Total cystectomy and subsequent urinary diversion to the prepuce or vagina in dogs with transitional cell carcinoma of the trigone area: A report of 10 cases (2005–2011)

Abstract

The cases of 10 dogs with transitional cell carcinoma of the urinary bladder that underwent total cystectomy were retrospectively reviewed to evaluate the feasibility and outcome of total cystectomy and ureteral transplantation to the prepuce or vagina. Dehiscence of ureterostomy (n = 2), pyelonephritis (n = 3), oliguria (n = 2), azotemia (n = 1), and ureteral obstruction (n = 1) were observed complications. The estimated median survival time was 385 days. This study demonstrates the feasibility of total cystectomy and subsequent urinary diversion to the prepuce or vagina in dogs. Compared to previous ureterocolonic anastomosis, this technique is associated with fewer gastrointestinal and neurologic complications.

Introduction

Transitional cell carcinoma (TCC) is the most common tumor of the urinary bladder of dogs (1–3). This malignant tumor is highly invasive and frequently occurs in the trigone region of the bladder, often causing partial or complete obstruction of urine flow. Most dogs die of tumor progression within 1 y (1).

In human medicine, initial treatment of invasive TCC in the bladder often involves total cystectomy because of the propensity to develop multiple recurrences in the bladder, a feature that is also observed in dogs (1,4). In dogs, however, surgical resection of the tumor is not routinely performed for the treatment of TCC. Possible reasons are that partial cystectomy in dogs with TCC has rarely been successful (3), and debulking of the tumor alone has not significantly prolonged survival (3,5). Furthermore, in a previous study in which total cystectomy followed by ureterocolonic anastomosis was performed in 10 dogs with TCC, the overall prognosis was poor with adverse effects such as severe vomiting and neurologic signs (6). Urinary incontinence is invariably associated with total cystectomy, another important reason why this procedure is infrequently performed in dogs.

Systemic medication has been the mainstay in veterinary medicine, and several chemotherapeutic regimens have been proposed as the first choice of treatment. Single-agent piroxicam is a widely accepted treatment with a mean survival time (MST) of 181 d (7); other non-steroidal anti-inflammatory drugs (NSAIDs), such as deracoxib (MST of 323 d) (8) and firocoxib (MST of 152 d) (9) are under trial. Chemotherapy alone (cisplatin, MST of 338 d) (9) or in combination with NSAIDs (mitoxantrone with piroxicam, MST of 291 d; carboplatin with piroxicam, MST of 161 d; and firocoxib with cisplatin, MST of 179 d) (9–11) results in similar survival periods, with most dogs succumbing to the disease < 1 y after diagnosis. Moreover, because of the tumor’s predisposition for the trigonal region of the bladder, many dogs eventually experience urinary obstruction and require additional therapeutic interventions to maintain urine flow (12). These reports illustrate the dire need for further therapies to prolong survival and maintain urine flow in dogs.

Total cystectomy is of interest because it enables radical resection of the tumor and prevents future urinary tract obstruction. We hypothesized that modification of the transplantation site of the ureters could be performed with less complication than total cystectomy with ureterocolonic anastomosis. The aim of this retrospective study was to evaluate the feasibility and clinical outcome of total cystectomy and urinary diversion to the prepuce or vagina for the treatment of TCC involving the trigone of the bladder.

Materials and methods

Case selection

The medical records of 11 274 dogs that had been referred to Department of Surgery, Veterinary Medical Center of the University of Tokyo (VMCUT), Tokyo, Japan, between April 2005 and December 2011 were reviewed, and 13 dogs that underwent total cystectomy were identified. Animals were included in this study if they had TCC in the trigonal area of the bladder accompanied by dysuria and underwent total cystectomy along with urinary diversion to the prepuce (males) or vagina (females). Ten dogs that met the criteria were investigated. Client approval was obtained for the surgery and postoperative care of the animals. Tumors were localized by 1 or more of double-contrast radiography, ultrasonography, and computed tomography (CT). The diagnosis of TCC was histologically confirmed prior to or following the surgery.

Preoperative evaluation

Medical records were reviewed to obtain: signalment; previous therapy; clinical signs; complete blood cell count and serum biochemistry; and findings from thoracic radiography, abdominal ultrasonography, and CT, which were used to stage the tumor (13). Ultrasound images of the kidney were analyzed, and hydronephrosis was diagnosed if dilation of the pelvis was accompanied by thinning of the renal cortex.

