Copper-associated hepatitis (CAH) has been well described in Labrador Retrievers. However, the association of CAH with proximal renal tubular dysfunction in this breed has not been characterized.
Copper-associated hepatitis (CAH) has been well described in Labrador Retrievers. However, the association of CAH with proximal renal tubular dysfunction in this breed has not been characterized.
To report clinical features, hepatic and renal histopathologic findings, tissue copper concentrations, and outcome of Labradors with CAH and proximal renal tubular disease.
Nine Labrador Retrievers with renal glucosuria and biopsy-confirmed CAH.
Clinical, clinicopathologic, and light microscopic findings were retrospectively reviewed. Rhodanine staining or atomic emission spectroscopy was performed on all hepatic samples and available renal tissue (4 dogs) to assess copper concentrations.
Eight dogs had a history of polyuria and polydipsia, and all dogs had increased serum bilirubin concentrations. Five dogs had hyperchloremic metabolic acidosis. Three dogs with acidemia had paradoxical alkalinuria. All renal specimens had increased copper concentrations. Renal tubular vacuolization, degeneration, and regeneration were observed on light microscopy. Four dogs died within 10 days of diagnosis. One dog survived 2 months; 4 dogs survived more than 1 year. In long-term survivors, including 2 that did not undergo immediate copper chelation, resolution of renal tubular dysfunction occurred within weeks to months.
Labrador Retrievers with CAH can develop clinical and laboratory evidence of renal tubular dysfunction in association with increased renal copper concentrations. Given the rarity of renal tubular disorders, detection of renal glucosuria and increased ALT activity in a Labrador Retriever is suggestive of CAH. Although renal tubular dysfunction may indicate advanced disease, successful long-term outcome is possible with a variety of therapies.
inductively coupled plasma atomic emission spectroscopy
Michigan State University Diagnostic Center for Population and Animal Health
Michigan State University Veterinary Teaching Hospital
polyuria and polydipsia
Copper-associated hepatitis (CAH) is a well-recognized disease in dogs. Since the original description of copper toxicosis in Bedlington Terriers in 1975, other affected breeds, including West Highland White Terriers, Doberman Pinschers, and Dalmatians, have been identified.[1-4] More recently, several reports have described CAH in Labrador Retrievers.[5-7] Although a specific genetic mutation has only been described in Bedlington Terriers, CAH is presumed to be hereditary in other affected breeds.[8, 9]
The pathophysiologic consequences of excess hepatic copper, as well as therapies to address this problem, are well described. Yet, little remains known about mechanisms of copper accumulation in non-Bedlington Terrier breeds.[2, 5, 8] Given that sex predisposition, signalment, and clinical course vary among breeds, it is probable that different genetic mutations are involved.[1-3, 8] Furthermore, it remains unknown whether copper can accumulate in other organs and result in dysfunction in those organs. Specific organ or tissue accumulation could vary depending on the mutation. There are a variety of copper storage disorders in humans. Wilson's disease is the best-described disorder in humans and is characterized by reduction or absence of ATP-7B gene expression resulting in copper accumulation in the brain, eyes, and kidneys in addition to the liver. In fact, approximately 40–50% of patients present for neurologic disease or signs unrelated to hepatic copper accumulation.[11, 12]
Recently, 2 case reports (not in Labrador Retrievers) have identified concurrent CAH and transient Fanconi syndrome that resolved with d-penicillamine chelation therapy.[13, 14] Renal copper was only quantitated in 1 dog and found to be abnormally high at 212 ppm (reference range, 17.5–60 ppm), and the semiquantitative copper determination (rhodanine-stained kidney sections) was negative in this dog. Another report documented Fanconi syndrome in a Labrador Retriever with transiently increased liver enzyme activity, but hepatic biopsy was not performed. In another series of 24 Labrador Retrievers with chronic hepatitis, 3 dogs were noted to have transient renal glucosuria. However, it remains unclear if these dogs had CAH. Furthermore, renal histopathology and follow-up information were not reported. Studies characterizing concurrent proximal renal tubular dysfunction and CAH in Labrador Retrievers have not been reported. The objective of this study was to describe clinical features, diagnostic findings, renal and hepatic histopathology, quantitative tissue copper concentrations, and outcomes in 9 Labrador Retrievers with concurrent CAH and proximal renal tubular dysfunction.
