Clinical and histologic features of acute-onset erythroderma in dogs with gastrointestinal disease: 18 cases (2005–2015)

Christine L. Cain Department of Clinical Sciences and Advanced Medicine, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA 19104.

Search for other papers by Christine L. Cain in
Current site
Google Scholar
PubMed
Close
 DVM
,
Charles W. Bradley II Department of Pathobiology, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA 19104.

Search for other papers by Charles W. Bradley II in
Current site
Google Scholar
PubMed
Close
 VMD
, and
Elizabeth A. Mauldin Department of Pathobiology, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA 19104.

Search for other papers by Elizabeth A. Mauldin in
Current site
Google Scholar
PubMed
Close
 DVM

Click on author name to view affiliation information

Abstract

OBJECTIVE To describe the clinical and histologic features of acute erythroderma in dogs with gastrointestinal disease.

DESIGN Retrospective case series.

ANIMALS 18 dogs with erythroderma and gastrointestinal disease.

PROCEDURES Medical records and biopsy specimens were reviewed. Information collected from medical records included signalment, clinical signs, physical examination and diagnostic test results, treatment, and outcome. The Naranjo algorithm was used to estimate the probability of an adverse drug reaction for each dog.

RESULTS All dogs had an acute onset of erythematous macules or generalized erythroderma. Histologic features of skin biopsy specimens had 3 patterns representing a progressive spectrum of inflammation. Most dogs had vomiting (n = 17) and hematochezia (10). Signs of gastrointestinal disease became evident before, after, or concurrent with the onset of skin lesions in 10, 3, and 5 dogs, respectively. Inflammatory bowel disease, pancreatitis, and adverse food reaction were diagnosed in 5, 3, and 3 dogs, respectively. The cause of the gastrointestinal signs was not identified for 8 dogs. Eight dogs had a Naranjo score consistent with a possible adverse drug reaction. Treatment of skin lesions included drug withdrawal (n = 15), antihistamines (16), and corticosteroids (14). Signs of gastrointestinal disease and skin lesions resolved at a mean of 4.6 days and 20.8 days, respectively, after onset.

CONCLUSIONS AND CLINICAL RELEVANCE Results indicated acute erythroderma may be associated with > 1 gastrointestinal disease or an adverse drug reaction in some dogs. Recognition of the clinical and histologic features of this syndrome is essential for accurate diagnosis.

Abstract

OBJECTIVE To describe the clinical and histologic features of acute erythroderma in dogs with gastrointestinal disease.

DESIGN Retrospective case series.

ANIMALS 18 dogs with erythroderma and gastrointestinal disease.

PROCEDURES Medical records and biopsy specimens were reviewed. Information collected from medical records included signalment, clinical signs, physical examination and diagnostic test results, treatment, and outcome. The Naranjo algorithm was used to estimate the probability of an adverse drug reaction for each dog.

RESULTS All dogs had an acute onset of erythematous macules or generalized erythroderma. Histologic features of skin biopsy specimens had 3 patterns representing a progressive spectrum of inflammation. Most dogs had vomiting (n = 17) and hematochezia (10). Signs of gastrointestinal disease became evident before, after, or concurrent with the onset of skin lesions in 10, 3, and 5 dogs, respectively. Inflammatory bowel disease, pancreatitis, and adverse food reaction were diagnosed in 5, 3, and 3 dogs, respectively. The cause of the gastrointestinal signs was not identified for 8 dogs. Eight dogs had a Naranjo score consistent with a possible adverse drug reaction. Treatment of skin lesions included drug withdrawal (n = 15), antihistamines (16), and corticosteroids (14). Signs of gastrointestinal disease and skin lesions resolved at a mean of 4.6 days and 20.8 days, respectively, after onset.

CONCLUSIONS AND CLINICAL RELEVANCE Results indicated acute erythroderma may be associated with > 1 gastrointestinal disease or an adverse drug reaction in some dogs. Recognition of the clinical and histologic features of this syndrome is essential for accurate diagnosis.

Aunique syndrome of severe eosinophilic dermatitis with edema has been described in dogs with and without concurrent signs of gastrointestinal disease, particularly vomiting and hematochezia.1–3 Skin lesions in those dogs include deeply erythematous macules, papules, and plaques, as well as wheals and facial edema.1–3 This cutaneous syndrome in dogs has been likened to Wells syndrome (eosinophilic cellulitis) in humans. Eosinophilic cellulitis is characterized by erythema and edema of the skin, often with prodromal to concurrent pruritus or pain, followed by the development of erythematous plaques that may wax and wane over months to years.4 It should be noted that the cellulitis in the name of this disorder refers to its clinical features of erythema and edema of the skin, which may mimic bacterial cellulitis. The described histologic features are similar between affected dogs and human patients and include dermal edema, moderate to severe eosinophilic dermatitis, and collagen flame figures.1–4

To our knowledge, the veterinary literature contains only 1 report2 describing moderate to severe eosinophilic dermatitis with edema in association with enteric disease in dogs. That was a case-series report2 that included dogs with and without signs of gastrointestinal disease. Given the acute onset of clinical signs for dogs affected by this syndrome as well as the types of skin lesions that develop (eg, markedly erythematous macules and plaques with or without central clearing, likely caused by dermal edema), clinicians unfamiliar with this syndrome may mistakenly diagnose cutaneous or urticarial vasculitis, erythema multiforme (macules with central clearing may mimic early nonulcerated target lesions characteristic of erythema multiforme), or a coagulopathy resulting in cutaneous ecchymoses solely on the basis of clinical signs. Since the case-series report2 describing moderate to severe eosinophilic dermatitis with edema in dogs with enteric disease was published in 2006, we have observed erythematous macules and patches, plaques, and wheals in dogs with signs of gastrointestinal disease and mild eosinophilic dermatitis. Therefore, the objective of the study reported here was to better elucidate the spectrum of clinical and histologic features, as well as treatment response and prognosis, for dogs with eosinophilic dermatitis and concurrent signs of gastrointestinal disease.

Materials and Methods

Case selection criteria

The electronic database of the Matthew J. Ryan Veterinary Hospital of the University of Pennsylvania was searched to identify dogs that had a skin biopsy performed, which yielded a morphological diagnosis of eosinophilic dermatitis between July 31, 2005, and October 31, 2015. The medical records of identified dogs were then reviewed to determine eligibility for study inclusion. Only dogs with a history of acute onset of skin lesions that persisted for > 48 hours and signs of gastrointestinal disease prior to, concurrent with, or following the onset of the skin lesions were included in the study.

Medical records review

For each dog that was eligible for study inclusion, information extracted from the medical record included signalment; month of onset and description of skin lesions; description of clinical signs of gastrointestinal disease and onset of those signs relative to the onset of the skin lesions; results of clinicopathologic tests, abdominal and thoracic diagnostic imaging, and histologic evaluation of biopsy specimens obtained from the skin and gastrointestinal tract (if performed); duration of clinical signs prior to collection of skin biopsy specimens; gastrointestinal disease diagnosis; drugs administered prior to and during hospitalization; other treatments; time required for resolution of signs of gastrointestinal disease and skin lesions; and follow-up information regarding recurrence of gastrointestinal disease or skin lesions (when available).

