Background Cutaneous vasculitis (CV) can be a clinical manifestation of a connective tissue or other autoimmune disease.

Objectives Our aim was to assess the frequency and clinical features of CV associated to connective tissue diseases.

Methods Study of a large series of 766 unselected patients with CV attended in the same tertiary-care hospital. CV was considered as secondary to connective tissue disease or other autoimmune disorder when occurred in the setting of any of these conditions.

Results In 35 (10 men/25 women; mean age±SD, 48±15 years) of 766 patients (4.57%) CV was associated to the following connective tissue disease or autoimmune disorder: systemic lupus erythematosus (SLE) (n=15), rheumatoid arthritis (n=9), Sjögren's syndrome (n=5), ankylosing spondylitis (n=2), relapsing polychondritis (n=1), Behçet's disease (n=1), sarcoidosis (n=1) and autoimmune hepatitis (n=1). All patients developed vasculitic skin lesions: palpable purpura (n=31), maculopapular erythema (n=4) and urticarial lesions (n=2). The most common CV locations were: lower limbs (n=30), upper limbs (n=12) and trunk (n=10). The mean duration of skin lesions was 13.83±7.62 days. Other clinical features were nephropathy (n=9), joint involvement (n=6), constitutional symptoms (n=4) and gastrointestinal involvement (n=1). Regarding laboratory parameters, elevated ESR was observed in 25 patients, hematuria in 4, anemia in 7 and proteinuria in 5 patients. Immunologic findings were as follows: positive rheumatoid factor (10 of 13 patients tested), positive antinuclear antibodies (15 of 19 patients tested), and decreased serum complement levels (11 of 20 patients tested). Twenty-five patients required some drug for CV: corticosteroids (n=22), cyclophosphamide (n=6), methotrexate (n=5), NSAIDs (n=4), azathioprine (n=4), chloroquine (n=3), hidroxichloroquine (n=2), dapsone (n=1) and rituximab (n=2).

Conclusions CV, although infrequent, is not exceptional in the setting of any connective tissue disease or autoimmune disorder.

Acknowledgement This study was supported by a grant from “Fondo de Investigaciones Sanitarias” PI12/00193 (Spain). This work was also partially supported by RETICS Programs, RD08/0075 (RIER) and RD12/0009/0013 from “Instituto de Salud Carlos III” (ISCIII) (Spain).

Disclosure of Interest None declared