Introduction
Copper is a vital micronutrient in mammals essential for numerous metabolic functions.1,2 The ability to serve as an electron donor or acceptor allows copper to function as a transition metal crucial to oxidation-reduction reactions.3,4 The liver is central to systemic copper homeostasis, a process regulated by a complex network of proteins.1,2,5 Dietary copper is enterically absorbed and circulates protein-bound, with hepatocyte uptake mediated by membrane copper transporters. In hepatocytes, copper binds with storage proteins, copper-dependent enzymes (ie, superoxide dismutase and cytochrome oxidase), and several chaperone proteins with excess copper exported in canalicular bile.1–5 Hepatic copper accumulation can be deleterious, driving oxidative injury that leads to apoptotic hepatocyte death and copper-associated hepatitis.3–6 Cellular injury prominently associates with lysosomal copper accumulation (releasing autodigestive enzymes) and injury to mitochondrial membranes and DNA, limiting energy generation and exhausting essential organelle and cytosolic antioxidant defenses (ie, glutathione and superoxide dismutase).2–5
Pathologic copper accumulation designated as a primary copper-associated hepatopathy reflects either genetic mutations disrupting copper homeostasis or excessive copper intake (ie, dietary or environmental). Secondary hepatic copper accumulation reflects copper accumulation subsequent to chronic cholestasis; this only occurs in species vulnerable to chronic cholestatic syndromes (ie, humans and domestic cats with chronic immune-mediated cholangiohepatitis but not dogs).7–11 Because ferret hepatobiliary disorders are similar to conditions common in the domestic cat, chronic cholestasis might contribute to hepatic copper accumulation in some ferrets.12
The best investigated genetic causes of primary copper-associated hepatopathy are Wilson disease in humans (ATP7β mutations), an animal model of this disorder (Long Evans cinnamon rat), and Bedlington Terrier copper-associated hepatopathy (COMMD1 mutation).5,13,14 The net impact of these mutations is impaired biliary canalicular copper egress. Primary copper-associated hepatopathy has been speculated to involve an ATP7β mutation in a few cats.15,16 A genetic cause of copper-associated hepatopathy in ferrets has been suspected in 4 animals, with centrilobular hepatic copper distribution in each case, similar to dogs and cats with primary copper-associated hepatopathy.7,10,17–21 Copper- associated hepatopathy has only recently been considered in ferrets. We speculate that this diagnosis has been overlooked because requisite copper-specific biopsy specimen staining and hepatic copper quantification are infrequently explored in this species.
Nutritional oversupplementation of copper subsequent to modification of food-grade copper supplements in commercial dog foods (mid 1990s) corresponds with an increased incidence of copper-associated hepatopathy in mixed-breed and purebred dogs.7 In predisposed breeds, naïve selection of hitchhiking or flanking genes adjacent to determinants of breed characteristics has likely influenced copper homeostasis, reducing physiologic tolerance to food-borne bioavailable copper.7,17,22–24 The scenario of diet-provoked liver injury in predisposed individuals has been experimentally documented in a rodent model of Wilson disease..6 Chronic copper ingestion secondary to environmental contamination was surmised in 1 cat housed on the site of an old copper mine (ie, presumed waterborne and self-grooming copper ingestion).10 Diet-related hepatic copper accumulation has been suspected in some bat species and sugar gliders but has not yet been investigated in ferrets with pathologic copper accumulation.25,26
Liver copper concentration is commonly measured by atomic absorption spectroscopy (AAS) by use of fresh, frozen, or neutral-buffered 10% formalin–fixed tissue.19,27,28 Alternatively, rhodanine-stained liver sections can be used to quantify liver copper by digital scanning.7,19,27 Advantageously, digital scanning can evaluate all histologic sections as opposed to randomly selected single samples forfeited for AAS or other bench copper analytic methods.7,19,27 Digital scanning mitigates the influence of irregular tissue copper distribution shown to confound reliability of single sample analyses.19 Indeed, digital scanning copper quantification has better correlation with rhodanine-based copper scores compared to AAS.19 Although retrospective study of hepatic copper concentrations can be orchestrated by bench or digital scanning methods, AAS and other bench analytic methods can deplete small samples, limiting further histologic investigations.19,28 Regardless of analytic method, hepatic copper concentrations should be reconciled with a standardized qualitative copper score that characterizes the severity and zonal distribution of copper and its proximity to necroinflammatory lesions.7 This action surveils for erroneous copper measurements.
The present study had several objectives, including validation of the rhodanine-based digital scanning method for hepatic copper quantification in ferret liver; determination of associations between signalment, clinical, or clinicopathologic features and hepatic copper accumulation that might discriminate animals with copper-associated hepatopathy; comparison of hepatic copper concentrations in ferrets with and without hepatobiliary disease; characterization of zonal distribution of accumulated copper in ferret hepatic samples; and assessment of the correlation between qualitative rhodanine-based copper scores (canine method) and hepatic copper concentration in ferret hepatic samples.
Materials and Methods
Case selection
Histologic reports were identified from necropsies with hepatic sample evaluations (n = 27) and liver biopsies (7) of ferrets from 2010 through 2018 (Northwest ZooPath, Zoo/Exotic Pathology Service, 31; Cummings School of Veterinary Medicine at Tufts University, 3). Study inclusion required access to a record documenting clinical and clinicopathologic features and availability of archived hepatic tissue for preparation of sections for H&E and rhodanine staining.
Data collection
Data transcribed from medical records included signalment, clinical signs, clinicopathologic features (hematologic findings, including Hct or PCV; WBC count; serum biochemical findings, including concentrations total protein, albumin, BUN, creatinine, glucose, total bilirubin, and cholesterol; and activities of alanine aminotransferase [ALT], aspartate aminotransferase, alkaline phosphatase [ALP], and γ-glutamyltransferase [GGT]), diet, and timing of biopsy (antemortem or postmortem). Too few radiographic or ultrasonographic imaging studies were described for meaningful analyses and thus were not included in the study.
