Clinicopathological, Ultrasonographic, and Histopathological Findings of Superficial Necrolytic Dermatitis With Hepatopathy in a Cat

Case Report

A 5-year-old, neutered male Maine coon was presented with a 1-week history of alopecia involving the ventral thorax and abdomen. The cat tested negative for feline leukemia virus and feline immunodeficiency virus as a kitten, and was kept completely indoors, sharing the household with another cat that was healthy. The cat was current on vaccines for feline herpesvirus - 1, calicivirus, panleukopenia, leukemia, and rabies. The owners reported no other abnormalities or clinical signs at home.

Ten months prior to presentation, the cat had an acute onset of vomiting, diarrhea, and anorexia. A serum biochemistry profile performed at that time revealed an elevated alanine aminotransferase (ALT, 278 U/L; reference range, 10 to 100 U/L). Abdominal radiographs were suggestive of an intestinal foreign body. An exploratory laparotomy was performed, and no foreign body was found. The liver appeared normal in size with a mottled appearance, but it was not biopsied. The cecum appeared thickened and irregular. Biopsy of the cecum revealed moderate subacute typhlitis. The cat was treated with orbifloxacin (3.5 mg/kg body weight per os [PO] q 24 hours for 7 days) and made an apparent complete recovery.

Upon presentation to the referring veterinarian, the cat was found to be in good body condition and weighed 6.72 kg. The cat was mildly pruritic. Bilateral, ventral alopecia was noted, beginning in the axillae and extending caudally down the medial thighs to both tarsi. The skin in these areas was moderately erythematous, and mild crusting was present. Mild gingivitis and dental calculus, hepatomegaly, and a grade II/VI left-sided holosystolic heart murmur were also found on physical examination.

A complete blood count (CBC), serum biochemistry profile, urinalysis, serum thyroxine concentration, deep and superficial skin scrapings, and dermatophyte culture were performed. Abnormalities on the CBC included a moderate neutropenia (1.29 × 10³/μL; reference range, 2.5 to 8.5 × 10³/μL). The serum biochemical profile revealed an elevated ALT (444 U/L; reference range, 10 to 100 U/L) and aspartate aminotransferase (AST, 213 U/L; reference range, 10 to 100 U/L). Urinalysis was unremarkable, and serum thyroxine concentration was within reference range. The skin scrapings and dermatophyte culture were negative.

An abdominal ultrasound and echocardiogram were performed. The abdominal ultrasound showed multiple, diffuse nodules throughout the liver, as well as distended hepatic vasculature. An echocardiogram was performed, and mild hypertrophy of the septal myocardium was noted with maintenance of normal systolic function. Because these echocardiographic findings were consistent with hypertrophic cardiomyopathy, the cat was treated with atenolol (1 mg/kg body weight, PO q 24 hours) and was referred to the Veterinary Hospital of the University of Pennsylvania (VHUP) for further evaluation of its hepatic and dermatological abnormalities.

Upon presentation to VHUP, the cat’s body weight and the character of the heart murmur were unchanged. Alopecia and excoriations were noted primarily on the trunk in a cranioventral distribution, especially prominent in the axillae and extending caudally to the inguinal area. A CBC, serum biochemistry profile, coagulation profile, serum bile acids, skin scrapings, Wood’s lamp examination, dermatophyte culture, abdominal ultrasound, and thoracic radiographs were performed. The CBC revealed a mild neutropenia (2.1 × 10³/μL; reference range, 2.5 to 12.5 × 10³/μL) and lymphopenia (0.8 × 10³/μL; reference range, 1.5 to 7.5 × 10³/μL). Serum biochemistry profile revealed elevations in ALT (353 U/L; reference range, 20 to 107 U/L), AST (132 U/L; reference range, 1 to 37 U/L), and total bilirubin (0.6 mg/dL; reference range, 0.1 to 0.5 mg/dL). Prothrombin and activated partial thromboplastin times, fibrin(ogen) degradation products, and platelet count were all within reference ranges. Serum bile acids were suggestive of mild liver dysfunction, with elevations in concentrations after fasting (3.8 μM/L; reference range, ≤2 μM/L) and postprandial (15.4 μM/L; reference range, ≤10 μM/L). Skin scrapings, Wood’s lamp examination, and dermatophyte culture were all negative. Ultrasonography of the liver showed a diffusely coarse echotexture with a reticular pattern, similar to what is observed in dogs with superficial necrolytic dermatitis (SND) with hepatopathy (i.e., hepatocutaneous syndrome).¹² No hepatic vascular abnormalities or discrete nodules were visualized. Although the ultrasonographic appearance of the liver was suggestive of the hepatopathy associated with SND, other differential diagnoses included inflammatory disease such as hepatitis, and infiltrative neoplasia such as lymphosarcoma. Thoracic radiographs were unremarkable.

