ATVB In FocusChemokines in Atherosclerosis, Thrombosis, and Vascular Biology

Series Editor: Christian Weber

Chemokines and Cardiovascular RiskPål Aukrust, Bente Halvorsen, Arne Yndestad, Thor Ueland, Erik Øie,

Kari Otterdal, Lars Gullestad, Jan K. Damås

Abstract—Based on the importance of inflammation in atherogenesis, recent work has focused on whether plasma markersof inflammation can noninvasively diagnose and prognosticate atherosclerotic disorders. Although several studiessupport an important pathogenic role of chemokines in atherosclerosis, potentially representing attractive therapeutictargets in atherosclerotic disorders, this does not necessarily mean that chemokines are suitable parameters for riskprediction. In fact, the ability to reflect upstream inflammatory activity, stable levels in individuals, and high stabilityof the actual protein (eg, long half-life and negligible circadian variation) are additional important criteria for an idealbiomarker in cardiovascular disease. Although plasma/serum levels of certain chemokines (eg, interleukin- 8/CXCL8and monocyte chemoattractant protein-1/CCL2) have shown to be predictive for future cardiac events in some studies,their role as clinical biomarkers is unclear, and their ability to predict subclinical atherosclerosis has been disappointing.Further prospective studies, including a larger number of patients, are needed to make any firm conclusion. Based onthe participation of several chemokines in atherogenesis, it is possible that in the future, combined measurements ofmultiple chemokines could reveal as a “signature of disease” that can serve as a highly accurate method to assess forthe presence of atherosclerotic disease. (Arterioscler Thromb Vasc Biol. 2008;28:1909-1919)

Key Words: chemokines � atherosclerosis � biomarkers � inflammation

A decade ago, lipid lowering therapy was expected toeliminate coronary artery disease (CAD) and other

forms of atherosclerosis by the end of the 20th century.However, that optimistic prediction has needed revision andcardiovascular diseases are expected to be the main cause ofdeath globally within the next 15 years, at least partly owingto a rapidly increasing prevalence in developing countries andEastern Europe and the rising incidence of obesity anddiabetes in the Western world.1 At present, cardiovasculardiseases cause 38% of all deaths in North America and are themost common cause of death in European men under 65 yearsof age and the second most common cause in women. Thesefacts force us to revisit cardiovascular disease and considernew strategies for prevention and treatment as well as fordisease and risk prediction. Commonly used risk algorithms,such as the Framingham Risk Score, and lipid parameters failto identify all affected individuals.1,2 Novel cardiovascularrisk factors that identify these missed individuals wouldgreatly improve overall care of patients. These risk markers

should also give prognostic information in patients withestablished cardiovascular disease as well as being predictorsof efficacy for various therapeutic interventions (eg, statinsand different forms of antithrombotic medications).1,2 Ideally,new risk markers should also reflect important pathogenicprocesses in atherogenesis and plaque destabilization, and theattempt to identify new biomarkers in atherosclerotic disor-ders is therefore linked to the study of the pathogenicmechanisms of atherosclerosis.

The Role of Chemokines in AtherogenesisAtherosclerosis is a chronic disease characterized by twofundamental hallmarks: lipid accumulation and inflamma-tion.1–3 The interaction between these two processes definesthe principal pathogenesis and distinguishes atherosclerosisfrom other chronic inflammatory disorders. Although theconcept that atherosclerosis is an inflammatory disease is nolonger controversial, the regulation of these inflammatoryprocesses as well as their pathogenic consequences is not

Original received April 21, 2008; final version accepted July 14, 2008.From the Research Institute for Internal Medicine (P.A., B.H., A.Y., T.U., E.Ø., KJ.O., J.K.D.), the Section of Clinical Immunology and Infectious

Diseases (P.A., J.K.D.), the Section of Endocrinology (T.U.), and the Department of Cardiology (E.Ø., L.G.), Rikshospitalet University Hospital,University of Oslo, Norway.

Correspondence to Pål Aukrust, Section of Clinical Immunology and Infectious Diseases, Medical Department, Rikshospitalet University Hospital,University of Oslo, Sognsvannsveien 20, 0027 Oslo, Norway. E-mail pal.aukrust@rikshospitalet.no

© 2008 American Heart Association, Inc.

Arterioscler Thromb Vasc Biol is available at http://atvb.ahajournals.org DOI: 10.1161/ATVBAHA.107.161240

1909

by guest on May 17, 2018

http://atvb.ahajournals.org/D

ownloaded from

fully understood. Moreover, although the participation ofinflammation in atherogenesis has become widely recog-nized, the identification and characterization of the differentactors and their relative importance are not fulfilled. How-ever, several lines of evidence suggest that chemokines areimportant actors in this immune-mediated process, leading toprogression of atherosclerosis and plaque destabilization(Figures 1 and 2).4,5

Chemokines are inflammatory cytokines characterized bytheir ability to cause directed migration of leukocytes intoinflamed tissue, and raised levels are found in atherosclerosis,both systemically and within the atherosclerotic plaques.4,5

Hence, there are several reports of enhanced expression ofCXC-chemokines (eg, IL-8 [IL-8/CXCL8], neutrophil-

activating peptide-2 [NAP-2/CXCL7], growth-relatedoncogene-� [GRO-�/CXCL1], interferon [INF]-�–inducible10 [IP-10/CXCL10], and CXCL16), CC-chemokines (eg,monocyte chemoattractant protein-1 [MCP-1/CCL2],leukotactin-1 [Lkn-1/CCL15], regulated on activation, nor-mal T cell expressed and secreted [RANTES/CCL5], CCL19,and CCL21) within human atherosclerotic lesions,6–14 andthere are also some studies suggesting that these chemokinesare colocalized with enhanced expression of their correspond-ing receptors.7,13,14 The role of chemokines in atherosclerosisis further supported by several studies showing that modifiedLDL particles are potent inducers of chemokines in variouscells such as macrophages and vascular smooth muscle cells(SMC), suggesting that chemokines may represent a link

Figure 1. The pathogenic role of chemokines in atherogenesis. Chemokines promote recruitment and migration of T cells and mono-cytes into to the atherosclerotic lesion. Within the plaque, chemokines induce activation of endothelial cell and different leukocyte sub-sets (eg, T cells) with subsequent release of inflammatory cytokines and chemokines, which in turn will further promote recruitment andactivation of leukocytes into the lesion. For T cell activation, the interaction between CXCR3 and their ligands (MIG, IP-10, and I-TAC)seems to be of particular importance, promoting an inflammatory T cell phenotype, which in turn will promote macrophage activation.Chemokines are also potent activators of macrophages in a more direct way, inducing not only enhanced release of inflammatory cyto-kines, but also increased production of tissue factor (TF), matrix metalloproteinases (MMP), and reactive oxygen species (ROS), andsome chemokines may also enhance lipid loading and foam cell formation (eg, GRO-�) through their ability to upregulate scavengerreceptors. These events will transform macrophages into an inflammatory, matrix degrading, procoagulant, and apoptosis-inducingphenotype. The interaction between MCP-1/CCR2 appears to be of particular importance for the promotion of an inflammatory mono-cyte/macrophage phenotype. Beyond their effects on T cells and macrophages, chemokines also interfere with vascular smooth musclecells (SMCs). Thus, while chemokines may promote vascular SMC migration into the lesion, an early event in atherogenesis, they couldalso transform these cells from a contractile to a proliferative/secretory phenotype, a hallmark of the vascular remodeling characterizingatherosclerosis. Taken together, chemokines may act on several levels within the lesion, contributing to various pathogenic loops beingimportant actors in the inflammatory arm of atherogenesis. GRO-� indicates growth-related oncogene-�/CXCL1; NAP-2, neutrophil-activating peptide-2/CXCL7; IL-8/CXCL8; MIG, monokine induced by IFN-�/CXCL9; IP-10, INF-�-inducible 10/CXCL10; I-TAC, INF-inducible T cell �/CXCL11; MCP-1, monocyte chemoattractant protein-1/CCL2; fractalkine/CX3CL1.