Medical records of Labrador Retrievers diagnosed with CAH at the Michigan State University Veterinary Teaching Hospital (MSU-VTH) between 2006 and 2012 were examined. Archived data used in a previous pathologic evaluation of 16 Labradors with CAH conducted by one of the authors (RCS) also were evaluated. Cases were excluded from review if hepatic histopathology, urinalysis, or urine sediment examination had not been performed. Cases were also excluded if the medical record and follow-up information were unavailable. Cases were included only if a paired serum biochemical profile and urinalysis were available that documented glucosuria in the absence of hyperglycemia. Information collected from the medical record included signalment, history, physical examination findings, diagnostic imaging and laboratory results, treatment, and outcome.
Hematologic, serum biochemical, and venous blood gas analyses were performed using automated analyzers at the Michigan State University Diagnostic Center for Population and Animal Health (MSU-DCPAH).1,2,3 Urine pH was determined by pH meter4 in 7 dogs and by routine urine dipstick5 analysis in 2 dogs (patients 5 and 7). Urine specific gravity was determined by refractometry.6 Protein was determined by the sulfosalicylic acid turbidimetric test in 7 dogs and by routine dipstick analysis in 2 dogs (patients 5 and 7). Detection of glucosuria and ketonuria was by urine dipstick analysis. Urine metabolic profiles, including determination of urine carbohydrate, cystine, organic acid, and amino acid concentrations, were performed at a commercial laboratory.7 Serum leptospiral microscopic agglutination antibody titers were performed at the MSU-DCPAH.
One author (RCS) examined hematoxylin and eosin (H&E) and rhodanine stained slides from all cases meeting the above criteria. A diagnosis of CAH was based on findings of centrilobular hepatitis and increased centrilobular copper accumulation.[5, 7, 8] The severity of hepatic copper accumulation was designated as none, minimal, mild, moderate, or marked, which corresponds to the staining grades 0, 1–2, 3, 4, and 5, respectively.8 This semiquantitative grading system has been evaluated and used in previous publications to diagnose CAH, with grades 3 through 5 being consistent with primary CAH.[3, 5, 16] Rhodanine stains also were applied to any available renal specimens to evaluate for copper accumulation.
For cases in which tissue copper concentrations had not been previously determined and liver and renal tissue were still available, quantitative copper analysis of formalin-fixed, deparaffinized tissue sections was performed using inductively coupled plasma atomic emission spectroscopy (ICP-AES) as described elsewhere. These procedures have been validated and are routinely used to evaluate copper concentrations according to standard operating procedures at the MSU-DCPAH.[7, 17]
Copper-associated hepatitis was diagnosed in 30 Labrador Retrievers. One case with a presumptive diagnosis based on aspiration cytology, 1 case lacking a urinalysis, and 1 case with an incomplete medical record and follow-up information were excluded from review. Twenty-seven cases with a complete medical record, follow-up information, and a paired serum biochemical profile and urinalysis were reviewed. Nine cases (33%) had glucosuria in the absence of hyperglycemia. Median age at diagnosis was 5 years (range, 3–9 years) and the median body weight was 35.4 kg (range, 25.5–50.6 kg). There were 5 spayed females and 4 castrated males.
Eight dogs (89%) had a history of polyuria and polydipsia (PU/PD) ranging from 3 days to >1 month. Gastrointestinal abnormalities were a major presenting concern with 9 (100%) dogs having both vomiting and either decreased appetite or complete anorexia. One dog had diarrhea (patient 6). In addition to PU/PD, anorexia, and vomiting, other common owner concerns were lethargy (8/9) and weight loss (4/9). The median duration of clinical signs before presentation was 21 days (range, 3–90 days).
Physical examination indicated dehydration (based on tacky mucous membranes or prolonged skin tent) in 4 dogs (patients 3, 6, 8, 9), icterus in 4 dogs (patients 1, 3, 6, 9), tachycardia in 2 dogs (patients 1 and 3), and hypersalivation in 1 dog (patient 5). One dog (patient 3) was laterally recumbent upon initial evaluation and had gingival and rectal bleeding.
