Association Between Novel Antiphospholipid Antibodies and Adverse Pregnancy Outcomes

OBJECTIVE: To investigate the value of anti-β2 glycoprotein-I domain 1 (aD1) and antiphosphatidylserine–prothrombin antibodies for predicting adverse pregnancy outcomes in an at-risk population and to describe the relationship among aD1, antiphosphatidylserine–prothrombin, lupus anticoagulant, and other antiphospholipid antibodies (aPL). METHODS: Data were obtained from a prospective cohort of pregnant patients with aPL, with systemic lupus erythematosus (SLE) (n=59) or without SLE (n=106), or SLE without aPL (n=100) (PROMISSE [Predictors of Pregnancy Outcome in Systemic Lupus Erythematosus and Antiphospholipid Syndrome] study; NCT00198068). Levels of aD1 and antiphosphatidylserine–prothrombin were quantified with the QUANTA Flash and QUANTA Lite systems, respectively, in sera collected at less than 18 weeks of gestation. Adverse pregnancy outcome was defined as delivery at before 34 weeks of gestation for preeclampsia or placental insufficiency or fetal death after 12 weeks of gestation. Receiver operating characteristic (ROC) analysis assessed the diagnostic properties of aD1 and antiphosphatidylserine–prothrombin for adverse pregnancy outcomes. Bivariate comparisons were made between each biomarker. Multivariable regression modeling of adverse pregnancy outcomes was performed, and backward selection determined variables for a final model for adverse pregnancy outcomes. Logistic regression of lupus anticoagulant quantified the association with aD1 and antiphosphatidylserine–prothrombin. The rate of adverse pregnancy outcomes was described by combined results of lupus anticoagulant, aD1, and antiphosphatidylserine–prothrombin immunoglobulin G (IgG). RESULTS: Of 265 individuals, 45 (17.0%) experienced adverse pregnancy outcomes. Area under the curve from ROC analysis for aD1 was 0.734 (95% CI, 0.664–0.805), for antiphosphatidylserine–prothrombin IgG was 0.83 (95% CI, 0.751–0.899), and for antiphosphatidylserine–prothrombin immunoglobulin M (IgM) was 0.612 (95% CI, 0.520–0.703). Markers associated with adverse pregnancy outcomes were aD1 ( P <.001), anticardiolipin IgG ( P <.001), β2-glycoprotein I IgG ( P =.003), antiphosphatidylserine–prothrombin IgG ( P <.001), antiphosphatidylserine–prothrombin IgM ( P =.03), and lupus anticoagulant ( P <.001). Backward selection identified lupus anticoagulant, aD1, and antiphosphatidylserine–prothrombin IgG for final adverse pregnancy outcome modeling: lupus anticoagulant odds ratio (OR) 7.0 (95% CI, 3.4–14.4), aD1 OR 12.1 (95% CI, 3.64–40.2), and antiphosphatidylserine–prothrombin IgG OR 11.4 (95% CI, 5.2–25.2). Both aD1 and antiphosphatidylserine–prothrombin IgG remained significant when lupus anticoagulant was removed from the model. Both aD1 and antiphosphatidylserine–prothrombin IgG performed the best in ruling in adverse pregnancy outcomes. With a likelihood ratio less than 0.1, aD1 or antiphosphatidylserine–prothrombin IgG performed well for ruling out adverse pregnancy outcomes. Both aD1 and antiphosphatidylserine–prothrombin IgG were associated with lupus anticoagulant positivity: OR 27.9 (95% CI, 12.1–64.0) if both were positive. Adverse pregnancy outcomes were highest in those with positive lupus anticoagulant, aD1, and antiphosphatidylserine–prothrombin IgG (47.6%). CONCLUSION: In pregnant individuals with aPL, with or without SLE, aD1 and antiphosphatidylserine–prothrombin IgG are significant independent predictors of adverse pregnancy outcomes and are strongly associated with lupus anticoagulant. Combined use may identify patients at greatest risk for severe obstetric complications.