Objective To evaluate the prevalence of different definitions of remission and their effect on damage in systemic lupus erythematosus (SLE).

Design and method We considered 293 caucasian SLE patients followed-up for 7 years (2009–2015): 253 (86.3%) were female, mean ±SD disease duration 11.1±7.8 years. Disease activity was assessed by clinical SLEDAI-2K (c-SLEDAI) and damage by SLICC/ACR Damage Index (SDI). We evaluate the effect of different definitions of remission (c-SLEDAI=0; c-SLEDAI ≤1; c-SLEDAI=0 and prednisone ≤5 mg/day; c-SLEDAI ≤1 and prednisone ≤5 mg/day; c-SLEDAI=0 and PGA <0.5; c-SLEDAI ≤1 and PGA <0.5; c-SLEDA I=0 and prednisone ≤5 mg/day and PGA <0.5; c-SLEDAI ≤1 and prednisone ≤5 mg/day and PGA <0.5) and different durations of remission (1, 2, 3, 4, ≥5 consecutive years) on SDI using multiple logistic regression analysis.

Results Frequency of remission achieved during the 7 year follow-up are reported in table 1 according to the different definitions.

The mean increase in SDI and the percentage of patients with increased of SDI from the baseline to the end of follow-up were significantly higher in unremitted and 1 year remitted patients compared with patients with 2-, 3-, 4- and ≥5 year remission, irrespective of the definition of remission. 5 year remitted patients had lower damage compared with 2 year (p<0.01) and 3 year (p<0.01) remitted patients. At multivariate analysis, a remission lasting at least 2 years was an independent predictor of no damage accrual only in the definitions including prednisone intake ≤5 mg/day and/or PGA <0.5 (table 2).

Conclusions The inclusion of PGA <0.5 in the definition reduces the frequency of remission only in the long-term (≥5 year). A sustained remission, regardless of its definition, is associated with a lower chronic damage development. The addition of prednisone ≤5 mg/day and/or PGA <0.5 to c-SLEDAI=0/≤1 increases the ability to predict the absence of damage accrual compared with cSLEDAI=0/≤1 without substantial differences among them.