Hematologic evaluations identified a median PCV of 45% (range, 24–49%) and 3 dogs (patients 3, 6, 8) were anemic (PCV <40%). An inflammatory leukogram characterized by leukocytosis with band neutrophilia was identified in the same 3 dogs (patients 3, 6, 8). All 9 dogs had normal platelet counts at initial evaluation. Coagulation testing was performed in 8 dogs, and prolongation of both the prothombin time (PT) and activated partial thromboplastin time (APTT) was present in 2 dogs (patients 3 and 6). Evaluation of serum biochemistry results identified increased alanine aminotransferase (ALT) activity in all dogs with a median value of 774 U/L (range, 115–2074 U/L; normal, 14–102 U/L). Increased serum alkaline phosphatase (ALP) activity was present in 6/9 (67%) dogs. The magnitude of ALT increase was greater than the ALP increase in all dogs. Total bilirubin concentration was increased in all 9 dogs with a median concentration of 1.2 mg/dL (range, 0.6–14.2 mg/dL; normal, 0.1–0.4 mg/dL). Seven dogs (patients 2, 4–9) had only mild increases in total bilirubin concentrations (<2.5 mg/dL). Median serum albumin concentration was 3.1 g/dL (normal, 2.8–4.0 g/dL) and was decreased in 3 dogs (patients 3, 6, 9). Three dogs (patients 6, 8, 9) had decreased blood urea nitrogen concentrations. A fasting ammonia concentration was moderately increased (48 mmol/L; normal, 0.0–21.0 mmol/L) in 1 dog (patient 6) and markedly increased (121 mmol/L; normal, 0.0–21.0 mmol/L) in 1 dog (patient 3). Mild hypoglycemia was present in 2 dogs (patients 3 and 8); 7 dogs were normoglycemic. Two dogs (patients 5 and 7) were mildly azotemic (serum creatinine concentrations, 2.0 and 2.3 mg/dL) and isosthenuric (urine specific gravity, 1.010 and 1.013). Serum microscopic agglutination antibody titers for leptospira serovars were interpreted as negative in 6 dogs and mildly increased in 2 dogs (serovars grippotyphosa: 200, pomona: 100 and serovars grippotyphosa: 200, canicola: 100) that had received leptospira vaccines. Convalescent titers were evaluated in 2 dogs and found to be negative. Fluorescent antibody testing on hepatic and renal tissue for leptospira antigens was negative in 1 dog.
Serum biochemical and urinalysis findings supported renal tubular dysfunction in all dogs. Seven of 7 dogs (patients 1–4, 6, 8, 9) had decreased or low normal serum TCO2 (≤20 mEq/L) concentrations with a normal anion gap. Six of 9 dogs (patients 3–6, 8, 9) were hyperchloremic. Serum phosphorous concentration ≤3 mg/dL was present in 6 dogs (patients 1–4, 8, 9). Two dogs (patients 1 and 3) had mild hypokalemia. Venous blood gas analyses were performed in 4 dogs (patients 3, 4, 6, 8). All 4 dogs had metabolic acidemia with pH values of 7.28, 7.34, 7.32, and 7.27 (normal, 7.36–7.44) and bicarbonate concentrations of 8.0, 13.5, 12.6, and 13.3 mEq/L (normal, 16.0–24.0 mEq/L), respectively. Three of these dogs (patients 4, 6, 8) had findings consistent with a normal anion gap hyperchloremic metabolic acidosis. One severely affected dog (patient 3) appeared to have a mixed disturbance consisting of a normal anion gap hyperchloremic metabolic acidosis and metabolic alkalosis. Partial pressure of carbon dioxide was low or low normal in all 4 dogs consistent with a compensatory respiratory alkalosis.
The median urine specific gravity was 1.013 (range, 1.007–1.037). Urine pH was neutral to alkaline in 8 of 9 dogs. In 2 dogs (patients 5 and 7), pH determinations were based on dipstick analysis (urine pH, 7.0). Urine pH in the 4 dogs with metabolic acidemia was neutral to alkaline (patients 3, 4, 6) or mildly acidic (pH, 6.4) (patient 8). Trace to moderate (300 mg/dL) proteinuria was detected in 8 of 9 dogs. Two proteinuric dogs had pyuria, 1 had microscopic hematuria, and 1 had pyuria and microscopic hematuria. The severity of glucosuria was variable with trace (patients 3 and 9), 250 mg/dL (patient 5), 500 mg/dL (patient 7), 1000 mg/dL (patients 1, 2, 6, 8), and >2000 mg/dL (patient 4) being detected. Trace (patients 6, 7, 9) and moderate (patient 8) ketonuria were detected in some dogs. Urine metabolic profiles6 were submitted in 2 dogs (patients 2 and 4) and disclosed generalized amino aciduria, cystinuria, glucosuria, and lactic aciduria. These findings confirmed Fanconi syndrome in both dogs. Aerobic bacterial urine culture was performed and found to be negative in 6 dogs. Urine sediment examination identified pyuria in 4 dogs (patients 1–3, 9), all of which had negative urine cultures. Microscopic hematuria was detected in 2 dogs (patients 2 and 4) (10–25 RBC/hpf). Moderate numbers (>10/lpf) of granular casts were detected in 3 dogs (patients 1, 2, 7).