Biopsy specimen review

For all study dogs, archived slides containing skin-punch biopsy specimens (6 to 8 mm in diameter) that were stained with H&E stain were independently reviewed by 2 board-certified veterinary pathologists (EAM and CWB). Additionally, archived skin biopsy specimens were stained with a Luna stain for specific assessment of specimens for the presence of eosinophils and extracellular eosinophil granules in the inflammatory infiltrate.5,6 For each specimen, the inflammatory infiltrate was graded on a scale from 0 to 3, where 0 = no inflammation, 1 = mild inflammation (inflammatory cells confined to the perivascular region and predominately within the superficial dermis), 2 = moderate inflammation (a moderate number of inflammatory cells distributed perivascularly to diffusely throughout the dermis), and 3 = severe inflammation (large number of inflammatory cells distributed diffusely throughout the dermis). Collagen flame figures (foci of hypereosinophilic collagen fibrils characterized by deposition of eosinophilic amorphous material and surrounding eosinophils) were also identified.

Drug score

All drugs and biologics administered to each dog were classified into the following categories: antimicrobials (including antifungals), antiemetics, H2 receptor antagonists, H1 receptor antagonists, proton pump inhibitors, other gastroprotectants, corticosteroids, NSAIDs, anthelmintics, and vaccines. The Naranjo algorithm7 (Appendix) was used to estimate the probability (drug score) of an adverse drug reaction for each dog. An adverse drug reaction was considered doubtful for dogs with a drug score < 1, possible for dogs with a drug score between 1 and 4, probable for dogs with a drug score between 5 and 8, and definitive for dogs with a drug score > 9.

Data analysis

Descriptive data were generated and reported.

Results

Dogs

Eighteen dogs (13 spayed females, 4 castrated males, and 1 sexually intact male) met the inclusion criteria and were enrolled in the study. Body weight ranged from 4.4 to 62 kg (9.7 to 136.4 lb), and age ranged from 11 months to 12 years. Dogs were classified as English Bulldog (n = 2), Labrador Retriever (2), Miniature Schnauzer (2), mixed-breed dog (2), Pug (2), and American Pit Bull Terrier, Basset Hound, Bichon Frise, Newfoundland, Papillon, Rough Collie, Shih Tzu, and Soft Coated Wheaten Terrier (1 each). The onset of clinical signs occurred during all months of the year, with the most common being August (n = 4 dogs) and April (3).

Clinical and histologic features of skin lesions

All 18 dogs had generalized erythroderma or macular erythema that was most pronounced over the ventral aspect of the abdomen or inguinal region (Figure 1). Four dogs had erythematous macules and patches with central areas of clearing (ie, targetoid macules). The central clearing in those targetoid macules was characterized by a zone of pallor (likely caused by dermal edema) and was not palpably raised.

Figure 1—
Figure 1—

Representative photographs of skin lesions that developed acutely on the ventral aspect of the abdomen and inguinal region of dogs with concurrent signs of gastrointestinal disease. A—An 8-year-old castrated male Pug with severe erythroderma that was most pronounced over the ventral aspect of the abdomen and inguinal region; hind limb edema is also present. B—A 5-year-old spayed female Soft Coated Wheaten Terrier with intense to violaceous erythroderma that was most pronounced over the ventral aspect of the thorax and abdomen. C—A 4-year-old castrated male English Bulldog with erythematous macules, plaques, and wheals with central clearing (ie, targetoid macules and plaques) in the inguinal region. D—An 8-year-old spayed female mixed-breed dog with intense to violaceous erythematous patches over the ventral aspect of the thorax and cranial portion of the abdomen; the hair has been clipped to reveal the skin lesions.

Citation: Journal of the American Veterinary Medical Association 251, 12; 10.2460/javma.251.12.1439

The histologic features of the skin biopsy specimens obtained from areas with lesions had 3 basic patterns (1, 2, and 3) that represented an increasing spectrum of inflammation. For specimens with > 1 histologic pattern, the most severe pattern was reported. Differentiation of neutrophils from degranulated eosinophils was difficult on H&E-stained sections, but extracellular eosinophil granules were easily identified on Luna-stained sections for all dogs (Figure 2).

Figure 2—
Figure 2—

Photomicrograph of a skin biopsy specimen obtained from a dog with acute-onset erythroderma and concurrent signs of gastrointestinal disease. The erythroderma was characterized by vascular ectasia within the superficial dermis and mild to moderate edema (histologic pattern 1). Eosinophil granules are present in macrophages (arrows) and the cytoplasm of granulocytes (arrowheads). Luna stain; bar = 50 μm.

Citation: Journal of the American Veterinary Medical Association 251, 12; 10.2460/javma.251.12.1439

Histologic pattern 1 was characterized by vascular ectasia within the superficial dermis and mild to moderate edema (Figure 3). Blood vessels were dilated and lined with hypertrophied endothelial cells. Inflammation was minimal to mild; was perivascular to interstitial; consisted of eosinophils, neutrophils, and macrophages; and was localized to the superficial dermis with some inflammatory cells percolating through the mid-dermis. Epidermal hyperplasia, when present, was very mild. Eight dogs had skin lesions with histologic pattern 1. Gross skin lesions for those dogs included erythematous macules or generalized erythroderma only (n = 4), wheals (2), facial edema (2), and limb edema (1). The skin lesions in those 8 dogs had been present for a median of 6.5 days (range, 1 to 10 days) prior to collection of the biopsy specimens.

Figure 3—
Figure 3—

Representative photomicrographs of skin biopsy specimens obtained from dogs with acute-onset erythroderma and signs of gastrointestinal disease that depict each of the 3 histologic patterns commonly observed. A—Skin specimen with histologic pattern 1 characterized by dermal edema with paucicellular infiltrate and vascular ectasia. H&E stain; bar = 200 μm. B—Higher magnification of the specimen in panel A that depicts a mild infiltrate of granulocytes. H&E stain; bar = 100 μm. C—Skin specimen with histologic pattern 2 characterized by mild acanthosis with dermal edema and moderate interstitial infiltrate of eosinophils and a focal collagen flame figure (arrow). H&E stain; bar = 200 μm. D—Higher magnification of the specimen in panel C that depicts eosinophilic infiltrate and a collagen flame figure (arrow). H&E stain; bar = 100 μm. E—Skin specimen with histologic pattern 3 characterized by marked superficial dermal edema with epidermal vesicles (arrow) and a dense dermal inflammatory infiltrate. H&E stain; bar = 400 μm. F—Higher magnification of the specimen in panel E that depicts the marked inflammatory cell infiltrate with a large collagen flame figure (arrows). H&E stain; bar = 100 μm.

