Iron, copper, and zinc are essential to numerous metalloenzymes and cellular processes involved with intermediary metabolism and xenobiotic detoxification in animals. The liver can serve as a repository for copper and iron, and excessive accumulation of these metals can contribute to tissue injury in various contexts. Importantly, copper and iron are considered transition metals because they fluctuate between ionic forms. In so doing, transition metals catalyze formation of reactive oxygen species from oxygen via the Fenton-driven Haber-Weiss reaction.1–3
The liver plays an important role in zinc metabolism because it constitutes an important pool of quickly exchanged zinc.4,5 Zinc is needed for proper liver function because it has important antioxidant, anti-inflammatory, and antiapoptotic properties and is essential for typical tissue replication and healing, immunoregulation, and ammonia detoxification. Tissue zinc insufficiency is recognized as a complication of liver injury or portosystemic shunting.4,5 Because a low liver zinc concentration can increase predisposition to certain hepatotoxins, may compromise tissue repair, reduces antioxidant protection, and can potentiate development of hyperammonemia and hepatic encephalopathy, measurement of zinc concentrations in liver tissue can be used when making decisions about therapeutic zinc supplementation.4,5
A liver biopsy specimen is estimated to constitute approximately 1/50,000 of the hepatic mass and is routinely interpreted as representing the health status of the entire organ.6 However, several studies in humans7–13 and 1 study in dogs and cats14 revealed the inaccuracy of needle-core biopsy for definitive diagnosis of liver disease and for therapeutic monitoring. Because of heterogeneity of lesion distribution, inadequate representation of acinar units, and collection of too few portal triads (< 11 to 15), needle-core biopsy specimens can be misleading and can yield erroneous results when analyzed.7–14 Specifically, histologic evaluation of needle-core biopsy specimens is reportedly inaccurate for detection of copper-storage hepatopathy and for sequential appraisal of treatment efficacy in humans with genetically confirmed Wilson's disease, even when copper-specific stains are applied.15,16 Further complicating analysis of small specimens is the heterogenous distribution of copper in liver tissue in humans15–20 and other animals.21–24
Hepatic copper concentrations in dogs (< 400 μg/g [ppm] of dry-weight liver tissue) are typically substantially higher than in most other animal species.25–27 Humans, for example, have a typical hepatic copper concentration of only 10 to 30 ppm.27,28 Hepatic copper concentrations > 2,000 ppm of dry-weight liver tissue in dogs are often associated with progressive liver injury. However, some dogs tolerate extraordinary increases in hepatic copper concentrations without developing liver damage.27,29
Hepatic iron concentrations in clinically normal dogs range from 400 to 1,200 ppm of dry-weight liver tissue. A high iron concentration is common in dogs with necroinflammatory liver disease. It is probable that this high iron concentration contributes to or makes tissues vulnerable to injury as the liver engages in its usual sentinel functions, protecting the systemic circulation from the toxins and enteric pathogens delivered in the portal circulation.1,3,30,31 Retention of transition metals in the liver may contribute to the considerable glutathione depletion that has been detected in dogs with spontaneous necroinflammatory and cholestatic liver injury.32
It is unknown whether measurement of metal concentrations in liver biopsy specimens from dogs is influenced by specimen size or lesion heterogeneity. In humans, a minimum of 3 to 5 mg of dry-weight tissue (equivalent to a minimum of approx 6 to 10 mg of wet tissue) is a standard requirement when performing metal analyses of needle-core biopsy specimens.15,33,34 Smaller specimens are considered unsatisfactory for determination of copper concentration because the distribution of copper within the tissue may be inhomogenous, particularly in architecturally remodeled liver containing regenerative nodules and extensive fibrosis.15 Regional variability in copper concentrations has been confirmed in many animals (eg, humans, pigs, copper-fed rats, neonatal calves, and weanling rats).15–24 In humans, variability in copper concentrations between specimens of the same liver has been detected in neonates, adults, and patients with cirrhosis or other necroinflammatory liver disorders.15–20 Such variability has the potential to complicate diagnostic conclusions and therapeutic recommendations, particularly when stained needle-core biopsy specimens are used for subjective assessments or when specimens are submitted for quantitative metal analyses.
In veterinary medicine, it is increasingly common for clinicians to request an analysis of metal concentrations in formalin-fixed, paraffin-embedded, xyleneextracted tissue after receipt of histologic descriptions suggesting that excessive hepatic copper or iron stores have contributed to liver injury. There is no information confirming that such measurements correlate well with measurements obtained from fresh, unfixed tissue. The purpose of the study reported here was to determine the influence of specimen size on metal concentrations in liver tissue as measured by use of atomic absorption spectrometry, the variability associated with measurement of metal concentrations in the same liver tissue when biopsy specimens from the same liver lobe are analyzed on different days, and the validity of measurements of copper, iron, and zinc concentrations in formalin-fixed, paraffin-embedded, xylene-extracted liver specimens.
Materials and Methods
Liver specimens—The study involved 2 parts. For part 1, fresh liver tissues were obtained from 2 dogs that died of chronic hepatitis and tissues were frozen at −80°C for 6 months. One set of tissues (liver 1) was obtained from a dog with copper-storage hepatopathy; high copper but unremarkable iron concentrations were verified by use of special stains and metal quantification at the time of diagnosis. The other set (liver 2) was obtained from a dog with chronic hepatitis that was not associated with excessive copper stores; unremarkable copper but high iron concentrations were verified similarly to the first set. Formalin-fixed, paraffin-embedded tissue sections from each liver were stained with rhodanine (for copper) and Prussian blue (for iron) and microscopically examined to assess the heterogeneity of tissue metal distribution.
Eight paired biopsy specimens, one simulating a wedge or laparoscopic biopsy specimen and the other simulating a needle-core biopsy specimen, were obtained from livers 1 and 2 and submitted for paired analyses on 8 consecutive days. The simulated needle-core biopsy specimens were obtained by use of an 18-gauge automated spring-triggered needle-core biopsy toola (22-mm-long and 1-mm-wide biopsy needle; 16-mm-long chamber). In addition, 8 specimens from each liver were obtained from formalin-fixed, paraffin-embedded tissue that was extracted with xylene. These deparaffinized specimens (deparaffinized-archived specimens) were also submitted for metal analyses on 8 consecutive days.
