intl.elsevierhealth.com/journals/thre

REGULAR ARTICLE

Endothelial function is impaired in patients withprimary antiphospholipid syndrome

Monika Stalc a,*, Pavel Poredos a, Polona Peternel a, Matija Tomsic b,Miran Sebestjen a, Tanja Kveder b

a Department of Vascular Disease, Zaloska c.7, Clinical Centre, SI-1000 Ljubljana, Sloveniab Department of Rheumatology, Clinical Centre, Ljubljana, Slovenia

Received 30 June 2005; received in revised form 8 September 2005; accepted 12 September 2005Available online 7 November 2005

0049-3848/$ - see front matter D 200doi:10.1016/j.thromres.2005.09.005

Abbreviations: APS, antiphospholipcoagulant antibodies; EC, endotheliallar cell adhesion molecule-1; sICAMplasminogen activator inhibitor-1; vprotein; ELISA, enzyme-linked immuno* Corresponding author. Tel.: +386 1E-mail address: monikastalc@emai

KEYWORDSPrimaryantiphospholipidsyndrome;Endothelial function;Flow mediated dilationof brachial artery;Atherosclerosis

Abstract

Introduction: The aim of this study was to evaluate endothelial function in patientswith primary antiphospholipid syndrome (PAPS).Patients and methods: Flow mediated (FMD) and glyceryl trinitrate (GTN) induceddilation of the right brachial artery were studied in 25 patients with PAPS and25 controls matched by age, sex and conventional risk factors for atheroscle-rosis. Fibrinogen, D-dimer, adhesion molecules, tissue plasminogen activator (t-PA) and plasminogen activator inhibitor-1 (PAI-1) antigens and activities weremeasured.Results: Mean (SD) FMD was significantly lower in PAPS than in controls (8F5% vs.15F6%, P b0.001); GTN-induced dilation did not differ between the groups. Therewas a correlation between the baseline diameter of the brachial artery andduration of the disease (�0.56, P b0.05) and between GTN induced dilation andduration of the disease (0.51, P b0.05). Concentrations of vascular cell adhesionmolecule-1 ( P b0.001), intracellular adhesion molecule-1 ( P b0.001) and fibrino-gen ( P b0.05) were higher in patients than in controls but no differences wereobserved for D-dimer, t-PA and PAI-1 antigens and activities. There was correlationbetween concentration of vascular cell adhesion molecule-1 and FMD (�0.35,P b0.05) and between intracellular adhesion molecule-1 and FMD (�0.41,P b0.05).

Thrombosis Research (2006) 118, 455—461

5 Elsevier Ltd. All rights reserved.

id syndrome; aPL, antiphospholipid antibodies; aCL, anticardiolipin antibodies; LA, lupus anti-cell; h2GPI, beta2 glycoprotein I; FMD, flow mediated vasodilation; sVCAM-1, soluble vascu--1, soluble intracellular adhesion molecule-1; t-PA, tissue plasminogen activator; PAI-1,WF, von Willebrand factor; TF, tissue factor; HDL, high density lipoproteinLDL, low density lipo-sorbent assay; ACE, angiotensin-converting enzyme; GTN, glyceryl trinitrate.522 80 32; fax: +386 1 522 80 70.l.si (M. Stalc).

M. Stalc et al.456

Conclusions: This study shows that endothelial function is impaired in patients withprimary APS, possibly contributing to accelerated atherosclerosis and thromboem-bolic complications in these patients.D 2005 Elsevier Ltd. All rights reserved.

