Hepatocutaneous syndrome is a complex, progressive, usually fatal disorder simultaneously affecting the skin and liver of dogs.1 Cutaneous lesions associated with HCS are characterized by a crusting, ulcerative, painful dermatosis that affects the mucocutaneous junctions, pressure points, and footpads with classic histologic features of superficial necrolytic dermatitis.1–3 Hepatic lesions are characterized by diffuse, severe, noninflammatory, degenerative vacuolar hepatopathy, leading to parenchymal collapse and formation of proliferative hepatocyte nodules.1
Previous reports4,5 have suggested that HCS is more common in male than female dogs, with a median age of onset of 10 years and a predilection for medium- and small-sized dogs. In a review4 of 110 cases, mixed-breed dogs were most common (28%), with breeds represented by ≥ 10 dogs including Shetland Sheepdog, Cocker Spaniel, and West Highland White Terrier. Reportedly, affected dogs are typically examined because of painful cutaneous lesions, lethargy, inappetence, and weight loss, and common clinicopathologic abnormalities include nonregenerative anemia and high hepatic enzyme activities (especially, high ALP activity). An inconsistent association with glucagon-secreting tumors, diabetes mellitus, and phenobarbital has been described.1,2,4,6,7 At the time of initial examination, affected dogs typically do not have icterus and may or may not have high serum bile acids concentrations.3,4,6 Hepatic ultrasonography usually reveals heterogenous hyperechoic parenchymal echogenicity with innumerable hypoechoic nodules (ie, a Swiss cheese or honeycomb-like pattern). Liver size is variable, ranging from large to small. The cause of HCS remains enigmatic, although an association with low plasma amino acids concentrations has been established.5 Although HCS has sometimes also been referred to as glucagonoma syndrome, hyperglucagonemia is inconsistent, which may reflect difficulties associated with quantifying circulating glucagon concentrations or a true lack of association between HCS and hyperglucagonemia.2 Although various treatments have been recommended, including IV administration of amino acid solutions and nutritional supplements, most dogs die within 6 months after developing cutaneous lesions.3–5
Although HCS has been reported to affect dogs of a wide variety of breeds, over the past decade, the authors have specifically identified numerous Shih Tzus with this condition. The purpose of the study reported here, therefore, was to characterize clinical, clinicopathologic, ultrasonographic, and histologic abnormalities in Shih Tzus with HCS. In addition, we wanted to determine whether there was a heritable component to HCS in this breed.
Methods and Materials
Case selection criteria
Medical records and records of the pathology service at the Cornell University College of Veterinary Medicine from 2000 through 2012 were searched to identify adult (ie, ≥ 5 years old) Shih Tzus in which a diagnosis of HCS had been made. Subsequently, breed enthusiasts and veterinarians who had cared for Shih Tzus with HCS were contacted to identify additional dogs in which HCS had been diagnosed between 1996 and 2014.
Dogs were eligible for inclusion in the study if they had a diffuse, erosive, crusting, painful dermatopathy consistent with HCS, a diffuse degenerative vacuolar hepatopathy associated with proliferative nodules characteristic of HCS, or both. Dogs with cutaneous lesions were included only if they had been examined by a board-certified veterinary dermatologist to verify that the lesions were consistent with HCS and the diagnosis had been confirmed by means of histologic examination of skin or liver biopsy specimens by a board-certified veterinary anatomic pathologist or veterinary dermatohistopathologist. Dogs were excluded if HCS was suspected on the basis of gross examination of cutaneous lesions but diagnostic testing to confirm the diagnosis had not been performed.
Medical records review
Medical records were reviewed to obtain information on signalment, clinical signs, results of clinicopathologic testing (eg, CBC, serum biochemical analyses [including pre- and postprandial serum total bile acids concentrations, serum sex hormone concentrations, and plasma amino acid concentrations], urinalysis, low-dose dexamethasone suppression testing, and ACTH stimulation testing), abdominal ultrasonographic findings (eg, hepatic echogenicity, presence of hepatic nodules [hypoechoic or hyperechoic], irregularity of the hepatic surface, adrenal gland dimensions and contour, presence of adrenal gland nodules, pancreatic and peripancreatic echogenicity, presence of pancreatic or peripancreatic nodules, and presence of abdominal effusion), results of histologic examination of skin and liver specimens, treatment, and outcome. Owners and referring veterinarians were contacted by telephone to obtain information on outcome and survival time.
Adrenal function testing and serum hormone concentrations
For ACTH stimulation testing, serum cortisol concentrations were measured before and 1 hour after IV administration of synthetic ACTHa,b (5 μg/kg [2.3 μg/lb]). For low-dose dexamethasone suppression testing, serum cortisol concentrations were measured before and 6 and 8 hours after IV administration of dexamethasone (0.015 mg/kg [0.007 mg/lb]). For both tests, serum cortisol concentrations were measured by means of chemiluminescent methodsc validated for use in dogs8,9 or radioimmunoassaysd validated for use in dogs.
Serum sex hormone (androstenedione, progesterone, 17-hydroxyprogesterone, and estradiol) concentrations and serum aldosterone concentration were measured by means of validated radioimmunological analyses in samples obtained before and 1 hour after IV administration of synthetic ACTH (5 μg/kg). Serum samples were shipped frozen by overnight delivery to the University of Tennessee College of Veterinary Medicine's Clinical Endocrinology Service for analysis.10
Plasma amino acid concentrations
Plasma amino acid concentrations were measured in heparinized plasma samples obtained after food had been withheld overnight. Concentrations were determined with an automated analyzer as described5 by the Amino Acid Laboratory at the University of California School of Veterinary Medicine.
Histologic examination of skin and liver specimens
When available, H&E-stained, 5-μm-thick sections of skin and liver biopsy specimens were reviewed. Skin biopsy specimens were reviewed by one of the authors (JPK) to characterize lesions and confirm that changes were consistent with HCS. Severity of parakeratotic hyperkeratosis, epidermal hyperplasia, and subcorneal epidermal laminar edema were scored on a scale from 0 to 3, where 0 = absent, 1 = mild, 2 = moderate, and 3 = severe. The presence or absence of suppurative luminal folliculitis or hidradenitis (evidence of bacterial infection), Demodex mites, follicular basement membrane mineralization, serous or cellular crusts, superficial perivascular to interstitial pleocellular dermatitis (variable numbers of lymphocytes, plasma cells, macrophages, neutrophils, and eosinophils), lymphocytic exocytosis, pigmentary incontinence (melanomacrophages or free melanin pigment in the superficial dermis), and keratinocyte apoptosis with satellitosis was also recorded. Keratinocyte apoptosis was considered to be present only if there were > 6 apoptotic or individually necrotic keratinocytes/6-mm section of skin.11 Satellitosis was defined as ≥ 1 lymphocyte surrounding at least 1 necrotic keratinocyte.