For the other part of the study (part 2), formalin-fixed, paraffin-embedded archived liver specimens that had been collected for diagnostic purposes from dogs evaluated at the College of Veterinary Medicine at Cornell University were selected from tissue archives on the basis of availability of results for assays of metal concentrations in fresh tissue. Twenty specimens, originally obtained from 1994 through 2007 and large enough to accommodate repeated metal analyses, were selected to represent dogs with various hepatopathies (degenerative hepatopathy [n = 3], copper storage hepatopathy [3], cholangiohepatitis [2], cirrhosis [1], and chronic hepatitis [10]). Ten of the specimens had a high copper concentration (reference limits, 0 to < 400 ppm) and 10 had a copper concentration within reference limits, 15 specimens had a high iron concentration (reference limits, 400 to 1,200 ppm) and 5 had an iron concentration within reference limits, and 8 specimens had an inadequate zinc concentration (reference limits, 120 to 280 ppm) and 12 had a zinc concentration within reference limits. All formalin-fixed, paraffin-embedded archived tissues underwent xylene extraction before submission for metal analyses. Paraffin-embedded liver tissue was sectioned at 5 μm and stained with H&E and Masson trichrome strains to detect nonuniform lobular architecture and tissue fibrosis that might influence tissue metal concentrations. Tissues with regenerative nodules or extensive fibrosis were assigned a histopathologic score of 1, and those lacking these features were assigned a score of 2.
Study design—Two sets of evaluations were performed. In one, the specimens of fresh frozen liver tissue and formalin-fixed, paraffin-embedded liver tissue were used to determine the influence of specimen size, interassay variability, and tissue fixation on assay results for metal concentrations. In the other, the archived specimens and results for fresh tissues collected at the time of liver biopsy in the 20 dogs with various types of liver disease were used to determine the effect of specimen storage conditions on metal assay results.
Extraction of formalin-fixed, paraffin-embedded tissues from archived liver specimens—Tissue extraction was performed by use of a modified version of a method described elsewhere.b Paraffin blocks of liver tissues were melted on a standard embedding unit, and a specimen (approx 50 to 100 mg) was removed by use of a sterile scalpel blade. Tissue was placed in a 20-mL glass beaker under a fume hood and submerged in 1 mL of xylene at room temperature (approx 25°C). Xylene was changed hourly for 3 hours. Afterward, tissues were bathed in 1 mL of 70% ethanol at room temperature and the ethanol was changed hourly for 2 hours. Specimens were subsequently placed in a 2-mL sterile Eppendorf tube and stored at room temperature until submitted for analysis. Analyses were completed serially, without operator knowledge of specimen identification.
Measurement of metal concentrations—Copper, iron, and zinc concentrations were determined by use of flame atomic absorption spectrometry.c,35 Seven standardizing solutions were analyzed during each copper analysis (0.05, 0.10, 0.20, 0.50, 1.00, 2.00, and 3.00 ppm of copper), 6 standardizing solutions were analyzed during each iron analysis (0.10, 0.20, 0.50, 1.00, 2.00, and 3.00 ppm of iron), and 5 standardizing solutions were analyzed during each zinc analysis (0.05, 0.10, 0.20, 0.50, and 1.00 ppm of zinc). Standardizing solutions were prepared from 1,000-ppm certified calibrating stock solutions.d
All liver specimens were placed in a weighed drying crucible and desiccated to constant weight at 85°C in a drying oven. The crucible and dry liver tissue were cooled in a desiccator and weighed to determine dry tissue weight; the tissue was subsequently ashed overnight in a muffle furnace at 600°C. After cooling, the gross ash weight was determined and the ash was dissolved in an appropriate volume of 3.6N HNO3 with sonication for 1 minute to promote full dissolution. The tissue dilutional factor was recorded. Mineral concentration was determined with a flame photometer.35 Fresh and deparaffinized-archived specimens were measured in the same laboratory with the same analytic method and equipment.e
Reference limits for liver metal concentrations were used to designate biopsy specimen metal concentrations as unremarkable or high for comparison of tissue metal status. A clinical treatment score was also assigned to each tissue specimen on the basis of its measured metal concentration to determine the influence of metal concentration on conventional treatment recommendations. Score categories were based on clinical criteria used by the authors in making treatment recommendations. Tissue copper concentrations < 400 ppm were considered unremarkable and assigned a score of 1, concentrations from 400 to 1,500 ppm were considered moderately high and assigned a score of 2, and concentrations > 1,500 ppm were considered markedly high and assigned a score of 3. Concentration scores were assigned with the consideration that dogs with copper values > 1,500 ppm would receive a recommendation for chelation treatment, antioxidant treatment, and restriction of exogenous copper intake, whereas chelation recommendations for dogs with moderately high values (400 to 1,500 ppm) would depend on clinical signs and histopathologic features of liver biopsy specimens. Dogs with a copper score of 2 would be minimally recommended to receive antioxidant treatment and a restriction of exogenous copper intake. Tissue iron concentrations ≤ 1,200 ppm were considered unremarkable and assigned a score of 1, whereas tissue iron concentrations > 1,200 ppm were considered high and assigned a score of 2. These scores were assigned with the consideration that long-term antioxidant treatment would be prescribed for dogs with liver iron concentrations > 1,200 ppm. Tissue zinc concentrations within reference limits (120 to 280 ppm) were assigned a severity score of 1 (adequate), whereas zinc concentrations indicative of zinc insufficiency (< 120 ppm) were assigned a clinical severity score of 2 (inadequate). These scores were assigned with the consideration that supplemental zinc acetate would have been prescribed for dogs with a clinical severity score of 2.
Statistical analysis—Distribution of specimen data was examined by constructing box-and-whisker plots and applying the Kolmogorov-Smirnov test. Mean ± SD values and CVs were calculated for results when a known concentration of metal was analyzed in each analytic run on different days to determine the influence of interassay variability on measurements of tissue metal concentrations.
For livers 1 and 2 from the dogs with chronic hepatitis, tissue weights were normally distributed for the simulated wedge, laparoscopic, and needle-core biopsy specimens; mineral concentrations were also normally distributed for serially analyzed, fresh frozen, and deparaffinized-archived tissues that had been analyzed on 8 consecutive days. Therefore, data for these specimens are expressed as mean ± SD. Results for large (wedge) and small (needle-core) biopsy specimens collected from livers 1 and 2 were compared by use of a paired t test; a value of P ≤ 0.05 was considered significant. Repeatability of tissue metal analyses and the potential influence of heterogenous lobular distribution of metals were determined by calculating the CVs for large (simulated wedge and laparoscopic) and small (simulated needle-core) biopsy specimens from fresh frozen liver tissue; CVs were also calculated for results of analyses of deparaffinized-archived tissues performed on different days. The influence of having a high versus low tissue metal concentration on repeatability of results was determined by calculating the CVs from data for liver 1 (with a high copper concentration) and liver 2 (with a high iron concentration).
