Abstract
A search from databases of four veterinary colleges and one private referral practice between January 1992 and April 1998 provided 20 cases diagnosed with primary renal neoplasia. Review of these cases revealed 19 primary renal tumours, excluding lymphoma. Of the 20 histologically reviewed cases, the diagnosis was amended in eight. There were 13 renal carcinomas (11 tubular and two tubulopapillary), three transitional cell carcinomas, one malignant nephroblastoma, one haemangiosarcoma and one adenoma. The haemangiosarcoma is, to our knowledge, the first reported case of this tumour type as a primary renal tumour in the cat. Most cats were presented for non-specific clinical signs such as anorexia and weight loss. One cat presented with tumour-associated polycythaemia which has not, to our knowledge, been reported previously. The metastatic rate for cats with complete staging was 64%, and 100% for transitional cell carcinomas.
Lymphoma is the most common renal tumour in cats, but is more often a metastatic than a primary renal neoplasm Hammer & LaRue (1995), Mooney et al (1987), Osborne et al (1971), Cotchin (1956). Reports of other primary feline renal tumours are rare and generally limited to case reports and surveys of feline tumours. One book chapter describes 19 primary renal tumours found in cats at the Angell Memorial Animal Hospital, with carcinoma being the most frequent diagnosis Carpenter et al (1987). As in the series of Carpenter et al, epithelial tumours (including renal adenocarcinoma, transitional cell (TCC), clear cell carcinoma, and squamous cell carcinoma) have been reported most commonly in the literature. Britt et al (1985), Caywood et al (1980), Madewell & Theilen (1987), Hammer & LaRue (1995), Steinberg & Thomson (1994), Martin de las Mulas, et al (1995), Yamazoe et al (1994). Other primary renal tumours in cats include nephroblastoma Cotchin (1956), Potkay & Garman (1969), Caywood et al (1980), Carpenter et al (1987), adenoma Flir (1952/53), Priester & McKay (1980), Nafe et al (1981), Carpenter et al (1987), Clark & Wilson (1988), sarcoma (including leiomyosarcoma) Lombard (1940), Priester & McKay (1980), Caywood et al (1980), Cotchin (1956), Carpenter et al (1987), Fitts (1960), leiomyoma Priester & McKay (1980) and carcinosarcoma Whitehead (1967). The purpose of this retrospective study was to characterise primary renal tumours, excluding lymphoma, in cats and to describe the signalment, clinical signs, and biological behaviour associated with these tumours.
Materials and methods
The databases of four veterinary college diagnostic laboratories and/or teaching hospitals (University of Missouri, Auburn University, University of Wisconsin-Madison, and Iowa State University) were searched to identify cases of primary renal tumours diagnosed in cats between January 1992 and April 1998. An additional case was submitted from a private referral practice. Diagnoses were confirmed by evaluation of histologic specimens by one pathologist (SET). When available, necropsy specimens were also reviewed histologically.
Tubular and tubulopapillary carcinomas were assigned histological grades according to several criteria including mitotic rate per 10 high power fields (0 = 0 mitotic figures; 1 = <10; 2 = 10–20; 3 = >20); nuclear grade (1 = uniform nuclei with inconspicuous or absent nucleoli; 2 = larger irregular nuclei with nucleoli evident at 400 ×; 3 = large nuclei with very irregular margins and prominent large nucleoli evident at 100 ×; 4 = similar to 3 except nuclei more bizarre with prominent, multiple, variably sized nucleoli and heavy, coarse chromatin); tissue invasion (1 = yes, 2 = no); vascular invasion (1 = yes, 2 = no); scirrhous reaction (0 = none; 1 = <20%; 2 = 20–50%; 3 = >50%); tumour necrosis (0 = none; 1 = <20%; 2 = 20–50%; 3 = >50%); and inflammation (0 = none; 1 = mild; 2-moderate; 3 = marked). Based on a total score, the tumours were graded as I—low (0–6), II—intermediate (7–12) and III—high (>12). The nuclear grading system was a modification of previously reported grading systems Skinner et al (1971) and Fuhrman et al (1982)
For all confirmed cases, data regarding signalment, presenting complaint, tumour staging, bloodwork and urinalysis abnormalities, and case outcome were collected by review of medical records and telephone interview of referring veterinarians, when applicable. The possibility of a sex-based predilection was examined (assuming an equal sex distribution in the general cat population) by calculating the binomial probability of the observed result Daniel (1999).
