I

Togapasw

M

THihl

cHSAUv

0d

Anti-Cardiolipin Antibodies and Endothelial Function in PatientsWith Coronary Artery Disease

brahim Marai, MDa,d, Michael Shechter, MDa,d, Pnina Langevitz, MDb,d, Boris Gilburd, MD, PhDc,d,Ardon Rubenstein, MDe, Eiji Matssura, PhDg, Yaniv Sherer, MDc,d, and Yehuda Shoenfeld, MDc,d,f,*

Endothelial dysfunction is considered an important marker in atherosclerosis, having aprognostic value. Antiphospholipid antibodies are considered prothrombotic and haverecently been reported to be associated also with atherosclerosis. This study was conductedto investigate a possible association of endothelial dysfunction with various antiphospho-lipid autoantibodies in healthy subjects and patients with cardiovascular disease. In asingle-center, prospective study, 2 groups were included. The study group included patientswith cardiovascular diseases (coronary disease and/or cerebrovascular disease) and healthysubjects without apparent heart disease who were referred to the endothelial functionlaboratory for the assessment of endothelial function. Flow-mediated dilatation, whichindicates endothelial function, and nitroglycerin-mediated vasodilatation, which indicatessmooth-muscle function, were measured. The 2 groups were evaluated for autoantibodies,including anticardiolipin (aCL; immunoglobulin G [IgG], immunoglobulin M [IgM], andimmunoglobulin A [IgA]), antinuclear antibody, anti–�2-glycoprotein I (IgG, IgM, andIgA), and oxidized low-density lipoprotein. One hundred seven subjects were included inthe study: 45 patients (42%) and 62 healthy controls (58%). Flow-mediated dilatation wassignificantly lower in patients compared with healthy controls (8.0 � 9.5% vs 8.0 � 13.5%,p � 0.012). In addition, nitroglycerin-mediated vasodilatation was nonsignificantly lowerin patients than in healthy controls (8.0 � 13.4% vs 11.0 � 16.7%, p � 0.084). The meanlevels of anti–�2-glycoprotein I (IgG, IgM, and IgA), aCL (IgM and IgA), antinuclearantibody, and oxidized low-density lipoprotein were not different between groups. How-ever, the mean level of IgG aCL was significantly higher in patients than in healthycontrols. In conclusion, in accordance with previous reports of an association between aCLand atherosclerosis, patients with cardiovascular disease had endothelial dysfunction andelevated levels of aCL. © 2008 Elsevier Inc. All rights reserved. (Am J Cardiol 2008;101:

1094–1097)

fcwmpdccadtaaidr(

tlBdwa

he aim of the present study was to explore the associationsf autoantibodies, including anticardiolipin (aCL), anti–�2-lycoprotein I (�2GPI), antinuclear antibody (ANA), andnti–oxidized low-density lipoprotein (oxLDL), which areresent in high prevalence in autoimmune diseases, mainlyntiphospholipid syndrome and systemic lupus erythemato-us, and endothelial function in cardiovascular patientsithout apparent autoimmune disease.

ethods and Results

his was a single-center (Sheba Medical Center, Telashomer, Israel), prospective study. The study population

ncluded 2 groups: patients with cardiovascular diseases andealthy controls. All patients were referred to the endothe-ial function laboratory for the assessment of endothelial

aHeart Institute, bRheumatic Disease Unit, and cDepartment of Medi-ine B and Center of Autoimmune Diseases, Sheba Medical Center, Telashomer; dSackler Faculty of Medicine, eMetabolic Unit, Tel Avivourasky Medical Centre, and fLaura Schwarz-Kipp Chair for Research inutoimmunity, Tel Aviv University, Tel Aviv, Israel; and gOkayamaniversity, Okayama, Japan. Manuscript received October 14, 2007; re-ised manuscript received and accepted December 9, 2007.

*Corresponding author: Tel: 972-3-5302652; fax: 972-3-5352855.

iE-mail address: shoenfel@post.tau.ac.il (Y. Shoenfeld).

