Antiphospholipid antibodies (aPLs) are a heterogeneous group of autoantibodies or alloantibodies with an affinity for anionic phospholipids [1]. aPLs occur in patients with antiphospholipid syndrome (APS), systemic lupus erythematosus (SLE), and various other rheumatic diseases. Furthermore, an elevated level of aPLs is recognized as a risk factor for thrombosis [2]. Several studies have suggested that the pathogenic mechanisms of aPLs are related to platelet activation, endothelial cell activation, and activation of the complement cascade [3, 4]; however, their pathogenic functions in patients with autoimmune diseases like SLE and primary APS are not fully understood.

aPLs have also been detected in numerous infectious diseases, including those caused by parvovirus B19, cytomegalovirus, varicella-zoster virus, human immunodeficiency virus (HIV), hepatitis B virus (HBV), hepatitis C virus (HCV), Helicobacter pylori, streptococci, and staphylococci [5-7]. The reported prevalence of aPLs in infected patients varies across studies [5, 8]. The variation in prevalence between studies may be partly due to differences in infection types, antibody types, and aPL measurement methods. For example, in HIV patients, a high prevalence of anticardiolipin antibodies (aCL Ab) (46.5%) [8] and lupus anticoagulant (LA) (43%) [9] was reported. In contrast, anti-β₂-glycoprotein I antibodies (β₂GPI Ab) were rarely detected in HIV patients [5].

The clinical significance of aPLs associated with various infections is controversial. In many studies, the presence of aPLs associated with infections has been regarded as non-pathogenic [5, 8]; however, in patients with various infections, thrombotic manifestations such as portal vein thrombosis and pulmonary embolism have been reported [10-13].

Recently, several studies showed a higher prevalence of aCL Ab in chronic viral hepatitis patients than in control individuals [14-17]. In the present study, we investigated the prevalence, persistence, clinical significance, and characteristics of aPLs in HBV-infected patients.

The total prevalence of aPLs (aCL Ab, β₂GPI Ab, or LA) in HBV-infected patients was 12.6% (18 of 143) (Table 1). Among these 18 patients, aPLs were detected in 15 of 97 chronic hepatitis B patients and 3 of 46 inactive HBsAg carrier patients. No clinical manifestations related to aPLs were identified. Among the 143 HBV-infected patients, 15 patients (10.5%) had a low to moderate level of aCL Ab (10 with IgM, 4 with IgG, and 1 with both isotypes) compared to only 1 of the 32 control subjects (3.1%). The difference between the groups was not significant (P=0.19). The median values for aCL Ab IgM and IgG isotypes were 28.1 MPL (range, 23.4-42.2) and 32.9 GPL (range, 20.5-47.2), respectively (Table 2 and Fig. 1).

The prevalence of β₂GPI Ab and the LA activity in the HBV-infected patient group was 2.1% (3 of 143) and 1.4% (2 of 143), respectively. In contrast, none of the healthy controls had elevated levels of β₂GPI Ab or LA; however, the difference between the groups was not significant (P=0.40). The isotype distribution for β₂GPI Ab in HBV-infected patients was 1 with IgM and 2 with IgG. Two HBV-infected patients simultaneously had 2 types of aPLs; 1 patient had IgM aCL Ab with cofactor dependency (β₂GPI Ab, IgG type), and the other had β₂GPI Ab (IgM type) and LA.

Follow-up specimens were obtained from 14 of the 17 patients with elevated levels of aCL Ab or β₂GPI Ab (11 with aCL Ab, 2 with anti-β₂GPI Ab, and 1 with both aCL Ab and β₂GPI Ab). The median follow-up duration was 30 weeks (range, 9-100 weeks). The aPLs persisted in 10 of the 14 patients (71.4%), and β₂GPI Ab persisted in all 3 patients with β₂GPI Ab (Table 3).

aPLs have been found not only in patients with autoimmune diseases like SLE, but also in patients with various infections [5, 8]. Molecular mimicry is the proposed mechanism for the development of aPLs in infections. Previous reports showed that antigenic determinants shared between the antigens of infectious agents and host tissue might trigger the immune response [20, 21]. However, the mechanisms that lead to the development of aPLs and their possible pathophysiological implications in patients with infections have not been well established.

The reported prevalence of aPLs in infectious patients is variable, in part because of a methodological problem. Currently, the ideal approach for standardized measurement for aCL Ab is under debate [22]. A discrepancy is observed among the different assay kits and methods, particularly in the lower range of antibody levels. Furthermore, one of the critical problems in the standardization process is the absence of a defined cut-off value for positivity. In previous reports, the prevalence of aCL Ab in patients with HBV infection was between 14% and 42% [14, 17, 23], and that of β₂GPI Ab was 2% [17] and 7.5% [14]. In the present study, the prevalence of aCL Ab and β₂GPI Ab was 10.5% and 2.1%, respectively, which was lower than that in previous reports. The lower prevalence of aCL Ab in the present study may be partly because of a higher cut-off value for positivity than that of previous reports. We used 20 GPL or MPL as the cut-off values, whereas 10 GPL or MPL were used as the cut-off values in the other reports [14, 23]. The prevalence of β₂GPI Ab in the present study was similar to those in previous reports.

In most cases, infection-associated aPLs appear temporarily and disappear within 2 or 3 months [24, 25], and except in rare cases, they are unrelated to thrombotic complications. Contrary to previous reports, in the present study, most aCL Ab and β₂GPI Ab showed persistence and were present over a 12-week period. We speculated that continuous antigenic stimulation might be associated with persistent aPL production, because in the present study, the prevalence of aCL Ab in chronic hepatitis B patients was higher than that in inactive HBsAg carrier patients (15% vs. 6.5%). Furthermore, it was shown that in a patient with HCV infection and aPLs, elimination of the virus was accompanied by the disappearance of aPLs, and when HCV infection relapsed, aPLs also reappeared [26]. Therefore, aPLs associated with other chronic infections caused by HIV or HCV may also show persistent positivity.

According to a previous report [27], IgM aPLs are associated with the clinical manifestations of APS less often than the IgG isotype is, and the diagnostic criteria for APS include only a medium or high titer of aCL Ab (>40 GPL or MPL) [28]. In the present study, IgM aCL Ab were found more frequently (11 of 15) than the IgG isotype (5 of 15), and the titers of aCL Ab were mostly low. These findings support that infection-associated aPLs rarely manifest with the clinical features of APS. Because the rates of aCL Ab according to isotype were not provided in previous reports of aPLs in patients with HBV infection, comparison with our results was not possible.

In conclusion, the most frequently detected aPLs in HBV-infected patients were IgM aCL Ab, which has a weak association with the clinical manifestations of APS. Unlike the transient presence of other infection-associated aPLs, most aPLs were persistently detected over a 12-week period in patients with HBV infection.

Distribution of anticardiolipin antibody titers in hepatitis B virus (HBV)-infected patients and normal controls, IgM isotype (A) and IgG isotype (B).

Table 1 Prevalence of antiphospholipid antibodies (aCL Ab, β₂GPI Ab, and lupus anticoagulant) in HBV-infected patients.

Table 2 Characteristics of HBV-infected patients with antiphospholipid antibodies and the types of antiphospholipid antibodies.

Table 3 Follow-up duration and results of aCL Ab and β₂GPI Ab testing in antiphospholipid antibody-positive HBV-infected patients.