Total cystectomy and ureteral transplantation

Propofol (1% propofol; Maruishi Pharmaceutical, Osaka, Japan), 10 mg/kg body weight (BW), IV, was used for induction of anesthesia and animals were maintained with isoflurane inhalation, sevoflurane inhalation or propofol infusion, 10 mg/kg BW per hour, IV. Cefazolin sodium (Rasenazolin; Nichi-Iko Pharmaceutical, Toyama, Japan), 20 mg/kg BW, IV or SC, was administered at the induction of anesthesia and continued q12h for 3 d after surgery unless otherwise specified. Following induction, the animal was placed in dorsal recumbency. A median laparotomy was performed to expose the bladder and adherent urological vasculature. Tumor invasion into the ureters was evaluated visually and with palpation. The ureters were cut at a macroscopically healthy site near the orifice of the bladder, while keeping the proximal ureters as long as possible. Next, the urethra was isolated from surrounding tissues until the urethrovestibular junction was identified with palpation in females or until it became inaccessible from inside the cavity at the peritoneal area in males. In female dogs, the urethra was ligated by a hemoclip (SLS-Clip; Vitalitec, Plymouth, Massachusetts, USA) close to the external urethral opening and resected, with care taken to not damage the vaginal branch of the internal iliac artery. In male dogs, the prostate gland was resected with the proximal urethra and both the penis and remaining distal urethra were removed with a perineal incision from outside the body. Finally, the whole urinary bladder and associated structures were resected.

Dogs had 1 of 3 procedures (Prepuce, Vagina A, or Vagina B) performed, according to differences in transplantation sites of the ureters and the relationship between the 2 ureters. All procedures were performed by a single surgeon (RN). The type of urinary diversion and the duration of the operation were recorded.

Prepuce group

In the 2 male dogs, the prepuce was chosen as the grafting site (Prepuce group, Figure 1A). After the penis and distal urethra were removed, dogs were castrated and 2 orifices were created at the mucosa of the prepuce by making a small incision with a skin biopsy trepan (2 to 3 mm diameter); the location of these orifices is best determined by considering the length of the remaining ureters and future coverage by the diaper. The abdominal wall was penetrated by hemostatic forceps with stab incisions and the ureters were withdrawn through the abdominal wall and preputial orifices. Ureteromucosal anastomosis was performed with 5-0 or 6-0 absorbable monofilament suture in a simple interrupted pattern. Following the procedure, urine is continuously excreted from the ureters.

Figure 1.

Representative images during and after total cystectomy and ureteral transplantation. A — Isolation of the urinary bladder and urethra in a female dog. B — Isolation of the penis and distal urethra with a peritoneal approach in a male dog. C — After removal of the penis and distal urethra, hemostatic forceps was inserted into the abdominal cavity through the opening at the preputial mucosa made by a small trepan, and abdominal wall. The distal tip of the resected ureter was grasped by the forceps and, as the forceps was withdrawn, it was led to outside of the body through the opening. D — Final appearance of the Prepuce technique. The ureter was sutured to the novel orifice. In this case, 1 ureter was transplanted after nephrectomy of the dysfunctional kidney. E, F — In female dogs, novel opening of the ureter was made at the vagina using a small trepan. In this case, the ureter was anastomosed to the vagina with end-to-side anastomosis. G — Final appearance of the Vagina A technique. In this dog, 1 ureter was anastomosed to another (end-to-side anastomosis) and then connected to the vagina with 1 orifice (end-to-end anastomosis) since the ureter was sufficiently dilated to enable end-to-end anastomosis, instead of end-to-side anastomosis, to the vagina. H — Final appearance of the Vagina B procedure. Both ureters were transplanted into the vagina with 2 separate orifices (end-to-side anastomosis). White arrowhead — urethra; black arrowhead — ureter; black arrow — novel opening at the prepuce; * — vagina.