When available, unstained sections of liver biopsy specimens were stained with reticulin (to evaluate hepatic cord width, sinusoidal organization, parenchymal collapse, and the presence of proliferative or regenerative foci), Masson trichrome stain (to characterize the extent and location of fibrillar collagen), Prussian blue stain (to identify the location and extent of iron sequestration), rhodanine stain (to identify hepatocellular cytosolic copper-protein aggregates), periodic acid–Schiff stain with and without amylase pretreatment (to determine whether hepatocellular cytosolic vacuoles contained material consistent with glycogen), or oil red O stain (to confirm the degree of microvesicular lipid vacuolation). Morphological features were assigned a grade from 0 to 3, where 0 = absent, 1 = mild, 2 = moderate, and 3 = severe, by 2 of the authors (SAC and SPM) working independently. Scores for the 2 observers were then averaged to obtain a final score.
Hepatic biopsy specimens were examined for the presence and severity of hepatocellular vacuolation, architectural remodeling (presence or absence of parenchymal collapse and regenerative or proliferative foci), bile duct proliferation, inflammatory infiltrates or macrophage clean-up reaction, ballooning degeneration, hepatocellular necrosis, and fibrosis. Iron sequestration in free macrophages, fixed sinusoidal macrophages (Kupffer cells), lipogranulomas (foamy macrophages with engulfed cellular debris), and hepatocytes was characterized and graded in sections stained with Prussian blue stain, with cell counts expressed on the basis of 20 portal tracts and 20 hepatic lobules. Hepatocellular copper accumulation was qualitatively graded in sections stained with rhodanine stain on a scale from 0 (not detected) to 5 (numerous copper granules in ≥ 75% of cells), as described.12
Statistical analysis
Data were examined for normality by means of box-and-whisker plots and the Kolmogorov-Smirnov test. Specific clinical features (eg, sex and clinical signs), hepatic ultrasonographic findings (estimated hepatic size [normal, small, or large], hepatic parenchymal nodules, irregular liver surface contour, ascites, bilateral vs unilateral adrenomegaly, adrenal nodules, and pancreatic abnormalities), and results of hormonal testing (low-dose dexamethasone suppression test, ACTH stimulation test, and serum aldosterone and sex hormone concentrations) were enumerated. Sex distribution of all dogs and of dogs with cutaneous lesions was compared with the expected equivalent distribution (ie, 50% males and 50% females) by use of 2 × 2 tables. Because hematologic and serum biochemical testing had been performed in various reference laboratories with overlapping but different reference intervals, results were summarized as numbers and percentages of dogs with results above, within, or below the respective reference interval.
Associations between age and serum ALP activity, between serum ALP activity and serum alanine aminotransferase activity, between serum ALP activity and serum aspartate aminotransferase activity, and between serum albumin concentration and serum creatinine concentration (with biochemical values represented as fold increase relative to the upper reference limit) were examined by means of the Spearman rank correlation method. Survival time after diagnosis was examined by means of the Kaplan-Meier method for all dogs, for male versus female dogs, and for dogs with versus without cutaneous lesions. Survival times for male versus female dogs and for dogs with versus without cutaneous lesions were compared by use of the Gehan-Wilcoxon test (reflecting differences in short-term survival) and the log-rank test (reflecting differences in long-term survival). Median survival time and its 95% CI were determined.
All statistical analyses were performed with standard software.e Values of P ≤ 0.05 were considered significant.
Results
A total of 36 adult (≥ 5 years old) Shih Tzus examined between 1996 and 2014 that were suspected to have HCS were identified. In 5 dogs, however, the diagnosis had been made on the basis of gross inspection of cutaneous lesions and the dogs were euthanized without additional diagnostic testing because the owners of these dogs had previously cared for related dogs with HCS. Thus, these 5 dogs were excluded; the remaining 31 dogs met the criteria for inclusion in the study.
Of the 31 dogs included in the study, 16 were male (2 sexually intact and 14 castrated) and 15 were female (2 sexually intact and 13 spayed). Median age at the time of diagnosis was 8 years (range, 5 to 14 years). Twenty-five dogs had cutaneous lesions consistent with HCS at the time of initial examination; the remaining 6 initially only had hepatic abnormalities, but 3 of the 6 subsequently (2, 6, and 12 months later) developed cutaneous lesions. In 18 dogs, examination of liver biopsy specimens revealed changes consistent with HCS. In 8 dogs in which liver biopsy was not performed, abdominal ultrasonography revealed lesions typical of HCS (eg, marked Swiss cheese pattern, parenchymal hypoechoic nodules, and irregular liver border). Sex distribution was not significantly imbalanced when considering all dogs (P = 0.80) or just the 28 dogs that developed cutaneous lesions (15 males and 13 females; P = 1.0).
Common clinical features included lethargy, inappetence, weight loss, lameness, footpad hyperkeratosis with painful oozing fissures, and crusting moist ulcerative lesions involving pressure points, mucocutaneous junctions, the flanks, and the lateral aspects of the thorax. For 23 of the 28 dogs with cutaneous lesions, the diagnosis was confirmed by means of histologic examination of skin biopsy specimens. In the remaining 5 dogs, cutaneous lesions were verified to be consistent with HCS by a board-certified veterinary dermatologist.
The 6 dogs that did not have cutaneous lesions initially were all examined because of persistent, unexplained, marked increases in serum ALP activity and ultrasonographic evidence of a severe nodular hepatopathy characterized as having a Swiss cheese or honeycomb-like appearance. Clinical signs in these 6 dogs included inappetence, vomiting, and lethargy. Three of the 6 also had neurobehavioral features of hepatic encephalopathy.
Hematologic, serum biochemical, and urinalysis findings
Results of hematologic and serum biochemical testing were available for 24 dogs (Table 1). The most common hematologic abnormalities were low PCV (11/24 [46%]) and low MCV (15/24 [63%]). Twenty of the 24 (83%) dogs had an RBC count within reference limits, and only 4 (17%) had a low RBC count. Monocytosis without a stress leukogram was observed in 12 of the 24 (50%) dogs, and thrombocytosis was observed in 17 (71%).
Results of clinicopathologic testing in 24 Shih Tzus with HCS.
| Variable | No. of dogs tested | No. (%) with high values* | No. (%) with low values* |
|---|---|---|---|
| PCV | 24 | 0 (0) | 11 (46) |
| RBC count | 24 | 0 (0) | 4 (17) |
| MCV | 24 | 0 (0) | 15 (63) |
| WBC count | 24 | 3 (13) | 0 (0) |
| Neutrophil count | 24 | 4 (17) | 0 (0) |
| Monocyte count | 24 | 12 (50) | 0 (0) |
| Platelet count | 24 | 17 (71) | 0 (0) |
| SUN | 23 | 3 (13) | 6 (26) |
| Creatinine | 22 | 0 (0) | 6 (27) |
| Glucose | 24 | 8 (33) | 0 (0) |
| Total protein | 24 | 0 (0) | 1 (4) |
| Albumin | 24 | 0 (0) | 5 (21) |
| Alanine aminotransferase | 24 | 10 (42) | 0 (0) |
| Aspartate aminotransferase | 17 | 13 (76) | 0 (0) |
| ALP | 24 | 24 (100) | 0 (0) |
| γ-Glutamyltranspeptidase | 16 | 9 (56) | 0 (0) |
| Total bilirubin | 21 | 4 (19) | 0 (0) |
| Cholesterol | 23 | 5 (22) | 1 (4) |
Relative to reference limits established by laboratories that performed the testing. MCV = Mean corpuscular volume.
