing for manipulation of the guide wire within the ureteral lumen to the level of the renal pelvis. The renal pelvis was rigorously lavaged, and the guide wire was replaced. The catheter was then removed over the guide wire, and the stent was placed as described. Once the stent was in place, the area adjacent the rupture was debrided and the ureter closed with a simple interrupted pattern of 6–0 polydioxanone suture material. In another dog, a percutaneous cystoscopy¹⁹ approach was planned to allow removal of large cystoliths and simultaneous ureteral access via a minimally invasive approach. A rigid cystoscope placed through a 6-mm port^s allowed direct access into the bladder. This is a minimally invasive technique applicable for small male dogs where retrograde cystoscopy is not feasible. The guide wire, ureteral catheter, and stent were advanced through the cystoscope, as described for a female dog.

Results

Signalment—Thirteen dogs with 14 obstructed ureters met the criteria for inclusion in the study. Breeds included Yorkshire Terrier (3), Bichon Frise (3), Lhasa Apso (2), and 1 each of the following: Jack Russell Terrier, Pug, Maltese, Cavalier King Charles Spaniel, and Bernese Mountain Dog. Three dogs were castrated males, and 10 were spayed females. The median age at diagnosis was 9.5 years (range, 2.5 to 14 years). Median body weight was 8.5 kg (18.7 lb; range, 2.2 to 41 kg [4.84 to 90.2 lb]).

History and initial examination findings—Historical diagnoses included recurrent urinary tract infection (n = 1), diabetic ketoacidosis (1), idiopathic epilepsy with increase in seizure frequency prior to diagnosis (1), chronic valvular disease (1), and immune-mediated thrombocytopenia (1). Initial clinical signs included anorexia (n = 11), vomiting (11), lethargy (8), diarrhea (8), urinary incontinence (2), polyuria or polydipsia (2), shaking (2), stranguria (1), and signs of abdominal pain (1). Physical examination findings included dehydration (n = 6), a palpable midabdominal mass (2), and bruising of the skin over the affected kidney (1). High body temperature was recorded in 6 dogs (median, 39.28°C [102.7°F]; range, 37.22° to 40.56°C [99° to 105°F). Ten dogs were tachycardic (median, 125 beats/min; range, 100 to 135 beats/min).

Clinicopathologic data—A CBC was performed in all 13 dogs prior to the procedure. Neutrophil count was high in 9 dogs (median, 13,500 neutrophils/μL; range, 6,188 to 39,700 neutrophils/μL [reference range, 2,500 to 8,500 neutrophils/μL]). Band neutrophil count was high in 9 dogs (median, 346.5 band neutrophils/μL; range, 0 to 2,520 band neutrophils/μL [reference range, 0 to 150 band neutrophils/μL]). Four dogs had evidence of toxic changes within the neutrophils on slide evaluation. Seven dogs had non-regenerative anemia (median Hct, 35.6%; range, 27% to 45% [reference range, 29% to 45%]). Ten dogs had thrombocytopenia (median, 61,000 platelets/dL; range, 1,000 to 243,000 platelets/dL [reference range, 160,000 to 450,000 platelets/dL]).

Serum biochemical analysis was performed in all 13 dogs prior to the procedure. Eight dogs had a high creatinine concentration (median, 2.1 mg/dL; range, 0.6 to 8.4 mg/dL [reference range, 0.6 to 1.6 mg/dL]) and 7 had a high BUN concentration (median, 30 mg/dL; range, 8 to 189 mg/dL [reference range, 12 to 35 mg/dL]). Five dogs were hyperphosphatemic (median, 8.8 mg/dL; range, 5.6 to 20 mg/dL [reference range, 2.5 mg/dL to 6 mg/dL]), and 8 were hypoalbuminemic (median, 2.2 mg/dL; range, 1.6 to 3.3 mg/dL [reference range, 2.5 to 3.9 mg/dL). Serum potassium concentrations were within reference limits in all dogs. Other abnormalities included high total bilirubin concentration in 3 dogs (median, 0.3 mg/dL; range, 0.1 to 2.0 mg/dL [reference range, 0.1 to 0.4 mg/dL]).

