Materials and Methods

Selection of cases—Medical records from the Matthew J. Ryan Veterinary Hospital of the University of Pennsylvania and The Animal Medical Center were reviewed for any patient that had a locking-loop PNC placed by 1 of 2 authors (ACB or CWW) from 2004 to 2009. Patients were included if a procedural report with the final outcome after PNC use was available for review. The reason for PNC placement and technique were extracted from the medical record, along with patient information, including signalment, history, and laboratory data (CBC, serum biochemical profile, preoperative urinalysis and microbiological data, and coagulation profiles). Findings of preprocedural imaging (radiography and abdominal ultrasonography) were evaluated. Information on renal pelvic size determined in a transverse dimension via ultrasonography was obtained from the medical records. Postprocedural data, including renal pelvis size, catheter patency, results of microbial cultures of urine samples, PNC indwelling time, other urinary diversion mechanisms applied, and reason for catheter removal, were obtained. All complications cited in the medical record were noted.

Procedure—Each patient was placed under general anesthesia by use of various protocols. After surgery, all patients were administered analgesics for the first 24 hours and then as needed on the basis of clinical signs.

Percutaneous approach—Each patient was placed in lateral recumbency with the affected kidney in the up position. Antimicrobials (cefazolin) were given (22 mg/kg [10 mg/lb], IV, at induction and q 2 h during the procedure) if patients were not currently being treated with antimicrobials. The area over the kidney was clipped of hair and aseptically prepared. A small stab incision was made in the skin for catheter penetration. Catheter placement was performed by use of either the modified Seldinger technique or a 1-stab trocar introduction technique.

Modified Seldinger technique—Ultrasound guidance was used to perform pyelocentesis with an over-the-needle standard IV catheter or renal access needle^c (dog, 18 gauge; cat, 22 gauge). Once the tip of the catheter was in the renal pelvis, the stylette was removed and an extension set and 3-way stopcock were attached to the catheter. Urine was drained and an equal amount of diluted contrast solution^d (50% contrast: 50% sterile saline [0.9% NaCl]) was infused into the renal pelvis to perform a pyelogram. Urine was submitted for microbial culture and antimicrobial susceptibility profiling. Under fluoroscopic guidance,^e an angle-tipped hydrophilic guidewire^f (dog, 0.035-inch-diameter wire; cat, 0.018-inch-diameter wire) was advanced through the catheter and coiled inside the renal pelvis. The catheter was removed over the wire, and the PNC (cat, 5F^g; dog, 6F^h) was advanced through the skin, into the renal parenchyma, and into the renal pelvis. For this process, the hollow cannula inside the PNC remained secure to keep the catheter rigid during renal penetration. Once the tip of the PNC was inside the renal pelvis, the cannula was immobilized as the catheter was advanced over the guidewire to form its loop. Once the loop of the pigtail was completely within the renal pelvis, the loop of the catheter was locked in place by pulling on the locking string at the catheter hub (Figures 1 and 2) and the string was carefully locked and secured. The cannula was removed from the catheter. The catheter was secured to the body wall by means of a purse-string and Chinese finger trap suture pattern and a second butterfly suture along the shaft of the catheter to the body wall. A sterile urine collection system was attached to the catheter for gravity drainage. A secure abdominal bandage was placed.

View Full Size

Photographs of a locking-loop PNC. A—The distal end of the catheter over the hollow metal stylette and a guidewire; notice the multiple fenestrations and locking string. B—After the metal stylette and guidewire are withdrawn, notice the pigtail shape forming. C—The string is being locked, creating a tight curl in the distal end of the catheter. D—Notice the string at the top of the catheter near the hub that is pulled to lock the distal pigtail (panel C) in place.