The cat was anesthetized, and ultrasound-guided percutaneous biopsy of the liver and punch biopsies of the skin were performed. In addition, a bone-marrow aspirate was performed because of the mild to moderate neutropenia. Liver histopathology revealed multiple foci of nodular regeneration and periportal zones of swollen hepatocytes with marked vacuolation of their cytoplasm and peripheralized nuclei. Aerobic and anaerobic cultures of liver tissue grew Staphylococcus intermedius sensitive to amoxicillin/clavulanic acid. Histopathology of skin biopsies taken from the ventral thorax and axillary regions revealed epidermal hyperplasia with focal compact hyperkeratosis, acanthosis, spongiosis, and occasional mast cells within the basal portion of the epidermis. Additional findings included superficial dermal fibrosis with mast cells and eosinophils diffusely scattered throughout the dermis. Bone-marrow aspirate cytopathology revealed normal myeloid and erythroid series and mild hemosiderosis.

Although the histopathology of the liver was compatible with the abnormalities found in dogs with SND and hepatopathy, the histopathological findings of the skin were compatible with an allergic dermatitis. Therefore, an elimination diet trial was started using a commercial diet.^a Amoxicillin/clavulanic acid (20 mg/kg body weight, PO q 12 hours for 3 weeks) was prescribed because of the bacterial growth from the liver tissue and presence of liver dysfunction. The cat’s alopecia, crusting, and erythema did not improve after 3 weeks on this therapy. The owners reported that the cat’s pruritus had intensified, and the cat was vomiting intermittently. Metoclopramide (0.4 mg/kg body weight, PO q 8 hours) was recommended for symptomatic relief of the vomiting. Chlorpheniramine (0.3 mg/kg body weight, PO q 12 hours) was recommended for symptomatic relief of pruritus. After several days without noting any clinical improvement, the owner declined additional diagnostics and therapeutics and elected euthanasia. Prior to euthanasia, plasma for determination of amino acid concentrations was collected with the owner’s consent. Amino acid assays were performed by use of an automated analyzer,^b and values were compared to those previously reported in 24 healthy cats.³

At the time of necropsy, the cat weighed 6.0 kg. Gross findings included ulceration and crusting of the oral mucocutaneous junction and ulceration of the skin at the base of each ear, medial canthus of the eye, interdigital spaces, ventral abdomen, and inguinal regions [Figure 1]. The liver was small (142.3 g; 2.4% body weight) and mottled dark red-black and yellow-brown on the surface. There were fine, white, slightly depressed lines dividing the parenchyma into an irregular reticular pattern [Figure 2]. The pancreas and intestinal tract appeared grossly normal. Histopathology of the liver revealed nodular regeneration with bands of vacuolated cells and multifocal bile duct hyperplasia [Figure 3]. Histopathology of skin from the inguinal region revealed multifocal parakeratosis, midepidermal spongiosis, and basal cell hyperplasia with superficial coccoid bacteria and yeast [Figure 4]. These findings were compatible with SND as it occurs in dogs. Skin biopsy specimens from the medial canthus, interdigital spaces, and oral mucocutaneous junctions showed multifocal, moderate to severe ulcerative crusting dermatitis with periappendageal lymphoid infiltration and multifocal squamous cell carcinoma in situ (i.e., Bowen’s disease); immunohistochemistry for papillomavirus was negative. All skin biopsy specimens revealed mild to moderate, chronic-active dermatitis with intraepithelial eosinophils and mast cells.

