Damage did not independently influence mortality in childhood systemic lupus erythematosus

To analyze clinical, laboratory and treatment features associated with death in a childhood-onset SLE population.

Patients with childhood-onset SLE followed at the State University of Campinas, Brazil, between 1980 and 2002 were included. Data on clinical and laboratory features of the disease were collected regularly. Logistic regression was used for analyzing association between clinical and laboratory features and death. Kaplan-Meyer tests were used to estimate the survival curves.

Of 61 patients identified, six were lost to follow-up during the first year of disease. The mean follow-up period of the remaining 55 patients was 3.25 years (SD=1.2). Mean SLICC/ACR-DI score was 4.9 (SD=3.4). Death occurred in 12 (21.8%) of 55 patients. Direct causes of death were: infection in six (50%), stroke in four (33.3%), and renal insufficiency in two (16.7%). Five patients (41.7%) died during the first 5 years of disease due to infection. Male gender (p=0.004; OR=9.1; 95% CI=7.6-21.0), infection (p=0.001; OR=4.2; 95% CI=1.6-15.2) and nephritis (p=0.02; OR=2.3; 95% CI=1.3-5.2) were independent factors associated with death in the multivariate analysis. The global survival rate adjusted for duration of disease was 93.9% in the first year of disease, 88.9% in the second year, 80.8% in the fifth year and 48.1% in 20 years of follow-up. When comparing survival curves, male gender, the presence of infection during the course of the disease and the presence of nephritis during follow-up had a worse survival.

Male gender, the presence of infection and nephritis were independent risk factors for death in our Brazilian cohort. Damage did not independently influence survival in this study.