Outcomes of total cystectomy with medical treatment in canine urothelial carcinoma of the bladder trigone

Abstract

Objective

This study aimed to evaluate outcomes and complications in dogs with urothelial carcinoma (UC) of the bladder trigone treated with total cystectomy using uretero-prepuce/vagina/cutaneous anastomosis combined with medical treatment.

Animals

Twenty-one dogs.

Procedure

Total cystectomy was completed as follows: The whole bladder and urethra were removed, and the ureters were anastomosed to the skin in 1 case and to the vagina in 9 cases in females. The entire bladder, prostate, urethra, and penis were removed, and the ureters were anastomosed to the prepuce in 11 males. Medical treatments were administered to all dogs after surgery.

Results

Short-term surgical complications included diarrhea (16 dogs), acute kidney injury (2 dogs), and dehiscence of the ureterovaginal anastomosis (3 dogs). Long-term surgical and medical complications included pyelonephritis (11 dogs) and chronic kidney disease (5 dogs). In the survival analysis, median survival time from the initial diagnosis in all cases was 481 d, which was longer than the previously reported median survival time in dogs with UC involving the trigone.

Conclusion and clinical relevance

Total cystectomy combined with medical treatment prolonged overall survival compared with medical, surgical, or stenting procedures in dogs with UC involving the trigone.

INTRODUCTION

Urinary bladder urothelial carcinoma (UC) is the most common tumor of the urinary tract in dogs, accounting for ~2% of all canine malignancies (1). Most canine UC extensively invades the bladder wall and mimic human invasive bladder cancer, based on pathological findings. In addition, canine UC frequently originates from or invades the bladder trigone, where complete resection by partial cystectomy is virtually impossible, resulting in obstruction of the ureter and/or urethra in many cases (2). Medical treatment is essential, and various protocols have been reported using NSAIDs, cytotoxic chemotherapeutic drugs, and molecular targeted drugs, alone or in combination (3). Recently, clinical efficacies were reported for several molecular targeted drugs, including a multi-kinase inhibitor (toceranib), an anti-CCR4 antibody (mogamulizumab), a tyrosine kinase inhibitor (lapatinib) targeting epidermal growth factor receptor and human epidermal growth factor receptor 2, and a v-Raf murine sarcoma viral oncogene homolog B (BRAF) inhibitor (vemurafenib) (4,5). Although medical treatments have improved outcomes in dogs with UC, ureter or urethral obstruction caused by tumor invasion remains a lethal problem in many cases.

In humans, total cystectomy is a common surgical procedure for patients with invasive bladder tumors and is endorsed by global guidelines (6). In dogs, the first case series of total cystectomy with ureterocolonic anastomosis was reported as a completely resectable surgery for UC invading bladder trigone. However, all dogs showed azotemia due to intestinal recycling of urea or infectious pyelonephritis (7). A novel total cystectomy surgery by uretero-prepuce/vagina/uterus/cutaneous anastomosis was reported and feasibility of these procedures was confirmed (8–10). However, the median survival times (MST) post-surgery reported in the 2 case series were 278.6 d (10) and 385 d (8), in which most cases were treated with surgery exclusively. Compared with previous reports of medical treatment alone [MST: 176 to 344 d (1)], total cystectomy did not improve outcomes markedly, mainly due to tumor recurrence and/or metastasis. This study aimed to evaluate the outcomes in dogs with UC of the bladder trigone treated with total cystectomy using uretero-prepuce/vagina/cutaneous anastomosis combined with medical treatment.

MATERIALS AND METHODS

Case selection and retrospective survey

Data were collected through a retrospective survey of the medical records of 21 dogs (10 females and 11 males) with UC of the urinary bladder invading the trigone that were treated with total cystectomy combined with medical treatment at the Veterinary Medical Center, The University of Tokyo, between April 2016 and November 2021. Clinical data, including sex, age, breed, body weight (BW), surgical technique, operation time, histopathological diagnosis, complications, medical treatments, and cause and date of death, were recorded based on electronic medical records. All pathological diagnoses were made by veterinary pathologists certified by the Japanese College of Veterinary Pathologists, Department of Veterinary Pathology, The University of Tokyo. Informed consent for the use of data in this study was obtained from the owners of all study dogs.

