Systemic sulfa-induced Sweet's syndrome



Sweet's syndrome is usually characterized as a cutaneous disorder, but may present as a systemic disease that is potentially life threatening. Therefore, early recognition and treatment are crucial. We present a case of histiocytoid-variant Sweet's syndrome that mimicked an antineutrophil cytoplasmic antibody–associated small-vessel vasculitis in its presentation.

We describe a 47-year-old man with systemic Sweet's syndrome presenting with inflammatory arthritis, rash, scleritis, and pulmonary infiltrates after an allergic reaction to trimethoprim/sulfamethoxazole prescribed for treatment of a sinus infection. The skin biopsy result revealed a diffuse infiltrate of neutrophils, papillary dermal edema, and no evidence of vasculitis. The patient was treated with systemic corticosteroids, resulting in complete resolution of systemic and cutaneous disease.

Case Report

A 47-year-old man presented with a migratory polyarthritis of 4 months' duration. His disease began when he developed fever, chills, nausea, and hypotension while receiving trimethoprim/sulfamethoxazole for treatment of a sinus infection. His initial symptoms resolved with discontinuation of the drug and the administration of antihistamines, but he soon developed pain and swelling of the ankles, knees, shoulders, elbows, wrists, and fingers, associated with morning stiffness lasting throughout the day, extreme fatigue, and low-grade fevers. The patient denied conjunctivitis, oral ulcers, sicca symptoms, xerostomia, Raynaud's phenomenon, low back pain, heel pain, and urethritis. There was no history of significant medical illness, and he did not smoke or drink alcoholic beverages.

On physical examination, the patient had synovitis of the ankles, wrists, metacarpophalangeal joints, and proximal interphalangeal joints. Initial laboratory results revealed an erythrocyte sedimentation rate of 75 mm/hour and a C-reactive protein level of 44.8 mg/liter. Rheumatoid factor, anti–cyclic citrullinated peptide antibody, antinuclear antibody, hepatitis B surface antigen and antibody, and hepatitis C antibody assays were negative. Radiographs of his hands and feet showed no abnormalities.

Ten days after his initial presentation, the patient was admitted to the hospital with shortness of breath, pleuritic chest pain, and dyspnea on exertion. Computed tomography of the chest showed patchy peribronchovascular areas of consolidation and ground-glass opacities. There was no evidence of pulmonary embolism. Bronchoscopy was performed, and bacterial and fungal cultures obtained by bronchoalveolar lavage were negative. The patient was treated with levofloxacin, which did not ameliorate his symptoms, and he was discharged to home.

Shortly thereafter, the patient was seen in our rheumatology clinic and was noted to have a unilateral right-sided scleritis and tender, red circinate plaques over his arms, shoulders, back, and legs. Human immunodeficiency virus antibody test and viral load assays were negative. Blood cultures were negative. There was no neutrophilia or eosinophilia on a white blood cell differential, and the patient's creatinine level was normal. Urinalysis revealed 3–10 red blood cells per high-power field with no protein or casts. The antineutrophil cytoplasmic antibody assay was negative for anti–serine proteinase 3, but antimyeloperoxidase antibodies were detected at >100 units/ml. Skin biopsy samples from the right arm and left ankle were obtained. Both biopsy samples showed a diffuse infiltrate of monocytoid cells, with immature myelocyte forms, occasional mature neutrophils, and mild papillary dermal edema without vasculitis (Figure 1). The cells comprising the infiltrate labeled with CD68, myeloperoxidase, and CD15 in varying degrees were consistent with myeloid lineage and supported the diagnosis of Sweet's syndrome. The nature of the infiltrate was consistent with the so-called mononuclear or histiocytoid variant of Sweet's syndrome. There was no deposition of IgG, IgM, IgA, or complement component C3 along the dermal–epidermal junction or in the vicinity of blood vessels on direct immunofluorescence studies.

Figure 1.