Citation: Journal of the American Veterinary Medical Association 251, 12; 10.2460/javma.251.12.1439

Histologic pattern 2 was characterized by more extensive inflammation and edema and a greater abundance of eosinophils than histologic pattern 1. The inflammation extended into the deep dermis, and the epidermis had moderate hyperplasia and a spongiotic appearance (Figure 3). Six dogs had skin lesions with histologic pattern 2. The biopsy specimens for 4 of those dogs contained collagen flame figures. Gross skin lesions for those dogs included erythematous wheals or plaques (n = 5), facial edema (2), and erythematous papules (2); none of the dogs had erythematous macules or generalized erythroderma alone. The skin lesions in those 6 dogs had been present for a median of 3 days (range, 1 to 6 days) prior to collection of the biopsy specimens.

Histologic pattern 3 was characterized by severe and diffuse dermal inflammation with spongiotic vesicles, eosinophilic or neutrophilic pustules, and ulcers (Figure 3). Other histologic lesions observed in this pattern included collagen flame figures and protein-rich dermal edema caused by substantial separation of collagen fibers by an eosinophilic matrix. Evidence of vascular damage was apparent and ranged from perivascular fibrin deposition with erythrocyte diapedesis to overt fibrinoid vasculitis. Four dogs had skin lesions with histologic pattern 3. The biopsy specimens for some of those dogs also contained focal ulcers. Gross skin lesions were extensive for those dogs, and none had erythematous macules or generalized erythroderma alone. One dog (an 8-year-old castrated male Pug; Figure 1) with histologic pattern 3 lesions had evidence of fibrinoid vasculitis. That dog had acute onset of vomiting, inappetence, severe generalized erythroderma, and hind limb edema. It had not been exposed to any drugs prior to the manifestation of the skin lesions and clinical signs of gastrointestinal disease, and the underlying cause of the gastrointestinal signs was not determined. The skin lesions in the remaining 3 dogs with histologic pattern 3 had been present for a median of 2 days (range, 1 to 3 days) prior to collection of the biopsy specimens.

Clinicopathologic findings

A CBC was performed for 15 of the 18 dogs, and hematologic abnormalities were detected for 14. The most common hematologic abnormalities were mild anemia (Hct range for affected dogs, 34.8% to 40.2%; reference range, 40.3% to 60.3%; n = 6 dogs), mild thrombocytopenia (platelet count range for affected dogs, 80,000 to 162,000 platelets/μL; reference range, 177,000 to 398,000 platelets/μL; 4), and lymphopenia (lymphocyte count range for affected dogs, 120 to 780 lymphocytes/μL; reference range, 900 to 5,500 lymphocytes/μL; 4) Two dogs had a left shift (band neutrophil count, 320 and 810 band neutrophils/μL; reference range, 0 to 200 band neutrophils/μL) in the absence of a mature neutrophilia. Two dogs had eosinophilia (eosinophil count, 1,650 and 1,750 eosinophils/μL; reference range, 100 to 1,490 eosinophils/μL), whereas 2 other dogs had eosinopenia (eosinophil count, 40 and 50 eosinophils/μL).

A serum biochemical analysis was performed for 16 of the 18 dogs, and biochemical abnormalities were detected for 14. The most common abnormality was mild to moderate hypoalbuminemia (albumin concentration range for affected dogs, 1.6 to 2.3 g/dL; reference range, 2.5 to 3.7 g/dL; n = 8 dogs) prior to initiation of IV fluid therapy. Of the 8 dogs with hypoalbuminemia, 4 had skin lesions with histologic pattern 1 and 4 had skin lesions with histologic pattern 3. Six dogs had abnormally increased activities of liver enzymes (alanine aminotransferase, alkaline phosphatase, γ-glutamyltransferase, and aspartate aminotransferase), 4 dogs had hypocholesterolemia, and 1 dog had abnormally increased amylase and lipase activities.

Diagnostic testing in addition to a CBC and serum biochemical analysis was performed for 14 of the 18 dogs. Fecal flotation for parasite detection was performed for 7 dogs, and results were negative for all dogs. Bacterial culture of a fecal sample was performed for 2 of those 7 dogs, and neither Salmonella spp nor Campylobacter spp was isolated from either dog. Resting cortisol analysis or an ACTH stimulation test was performed for 7 dogs, and results were within reference limits for all tests. A pancreas-specific lipase immunoreactivity test was performed for 7 dogs, and 3 had abnormally increased immunoreactivity. Urinalysis and bacterial culture of a urine sample were performed for 6 dogs. One of those dogs had hematuria and bacteriuria detected on urinalysis and Escherichia coli isolated on culture. Coagulation profiles (prothrombin time, partial thromboplastin time, and D-dimer concentrations) were performed for 5 dogs. One dog had D-dimer concentrations that were moderately increased from the reference range, and 1 dog had a mildly prolonged partial thromboplastin time. Both of those dogs had skin lesions with histologic pattern 3. An ELISA for detection of antibodies against Borrelia burgdorferi, Ehrlichia spp, and Anaplasma spp was performed for 3 dogs, and negative results were recorded for all 3 dogs. Pre- and postprandial bile acid concentrations were measured for 1 dog, and results were within reference limits. Arthrocentesis and lymph node aspiration for cytologic evaluation were performed for a 6-year-old spayed female Rough Collie with pyrexia, difficulty rising, vomiting, regurgitation, and inappetence, and results were consistent with immune-mediated polyarthritis and reactive lymphadenopathy.

Abdominal and thoracic imaging

Abdominal ultrasonography was performed for 13 of the 18 dogs. No ultrasonographic abnormalities were detected for 2 of those dogs. Ultrasonographic findings consistent with gastroenteritis and colitis, such as thickening of the gastrointestinal wall, fluid-filled intestinal loops, hypomotility, and abdominal lymphadenopathy, were recorded for 10 dogs. Three dogs had abnormal echogenicity of the pancreas or peripancreatic fat, which was consistent with pancreatitis. Two of those 3 dogs had a pancreas-specific lipase immunoreactivity test performed and had abnormally increased immunoreactivity.

Thoracic radiographs were obtained for 9 of the 18 dogs. Eight of those dogs had either no abnormalities or only mild nonspecific changes. One dog had a soft tissue opacity present in the left cranial lung lobe that was consistent with a primary lung tumor.

Gastrointestinal disease

Of the 18 dogs, 10 and 3 developed clinical signs of gastrointestinal disease before and after the onset of ski n lesions, respectively, and 5 developed signs of gastrointestinal disease the same day that the skin lesions became evident. For the 10 dogs that developed signs of gastrointestinal disease before the skin lesions became evident, the onset of those signs ranged from 1 to 14 days (mean, 5.4 days) before manifestation of the skin lesions. For the 3 dogs that developed skin lesions before signs of gastrointestinal disease, the skin lesions manifested 1 and 2 days prior to signs of gastrointestinal disease for 2 and 1 dogs, respectively.

Clinical signs associated with gastrointestinal disease varied among dogs. The most frequently observed clinical sign of gastrointestinal disease was vomiting (n = 17 dogs) followed by hematochezia (10).