To verify that low MCV was not a unique breed-related feature of Shih Tzus, MCV and RBC counts for 30 (11 males and 19 females) randomly selected Shih Tzus examined at the Cornell University Hospital for Animals for illnesses not associated with disorders in erythron mass or RBC cell size (eg, anemia, renal failure, or bleeding tendencies) or for hepatobiliary or cutaneous disorders similar to those associated with HCS were examined. For these 30 dogs, median MCV was 67 fL (range, 62 to 71 fL; reference interval, 58 to 79 fL) and median RBC count was 9.5 × 106/μL (range, 4.0 × 106/μL to 16.0 × 106/μL; reference interval, 4.8 × 106/μL to 9.3 × 106/μL).
All 24 dogs in which serum ALP activity had been measured had high activities, with values ranging from 3.0 to 36 (median, 8.5) times the upper reference limit. Ten of 24 (42%) dogs had a high serum alanine aminotransferase activity, with values in all 24 dogs ranging from 0.6 to 6.9 (median, 2.3) times the upper reference limit; 13 of 17 (76%) dogs had a high serum aspartate aminotransferase activity, with values in all 17 dogs ranging from 0.6 to 6.6 (median, 1.8) times the upper reference limit; and 9 of 16 (56%) dogs had a high serum γ-glutamyltranspeptidase activity, with values in all 16 dogs ranging from 0.5 to 5.5 (median, 1.3) times the upper reference limit. Alkaline phosphatase isoform analyses were performed in only 1 dog, and in this dog, the glucocorticoid isozyme predominated.
Eight of the 24 (33%) dogs had hyperglycemia, and insulin-dependent diabetes mellitus was subsequently diagnosed in 4 of these 8 dogs. Serum total bile acids concentration was measured in 9 dogs, and 5 of the 9 had values exceeding the upper reference limit. Biochemical evidence of hepatic failure was sparse, with only a single dog having a serum albumin concentration < 2.0 gm/dL.
Results of urinalyses were available for 18 dogs. Two dogs had ammonium biurate crystalluria reflecting hyperammonemia that may have been caused by acquired hepatic insufficiency or acquired portosystemic shunting (one of these had a history of a congenital portosystemic shunt that had been successfully ligated 8 years earlier). For the remaining dogs, results of urinalysis were unremarkable.
Significant differences in clinicopathologic variables were not found between male and female dogs or between dogs with versus without cutaneous lesions (all P > 0.31). There was no significant association between age and serum ALP activity (expressed as fold increase relative to the upper reference limit; r = 0.13; P = 0.64), between serum ALP activity and serum alanine aminotransferase activity (r = −0.11; P = 0.68), between serum ALP activity and serum aspartate aminotransferase activity (r = 0.39; P = 0.24), or between serum albumin concentration and serum creatinine concentration (r = 0.13; P = 0.64).
Hormonal testing
Six dogs underwent low-dose dexamethasone suppression testing and 12 underwent ACTH stimulation testing. Results were consistent with hyperadrenocorticism in 2 of the 6 dogs that underwent low-dose dexamethasone suppression testing and in 5 of the 12 that underwent ACTH stimulation testing.
Serum sex hormone and aldosterone concentrations were measured before and after ACTH administration in 6 dogs, and ≥ 1 concentration was ≥ 2X the upper reference limit in 5 of the 6. Prior to ACTH administration, hormone concentrations ≥ 2X the upper reference limit were documented for androstenedione in 4 dogs, for progesterone in 3 dogs, and for 17-hydroxyprogesterone in 1 dog. Following ACTH administration, hormone concentrations ≥ 2X the upper reference limit were documented for androstenedione in 5 dogs, for progesterone in 3 dogs, and for 17-hydroxyprogesterone in 4 dogs.
Plasma amino acid concentrations
Preprandial plasma amino acid concentrations in 7 Shih Tzus with HCS (all had classic histologically confirmed hepatic lesions, and 2 lacked cutaneous lesions) were similar to published plasma amino acid profiles from dogs of other breeds with HCS.7 Low plasma amino acid concentrations were noted for 17 amino acids, with a > 50% reduction confirmed for alanine, arginine, asparagine, citrulline, glutamine, lysine, methionine, ornithine, proline, serine, and threonine. On observation, the 2 dogs lacking cutaneous lesions had less severe decreases in plasma amino acid concentrations.
Abdominal ultrasonography
Twenty-two dogs underwent abdominal ultrasonography. Subjectively, the liver was classified as being of normal size in 16, large in 3, and small in 3. Nineteen of the 22 (86%) dogs had a diffusely hyperechoic or complex heteroechoic appearance to the hepatic parenchyma, and 14 (64%) had a Swiss cheese or honeycomb-like pattern typical of HCS. One dog with a small, diffusely nodular liver had abdominal effusion and acquired portosystemic shunts confirmed by means of color flow Doppler ultrasonography. Results of histologic examination of liver biopsy specimens from this dog were suggestive of sinusoidal hypertension secondary to a severe degenerative vacuolar hepatopathy. In 3 dogs, the gallbladder was ultrasonographically abnormal. One dog had a large gallbladder (> 1.2 mL/kg of body weight), 1 had a mineralized gallbladder wall, and 1 had a gallbladder mucocele. One dog had a hepatic mass confirmed histologically to be a hepatocellular carcinoma.
Adrenal gland size was recorded in 16 of the 22 dogs that underwent abdominal ultrasonography. Three of the 16 dogs had adrenomegaly (largest dimension measured in the transverse plane > 0.6 cm), which was bilateral in 1 dog and unilateral in the other 2. Two additional dogs had nodules in the caudal pole of the adrenal glands.
Skin biopsy specimens
For 16 of the 31 dogs, H&E-stained, 5-μm-thick sections of skin biopsy specimens were available for review. Between 2 and 8 specimens were evaluated for each dog.
All 16 dogs for which skin biopsy specimens were available had histologic features consistent with HCS (Figure 1), including locally extensive to diffuse parakeratotic hyperkeratosis (severe in 9, moderate in 4, and mild in 3) and epidermal hyperplasia (severe in 6, moderate in 7, and mild in 3). Laminar subcorneal epidermal edema was seen in specimens from 14 of the 16 dogs (severe in 7, moderate in 4, and mild in 3). Suppurative luminal folliculitis, hidradenitis, or both was seen in specimens from 5 dogs; concurrent demodicosis was identified in 3 dogs, and follicular basement membrane mineralization was seen in 8 dogs. All 16 dogs had superficial perivascular to interstitial pleocellular inflammation (mild to moderate in severity). Lymphocytic exocytosis and cellular to serocellular crusts were observed in 15 dogs, and pigmentary incontinence was seen in 10. Keratinocyte apoptosis with satellitosis was evident in 4 dogs. A single dog had moderate eosinophilic superficial dermal interstitial and eosinophilic epidermal inflammation suggestive of an adverse cutaneous drug eruption or sterile eosinophilic pustulosis. In another dog, cutaneous vasculitis and ulceration were evident in biopsy specimens collected from the nasal planum and haired skin. In yet another dog, multilevel apoptosis, satellitosis, and interface dermatitis suggestive of erythema multiforme were seen in 2 of 3 skin specimens. This dog did not have subcorneal laminar edema but did have moderate diffuse parakeratotic hyperkeratosis supporting the diagnosis of HCS.