Survey abdominal radiographs were performed in 5 dogs and disclosed hepatomegaly in 2 dogs (patients 1 and 6), microhepatica in 2 dogs (patients 8 and 9), and normal liver size in 1 dog (patient 2). Renal parenchyma appeared radiographically normal in all 5 dogs. Abdominal ultrasonography was performed in 8 dogs. Common hepatic sonographic abnormalities included parenchymal heterogeneity in 5 dogs (patients 3–6, 9) and microhepatica in 5 dogs (patients 3, 4, 5, 8, 9). One dog (patient 8) had hyperechoic parenchyma; 2 dogs (patients 1 and 2) had sonographically normal livers. Hyperechoic renal cortices were observed in 4 dogs (patients 2, 3, 4, 9). The kidneys were sonographically normal in the other 4 dogs.
Liver tissue specimens were obtained by laparoscopy (4 dogs), laparotomy (2 dogs), or at the time of postmortem examination (3 dogs). Inflammation was present in all samples and was described as moderate (4 dogs) or severe (5 dogs). Inflammatory cell infiltrates were mixed (lymphocytic, neutrophilic, and macrophagic) (patients 2–4, 8, 9), mixed with neutrophils and macrophages (patients 5–7), and lymphoplasmacytic (patient 1). Inflammation was localized primarily to the centrilobular region in 7 of 9 cases (Fig 1). In the remaining dogs (patients 1 and 5), localization of inflammatory lesions was difficult because of lobular collapse, but thought to be centrilobular. Six dogs (patients 3, 5–9) had moderate to marked, primarily centrilobular, necrosis. Six dogs (patients 1–5, 9) had bridging fibrosis and 3 dogs (patients 1, 4, 5) had cirrhosis. Rhodanine staining identified excess copper accumulation in all cases, and was graded as mild to moderate (2/9), moderate (2/9), or marked (5/9) (Fig 2). The median liver quantitative copper concentration was 2239 ppm (dry matter basis) (range, 1200–5331 ppm; normal, <400 ppm). In 2 dogs (patients 6 and 9) in which quantitative copper determinations were not performed, rhodanine staining demonstrated marked centrilobular copper accumulation and hepatitis that were consistent with CAH. One dog (patient 1) with only a moderate increase in hepatic quantitative copper (1200 ppm) had severe bridging fibrosis and nodular regeneration. However, a rhodanine stain identified moderate copper accumulation (grade 3+) and associated hepatitis consistent with CAH.[3, 5, 15]
Renal tissue specimens were obtained by laparotomy (2 dogs) or at the time of postmortem examination (3 dogs). Light microscopic findings consisted of multifocal, segmental to diffuse, proximal tubular degeneration, necrosis, and regeneration in all kidney specimens (Fig 3). Proximal tubular epithelial degeneration was characterized by vacuolation with vacuoles being round, discrete, and of variable size. The proximal tubular necrosis was acute and characterized by cytoplasmic hypereosinophilia and nuclear pyknosis or karyolysis. Occasionally, tubular epithelial cells were attenuated. Tubular regeneration was characterized by epithelial cell hypertrophy with cytoplasmic basophilia, increased mitoses, and occasionally a haphazard arrangement of the cells. Multifocal proximal tubule eosinophilic granular casts were observed in 3 dogs. Minimal to mild, multifocal, interstitial inflammation was observed in all renal specimens and was characterized as primarily lymphoplasmacytic and occasionally neutrophilic. Severe tubular necrosis, moderate, multifocal hemorrhage, and cast formation were observed in 1 severely affected dog (patient 3) with coagulation abnormalities.