Antiphospholipid syndrome (APS) is characterizedby recurrent arterial and/or venous thrombosisand/or foetal loss in the presence of circulatingantiphospholipid antibodies (aPL) (i.e. anticardio-lipin (aCL) and/or lupus anticoagulant (LA) anti-bodies) [1]. Antibody mediated endothelial cell(EC) injury has been identified as a factor poten-tially involved in the pathogenesis of thrombosis inAPS. A relationship between aPL and endothelialcells activation and/or damage has been repro-duced in in vitro and in vivo experimental animalmodels [2—4]. Several studies have shown thatantibodies reacting with beta 2 glycoprotein I(h2GPI), the plasma cofactor for aPL, activate EC.Anti-h2GPI were found to induce an endothelialpro-adhesive phenotype mediated by adhesionmolecule up-regulation, and to increase the secre-tion of pro-inflammatory cytokines [2—4].

Endothelial cell damage may result in endothe-lial dysfunction, which is characterized by animbalance between relaxing and contracting fac-tors, between anticoagulant and procoagulantmediators or between growth inhibiting and growthpromoting factors [5]. Endothelial dysfunctionplays an important role in the pathogenesis ofatherosclerosis [6]. It is an early event in experi-mental studies of atherogenesis, preceding forma-tion of plaques [7].

Endothelial dysfunction can be assessed, bybiochemical analysis of systemic markers of endo-thelial damage and activation or, clinically, in theperipheral circulation, by monitoring the vasodila-tory response evoked by increased blood-flow shear(flow mediated vasodilation—FMD) [5,8]. Dimin-ished FMD in the brachial artery is an indicator ofcardiovascular risk and is considered to be a markerof atherosclerosis [9].

In spite of much in vitro evidence, few studieshave addressed the issue as to whether a compa-rable endothelial perturbation might be detectablein APS patients.

Circulating markers of endothelial functionwere measured in patients with APS but resultsdiffer. Most studies have not differentiatedpatients with primary and secondary APS. In-creased vascular cell adhesion molecule-1(sVCAM-1) was found in the plasma of patientswith primary APS and recurrent thrombosis [10].Endothelium derived microparticles were in-creased in patients with APS [11,12]. Plasma levels

of endothelin 1 peptide, the most potent endo-thelium derived contracting factor, correlatedsignificantly with a history of arterial thrombosisin APS [13]. The in vivo up-regulation of the tissuefactor (TF) pathway and the imbalance of cyto-kines have been proposed as potential mechanismsof thrombosis in the APS. Higher levels of circu-lating TF and elevated levels of interleukin 6 andtumor necrosis factor were found in patients withprimary APS [14]. On the other hand there was nodetectable increase in markers of endothelial cellsactivation (adhesion molecules, endothelin-1, in-terleukin-6, soluble P-selectin) in patients withprimary APS [15]. There is little information onFMD in patients with primary APS [16].

The aim of our study therefore was to investi-gate several in vivo parameters of endothelialdysfunction and their interrelationship in patientswith primary APS, by measuring FMD and circulatingmarkers of endothelial damage. Some circulatingmarkers of haemostatic system activation were alsoassessed.

Materials and methods

25 patients (15 men, 10 women, mean (FSD) age47F13, range 26 to 76) with primary APS accordingto the international criteria for APS were includedin the study [1]. Clinical manifestations of patientswith primary APS were stroke in six and venousthrombosis in 19 patients. The mean (SD) diseaseduration was 8 (6) years. All patients were treatedwith oral anticoagulants. Eight patients and twocontrols were treated with angiotensin-convertingenzyme (ACE) inhibitors and two patients and onecontrol with statins.

Each patient was matched by sex, age andclassical risk factors for atherosclerosis with onecontrol subject who underwent a similar protocol.All participants underwent a full medical history andphysical examination. Cardiovascular risk factorssuch as arterial hypertension, diabetes mellitus,smoking, obesity and hyperlipidemia were investi-gated. A total cholesterol higher than 5mmol/l and/or with normal values, but taking hypolipemicdrugs, was indicative of hypercholesterolemia.

Blood was sampled from all subjects in themorning after an overnight fast for the followinglaboratory tests: cholesterol (total and fractions),

Table 1 Distribution of risk factors for patients andcontrols

Primary APS,n =25

Controls,n =25

P Value

Age (years), mean (SD) 47 (13) 47 (13) NSHypertension 5 3 NSHyperlipidemia 7 6 NSDiabetes mellitus 1 1 NSSmoking 1 1 NSBMIN27 2 2 NS

Endothelial function is impaired in patients with primary antiphospholipid syndrome 457

triglycerides, glucose, intracellular adhesion mole-cule-1 (sICAM-1), sVCAM-1, fibrinogen, D-dimer,plasminogen activator inhibitor-1 (PAI-1) and tissueplasminogen activator (tPA) antigens and activities.

Concentrations of serum glucose, total choles-terol, high density lipoprotein (HDL) cholesteroland triglycerides were determined by standardcolorimetric assays (Ektachem 250 Analyzer, East-man Kodak Company, Rochester, USA). Low densitylipoprotein (LDL) cholesterol was calculated fromFriedewald’s formula.

Soluble ICAM-1 and VCAM-1 were measured byELISA (both R&D Systems, Abingdon,UK).

Fibrinogen was determined by a modified Claussmethod using MultifibrenR U (Dade Behring) on anautomated coagulation analyser (BCT, Dade Behr-ing, Marburg, Germany).

t-PA activity was detected with an immunoactiv-ity assay, ChromolizeR tPA and PAI-1 activity withChromolizeR PAI-1 (both Trinity Biotech Plc, CoWicklow, Ireland).

t-PA and PAI-1 antigens were detected by ELISA(Imulysek t-PA and Imulyse PAI-1, respectively,both Biopool, Ume3, Sweden).

D-dimer was determined by ELISA (AsserachromRD-dimer, Diagnostica Stago, Asnieres, France).

Endothelium-dependent FMD and glyceryl trini-trate (GTN)-induced (endothelium-independent)dilation of the right brachial artery were studiedusing a high resolution B mode ATL 5000 ultrasoundsystem with a 7 MHz linear array transducer, asdescribed [7]. The subjects rested in the supineposition for 10 min before hemodynamic measure-ments were performed. The right brachial arterywas scanned in the longitudinal section 2 to 15 cmabove the elbow to find the clearest images of theanterior and posterior wall layers. The meanarterial diameter was measured at the end ofdiastole, which was determined by simultaneousmonitoring of the electrocardiogram. At least threecardiac cycles were analyzed for each scan and themeasurements averaged. The flow velocity wasmeasured at a fixed incident angle of 608 to thevessel with the range gate of 1.5 mm located in thecentre of the artery. The baseline (resting) bloodflow was estimated by multiplying the velocity timeintegral of the Doppler flow signal (corrected forincident angle) by the vessel cross-sectional area.Hyperaemic flow increase was induced by inflationof a blood pressure tourniquet, placed around theforearm, to a pressure of 300 mm Hg for 4.5 min.Hyperaemic flow (with increased flow producing anendothelium-dependent stimulus for vasodilation)was recorded for the first 15 s and diametermeasurements were taken 45—60 s after cuffdeflation. The endothelium-dependent dilation

was expressed as the percentage change of thediameter after reactive hyperaemia relative to thebaseline scan. A period of 10 min was allowed forvessel recovery, after which a further resting scanwas taken. Endothelium independent dilation wasprovoked by sublingual administration of 400 Ag ofGTN. The final scan was performed 4.5 min later.Endothelium-independent dilation was expressedas the percentage change in the diameter afterGTN administration relative to the baseline scan.

The ratio between the endothelium-dependentand endothelium independent dilation was cal-culated for determining the endothelium-depen-dent and endothelium-independent vasodilatoryresponses.

To assess the reproducibility of measurements,20 subjects were selected at random for repeatedvascular studies. The correlation coefficient be-tween the absolute differences and mean values ofpaired measurements was 0.92, P b0.05.

All participants gave written consent for allprocedures. The study was approved by the Stateethical committee of Slovenia.

