OBJECTIVES: To evaluate the incidence of extracutaneous manifestations and to identify predictive factors for renal involvement in adult patients with Schönlein-Henoch purpura. DESIGN: Retrospective study with a comparative analysis of patients with and without renal involvement. SETTING: Patients who were attending the dermatologic department of an academic medical center. PATIENTS: In patients with purpura of the lower limbs and cutaneous vascular IgA deposits for which cases were recorded from 1985 to 1993, the following selection criteria were used: age older than 15 years and absence of thrombocytopenia, of IgA deposits in the basement membrane zone, and of a known hematologic or connective tissue disorder. MAIN OUTCOME MEASURES: Clinical and biological data, results of histological studies, and findings from direct immunofluorescence studies of skin biopsy specimens were compared in patients with and without renal involvement. RESULTS: Fifty-seven patients were included: 23% had an IgA glomerulonephritis confirmed by results of a renal biopsy, and a further 26% showed abnormalities on urine microscopy. Joint and gastrointestinal involvement was noted in, respectively, 33% and 19% of the patients. A comparative analysis of patients with and without renal involvement failed to reveal significant differences with regard to age, sex, the presence of bullous or necrotic cutaneous lesions, gastrointestinal or joint involvement, histological features, and findings from direct immunofluorescence studies. An IgA glomerulonephritis was significantly associated with purpura above the waist (P = .03), a recent infectious history (P = .02), pyrexia (P = .01), and biological markers of inflammation (P = .006). CONCLUSIONS: Despite a lower incidence of systemic involvement compared with that in other published series, the incidence of renal involvement remained high (ie, between 23% and 49%). A recent infectious history, pyrexia, the spread of purpura to the trunk, and biological markers of inflammation were predictive factors for renal involvement.
OBJECTIVES: To evaluate the incidence of extracutaneous manifestations and to identify predictive factors for renal involvement in adult patients with Schönlein-Henoch purpura. DESIGN: Retrospective study with a comparative analysis of patients with and without renal involvement. SETTING:Patients who were attending the dermatologic department of an academic medical center. PATIENTS: In patients with purpura of the lower limbs and cutaneous vascular IgA deposits for which cases were recorded from 1985 to 1993, the following selection criteria were used: age older than 15 years and absence of thrombocytopenia, of IgA deposits in the basement membrane zone, and of a known hematologic or connective tissue disorder. MAIN OUTCOME MEASURES: Clinical and biological data, results of histological studies, and findings from direct immunofluorescence studies of skin biopsy specimens were compared in patients with and without renal involvement. RESULTS: Fifty-seven patients were included: 23% had an IgA glomerulonephritis confirmed by results of a renal biopsy, and a further 26% showed abnormalities on urine microscopy. Joint and gastrointestinal involvement was noted in, respectively, 33% and 19% of the patients. A comparative analysis of patients with and without renal involvement failed to reveal significant differences with regard to age, sex, the presence of bullous or necrotic cutaneous lesions, gastrointestinal or joint involvement, histological features, and findings from direct immunofluorescence studies. An IgA glomerulonephritis was significantly associated with purpura above the waist (P = .03), a recent infectious history (P = .02), pyrexia (P = .01), and biological markers of inflammation (P = .006). CONCLUSIONS: Despite a lower incidence of systemic involvement compared with that in other published series, the incidence of renal involvement remained high (ie, between 23% and 49%). A recent infectious history, pyrexia, the spread of purpura to the trunk, and biological markers of inflammation were predictive factors for renal involvement.