In rhodanine-stained kidney sections from 2 dogs, there were moderate (patient 7) to marked (patient 3) numbers of copper granules within the proximal tubular epithelial cells and tubule lumens, with the distal tubules being affected less severely (Fig 4). Intracytoplasmic copper granules often were apically oriented within the epithelial cells. Copper deposition was primarily localized to the corticomedullary junction and medullary areas. Quantitative renal copper concentrations were performed and found to be increased in 4 dogs (patients 2, 3, 6, 7) at 624, 775, 698, and 261 ppm (dry matter basis9) (normal, 17.5–60 ppm).
Four dogs (patients 3, 6–8) died or were euthanized within 10 days of diagnosis. One dog (patient 6) was stable at presentation, but declined rapidly after surgical hepatic and renal biopsies. The other 3 dogs were lethargic and dehydrated at initial evaluation and deteriorated clinically during hospitalization. Treatment of these dogs was primarily symptomatic and supportive and included IV administration of fluids, antibiotics (ampicillin), and antacids (4 dogs); anti-emetics (3 dogs); vitamin K1 (2 dogs); and lactulose (1 dog). Two of these dogs received colloids IV because of hypoalbuminemia and hypotension, and 2 received fresh frozen plasma transfusions. None of these 4 dogs received steroids, antioxidants, ursodeoxycholic acid, or treatment to decrease tissue copper. All dogs with anemia, inflammatory leukograms, coagulopathy, and 2 of 3 dogs with hypoalbuminemia were among the dogs that died near the time of diagnosis.
One dog (patient 9) with an intermediate survival time of 2 months was treated with antacids, vitamin E, ursodeoxycholic acid,10 and antiemetics. This dog was clinically stable for nearly 2 months, but was euthanized for relapsing gastrointestinal signs.
Four dogs (patients 1, 2, 4, 5) experienced long-term (>1 year) survival, 3 (patients 1, 2, 4) of which were alive at the time of writing. Treatment of these dogs was variable. All 4 dogs were fed a low copper diet (Hill's l/d11 or Royal Canin Hepatic LS12) and treated with a short course (<3 weeks) of antibiotics. In addition, all of these dogs received antioxidants or nutraceuticals, including s-adenosyl-methionine (4 dogs), silymarin (3 dogs), and vitamin E (1 dog). Three dogs (patients 2, 4, 5) were treated with d-penicillamine (dosage range, 10–14 mg/kg PO q12 h) for 1–4 months. Three dogs were treated with ursodeoxycholic acid10 (patients 1, 2, 4) (dosage range, 7.5–12 mg/kg PO q12–24 h) and a tapering course of prednisolone (patients 1, 4, 5) (initial dosage range, 0.8–1.8 mg/kg/day). Sequential serum biochemical profiles disclosed normal or decreased ALT activity and bilirubin concentration. The renal azotemia in patient 5 had normalized at a 1-month follow-up evaluation. Liver biopsies were repeated in 1 dog (patient 2) after receiving d-penicillamine for 4 months. Quantitative copper concentrations were decreased by over 50% (1075 ppm, originally 2448 ppm), and the dog was transitioned to oral zinc gluconate.
Resolution of glucosuria and proteinuria was evident in all 4 long-term survivors at 1, 3, 4, and 6 months based on normal urinalysis findings. One dog (patient 4) had no detectable protein in its urine by sulfosalicylic acid precipitation (previously, trace) and only trace glucosuria (previously, >2000 mg/dL) within 5 days of starting ursodeoxycholic acid,10 Denamarin,13 and dietary therapy. This dog then was started on prednisolone and d-penicillamine, and urinalysis results were normal several weeks later. Two other dogs had resolution of proteinuria and glucosuria with normalization of serum TCO2 without (patient 1) or before (patient 5) chelation therapy. The remaining long-term survivor (patient 2) initially was treated with d-penicillamine and also experienced resolution of proteinuria and glucosuria. Renal biopsies were not repeated in any dog.
Copper-associated hepatitis is a common liver disease in Labrador Retrievers.[5-8] Diagnosis requires hepatic histopathology with either semiquantitative or quantitative copper determinations. Both the amount and location of copper accumulation are important in reaching a diagnosis. Eight dogs in our series had classic CAH characterized by centrilobular hepatitis and copper accumulation. One dog had moderate increases (1200 ppm) in quantitative copper. However, the presence of fibrosis decreases the relative copper concentration, and regenerative nodules often do not accumulate copper. This dog had severe fibrosis and nodular regeneration. It is possible that the tissue analyzed quantitatively contained fibrosis or regenerative nodules, which may explain the discrepancy between rhodanine staining and quantitative copper concentrations. This case still was considered to be CAH given the breed, rhodanine grading, and association of centrilobular inflammatory changes with hepatocellular copper.