Citation: Journal of the American Veterinary Medical Association 241, 3; 10.2460/javma.241.3.348

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Fluoroscopic images during percutaneous nephrostomy catheter placement in a cat via the modified Seldinger technique. Please note sterile leaded gloves are worn by the clinician for radiation protection. A—A renal access needle is inside the renal pelvis after renal puncture via ultrasound guidance. The needle is being held with a hemostat during an antegrade ureteropyelogram documenting renal pelvic and proximal ureteral dilation. B—A guidewire is coiled inside the renal pelvis through the renal access needle. C—The PNC is advanced over the guidewire with the stiff hollow stylette in place. D—The catheter is advanced off the stylette to form a curl inside the renal pelvis as the guidewire is removed. E—The lock is formed by pulling the locking-string mechanism tightly, curling the pigtail inside the renal pelvis.

Citation: Journal of the American Veterinary Medical Association 241, 3; 10.2460/javma.241.3.348

• Download Figure
• Download figure as PowerPoint slide

One-stab trocar induction technique—The 1-stab trocar introduction technique was performed with the sharp stylette through the hollow cannula in the PNC. Via ultrasonographic or fluoroscopic guidance, the locking-loop PNC with the hollow trocar and sharp stylette (Figure 1) was directly advanced through the body wall and punctured through the greater curvature of the kidney and into the renal pelvis, being careful to avoid the ureteropelvic junction. The sharp stylette was removed once the catheter was seen inside the renal pelvis, and urine was drained through the hollow cannula. The hollow cannula was injected with contrast material by means of fluoroscopic guidance to ensure proper location. The loop was advanced off the hollow cannula into the pelvis, and the string was slowly pulled to help form the loop inside the renal pelvis and lock the pigtail. Once a full loop was made, the hollow cannula was withdrawn. The catheter was secured to the body wall as described.

Ventral midline laparotomy—Because of the nature of the smaller, more mobile kidneys of feline patients, placement via a ventral midline laparotomy was typically performed (n = 15/18 feline kidneys). Either the modified Seldinger technique or the 1-stab trocar introduction technique was performed via intraoperative fluoroscopy during surgery. Once in place, locked, and secured, the hub of the PNC was passed through the body wall by means of a combination of sharp and blunt dissection to exteriorize the catheter. A nephropexy was performed to secure the kidney to the body wall prior to closure with 4 deep box sutures of 3-0 or 4-0 polydioxanone suture material. The abdomen was closed routinely, and the catheter was secured to the outside of the body wall as described. The catheter was tested with contrast material prior to closure to ensure no leakage was seen.

Postprocedural management—All hospitalized patients had a closed urine collection system attached to the PNC. This was made of a sterile fluid line and a sterile empty fluid bag (dog, 1,000 mL; cat, 100 mL). The system was drained, and urine was quantified every 2 hours in a sterile manner by use of alcohol swabs and a sterile needle and syringe. All catheters and urine collection lines were cleaned with chlorhexidine scrub 6 times daily to prevent bacterial colonization of the catheter near the insertion site, and the abdominal wrap was changed and tube exit site evaluated every 24 hours during the time patients were hospitalized.

PNC removal—The PNCs were removed under fluoroscopic guidance. The proximal portion of the catheter was aseptically scrubbed. The entire catheter was cut 5 to 10 cm from the insertion site, releasing the locking string mechanism. By use of a sterile technique, an appropriately sized guidewire (5F catheter, 0.018-inch-diameter wire; 6F catheter, 0.035-inch-diameter wire) was advanced through the cut end of the catheter, and under fluoroscopic guidance, the loop was seen to uncurl. The PNC was carefully removed over the guidewire, with fluoroscopic guidance to ensure any associated ureteral stent was not entrapped and the curl was straightening without resistance. Afterward, the remaining locking string was identified to be sure it did not remain in the insertion site and an abdominal wrap was applied for 12 to 24 hours.