002-9149/08/$ – see front matter © 2008 Elsevier Inc. All rights reserved.oi:10.1016/j.amjcard.2007.12.010

unction. Consecutive subjects who fulfilled the inclusionriteria were enrolled in the study. The inclusion criteriaere as follows: (1) for healthy subjects, no histories ofyocardial infarction, coronary bypass, coronary angiogra-

hy with angioplasty and/or stenting, cerebrovascular acci-ent, or peripheral vascular disease; and (2) for patients withardiovascular disease, histories of myocardial infarction,oronary bypass, coronary angiography with angioplastynd/or stenting, heart failure secondary to coronary arteryisease, or cerebrovascular accident, without apparent au-oimmune disease (i.e., systemic lupus erythematosus orntiphospholipid syndrome). The exclusion criteria were (1)trial fibrillation; (2) sinus bradycardia without pacemakermplantation, sick sinus syndrome, or secondary or third-egree atrioventricular block; (3) intolerance to nitrates; (4)enal failure with creatinine �4 mg/dl or liver failure; and5) drug and/or alcohol addiction.

After clinical histories and physical examinations, endo-helial function was assessed according to recently pub-ished American College of Cardiology guidelines.1,2

riefly, endothelial function in the form of endothelium-ependent brachial artery flow-mediated dilatation (FMD)as measured as previously described. Briefly, FMD was

ssessed in the subject’s right arm in the recumbent position

n a temperature-controlled room (22°C), after a 10-minute

www.AJConline.org

etsnwtWskoum

dassttsnatwda

bAtsuwtmtdccaasCbVbpvtal

wmgaTmbwe

twa�I

Stfcm

65

TB

P

ABTL

TH

HH

GSDPMBFN

TB

P

WHHSDFUUU

UU

U

d

1095Coronary Artery Disease/Anti-Cardiolipin and Endothelial Function

quilibration period, by a single ultrasonographer blinded toreatment assignment. Using 15.6-MHz linear-array ultra-ound (Hewlett-Packard SONOS 5500 CV; Agilent Tech-ologies, Inc., Andover, Massachusetts), the brachial arteryas longitudinally imaged approximately 5 cm proximal to

he antecubital crease, where the clearest image was noted.hen a reasonable image was obtained, the surface of the

kin was marked, and the arm and the ultrasound probe wereept in the same position by the ultrasonographer through-ut the study. An electrocardiogram was monitored contin-ously, and blood pressure was taken in the left arm everyinute throughout the study.After a 2-minute baseline period, a frozen 3-cm longitu-

inal image of the vessel without color flow was obtainednd frozen for 5 seconds. The image was then unfrozen andwitched to pulsed-wave Doppler for 5 seconds at a sweeppeed of 50 mm/s. A pneumatic tourniquet placed aroundhe forearm proximal to the target artery was inflated afterhe baseline phase to a pressure of 50 mm Hg higher than theubject’s systolic blood pressure (or until no blood flow wasoticed through the brachial artery by the Doppler probe),nd this pressure was held for 5 minutes. Increased flow washen induced by sudden cuff deflation. A continuous scanas performed at cuff deflation and 60 and 90 seconds aftereflation, with frozen and Doppler measurements recordedt similar intervals to the baseline phase.

Thirteen minutes after cuff deflation, a second 2-minuteaseline rest scan was recorded to confirm vessel recovery.fter the administration of a sublingual nitroglycerin (NTG)

ablet (Nitrostat 0.4 mg; Pfizer, Inc., New York, New York),canning was performed continuously for 5 minutes. Theltrasound images were recorded on an S-VHS videotapeith an SLV-RS7 videocassette recorder (Sony Corpora-

ion, Tokyo, Japan). The diameter of the brachial artery waseasured from the anterior to the posterior interface be-

ween the media and adventitia (the “m line”) at a fixedistance. The mean diameter was calculated from 4 cardiacycles synchronized with the R-wave peaks on the electro-ardiogram. All measurements were made at end-diastole tovoid possible errors caused by variable arterial compli-nce. The internal diameter was calculated using Prosoundoftware (University of Southern California, Los Angeles,alifornia) using a Horita data translation image processingoard (DT2862-60Hz; Horita Company, Inc., Missioniejo, California). The percentage diameter change causedy endothelium-dependent FMD and the endothelium-inde-endent percentage change from baseline in NTG-mediatedasodilatation are expressed as the percentage change rela-ive to that at the initial rest scan. The intraobserver vari-bility for repeated measurements is 0.0 � 0.07 mm in ouraboratory.