Vagina A and Vagina B groups

Female dogs received either the Vagina A or Vagina B procedure. In intact female dogs, ovariohysterectomy was performed prior to total cystectomy. Initially, 1 ureter was anastomosed to the second ureter (end-to-side anastomosis) so that the 2 ureters formed a Y with 2 proximal ends and 1 distal end. The distal end of the Y was connected to 1 orifice in the vagina created by making a small incision with a skin biopsy trepan (2 to 3 mm diameter) (Vagina A). This method was modified subsequently so that the distal ends of the free ureters were anastomosed to the vagina independently with 2 separate orifices (Vagina B). In both groups, ureteral suture to the orifice was as described for the Prepuce group. Following the procedure, urine streams through the vagina and is discharged from the vulva. An indwelling ureteral catheter was not placed, as long as there was freedom from obstruction.

Complications and follow-up

Information on complications, perioperative management, hospitalization period, histopathological findings, and follow-up serum biochemistry was obtained from the medical records. Data from follow-up imaging in dogs which had preoperative hydronephrosis were collected. We also compiled information on the long-term outcome of dogs, including adjunctive anti-neoplastic treatment, tumor recurrence, distant metastasis, cause of death, and survival time, by conducting telephone interviews of referral veterinarians.

Statistics

For normally distributed values (determined by the Shapiro-Wilk normality test), mean values and standard deviations were calculated. Median values and/or range were described for others. Kaplan-Meier survival curve was generated and MST was estimated. A P-value < 0.05 was considered significant. All statistical analyses were done using software (R, R Development Core Team, Vienna, Austria and StatView, SAS Institute Japan, Tokyo, Japan).

Results

Dogs and preoperative examination

Breeds of dogs included Shetland sheepdogs (n = 3), beagles (n = 2), and 1 each of the following breeds: Maltese, pug, miniature dachshund, shiba inu, and miniature pinscher. There were 2 intact males, 3 intact females, and 5 spayed females. The mean age of the dogs was 11.5 ± 1.2 y (mean ± SD), and the mean body weight was 7.4 ± 1.9 kg (range: 3.3 to 10 kg). All 10 dogs had dysuria at initial presentation, including stranguria (n = 9), urinary incontinence (n = 1), hematuria (n = 4), and pollakiuria (n = 1). Four dogs had anorexia, which was likely induced by azotemia.

The clinical stage of the tumors at the time of surgery was T2N0M0 in 8 dogs; T3N0M0 in 1 dog (dog 1) with prostate involvement, which was confirmed histologically after the surgery; and T2N0M1 in 1 dog (dog 2) with multiple lung nodules that were seen on radiographs. Because enlargement of the internal iliac lymph nodes was not observed in the diagnostic images of any dog, the tumors were considered as nonmetastatic, and biopsy for tumor staging was not performed. Presence of the tumor in the trigonal area of the urinary bladder was confirmed by ultrasonography in all the dogs.

Results of preoperative renal ultrasonography and serum biochemical analysis were available for all dogs. Four dogs were azotemic, with blood urea nitrogen (BUN) and serum creatinine ranging from 12.9 to 55.7 mmol/L [reference interval (RI): 3.6 to 10 mmol/L] and 176.8 to 574.6 μmol/L (RI: 44.2 to 132.6 μmol/L), respectively (Table 1). Five dogs had hydronephrosis and hydroureters (Table 2).

Table 1.

Serum BUN and creatinine concentrations of patients

Dog number	BUN/creatininea
	Preoperative	At discharge	At follow-up
1	1.9/35.4	7.6/88.4	12.2/70.7 (day 264)
2	26.6/335.9	35.2/212.2	26.8/221 (day 130)
3	12.8/35.4	5.3/26.5	Not available
4	3.4/79.6	8.3/79.6	8.6/79.6 (day 470)
5	6.0/35.4	11.0/44.2	32.2/53.0 (day 672)
6	17.4/97.2	8.3/70.7	13.0/70.7 (day 156)
7	12.9/176.8	4.9/141.4	10.3/123.8 (day 69)
8	14.1/150.3	7.2/106.9b	Not available
9	24.0/194.5	10.8/168.0	7.9/53.0 (day 364)
10	55.7/574.6	20.8/176.8	14.1/79.6 (day 85)

^aBUN — Blood urea nitrogen (mmol/L), creatinine (μmol/L).

^bValues from blood before death from acute pancreatitis on day 3.

Table 2.

