Estimation of anticardiolipin antibodies, anti-beta2 glycoprotein I antibodies and lupus anticoagulant in a prospective longitudinal study of children with juvenile idiopathic arthritis

OBJECTIVE: Anticardiolipin antibodies (aCL) have been frequently detected in juvenile idiopathic arthritis (JIA), but have not been associated with disease activity or clinical features of the antiphospholipid syndrome (APS). Our aim was to determine aCL and anti-beta2 glycoprotein I (anti-beta2GPI) antibody levels and lupus anticoagulant (LA) in serial samples from children with JIA and to investigate the clinical significance of these antibodies. METHODS: The values of aCL, anti-beta2GPI and LA were prospectively followed in 28 children with JIA from disease onset. aCL and anti-beta2GPI were assayed by an ELISA method. Two monoclonal beta2GPI-dependent aCL (HCAL and EY2C9) were used as calibrators. LA was determined by a modified dilute Russell viper venom time test. RESULTS: Thirteen (46.4%) children with JIA were already positive for aCL at their first referral to our center. During the follow-up, the frequency of aCL decreased from 46.4% to 28.6%; however, it remained significantly higher compared with healthy children. In contrast, for anti-beta2GPI the difference in the frequency between the children with JIA and healthy children was not statistically significant. Serial determination of aPL levels in JIA patients revealed frequent fluctuations. Positive aCL persisted over time in 6 (21.4%) children with JIA, 6 (21.4%) children were initially positive for aCL, but became later negative, and 3 (10.7%) children were initially negative for aCL and became later positive. Persistently positive anti-beta2GPI were observed during the follow-up only in one patient, while none of the patients was persistently positive for LA. No association between aCL, anti-beta2GPI or LA and disease activity could be established. No patient with positive aCL, anti-beta2GPI or LA showed any clinical feature of APS. CONCLUSION: The discrepancy between the presence of aCL and anti-beta2GPI might indicate that the production of aCL in JIA is associated with an infectious trigger. Furthermore, the low frequency of anti-beta2GPI and LA could explain the limited prothrombotic potential of aPL observed in JIA. However, we found a distinct group of JIA patients with persistently positive aCL, the clinical implications of which are at the present time unknown.