Follow-up—Microbial culture of urine samples from the nephrostomy bag was typically performed 3 days after PNC placement or prior to tube removal if that was < 3 days. If the PNC remained in > 3 days, microbial culture of urine samples was recommended at 7 days and every 2 weeks until the catheter was removed. Microbial culture of urine samples obtained through the PNC was typically performed at the time of PNC removal and via cystocentesis 2 weeks later. Renal pelvis dimensions were measured via abdominal ultrasonography to obtain a transverse diameter prior to catheter removal in most patients. The serum BUN, creatinine, sodium, and potassium concentrations were monitored daily during hospitalization and weekly if patients were discharged from the hospital with their PNC in place. The outcome of each patient was recorded.

Results

Clinical data—Twenty-two PNCs were placed (16 cats [18 kidneys] and 4 dogs [4 kidneys]; Table 1). Median body weight of dogs and cats combined was 4.5 kg (9.9 lb) and ranged from 2.5 to 33 kg (5.5 to 72.6 lb); median age was 9 years (range, 2 to 14 years). There were 11 domestic shorthair cats (6 castrated males; 5 spayed females), 3 domestic longhair cats (3 spayed females), 1 castrated male Siamese, and 1 spayed female Bengal cat. There was 1 castrated male Shetland Sheepdog, 1 castrated male Lhasa Apso, 1 spayed female French Bulldog, and 1 spayed female Spitz. One cat had a peritoneal dialysis catheter in place because of persistent ureteral obstruction after a ureterotomy had been performed at a referring facility. One cat with ureterolithiasis was uninephric as a result of previously being a renal transplant donor. Reasons for PNC placement included ureterolithiasis (15 feline kidneys), ureteral stricture (3 feline kidneys), malignant ureteral obstruction (2 canine kidneys), and following a PCNL for nephrostomy access-tract closure during nephropexy development (2 canine kidneys).

Eighteen of 20 patients had azotemia prior to PNC placement. Median preprocedural serum creatinine concentration was 7.4 mg/dL (range, 1.1 to 15 mg/dL; cat reference range, 0.8 to 1.9 mg/dL; dog reference range, 0.4 to 1.8 mg/dL), median BUN concentration was 75 mg/dL (range, 31 to > 240 mg/dL; cat reference range, 15 to 34 mg/dL; dog reference range, 7 to 27 mg/dL), and median potassium concentration was 4.8 mEq/L (range, 3.2 to 6.5 mEq/L; cat reference range, 3.4 to 5.6 mEq/L; dog reference range, 3.6 to 5.5 mEq/L). Eight of 20 patients had microorganisms cultured from urine samples obtained via cystocentesis (n = 7) or urethral catheterization (1) prior to PNC placement; each was given appropriate antimicrobials at least 48 hours prior to the procedure. Ten patients had a coagulation profile obtained; all results were within reference ranges.

Imaging—All patients had diagnostic abdominal ultrasonography performed prior to PNC placement. As determined on the basis of transverse renal imaging, the renal pelvic dimension had a median diameter of 13 mm (range, 4.2 to 34 mm). Abdominal radiography confirmed ureterolithiasis in 15 patients (14 cats and 1 dog) and nephrolithiasis in 18 patients (16 cats and 2 dogs). Two cats had bilateral ureteral obstructions. One cat was uninephric with associated ureterolithiasis. Two dogs had a trigonal mass obstructing the ureter at the ureterovesicular junction.

Procedures—In 18 feline kidneys, 3 had PNC placement via a percutaneous approach (modified Seldinger technique) and 15 had PNC placement via a ventral midline laparotomy (modified Seldinger technique [n = 13] and 1-stab trocar introduction technique [2]). Eleven cats had PNCs placed in the left kidney, 3 had PNCs placed in the right kidney, and 2 cats had PNCs placed bilaterally. All 3 cases of percutaneous PNC placement in kidneys of cats were left sided. Of 18 feline kidneys, 17 had a 5F PNC placed and 1 had a 6F PNC placed. In the 4 canine kidneys, PNC placement was performed on the right kidney percutaneously by means of the modified Seldinger technique and a 6F catheter in all cases.