1910 Arterioscler Thromb Vasc Biol November 2008

by guest on May 17, 2018

http://atvb.ahajournals.org/D

ownloaded from

between lipids and inflammation in atherogenesis.15,16 Inaddition to being potent chemoattractants, several otherleukocyte/macrophage responses such as cell proliferation,enzyme secretion, induction of reactive oxygen species, andpromotion of foam cell formation have been observed in vitroafter chemokine stimulation.17–19 Moreover, beyond theireffects on leukocytes, chemokines may also interfere withSMC migration and growth as well as platelet activation.19–21

Some of these responses may clearly be relevant to athero-genesis and plaque destabilization, and the importance ofchemokines in these processes is supported by several studiesin gene-modified mice. Thus, targeted disruption of the genesfor CCL2, CCR2 (ie, CCL2 receptor), CXCR2 (ie, CXCL1,CXCL7, and CXCL8 receptor), CXCR6 (ie, CXCL16 recep-tor), and CX3CR1 (ie, fractalkine/CX3CL1 receptor) signif-

icantly decreases atherosclerotic lesion formation and lipiddeposition in mice prone to develop atherosclerotic-likelesions.22–27 Furthermore, deletion of CX3CL1 in CCR2�/�

ApoE�/� mice dramatically reduces the development of ath-erosclerosis, providing in vivo evidence for independent rolesfor CCR2 and CX3CL1 in atherogenesis,28 indicating thatsuccessful therapeutic strategies may need to target multiplechemokines or chemokine receptors. This view was sup-ported by Combadiere and coworkers showing that combinedinhibition of CCL2, CX3CR1, and CCR5 in ApoE-deficientmice leads to an additive reduction in atherosclerosis.29

However, the role of chemokines in atherogenesis is far fromcompletely understood. In fact, studies in gene-modified miceindicate antiatherogenic rather than proatherogenic effects ofsome of these mediators (eg, deficiency in CXCL16 or CCR1

Figure 2. The pathogenic role of chemokines in plaque destabilization. Infiltration and activation of leukocyte subsets into the athero-sclerotic plaque play an important role in triggering of acute coronary syndromes (ACS) with chemokines as important actors. Thesechemotactic cytokines may promote not only by recruiting activated leukocytes into the lesion, but also by directly contributing toplaque rupture and thrombus formation through mechanisms such as enhancement of the matrix degrading potential in macrophagesand inducing tissue factor (TF) in vascular SMCs and endothelial cells. Chemokines could also promote plaque rupture by inducing oxi-dative stress and apoptosis in endothelial cells and vascular SMCs within the lesion. Moreover, although the role of platelets in throm-bus formation is well established, platelet-mediated inflammation could also contribute to the development of ACS. Activated platelets,in contact with an inflammatory endothelium, will release several chemokines (eg, platelet factor-4, RANTES, GRO-�, ENA-78, andNAP-2), which will promote inflammatory responses in neighboring leukocytes and endothelial cells. Notably, platelets may themselvesrespond to chemokines that are released from these adjacent cells, through interaction with platelet-expressed chemokine receptors,representing a pathogenic loop in plaque destabilization, involving chemokine-mediated inflammation and thrombus formation. GRO-�indicates growth-related oncogene-�/CXCL1; NAP-2, neutrophil-activating peptide-2/CXCL7; platelet factor 4/CXCL4; ENA-78, epithe-lial neutrophil activating protein 78/CXCL5; MIP-1�, macrophage inflammatory peptide-1�/CCL3; RANTES, regulated on activation, nor-mal T cell expressed and secreted/CCL5; �-TG, �-thromboglobulin; MMP, matrix metalloproteinases.

Aukrust et al Chemokines in Cardiovascular Risk Prediction 1911

by guest on May 17, 2018

http://atvb.ahajournals.org/D

ownloaded from

accelerates atherosclerosis),30,31 suggesting that chemokinesalso may exert atheroprotective properties, at least whenoperating at a physiological level. Moreover, studies in genemodified mice should be interpreted with some caution,focusing on atheroclerotic lesion in aorta and not in thecoronary circulation. Also, the total lack of one particularchemokine or chemokine receptor may not necessarily giveinsight into the situation in CAD patients with moderatelyupregulation of these mediators.

The Role of Chemokines inPlaque Destabilization

Infiltration and activation of leukocytes into the atheroscle-rotic plaque may also be involved in triggering of acutecoronary syndromes (ACS).1–3 Again, chemokines may playan important role in this immune-mediated plaque destabili-zation, not only by recruiting activated leukocytes into thelesion, but also by directly contributing to plaque rupture andthrombus formation through mechanisms such as enhance-ment of the matrix degrading potential in macrophages,inducing tissue factor in vascular SMCs and endothelial cells,and by inducing oxidative stress and apoptosis within theatherosclerotic lesion.18,32,33 Moreover, although the role ofplatelets in thrombus formation is well established, platelet-mediated inflammation could also contribute to the develop-ment of ACS.21,34,35 Thus, by releasing a wide range ofinflammatory mediators on activation, including several che-mokines (eg, CCL5, CXCL1, epithelial neutrophil activatingpeptide-78 [ENA-78/CXCL5], and CXCL7), platelets maypromote inflammatory responses in neighboring leukocytesand endothelial cells, and notably, platelets may themselvesrespond to inflammatory mediators that are released fromthese adjacent cells, at least partly through interaction withplatelet-expressed chemokine receptors.21,35 Consequently,chemokines could be identified as potential important medi-ators not only in atherogenesis, but also in plaque destabili-zation (Figure 2). Such a notion was recently supported in astudy by Lutgens et al showing that inhibiting CCL2 andMCP-5/CCL12 induced a stable plaque phenotype inApoE�/� mice.36 However, although there are some reportssuggesting that ApoE�/� and LDL receptor (LDLR)�/� micemay spontaneously develop plaques in certain parts of thearterial tree showing features suggestive of plaque rupture,there is an urgent need for representative animal modelswhere prospective examination of the events leading up toplaque rupture and the rupture process itself can beperformed.37

Inflammatory Mediators as Biomarkers inCardiovascular Disease

Despite the important role of cholesterol in atherosclerosis,many individuals who experience myocardial infarction (MI)have cholesterol concentrations at or below recommendedlevels.1,2 Moreover, many patients who present with acute MIare receiving drug therapy for dyslipidemia, despite LDLconcentrations at currently mandated targets or below.38 Thisconvergence of clinical findings highlights the necessity ofimproving our ability to predict cardiovascular risk. Inconsideration of the important role of inflammation in athero-

genesis, recent work has focused on whether plasma markersof inflammation can noninvasively diagnose and prognosti-cate CAD and other forms of atherosclerosis.39,40 C-reactiveprotein (CRP), the prototype inflammatory marker, has cer-tainly been the best studied of these “new” biomarkers. As adownstream marker, CRP provides functional integration ofupstream cytokine activation, and CRP has proven remark-ably robust as a marker of cardiovascular risk and givespredictive value beyond that of traditional risk factors inapparently healthy individuals as well as in CAD pa-tients.41–43 Also, in patients with ACS, CRP levels seem togive prognostic information on mortality.44 Finally, measur-ing CRP in CAD patients may be useful in monitoringresponses to various intervention such as statin therapy.45

However, although CRP is a stable and reliable marker ofinflammation and may mirror several aspects of the immu-nopathogenic mechanisms in CAD, the inflammatory pro-cesses that underlie atherosclerosis are mediated by a multi-tude of cytokines and are unlikely to be reflected by CRPlevels alone.46 Thus, CRP seems not to predict early or lateoccurrence of MI in patients with ACS.44 Moreover, otherinflammatory mediators have also been evaluated as biomar-kers in CAD, and some of these have been found to giveprognostic information beyond that of CRP (eg, IL-6, IL-18,intracellular adhesion molecule [ICAM]-1, osteoprotegerin,and soluble CD40 ligand [sCD40L]).47–51 Finally, althoughsome in vitro studies have claimed that human CRP hasproatherogenic effects, the pathogenic role of CRP in athero-genesis is at least controversial.52

What Is a Reliable Biomarker inCardiovascular Disease?

A biomarker for cardiovascular disease should reflect impor-tant pathophysiological processes in atherogenesis and plaquedestabilization, and some biomarkers have failed becausethey are involved in only one pathway in a multiple-pathwaydisease or they reflect epiphenomena independent of thedisease process. However, although chemokines seem to playa central pathogenic role in atherogenesis, this does notnecessarily mean that chemokines are suitable parameters forrisk prediction. In fact, the leading role of CRP as aninflammatory biomarker in cardiovascular disease is notprimarily based on its pathogenic role in these disorders, butrather on its ability to reflect upstream inflammatory activity.The fact that several different inflammatory markers withdifferent biological activities contribute to the statistical riskfor CAD makes it unlikely that CRP or any of the othermarkers actually causes the disease. Instead, these markershave the ability to reflect the local inflammatory process inthe artery and, perhaps, in other tissues and organs withrelevance to atherogenesis such as adipose tissue in patientswith metabolic syndrome and diabetes mellitus.1 Moreover, aproposed biomarker should not only provide independentinformation on cardiovascular risk, but also be easy tomeasure using inexpensive and standardized commercialassays with low variability that do not require specializedplasma collection or assay techniques. In this regard CRP hasproved most robust, as it is an excellent analyte with astandardized assay, has negligible diurnal variation, does not

