Sweet syndrome has been known as acute febrile neutrophilic dermatosis1 since Dr. Robert Douglas Sweet reported a series of observations in 8 women who presented the 4 essential characteristics of the disease: fever, neutrophilic leukocytosis, massive dermal neutrophilic infiltrate without vasculitis, and acute eruption in the form of generalized edematous and erythematous nodules or plaques. The diagnostic criteria still used today are those proposed in 1987 by Su and Liu,2 with slight modifications made by Von den Driesch et al.3 between 1989 and 1990.

Given that Sweet syndrome is an acute process that can arise from a variety of causes and associations, robust data on its incidence and prevalence in the general population are lacking.4 Etiology influences the distribution by sex,5 that is, parainflammatory, idiopathic, and drug-induced cases are more commonly observed among females, whereas malignancy-associated and infantile cases are more common among males. Etiology also influences distribution by age, with idiopathic and inflammatory cases being more common in individuals aged 20–60 years.

Depending on associated diseases or etiology, Sweet syndrome is classified as idiopathic, parainflammatory (including infections and inflammatory diseases), malignancy-associated, drug-induced, and pregnancy-associated.6

Our objectives when evaluating the present case series were to define the profile of patients with Sweet syndrome in the study population and to identify subgroupings based on etiology, associated symptoms, triggers, sex, age at onset, time of the year, disease duration, recurrences, treatment, and abnormal laboratory findings. The characteristics of this series were compared with those reported in the literature.

We performed a retrospective study of patients attended at the dermatology department of Hospital Universitario Dr. Peset in Valencia, Spain, between 2001 and 2009. The hospital serves a population of 379000. The sample comprised 24 patients who fulfilled the diagnostic criteria of Su and Liu2 (Table 1).

Data were recorded in a data table (Excel, Microsoft Office 2007) and analyzed using SPSS version 17.

The characteristics of the study sample are shown in Table 2. Women accounted for 54.2% of the sample and men 45.8% (P=.563). Mean (SD) age was 56.45 (18.40) years for men and 52.00 (18.68) years for women (P=.564). Patients aged 70–79 years comprised the largest age group.

Sweet syndrome was most often associated with infection or inflammation (P=.002).

In this series, 81.3% of patients had fever (P=.007), 29.2% arthralgia (P=.003), 4.1% eye problems (P=.001), and 8.3% involvement of the oral mucosa (P=.001).

Corticosteroids were administered to 16 patients (66.6%): 12 were prescribed a systemic regimen and 4 a topical form. Two patients (8.3%) received systemic treatment, and the remainder (25%) required no treatment.

The most frequently affected site was the arms (66.6%), followed by the legs and trunk in equal proportions (54.4%); the least often affected sites were the face and neck (16.7%) and the dorsum of the hands (12.8%). Lesions most commonly took the form of erythematous and edematous plaques (91.7%); 45.8% of patients had vesicular-like or bullous lesions and 20.8% had pustular forms.

Lesions took longer to heal and recurrences of lesions continued longer in patients with malignancy-associated disease. Lesions also took longer to heal in patients with idiopathic disease. In contrast, lesions healed fastest in patients whose disease was related to an infectious or inflammatory cause (P=.042).

In patients aged 30–49 years, idiopathic disease and infectious or inflammatory causes predominated, whereas in patients aged 60–79 years the 4 forms were more equally distributed. Malignancy-associated Sweet syndrome was mostly observed in the eldest patients.

A comparison of our findings with those of other published series7–12 revealed the following differences for our series:

• 1.

  Smaller variation in the ratio of women to men

• 2.

  Greater frequency of the parainflammatory etiology than of the idiopathic etiology

• 3.

  Longer duration of disease in patients with malignancy-associated and idiopathic forms than in those with parainflammatory disease

• 4.

  One case associated with infliximab (first reported case)

We found that most cases of Sweet syndrome associated with infection of the upper respiratory tract occurred during winter, although it was noteworthy that a Portuguese series9 identifies a seasonal peak in autumn in this subgroup; the difference would be attributable to differences in climate between Valencia (Mediterranean) and Portugal (Atlantic). In the Mediterranean area, winter temperatures are mild and equivalent to autumn temperatures in Portugal. Therefore, the seasonal distribution of infectious Sweet syndrome should be interpreted with respect to the climate of the area under study. We observed that 50% of patients with malignancy-associated disease had a solid tumor, in contrast with findings reported in the literature, in which hematologic tumors clearly predominate.13,14 The remaining characteristics of this subgroup were similar to those described elsewhere.13–15 Three patients in this subgroup (75%) had vesicular and/or bullous lesions. We observed only 1 case of drug-induced Sweet syndrome, involving infliximab in a 51-year-old man. This is the first such case described to date. The patient was receiving the drug for psoriatic arthritis. Symptoms of Sweet syndrome appeared after the 2 most recent infusions. The drug was withdrawn, and no further recurrences were observed. We recorded no cases of pregnancy-associated Sweet syndrome.

Oral corticosteroids were administered to 50% of patients at dosages of 0.5–1mg/kg for periods ranging from 2 weeks to 3.5 months. Recurrences after interruption of treatment were recorded in only 2 of the patients who received corticosteroids, i.e., 16.6%, a percentage which is lower than those reported for other series.7,8 Topical corticosteroids were sufficient in 4 cases, sometimes in combination with a topical antibiotic. Drawing poultices were necessary to treat the bullous lesions in 2 cases. We did not use any of the other systemic treatments reported in the literature.

The present study is limited by its small size, which is a consequence of the low frequency of the condition described. As this was a retrospective study of features of an acute disease in a patient series, the classification of patients into specific etiologic groups could vary from classifications that would have been made had poststudy outcomes been recorded. Finally, we believe that the total number of cases of Sweet syndrome in the population could have been greater than the number we recorded: as this is an acute condition that resolves spontaneously in most cases (even without treatment), it could be underreported.

The authors declare that they have no conflicts of interest.