Statistical analysis

Data are present as means (FSD) or medians (min—max). Student’s t test was performed to compareparametric variables between cases and controls.Mann—Whitney U test was performed for non-parametric parameters. Pearson’s correlation co-efficient was used to examine the relation betweenbrachial FMD and several study variables. Levels ofP b0.05 were considered significant. Data wereanalysed by the Statistica 6.0 computer program(Stat Soft Inc., USA).

Results

There was no difference between groups withrespect to diabetes mellitus, smoking, hyperten-sion, body mass index or hypercholesterolemia(Table 1). There was no difference of total and

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30B

rach

ial a

rter

y di

latio

n (%

)

P-APS Controls P-APS Controls

P < 0.05

NS

Figure 1 Flow-mediated (left) and glyceryltrinitrate-mediated (right) dilation of brachial artery in patientswith primary antiphospholipid syndrome (P-APS) andcontrols after correction for the difference in baselinediameter of the brachial artery.

M. Stalc et al.458

LDL cholesterol between groups or of baselinediameter of brachial artery between groups(0.39F0.06 vs. 0.37F0.06 cm; P =0.16).

FMD was significantly lower in patients withprimary APS than in control subjects (8F5% vs.15F6%, P b0.001), as was GTN-induced dilation ofthe brachial artery (16F6% vs. 21F7%, P b0.05).After correction for the difference in baselinediameter of the brachial artery, FMD was stillsignificantly lower in patients (P b0.05), but GTN-induced dilation did not differ between the groups(Fig. 1).

The ratio (FSD) between FMD and GTN-induceddilation was significantly lower in patients withprimary APS than in controls (0.48F0.3 vs.0.71F0.18; P b0.01).

Table 2 shows measured circulating markers ofendothelial activation/damage and markers ofhaemostatic system activation in patients andcontrols. Concentrations of sVCAM-1 (P b0.001),sICAM-1 (P b0.001) and fibrinogen (P b0.05) weresignificantly higher in patients than in controls.

Table 2 Distribution of some circulating markers of endosystem activation in patients and controls

Primary AP

Fibrinogen (g/L) median (min—max) 3.2 (2.1—D-dimer (ng/mL) median (min—max) 190 (58—2PAI-1 activity (rel) median (min—max) 2.0 (0.0—PAI-1antigen (ng/mL) median (min—max) 9.1 (1.4—tPA activity (rel) median (min—max) 1.15 (0.2—tPA antigen (ng/mL) median (min—max) 8.6 (0—14sVCAM-1 (ng/mL) median (min—max) 714 (477—sICAM-1 (ng/mL) median (min—max) 301 (220—

There was no difference between groups withrespect to D-dimer, t-PA and PAI-1 antigen andactivity.

We found no significant correlation of FMD withage, lipid parameters, t-PA and PAI-1 antigen andactivity, fibrinogen, D-dimer, duration of disease orbaseline diameter of the brachial artery. There wasa significant inverse correlation between concen-tration of vascular cell adhesion molecule-1 andFMD (�0.35, P b0.05) and between intracellularadhesion molecule-1 and FMD (�0.41, P b0.05).

There was also a significant inverse correlationbetween baseline diameter of the brachial arteryand duration of disease (�0.56, P b0.05) andbetween GTN induced dilation and duration ofdisease (0.51, P b0.05).

Discussion

Normal endothelial function includes control overthrombosis and thrombolysis, platelets and leuko-cyte interaction with the vessel wall, and regula-tion of vascular tone and smooth muscle cellproliferation [5]. Several in vitro studies andstudies on animal models have shown that incuba-tion of endothelial cells with aPL from APS patientsgenerates different effects on endothelial functionvia h2GPI. As a whole this might cooperate insustaining endothelial perturbation that has beensuggested to have a pivotal pathogenetic role inAPS associated thrombosis [2—4]. A few studieshave addressed this issue in vivo in APS patients,but without definite conclusions.