Photomicrograph of a section of affected skin from a Shih Tzu with HCS. Notice the typical red, white, and blue epidermal layering, with the red layer reflecting parakeratotic hyperkeratosis and crusting (asterisk); the white layer representing a linear distribution of pale, swollen, edematous keratinocytes of the stratum granulosum and stratum spinosum (arrow); and the blue layer reflecting hyperplasia of the stratum basale (bracket). H&E stain; bar = 200 gm.
Citation: Journal of the American Veterinary Medical Association 248, 7; 10.2460/javma.248.7.802
Liver biopsy specimens
For 15 of the 31 dogs, H&E-stained, 5-μm-thick sections of skin biopsy specimens were available for review. The typical histologic appearance consisted of random multifocal-to-coalescing foci of hepatocytes with severe ballooning degeneration (often associated with glycogen vacuolation and, less commonly, lipid vacuolation) admixed with expansive foci of proliferative hepatocytes and areas of parenchymal collapse. These changes imparted a map-like pattern to hepatic sections (Figure 2). Liver lobules were variably affected, with occasional areas spared.
Low- and high-magnification photomicrographs of hepatic biopsy specimens from a Shih Tzu with HCS. In the low-magnification photomicrograph (A), notice the map-like appearance of proliferative foci contrasting against severely vacuolated (glycogen and lipid; light pink-gray) hepatocytes. In the high-magnification photomicrograph (B), notice the glycogen-type hepatocellular vacuolation and extreme distention of many hepatocytes characteristic of ballooning degeneration. H&E stain; bars = 3 mm (lower magnification) and 200 μm (higher magnification).
Citation: Journal of the American Veterinary Medical Association 248, 7; 10.2460/javma.248.7.802
Glycogen-type hepatocyte vacuolation was seen in specimens from 14 of the 15 dogs (severe in 9 and mild in 5). Vacuoles were clear, rough-bordered, and separated by thin wisps of cytoplasm extending from the hepatocyte membrane to the nucleus and usually did not displace or compress the nucleus. Examination of sections stained with periodic acid–Schiff stain, with and without amylase pretreatment, confirmed that the vacuolar contents were consistent with glycogen (Figure 3).
Photomicrographs of serial sections of a hepatic biopsy specimen from a Shih Tzu with HCS. Sections were stained with periodic acid–Schiff stain without (A) and with (B) amylase pretreatment. In the section without amylase pretreatment, glycogen stains dark magenta. In the other section, pretreatment with amylase has removed the glycogen, confirming the glycogen nature of the cytosolic vacuolation. Periodic acid–Schiff stain; bars = 200 μm.
Citation: Journal of the American Veterinary Medical Association 248, 7; 10.2460/javma.248.7.802
Five dogs had micro- and macrovesicular hepatocyte lipid vacuolation (severe in 3 and moderate in 2). Vacuoles were round and clear with discrete, sharp margins and were most often seen at the margins of proliferative nodules (Figure 4). Oil red O staining of frozen liver sections from 2 dogs confirmed that the vacuole contents were consistent with lipid.
Photomicrographs of hepatic biopsy specimens from a Shih Tzu with HCS. A—Notice the hepatocellular cytosolic microvesicular lipidosis. H&E stain; bar = 200 μm. B—Oil red O staining of a frozen section not embedded in paraffin confirms the neutral fat content of the hepatocellular vacuoles, with fat appearing as orange-red globules on the surface of hepatocytes. Oil red O stain; bar = 100 μm.
Citation: Journal of the American Veterinary Medical Association 248, 7; 10.2460/javma.248.7.802
Proliferative foci of hepatocytes composed of disorganized double-wide hepatic cords were seen in 14 dogs (severe in 10, moderate in 3, and mild in 1). Expansive proliferative nodules lacked the connective tissue mantle characteristic of regenerative nodules associated with necroinflammatory liver disorders (Figure 5). Interdigitation of proliferative and vacuolated hepatocytes at the edge of proliferative foci created a jagged, moth-eaten margin.
Photomicrograph of a hepatic biopsy specimen from a Shih Tzu with HCS. Notice the jagged interface between severely vacuolated hepatocytes and proliferative foci. Proliferative foci consist of small angular hepatocytes with a disorganized hepatic cord structure. H&E stain; bar = 200 μm.
Citation: Journal of the American Veterinary Medical Association 248, 7; 10.2460/javma.248.7.802
Bile duct proliferation was observed in 6 of the 15 dogs. Proliferation was mild in 5 of the 6 dogs but was moderate in 1 dog with cholangiohepatitis associated with a gallbladder mucocele. Individual hepatocyte necrosis was rare (2/15). Reticulin staining revealed parenchymal collapse that lacked zonal tropism in 9 dogs (severe in 1, moderate in 4, and mild in 4) and also reduced reticulin substructure within proliferative foci. Four of the 9 dogs with parenchymal collapse had portal-to-portal bridging, 2 had central-to-central bridging, 1 had central-to-central and portal-to-portal bridging, and 2 had random foci of ballooning degeneration. Occasionally, central veins (hepatic venules) were entrapped in bridging partitions formed by amalgamations of extinct parenchyma and slender collagen tendrils. Masson trichrome staining confirmed minimal thin dissecting tendrils of fibrillar collagen in areas of parenchymal collapse or regeneration in 6 dogs. When present, thin collagen fibrils dissected along the space of Disse in a chicken-wire pattern around hepatocytes undergoing ballooning degeneration. Tiny thin collagen tendrils often accompanied minimal mixed inflammatory infiltrates (lymphocytes, plasma cells, macrophages, and neutrophils) in portal (n = 6 dogs) or centrilobular (2) regions or both. However, most of these dogs had only scant inflammatory infiltrates, with only 2 having an association with bridging fibrosis.
Histologic features reflective of portal vein hypoperfusion most likely secondary to developmental malformations of the portal vein were seen in 6 dogs; however, none of these dogs had ascites, acquired portosystemic shunts (as determined by abdominal ultrasonography), or ammonium biurate crystalluria. Features consistent with portal vein hypoperfusion included lobular atrophy (eg, close approximation of portal and centrilobular regions and miniaturized portal tracts) and arteriolar portal tract perfusion (eg, increased arterial cross sections in portal tracts and small isolated arteries in the hepatic parenchyma). One dog with portal vein hypoperfusion had had a congenital extrahepatic portosystemic shunt attenuated at 1 year of age. The remaining dogs with portal vein hypoperfusion were suspected to have had congenital hepatoportal microvascular dysplasia.
Rhodanine staining confirmed centrilobular hepatocellular copper accumulation in 11 of 14 dogs (4 dogs with a score of 1,4 dogs with a score of 2, and 3 dogs with a score of 3 on a scale from 0 to 5). The 3 dogs with moderate copper accumulation (score of 3) had lipofuscin-laden macrophages in centrilobular regions closely associated with hepatocytes that contained copper.