Plasma amino acid profile results showed decreased concentrations in 20 of 21 amino acids measured [see Table], compatible with findings in dogs with SND and hepatopathy. Decreases in these amino acid concentrations ranged from 20% to 78% of mean values for normal cats, with threonine, methionine, proline, glycine, arginine, and lysine most severely depleted. Tryptophan was within reference ranges.

Discussion

Necrolytic migratory erythema (NME) is a skin eruption first reported in humans with alpha cell pancreatic tumors and elevated plasma glucagon concentrations (i.e., glucagonoma syndrome). A small number of humans with NME do not have a glucagonsecreting tumor.⁴ In the veterinary literature, the terms superficial necrolytic dermatitis (SND) and metabolic epidermal necrosis (MEN) are also used to describe the skin eruption.^5–8 In dogs, the vast majority do not have a demonstrable glucagonoma and instead have a hepatopathy with characteristic histopathological and ultrasonographic features (i.e., hepatocutaneous syndrome). The etiology of the hepatic and cutaneous lesions and their possible relationships are not known.⁵⁹¹⁰ Therefore, the authors prefer the use of the term “superficial necrolytic dermatitis with hepatopathy” to the commonly used term “hepatocutaneous syndrome,” since the former is merely descriptive without implying causality or a relationship between the skin and liver lesions.

The histopathological appearance of the skin of this cat is similar to SND in dogs⁵ and a case of SND associated with a pancreatic tumor in a cat.¹¹ Further, the histopathological appearance of the liver in this cat is similar to what is found in the majority of dogs with SND. Superficial necrolytic dermatitis appears to be very rare in the cat, with only three cases reported thus far in the English language literature. These reports describe a cat with pancreatic carcinoma,¹¹ a cat with thymic amyloidosis,¹² and a cat with hepatopathy and cutaneous lesions.¹³

In dogs, the main cutaneous signs of SND are an erosive, crusting, and scaling dermatopathy distributed symmetrically over the face, distal paws, and inguinal area, occurring in areas of constant friction.¹⁰ In one cat with hepatopathy, lesions were present on the inguinal and ventral abdominal skin as well as interdigital areas.¹³ The appearance of this cat’s lesions was similar to that observed in dogs and the one reported cat. However, although there were lesions present on the face and interdigital spaces, only the lesions from the inguinal area appeared grossly and histopathologically consistent with SND. This is similar to the previously reported case of feline SND secondary to pancreatic carcinoma, in which there were no lesions on the footpads or face; yet this case is different from the one cat with hepatopathy, whose interdigital lesions, but not truncal lesions, were consistent with SND. It is possible that the distribution of lesions in cats is unlike dogs because of differences in areas of friction. It has been previously suggested that the lesions seen in cats may be associated with self-grooming.¹¹ In addition, the presence of characteristic histopathological findings of SND in only one of the skin specimens supports the practice of obtaining multiple biopsy specimens, because the characteristic features of SND may not be uniformly present in affected skin.¹⁴

The cat the authors describe in this report was pruritic. However, pruritus is not a consistent feature of SND in dogs, and pruritus was not described in two cases of feline SND reported previously.¹¹¹³ The discrepancy in clinical signs reported may be related to the probable allergic dermatitis in this cat. The etiology of the allergic disease in this cat was not determined, and a longer hypoallergenic diet trial of at least 10 weeks would have been necessary to rule out food allergy.¹⁵ In addition, the inflammatory changes suggestive of allergic dermatitis may have occurred in this cat secondary to bacterial or yeast infections. Pruritus is a common feature of cholestatic liver disease in humans.¹⁶ This phenomenon has not been reported in the veterinary literature; however, it is possible that the hepatopathy could have contributed to pruritus in this cat.

Multifocal squamous cell carcinoma (SCC) in situ is an uncommon neoplasia in older cats. Squamous cell carcinoma in situ may be induced by chronic sun exposure of lightly pigmented skin, similar to actinic keratosis of humans. Multicentric SCC in situ that resembles human Bowen’s disease has been previously reported in cats¹⁷ and may be associated with papillomavirus infection. This cat did have SCC in situ on the medial canthus, and this lesion could have been sun-induced. However, although tests for papillomavirus were negative, SCC in situ lesions in this indoor cat were present on areas with little sun exposure and were more typical of Bowen’s disease.