Gastrointestinal biopsy specimens were obtained by endoscopy or laparoscopy for 5 of the 18 dogs, and histologic results were consistent with inflammatory bowel disease for all 5 dogs. An 8-year-old spayed female Miniature Schnauzer had moderate to severe lymphoplasmacytic gastritis and duodenitis. A 9-year-old castrated male Miniature Schnauzer had mild to moderate lymphoplasmacytic and eosinophilic gastritis, duodenitis, and colitis. A 12-year-old spayed female Basset Hound had moderate to severe eosinophilic gastritis and duodenitis with lymphangiectasia (this was the dog that had radiographic evidence of a primary lung tumor). A 2-year-old spayed female American Pit Bull Terrier had moderate lymphoplasmacytic and eosinophilic enteritis. An 8-year-old spayed female mixed-breed dog had moderate lymphoplasmacytic gastritis and enteritis with lymphangiectasia.

Three dogs were determined to have pancreatitis, including the 8-year-old spayed female Miniature Schnauzer with moderate to severe lymphoplasmacytic gastritis and duodenitis. Presumed esophagitis was diagnosed in the 6-year-old spayed female Rough Collie with immune-mediated polyarthritis and lymphadenopathy; that dog also had a lower urinary tract infection.

An adverse food reaction was definitively diagnosed for 3 dogs. Two of those dogs, an 8-year-old spayed female mixed-breed dog that underwent a diet change 2 days prior to the onset of vomiting and diarrhea and the 2-year-old spayed female American Pit Bull Terrier that had moderate lymphoplasmacytic and eosinophilic enteritis, had a prior history of adverse food reaction. Adverse food reaction was diagnosed in the remaining dog, a 1-year-old spayed female Pug with a history of recurrent otitis externa and pruritus, following a diet trial and dietary rechallenge. Adverse food reaction was suspected in another dog, a 4-year-old spayed female English Bulldog, and a diet trial with a hydrolyzed protein diet was prescribed.

The definitive cause of the signs of gastrointestinal disease remained unknown for 8 of the 18 dogs. For one of those dogs, food from a new bag was fed beginning 2 days prior to the onset of vomiting and skin lesions. The owner of another dog reported that it had sustained multiple bee or wasp stings prior to the onset of vomiting and skin lesions.

Drug score

Sixteen of 18 dogs were initially treated with multiple medications including gastroprotectants (antiemetics, H2 receptor antagonists or proton pump inhibitors, and sucralfate) and antimicrobials (metronidazole, β-lactams, doxycycline, or fluoroquinolones). Of the 2 remaining dogs, 1 was initially treated with metronidazole alone beginning the day before skin lesions manifested and the other was initially treated with cephalexin alone beginning the day after skin lesions manifested.

Ten of the 18 dogs had a drug score of 0, which indicated that an adverse drug reaction was doubtful. The remaining 8 dogs had a drug score consistent with a possible adverse drug reaction (drug score, 1 to 4). The drug score was 3 or 4 for 4 of those dogs. One was a 4-year-old spayed female Newfoundland that was vaccinated against B burgdorferi (Lyme disease), Bordetella bronchiseptica, Leptospirosis spp, and rabies 13 days prior to the manifestation of erythematous macules and patches with pinnal edema, vomiting, and hematochezia. A 1-year-old spayed female Pug developed erythematous macules and exudative plaques 7 days after the onset of vomiting, hematochezia, and inappetence and had been treated with metronidazole, ampicillin, dolasetron, ranitidine, and sucralfate prior to the manifestation of the skin lesions. A 5-year-old spayed female Soft Coated Wheaten Terrier developed erythematous macules and plaques with hind limb edema 11 days after the onset of vomiting, hematochezia, and inappetence and had been administered metronidazole, procaine penicillin G, enrofloxacin, aminopentamide, maropitant, ondansetron, metoclopramide, and famotidine prior to the manifestation of the skin lesions. That dog also had received vaccines against Lyme disease, B bronchiseptica, and rabies and a multivalent vaccine against canine distemper virus, adenovirus, parvovirus, and parainfluenza virus and Leptospirosis spp 18 days before the skin lesions became evident. An 8-year-old spayed female mixed-breed dog developed generalized erythematous macules and patches 14 days after the onset of vomiting and had been treated with metronidazole, maropitant, and famotidine before the onset of skin lesions. Maropitant and famotidine were discontinued when the skin lesions became evident, but metronidazole was continued for an additional 7 days.

The drug score was 1 or 2 for the remaining 4 dogs with possible adverse drug reactions. One of those dogs was a 7-year-old spayed female Papillon that was suspected to have a cutaneous reaction to one of the medications (metronidazole, maropitant, famotidine, omeprazole, or a combination heartworm preventive containing ivermectin and pyrantel pamoate) it had received. That dog had a recurrence of signs of gastrointestinal disease 1 month after the initial episode and was treated with metronidazole, maropitant, and famotidine with no recurrence of skin lesions. It was not exposed again to omeprazole or the combination heartworm preventive containing ivermectin and pyrantel pamoate, and was not receiving any medications, such as antihistamines or corticosteroids, that may have affected the development of skin lesions or eosinophilic dermatitis at the time of drug reintroduction.

Treatment

All drugs administered prior to the manifestation of skin lesions were discontinued for 15 of the 18 dogs. One dog continued to receive enrofloxacin, which was initiated 2 days prior to the onset of skin lesions. One dog continued to receive famotidine, which was initiated 3 days prior to the onset of skin lesions. For the remaining dog, treatment with metronidazole was initiated 10 days prior to and continued for 7 days after the skin lesions became evident. The most commonly prescribed systemic medications for treatment of the skin lesions and gastrointestinal disease were antihistamines (cetirizine, diphenhydramine, hydroxyzine, or chlor-pheniramine; n = 16 dogs), corticosteroids (prednisone; 14), gastroprotectants (famotidine, omeprazole, ondansetron, metoclopramide, or sucralfate; 13), antimicrobials (cephalexin, cefpodoxime, clindamycin, doxycycline, enrofloxacin, fluconazole, or tylosin; 8), and anthelmintics (fenbendazole or praziquantel-pyrantel pamoate-febantel; 2).

Eight of 18 dogs received corticosteroids prior to the skin biopsy. Five of those dogs received only a single dose of prednisone prior to the skin biopsy. Prednisone administration was initiated 2 days prior to the skin biopsy for 2 dogs and 4 days prior to the skin biopsy for 1 dog. The mean starting dose of prednisone was 1.1 mg/kg/d (0.5 mg/lb/d; range, 0.6 to 2.8 mg/kg/d [0.3 to 1.3 mg/lb/d]). Of the 8 dogs that received corticosteroids prior to skin biopsy, 4, 3, and 1 had skin lesions with histologic pattern 1, 2, and 3, respectively.