Acquired diseases of the proximal renal tubule are rare in veterinary medicine. Primary renal glucosuria, Fanconi syndrome, and type II renal tubular acidosis have all been described.[15, 20, 21] Fanconi syndrome is characterized by failure of normal proximal tubular reabsorptive function resulting in excessive loss of glucose, amino acids, water, bicarbonate, phosphate, and other electrolytes. Renal glucosuria often is the initial finding before the development of overt Fanconi syndrome.[19, 22] Osmotic diuresis results in PU/PD in many of these dogs. With disease progression, hyperchloremic metabolic acidosis and renal azotemia often develop. Although substantial information is available on Fanconi syndrome in Basenjis, there is little published information characterizing acquired renal tubulopathies. Several reports have described acquired, sometimes transient tubulopathies in dogs in association with a variety of diseases, including acute kidney injury, renal toxins, primary hypoparathyroidism, and CAH.[13, 14, 24, 25]
Eight dogs in our series had PU/PD. A previous report of Labrador Retrievers with hepatitis only documented PU/PD in 2 of 24 dogs. On the basis of our inclusion criteria (renal tubular dysfunction), the presence of PU/PD is not unexpected. Glucosuria resulting in an osmotic diuresis is 1 explanation. However, PU/PD is known to occur in various liver diseases in which there is no evidence of renal tubular dysfunction. Although exact mechanisms are unknown, renal medullary washout secondary to decreased blood urea nitrogen concentrations, stimulation of thirst receptors secondary to hepatic encephalopathy, and increased endogenous cortisol production have all been proposed. Three dogs in our series had decreased blood urea nitrogen concentrations, 2 dogs had increased fasting ammonia concentrations, and 2 dogs were azotemic. Furthermore, the trace glucosuria observed in 2 dogs was unlikely to induce substantial osmotic diuresis. As such, the PU/PD in our series of dogs is likely multifactorial. Nevertheless, PU/PD is an important historical finding that could be indicative of underlying renal tubular dysfunction.
The presence of renal glucosuria localized disease to the proximal renal tubules in all dogs in our report. The findings of low to low normal serum phosphorous concentration (≤3.0 mg/dL; 6/7), mild proteinuria (8/9), and low to low normal serum TCO2 (≤20 mEq/L; 7/7) support more global dysfunction of the proximal renal tubules and are suggestive of Fanconi syndrome. The presence of hyperchloremic metabolic acidemia documented in 4 dogs in our series also is a classic finding associated with both proximal renal tubular acidosis (type II) and progressive Fanconi syndrome. However, confirmatory testing for Fanconi syndrome only was performed in 2 dogs. In previously reported cases of concurrent Fanconi syndrome and CAH in dogs, proteinuria and glucosuria also were detected. However, acid base status and urine pH were only reported in 1 dog (hyperchloremic metabolic acidemia with alkalinuria).[13, 14] In our series, we documented metabolic acidemia in all dogs in which blood pH determinations were made. Interestingly, the majority of dogs (78%) in this study paradoxically had neutral to alkaline urine despite evidence of systemic acidosis (low normal or decreased serum TCO2). In an animal with only proximal tubular dysfunction, distal urinary acidification mechanisms are still intact, and an acidic urine pH can be achieved, especially in the face of acidemia. This is consistent with distal renal tubular acidosis (type I). As such, it is possible that some dogs had defects in both proximal and distal tubular function. The documented copper accumulation in the renal medulla supports this possibility. The normal anion gap hyperchloremic metabolic acidosis present in the majority of dogs in our study is consistent with renal tubular acidosis. However, it is probable that some dogs had mixed acid base disturbances as evidenced by 1 dog (patient 3) that had hyperchloremic metabolic acidosis and metabolic alkalosis. Furthermore, the presence of renal tubular acidosis with concomitant vomiting, diarrhea, dehydration, and azotemia could contribute to mixed acid base disturbances in other dogs. Despite this possibility, the predominant clinical and laboratory picture was that of a hyperchloremic (normal anion gap) metabolic acidosis consistent with renal tubular acidosis.
