Naturally-occurring canine transitional cell carcinoma of the urinary bladder A relevant model of human invasive bladder cancer

Urologic Oncology 5 (2000) 47–59

Review article

Naturally-occurring canine transitional cell carcinoma of the urinary bladder A relevant model of human invasive bladder cancer Deborah W. Knapp, D.V.M.a,*, Nita W. Glickman, M.P.H.c, Dennis B. DeNicola, Ph.D.b, Patty L. Bonney, B.S.a, Tsang L. Lin, D.V.M.b, Lawrence T. Glickman, V.M.D.b a

Department of Veterinary Clinical Sciences, School of Veterinary Medicine, 1248 Lynn Hall, West Lafayette, IN 47907, USA b Department of Veterinary Pathobiology, School of Veterinary Medicine, 1248 Lynn Hall, West Lafayette, IN 47907, USA c Research Programs, Purdue University, School of Veterinary Medicine, 1248 Lynn Hall, West Lafayette, IN 47907, USA Received 25 January 1999

Abstract Invasive bladder cancer results in over 10,000 deaths yearly in the United States alone. More effective therapy for invasive bladder cancer is clearly needed. As new cellular and molecular targets for therapy are identified, relevant animal models are needed to test new therapeutic strategies aimed at these targets prior to human clinical trials. The purpose of this review is to characterize spontaneous invasive transitional cell carcinoma of the urinary bladder (TCC) in dogs, to summarize the similarities and differences between canine and human invasive TCC, and to describe how canine TCC could serve as a relevant model of human invasive bladder cancer. Information was summarized from 102 dogs with TCC evaluated and treated at the Purdue University Veterinary Teaching Hospital, from a review of the Veterinary Medical Data Base, and from reports in the literature. Canine TCC was found to be very similar to human invasive bladder cancer in histopathologic characteristics, molecular features, biological behavior including metastasis, response to medical therapy, and prognosis. Differences between canine and human TCC were few, but included gender predilection with a male:female ratio of 2.8:1 in humans versus a male:female ratio of 0.5:1 in dogs. The location of the TCC within the bladder also differed: Most canine TCC was trigonal in location, whereas more than 50% of human TCC was in the lateral and posterior walls of the bladder. Considering the great similarity between invasive bladder cancer in humans and dogs, spontaneous canine TCC can be considered a relevant animal model of human invasive bladder cancer. © 1999 Elsevier Science Inc. All rights reserved. Keywords: Bladder cancer; Animal models; Canine

Bladder cancer is diagnosed in over 50,000 people yearly in the United States alone [1]. Although the superficial form of bladder cancer typically responds favorably to transurethral resection and intravesical therapy, invasive bladder cancer [transitional cell carcinoma (TCC)] is often more resistant to therapy, metastasizes widely, and results in over 10,000 deaths yearly in the United States [1]. A better understanding of the molecular processes involved in invasive bladder cancer growth and progression will facilitate identification of new targets for therapy [2]. Relevant animal models of invasive bladder cancer are needed to test new therapies prior to clinical trials in humans. Although in vitro * Corresponding author. Tel.: 1011-765-494-9900; fax: 1011-765496-1108. E-mail address: [email protected] (D.W. Knapp) There is not a financial relationship between any of the authors and the subject matter.

model systems provide valuable information on cellular and molecular processes, in vivo models with intact physiologic processes such as the immune system, angiogenesis pathways, and multiple organ interactions are clearly needed. The ideal animal model for human cancer would be one that is readily available and inexpensive, and in which the cancer in the animal mimics the specific form of human cancer in histopathologic characteristics, cellular and molecular features, biological behavior, and response to therapy. Most importantly, the antitumor response and the adverse effects of a specific therapy in the animal model should be predictive of the tumor response and adverse effects in humans. Numerous experimental rodent models of bladder cancer have been investigated. Chemical carcinogens such as N-butyl-N-(4-hydroxybutyl)nitrosamine and others have been used to induce bladder tumors in rats and mice [3–5]. Following N-butyl-N-(4-hydroxybutyl)nitrosamine expo-

1078-1439/00/$ – see front matter © 1999 Elsevier Science Inc. All rights reserved. PII: S1078-1439(99)00 0 0 6 - X

48

D.W. Knapp et al. / Urologic Oncology 5 (2000) 47–59

sure, mice develop nodular invasive carcinoma preceded by carcinoma in situ, and rats develop polypoid exophytic masses with late invasion. Although these rodent models provide useful information concerning the risk of chemical exposure and bladder carcinogenesis, they have limitations due to low rates of metastic spread [6]. Developing effective strategies to treat or prevent metastatic disease is currently one of the most pressing challenges in cancer therapy. Another type of rodent model is produced by the orthotopic implantation of human bladder cancer cells in athymic mice [7,8]. By repeatedly cycling the cancer cells through immunodeficient mice, cell lines have been developed that produce primary tumors in 100% of mice and lung metastases in 30% of mice [7]. Another, often overlooked animal model is spontaneous cancer in pet dogs. The advantages of this model system have been reviewed [9] and include: (1) similarities between specific types of canine cancer and human cancer in histopathologic appearance, biological behavior, and response to therapy, (2) similar drug metabolism between dogs and humans, (3) less constraints in testing new therapies because “standard” therapy is not well defined for many canine cancers, (4) the compressed lifespan of the dog, which makes completion of clinical studies possible in a timely manner, (5) the fact that dogs share the environment with their owners and thus have similar exposures to water, passive cigarette smoke, and insecticides, (6) the larger size of the dog compared with rodents, which makes many medical procedures technically feasible, and (7) the concept that novel interventional strategies developed in vitro or in laboratory animal studies can be tested in vivo in spontaneous tumor-bearing dogs, whereas similar studies may be unacceptable or less feasible in humans with cancer, especially when a standard, yet only partially effective, treatment exists [9]. The study of spontaneous animal cancer is also more acceptable to a society concerned with animal welfare and inducing disease in animals for research. In addition, any comparative studies performed in pet animals clearly benefit both animals and humans. One form of naturally-occurring cancer in pet dogs with great similarity to human cancer is invasive TCC of the urinary bladder. In this article, we review our current knowledge of canine TCC in regards to frequency, etiology, histopathologic characteristics, cellular and molecular features, response to therapy, and prognostic factors. We discuss why canine TCC has great potential as a model of human invasive bladder cancer. Information presented in this article is derived from several sources including: (1) a series of 102 dogs with TCC treated in clinical trials in the Purdue Comparative Oncology Program (PCOP) at the Purdue University Veterinary Teaching Hospital (PUVTH), (2) analysis of records in the Veterinary Medical Data Base (VMDB) [10], and (3) reports in the literature. All clinical trials in dogs conducted by the PCOP have been approved by the Purdue Animal Care and Use Committee.

1. Spontaneous canine transitional cell carcinoma of the urinary bladder 1.1. Frequency Although the true incidence of canine TCC is not known, TCC is the most common form of urinary tract cancer in the dog and comprises 1.5 to 2% of all canine cancers [11,12]. Bladder cancer occurs less frequently than several other canine malignancies, but numerous cases of TCC are diagnosed each year by veterinarians. An estimated 54 million dogs in the United States receive routine veterinary care [9], and cancer is the leading cause of death in older pet dogs. In a necropsy study including over 2,000 animals, 45% of dogs over the age of 10 years had cancer [13]. Therefore, even unusual forms of cancer such as TCC affect numerous dogs yearly. The hospital prevalence or proportionate morbidity of bladder cancer at university-based veterinary hospitals appears to be increasing. A search of the VMDB [10] from 1975 through 1995 showed a continuous increase in the prevalence of bladder cancer, with prevalence defined as the number of dogs with bladder cancer divided by the total number of individual dogs seen for any reason at the same participating university veterinary hospitals (Fig. 1). 1.2. Etiology and risk factors The etiology of naturally-occurring canine bladder cancer is most likely multifactorial. Risk factors that have been identified include exposure to topical insecticides for flea and tick control, exposure to marshes that have been sprayed for mosquito control, obesity, possibly cyclophosphamide administration, female gender, and breed (e.g., Scottish Terrier) [14–16]. In addition to these risk factors, a 1981 study showed a significant positive correlation between the proportional morbidity ratios for canine bladder cancer and the overall level of industrial activity in the host county of the veterinary hospital [17]. Mortality from blad-

Fig. 1. Hospital prevalence of canine bladder cancer. Veterinary Medical Data Base (n 51,290).

