Dianhvct(a

FUS§H

2

Journal of the American College of Cardiology Vol. 51, No. 22, 2008© 2008 by the American College of Cardiology Foundation ISSN 0735-1097/08/$34.00P

CLINICAL RESEARCH Interventional Cardiology

Differential Effects of Drug-Eluting Stentson Local Endothelium-Dependent Coronary Vasomotion

Michalis I. Hamilos, MD,* Miodrag Ostojic, MD, PHD,† Branko Beleslin, MD, PHD,†Dragan Sagic, MD, PHD,‡ Ljubco Mangovski, MD,‡ Sinisa Stojkovic, MD, PHD,†Milan Nedeljkovic, MD, PHD,† Dejan Orlic, MD,† Bratislav Milosavljevic, MD,‡Dragan Topic, MD,‡ Nevena Karanovic, MD, PHD,§ William Wijns, MD, PHD,*on behalf of the NOBORI CORE Investigators

Aalst, Belgium; and Belgrade, Serbia

Objectives The aim of our study was to compare coronary vasomotion after implantation of a second-generation biolimusA9-eluting stent (BES) and of a sirolimus-eluting stent (SES).

Background Drug-eluting stents (DES) have been associated with impaired local coronary vasomotion, delayed endothelializa-tion, and increased late thrombotic risk. New DES with different drugs, pharmacokinetics, and polymers havebeen developed.

Methods Nineteen patients with a BES and 15 patients with a SES were studied 9 months after stent implantation.Endothelium-dependent and -independent coronary vasomotion were tested proximally and distally to the stentas well as at a reference segment during right atrial pacing at increasing heart rates. Quantitative coronaryangiographic measurements were performed offline.

Results Of the patients with BES, 2 showed vasoconstriction with increased heart rate and 17 showed vasodilatation. Ofthe patients with a SES, 9 showed vasoconstriction while 6 showed vasodilatation. The SES showed significantvasoconstriction at both the proximal (�2.3 � 10% vs. 7.9 � 10%) and the distal (�5.4 � 9% vs. 6.1 � 8%)segments to the stent compared with the BES (p � 0.003 for proximal, p � 0.001 for distal segment).Endothelium-independent vasomotion after intracoronary nitrates did not differ significantly between the2 groups (p � NS for proximal and distal segment).

Conclusions Unlike the case with the SES, endothelium-dependent vasomotion at adjacent stent segments seems to be pre-served after BES implantation. This result may be explained by the different drug release kinetics, DES design,or characteristics of polymer used in the stent system. (J Am Coll Cardiol 2008;51:2123–9) © 2008 by theAmerican College of Cardiology Foundation

ublished by Elsevier Inc. doi:10.1016/j.jacc.2007.12.059

aemdNs(s

dmSr

rug-eluting stents (DES) are considered a breakthroughn the treatment of coronary artery disease because of theirbility to reduce rates of early restenosis and subsequenteed for target lesion reintervention. However, delayedealing has been associated with an increased risk of late andery late stent thrombosis (1–3). Moreover, paradoxicaloronary vasoconstriction at coronary segments adjacent tohe stent has been reported for both sirolimus-eluting stentsSES) and paclitaxel-eluting stents (PES) up to 12 monthsfter implantation (4–7). This observation may be attribut-

rom the *Cardiovascular Centre, Aalst, Belgium; †Department of Cardiology,niversity Institute for Cardiovascular Diseases, Clinical Center of Serbia, Belgrade,erbia; ‡Institute for Cardiovascular Disease Dedinje, Belgrade, Serbia; and theMinistry of Health of Serbia, Belgrade, Serbia. Dr. Hamilos is supported by aellenic Society of Cardiology fellowship.

rManuscript received October 8, 2007; revised manuscript received November 20,

007, accepted December 8, 2007.

ble to delayed endothelialization caused by the drug and/orndothelial dysfunction caused by polymer-induced inflam-ation or hypersensitivity reaction. New DES have been

eveloped with different stent designs, drugs, and polymers.obori (Terumo Corporation, Tokyo, Japan) is one such

econd-generation DES that uses a bioresorbable polymerpolylactic acid) from which biolimus A9, an analogue ofirolimus, is eluted.