Tissue staining
Formalin-fixed, paraffin-embedded hepatic samples were retrieved from the archives of referring institutions (Northwest ZooPath, Zoo/Exotic Pathology Service and Cummings School of Veterinary Medicine), sectioned at 5 μm, stained with H&E, and examined for the presence or absence of hepatic disease. Samples classified as hepatobiliary disease were categorized into 3 groups: necroinflammatory (NI) liver disease, hepatocellular carcinoma (HCCA), or non-necroinflammatory (NNI) liver disease. A fourth group was designated as lacking hepatobiliary disease. In the studied population, features diagnostic for NI liver disease included presence of suppurative or nonsuppurative ductcentric inflammation classified as cholangitis and as cholangiohepatitis if inflammatory infiltrates breached the limiting plate causing regional hepatocyte injury, hepatocellular necrosis (focal or multifocal necrotic or apoptotic hepatocytes [exceeding 1 dead cell/ten 400X fields of view]), cholecystitis (mucosal or mural gallbladder inflammatory infiltrates [suppurative or nonsuppurative]), or ductal plate malformation (DPM) complicated by inflammatory infiltrates (dispersed within fibroductal DPM trabeculae). Diagnosis of DPM required identification of malformative bile duct silhouettes or cystic ductal structures embedded in exuberant extracellular matrix or fibroductal trabeculae populated by enumerable malformative bile duct profiles.29 Diagnosis of HCCA was based on identification of a grossly evident mass lesion populated by cuboidal to pleomorphic hepatocytes with variable anisocytosis, anisokaryosis, polynucleation, prominent or multiple nucleoli, loss of sinusoidal and hepatic cord structure or with cells arranged in acini or sheet configurations and regionally compressive mass impact on adjacent nonneoplastic parenchyma. Histologic features designated as NNI liver disease in the studied population included diffuse hepatic lipidosis without inflammatory foci, metastatic neoplastic lesions, DPM lacking inflammatory infiltrates, and extreme infiltrative extramedullary hematopoiesis. Samples designated as lacking hepatobiliary disease had no evidence of architectural remodeling, hepatocyte necrosis, cholangitis, cholangiohepatitis, HCCA, metastatic neoplasia, canalicular cholestasis, or mechanical bile duct obstruction. Biopsies with only minor and inconsistent portal lymphocytic infiltrates (ie, < 15 inflammatory cells/medium-sized portal tract), without portal tract expansion by fibrosis or inflammatory infiltrates, with absence of a ductcentric orientation or ductal exocytosis, and without limiting plate interruption were considered secondary nonspecific phenomenon. Such biopsies were classified as lacking hepatobiliary disease. This process was considered to represent inflammatory cells trafficking the portal region from the splanchnic circulation and has previously been cited as a common finding in ferrets.30,31
Rhodanine staining was applied for copper discrimination. Independent review of H&E-stained histologic sections was completed by 3 of the authors (DRR, MMG, and SAC) with findings reconciled into a consensus diagnosis. Histologic lesions in hepatic sections were transcribed to a spreadsheet and thereafter were subcategorized into the 4 hepatobiliary disease groups: NI hepatic disease, HCCA, NNI hepatic disease, and no hepatobiliary disease. Ultimately, ferrets with hepatobiliary disease also were consolidated into 1 group for comparisons with ferrets lacking hepatobiliary disease.
Rhodanine-stained hepatic sections inspected by one of the authors (SAC) were used to assign a qualitative copper score based on a system developed for and routinely applied to canine hepatic samples.17,19,27 The acinar zonal distribution of stainable copper was recorded from these samples.17,19,27 Hepatic copper concentration was measured by AAS (Fort Collins Diagnostic Laboratory, Colorado State University,) or rhodanine-based digital scanning and expressed as μg/dL of dry weight liver (DWL). Samples undergoing AAS analysis used for digital scanning validation had evenly dispersed liver copper on microscopic inspection of rhodanine-stained sections. These samples weighed ≥ 25 mg and included fresh, formalin-fixed, and paraffin-embedded xylene-extracted tissues. The digital method was validated for ferret liver by use of a Leica-Aperio proprietary positive pixel algorithm10,17,27 (Aperio SanScope CS and Aperio version 8.1; Aperio Technologies Ltd) and linear regression developed from 17 dually analyzed samples spanning 63 to 1,720 µg/g of DWL by AAS analysis.
Statistical analysis
Although sex-specific reference intervals are reported for hematologic and biochemical parameters for male and female ferrets, there is considerable overlap in reported ranges.32 We applied reference intervals for male ferrets arbitrarily. Because serum biochemical assessments were completed by use of different analytic methods and equipment, values were compared to the reference range for male ferrets. Subsequently, values were reported as a fold-change against the highest or lowest values of the respective reference intervals as appropriate for the considered analyte.
Validation of the rhodanine-based digital scanning method was completed by use of linear regression of AAS-measured hepatic copper against digitally scanned data with determination of linear fit (R2), a simultaneous (Working-Hotelling) 95% CI, and 95% predictive interval. Validation was completed by use of 17 samples with homogenous microscopic copper distribution that were measured by AAS grouped as low (63 to 189 µg/g of DWL), medium (202 to 310 µg/g of DWL), high (726 to 828 µg/g of DWL), and extreme (1,450 to 1,720 µg/g of DWL) copper concentrations.
Age, clinicopathologic, and hepatic copper data were evaluated for normality of distribution by use of histograms and the Kolmogorov-Smirnov test, revealing a nonparametric data distribution. Consequently, data are consistently reported in this manuscript as median and range values. The proportion of ferrets with laboratory parameters below, within, or above reference limits was determined. Sex distribution per disease category was compared against a hypothetical population having an equivalent sex distribution by use of 2 X 2 tables and the Fisher exact test. Sex distributions between ferrets with and without hepatic disease were similarly compared against each other.