Prussian blue staining for iron was performed on hepatic biopsy specimens from 13 dogs. When portal tracts were examined, 5 of the 13 dogs had no macrophages with stainable iron in the portal tracts, 2 had ≤ 50 macrophages with stainable iron/20 portal tracts, and 6 had 50 to 100 macrophages with stainable iron/20 portal tracts. Enumeration of iron-laden lipogranulomas revealed that 5 dogs had no iron-laden lipogranulomas, 2 had ≤ 50 lipogranulomas with stainable iron/20 liver lobules, and 6 had > 50 lipogranulomas with stainable iron/20 liver lobules. Moderate to marked panlobular mesenchymal iron accumulation was observed in 3 dogs. Two of these had been treated with injectable iron dextran on recommendation of a veterinary nutritionist who suspected iron deficiency because the dog had microcytosis and low iron saturation.
Pedigree analyses
Detailed pedigree information was available for 18 of the 31 dogs, and 12 of these 18 dogs were found to have a common ancestry (Figure 6). Of the remaining 6 dogs for which detailed pedigree information was available, 4 were from unrelated kindreds in the United States, 1 was from an unrelated kindred of Canadian ancestry, and 1 was from an unrelated kindred of European ancestry.
Pedigree of a Shih Tzu kindred that included 12 dogs with HCS. Notice the common founders (1,2) for all 12 affected dogs. Circle = Female. Box = Male. Black symbol = Affected dog.
Citation: Journal of the American Veterinary Medical Association 248, 7; 10.2460/javma.248.7.802
For an additional 8 dogs, detailed pedigree information was not available, but all 8 were from the same breeding kennel that produced the 12 dogs known to have a common ancestry. For the remaining 5 dogs, no information on pedigree was available.
Treatment
Seventeen dogs were treated for cutaneous lesions and severe degenerative vacuolar hepatopathy. All 17, except 1 in end-stage liver failure, were fed a normal maintenance diet or a high-protein diet. Twelve dogs received IV amino acid infusions ranging in concentration from 3.0% to 10%.f–h Infusion intervals were tailored on the basis of response of the cutaneous lesions but typically ranged from weekly to monthly. Protein supplements were provided for 16 dogs as protein powder, egg yolks, or combined individual amino acid and metabolic supplements recommended by veterinary specialists (dermatologists, internists, and nutritionists). Zinc methionine was prescribed for 12 dogs, and essential fatty acid supplements containing omega-3 fatty acidsi,j were prescribed for 12 dogs. Fourteen dogs were prescribed antioxidant nutraceuticals, including s-adenosylmethioninek and s-adenosylmethionine with silibinin-phosphatidylcholine complex.l Systemic antimicrobial administration and topical foot soaks were recommended for most dogs, and tramadolm was often dispensed for relief from discomfort caused by fissured footpads. Three dogs were treated for suspected hyperadrenocorticism with mitotanen (n = 2) or trilostaneo (1).
Outcome
One dog was lost to follow-up. Median survival time was 3.0 months (range, 1 to 36 months), with 50% of dogs still alive at 2.5 months (95% CI, 2 to 5 months; Figure 7). Most dogs were euthanized because of severe signs of footpad pain or secondary cutaneous infections resulting in poor life quality; however, several owners had previously owned related Shih Tzus with HCS and elected euthanasia immediately after a diagnosis of HCS was made. Survival time was not significantly (P > 0.51) different between males and females, and subjectively, survival time did not appear to differ between dogs with and without cutaneous lesions. However, the longest survival times were recorded for 3 dogs with hepatopathy without cutaneous lesions or with only mild cutaneous lesions. The owner of 1 dog with hepatopathy without cutaneous lesions elected euthanasia 2 weeks after the diagnosis was made because of the expected poor prognosis. Subjectively, treatment of suspected hyperadrenocorticism did not appear to alter survival time, compared with survival time for dogs that were not treated for suspected hyperadrenocorticism.
Survival curve (survival time after definitive diagnosis) for 28 Shih Tzus with HCS. Two dogs were still alive, and 1 dog was lost to follow up.
Citation: Journal of the American Veterinary Medical Association 248, 7; 10.2460/javma.248.7.802
Discussion
Results of the present study indicated that clinical, clinicopathologic, ultrasonographic, and histologic abnormalities in Shih Tzus with HCS were similar to those previously reported for affected dogs of other breeds. In addition, our findings suggested that HCS may have a heritable component in Shih Tzus, although the condition was also identified in Shih Tzus without any relatives known to be affected.
In the present study, analysis of detailed pedigree information for 18 Shih Tzus with HCS revealed an apparent founder effect and lack of sex predilection, with heritability most consistent with an autosomal recessive tract. However, the frequency of this condition was unclear because information regarding litter size and when signs initially emerged was unavailable. Unfortunately, detailed pedigrees were not available for all dogs in the study. Nevertheless, it was clear that HCS extended beyond a single kindred because 4 dogs were from unrelated kindreds in the United States, 1 was from an unrelated kindred of Canadian ancestry, and 1 was from an unrelated kindred of European ancestry. Importantly, we could not estimate the incidence of HCS among Shih Tzus because we networked with breeders, owners, and referring veterinarians who had cared for affected dogs, soliciting case submissions after recognition of the initial 4 dogs.
Median age at the time of diagnosis for dogs in the present study (8 years) was slightly younger than that reported for dogs of other breeds with HCS (10 to 11 years).1,5,6 Although HCS has been reported to be more common in small- and medium-sized dogs, to our knowledge, a heritable component has not been suggested previously, possibly because previous retrospective case series4,5 contained small numbers of Shih Tzus.
Clinical findings for Shih Tzus with HCS in the present study were similar to those described for affected dogs of other breeds, including lethargy, inappetence, weight loss, lameness, footpad hyperkeratosis with painful oozing fissures, and crusting moist ulcerative lesions involving pressure points and mucocutaneous junctions. Clinical signs of liver failure and hepatic encephalopathy, however, were uncommon. Typical hepatic abnormalities included severe, diffuse, degenerative glycogen or microvesicular lipid vacuolation with proliferative hepatocellular foci and absent to minimal collagen deposition. Hepatocyte necrosis, biliary hyperplasia, accumulation of lipofuscin-laden macrophages, and inflammatory infiltrates were absent, minimal, inconsistent, or focal.
Hepatocytes with stainable cytosolic copper-protein aggregates were observed in 11 of the 14 dogs in the present study for which liver sections were available for rhodanine staining. Moderate hepatic copper accumulation was confirmed in 3 dogs and may have contributed to oxidative liver injury in these dogs. We were not surprised to identify copper accumulation in a subset of dogs. A previous study12 of Labrador Retrievers demonstrated an association between copper accumulation and modifications in copper dietary supplementation suggested by the Association of American Feed Control Officials in 1997. All but a single biopsy specimen in the present study were collected from dogs after 1997; consequently, the copper accumulation in these dogs may have reflected increased dietary copper exposure. Stainable iron was confirmed by means of Prussian blue staining in the portal tracts of 8 of 13 dogs, likely reflecting macrophage scavenging of heme released from hepatocyte cytochromes as cells succumbed to vacuolar degeneration. It is widely acknowledged that iron accumulation in Kupffer cells and copper accumulation in hepatocytes increase the risk of hepatic oxidant injury through multifactorial mechanisms.13 Thus, it is plausible that accumulation of these transition metals may have contributed to hepatic injury.