For the fresh and deparaffinized-archived liver specimens from 20 dogs with various hepatobiliary disorders, metal concentrations were not normally distributed. Differences in liver metal concentrations between fresh and deparaffinized-archived tissues were determined by use of the Wilcoxon signed rank test. The relationship between metal concentrations in fresh and deparaffinized-archived biopsy specimens was investigated by use of the Pearson correlation coefficient (r). A 2-sided value of P ≤ 0.05 was used to indicate significance.
Tissue metal status (unremarkable or high) was compared between paired fresh and deparaffinized-archived biopsy specimens by use of the Wilcoxon signed rank test; a 2-sided value of P ≤ 0.05 was applied to designate significance. To determine whether differences in metal status between paired fresh and deparaffinized-archived biopsy specimens might influence treatment recommendations, clinical treatment scores were compared by use of the Wilcoxon rank sum test; a 2-sided value of P < 0.05 was applied to designate significance.
The influence of nonuniform lobular hepatic architecture on tissue metal concentrations was determined by inspecting the CVs of tissue metal concentrations from specimens for which data were stratified by the presence or absence of regenerative nodules or bridging fibrosis. All statistical analyses were performed by use of a commercial software program.f
Results
Liver specimens—All liver specimens were analyzed on 12 days. Mean ± SD values for metal standards measured on each day were 113 ± 4 ppm (CV, 3.7%) for copper, 572 ± 22 ppm (CV, 3.9%) for iron, and 191 ± 7 ppm (CV, 3.9%) for zinc. Serially measured (each of 8 days) fresh frozen and deparaffinized-archived tissue specimens from the 2 dogs with chronic hepatitis (livers 1 and 2) had a higher CV than did the metal standards, reflecting differences in acinar metal distribution in regions from which specimens were obtained (ie, fibrous connective tissue and regenerative nodules; Figures 1 and 2) or small errors in determination of desiccated specimen weights (Table 1).
Photomicrographs of liver specimens from a dog with copper-storage hepatopathy, which were used in the study to evaluate the effects of specimen size, tissue fixation, and assay variation on liver metal concentrations. A—Application of rhodanine stain, which stains copper red-orange, reveals a high copper concentration. Bar = 500 μm. B—Margin of a regenerative nodule containing macroscopically vacuolated hepatocytes and marginated copper deposits. Bar = 200 μm. Both photomicrographs reveal heterogenous copper distribution within liver tissue, which can influence metal measurements when only single, small specimens are evaluated.
Citation: American Journal of Veterinary Research 70, 12; 10.2460/ajvr.70.12.1502
Photomicrographs of liver specimens from a dog with copper-storage hepatopathy, which were used in the study to evaluate the effects of specimen size, tissue fixation, and assay variation on liver metal concentrations. A—Application of rhodanine stain, which stains copper red-orange, reveals a high copper concentration. Bar = 500 μm. B—Margin of a regenerative nodule containing macroscopically vacuolated hepatocytes and marginated copper deposits. Bar = 200 μm. Both photomicrographs reveal heterogenous copper distribution within liver tissue, which can influence metal measurements when only single, small specimens are evaluated.
Citation: American Journal of Veterinary Research 70, 12; 10.2460/ajvr.70.12.1502
Photomicrographs of liver specimens from a dog with copper-storage hepatopathy, which were used in the study to evaluate the effects of specimen size, tissue fixation, and assay variation on liver metal concentrations. A—Application of rhodanine stain, which stains copper red-orange, reveals a high copper concentration. Bar = 500 μm. B—Margin of a regenerative nodule containing macroscopically vacuolated hepatocytes and marginated copper deposits. Bar = 200 μm. Both photomicrographs reveal heterogenous copper distribution within liver tissue, which can influence metal measurements when only single, small specimens are evaluated.
Citation: American Journal of Veterinary Research 70, 12; 10.2460/ajvr.70.12.1502
Photomicrographs of liver specimens from a dog with chronic hepatitis, which were used to evaluate the effects of specimen size, tissue fixation, and assay variation on liver metal concentrations. A—A large regenerative nodule extends from the hepatic parenchyma, in which necroinflammatory lesions are highlighted by Prussian blue stain, which stains iron blue (lower left of field). Bar = 500 μm. B—A large amount of stainable iron is evident in an area of active inflammation. Prussian blue stain; bar = 200 μm. C—A paucity of stainable iron is evident within the large regenerative nodule. Prussian blue stain; bar = 200 μm. All photomicrographs reveal heterogenous iron distribution within the liver, which can influence metal measurements when only single, small specimens are evaluated.
Citation: American Journal of Veterinary Research 70, 12; 10.2460/ajvr.70.12.1502
Photomicrographs of liver specimens from a dog with chronic hepatitis, which were used to evaluate the effects of specimen size, tissue fixation, and assay variation on liver metal concentrations. A—A large regenerative nodule extends from the hepatic parenchyma, in which necroinflammatory lesions are highlighted by Prussian blue stain, which stains iron blue (lower left of field). Bar = 500 μm. B—A large amount of stainable iron is evident in an area of active inflammation. Prussian blue stain; bar = 200 μm. C—A paucity of stainable iron is evident within the large regenerative nodule. Prussian blue stain; bar = 200 μm. All photomicrographs reveal heterogenous iron distribution within the liver, which can influence metal measurements when only single, small specimens are evaluated.
Citation: American Journal of Veterinary Research 70, 12; 10.2460/ajvr.70.12.1502
Photomicrographs of liver specimens from a dog with chronic hepatitis, which were used to evaluate the effects of specimen size, tissue fixation, and assay variation on liver metal concentrations. A—A large regenerative nodule extends from the hepatic parenchyma, in which necroinflammatory lesions are highlighted by Prussian blue stain, which stains iron blue (lower left of field). Bar = 500 μm. B—A large amount of stainable iron is evident in an area of active inflammation. Prussian blue stain; bar = 200 μm. C—A paucity of stainable iron is evident within the large regenerative nodule. Prussian blue stain; bar = 200 μm. All photomicrographs reveal heterogenous iron distribution within the liver, which can influence metal measurements when only single, small specimens are evaluated.