Initial comparisons of age, clinical features, clinicopathologic parameters, and hepatic copper concentrations between disease groups were performed by use of Kruskal-Wallis ANOVA. Because limited differences were found, the Wilcoxon rank sum test was used to compare individual categories and a combination hepatobiliary disease group against ferrets lacking hepatobiliary disease. The Wilcoxon rank sum test was also used to compare variables between ferrets with and without hepatic copper concentrations > 1,000 µg/g of DWL, which was considered an extreme increase in liver copper concentration.
Correlations between age and hepatic copper concentration and between rhodanine-based liver copper scores and hepatic copper concentration determined by AAS and by rhodanine digital scanning were determined by calculating the Spearman rank correlation coefficient (ρ). Values of P < 0.05 (2-tailed test) were considered significant, and SAS (version 9.4; SAS Institute Inc), Statistix (version 9; Analytical Software), and Analyse-it (Analyse-it Software Ltd) were used for all analyses. Dietary intake was compared, by inspection, to a list of diets recently analyzed for copper concentrations that are commonly fed to ferrets.33
Results
Validation of rhodanine-based digital copper quantification
Validation of the rhodanine-based digital scanning method of copper quantification in ferret hepatic samples (Figure 1) resulted in an equation for the calculation of liver copper concentration (110 + 75,580X, where 110 represents the constant and X represents positive pixel summation; R2 = 0.98; P < 0.001).
Results of linear regression for rhodanine-based digital quantification of hepatic copper concentrations in ferrets. Linear fit = 110 + 75,580X (solid black line; R2 = 0.98; P < 0.001). The dotted blue lines denote the 95% CI, and the dashed red line denotes the 95% prediction interval.
Citation: American Journal of Veterinary Research 83, 4; 10.2460/ajvr.21.08.0131
Samples
Of 34 ferrets that met the inclusion criteria, 8 were from zoological institutions and 26 were companion animals. Of 34 ferret liver samples studied, 27 were collected postmortem and 7 were collected antemortem. Among antemortem samples, 4 were collected because of clinicopathologic evidence of hepatic disease, whereas 3 were collected because of gross hepatic abnormalities observed during exploratory laparotomy in patients with health concerns unrelated to hepatic disease.
Hepatobiliary diagnosis
Diagnoses among ferrets with hepatobiliary disease included 7 with NI hepatic disease (5 with cholangiohepatitis, and 2 with cholecystitis), with 4 having a concurrent DPM; 4 with HCCA; and 14 with NNI hepatic disease (8 with secondary or metastatic neoplasia, 4 with DPM without inflammation, 1 with hepatic lipidosis, and 1 with extreme hepatic extramedullary hematopoiesis). The group of 9 ferrets lacking histologic evidence of hepatobiliary disease included 3 ferrets with pancreatic islet cell tumors, 2 with adrenal adenomas, and 1 each with acute interstitial pneumonia and gastritis, enteric lymphoma, mesenteric torsion, systemic ferret coronavirus infection, intestinal coccidiosis with severe gastric and duodenal ulcers and septic shock, thymoma causing myasthenia gravis, and gastrointestinal carcinoma (several of these ferrets had more than 1 comorbidity). Antemortem liver samples were collected from 2 ferrets with HCCA, 1 ferret with NI liver disease, 2 ferrets with NNI liver disease, and 2 ferrets lacking hepatobiliary disease.
Ferrets with hepatobiliary disease were significantly (P < 0.001) older than ferrets lacking hepatobiliary disease, with a median age of 6.0 years (range, 3.0 to 9.0 years) versus 4.0 years (range, 0.2 to 6.0 years), respectively. While each categorical hepatic disease group was also significantly (all P ≤ 0.01) older than ferrets lacking hepatobiliary disease, there were no significant age differences between hepatobiliary disease categories. There was also no significant age difference between ferrets with and without a hepatic copper concentration > 1,000 µg/g of DWL, with a median age of 6.0 years (range, 1.0 to 9.0 years) versus 5.8 years (range, 0.2 to 9.0 years), respectively. There was no significant sex predisposition among ferrets with hepatobiliary disease (10 males [8 castrated and 2 sexually intact] and 15 females [10 spayed and 5 sexually intact]) and no significant difference in reproductive status from a normally distributed population. However, because all ferrets without hepatobiliary disease were male (6 castrated and 3 sexually intact), there was a significant (P = 0.004) difference in sex distribution between ferrets with and without hepatobiliary disease due to a male control population bias. Seven ferrets (7/34 [21%]) with a hepatic copper concentration > 1,000 µg/g of DWL included 4 castrated males and 3 females (2 spayed and 1 sexually intact); thus, there was no evidence of sex predilection for extreme hepatic copper accumulation.
Common clinical features in ferrets with hepatobiliary disease included lethargy (10/25 [40%]), weight loss (7/25 [28%]), and weakness (7/25 [28%]; Supplementary Table S1). Four ferrets developed abdominal effusion, 2 were jaundiced, and 2 had no clinical signs. No significant differences in clinical features existed between ferrets with HCCA or NI or NNI hepatobiliary disease with the exception of a significantly (P = 0.03) greater proportion of ferrets with NI hepatobiliary disease demonstrating weight loss, compared with ferrets with NNI hepatobiliary disease, and a significantly (P = 0.02 each) higher proportion of ferrets with HCCA demonstrating abdominal pain, compared with NI and NNI hepatobiliary disease. Clinical features in 7 ferrets with a hepatic copper concentration >1000 µg/g of DWL included weakness (n = 3), diarrhea (2), lethargy (2), and weight loss, hyporexia, abdominal pain, jaundice, and no clinical signs (1 each). Thus, clinical features could not be used to predict extreme hepatic copper accumulations.