D.W. Knapp et al. / Urologic Oncology 5 (2000) 47–59

49

Table 1 Body conformation, exposure to flea and tick dips, and the risk of bladder cancer in pet dogs [14] Exposure to flea and tick dips Gender

Body conformation

No [odds ratio (95% CI)]

Yes [odds ratio (95% CI)]

Male Male Female Female

Thin or average Overweight or obese Thin or average Overweight or obese

1.0* 2.2 (0.5–10.0) 1.0* 1.8 (0.5–4.8)

0.7 (0.3–2.6) 3.5 (0.4–32.9) 1.3 (0.3–3.0) 28.5 (3.1–237.2)

* Reference category.

der cancer among Caucasian men and women in the same counties showed a similar correlation with industrial activity [17], thus suggesting a role for environmental pollution. In addition to spontaneous TCC, bladder tumors can be experimentally induced in dogs in a laboratory setting with chemical carcinogens such as N-butyl-n-(4-hydroxybutyl) nitrosamine [18]. 1.2.1. Insecticides, obesity, and gender In a case-control study, 59 pet dogs with TCC and 71 age- and breed size-matched control pet dogs with other chronic disease or neoplasms were studied to determine if an association existed between bladder cancer and exposure to sidestream cigarette smoke and chemicals in the home, use of topical insecticides, and obesity [14]. Although bladder cancer risk was unrelated to sidestream cigarette smoke and household chemical exposure, the risk was significantly increased by topical application of flea and tick dip (Table 1). This risk was dose-related and was enhanced in overweight or obese dogs (Table 2). In addition, living within 1 mile of a marsh that was sprayed with insecticide was associated with an increased risk of TCC in these pet dogs (Table 3). Interestingly, when the types of insecticides (e.g., cholinesterase inhibitors, pyrethrins, and others) were classified, no one specific chemical type accounted for the increased risk. The authors speculated that it was the “inert” ingredients, which accounted for more than 95% of the total product, that were the probable bladder carcinogens in these products. These “inert” ingredients consisted of solvents such as benzene, toluene, xylene, and petroleum distillates. Benzene is one of several compounds associated with an increased risk of human bladder cancer [19]. Furthermore, some of these “inert” ingredients are lipophilic and are

stored in fat. This could explain the increased risk of TCC in obese and female dogs [14]. The lack of an association between sidestream cigarette smoke and the risk of canine TCC [14] was not surprising. Although smoking is thought to contribute to 50% or more of human bladder cancer in men and 33% or more of bladder cancer in women, the risk of TCC associated with exposure to sidestream smoke is less well defined. The risk of bladder cancer increases linearly (two- to three-fold) for persons who smoke at least 10 cigarettes per day, and then increases substantially in people who smoke 40 to 60 cigarettes per day [20,21]. It is highly likely that the dogs’ exposure to carcinogens in sidestream cigarette smoke was insufficient to contribute to the development of TCC. Multiple studies have confirmed the increased risk of bladder cancer in female dogs. The female:male ratio in a series of 102 dogs with TCC treated at the PUVTH was 1.7:1. Similarly, a study of the VMDB in which dogs with TCC were compared with institution-matched and agematched control dogs without bladder cancer confirmed the increased risk of TCC in female dogs (Table 4). Neutered dogs of both genders were at increased risk compared with sexually intact dogs of the same gender, although the age at surgical neutering was not known. 1.2.2. Cyclophosphamide exposure Cyclophosphamide causes sterile hemorrhagic cystitis in humans and pet dogs [15,16,22]. When used as an oncolytic or immunosuppressive agent in humans, cyclophosphamide increases the risk of bladder cancer ninefold [22]. TCC also has been reported in a small number of pet dogs following cyclophosphamide treatment for other malignancies [15,16].

Table 2 Body conformation, number of flea and tick dips, and risk of bladder cancer in pet dogs [14] Adult body conformation Dip applications per year

Thin [odds ratio (95% CI)]

Average [odds ratio (95% CI)]

Overweight/obese [odds ratio (95% CI)]

0 1–2 >2

1.0* 2.0 (0.5–5.1) 4.4 (1.5–13.1)

2.0 (0.4–11.1) 3.9 (0.3–56.2) 8.8 (0.5–145.3)

5.5 (0.9–32.3) 11.0 (0.7–163.7) 24.5 (1.4–423.2)

* Reference category.

50

D.W. Knapp et al. / Urologic Oncology 5 (2000) 47–59

Table 3 Exposure to marsh (living within 1 mile of marsh), obesity, and risk of bladder cancer in pet dogs [14] Exposure to marsh Body conformation

No [odds ratio (95% CI)]

Yes [odds ratio (95% CI)]

Thin or average Overweight or obese

1.0* 2.7 (0.5–12.2)

3.1 (0.7–18.6) 13.2 (1.9–32.3)

* Reference category.

1.2.3. Breed predisposition The risk of canine TCC varies considerably between different breeds (Table 5), with the Scottish Terrier at 19-fold increased risk compared with mixed breeds. There is no association, however, between breed and the biological behavior of the tumor or its response to therapy. The cause of the breed-associated risk is not known, but in all likelihood, it represents genetic predisposition to bladder cancer such as differences in metabolic activation and detoxification pathways. 1.3. Histopathologic characteristics Tumor tissue samples from 74 dogs with TCC treated at the PUVTH were classified using a system developed by Mostofi et al. [23] and modified by Valli et al. [24] and DeNicola (Table 6). The majority of tumors were papillary infiltrative TCC of intermediate to high grade (Table 6), with great similarity between canine and human TCC (Fig. 2). These findings were similar to those reported by Valli et al., except for a slightly higher percentage of dogs with carcinoma in situ in Valli et al. [24]. 1.4. Other cellular and molecular features In a report by Clemo et al. [25] 34 (79%) of 43 canine TCCs were aneuploid. The immunoreactivity of canine TCC (n 5 51 cases) to monoclonal antibodies (B72.3, CC49, CC83) for tumor-associated glycoprotein 72 (TAG-72) also was reported. Of 51 samples of TCC, 53% had immunoreactivity to Mab B72.3, whereas tissues from normal ca-

Table 4 Gender and risk of bladder cancer in pet dogs Gender

TCC Cases

Controls

Total

Odds ratio (95%CI)

Male Female Intact female Neutered female Intact male Neutered male

432 856 130 726 174 258

630 637 285 352 462 168

1,062 1,493 415 1,078 636 426

1.0* 1.96 (1.67–2.30) 1.0* 4.52 (3.54–5.77) 1.0* 4.08 (3.14–5.29)

Note: Summary of data from the Veterinary Medical Data Base comparing dogs with transitional cell carcinoma (TCC) to institution and age matched control dogs without TCC. * Reference category.

Table 5 Breed and risk of bladder cancer in pet dogs Breed

Odds ratio

95% Confidence interval

Mixed breed All pure breeds Scottish Terrier Beagle Shetland Sheepdog Wire Hair Fox Terrier West Highland White Terrier Miniature Poodle Miniature Schnauzer Doberman Pinscher Labrador Retriever Golden Retriever German Shepherd

1.0* 0.74 18.09 4.15 4.46 3.20 3.02 0.86 0.92 0.51 0.46 0.46 0.40

0.62–0.88 7.30–44.86 2.14–8.05 2.48–8.03 1.19–8.63 1.43–6.40 0.55–1.35 0.54–1.57 0.30–0.87 0.30–0.69 0.30–0.69 0.26–0.63

Note: Summary of data from 1,290 dogs with transitional cell carcinoma (TCC) and 1,290 institution- and age-matched control dogs without TCC in the Veterinary Medical Data Base. * Reference category.