Citation: American Journal of Veterinary Research 70, 12; 10.2460/ajvr.70.12.1502
Influence of metal status and tissue specimen size on serial determinations (8 consecutive days) of copper, iron, and zinc in fresh frozen liver tissue (simulated needle-core and simulated wedge biopsy specimens) and deparaffinized-archived tissue from 2 dogs with chronic hepatitis.
| Metal | Liver 1 specimen type | Liver 2 specimen type | ||||
|---|---|---|---|---|---|---|
| Fresh frozen needle-core | Fresh frozen wedge | Deparaffinized | Fresh frozen needle-core | Fresh frozen wedge | Deparaffinized | |
| Copper | ||||||
| No. of specimens | 8 | 8 | 7 | 8 | 8 | 7 |
| Mean ± SD concentration | 1,841 ± 217a,c | 2,349 ± 439b,c | 4,400 ± 225a,b | 166 ± 60 | 197 ± 71 | 188 ± 94 |
| CV (%) | 11.8 | 50.0 | 5.1 | 35.9 | 36.1 | 50.0 |
| Iron | ||||||
| No. of specimens | 8 | 8 | 7 | 8 | 8 | 7 |
| Mean ± SD concentration | 624 ± 23 | 603 ± 129 | 803 ± 275 | 6,514 ± 1,422d,f | 8,163 ± 1,728e,f | 4,095 ± 1,742d,e |
| CV (%) | 3.6 | 21.4 | 34.2 | 21.8 | 21.2 | 42.6 |
| Zinc | ||||||
| No. of specimens | 8 | 8 | 7 | 8 | 8 | 7 |
| Mean ± SD concentration | 127 ± 7g | 114 ± 10h | 192 ± 16g,h | 110 ± 15i | 97 ± 11i | 111 ± 30 |
| CV (%) | 5.5 | 9.1 | 8.5 | 13.7 | 11.0 | 27.1 |
Significant (P ≤ 0.05) differences between values are indicated by shared superscript letters: a,b,e,g,hP < 0.001; dP = 0.007; cP = 0.01; and f,iP = 0.05.
The liver of 1 dog had a high copper and low iron concentration (liver 1), and that of the other dog had a low copper and high iron concentration (liver 2).
For livers 1 and 2, mean wet weight of the simulated needle-core biopsy specimens (13.9 ± 2.3 mg; CV, 16.9%) was significantly (P < 0.001) lower than that of the simulated wedge biopsy specimens (47.3 ± 19.9 mg; CV, 42.1%). Similarly, mean dry tissue weight of the simulated needle-core biopsy specimens (2.9 ± 1.0 mg; CV, 34.0%) was significantly (P < 0.001) lower than that of the simulated wedge biopsy specimens (12.3 ± 6.1 mg; CV, 49.6%). Dry weight of needle-core specimens (20.4 ± 5.3%; CV, 26.1%) constituted a significantly (P = 0.03) lower percentage of wet tissue weight, compared with that of the simulated wedge specimens (26.3 ± 9.2%; CV, 34.9%), suggesting a greater degree of desiccation in the needle-core specimens.
Metal concentrations in livers from dogs with chronic hepatitis—Copper concentrations were not significantly (P = 0.36) different between needle-core and wedge biopsy specimens from liver 2 (low copper concentration tissue [range, 94 to 257 ppm]). However, copper concentrations in needle-core biopsy specimens were significantly (P = 0.01) lower than in wedge biopsy specimens from liver 1 (high copper concentration tissue [range, 1,530 to 3,200 ppm]). There were no differences between copper concentrations in liver 2 between deparaffinized-archived and frozen specimens (needle-core, P = 0.51; wedge, P = 0.87). However, deparaffinized-archived specimens from liver 1 contained significantly (P < 0.001) higher copper concentrations than did needle-core or wedge biopsy specimens. Metal assays of needle-core biopsy specimens underestimated tissue copper concentrations in liver 1, with results indicating a mean of 81 ± 17% of the copper concentration in wedge specimens and a mean of 42 ± 5% of the copper concentration in deparaffinized-archived specimens. However, given the extremely high copper concentration in the tissue from liver 1, differences in copper concentrations would not have changed clinical severity scores because all specimens contained > 1,500 ppm of copper.
Whereas there was no significant (P = 0.67) difference between iron concentrations in needle-core and wedge biopsy specimens in liver 1 (low iron concentration tissue [range, 381 to 844 ppm]), liver 2 (high iron concentration tissue [range, 5,130 to 10,100 ppm]) had a significantly (P = 0.05) lower iron concentration in needle-core versus wedge biopsy specimens. Iron concentrations in the frozen needle-core and wedge tissue biopsy specimens from liver 2 were significantly higher than in deparaffinized-archived biopsy specimens (P = 0.004 and P = 0.03, respectively; Table 1). When metal concentrations were measured in needle-core biopsy specimens from liver 2, the assay detected a mean of 83 ± 27% of the iron concentration in wedge biopsy specimens and a mean of 174 ± 53% of the iron concentration in deparaffinized-archived tissue, whereas wedge biopsy specimens contained a mean of 228 ± 95% of the iron concentration in deparaffinized-archived tissue. The source of the differences was investigated and confirmed by staining liver 2 tissue with Prussian blue and microscopically examining the tissue for inhomogenous distribution of iron. Nevertheless, because all measurements revealed iron concentrations > 2,360 ppm (median, 4,120 ppm; range, 2,360 to 7,100), the assigned clinical severity score for each analyzed specimen was consistent (score = 2).
Zinc concentrations in serially measured specimens were not significantly (r = 0.056; P = 0.81) correlated; however, the range of zinc concentrations was limited (27 to 285 ppm). Needle-core biopsy specimens from fresh frozen tissue contained significantly (P = 0.002) higher zinc concentrations than did wedge specimens in both livers, with the needle-core specimens containing a mean of 112 ± 13% (liver 1) and 114 ± 14% (liver 2) of zinc concentrations in wedge biopsy specimens.
Significantly (P < 0.001) higher zinc concentrations were detected in deparaffinized-archived specimens from liver 1, compared with concentrations in fresh frozen tissue; needle-core specimens contained a mean of 67 ± 8% of zinc concentrations in deparaffinized-archived specimens, and wedge specimens contained a mean of 59 ± 7%. However, there were no significant (all P ≥ 0.4) differences in zinc concentrations between frozen and deparaffinized-archived specimens from liver 2.