Clinicopathologic features broadly overlapped among ferrets with hepatobiliary disease (Table 1; Figure 2) and failed to distinguish disease categories or ferrets with hepatic copper accumulation. There were no relevant differences in hematologic parameters between ferrets with and without hepatobiliary disease. Among hepatic disease subcategories, significant (P = 0.04) differences in median fold increase in serum ALT activity existed between ferrets with NI and NNI hepatobiliary disease with a median fold increase of 2.2 (range, 0.6 to 4.5) versus 1.1 (range, 0.2 to 3.2), respectively, and in serum BUN concentration between ferrets with HCCA and NNI hepatobiliary disease with a median fold increase of 1.1 (range, 0.8 to 1.3) versus 0.5 (range, 0.3 to 1.0), respectively. However, observed differences were too small and inconsistent among individuals to provide predictive diagnostic utility. Compared with ferrets without hepatobiliary disease, those with NI hepatic disease had a significantly (P = 0.02) higher median fold increase in serum creatinine concentration (0.9 [range, 0.4 to 3.1] vs 0.4 [range, 0.3 to 0.4], respectively) and significantly (P = 0.03) higher median fold increase in serum ALP activity (0.8 [range, 0.1 to 8.6] vs 0.2 [range, 0.1 to 0.7], respectively). However, the higher ALP activity could not provide predictive utility for detection of NI hepatobiliary disease because of its imperfect occurrence.
Relevant clinicopathologic abnormalities and hepatic copper concentrations for ferrets with and without hepatobiliary disease. Clinicopathologic abnormalities are reported as number of ferrets with abnormal values over the number tested.
| Ferret group | ↑ BUN | ↑ ALT | ↑ ALP | ↑ GGT | ↑ TB | Median (range) copper concentration* |
|---|---|---|---|---|---|---|
| Necroinflammatory liver disease | 2/4 | 4/5 | 1/5 | 3/3 | 2/5 | 151 (112–2,061) |
| Hepatocellular carcinoma | 2/3 | 2/4 | 3/4 | 4/4 | 0/5 | 5,308 (193–11,411) |
| Non-necroinflammatory liver disease | 0/10 | 5/10 | 3/10 | 3/4 | 4/10 | 169 (115–1,329) |
| All hepatobiliary disease | 4/17 | 11/19 | 7/19 | 10/11 | 6/19 | 172 (112–11,411) |
| No hepatobiliary disease | 0/3 | 0/4 | 0/4 | 2/2 | 1/4 | 271 (131–2,867) |
= Parameter exceeded the relevant upper reference limit. *Values represent µg/g of dry weight liver.
ALP = Alkaline phosphatase activity. ALT = Alanine aminotransferase activity. GGT = γ-Glutamyltransferase activity. TB = Total bilirubin concentration.
Dot plots of serum analytes and hepatic copper concentrations in ferrets with (n = 25) and without (9) hepatobiliary disease. ALP = Alkaline phosphatase activity. ALT = Alanine aminotransferase activity. GGT = γ-Glutamyltransferase activity. HCCA = Hepatocellular carcinoma (n = 5). NI = Necroinflammatory liver disease (n = 7). NNI = Non-necroinflammatory liver disease (n = 13).
Citation: American Journal of Veterinary Research 83, 4; 10.2460/ajvr.21.08.0131
Comparison of clinicopathologic parameters in ferrets with (3 subcategories combined) and without hepatobiliary disease demonstrated a significantly (P = 0.04) higher fold increase in median serum creatinine concentration (0.7 [range, 0.2 to 3.1] vs 0.4 [range, 0.3 to 0.4], respectively), significantly (P = 0.02) higher median fold increase in serum activity of ALP (0.8 [range, 0.2 to 8.6] vs 0.2 [range, 0.2 to 0.7], respectively), and significantly (P = 0.03) higher median fold increase in serum GGT activity (14.2 [range, 0.6 to 101.6] vs 3.2 [range, 0.6 to 4.2]), respectively) in ferrets with hepatobiliary disease. Differences in hepatic enzyme activity were substantial enough to provide diagnostic utility for suspecting hepatobiliary disease but were not specific for copper accumulation.
Seven ferrets with a hepatic copper concentration > 1,000 µg/g of DWL (3 with HCCA) had a significantly (P = 0.04) higher median fold increase in serum ALT activity, compared with ferrets without severe hepatic copper accumulation (3.3 [range, 0.9 to 6.5] vs 0.6 [range, 0.2 to 3.3], respectively). However, in a clinical circumstance the magnitude of change would not be sufficient to differentiate the underlying hepatic disorder without liver biopsy.
Histologic features of hepatobiliary disease
The most common histologic feature among ferrets with hepatobiliary disease (19/25 [76%]) was a variable (mild to severe) lymphocytic infiltrate constrained within the margins of portal tracts. Three ferrets with NI hepatic disease had lymphocytic infiltrates that breached the limiting plate consistent with cholangiohepatitis. Lymphocytic portal infiltrates were absent in 6 ferrets with hepatobiliary disease (4 with NNI hepatic disease and 2 with HCCA), documenting the inconsistency of finding relevant portal lymphocytic infiltrates. Seven ferrets classified as lacking hepatobiliary disease displayed mild lymphocytic portal infiltrates consistent with lymphocytes moving to the liver from the splanchnic circulation. Finding necrotic or apoptotic hepatocytes was rare; only 6 of 25 (24%) ferrets with hepatobiliary disease and only 2 ferrets with a hepatic copper concentration > 1,000 µg/g of DWL demonstrated necrotic or apoptotic hepatocytes. Necrotic or apoptotic hepatocytes in these ferrets did not obviously coordinate with regional copper accumulation.
Other hepatic histologic features that were variably encountered among studied ferrets included multifocal and rarely diffuse glycogen-type hepatocyte vacuolation (6 with and 4 without hepatobiliary disease), foci of micro- or macrovesicular hepatocyte lipid vacuolation (8 with hepatobiliary disease, including 1 with diffuse hepatic lipidosis and 2 ferrets lacking hepatobiliary disease), and prominent extramedullary hematopoiesis in portal tracts (5 ferrets with hepatobiliary disease, 3 of these also displayed centrilobular extramedullary hematopoiesis).