The PNC was placed during the primary treatment of ureteral obstruction in 13 of 22 (59%) kidneys and for azotemia stabilization and renal pelvis decompression prior to definitive treatment of the ureteral obstruction in 9 (41%) kidneys (all of which were feline kidneys). Nephrostomy catheters were placed intraoperatively during the primary treatment if there was concern of postoperative urine leakage to allow for renal pelvis decompression. This resulted in most patients having > 1 drainage mechanism (PNC and primary treatment technique [ureterotomy or ureteral stent placement]).

Placement of a PNC typically occurs during the primary treatment procedure when immediate patency is uncertain or temporary urine leakage is anticipated, regardless of the method of correction (traditional surgery or stent). In this group of patients, a PNC was used to prevent urine leakage after ureterotomy or ureteral stent placement (n = 8), to maintain renal pelvis decompression and prevent urine leakage when a renal pelvis rupture was discovered (likely due to the complete ureteral obstruction; 2), to drain and flush an obstructive pyelonephritis due to TCC (1), and as a standard procedure after a PCNL to prevent urine leakage after a large sheath is placed in the kidney percutaneously for intracorporeal nephrolithotripsy (2).

The median time for PNC placement prior to definitive correction of the ureteral obstruction (n = 9) was 40 minutes (range, 35 to 75 minutes); 2 PNC placements were bilateral and represented the longest procedure times in this group. For those that had the catheter placed intraoperatively during ureteral stent placement, ureterotomy, or PCNL, the procedure time was not possible to discern from the medical records because the exact time of PNC placement was not recorded. Nineteen of 20 patients had a urine sample submitted for microbial culture at the time of the pyelocentesis during PNC placement; only 1 of the 19 patients had microorganisms cultured from urine samples.

Complications—One feline patient had evidence of urine leakage into the subcutaneous space 3 days after PNC placement without associated uroabdomen. This was corrected by catheter exchange and purse-string suture at the subcutaneous exit site. During the catheter exchange, the indwelling double pigtail ureteral stent, which was placed for permanent fixation at the time of PNC placement, was accidentally entrapped by the pigtail loop of the PNC and the stent was withdrawn through the renal parenchyma. This complication required a small cystotomy.

A second patient was discharged from the hospital with the PNC after a ureteral stent was placed for TCC as a result of an associated renal pelvis rupture. Because of owner concerns, hospitalization was not considered an option for this patient. Five days after placement, the catheter was accidentally dislodged in the crate at home. The catheter remained intact (locking loop still curled), and there was no known complication associated with this event (no evidence of abdominal or subcutaneous effusion or bleeding). The ureteral stent remained in place, maintaining urinary tract patency with this patient surviving this episode. Finally, 1 patient required 2 attempts at renal pelvis access to improve the angle for the locking loop to curl. Complications related to the PNC in all 3 patients resolved.

Postoperative clinical data—Improvement in renal function on the basis of creatinine concentrations was documented within 12 to 72 hours after PNC placement in 16 of 17 patients that were alive after surgery (3 of 20 patients not included because of death during anesthetic recovery [1 patient] or death because of CHF [2] within 72 hours). Some degree of postprocedural azotemia remained in 9 of 17 patients that were previously azotemic, with a median creatinine concentration of 2.5 mg/dL (range, 1.1 to 7.2 mg/dL), compared with 7.4 mg/dL prior to PNC placement.

There was evidence of gross hematuria from the PNC kidney in 14 of 19 patients, all of which resolved within 12 to 36 hours (1 of 20 patients not included because of death during anesthetic recovery). No signs of pain at the nephrostomy site were recorded after the first 24 hours for 18 of these 19 patients. One patient that had persistent signs of discomfort at the PNC site was the patient with urine leakage into the subcutaneous space. There was no mention of discharge from the PNC exit site in any case, including the one with urine leakage. Urine output from the PNC ranged from 2.3 to 35 mL/kg/h (1.04 to 15.9 mL/lb/h).