nine urinary bladders and hyperplastic/inflamed urinary bladders did not. However, as is the case with human TCC, immunoreactivity to B72.3 was not specific for canine TCC [25–27]. Immunoreactivity to other canine carcinomas was noted as well. Studies of tumor suppressor gene and oncogene expression in canine TCC are very limited. In a small number of canine TCC tissue samples examined in our laboratory, marked immunoreactivity to p53 antibodies was observed (D Knapp, T Lin, unpublished data). Positive immunoreactivity is thought to occur most frequently with the presence of mutated p53 protein, which has a prolonged half life, thus facilitating its detection. A canine TCC cell line [28] also had immunoreactivity for p53 antibodies on Western blot (D Knapp, K Coffman, unpublished data). In contrast, Gamblin et al. [29] reported only small numbers of TCC cell nuclei having immunoreactivity for p53. Basic fibroblast growth factor (bFGF) is a potent proangiogenic peptide that has been found in high concentrations in the urine of humans with urologic and nonurologic malignancies [30,31]. The concentration of bFGF in the urine of dogs with TCC was significantly higher than that of normal dogs or dogs with bacterial cystitis [32]. In most dogs with TCC, the urine bFGF concentration was over sevenfold higher than in normal dogs. It has been hypothesized that the high bFGF concentration in the urine of human TCC patients is tumor cell-derived bFGF that is cleared from the circulation by glomerular filtration [31]. Expression of vascular endothelial growth factor (VEGF) and other angiogenic factors that are important in human TCC [33] have not yet been studied in canine TCC. Interest in arachidonic acid and its metabolites in TCC has grown with the documentation of tumor remission induced by a nonsteroidal antiinflammatory drug (NSAID; cyclooxygenase inhibitor) in spontaneous canine TCC [34] as well as the chemopreventative effects of these drugs in

D.W. Knapp et al. / Urologic Oncology 5 (2000) 47–59

51

Table 6 Histopathologic features of canine TCC based on 74 dogs examined at the Purdue University Veterinary Teaching Hospital Histopathologic feature Histologic classification Papilloma, transitional Papilloma, squamous Carcinoma, transitional, various elements, not squamous or glandular Carcinoma, transitional, various elements, squamous differentiation Carcinoma, transitional, various elements, glandular differentiation Carcinoma, squamous Carcinoma, adenocarcinoma Undifferentiated Growth pattern Papillary, noninfiltrative Papillary, infiltrative Nonpapillary, noninfiltrative Nonpapillary, infiltrative Undeterminable* Extent of infiltration None Lamina propria* Muscle* Perivesicular Undeterminable* Histologic grade 1 2 3 Vascular invasion No Yes Lymphoid reactivity around the tumor None Few cells Moderate Marked Lymphoid reactivity within the tumor None Few cells Moderate Marked

Number (%) 0 (0) 0 (0) 58 (78.4) 3 (4.1) 12 (16.2) 1 (1.4) 0 (0) 0 (0) 1 (1.4) 33 (44.6) 0 (0) 12 (16.2) 28 (37.8) 1 (1.4) 32 (43.2) 18 (24.3) 4 (5.4) 19 (25.7) 2 (2.7) 60 (81.1) 12 (16.2) 64 (86.5) 10 (13.5) 27 (36.5) 28 (37.8) 17 (23.0) 2 (2.7) 25 (33.8) 45 (60.8) 4 (5.4) 0 (0)

* Some of the tumor samples obtained by cystoscopy or catheter biopsy were too limited in size or depth to determine the growth pattern and extent of infiltration. It is likely that muscle invasion was present in some samples listed as invasion of only lamina propria. Although second biopsies to confirm the depth of invasion are common in staging human transitional cell carcinoma (TCC), this was not performed in most dogs, because the type of treatment would not have been altered with the additional information.

chemically induced bladder tumors in rodents [35–37]. Prostaglandin E2 (PGE2), a metabolite of arachidonic acid metabolism, is important in cancer due to its (1) inflammatory effects with recruitment of numerous growth factors into the inflamed environment, (2) potent immunosuppressive effects, (3) effects on angiogenesis, and (4) other actions involving cellular processes [38]. Excessive concentrations of PGE2 have been measured in canine TCC tissue. The mean PGE2 concentrations in samples of normal canine bladder tissue and canine TCC were 62 ng/g (n 5 6) and

Fig. 2. Human (A) and canine (B) transitional cell carcinoma of the urinary bladder, grade 3 (bar 5 22 mm).

1,271 ng/g (n 5 5) of tissue, respectively (D Knapp, S Mohammed, unpublished data). A canine TCC cell line [28] also produced high concentrations of PGE2 (D Knapp, S Mohammed, unpublished data). Similarly, PGE2 concentrations in plasma and in supernatants from stimulated peripheral blood monocytes from dogs with TCC were increased compared with those of normal dogs [34]. Recently, two isoforms of cyclooxygenase (cox-1 and cox-2), the key enzyme in prostaglandin synthesis, have been identified [39]. Recent studies have shown that the normal canine urinary bladder epithelium expresses cox-1, but not cox-2 [40]. In contrast, 21 of 21 canine TCC samples overexpressed cox-2. TCC also expressed cox-1 similar to that on normal canine bladder epithelium [40]. 1.5. Clinical signs, diagnosis and staging, and biological behavior of canine transitional cell carcinoma Canine TCC is typically a disease of older dogs. The mean age at diagnosis of 102 dogs with TCC presented to the PUVTH was 11.05 years. Patronek and Glickman [41] developed a formula to convert chronological years of a dog’s life to human physiologic equivalent years, taking into account the effects of body size and breed on life span

52

D.W. Knapp et al. / Urologic Oncology 5 (2000) 47–59

Table 7 Clinical stage (TNM) of canine bladder cancer [44] T: Primary tumor Tis T0 T1 T2

Carcinoma in situ No evidence of primary tumor Superficial papillary tumor Tumor invading the bladder wall, with induration Tumor invading neighboring organs (prostate, uterus, vagina, and pelvic canal)

T3 N: Regional lymph node (internal and external iliac lymph node) N0 N1 N2 M: Distant metastases M0 M1

No regional lymph node involved Regional lymph node involved Regional lymph node and juxta regional lymph node involved No evidence of metastasis Distant metastasis present

of the dog. Using this formula, the mean physiologic age (6 standard deviation) at diagnosis for the 102 dogs with TCC was 60.04 6 8.71 human equivalent years. The mean body weight in this series of dogs was 15.67 6 10.25 kg (range 3.0–51.0 kg), and the female to male ratio was 1.7:1. Common presenting signs included hematuria, stranguria, and other forms of dysuria, and less commonly lameness, lethargy, and weight loss. Similarly, Norris et al. [42] reported dysuria (84%), grossly visible hematuria (50%), and pollakiuria (37%) as the most frequent clinical signs in 115 dogs with bladder or urethral tumors. The diagnosis of canine bladder cancer requires histopathologic examination of tissues obtained by cystotomy, cystoscopy, or catheter biopsy [43]. Clinical staging of canine bladder cancer is performed with complete physical examination, radiography of the thorax and abdomen, and imaging of the bladder (contrast cystography, ultrasonography, or computed tomography). A TNM classification scheme for canine bladder cancer has been defined by the World Health Organization (WHO; Table 7) [44].

Table 8 TNM stage at diagnosis (n5102) and at death (n580) of TCC in dogs

Tis T0 T1 T2 T3 N0 N1 and N2a M0 M1 a

Number at diagnosis (%)

Number at death (%)

0 (0) 0 (0) 2 (2) 80 (78) 20 (20) 86 (84) 16 (16) 88 (86) 14 (14)

0 (0) 2 (2) 0 (0) 48 (60) 30 (37) 48 (60) 32 (40) 41 (51) 39 (49)

Stages N1 and N2 were combined because imaging procedures and necropsy reports did not always allow differentiation of these stages.