1912 Arterioscler Thromb Vasc Biol November 2008

by guest on May 17, 2018

http://atvb.ahajournals.org/D

ownloaded from

depend on food intake, and has a long half-life, in addition toa remarkable dynamic range.53,54 It is easily measured, andstandardized high-sensitivity immunoassays provide similarresults in fresh, stored, or frozen plasma, reflecting thestability of the protein. All these characteristics are importantcriteria for an ideal biomarker in cardiovascular disease(Table 1). Unfortunately, measurements of chemokines inplasma/serum are hampered with some methodological prob-lems which may limit their use as biomarkers in CAD andother cardiovascular disease (Table 2). First, platelets are animportant cellular source for several chemokines such asCCL5, CXCL5, and CXCL7,21,34,35 and these parametersshould therefore ideally be measured in platelet-poor orplatelet-free plasma which is not always available. Thus,measurement in serum may underestimate the concentrationin high-risk patients (ie, less release of chemokines fromplatelets during coagulation ex vivo attributable to degranu-lation and platelet activation in vivo). Moreover, freeze andthaw cycles of ordinary plasma may results in ex vivo releaseof chemokines from platelets which again may depend on theplatelet counts within the stored plasma sample and thedegree of platelet degranulation in vivo. Second, plasma/serum levels of several chemokines are in general low, oftenjust above or even below the detection limit of the variousenzyme immunoassays, at least partly reflecting their shorthalf-lives in circulation. These aspects may confound theirbenefit in routine risk prediction in particularly in those withthe lowest levels (ie, apparently healthy individuals). Forexample, whereas serum levels of CXCL8 in CAD patientsare mostly �10 pg/mL, with even lower levels in healthycontrols, CRP levels in these individuals will mostly be �1�g/mL. Third, even when using platelet-free plasma, precau-tion is needed to avoid degradation of chemokine levels exvivo after blood collection or during storage. Thus, the tubesshould ideally be placed on ice, centrifuged within 30 to 60minutes, stored at �70 to 80°C, and the samples should bethawed fewer than 3 times. Fourth, heparin administrationmay influence plasma/serum levels of chemokines throughenhanced release of heparan sulfate-bound chemokines in thevessel wall,55 which may limit their use as prognostic markersin ACS. Finally, several chemokines are involved in athero-genesis and plaque destabilization, at least partly operatingthrough different mechanisms and at different step in thedisease process. It is perhaps unlikely that one particular ofthese mediators could reflect the entire activity of this rather

complicated chemokine network, being mediators that areoperating rather upstream in the inflammatory cascade.

Chemokines as Markers for Subclinical CADMigration of monocytes into the arterial wall is an early eventin atheroma formation.1–3 Based on the suggested importantrole of chemokines in this process, and in particular CXCL8and CCL2,6,11,23,24 they could potentially be early markers ofCAD, and a few studies have tested this hypothesis (Table 3).In a nested case–control study in the prospective EPIC-Norfolk population study, baseline CXCL8 concentrationsamong 785 apparently healthy individuals, in whom fatal ornonfatal CAD developed during follow-up (average of �6years), was significantly higher than in 1570 matched con-trols (3.5 pg/mL versus 3.1 pg/mL, P�0.001).56 The authorsshowed that people in the highest CXCL8 quartile had a fullyadjusted odds ratio of 1.77 (95% CI, 1.21 to 2.60) comparedwith those in the lowest quartile (P�0.001). The odds ratiofor future CAD was still significant after adjustment fortraditional risk factors and after additional adjustment forCRP and total leukocyte count. Although the authors con-clude that CXCL8 could represent a novel biomarker forCAD in apparently healthy individuals, there was a consid-erable overlap between the two study groups, and as dis-cussed above, the levels were mostly just above the detectionlimit of the assay (ie, 2.5 pg/mL).

Deo et al observed strong associations with CAD riskfactors such as older age, gender, hypertension, diabetes, andrenal insufficiency after measuring CCL2 levels in subjectsfrom the Dallas Heart Study (3499 subjects �65 years old).57

In this study, CCL2 levels were associated with coronaryartery calcification (CAC), as a marker of subclinical athero-sclerosis, in multivariable analyses adjusting for traditionalcoronary risk factors. However, when further adjustment wasmade for age, CCL2 was no longer independently associatedwith the presence of subclinical atherosclerosis. The authorsconclude that CCL2 may not be useful as a clinical tool thatis additive to the assessment of age, traditional risk factors, orCRP for the detection of subclinical atherosclerosis. Morerecently, Tang et al investigated the association betweenCCL2 levels and CAC in a large population-based sample

Table 1. Criteria for a Reliable Biomarker in CardiovascularDisease

● Clear demonstration of incremental prognostic value over traditional riskmodels.

● Ability to reflect several pathogenic processes in atherogenesis andplaque destabilization.

● Stable levels in individuals. Not dependent on food intake with negligiblecircadian variation.

● High stability of the actual protein (e.g., long half-life).

● Easy to measure using inexpensive and standardized commercial assayswith low variability.

● No requirement for specialized plasma/serum collection or storage.

Table 2. Challenges When Using Chemokines as RiskPredictors in Cardiovascular Disease

● Several chemokines are involved in atherogenesis, operating throughdifferent mechanisms and at different levels. No uniform chemokinemarker that reflect all these processes.

● Several chemokines are found at low levels in serum/plasma; oftenbelow the detection limit of the actual assays.

● Short half-life. Unstable.

● Strict and bothersome demands for blood sampling protocol and storage(e.g., the tubes should be placed on ice, centrifuged within 30–60minutes, stored at �70–80oC, and thawed less than three times).

● Release from platelets ex vivo during coagulation (serum) or duringfreeze and thaw cycles (ordinary plasma) may be an importantconfounder. Platelet-poor or platelet-free EDTA or citrated plasma shouldbe preferred.

● Plasma/serum levels are influenced by the use of heparin in vivo.

Aukrust et al Chemokines in Cardiovascular Risk Prediction 1913

by guest on May 17, 2018

http://atvb.ahajournals.org/D

ownloaded from

free of clinical CAD (2246 whites and 470 blacks, mean age55 years).58 Although CCL2 was significantly and positivelyassociated with the presence of CAC after controlling for ageand gender, with the same pattern in both white and blacks,the observed association was attenuated and no longer statis-tically significant after additionally adjusting for other CADrisk factors, further underscoring the limitation of CCL2 as anindependent risk predictor in apparently healthy individuals.In line with this, Thakore et al could not detect any associa-tion between CCL2 levels and carotid intimal medial thick-ness (IMT) or stenosis in the Offspring Cohort of theFramingham Heart Study (n�2885, 53% women, mean age59 years).59 Finally, Herder et al investigated CCL2, CXCL8,and CXCL10 serum levels in a case–cohort design, based ondata from 381 individuals with (294 men, 87 women, meanage 57.3 years) and 1977 individuals without (1006 men, 971women, mean age 52.3 years) incident CAD from theprospective, population-based MONICA/KORA Augsburgstudy (1984 to 2002).60 The mean follow-up time was 11.0years. Although baseline concentrations were significantlyhigher in cases compared with noncases (P�0.001 for allchemokines), only CCL2 and CXCL8 remained associatedwith CAD risk after adjustment for age and sex. Moreover,after adjustment for further cardiovascular and immunologicrisk factors, the observed associations became nonsignificant.Taken together, although some significant findings, the abil-

ity of serum/plasma levels of chemokines to predict subclin-ical CAD is so far disappointing.

Chemokines as Predictors of CardiovascularEvents in Patients With Overt CAD

In contrast to these somewhat disappointing studies on theability of CCL2 to predict subclinical CAD, CCL2 may seemto be of some value in predicting cardiovascular event inpatients with overt CAD (Table 3). In a substudy of the OralGlycoprotein IIb/IIIa Inhibition with Orbofiban in Patientswith Unstable Coronary Syndromes (OPUS-TIMI) 16 trial,the authors showed that among 2279 patients with ACS,CCL2 levels above the 75th percentile (238 pg/mL) wereassociated with an almost 2-fold increased risk of death or MIduring follow-up (10 months) after adjustment for standardrisk predictors including CRP.61 The authors suggest thatCCL2 could be attractive as a surrogate biomarker in thesepatients and merits further study as a potential therapeutictarget. More recently, the authors extended these findings byperforming serial measurements of CCL2 in a large andwell-characterized population of patients stabilized early afterACS (4244 individuals, 21 to 80 years, follow-up time 6months to 2 years).62 The authors show that higher baselinelevels of CCL2 were associated with an increased risk forlong-term death and major adverse cardiac outcomes, inde-pendent of standard measurements such as creatinine clear-

Table 3. Chemokine Plasma Levels and Risk for Cardiovascular Disease

Chemokine Population Numbers Follow Up Assoc. disease/event Study Design Outcome

IL-8/CXCL856 Healthyindividuals

785 cases1570 controls

6 years Fatal and nonfatal CAD Nested case-control

Higher levels of CXCL8 in cases

MCP-1/CCL257 Healthyindividuals

3499 Subclinical atherosclerosis Cross sectional CCL2 assoc. with CAC

MCP-1/CCL258 Healthyindividuals

2246 whites470 Afr. Am.