Enzyme activity was evaluated in 10 of the 13 dogs. All 10 of these dogs had high alkaline phosphatase activity (median, 258 U/L; range, 172 to 819 U/L [reference range, 5 to 130 U/L]), and 1 had high alanine aminotransferase activity (median, 63 U/L; range, 15 to 193 U/L [reference range, 12 to 101 U/L]).

Urinalysis was performed for 10 of the 13 dogs. Abnormalities on urinalysis included hematuria in 9 dogs, pyuria in 7 dogs, and bacteriuria in 5 dogs. Urine was alkaline (pH > 7.0) in 2, acidic (pH < 7.0) in 6, and neutral (pH, 7.0) in 2.

Bacteriologic culture of urine obtained via cystocentesis was performed for all 13 dogs prior to the procedure, and results were positive in 7 of 13. Five of the 6 dogs with negative results of bacteriologic culture of urine from the bladder were concurrently receiving antimicrobial treatment. Bacterial species isolated included Escherichia coli (4 dogs) and Staphylococcus spp (3). In 1 dog with a peritoneal and retroperitoneal effusion, results of bacteriologic culture of this fluid were positive, whereas results of bacteriologic culture of the urine sample obtained via cystocentesis in this patient were negative. On the basis of initial physical assessment and clinicopathologic data, all 13 dogs were found to meet criteria for diagnosis of sepsis.

Diagnostic imaging—Abdominal radiographs were available for review in 12 of 13 dogs. Eleven of 12 dogs had uroliths radiographically visible within the urinary tract (kidneys [n = 8; 4 bilateral and 4 unilateral], ureters [3; 2 unilateral and 1 bilateral], or bladder [3]). Of the 8 dogs with uroliths radiographically visible within the kidneys, 7 had nephroliths on the side of the obstructed ureter and 5 had nephroliths on the contralateral side. One patient had a ureterolith visible in the nonobstructed contralateral ureter. Additional radiographic findings included unilateral renomegaly (n = 3), unilateral small kidney (2), hepatomegaly (2), and decreased serosal detail (3).

Abdominal ultrasonography was performed in all 13 dogs (Figure 3). Twelve dogs had a unilateral ureteral obstruction, and 1 dog was bilaterally obstructed. The causes of obstruction for the 14 ureters were ureteroliths in 13 and a suspected stricture in 1. Of the 14 ureteral obstructions, 8 occurred on the right and 6 on the left. The location of the obstruction was the proximal ureter in 4 dogs, midureter in 6 dogs, and distal ureter in 4 dogs. The median renal pelvis diameter in transverse was 24 mm (range, 6 to 48 mm). The median obstructed ureteral diameter proximal to the obstruction was 9 mm (range, 6 to 14 mm). Two of 13 dogs had evidence of echogenic debris in the renal pelvis suggestive of purulent material on ultrasonography. Ureteroliths (n = 1) or nephroliths (5) were noted on the contralateral side in 6 dogs. Additional findings on abdominal ultrasonography included hyperechogenicity of the retroperitoneal fat (n = 4), peritoneal effusion (2), mesenteric lymphadenopathy (2), diffuse hepatopathy (2), bilateral adrenomegaly (2), perinephric fluid (1), and retroperitoneal fluid (1). When the results of abdominal ultrasonography and abdominal radiography were combined, 8 of 12 unilaterally obstructed patients were found to have bilateral disease, with stones in the contralateral kidney or ureter.

View Full Size

Ultrasonographic images of the left kidney of the patient in Figure 1 acquired before (A) and after (B) stent placement. A—Sagittal image showing obstructive hydronephrosis of the left kidney with echogenic debris within the dilated renal pelvis. B—Sagittal ultrasonographic image of the left kidney in the same dog 7 days after ureteral stent placement for treatment of obstructive pyonephrosis. Notice the resolution of hydronephrosis and the presence of a ureteral stent within the renal pelvis.