A, Diffuse dermal infiltrate of neutrophils and mononuclear cells, with neutrophilic nuclear dust (original magnification × 100). B, CD68 immunoperoxidase stain labels the majority of cells within the infiltrate, indicating monocyte/macrophage lineage (original magnification × 100). C, Myeloperoxidase immunoperoxidase stain labels a subset of the cells, indicating myeloid lineage more specifically (original magnification × 200).

The patient was started on 60 mg of prednisone per day, and his symptoms resolved. A chest radiograph demonstrated resolution of his pulmonary infiltrates. Peripheral blood smear remained normal. The patient's steroids were tapered and discontinued after 6 months. After more than 1 year since presentation, he is not receiving any immunosuppressive medication, and has had no recurrence of arthritis, rash, episcleritis, pulmonary infiltrates, or positive antimyeloperoxidase antibodies.


The definitive diagnosis of Sweet's syndrome depends on the presence of 2 major criteria and at least 2 of 4 minor criteria. The 2 major criteria are the abrupt onset of tender or painful plaques and a neutrophilic dermal infiltration determined by biopsy, without evidence of vasculitis. The 4 minor criteria are a preceding fever or infection; constitutional manifestations including fever, arthralgias, conjunctivitis, or malignancy; leukocytosis; and a clinical improvement with corticosteroids but not with antibiotics (1). Our patient clearly met the criteria for an unambiguous diagnosis.

Sweet's syndrome, also known as benign febrile neutrophilic dermatosis, is usually considered a cutaneous disease, but it can occur as a systemic disorder (2) with involvement of other organs, including the lungs, joints, and eyes, as happened in our case. Vignon-Pennamen (3) reviewed the extracutaneous manifestations in 27 patients with biopsy-proven Sweet's syndrome. Twenty-five (93%) had culture-negative pulmonary infiltrates. Joint involvement occurred in ∼50% and ocular involvement including conjunctivitis, scleritis, and iritis varied in frequency from 30% to 75% of patients.

The pathogenesis of Sweet's syndrome is unknown, but the disease has been associated with prior infection, autoimmune disease, malignancy, and drug reactions. Although numerous drugs have been implicated in the causation of Sweet's syndrome, the sulfonamides and related agents have been most closely linked to the disorder (4–6). Given the temporal relationship between the administration of trimethoprim/sulfamethoxazole and the onset of disease, we believe that our patient had sulfonamide-induced Sweet's syndrome. Because of the presence of antimyeloperoxidase antibodies, our initial clinical diagnosis was drug-induced microscopic polyangiitis. Although several investigators have reported the presence of antibodies to neutrophilic cytoplasmic antigens in patients with Sweet's syndrome, we believe this to be the first case report of specific antimyeloperoxidase antibodies in a patient with Sweet's syndrome (7–10). The skin biopsy, however, excluded any form of vasculitis and revealed the proper diagnosis. The most common characteristic microscopic features of Sweet's syndrome include a diffuse infiltrate of neutrophils, neutrophilic nuclear dust, and often eosinophils, with a variable degree of subepidermal edema. Vasculitis is rarely a feature, and may be secondary. This patient's biopsy samples revealed a histologic variant of Sweet's syndrome, the so-called mononuclear or histiocytoid variant, in which besides mature segmented neutrophils, there is a population of more immature myeloid or histiocytic cells (11). In some cases, this latter population may predominate, making immunoperoxidase studies necessary to identify the lineage of the cells.

We recommend that the diagnosis of Sweet's syndrome be considered in any patient with arthritis, pulmonary infiltrates, and ocular disease who develops a rash, regardless of the presence of antimyeloperoxidase antibodies. A skin biopsy is required to make the diagnosis.


Dr. Margaretten had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Margaretten, Fye.

Acquisition of data. Margaretten, Ruben.

Analysis and interpretation of data. Margaretten, Ruben, Fye.

Manuscript preparation. Margaretten, Ruben, Fye.