Metal concentrations in liver specimens from dogs with various hepatopathies—Copper concentrations in deparaffinized-archived liver tissue were not significantly (P = 0.23) different from values in fresh tissue collected at the time of liver biopsy in the 20 dogs with various hepatopathies (Figure 3). Copper concentrations in these 2 types of specimens were significantly (r = 0.94; P < 0.001) correlated with each other.
Concentrations of copper (A), iron (B), and zinc (C) in fresh and archived liver tissues from 20 dogs with various hepatopathies. Reference limits are < 400 μg/g (ppm) of dry-weight tissue for copper, 400 to 1,500 μg/g of dry-weight tissue for iron, and 120 to 280 μg/g of dry-weight tissue for zinc. Striated bar on each graph depicts cutoff values used to indicate high copper or iron concentrations and low zinc concentration.
Citation: American Journal of Veterinary Research 70, 12; 10.2460/ajvr.70.12.1502
Concentrations of copper (A), iron (B), and zinc (C) in fresh and archived liver tissues from 20 dogs with various hepatopathies. Reference limits are < 400 μg/g (ppm) of dry-weight tissue for copper, 400 to 1,500 μg/g of dry-weight tissue for iron, and 120 to 280 μg/g of dry-weight tissue for zinc. Striated bar on each graph depicts cutoff values used to indicate high copper or iron concentrations and low zinc concentration.
Citation: American Journal of Veterinary Research 70, 12; 10.2460/ajvr.70.12.1502
Concentrations of copper (A), iron (B), and zinc (C) in fresh and archived liver tissues from 20 dogs with various hepatopathies. Reference limits are < 400 μg/g (ppm) of dry-weight tissue for copper, 400 to 1,500 μg/g of dry-weight tissue for iron, and 120 to 280 μg/g of dry-weight tissue for zinc. Striated bar on each graph depicts cutoff values used to indicate high copper or iron concentrations and low zinc concentration.
Citation: American Journal of Veterinary Research 70, 12; 10.2460/ajvr.70.12.1502
Iron concentrations in deparaffinized-archived liver tissue were also not significantly (P = 0.42) different from values in fresh specimens collected at the time of liver biopsy. Iron concentrations in these 2 types of specimens were significantly (r = 0.58; P = 0.007) correlated with each other. However, in 9 of 20 (45%) pairs of liver specimens, a difference > 1,000 ppm of iron existed between fresh and deparaffinized-archived tissues.
Zinc concentrations in deparaffinized-archived liver tissue were not significantly (P = 0.39) different from those in fresh tissue at the time of liver biopsy (Figure 1). However, there were 10 dogs that had a low zinc concentration (< 120 ppm) in fresh frozen tissue but a zinc concentration within the reference limit in deparaffinized-archived tissue.
Influence of tissue processing on clinical severity scores—Clinical severity scores did not differ significantly (all P ≥ 0.07) between fresh and deparaffinized-archived liver tissue when data were stratified by degree of copper or iron concentration (unremarkable [< 400 ppm] vs moderately high [400 to 1,500 ppm] vs markedly high [> 1,500 ppm] for copper; unremarkable [≤ 1,200 ppm] vs high [> 1,200 ppm] for iron). However, a significant difference in severity scores for fresh versus deparaffinized-archived tissues was detected when data were stratified by degree of zinc concentration (adequate [120 to 280 ppm] vs inadequate [< 120 ppm]), in which more fresh tissue specimens contained a low concentration of zinc, compared with deparaffinized-archived tissues (P = 0.009). A greater proportion of fresh tissue specimens (12/20 [60%]) had a low zinc concentration than did deparaffinized-archived specimens (3/20 [15%]). Liver tissues of 10 of 20 (50%) dogs with a higher zinc concentration in deparaffinized-archived tissue than in fresh tissue had a change in severity score from low to adequate.
In liver tissues from 5 of 20 (25%) dogs, higher copper and iron concentrations were detected in deparaffinized-archived tissue, compared with concentrations measured at the time of liver biopsy. In liver tissues from 4 (20%) dogs, lower copper and iron concentrations were detected in deparaffinized-archived liver tissue, compared with concentrations measured at the time of liver biopsy. In liver tissues from 3 (15%) dogs, the clinical severity score for copper changed when deparaffinized-archived rather than fresh tissues were used for the metal assay. In 1 liver, the score decreased (from moderately high to unremarkable), and in 2 other livers, the score increased (from unremarkable to moderately high). The clinical severity score for iron also changed in 2 (10%) dogs when deparaffinized-archived rather than fresh tissue was used. In 1 liver, the score changed from unremarkable to high, and in the other liver, the opposite was true.
Influence of liver architecture on metal concentrations—Liver tissues from 7 of the 20 dogs with various hepatopathies were judged to have nonuniform lobular architecture when evaluated histologically. The CVs for copper and iron concentrations were higher in these tissues, compared with tissues with uniform lobular architecture. The CVs for metal measurements in fresh liver biopsy specimens with nonuniform or uniform hepatic architecture were 107% versus 62%, respectively, for copper; 60% versus 44%, respectively, for iron; and 60% versus 47%, respectively, for zinc. The CVs for metal measurements in deparaffinized-archived biopsy specimens with nonuniform or uniform hepatic architecture were 104% versus 84%, respectively, for copper; 74% versus 38%, respectively, for iron; and 12% versus 38%, respectively, for zinc.
Discussion
The findings of the present study supported the belief that small biopsy size can importantly influence the tissue metal concentrations detected in livers of dogs with hepatopathy. They also confirmed that hepatic copper and iron concentrations can be reasonably ascertained by use of deparaffinized-archived tissue. In addition, our data suggested that specimens stored > 10 years can be used in evaluations of liver copper and iron status, expanding on findings of another study36 in which formalin-fixed tissue was used. However, we found that deparaffinized-archived liver tissue is unreliable for the assessment of hepatic zinc sufficiency. Although zinc concentrations were significantly affected (ie, increased) by tissue fixation, processing, or deparaffinization, the exact cause of this phenomenon was not investigated (eg, contamination of formalin, xylene, or paraffin). An investigation37 of copper, iron, and zinc concentrations in formalin solutions used in our hospital and necropsy facility revealed < 1 mg/L of these metals in the solutions, which would suggest that formalin solution was not a source of contamination in the present study. The detection of a significantly higher zinc concentration in the deparaffinized-archived liver specimen from liver 1, compared with concentration in fresh frozen tissue, also may have reflected an inhomogenous tissue distribution of zinc, as revealed histologically for copper and iron (Figure 1). We also detected significantly higher zinc concentrations in needle-core versus wedge biopsy specimens from the same fresh frozen liver tissue. Whereas each tissue specimen was gently removed from the cutting chamber of the needle-core biopsy device without scraping instrument metal, contamination cannot be discounted as a confounding variable. However, a study38 of the influence of contamination from needle metal on tissue zinc measurements failed to reveal such an influence.