Hepatic copper in ferrets with and without hepatobiliary disease
The range of hepatic copper concentrations broadly overlapped among ferrets with and without hepatobiliary disease (Table 1; Figure 2). Hepatic copper concentration exceeded feline (< 180 µg/g of DWL) and canine (< 400 µg/g of DWL) reference limits in 19 (56%) and 9 (27%) of 34 ferrets, respectively. Hepatic copper concentrations > 1,000 µg/g of DWL occurred in 5 ferrets with and 2 without hepatobiliary disease (7/34 [21%]). Among ferrets with hepatobiliary disease, the median hepatic copper concentration in neoplastic tissue in HCCA (5,038 [range, 193 to 11,411 µg/g of DWL]) was significantly higher than that of the NI (151 [range, 112 to 2,061 µg/g of DWL]; P = 0.04) and NNI hepatobiliary disease (169 [range, 115 to 1,329 µg/g of DWL]; P = 0.008) categories. However, median hepatic copper concentration was not significantly different between ferrets with and without liver disease (172 [range, 112 to 11,411 µg/g of DWL] vs 271 [range, 131 to 2,867 µg/g of DWL], respectively).
Hepatic copper zonal distribution
When visible with rhodanine staining, copper distribution was predominantly centrilobular. However, in ferrets with severe copper accumulation (qualitative copper score of 4 to 5 with copper concentration > 1,000 µg/g of DWL), copper also often had a lesser distribution in midzonal (zone 2) and periportal (zone 1) regions (Figure 3). A marked tropism for hepatic copper distribution was documented in 2 ferrets with HCCA in which neoplastic tissue harbored 8,274 and 11,411 µg/g of DWL, compared with 204 and 584 µg/g of DWL in nonneoplastic liver.
Photomicrographs of rhodanine-stained liver sections from a ferret with primary copper hepatopathy (A) and a ferret with hepatocellular carcinoma (B). A—Vivid orange-red cytosolic copper accumulation is visible in centrilobular (CV; zone 3) hepatocytes, with comparably devoid portal tracts (PT). Bar = 500 µm. B—Notice the severe hepatocyte copper accumulation within the neoplasm (orange-red stippled lesion), compared with the starkly less affected adjacent nonneoplastic tissue (confluent pale-blue staining). Bar = 9 mm.
Citation: American Journal of Veterinary Research 83, 4; 10.2460/ajvr.21.08.0131
Correlation of rhodanine-based copper scoring and hepatic copper concentration
Rhodanine-based histologic copper scores (scoring of 0 to 5)17,19 were strongly associated with hepatic copper concentrations determined by AAS and digital scanning (ρ = 0.92 and 0.88, respectively; P < 0.001). Seventeen liver biopsies with homogenous copper distribution had AAS copper quantification (validation samples), and all liver biopsies had digital copper quantification.
Nine ferrets (4 with NNI hepatobiliary disease, 2 with NI hepatobiliary disease, 1 with HCCA, and 2 without hepatobiliary disease) were fed several diets with copper concentrations > 6.12 mg of copper/1,000 kcal.33 However, 7 of these 9 ferrets were also fed an unspecified blend of different diets, including some with unknown copper concentrations. This situation compromised comparisons among ferrets based on categorized dietary intake.
Discussion
We achieved targeted objectives declared for this study including validation of a rhodanine-based digital scanning method for copper quantification that is routinely used in diagnosis of copper-associated hepatopathy in dogs and cats. We confirmed a strong correlation between a rhodanine-based copper scoring system and hepatic copper concentrations determined by digital scanning. These achievements provide an investigative platform for future prospective studies of hepatic copper accumulation in ferrets, where small sample size can restrict histologic evaluations. As there are no established reference limits for hepatic copper concentrations in clinically healthy ferrets and only a few case reports20,21 describing copper-related hepatic injury in this species, the digital scanning methodology should permit expansion of this limited database.
We documented a wide range of hepatic copper concentrations in ferrets with and without hepatobiliary disease but were unable to distinguish clinical features exclusively implicating pathologic hepatic copper accumulation at levels considered toxic to dogs and cats (values > 600 µg/g of DWL).7,10 Considering the wide cross-species vulnerability to copper hepatotoxicosis and the absence of conspicuous features implicating liver injury associated with liver copper accumulation in ferrets, we speculate that ferrets may have an innate greater tolerance for hepatic copper accumulation than is recognized in dogs or cats.7,34 Indeed, among 34 ferrets with and without hepatobiliary disease that were studied, we demonstrated a > 50% incidence of hepatic copper concentrations exceeding the upper reference limit for cats and 29% exceeding the upper reference limit for dogs.7,10 Although there was no significant difference in median hepatic copper concentration between ferrets with and without hepatobiliary disease, most ferrets with copper concentrations > 1,000 µg/g of DWL had some form of liver disease (including HCCA). We were unable to define an upper reference limit for clinically healthy ferrets from the present study because ferrets without hepatobiliary disease had nonhepatic health issues that led to liver sample collection, most commonly at necropsy. Thus, ferrets without hepatobiliary disease usually had some nonhepatic organ disturbance or multisystemic disorder that might have influenced hepatic copper accumulation (ie, inflammatory cytokines or oxidative injury impacting the liver, even though such impact was not histologically evident). We also could not determine whether sex influenced hepatic copper accumulation because all ferrets lacking hepatobiliary disease were male, biasing comparisons. A study specifically designed to establish reference limits for hepatic copper concentrations in clinically healthy ferrets over a spectrum of sex and age distributions is needed and could be pursued by use of the rhodanine-based copper method validated herein.
Ferrets with hepatobiliary disease displayed nonspecific clinical signs, with lethargy and weight loss most common but not exclusive to liver disease. Thus, clinical features did not discriminate ferrets with hepatobiliary disease from those without hepatobiliary disease. Although ferrets with hepatobiliary disease were significantly older than ferrets without hepatobiliary disease, the small population studied and lack of samples from healthy ferrets precluded conclusions that older age influenced development of hepatic disease, accrual of hepatic copper, or clinicopathologic metrics. Among clinicopathologic features, significantly higher normalized serum creatinine concentration and cholestatic enzyme (ALP and GGT) activities occurred in ferrets with hepatobiliary disease. While the finding of increased cholestatic hepatic enzymes is consistent with our expectations (ie, ferrets develop cat-like cholangitis and cholangiohepatitis), the small number of ferrets without hepatobiliary disease in this study limited strong conclusions. As ferrets with hepatobiliary disease were significantly older than ferrets without hepatobiliary disease, it is possible that older age in diseased animals may have influenced development of renal dysfunction. However, a broader population study should be conducted to further investigate this association.