Citation: Journal of the American Veterinary Medical Association 246, 2; 10.2460/javma.246.2.216

• Download Figure
• Download figure as PowerPoint slide

Retrograde ureteropyelography was performed in all dogs at the time of stent placement, confirming ureteral obstruction in all 14 obstructed ureters with hydroureter and hydronephrosis proximal to the obstructive lesions. Extravasation of contrast material indicating rupture of the ureter was observed in 1 dog prior to interventional manipulation or stent placement.

Additional diagnostic testing—Urine samples were obtained from the upper urinary tract in 9 of the 13 dogs. Eight were from the renal pelvis during stent placement, and 1 was from retroperitoneal or abdominal effusion in the patient with a ureteral rupture. Results of bacterial cultures were positive in 7 of 9 samples. Isolates included E coli (n = 5), Staphylococcus sp (1), and Klebsiella sp (1). All 9 dogs were receiving antimicrobial treatment at the time samples were obtained.

One dog had a combined bacteriologic culture of the bladder mucosa and a struvite cystolith obtained during stent placement. Microbial culture and antimicrobial susceptibility testing of this sample identified a Staphylococcus sp with the same antimicrobial susceptibility pattern as the Staphylococcus sp isolated from the urine obtained via cystocentesis prior to stent placement.

Bacteriologic culture of urine obtained from both the urinary bladder and the upper urinary tract of 9 of the 13 dogs. Of these 9 dogs, 1 had positive results of bacteriologic culture of urine from the bladder but negative results of bacteriologic culture of urine from the renal pelvis and ureter, and 4 had positive results of bacteriologic culture of urine from the renal pelvis and ureter but negative results of bacteriologic culture of urine from the bladder. Results of bacteriologic culture of urine from both locations matched in the remaining 4 dogs. Overall, including results of bacteriologic culture of urine obtained from both the upper and lower urinary tract, a bacterial infection was diagnosed in 11 of 13 dogs.

Stone analysis was performed in 5 dogs. Uroliths were determined to be calcium based in 3 dogs (100% calcium oxalate monohydrate [n = 1] or 50% calcium oxalate monohydrate and 50% calcium phosphate [2]) and struvite based (95% struvite and 5% calcium phosphate) in 1 dog. One additional dog had a cystolith of mixed composition of struvite (65%) and urate (35%). Seven additional dogs had evidence of stone disease on imaging, but no stone was retrieved for analysis. Of these 7 dogs, 2 had a previous history of calcium oxalate cystolithiasis.

Treatment—Successful retrograde stent placement with cystoscopic and fluoroscopic guidance was achieved in 11 of 13 dogs. Included in these was the small male dog (2.2 kg) in which a percutaneous keyhole cystoscopy approach, rather than a retrograde urethroscopy approach, was used to gain access to the UVJ.¹⁹ There was 1 endoscopic failure in which retrograde stent placement with cystoscopic and fluoroscopic guidance was not achieved. Following ureteral cannulation, the wire was unable to be negotiated around a completely obstructive stone. The wire and stent were then successfully placed with surgical assistance in an antegrade manner through a nephrostomy puncture. One dog did not have endoscopy attempted because of known septic retroperitonitis. This dog underwent a planned laparotomy including surgically assisted stent placement through a ureteral tear followed by debridement of a retroperitoneal abscess and placement of a Jackson-Pratt drain. Drainage and lavage of the renal pelvis was successful in all dogs.

Intraoperative complications—Two intraoperative complications were noted. In one dog, a guide wire could not be passed around the obstructive urolith. Retrograde ureteropyelography revealed contrast extravasation near the obstruction indicating ureteral wall penetration. When this was discovered, the procedure was converted to a laparotomy and a single ureterotomy was performed over the obstruction to facilitate removal of the obstructive ureterolith. The ureter appeared discolored and thin near the site of the obstruction. To avoid causing further damage to the ureter, guide wire access was achieved through a nephrostomy puncture. The stent was then placed in a retrograde fashion over the guide wire and the ureterotomy incision was closed with 6–0 polydioxanone in a simple interrupted pattern. Ureteropyelography was repeated and confirmed no leakage of contrast from the nephrostomy or ureterotomy sites. Thorough lavage of the retroperitoneal space and abdominal cavity was performed, and a Jackson-Pratt drain was placed prior to abdominal closure. Despite these measures, this dog's azotemia and signs of sepsis continued to progress after surgery. Thirty-six hours after surgery, this patient developed pulmonary edema attributed to acute respiratory distress syndrome and decompensated septic shock. The owners elected euthanasia 48 hours after surgery.

In another dog, an intraoperative complication occurred when a stent was occluded by the outer coating of a guide wire. Placement of the stent required the use of a stiff guide wire because of excessive ureteral tortuosity. There was excessive friction between the outer coating of the guide wire and the stent, and as the wire was withdrawn, the outer coating of the guide wire sheared off within the lumen of the stent. This complication was not recognized at the time of the procedure. In the first 24 hours, the dog's azotemia and signs of sepsis progressed, and the dilation of the renal pelvis did not resolve. Abdominal radiographs obtained 48 hours after the procedure confirmed the presence of material within the lumen of the stent causing an occlusion. Three days after the initial procedure, the dog underwent a second procedure to exchange the stent. This dog had immediate improvement in renal function and complete resolution of renal pelvic dilation within 24 hours. Signs of sepsis, which included peripheral edema, band neutrophilia, and hypoalbuminemia resolved over the following 7 days.

Perioperative outcome—Eleven dogs were reevaluated within 14 days after the procedure. At this time, of 8 patients that had blood work performed 3 remained azotemic (median BUN concentration, 52 mg/dL [range, 15 to 120 mg/dL]; creatinine concentration, 2.1 mg/dL [range, 0.7 to 4.3 mg/dL]) and 5 remained anemic (median PCV, 34% [range, 22% to 47%]). Thrombocytopenia had resolved in all dogs (509,000 platelets/μL; range, 220,000 to 691,000 platelets/μL). Of 9 patients with inflammatory leukograms, 3 continued to have high total WBC count (median, 12,000 WBCs/μL; range, 7,300 to 24,000 WBCs/μL), and 2 had band neutrophils noted on a blood smear.

Short-term outcome—Eight dogs were seen for recheck examination 14 to 30 days after the procedure. Two remained mildly anemic (median PCV, 34.5%; range, 34% to 35%). All abnormalities in WBC count, platelet count, albumin concentration, and total bilirubin concentration had normalized. Five dogs continued to have high alkaline phosphatase activity (median, 355 U/dL; range, 172 to 819 U/dL). Two of the 8 initially azotemic dogs continued to have high creatinine concentration (median, 2.2 mg/dL; range, 1.8 to 4.3 mg/dL). One dog had signs of intermittent dysuria that resolved by 30 days. Ultrasonography of the affected kidney revealed decreased dilation of the renal pelvis in 6 of 6 dogs in which this was reevaluated (median renal pelvis measurement in transverse dimension, 6.5 mm; range, 3 to 10 mm).

Long-term outcome—Long-term follow-up data (> 30 days) was available for all dogs discharged from the hospital after stent placement (n = 12). At the time of last follow-up (median, 480 days; range, 120 to 1,410 days), 5 dogs remained azotemic (median creatinine concentration, 1.5 mg/dL; range, 0.6 to 5.0 mg/dL). Minimal renal pelvic dilation on ultrasonography persisted in 6 of 6 dogs (median transverse measurement, 7 mm; range, 3 to 13 mm). Six of 6 dogs serially evaluated with a urinalysis had persistent active sediments on follow-up urinalysis including hematuria, pyuria, and proteinuria, which was not associated with a urinary tract infection as determined on the basis of negative results of bacteriologic culture of urine obtained from the bladder at the time of the urinalysis.

Recurrent urinary tract infections were reported in 7 of 12 dogs during the follow-up period. Three dogs had multiple infections (either 3 or 4 episodes), including 1 patient that had recurrent staphylococcal infections; 1 dog with separate Streptococcus sp, Proteus sp, and E coli infections; and 1 dog with separate Staphylococcus sp, Enterococcus sp, and Enterobacter sp infections. Two other dogs were determined to be infected with 2 bacterial species simultaneously (Enterococcus sp with Proteus sp and Staphylococcus sp with Pasteurella sp). The 2 remaining dogs each had a single infection with an Enterobacter sp and E coli. All dogs were infection free on posttreatment bacteriologic culture of urine at last follow-up.