The median renal pelvis diameter after PNC placement was 2 mm (range, 1 to 4 mm) on the basis of ultrasonography. The catheters were left indwelling in patients a median of 7 days (range, 1 to 28 days), depending on the reason and mechanism of placement. For the patients that had a nephropexy performed, the catheter remained in place for a median of 5 days (range, 1 to 21 days). For the patients that had the catheter placed percutaneously, the median indwelling time was 21 days (range, 5 to 28 days) to allow a nephropexy to form. Twelve cats had a transurethral catheter placed at the time of PNC placement to help monitor urine output from the contralateral kidney due to the presence of oliguria or anuria preoperatively.

Five patients (3 dogs and 2 cats) were discharged from the hospital with the PNC. Two cats were discharged to the local veterinarian with the PNC draining to determine whether renal function improved prior to considering definitive treatment (1 stricture and 1 multiple calculi in the kidneys and ureters). In the authors’ opinion, a serum creatinine concentration of < 4 mg/dL was the aim. Two dogs had their catheters capped off and wrapped because the PNCs were only in place following the PCNL for nephropexy formation. One dog had the catheter draining at home; this was the catheter that was accidentally dislodged 5 days after surgery while the dog was running with a patent indwelling ureteral stent. This patient had a renal pelvis rupture from the obstruction, so constant drainage was desired, but hospitalization was declined.

Six of 17 patients had microorganisms cultured from urine samples within 2 weeks after PNC placement (median, 7 days; range, 3 to 14 days). Fifteen of 17 patients had microbial cultures of urine samples obtained from the PNC prior to removal, and 2 of 17 from a cystocentesis after catheter removal. Of the 6 patients with microorganisms cultured from urine samples, 4 had urine samples obtained from the PNC collection system at the time of removal and 2 from a cystocentesis 3 and 10 days after PNC removal. Of the isolates, 3 were Escherichia coli, 2 were Enterococcus spp, and 1 was a combination of E coli and Enterococcus spp. All positive samples had a growth of > 100,000 colony forming units. Five isolates were multidrug resistant. All E coli isolates were susceptible to amikacin and imipenem, and all Enterococcus spp were susceptible to amikacin, chloramphenicol, nitrofurantoin, and vancomycin. One E coli isolate had a broad susceptibility pattern. Twelve cats had a transurethral catheter placed at the time of PNC placement; 5 were of the 6 patients that had microorganisms cultured from urine samples obtained after PNC.

Follow-up—Fifteen of 20 patients were discharged from the hospital. No known cause of death was directly related to the placement or presence of a PNC. One patient died during recovery under anesthesia. This was likely associated with the ruptured kidney found at the time of surgery. Four patients died (n = 2) or were euthanized (2) because of nonresponsive CHF after surgery (3 to 7 days after catheter placement). All 4 of these patients had a severe postobstructive diuresis requiring high-volume fluid replacement (7 to 25 mL/kg/h [3.2 to 11.4 mL/lb/h]), and all of these patients had improvement in their serum creatinine concentrations immediately after PNC placement prior to death (median, 1.5 mg/dL; range, 1.3 to 2.9 mg/dL), compared with before (median, 11 mg/dL; range, 5.1 to 15 mg/dL). A diagnosis of CHF was made on the basis of some combination of the following: thoracic radiography (pleural effusion [n = 3], pulmonary edema [3], and enlarged pulmonary veins [4]) and echocardiographic findings (enlarged right and left atrium [4]), elevated central venous pressures (4), and weight gain (4).

Most patients had improvement in their azotemia. The 1 patient that did not improve in the first few days was discharged without PNC complication; renal function never improved, and euthanasia was performed 42 days after PNC placement.