The clinicopathological features of SND with hepatopathy were previously reported in one cat.¹³ In the prior report, serum bile acid, ALT, AST, and alkaline phosphatase (ALP) concentrations were elevated. In the cat of this study, serum fasting and postprandial bile acid concentrations, ALT, and AST were elevated, although ALP (as well as gamma glutamyl transferase [GGT]) was normal. Elevated ALP is consistently found in dogs with SND and hepatopathy.⁵ This may be due to differences in isoenzyme release versus isoenzyme induction between dogs and cats, or less likely due to differences in disease severity, etiology, or other features of the hepatopathy. The cause of the consistent mild to moderate neutropenia was not found in this cat. Bone-marrow cytopathology did not reveal any abnormalities of myelopoiesis, nor did the cat appear clinically affected by the neutropenia (e.g., fever, sepsis, etc.).

The ultrasonographic appearance of the liver in dogs with SND has been previously described as a hyperechoic, reticular pattern surrounding hypoechoic areas in a “Swiss cheese-like” appearance.¹² Such a reticular pattern was observed in this cat, and this ultrasonographic appearance corresponds to the gross and microscopic features of parenchymal collapse surrounding areas of nodular hyperplasia that are characteristic of hepatocutaneous syndrome. Possibly, the initial abdominal ultrasound performed on this cat prior to referral to VHUP that identified multiple nodules, either visualized areas of nodular hyperplasia, or possibly misinterpreted as nodular lesions the multiple areas in the liver demarcated by the hyperechoic network representative of parenchymal collapse.

The dermatohistopathological findings of parakeratosis, midepidermal spongiosis, and basal cell hyperplasia, provide the characteristic “red, white, and blue” sign of SND observed in this cat. These findings are compatible with those of other species, including the human, dog, and black rhinoceros, in which SND has been described.^24101118

Although a documented cause of liver disease has been associated with SND, such as alcoholic liver disease in humans⁴ and toxic hepatopathy in dogs,¹⁹²⁰ most cases of canine SND that are associated with hepatopathy are of unknown etiology. Numerous theories have been proposed regarding the pathogenesis of SND with hepatopathy. Although a glucagon-secreting tumor is not present in these patients, elevations in glucagon or different immunoreactive fractions of glucagon may be present. It is thought that because of the liver’s role in degrading glucagon, liver dysfunction results in abnormal ratios of circulating immunoreactive glucagons.⁹ Elevated glucagon concentrations may result in increased gluconeogenesis, amino acid deficiency, and depletion of epidermal proteins, leading to the skin lesions of SND.²¹ Therefore, the pathogenesis of SND in patients with glucagonoma and those that have SND with hepatopathy may be similar. However, glucagon levels are inconsistently elevated in humans that have NME with hepatopathy and are only mildly elevated in dogs that have SND with hepatopathy, and clinical signs do not appear to correlate with glucagon concentrations.⁴⁹ Differences between systemic glucagon concentration and portal glucagon concentration, which would more directly affect hepatic gluconeogenesis, likely exist.²² In addition, an undetectable hormone with glucagon-like activity may be responsible for the development of SND with hepatopathy.⁹ Other theories including zinc deficiency, fatty acid deficiencies, or a combination of vitamin, mineral, fatty acid, and amino acid deficiencies as the cause of the lesions of SND with hepatopathy have also been proposed.⁴

Plasma amino acid deficiency was documented in the cat of this report. The depression of plasma amino acid concentrations in this cat appears to be more severe and more extensive than what has been previously described in dogs.¹⁰ Such severe hypoaminoacidemia may suggest the influence of hyperglucagonemia. Unfortunately, plasma glucagon concentration was not determined in this cat.

Superficial necrolytic dermatitis with hepatopathy has a poor prognosis, with a mean survival of 1.6 months after diagnosis in dogs;⁹ this is similar to the cat reported here. However, additional therapies such as hyperalimentation, zinc and fatty acid supplementation, and intravenous amino acid therapy were declined by the owner of this cat prior to euthanasia, and the response to such therapies of cats that have SND with hepatopathy remains unknown.

Acknowledgments

The authors acknowledge the support of Dr. Stan Marks of the University of California-Davis for performing the amino acid analysis for this case.