Initial diagnosis and preoperative evaluation

In all cases, physical examination, complete blood (cell) count, serum chemical analysis, urinalysis, abdominal ultrasound, and thoracic and abdominal radiography were completed at the initial visit and after surgery. Preoperative diagnosis was based on the cytopathology of urine sediment and the urinary BRAF gene mutation examination, and postoperative diagnosis was based on the histopathology of the resected tissues in all cases. The BRAF^V595E gene mutation was examined by digital PCR assay, as previously described (11). Lung, lymph node, and bone metastases were evaluated post-surgery using ultrasound and radiography in all cases and using computed tomography (CT) in selected cases (19 cases). For the CT scans, an IV infusion of lactated Ringer’s solution (Lactated Ringer’s Solution “FUSO”; Fuso Pharmaceutical Industries, Osaka, Japan) was started 1 h before the scan, at 3 to 5 mL/kg per hour. During the scan, iohexol (IOHEXOL300 injection; Fuji Pharma, Toyama, Japan), 2mL/kg, was administered rapidly, and 3-phase computed tomographic angiography was completed. After receiving a CT scan, 4 dogs underwent surgery immediately and 15 dogs were recovered from anesthesia and continued receiving IV infusion for about 1 h.

The clinical stages were defined based on the World Health Organization tumor, node, metastasis classification of tumors in domestic animals (12). The T1 and T2 stages were not separately defined and were denoted as T1-2.

Medical treatment

Preoperative medical treatments were applied in all cases. Treatments included NSAIDs alone (piroxicam) in 10 dogs, NSAIDs (piroxicam) with lapatinib in 5, NSAIDs (piroxicam) with sorafenib in 2, NSAIDs (piroxicam) with mogamulizumab in 2, NSAIDs (piroxicam) with toceranib in 1, and NSAIDs (piroxicam) with mitoxantrone and sorafenib in 1. Clinical responses were evaluated according to the RECIST criteria (13).

Postoperative medical treatments were applied in 20 dogs. Treatments included NSAIDs alone (piroxicam) in 6 dogs; NSAIDs (piroxicam in 5 dogs and firocoxib in 1) with lapatinib in 6; NSAIDs (piroxicam) with mogamulizumab in 4; and sorafenib alone, toceranib alone, NSAIDs (firocoxib) with toceranib, and NSAIDs (piroxicam) with mitoxantrone in 1 dog each.

Analgesia/anesthesia and surgery

Before induction of anesthesia, atropine (Atropine sulfate injection, 0.5 mg; Nipro Es Pharma, Osaka, Japan), 10 μg/kg BW; fentanyl (Fentanyl injection, 0.25 mg; Terumo, Tokyo, Japan), 3 to 5 μg/kg BW; and ketamine (Ketalar for intramuscular injection, 500 mg; Daiichi Sankyo Propharma, Tokyo, Japan), 0.3 to 0.5 mg/kg BW, were administered IV as premedication. Propofol (1% propofol; Maruishi Pharmaceutical, Osaka, Japan), 6 to 8 mg/kg BW, IV, was administered for induction of anesthesia. Anesthesia was maintained using isoflurane (ds isoflurane; Bussan Animal Health, Osaka, Japan). For intraoperative analgesia, dogs received fentanyl constant-rate infusion (CRI), 10 to 20 μg/kg per hour; ketamine CRI, 1 to 5 mg/kg per hour; and epidural bupivacaine (Marcain Injection 0.5%; Sandoz Pharma KK, Tokyo, Japan), 1 mg/kg BW, with morphine (Morphine hydrochloride injection 10 mg; Daiichi Sankyo Propharma, Tokyo, Japan), 0.1 mg/kg BW. Anesthetic management was employed with a mean target blood pressure > 60 mmHg, using constant infusion of crystalloid liquid and ephedrine (Ephedrine Hydrochloride 40 mg; Nichi-Iko Pharmaceutical, Toyama, Japan), 0.1 mg/kg BW, IV, and norepinephrine CRI (Noradrenalin injection 1 mg; Alfresa Pharma, Osaka, Japan), 0.05 to 0.2 μg/kg per minute BW, if needed. The SpO₂ was maintained at > 95% and ETCO₂ was controlled within the range of 35 to 55 mmHg using spontaneous or controlled breathing. The body temperature was maintained between 36 and 38°C. For additional postoperative analgesia, fentanyl CRI, 3 to 5 μg/kg per hour, was administered, depending on the pain status. Cefazolin (Rasenazolin; Nichi-Iko Pharmaceutical, Toyama, Japan), 20 mg/kg BW, was administered IV q2h during surgery and q8h for 2 d after surgery.