Subclinical atherosclerosis Cross sectional No assoc. between CCL2 and CACwhen adjusting for other risk factors

MCP-1/CCL259 Healthyindividuals

2885 Subclinical atherosclerosis Cross sectional No assoc. between CCL2 and carotidIMT

MCP-1/CCL260 CAD 381 cases1977 controls

11 years CAD Case-cohort No assoc. between CCL2 or CXCL8and CAD when adjusting for other risk

factors

MCP-1/CCL261 ACS 2279 10 mo Death or MI Prospective CCL2 above 75th percentile assoc.with increased risk of an event

MCP-1/CCL262 ACS,stabilized

4244 6 mo �2years

Death, MI, stroke, HF Prospective.Serial MCP-1measures

CCL2 levels at baseline and at 4 moassoc. with an event

MCP-1/CCL263 ACS 183 13 mo Composite (death, MI,unstable angina, revasc.)

Prospective CCL2 predicted a new coronary event

Eotaxin-3/CCL2664 CAD 1026 2.7–4.1years

Cardiovascular death, MI Prospective Lower CCL26 levels predicted futureevents

RANTES/CCL5,IP-10/CXCL10,IL-8/CXCL868

CAD 312 cases472 controls

Angiographically confirmedand stable CAD

Case control Higher levels of CXCL8 and CXCL10,lower CCL5 in cases

RANTES/CCL569 Coronaryangiography

389 24 mo Cardiovascular death, MI Prospective Low CCL5 levels were an predictorof cardiac mortality

RANTES/CCL570 ACS 54 Refractory ischemia Cross sectional Higher CCL5 in ischemic versusstabilized patients

IL-8, interleukin-8; MCP-1, monocyte chemoattractant protein-1; IP-10, interferon (INF)-inducible protein of 10 kd; CAD, coronary artery disease; ACS, acutecoronary syndrome; MI, myocardial infarction; IMT, intimal medial thickness; CAC, coronary artery calcium score; HF, heart failure.

1914 Arterioscler Thromb Vasc Biol November 2008

by guest on May 17, 2018

http://atvb.ahajournals.org/D

ownloaded from

ance and LDL cholesterol, as well as biomarkers such as CRPand B-type natriuretic peptide. In addition, CCL2 levelsmeasured 4 months after ACS also provided independentprognostic value. Thus, while the authors confirmed that anCCL2 threshold of 238 pg/mL identifies patients at increasedrisk for adverse events in both the acute and the chronicsetting after ACS, they also showed that patients with levelspersistently above this threshold seemed to be at particularlyhigh risk for death compared with those with persistentlylower levels, or levels that were only transiently above thisthreshold. In contrast, and somewhat disappointingly, ele-vated CCL2 levels did not identify patients who derivedincremental benefit from intensive statin therapy. The abilityof CCL2 to predict coronary events in patients with overtCAD has also been reported by Kervinen et al showing in amuch smaller study (n�183) of unselected ACS patients,with a very high rate (64%) of coronary events (ie, cardiacdeath, recurrent MI, unstable angina, or revascularization)during follow-up (13 months), that increased plasma levels ofCCL2, as well as the T cell marker soluble IL-2 receptor,were helpful for predicting new coronary events independentof other inflammatory mediators (ie, CRP and IL-6).63

Eotaxins (eotaxin-1/CCL11, eotaxin-2/CCL24, andeotaxin-3/CCL26) are members of the CC chemokine branchthat mainly acts on CCR3-bearing cells like eosinophils,basophils, and lymphocytes of the T helper cell type 2phenotype.4,17 There are few data on the role of thesechemokines in CAD, but recently, Falcone et al reported thatlower CCL26 concentrations were predictive of future car-diovascular events, whereas both CCL11 and CCL24 showedno association with risk, in a study population with confirmedCAD (n�1,026, 841 with stable and 185 with unstableangina) and with 105 cardiac events during follow-up (2.7 to4.1 years; Table 3).64 The highest risk of future cardiovascu-lar events was observed in subjects with combined elevationof CRP and reduction of CCL26, and receiver-operatingcharacteristic curves analysis suggested a superior prognosticvalue of CCL26 compared with CRP for predicting cardiacevents. However, although there are some reports of potentialantiinflammatory effects of CCL26,65 the reason for theassociation between low CCL26 levels and cardiac eventsremain obscure. The authors, as well as Kaehler et al, havepreviously reported an association between high CCL11levels and documented CAD,66,67 further underscoring thecomplexity in relation to the role of eotaxins in CAD.

CCL5: High and Low Levels Are AssociatedWith Disease Progression

Although most focus has been drawn against high levels ofvarious chemokines, some studies have suggested that lowlevels of CCL5 may be associated with disease progression inatherosclerosis (Table 3). Thus, Rothenbacher et al reportedthat in 312 patients (aged 40 to 68 years) with angiographi-cally confirmed and stable CAD and 472 age- and gender-matched controls, low levels of CCL5, as well as upregula-tion of CXCL10 and CXCL8, were associated with CAD,with no clear disease association for CCL2, macrophageinflammatory protein � (MIP-1�/CCL3), or CCL11.68 Whiledata from such case-controlled study, with no longitudinal

follow-up, should be interpreted with caution, Cavusoglu et almeasured baseline plasma levels of CCL5 in 389 malepatients undergoing coronary angiography at a VeteransAffairs Medical Center.69 The patients were followed-upprospectively for 24 months for the occurrence of cardiacmortality and MI. In the entire cohort of patients, lowbaseline CCL5 levels were an independent predictor ofcardiac mortality. Furthermore, when patients were risk-stratified into those with and without an ACS, CCL5 was anindependent predictor of both cardiac mortality and MI inthose without an ACS. An additional group in whom lowCCL5 levels were shown to be strongly predictive of eventswas the diabetic subset, possibly reflecting the greater ath-erosclerotic burden known to be present in this population.The authors hypothesize that the low CCL5 level in thosewith increased frequency of clinical events could reflectincreased deposition of CCL5 on the vascular endotheliumleading to greater CCR5 stimulation. However, the predictivevalue of low CCL5 levels may seem in contrast with a recentstudy by Kraaijeveld et al showing in a small population ofpatients with ACS (n�54) that high plasma levels of CCL5were significantly elevated in patients with refractory ische-mic symptoms versus stabilized patients.70 There may beseveral reasons for these apparently discrepancies. First,platelets are the most important cellular source of circulatingCCL5 levels, and the blood sampling protocol may havemajor influence on the estimated CCL5 level. Thus, althoughKraaijeveld et al used platelet-free plasma, there is nodetailed information on the blood sampling protocol in theother studies. Importantly, release from platelet ex vivoduring freeze and thaw cycles may be decreased in those withactive disease because of platelet degranulation in vivo,leading to underestimation of the actual CCL5 level. Duringmeningococcal septicemia low serum levels of CCL5 arereported as a prominent feature,71 reflecting decreased releaseof CCL5 during ex vivo coagulation attributable to markedlyenhanced release of CCL5 from activated platelets in vivo,illustrating that a decreased measurable CCL5 level undercertain circumstances could reflect high rather than lowCCL5 level in vivo. On the other hand, the use of heparin inACS may lead to incidentally increased CCL5 levels attribtu-able to enhanced release of heparan sulfate-bound chemo-kines in the vessel wall.55 However, the inability of CCL5levels to predict long-term adverse cardiovascular outcomesin ACS may also reflect that the acute spike in CCL5 levelsthat occur in this particular setting is an unreliable marker ofcoronary inflammation and long-term prognosis. Neverthe-less, these issues clearly illustrate the methodological chal-lenges when using chemokines, and in particular those thatare released from activated platelets, as risk predictors in CAD.

“New” Chemokines as Risk PredictorExcept for CCL5, the circulating levels of chemokines thathave been studied as risk markers in CAD are rather low (ie,�500 pg/mL), and as for CXCL8, the levels are just abovethe detection limit of the assays, which weaken their impactas risk markers. Recently, Kraaijeveld et al showed that veryhigh plasma levels (�100 ng/mL) of pulmonary andactivation-regulated chemokine (PARC/CCL18) in ACS pa-

Aukrust et al Chemokines in Cardiovascular Risk Prediction 1915

by guest on May 17, 2018

http://atvb.ahajournals.org/D

ownloaded from

tients (n�54) were indicative of refractory symptoms, sug-gesting that this chemokine should be evaluated as a riskpredictor in a larger population of CAD patients.70 Moreover,based on the link between inflammatory mediators and thecleavage of the membrane-bound to the soluble form ofCXCL16, it has been suggested that soluble CXCL16 couldserve as a reliable, stable, and robust marker of inflammation,being detectable in plasma/serum in concentrations regularly�1 ng/mL. Indeed, raised levels of soluble CXCL16 havebeen reported in some inflammatory conditions such asrheumatoid arthritis and systemic lupus erythematosus.72,73

However, conflicting data exist on plasma levels of CXCL16in CAD. Thus, whereas Sheikine et al found decreasedCXCL16 levels in both stable and unstable angina,74 othershave reported increased CXCL16 levels in CAD patients,75,76

and in the study by Lehrke et al, particularly high levels werefound in those with unstable disease.75 Hence, although somepotential advantages (eg, relatively high serum/plasma levels,minor influence on freeze/thaw cycles, stable during storage;Aukrust P and Damås JK, unpublished data 2007), we are inneed of bigger, carefully designed clinical studies evaluatingCXCL16 as a potential biomarker of atherosclerotic vesseldisease. At present, prospective studies on the association ofCXCL16 with the risk of developing CAD or clinical eventsin those with overt CAD are lacking. Furthermore, recentstudies have shown raised plasma levels of the so-calledregulatory chemokines CCL19 and CCL21 in CAD withparticularly high levels in those with unstable disease,13 andErbel et al reported enhanced expression of CCL19 andCCL21 in unstable carotid and coronary plaques.77 However,at present, there are no prospective data on CCL19 andCCL21 as risks predictors in CAD. Finally, Ardigo et al haverecently used a multimarker proteomic approach, hypothesiz-ing that measuring serum levels of vascular derived inflam-matory biomarkers could reveal a “signature of disease” thatcan serve as a highly accurate method to assess for thepresence of CAD.78 By using this approach, they identified anoptimal combined chemokine model as the best predictor ofCAD (CCL11, CXCL10, CCL2, MCP-2/CCL8, MCP-3/CCL7, MCP-4/CCL13, and CCL3), far superior to CRP.Although few patients were examined in this cross-sectionalstudy (49 cases and 44 controls), measuring of global patternsof cytokine and chemokine expression plausibly may yieldmore relevant biological information than individual proteinassays. As several chemokines are involved in atherogenesis,at least partly through different mechanisms and at differentlevels, a combination of serum levels of multiple chemokinescould potentially identify subjects with clinically significantatherosclerotic heart disease with a very high degree ofaccuracy. However, at present such an approach is unrealisticfrom a practical and clinical point of view.

Genetic Variation in the Chemokine Genes asRisk Factor in Atherogenesis

Epidemiological genetic studies in gene related to lipidmetabolism have been of major importance for our under-standing of the role of these factors in atherogenesis. Similar,polymorphism studies in chemokine or chemokine receptorgenes, trying to relate these genetic variations to increased

risk for cardiovascular disease, are of importance for thestudy of the pathogenic role of these mediators in theatherosclerotic process.79 Such analyses could also be ofimportance for risk stratification, in particular in the comingyears where molecular methodology will be more accessible.Several chemokine/chemokine receptor polymorphisms havebeen linked to increased risk of cardiovascular disease suchas polymorphisms in the gene for CCL2 and CCL5 and theircorresponding receptors CCR2 and CCR5, respectively, aswell as polymorphisms in the CX3CL1 receptor gene,CX3CR1. These issues have recently been reviewed80 andwill not be discussed in further detail in this review. How-ever, although the results from some of these studies mayprovide new conceptual, diagnostic, and therapeutic ap-proaches to vascular diseases, the single nucleotide poly-morphisms (SNPs) studies also have several limitations.81

Although the standards and quality seem to be improving,there is nevertheless a risk that SNP-based association anal-yses will squander academic trust and scientific resourcesowing to unsatisfactory analysis. Moreover, some of thechemokine/chemokine receptor genotypes are rare, makingepidemiological conclusions difficult except in the largestcohorts. However, the strength of such studies will greatlyimprove if the actual polymorphism could be related tophenotypic alterations with relevance to atherogenesis. As anelegant and path-breaking example, McDermott et al showedthat CX3CL1-dependent cell–cell adhesion under conditionsof physiological shear was severely reduced in cells express-ing the CX3CR1 polymorphism which causes amino acidchange from threonine to methionine at position 280(CX3CR1-M280).82 This was associated with marked reduc-tion in the kinetics of CX3CL1 binding as well as reducedCX3CL1-induced chemotaxis of primary leukocytes fromdonors homozygous for CX3CR1-M280. Importantly, theseauthors also showed that CX3CR1-M280 is independentlyassociated with a lower risk of atherosclerotic cardiovasculardisease in the Offspring Cohort of the Framingham HeartStudy, a long-term prospective study of the risks and naturalhistory of this disease (204 cases and 1655 controls).

Chemokines as Predictor Restenosis andCardiac Allograft Vasculopathy

In addition to atherosclerosis, chemokines are also involvedin the pathogenesis of other and related cardiovasculardisorders such as the arterial remodeling process character-izing restenosis after PCI, with and without stenting, andcardiac allograft vasculopathy (CAV) after heart transplanta-tion.83 The arterial vessel wall response to a variety of injuriesconsists in structural changes, which can result in luminalnarrowing and aggravation of the underlying disease. Thisarterial remodeling is characterized by neointima formationand medial thickening, inflammatory cell recruitment, andendothelial dysfunction. Chemokines and the correspondingreceptors have been shown to participate at every step of theremodeling process.83 Thus, CCL2 may promote monocyteinfiltration of the injured vessel wall and can stimulateproliferation of SMC in models of restenosis, and indeed,there are some reports suggesting that CCL2 can predictrestenosis after PCI, with and without accompanying stent-

1916 Arterioscler Thromb Vasc Biol November 2008

by guest on May 17, 2018

http://atvb.ahajournals.org/D

ownloaded from

ing. Hence, Cipollone et al showed that plasma levels ofCCL2, but not of CCL5 and CXCL8, were associated withrestenosis during 6 months of follow-up in 50 CAD patients(15 with stable and 35 with unstable disease) who underwentPCI.84 Interestingly, those with restenosis were characterizedby persistently raised CCL2 levels without any increase inbaseline levels. There are also a few studies, examining arelatively low number of patients (�50), who suggest thatplasma levels of CCL2 may predict in-stent restenosis.85,86

Again, serial measurements seem to be superior to measure-ments at baseline or immediately after the procedure. Reste-nosis seems also to involve platelet-related mechanisms,87

and in contrast to Cipollone et al,84 Inami and colleaguessuggested that serial measurements of platelet-derived CCL5could predict restenosis after PCI in stable angina patients,but again, few patients (n�52) were examined.88 Finally,although inflammation and vascular remodeling appear toplay an important pathogenic role in the development ofCAV, few studies have examined the predictive value ofchemokine measurements, and the results are somewhatconflicting.89,90 These studies are also hampered by method-ological flaws by using coronary angiography instead ofintravascular ultrasound (IVUS) for CAV diagnosing andevaluation.

ConclusionsThe role of chemokines as clinical biomarkers is at presentunclear, and their ability to predict subclinical atherosclerosishas been disappointing. Further prospective studies, includinga larger number of patients, are needed to make any firmconclusion (Table 4). Such studies should also include serialmeasurements in addition to baseline measurements, andbased on previous research, studies in patients with estab-lished CAD or ACS should be given priority as opposed tostudies in apparently healthy individuals. Studies that exam-ine the relationship between chemokine levels and the degreeof atherosclerosis and plaque instability as assessed by new

imaging techniques, including molecular imaging, will alsobe needed. While the demonstration of an association be-tween chemokines and CAD is a necessary first step, suchstudies do not establish the full clinical utility of a biomarker,which is a more demanding process that requires validation inmultiple cohorts, a reliable and cost-effective assay, and cleardemonstration of incremental prognostic value over tradi-tional risk models. If successful, such new biomarkers will beuseful indicators for better risk assessment, diagnosis, andprognosis, as well as for monitoring pharmacological treat-ments for atherosclerosis. Based on the participation ofseveral chemokines in atherogenesis, it is possible that in thefuture, combined measurements of multiple chemokinescould reveal as a “signature of disease” that can serve as ahighly accurate method to assess for the presence of athero-sclerotic disease. Finally, it is also important to underscorethat the lack of function as a biomarker does not exclude animportant pathogenic role of chemokines in atherogenesisand plaque destabilization, and accordingly, does not negatethe potential value of chemokines as novel targets for therapyin atherosclerotic disorders.

Sources of FundingThis work was supported by grants from the Norwegian Council ofCardiovascular Research, Research Council of Norway, the Univer-sity of Oslo, Medinnova Foundation, Rikshospitalet UniversityHospital, and Helse Sør-Øst.

DisclosuresNone.

References1. Hansson GK. Inflammation, atherosclerosis, and coronary artery disease.

N Engl J Med. 2005;352:1685–1695.2. Packard RR, Libby P. Inflammation in atherosclerosis: from vascular biology

to biomarker discovery and risk prediction. Clin Chem. 2008;54:24–38.3. Davies MJ. The composition of coronary-artery plaques. N Engl J Med.