Seventeen dogs received > 1 systemic medication. The most frequently administered antihistamines were cetirizine (mean dose, 0.9 mg/kg [0.4 mg/lb; range, 0.6 to 1.1 mg/kg], PO, q 24 h; n = 9 dogs) and diphenhydramine (mean dose, 2.3 mg/kg [1.0 mg/lb; range, 1.7 to 2.7 mg/kg {0.8 to 1.2 mg/lb}], PO, q 8 to 12 h; 6 dogs). Six dogs were treated with dietary modifications; 4 dogs were switched to prescription novel-protein or hydrolyzed diets, and 2 dogs were switched to prescription low-fat diets. Three dogs received adjunct topical treatment for the skin lesions; 2 dogs were bathed with chlorhexidine-based shampoo, and 1% silver sulfadiazine cream was applied to the lesions of 1 dog. The mean treatment duration for skin lesions was 28 days (range, 12 to 42 days).

The 8-year-old spayed female Miniature Schnauzer with moderate to severe lymphoplasmacytic gastritis and duodenitis and pancreatitis was discharged from the hospital and prescribed to receive cetirizine and a prescription low-fat diet. The dog began vomiting and developed diarrhea 1 day later and was readmitted to the hospital. It was treated with IV fluid therapy, colloid support, diphenhydramine, sucralfate, esomeprazole, and dexamethasone. Eight days after the dog was first admitted to the hospital, it developed respiratory distress secondary to congestive heart failure and was euthanized.

Treatment outcome

Signs of gastrointestinal disease resolved 1 to 13 days (mean, 4.6 days) following initial examination at the hospital. Follow-up information was available for 18 dogs. Six of those dogs had a recurrence of vomiting, diarrhea, or hematochezia severe enough to warrant veterinary care between 3 and 334 days after initial resolution of the signs of gastrointestinal disease.

The time required for resolution of the skin lesions was difficult to determine because a recheck appointment was not scheduled at a standard time following hospital discharge for all dogs. For 14 of the 18 dogs, skin lesions had completely resolved between 8 and 30 days (mean, 20.8 days) after they first became evident. Three dogs had partial resolution or substantial improvement of skin lesions at 3, 7, and 8 days after onset. The skin lesions had not resolved for the dog that was euthanized because of respiratory distress secondary to congestive heart failure. Skin lesions did not recur in any of the 6 dogs that had a recurrence of signs of gastrointestinal disease. A spayed female Labrador Retriever that was initially treated at 11 months old for acute onset of vomiting, hematochezia, and erythematous macules and wheals was reevaluated 3 years later for a 1-week history of acute onset of crusts and erosion on the limbs that progressed to generalized erythroderma, facial edema, and pruritus; however, signs of gastrointestinal disease did not recur. Multiple skin biopsy specimens were obtained from that dog for histologic evaluation, which revealed eosinophilic dermatitis with extensive dermal edema, ulcers, and perivascular fibrin accumulation around superficial dermal blood vessels.

Discussion

Results of the present study indicated that dogs with a unique syndrome characterized by acute onset of eosinophilic dermatitis associated with gastrointestinal disease (cutaneous syndrome) typically developed macular erythema that was most pronounced, or perhaps most noticeable, over the ventral aspect of the abdomen and inguinal region. Some of the dogs of the present study also developed targetoid macules (flat lesions with a zone of central pallor likely associated with dermal edema), but those lesions were not as common as diffuse, and often intense, erythematous macules and generalized erythroderma. These findings were similar to those of the 2006 case-series report by Mauldin et al2 of 29 dogs with severe eosinophilic dermatitis. Two of the inclusion criteria for that study2 were moderate to severe eosinophilic dermatitis and acute onset of severe erythroderma. Although the dogs enrolled in the present study had to have signs of gastrointestinal disease, the other inclusion criteria were intentionally broadened from those used in the other case-series report2 to elucidate the potential variability in the clinical signs and histologic features associated with this cutaneous syndrome. Other skin lesions described for some of the dogs of the present study included wheals and facial edema, but unlike angioedema or urticaria associated with an acute hypersensitivity reaction, those lesions persisted for > 48 hours.

Eight of the 18 dogs of the present study had hypoalbuminemia. Hypoalbuminemia was a common finding in dogs of other studies.1,2 Although hypoalbuminemia may have been a consequence of the underlying gastrointestinal disease, it has also been described in dogs with similar skin lesions and no signs of gastrointestinal disease. In those dogs, hypoalbuminemia was attributed to protein leakage from small blood vessels in the skin.1 Albumin is a negative acute-phase protein and may decrease with inflammation, which may have contributed to the hypoalbuminemia for some dogs of this study.

The dogs of the present study had skin lesions that were characterized by 3 distinct histologic patterns, which represented a progressive spectrum that ranged from minimal to mild infiltration of inflammatory cells (pattern 1) to severe eosinophilic inflammation (pattern 3). It is important for pathologists to recognize that these 3 patterns can all be associated with this cutaneous syndrome, and that those histopathologic features correlate with an acute type I hypersensitivity reaction but not with a typical urticarial reaction or atopic dermatitis. In dogs, histopathologic changes typically associated with atopic dermatitis include epidermal hyperplasia, spongiosis, and a mixed superficial perivascular dermatitis characterized by an infiltrate of mononuclear cells, mast cells, and a relative paucity of eosinophils.8 Eight dogs of the present study had skin lesions with histologic pattern 1, and those dogs represented the greatest diagnostic challenge for the examining pathologist because eosinophils in the stained biopsy specimens were often sparse and degranulated eosinophils were easily mistaken for neutrophils. Superficial dermal edema was the most prominent feature in those lesions, and Luna staining was helpful for identification of degranulated eosinophils. The subtle histopathologic changes associated with pattern 1 lesions highlight the importance of correlating histologic findings with clinical signs by both the pathologist and clinician. This unique cutaneous syndrome has distinct clinical signs and should be suspected for any dog with compatible clinical lesions even in the absence of histologic evidence of severe eosinophil infiltration or collagen flame figures in skin biopsy specimens.

In the present study, there was no correlation between the histologic pattern for skin biopsy specimens and clinical appearance of skin lesions. However, 4 of the 8 dogs with histologic pattern 1 had only erythematous macules or generalized erythroderma, whereas all 6 dogs with histologic pattern 2 and 4 dogs with histologic pattern 3 had erythematous wheals or plaques (with or without exudation) or regional edema in addition to erythematous macules or generalized erythroderma. On average, skin biopsy specimens from dogs with pattern 2 or 3 lesions were obtained 2 to 3 days sooner after manifestation of skin lesions than were those from dogs with pattern 1 lesions. This was most likely because dogs with pattern 2 or 3 lesions were more likely to have more prominent or severe clinical lesions (wheals, plaques, or facial edema) than dogs with pattern 1 lesions. However, because of the retrospective nature of this study, it is possible that the description of the skin lesions in the medical records and biopsy specimen submission forms were incomplete for some dogs. Furthermore, the histologic pattern may have been influenced by skin biopsy site selection. For example, plaques (ie, palpably raised lesions) are more likely to have severe histopathologic changes, such as a pronounced inflammatory infiltrate within the dermis, than are macules (ie, palpably flat lesions), which often only have mild eosinophilic dermatitis with edema. Also, 8 dogs of the present study received corticosteroids prior to collection of the skin biopsy specimens, although 5 of those 8 received only 1 dose of prednisone prior to the biopsy. Corticosteroid administration may have suppressed eosinophilic inflammation in those dogs, which could have affected the histologic findings for the skin biopsy specimens. Prospective studies in which lesion descriptions and scoring systems, biopsy procedures (time of specimen acquisition after lesion onset and number of specimens acquired from each site), and treatments (eg, antihistamines and corticosteroids) administered before skin biopsy are standardized are necessary to validate the findings of the present study.