Total cystectomy using uretero-prepuce/vagina/cutaneous anastomosis and its variant methods was as previously described (8,10). In females, the whole bladder and urethra were removed, and ureters were anastomosed to the skin in 1 dog and to vagina in 9 dogs. In 4 of 9 dogs, a ureter was anastomosed to the vaginal end (end-to-end anastomosis) and the other was anastomosed to a hole made with a trepan on the side of the vagina (end-to-side anastomosis). In another 4 of 9 dogs, the left and right ureters were anastomosed into a single “V” shape, which was then anastomosed to the vaginal end. In 1 of 9 dogs, both ureters were transplanted into the vagina with 2 separate holes made by a trepan (end-to-side anastomosis). In males, the entire bladder, prostate, urethra, and penis were removed, and ureters were anastomosed to the prepuce in all cases. To anastomose the ureter and prepuce, prepuces were punched using a dermal punch in all cases. Abdominal walls were punched using a dermal punch in 7 dogs, a 1 to 2-centimeter incision of the rectus abdominis muscle was made in 3 dogs, and the method was unrecorded in 1 dog. Histopathological margins of the resected tissue were evaluated in all cases, except for male dogs in which the urethra and penis were completely resected. Scheduling for surgery was decided based on tumor progression owing to a lack of response to preoperative medical treatment, and on owner and veterinarian decisions. The group that underwent surgery within 2 wk of the initial diagnosis was defined as the “early surgery” group, and the group that underwent surgery > 2 wk after the initial diagnosis was defined as the “late surgery” group. Incontinence was managed using a disposable diaper changed several times daily.

Definitions of complications

Medical complications were evaluated according to the VCOG-CTCAE criteria (14) in the pre- and postoperative periods.

Complications that occurred before surgery were defined as preoperative medical complications. Data on preoperative medical complications in the dogs treated with sorafenib could not be collected due to data-masking requirements for the prospective clinical trial.

Complications observed after surgery until discharge from the Veterinary Medical Center, The University of Tokyo were defined as short-term surgical complications; those after discharge were defined as long-term surgical and medical complications. For postoperative complications, the following additional criteria were used. According to the International Renal Interest Society (IRIS) staging guidelines (15), postoperative oliguria or anuria indicates acute kidney injury (AKI). Postoperative pyelonephritis was diagnosed based on ultrasound findings of the kidneys, urinalysis, bacterial culture of the urine, and clinical signs. Hydronephrosis and ureteral obstruction were diagnosed using ultrasonography. Elevated blood creatinine concentration was defined as chronic kidney disease (CKD), based on the IRIS staging guidelines, using the most abnormal blood examination data during the observation period. Dogs diagnosed with CKD before surgery were excluded from the postoperative CKD group.

Statistical analyses

Survival curves and MST were calculated using the Kaplan–Meier method, and univariate analyses used the log-rank test. The MST from the initial diagnosis were compared by the types of preoperative medical treatments, sex, and timing of surgery; MST from the surgery were compared by sex. Length of operation time by sex was compared using the t-test. The incidence of AKI according to whether preoperative CT was taken on the day of surgery or on a different day was analyzed using Fisher’s exact test. Statistical significance was set at P < 0.05. Statistical analyses used SAS Viya software (SAS Institute Austria, Vienna, Austria) and StatView software (SAS Institute Japan, Tokyo, Japan).