The principal finding of this cross-sectional studyprovides evidence for an endothelial dysfunction inpatients with primary APS. We found an impairedendothelium-dependent vasodilatory response ofconduit peripheral arteries to provoked endogenicrelease of nitric oxide by increased flow duringreactive hyperaemia in patients with primary APS,as well as increased plasma levels of solubleadhesion molecules.

thelial activation/damage and markers of haemostatic

S Controls P value

4.7) 2.6 (2.1—3.8) P b0.05106) 241 (124—1132) NS228.0) 3.7 (0—27) NS4.7) 10.4 (0—41.6) NS2.2) 1.2 (0.3—2) NS.1) 5.6 (3—12.9) NS1176) 535 (407—790) P b0.001431) 238 (187—360) P b0.001

Endothelial function is impaired in patients with primary antiphospholipid syndrome 459

Patients did not differ from controls withrespect to the traditional risk factors for athero-sclerosis, suggesting that in this group of patientsendothelial dysfunction measured by FMD ispredominantly the consequence of the presenceof primary APS itself. Lipid parameters did notinfluence FMD. Possibly this is because patientsdid not have a significant dyslipidemia or weretreated with adequate hypolipaemic drugs. Only afew patients exhibited increased levels of LDLcholesterol.

It is considered that FMD is inversely related toarterial diameter. However, when the influence ofbaseline diameter on flow induced dilation wasevaluated, patients still had significantly reducedendothelium-dependent vasodilation, but therewere no differences in endothelium independentvasodilation.

All patients were treated with warfarin. As far aswe are aware there are no studies of warfarin thatdescribe effects on EC activation.

The results of our study are in agreement withthose of Mercanoglu et al. who found markedlydecreased FMD in 20 patients with primary APScompared to age, sex and traditional risk factorsfor atherosclerosis matched controls [16].

In contrast, Meroni et al. found no significantdifference between 20 primary and 6 systemiclupus erythematosus (SLE) APS and controls regard-ing FMD [17]. The distribution of atherosclerosisrisk factors in patients and controls was notdescribed. They might have masked the atherogen-ic potential of aPL. In their study SLE APS patientswere also included so the two studies cannot becompared.

We found that the appearance of endothelialdysfunction is not related to the duration ofprimary APS. As the duration of the disease doesnot influence the endothelial function, it appearsthat primary APS affects endothelial dysfunctionindependently of its duration, which is in accor-dance with the assertion that primary APS isdirectly involved in the pathogenesis of endothelialdysfunction [3].

In our study, the diameter of the brachial arterywas found to be smaller with longer duration ofprimary APS. This could be a consequence of adysregulation of vascular tonus towards a vasocon-striction that might be related to decreased releaseof nitric oxide, or to increased production ofvasoconstrictor substances with duration of thedisease.

In a disease that affects the vessel wall, damageto the muscular layer is also expected. In ourprimary APS patients, endothelium-independentvasodilation (mediated by GTN) was diminished to

a lesser extent than was FMD. After correction forthe difference in baseline diameter of the brachialartery, GTN-induced dilation was similar in the twogroups. This finding implies that primary APS doesnot influence the sensitivity of vascular smoothmuscle to nitric oxide or its functional capability.

However the correlation observed between du-ration of the disease and GTN induced dilationsuggests that there are probably some changes inthe vascular smooth muscle cell that progress withthe duration of disease and that a time lag is neededfor manifestation of smooth muscle damage.

The ratio between endothelium-dependent andendothelium-independent dilation was lower inpatients with primary APS than in controls. Thismeans that brachial artery dilation is altered,mainly because of impaired endothelial function.We can conclude that primary APS has a predom-inantly negative effect on endothelial function.