Results of imaging of the kidney and ureter

Dog number	Preoperative imaging		Follow-up imaging
	Hydronephrosis	Hydroureter
2	Bilateral	RU: 6.2 mm LU: 8.2 mm	Not available
7	Unilateral (Right)	RU: 7.1 mm	RU: not detectable (day 16)
8	Unilateral (Left)	LU: 20.0 mm	Not available
9	Bilateral	RU: 12.0 mm LU: 6.6 mm	RU, LU: not detectable (day 14)
10	Bilateral	RU: 7.0 mm LU: 4.5 mm	RU, LU: not detectable (day 16)

RU — right ureter. LU — left ureter.

Total cystectomy and urinary diversion

In 1 male dog, tumor involvement of the prostate gland was observed during the surgery. In the other male dog with bilateral hydronephrosis, left nephrectomy accompanied total cystectomy because preoperative ultrasonography determined that the left kidney was dysfunctional owing to the lack of cortical blood flow. Of the 8 female animals, the Vagina A procedure was performed in 2 dogs (Figure 1B), and the Vagina B procedure was performed in 6 dogs (Figure 1C). Mean operation time was 160 ± 29 min (mean ± SD). The procedures, the operation times, and the hospitalization periods are summarized in Table 3.

Table 3.

Procedure, complications, operation time, and hospitalization period

Dog number	Procedure	Perioperative complications	Operation time (min)	Hospitalization time (days)
1	Prepuce (ureterocolonic)a	Dehiscence, azotemiab, acidosisb, vomitingb, ataxiab	150	15
2	Prepuce	None	200	25
3	Vagina A	None	160	7
4	Vagina A (Vagina A)a	Dehiscence, azotemia	190	13
5	Vagina B	None	160	11
6	Vagina B	None	150	7
7	Vagina B	Oliguria	150	4
8	Vagina B	Oliguria, acute pancreatitis	200	3c
9	Vagina B	None	120	4
10	Vagina B	Ureteral obstruction	120	6

^aUreterostomy technique in second operation.

^bComplications associated with second operation.

^cDog 9 died on day 3.

Histopathology

All resected specimens were evaluated histopathologically and confirmed as TCC of the urinary bladder. Muscular layer involvement was observed in all dogs; therefore, tumors were graded as T2 according to the WHO classification, except for dog 1 whose prostate gland was histologically invaded by the tumor and thus was graded as T3. Descriptions of the surgical margins of the urethra were available for 3 female dogs (dogs 3, 4, 9); urethral margins were contaminated by tumor cells in all 3 dogs. Descriptions of the surgical margins of the ureter were available for 3 dogs (dogs 5, 6, 9); ureteral margins were free from tumor cells in all 3 dogs.

Complications

Dehiscence of ureters

Dehiscence of the grafted ureters occurred in 2 dogs. Dog 1 (Prepuce group) had abdominal leakage of the urine 1 d after total cystectomy, thus necessitating another urinary diversion. Because necrosis of the anastomosed ends of both ureters was observed during the operation, and the debrided ureters were no longer accessible to the abdominal wall, they were anastomosed separately to the descending colon, as in previous reports (6,14). We immediately began prophylactic administration of cefazolin sodium (20 mg/kg BW, IV, q12h), kanamycin (Kanamycin dry syrup 20%, Meiji Seika Pharma, Tokyo, Japan), 10 mg/kg BW, PO, q12h, enrofloxacin (Bytril, Bayel Yakuhin, Osaka, Japan), 5 mg/kg BW, SC, q12h and lactulose (Monilac; Chugai Pharmaceutical, Tokyo, Japan), 1 mg/kg BW, PO, q12h, to guard against retrograde infection of the urinary tract and absorption of urine from the large intestine. However, 5 d after ureterocolonic anastomosis, azotemia (BUN, 14.9 mmol/L; creatinine, 159.1 μmol/L), intermittent vomiting, and ataxia were observed, and recycling of urine from the colon was suspected. Activated charcoal (Covalzin, Novartis Animal Health, Tokyo, Japan), 40 mg/kg BW, PO, q12h, and famotidine (Famotidine, Nichi-Iko Pharmaceutical, Toyama, Japan), 1 mg/kg BW, IV, q12h, were added to daily treatments. Fluid therapy (Physio 70 Injection; Otsuka Pharmaceutical, Tokyo, Japan), 3 mL/kg BW per hour, was started as well. The BUN (19.1 mmol/L) and creatinine (185.6 μmol/L) peaked at day 6, and the azotemia gradually resolved. From day 7, the dog was maintained with flomoxef sodium (Flumarin; Shionogi & Co., Osaka, Japan), 72 mg/kg BW/day by infusion, kanamycin (Meiji Seika Pharma), 10 mg/kg BW, PO, q12h, and activated charcoal (Covalzin), 40 mg/kg BW, PO, q12h. On day 8, a venous blood gas analysis [(HCO₃), 17.0 mmol/L; RI: 18.8 to 25.6 mmol/L and base excess, −10 mmol/L; RI: −2.1 ± 2.3 mmol/L] suggested metabolic acidosis, and sodium bicarbonate, 0.05 g/kg BW, PO, q12h, was administered until discharge. At day 25, the dog appeared to have recovered from postoperative complications of ureterocolonic anastomosis based on clinical signs and blood values (BUN, 7.6 mmol/L; creatinine, 88.4 μmol/L) and was discharged from the hospital with a prescription for ofloxacin (Wellmate, Meiji Seika Pharma), 5 mg/kg q12h, PO. Elevation of serum ammonia levels (54.6 μmol/L, RI: 9.4 to 44.0 mmol/L) was detected at day 1 only.