Following the WHO classification scheme, the TNM stage at diagnosis of 102 cases of TCC at the PUVTH was determined (Table 8). Lymph node and distant metastases were present in 16% and 14% of dogs, respectively. At the time of diagnosis, 10% of dogs had both nodal and distant metastases. The TNM stage at the time of death was also available for 80 dogs (Table 8), and 49% of dogs had distant metastases at death. The TNM stage at diagnosis in this series was similar to that reported by Norris et al. except for a small number of dogs with Tis and T1 lesions identified by Norris et al. [42]. Of the 102 dogs evaluated at the PUVTH, the TCC involved the urethra as well as the bladder in 57 dogs (56%). Of 38 male dogs, the prostate was involved in 11 dogs (29%). The majority of the canine bladder tumors were located in the trigone region of the bladder. As shown in Table 8, distant metastases were noted in 14% of dogs at diagnosis, and in 49% of dogs at the time of death. When the primary tumor is not controlled, death due to urinary tract obstruction occurs in many dogs with TCC prior to the development of lethal metastasis. When the primary tumor can be controlled, metastatic disease occurs more frequently. The cause of death, which was known for 85 of the 102 dogs, was the primary tumor in 52 dogs (61%), metastatic disease in 12 dogs (14%), and nontumor related causes in 21 dogs (25%). Fifty of the 102 dogs with TCC examined at the PUVTH underwent postmortem examination. Sites of TCC metastases in these 50 dogs included the lung (14 dogs, 28%), regional lymph nodes (13 dogs, 26%), liver (9 dogs, 18%), kidney (2 dogs, 4%), spleen (2 dogs, 4%), prescapular lymph node (2 dogs, 4%), and uterus (2 dogs, 4%), and one case each (2%) had metastases in mesenteric lymph nodes, cecum, bronchial lymph nodes, vertebrae, ilium, colon, abdominal wall, diaphragm, renal lymph node, and oral mucosa. 1.6. Response to therapy Therapy of canine TCC is usually less aggressive or intense than that used in human TCC. This is due to the fact that most owners will not tolerate signs of moderate or severe toxicity in their dog, especially when the probability of cure is low. 1.6.1 Surgery Canine TCC is difficult to remove surgically due to the trigonal location of the tumor, frequency of urethral involvement, and metastases in 20% or more of dogs at the time of diagnosis. Complete cystectomy, which may be routine in human bladder cancer patients, has not been attempted to any extent in the dog. Conduits with external collection devices and continent reservoirs drained by intermittent catheterization would not be acceptable to many owners. Enterocystoplasty with cystectomy and subtotal intracapsular prostatectomy has been described in male nontumor bearing dogs, but the hourly micturition observed in dogs undergoing this procedure would be unacceptable for

D.W. Knapp et al. / Urologic Oncology 5 (2000) 47–59

most owners of house dogs [45]. Partial cystectomy was reported in dogs with bladder cancer, but local and distant tumor relapse occurred in 8 of 10 dogs [46]. Ureterocolonic anastomosis was reported in 10 dogs with TCC, but complications were frequent and included hyperammonemia, hyperchloremic metabolic acidosis, uremia, and pyelonephritis [47]. Vaginourethroplasty of localized canine urethral tumors [48] and ileocystoplasty of canine bladder tumors [49] have been reported, but these procedures are not useful for tumors involving the trigone region of the bladder, which is the area most frequently affected by canine TCC. For palliation of urinary obstruction and to allow time for other therapies to work, prepubic cystostomy catheters that bypass urethral obstruction have been used in a small number of dogs [50]. When performing tumor “debulking” with or without partial cystectomy, surgical cure is virtually impossible. In a series of 102 dogs with TCC examined at the PUVTH, “complete resection” of the primary tumor (with histopathologically tumor free margins) was accomplished in only two dogs. For these two dogs rendered “tumor free,” extensive local recurrence was noted in one dog 8 months postsurgery. In the second dog, tumor regrowth in the bladder was not detected, but distant relapse occurred within 4 months of surgery. It is not surprising that survival after surgery is short. An analysis of data from the PCOP Tumor Registry [51] revealed a median survival time of only 106 days for 42 dogs with TCC that underwent surgical debulking. Similarly, Norris et al. [42] reported postsurgery median survival of 125 days in 23 dogs with TCC. A point worth noting is that in relapse after surgery, new lesions are frequently observed distant from the original site, raising the question of whether the “field effect” [1] or tumor seeding within the bladder is important in canine bladder cancer. 1.6.2. Radiation therapy Radiation therapy has been used infrequently in canine TCC [52,53]. Walker and Breider [52] reported the use of intraoperative radiation therapy in 13 dogs with bladder tumors including 11 dogs with TCC. Following surgical exposure and partial tumor removal, the remaining bladder was

53

treated with 2,188 to 2888 cGy delivered from a 137Cs teletherapy machine. The median survival time posttherapy was 15 months for dogs with TCC, and recurrence (local or distant) was noted in six dogs. Side effects included pollakiuria, urinary incontinence, cystitis, stranguria, and hydronephrosis; four dogs were euthanatized due to side effects of therapy. 1.6.3. Chemotherapy Spontaneous canine TCC is a good model system for studying a relatively chemotherapy resistant solid tumor. The response of canine TCC to chemotherapy has been similar to that of single agent chemotherapy in human TCC. Table 9 summarizes the response of canine TCC to chemotherapy. Aggressive multiple agent protocols have not been studied in the dog due to the unwillingness of pet owners to accept the toxicity associated with that type of protocol, especially with little chance of cure. 1.6.4. Nonsteroidal antiinflammatory drug therapy Our interest in NSAID therapy began when dogs with various forms of spontaneous cancer had remission while receiving the NSAID piroxicam (Feldene, Pfizer Inc., New York, New York) for pain control, and no other therapy [54]. In a phase I study of piroxicam in 62 dogs with various histopathologically confirmed, measurable tumors, gastrointestinal toxicity was dose-related and dose limiting, but antitumor activity occurred at the lower, less toxic doses [54]. Partial remission (PR; $50% reduction in tumor volume) occurred in 8 dogs including 3 of 10 dogs with TCC. A phase II clinical trial of piroxicam in dogs with histologically confirmed, measurable, nonresectable TCC was performed [34]. As in most clinical trials at the PUVTH, the dogs lived at home with their owners and were evaluated at the PUVTH at monthly intervals. These evaluations included physical examination, radiography of the thorax and abdomen, cystography, complete blood count, serum biochemical profile, and urinalysis. Piroxicam was given orally at a dosage of 0.3 mg/kg q24h, a dosage that had resulted in serum piroxicam concentrations of 15 to 30 mM in the earlier phase I trial [54]. Tumor response in 34 dogs included 2 complete remissions (CR; complete resolution of all clinical

Table 9 Results of chemotherapy of canine TCC Drug

Dosage (mg/m2)

Number treated

Number CR/PR

% CR1PR

Survival (days)

Reference number

Cisplatin

60 50 25–50 300 2.5–5 30 0.5–1.1

25* 15 8 14 6 5* 6*

0/3 0/3 0/1 0/0 0/1 0/1 0/1

12% 20% 25% 0% 17% 20% 17%

130 132 NA 132** NA NA NA

[60] [61] [62] [63] [64] [65] [66]

Carboplatin Mitoxantrone Adriamycin Actinomycin D

* Also includes additional dogs treated by same protocol at the Purdue University Veterinary Teaching Hospital (part of the series of 102 dogs). ** After failing carboplatin therapy, dogs were then treated with piroxicam. CR, complete remission, complete resolution of all radiographic, ultrasonographic and clinical evidence of transitional cell carcinoma (TCC); PR, partial remission, $50% reduction in tumor volume.

54

D.W. Knapp et al. / Urologic Oncology 5 (2000) 47–59

Table 10 Factors associated with TNM stage at diagnosis of canine transitional cell carcinoma TNM stage

Factor

P-value

Number dogs stage N1/total dogs of that T stage (%) 9/82 (11%) 7/20 (35%) Number dogs stage M1/total dogs of that T stage (%) 7/82 (8%) 7/20 (35%) Number male dogs/total of that T stage (% male) 26/82 (32%) 16/20 (60%) Number dogs with prostate involved/total dogs of that T stage (%)** 0/82 (0%) 11/20 (55%) Median age at diagnosis, years/human equivalent years 11.19 years/59 years, n586 9.74 years/55years, n516 Number dogs stage T3/total dogs of that N stage (%) 13/86 (15%) 7/16 (44%) Number dogs stage M1/total dogs of that N stage (%) 5/86 (6%) 9/16 (56%) Number dogs with prostate involved/total dogs of that N stage (%) 7/86 (8%) 4/16 (25%) Number dogs stage T3/total dogs of that M stage (%) 13/88 (15%) 7/14 (50%) Number dogs stage N1/total dogs of that M stage (%) 7/88 (8%) 9/14 (64%) Number dogs with prostate involved/total dogs of that M stage (%) 7/88 (8%) 4/14 (28%)