This cutaneous syndrome in dogs has been likened to eosinophilic cellulitis (Wells syndrome) in humans.1,2 In human patients with eosinophilic cellulitis, the histopathologic changes observed vary on the basis of the stage of the lesion at the time the biopsy specimen was obtained. Early lesions are characterized by dermal edema and eosinophil infiltration, whereas resolving lesions often contain low numbers of eosinophils and high numbers of macrophages surrounding collagen flame figures.4 In the present study, some of the histopathologic changes observed in specimens with severe histologic pattern 2 overlapped those observed in specimens with mild histologic pattern 3, which suggested that the 3 histologic patterns described represent varying degrees of severity of the same clinical entity rather than distinct clinical entities.

Results of the present study suggested that this cutaneous syndrome is associated with > 1 gastrointestinal disease, including pancreatitis, inflammatory bowel disease, and adverse food reaction. The definitive cause of the gastrointestinal signs was not determined for 8 of the 18 dogs of this study, and it is possible that those dogs had 1 or more gastrointestinal diseases because the diagnostic workup varied from dog to dog and some diagnostic procedures, such as limited-ingredient diet trials and gastrointestinal tract biopsies, were performed for only a few dogs. Although eosinophilic cellulitis in human patients has not been associated with pancreatitis, it has been reported in a patient with inflammatory bowel disease (ulcerative colitis).9,10 Five of the 18 dogs of the present study had some type of inflammatory bowel disease. In humans, eosinophilic cellulitis has been associated with a number of different triggers such as insect bites or stings,4,11 viral and bacterial infections, and endoparasitism including ascariasis and giardiasis.12–15 One of the dogs of the present study was reported to have sustained multiple bee or wasp stings prior to the onset of vomiting and erythroderma. Similar to the case-series report by Mauldin et al,2 none of the 7 dogs that had fecal flotations performed had positive results for parasite ova, and neither of the 2 dogs that had bacterial culture of a fecal sample performed had growth of gastrointestinal pathogens, such as Campylobacter spp and Salmonella spp. However, those results should be interpreted cautiously because of the small number of dogs evaluated, and bacterial or viral infection or endoparasitism cannot be ruled out as a potential cause of gastrointestinal disease in dogs with this cutaneous syndrome on the basis of the results of this study.

In humans, eosinophilic cellulitis is associated with neoplasia, such as hematologic cancers and carcinomas16–20; therefore, patients with eosinophilic cellulitis are often screened for the presence of underlying malignancies. In dogs, a definitive association between this cutaneous syndrome and neoplasia has not been identified; however, 1 dog in the present study had radiographic evidence of a primary lung tumor. That dog was subsequently determined to have inflammatory bowel disease on the basis of histologic evaluation of biopsy specimens from the gastrointestinal tract that were obtained by endoscopy, and the contribution of the primary lung tumor to the development of skin lesions was unknown. In the Mauldin et al2 case-series report, a 4-year-old Doberman Pinscher had a low-grade pulmonary carcinoma excised 8 months before skin lesions became evident.

One of the dogs of the present study was examined because of pyrexia, vomiting, regurgitation, inappetence, and difficulty rising in addition to skin lesions. Immune-mediated polyarthritis was diagnosed in that dog on the basis of results of various diagnostic tests, including arthrocentesis and cytologic evaluation of joint fluid. Some human patients with eosinophilic cellulitis develop pyrexia and joint pain in addition to skin lesions,4 and the condition has been associated with immune-mediated diseases, such as systemic lupus erythematosus.21 Collectively, the results of the present study suggested that this cutaneous syndrome is a reaction that can be associated with a number of different underlying conditions.

In human patients, eosinophilic cellulitis is associated with the administration of various medications including antimicrobials, NSAIDs, tumor necrosis factor antagonists, and vaccines.22 In the Mauldin et al2 case-series report, 7 of 22 dogs with gastrointestinal disease and skin lesions had a positive drug score. In the present study, 8 of 18 dogs had drug scores that were indicative of a possible adverse drug reaction as determined by use of the Naranjo algorithm (drug score, 1 to 4). However, 4 of those 8 dogs had very low scores (drug score, 1 or 2), and none of the 18 study dogs had drug scores consistent with a probable (drug score, 5 to 8) or definitive (drug score, ≥ 9) adverse drug reaction. Accurate assignment of drug scores to the dogs of this study was difficult because of its retrospective nature. Most dogs were treated with multiple drugs, and re-exposure to a drug after it had been discontinued was rare. Also, none of the dogs were treated by discontinuing administration of all drugs only. Although the attending clinician for some dogs strongly suspected an adverse drug reaction (particularly when metronidazole or a β-lactam antimicrobial was being administered), it was considered unlikely after careful review of the medical records. For example, an adverse reaction to metronidazole was strongly suspected for a 7-year-old Papillon; however, that dog was treated with multiple other medications including gastroprotectants and a heartworm preventative that contained ivermectin and pyrantel pamoate before the manifestation of skin lesions. That dog had a recurrence of vomiting and diarrhea 1 month after resolution of the skin lesions and was again treated with metronidazole, maropitant, and famotidine. It did not redevelop skin lesions; thus, an adverse reaction to metronidazole as a cause of the skin lesions was not supported. A Basset Hound in the Mauldin et al2 report was similarly rechallenged with metronidazole and did not redevelop skin lesions. The role of metronidazole in eosinophilic dermatitis in dogs has been of particular interest for the veterinary dermatology community. An adverse reaction to metronidazole could not be ruled out for an 8-year-old mixed-breed dog of the present study. That dog was initially treated with famotidine, maropitant, and metronidazole for signs of gastrointestinal disease. Maropitant and famotidine were discontinued when the dog developed skin lesions, but the metronidazole was continued for 7 days after the skin lesions became evident. During that time, the skin lesions (erythematous macules with central clearing) continued to progress and did not regress when metronidazole was discontinued. The dog was not rechallenged with metronidazole; therefore, an adverse reaction to metronidazole could not be ruled out and was purely speculative.