See page 2139

The aim of our study was to asses the endothelium-ependent and -independent coronary vasomotion 9onths after either biolimus A9-eluting stent (BES) or

ES implantation, using right atrial pacing at increasingate steps. Cardiac pacing increases myocardial oxygen

equirements to be matched by augmented coronary blood

Plsuastsagmramsipa

pacgapavSlna

roivSrAw2bb

tpnvvtvir

fpibsQpdavvdfd

SgapcrnlsuigsssgwpidvetdvptmMm

2124 Hamilos et al. JACC Vol. 51, No. 22, 2008Endothelium and DES June 3, 2008:2123–9

flow. The resulting increase inshear stress dilates human coro-nary epicardial arteries in thepresence of functional endothe-lium. Endothelial cell dysfunc-tion and the subsequent reducednitric oxide activity are associatedwith paradoxical constriction ofepicardial arteries with cardiacpacing (8–11).

Methods

atients with BES (n � 24) and with SES (n � 19) havingesions treated with a single stent and who consented to thetudy protocol were included in the study. All of themnderwent follow-up coronary angiography 9 � 1 monthsfter stent implantation as a part of the NOBORI COREtudy. The NOBORI CORE study is a prospective, mul-icenter, comparative study of Nobori and Cypher DESystems in which all patients were scheduled to undergongiographic follow-up. The primary end point was angio-raphic in-stent late loss at 9 months post-procedure. Theain secondary end points were major cardiac adverse event

ate at 1 and 9 months and yearly up to 5 years as well asssessment of endothelial function by atrial pacing at 9onths. The NOBORI CORE study was conducted in 5

tudy sites, but endothelial function assessment was plannedn only 2 sites enrolling the largest number of patients. Allatients consented to undergo the study protocol, which waspproved by the local ethics committee.

All stents were implanted in de novo lesions. Twoatients (1 in the BES and 1 in the SES group) withngiographically significant in-stent restenosis were ex-luded from the study protocol. In 2 patients from eachroup the quality of angiography was not sufficient to allowccurate measurement of vessel diameter changes, and thoseatients were also excluded. Finally, excluded from thenalysis were patients having constriction of the referenceessel (2 in the BES group and 1 in the SES group).tudy protocol. All vasoactive drugs were discontinued at

east 24 h before catheterization except for sublingualitroglycerin, which was withheld for at least 1 h. Long-cting beta-blockers were stopped 48 h before study.

Diagnostic left heart catheterization and coronary arte-iography were performed by a standard percutaneous fem-ral approach. A 6-F diagnostic catheter was introducednto the left main or right coronary artery, depending on theessel studied. A 5-F bipolar pacing wire (St. Jude Medical,t. Paul, Minnesota) was placed against the high lateralight atrial wall. The pacing study was then performed.fter control conditions were established, rapid atrial pacingas conducted at 20 beats/min above baseline heart rate formin, followed by increments in the pacing rate of 20

eats/min for 2 min each until a final pacing rate of 150

Abbreviationsand Acronyms

BES � biolimus A9-elutingstent(s)

DES � drug-eluting stent(s)

PES � paclitaxel-elutingstent(s)

QCA � quantitativecoronary angiography

SES � sirolimus-elutingstent(s)

eats/min was reached, angina was produced, or atrioven- l

ricular Wenckebach block developed. After rapid atrialacing, a 2-min recovery period was allowed. Intracoronaryitroglycerin (200 �g total dose) was administered. Atrio-entricular Wenckebach block was not treated with intra-enous atropine in order not to influence coronary vasomo-ion and blood flow responses. Pacing from the rightentricle was performed up to 150 beats/min in the patientsn whom atrioventricular Wenckebach block developed atates below 110 beats/min.

Serial contrast injections of the study vessel were per-ormed at baseline, at the end of a 2-min period at eachacing rate, immediately after ending the pacing, and afterntracoronary nitroglycerin administration. Heart rate andlood pressure were digitally recorded during the entiretudy protocol.