Endothelial dysfunction can also be assessed bybiochemical analysis of systemic markers of endo-thelial damage and activation [8]. aPL have beenshown to activate cultured EC, as demonstrated byenhanced expression of adhesion molecules [3]. Wefound that plasma levels of soluble adhesionmolecules (sICAM-1 and sVCAM-1) were significantlyhigher in patients than in controls and correlatedwith FMD. These findings support the hypothesisthat endothelium is activated in our group ofpatients with primary APS. These results are inagreement with those of Kaplanski et al. who foundmarkedly increased sVCAM-1 concentrations inprimary APS [10]. No differences in the levels ofsoluble adhesion molecules were found in two otherstudies [15,17]. Such results might be attributed todifferent study protocols. Some studies have notdifferentiated patients with primary and secondaryAPS. In all studies patients were compared to ageand sex matched controls, but traditional riskfactors for atherosclerosis and drugs that caninfluence endothelial function (i.e. statins, ACEinhibitors) were not taken into consideration.Number of patients with multiple thromboticevents in a study group might also be importantas sVCAM-1 concentrations were especially in-creased in patients with repeated thromboticevents [10].

Measurements of levels of haemostatic factorsand activation products are important in investiga-tions of the hypercoagulable state. Impaired fibri-nolysis has been reported in patients with APS [18].We investigated t-PA and PAI-1 levels but nodifferences were found between patients andcontrols. This is in agreement with results reportedby Meroni et al. [17] which showed no increase int-PA levels.

M. Stalc et al.460

In several studies an association between fibrin-ogen levels and subsequent cardiovascular diseaserisk was shown [19]. We measured plasma fibrino-gen and found that its concentration was higher inpatients than in controls, in keeping with a previousstudy [20]. Increased fibrinogen could be anadditional risk factor for atherothrombotic compli-cations in APS patients.

D-dimer is also a strong predictor of cardiovas-cular events in the general population and inpatients with cardiovascular events [21], and itsincrease was expected in our study population aswell. It is significantly decreased during oralanticoagulant treatment [22]. This could be oneof the reasons why, in our study, there were nodifferences between patients and controls since allpatients were treated with warfarin.

This study showed functional endothelial dam-age in patients with primary APS manifested byimpaired endothelial function in peripheral circu-lation and increased levels of systemic markers ofendothelial cell activation. As endothelial dysfunc-tion is most probably involved in atherogenesis,our study suggests that patients with primary APSare at risk for atherosclerotic disease, in spite ofabsence of classic risk factors for atherosclerosis.The role of primary APS as an independent riskfactor for atherosclerosis is unclear. Recently, highprevalence of increased carotid artery intimamedia thickness (IMT) was found in two studiesof patients with primary APS [23,24]. Abnormalankle brachial index (ABI) was more common inprimary APS than in healthy controls, suggesting asubclinical atherosclerotic process in thesepatients [25].

Endothelial function of the brachial artery canbe measured in APS patients non-invasively withbrachial ultrasound. It may help identify patientswith, or at risk of, atherosclerotic disease. Thesedata are of particular interest from the clinicalpoint of view, because patients with primary APSshould be considered at high risk of atherosclerosis.

It would be interesting and clinically importantto evaluate whether interventions known to influ-ence endothelial function [26—30] can modify theimpaired endothelial function in primary APSpatients. As endothelial activation induced byanti-h2GPI antibodies is prevented by statins [31]and TF expression on endothelial cells is inhibitedby statins [32] it seems rational for using them inAPS [33].

The fact that our primary APS patients withthromboembolic complications had endothelialdysfunction raises the question whether endotheli-al dysfunction is involved also in pathogenesis ofthromboembolic disease in APS.

Further studies with larger number of patientsare needed to clarify the activation and/or damageof endothelium and to identify the mechanismsinvolved in the aetiopathogenesis of thromboem-bolic and atherosclerotic complications in patientswith primary APS.