Results
Case records and histopathology specimens were reviewed for 20 cats. A diagnosis of primary renal tumour (excluding lymphoma) was confirmed in 19 cats. Data are summarised in Table 1. The age of cats ranged from 6 to 16 years (mean 11 years; median 11 years). Twelve of the cats were male, with eight castrated and four intact. All female cats were spayed. The probability of seven or fewer of the 19 cats being female calculated using the binomial distribution was not statistically significant (P = 0.1796). Breeds represented included domestic short hair (n = 12), domestic long hair (n = 2), Himalayan (n = 3), Persian (n = 1) and Manx (n = 1). Tumours occurred unilaterally in 18 of 19 cats.
Presenting clinical signs included anorexia/weight loss (n = 9), depression/lethargy (n = 3), neurological signs (n = 2), haematuria (n = 2), abdominal distension (n = 1), abdominal pain (n = 1), blindness (n = 1), dyspneoa (n = 1) and acute death (n = 1). Three cats had no associated clinical signs and tumours were incidental findings.
Serum biochemical profile results were available for 10 cats. Abnormalities included azotaemia (n = 8; three of which had increases in creatinine only), hypophosphataemia (n = 2), hyperphosphataemia (n = 1), hyperkalaemia (n = 2), hypoalbuminaemia (n = 2; associated with panhypoproteinaemia in one), hypoglycaemia (n = 1), hypernatraemia (n = 2), hypocalcaemia (n = 2), hypercholesterolaemia (n = 1), and increased creatine phosphokinase (CPK) activity (n = 2). Urinalysis from seven of the azotaemic cats revealed specific gravities ranging from 1.010 to 1.042, with four < 1.035. Haematuria (6/9) and proteinuria (4/9) were also noted. Complete blood counts (CBC) were available for 11 cats. Abnormalities included anaemia (5/11) and polycythaemia (1/11). The polycythaemia occurred in a cat with a history of seizures that resolved following nephrectomy of a low-grade carcinoma.
Initial diagnoses included carcinoma (n = 12), TCC (n = 3), adenoma (n = 4) and haemangiosarcoma (n = 1). Eight tumours were reclassified after histopathologic review. Three carcinomas were reclassified as TCC, nephroblastoma and lymphoma, respectively, and the lymphoma case was dropped from the study. Two adenomas were determined to be carcinomas after histopathological review and a third was reclassified as a TCC. Two tumours that were originally diagnosed as TCCs were later reclassified as high-grade carcinomas.
The tubular (n = 11) and tubulopapillary (n = 2) carcinomas were the most abundant and were graded as described previously. There were three grade I tumours, two grade II tumours, seven grade III tumours and one with too little tissue to determine tumour grade among the tubular and tubulopapillary tumours. The three low-grade (I) tumours lacked evidence of tissue or vascular invasion, in contrast with the intermediate-grade (II) and high-grade (III) carcinomas that all (9/9) had evidence of tissue invasion. Five (56%) of the grade II and III carcinomas had evidence of vascular invasion. All (7/7) of the high grade tumours had the highest nuclear histology (grade 4), and four of these seven tumours also exhibited the highest mitotic rate (grade 3≥20 per 10 high power fields). The tubular and tubulopapillary carcinomas were composed of cuboidal to columnar to polygonal cells with mild (low grade) to marked (high grade) cellular atypia. Cytoplasm was clear and vacuolated or eosinophilic and granular. The cells were arranged in tubular or tubular and papillary structures with variable amounts of fibrous connective tissue stroma. The higher-grade tumours tended to have greater amounts of fibrous connective tissue and also tended to be more necrotic.
Primary renal tumours in 19 cats
DSH=domestic short hair, M=male, MC = male castrated, FS=female spayed, NA=not available, LTF=lost to follow-up, HSA = haemangiosarcoma, TCC=transitional cell carcinoma.
There was one clearly benign adenoma in this group of feline renal tumours. The tumour was 2 mm, well circumscribed, non-encapsulated and composed of uniform cuboidal cells arranged in tubules. Two other tumours that had been diagnosed originally as adenomas were large (>2 cm) and contained areas of atypia consistent with malignancy. These were reclassified as low-grade carcinomas.
Two of the three transitional cell carcinomas induced a significant scirrhous reaction. Invasion of the distal tubules and collecting ducts was also a feature of these tumours. All three of the transitional cell carcinomas had evidence of metastasis.
