Current Rheumatology Reports 2001 3: 277-285
Georgios N Dalekos MD, , Kalliopi Zachou MD, and Christos Liaskos MD
Larisa Medical School, 22 Papakiriazi str, University of Thessaly, Larisa, 412 22, Greece.
Infectious agents have been implicated in the induction of antiphospholipid (aPL) antibodies and the development of the antiphospholipid syndrome (APS). This review focuses on the types of aPL antibodies detected in infections and addresses whether these antibodies are of clinical importance in patients with infections. Hepatitis C virus (HCV) infection is given special attention because this virus has the propensity to induce various autoimmune phenomena. Several aspects are emphasized that should be considered carefully when interpreting results. Most of the published data agree that thrombophilia is not observed in patients with infections (including HCV) because aPL antibodies are mostly the natural or nonpathogenic type. Thus, we do not recommend routinely testing for HCV in patients with APS. However, not all infection-associated aPL antibodies are cofactor independent. For instance, infections are increasingly recognized as a major precipitating condition of the catastrophic variant of APS, perhaps via mechanisms of molecular mimicry. Therefore, it may be possible to prevent this devastating evolution if the infectious process is promptly recognized and exhaustively treated.
The antiphospholipid syndrome (APS) is an autoimmune disorder characterized by arterial or venous thrombosis, recurrent fetal losses, and usually moderate thrombo-cytopenia in association with the presence of antiphospholipid (aPL) antibodies . Neurologic and behavioral disturbances, livedo reticularis, and renal, liver, or valvular heart dysfunction have also been reported . In 1992, the term catastrophic APS was proposed by Asherson et al.  to describe an unusual variant of APS characterized by acute multi-organ failure including kidney damage, acute pulmonary manifestations, and a high incidence of central nervous system involvement, with histopathologic evidence of multiple microthrombosis. Consensus preliminary criteria for the diagnosis of APS have been reported recently . However, major thrombogenic conditions should be excluded before establishing the diagnosis of APS (1).
What stimuli induce APS? The catastrophic variant of APS may occur in patients susceptible to infections [2,4,5. Historically, aPL antibodies have been related to infections. The venereal disease research laboratory (VDRL) test, originally used to diagnose syphilis, has diagnosed it in some patients with APS (false-positive VDRL test) [1,6]. Since then, aPL antibodies have been reported to occur in several conditions other than APS, including infections, malignant diseases, inflammatory bowel disease, older age, drug use, and chronic mental illness [6,7,8,9,10].
This review focuses on the types of aPL antibodies detected in patients with various infections and addresses whether these autoantibodies are clinically important in such patients. Hepatitis C virus (HCV) infection is emphasized because it can induce several autoimmune phenomena [11,12,13].
The aPL antibodies are an heterogeneous family of immuno-globulins reactive with complexes of phospho-lipids and plasma proteins (cofactors) [1,3,6,14]. Anti-cardiolipin (aCL) antibodies and the lupus anticoagulant (LA) are the most widely studied and are considered the laboratory tests of choice for the diagnosis of APS, according to the recent Sapporo criteria .
It has been suggested that cofactors such as beta2-glycoprotein I (beta2-GPI), annexin V, protein S, and prothrombin must be present for the aPL antibodies to bind to phospholipids in patients with APS (cofactor dependent, autoimmune, or pathogenic aPL antibodies) [1,6,9]. Autoimmune aPL antibodies seem to be correlated with APS manifestations [15,16,17], whereas the non-pathogenic aPL antibodies detected in several conditions other than APS appear to bind directly to phospholipids (cofactor independent) [6,9,10,11].
Anticardiolipin antibodies are detected by solid-phase enzyme-linked immunosorbent assays (ELISA) [6,10,11]. The aCL test is sensitive but not specific; it may have positive results for several disorders other than APS 7,8,9,10, 11,14]. However, because patients with APS usually have higher levels of aCL antibodies, the diagnosis can be made more specific by using higher cut-off points [1,3,14]. In addition, new ELISA techniques have been developed using other antigens such as â2-GPI or a mixture of phospholipids [1,6,14]. These new tests are more specific than the aCL ELISA for the diagnosis of APS and seem to correlate well with thromboembolic events .
Compared with ELISA, assays for LA detection are considered more specific for the diagnosis of APS and are better correlated with the clinical manifestations of APS [1,3,14]. However, there are certain difficulties in clinical practice because thromboplastin reagents sensitive to LA are not widely available, LA assays are insufficiently sensitive, and the test results may be unreliable in patients on warfarin [1,3,14]. According to the Standardization Subcommittee on LA/phospholipid-dependent antibodies, LA is detected when at least one of the clotting times are prolonged, as shown in 2, and not corrected by the mixing procedure [1,3,14,18].