Similar to hepatic copper distribution in dogs and cats with primary copper-associated hepatopathy, hepatic copper in ferrets accumulates predominantly in centrilobular regions, with lesser amounts in the intermediate zone (ie, zone 2).10,17,19,27 Cats with cholestatic disease, unlike dogs, may accumulate rhodanine-stainable copper in periportal regions similar to humans with chronic cholestasis.8–10 Presumably this reflects the chronicity of cholestatic hepatic injury in the nonsuppurative feline cholangitis or cholangiohepatitis syndrome, with which survival for years is common. Chronic survival of humans with cholestatic disorders also similarly influences a periportal copper distributional pattern.8,9 However, for dogs in which chronic portal triad-focused injury and chronic cholestasis are comparably uncommon, periportal copper distribution is exceedingly rare.7,11 Despite inclusion of ferrets with cholangiohepatitis in the present study, we found no evidence that cholestasis influenced periportal copper accumulation, albeit the studied population was small.
Lymphocytic infiltrates within the confines of the portal tract without cholangiocyte targeting were common in ferrets with and without hepatobiliary disease. Although this has been proposed as a common and normal histologic feature of ferret liver, some authors theorize that such infiltrates reflect chronic lymphocytic portal hepatitis associated with ascending inflammation from the alimentary tract or pancreas.12,31
It is possible that excessive dietary or water-borne copper intake may contribute to hepatic copper accumulation in ferrets, as reported in rabbits, dogs, and cats.7,10,17,24,35 Unlike the situation in dogs and cats, there are no Association of American Feed Control Official guidelines specifying ferret dietary copper recommendations. Findings of a recent study33 analyzing 10 different ferret diets calculated an approximate 7-fold greater copper intake than the Association of American Feed Control Official–recommended minimum intake for cats (1.25 mg copper/1000 kcal, as fed).36,37
Because diets specified for mink are commonly fed to ferrets, it is relevant to acknowledge that such diets deliver a copper content similar to products formulated for felines.33,37 Thus, considering our finding that > 50% of studied ferrets had hepatic copper concentrations exceeding dog or cat reference limits and with extreme concentrations (> 1,000 µg/g of DWL) in 21% of the studied population, an investigation of ferret copper requirements is necessary to adjudicate the safety of current husbandry practices. Unfortunately, inconsistency of allocated foods and use of blended diets in ferrets in the present study precluded inspection for associations between dietary formulations and hepatic copper accrual.
Although the prevalence of HCCA in ferrets is unknown, it has been proposed to be relatively common.38,39 In the present study, the prevalence of HCCA was 15%. It was intriguing that hepatic copper concentrations were markedly increased in 3 of 4 ferrets with HCCA (1,797, 8,278, and 11,411 µg/g of DWL). Finding an association between HCCA and hepatic copper accumulation is not unprecedented, as this has been noted in some humans with nonoptimally managed Wilson disease, in the rodent model of this disease (Long-Evans cinnamon rats), and in a few dogs, cats, and sugar gliders with copper-associated hepatopathy.10,26,40–47 It is unknown whether evolution of HCCA alters hepatocyte phenotype allowing copper accumulation or neoplastic transformation emerges in a copper-laden microenvironment (due to chronic oxidative injury, impaired enzyme activity, nucleoprotein strand breaks, and chronic inflammation).46,47 Two ferrets with HCCA had the highest hepatic copper concentrations in the present study. In these, neoplastic hepatocytes showed propensity for copper accumulation implicating phenotypic diversity of neoplastic cells, a phenomenon previously noted in a small number of cats and dogs.10,44,45
Weaknesses of the present study included the limited statistical power imposed by the small sample size, inclusion of only male ferrets with a limited age spectrum in the group of animals lacking hepatobiliary disease, and unavailability of a reference interval for hepatic copper concentrations in clinically healthy ferrets. Because the included ferrets were initially evaluated by a variety of veterinary and zoological institutions with different in-house chemistry analyzers or outside laboratory access, not all ferrets had clinicopathologic assessments. Variability between laboratory analyses (ie, analytic methods and equipment) compromised direct comparison of laboratory results, necessitating normalization of diagnostic parameters based on upper reference limits.
In conclusion, the present study revealed a spectrum of hepatic copper concentrations in ferrets with and without hepatobiliary disease and a propensity for centrilobular copper accumulation. We documented hepatic copper concentrations in 50% of studied ferrets exceeding the upper reference limits for hepatic copper in canine or feline liver but found that clinical or clinicopathologic features did not exclusively implicate pathologic liver copper accrual. The curious association between severe copper accumulation and HCCA in ferrets should be more fully investigated, and studies are needed to establish reference limits for hepatic copper concentrations in clinically healthy male and female ferrets among a spectrum of ages. The potential exists for an association between hepatic copper accrual and dietary copper intake in ferrets, and this possibility warrants further appraisal.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org
Acknowledgments
Funded by the Companion Animal Health Fund at the Cummings School of Veterinary Medicine at Tufts University.
The authors declare that there were no conflicts of interest.
The authors thank Emily Morrison for her assistance with statistical analysis and Cathy Minogue of Northwest ZooPath for data retrieval.
Dr. Reavill is deceased.