One female dog treated with the Vagina A procedure (dog 4) developed ascites and had no urine in the vagina several hours after the first surgery. Exploratory laparotomy revealed obstruction and leakage at the site of anastomosis of the 2 ureters, which was reconstructed by performing a second Vagina A procedure to rejoin the ureters. Elevation of BUN (19.2 mmol/L) and creatinine (389 μmol/L) occurred 3 d after the second surgery, and the dog was maintained with single administration of IV mannitol (20% Mannitol Inj. “YD”, Yoshindo, Toyama, Japan), 0.25 mg/kg, and fluid transfusion (Physio 70 Injection), 3 mL/kg BW per hour. The azotemia normalized within 24 h, and the dog was discharged on day 12 from the second surgery.

Ureteral obstruction

Bilateral ureteral obstruction was suspected in dog 10 (Vagina B group) 1 day after the surgery because of accumulation of urine in the renal pelvis and absence of urine from the vagina. For 2 days, urine was drained from the left kidney through a nephrostomy tube, which had been placed prior to total cystectomy for temporary relief of hydronephrosis, until a permanent bilateral ureteral stent was placed on day 3 after the surgery (12). Thereafter, urine excretion was observed, and the dog was discharged on day 5.

Oliguria

Two dogs (dogs 7 and 8, Vagina B group) developed acute oliguria and underwent peritoneal dialysis and received a diuretic agent. Because both dogs showed none of ascites, obstruction of the site of the ureterostomy, or accumulation of urine in the renal pelvis on ultrasonographic examination, acute oligouria was attributed to acute renal failure that may have resulted from decreased renal blood flow associated with the anesthesia and surgery.

Dog 7 developed oliguria immediately after surgery and was started on a continuous rate infusion of human atrial natriuretic peptide (hANP for Injection, Daiichi Sankyo, Tokyo, Japan), 0.05 μg/kg BW per hour, and supported by abdominal perfusion with 1 L of a peritoneal dialysis solution (Midpeliq L 135; Terumo Corporation, Tokyo, Japan) until urination resumed. Peritoneal dialysis was performed 3 times. Urination was observed 12 h after initiation of the therapy, and hANP administration was discontinued. The dog was monitored for another 2 d and discharged.

Dog 8 was diagnosed with acute renal failure 10 h after the operation based on lack of urine production. The dog was administered hANP, (0.1 μg/kg BW per hour), and perfused with 200 mL of peritoneal dialysis solution (Midpeliq L 135). This dog received the peritoneal infusion 3 times until urination resumed 8 h after initiation of hANP therapy. Thereafter, urine production was maintained by hANP, 0.03 μg/kg BW per hour; however, the dog died 3 days after the operation with concomitant acute pancreatitis, which was suspected by elevation of serum pancreatic lipase immunoreactivity (Spec cPL Test; IDEXX Laboratories Japan, Tokyo Japan), > 1000 μg/L; reference value: < 200 μg/L.