RESULTS

Clinical characteristics of the cohort

Twenty-one dogs that underwent total cystectomy combined with medical therapy were included in this study. The dogs included 10 females (7 neutered and 3 intact) and 11 males (9 neutered and 2 intact) with a median age of 11 y (range: 7 to 14 y) and a median BW of 6.2 kg (range: 2.6 to 13.6 kg). Shetland sheepdog and Pomeranian (3 each) were the most common breeds; followed by French bulldog, Jack Russell terrier, and miniature dachshund (2 each); and toy poodle, Scottish terrier, Yorkshire terrier, Welsh corgi, miniature schnauzer, Chihuahua, papillon, mongrel, and Cavalier (1 each). A BRAF^V595E gene mutation was observed in 12/20 (60.0%) dogs. The clinical stages at initial diagnosis were T1-2N0M0 in 18 dogs, T3N0M0 in 2, and T1-2N1M0 in 1. A summary of case data is shown in Table S1 (available online from: Supplementary Materials).

Outcomes and complications of preoperative medical treatment

Clinical responses to preoperative medical treatment were evaluated, except for 6 dogs that underwent surgery within 2 wk of the initial diagnosis (early surgery group). Among the 4 dogs that received NSAIDs alone, 2 had progressive disease and 2 had stable disease but progressed and underwent surgery at a median of 62 d (range: 43 to 82 d) from the initial diagnosis. The other 5 dogs that received NSAIDs and lapatinib showed a partial response for a median of 169 d (range: 54 to 394 d) but then progressed and underwent surgery at a median of 232 d (range: 62 to 475 d) from the initial diagnosis. Two dogs that received NSAIDs and mogamulizumab showed stable disease for 225 and 177 d, but the disease progressed and they underwent surgery on 265 and 182 d, respectively, after initial diagnosis. Of the other 2 dogs that received NSAIDs and sorafenib, 1 showed partial response for 146 d and the other showed stable disease to progressive disease; these dogs underwent surgery at 396 and 153 d, respectively, after initial diagnosis. A dog that received NSAIDs, mitoxantrone, and sorafenib showed stable disease after 2 rounds of treatment with mitoxantrone, but was switched to sorafenib owing to myelosuppression. Although partial remission was achieved for 151 d with sorafenib combined with NSAIDs, the disease subsequently progressed, and surgery was done 477 d after initial diagnosis. Another dog that received NSAIDs and toceranib showed stable disease for 619 d with NSAIDs alone, but then progressed, and toceranib was added. However, the dog showed disease progression, and surgery was done 670 d after initial diagnosis. There were no statistically significant differences in the clinical responses to each preoperative medical treatment (P = 0.1859). A summary of these data is shown in Table S2 (available online from: Supplementary Materials).

Complications of preoperative medical treatment included vomiting or diarrhea in 4 dogs [all Grade (G) 2], elevated BUN (G1, G1, G2, G3) or creatinine (G2) in 4, elevated alanine aminotransferase (G2, G3) in 2, and skin hypopigmentation (G1) and bone marrow suppression (G3) in 1 dog each.

Surgical records and short-term surgical complications

The median operation time was 219 min (range: 170 to 306 min) for females and 228 min (range: 111 to 277 min) for males, with no significant difference in operation time between sexes (P = 0.8579). Of the 10 females, the resection margin of the ureter was complete in 9. In 1 dog, the resected margins were incomplete. The resected margin of the urethra was complete in 5 and incomplete in 5 dogs. No recurrence at the urethral resection site was observed in any case. Of the 11 males, the ureteral resection margin was complete in 8. In 2 dogs, the resected margins were incomplete, and tumor recurrence was observed in 1. In the other dog, the resection margin was unrecorded (Table S3, available online from: Supplementary Materials).

The short-term surgical complications are summarized in Table 1. Short-term surgical complications were observed in 9 of 10 female and 9 of 11 male dogs. Diarrhea was observed in 8 female and 8 male dogs, and a median of 14 d (range: 7 to 105 d) were required for recovery.