T1 and T2 T3 T1 and T2 T3 T1 and T2 T3 T1 and T2 T3 N0 N1 and N2 N0 N1 and N2 N0 N1 and N2 N0 N1 and N2 M0 M1 M0 M1 M0 M1

0.015*

0.001*

0.037*

0.001*

0.03***

0.015*

0.0001*

0.068*

0.001*

0.001*

0.042*

* Fisher’s Exact test (two-tail). ** The association between T stage and prostate involvement is expected because tumors with prostatic involvement, by definition, are substage T3. *** Wilcoxan two-sample test.

and radiographic evidence of TCC), 4 PR, 18 stable disease (SD; ,50% change in tumor volume), and 10 progressive disease (PD; $50% increase in tumor volume or the development of new tumor lesions). The two dogs with complete remission lived 2.1 years (7.6 human equivalent years) and 3.3 years (12.2 human equivalent years), respectively, and were tumor free on postmortem examination. Piroxicam therapy was generally well tolerated, with gastrointestinal

Table 11 Association between TNM stage at diagnosis and survival of dogs with transitional cell carcinoma Tumor stage

Median survival No. of dogs (days)

T1 or T2 T3 N0 N1 M0 M1

82 20 86 16 88 14

218 118 234 70 203 105

Wilcoxon two-sample test, P-value 0.0167 0.0001 0.0163

toxicity noted in six dogs and renal papillary necrosis in two dogs. The median survival was 180 days (1.8 human equivalent years). Piroxicam has been used in additional dogs with TCC in our hospital, and currently, tumor responses in 55 dogs have included 2 CR, 7 PR, 32 SD, and 14 PD. The response of canine TCC to piroxicam has led to a phase II clinical trial of piroxicam in humans with carcinoma in situ of the urinary bladder (precursor lesion to invasive bladder cancer [1,55]). 1.7. Prognostic factors Using data from the series of 102 dogs with TCC at the PUVTH, analyses were performed to look for an association between TNM stage at diagnosis, development of metastasis between diagnosis and death, response to systemic therapy, and survival and the following potential prognostic factors at diagnosis: age, body weight, gender, neuter status, TNM stage, urethral involvement, prostate involvement, histopathologic characteristics (listed in Table 6), and type of therapy. The associations that were significant or ap-

D.W. Knapp et al. / Urologic Oncology 5 (2000) 47–59

proached significance are listed in Tables 10 and 11. A higher T stage at diagnosis was associated with advanced N and M stages, male gender, and prostatic involvement. The association between T stage and prostate involvement was expected because tumors with prostatic involvement are, by definition, substage T3. Factors associated with higher N stage were younger age at diagnosis, higher T and M stages, and prostatic involvement (prostatic involvement approached significance). Factors associated with M1 stage (vs. M0 stage) were higher T and M stages and prostatic involvement. The TNM stage at both diagnosis and death was known in 80 of the 102 dogs with TCC treated at the PUVTH. Of 64 dogs that had stage N0 tumors at diagnosis, 16 dogs (25%) developed lymph node metastasis between diagnosis and death. Of 66 dogs with stage M0 tumors at diagnosis, distant metastasis (M1) developed in 25 dogs (38%) between diagnosis and death. In addition, 34 (57%) of 60 dogs with stage N0,M0 tumors at diagnosis progressed to stage N1/2 or M1 at death. Two factors associated with the development of metastasis between diagnosis and death were vascular invasion and urethral involvement of the tumor. Vascular invasion was present in 5 (42%) of 12 biopsies from dogs that developed lymph node metastasis and in 5 (10%) of 48 biopsies from dogs that did not develop lymph node metastasis (Fisher’s exact test, two-tail, P , 0.021). Vascular invasion approached significance in association with the development of distant metastasis. Vascular invasion was present in 4 (10%) of 40 biopsies from dogs with development of distant metastases and in 6 (32%) of 19 biopsies from dogs that did develop distant metastases (Fisher’s exact test two-tail, P , 0.062). The presence of urethral involvement of the TCC was associated with the development of distant metastases, with urethral involvement in 19 (76%) of 25 dogs and in 27 (50%) of 54 dogs that did and did not develop distant metastasis, respectively (Fisher’s exact test two-tail, P , 0.049). The response to chemotherapeutic agents and piroxicam were similar. Two factors—T stage at diagnosis and histologic classification—were associated with the response (remission vs. no remission) with chemotherapy or piroxicam therapy. Remission occurred in 22 (30%) of 73 dogs with tumors stage T1/T2 and in 1 (5%) of 19 dogs with stage T3 tumors at diagnosis (Fisher’s exact test two-tail, P , 0.035). Remission was noted in 14 (27%) of 51 dogs with TCC with no glandular differentiation, and in 0 of 12 dogs with TCC with glandular differentiation (Fisher’s exact test two-tail, P , 0.05). Shorter survival was significantly associated with more advanced TNM stage at diagnosis (Table 11). There was a negative association between survival and prostate involvement, and a positive association (which approached significance) between survival and debulking surgery. The medium survival time was 211 days for 91 dogs without prostatic involvement, compared with 124 days for 11 dogs with prostatic involvement (Wilcoxan two-sample test, P , 0.063). Although surgery is rarely curative for TCC, the me-

55

dian survival for 25 dogs undergoing debulking surgery plus medical therapy (chemotherapy or piroxicam) was 272 days, compared with 195 for 42 dogs undergoing surgery for biopsy only plus medical therapy, and 150 days for 36 dogs receiving medical therapy only (chi square, P , 0.0631). 2. Comparison of canine and human invasive transitional cell carcinoma The similarities and differences between canine and human invasive TCC are summarized in Table 12. As demonstrated in Table 12, canine and human invasive TCC are extremely similar. One of the few differences between canine and human TCC is the difference in gender predilection, with men being at greater risk that women, and female dogs being at greater risk than male dogs. The reason for this difference is not known, but could involve several factors. Smoking and occupational exposures are thought to be responsible for much of the increased risk of bladder cancer in men [1], and these risks would generally not be relevant to canine TCC. Male dogs urinate more frequently for territorial marking than do female dogs; this could result in less exposure time of the bladder epithelium of male dogs (who live outdoors) to carcinogens in urine. In addition, female dogs have increased body fat compared with male dogs, and therefore increased storage of lipophilic environmental carcinogens. The increased risk of TCC in neutered dogs of both genders is interesting, but is not explained at this time. Similarly, it is now known why the location of the tumor within the bladder differs between dogs and humans.

3. Discussion and future research There are great similarities between canine and human invasive TCC. Spontaneous canine TCC provides a model that complements existing rodent models. Rodent models are well suited to the study of tumor initiation and promotion in response to chemical carcinogens and of interventional strategies aimed at these relatively early processes [6]. With the current knowledge of canine TCC, it is a model better suited for the study of later events in TCC progression such as drug resistance and metastasis. In addition, spontaneous canine TCC arises in a genetically diverse, aged (but still immunocompetent) population that is similar to that in humans. On a practical level, widespread availability of subjects and lower expense are important features of rodent models. Although this review focuses on invasive bladder cancer, it should be noted that superficial bladder cancer in the dog is infrequent and has not been studied. Superficial bladder cancer comprises 80% of human bladder cancers [1], but it rarely occurs in the dog. Different pathways (different genetic “hits”) are involved in the development of superficial and invasive bladder cancer in humans [1,56]. It appears likely that the pathways necessary for superficial bladder

56

D.W. Knapp et al. / Urologic Oncology 5 (2000) 47–59

Table 12 Similarities and differences between canine and human invasive TCC

Similarities % of all cancers Change in occurrence Age at diagnosis Risk factors Subpopulations at risk Environment Benzene Cyclophosphamide Histopathology Other cellular features DNA ploidy Tumor assoc. glycoprotein Urine bFGF concentration P53 (mutated) expression* Cyclooxygenase expression Clinical signs Metastasis at diagnosis Sites of metastasis Response to single agent chemotherapy*** Cisplatin Carboplatin Prognostic factors TNM stage Prostate involvement Differences Male:female ratio Tumor location within the bladder

Canine TCC

Human TCC

1.5–2% Increase university hospital prevalence 11 yr (60 human equiv. yr [41])

2% 0.7% increase annual incidence [67] 65 yr [1]