Renal histopathologic abnormalities in our series of dogs were similar to those in 2 previous reports, but it remains unclear from these reports if increased renal copper was associated with the renal tubulopathy.[13, 14] Of the 4 previously reported cases, increased renal copper was documented in 2 dogs. Staining for renal copper was negative in 1 dog; 1 dog did not undergo renal biopsy.[13, 14] In our series of dogs, increased copper deposition was observed in renal tissues of all 4 dogs in which kidney specimens were available. Our observations suggest that renal tubular abnormalities are due, at least in part, to renal copper accumulation. The decision to perform rhodanine staining or ICP-AES was based on the amount of tissue available. Quantitative copper analysis was deemed most important if only small, but sufficient, amounts of tissue were available. Given that the normal amount of renal copper is low (17.5–60 ppm, dry matter basis), histochemical staining methods (rubeanic acid or rhodanine) may not consistently detect increased renal copper. In liver tissue, histochemical staining usually is not positive until copper concentration exceeds 300–400 ppm.[3, 8]
Abnormalities of renal tubular function have been identified in humans with Wilson's disease and Menkes disease, an X-linked recessive condition caused by mutations in the ATP-7A gene.[27-30] Although renal tubular dysfunction in Labradors with CAH and humans with Wilson's and Menkes disease is associated with renal copper accumulation,[28, 30] the exact pathogenic mechanism of copper-induced injury is unclear. The proximal tubule is the site of nephrotoxicity caused by many metals. Copper is known to increase the rate of the normally slow-occurring Haber–Weiss reaction, leading to enhanced free radical formation and subsequent oxidative damage, cellular necrosis, and inflammation. Copper also may alter Na-K-ATPase activity in tubular epithelial cells. It is likely that these prooxidant and enzyme altering effects of copper are involved in the pathogenesis of copper-induced renal tubular disease. In rodent models of Wilson's and Menkes disease, proximal tubular epithelial copper accumulation was observed under both acute and chronic copper loading conditions, whereas other components of the nephron were relatively spared.[32-34] In 1 study, proximal tubular cell disarray and irreversible nuclear damage were correlated with increasing intranuclear copper accumulation. In the same study, decreases in copper and tubular recovery were associated with lysosomal sequestration and excretion of copper into the tubular lumen. However, in a different model of Wilson's disease, it was suggested that renal dysfunction was independent of increased renal copper. Their findings suggested that copper-metallothionein complexes released from the damaged liver were freely filtered at the glomerulus and then induced dysfunction of brush border membranes of the proximal tubular cells after the complexes split into free copper and amino acids. Similarly, copper-metallothionein complexes also were demonstrated in excess in the proximal tubular cells in a model of Menkes disease. However, the finding of metallothionein mRNA in the kidney suggested that formation of these complexes occurred in situ. Although proposed mechanisms of copper-induced renal injury vary with different models, most reports are in agreement that copper plays a causative role, and resolution of disease depends on reduction of tissue copper.[27-29] The source of renal copper in our dogs remains unknown. However, hepatic release of excess copper is an appealing hypothesis in these Labradors given the degree of hepatic necrosis noted in 6 dogs. Nevertheless, we cannot exclude the possibility that chronic renal copper accumulation observed in our dogs resulted from processes independent of hepatic copper or that multiple mechanisms may be involved. Additional studies are needed to determine the source of renal copper in dogs with CAH.
On the basis of the previous descriptions of concurrent Fanconi syndrome and CAH, it would seem that copper chelation is essential for resolution of renal tubular disease.[13, 14] Studies of Wilson's disease also support this assumption. However, in our series of dogs, resolution of urinalysis abnormalities (proteinuria and glucosuria) occurred in 2 dogs without copper chelation. In another surviving dog, proteinuria and glucosuria had nearly resolved by the time copper chelation had commenced. These findings suggest that renal tubular dysfunction may resolve with or without copper chelation. There are several plausible explanations including dietary intervention and effects of other pharmacologic agents. All dogs were fed a low copper diet which has been shown to decrease hepatic copper concentrations in Labradors with CAH. However, gradual reduction by dietary restriction is unlikely to be the sole explanation. All dogs were treated with antioxidants, which potentially could ameliorate the oxidative damage induced by excess tissue copper. The use of steroids in 2 of the dogs may have decreased hepatic inflammation and decreased the amounts of copper released from the damaged liver. Alternatively, steroids may have directly decreased renal inflammation, but there was only mild inflammation initially. Finally, administration of ursodeoxycholic acid to 3 dogs may have either altered tissue copper concentrations or blunted the deleterious effects of excess tissue copper by increasing bile flow, inhibiting apoptosis, modulating immune responses, or increasing glutathione and metallothionein production. Although treatment was varied and sometimes nonspecific, some type of treatment is likely necessary for the resolution of copper-induced tubular disease. This hypothesis is supported by the observation that 3 of these dogs had been glucosuric for weeks, and improvement was not seen until medical therapy was initiated. However, the most important aspect of treatment remains unknown as all dogs received multiple therapeutic interventions. None of the surviving dogs required alkali administration or other specific treatments for their tubulopathy.
