Proteinuria in canine patients with lymphoma

Abstract

Objectives

To determine if proteinuria is more common in dogs with lymphoma when compared with healthy dogs and to assess the severity and frequency of proteinuria in dogs with lymphoma.

Methods

Determination of urine protein:creatinine ratio in 32 dogs with lymphoma compared with 30 healthy dogs.

Results

Canine patients with lymphoma are more likely to be proteinuric compared with healthy dogs. Proteinuria is common in dogs with lymphoma, although in most cases it is not severe. The presence of proteinuria is not linked with the stage or substage of lymphoma.

Clinical Significance

Mild proteinuria is a common finding in dogs with lymphoma. The clinical impact of the proteinuria is probably low.

Introduction

Proteinuria is defined as the presence of an elevated amount of protein in the urine (Lees and others 2005). Proteinuria can be considered as prerenal, renal or postrenal in origin. Renal proteinuria can be classified as functional or pathologic and can also be considered as tubular or glomerular in origin. Renal proteinuria can be used as a marker of kidney dysfunction in dogs (Lees and others 2005).

It has been documented in human medicine that structural glomerular diseases and impaired glomerular permselectivity can be sequelae of lymphoma (Pedersen and Johensen 2005, Mallouk and others 2006). Lymphoma is a common haematopoietic malignancy in dogs. The standard of care treatment is systemic multi-drug chemotherapy that can be prolonged. Treated patients may be faced with several potential systemic problems due either to the chemotherapy or to the lymphoma itself. Identifying potential renal complications could be important to better manage the overall treatment of these patients. To the best of the authors’ knowledge, no prospective studies have been performed to assess the prevalence and severity of renal proteinuria in dogs with lymphoma.

The hypothesis was that renal proteinuria is common in dogs with lymphoma.

Materials And Methods

This was a prospective case–control hospital-based study conducted partly at Veterinary Referral and Critical Care (VRCC), a private referral hospital, and partly at Marks Farm Veterinary Centre, a first opinion practice. Thirty-two dogs with histological or cytological diagnosis of lymphoma were prospectively enrolled in the study. Complete haematology, biochemistry, electrolytes, urine analysis, urine protein:creatinine ratio (UPC) and appropriate imaging studies for disease staging (radiographs and ultrasound) were performed in each case. Additional tests (e.g. coagulation profiles, infectious disease serology and bone marrow aspirate) were performed in cases as clinical status dictated. Immunophenotyping was performed by immunocytochemistry, flow cytometry or immunohistochemistry when possible.

A control group of 30 healthy dogs was enrolled. The animals came from the same population as the diseased cases and had the same age range. Twenty-four dogs were presented for a routine health check. The six remaining dogs were blood donors routinely used in the hospital. The dogs were considered healthy after a normal physical examination, haematology, biochemistry and urine analysis.

Patients in the lymphoma group that had another disease known to cause proteinuria at the time of the diagnosis were excluded from the study. Patients with lymphoma of the urinary tract were excluded as it would have been difficult to assess the origin of the proteinuria. Patients with signs of dysproteinaemia (e.g. high globulin) were excluded. Patients with postrenal proteinuria, defined as presence of white blood cells (WBCs) and or red blood cells (RBCs) [>3 WBC/high power field (HPF) and/or >5 RBC/HPF] in the urine sediment or presence of bacteria in the sediment, were excluded.

None of the patients in the lymphoma group or in the control group was on any treatment for the tumour or for any unrelated condition at the time of the enrolment in the study.

All dogs with lymphoma received a 15-week induction of a COAP (cyclosphosphamide, vincristine, doxorubicin and prednisolone) based protocol. Several different rescue chemotherapy protocols were employed after the 15-week induction of COAP.

The urine analyses were performed within 30 minutes of urine collection. Urine sediment examination was performed at the same time. The UPC was performed at a reference laboratory (IDEXX, UK). The protein concentration was measured using pyrogallol red and creatinine by the Jaffe procedure. Both procedures were performed using an Olympus AU2700 analyser (Olympus).

The urine was collected by cystocentesis in the study group and by voiding in the control group.

Data were entered in an ad hoc database and analysed by using Stata 11 · 2 MP (Stata Corp). The age distribution in the two groups was tested by a t test, after having checked for normality by the Shapiro-Wilk test. As UPC values were not normally distributed, a non-parametric test was used (two-sample Wilcoxon rank-sum test). The level of proteinuria was categorised as either normal UPC ≤ 0 · 5 or pathological UPC > 0 · 5. To verify the presence of an association between the stage and the substage of patients with lymphoma and the magnitude of UPC, Fisher's test was performed.

