A nine-year-old girl was referred for management of refractory ulcerative Colitis (UC) that had been diagnosed one year previously. She had no history of reported drug allergies and had been prednisolone-dependent (2 mg/kg/day) for much of the preceding year, with her disease flaring after attempts to reduce the prednisolone dosage. During the hospitalization, she received mesalazine treatment (48 mg/kg/day) without prednisolone, but it was ineffective. She underwent azathioprine therapy (1 mg/kg/day) and 10 days later was hospitalized for fever (temperature of 39.2ºC), skin rash and hematochezia.

Physical examination was significant for numerous erythematous, painful, 1-3 mm vesicular lesions with central pustules on her face and both arms (Fig. 1). Laboratory results showed an elevated white blood cell count (13,470/µL) with neutrophilia, microcytic hypochromic anemia (hemoglobin 9.2 g/dL, MCV 80.9 fl, MCH 24.8 pg), hyponatremia (132 mmol/L), and a markedly raised erythrocyte sedimentation rate (89 mm/hr) and C-reactive protein level (3.13mg/dL). Anti-nuclear antibody was negative, but c-type anti-neutrophil cytoplasmic antibody (c-ANCA) was positive. Her thiopurine methyltransferase (TPMT) activity was normal (18.2 U/ml RBC, reference range: 15.1-26.4 U/mL RBC). Blood cultures and urinalysis were obtained and the patient was started on cefotaxime (100 mg/kg/day) for a possible infectious etiology.

Fig. 1 Pustular and crust lesions surrounded by erythema appeared on face (a) and arm (b) 10 days after administration of azathioprine. (c) Skin biopsy of pustular lesion shows massive neutrophilic infiltration in entire dermis (H&E, x400).

Two days after the cefotaxime treatment, the patient still had a fever. The patient then received metronidazole (30 mg/kg/day) for a week and prednisone (1.7 mg/kg/day). However, she was febrile with shaking chills and nausea. Cultures taken from blood and urine prior to antibiotic therapy were sterile. Biopsy specimens and tissue cultures were taken from the pustular lesions. Pathologic evaluation of skin biopsy showed massive neutrophilic infiltrate in the entire dermis (Fig. 1). Tissue culture results were negative for bacterial or fungal infection. Based on the clinical course, a diagnosis of sweet syndrome was made. As the patient’s fever had not subsided in spite of the administration of antibiotics and steroid, we presumed that the sweet syndrome was caused by the azathioprine and was not due to inflammatory bowel disease. The azathioprine therapy was discontinued. Within 48 hours, the patient’s fever abated and her skin lesions improved. Following this improvement, the prednisolone dose was reduced to 0.4 mg/kg per day without a recurrence of her symptoms.

At follow-up after two weeks, there had been no recurrences of her symptoms, and her UC was comparatively well controlled by prednisolone and mesalazine treatment.

The criteria for drug-induced SS have been reviewed by many authors [3,5] and include abrupt onset of painful erythematous plaques or nodules, histopathologic evidence of a dense neutrophilic infiltrate without evidence of leukocytoclastic vasculitis, pyrexia (temperature >38^ºC), and a temporal relationship between drug ingestion and clinical presentation, as well as resolution after withdrawal. Our patient meets most of these criteria.

ANCAs have been described in some cases, and may be pathogenically relevant through the activation of neutrophils [6]. In our case, c-ANCA was positive. Kemmett, et al. [7] reported the presence of c-ANCA in six of the seven patients with sweet syndrome and speculated whether ANCA may be helpful in establishing the diagnosis of sweet syndrome .

Azathioprine is a widely used immunosuppressive agent that has been used increasingly as a steroid-sparing agent for the treatment of Crohn’s disease and UC. Azathioprine rarely causes a hypersensitivity syndrome which is characterized by fever, headache, arthralgias, and rash, with possible cardiovascular, renal, lung, and hepatic involvement [8]. Skin lesions include erythematous or maculopapular eruptions, vesicules or pustules, urticaria, purpuric lesions, erythema multiforme, or erythema nodosum. A case of acute generalized exantematous pustules induced by azathioprine like our case also has been reported [9]. Diagnosis is often missed or delayed, as the clinical features are often misinterpreted as either sepsis or an exacerbation of the underlying disease state. According to previous studies [10], TPMT activity was not predictive of this type of adverse effect.

The morphology of these skin lesions can mimic that of several other mucocutaneous and systemic conditions. The differential diagnosis includes infectious and inflammatory disorders, neoplastic conditions, reactive erythemas, vasculitis. Skin lesions and negative cultures help in the diagnosis. In addition, negative test results for autoimmune diseases are important for diagnosis. In our case, an infection focus or signs of an autoimmune disease could not be detected. Clinical and histopathologic findings supported the drug-induced sweet syndrome and cessation of the drug caused a rapid regression in symptoms. In patients without prior exposure to azathioprine, signs and symptoms usually begin approximately two weeks from the initial azathioprine exposure [1], which began after 10 days in this child.

We believe that azathioprine-induced sweet syndrome may be under-diagnosed because it can easily be misinterpreted as inflammatory bowel disease-related skin changes.

Contributors: All authors contributed to case work-up and drafting the manuscript.

Funding: None.

Competing interests: None stated.

1. Garey KW, Streetman DS, Rainish MC. Azathioprine hypersensitivity reaction in a patient with ulcerative colitis. Ann Pharmacother. 1998;32:425-8.

2. Paoluzi OA, Crispino P, Amantea A, Pica R, Iacopini F, Consolazio A, et al. Diffuse febrile dermatosis in a patient with active ulcerative colitis under treatment with steroids and azathioprine: a case of Sweet’s syndrome. Case report and review of literature. Dig Liver Dis. 2004;36:361-6.

3. El-Azhary RA, Brunner KL, Gibson LE. Sweet syndrome as a manifestation of azathioprine hypersensitivity. Mayo Clin Proc. 2008;83:1026-30.

4. Yiasemides E, Thom G. Azathioprine hypersensitivity presenting as a neutrophilic dermatosis in a man with ulcerative colitis. Australas J Dermatol. 2009;50:48-51.

5. Su WP, Liu HN. Diagnostic criteria for Sweet’s syndrome. Cutis. 1986;37:167-74.

6. Sarkany RP, Burrows NP, Grant JW, Pye RJ, Norris PG. The pustular eruption of ulcerative colitis: a variant of Sweet’s syndrome? Br J Dermatol. 1998;138:365-6.

7. Kemmett D, Harrison DJ, Hunter JA. Antibodies to neutrophil cytoplasmic antigens: serologic marker for Sweet’s syndrome. J Am Acad Dermatol. 1991;24: 967-9.

8. El-Azhary RA. Azathioprine: current status and future considerations. Int J Dermatol. 2003;42:335-41.

9. Elston GE, Johnston GA, Mortimer NJ, Harman KE. Acute generalized exanthematous pustulosis associated with azathioprine hypersensitivity. Clin Exp Dermatol. 2007;32:52-3.