The prevalence of renal tubular disease and its relation to the prognosis of CAH in dogs is unknown. On the basis of increases in bilirubin concentrations in all 9 dogs, it is possible that development of renal tubular dysfunction is associated with more severe or advanced liver disease. This is further supported by the fact that all dogs had overt clinical signs relating to liver disease as opposed to serendipitous discovery of increased liver enzyme activity. An excellent long-term prognosis has been previously reported for dogs with concurrent CAH and Fanconi syndrome.[13, 14] This somewhat contrasts with our series in which 4 of 9 dogs died near the time of diagnosis. However, some dogs did have long-term survival with resolution of clinical renal and hepatic disease. Additional studies are needed to more accurately define the prognostic importance of renal tubular dysfunction in dogs with CAH. Even if renal tubular abnormalities correlate with more advanced disease, a favorable prognosis still is possible in affected dogs.
In summary, concurrent renal tubular dysfunction and CAH is likely more common than previously recognized (33% of the MSU-VTH Labradors reviewed for inclusion in this manuscript), although the overall incidence remains unknown. Detection of renal tubular abnormalities in a Labrador Retriever with biochemical evidence of liver disease is strongly suggestive of CAH and indicates the need for hepatic histopathology. Although renal tubular lesions are present in affected dogs and appear to be associated with copper accumulation, the necessity of renal histopathology is less clear as it is unlikely to alter patient management. The exact mechanism of renal copper accumulation and injury remains unknown and requires additional study. Despite the association of renal tubular dysfunction with more advanced disease, successful long-term outcome is possible in these dogs with a variety of therapies with or without d-penicillamine.
The authors acknowledge Dr Andreas Lehner for his assistance with quantitative copper determinations. This study was supported by the Michigan State University College of Veterinary Medicine Trinket Fund.
Conflict of Interest Declaration: Authors disclose no conflict of interest.
Advia 120 Hematology System, Siemens Healthcare, Deerfield, IL
Olympus AU640e, Olympus America Inc, Center Valley, PA
Nova Biomedical Stat Profile, Nova Biomedical, Waltham, MA
IQ pH meter, IQ150-77, IQ Scientific Instruments, Carlsbad, CA
Multistix Reagent Strips, Siemens Healthcare Diagnostics Inc, Tarrytown, NY
Reichert TS meter, 10406, Cambridge Instruments Inc, Buffalo, NY
Urine metabolic profiles were performed by the Metabolic Genetic Disease Testing Laboratory, School of Veterinary Medicine, University of Pennsylvania
0, no copper; 1, solitary liver cells and/or reticulohistiocytic cells containing some copper positive granules; 2, small groups or areas of liver cells contain small to moderate amounts of copper positive granules; 3, larger groups of liver cells contain moderate amounts of copper positive granules, sometimes associated with copper containing macrophages; 4, large areas of liver cells and cells of the reticulohistiocytic system (RHS) with many copper positive granules, usually associated with copper containing macrophages; 5, diffuse panlobular presence of liver cells and cells of the RHS with many copper positive granules, usually associated with copper containing macrophages
Quantitative copper concentrations were determined on a dry matter basis for all hepatic and renal specimens. However, the MSU-DCPAH only has reference ranges for renal copper concentrations on a wet matter basis (5–15 ppm). In converting from wet matter to dry matter, analytical chemists and toxicologists at the MSU-DCPAH suggested using correction factors ranging from 3.3 to 5, resulting in an approximate reference range of 16–75 ppm. This range is similar to the reference renal copper concentrations (17.5–60 ppm) reported by Appleman et al.
Ursodiol, Watson Laboratories, Corona, CA
Hill's Pet Nutrition Inc, Topeka, KS
Royal Canin USA Inc, St. Charles, MO
Denamarin, Nutramax Laboratories Inc, Edgewood, MD