Results

Of the 32 dogs in the lymphoma group (Table 1), 18 dogs were female (4 entire and 14 neutered) and 14 dogs were male (5 entire and 9 neutered). Ages ranged from 2 to 12 years (median age was 7 years). There were five Rottweilers, three Labrador retrievers, two Border collies, three Jack Russell terriers, two crossbreds, two golden retrievers, two West Highland white terriers and one each of the following breeds: English springer spaniel, Staffordshire bull terrier, flat coated retriever, Irish water spaniel, boxer, Irish setter, whippet, American pit bull terrier, Dogue de Bordeaux, Dalmatian, bullmastiff, Sealyham terrier and Shih tzu.

Table 1. Lymphoma group information about type of lymphoma and UPC values

Patient ID	Age (years)	Sex	Breed	Stage/substage	Immunophenotype	Grade	Diagnostic method	UPC
UPC Urine protein:creatinine ratio, F Female, N Neutered, M Male, Int Intermediate, C Cytology, H Histology
1	3	FN	Shih tzu	IVb	B	No grade	H	3 · 9
2	6	MN	Rottweiler	IIIa	Not available	Int-high	H	0 · 098
3	9	MN	West Highland white terrier	IVb	B	No grade	C	0 · 843
4	4	M	Rottweiler	IVa	Not available	No grade	C	0 · 18
5	6	F	Bullmastiff	Vb	Not available	No grade	C	0 · 14
6	7	MN	Dalmatian	Vb	Not available	No grade	C	0 · 371
7	9	FN	Golden retriever	Vb	Not available	No grade	C	0 · 5
8	12	FN	Labrador retriever	Vb	Not available	No grade	C	0 · 1
9	9	MN	Flat coated retriever	Vb	Not available	No grade	C	0 · 12
10	6	FN	Labrador retriever	Vb	Not available	No grade	C	0 · 14
11	2	FN	Dogue de Bordeaux	IVb	Not available	No grade	C	0 · 15
12	6	MN	Boxer	IVa	T	No grade	C	0 · 077
13	9	FN	Sealyham terrier	IIIa	T	Int grade	H	1 · 132
14	4	F	Rottweiler	IVa	B	High grade	H	0 · 1
15	11	M	American Staffordshire terrier	IIIa	B	Int-high	H	1 · 5
16	11	F	Border collie	Va	B	Low-inter	H	0 · 173
17	10	M	Border collie	IIIa	B	High grade	H	0 · 1
18	11	M	West Highland white terrier	Vb	B	High grade	H	1 · 8
19	4	M	Whippet	IVa	B	High grade	C	0 · 22
20	7	FN	Rottweiler	Va	B	High grade	H	0 · 063
21	6	FN	Labrador retriever	IVb	B	High grade	C	0 · 257
22	10	MN	English springer spaniel	IVb	B	High grade	H	0 · 69
23	10	FN	Staffordshire bull terrier	Vb	B	Inter grade	H	0 · 095
24	9	FN	Crossbreed	IVa	B	High grade	H	0 · 129
25	12	MN	Jack Russell terrier	IVb	B	High grade	H	0 · 176
26	8	FN	Crossbreed	Vb	B	No grade	C	1 · 113
27	9	FN	Golden retriever	IIIb	B	High grade	C	0 · 091
28	6	FN	Irish setter	IIIb	Not available	High grade	C	0 · 203
29	4	MN	Jack Russell terrier	Vb	B	High grade	C	0 · 1
30	7	MN	Irish water spaniel	IVa	B	High grade	H	0 · 067
31	5	FN	Rottweiler	Vb	Not available	No grade	C	0 · 1
32	4	F	Jack Russell terrier	Va	Not available	No grade	C	0 · 38

In the control group of 30 dogs (Table 2), there were 14 females (11 neutered and 3 entire) and 16 males (14 neutered and 2 entire). Ages ranged from 5 to 13 years (mean 7 · 8 years). Median age in the control group was 7 years. Fifteen different breeds were represented including six greyhounds, three golden retrievers, three crossbreeds, three Jack Russell terriers, three Labrador retrievers, two Border terriers, two Dalmatians and one each of the following breeds: Bernese mountain dog, Border collie, cocker spaniel, Doberman, springer spaniel, Staffordshire bull terrier, West Highland white terrier and German shepherd dog.