1997;336:1312–1314.4. Charo IF, Ransohoff RM. The many roles of chemokines and chemokine

receptors in inflammation. N Engl J Med. 2006;354:610–621.5. Braunersreuther V, Mach F, Steffens S. The specific role of chemokines

in atherosclerosis. Thromb Haemost. 2007;97:714–721.6. Apostolopoulos J, Davenport P, Tipping PG. Interleukin-8 production by

macrophages from atheromatous plaques. Arterioscler Thromb Vasc Biol.1996;16:1007–1012.

7. Smith C, Damås JK, Otterdal K, Øie E, Sandberg WJ, Yndestad A,Waehre T, Scholz H, Endresen K, Olofsson PS, Halvorsen B, GullestadL, Frøland SS, Hansson GK, Aukrust P. Increased levels of neutrophil-activating peptide-2 in acute coronary syndromes: possible role of plate-let-mediated vascular inflammation. J Am Coll Cardiol. 2006;48:1591–1599.

8. Breland UM, Halvorsen B, Hol J, Oie E, Paulsson-Berne G, Yndestad A,Smith C, Otterdal K, Hedin U, Wæhre T, Sandberg WJ, Frøland SS,Haraldsen G, Gullestad L, Damås JK, Hansson GK, Aukrust P. Apotential role of the CXC Chemokine GRO{alpha} in atherosclerosis andplaque destabilization. Downregulatory effects of statins. ArteriosclerThromb Vasc Biol. 2008;28:1005–1011.

9. Cheng C, Tempel D, van Haperen R, de Boer HC, Segers D, Huisman M,van Zonneveld AJ, Leenen PJ, van der Steen A, Serruys PW, de Crom R,Krams R. Shear stress-induced changes in atherosclerotic plaque compo-sition are modulated by chemokines. J Clin Invest. 2007;117:616–626.

10. Wuttge DM, Zhou X, Sheikine Y, Wågsater D, Stemme V, Hedin U,Stemme S, Hansson GK, Sirsjo A. CXCL16/SR-PSOX is an interferon-gamma-regulated chemokine and scavenger receptor expressed in athero-sclerotic lesions. Arterioscler Thromb Vasc Biol. 2004;24:750–755.

11. Takeya M, Yoshimura T, Leonard EJ, Takahashi K. Detection ofmonocyte chemoattractant protein-1 in human atherosclerotic lesions by

Table 4. Types of Studies Needed to Move ChemokineMeasurements into the Clinic

● Studies in patients with established CAD or patients with acute coronarysyndromes should be given priority.

● Prospective studies including a large number of patients, focusing onclinical end-point.

● Studies that examine the relationship between chemokine levels and thedegree of atherosclerosis and plaque instability as assessed by newimaging techniques, including molecular imaging, will also be needed.

● Studies that investigate the ability of combined measurements of severalchemokines to give prognostic information (signature of disease).

● Studies that include serial measurements in addition to baselinemeasurements.

● Studies that examine the predictive value of “new” chemokines with highserum/plasma concentrations and with presumably more stability(e.g. CXCL16, CCL18, and CCL21).

● Studies that examine the ability of chemokines to monitor responses totherapeutic intervention trials.

● Studies should compare the predictive value of chemokines with thepredictive value of established biomarkers including CRP.

Aukrust et al Chemokines in Cardiovascular Risk Prediction 1917

by guest on May 17, 2018

http://atvb.ahajournals.org/D

ownloaded from

an anti-monocyte chemoattractant protein-1 monoclonal antibody. HumPathol. 1993;24:534–539.

12. Lee WH, Kim SH, Jeong EM, Choi YH, Kim DI, Lee BB, Cho YS, KwonBS, Park JE. A novel chemokine, Leukotactin-1, induces chemotaxis,pro-atherogenic cytokines, and tissue factor expression in atherosclerosis.Atherosclerosis. 2002;161:255–260.

13. Damås JK, Smith C, Øie E, Fevang B, Halvorsen B, Waehre T, BoullierA, Breland U, Yndestad A, Ovchinnikova O, Robertson AK, SandbergWJ, Kjekshus J, Tasken K, Frøland SS, Gullestad L, Hansson GK,Quehenberger O, Aukrust P. Enhanced expression of the homeostaticchemokines CCL19 and CCL21 in clinical and experimental atheroscle-rosis: possible pathogenic role in plaque destabilization. ArteriosclerThromb Vasc Biol. 2007;27:614–620.

14. Mach F, Sauty A, Iarossi AS, Sukhova GK, Neote K, Libby P, Luster AD.Differential expression of three T lymphocyte-activating CXC che-mokines by human atheroma-associated cells. J Clin Invest. 1999;104:1041–1050.

15. Cushing SD, Berliner JA, Valente AJ, Territo MC, Navab M, Parhami F,Gerrity R, Schwartz CJ, Fogelman AM. Minimally modified low densitylipoprotein induces monocyte chemotactic protein 1 in human endothelialcells and smooth muscle cells. Proc Natl Acad Sci U S A. 1990;87:5134–5138.

16. Terkeltaub R, Banka CL, Solan J, Santoro D, Brand K, Curtiss LK.Oxidized LDL induces monocytic cell expression of interleukin-8, achemokine with T-lymphocyte chemotactic activity. ArteriosclerThromb. 1994;14:47–53.

17. Rollins BJ. Chemokines. Blood. 1997;90:909–928.18. Barlic J, Murphy PM. Chemokine regulation of atherosclerosis. J Leukoc

Biol. 2007;82:226–236.19. Schecter AD, Berman AB, Taubman MB. Chemokine receptors in

vascular smooth muscle. Microcirculation. 2003;10:265–272.20. Kodali RB, Kim WJ, Galaria II, Miller C, Schecter AD, Lira SA,

Taubman MB. CCL11 (Eotaxin) induces CCR3-dependent smoothmuscle cell migration. Arterioscler Thromb Vasc Biol. 2004;24:1211–1216.

21. Weber C. Platelets and chemokines in atherosclerosis: partners in crime.Circ Res. 2005;96:612–616.

22. Gu L, Okada Y, Clinton SK, Gerard C, Sukhova GK, Libby P, Rollins BJ.Absence of monocyte chemoattractant protein-1 reduces atherosclerosisin low density lipoprotein receptor-deficient mice. Mol Cell. 1998;2:275–281.

23. Boring L, Gosling J, Clearl M, Charo IF. Decreased lesion formation inCCR2�/� mice reveals a role for chemokines in the initiation of athero-sclerosis. Nature. 1998;394:894–897.

24. Boisvert WA, Santiago R, Curtiss LK, Terkeltaub RA. A leukocytehomologue of the IL-8 receptor CXCR-2 mediates the accumulation ofmacrophages in atherosclerotic lesions of LDL receptor-deficient mice.J Clin Invest. 1998;101:353–361.

25. Galkina E, Harry BL, Ludwig A, Liehn EA, Sanders JM, Bruce A, WeberC, Ley K. CXCR6 promotes atherosclerosis by supporting T-cell homing,interferon-gamma production, and macrophage accumulation in the aorticwall. Circulation. 2007;116:1801–1811.

26. Combadiere C, Potteaux S, Gao JL, Esposito B, Casanova S, Lee EJ,Debre P, Tedgui A, Murphy PM, Mallat Z. Decreased atheroscleroticlesion formation in CX3CR1/apolipoprotein E double knockout mice.Circulation. 2003;107:1009–1016.

27. Lesnik P, Haskell CA, Charo IF. Deceased atherosclerosis in CX3CR1�/�

mice reveals a role for fractalkine in atherogenesis. J Clin Invest. 2003;111:333–340.

28. Saederup N, Chan L, Lira SA, Charo IF. Fractalkine deficiency markedlyreduces macrophage accumulation and atherosclerotic lesion formation inCCR2�/� mice: evidence for independent chemokine functions in athero-genesis. Circulation. 2008;117:1642–1648.

29. Combadiere C, Potteaux S, Rodero M, Simon T, Pezard A, Esposito B,Merval R, Proudfoot A, Tedgui A, Mallat Z. Combined inhibition ofCCL2, CX3CR1, and CCR5 abrogates Ly6Chi and Ly6Clo monocytosisand almost abolishes atherosclerosis in hypercholesterolemic mice.Circulation. 2008;117:1649–1657.

30. Potteaux S, Combadiere C, Esposito B, Casanova S, Merval R, ArdouinP, Gao JL, Murphy PM, Tedgui A, Mallat Z. Chemokine receptor CCR1disruption in bone marrow cells enhances atherosclerotic lesion devel-opment and inflammation in mice. Mol Med. 2005;11:16–20.

31. Aslanian AM, Charo IF. Targeted disruption of the scavenger receptorand chemokine CXCL16 accelerates atherosclerosis. Circulation. 2006;114:583–590.

32. Gerszten RE, Mach F, Sauty A, Rosenzweig A, Luster AD. Chemokines,leukocytes, and atherosclerosis. J Lab Clin Med. 2000;136:87–92.