The definitive procedure included a ureteral stent placement in 16 ureters (with or without a ureterotomy or ureteral reimplantation) and a ureterotomy alone in 1. One patient had a ureterotomy performed, and a PNC was placed at the same time. This patient developed urine leakage and re-obstruction at the ureterotomy site when the PNC was capped off. Antegrade pyelography confirmed obstruction to flow at the previous stone site with leakage of contrast material. This was suspected to be due to a postsurgical stricture, edema, or a stone-induced stricture. A ureteral stent was placed 7 days after the ureterotomy. Two patients that had a PCNL performed did not require further treatment once the nephrostomy site was healed and the nephrolith was successfully removed. All PNCs that were electively removed (n = 15) were removed under fluoroscopic guidance over an appropriately sized guidewire. This was performed without complication in all except the patient that developed urine leakage and required catheter exchange. The remaining catheters were removed as a result of dislodgment (n = 1) or removed at the time of death (6).

Discussion

The present study supports the use of a locking-loop PNC in canine and feline patients and documents that PNCs are associated with few serious complications. When complications occur, they can be detrimental. Timely and effective renal pelvis decompression in patients with ureteral obstructions can ultimately preserve remaining renal function.^10–13,16 If definitive treatment is possible (eg, via ureterotomy, ureteral reimplantation, ureteral stenting, or ureteral bypass)^2,19,b,i and the patient is stable enough to handle a longer anesthesia period, then definitive correction is recommended over the placement of a PNC alone, in the authors’ experience. Unfortunately, in many circumstances, because of patient instability or procedural circumstance, temporary drainage is considered necessary either prior to or during definitive treatment.

The most common reason for PNC placement in the present study was renal decompression secondary to ureterolithiasis in cats. Another potential use of PNCs was to provide scaffolding for the formation of a nephropexy after an intracorporeal PCNL in dogs. Nephrostomy catheters are rarely considered necessary if there is a patent ureter following an appropriate intervention. After ureteral surgery, the ureter can become edematous and primary closure of the ureter can be technically difficult, resulting in either luminal occlusion or urine leakage. If urine leakage or continued obstruction exists, it can be life-threatening but is often temporary, as the postoperative edema and swelling resolves. When this occurs, having a nephrostomy catheter for drainage as the surgical site heals can be beneficial, as was the case in 4 cats of this report. Patency was documented on a ureteropyelogram through the PNC prior to removal. Because of the concern for the inability to provide appropriate drainage, a PNC was used for severe obstructive pyelonephritis. This diagnosis was made on the basis of purulent material and a positive microbial culture of urine samples removed from the renal pelvis during the pyelocentesis. Having a catheter in the renal pelvis for drainage and flushing may be necessary if the infected material is too thick to passively drain through the ureter or ureteral stent. In 1 patient with an obstructive ureteral TCC, there was concurrent severe pyelonephritis and the material was thick and inspissated in nature. A PNC was placed to provide adequate drainage. The PNC was removed once the urine was grossly normal.

In the present study, improvement in renal function on the basis of serum creatinine concentrations was documented in 16 of 17 patients within the first 3 days (12 to 72 hours) after PNC placement. Median serum creatinine concentration was 2.5 mg/dL after PNC placement, compared with 7.4 mg/dL before placement. Patients that received a PNC were typically more unstable (ie, higher serum creatinine and potassium concentrations or severely hypotensive under anesthesia) than most patients that were able to withstand a definitive treatment. With the placement of a PNC before definitive treatment, renal stabilization is possible prior to more prolonged anesthesia, which could be a safer alternative for certain patients. One patient had worsening of serum BUN and creatinine concentrations immediately after PNC placement, which took > 1 month to return to baseline. Although results of postmortem examination were not available, it might be speculated that the decrease in renal function was secondary to the anesthetic event or directly related to PNC placement. The latter was considered unlikely because this procedure is typically considered to cause minimal renal damage.^24–26

Catheter placement was associated with short procedure times in the present study. Unfortunately, documentation of the exact duration of catheter placement was not available in the anesthesia record when the catheters were placed during another definitive procedure, which included most cases. The longest procedure time for PNC placement was 75 minutes in a patient that had bilateral PNCs placed via a ventral midline laparotomy prior to any intervention. The median procedure time was 37.5 minutes for unilateral PNC placement. This would suggest that the placement of PNCs either percutaneously or via a ventral midline laparotomy can be quicker than the procedures associated with more definitive treatment.