Table 2. Control group with UPC values

Patient ID	Age (years)	Sex	Breed	UPC
UPC Urine protein:creatinine ratio, M Male, N Neutered, F Female, GSD German shepherd dog, BMD Bernese mountain dog
1	7	MN	Labrador retriever	0
2	10	MN	Golden retriever	0 · 1
3	8	FN	Golden retriever	0 · 1
4	7	FN	English springer spaniel	0 · 1
5	8	FN	Golden retriever	0 · 1
6	8	FN	Dalmatian	0 · 2
7	7	F	Cocker spaniel	0 · 1
8	13	FN	Labrador retriever	0 · 11
9	8	MN	Jack Russell terrier	0 · 1
10	7	M	Crossbreed	0 · 089
11	7	FN	GSD	0 · 1
12	7	FN	Greyhound	0 · 1
13	6	MN	Greyhound	0 · 1
14	9	MN	Greyhound	0 · 4
15	8	MN	Greyhound	0 · 161
16	6	MN	Greyhound	0 · 038
17	10	FN	Greyhound	0 · 068
18	7	F	BMD	0 · 2
19	14	MN	Crossbreed	0 · 1
20	5	M	BMD	0 · 2
21	5	FN	Border terrier	0 · 1
22	6	MN	Border collie	0 · 1
23	6	FN	Crossbreed	0 · 1
24	5	F	Border terrier	0 · 119
25	9	MN	Jack Russell terrier	0 · 1
26	6	MN	Doberman	0 · 1
27	6	MN	Staffordshire bull terrier	0 · 2
28	13	MN	West Highland white terrier	0 · 11
29	11	FN	Dalmatian	0 · 1
30	6	MN	Crossbreed	0 · 1

There was no significant difference in the age distribution between the two groups (t = 0 · 69; P = 0 · 49).

In the lymphoma group, 14 dogs were diagnosed following histology and in 18 cases cytology was sufficient for diagnosis. Thirteen of the lymphomas were classified as high grade, 2 as intermediate to high grade, 3 as intermediate grade and in 14 cases a grade was not given by the pathologist or clinical pathologist. In 18 cases, lymphoma was classified as B-cell, 2 cases as T-cell lymphoma and in 12 cases immunophenotype was not determined. Staging was performed according to a modified WHO staging scheme: for the purpose of this study, patients were classified as stage V lymphoma when abnormal circulating lymphoid cells were identified in the blood smear or in the bone marrow aspirate (if performed). According to this staging scheme: 6 cases were staged III (4 substage a and 2 substage b), 12 were staged IV (6 substage a and 6 substage b) and 14 were staged V (3 substage a and 11 substage b).

The UPC ranged from 0 · 063 to 3 · 9 [median: 0 · 16; interquartile range (IQR): 0 · 1 to 0 · 44] in the lymphoma group. The UPC varied from 0 to 0 · 4 (median: 0 · 1; IQR: 0 · 10 to 0 · 11) in the control (healthy dogs) group.

There was a significant statistical difference (P = 0 · 0006) in the magnitude of proteinuria between the lymphoma group and the control group (Fig 1). No association was identified between the stage and the substage of patients with lymphoma and the magnitude of UPC (Fisher's exact test = 0 · 195 and 0 · 389, respectively).

Figure 1. Boxplot representing the significant difference in UPC levels between the cases of lymphoma and the controls. The box contains the middle half of the data and is bisected by a line representing the median with the whiskers representing the typical range and closed circles outliers

Discussion

Proteinuria has been reported in canine patients with different tumour types (Pressler and others 2003, Whittemore and others 2006, Wehner and others 2008) including lymphoma.

Lymphoma is a heterogeneous group of tumours a rising from the lymphoreticular system (Vail and Young 2007). Multi-agent systemic chemotherapy is the standard of care treatment. Affected patients may receive chemotherapy for prolonged periods of time with possible systemic complications. It would be important to detect any systemic complications of the lymphoma, such as clinically significant proteinuria in order to better manage the treatment of these patients. As such, the aim of this study was to investigate the presence, frequency, severity and clinical impact of proteinuria in dogs with lymphoma.

The results of the study demonstrated an expected overall increased proteinuria in dogs with lymphoma compared with an age-matched population of healthy dogs. Of 32 dogs with lymphoma, however, only eight had UPC values higher or equal to 0 · 5. This value has been suggested (Stockham and Scott 2008) as the lower-end value for borderline proteinuria (0 · 5 to 2 · 0). Therefore, only one dog in four (25%) had mild proteinuria and only one dog had proteinuria higher than 2 · 0, a value more suspicious of structural glomerular disease.

Only one other study (Pressler and others 2003) addressed the prevalence and severity of proteinuria in dogs with lymphoma. In that study, 4 of 19 (21%) dogs with lymphoma had a UPC > 0 · 5, and therefore the frequency of proteinuria was similar to this study supporting that proteinuria is common in dogs with lymphoma.

