Profiles of criteria and non-criteria anti-phospholipid autoantibodies are associated with clinical phenotypes of the antiphospholipid syndrome.

Background: Specific anti-phospholipids antibodies (aPLs) are used as classification criteria of the antiphospholipid syndrome (APS). These aPLs, although essential for diagnosis, do not predict disease phenotypes, which may require specific therapies. Non-criteria aPLs are rarely evaluated and their role is yet to be defined. In the current study, we aimed to examine the association between criteria and non-criteria aPLs and APS phenotypes.

Methods: Serum samples from 188 subjects, 130 APS patients and 58 controls were analyzed for the presence of 20 aPLs (IgG and IgM isotypes to cardiolipin (CL), beta2-glycoprotein1 (β2GP1), phosphatidic acid (P-acid), phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), phosphatidylinositol (PI), phosphatidylserine (PS), annexin-5 (AN) and prothrombin (PT) using a line immunoassay (GA Generic Assays, Germany). Sero-positivity to the different aPLs/aPLs profiles was correlated to APS phenotypes (i.e. arterial thrombosis, CNS manifestations, venous thrombosis, relapsing disease, obstetric morbidity).

Results: In this cohort, arterial thrombosis was associated with accumulative number of ≥ 7/20 aPLs evaluated (OR 4.1; CI 95% 1.9-96, p = 0.001) as well as the sole presence of aPT (IgG) (OR 2.3;CI 95% 1.1-5.1, p = 0.03). CNS manifestations were linked with a profile of 4 aPLs (IgG): aPT, aPG, aPI and aAN (OR 2.6;CI 95% 1.1-6.3, p = 0.03). Symptom-free period of ≥ 3 years was linked with lower number of aPLs and the presence of aPI (IgG) (OR 3.0;CI 95% 1.08-8.1, p < 0.05) or aAN (IgG) (OR 3.4;CI 95% 1.08-10.9, p < 0.05). APS related pregnancy morbidity correlated with a profile of 2 aPLs (IgG): aCL and aPS (OR 2.9; CI 95% 1.3-6.5, p < 0.05) or the sole presence of aAN (IgG) (OR 2.8; CI 95% 1.02-8, p = 0.05).

Conclusion: In this study, we observed an association between specific criteria/non-criteria aPLs or aPLs profiles and clinical phenotypes of APS. Our data suggest that examination of a wider variety of aPLs may allow better characterization of APS.