All 16 dogs in the present study for which skin biopsy specimens were available had the distinctive red, white, and blue epidermal layering associated with HCS, with the red layer reflecting parakeratotic hyperkeratosis and crusting, the white layer representing linear distribution of pale, swollen, edematous keratinocytes of the stratum granulosum and stratum spinosum, and the blue layer reflecting hyperplasia of the stratum basale. The absence of laminar subcorneal epidermal edema in biopsy specimens from 2 dogs (footpad and haired skin samples) was not surprising because this feature may be absent in dogs with chronic HCS, with parakeratotic hyperkeratosis predominating.14 Superficial perivascular to interstitial pleocellular inflammation is common in skin specimens from dogs with HCS and was present in all 16 dogs in the present study for which skin biopsy specimens were available for review.
Lymphocytic exocytosis was seen in 15 of 16 dogs in the present study, and exocytosis of inflammatory cells through epidermal intercellular spaces is a common nonspecific feature of inflammatory dermatoses. For example, lymphocytic exocytosis has been documented in association with interface dermatitis, Malassezia dermatitis, atopic dermatitis, contact hypersensitivity, seborrheic disorders, and some ectoparasitisms.11 Pigmentary incontinence, defined as free melanin granules within the superficial dermis and dermal macrophages, was identified in 10 of 16 dogs in the present study and is associated with any process that damages the stratum basale and basement membrane zone (eg, interface dermatoses and many other conditions, such as Malassezia dermatitis).11 Five dogs had suppurative luminal folliculitis, hidradenitis, or both, consistent with bacterial folliculitis developing secondary to HCS. Keratinocyte apoptosis with satellitosis was evident in 4 dogs and has been previously described in dogs with HCS. This is a feature of cell-mediated cytotoxicosis and has been proposed to represent erythema multiforme superimposed on HCS.14 Alternatively, erythema multiforme is proposed to reflect protein deficiency or imbalance in older dogs14 and may have contributed to the cutaneous lesions. It was interesting that 3 dogs had concurrent demodicosis, as this association has not previously been recognized in dogs with HCS. Adult-onset demodicosis is known to occur in dogs with metabolic, immunosuppressive, or debilitating disorders (eg, hyperadrenocorticism, hypothyroidism, malignant neoplasia, leishmaniasis, and immunosuppressive treatments for cancer or autoimmune diseases).15 Eight dogs had basement membrane mineralization involving at least 1 hair follicle, a common feature associated with glucocorticoid administration, hyperadrenocorticism, or aging in dogs. Considering that results of adrenal function testing were abnormal in 7 dogs, it is possible that additional endocrine-related changes (eg, epidermal, follicular, and sebaceous gland atrophy) may have been observed if additional biopsy specimens had been collected from dorsal truncal regions.
The most common hematologic abnormalities for dogs in the present study were microcytosis (15/24) and low PCV (11/24); however, most dogs (20/24) had an RBC count within reference limits. Finding microcytosis was not unexpected, as this feature has previously been reported in 40% to 67% of dogs with HCS in previous case series.6,16–20 Microcytosis can reflect iron deficiency and is associated with portosystemic shunting in dogs but not with anemia of chronic disease.21–24 However, it was unlikely that microcytosis reflected iron deficiency in dogs in the present study because it was not responsive to iron dextran injection in 2 dogs. In fact, parenterally administered iron dextran was associated with marked iron sequestration in fixed sinusoidal macrophages (Kupffer cells) in both of these dogs. We determined that microcytosis was not a breed-related phenomenon by inspecting hematologic data for 30 randomly selected Shih Tzus from our hospital database, excluding dogs with illnesses that might have influenced erythron mass or RBC size. Observational studies21–23 of microcytosis in dogs with portosystemic shunting have not elucidated an underlying cause. Microcytosis has also been identified in dogs with dietary vitamin B6 (pyridoxine) deficiency.25 However, vitamin B6 deficiency has not been explored as a cause of microcytosis in spontaneously ill dogs. Among other things, vitamin B6 deficiency can compromise collagen synthesis and may enhance transcription of the glucocorticoid receptor.26,27 These effects could potentially contribute to cutaneous lesions and systemic features of HCS. For example, reduced collagen synthesis might promote cutaneous wounds by delaying healing, and increased glucocorticoid receptor transcription might augment glycogen-type hepatocellular vacuolation and cause increases in ALP activity.
Twelve of 24 dogs in the present study had monocytosis without other hematologic changes typical of a glucocorticoid-induced stress leukogram. The monocytosis, therefore, might have been a result of systemic inflammation initiated by the cutaneous lesions themselves or by secondary infection of the cutaneous lesions. Thrombocytosis was identified in 17 of 24 dogs and might have reflected systemic inflammation resulting in increased hepatic thrombopoietin production or an endogenous catecholamine stress response causing splenic contraction, because no other causal disorders were identified, such as splenic disease or dysfunction, paraneoplastic effects, blood loss, bone marrow disease, leukemia, or iron deficiency.28–30 To our knowledge, monocytosis and thrombocytosis have not previously been characterized as hematologic features of HCS in dogs.
The major serum biochemical abnormalities among dogs in the present study included marked increases in serum ALP activity with lesser increases in serum alanine aminotransferase, aspartate aminotransferase, and γ-glutamyltranspeptidase activities, as has previously been reported for dogs with HCS.1,3 The cause of the high serum ALP activity in dogs with HCS remains enigmatic because there is no evidence of biliary structural, mechanical, or metabolic cholestasis, with relatively few patients progressing to hyperbilirubinemia. Considering these factors, an induction phenomenon seems most plausible and consistent with a recent suggestion that ALP plays a protective physiologic role similar to the acute phase immune and inflammatory responses.31 Yet, this hypothesis might not account for the high serum ALP activity in dogs lacking cutaneous lesions. Evaluation of ALP isoforms is generally not used to differentiate between diseases in dogs because the glucocorticoid-associated isoform typically predominates in dogs with chronic illness.32–34 A retrospective study35 of dogs with glycogen-type vacuolar hepatopathy suggested an association between ALP activity and stressful or cytokine-inducing systemic non-adrenal disease or neoplasia. There is no evidence of osteomyelitis associated with pedal inflammation in dogs with HCS that might initiate bone ALP activity, and the magnitude of enzyme activity exceeds that expected with bone remodeling or lysis.34
Hyperglycemia was identified in 8 dogs in the present study but was associated with insulin-dependent diabetes mellitus in only 4. Percentages of cases with insulin-dependent diabetes mellitus in previous studies of dogs with HCS have been higher (18%,6 27%,5 and 41%1), and a relationship among hyperglycemia, diabetes mellitus, and HCS in dogs has been suspected since 19867 with a less consistent relationship between HCS and degenerative (glycogen-type and microvesicular lipid type) vacuolar hepatopathy and glucagon-secreting or neuroendocrine neoplasia suspected since 1990.18 While often discussed, the relationship between hyperglucagonemia or neuroendocrine neoplasia and HCS has only been confirmed in < 10% (8/87) of dogs reported since then.6,18–20,36–42 Confident diagnosis of hyperglucagonemia remains problematic because of the lack of a well-validated canine assay and complexities caused by variable circulating molecular forms of glucagon.2,7 Thus, the cause of hyperglycemia and insulin-dependent diabetes mellitus in dogs with HCS remains unexplained. No causal factors were identified in the dogs of the present report with insulin-dependent diabetes or in dogs with transient hyperglycemia, in that none of the dogs had a history of chronic recurrent or severe pancreatitis or ultrasonographic evidence of a pancreatic mass lesion. While hepatic insufficiency is associated with diabetes mellitus in human beings, there is no evidence that this also occurs in dogs with hepatic disorders other than HCS.
