Objective: To determine the degree to which changes in anti-double-stranded DNA (anti-dsDNA), as determined by Crithidia and enzyme-linked immunosorbent assays (ELISAs), precede or coincide with changes in systemic lupus erythematosus (SLE) activity, as measured by 5 clinical indices, the physician's global assessment (PGA), modified SLE Disease Activity Index (M-SLEDAI), modified Lupus Activity Index (M-LAI), Systemic Lupus Activity Measure (SLAM), and the modified British Isles Lupus Assessment Group (M-BILAG).

Methods: Disease activity and anti-dsDNA were measured monthly in 53 SLE patients who were followed up for 1 year. Lupus flare was defined as an increase in PGA of > or = 1.0, M-SLEDAI > or = 3, M-LAI > or = 0.1, SLAM > or = 3, and M-BILAG > or = 4 within a 1-month period. Flare rates were calculated for groups, which were defined by "previous" (1 month prior to the flare) or "concurrent" (at the time of the flare) changes in anti-dsDNA. Logistic regression models were used to determine the significance of the association between recent changes in anti-dsDNA and flare, controlling for the prednisone dosage.

Results: Flares occurred at 12% of visits, based on the PGA measure of disease activity. Using the other indices, flare rates were 19% (M-SLEDAI), 25% (M-LAI), 13% (SLAM), and 12% (M-BILAG). A concurrent decrease in anti-dsDNA (ELISA) was associated with significantly higher flare rates based on PGA (18 of 84, 21%; P = 0.0014), M-SLEDAI (27 of 89, 30%; P = 0.0019), M-LAI (37 of 89, 42%; P = 0.0001), and M-BILAG (19 of 89, 21%; P = 0.0264) scores. Flare rates were also significantly higher after a previous increase in anti-dsDNA (ELISA) based on M-SLEDAI (26 of 93, 30%; P = 0.0022) and M-LAI (34 of 93, 37%; P = 0.0117) scores. Flare rates tended to be lowest when there was a concurrent increase in anti-dsDNA (ELISA). Analysis of specific organ systems showed that a concurrent decrease in anti-dsDNA (ELISA) was significantly associated with increases in renal disease activity. Similar results were obtained using the Crithidia assay.

Conclusion: A previous increase in anti-dsDNA levels occurred before SLE flares, as measured by the M-SLEDAI and M-LAI only. However, during lupus flares, including the subset of renal flares, anti-dsDNA levels frequently decreased. We hypothesize that this decrease in anti-dsDNA represents deposition in tissue at the time of flare.