The loop of the PNC is approximately 10 to 12 mm for the 5F catheter and 15 mm for the 6F catheter, making placement more difficult in a small renal pelvis. Seven patients had a renal pelvis < 10 mm in diameter, and in these cases, catheters were technically more challenging to place so that the loop was appropriately placed inside the renal pelvis. On the basis of experience, the authors suggest placement of PNCs via a ventral midline laparotomy in cats and patients with a renal pelvis < 10 mm in diameter. Most cats have very mobile kidneys, and safely placing the PNC without manually securing the kidney makes percutaneous placement more difficult. The authors prefer the modified Seldinger technique over the 1-stab trocar introduction technique whenever possible.

Previously reported complications for nephrostomy catheters include hemorrhage, urine leakage, pneumothorax, dislodgement, and infection.^2,8,20,21,27 Although complications are possible with PNC, they are minimized by careful placement, securing techniques, the locking-string mechanism, and careful catheter management. To prevent urine leakage, it is important to ensure all of the fenestrations in the loop are situated within the renal pelvis prior to locking the string. This can be done by testing the catheter location within the renal pelvis via pyelography. Also, in the authors’ opinion, a nephropexy in feline patients ensures more stability because feline kidneys are very mobile, compared with those of dogs or humans. Numerous security sutures around the catheter to the skin can prevent catheter movement, and a secure abdominal wrap may prevent contamination and dislodgement. Other potential complications include creating a pneumothorax during puncture, renal hemorrhage during nephropexy or catheter insertion, and penetrating the catheter wall with securing sutures. Such complications did not occur in the present study but are certainly possible. It is also important when removing a PNC to monitor the loop carefully by means of fluoroscopy. This will allow one to ensure that removal is appropriate and the curl is unlocked prior to removal. If there is a double pigtail ureteral stent present and the pigtail of the stent is within the renal pelvis, then the loops of the PNC and the stent can entrap, as in 1 case described in the present report. Allowing patients to be managed at home with their nephrostomy catheters could increase the risk of an adverse event, such as dislodgement. This did not seem to be an issue in those PNCs that were covered and capped under an abdominal wrap for several weeks after PCNL, although there were few cases that were discharged with their catheters (5 of 20 cases).

One patient had evidence of renal rupture due to a complete ureteral obstruction prior to PNC placement. The PNC seemed to initially decompress the renal pelvis and control the hemorrhage; however, upon anesthetic recovery, this patient seemed to have another acute bleeding episode and, despite aggressive resuscitation, ultimately died. Unfortunately, a postmortem examination was not permitted by the owner; the suspected cause of deterioration was persistent renal hemorrhage. The possibility that the PNC was associated with further hemorrhage cannot be ruled out.

Eight of 20 patients in the present study had evidence of a urinary tract infection prior to PNC placement, and all were treated with appropriate antimicrobials for at least 48 hours prior to the procedure. Six patients had microorganisms cultured from urine samples after PNC placement, only half of which (n = 3) had positive microbial culture results prior to PNC placement. Only 1 of 3 had the same isolate identified before and after PNC placement. One patient with a subclinical Enterococcus infection was unable to be cleared of the infection, living an additional 16 months without signs of a lower urinary tract infection. Infections were found only in patients that had a PNC indwelling for > 5 days, and none were seen in patients that had the catheters removed sooner. Twelve of 19 patients had an indwelling transurethral catheter after surgery; most of the patients with positive microbial culture results of after PNC placement were those with an indwelling transurethral catheter (5 of 6 patients). Which externalized catheter was responsible for the urinary infection is not clear, and one might expect the urethral catheter would be more responsible because it is easily contaminated by fecal matter, which is consistent with the literature.^28–30 The shorter the duration that a transurethral catheter remains indwelling, the lower the risk of urinary tract infections, with a recommendation of < 3 days.^28–30 One might speculate the same to be true for PNCs. The authors’ current recommendation is to avoid transurethral catheterization if possible in any patient; however, if necessary, removal of any catheter at the earliest possible time is suggested, and maintaining clean technique with a closed collection system in all catheter manipulation is important.

After surgery, 1 patient had signs of discomfort that persisted at the PNC site after 24 hours. This was the same patient that had urine leaking into the subcutaneous space. Careful attention should be paid to patients that have overt signs of pain and subcutaneous bruising beyond standard laparotomy-associated discomfort after surgery. Diagnostic imaging with contrast material for urine leakage is highly recommended.

There is no clear evidence of how long a PNC should remain indwelling for a seal to form and prevent urine leakage. On the basis of recommendations for other viscous organs (stomach after percutaneous endoscopic gastrostomy catheter placement and urinary bladder after cystostomy catheter placement), 3 to 6 weeks should be adequate for a fibrous seal to form. The authors recommend the catheters remain indwelling until either the obstruction is relieved (if a concurrent surgical nephropexy was performed) or a catheter-induced nephropexy is formed.

One study³¹ evaluated clinically normal dogs in which a nephrostomy catheter was placed after a ureteral obstruction (via an artificial stone) was created at the ureteropelvic junction. In that study, more dogs with a PNC passed the artificial calculi, compared with dogs without a PNC. This suggests that patients may spontaneously pass calculi after nephrostomy catheter placement, and in theory, this may be another potential benefit of renal pelvic drainage prior to a definitive treatment. Decompression of the renal collection system may allow for ureteral relaxation and stone passage without the need for further intervention.

In the present study, no known cause of death was directly related to the placement or presence of the PNC. Interestingly, 4 patients died of or were euthanized because of presumable CHF that was nonresponsive to aggressive medical management. Unfortunately, this can occur, particularly in cats requiring high rates of fluid administration IV These rates are often necessary to prevent hypovolemia and subsequent hypotension and severe dehydration in cats that have severe postobstructive diuresis and excessive fluid losses. There were not enough patients that had this occur in the present study to make generalizations, but it does seem that the patients that developed CHF were those that received the highest fluid rates (typically on the basis of high urine output), had the highest preoperative creatinine concentrations (median, 11 mg/dL), and had a dramatic and fast decrease in the degree of azotemia within 36 hours. This finding requires further investigation for markers and predictors of CHF, and care should be taken with fluid therapy during postobstructive diuresis. The authors currently allow urine output to remain above fluid inputs throughout hospitalization while maintaining appropriate cardiovascular stability to prevent this phenomenon.

Most of the patients in the present study had a definitive treatment either during or after the PNC was placed. All catheters in surviving animals were able to be removed under fluoroscopic guidance, with 1 catheter removal resulting in ureteral stent dislodgement. After appropriate drainage, each renal pelvis was smaller in diameter, compared with before PNC placement.

The limitations of the present study include the retrospective nature, paucity of data regarding procedure times because of PNC placement during other procedures, and various reasons for PNC placement, making it difficult to compare outcomes. The goal of the present study was to describe the use of locking-loop PNCs and any complications and clinical outcomes associated with their use. In summary, as determined on the basis of results of the present study, the use of locking-loop PNCs appears safe, secure, and effective, provided that appropriate techniques are followed. These catheters need to be carefully placed, monitored, and secured to prevent complications. The procedure can be performed either percutaneously or via a ventral midline laparotomy with fluoroscopy, with or without ultrasonographic guidance. Catheters were typically placed via a ventral midline laparotomy in cats and percutaneously in dogs. Careful monitoring for urinary tract infections is necessary, and the avoidance of concurrent transurethral urinary catheters should be considered.