Association between antiphospholipid antibodies and recurrent fetal loss in women without autoimmune disease: a metaanalysis

OBJECTIVE: To assess the strength of association between recurrent fetal loss (RFL) and presence of antiphospholipid antibodies (aPL) in women without autoimmune disease, and to examine whether magnitude of association varies according to type or titer of antibody and timing of fetal loss. METHODS: We searched Medline and Current Contents for articles published between 1975 and 2003 with terms denoting early (less than 13 weeks) and late (less than 24 weeks) RFL associated with various aPL. Published case-control, cohort, and cross-sectional studies rated moderate or strong were included in our metaanalysis. Pooled odds ratios with 95% CI were generated using the random-effects models with Cochrane Review Manager software. RESULTS: Our analysis included 25 studies. Lupus anticoagulant (LAC) was associated with late RFL (OR 7.79, 95% CI 2.30-26.45); the association of LAC was stronger than that of any other aPL. IgG anticardiolipin antibodies (aCL), when combining all titers, were associated with both early (OR 3.56, 95% CI 1.48-8.59) and late RFL (OR 3.57, 95% CI 2.26-5.65). Restricting analysis to include only women with moderate to high titers increased the strength of association (OR 4.68, 95% CI 2.96-7.40). It was not possible to extract data on isolated low IgG aCL positivity. IgM aCL were associated with late RFL (OR 5.61, 95% CI 1.26-25.03). There was no association found between early RFL and anti-Beta2-glycoprotein I antibodies (OR 2.12, 95% CI 0.69-6.53). CONCLUSION: The magnitude of the association between aPL and RFL varies according to type of aPL. More data on the relationship between recurrent fetal loss and isolated IgM aCL as well as with low titer IgG aCL would be useful. The place of testing for anti-Beta2-glycoprotein I antibodies remains to be determined.