Abdominal ultrasonography was an integral diagnostic test for dogs in the present study, with 14 of 22 dogs having the characteristic Swiss cheese or honeycomb-like ultrasonographic pattern that has been reported previously.16,17 Liver size was variable, but was rarely small, and abdominal effusion was rare.
Measurement of plasma amino acid concentrations in 7 dogs of the present study revealed the low amino acid concentrations reportedly typical of dogs with HCS.5 The etiopathogenesis of hypoaminoacidemia remains undetermined, although it has been hypothesized to reflect amino acid depletion secondary to hyperglucagonemia and hepatic insufficiency. Plasma amino acid concentrations in dogs with HCS dramatically differ from concentrations in dogs with experimentally induced acute severe liver injury, congenital portosystemic shunts, or spontaneous chronic necroinflammatory hepatitis.43,44 Subjectively, amino acid depletion appeared to be less severe in dogs in the present study with hepatic lesions that lacked concurrent skin lesions. Unfortunately, the small number of cases precluded statistical evaluation of this observation.
Assessment of adrenal function was performed in some dogs in the present study because the primary care clinician suspected hyperadrenocorticism or because adrenomegaly or adrenal gland nodules were discovered during abdominal ultrasonography. Overall, results were considered consistent with hyperadrenocorticism in 7 of the 13 dogs that were tested. Evaluation of serum sex hormone concentrations before and after exogenous ACTH administration in 6 dogs confirmed high sex hormone concentrations in the absence of hypercortisolemia in 5 of the 6, with each dog having at least 1 hormone concentration more than twice the upper reference limit. These findings suggested that adrenal gland hyperfunction may contribute to vacuolar hepatopathy and ALP induction in some dogs. However, hormone concentration abnormalities varied widely among dogs, concentrations were variable following ACTH administration, and values did not appear subjectively to predict severity of clinical signs or survival time. We postulate that stress associated with debilitating painful cutaneous and footpad lesions and changes in hormone catabolism due to evolving hepatic insufficiency may have contributed to the hormonal testing findings. Two dogs treated with mitotane and 1 dog treated with trilostane for suspected hyperadrenocorticism had survival times similar to those for untreated dogs. Thus, on the basis of this small number of cases, adrenal modulation did not appear to alter the clinical disease course.
Treatment of dogs in the present study included conventionally recommended therapies.3 Irrespective of treatment, cutaneous lesions relentlessly progressed or persisted, leading to a 50th percentile survival time of only 2.5 months (95% CI, 2 to 5 months). One dog with cutaneous lesions responded to amino acid infusions with cutaneous lesions remitting for 10 months, and 2 dogs without cutaneous lesions survived for 17 and 36 months without amino acid infusions.
In conclusion, the present study characterized the clinical, clinicopathologic, and histologic features of HCS in a single breed of dog and suggested that HCS in this breed likely was heritable. Because of the retrospective nature of the study, we were unable to characterize clinicopathologic features, plasma amino acid concentrations, liver morphology, or survival age in unaffected parents or litter mates. Evaluation of pedigrees for a subset of the dogs in this study revealed consanguineous breeding within 3 to 5 generations in a single large kindred with an apparent founder effect. However, we also documented affected Shih Tzus from unrelated kindreds, suggesting that this disorder may be sporadically encountered in this breed in routine clinical practice.
Acknowledgments
No third-party funding or support was received in connection with this study or the writing or publication of the manuscript.
ABBREVIATIONS
| ALP | Alkaline phosphatase |
| HCS | Hepatocutaneous syndrome |
Footnotes
Cortrosyn, Amphastar Pharmaceuticals, Rancho Cucamonga, Calif.
Cortrosyn, Organon Inc, West Orange, NJ.
Immulite 1000, Siemens Medical Solutions Diagnostics Inc, Siemens Healthcare Diagnostics, Deerfield, Ill.
Cortisol Kit, Siemens Healthcare Diagnostics Inc, Deerfield, Ill.
Statistix, version 9, Analytical Software, Tallahassee, Fla.
ProcalAmine 3.0%, B. Braun Medical Inc, Bethlehem, Pa.
Aminosyn 8.5%, Abbott Nutrition, Lake Forest, Ill.
Aminosyn 10.0%, Abbott Nutrition, Lake Forest, Ill.
Wellactin, Nutramax Laboratories Veterinary Sciences Inc, Lancaster, SC.
Omega-3 Freeform Snip Tips, DVM Pharmaceuticals Inc, Miami, Fla.
Denosyl, Nutramax Laboratories Veterinary Sciences Inc, Lancaster, SC.
Denamarin, Nutramax Laboratories Veterinary Sciences Inc, Lancaster, SC.
Tramadol, Janssen Pharmaceuticals Inc, Titusville, NJ.
Lysodren, Bristol-Myers Squibb Co, Princeton, NJ.
Trilostane, Dechra Veterinary Products LLC, Overland Park, Kan.
References
1. Gross TL, Song MD, Havel PJ, et al. Superficial necrolytic dermatitis (necrolytic migratory erythema) in dogs. Vet Pathol 1993; 30:75–81.
2. Turek MM. Cutaneous paraneoplastic syndromes in dogs and cats: a review of the literature. Vet Dermatol 2003; 14:279–296.
3. Byrne KP. Metabolic epidermal necrosis-hepatocutaneous syndrome. Vet Clin North Am Small Anim Pract 1999; 29:1337–1355.
4. Outerbridge CA. Hepatocutaneous syndrome. In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine. 7th ed. St Louis: Saunders Elsevier, 2010;112–115.
5. Outerbridge CA, Marks SL, Rogers QR. Plasma amino acid concentrations in 36 dogs with histologically confirmed superficial necrolytic dermatitis. Vet Dermatol 2002; 13:177–186.
6. March PA, Hillier A, Weisbrode SE, et al. Superficial necrolytic dermatitis in 11 dogs with a history of phenobarbital administration (1995–2002). J Vet Intern Med 2004; 18:65–74.
7. Walton DK, Center SA, Scott DW, et al. Ulcerative dermatosis associated with diabetes mellitus in the dog. A report of four cases. J Am Anim Hosp Assoc 1986; 22:79–88.
8. Reimers TJ, Salerno VJ, Lamb SV. Validation and application of solid-phase chemiluminescent immunoassays for diagnosis of endocrine diseases in animals. Comp Haematol Int 1996; 6:170–175.
9. Scott-Moncrieff JC, Koshko MA, Brown JA, et al. Validation of a chemiluminescent enzyme immunometric assay for plasma adrenocorticotropic hormone in the dog. Vet Clin Pathol 2003; 32:180–187.