References

[1] Wilson WA, Gharavi AE, Koike T, Lockshin MD, Branch DW,Piette JC, et al. International consensus statement onpreliminary classification criteria for definite antiphospho-lipid syndrome: report of an international workshop.Arthritis Rheum 1999;42:1309–11.

[2] Riboldi P, Gerosa M, Raschi E, Testoni C, Meroni PL.Endothelium as a target for antiphospholipid antibodies.Immunobiology 2003;207:29–36.

[3] Pierangeli SS, Colden-Stanfield M, Liu X, Barker JH,Anderson GL, Harris EN. Antiphospholipid antibodies fromantiphospholipid syndrome patients activate endothelialcells in vitro and in vivo. Circulation 1999;99:1997–2002.

[4] Amengual O, Atsumi T, Khamashta MA. Tissue factor inantiphospholipid syndrome: shifting the focus from coagu-lation to endothelium. Rheumatology (Oxford) 2003;42:1029–31.

[5] Poredos P. Endothelial dysfunction in the pathogenesisof atherosclerosis. Clin Appl Thromb Hemost 2001;7:276–80.

[6] Vogel RA. Measurement of endothelial function by brachialartery flow-mediated vasodilation. Am J Cardiol 2001;88:31E–4E.

[7] Celermajer DS, Sorensen KE, Gooch VM, Spiegelhalter DJ,Miller OI, Sullivan ID, et al. Non-invasive detection ofendothelial dysfunction in children and adults at risk ofatherosclerosis. Lancet 1992;340:1111–5.

[8] Verma S, Buchanan MR, Anderson TJ. Endothelial functiontesting as a biomarker of vascular disease. Circulation2003;108:2054–9.

[9] Celermajer DS. Endothelial dysfunction: does it matter? Is itreversible? J Am Coll Cardiol 1997;30:325–33.

[10] Kaplanski G, Cacoub P, Farnarier C, Marin V, Gregoire R,Gatel A, et al. Increased soluble vascular cell adhesionmolecule 1 concentrations in patients with primary orsystemic lupus erythematosus-related antiphospholipidsyndrome: correlations with the severity of thrombosis.Arthritis Rheum 2000;43:55–64.

[11] Combes V, Simmon AC, Grau GE. In vitro generation ofendothelial microparticles and possible prothromboticactivity in patients with lupus anticoagulant. J Clin Invest1999;104:93–102.

[12] Dignat-George F, Camoin-Jau L, Sabatier F, Arnoux D,Anfosso F, Bardin N, et al. Endothelial microparticles: apotential contribution to the thrombotic complications ofthe antiphospholipid syndrome. Thromb Haemost 2004;91:667–73.

[13] Atsumi T, Khamashta MA, Haworth RS, Brooks G, AmengualK, Ichikawa K, et al. Arterial disease and thrombosis in theantiphospholipid syndrome: a pathogenic role for endothe-lin 1. Arthritis Rheum 1998;41:800–7.

[14] Forastiero RR, Martinuzzo ME, de Larranaga GF. Circulatinglevels of tissue factor and proinflammatory cytokines inpatients with primary antiphospholipid syndrome or lep-rosy related antiphospholipid antibodies. Lupus 2005;14:129–36.

Endothelial function is impaired in patients with primary antiphospholipid syndrome 461

[15] Williams FM, Parmar K, Hughes GR, Hunt BJ. Systemicendothelial cell markers in primary antiphospholipid syn-drome. Thromb Haemost 2000;84:742–6.

[16] Mercanoglu F, Erdogan D, Oflaz H, Kucukkaya R, Selcukbir-icik F, Gul A, et al. Impaired brachial endothelial function inpatients with primary anti-phospholipid syndrome. Int JClin Pract 2004;58:1003–7.

[17] Meroni PL, Borghi MO, Raschi E, Ventura D, Sarzi Puttini PC,Atzeni F, et al. Inflammatory response and the endotheli-um. Thromb Res 2004;114:329–34.