References
- 1. ↑
Boal AK, Rosenzweig AC. Structural biology of copper trafficking. Chem Rev. 2009;109(10):4760–4779. doi:10.1021/cr900104z
- 2. ↑
Kim BE, Nevitt T, Thiele DJ. Mechanisms for copper acquisition, distribution and regulation. Nat Chem Biol. 2008;4(3):176–185. doi:10.1038/nchembio.72
- 3. ↑
Gaetke LM, Chow CK. Copper toxicity, oxidative stress, and antioxidant nutrients. Toxicology. 2003;189(1-2):147–163. doi:10.1016/s0300-483x(03)00159-8
- 4. ↑
Gaetke LM, Chow-Johnson HS, Chow CK. Copper: toxicological relevance and mechanisms. Arch Toxicol. 2014;88(11):1929–1938. doi:10.1007/s00204-014-1355-y
- 5. ↑
Członkowska A, Litwin T, Dusek P, et al. Wilson disease. Nat Rev Dis Primers. 2018;4(1):21. doi:10.1038/s41572-018-0018-3
- 6. ↑
Siaj R, Sauer V, Stöppeler S, et al. Dietary copper triggers onset of fulminant hepatitis in the Long-Evans cinnamon rat model. World J Gastroenterol. 2012;18(39):5542–5550. doi:10.3748/wjg.v18.i39.5542
- 7. ↑
Center SA, Richter KP, Twedt DC, Wakshlag JJ, Watson PJ, Webster CRL. Is it time to reconsider current guidelines for copper content in commercial dog foods? J Am Vet Med Assoc. 2021;258(4):357–364. doi:10.2460/javma.258.4.357
- 8. ↑
Goldfischer S, Popper H, Sternlieb I. The significance of variations in the distribution of copper in liver disease. Am J Pathol. 1980;99(3):715–730.
- 9. ↑
Miyamura H, Nakanuma Y, Kono N. Survey of copper granules in liver biopsy specimens from various liver abnormalities other than Wilson’s disease and biliary diseases. Gastroenterol Jpn. 1988;23(6):633–638. doi:10.1007/BF02782948
- 10. ↑
Hurwitz BM, Center SA, Randolph JF, et al. Presumed primary and secondary hepatic copper accumulation in cats. J Am Vet Med Assoc. 2014;244(1):68–77. doi:10.2460/javma.244.1.68
- 11. ↑
Azumi N. Copper and liver injury–experimental studies on the dogs with biliary obstruction and copper loading. Hokkaido Igaku Zasshi. 1982;57(3):331–349.
- 12. ↑
Hoefer HL. Gastrointestinal diseases. In: Quesenberry KE, Orcutt CJ, Mans C, Carpenter JW, eds. Ferrets, Rabbits, and Rodents: Clinical Medicine and Surgery. 4th ed. Elsevier; 2020:27–38.
- 13. ↑
van de Sluis BJ, Breen M, Nanji M, et al. Genetic mapping of the copper toxicosis locus in Bedlington terriers to dog chromosome 10, in a region syntenic to human chromosome region 2p13-p16. Hum Mol Genet. 1999;8(3):501–507. doi:10.1093/hmg/8.3.501
- 14. ↑
Sarkar B, Roberts EA. The puzzle posed by COMMD1, a newly discovered protein binding Cu(II). Metallomics. 2011;3(1):20–27. doi:10.1039/c0mt00031k
- 15. ↑
Asada H, Chambers JK, Kojima M, et al. Variations in ATP7B in cats with primary copper-associated hepatopathy. J Feline Med Surg. 2020;22(8):753–759. doi:10.1177/1098612X19884763
- 16. ↑
Asada H, Kojima M, Nagahara T, et al. Hepatic copper accumulation in a young cat with familial variations in the ATP7B gene. J Vet Intern Med. 2019;33(2):874–878. doi:10.1111/jvim.15399
- 17. ↑
Johnston AN, Center SA, McDonough SP, Wakshlag JJ, Warner KL. Hepatic copper concentrations in Labrador Retrievers with and without chronic hepatitis: 72 cases (1980–2010). J Am Vet Med Assoc. 2013;242(3):372–380. doi:10.2460/javma.242.3.372
- 18.
Fox JG, Zeman DH, Mortimer JD. Copper toxicosis in sibling ferrets. J Am Vet Med Assoc. 1994;205(8):1154–1156.
- 19. ↑
Miller AJ, Center SA, Randolph JF, Friesen CH, Miller AD, Warner KW. Disparities in hepatic copper concentrations determined by atomic absorption spectroscopy, inductively coupled plasma mass spectrometry, and digital image analysis of rhodanine-stained sections in dogs. J Am Vet Med Assoc. 2021;258(4):395–406. doi:10.2460/javma.258.4.395
- 20. ↑
Setsunan University, Faculty of Pharmaceutical Sciences, Department of Pathology. Wednesday Slide Conference 11, Case III – 2004903384 (AFIP 2937492). Armed Forces Institute of Pathology Department of Veterinary Pathology; 2005:5–9. Accessed May 30, 2021. https://www.askjpc.org/wsco/wsc/wsc04/04wsc11.pdf.
- 21. ↑
Graham JE, Heinze C, Beamer G, Webster CL. Management of copper toxicosis in a ferret (Mustela putorius furo). Abstract in: Proceedings of the Association of Exotic Mammal Veterinarians. Association of Exotic Mammal Veterinarians; 2016:441.
- 22. ↑
Fieten H, Hooijer-Nouwens BD, Biourge VC, et al. Association of dietary copper and zinc levels with hepatic copper and zinc concentration in Labrador Retrievers. J Vet Intern Med. 2012;26(6):1274–1280. doi:10.1111/j.1939-1676.2012.01001.x
- 23.