Pyelonephritis

Pyelonephritis was suspected in dogs 1, 4, and 5. Dog 1 (Prepuce group) was presented to the referral veterinarian on day 170 from the total cystectomy with leukocytosis (42 000 cells/μL, 85% neutrophils; RI: 6000 to 17 000 cells/μL), elevated C-reactive protein (CRP) concentration (1619 nmol/L; reference value: < 95.2 nmol/L), and dilation of the renal pelvis. Signs of infection or other inflammatory disease in other parts of the body were not detected by physical examination and ultrasonography. The dog was successfully treated with antibiotics and fluid infusion by the referral veterinarian.

In dog 4 (Vagina A), elevated CRP concentration (> 1905 nmol/L) with dilation of the left ureter and left kidney was suspicious for pyelonephritis at day 120. The dog was initially maintained with antibiotics and fluids by the referral veterinarian. However, due to progression of hydronephrosis and dysfunction of the left kidney, which were diagnosed by ultrasound Doppler method and contrast enhanced CT scanning, a left nephrectomy was ultimately performed at the VMCUT. Urine in the dilated left pelvis was infected with bacteria, and no obvious obstruction of the urinary tract was observed during surgery. Histological diagnosis was chronic hyperplastic ureteritis and sclerotic atrophy of the kidney.

Dog 5 (Vagina B) had a dilated left renal pelvis (5.8 mm), leukocytosis (20 400 cells/μL), and elevated CRP concentration (166.7 nmol/L) at day 373 after surgery. Elevated BUN (22.0 mmol/L) and creatinine (150.3 μmol/L) were observed as well. Obstruction of the urinary tract was ruled out by ultrasonographic examination; therefore, pyelonephritis was suspected. Although the dog was treated with antibiotics, and the pyelonephrosis was seemingly resolved in approximately 1 month (WBC, 20 700 cells/μL; CRP, 9.5 nmol/L; BUN, 28.2 mmol/L; and creatinine, 88.4 μmol/L), the dog developed chronic renal failure and required life-long supportive care.

Follow-up for renal function

Serum BUN and creatinine levels had decreased by the time of discharge in all 4 dogs that had azotemia at the time of surgery (dogs 2, 8, 9, 10), although they were not all within the normal reference range (Table 1). In follow-up ultrasonography, 3 of the 5 dogs with hydronephrosis and hydroureters (unilateral, dog 7; bilateral, dogs 9 and 10) showed improvement within 1 mo after the surgery. The left kidney of dog 10 was an exception; based on the lack of blood flow accompanied by thinning of the cortex, it was diagnosed as completely dysfunctional. Dog 8 died of acute pancreatitis 3 d after the surgery. Dog 2 was lost to follow-up at day 131, and follow-up imaging was not available (Table 2).

Prognosis

Two dogs (6 and 7) were treated preoperatively with piroxicam (Baxo; Taisho Toyama Pharmaceutical, Tokyo, Japan) (0.3 mg/kg BW, q24h orally for 14 d and 42 d, respectively), and 4 dogs received postoperative adjunctive therapy. The treatment regimen varied depending on the condition of each dog (Table 4).

Table 4.

Adjunctive therapy and prognosis of the patients

Dog number	Postoperativeadjunctive therapy	Tumor progression	Outcome
1	Meloxicam (0.1 mg/kg BW per day, PO, 14 d)	Metastasis (day 425)	Died (TCC, day 433)
3	None	Recurrence (day 150)	Euthanized (TCC, day 176)
4	None	Recurrence (day 365)	Died (TCC, day 508)
5	Mitoxantrone (5.0 mg/m2, IV, twice)	None	Died (kidney failure, day 718)
6	None	None	Died (heart failure, day 467)
7	Piroxicama (0.3 mg/kg BW per day, PO, 14 d), Cisplatina (50 mg/m2, IV, once)	Recurrence (day 147)	Died (TCC, day 216)
9	Metronomic chemotherapy (370 d), Radiationa (8 Gy, 4 times)	Metastasis (day 291)	Died (TCC, day 385)
10	None	—	Euthanized (MGT, day 369)

^aTreatment after disease progression.

^bPiroxicam (0.3 mg/kg BW per day, PO) and cyclophosphamide (27.8 mg/m², PO, once every 2 d); MGT — mammary gland tumor.