The pathogenesis of this unique cutaneous syndrome in dogs is unknown. Similarly, the pathogenesis of eosinophilic cellulitis has not been completely elucidated in human patients. Cytokine production, particularly IL-5, by clonal populations of CD4+CD7- T cells and degranulation of CD25+ eosinophils following IL-2 stimulation have been described in some human patients with eosinophilic cellulitis.23–26 Analogous imbalances in T-cell populations or cytokine production may be involved in the pathogenesis of the cutaneous syndrome in dogs, and the underlying gastrointestinal conditions in affected dogs may contribute to changes in the circulating or cutaneous cytokine milieu. For example, acute pancreatitis may be accompanied by massive activation of cytokines and chemokines (ie, cytokine storm), including IL-5, which is integral for eosinophil activation and differentiation.27,28 In 1 study,29 expression of IL-5 and other cytokines in the duodenal mucosa was greater in German Shepherd Dogs with inflammatory bowel disease than in clinically normal dogs; however, those results were not substantiated by the findings of another study30 in which cytokine expression in the duodenal mucosa of healthy dogs and dogs with inflammatory bowel disease was measured by a real-time reverse-transcriptase PCR assay. The pathogenesis of eosinophilic cellulitis in human patients may differ from that of this cutaneous syndrome in dogs. For example, peripheral eosinophilia is rare in dogs with this cutaneous syndrome but fairly common in human patients with eosinophilic cellulitis.31,32 Also, human patients with eosinophilic cellulitis can develop chronic waxing and waning of the disease.4 Although follow-up was limited for many dogs of the present study, the medical records for the 6 dogs that had recurrence of signs of gastrointestinal disease did not indicate that the skin lesions had recurred. One dog in this study had a recurrence of acute-onset eosinophilic dermatitis 3 years after the initial episode; however, it did not have a recurrence of signs of gastrointestinal disease. The inciting cause of the skin lesions in that dog was not identified for either episode, which further suggested that the pathogenesis of this cutaneous syndrome is multifactorial.

On the basis of the results of the present study, the prognosis appeared favorable for dogs with this cutaneous syndrome. Most dogs evaluated in this study received > 1 systemic medication, with antihistamines and corticosteroids most commonly prescribed specifically for treatment of skin lesions. Corticosteroids are considered the first-line treatment of choice for human patients with eosinophilic cellulitis,33,34 although antihistamines are used as adjunct or occasionally sole treatments.35,36 For the dogs of this study, cetirizine was the most commonly prescribed antihistamine. Cetirizine is an H1 receptor antagonist, and it and its active enantiomer, levocetirizine, have direct inhibitory effects on eosinophils of humans and mice by decreasing the production of cytokines and chemokines that serve as eosinophil chemoattractants,37,38 inhibiting eotaxin-induced eosinophil migration through endothelial cells,39 and decreasing survival of eosinophils in the presence of IL-5.40 Those antieosinophilic effects may make cetirizine particularly useful for the treatment of dogs with this cutaneous syndrome as well as human patients with eosinophilic cellulitis.35

For the dogs of the present study, vomiting and diarrhea appeared to resolve quicker than did the skin lesions (mean time to resolution, 4.6 vs 20.8 days, respectively); however, because of the retrospective nature of the study and lack of standardized recheck times, it was difficult to determine the exact time required for the skin lesions to resolve. The signs of gastrointestinal disease had resolved before hospital discharge for most dogs, whereas the skin lesions were not recorded as resolved until the patient was rechecked. It is likely the skin lesions resolved between hospital discharge and the recheck examination, but that information was not available in the medical record. To better elucidate the clinical course of this cutaneous syndrome in dogs, prospective studies should be performed in which lesion scoring systems and times for recheck examinations are standardized.

In dogs, a unique syndrome characterized by acute-onset erythroderma and concurrent signs of gastrointestinal disease, particularly vomiting, diarrhea, and hematochezia, might represent a reaction pattern to various underlying pathological conditions of the gastrointestinal tract including inflammatory bowel disease, pancreatitis, and adverse food reaction and may occasionally be associated with drug administration. Results of the present study indicated that the histologic features of this syndrome are more variable than previously described and range from mild superficial perivascular eosinophilic dermatitis with dermal edema to severe eosinophilic dermatitis with collagen flame figures and vascular damage. That spectrum of histopathologic changes likely corresponds to lesion severity at the time the biopsy specimen was obtained. Most dogs of the present study responded favorably to treatment for skin lesions, particularly discontinuation of all medications except systemic corticosteroids and antihistamines. Further research is necessary to elucidate the pathogenesis of this unique cutaneous syndrome.

Acknowledgments

No third-party funding or support was received in connection with this study or the writing or publication of the manuscript.

Presented in abstract form at the North American Veterinary Dermatology Forum, Nashville, Tenn, April 2015.

ABBREVIATIONS

IL

Interleukin

References

  • 1. Holm KS, Morris DO, Gomez SM, et al. Eosinophilic dermatitis with edema in nine dogs, compared with eosinophilic cellulitis in humans. J Am Vet Med Assoc 1999;215:649653.

    • Search Google Scholar
    • Export Citation
  • 2. Mauldin EA, Palmeiro BS, Goldschmidt MH, et al. Comparison of clinical history and dermatologic findings in 29 dogs with severe eosinophilic dermatitis: a retrospective analysis. Vet Dermatol 2006;17:338347.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 3. Giannoulopoulos GD, Farmaki R, Koutinas CK, et al. Eosinophilic dermatitis with edema (Wells'-like syndrome) possibly triggered by cooked fish in a dog. J Hell Vet Med Soc 2011;62:320326.

    • Search Google Scholar
    • Export Citation
  • 4. Long H, Zhang G, Wang L, et al. Eosinophilic skin diseases: a comprehensive review. Clin Rev Allergy Immunol 2016;50:189213.

  • 5. Luna LG. Manual of histologic staining methods of the armed forces institute of pathology. 3rd ed. New York: McGraw-Hill Book Co, 1968;111112.

    • Search Google Scholar
    • Export Citation
  • 6. Gomes P, Torres SM, Plager DA, et al. Comparison of three staining methods to identify eosinophils in formalin-fixed canine skin. Vet Dermatol 2013;24:323328.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7. Naranjo CA, Busto U, Sellers EM, et al. A method for estimating the probability of adverse drug reactions. Clin Pharmacol Ther 1981;30:239245.

  • 8. Olivry T & Hill PB. The ACVD task force on canine atopic dermatitis (XVIII): histopathology of skin lesions. Vet Immunol Immunopathol 2001;81:305309.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9. Sakaria SS, Ravi A, Swerlick R, et al. Wells' syndrome associated with ulcerative colitis: a case report and literature review. J Gastroenterol 2007;42:250252.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10. Utikal J, Peitsch WK, Kemmler N, et al. Bullous eosinophilic cellulitis associated with ulcerative colitis: effective treatment with sulfasalazine and glucocorticoids. Br J Dermatol 2007;156:764766.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11. Lin HL, Lin JN, Chen CW, et al. Eosinophilic cellulitis after honeybee sting. J Formos Med Assoc 2009;108:964966.

  • 12. Tsuda S, Tanaka K, Miyasato M, et al. Eosinophilic cellulitis (Wells' syndrome) associated with ascariasis. Acta Derm Venereol 1994;74:292294.

    • Search Google Scholar
    • Export Citation
  • 13. Bassukas ID, Gaitanis G, Zioga A, et al. Febrile “migrating” eosinophilic cellulitis with hepatosplenomegaly: adult toxocariasis—a case report. Cases J 2008;1:356.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14. Hurni MA, Gerbig AW, Braathen LR, et al. Toxocariasis and Wells' syndrome: a causal relationship? Dermatology 1997;195:325328.

