A hallmark of autoimmune conditions is the strong association of many of these diseases with genes in the major histocompatibility complex (MHC) region. Many HLA antigens have been associated with aPL, primary APS (PAPS), and APS associated with other autoimmune diseases (Tables 4.2 and 4.3), producing a complicated and confusing dataset for APS. Many of these reports are small case series, occasionally even single multiplex families, and testing for aPL is frequently incompletely documented (e.g., single test performed, positive cutoff values not provided, and positive results not repeated to confirm). Confirmation of meeting classification criteria for APS is also limited as many of the studies were performed before the existence of criteria. Some of these studies also lack, or fail to specify, appropriate ethnic- or gender-matched controls for the populations studied, where gender or ethnicity might account in part for the frequency of associations.

β₂-glycoprotein I is a phospholipid-binding protein that has been identified as a major antigen in patients with APS. The protein sequence for this plasma protein was published in 1984 [40], and the complementary DNA sequence in 1991 [41, 42]. Several polymorphisms have been identified in the protein [43], including three in the phospholipid-binding fifth domain (Val/Leu²⁴⁷, Cys/Gly³⁰⁶, and Trp/Ser³¹⁶). Studies have investigated the relationship between polymorphisms in β₂GPI and anti-β₂-glycoprotein-I (aβ₂GPI) in various ethnic groups. Gushiken and colleagues found no association between aPL (LA or aβ₂GPI) and the Cys/Gly³⁰⁶ and Trp/Ser³¹⁶ polymorphisms which disrupt the ability of the protein to bind to anionic phospholipids [44], in patients with SLE and/or APS [45]. Similarly, Camilleri and colleagues found no relationship between the Trp/Ser³¹⁶ polymorphism and aPL in patients with thrombosis [46]. Palomo and colleagues did find a significant relationship between this polymorphism and venous and arterial thromboses, but not with aPL, in Chilean patients [47]. Similar findings were reported by Pardos-Gea et al.; i.e., polymorphisms in Trp/Ser³¹⁶, by means of statistically significant associations for an increased S allele and T/S genotype in Spanish Caucasian patients, might play a role in the pathogenic development of PAPS, but not via increased production of aβ₂GPI or other aPL [48].

The Val/Leu²⁴⁷ polymorphism has been more extensively studied. This variant causes a conformational change in β₂GPI [49] hypothesized to result in the exposure of cryptic epitopes [50], theoretically providing a likely target for autoantibodies. Initial studies identified a relationship between this polymorphism and aβ₂GPI in patients with APS, although one study identified this association in Caucasian patients with primary but not secondary APS [51], and a second study identified the association in Asian patients with APS but not Caucasian or African American patients [52]. To evaluate not only potential differences related to ethnicity or the presence of primary versus secondary APS, a recent meta-analysis by Chamorro et al. of eight previous studies, comprising 488 patients meeting classification criteria for APS and 923 controls from eight countries, evaluated the association of APS, PAPS, aβ2GPI, and/or thrombosis with the Val/Leu²⁴⁷ polymorphism [53], determining that patients with APS had a significantly higher prevalence of the Val/Val genotype when compared with controls, with an enhanced association in those with aβ2GPI. No associations were found for the Val/Val phenotype and thrombosis, ethnicity (Caucasians with APS), or PAPS, but these analyses were limited by the small number of studies reporting these data. A separate meta-analysis evaluating the Val/Leu²⁴⁷ polymorphism included 10 studies with a total of 1507 patients and 1450 controls [54]. A limitation of this meta-analysis was that it did not confirm that patients met classification criteria for APS as done in the Chamorro meta-analysis. Only six studies were common to both meta-analyses. Despite the differences, the latter study confirmed the findings of the former except the latter also demonstrated an association of the polymorphism with thrombosis.

Several other genes have been studied in focused attempts to identify relationships with aPL and APS (Table 4.4). Fredi and colleagues studied 169 Italian patients with PAPS for polymorphisms in non-HLA genes that are associated with increased susceptibility for lupus, including IRF5, STAT4, and BLK [63] and found a strong genetic association with PAPS for STAT4 and BLK and a weak association for IRF5. In a similar study, 48 single nucleotide polymorphisms (SNP) from 40 candidate loci for lupus were typed in a cohort of 208 Italian and Hungarian SLE patients and 152 controls [65]. No associations between secondary APS and IRF5, STAT4, or BLK were found; weak associations that did not persist after multivariate analysis were found with NCF2 and TYRP1. In contrast, in separate studies, a STAT4 polymorphism was associated with primary as well as secondary APS [58], and a different STAT4 polymorphism was associated with aPL and ischemic cerebrovascular events in Swedish patients with lupus [59]. These conflicting data suggest that while genetic variants that increase the risk for the development of PAPS may overlap with those that increase the risk for the development of SLE, larger studies are needed to determine both those genes which truly contribute to heritable risk and why different disease phenotypes manifest.