Breeds (e.g., Scottish Terrier) Increased risk in urban area [17] Increased risk with exposure to insecticides containing benzene and other “inert” ingredients [14] TCC reported in small number of dogs receiving cyclophosphaide for other malignancies [15,16] Invasive TCC of intermediate to high grade (>90% grades 2 and 3)

Races (e.g., African Americans) [1] Increased risk in urban area [1] Increased risk with exposure to benzene and polycyclic aromatic hydrocarbons [1] Increased risk with exposure [22]

79% aneuploid [25]

Invasive TCC of intermediate to high grade (70% grades 2 and 3) [1,68]

Immunoreactivity to TAG-72 antibodies [25] Increased in TCC [32] p53 detected in small number of canine TCC tissue samples and in TCC cell line Cox-2 overexpressed, increased PGE2 concentrations in tumor tissue and in plasma [34] Hematuria, dysuria, urinary tract infection most common; bone pain infrequent 20% of dogs** Regional nodes and lung most common

Aneuploidy correlates with advanced stage and grade [69] Immunoreactivity TAG-72 antibodies [27] Increased in TCC [30,31] p53 expression correlates with advanced stage and grade and poor survival [70–72] Cox-2 overexpressed in tumor tissue in preliminary report [73] Hematuria, urinary tract infection most common; bone pain less common [1] 5–20% of patients [1] Regional nodes and lung most common [1]

12–20% [60,61] <10%****

17–34% [1] 15% [1]

Advanced TNM stage associated with decreased survival Associated with distant metastasis at diagnosis and with decreased survival

Advanced TNM stage associated with decreased survival [1,74] Associated with decreased survival, especially when stroma of gland is involved [75]

0.5:1 Majority trigonal

2.8:1 [1] Lateral wall (37%), posterior wall (18%), trigone and neck (23%), other sites (22%) [76]

* When mutated, p53 has a prolonged half-life, which allows it to be detected by immunohistochemistry (IHC) or Western blotting. It is estimated that IHC is 90% accurate in detecting p53 mutations in human transitional cell carcinoma (TCC). [70] ** Of the 102 dogs with TCC evaluated at the Purdue University Veterinary Teaching Hospital, 20% had metastasis at diagnosis. An earlier study reports 37% of dogs had metastasis at diagnosis [42]. *** Comparison to multiagent protocols is not possible because these have not been used in the dog due to toxicity. **** In a phase II clinical trial of carboplatin in dogs with TCC, remission did not occur in any dogs [63]. Anecdotal reports of response of canine TCC to carboplatin have been made.

cancer development are rarely present or are not functional in the dog. It is unlikely that superficial bladder cancer occurs in the dog, but goes undetected. Animals models are needed to help answer several key questions with respect to the invasive bladder cancer field including: (1) How can bladder cancer be prevented? (2) Why do people with no known risk factors develop TCC, or in other words, what are the yet to be identified risk factors? (3) How can progression of carcinoma in situ to invasive TCC be prevented? (4) How can metastasis of TCC be prevented? (5) How can existing metastatic TCC be treated more effectively? Canine bladder cancer provides a useful model to help answer many of these important questions.

Canine bladder cancer may be helpful in the search for previously unidentified risk factors for human bladder cancer. The case control study demonstrating risk of obesity and flea control products [14] is an example of how a study in dogs suggests a causal mechanism for a human cancer that can now be tested. Such epidemiologic findings in dogs point to the need to measure the concentration of the “inert” substances in the flea control product in adipose tissue of people with bladder cancer, particularly nonoccupationally exposed, nonsmoking patients. In addition, the people who applied the flea dips to the dogs in this study were heavily exposed in the process. In follow-up contact with all the individuals, some groomers and veterinary technicians re-

D.W. Knapp et al. / Urologic Oncology 5 (2000) 47–59

ported applying more than 100 dips per week over many years with little or no protective clothing. An occupational cohort study of veterinary technicians and groomers should be considered. Another area in which dogs may help identify risk factors for bladder cancer relates to breed predisposition for TCC. Genetic factors associated with TCC are important because many human TCC patients have no known exposure to environmental risk factors for TCC (i.e., no exposure to cigarette smoke, occupational exposure, etc.). Studies of breeds at high risk for bladder cancer may help to identify genetic factors important in bladder cancer initiation and development. This work is especially timely given current efforts to map the canine and human genomes [57,58]. Other questions the canine bladder cancer model could help answer relate to how to more successfully treat, or ideally prevent, metastatic TCC. In the series of 102 dogs with TCC, distant metastasis was present in 14% of dogs at diagnosis and 49% of dogs at death. Furthermore, as more effective control of the primary tumor(s) occurs and urinary tract obstruction is averted, a higher percentage of dogs will ultimately develop metastasis, providing ideal subjects for studies of antimetastatic therapy. The study of NSAID therapy in canine TCC is a good example of the type of study that can be conducted in the canine model system. The phase II trial of piroxicam in dogs with TCC [34] was very well accepted by pet owners and veterinarians, was successfully completed in a timely fashion, and demonstrated the antitumor activity of NSAIDs against invasive bladder cancer. In addition, preliminary results of a follow-up study suggest much greater antitumor effects when piroxicam is combined with chemotherapy [59]. The mechanisms of NSAID antitumor activity can be studied in the canine TCC model system. The studies of NSAIDs in dogs with TCC have identified a potentially important bladder cancer therapy for evaluation in human clinical trials. Success of NSAIDs in human trials will, among other things, provide further confirmation of the utility of the canine TCC model. Further characterization of canine TCC should include molecular features such as expression of tumor suppressor genes and oncogenes, angiogenic and anti-angiogenic factors, and cyclooxygenase. Carcinoma in situ should be studied further in the dog because it is an important precursor lesion to human invasive bladder cancer [1,55]. Carcinoma in situ is found (adjacent to or remote from the primary tumor) in over 50% of human bladders with multiple papillary tumors, and carcinoma in situ alone without concurrent exophytic tumor constitutes 1 to 2% of newly detected cases of human bladder cancer [1]. In a series of canine bladder tumor cases reported by Valli et al. [24], carcinoma in situ was infrequently observed in the dog. This is in lieu of the fact that histopathologic examination of the majority of canine bladder cancer cases studied to date frequently has been limited to grossly visible mass lesions in the bladder. Examination of the complete urinary bladder from dogs

57

with TCC and dogs at risk for TCC (such as elderly Scottish Terriers) is indicated to further characterize carcinoma in situ. To fully utilize spontaneous cancer in dogs as model for human cancer, veterinary and comparative oncologists must develop strategies to enroll larger number of pet dogs into clinical studies. With 54 million pet dogs in the United States receiving routine veterinary care, and cancer being the major cause of death in older dogs, the cancer-bearing pet dog population is a vast resource that has barely been tapped [9]. For example, at the PUVTH, approximately 12 to 15 dogs with TCC have entered clinical trials each year for the last 5 years. Although these numbers of dogs with TCC have allowed completion of important studies, identification of larger numbers of dogs with TCC in a timely fashion is needed for more widespread availability of this bladder cancer model. Fortunately, the demand for advanced care of cancer and other serious illnesses in pet dogs is rapidly growing in the United States. Identifying and enrolling larger numbers of dogs with TCC in clinical studies will involve continued close collaborative relationships between centers performing the clinical trials, primary care veterinarians, and veterinary specialists in private and corporate practices. Veterinary specialty practices have been established in most major cities in the United States and constitute the most rapidly growing sector of specialized veterinary care. Along with identifying dogs for clinical studies, financial support for the expenses incurred in these trials must be addressed. Pet health insurance is becoming more common in the United States, but the vast majority of dogs are not currently covered by an insurance plan. Although some pet owners have the financial resources to pursue costly evaluation and treatment of their pet dogs, many pet owners cannot afford cancer therapy. For example, tests to obtain a diagnosis and initial staging alone, without treatment, can cost $500 to $1,000 at most university and private or corporate referral veterinary hospitals in the United States. To accrue tumor-bearing dogs for studies in a timely fashion, the cost of cancer staging and treatment should be covered by the researchers through grants. Compared with human cancer trials, the costs of cancer trials in animals are modest, yet these trials can yield extremely valuable comparative information regarding human cancer etiology, treatment, and prevention. In conclusion, spontaneous canine TCC is very similar to human TCC in histopathologic characteristics, molecular features, biological behavior, response to medical treatment, and prognosis. While further characterization of canine TCC is continuing, the information currently available strongly supports the use of dogs with TCC as a useful and relevant model of human invasive bladder cancer. References [1] Scher HI, Shipley WU, Herr HW. Cancers of the genitourinary system. Biology of genitourinary cancers. Cancer of the bladder. In: De

58

[2]

[3]

[4]

[5]

[6] [7]

[8]

[9]

[10] [11]

[12]

[13] [14]

[15]

[16]

[17] [18]

[19]

[20]

[21]

[22]

D.W. Knapp et al. / Urologic Oncology 5 (2000) 47–59 Vita VT, Hellman S, Rosenberg SA, editors. Principles and Practice of Oncology, 5th Ed. Philadelphia: Lippincott-Raven, 1997. pp. 1300–22. Mitchell SS, Satterwhite DJ, Moses HL. Molecular insights into altered cell cycle regulation and genitourinary malignancy. Urol Oncol 1995;1:3–17. Becci PJ, Thompson HJ, Strum JM, Brown CC, Sporn MB, Moon RC. N-Butyl-N-(4-hydroxybutyl)nitrosamine-induced urinary bladder cancer in C57BL/6 X DBA/2 F1 mice as a useful model for study of chemoprevention of cancer with retinoids. Cancer Res 1981;41: 927–32. McCormick DL, Roman SS, Becci PJ, Moon RC. Influence of total dose and dose schedule on induction of urinary bladder cancer in the mouse by N-butyl-N-(4-hydroxybutyl)nitrosamine. Carcinogenesis 1981;2:251–54. Moon RC, Kelloff GJ, Detrisac CJ, Steele VE, Thomas CF, Sigman CC. Chemoprevention of OH-BBN-induced bladder cancer in mice by piroxicam. Carcinogenesis 1993;14:1487–9. Oyasu R. Epithelial tumours of the lower urinary tract in humans and rodents. Food Chem Toxicol 1995;33:747–55. Dinney CPN, Fishbeck R, Singh RK, et al. Isolation and characterization of metastatic variants from human transitional cell carcinoma passaged by orthotopic implantation in athymic nude mice. J Urol 1995;154:1532–8. Chang SG, Kim JI, Jung JC, et al. Antimetastatic activity of the new platinum analog {Pt(cis-dach) (DPPE) ? 2NO3} in a metastatic model of human bladder cancer. Anticancer Research 1997;17:3239–42. Knapp DW, Waters DJ. Naturally occurring cancer in pet dogs: important models for developing improved cancer therapy for humans. Mol Med Today 1997;3:8–11. Priester WA, McKay FW. The occurrence of tumors in domestic animals. Monogr Natl Cancer Inst 1980;54:1–10. Withrow SJ. Tumors of the urinary system. In: Withrow SJ, MacEwan EG, editors. Small Animal Clinical Oncology, 2nd ed. Philadelphia: WB Saunders, 1996. pp. 380–92. Priester WA, McKay FW. The Occurrence of Tumors in Domestic Animals. National Cancer Institute Monograph 54. Prepared by Epizoology Section, Clinical Epidemiology Branch, Field Studies and Statistics Program, Division of Cancer Cause and Prevention. Bethesda, National Cancer Institute, 1986. Bronson RT. Variation in age at death of dogs of different sexes and breeds. Am J Vet Res 1982;43:2057–9. Glickman LT, Schofer FS, McKee LJ. Epidemiology study of insecticide exposures, obesity, and risk of bladder cancer in household dogs. J Toxicol Environ Health 1989;28:407–14. Weller RE, Wold AM, Dyjido A. Transitional cell carcinoma of the bladder associated with cyclophosphamide therapy in a dog. J Amer Anim Hosp Assoc 1979;5:733–6. Macy DW, Withrow SJ, Hoopes J. Transitional cell carcinoma of the bladder associated with cyclophosphamide administration. J Amer Anim Hosp Assoc 1983;19:965–9. Hayes HM, Jr, Hoover R, Tarone R. Bladder cancer in pet dogs: a sentinel for environmental cancer? J Epidemiol 1981;114:229–33. Okajima E, Hiramatsu T, Hirao K, et al. Urinary bladder tumors induced by n-butyl-n(4-hydroxybutyl) nitrosamine in dogs. Cancer Res 1981;41:1958–66. Steineck G, Plato N, Gerhardsson M, Norell SE, Hogstedt C. Increased risk of urothelial cancer in Stockholm during 1983–87 after exposure to benzene and exhausts. Int J Cancer 1990;45:1012–17. IARC. Overall evaluations of carcinogenicity: an updating of IARC monographs, volumes 1 to 42. In: IARC Monogr Eval Carcinog Risks Hum, Suppl. 7:1–440. Lyon: IARC, 1987. Hartage P, Silverman D, Hoover R, et al. Changing cigarette habits and bladder cancer risk: a case-control study. J Natl Cancer Inst 1987; 78:1119–25. Levine LA, Richie JP. Urologic complications of cyclophosphamide. J Urol 1989;141:1063–9.

[23] Mostofi K, Ito N, Weinstein R. Pathology: the need for standardization of pathological examination and reporting. In: Developments in Bladder Cancer. New York: Alan R. Liss, Inc., 1986. pp. 67–83. [24] Valli VE, Norris A, Jacobs RM, et al. Pathology of canine bladder and urethral cancer and correlation with tumour progression and survival. J Comp Pathol 1995;113:113–30. [25] Clemo FAS, DeNicola DB, Carlton WW, Walker E, Morrison WB. Immunoreactivity of canine transitional cell carcinoma of the urinary bladder with monoclonal antibodies to tumor-associated glycoprotein 72. Vet Pathol 1995;32:155–61. [26] Clemo FAS, DeNicola DB, Delaney LJ. Immunohistochemical evaluation of canine carcinomas with monoclonal antibody B72.3. Vet Pathol 1993;30:140–5. [27] Molinolo A, Simpson JF, Thor A, Schlom J. Enhanced tumor binding using immunohistochemical analyses by second generation anti-tumorassociated glycoprotein 72 monoclonal antibodies versus monoclonal antibody B72.3 in human tissue. Cancer Res 1990;50:1291–8. [28] Knapp DW, Chan TCK, Kuczek T, Reagen WJ, Park B. Evaluation of in vitro cytotoxicity of nonsteroidal anti-inflammatory drugs against canine tumor cells. Am J Vet Res 1995;56:801–5. [29] Gamblin RM, Sagartz, JE, Couto CG. Overexpression of p53 tumor suppressor protein in spontaneously arising neoplasms of dogs. Amer J Vet Res 1997;58:857–63. [30] Chodak GW, Hospelhorn V, Judge SM, Mayforth R, Koeppen H, Sasse J. Increased levels of fibroblast growth factor-like activity in urine from patients with bladder or kidney cancer. Cancer Res 1988; 48:2083–8. [31] Nguyen M, Watanabe H, Budson AE, Richie JP, Hayes DF, Folkman J. Elevated levels of an angiogenic peptide, basic fibroblast growth factor, in the urine of patients with a wide spectrum of cancers. J Natl Cancer Inst 1994;86:356–61. [32] Allen DK, Waters DJ, Knapp DW, Kuczek T. High urine concentrations of basic fibroblast growth factor in dogs with bladder cancer. J Vet Intern Med 1996;10:231–4. [33] Campbell SC, Volpert OV, Ivanovich M, Bouck NP. Molecular mediators of angiogenesis in bladder cancer. Cancer Res 1998;58:1298– 304. [34] Knapp DW, Richardson RC, Chan TCK, et al. Piroxicam therapy in 34 dogs with transitional cell carcinoma of the urinary bladder. J Vet Intern Med 1994;8:273–8. [35] Shibata MA, Hasegawa R, Shirai T, Takesada Y, Fukushima S. Chemoprevention by indomethacin of tumor promotion in a rat urinary bladder carcinogenesis model. Int J Cancer 1993;55:1011–7. [36] Cohen SM, Zenser TV, Murasaki G, et al. Aspirin inhibition of N-[4(5-Nitro-2furyl)-2-thiazolyl] formamide-induced lesions of the urinary bladder correlated with inhibition of metabolism by bladder prostaglandin endoperoxide synthetase. Cancer Res 1981;41:3355–9. [37] Rao KV, Detrisac CJ, Steele VE, Hawk ET, Kelloff GJ, McCormick DL. Differential activity of aspirin, ketoprofen and sulindac as cancer chemopreventive agents in the mouse urinary bladder. Carcinogensis 1996;17:1435–8. [38] Marnett LJ. Aspirin and the potential role of prostaglandins in colon cancer. Cancer Res 1992;52:5575–89. [39] Yokoyama C, Tanabe T. Cloning of human gene encoding prostaglandin endoperoxide synthase and primary structure of the enzyme. Biochem Biophys Res Commun 1989;165:888–94. [40] Mohammed SI, Khan KN, Koki AT, et al. Expression of cyclooxygenase-2 in invasive transitional cell carcinoma of the urinary bladder in humans and dogs [abstract]. Proc Amer Assoc Cancer Res 1999;40. Abstract 1576. [41] Patronek GJ, Waters DJ, Glickman LT. Comparative longevity of pet dogs and humans: implications for gerontology research. J Gerontol A Biol Sci Med Sci 1997;52:B171–8. [42] Norris AM, Laing EJ, Valli VE, et al. Canine bladder and urethral tumors: a retrospective study of 115 cases (1980–1985). J Vet Intern Med 1992;6:145–53. [43] Knapp DW. Medical therapy of canine transitional cell carcinoma of