  • 15. Eren M & Açikalin M. A case report of Wells' syndrome in a celiac patient. Turk J Gastroenterol 2010;21:172174.

  • 16. Nakazato S, Fujita Y, Hamade Y, et al. Wells' syndrome associated with chronic myeloid leukaemia. Acta Derm Venereol 2013;93:375376.

  • 17. Gambichler T, Othlinghaus N, Rotterdam S, et al. Impetiginized Wells' syndrome in a patient with chronic lymphocytic leukaemia. Clin Exp Dermatol 2009;34:274e275.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18. Spinelli M, Frigerio E, Cozzi A, et al. Bullous Wells' syndrome associated with non-Hodgkin's lymphocytic lymphoma. Acta Derm Venereol 2008;88:530531.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 19. Rajpara A, Liolios A, Fraga G, et al. Recurrent paraneoplastic Wells' syndrome in a patient with metastatic renal cell cancer. Dermatol Online J 2014;20:14.

    • Search Google Scholar
    • Export Citation
  • 20. Hirsch K, Ludwig RJ, Wolter M, et al. Eosinophilic cellulitis (Wells' syndrome) associated with colon carcinoma. J Dtsch Dermatol Ges 2005;3:530531.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21. Yin G & Xie Q. Systemic lupus erythematosus associated with Wells' syndrome. Rheumatol Int 2012;32:10871089.

  • 22. Heelan K, Ryan JF, Shear NH, et al. Wells syndrome (eosinophilic cellulitis): proposed diagnostic criteria and a literature review of the drug-induced variant. J Dermatol Case Rep 2013;7:113120.

    • Search Google Scholar
    • Export Citation
  • 23. Yagi H, Tokura Y, Matsushita K, et al. Wells' syndrome: a pathogenic role for circulating CD4+ CD7- T cells expressing interleukin-5 mRNA. Br J Dermatol 1997;136:918923.

    • Search Google Scholar
    • Export Citation
  • 24. España A, Sanz ML, Sola J, et al. Wells' syndrome (eosinophilic cellulitis): correlation between clinical activity, eosinophil levels, eosinophil cation protein and interleukin-5. Br J Dermatol 1999;140:127130.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 25. Simon HU, Plötz S, Simon D, et al. Interleukin-2 primes eosinophil degranulation in hypereosinophilia and Wells' syndrome. Eur J Immunol 2003;33:834839.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26. French LE, Shapiro M, Junkins-Hopkins JM, et al. Eosinophilic fasciitis and eosinophilic cellulitis in a patient with abnormal circulating clonal T cells: increased production of interleukin 5 and inhibition by interferon alpha. J Am Acad Dermatol 2003;49:11701174.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27. Mansfield C. Pathophysiology of acute pancreatitis: potential application from experimental models and human medicine to dogs. J Vet Intern Med 2012;26:875887.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 28. Pietruczuk M, Dabrowska MI, Wereszczynska-Siemiatkowska U, et al. Alteration of peripheral blood lymphocyte subsets in acute pancreatitis. World J Gastroenterol 2006;12:53445351.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 29. German AJ, Helps CR, Hall EJ, et al. Cytokine expression in mucosal biopsies from German Shepherd Dogs with small intestinal enteropathies. Dig Dis Sci 2000;45:717.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30. Peters IR, Helps CR, Calvert EL, et al. Cytokine mRNA quantification in duodenal mucosa from dogs with chronic enteropathies by real-time reverse transcriptase polymerase chain reaction. J Vet Intern Med 2005;19:644653.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 31. Ferreli C, Pinna AL, Atzori L, et al. Eosinophilic cellulitis (Wells' syndrome): a new case description. J Eur Acad Dermatol Venereol 1999;13:4145.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32. Falagas ME & Vergidis PI. Narrative review: diseases that masquerade as infectious cellulitis. Ann Intern Med 2005;142:4755.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 33. Sinno H, Lacroix JP, Lee J, et al. Diagnosis and management of eosinophilic cellulitis (Wells' syndrome): a case series and literature review. Can J Plast Surg 2012;20:9197.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 34. Chung CL & Cusack CA. Wells syndrome: an enigmatic and therapeutically challenging disease. J Drugs Dermatol 2006;5:908911.

  • 35. Aroni K, Aivaliotis M, Liossi A, et al. Eosinophilic cellulitis in a child successfully treated with cetirizine. Acta Derm Venereol 1999;79:332.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36. Verma P, Singal A & Sharma S. Idiopathic bullous eosinophilic cellulitis (Wells' syndrome) responsive to topical tacrolimus and antihistamine combination. Indian J Dermatol Venereol Leprol 2012;78:378380.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 37. Hasala H, Janka-Junttila M, Moilanen E, et al. Levocetirizine and cytokine production and apoptosis of human eosinophils. Allergy Asthma Proc 2007;28:582591.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 38. Kanei A, Asano K, Kanai K, et al. Inhibitory action of levocetirizine on the production of eosinophil chemoattractants RANTES and eotaxin in vitro and in vivo. In Vivo 2014;28:657666.

    • Search Google Scholar
    • Export Citation
  • 39. Thomson L, Blaylock MG, Sexton DW, et al. Cetirizine and levocetirizine inhibit eotaxin-induced eosinophil transendothelial migration through human dermal or lung microvascular endothelial cells. Clin Exp Allergy 2002;32:11871192.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 40. Sedgwick JB & Busse WW. Inhibitory effect of cetirizine on cytokine-enhanced in vitro eosinophil survival. Ann Allergy Asthma Immunol 1997;78:581585.

    • Crossref
    • Search Google Scholar
    • Export Citation

Appendix

The Naranjo algorithm used for estimation of the probability of an adverse drug reaction in each of 18 dogs with acute onset of erythroderma and signs of gastrointestinal disease that were examined at a veterinary teaching hospital between July 31, 2005, and October 31, 2015.

QuestionYesNoDo not know
Are there previous reports of this adverse event?100
Did this adverse event appear after the suspected drug was administered?2−10
Did the adverse event improve when the drug was discontinued or a specific antagonist was administered?100
Did the adverse event reappear when the drug was readministered?2−10
Are there alternative causes (other than the drug) that could have solely caused this adverse event?−120
Did the adverse event reappear when a placebo was given?−110
Was the drug detected in the blood (or other fluids) in a concentration known to be toxic?100
Was the adverse event more severe when the dose was increased or less severe when the dose was decreased?100
Did the patient have a similar adverse event in response to the same or similar drugs in a previous exposure?100
Was the adverse event confirmed by any objective evidence?100

The scores for all 10 questions were summed. An adverse drug reaction was considered doubtful when the Naranjo (drug) score was < 1, possible when the drug score was between 1 and 4, probable when the drug score was between 5 and 8, and definitive when the drug score was ≥ 9.

All Time Past Year Past 30 Days
Abstract Views 1118 0 0
Full Text Views 2238 1581 210
PDF Downloads 1383 595 49
Advertisement