Tumor recurrence was diagnosed in dogs 3, 4, and 7, based on findings of urinary obstruction, hematuria, and a skin lesion at the inguinal area, respectively. Recurrence in the latter 2 dogs was confirmed cytologically. In dog 3, because a positive urethral margin was reported in the pathological findings and nearly 5 months had passed since the neoureterostomy, tumor recurrence was our primary differential diagnosis for the cause of urethral obstruction; however, further diagnostics were declined by the client and other causes of obstruction such as stricture at the neoureterostomy site and calculus formation, could not be ruled out. Distant metastasis resulting in bone lysis observed with routine radiological examination occurred in 2 dogs (1 and 9). These data are summarized in Table 4. Necropsy was not performed on any of the dogs. Estimated median survival time was 385 d.

Discussion

Characteristics of the dogs in this study were similar to those of dogs in previous studies with respect to age, breed, and clinical stage at diagnosis; however, the male:female ratio was 1:4 in our study compared with ratios of 1:1.1 and 1:≈2.1 in other studies (1,8,9,15,16). The difference may be due to the small number of subjects in this study.

For retention of urine storage capability and urinary continence in some techniques, reconstruction of the urinary tract is performed mainly via ureterocutaneostomy with or without ileal conduit or neobladder reconstruction using autologous ileum mucosa transplantation in humans (17). Complication rates of 25% to 57% have been reported and frequently involve digestive system complications, such as ileus and hepatic failure (18–20). One study which analyzed short-term (within 30 d) complications of radical cystectomy in 6577 human patients, reported in-hospital mortality rate and overall complication rate other than death as 2.57% and 28.4%, respectively; the complications were primarily digestive (16.1%), cardiac (4.12%), respiratory (3.8%), urinary (2.92%), and wound-related (4.3%).

In contrast, in dogs, total cystectomy has been performed with ureterocolonic anastomosis (6,14). Only 1 article described complications in a case series (6). That study reported perioperative complications including neurologic disorders such as ataxia, depression, and seizures (4/10 cases); gastrointestinal signs such as vomiting (5/10 cases); hyperchloremic metabolic acidosis (5/10 cases); ureteral stent dislodgement (3/10 cases); elevated creatinine levels (5/7 cases, 1–5 mo after surgery); and pyelonephritis (5 kidneys; 2 resulted from obstruction of the urinary tract by the tumor or retained stent). Five of 10 dogs were euthanized solely due to neurologic or gastrointestinal signs.

In our study, ureterovaginal or ureteropreputial anastomosis was used, and no dog developed neurologic or gastrointestinal signs. Since complications are often attributed to urine stasis in the colon after ureterocolonic anastomosis, our novel procedure may have the advantage of preventing reabsorption of urine and breakdown products from the digestive tract and, therefore, decreasing the occurrence of neurologic and gastrointestinal complications. One dog in our study underwent ureterocolonic anastomosis after failure of the initial ureteropreputial anastomosis and developed neurologic and gastrointestinal signs with concurrent acidosis and hyperammonemia, an additional risk of ureterocolonic anastomosis.

After the total cystectomy, serum creatinine levels at follow-up (69–672 d) were within the normal range in 7/8 cases. One dog with creatinemia at follow-up (dog 2 at day 130, creatinine 221 μmol/L) had bilateral hydronephrosis at the time of initial presentation (creatinine 336.0 μmol/L), so renal function might have been irreversibly disrupted at the time of surgery, which contributed to the persistently elevated creatinine level. One of the problems associated with ureterocolonic anastomosis in a previous study (6) was elevated serum creatinine levels in 5/7 dogs. Creatinemia may develop after urinary diversion as a result of chronic renal failure associated with causes such as retrograde infection and postrenal obstruction of the urinary tract, or the recycling of creatinine absorbed from the colon. The rates of occurrence of pyelonephritis (3/19 kidneys in total) or obstruction of the urinary tract (1/8 follow-up cases) in our study were similar to those (pyelonephritis; 5/19 kidneys in total and obstruction of the urinary tract; 1/7 follow-up cases) in the previous study (6); eliminating the recycling of creatinine from the colon might therefore be a benefit of our procedure. The same discussion may apply to serum BUN levels.

