RESUMEN
Objective The objective of this study was to compare demographic data, clinical/laboratorial features and disease activity at diagnosis in three different groups with distinct time intervals between onset of signs/symptoms and disease diagnosis. Methods A multicenter study was performed in 1555 childhood-onset systemic lupus erythematosus (American College of Rheumatology criteria) patients from 27 pediatric rheumatology services. Patients were divided into three childhood-onset systemic lupus erythematosus groups: A: short time interval to diagnosis (<1 month); B: intermediate time interval (≥1 and <3 months); and C: long time interval (≥3 months). An investigator meeting was held to define the protocol. Demographic data, SLICC classification criteria and SLEDAI-2 K were evaluated. Results The number of patients in each group was: A = 60 (4%); B = 522 (33.5%); and C = 973 (62.5%). The median age at diagnosis (11.1 (4.2-17) vs. 12 (1.9-17.7) vs. 12.5 (3-18) years, P = 0.025) was significantly lower in group A compared with groups B and C. The median number of diagnostic criteria according to SLICC (7 (4-12) vs. 6 (4-13) vs. 6 (4-12), P < 0.0001) and SLEDAI-2 K (18 (6-57) vs. 16 (2-63) vs. 13 (1-49), P < 0.0001) were significantly higher in group A than the other two groups. The frequency of oral ulcers in the palate (25% vs. 15% vs. 11%, P = 0.003), pleuritis (25% vs. 24% vs. 14%, P < 0.0001), nephritis (52% vs. 47% vs. 40%, P = 0.009), neuropsychiatric manifestations (22% vs. 13% vs. 10%, P = 0.008), thrombocytopenia (32% vs. 18% vs. 19%, P = 0.037), leucopenia/lymphopenia (65% vs. 46% vs. 40%, P < 0.0001) and anti-dsDNA antibodies (79% vs. 66% vs. 61%, P = 0.01) were significantly higher in group A compared with the other groups. In contrast, group C had a less severe disease characterized by higher frequencies of synovitis (61% vs. 66% vs. 71%, P = 0.032) and lower frequencies of serositis (37% vs. 33% vs. 25%, P = 0.002), proteinuria >500 mg/day (48% vs. 45% vs. 36%, P = 0.002) and low complement levels (81% vs. 81% vs. 71%, P < 0.0001) compared with groups A or B. Conclusions Our large Brazilian multicenter study demonstrated that for most childhood-onset systemic lupus erythematosus patients, diagnosis is delayed probably due to mild disease onset. Conversely, the minority has a very short time interval to diagnosis and a presentation with a more severe and active multisystemic condition.