TABLE 1.

Short-term postoperative complications.

Female (n = 10)	Number of cases	Male (n = 11)	Number of cases
Diarrhea	8	Diarrhea	8
Acute kidney injury	2	Pyelonephritis	1
Ureteral anastomosis dehiscence	3	Ureteral obstruction	1
Euthanasia (acute kidney injury)	1

There was AKI in 2 of the 19 dogs that underwent contrast-enhanced CT scanning before surgery. There was a statistically significant difference in incidence of AKI between dogs for which CT was used on the day of surgery (2 of 4) and dogs that were scanned > 1 d before surgery (0 of 15) (Fisher’s exact test, P = 0.0351). Both dogs with AKI required peritoneal dialysis: 1 recovered and 1 was euthanized because there was no improvement.

Dehiscence of the ureterovaginal anastomosis was observed in 3 female dogs. Three cases of dehiscence of the ureterovaginal anastomosis were observed in 2 neutered dogs and 1 intact dog. Among these cases, 2 required reanastomosis surgery. Pyelonephritis was observed in 1 male. Dogs with pyelonephritis were treated with antibiotics, which resulted in improvement. Obstruction of the ureteral orifice caused by debris-like material was observed in 1 male. The obstruction was relieved by catheter insertion.

Outcomes and complications of long-term surgical and medical treatment

Long-term surgical and medical complications were evaluated in 20 dogs that survived the perioperative period (Table 2). Complications included gastrointestinal (G2 vomiting or diarrhea) in 3 dogs, and pancreatitis (G2), aspiration pneumonia (G3), erythema multiforme (G1), and elevated alanine aminotransferase (G3) in 1 each. Urinary tract-related complications were frequently observed; pyelonephritis was observed in 11/20 (55.0%) dogs and CKD in 5/18 (27.8%) dogs. Two were not included in this criterion because of preoperative CKD. Based on antimicrobial susceptibility testing, pyelonephritis responded well to medical treatment using antibiotics; however, disease recurrence was observed in 4 dogs. Chronic kidney disease was categorized as IRIS Stage 2 in 2 dogs and Stage 3 in 3.

TABLE 2.

Long-term surgical and medical complications.

Complication	Female (n = 9)	Male (n = 11)	Total incidence
Pyelonephritis	6	5	55.0%
Chronic kidney disease	2	3	27.8%a
Gastrointestinal disorders (vomiting or diarrhea)	2	1	15.0%
Aspiration pneumonia	1	0	5.0%
Elevated alanine aminotransferase	0	1	5.0%
Erythema multiforme	1	0	5.0%

^aIn males, there were 2 cases with preoperative chronic kidney disease (n = 9).

In the survival analysis, the MST from the initial diagnosis in all cases was 481 d (Figure 1 A). There was no significant difference in MST from the initial diagnosis between sexes [female (n = 10) versus male (n = 11): 481 d (range: 77 to 963 d) versus 530 d (range: 279 to 1833 d), P = 0.4176; Figure 1 B]. Comparing the early surgery group and the late surgery group, the MST from initial diagnosis in the late surgery group tended to be longer than that in the early surgery group, though the difference was not significant [early surgery group (n = 6) versus late surgery group (n = 15): 365 d (range: 77 to 963 d) versus 691 d (range: 132 to 1833 d), P = 0.2435; Figure 1 C]. The postoperative MST of the 21 dogs was 307 d (range: 3 to 1777 d) (Figure 1 D). There was no significant difference in postoperative MST between sexes [female (n = 10) versus male (n = 11): 357 d (range: 3 to 951 d) versus 307 d (range: 76 to 1777 d), P = 0.6977; Figure 1 E).

FIGURE 1.

Survival analyses of the dogs treated with total cystectomy combined with medical treatment. Kaplan–Meier curves depicting survival time from the initial diagnosis for all dogs (A) and the effects of sex (B) and surgery timing (C); and postoperative survival time for all dogs (D) and the effect of sex (E).

MST — Median survival time.