Three dogs were reported to have transient dysuria at some time during long-term follow-up, none of which was associated with a urinary tract infection. All episodes of dysuria resolved without treatment within 1 to 2 days. One patient had recurrence of urolithiasis 14 months after stent placement and was successfully treated with a dissolution diet and appropriate antimicrobials. Overall, 6 of 12 dogs that had follow-up for > 30 days were still alive at the time of final data collection. Two dogs were euthanized because of progression of renal disease, both at 16 months after treatment. The 3 other dogs died of unrelated causes at 5 months, 12 months, and 4.5 years after the procedure. Follow-up time ranged from 2 to 1,460 days (median, 480 days) for all dogs.

Complications—Major intraoperative complications occurred in 2 dogs. One of these dogs had ureteral penetration with a guide wire and was unable to have a stent placed endoscopically resulting in the need for surgical conversion. This dog developed signs of decompensated sepsis after surgery, including progressive azotemia and pulmonary edema within 36 hours after the procedure. In the second dog, the stent was occluded by the outer coating of a guide wire that occurred the time of the stent placement. This complication necessitated placement of a second stent to establish drainage of the occluded ureter. Two other dogs were noted to have minor short-term complications. One dog was noted to have a hematoma in the bladder, which resolved without intervention within 30 days, and the other dog developed transient dysuria, not associated with a urinary tract infection, which resolved without intervention within 2 days.

The only major long-term (> 30 days) complication requiring an additional procedure was in a dog that developed encrustation of the stent leading to reocclusion of the ureter 5 months after stent placement. Mild mineralization of the stent was noted 90 days after stent placement, at which time the dog was prescribed an acidifying diet in an attempt to dissolve the mineral adhered to the stent. Two months later, the patient was returned to the hospital for routine recheck examination, and ultrasonography revealed progressive hydronephrosis and hydroureter. Endoscopic stent exchange was performed on an outpatient basis. The mineralized material on the stent was determined to be composed of calcium hydrogen phosphate dihydrate (90%) and calcium phosphate (10%) mineral. The dog was transitioned to a neutralizing diet, and potassium citrate (75 mg/kg, PO, q 12 h) was added to the treatment regimen to prevent further encrustation.

Minor long-term complications included transient dysuria, which resolved spontaneously over 1 to 2 days in 3 of 12 dogs. No concurrent urinary tract infections were documented in these cases. Another dog was noted to have encrustation of the stent on abdominal radiographs 16 months after the procedure, which did not alter the patency of the ureter. The only other minor complication seen over the period of long-term follow-up was tissue proliferation at the UVJ seen on abdominal ultrasonography in 5 dogs, which was not associated with clinical signs in any of these dogs.

Discussion

Results of the present study suggest that ureteral stenting is an effective renal-sparing treatment for obstructive pyonephrosis in dogs. In 11 of 12 dogs in this case series, renal pelvic lavage and ureteral stent placement under endoscopic and fluoroscopic guidance was accomplished when attempted. No dog required removal of an infected or obstructed kidney, and it is suggested that, when possible, ureteronephrectomy should be avoided. A ureteral obstruction secondary to obstructive ureterolithiasis and associated pyonephrosis should be considered in dogs with hydroureter, hydronephrosis, and systemic signs of infection.

Twelve of 13 dogs in this study were of breeds predisposed to the formation of calcium-based uroliths, including Shih Tzu, Bichon Frise, Miniature Schnauzer, Lhasa Apso, Miniature Poodle, and Yorkshire Terrier.^20–22 The propensity for calcium-based urolith formation is known to be a life-long problem in many of these breeds, and recurrence of stones should be expected. The etiology of calcium urolith formation is complex and not fully understood in canine patients. It has been associated with a variety of systemic conditions, including hypercalcemia, hypercalciuria, hyperoxaluria, and hyperlipidemia as well as decreased solubility of calcium-containing compounds that may occur in acidic urine.^21,223 These dogs are at risk for recurrent stone formation and the potential for obstruction to the contralateral kidney. In patients in the present study, 8 of 13 dogs had evidence of nephroliths, ureteroliths, or both on the contralateral side at the time of diagnosis on the basis of results of radiography or ultrasonography. This suggests an increased risk of obstruction in the contralateral kidney in the future in these dogs. In light of this, care should be taken to preserve and maximize the function of both kidneys whenever possible.

In the patients in this study, initial clinical signs were frequently nonspecific and included lethargy, shaking, inappetence, vomiting, diarrhea, and abdominal pain. Only 4 of 13 dogs were initially examined for polyuria, polydipsia, incontinence, stranguria, pollakiuria, gross hematuria, or flank pain, which would have drawn the focus of the initial diagnostic workup to the urinary tract. Similarly, physical examinations often failed to identify signs of disease within the urinary tract. Despite considerable enlargement and renal asymmetry visualized on abdominal ultrasonography, renomegaly was suspected or described on physical examination in only 2 dogs.

Clinicopathologic abnormalities used to define sepsis in the human⁵ and veterinary⁶ medical literature were seen frequently in this population of 13 dogs, and these included fever (n = 6), high heart rate (10), and leukocytosis (9).^5,6 Other abnormalities associated with sepsis such as thrombocytopenia (10), hyperbilirubinemia (3), hypoalbuminemia (8), and high alkaline phosphatase activity (10/10) were also common. All 13 dogs had ≥ 2 of these abnormalities, which suggests that sepsis is common in this patient population. These dogs should be considered at risk for anesthetic complications, coagulopathies, hypotension, and organ dysfunction due to hypoperfusion. Guidelines in human⁵ and veterinary⁶ literature describe rapid identification and removal of the septic nidus as the single most important initial step in treating septic patients. In patients with sepsis secondary to pyonephrosis, endoscopic renal pelvic lavage and ureteral stent placement may offer several advantages over traditional surgery in potentially unstable patients, including avoiding open surgery, maintaining the integrity of a closed urinary tract, establishing drainage of the abscess without the need for urinary system puncture, and avoiding removal of a potentially salvageable kidney.

In this study, thrombocytopenia was one of the most common findings, occurring in 7 of 13 dogs. Thrombocytopenia was found by Snyder et al¹⁶ in 7 of 16 group dogs undergoing surgical intervention for treatment of ureteral obstruction, but that case series was not restricted to patients with infected pyonephrosis, as was this series. Guidelines derived from studies²³ in human patients suggest prophylactic platelet transfusions be administered preoperatively for patients with platelet counts < 50,000 to 100,000 platelets/dL. This is an especially important consideration in septic and uremic patients, given that both of these conditions have been associated with thrombocytopenia and thrombocytopathies.^5,24,25 This common finding further promotes the use of a noninvasive techniques in these patients to avoid any bleeding diathesis. Of the 3 dogs in this study that underwent laparotomy, 1 was moderately thrombocytopenic with 52,000 platelets/dL. Although excessive hemorrhage was not noted in this dog during surgery, this was the only dog that was euthanized because of progressive sepsis. Additionally, this dog had the highest serum creatinine concentration prior to surgery (8.4 mg/dL).

In the present study, azotemia was common (8/13) and occurred in most of the dogs with unilateral ureteral obstruction. Because > 75% loss of renal function needs to occur before a dog becomes azotemic, this finding suggests the contralateral kidney was not functioning normally at the time of initial examination.²⁶ These elevations in the creatinine concentration could have represented prerenal azotemia, sepsis, bilateral pyelonephritis, chronic kidney disease, or a combination of these factors. Because these patients are typically examined on an emergency basis, the etiology of the dysfunction in the contralateral kidney cannot be fully investigated before action must be taken to address the source of sepsis in these cases. This further supports the desirability for a renal-sparing treatment to salvage the obstructed kidney, avoiding ureteronephrectomy in these patients.

In this study, 11 of 13 dogs had positive results of bacteriologic culture of urine from the lower or upper urinary tract, whereas, in a study by Snyder et al,¹⁶ 11 of 16 dogs with a ureteral obstruction had positive results of bacteriologic culture of urine. However, it should be noted that our study focused only on dogs with pyonephrosis and not those with ureteral obstructions in general. In the present study, 7 dogs being treated with antimicrobials had negative results of bacteriologic culture of urine, including samples obtained from either the bladder or renal pelvis. Treatment with antimicrobials prior to sampling may have decreased the number of urinary tract infections documented. The frequency of positive results of bacteriologic culture of urine indicates the need to institute broad-spectrum antimicrobial treatment as early as possible when a ureteral obstruction is documented.

In 5 of the 9 dogs that had bacteriologic culture of urine from multiple locations, results were discordant (positive for one location and negative for another). One dog had bacteria isolated from just the urinary bladder and not the renal pelvis, and 4 dogs had positive results of bacteriologic culture of urine from just the renal pelvis or retroperitoneal fluid and not the urinary bladder. Snyder et al¹⁶ similarly reported that 4 of 13 dogs with bacteriologic culture of urine from multiple locations had discordant results. In that study,¹⁶ dogs with negative results of bacteriologic culture of urine obtained from the bladder had positive results of bacteriologic culture of uroliths or urine obtained from the upper urinary tract. It seems to be a consistent finding that bacteriologic culture of samples from multiple locations within the urinary tract corresponds to a higher likelihood of obtaining a positive result when infection is present.

Struvite uroliths have been associated with alkaline urine produced by infections with urease-producing bacteria, including bacteria of the genera Staphylococcus, Proteus, Klebsiella, Corynebacteria, and Mycoplasma. In this case series, there were 4 infections with urease-producing bacteria (Staphylococcus spp in 3 and a Klebsiella sp in 1). Urine pH in these dogs was alkaline (pH, 7.5 to 8.0) in 2 and neutral (pH, 7.0) in 1; urine pH was not measured in 1 dog. Of the 2 dogs in this study confirmed to have struvite stones, one was determined to be infected with a Staphylococcus sp, but interestingly, the other was determined to be infected with E coli. Stone analysis was not performed for the other dogs with urease-producing infections because no cystoliths were present at the time of the procedure for endoscopic retrieval. Our findings illustrate the relationship between urease-producing organisms, urine pH, and struvite urolithiasis, encouraging careful medical management once ureteral decompression is performed when indicated. Minimally invasive endoscopic placement of stents was successful in 11 of the 12 dogs in which it was attempted. Ultimately, stent placement was achieved in all dogs in this series. This is similar to the placement success rate in humans.^7–10

There were 2 major intraoperative complications necessitating surgical conversion in one case and stent exchange in the other. Only 1 major long-term complication occurred, resulting in reobstruction of the ureter due to stent encrustation. Endoscopic stent exchange was performed on an outpatient basis in this dog.

The main complications reported to be associated with ureteral stents are dysuria, stent migration, ureteral reocclusion, stent encrustation, and a ureteral reaction to a stent.^{7–10,13,15,28} In the present study, chronic dysuria from the stent was not seen in any dog. This has been reported to occur in 50% to 80% of humans and 20% to 40% of cats.^8–10,28 In cats, it has been our clinical experience that most of the signs associated with pollakiuria and stranguria resolve with steroid treatment, but this was not needed in any dog in this study. Stent migration is another reported complication in humans and cats^10,28 but also did not occur in any dog in this study. Ureteral reocclusion occurred in 1 dog and seemed to be associated with stent encrustation, which is not a common phenomenon seen in dogs or cats in our clinical experience but is very common in people.^27,28 The one stent encrustation in this series occurred in the only dog that did not have evidence of uroliths at the time of initial stent placement. When this was initially noted on ultrasonography, an acidifying diet was prescribed in an attempt to dissolve the mineral. Given that the mineralized material is most commonly calcium based in humans, acidification of the urine could have encouraged further mineralization of the stent, ultimately resulting in its occlusion at 5 months.²⁹ Once this stent was exchanged and a more appropriate neutralizing stone diet was instituted, this dog remained free of clinical signs and with no evidence of encrustation (follow-up, 540 days). Another minor complication noted after stent placement include proliferative tissue seen at the UVJ around the stent entry site into the bladder in 5 of 12 dogs; however this had no known clinical impact. Finally, many dogs had persistently active sediments on follow-up urinalysis, but these were not associated with a urinary tract infection as determined on the basis of negative results of bacteriologic culture of urine at the time of urinalysis.