Detection of aPL Antibodies in Various Infection
Possible mechanisms for antiphospholipid antibody production in HCV infection
Numerous infectious agents such as HIV, HCV, parvovirus B19, spirochetes, mycoplasma species, cytomegalovirus (CMV), Epstein-Barr virus (EBV), Salmonella typhi, and other bacteria have been implicated in the induction of aPL antibodies and, in some cases, for the development of APS [4,5,6,9,11,13,14,15,16].
Antiphospholipid antibodies in HIV infection
HIV infection is associated with several abnormalities of B-cell function that result in the development of various autoimmune phenomena of unknown but pathogenic significance. Under this context, a high prevalence of aCL antibodies (range 12%-67%) has been reported in patients with HIV [9,19,20,21]. In contrast, anti-â2-GPI antibodies are rarely detected in patients with HIV, which suggests that aCL antibodies in such patients have characteristics of the natural type, not the pathogenic [9,14,19,20,21]. aCL antibodies in HIV-positive patients have been associated with the presence of cerebral perfusion defects and transient neurologic deficits, but the association with thrombosis was uncommon even in patients with high aCL titers . This corresponds with a recent report by de Larranaga et al.  who describe an unexpectedly high prevalence of LA (72%) without any clinical evidence of thrombophilia in 61 HIV-positive patients.
Antibodies against phospholipids other than cardiolipin such as phosphatidyl serine (aPS), phosphatidyl inositol (aPI), and phosphatidyl choline (aPC) have also been detected [9,19,22,23]. The aPS and aPC antibodies are the most commonly detected. Petrovas et al.  demonstrated that aPL reactivity does not correlate with disease duration or stage of HIV infection. They also showed that aPL antibodies were not associated with a history of Pneumocystis carinii pneumonia or coinfection with HCV . A significant decrease of aPL binding after urea and NaCl treatment was observed in the aPL-positive sera of patients infected with HIV compared with that of patients with APS, indicating that aPL antibodies from patients with HIV have low resistance to dissociating agents . The latter finding probably reflects low antibody avidity as described recently by Vlachoyiannopoulos et al. . According to this report, the high resistance of antibody/antigen binding to dissociating agents was associated with the occurrence of thrombotic events, offering additional data in identifying patients with APS . Thus, the low resistance to dissociating agents categorizes the HIV-related aPL antibodies as nonthrombogenic.
In conclusion, a high prevalence of different aPL antibodies has been reported in patients with HIV infection [9,14,19, 20,21], which suggests a B-cell polyclonal activation that is probably parallel with the activation against HIV-1-specific determinants. The aPL antibodies have low avidity, which could be attributed to a potential exposure of normally hidden phospholipids after the disruption of cell membranes during apoptosis. Recently, Silvestris et al. , by using fluorescence--activated cell-sorting analysis, have shown that in vitro apoptosis of T cells was increased in patients with high levels of aPS antibodies. Because phospholipids are exteriorized by apoptotic lymphocytes, their persistence may stimulate antibodies, which helps macrophages clear dead cells by an enhanced antibody-dependent cellular cytotoxicity. This interpretation could explain the absence of thrombophilia in HIV-positive patients with aPL antibodies. However, the increased oxidative stress observed in HIV infection may further explain the formation of aPL antibodies as a result of neoepitope formation by oxidized phospholipids.
Antiphospholipid antibodies in hepatitis C virus
Hepatitis C virus is a hepatotropic and lymphotropic virus that has been found to be associated with various diseases and syndromes known as extrahepatic manifestations of this chronic disease . HCV tends to induce nonspecific autoimmune reactions, as demonstrated by the high prevalence of various non-organ-specific autoantibodies, usually in low titers. The non-organ-specific antibodies include the antinuclear, smooth muscle, antineutrophil, and liver-kidney microsomal antibodies [11,12,20,26]. aCL antibodies have also been reported in patients with HCV [11,13,20,27,30,31,132,33].
Prevalence of aCL Antibodies in HCV Infection
The prevalence of these antibodies ranges from 3.3% to 46% (3). The high variation in aCL prevalence can be attributed to the laboratory methods used and to the design of the studies. For instance, it is important whether a commercial ELISA kit or a homemade, standardized, internationally accepted ELISA was used to determine aCL antibodies. In a multicenter study, an evaluation of nine commercial kits used to quantify aCL antibodies showed variations of 31% to 60% and 6% to 50% for the IgG and IgM aCL isotypes, respectively . The discrepancy of the results is more common in samples from patients with low levels of aCL antibodies, which is the case in most of the patients with aCL-positive HCV. Most studies agree that the mean levels of IgG and IgM isotypes of these antibodies are low to moderate in patients with HCV and significantly lower than those found in patients with APS [11,13,27,32,34,38].