Tumor-related death was observed in 11/16 (68.8%) dogs, and 5 were alive during the observation period (survival times from initial diagnosis: 77, 267, 384, 397, and 743 d). Chronic kidney disease-related death was observed in 2 dogs, worsening heart and kidney disease in 1, and perioperative death in 1 (euthanasia due to AKI unresponsive to treatment). Among the non-tumor-related deaths, 3 dogs survived > 2 y after surgery; post-surgery survival times of those dogs were 836, 951, and 1777 d (Table 3). Those cases were clinical stage T1-2N0M0 at initial diagnosis, and no recurrences or metastases were observed during the observation period.

TABLE 3.

Characteristics of dogs with longer survival times.

Case No.a	Breed	Age	TNM stage at surgery	Preoperative treatment	Postoperative treatment	Surgical margin	Short-term complications	Long-term complications	Cause of death	Postoperative survival time (d)
1	Scotch terrier	9 y 3 mo	T2N0M0	Piroxicam	Piroxicam	Ureter: clear Urethra: dirty	None	Chronic kidney disease	Renal failure	951
4	Welsh corgi	10 y 9 mo	T2N0M0	Piroxicam Lapatinib	Piroxicam Lapatinib	Ureter: clear Urethra: clear	Ureteral anastomosis dehiscence Diarrhea	Pyelonephritis	Renal failure	836
11	Chihuahua	7 y 5 mo	T1N0M0	Piroxicam	Piroxicam Mitoxantrone	Ureter: clear	Diarrhea	None	Heart and renal failure	1777

TNM — World Health Organization tumor, node, metastasis classification of tumors in domestic animals.

^aCase No. corresponds to Case No. in Table S1 (available online from: Supplementary Materials).

DISCUSSION

In this study, we retrospectively evaluated the outcomes and complications of dogs with UC of the bladder trigone treated with a combination of total cystectomy and medical treatments. We showed that the MST from initial diagnosis in the 21 dogs was 481 d. In previous reports, the MST of dogs treated with total cystectomy alone were 278.6 d (10) and 385 d (8). However, the MST (481 d) from initial diagnosis of the dogs treated with total cystectomy and medical therapy could not be directly compared to the MST from the other 2 studies, as neither accurately identified the preoperative medical management and survival time before surgery. In reports on medical treatment alone, the MST of dogs treated with piroxicam alone was 244 d (1), that of dogs treated with piroxicam combined with carboplatin was 161 d (16), and that of dogs treated with piroxicam combined with doxorubicin was 168 d (17). In reports on local treatment combined with medical treatment, the MST of dogs treated with urethral stents and medical treatment was 153 d (18) and that of dogs treated with ureteral stents and medical treatment was 285 d (19). Therefore, our results suggested that total cystectomy combined with medical treatment prolonged overall survival times compared with surgery or medical treatment alone or stenting procedures and medical treatment in dogs with UC involving the trigone. Moreover, there was no significant difference in the MST from the initial diagnosis between male and female dogs, suggesting that total cystectomy combined with medical treatment was beneficial regardless of sex.

Furthermore, the duration before surgery was 188 d in this study, similar to the reported MST in dogs receiving medical treatment alone (1,16,17). In general, urinarytract obstruction is the most common cause of death in dogs receiving only medical treatment (20). Moreover, the most frequent reason for undergoing total cystectomy in this study was obstruction of the urinary tract. Total cystectomy relieved the obstruction of the urinary tract caused by urinary-tract alteration. Therefore, in this study, the longer survival in dogs could suggest that the medical treatment prolonged survival during the periods before and after surgery, as well as during the preoperative medical treatments. Further, the postoperative medical treatments and the total cystectomy prevented death caused by urinary obstruction.