[18] Yasuda S, Atsumi T, Ieko M, Koike T. Beta2-glycoprotein I,anti-beta2-glycoprotein I, and fibrinolysis. Thromb Res2004;114:461–5.

[19] Eidelman RS, Hennekens CH. Fibrinogen: a predictor ofstroke and marker of atherosclerosis. Eur Heart J 2003;24:499–500.

[20] de Pablo P, Ramirez A, Cortina E, de la Pena A, Zamora J,Izaguirre R, et al. Increased fibrin polymerization rate inpatients with primary antiphospholipid syndrome andsystemic lupus erythematosus. Clin Appl Thromb Hemost2003;9:221–5.

[21] Stegnar M, Vene N, Bozic M. Do haemostasis activationmarkers that predict cardiovascular disease exist? Patho-physiol Haemost Thromb 2003;33:302–8.

[22] Vene N, Mavri A, Kosmelj K, Stegnar M. High D-dimer levelspredict cardiovascular events in patients with chronic atrialfibrillation during oral anticoagulant therapy. ThrombHaemost 2003;90:1163–72.

[23] Medina G, Casaos D, Jara LJ, Vera-Lastra O, Fuentes M,Barile L, et al. Increased carotid artery intima-mediathickness may be associated with stroke in primary antipho-spholipid syndrome. Ann Rheum Dis 2003;62:607–10.

[24] Ames PR, Margarita A, Sokoll KB, Weston M, Brancaccio V.Premature atherosclerosis in primary antiphospholipidsyndrome: preliminary data. Ann Rheum Dis 2005;64:315–7.

[25] Baron MA, Khamashta MA, Hughes GR, D’Cruz DP. Preva-lence of an abnormal ankle-brachial index in patients with

primary antiphospholipid syndrome: preliminary data. AnnRheum Dis 2005;64:144–6.

[26] O’Driscoll G, Green D, Taylor RR. Simvastatin, an HMG-coenzyme A reductase inhibitor improves endothelialfunction within 1 month. Circulation 1997;95:1126–31.

[27] Egashira K, Hirooka Y, Kai H, Sugimachi M, Suzuki S, Inou T,et al. Reduction in serum cholesterol with pravastatinimproves endothelium-dependent coronary vasomotion inpatients with hypercholesterolemia. Circulation 1994;89:2519–24.

[28] Sebestjen M, Zegura B, Keber I. Both cerivastatin andfenofibrate improve arterial vasoreactivity in patients withcombined hyperlipidaemia. J Intern Med 2002;251:77–85.

[29] Mancini GB, Henry GC, Macaya C, O’Neill BJ, Pucillo AL,Carere RG, et al. Angiotensin-converting enzyme inhibitionwith quinapril improves endothelial vasomotor dysfunctionin patients with coronary artery disease. The TREND (Trialon Reversing ENdothelial Dysfunction) Study. Circulation1996;94:258–65.

[30] Meroni PL, Raschi E, Testoni C. Endothelium as a target foranti-phospholipid antibodies and for therapeutical inter-vention. Autoimmun Rev 2002;1:55–60.

[31] Meroni PL, Raschi E, Testoni C, Tincani A, Balestrieri G,Molteni R, et al. Statins prevent endothelial cell activationinduced by antiphospholipid (anti-beta2-glycoprotein I)antibodies: effect on the proadhesive and proinflammatoryphenotype. Arthritis Rheum 2001;44:2870–8.

[32] Ferrara DE, Swerlick R, Casper K, Meroni PL, Vega-OstertagME, Harris EN, et al. Fluvastatin inhibits up-regulation oftissue factor expression by antiphospholipid antibodies onendothelial cells. J Thromb Haemost 2004;2:1558–63.

[33] Meroni PL, Luzzana C, Ventura D. Anti-inflammatory andimmunomodulating properties of statins. An additional toolfor the therapeutic approach of systemic autoimmunediseases? Clin Rev Allergy Immunol 2002;23:263–77.