10. Frank LA, Rohrbach BW, Bailey EM, et al. Steroid hormone concentration profiles in healthy intact and neutered dogs before and after cosyntropin administration. Domest Anim Endocrinol 2003; 24:43–57.
11. Scott DW. Diagnostic methods. In: Scott DW, Miller WH, Griffin CE, eds. Muller & Kirk's small animal dermatology. 6th ed. Philadelphia: Saunders, 2001;71–76.
12. Johnston AN, Center SA, McDonough SP, et al. Hepatic copper concentrations in Labrador Retrievers with and without chronic hepatitis: 72 cases (1980–2010). J Am Vet Med Assoc 2013; 242:372–380.
13. Videla LAI, Fernández V, Tapia G, et al. Oxidative stress-mediated hepatotoxicity of iron and copper: role of Kupffer cells. Biometals 2003; 16:103–111.
14. Gross TL. Necrotizing diseases of the epidermis. In: Gross TL, Ihrke PJ, Walder EJ, et al, eds. Skin diseases of the dog and cat: clinical and histopathologic diagnosis. 2nd ed. Ames, Iowa: Blackwell, 2005;75–104.
15. Miller WH. Parasitic skin disease. In: Miller WH, Griffin CE, Campbell KL, eds. Muller & Kirk's small animal dermatology. 7th ed. St Louis: Elsevier, 2013;284–342.
16. Jacobson LS, Kirberger RM, Nesbit JW. Hepatic ultrasonography and pathological findings in dogs with hepatocutaneous syndrome: new concepts. J Vet Intern Med 1995; 9:399–404.
17. Nyland TG, Barthez PY, Ortega RM, et al. Hepatic ultrasonographic and pathologic findings in dogs with canine superficial necrolytic dermatitis. Vet Radiol Ultrasound 1996; 37:200–205.
18. Allenspach K, Arnold P, Glaus T, et al. Glucagon-producing neuroendocrine tumour associated with hypoaminoacidaemia and skin lesions. J Small Anim Pract 2000; 41:402–406.
19. Miller WH, Anderson WI, McCann JP. Necrolytic migratory erythema in a dog with a glucagon-secreting endocrine tumor. Vet Dermatol 1991; 2:179–182.
20. Cave TA, Evans H, Hargreaves J, et al. Metabolic epidermal necrosis in a dog associated with pancreatic adenocarcinoma, hyperglucagonaemia, hyperinsulinaemia, and hypoaminoacidaemia. J Small Anim Pract 2007; 48:522–526.
21. Bunch SE, Jordan HL, Sellon RK, et al. Characterization of iron status in young dogs with portosystemic shunt. Am J Vet Res 1995; 56:853–858.
22. Laflamme DP, Mahaffey EA, Allen SW, et al. Microcytosis and iron status in dogs with surgically induced portosystemic shunts. J Vet Intern Med 1994; 8:212–216.
23. Simpson KW, Meyer DJ, Boswood A, et al. Iron status and erythrocyte volume in dogs with congenital portosystemic vascular anomalies. J Vet Intern Med 1997; 11:14–19.
24. Feldman BF, Kaneko JJ, Farver TB. Anemia of inflammatory disease in the dog: clinical characterization. Am J Vet Res 1981; 42:1109–1113.
25. Fouts PJ, Helmer OM, Lepkovsky S, et al. Production of microcytic hypochromic anemia in puppies on synthetic diet deficient in rat antidermatitis factor (vitamin B6). J Nutr 1938; 16:197–207.
26. Allgood VE, Powell-Oliver FE, Cidlowski JA. The influence of vitamin B6 on the structure and function of the glucocorticoid receptor. Ann N Y Acad Sci 1990; 585:452–465.
27. Merrill AH, Jr., Henderson JM. Diseases associated with defects in vitamin B6 metabolism or utilization. Annu Rev Nutr 1987; 7:137–156.
28. Neel JAI, Snyder L, Grindem CB. Thrombocytosis: a retrospective study of 165 dogs. Vet Clin Pathol 2012; 41:216–222.
29. Hollen CW, Henthorn J, Koziol JA, et al. Elevated serum interleukin-6 levels in patients with reactive thrombocytosis. Br J Haematol 1991; 79:286–290.
30. Beguin Y. Erythropoietin and platelet production. Haematologica 1999; 84:541–547.
31. Pike AF, Kramer NI, Blaauboer BJ, et al. A novel hypothesis for an alkaline phosphatase “rescue” mechanism in the hepatic acute phase immune response. Biochim Biophys Acta 2013: 1832;2044–2056.
32. Hoffmann WE. Diagnostic value of canine serum alkaline phosphatase and alkaline phosphatase isoenzymes. J Am Anim Hosp Assoc 1977; 13:237–241.
33. Solter PF, Hoffmann WE, Hungerford LL, et al. Assessment of corticosteroid-induced alkaline phosphatase isoenzyme as a screening test for hyperadrenocorticism in dogs. J Am Vet Med Assoc 1993; 203:534–538.
34. Center SA. Interpretation of liver enzymes. Vet Clin North Am Small Anim Pract 2007; 37:297–333.
35. Sepesy LM, Center SA, Randolph JF, et al. Vacuolar hepatopathy in dogs: 336 cases (1993–2005). J Am Vet Med Assoc 2006; 229:246–252.
36. Gross TL, O'Brien TD, Davies AP, et al. Glucagon-producing pancreatic endocrine tumors in two dogs with superficial necrolytic dermatitis. J Am Vet Med Assoc 1990; 197:1619–1622.
37. Yoshida M, Barata K, Ando-Lu J, et al. A case report of superficial necrolytic dermatitis in a Beagle dog with diabetes mellitus. Toxicol Pathol 1996; 24:498–501.
38. Torres S, Johnson K, McKeever P, et al. Superficial necrolytic dermatitis and a pancreatic endocrine tumour in a dog. J Small Anim Pract 1997; 38:246–250.
39. Hill PB, Auxilia ST, Munro E, et al. Resolution of skin lesions and log-term survival in a dog with superficial necrolytic dermatitis and liver cirrhosis. J Small Anim Pract 2000; 41:519–523.
40. Oberkirchner U, Linder KE, Zadrozny L, et al. Successful treatment of canine necrolytic migratory erythema (superficial necrolytic dermatitis) due to metastatic glucagonoma with octreotide. Vet Dermatol 2010; 21:510–516.
41. Bach JF, Glasser SA. A case of necrolytic migratory erythema managed for 24 months with intravenous amino acid and lipid infusions. Can Vet J 2013; 54:873–875.
42. Isidoro-Ayza M, Lloret A, Bardagí M, et al. Superficial necrolytic dermatitis in a dog with an insulin-producing pancreatic islet cell carcinoma. Vet Pathol 2014; 51:805–808.
43. Strombeck DR, Harrold D, Rogers Q, et al. Plasma amino acid, glucagon, and insulin concentration in dogs with nitrosamine-induced hepatic disease. Am J Vet Res 1983; 44:2028–2036.
44. Strombeck DR, Rogers Q. Plasma amino acid concentrations in dogs with hepatic disease. J Am Vet Med Assoc 1978; 173:93–96.