After stent placement, subsequent urinary tract infection was reported in 7 of 12 dogs discharged from the hospital. In all of these dogs, infections resolved with a 6- to 8-week course of appropriate antimicrobial treatment. Three dogs had recurrent infections with the same organism as that found at the time of stent placement. One additional dog had consecutive infections with a different organism than that found at the time of stent placement but the same as that found during the follow-up period. Because of this finding, urinary tract infections in dogs with ureteral stents should be considered complicated and treated with a longer course of antimicrobials (chosen based on appropriate bacteriologic culture of urine and antimicrobial susceptibility testing). We recommend a 6- to 8-week course of treatment with an appropriate antimicrobial, with bacteriologic culture of urine performed 1 week after starting treatment, 1 week before stopping treatment, and 1 week after stopping treatment. The results of this study indicate that the risk of future infections in dogs with indwelling stents should be considered, with appropriate monitoring, but should not preclude the use of this treatment modality. The consideration of biofilm formation needs to be further investigated.

This study had several limitations. The retrospective nature of the study, small sample size, variability in client compliance, and short follow-up period make it difficult to make general claims of success or failure for certain procedures. This series can alert the veterinary community that there is an effective, minimally invasive, renal-sparing treatment option for this disease. Unlike many retrospective studies, all dogs were treated by the same clinicians (ACB and CWW), and all follow-up was also performed by the same individuals. Therefore, most clinical decisions would be expected to be consistent, which we suggest would make outcomes more comparable than for a typical retrospective study based on a search of medical records where patients may be treated by many clinicians. Future prospective multicenter studies are warranted.

The patients described in the present study had pyonephrosis associated with a ureteral obstruction, most commonly due to ureteroliths. It may be difficult to extrapolate findings to dogs with other causes of ureteral obstructions. Additionally, the severity and duration of clinical signs of illness in these dogs were variable. Regardless of this, resolution or considerable reduction in azotemia was noted in all patients followed long term.

Recovery of renal function and the risk of developing chronic kidney disease following ureteral obstructions are impacted by the duration of obstruction.^30,31 In 1 study³⁰ of 5 healthy dogs with an experimentally induced ureteral occlusion, it was found that after 7 days of obstruction the patient's glomerular filtration rate was permanently diminished by 35%, and when the obstruction lasted for 14 days, the glomerular filtration rate was diminished by 54%. The duration of obstruction was not known for any patient in this study, so it is possible that those patients with long-standing obstructions were predisposed to permanent loss of renal function and at higher risk for the development of chronic kidney disease. Two patients in this series developed severe chronic kidney disease and were euthanized 16 months after the procedure, without recurrence of ureteral obstruction. Therefore, decompression of the upper urinary tract should be performed as soon as possible once an obstruction is identified and a ureteronephrectomy should be avoided if possible.

Ureteral stenting for treatment of patients with pyonephrosis may be a valuable therapeutic option, which compares favorably to traditional surgical techniques such as pyelotomy, ureterotomy, or ureteronephrectomy.^1,2,16–18 This minimally invasive procedure offers the ability to maintain the integrity of the urinary tract while preserving the kidney and provide long-term protection from future obstructions due to ureteroliths. The advantages of stenting versus traditional surgical techniques are particularly important in dogs with concurrent risks of sepsis, thrombocytopenia,³⁴ chronic kidney disease, and future urolith formation.^{5,6,20,21,24,30,31} Endoscopic placement of a ureteral stent is technically challenging, and advanced training in endoscopic interventional procedures is recommended prior to attempting to place a ureteral stent. This procedure could be considered when a diagnosis of obstructive pyonephrosis is made.