However, this study did not clarify the optimal timing of surgical intervention. Theoretically, early radical surgery can completely remove gross tumor lesions and prevent systemic metastases. In contrast, surgery-induced metastasis owing to the induction of angiogenesis and suppression of antitumor immunity has been reported in clinical and preclinical settings (21). Moreover, surgery-induced metastases have been reported in canine UC (21). In fact, the dogs with complete margins in this cohort also had tumor-related deaths. The advantages and disadvantages of radical surgery complicate the optimal timing of total cystectomy in dogs with UC involving the trigone. Because the timing of surgery varied in each dog in this cohort, we evaluated the MST between dogs that underwent early surgery (365 d) and those that underwent late surgery (691 d; P = 0.2435). The late surgery group tended to have longer survival compared to the early surgery group, although the difference was not significant, perhaps because of low statistical power due to the small numbers. Further prospective analyses are needed to clarify the optimal timing of surgical intervention in dogs with UC involving the trigone.

As for short-term surgical complications, diarrhea might have been due to pelvic-cavity damage and inflammation associated with surgical manipulations, because the clinical signs disappeared within 15 d in most dogs. To the best of our knowledge, there is no report of diarrhea as a short-term complication in humans undergoing minimally invasive total cystectomy. Therefore, episodes of diarrhea may be prevented by the introduction of minimally invasive surgery, such as robot-assisted surgery, in dogs undergoing total cystectomy (22,23). In 2 dogs with AKI, contrast-enhanced CT was done on the day of surgery and the difference in having AKI was statistically significant (P = 0.0351). The cause of the relationship between AKI and contrast-enhanced CT on just the day of surgery was unknown, but there are several possible reasons. A CT scan just before surgery may prolong the anesthesia time and induce kidney injury. The CT contrast agent was reported to induce renal injury in humans (24), and CT contrastagent injection just before surgery may have induced kidney injury in this cohort. Although further studies regarding the timing of contrast-enhanced CT scans in dogs treated with total cystectomy are needed, contrast-enhanced CT scans just before surgery may be best avoided.

Another common short-term surgical complication was dehiscence of the ureteral anastomosis in 3/24 (12.5%) cases, which was also observed in previous case series (8). In this study, all dehiscences were disruptions of the ureterovaginal anastomosis in female dogs. In female dogs, it is necessary to manipulate the ureter in the abdominal cavity within a narrow surgical field and anastomose the 2 thin ureters with different diameters to the vagina. Moreover, 2 dogs had previously undergone ovariohysterectomy, and the short remaining vagina may have caused tension at the anastomotic site. Therefore, retroperitoneal detachment and caudal transposition of the kidneys may be useful in such cases (25).

Regarding long-term surgical and medical complications, most dogs had pyelonephritis (13/23, 56.5%). Although dogs tolerated disposable diapers, urine retention or fecal contamination in the diaper could be sources of retrograde contamination. Therefore, frequent changing of disposable diapers and careful hygiene management of the ureterostomy site are important considerations to prevent pyelonephritis during long-term care.

Another common long-term surgical and medical complication was CKD (6/21, 28.5%) causing death in 17.6% dogs. In humans, deterioration of renal function was a major postoperative long-term complication, observed in 72% patients who underwent total cystectomy (26). The major causes of postoperative CKD in humans are hydronephrosis, pyelonephritis, and ureteral strictures. In this study, ≥ 1 episodes of postoperative pyelonephritis were observed in 56.5% of dogs; therefore, pyelonephritis might contribute to CKD progression. Moreover, almost all dogs were treated with NSAIDs, which significantly decreased glomerular filtration rate and caused renal damage in dogs (27). Use of NSAIDs may also contribute to CKD development. Although NSAIDs are widely used as evidence-based treatments in canine UC (1,28), more careful use, such as conducting case selection and periodic blood and urine examination, is important to reduce long-term complications and improve outcomes.

This study had several limitations. First, the small number of analyzed cases from the 6-year period of the retrospective study might not have been sufficient to reliably clarify the clinical effects of each parameter in detail, owing to low statistical power. Second, clinical procedures, such as disease staging, timing of surgery, and medical treatment protocol, varied because of the retrospective setting.

To summarize, the combination of total cystectomy and medical treatment prolonged the survival of dogs with UC involving the trigone. Moreover, this study identified several frequent complications, including diarrhea, AKI, and dehiscence, as short-term surgical complications; and pyelonephritis and CKD as long-term surgical or medical complications.