Different causes of renal proteinuria are possible in these patients. Functional proteinuria is defined as transient and is usually mild and implies no structural or functional lesions in the kidney. Typically this proteinuria has been linked with exercise (Gary and others 2004) or systemic inflammatory response syndrome (Schafer and others 2011), but none of our patients had signs compatible with SIRS or underwent any form of exercise before the collection of urine. Pathologic renal proteinuria can be divided into glomerular causes, where there is increased permeability of the glomerular capillary wall and tubular causes where there is impaired reabsorption of the protein by the epithelial cells of the tubuli. It has been documented in human medicine that structural glomerular disease (e.g. glomerulonephritis) can be a sequela of lymphoma (Iwanaga and others 2006, Mallouk and others 2006, Alshayeb and Wall 2009), but also impaired glomerular permselectivity is seen in lymphoma cases (Pedersen and Johensen 2005). As the overall level of proteinuria was low in this study, it is more likely that impaired glomerular permselectivity is the cause of the proteinuria in most of these patients.

In our group of dogs with lymphoma, none of the patients developed, during the chemotherapy, signs of overt kidney failure or nephrotic syndrome or any complication clearly linked with the proteinuria. However, eight cases in the study were proteinuric and it is possible that proteinuria may represent a clinical issue in a subgroup of dogs with lymphoma. However, because of the low number of proteinuric patients, this study cannot satisfactorily assess this hypothesis.

The clinical impact of the proteinuria appears negligible in patients with lymphoma.

There was no relationship between stage or substage and proteinuria.

Two different types of urine collection were employed in this study; the urine collection in the lymphoma group was performed by ultrasound-guided cystocentesis in all cases; meanwhile, the urine samples in the control group were collected by free catch method. A study conducted by Beatrice and others (2010) demonstrated that proteinuria measured as UPC would not change in samples collected by free catch or by cystocentesis in the same patient; therefore, using these two different techniques of collection was unlikely to introduce any significant bias in the results.

This study has several limitations. Possibly, the main limitation regards the definition of proteinuria.

The American College of Veterinary Internal Medicine (ACVIM) consensus group study suggests obtaining three samples for UPC taken two or more weeks apart to confirm consistent renal proteinuria. However, this was not performed in this study as these patients once diagnosed quickly underwent multi-agent chemotherapy including glucocorticoids. This class of drugs has been shown to cause significant proteinuria in dogs (Waters and others 1997, Schellenberg and others 2008) and any interference caused by this drug would not allow assessment of the origin of the proteinuria during treatment. Multiple UPCs were therefore not performed in these patients as delay of chemotherapy to collect multiple urine samples would not be considered ethical.

None of the patients had a kidney biopsy performed. Hence, the exact cause of the proteinuria (glomerular versus tubular disease, glomerular disease versus altered permselectivity) could not be ascertained.

The study group included a relatively heterogeneous group of lymphomas, although low-grade and stage I to II lymphomas were excluded. It is widely accepted (Vail and Young 2007, Valli 2007) that lymphoma encompasses a large spectrum of diseases with different biological behaviour and possible different prognoses. As immunophenotype and histology were not available in all cases a specific classification of lymphoma was not possible. Moreover, the definition of stage V in this study was often based on peripheral blood smear abnormalities, and a bone marrow aspirate was not performed in all cases. It is possible that stage migration would have occurred in some cases if we had performed a bone marrow aspirate (Flory and others 2007) in all cases.

It is possible that specific forms of lymphoma or different stages, assessed differently, are associated with different degrees of proteinuria. This study, because of the previously mentioned limitations, cannot provide definitive information on this issue.

In this study, an attempt was made to rule out concurrent disease or treatments potentially capable of causing proteinuria. Infectious diseases, vector-borne diseases in particular, are a well-recognised cause of proteinuria. A control group from the same geographic area was used to decrease the risk of interference with geographically linked infectious diseases. In the southeast of England, where this study was performed, there are no epidemiologic studies available for vector-borne diseases; however, clinical disease due to Borrelia burgdorferi and Anaplasma phagocytophylum are uncommon (Shaw and others 2005). Ehrlichia canis, Dirofilaria immitis and Leishmania spp. are capable of causing proteinuria (Vaden and Grauer 2011) but are not endemic diseases in the British Isles. There were no physical or clinicopathological findings suspicious of infectious diseases in the cases and none of these dogs had left the UK. The presence of occult diseases in the study dogs cannot be ruled out but most were followed throughout the chemotherapy protocol (induction and rescue) and none revealed any signs suspicious of infectious or other chronic inflammatory disease.

Systemic hypertension is another potential cause of proteinuria (Finco 2004, Stephien 2010), although this had been documented only in experimentally induced renal failure. Systemic blood pressure was not recorded in the patients of this study, but none was diagnosed with a disease recognised as likely to cause systemic hypertension (e.g. renal failure and diabetes mellitus). Occult diseases causing hypertension (e.g. early renal failure) or idiopathic hypertension were not specifically ruled out.

In conclusion, this study suggests that proteinuria is common in dogs with lymphoma although it tends to be mild. The clinical impact of the proteinuria is probably low.

Conflict of interest

None of the authors of this article has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper.