Predictive Value of Oxidized Low-Density Lipoprotein/β 2 -Glycoprotein-I Complexes (oxLDL/β 2 GPI) in Nonautoimmune Atherothrombosis

Abstract

Introduction:

Lipid oxidation is a definite feature of atherosclerosis, and oxidized low-density lipoprotein (oxLDL) is not only highly immunogenic but toxic to several cell types. Beta-2-glycoprotein-I (β₂GPI) dampens oxLDL toxicity by forming binary oxLDL/β₂GPI complexes. We evaluated whether circulating oxLDL/β₂GPI complexes are associated to atherosclerosis-related events (ARE) and to venous thromboembolism (VTE).

Methods:

In a cross-sectional case–control study, cases were (a) 57 consecutive patients (male/female [M/F] 33/24, mean age 57 [10] years) attending a thrombosis unit for ARE (myocardial infarction [MI] n = 20, peripheral vascular disease n = 7, and ischemic strokes n = 30); (b) 52 consecutive patients (M/F 22/30, mean age 55 [17] years) attending the same unit for unprovoked (VTE); (c) normal controls comprised 90 participants (M/F 35/55, mean age 41 [15] years); and (d) oxLDL/β₂GPI complexes were measured by immunoassay and resulting levels divided into quartiles.

Results:

The odds ratio (OR) of ARE was greater in the fourth and second quartiles than in the first quartile (8.5 and 6.0, respectively); the OR of developing MI was greatest in the fourth quartile (17.8). By multivariable analysis with age, sex, smoking, lipid status, statin, and ARE phenotypes as independent variables and oxLDL/β₂GPI as the dependent variable, only MI predicted oxLDL/β₂GPI (P < .0001).

Conclusions:

OxLDL/β₂GPI may be regarded as a marker of ARE, in particular of MI.

Keywords

thrombosis atherosclerosis

Introduction

The notion that oxidized low-density lipoprotein (oxLDL) plays a pathogenic role in early atherosclerosis¹ was followed by the discovery that oxLDL binds in vitro to beta-2-glycoprotein-I (β₂GPI) forming stable (nondissociable) but similarly pathogenic oxLDL/β₂GPI binary complexes²; indeed, these complexes primarily formed in atherosclerotic lesions are released into the circulation where can be measured reliably. Plasma levels of oxLDL/β₂GPI complexes have been shown to predict adverse outcomes and severity of coronary artery disease.³ The pathogenic role of oxLDL/β₂GPI complexes is further supported by their presence in atherosclerotic lesions by in vivo imaging techniques,⁴ where they co-localize with immune-inflammatory mononuclear cells and CD4+ lymphocytes.² The intracellular accumulation of these complexes by mononuclear cells is mediated by scavenger and specific antibodies via Fcγ receptors.^5,6 These findings support not only a role of these complexes in the initiation but, importantly, in the progression of atherothrombotic disease. The oxLDL/β₂GPI complexes have been described in arterial disease in the autoimmune^6,7 and in the nonautoimmune settings,^3,8,9 though there are very scanty data on venous thromboembolism (VTE).⁸ To expand on the topic, we evaluated oxLDL/β₂GPI complexes in patients with different types of atherosclerosis-related events (ARE) and VTE; because the study was carried out in a hospital-based thrombosis unit where homocysteine (HC) was measured as part of the thrombophilia screen, we also evaluate whether oxLDL/β₂GPI bore any relation to plasma HC, known to favor LDL oxidation,¹⁰ and to the activated partial thromboplastin time ratio (aPTTr).

Method

Study Type and Patients

This was a cross-sectional case–control study. Cases were consecutive patients attending the Haemostasis Clinic and the Internal Medicine wards of the Ospedale Riuniti di Foggia for ARE and spontaneous VTE. Inclusion criteria were thrombotic occlusion in any vascular district, peripheral vascular disease, cardiovascular disease, and cerebrovascular disease: In the case of acute occlusions, thrombophilia screening and other specific tests were performed between 3 and 6 months from the event to account for an acute-phase state, whereas blood samples were taken at the time of attendance if an acute event was excluded. Exclusion criteria were active or historical cancer, acute and chronic infectious, inflammatory and autoimmune disorders, acute and chronic renal, hepatic and respiratory disorders, primary/secondary antiphospholipid syndrome, insulin dependent and independent diabetes mellitus, and pregnancy. Controls were of 2 types: (1) hospital controls, participants attending the hospital for various symptoms and diagnosed with minor ailments and (2) healthy controls, patients attending the hemostasis clinic as part of family or personal thrombophilia screening for various reasons. Both control groups underwent the same exclusion criteria as patients. Consecutive patients and controls were invited to participate; consenting patients were enrolled in the study that was carried out according to the declaration of Helsinki and approved by the Ethics Committee of the Ospedale Riuniti of Foggia (Foggia, Italy; (99/CE/2015).

Blood Sampling

All participants were seen between 8.00 and 11.00 am, and their blood samples collected after a 20-minute resting period by neat venepuncture of an antecubital vein in plastic tubes containing 1/10 volume of 0.129 M trisodium citrate. All samples were immediately centrifuged at 2000 g for 10 minutes at 4°C and aliquots were stored at −70°C until use.

Thrombophilia Screening

In all participants, the following were measured: aPTTr, antithrombin, protein C (chromogenic assay, Behring, Marburg, Germany), free protein S antigen (enzyme-linked immunosorbent assay [ELISA]; Diagnostica Stago, Asnieres, France), and plasma HC (Bio-Rad, Oslo, Norway). Normal plasma obtained from 65 blood donors (male/female [M/F] 30/35, mean age 45 [16] years) served as control plasma for aPTTr and natural anticoagulants and to generate reference ranges (mean [2 standard deviation]). The dilute Russel viper venom time (screen and confirm) and the aPTT (with Hemosil RecombiPlasTin 2G as activator) were performed an ACLTOP500 coagulometer (all from Instrumentation Laboratory, Monza, Italy). Immunoglobulin G (IgG) anticardiolipin antibodies were measured by ELISA (Byk Gulden, Italy). Plasma HC was measured by ELISA, where free HC is obtained by hydrolysis of protein-bound HC and then enzymatically converted to S-adenolsyl-l-homocysteine (SAH) prior to the immunoassay. The latter is based on competition for binding sites on a monoclonal anti-SAH antibody between SAH in the sample and SAH immobilized on the walls of the microtiter plate. Factor V Leiden, prothrombin G20210A, and methylene-tetrahydrofolate reductase C677T polymorphisms were detected as previously described.¹¹

Lipid Profile

The lipid profile was laboratory grade and included total cholesterol (CHO), high-density lipoprotein (HDL), and triglycerides with LDL calculated according to Friedewald equation; an abnormal lipid profile was defined as CHO >200 mg/dL, LDL above 130 mg/dL, HDL <40 mg/dL for men, <50 mg/dL for women, and triglycerides >200 mg/dL or a combination of them.

oxLDL/β₂GPI Immunoassay

Plasma oxLDL/β₂GPI complexes were measured as previously described.⁹ Briefly, IgG2b murine monoclonal antibody (3H3) specific for human β₂GPI was coated onto 96-microwell plates and used to capture oxLDL/β₂GPI complexes via its reactivity with β₂GPI. Diluted patient sample was added to the appropriate microwells for incubation at room temperature for 1 hour. After washing, biotinylated 2E10 antibody (IgG murine monoclonal antihuman Apo B-100) was added to the microwells and incubated for 30 minutes, followed by Streptavidin–horseradish peroxidase for 30 minutes. Color was developed with tetramethylbenzidine/H₂O₂ for 30 minutes and the reaction stopped with 0.36 N sulphuric acid. Optical density was read at a wavelength of 450 nm (650 nm reference). The oxLDL/β₂GPI complex concentration (expressed in U/mL) was calculated against a reference curve built with 3-fold serial dilutions of a reference preparation.

Diagnosis of ARE and VTE

Both ARE and VTE were diagnosed by angiography, magnetic resonance imaging (MRI) and angio-MRI, electrocardiography, and serial cardiac markers measurements as indicated; VTE was diagnosed by Doppler ultrasound, computed tomography (CT) scan, and/or CT angiography as required.

Statistical Analysis

For categorical variables, χ² test or Fisher exact test was used; for continuous variables, paired t test and Wilcoxon signed rank sum test (when applicable) were used. Association between variables was assessed by univariate analysis (Pearson ρ) and the assumptions of the univariate analysis were tested by multivariable regression. Cochran-Armitage Trend test was used to determine the trend between oxLDL/β₂GPI quartiles and odds ratio (OR). Statistical analysis was performed using the JMP 11.2.0 program from SAS Institute Inc, North Carolina.

Results

Participants

The demographics of the patient population are shown in Table 1. The patients with ARE were divided into 3 groups: (1) those with myocardial infarction (MI; n = 20, of which 16 MI proper [twice in 3 patients] and 4 non-ST segment elevation MIs; (2) those with chronic peripheral vascular diseases with an acute occlusion (n = 5) and acute renal artery occlusions (n = 2); and (3) those with ischaemic strokes (n = 30; twice in 1 patient). The patients with VTE had been diagnosed with unprovoked occlusions in the following sites: above knee (n = 29), inferior vena cava (n = 2), central retinal vein (n = 3), cerebral sinuses (n = 3), pulmonary artery (n = 9; twice in 1 patient), pulmonary artery + femoral vein (n = 4), and portal vein (n = 2).

Table 1.

Demographics and Laboratory Parameters of All Participants.

		CTR	VTE	ARE	P
N		104	52	57
Age	mean [SD]	41 [15]	55 [17]	56 [11]	<.0001
M/F		45/59	22/30	33/24
Age at event	mean [SD]		47 [16]	47 [12]	Ns
MTHFR+/+		18	13	15
FVL+/−		3	3	5
PT+/−		2	3	2
MTHFR+/+ and PT+/−		2	4	1
MTHFR+/+ and VL+/−		1	1	0
PS		1	1	1
AT		0	1	0
Lipid		2	3^a	11^a
HTN		2	3	17
Smoke		4	2	16
BMI		1	2	2
Statin		1	2	7
ACE		2	1	7
β-Blocker		0	1	5
oxLDL/β₂GPI	mean [SD]	195 [613]	241 [51]	261 [94]	<.0001
HC (μmol/L)	mean [SD]	12 [10]	13 [6.7]	15 [0.93]	ns
aPTTr	mean [SD]	1.002 [0.09]	0.94 [0.09]	0.95 [0.09]	.02

Abbreviations: CTR, controls; VTE, venous thromboembolism; ARE, atherosclerosis-related event; SD, standard deviation; MTHFR, methylene tetrahydrofolate reductase; FVL, factor V Leiden; PT, prothrombin mutation; PS, protein S; AT, antithrombin; HTN, diastolic hypertension; BMI, body mass index; ACE, angiotensin-converting enzyme: β₂GPI/oxLDL, beta-2-glycoprotein-I/oxidized low-density lipoprotein; HC, homocysteine; aPTTr, activated partial thromboplastin ratio.

^aDone on 22 VTE and 56 ARE participants.

The hospital and normal controls were merged in 1 group (Table 1). The hospital controls included 55 participants (M/F 22/33) diagnosed with the following conditions: 7 ovarian cysts, 9 headaches, 6 diastolic hypertension, 1 cataract, 2 Raynaud phenomenon, 1 patent foramen ovale, 2 lower limb edema, 9 early miscarriages, 1 bowel fistula, 3 hirsutism, 1 mitral valve prolapse, 3 menorrhagia, 1 eczema, 1 sternal trauma, 1 left ventricular dysfunction, 1 hypoxic gliosis, 2 peripheral neuropathy, 1 epilepsy, 1 colonic polyps, 1 goitre, and 1 upper limb paresthesias. The normal controls included 49 participants (M/F 23/26) seen for the following reasons: 2 liver donors, 1 kidney donor, and 46 for a thrombophilia screen: 7 before starting oral contraception and 39 were first-degree family members of a proband with thrombosis and inherited thrombophilia.

oxLDL/β₂GPI, HC, and aPTT Across Groups

Plasma concentrations of oxLDL/β₂GPI were progressively higher from controls, through patients with VTE to patients with ARE (P < .0001), HC showed a similar but nonsignificant trend, whereas aPTTr was lower in the VTE group (P = .02; Table 1).

oxLDL/β₂GPI, ARE, and VTE

The oxLDL/β₂GPI levels were divided into 4 quartiles: quartile 1, 20 to 175.6 U/mL; quartile 2, 175.7 to 227.8 U/mL; quartile 3, 227.9 to 251.6 U/mL; and quartile 4, 251.7 to 482.2 U/mL. The odds of having any vascular event was greater in the fourth and second quartiles than in the first quartile (9.9 and 4.1, respectively; Figure 1A); this is a reflection of the greater odds of having ARE in the fourth and second quartiles than in the first quartile (8.5 and 6.0, respectively; Figure 1C); among the ARE phenotypes, the odds of having MI was maximal in the fourth quartile (17.8; Figure 1C). The odds of having VTE was greater in the fourth quartile with a gradual reduction to the second quartile (11.4; Figure 1B).

Figure 1.

Odds ratio (OR) of any vascular event (A), atherosclerosis-related events (ARE) (B), spontaneous venous thromboembolism (VTE) (C), and myocardial infarction (MI) (D) by oxLDL/β₂GPI quartiles. oxLDL/β₂GPI indicates oxidized low-density lipoprotein/beta-2-glycoprotein-I.

Predictors of oxLDL/β₂GPI in the ARE Group

By univariate analysis, oxLDL/β₂GPI correlated to lipids status (r = 0.25, P = .056), subgroup (r = 0.62, P < .0001), and aPTTr (r = −0.32, P = .02); by multivariable analysis with age, sex, lipid status, and subgroup as independent variables and oxLDL/β₂GPI as the dependent variable, only the MI subgroup independently predicted oxLDL/β₂GPI (Table 2).

Table 2.

Predictors of β₂GPI/oxLDL in Atherosclerosis-Related Event Group.

Independent Variables	Coefficient	Standard Error	r _partial	t	P
Age	−1.1595	0.9477	−.1673	−1.223	.2267
Sex	−4.7850	20.7009	−.03204	−0.231	.8181
Lipid	12.0370	29.0738	.05732	0.414	.6806
Myocardial infarction	−65.8969	12.1624	−.6007	−5.418	<.0001

Abbreviation: β₂GPI/oxLDL, beta-2-glycoprotein-I/oxidized low-density lipoprotein.

Predictors of aPTTr in the ARE Group

By univariate analysis, aPTT correlated negatively to age (r = −.27, P = .054) and to oxLDL/β₂GPI (r = −.32, P = .02); by multivariable analysis, age, sex, oxLDL/β₂GPI, and lipid status entered a multivariable regression model as independent variables and aPTTr as the dependent variable: Only age and oxLDL/β₂GPI independently predicted aPTTr (Table 3). No relationship was found between oxLDL/β₂GPI and any variable in patients with VTE.

Table 3.

Predictors of aPTTr in Atherosclerosis-Related Event Group.

Independent Variables	Coefficient	Standard Error	r _partial	t	P
Age	−.002307	.001087	−.3047	−2.122	.0395
Sex	−.01035	.02576	−.06047	−0.402	.6897
Lipid	.03521	.03832	.1372	0.919	.3632
β₂GPI/oxLDL	−.0003503	.0001393	−.3544	−2.514	.0157

Abbreviations: aPTTr, activated partial thromboplastin ratio; β₂GPI/oxLDL, beta-2-glycoprotein-I/oxidized low-density lipoprotein.

Discussion

Our study shows that the mean plasma concentration of oxLDL/β₂GPI complexes is elevated in ARE and VTE. Patients in the upper quartiles were equally likely to suffer ARE or VTE: However, when the data were analyzed according to ARE subtype, patients in the upper quartile were much more likely to have suffered MI than other arterial phenotypes, though different MI severity may account for the bimodal significance of the quartiles; the multivariable regression confirmed MI as an independent predictor of oxLDL/β₂GPI complexes. Given the relationship between plasma HC and oxidative stress,¹⁰ we sought a possible relation between plasma HC and oxLDL/β₂GPI: In the ARE group, neither HC nor hypertension, nor smoking, nor statin use was associated with oxLDL/β₂GPI complexes though lipid status did, but only in univariable analysis. Likely the predictive effect of MI toward oxLDL/β₂GPI may be explained by the vascular inflammation¹² that predisposes to oxLDL/β₂GPI complex formation.¹³

Notably, in the ARE group, we found that oxLDL/β₂GPI negatively predicted the aPTTr: Oxidative modifications of LDLs have been associated either with a lower aPTTr¹⁴ or with a higher aPTTr,¹⁵ but this is the first time such an association is found with the oxLDL/β₂GPI complex; β₂GPI binds to oxLDL in the attempt to quench LDL oxidation, and oxidation of certain lipoprotein phospholipids supports thrombin generation,¹⁶ hence β₂GPI quenching should limit thrombin generation and coagulation activation and be more in keeping with a greater aPTTr¹⁵; on the other hand, even oxLDL/β₂GPI complexes may have residual oxidative capacity as indicated by C-reactive protein binding to oxLDL/β₂GPI in an attempt to further quench its oxidative potential.¹⁷ Alternatively or additionally, the relation between oxLDL/β₂GPI and a lower aPTTr may be explained by increased plasma level of clotting factors as part of the vascular inflammation accompanying MI.¹⁸

With regard to VTE, this is the first report of a graded OR between oxLDL/β₂GPI quartiles and VTE; moreover, a closer overview of our data shows that the ORs for VTE and ARE (excluding MI) are quite similar, lending support to the idea that arterial and venous disease might share common pathogenic pathways among which oxidative and inflammatory mechanisms. According to a meta-analysis including 21 studies and 63 552 patients, cardiovascular risk factors associate with VTE¹⁹; moreover, residual VTE is associated with a higher prevalence of subclinical atherosclerosis,²⁰ indicating a bidirectional relationship into which oxLDL/β₂GPI complexes may well find its niche.

Limitations

We consider the following as limitations: (1) the cross-sectional nature of the study; (2) the lack of electrocardiographic screening in most control participants, as we cannot exclude that some of them might have subclinical ischemic heart disease though this would blunt rather than accentuate differences between groups; (3) the lack of lipid profiles in a proportion of patients, mostly in the VTE group, as participants attended a hamostasis unit and we did not test them; and (4) the lack of inflammatory markers or other clotting factors behaving as acute-phase reactants that would have allowed further insight into the relationship between oxLDL/β₂GPI and aPTTr of the ARE group.

Conclusions

The oxLDL/β₂GPI may represent a marker for MI, whereas the graded association with VTE should be further explored taking into account the extension of the venous occlusions⁸ and we describe for the first time a link between oxLDL/β₂GPI complexes and a low aPTTr. The bidirectional relationship oxLDL/β₂GPI with ARE and VTE suggests that the link between arterial and venous disease may rely on similar oxidative/inflammatory and eventually coagulation pathways that require more elucidation on the venous side.

Footnotes

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Dr L. R. Lopez is chief medical officer of Corgenix.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: oxLDL/β₂GPI kits were a kind gift from Corgenix.

References

Gao

Zhao

Wang

. Association between circulating oxidized LDL and atherosclerotic cardiovascular disease: a meta-analysis of observational studies. Can J Cardiol. 2017;33(12):1624–1632.

Kobayashi

Kishi

Atsumi

. Circulating oxidized LDL forms complexes with beta2-glycoprotein I: implication as an atherogenic autoantigen. J Lipid Res. 2003;44(4):716–726.

Greco

Conti-Kelly

Anthony

. Oxidized-LDL/beta-2-glycoprotein I complexes are associated with disease severity and increased risk for adverse outcomes in patients with acute coronary syndromes. Am J Clin Pathol. 2010;133(5):737–743.

Sasaki

Kobayashi

Kita

. In vivo distribution of single chain variable fragment (scFv) against atherothrombotic oxidized LDL/B2-glycoprotein I complexes into atherosclerotic plaques of WHHL rabbits: implication for clinical PET imaging. Autoimmun Rev. 2016;16(2):159–167.

Lopez

Kobayashi

Merrill

Matsuura

Lopez

. IgG autoantibodies against beta2-glycoprotein I complexed with a lipid ligand derived from oxidized low-density lipoprotein are associated with arterial thrombosis in antiphospholipid syndrome. Clin Dev Immunol. 2003;10(2-4):203–211.

Matsuura

Kobayashi

Matsunami

Lopez

. The immunology of atherothrombosis in the antiphospholipid syndrome: antigen presentation and lipid intracellular accumulation. Autoimmun Rev. 2009;8(6):500–505.

Lopez

Salazar-Paramo

Palafox-Sanchez

Hurley

Matsuura

Garcia-De La Torre

. Oxidized low-density lipoprotein and beta2-glycoprotein I in patients with systemic lupus erythematosus and increased carotid intima-media thickness: implications in autoimmune-mediated atherosclerosis. Lupus. 2006;15(2):80–86.

Berger

Rockman

Guyer

Lopez

. Proatherogenic oxidized low-density lipoprotein/β2-glycoprotein I complexes in arterial and venous disease. J Immunol Res. 2014;2014:234316.

Bliden

Chaudhary

Lopez

. Oxidized low-density lipoprotein-β2-glycoprotein I complex but not free oxidized LDL is associated with the presence and severity of coronary artery disease. Am J Cardiol. 2016;118(5):673–678.

10.

Voutilainen

Morrow

Roberts

LJ II

. Enhanced in vivo lipid peroxidation at elevated plasma total homocysteine levels. Arterioscler Thromb Vasc Biol. 1999;19(5):1263–1266.

11.

Amitrano

Guardascione

Ames

PRJ

. Thrombophilic genotypes, natural anticoagulants, and plasma homocysteine in myeloproliferative disorders: relationship with splanchnic vein thrombosis and arterial disease. Am J Hematol. 2003;72(2):75–81.

12.

Teague

Ahlman

Alavi

. Unravelling vascular inflammation: from immunology to imaging. J Am Coll Cardiol. 2017;70(11):1403–1412.

13.

Matsuura

Lopez

Shoenfeld

Ames

PRJ

. β2-glycoprotein-I and oxidative inflammation in early atherogenesis: a progression from innate to adaptive immunity?

Autoimmun Rev. 2012;12(2):241–249.

14.

Bai

Liu

Deng

. Plasma very-low-density lipoprotein, low-density lipoprotein, and high-density lipoprotein oxidative modification induces procoagulant profiles in endogenous hypertriglyceridemia. Free Radic Biol Med. 2006;40(10):1796–1803.

15.

Kumagai

Hoshi

Tsutsumi

Ebina

Yokota

. Inhibition of plasma coagulation through interaction between oxidized low-density lipoprotein and blood coagulation factor VIII. Biol Pharm Bull. 2005;28(6):952–956.

16.

Rota

McWilliam

Baglin

Byrne

. Atherogenic lipoproteins support assembly of the prothrombinase complex and thrombin generation: modulation by oxidation and vitamin E. Blood. 1998;91(2):508–515.

17.

Tabuchi

Inoue

Usui-Kataoka

. The association of C-reactive protein with an oxidative metabolite of LDL and its implication in atherosclerosis. J Lipid Res. 2007;48(4):768–781.

18.

Kucharska-Newton

Couper

Pankow

. Hemostasis, inflammation, and fatal and nonfatal coronary heart disease: long-term follow-up of the atherosclerosis risk in communities (ARIC) cohort. Arterioscler Thromb Vasc Biol. 2009;29(12):2182–2190.

19.

Ageno

Becattini

Brighton

Selby

Kamphuisen

. Cardiovascular risk factors and venous thromboembolism: a meta-analysis. Circulation. 2008;117(1):93–102.

20.

Prandoni

Ciammaichella

Mumoli

; Veritas investigators. An association between residual vein thrombosis and subclinical atherosclerosis: cross sectional study. Thromb Res. 2017;157:16–19.

Predictive Value of Oxidized Low-Density Lipoprotein/β 2 -Glycoprotein-I Complexes (oxLDL/β 2 GPI) in Nonautoimmune Atherothrombosis

Abstract

Introduction:

Methods:

Results:

Conclusions:

Keywords

Introduction

Method

Study Type and Patients

Blood Sampling

Thrombophilia Screening

Lipid Profile

oxLDL/β2GPI Immunoassay

Diagnosis of ARE and VTE

Statistical Analysis

Results

Participants

oxLDL/β2GPI, HC, and aPTT Across Groups

oxLDL/β2GPI, ARE, and VTE

Predictors of oxLDL/β2GPI in the ARE Group

Predictors of aPTTr in the ARE Group

Discussion

Limitations

Conclusions

Footnotes

Declaration of Conflicting Interests

Funding

References

oxLDL/β₂GPI Immunoassay

oxLDL/β₂GPI, HC, and aPTT Across Groups

oxLDL/β₂GPI, ARE, and VTE

Predictors of oxLDL/β₂GPI in the ARE Group