D.W. Knapp et al. / Urologic Oncology 5 (2000) 47–59

[44] [45]

[46]

[47]

[48] [49]

[50]

[51]

[52] [53]

[54]

[55] [56] [57]

[58] [59]

[60]

the urinary bladder. In: Bonagura JD, editor. Kirk’s Current Veterinary Therapy XII. Philadelphia: WB Saunders, 1995. pp. 1016–8. Owen LN. TNM Classification of Tumours in Domestic Animals, 1st ed. Geneva: World Health Organization, 1980. Fries CL, Binnington AG, Valli VE, Connolly JG, Holmberg DL, Pennock P. Enterocystoplasty with cystectomy and subtotal intracapsular prostatectomy in the male dog. Vet Surg 1991;20:104–12. Stone EA, George TF, Gilson SD, Page RL. Partial cystectomy for urinary bladder neoplasia: surgical technique and outcome in 11 dogs. J Small Anim Pract 1996;37:480–5. Stone EA, Withrow SJ, Page RL, Schwarz PD, Wheeler SL, Seim HB. Ureterocolonic anastomosis in ten dogs with transitional cell carcinoma. Vet Surg 1988;17:147–53. White RN, Davies JV, Gregory SP. Vaginourethroplasty for treatment of urethral obstruction in the bitch. Vet Surg 1996;25:503–10. Schwartz PD, Egger EL, Klause SE. Modified “cup-patch” ileocystoplasty for urinary bladder reconstruction in a dog. J Am Vet Med Assoc 1991;198:273–7. Smith JD, Stone EA, Gilson SD. Placement of a permanent cystostomy catheter to relieve urine outflow obstruction in dogs with transitional cell carcinoma. J Am Vet Med Assoc 1995;206:496–9. Lengerich EJ, Teclaw RF, Mendlein M, Mariolis P, Garbe PL. Pet populations in the catchment area of the Purdue Comparative Oncology Program. J Am Vet Med Assoc 1992;200:51–6. Walker M, Breider M. Intraoperative radiotherapy of canine bladder cancer. Vet Radiol 1987;28:200–4. McCaw DL, Lattimer JC. Radiation and cisplatin for treatment of canine urinary bladder carcinoma, a report of two case histories. Vet Radiol 1988;29:264–8. Knapp DW, Richardson RC, Bottoms GD, Teclaw R, Chan TCK. Phase I trial of piroxicam in 62 dogs bearing naturally occurring tumors. Cancer Chemother Pharmacol 1992;29:214–8. Husdon MA, Herr HW. Carcinoma in situ of the bladder. J Urol 1995;153:564–72. Saran KK, Gould D, Godec CJ, Verma RS. Genetics of bladder cancer. J Mol Med 1996;74:441–5. Ostrander EA, Giniger E. Insights from model systems. Semper fidelis: what man’s best friend can teach us about human biology and disease. Am J Hum Genet 1997;61:475–80. Uddhav K, Ketan S. Advances in the human genome project. A review. Molec Biol Rept 1998;25:27–43. Knapp DW, DeNicola DB, Widmer WR, et al. In vivo synergistic activity between cisplatin and piroxicam in canine transitional cell carcinoma of the urinary bladder, a model of human invasive bladder cancer. Proc Annu Meet Am Assoc Cancer Research 1997;38:430. Chun R, Knapp DW, Widmer WR, Glickman NW, DeNicola DB, Bonney P. Cisplatin treatment of transitional cell carcinoma of the urinary bladder in dogs: 18 cases (1983–1993). J Am Vet Med Assoc 1996;209:1588–91.

59

[61] Moore AS, Cardona A, Shapiro W, Madewell BR. Cisplatin (cisdiamminedichloroplatinum) for treatment of transitional cell carcinoma of the urinary bladder or urethra. A retrospective study of 15 dogs. J Vet Intern Med 1990;4:148–52. [62] Shapiro W, Kitchell BE, Fossum TN, et al. Cisplatin for treatment of transitional cell carcinoma and squamous cell carcinomas in dogs. J Am Vet Med Assoc 1988;193:1530–3. [63] Chun R, Knapp DW, Widmer WR, et al. Phase II clinical trial of carboplatin in canine transitional cell carcinoma of the urinary bladder. J Vet Intern Med 1997;11:279–83. [64] Ogilvie GK, Obradovich JE, Elmslie RE, et al. Efficacy of mitroxantrone against various neoplasms in dogs. J Am Vet Med Assoc 1991;198:1618–21. [65] Ogilvie GK, Reynolds HA, Richardson RC, et al. Phase II evaluation of doxorubicin treatment of various canine neoplasms. J Am Vet Med Assoc 1989;195:1580–3. [66] Hammer AS, Couto CG, Ayl RD, Shank KA. Treatment of tumorbearing dogs with actinomycin D. J Vet Intern Med 1994;8:236–9. [67] Hankes BF, Silverman DE. SEER Cancer Statistics Review, 1973– 1990. NIH Pub No. 93-2789. Bethesda: National Cancer Institute, 1993. [68] Cummings KB. Uroepithelial tumors. In: Moossa AR, Robson MC, Schimpff SC, editors. Comprehensive Textbook of Oncology. Baltimore: Williams and Wilkins, 1991. pp. 1067–87. [69] Blomjous CEM, Schipper NW, Baak JPA, van Galen EM, de Voogt HJ, Meyer CJ. Retrospective study of prognostic importance of DNA flow cytometry on urinary bladder carcinoma. J Clin Pathol 1988;41: 21–5. [70] Linehan WM, Cordon-Cardo C, Isaacs W. Cancers of the genitourinary system. Molecular biology of genitourinary cancers. In: DeVita VT, Hellman S, Rosenberg SA, editors. Principles and Practice of Oncology, 5th Ed. Philadelphia: Lippincott-Raven, 1997. pp. 1253–70. [71] Esrig D, Elmajian D, Groshen S, et al. Accumulation of nuclear p53 and tumor progression in bladder cancer. N Engl J Med 1994;331: 1259–64. [72] Esrig D, Spruck CH, Nichols PW, et al. p53 Nuclear protein accumulation correlates with mutations in the p53 gene, tumor grade, and stage in bladder cancer. Am J Pathol 1993;143:1389–97. [73] Koki AT, Edwards DA, Dannenberg AJ, et al. Cyclooxygenase-2 is expressed in animals and human cancers. Proc Annu Meet Am Assoc Cancer Research 1998;39:150. [74] Whitmore WF. Management of invasive bladder neoplasms. Semin Urol 1983;1:34–41. [75] Wood DP, Jr, Montie JE, Pontes JE, Levin HS. Identification of transitional cell carcinoma of the prostate in bladder cancer patients: a prospective study. J Urol 1989;142:83–5. [76] Stephenson WT, Holmes FF, Noble MJ, Gerald KB. Analysis of bladder carcinoma by subsite. Cystoscopic location may have prognostic value. Cancer 1990;66:1630–5.