After the endothelial function study was completed,hile the subjects were still fasting, blood samples for theeasurement of aCL (immunoglobulin G [IgG], immuno-

lobulin M [IgM], and immunoglobulin A [IgA]), ANA,nti-�2GPI (IgG, IgM, and IgA), and oxLDL were taken.he blood samples were centrifuged immediately for 15inutes at 3,000 rpm. The serum was stored at �20°C. All

lood samples were tested at the end of the study. Every testas done in the same laboratory by the same operator using

nzyme-linked immunosorbent assay. The cut-off values for 4

he tested antibodies (determined using normal controls)ere as follows: for aCL, IgG �11 IU/ml, IgM �10 IU/ml,

nd IgA �13 IU/ml; for anti-�2GPI, IgG �20 IU/ml, IgM10 IU/ml, and IgA �10 IU/ml; ANA �10 IU/ml; and 26

U/ml �oxLDL �105 IU/ml.All clinical and laboratory data were summarized using

PSS version 11 (SPSS, Inc., Chicago, Illinois). Chi-squareests were used for categorical parameters, Student’s t testsor comparisons of the 2 groups, and linear regression forontinuous parameters. Statistical significance was deter-ined at p �0.05.One hundred seven subjects were included in the study,

7 (63%) men and 40 (37%) women. The average age was7 � 11 years (range 17 to 81). The patient group included

able 1aseline data in healthy controls and patients

arameterControls(n � 64)

Patients(n � 46) p Value

ge (yrs) 55 � 11 60 � 12 0.023ody mass index (kg/m2) 27 � 4 27 � 3 0.936otal cholesterol (mg/dl) 209 � 35 179 � 39 0.000ow-density lipoprotein

cholesterol (mg/dl)132 � 28 107 � 29 0.000

riglycerides (mg/dl) 123 � 71 160 � 96 0.023igh-density lipoproteincholesterol (mg/dl)

52 � 14 41 � 9 0.000

omocysteine (�mol/L) 13 � 3 14 � 6 0.065igh-sensitivity C-reactive protein(mg/L)

7 � 4 10 � 5 0.716

lucose (mg/dl) 95 � 21 120 � 52 0.002ystolic blood pressure (mm Hg) 137 � 17 140 � 23 0.480iastolic blood pressure (mm Hg) 79 � 8 79 � 10 0.773ulse pressure (mm Hg) 58 � 13 61 � 20 0.305ean arterial pressure (mm Hg) 99 � 10 99 � 13 0.797aseline brachial artery (mm) 5.43 � 1.00 6.11 � 1.00 0.001MD 13.5 � 8.0 9.5 � 8.0 0.012TG-mediated vasodilatation 16.7 � 11.0 13.4 � 8.0 0.084

able 2aseline clinical data in healthy controls and patients

arameterControls(n � 64)

Patients(n � 46) p Value

omen/men 34 (53%)/30 (47%) 8 (17%)/38 (83%) 0.000ypertension* 23 (36%) 22 (48%) 0.242ypercholesterolemia 30 (47%) 25 (54%) 0.562moking 9 (14%) 8 (17%) 0.790iabetes mellitus 4 (6%) 11 (24%) 0.011amily history 36 (56%) 16 (35%) 0.034se of aspirin 25 (39%) 26 (57%) 0.083se of statins 20 (31%) 27 (59%) 0.006se of calciumchannel blockers

8 (13%) 6 (13%) 1

se of diuretics 7 (11%) 13 (28%) 0.025se of angiotensin-converting enzymeinhibitors

13 (20%) 15 (33%) 0.184

se of � blockers 10 (16%) 19 (41%) 0.004

* Hypertension was defined as systolic blood pressure �140 mm Hg oriastolic blood pressure �90 mm Hg.

5 patients (42%), and the control group included 62 sub-

jasjms

1lp

abattiwwcacwa

D

IalAsiastbarhhp(swvapdtwmcMmipsh

edowmgp

TT

A

IIIAAAAo

TCa

A

IIIAAAAo

TPh

A

I

I

I

A

o

li

s

1096 The American Journal of Cardiology (www.AJConline.org)

ects (58%). The clinical and laboratory data for the patientnd healthy groups are listed in Tables 1 and 2. FMD wasignificantly lower in patients compared with healthy sub-ects (9.5 � 8.0% vs 13.5 � 8.0%, p � 0.012). NTG-ediated vasodilatation in patients was less than in healthy

able 3iters of antibodies in healthy controls and patients

ntibody (IU/ml)Controls(n � 62)

Patients(n � 45) p Value

gG aCL 1.7 � 1.0 2.6 � 2.0 0.000gM aCL 7 � 8 6 � 5 0.303gA aCL 4 � 2 4 � 2 0.156nti-�2GPI IgG 4 � 2 6 � 16 0.425nti-�2GPI IgM 4 � 4 3 � 3 0.215nti-�2GPI IgA 16 � 51 9 � 19 0.358NA 4 � 3 5 � 6 0.241xLDL 35 � 17 36 � 20 0.748

able 4orrelation of percentage flow-mediated dilatation with levels ofntibodies in all study subjects (healthy controls and patients)

ntibody (IU/ml)

FMD

r p Value

gG aCL �0.006 0.952gM aCL 0.024 0.807gA aCL �0.066 0.496nti-�2GPI IgG �0.063 0.518nti-�2GPI IgM �0.049 0.617nti-�2GPI IgA 0.016 0.869NA 0.145 0.135xLDL 0.154 0.113

able 5ercentage flow-mediated dilatation in all study subjects (patients andealthy controls) with normal and high levels of antibodies*

ntibody No. FMD p Value

gM aCL 0.624Normal 92 11.4 � 8.0High 15 12.5 � 7.0

gM anti-�2GPI 0.688Normal 97 11.6 � 8.0High 10 10.5 � 6.0

gA anti-�2GPI 0.755Normal 88 11.6 � 9.0High 19 11 � 6.0NA 0.424Normal 101 11.4 � 8.0High 6 14 � 8.0xLDL† 1Normal 24 12.2 � 11High 80 10.9 � 7

* The antibodies IgG aCL, IgA aCL (there were no subjects with highevels), and IgG anti-�2GPI (only 1 subject with a high level) are notncluded in the table.

† Subjects with high levels of oxLDL were excluded because of theirmall number (n � 3).

ubjects, but not statistical significantly (13.4 � 8.0% vs d

6.7 � 11.0%, p � 0.084; Table 1). FMD was significantlyower in men than in women (10.4 � 8.0% vs 14.0 � 9.0%,� 0.024).The average levels of anti-�2GPI (IgG, IgM, and IgA),

CL (IgM and IgA), ANA, and oxLDL were not differentetween healthy controls and patients (Table 3). The aver-ge level of IgG aCL was significantly higher in patientshan in healthy controls. There were no differences betweenhe 2 groups when the levels of antibodies were classifiednto 2 categories, normal and high. In addition, oxLDL,hich was classified into 3 groups (low, normal, and high)as not different between healthy controls and patients. No

orrelations were found between FMD and the levels of thentibodies in all study subjects (Table 4). In addition, noorrelations were detected between FMD and the antibodieshen we compared FMD in subjects with normal levels of

ntibodies and those with high levels of antibodies (Table 5).

iscussion

n this study, we examined various autoantibodies and theirssociations with endothelial function. There was no corre-ation between the inflammatory markers (aCL, anti-�2GPI,NA, and oxLDL) and endothelial function in all study