33. Aukrust P, Berge RK, Ueland T, Aaser E, Damås JK, Wikeby L,Brunsvig A, Muller F, Forfang K, Frøland SS, Gullestad L. Interactionbetween chemokines and oxidative stress: possible pathogenic role inacute coronary syndromes. J Am Coll Cardiol. 2001;37:485–491.

34. Weyrich AS, Zimmerman GA. Platelets: signaling cells in the immunecontinuum. Trends Immunol. 2004;25:489–495.

35. von Hundelshausen P, Weber C. Platelets as immune cells: bridginginflammation and cardiovascular disease. Circ Res. 2007;100:27–40.

36. Lutgens E, Faber B, Schapira K, Evelo CT, van Haaften R, Heeneman S,Cleutjens KB, Bijnens AP, Beckers L, Porter JG, Mackay CR, Rennert P,Bailly V, Jarpe M, Dolinski B, Koteliansky V, de Fougerolles T, DaemenMJ. Gene profiling in atherosclerosis reveals a key role for smallinducible cytokines: validation using a novel monocyte chemoattractantprotein monoclonal antibody. Circulation. 2005;111:3443–3452.

37. Jackson CL, Bennett MR, Biessen EA, Johnson JL, Krams R. Assessmentof unstable atherosclerosis in mice. Arterioscler Thromb Vasc Biol. 2007;27:714–720.

38. Libby P, Aikawa M. Stabilization of atherosclerotic plaques: new mech-anisms and clinical targets. Nat Med. 2002;8:1257–1262.

39. Tsmikas S, Willerson JT, Ridker PM. C-reactive protein and otheremerging blood biomarkers to optimize risk stratification of vulnerablepatients. J Am Coll Cardiol. 2006;47(suppl. C):19–31.

40. Fichtlscherer S, Heeschen C, Zeiher AM. Inflammatory markers andcoronary artery disease. Curr Opin Pharmacol. 2004;4:124–131.

41. Ridker PM. Clinical application of C-reactive protein for cardiovasculardisease detection and prevention. Circulation. 2003;107:363–369.

42. Ridker PM, Hennekens CH, Buring JE, Rifai N. C-reactive protein andother markers of inflammation in the prediction of cardiovascular diseasein women. N Engl J Med. 2000;342:836–843.

43. Zebrack JS, Muhlestein JB, Horne BD, Anderson JL; Intermountain HeartCollaboration Study Group. C-reactive protein and angiographic coronaryartery disease: independent and additive predictors of risk in subjects withangina. J Am Coll Cardiol. 2002;39:632–637.

44. Morrow DA, Rifai N, Antman EM, Weiner DL, McCabe CH, CannonCP, Braunwald E. C-reactive protein is a potent predictor of mortalityindependently of and in combination with troponin T in acute coronarysyndromes: a TIMI 11A substudy. Thrombolysis in MyocardialInfarction. J Am Coll Cardiol. 1998;31:1460–1465.

45. Morrow DA, de Lemos JA, Sabatine MS, Wiviott SD, Blazing MA, ShuiA, Rifai N, Califf RM, Braunwald E. Clinical relevance of C-reactiveprotein during follow-up of patients with acute coronary syndromes in theAggrastat-to-Zocor Trial. Circulation. 2006;114:281–288.

46. Rao M, Jaber BL, Balakrishnan VS. Inflammatory biomarkers and car-diovascular risk: association or cause and effect? Semin Dial. 2006;19:129–135.

47. Biasucci LM, Liuzzo G, Fantuzzi G, Caligiuri G, Rebuzzi AG, GinnettiF, Dinarello CA, Maseri A. Increasing levels of interleukin (IL)-1Ra andIL-6 during the first 2 days of hospitalization in unstable angina areassociated with increased risk of in-hospital coronary events. Circulation.1999;99:2079–2084.

48. Hulthe J, McPheat W, Samnegård A, Tornvall P, Hamsten A, Eriksson P.Plasma interleukin (IL)-18 concentrations is elevated in patients withprevious myocardial infarction and related to severity of coronary ath-erosclerosis independently of C-reactive protein and IL-6. Atherosclerosis.2006;188:450–454.

49. Ridker PM, Hennekens CH, Roitman-Johnson B, Stampfer MJ, Allen J.Plasma concentration of soluble intercellular adhesion molecule 1 andrisks of future myocardial infarction in apparently healthy men. Lancet.1998;351:88–92.

50. Ueland T, Jemtland R, Godang K, Kjekshus J, Hognestad A, Omland T,Squire IB, Gullestad L, Bollerslev J, Dickstein K, Aukrust P. Prognosticvalue of osteoprotegerin in heart failure after acute myocardial infarction.J Am Coll Cardiol. 2004;44:1970–1976.

51. Heeschen C, Dimmeler S, Hamm CW, van den Brand MJ, Boersma E,Zeiher AM, Simoons MLed; CAPTURE Study Investigators. SolubleCD40 ligand in acute coronary syndromes. N Engl J Med. 2003;348:1104–1111.

52. Kovacs A, Tornvall P, Nilsson R, Tegner J, Hamsten A, Bjorkegren J.Human C-reactive protein slows atherosclerosis development in a mousemodel with human-like hypercholesterolemia. Proc Natl Acad Sci U S A.2007;104:13768–13773.

53. Ockene IS, Matthews CE, Rifai N, Ridker PM, Reed G, Stanek E.Variability and classification accuracy of serial high-sensitivity

1918 Arterioscler Thromb Vasc Biol November 2008

by guest on May 17, 2018

http://atvb.ahajournals.org/D

ownloaded from

C-reactive protein measurements in healthy adults. Clin Chem. 2001;47:444–450.

54. Meier-Ewert HK, Ridker PM, Rifai N, Price N, Dinges DF, MullingtonJM. Absence of diurnal variation of C-reactive protein concentrations inhealthy human subjects. Clin Chem. 2001;47:426–430.

55. Ranjbaran H, Wang Y, Manes TD, Yakimov AO, Akhtar S, Kluger MS,Pober JS, Tellides G. Heparin displaces interferon-gamma-inducible che-mokines (IP-10, I-TAC, and Mig) sequestered in the vasculature andinhibits the transendothelial migration and arterial recruitment of T cells.Circulation. 2006;114:1293–1300.

56. Boekholdt SM, Peters RJ, Hack CE, Day NE, Luben R, Bingham SA,Wareham NJ, Reitsma PH, Khaw KT. IL-8 plasma concentrations and therisk of future coronary artery disease in apparently healthy men andwomen: the EPIC-Norfolk prospective population study. ArteriosclerThromb Vasc Biol. 2004;24:1503–1508.

57. Deo R, Khera A, McGuire DK, Murphy SA, Meo Neto Jde P, MorrowDA, de Lemos JA. Association among plasma levels of monocyte che-moattractant protein-1, traditional cardiovascular risk factors, and sub-clinical atherosclerosis. J Am Coll Cardiol. 2004;44:1812–1818.

58. Tang W, Pankow JS, Carr JJ, Tracy RP, Bielinski SJ, North KE, HopkinsPN, Kraja AT, Arnett DK. Association of sICAM-1 and MCP-1 withcoronary artery calcification in families enriched for coronary heartdisease or hypertension: the NHLBI Family Heart Study. BMC Car-diovasc Disord. 2007;7:30.

59. Thakore AH, Guo CY, Larson MG, Corey D, Wang TJ, Vasan RS,D’Agostino RB Sr, Lipinska I, Keaney JF Jr., Benjamin EJ, O’DonnellCJ. Association of multiple inflammatory markers with carotid intimalmedial thickness and stenosis (from the Framingham Heart Study). Am JCardiol. 2007;99:1598–1602.

60. Herder C, Baumert J, Thorand B, Martin S, Lowel H, Kolb H, Koenig W.Chemokines and incident coronary heart disease: results from theMONICA/KORA Augsburg case-cohort study, 1984–2002. ArteriosclerThromb Vasc Biol. 2006;26:2147–2152.

61. de Lemos JA, Morrow DA, Sabatine MS, Murphy SA, Gibson CM,Antman EM, McCabe CH, Cannon CP, Braunwald E. Associationbetween plasma levels of monocyte chemoattractant protein-1 andlong-term clinical outcomes in patients with acute coronary syndromes.Circulation. 2003;107:690–695.

62. de Lemos JA, Morrow DA, Blazing MA, Jarolim P, Wiviott SD, SabatineMS, Califf RM, Braunwald E. Serial measurement of monocyte chemoat-tractant protein-1 after acute coronary syndromes: results from the A to Ztrial. J Am Coll Cardiol. 2007;50:2117–2124.

63. Kervinen H, Manttari M, Kaartinen M, Makynen H, Palosuo T, Pulkki K,Kovanen PT. Prognostic usefulness of plasma monocyte/macrophage andT-lymphocyte activation markers in patients with acute coronary syn-dromes. Am J Cardiol. 2004;94:993–996.