In 3 kidneys with pyelonephritis, there were no recognized complicating factors, such as a history of hydronephrosis or hydroureter, and no evidence of postrenal obstruction. However, it is possible that there was a functional postrenal obstruction, which we did not detect. In dog 4 in particular, unilateral hydronephrosis and concurrent infection without evidence of substantial obstruction was confirmed during the operation. In this case, the Vagina A method was used, and the 2 ureters were connected with an end-to-side anastomosis midway to the vagina. Thus, if 1 kidney was hyperfunctional and produced more urine, there would be an imbalance in the fluid pressure between 2 interfluent ureters. This imbalance could cause a problem with the discharge of urine from 1 side of the ureter and eventually, pelvic accumulation of urine. This observation combined with the previous failure of the Vagina A procedure in dog 4 led to our modification and replacement of this procedure with the Vagina B method. In addition, we did not necessarily place a ureteral stent in this case series, which may facilitate prevention of obstruction of the ureters after ureterostomy especially in small breed dogs. The procedure should be modified for future studies.

In dog 8, in-hospital death was thought to be due to acute renal failure and pancreatitis. Acute renal failure might result from low blood supply to the kidney during surgery and anesthesia. Since there was peritoneal dialysis treatment, pancreatitis was considered to be the main contributor to in-hospital death although the precise cause of death could not be specified. It is possible that the nausea and panting observed at presentation were mistakenly attributed to azotemia instead of acute pancreatitis. In this case, the total cystectomy would have contributed to a secondary response, likely resulting in a fatal situation such as systemic inflammatory response syndrome and disseminated intravascular coagulation. Therefore, complete preoperative evaluation and intensive postoperative monitoring are essential to improve in-hospital morbidity.

Dehiscence of the ureterostomy site was observed in 1 Prepuce group dog and 1 Vagina A dog. These dogs were among the first to undergo total cystectomy early in the introduction of the procedure, so technical or procedural immaturities may have contributed to the dehiscence. Moreover, we did not use an indwelling ureteral catheter in this case series. Additional temporary support of the novel ureteral tract by a catheter would be helpful to prevent dehiscence.

Postoperative azotemia was diagnosed in 2 dogs. One dog became azotemic immediately following dehiscence and urine leakage into the abdomen. This was likely due to urine absorption from the peritoneum, which could be prevented by proper surgery. Azotemia in 1 dog might have occurred as a result of absorption of urine from the colon because the azotemia was evident 4 days after ureterocolonic anastomosis. We did not consider this latter case as a complication associated with urinary diversion to the prepuce.

In conclusion, our novel method may be advantageous for the prevention of severe neurologic and gastrointestinal complications, which were the most common signs leading to euthanasia of the dogs in a previous study. This technique possibly could maintain kidney function for a longer term, although it still has high morbidity and complication rates compared with techniques used in human medicine.

Statistical analyses were challenged by the small number of cases and the heterogeneity of treatment among the dogs, due to choices by veterinarians and owners; however, we did obtain a competitive estimated median survival time of 385 d. The results of this study indicate the importance of radical removal of the organs lined with transitional epithelium, namely, the urinary bladder and urethra, which is supported by the notion of field effect or field cancerization (21,22). Including 3 female dogs which had urethral margins contaminated by invading tumor cells, 5 dogs developed tumor recurrence or distant metastasis. Therefore, all cases should be subjected to histopathological examination of surgical margins. Especially in female dogs, our procedure did not enable complete resection of the urethra, leaving a short portion of the urethra at the urethrovesicular junction. It is suggested that we should adopt a method that enables us to resect the urethrovesicular junction and cranial vagina more extensively, such as the part of the vaginourethroplasty reported in a previous study to treat infiltrative urethral disease (23). Moreover, pubic osteotomy may improve surgical access to this area and facilitate radical resection particularly in large breed dogs, although all patients enrolled in this study were less than 10 kg and the urethrovesicular junction was feasible without an osteotomy.

Our data indicate that we should administer systemic chemotherapy as soon as possible after the surgery. The possibility of glove contamination by tumor cells and dissemination during the operation were not excluded since TCC of the bladder is such an aggressive tumor a needle biopsy could spread tumor cells (1). Surgical gloves and instruments should be changed after en bloc resection of the bladder and the tumor.

This study is limited by a small heterogeneous dog population, incomplete long-term follow-up, and lack of information about the quality of life for both the dogs and the clients due to the retrospective nature of this study. Because the procedure requires the use of a diaper for the dog, client inquiry should be performed prospectively.