ubjects (healthy controls and patients). The titers of thesenflammatory markers were not different between patientsnd healthy subjects, except for IgG aCL; its levels wereignificantly higher in patients. The lack of correlation be-ween antibodies and endothelial function in this study cane explained by the fact that most of the healthy controlsnd patients had levels of autoantibodies within the normalange. The number of healthy subjects and patients who hadigh levels of antibodies was small. In addition, no patientsad autoimmune inflammatory diseases such as antiphos-holipid syndrome and systemic lupus erythematosuswhich involve a high prevalence of antibodies and athero-clerosis).3 All patients were in stable condition, and it isell known that levels of inflammatory markers in cardio-ascular patients with unstable conditions or acute eventsre higher than in patients with stable disease. Liuzzo et al4reviously showed that C-reactive protein levels are higheruring index admission in patients with unstable anginahan in patients with stable angina. Serial studies in patientsith unstable angina have demonstrated that inflammatoryarker (C-reactive protein and interleukin-6) levels de-

rease by approximately 50% in the following weeks.5oreover, statins can enhance the decrease in inflammatoryarkers in patients with acute coronary syndromes.5 Thus,

t is possible that the titers of inflammatory markers inatients were not higher than in healthy controls because oftatin treatment, which was used more by patients than byealthy subjects.

Another finding in this study entails the significantlylevated levels of aCL in patients with coronary arteryisease compared with healthy subjects. We have previ-usly reported in a few studies that elevated levels of aCLere associated with enhanced atherosclerosis in variouseasures evaluated in the general population: coronary an-

iography and calcium score calculated using helical com-uted tomography.6,7 Another recent study by our group

isclosed enhanced early carotid atherosclerosis in patients

wllfta

1

2

3

4

5

6

7

8

1097Coronary Artery Disease/Anti-Cardiolipin and Endothelial Function

ith rheumatoid arthritis who had medium to high elevatedevels of aCL.8 Hence, our finding of significantly elevatedevels of aCL in patients with abnormal endothelial dys-unction is not surprising but rather reassuring with regardo the possible use of aCL as a marker of various forms oftherosclerosis.

. Corretti MC, Anderson TJ, Benjamin EJ, Celermajer D, Charbonneau F,Creager MA, Deanfield J, Drexler H, Gerhard-Herman M, Herrington D, etal; International Brachial Artery Reactivity Task Force. Guidelines for theultrasound assessment of endothelial-dependent flow-mediated vasodilata-tion of the brachial artery. J Am Coll Cardiol 2002;39:257–265.

. Shechter M, Matetzky S, Feinberg MS, Chouraqui P, Rotstein Z, HodH. External counterpulsation therapy improves endothelial function inpatients with refractory angina pectoris. J Am Coll Cardiol 2003;42:2090–2095.

. Sherer Y, Shoenfeld Y. Antiphospholipid syndrome, antiphospholipidantibodies, and atherosclerosis. Curr Atheroscler Rep 2001;3:328–333.

. Liuzzo G, Biasucci LM, Gallimore JR, Grillo RL, Rebuzzi AG, Pepys

MB, Maseri A. The prognostic value of C-reactive protein and serum

amyloid a protein in severe unstable angina. N Engl J Med 1994;331:417–424.

. Kinlay S, Schwartz GG, Olsson AG, Rifai N, Leslie SJ, Sasiela WJ,Szarek M, Libby P, Ganz P. Myocardial Ischemia Reduction WithAggressive Cholesterol Lowering Study Investigators. High-dose ator-vastatin enhances the decline in inflammatory markers in patients withacute coronary syndromes in the MIRACL study. Circulation 2003;108:1560–1566.

. Sherer Y, Tenenbaum A, Praprotnik S, Shemesh J, Blank M, FismanEZ, Harats D, George J, Levy Y, Peter JB, et al. Coronary artery diseasebut not coronary calcification is associated with elevated levels ofcardiolipin, beta-2-glycoprotein-I, and oxidized LDL antibodies. Car-diology 2001;95:20–24.

. Sherer Y, Shemesh J, Tenenbaum A, Praprotnik S, Harats D, FismanEZ, Blank M, Motro M, Shoenfeld Y. Coronary calcium and anti-cardiolipin antibody are elevated in patients with typical chest pain.Am J Cardiol 2000;86:1306–1311.

. Sherer Y, Gerli R, Gilburd B, Bartoloni Bocci E, Vaudo G, MannarinoE, Shoenfeld Y. Thickened carotid artery intima-media in rheumatoidarthritis is associated with elevated anti-cardiolipin antibodies. Lupus

2007;16:259–264.