Day-to-day variation in tissue metal concentrations determined by atomic absorption spectrometry was negligible in the present study when certified standards were incorporated in each analytic run. Differences among tissue metal concentrations determined on different analytic days were therefore presumed to reflect inhomogenous tissue metal distribution, variation in hepatic architecture of the specimens collected, and, potentially, small errors in recorded tissue weights. Whereas such tissue-associated variables yielded differences in measured metal concentrations of up to 50% in liver tissue with a high copper concentration (liver 1), they had no effect on clinical severity scores.
Results of the present study suggested that analysis of hepatic metal concentrations can be achieved with atomic absorption spectrometry when wet tissue specimens as small as 12 mg are used. Yet because we found that needle-core biopsy specimens may yield significantly lower measurements of copper concentration than larger specimens, small needle-core specimens may adversely influence definitive diagnosis or assessments of disease progression or treatment response when sequential specimens are collected. Through results of serial analyses of liver tissue containing a high copper concentration (liver 1), it was evident that needle-core specimens yielded measurements that were 81% and 43% of the copper concentrations measured in larger wedge or deparaffinized-archived specimens, respectively. It was also evident that a wide range of tissue metal concentrations can be obtained when large tissue specimens (ie, wedge biopsy specimens) are serially analyzed, reflecting a greater variety in the acinar architecture of tissues obtained, including large regenerative nodules and regions in which parenchyma is displaced or replaced by inflammation or fibrous connective tissue (Figures 1 and 2). This finding is consistent with conclusions made in other studies.15,16
Regenerative nodules contain comparatively less iron or copper than healthy hepatic parenchyma.16,19,27,39–42 Tissues densely infiltrated by fibrous connective tissue, as encountered with chronic necro-inflammatory liver disease, have comparatively smaller concentrations of copper and iron than regions actively engaged with metal-induced or -associated inflammation.15,19,40–42 Whereas some variation in liver copper and iron measurements was identified in the deparaffinized-archived tissue from livers of dogs with various hepatopathies, differences were generally not large enough to change clinical assessment scores and thus treatment recommendations. Findings in the present study also supported the hypothesis that nonuniform lobular architecture is associated with larger variations in replicate tissue metal determinations than is uniform lobular architecture.
In dogs, a substantial quantity of hepatic copper often initially accumulates in acinar zone 3 (centrilobular or periacinar location).27 Reconciliation of zonal distribution of tissue lesions and metal accumulation can assist in determining the role of copper in liver injury. However, this can be problematic when needle-core specimens are evaluated. Needle-core biopsy methods result in specimens that are underrepresentative of the acinar architecture of the liver, accounting for their inconsistent results when assessing histologic details.6,14,19,41,42 This problem is not easily solved through acquisition of multiple specimens because serial collection poses a risk of iatrogenic trauma with each biopsy attempt. The accuracy of diagnoses based on analysis of needle-core specimens is further compromised by the fact that most biopsy specimens are retrieved from left liver lobes in an attempt to avoid iatrogenic trauma to adjacent viscera and porta hepatis vasculature and biliary structures. Consequently, needle-core biopsy specimens from different portions of the liver are infrequently collected, complicating assessment of disorders with heterogenous liver lobe involvement. Because blind collection of liver biopsy specimens is hazardous and results in retrieval of specimens from unidentified locations, it is preferable to use ultrasonography-guided methods. Even then, the location from which a needle-core biopsy specimen is obtained is usually restricted to the left or “safer” liver lobes. Biopsy specimen collection from several liver lobes, including regions with gross lesions and apparently healthy tissue, optimizes accuracy during histologic characterization and is fundamental for making appropriate therapeutic recommendations. Such specimens are essential for identification of disorders associated with pathological metal accumulation because metal concentrations should be measured in grossly healthy tissue rather than just regenerative nodules.
Many dogs with necroinflammatory liver disease develop a high tissue iron concentration.31,43 Yet, to our knowledge, the pathophysiological importance of high hepatic iron concentration in dogs with liver disease has not been investigated. Given the results of experimental work in other species,1,3,18,30,44–47 we clinically recommend that dogs with a hepatic iron concentration > 1,200 ppm and histologic evidence of hepatocellular injury or remodeling receive antioxidants to counteract the role of iron as a transition metal promoting tissue injury. By stratifying the study data into 2 categories of hepatic iron concentrations (≤ 1,200 and > 1,200 ppm), we found that only 11% of dogs with paired measurements of iron concentration in fresh and deparaffinized-archived tissue changed category on the basis of results for deparaffinized-archived specimens. Iron is typically located within hepatic Kupffer cells and may be distributed diffusely within a liver lobe. However, dense concentrations often accumulate within or marginate regions of active inflammation as a consequence of iron sequestration in macrophages and lipogranulomatous foci. Thus, iron stores associated with necroinflammatory liver lesions usually have an inhomogenous acinar distribution that can markedly affect the measured iron concentration. Results for the histologic evaluation of tissue from the liver with a high iron concentration (liver 2; Figure 2) that was used for serial metal analyses suggested a wide variation in iron distribution that could complicate tissue metal quantification. However, despite large variations in iron concentrations in fresh liver specimens, compared with variations in deparaffinized-archived specimens, the iron concentration in deparaffinized-archived tissue would not have altered clinical recommendations for antioxidants in most dogs.
The significantly lower concentration of iron in deparaffinized-archived specimens relative to that in fresh frozen specimens underscored several important considerations relevant to measurement of tissue metal concentrations. First, it is important to obtain and assess representative liver specimens. Second, it is useful to evaluate > 1 liver specimen with qualitative staining; this can be inexpensively performed by orienting sections from several liver biopsy specimens on 1 slide for staining. Third, it is important to reconcile quantitative metal data with histologic features (including tissue metal stains) in liver specimens.
Distinct patterns of hepatic copper distribution have been attributed to underlying metabolic factors, primary liver disease, and exogenous copper loading in various species and disorders.15,19,39,40,48–51 Several other studies15,19,39,40,48,49 have revealed unequal hepatic metal sequestration among regions of the liver in copper-associated hepatopathies. Similarly, variability in tissue copper concentrations in liver 1 in our study seemingly contributed to the wide range of copper concentrations detected in wedge biopsy specimens from that liver.