Fieten H, Biourge VC, Watson AL, Leegwater PA, van den Ingh TSGAM, Rothuizen J. Dietary management of Labrador Retrievers with subclinical hepatic copper accumulation. J Vet Intern Med. 2015;29(3):822–827. doi:10.1111/jvim.12574
- 24. ↑
Strickland JM, Buchweitz JP, Smedley RC, et al. Hepatic copper concentrations in 546 dogs (1982–2015). J Vet Intern Med. 2018;32(6):1943–1950. doi:10.1111/jvim.15308
- 25. ↑
Hoenerhoff M, Williams K. Copper-associated hepatopathy in a Mexican fruit bat (Artibeus jamaicensis) and establishment of a reference range for hepatic copper in bats. J Vet Diagn Invest. 2004;16(6):590–593. doi:10.1177/104063870401600619
- 26. ↑
Yuschenkoff D, Graham J, Barton BA, Garner MM. Evaluation of the clinical presentation and histologic lesions of hepatic copper accumulation in sugar gliders (Petaurus breviceps). J Exot Pet Med. 2021;39:51–56. doi:10.1053/j.jepm.2021.06.007
- 27. ↑
Center SA, McDonough SP, Bogdanovic L. Digital image analysis of rhodanine-stained liver biopsy specimens for calculation of hepatic copper concentrations in dogs. Am J Vet Res. 2013;74(12):1474–1480. doi:10.2460/ajvr.74.12.1474
- 28. ↑
Johnston AN, Center SA, McDonough SP, Warner KL. Influence of biopsy specimen size, tissue fixation, and assay variation on copper, iron, and zinc concentrations in canine livers. Am J Vet Res. 2009;70(12):1502–1511. doi:10.2460/ajvr.70.12.1502
- 29. ↑
Pillai S, Center SA, McDonough SP, et al. Ductal plate malformation in the liver of Boxer dogs: clinical and histological features. Vet Pathol. 2016;53(3):602–613. doi:10.1177/0300985815610567
- 30. ↑
Quaglia A, Burt AD, Ferrell LD, Portmann BC. Systemic disease. In: Burt A, Portmann BC, Ferrell LD, eds. MacSween’s Pathology of the Liver. 6th ed. Churchill Livingston Elsevier; 2012:935–986.
- 31. ↑
Huynh M, Laloi F. Diagnosis of liver disease in domestic ferrets (Mustela putorius). Vet Clin North Am Exot Anim Pract. 2013;16(1):121–144. doi:10.1016/j.cvex.2012.10.003
- 32. ↑
Morrisey JK, Johnston MS. Ferrets. In: Carpenter JW, Marion CJ, eds. Exotic Animal Formulary. 5th ed. Elsevier; 2018:549–550.
- 33. ↑
Torkelson MR, Heinze CR, Graham JE. Copper concentration in ten popular commercial ferret diets. Abstract in: Proceedings of the Exotics Conference. 2019;407–408.
- 34. ↑
Hill GM, Shannon MC. Copper and zinc: nutritional issues for agricultural animal production. Biol Trace Elem Res. 2019;188(1):148–159. doi:10.1007/s12011-018-1578-5
- 35. ↑
Ramirez CJ, Kim DY, Hanks BC, Evans TJ. Copper toxicosis in New Zealand White rabbits (Oryctolagus cuniculus). Vet Pathol. 2013;50(6):1135–1138. doi:10.1177/0300985813490756
- 36. ↑
Doong G, Keen CL, Rogers Q, Morris J, Rucker RB. Selected features of copper metabolism in the cat. J Nutr. 1983;113(10):1963–1971. doi:10.1093/jn/113.10.1963
- 37. ↑
Committee on Animal Nutrition. In: Nutrient Requirements of Mink and Foxes. 2nd ed. National Academies Press; 1982:7–17.35. Accessed May 30, 2021. https://doi.org/10.17226/1114.
- 38. ↑
Williams BH, Wyre NR. Neoplasia in ferrets. In: Quesenberry K, Mans C, Orcutt C, eds. Ferrets, Rabbits, and Rodents: Clinical Medicine and Surgery. 4th ed. Elsevier; 2021:102–103.
- 39. ↑
Garner MM, Powers L. Diseases of domestic ferrets. In: Proceedings of the Association of Avian Veterinarians Conference. Association of Avian Veterinarians; 2010:209–219.
- 40. ↑
van Meer S, de Man RA, van den Berg AP, et al. No increased risk of hepatocellular carcinoma in cirrhosis due to Wilson disease during long-term follow-up. J Gastroenterol Hepatol. 2015;30(3):535–539. doi:10.1111/jgh.12716
- 41.
Pfeiffenberger J, Mogler C, Gotthardt DN, et al. Hepatobiliary malignancies in Wilson disease. Liver Int. 2015;35(5):1615–1622. doi:10.1111/liv.12727
- 42.
Gunjan D, Shalimar DM, Nadda N, et al. Hepatocellular carcinoma: an unusual complication of longstanding Wilson disease. J Clin Exp Hepatol. 2017;7(2):152–154. doi:10.1016/j.jceh.2016.09.012
- 43.
Jong-Hon K, Togashi Y, Kasai H, Hosokawa M, Takeichi N. Prevention of spontaneous hepatocellular carcinoma in Long-Evans cinnamon rats with hereditary hepatitis by the administration of D-penicillamine. Hepatology. 1993;18(3):614–620.
- 44. ↑
Goussev SA, Center SA, Randolph JF, Kathrani A, Butler BP, McDonough SP. Clinical characteristics of hepatocellular carcinoma in 19 cats from a single institution (1980–2013). J Am Anim Hosp Assoc. 2016;52(1):36–41. doi:10.5326/JAAHA-MS-6289
- 45. ↑
Harro CC, Smedley RC, Buchweitz JP, Langlois DK. Hepatic copper and other trace mineral concentrations in dogs with hepatocellular carcinoma. J Vet Intern Med. 2019;33(5):2193–2199. doi:10.1111/jvim.15619
- 46. ↑
Nair J, Sone H, Nagao M, Barbin A, Bartsch H. Copper-dependent formation of miscoding etheno-DNA adducts in the liver of Long Evans cinnamon (LEC) rats developing hereditary hepatitis and hepatocellular carcinoma. Cancer Res. 1996;56(6):1267–1271.
- 47. ↑
Rosencrantz RA, LeCompte L, Yusuf Y. Beneath the copper-pediatric Wilson’s disease cirrhosis and hepatocellular carcinoma: a case report with literature review. Semin Liver Dis. 2015;35(4):434–438. doi:10.1055/s-0035-1567828
