2-Glycoprotein I-dependent anticardiolipin antibodies as a predictor of adverse pregnancy outcomes in healthy pregnant women

Our aim was to elucidate prospectively whether beta 2-glycoprotein I-dependent anticardiolipin antibodies (beta 2GPI-dependent aCL; autoimmune type) can predict an adverse pregnancy outcome in healthy pregnant women and whether beta 2GPI-dependent aCL should be applied for routine screening of the pregnant population. A prospective cohort study was performed on 1600 healthy pregnant women from whom blood samples were obtained at about week 10 of gestation. We used a modified enzyme-linked immunosorbent assay with which to divide the subjects into three study groups: beta 2GPI-independent aCL positive, beta 2GPI-dependent aCL positive and aCL negative. Their subsequent pregnancy outcomes were ascertained and the three study groups were compared statistically for the following poor pregnancy outcomes: intrauterine fetal death (IUFD) after 12 gestational weeks, intrauterine growth retardation (IUGR) and pre-eclampsia. The total number of patients eligible for this study was 1125. The prevalence of beta 2GPI-dependent aCL positive was eight (0.7%), beta 2GPI-independent aCL positive was 17 (1.5%) and aCL negative was 1100 (97.8%). Beta 2-GPI-dependent aCL positivity was significantly associated with poor pregnancy outcome: 25.0% of beta 2GPI-dependent aCL-positive and 0.5% of aCL-negative patients experienced IUFD [relative risk 52.4; 95% confidence interval (CI), 12.7-216.3; P = 0.0009]; 37.5% of beta 2GPI-dependent aCL-positive and 2.9% of aCL-negative patients experienced IUGR (relative risk 18.4; 95% CI, 4.6-74.0; P = 0.001); and 50.0% of beta 2GPI-dependent aCL-positive and 4.0% aCL-negative patients experienced pre-eclampsia (relative risk 22.1; 95% CI, 5.7-85.7; P = 0.0002). In contrast, beta 2GPI-independent aCL did not show any significant association with such adverse pregnancy outcomes. beta 2GPI-dependent aCL are significantly highly associated with adverse pregnancy outcomes in healthy pregnant women and can be used for prediction purposes, whereas beta 2GPI-independent aCL cannot. Our results suggest that routine screening for beta 2GPI-dependent aCL should be introduced for the general pregnant population.