The etiology of copper-associated hepatopathy remains unknown in breeds of dogs other than the Bedlington Terrier. Thus, there remains great controversy regarding the role of copper as a primary cause of liver injury in this species, considering that copper retention can be an epiphenomenon of liver disease (ie, hepatocellular injury or cholestasis).27,29,36,52 Because healthy dogs have a higher hepatic copper concentration than many other species, defining a pathological copper concentration is difficult.25 Further, evaluation of tissue staining for copper is qualitative and subjective, reflecting the opinion of the examiner, adequacy of the acinar structure within the specimen, and quality or success of the staining procedure.27 In such situations, it is standard practice to concurrently stain a positive control tissue to rule out a false-negative stain reaction. Before declaring the importance of copper status in a canine liver biopsy specimen, size of the specimen must be appraised and the histologic features and zonal location of copper must be considered. Careful histologic assessment will help determine whether representative tissue was obtained. Finally, results of qualitative assessment of copper concentration based on tissue staining should be reconciled with quantitative copper measurements.16,19,27,39,40,49 The quality of the biopsy specimen and of the copper stain must be thoughtfully considered when serial biopsy specimens are used to determine disease progression or treatment success. Findings in the present study and another study14 suggest that diagnoses and results of serial evaluations may be compromised when needle-core specimens are used for determining histologic status and copper quantification.
In the authors' opinion, dogs with livers containing a copper concentration > 1,500 ppm associated with histologic evidence of hepatocellular injury or serially confirmed high serum transaminase activities are candidates for chelation treatment and restriction of exogenous copper intake from food and water. In the present study, clinical severity scores did not change in any dog on the basis of results from deparaffinized-archived versus fresh liver tissues, making use of deparaffinized-archived tissues a clinically viable option for metal quantification when determination of hepatic metal concentrations is an afterthought. Our findings supported the supposition that differences in copper distribution within the liver can complicate an initial diagnosis and comparisons between results of serially analyzed specimens when needle-core biopsy specimens are used.
Our extensive clinical experience suggests that optimal and accurate characterization of hepatobiliary disease can be achieved by collecting a minimum of 3 biopsy specimens, each from a different liver lobe. When large regenerative nodules are obtained, it is also important to obtain specimens of adjacent, apparently healthy liver tissue. The best tissue specimens are collected by use of surgical wedge or laparoscopic cup biopsy methods.14
Findings in the study reported here indicated that retrospective quantification of hepatic copper and iron in deparaffinized-archived tissue may provide useful clinical information. These measurements can be requisitioned after subjective histologic assessment of hepatic metal status. However, this is only possible when biopsy methods harvest enough representative tissue for subsequent testing. We maintain that it is preferable to measure transition metals in portions of fresh biopsy specimens rather than formalin-fixed, paraffin-embedded specimens because tissue zinc concentration also may be accurately determined when fresh specimens are used.
ABBREVIATION
| CV | Coefficient of variation |
Bard Urological Division, Covington, Ga.
Braselton WE, Slanker MR, Stuart KJ, et al. Comparison of element concentrations determined in fresh, formalin fixed and paraffin embedded tissue samples (abstr), in Proceedings. 40th Meet Am Assoc Vet Lab Diagn 1997;74.
Varian AA-1275, Varian Inc, Palo Alto, Calif.
Certified standardizing stock solutions (ferric nitrate, zinc oxide, copper nitrate) in nitric acid, 1,000 ppm ± 1%, Fisher Scientific, Pittsburgh, Pa.
Diagnostic Laboratory, College of Veterinary Medicine, University of Colorado, Fort Collins, Colo.
Statistix, version 9.0, Analytical Software, Tallahassee, Fla.
References
- 1.
Pietrangelo A. Iron, oxidative stress and liver fibrogenesis. J Hepatol 1998;28:8–13.
- 2.
Letelier ME, Lepe AM, Faúndez M, et al.Possible mechanisms underlying copper-induced damage in biological membranes leading to cellular toxicity. Chem Biol Interact 2005;151:71–82.
- 3.
Ramm GA, Ruddell RG. Hepatotoxicity of iron overload: mechanisms of iron-induced hepatic fibrogenesis. Semin Liver Dis 2005;25:433–449.
- 4.
McClain CJ, Marsano L, Burk RE, et al. Trace metals in liver disease. Semin Liver Dis 1991;11:321–339.
- 5.
Stamoulis I, Kouraklis G, Theocharis S. Zinc and the liver: an active interaction. Dig Dis Sci 2007;52:1595–1612.
- 7.
Regev A, Berho M, Jeffers LJ, et al. Sampling error and intraobserver variation in liver biopsy in patients with chronic HCV infection. Am J Gastroenterol 2002;97:2614–2618.
- 8.
Abdi W, Millan JC, Mezey E. Sampling variability on percutaneous liver biopsy. Arch Intern Med 1979;139:667–669.
- 9.
Holund B, Poulsen H, Schlichting P. Reproducibility of liver biopsy diagnosis in relation to the size of the specimen. Scand J Gastroenterol 1980;15:329–335.
- 10.
Schlichting P, Holund B, Poulsen H. Liver biopsy in chronic aggressive hepatitis. Diagnostic reproducibility in relation to size of specimen. Scand J Gastroenterol 1983;18:27–32.
- 11.
Poniachik J, Bernstein DE, Reddy KR, et al. The role of laparoscopy in the diagnosis of cirrhosis. Gastrointest Endosc 1996;43:568–571.
- 12.
Cholongitas E, Senzolo M, Standish R, et al. A systematic review of the quality of liver biopsy specimens. Am J Clin Pathol 2006;125:710–721.
- 13.
Van Leeuwen DJ, Balabaud C, Crawford JM, et al. A clinical and histopathologic perspective on evolving noninvasive and invasive alternatives for liver biopsy. Clin Gastroenterol Hepatol 2008;6:491–496.
- 14.↑
Cole TL, Center SA, Flood SN, et al. Diagnostic comparison of needle and wedge biopsy specimens of the liver in dogs and cats. J Am Vet Med Assoc 2002;220:1483–1490.
- 15.↑
Faa G, Nurchi V, Demelia L, et al. Uneven hepatic copper distribution in Wilson's disease. J Hepatol 1995;22:303–308.
- 16.