64. Falcone C, Minoretti P, D’Angelo A, Buzzi MP, Coen E, Emanuele E,Aldeghi A, Olivieri V, Geroldi D. Markers of eosinophilic inflammationand risk prediction in patients with coronary artery disease. Eur J ClinInvest. 2006;36:211–217.

65. Van Coillie E, Van Damme J, Opdenakker G. The MCP/eotaxin sub-family of CC chemokines. Cytokine Growth Factor Rev. 1999;10:61–86.

66. Emanuele E, Falcone C, D’Angelo A, Minoretti P, Buzzi MP, Bertona M,Geroldi D. Association of plasma eotaxin levels with the presence andextent of angiographic coronary artery disease. Atherosclerosis. 2006;186:140–145.

67. Kaehler J, Tuleweit A, Steven D, Krempl T, Haar A, Carstensen M,Koester R, Terres W, Meinertz T. Association between eotaxin (CCL11),C-reactive protein, and antimicrobial antibodies in patients undergoingcoronary angioplasty. J Investig Med. 2006;54:446–454.

68. Rothenbacher D, Muller-Scholze S, Herder C, Koenig W, Kolb H. Dif-ferential expression of chemokines, risk of stable coronary heart disease,and correlation with established cardiovascular risk markers. ArteriosclerThromb Vasc Biol. 2006;26:194–199.

69. Cavusoglu E, Eng C, Chopra V, Clark LT, Pinsky DJ, Marmur JD. Lowplasma RANTES levels are an independent predictor of cardiac mortalityin patients referred for coronary angiography. Arterioscler Thromb VascBiol. 2007;27:929–935.

70. Kraaijeveld AO, de Jager SC, de Jager WJ, Prakken BJ, McColl SR,Haspels I, Putter H, van Berkel TJ, Nagelkerken L, Jukema JW, BiessenEA. CC chemokine ligand-5 (CCL5/RANTES) and CC chemokineligand-18 (CCL18/PARC) are specific markers of refractory unstableangina pectoris and are transiently raised during severe ischemicsymptoms. Circulation. 2007;116:1931–1941.

71. Møller AS, Bjerre A, Brusletto B, Joø GB, Brandtzaeg P, Kierulf P.Chemokine patterns in meningococcal disease. J Infect Dis. 2005;191:768–775.

72. Kageyama Y, Torikai E, Nagano A. Anti-tumor necrosis factor-alphaantibody treatment reduces serum CXCL16 levels in patients with rheu-matoid arthritis. Rheumatol Int. 2007;27:467–472.

73. Wu T, Xie C, Wang HW, Zhou XJ, Schwartz N, Calixto S, Mackay M,Aranow C, Putterman C, Mohan C. Elevated urinary VCAM-1,P-selectin, soluble TNF receptor-1, and CXC chemokine ligand 16 inmultiple murine lupus strains and human lupus nephritis. J Immunol.2007;179:7166–7175.

74. Sheikine Y, Bang CS, Nilsson L, Samnegård A, Hamsten A, Jonasson L,Eriksson P, Sirsjo A. Decreased plasma CXCL16/SR-PSOX concen-tration is associated with coronary artery disease. Atherosclerosis. 2006;188:462–466.

75. Lehrke M, Millington SC, Lefterova M, Cumaranatunge RG, Szapary P,Wilensky R, Rader DJ, Lazar MA, Reilly MP. CXCL16 is a marker ofinflammation, atherosclerosis, and acute coronary syndromes in humans.J Am Coll Cardiol. 2007;49:442–449.

76. Smith C, Halvorsen B, Otterdal K, Wæhre T, Yndestad A, Fevang B,Sandberg WJ, Breland UM, Frøland SS, Øie E, Gullestad L, Damås JK,Aukrust P. High levels and inflammatory effects of soluble CXC ligand16 (CXCL16) in coronary artery disease: down-regulatory effects ofstatins. Cardiovasc Res. 2008;79:195–203.

77. Erbel C, Sato K, Meyer FB, Kopecky SL, Frye RL, Goronzy JJ, WeyandCM. Functional profile of activated dendritic cells in unstable athero-sclerotic plaque. Basic Res Cardiol. 2007;102:123–132.

78. Ardigo D, Assimes TL, Fortmann SP, Go AS, Hlatky M, Hytopoulos E,Iribarren C, Tsao PS, Tabibiazar R, Quertermous T; ADVANCE Inves-tigators. Circulating chemokines accurately identify individuals with clin-ically significant atherosclerotic heart disease. Physiol Genomics. 2007;31:402–409.

79. Bursill CA, Channon KM, Greaves DR. The role of chemokines inatherosclerosis: recent evidence from experimental models and popu-lation genetics. Curr Opin Lipidol. 2004;15:145–149.

80. Aukrust P, Yndestad A, Smith C, Ueland T, Gullestad L, Damås JK.Chemokines in cardiovascular risk prediction. Thromb Haemost. 2007;97:748–754.

81. Hegele RA. SNP judgments and freedom of association. ArteriosclerThromb Vasc Biol. 2002;22:1058–1061.

82. McDermott DH, Fong AM, Yang Q, Sechler JM, Cupples LA, MerrellMN, Wilson PW, D’Agostino RB, O’Donnell CJ, Patel DD, Murphy PM.Chemokine receptor mutant CX3CR1–M280 has impaired adhesivefunction and correlates with protection from cardiovascular disease inhumans. J Clin Invest. 2003;111:1241–1250.

83. Schober A, Zernecke A. Chemokines in vascular remodeling. ThrombHaemost. 2007;97:730–737.

84. Cipollone F, Marini M, Fazia M, Pini B, Iezzi A, Reale M, Paloscia L,Materazzo G, D’Annunzio E, Conti P, Chiarelli F, Cuccurullo F, MezzettiA. Elevated circulating levels of monocyte chemoattractant protein-1 inpatients with restenosis after coronary angioplasty. Arterioscler ThrombVasc Biol. 2001;21:327–334.

85. Oshima S, Ogawa H, Hokimoto S, Nakamura S, Noda K, Saito T,Soejima H, Takazoe K, Ishibashi F, Yasue H. Plasma monocyte chemoat-tractant protein-1 antigen levels and the risk of restenosis after coronarystent implantation. Jpn Circ J. 2001;65:261–264.

86. Sako H, Miura S, Iwata A, Nishikawa H, Kawamura A, Matsuo K, ShiraiK, Saku K. Changes in CCR2 chemokine receptor expression and plasmaMCP-1 concentration after the implantation of bare metal stents versussirolimus-eluting stents in patients with stable angina. Intern Med. 2008;47:7–13.

87. Mitra AK, Agrawal DK. In stent restenosis: bane of the stent era. J ClinPathol. 2006;59:232–239.

88. Inami N, Nomura S, Manabe K, Kimura Y, Iwasaka T. Platelet-derivedchemokine RANTES may be a sign of restenosis after percutaneouscoronary intervention in patients with stable angina pectoris. Platelets.2006;17:565–570.

89. Gullestad L, Simonsen S, Ueland T, Holm T, Aass H, Andreassen AK,Madsen S, Geiran O, Frøland SS, Aukrust P. Possible role of proinflam-matory cytokines in heart allograft coronary artery disease. Am J Cardiol.1999;84:999–1003.

90. Beaudeux JL, Dorent R, Bernard M, Reagan M, Foglietti MJ, Gand-jbakhch I, Cacoub P. Circulating monocyte chemoattractant protein-1 inheart transplant recipients with cardiac allograft vasculopathy. ClinBiochem. 2002;35:77–79.

Aukrust et al Chemokines in Cardiovascular Risk Prediction 1919

by guest on May 17, 2018

http://atvb.ahajournals.org/D

ownloaded from

Gullestad and Jan K. DamåsPål Aukrust, Bente Halvorsen, Arne Yndestad, Thor Ueland, Erik Øie, Kari Otterdal, Lars

Chemokines and Cardiovascular Risk

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 2008 American Heart Association, Inc. All rights reserved.

Greenville Avenue, Dallas, TX 75231is published by the American Heart Association, 7272Arteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/ATVBAHA.107.1612402008;28:1909-1919; originally published online July 31, 2008;Arterioscler Thromb Vasc Biol. 

http://atvb.ahajournals.org/content/28/11/1909World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://atvb.ahajournals.org//subscriptions/

at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:

  http://www.lww.com/reprints

Information about reprints can be found online at: Reprints: 

document. Question and AnswerPermissions and Rightspage under Services. Further information about this process is available in the

which permission is being requested is located, click Request Permissions in the middle column of the WebCopyright Clearance Center, not the Editorial Office. Once the online version of the published article for

can be obtained via RightsLink, a service of theArteriosclerosis, Thrombosis, and Vascular Biologyin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on May 17, 2018

http://atvb.ahajournals.org/D

ownloaded from