One tumour originally diagnosed as a renal carcinoma was reclassified as a nephroblastoma. This tumour contained all three components required for the diagnosis of nephroblastoma: blastemal, epithelial and mesenchymal elements. Blastemal cells were polygonal with oval, usually hyperchromic nuclei, indistinct nucleoli and scant amounts of eosinophilic cytoplasm. Arrangements of the blastemal cells varied from area to area within the tumour and included compact cords and nests, sheets, islands, trabeculae arranged haphazardly with no particular pattern. The mesenchymal component was scant in comparison to the other two elements and was also highly variable with areas of mature fibrous connective tissue as well as areas with a primitive loose myxoid appearance. The epithelial cells were cuboidal to columnar and formed tubular structures with and without a papillary component. Primitive glomerular structures were present infrequently. In addition, there were nests of epithelial cells with abundant clear cytoplasm and cells with abundant, eosinophilic, granular cytoplasm. The mitotic rate was highest (one to five mitoses per high power field) in the blastemic areas and nearly absent in the epithelial areas. This tumour had metastasised to the peritoneum.
The single mesenchymal tumour, a haemangiosarcoma, was composed of plump, spindle-shaped cells arranged in densely packed sheets and also lining blood filled clefts and channels. The cells had oval to irregularly shaped nuclei with moderate variation in size. The chromatin was coarsely clumped or marginated, and the nucleoli were usually multiple and varied in size and shape. The cytoplasm was scant and eosinophilic. There were up to 11 mitoses per high power field, and tumour necrosis was extensive. Tumour cells had a predilection for the tissue surrounding large calibre arteries, but there was no evidence of vascular invasion in the sections examined. Metastatic lesions were not found.
Staging was performed on 14 cats. One cat had stage T1N0M0 disease, seven cats were staged T2N0M1, three were T2N0M0, two were T4N0M1, and one was T4N0M0. Metastatic disease was recognised in 9/14 (64%) of the staged cases. Metastasis occurred in all of the TCC cases (n = 3), five of the carcinoma cases and in the cat with nephroblastoma. Four of the five carcinoma cases that metastasised involved high-grade tumours and one was an intermediate-grade tumour. Metastatic lesions occurred in the lung (n = 6; two TCCs and four intermediate or high-grade carcinomas), liver (n = 2; TCC and high-grade carcinoma), abdomen/peritoneum (n = 3; nephroblastoma and two TCCs), adrenal gland (n = 1; carcinoma) and ureter (n = 1; TCC). It is possible that the TCC in the ureter was a second primary tumour, rather than a metastatic lesion.
Deaths were due to euthanasia (n = 9), septicaemia (n = 1), haemorrhage (n = 2), post-operative complications (n = 1) and hypertrophic cardiomyopathy (n = 2). One cat is still alive following nephrectomy for a low-grade carcinoma. In this case the presenting complaint of seizures and the laboratory finding of polycythaemia resolved after surgery. Case outcome or cause of death is unknown for five cats. Adjunctive therapy was attempted in only one case in this series. A cat with a high-grade carcinoma was treated with one dose of doxorubicin (1 mg/kg IV) when recurrence was noted 3 months postnephrectomy. No metastatic lesions were noted in this cat, but euthanasia was performed because of continued anorexia and weight loss.
Discussion
Primary renal tumours are rare in cats, as evidenced by the small number of cases from four veterinary colleges and one referral clinic in a nearly 6-year period. In this retrospective study, there were 13 renal carcinomas (11 tubular and two tubulopapillary), three transitional cell carcinomas, one malignant nephroblastoma, one haemangiosarcoma and one adenoma. The haemangiosarcoma is, to our knowledge, the first reported case of this tumour type as a primary renal tumour in the cat. The fact that eight of 20 specimens (40%) that had histopathological review as part of this study were rediagnosed as other tumour types is in keeping with the experience of others. Carpenter et al cited the misdiagnosis of carcinoma in a cat with lymphoma that was published in a peer-reviewed journal. They also mention the difficulty in differentiating adenomas and carcinomas, as we discuss below Carpenter et al (1987). Clearly, the paucity of descriptions of these feline tumours in the literature makes their diagnosis more difficult. It may be advisable, therefore, to have more than one pathologist review slides when a diagnosis of a primary feline renal tumour is made.