Ferenci P, Steindl-Munda P, Vogel W. Diagnostic value of quantitative hepatic copper determination in patients with Wilson's disease. Clin Gastroenterol Hepatol 2005;3:811–818.
- 17.
Diaz G, Faa G, Fami AMG, et al. Copper distribution within and between newborn livers. J Trace Elem Electrolytes Health Dis 1990;4:61–64.
- 18.
Milman N, Laursen G, Podenphant J, et al. Trace elements in normal and cirrhotic human liver tissue. Iron, copper, zinc, selenium, manganese, titanium and lead measured by X-ray fluorescence spectroscopy. Liver 1986;6:111–117.
- 19.
Faa G, Diaz G, Farci AMG, et al. Variability of copper levels in biopsy tissue from a cirrhotic liver. J Trace Elem Electrolytes Health Dis 1990;4:49–50.
- 20.
Faa G, Liguori C, Columbano A, et al. Uneven copper distribution in the human newborn liver. Hepatology 1987;7:838–842.
- 21.
Cassidy J, Eva JK. The variations in the concentrations of copper and iron within and between the lobes of pig's liver. Proc Nutr Soc 1958;17:30.
- 22.
Howell JS. Histochemical demonstration of copper in copperfed rats and in hepatocellular degeneration. J Pathol Bacteriol 1959;77:473–483.
- 23.
Bingley JB, Dufty JH. Distribution of copper in the tissues of the bovine neonate and dam. Res Vet Sci 1972;13:8–14.
- 24.
Haywood S. The non-random distribution of copper within the liver of rats. Br J Nutr 1981;45:295–300.
- 25.↑
Su LC, Owen CA, Zollman PE, et al. A defect of biliary excretion of copper in copper-laden Bedlington terriers. Am J Physiol 1982;343:G231–G236.
- 26.
Thornburg LP, Rottinghaus G, McGowan M, et al. Hepatic copper concentrations in purebred and mixed-breed dogs. Vet Pathol 1990;27:81–88.
- 27.↑
Thornburg LP. A perspective on copper and the liver disease in the dog. J Vet Diagn Invest 2000;12:101–110.
- 28.
Linder MC. Introduction and overview of copper as an element. In: Linder MC, ed. Biochemistry of copper. New York: Plenum Press, 1991;1–15.
- 29.
Thornburg LP, Rottinghaus G, Dennis G, et al. The relationship between hepatic copper content and morphologic changes in the liver of West Highland White Terriers. Vet Pathol 1996;33:656–661.
- 30.
She H, Xiong S, Lin M, et al. Iron activates NF-kB in Kupffer cells. Am J Physiol Gastrointest Liver Physiol 2002;283:G719–G726.
- 31.
Center SA. Metabolic, antioxidant, nutraceutical, probiotic, and herbal therapies relating to the management of hepatobiliary disorders. Vet Clin North Am Small Anim Pract 2004;34:67–172.
- 32.↑
Center SA, Warner KL, Erb HN. Liver glutathione concentrations in dogs and cats with naturally occurring liver disease. Am J Vet Res 2002;63:1187–1197.
- 33.
Sternlieb I. Copper and the liver. Gastroenterology 1980;78:1615–1628.
- 34.
Danks DM. Copper and liver disease. Eur J Pediatr 1991;150:142–148.
- 35.↑
Association of Official Analytical Chemists (AOAC). In: Helrich K, ed. Official methods of analysis. 15th ed. Arlington, Va: AOAC, 1990.
- 36.↑
Thornburg LP, Beissenherz M, Dolan M, et al. Histochemical demonstration of copper and copper-associated protein in the canine liver. Vet Pathol 1985;22:327–332.
- 37.↑
Bischoff K, Lamm C, Erb HN, et al. The effects of formalin fixation and tissue embedding of bovine liver on copper, iron, and zinc analysis. J Vet Diagn Invest 2008;20:220–224.
- 38.↑
Nooijen JL, van den Hamer CJA, Houtman JPW, et al.Possible errors in sampling percutaneous liver biopsies for determination of trace element status: application to patients with primary biliary cirrhosis. Clin Chim Acta 1981;113:335–338.
- 39.
Ferenci P. Wilson's disease. Clin Gastroenterol Hepatol 2005;3:726–733.
- 40.
Goldfischer S, Popper H, Sternlieb I. The significance of variations in the distribution of copper in liver disease. Am J Pathol 1980;99:715–730.
- 41.
Guido M, Colloredo G, Fassan M, et al. Clinical practice and ideal liver biopsy sampling standards: not just a matter of centimeters. J Hepatol 2006;44:818–826.
- 42.
Bedossa P, Dargere D, Paradise V. Sampling variability of liver fibrosis in chronic hepatitis. Hepatology 2003;38:1449–1457.
- 43.
Schultheiss PC, Bedwell CL, Hamar DW. Canine liver iron, copper, and zinc concentrations and association with histologic lesions. J Vet Diagn Invest 2002;14:396–402.
- 44.
Pietrangelo A, Gualdi R, Casalgrandi G, et al. Molecular and cellular aspects of iron-induced hepatic cirrhosis in rodents. J Clin Invest 1995;95:1824–1831.
- 45.
Pietrangelo A. Hereditary hemochromatosis—a new look at an old disease. N Engl J Med 2004;350:2383–2397.
- 46.
Bonkovsky HL, Banner BF, Rothman AL. Iron and chronic viral hepatitis. Hepatology 1997;25:759–768.
- 47.
Alla V, Bonkovsky H. Iron in nonhemochromatotic liver diseases. Semin Liver Dis 2005;25:461–472.
- 48.
Jain S, Scheuer PJ, Archer B, et al. Histological demonstration of copper and copper-associated protein in chronic liver disease. J Clin Patho1 1978;31:784–790.
- 49.
Theodossi A, Skene AM, Portmann B, et al. Observer variation in assessment of liver biopsies including analysis by Kappa statistics. Gastroenterology 1980;79:232–241.
- 50.
Thornburg LP, Rottinghaus G, Gage H. Chronic liver disease associated with high hepatic copper concentration in a dog. J Am Vet Med Assoc 1986;188:1190–1191.
- 51.
Webb CB, Twedt DC, Meyer DJ. Copper-associated liver disease in Dalmatians: a review of 10 dogs (1998–2001). J Vet Intern Med 2002;16:665–668.
- 52.
Hoffmann G, van den Ingh TS, Bode P, et al.Copper-associated chronic hepatitis in Labrador Retrievers. J Vet Intern Med 2006;20:856–861.