Considering or ignoring the nonspecific binding of the antibodies in the ELISA plate can alter aCL results. This phenomenon is observed more frequently with the IgM aCL isotype that is found in a significant proportion of patients with HCV infection. This problem is minimized by subtracting the background values developed in the no-antigen wells (coated with ethanol without cardiolipin) in all assays or by adopting stringent cut-off points for aCL positivity (eg, mean of healthy plus 4 or 5 standard deviations [SD]). Furthermore, not all the studies clearly demonstrated the cofactor dependence of aCL antibodies. In this setting, the more straightforward assay for anti-â2-GPI antibodies is preferred because modified aCL assays (using albumin- or gelatin-based solutions) may have several inherent problems, including the contribution to aCL binding by endogenous â2-GPI in patient serum.
Etiopathogenetic Relationship Between HCV and APS
Data indicate that there is no relationship between HCV and APS (3). Most studies reported no clinical manifestations of APS in aCL- and HCV-positive patients before or during the follow-up visit [11,13·,20··,27,30·,31, 32·,34,35,38]. In addition, in most of the studies, the aCL antibodies detected in patients with HCV infection were of the nonthrombogenic type, which means they were co-factor independent [20··,27,30·,32·,34-36,38]. This finding, along with the absence of LA activity, justifies the lack of clinical manifestations of APS in HCV-infected patients [13·,27,32·,36].
Prieto et al.  found that aCL positivity in 100 patients with HCV infection was significantly associated with cirrhosis-related portal hypertension, thrombo-cytopenia, and prior thrombotic episodes, whereas occult HCV infection was present in six of 36 patients (16.7%) with thrombotic disorders and aCL antibodies (primary APS). Selection bias of the patients most likely accounts for part of this discrepancy. In this study, 20 of 80 patients (25%) who had undergone a liver biopsy had cirrhosis. Cirrhosis, whatever its cause, is associated frequently with the presence of nonpathogenic aCL antibodies [27,29]. However, the finding of HCV markers in some (six of 36) of the patients with primary APS is puzzling and deserves comment. Three of these six patients had low titer aCL (eg, 40 GPL, the cut-off being at 23 GPL units) . As noted earlier, low titer aCL antibodies may represent a transient epiphenomenon and are not accurately determined in the poorly standardized assays . In addition, LA activity and nonspecific binding in aCL ELISA and the cofactor dependence of aCL antibodies were undetermined . However, in the study of Prieto et al. , the presence of HCV in some patients with APS may be related to transmission routes in the course of APS because repetitive in-hospital therapies may be needed in this syndrome. Recently, Baid et al.  reported a significant association of aCL antibodies with renal thrombotic microangiopathy in a few HCV-positive recipients of renal allograft. Again, there is no information regarding the cofactor dependence of aCL antibodies. In addition, the cut-off value for aCL positivity was low (mean 2 SD) .
We prospectively investigated the possible etiopatho-genetic link between APS and HCV infection in a bidirectional model, investigating aCL antibodies and aCL-related manifestations in patients infected with HCV and vice versa. We also investigated by ELISA and polymerase chain reaction (PCR) the presence of HCV markers in patients with well-defined APS. We were unable to reveal a trend for such an etiopathogenetic association in this study . A long-term follow-up of aCL-positive, anti-â2-GPI-negative patients with HCV infection is under way in our center (mean follow-up until now is 26 months). Thrombotic episodes not statistically significant have been observed in these patients compared with those who had negative test results for aCL antibodies (Zachou K and Dalekos GN, unpublished data). A recent study by Mangia et al.  conducted in 372 patients with several chronic liver diseases reported that aCL antibodies were significantly more frequent in patients with hepatitis B virus infection or alcoholic liver disease than those with HCV. Munoz-Rodriguez et al.  showed that the prevalence of HCV in 88 patients with APS was low (2.2%) and similar to that observed in healthy individuals of the same area.
Mechanisms for aCL Production in HCV
The mechanisms by which HCV infection induces aCL antibodies has not been elucidated. Several possibilities can be proposed (4). It is possible that helper T cells reactive to viral antigens that are exposed on the surface of infected hepatocytes assist the autoreactivity of B cells, leading to chronic B-cell stimulation. This theory seems reasonable because several articles have shown the presence of low titers of at least one non-organ-specific autoantibody in most patients with HCV [11,20]. However, Cacoub et al. [20,38] and Sthoeger et al.  demonstrated that polyclonal B-cell activation by the HCV does not appear to be related to the high prevalence of aCL antibodies. Evidence supporting this include: 1) there was no close relationship between the IgG aCL antibody titer and the total IgG levels , 2) anti-â2-GPI and anti-nucleosome antibodies were not detected in 46 patients with HCV , and 3) only some non-organ-specific autoantibodies in a panel of 17 auto-antibodies tested were found positive among 321 patients with HCV .
Another possible mechanism by which HCV infection induces aCL antibodies is HCV-induced endothe-lial and hepatic damage [13,27,30,35,38]. Hepatic damage, along with hepatocellular apoptosis, may lead to the expression of cell surface phospholipids and the induction of proinflammatory cytokines, which can generate more aCL antibodies. aCL antibodies are detected more frequently in patients with mixed cryoglobulinemia, whatever their HCV status, suggesting that endothelial damage induced by cryoglobulins may have a role in the production of aCL antibodies . Sthoeger et al.  have shown the presence of cardiolipin reactivity in some cryoglobulins obtained from aCL- and HCV--positive patients, suggesting that aCL antibodies may react with HCV epitopes that can be localized in the cryoprecipitates.
Interferon-á(IFN-á) therapy in patients with chronic HCV infection has been implicated in the induction of non-organ-specific autoantibodies [12,13,20,27,31]. However, in most studies, the presence or the induction of these antibodies did not influence the response rate of chronic hepatitis C to IFN-ánor has true APS been reported in such patients.
Molecular mimicry and several associations of self--antigens with HCV proteins have been considered as possible triggering mechanisms for autoantibody formation [12,13,26]. However, to the best of our knowledge, neither molecular mimicry nor an association of cardiolipin autoantigen with HCV proteins has been established . Ordi-Ros et al.  purified aCL antibodies by affinity chromatography and then the antibodies probed in a third generation recombinant immunoblot assay (RIBA 3.0) to determine their cross-reaction with HCV proteins. No reactivity was demonstrated with any of the HCV antigens.
Two recent studies suggest that HIV, not HCV, plays an important role in the development of aCL antibodies [20,39]. Cacoub et al.  reported that the prevalence of aCL antibodies in patients who were HCV- and HIV-positive was significantly higher (42%) than that found in patients who were HCV-positive and HIV--negative (21%), suggesting a synergistic effect of dual virus infection in the induction of aCL antibodies. Furthermore, after multivariate logistic regression analysis they found that only aCL positivity, thrombocytopenia, and arthralgia were significantly associated with HIV positivity in patients already infected with HCV . Clinically, these findings may be important to identify HIV infection among patients with HCV who have one or more of these parameters.
Taking into account the results of several studies, the prevalence of aCL antibodies in patients with HCV infection is found at a higher proportion than that detected in the control groups. In most cases, the titers of these antibodies are low, co-factor independent, and unrelated to APS manifestations. However, two recent independent studies [13,41] did not indicate that HCV infection markers occur significantly more frequently among patients with APS. A positive correlation between HCV and APS seems unlikely; therefore, we do not recommend testing for HCV in patients with APS. We also do not recommend strict follow-up for the possibility of APS development in patients with HCV infection. However, multicenter studies with long-term follow-up (10 years) of aCL- and HCV-positive patients may be helpful to address whether these autoantibodies are of clinical importance. The latter suggestion seems to be supported by a recent report by Shah et al. . These investigators observed that during a 10-year follow-up period, 50% of their patients (11/21) with connective tissue diseases developed the syndrome, although initially they had increased levels of aCL anti-bodies without clinical manifestations of APS .
Antiphospholipid antibodies in other infection
Increased prevalence of aPL antibodies in several bacterial, parasitic, and viral infections other than HIV and HCV have been reported [6,14,15, 16,21,43]. In most of these studies, aPL antibodies were not associated with the clinical features of APS, but they were cofactor independent [21,43].
Recently, McNally et al.  reported that patients with infections (11 with active syphilis, 63 with active tuberculosis, and 42 with urinary tract infection caused by Klebsiella; n 114) were observed for the presence of aCL and anti-â2-GPI antibodies. aCL antibodies were detected in 6%, 5%, and 64% of patients with tuberculosis, Klebsiella infection, and syphilis, respectively. However, none of the patients had apparent thrombotic complications or tested positive for anti-â2-GPI antibodies; one had IgG aCL antibodies more than 31 GPL unitsmL . Three other studies report similar findings [21,44,45].
Sorice et al.  conducted a study to clarify the antigen specificity of aPL antibodies in patients with infectious mononucleosis (IM). They found that 30% of patients with IM were positive for pure aCL antibodies. However, these antibodies were often present with anti-cofactor protein antibodies. However, anti-â2-GPI antibodies and anti-prothrombin antibodies had a low prevalence rate (0% and 6.5%, respectively) compared with that detected in patients with APS (33.3% and 33.3%, respectively; P= 0.002 and P 0.05, respectively). The authors also found antibodies against annexin V and protein S cofactors in approximately 25% of the patients with IM . The occurrence of these antibodies in patients with IM did not differ significantly from that found in patients with APS, enhancing the concept that anti-â2-GPI and antipro-thrombin antibodies represent specific and independent factors associated with the clinical manifestations of APS. aCL and anti-cofactor protein antibodies disappeared after 12 to 15 months of EBV infection .
Within the last year, a high prevalence (56%) of aPL antibodies was reported in CMV infection among patients who received unrelated bone marrow and cord blood allogenic stem cell transplantation for hematologic malignancies . No association was observed between these antibodies and clinical expression of autoimmunity . However, not all infection-associated aCL antibodies are â2-GPI independent, as shown recently in a small series of 12 patients with acute parvovirus B19 infection . Furthermore, two cases of APS associated with CMV infection and another one with Mycoplasma penetrans in an anti-HIV-negative patient were reported recently [49-51].
Catastrophic APS and infection
Several triggering factors have been involved in the pathogenesis of the catastrophic variant of APS. Recent reports suggest that infections, whether clinical or subclinical, are increasingly recognized as a major precipitating condition [2,4,5]. Bacterial infections have been associated with the development of catastrophic APS in six of eight patients described by Rojas-Rodriguez et al. , and in three of 11 patients (27.3%) with identified precipitating factors in a recent series . In addition, S. typhi was recently reported to be the triggering factor for another case of catastrophic APS . With S. typhi, the syndrome occurred 3 weeks after the onset of typhoid fever, which was successfully treated with ofloxacin. This patient responded dramatically to plasmapheresis . The authors suggest that antigens of S. typhi, such as lipopolysaccharides, may have immunologic and prothrombotic effects. Rojas-Rodriguez et al.  identified Staphylococcus aureus, Streptococcus pyogenes, Escherichia coli, and M. penetrans. Most of these bacteria may act as superantigens, leading to a disproportionate immune response. Therefore, it may be possible to prevent this devastating evolution if the infectious process is promptly recognized and exhaustively treated.
With the use of several monoclonal anti-â2-GPI antibodies and the phage display technique of a hexapeptide library, three peptides were identified. These peptides were recognized by aCL antibodies of patients with APS and were shown to prevent experimental APS induced by active immunizations with â2-GPI [4,15,16,17]. However, similar sequences of the three peptides were found on membranes of diverse viruses, bacteria, and parasites. Thus, it seems reasonable that by molecular mimicry, some of the infectious agents may induce nonpathogenic aPL antibodies and pathogenic anti-â2-GPI antibodies. Gharavi et al. [15,16,17] have added further support to this notion. They identified seven proteins with sequence homology to GDKV and GDKV2 proteins that are the major phospholipid binding site on â2-GPI. The peptide sequences are found in many human viruses to which people are commonly exposed. Recently, Blank et al.  demonstrated that ciprofloxacin, which has the ability to induce interleukin-3 production, can immunomodulate and benefit mice induced with the experimental APS.
According to the above mentioned, infections could be an etiology of the catastrophic variant of APS, especially in persons who already have other components of the mosaic autoimmunity, such as the genetic predisposition, immune system defects, or abnormalities of the hormonal milieu.
The aPL antibodies, with their protein cofactors, are essential for the diagnosis of APS. However, aPL antibodies of different kinds are detected in infections, although usually in lower titers. aPL antibodies found in APS can be differentiated from those found in infections by determining LA activity and anti-â2-GPI antibodies. LA activity and anti-â2-GPI are present in APS, correlating with thrombotic events, but not in most cases of infectious diseases (nonthrombogenic aPL).
It may be possible that not all of the nonthrombogenic aPL antibodies are innocent. These types of antibodies may be switched on by a second hit to the autoimmune type through several mechanisms. Alternatively, based on the clinical experience indicating that not all aPL antibodies are pathogenic, probably only a few aPL antibodies induced by certain viral or bacterial products would be pathogenic in genetically predisposed individuals. The exact mechanisms by which infectious agents induce the development of aPL antibodies remains to be ascertained.
To identify an aPL antibody, several aspects such as laboratory criteria, study design (retrospective versus prospective), and the types of analyses of the studies should be considered carefully and need to be emphasized before general conclusions are drawn. It is essential for every labora-tory to evaluate many normal people to rigorously define the normal ranges of these assays and the frequency of outlying values. In particular, the levels of aCL anti-bodies are not normally distributed in healthy persons. This distribution is best analyzed by the percentile method. Unfortunately, many of the studies mentioned in this review expressed their results in terms of mean SD, which is an unsuitable form of analysis of this type of distribution. If this expression is used, a stringent cut-off point should be selected to preclude false-positive results.
As noted earlier, the discrepancies in the prevalence of aCL antibodies detected in patients with infections may occur because of the methods used. Thus, one should consider whether the ELISA method used is a commercial kit or a homemade standardized ELISA, what solutions have been used, and whether the nonspecific binding in the ELISA plate has been considered. Even in such conditions, the adoption of stringent cut-off points may minimize the problem.
Most of the studies that reported the association of infections with the presence of aPL antibodies or the whole clinical spectrum of APS are retrospective or are case reports. Few studies have prospectively evaluated the patients in a long-term follow-up study. This could be an attractive strategy, particularly in chronic viral infections.
Papers of particular interest have been highlighted as:
• of special interest
•• of outstanding interest
1. MA Khamashta, GRV Hughes: Antiphospholipid antibodies and antiphospholipid
Curr Opin Rheumatol 1995, 7: 389-394
2. RA Asherson, R Cervera, JC Piette: The catastrophic antiphospholipid
syndrome: clinical and laboratory features of 50 patients.
Medicine 1998, 77: 195-207
• This paper summarizes the clinical and laboratory characteristics of 50 patients with catastrophic APS. Infections were recorded as the precipitating factor contributed to the development of catastrophic APS in three of 11 patients (27.2%) with identified factors.
3. WA Wilson, AE Gharavi, T Koike: International consensus statement
on preliminary classification criteria for definite antiphospholipid syndrome.
Arthritis Rheum 1999, 42: 1309-1311
4. R Asherson, Y Shoenfeld: The role of infection in the pathogenesis
of catastrophic antiphospholipid syndrome: molecular mimicry?
J Rheumatol 2000, 27: 12-14
5. J Rojas-Rodriguez, MG Carrasco, MR Casals: Catastrophic antiphospholipid
syndrome: clinical description and triggering factors in 8 patients.
J Rheumatol 2000, 27: 238-240
• The authors described eight patients with catastrophic APS, focusing on the possible extrinsic factors that may trigger this condition. Infections of the skin and the respiratory, urinary, and gastrointestinal tract were recorded in 75% of patients, suggesting that prevention of this devastating form of APS may be possible if the infectious process is promptly recognized.
6. HP McNeil, SA Krilis: Antiphospholipid antibodies.
Aust NZ J Med 1991, 21: 463-475
7. MN Manoussakis, AG Tzioufas, MP Silis: High prevalence of anti-cardiolipin
and other autoantibodies in a healthy elderly population.
Clin Exp Immunol 1987, 69: 557-565
8. SG Yannitsi, MN Manoussakis, AK Mavridis: Factors related to the
presence of anticardiolipin and other auto-antibodies in patients with
chronic mental disorders.
Biol Psychiatr 1990, 27: 747-756
9. C Petrovas, PG Vlachoyiannopoulos, T Kordossis: Anti-phospholipid
antibodies in HIV infection and SLE with or without anti-phospholipid syndrome:
comparisons of phospholipid specificity, avidity and reactivity with b2-GPI.
J Autoimmun 1999, 13: 347-355
•• This study demonstrates that HIV infection is associated with a high prevalence of antibodies against various phospholipids but not b2-GPI, irrespective of the disease stage, a history of Pneumocystis carinii infection, or co-infection with HCV. The aPL antibodies had low resistance to dissociating agents, a finding that probably reflects low antibody avidity. These results suggest that HIV infection-related aPL antibodies are nonthrombogenic.
10. GN Dalekos, MN Manoussakis, AC Goussia: Soluble interleukin-2 receptors,
antineutrophil cytoplasmic antibodies, and other autoantibodies in patients
with ulcerative colitis.
Gut 1993, 34: 658-664
11. GN Dalekos, MN Manoussakis, E Zervou: Immunologic and viral markers
in the circulation of anti-HIV negative addicts.
Eur J Clin Invest 1993, 23: 219-225
12. GN Dalekos, H Wedemeyer, P Obermayer-Straub: Epitope mapping of
cytochrome P450 2D6 autoantigen in patients with chronic hepatitis C during
J Hepatol 1999, 30: 366-375
13. GN Dalekos, K Kistis, D Boumba: Increased incidence of anticardiolipin
antibodies in patients with hepatitis C is not associated with aetiopathogenetic
link to antiphospholipid syndrome.
Eur J Gastroenterol Hepatol 2000, 12: 67-74
• The authors prospectively investigated the possible etiopathogenetic link between APS and HCV infection in a two-directional model, which means investigation of aCL antibodies and aCL-related manifestations in HCV-infected patients and vice versa, the presence of HCV markers by ELISA, and polymerase chain reaction in patients with well-defined APS. They did not reveal a trend for such an etiopathogenetic association.
14. EN Harris, SS Pierangeli, AE Gharavi: Diagnosis of the antiphospholipid
syndrome: a proposal for use of laboratory tests.
Lupus 1998, 7: 144-148
15. AE Gharavi, SS Pierangeli, EN Harris: New developments in viral
peptides and APL induction.
J Autoimmun 2000, 15: 227-230
•• This study shows the pathogenic effects of aPL antibodies induced by immunization with a human CMV-derived aPL-binding synthetic peptide, suggesting that aPL antibodies in patients with APS may be induced by b2-GPI-like Pl-binding products of common human bacteria or viruses.
16. AE Gharavi, SS Pierangeli: Origin of antiphospholipid antibodies:
induction of aPL by viral peptides.
Lupus 1998, 7: S52-S54
17. AE Gharavi, SS Pierangeli, M Colden-Stanfield: GDKV-induced antiphospholipid
antibodies enhance thrombosis and activate endothelial cells in vivo and
J Immunol 1999, 163: 2922-2927
18. JT Brandt, DA Triplett, B Alving: Criteria for the diag-nosis of
lupus anticoagulants: an update.
Thromb Haemostat 1995, 74: 1185-1190
19. F Silvestris, MA Frassanito, P Cafforio: Antiphospha-tidylserine
antibodies in human immunodeficiency virus-1 patients with evidence of
T-cell apoptosis and mediate antibody-dependent cellular cytotoxicity.
Blood 1996, 87: 5185-5195
20. P Cacoub, C Renou, E Rosenthal: Extrahepatic manifestations associated
with hepatitis C virus infection.
Medicine 2000, 79: 47-56
•• The prevalence of aCL antibodies in HCV-positive/HIV-positive patients was significantly higher than that found in HCV-positive/HIV-negative patients, suggesting a synergistic effect of dual virus infection in the induction of aCL antibodies. After multivariate logistic regression analysis, the authors found that only aCL positivity, thrombocytopenia, and arthralgia were significantly associated with HIV positivity in patients infected with HCV.
21. GF de Larranaga, RR Forastiero, LO Carreras: Different types of
antiphospholipid antibodies in AIDS: comparison with syphilis and the antiphospholipid
Thromb Res 1999, 96: 19-25
22. C Gonzalez, A Leston, B Garcia-Berrocal: Antiphospha-tidylserine
antibodies in patients with autoimmune disease and HIV-infected patients:
effects of Tween 20 and relationship with antibodies to beta2-glycoprotein
J Clin Lab Anal 1999, 13: 59-64
23. N Abuaf, S Laperche, B Rajoely: Autoantibodies to phospholipids
and to the coagulation protein in AIDS.
Thromb Haemostat 1997, 77: 856-861
24. BJ Brew, J Miller: Human immunodeficiency virus type 1-related transient
Am J Med 1996, 101: 257-261
25. P Vlachoyiannopoulos, C Petrovas, M Tektonidou: Antibodies to b2-glycoprotein-I:
urea resistance, binding specificity, and association with thrombosis.
J Clin Immunol 1998, 18: 380-391
•• This study evaluated whether the urea resistance of anti-b2-GPI antibodies is an additional parameter associated with APS. Their results support the notion that resistance to dissociating agents such as urea can be used as a serum characteristic in addition to antibody titer to identify patients with APS.
26. GN Dalekos, P Obermayer-Straub, T Maeda: Antibodies against cytochrome
P4502A6 (CYP2A6) in patients with chronic viral hepatitis (CVH) are mainly
linked to hepatitis C virus (HCV) infection.
Digestion 1998, 59: 36-37
27. V Leroy, J Arvieux, MC Jacob: Prevalence and significance of anticardiolipin,
anti-b2-glycoprotein I and anti-prothrombin antibodies in chronic hepatitis
Br J Haematol 1998, 101: 468-474
28. J Prieto, JR Yuste, O Beloqui: Anticardiolipin antibodies in chronic
hepatitis C: implication of hepatitis C virus as the cause of the antiphospholipid
Hepatology 1996, 23: 199-204
29. C Biron, H Andreani, P Blanc: Prevalence of antiphospholipid antibodies
in patients with chronic liver disease related to alcohol or hepatitis
C virus: correlation with liver injury.
J Lab Clin Med 1998, 131: 243-250
30. ZM Sthoeger, M Fogel, A Smirov: Anticardiolipin auto-antibodies
in serum samples and cryoglobulins of patients with chronic hepatitis C
Ann Rheum Dis 2000, 59: 483-486
• This study showed a high prevalence of IgG aCL antibodies in HCV patients but with no b2-GPI dependence. Furthermore, it demonstrated the localization of aCL antibodies in some cryoprecipitates, suggesting reactivity of these antibodies with HCV epitopes, which can be found in the cryoprecipitates.
31. J Matsuda, N Saitoh, M Gotoh: High prevalence of anti-phospholipid
antibodies and anti-thyroglobulin antibody in patients with hepatitis C
virus infection treated with interferon-a.
Am J Gastroenterol 1995, 90: 1138-1141
32. J Ordi-Ros, J Villarreal, F Monegal: Anticardiolipin antibodies
in patients with chronic hepatitis C virus infection: characterization
in relation to antiphospholipid syndrome.
Clin Diagn Lab Immunol 2000, 7: 241-244
• This study reported a low prevalence of aCL antibodies in 243 patients with HCV (3.3%). In addition, no cross-reactivity between aCL antibodies and HCV antigens was found, suggesting that aCL antibodies in HCV infection are an epiphenomenon without any clinical significance.
33. S Baid, M Pascual, WW Williams: Renal thrombotic microangiopathy
associated with anticardiolipin antibodies in hepatitis C-positive renal
J Am Soc Nephrol 1999, 10: 146-153
34. M Harada, Y Fujisawa, S Sakisaka: High prevalence of anticardiolipin
antibodies in hepatitis C virus infection: lack of effects on thrombocytopenia
and thrombotic complications.
J Gastroenterol 2000, 35: 272-277
35. A Mangia, M Margaglione, I Cascavilla: Anticardiolipin antibodies
in patients with liver disease.
Am J Gastroenterol 1999, 94: 2983-2987
36. P Giordano, M Galli, GC Del Vecchio: Lupus anticoagulant, anticardiolipin
antibodies and hepatitis C virus infection in thalassaemia.
Br J Haematol 1998, 102: 903-906
37. A Al-Saeed, M Makris, RG Malia: The development of antiphospholipid
antibodies in haemophilia is linked to infection with hepatitis C.
Br J Haematol 1994, 88: 845-848
38. P Cacoub, L Musset, Z Amoura: Anticardiolipin, anti-b2-glycoprotein
I, and antinucleosome antibodies in hepatitis C virus infection and mixed
J Rheumatol 1997, 24: 139-144
39. M Gotoh, J Matsuda: Human immunodeficiency virus rather than hepatitis
C virus infection is relevant to the development of an anti-cardiolipin
Am J Hematol 1995, 50: 220-222
40. G Reber, J Arvieux, E Comby: Multicenter evaluation of nine commercial
kits for quantification of anticardiolipin antibodies. The Working Group
on Methodologies in Haemostasis from the GEHT (Group d'Etudes sur l'Hemostase
et la Thrombose).
Thromb Haemost 1995, 73: 444-452
41. F Munoz-Rodriguez, D Tassies, J Font: Prevalence of hepatitis C
virus infection in patients with antiphospholipid syndrome.
J Hepatol 1999, 30: 770-773
• In this study, the prevalence of HCV infection markers in 88 patients with APS was low and similar to that of healthy individuals, suggesting that HCV does not seem to be involved in the pathogenesis of this syndrome.
42. NM Shah, MA Khamashta, T Atsumi: Outcome of patients with anticardiolipin
antibodies: a 10 year follow-up of 52 patients.
Lupus 1998, 7: 3-6
43. T McNally, G Purdy, IJ Mackie: The use of an anti-b2--glycoprotein-I
assay for discrimination between anticardiolipin antibodies associated
with infection and increased risk of thrombosis.
Br J Haematol 1995, 91: 471-473
44. JE Hunt, HP McNeil, GJ Morgan: A phospholipid-beta 2-glycoprotein
I complex is an antigen for anticardiolipin antibodies occurring in autoimmune
disease but not with infection.
Lupus 1992, 1: 75-81
45. RR Forastiero, ME Martinuzzo, LC Kordich: Reactivity to beta 2 glycoprotein
I clearly differentiates anticardiolipin antibodies from antiphospholipid
syndrome and syphilis.
Thromb Haemost 1996, 75: 717-720
46. M Sorice, V Pittoni, T Griggi: Specificity of anti-phospholipid
antibodies in infectious mononucleosis: a role for anti-cofactor protein
Clin Exp Immunol 2000, 120: 301-306
47. A Mengarelli, C Minotti, G Palumbo: High levels of antiphospholipid
antibodies are associated with cytomegal-ovirus infection in unrelated
bone marrow and cord blood allogeneic stem cell transplantation.
Br J Haematol 2000, 108: 126-131
48. S Louizou, JK Cazabon, MJ Walport: Similarities of specificity and
cofactor dependence in serum antiphospholipid antibodies from patients
with human parvovirus B19 infection and from those with systemic lupus
Arthritis Rheum 1997, 40: 103-108
49. I Uthman, Z Tabbarah, AE Gharavi: Hughes syndrome associated with
Lupus 1999, 8: 775-777
50. JA Labarca, M Rabaggliati, FJ Radrigan: Antiphospholipid syndrome
associated with cytomegalovirus infection: case report and review.
Clin Infect Dis 1997, 24: 197-200
51. A Yapez, L Cedillo, O Neyrolles: Mycoplasma penetrans bacteremia
and primary antiphospholipid syndrome.
Emerg Infect Dis 1999, 5: 164-167
52. G Hayem, N Kassis, P Nicaise: Systemic lupus erythematosus-associated
catastrophic antiphospholipid syndrome occurring after typhoid fever: a
possible role of Salmonella lipopolysaccharide in the occurrence of diffuse
Arthritis Rheum 1999, 42: 1056-1061
53. M Blank, J George, P Fishman: Ciprofloxacin immunomodulation of
experimental antiphospholipid syndrome associated with elevation of interleukin-3
and granulocyte-macrophage colony-stimulating factor expression.
Arthritis Rheum 1998, 41: 224-232