Impact of serological activity on flare following clinically inactive disease and remission in childhood-onset systemic lupus erythematosus.

OBJECTIVES:To assess the association between serological activity (SA) and clinical inactivity in SLE and to investigate whether SA predicts flare after clinically inactive disease (CID) and remission. METHODS:Longitudinal data of children from 3 paediatric rheumatology referral centres were retrospectively reviewed. CID was defined as clinical SLEDAI = 0 in patients with a prednisolone dose < 15 mg/day. A modified DORIS remission on treatment criteria was used to determine remission. RESULTS:Of the 124 patients included, 89.5% displayed SA at onset. Through follow-up, the rate of SA decreased to 43.3% at first CID and 12.1% at remission. Among patients with CID, 24 (20.7%) experienced a moderate to severe flare before the attainment ofremission. While previous proliferative lupus nephritis (OR : 10.2, p: 0.01) and autoimmune haemolytic anaemia (OR : 6.4, p: 0.02) were significantly associated with an increased odds of flare after CID, SA at CID was not associated with flare. In contrast, 21 (19.6%) patients experienced a flare in a median of 18 months after remission. Hypocomplementemia (OR : 9.8, p: 0.02) and a daily hydroxychloroquine dose < 5 mg/kg (OR : 5.8, p: 0.02) at remission significantly increased the odds of flare. CONCLUSION:SA increases the odds of flare at remission but not at CID. Suboptimal dosing of hydroxychloroquine should be avoided, especially in children with SA in remission to lower the risk of flares.