Discrimination between carcinomas and adenomas can be difficult, and two of the carcinomas and one TCC were initially diagnosed as adenomas. This is also controversial in human pathology. According to human renal pathology literature, adenomas are often incidental findings usually <6 mm in diameter Olsen (1984). Some authors include tumours up to 1 to 2 cm, but also admit that these tumours usually include some areas with malignant features Mostofi and Davis (1998). While size is not a particularly accurate predictor of biological behaviour, one human study showed that 65% of carcinomas greater than 3 cm in diameter metastasised Eble (1990). In the same study, less than 5% of tumours less than 3 cm metastasised. Nuclear grade is considered a more important prognostic indicator in human renal carcinomas Fuhrman et al (1982).
Nephroblastomas are uncommon in most species. We were able to identify 13 feline nephroblastomas in the literature Cotchin (1956), Fitts (1960), Potkay & Garman (1969), Caywood et al (1980), Carpenter et al (1987). Also known as Wilm's tumours (human) and embryonal nephromas and adenosarcomas, these tumours are thought to arise from metanephric blastema Beckwith (1983). Nephroblastomas are composed of three elements: blastemal, mesenchymal and epithelial cells. Triphasic or mixed tumours have fairly equal amounts of each element. Any of the three elements can predominate, and tumours with a predominate element are called monophasic. Histopathological descriptions of previously reported feline nephroblastomas are fairly scant. One was diagnosed as an embryonal sarcoma with mesenchymal elements predominating Fitts (1960). Primitive tubules and glomeruli were described in this tumour. A survey of feline tumours in Great Britain identified 13 renal tumours (including lymphoma), and one was identified as an embryonal nephroma of adenosarcoma type. No other description was provided Cotchin (1956). Another report of a malignant nephroblastoma in a 2-year-old male cat indicated that the histological appearance of the tumour was typical of a nephroblastoma with pseudoglomeruli and embryonic tubules Potkay & Garman (1969). The report of nine nephroblastomas by Caywood et al (1980) did not provide specific histological description of these tumours. One nephroblastoma in a 1.5-year-old cat reported by Carpenter et al (1987) was described as having ill-defined tubules lying within a fine fibrovascular stoma. The nephroblastoma in our study contained all three elements, blastemal, epithelial and mesenchymal. The blastemal and epithelial elements were nearly equal in amount, but the mesenchymal element was fairly scant. The tumour was malignant, and nodules morphologically identical to the renal tumour were found scattered throughout the omentum at necropsy examination. Although these tumours are reported to occur in young animals, the cat in our study was a 7-year-old DSH and the age range in a previous series was from 2 to 8 (mean = 6) Caywood et al (1980).
Although more male cats were affected than female cats in this study, a significant male predisposition was not demonstrated. However, there was no reference population data with which the gender distribution of this study can be compared and we have no assurance that a 50/50 sex ratio was present in the population at risk. A significant male predisposition has not been demonstrated in other feline studies, but is reported for dogs and people Caywood et al (1980), Carpenter et al (1987). The high metastatic rate (64% of staged cases) in this group of cats is similar to that reported previously (56%) Carpenter et al (1987), and suggests that primary renal tumours are aggressive in cats, although an apparent cure was achieved with surgery in at least one case. Transitional cell carcinomas appear especially aggressive, metastasising in all cases in our report, as well as in all (8/8) cases reported in a previous study Carpenter et al (1987). One case of TCC in our series that was presumed to have metastasised to the ureter may actually have had two separate primary tumours of the uroepithelium. Clinical signs were vague, such as anorexia and weight loss, as reported by others Caywood et al (1980), Carpenter et al (1987). While polycythaemia is reported in other species with renal tumours, we had only one case with polycythaemia in this series and did not find previous reports of feline primary renal tumour-related polycythaemia in the literature. It is of interest to note that the cat with polycythaemia had a 3-year history of seizures that resolved following nephrectomy for the renal carcinoma. Serum biochemical analysis revealed azotaemia in the majority of cases, but it often had a prerenal component and did not generally aid in diagnosis. Haematuria, a common finding in human renal cancer patients, was also found in the majority (6/9) of cases for which urine was examined and may be a useful indicator of renal neoplasia. However, hematuria was noted in only three of 18 (17%) of the cases reported by Carpenter et al (1987). Other bloodwork and urinalysis results for our series of cats were inconsistent and not specific for renal neoplasia.
As most tumours are unilateral, nephrectomy may be a possible treatment option, provided metastasis has not occurred. The metastatic sites identified in this group of cases indicate that three-view thoracic radiographs and abdominal ultrasound are warranted as routine staging procedures. The role of adjunctive therapy such as chemotherapy or immunotherapy is not established.