uantitative coronary angiography. Coronary angiogra-hy was performed on a digital X-ray system (AxiomatisSC, Siemens Medical Solutions, Malvern, Pennsylvania)t 25 frames/s. An appropriate view that permitted clearisualization of both the target artery and the referenceessel segment was selected. The angle of the view, theistance from the X-ray focus to the object, and the distancerom the object to the image intensifier were kept constanturing the study.The computer-based ACOM.PC 5.01 (Siemens Medical

olutions) was used for off-line quantitative coronary an-iography (QCA) analysis. All measurements were done byn independent observer who was blinded to the studyrotocol and type of stent implanted. Percent changes werealculated in all patients using the baseline angiogram aseference. In both groups, a reference angiographicallyormal segment in another vessel not related to the stented

esion as well as the stented segment and its adjacentegments proximal and distal were assessed (at least 5 mmp to 15 mm proximal and distal to the stent edges). If thentervened vessel was the right coronary artery, an angio-raphically normal segment as far as possible from thetented vessel segment was taken as reference. Because oftent placement in an ostial vessel segment, the proximalegments could not be evaluated in 4 patients in the BESroup and in 2 patients in the SES group. All measurementsere done exactly at the same segment, at baseline, at everyacing step, immediately after stopping pacing and afterntracoronary nitrates. A progressive reduction in vesseliameter was termed as vasoconstriction, and an increase asasodilatation. Patients with vasoconstriction in the refer-nce segment were not included in the final analysis becausehey were considered to have a generalized endothelialisease. Normalization of vessel diameters for the maximalasodilated state was performed by dividing resting andacing data by the vessel diameter after nitrates administra-ion. In our laboratory, coronary artery diameter measure-ents performed with the ACOM.PC 5.01 (Siemensedical Solutions) have interobserver variability of 0.11m, and intraobserver variability of 0.08 mm for mean

umen diameter on repeated analysis of the same frame.

SiCsqFbppss

R

TatmiawedbHmrmMd(w(MwmCd

dsttopsit1rt9bv1(s3sgIddmts

D

OamddfitktEtw

P

B

H

A

2125JACC Vol. 51, No. 22, 2008 Hamilos et al.June 3, 2008:2123–9 Endothelium and DES

tatistical analysis. Statistical analysis was performed us-ng the SAS 9.1 software (SAS Institute Inc., Cary, Northarolina). Continuous data are summarized as mean �

tandard deviation. Unpaired or paired t test were ade-uately used to analyze differences in continuous variables.isher and chi-square tests were used to analyze differencesetween categorical variables. A general linear model waserformed to adjust the vasomotor response to the baselinearameters and to predict the vasomotor response. Alltatistical comparisons were performed at the 5% level ofignificance.

esults

he baseline characteristics of the patients included in thenalysis are comparable (Table 1). Mean time from implan-ation was 9.12 � 1.0 months for SES and 9.9 � 0.9onths for BES. In the SES group, 8 stents were implanted

n the left anterior descending artery, 5 in the left circumflexrtery, and 2 in the right coronary artery (numbers for BESere 10, 3, and 6, respectively; p � NS). One patient in

ach group required right ventricular pacing because ofeveloping atrioventricular block at heart rate below 110eats/min.emodynamic data. Baseline and maximal pacing rate,ean arterial pressure at baseline and at maximal pacing

ate, and rate-pressure product both at baseline and ataximal pacing rate were similar in the 2 groups (Table 2).ean arterial blood pressure did not change significantly

uring pacing compared with baseline in the BES groupfrom 95 � 2.9 mm Hg to 95.4 � 3.8 mm Hg, p � NS),hereas it was increased significantly in the SES group

from 96 � 3.5 mm Hg to 101 � 4.5 mm Hg, p � 0.011).ean arterial pressure changes from baseline to peak pacing

ere not statistically significant between SES and BES (4.6m Hg vs. 1.0 mm Hg, p � 0.22 respectively).oronary vasomotor response to pacing and nitrates. Meanata for the 2 patient groups with regard to mean vessel

atient Characteristics

Table 1 Patient Characteristics

SES(n � 15)

BES(n � 19) p Value

Age, yrs 54 � 8 56 � 9 0.55

Gender, male/female 12/3 15/4 1

Number of diseased vessels 1.7 1.6 0.81

Hypertension 13 15 0.67

Diabetes 4 4 0.67

Hyperlipidemia 14 17 1

Smoking 9 11 1

Family history of coronary artery disease 8 15 0.15

Nitrates 4 7 0.71

Angiotensin-converting enzyme inhibitors 7 9 1

Calcium-channel blockers 9 2 0.003

Statins 15 18 1

Beta-blockers 14 18 1

lES � biolimus A9-eluting stent(s); SES � sirolimus-eluting stent(s).

iameter of the proximal, stented, distal, and referenceegments are shown in Table 3. For the segment proximalo the stent, normal vasomotion (vasodilatation) was main-ained in the BES group, whereas vasoconstriction wasbserved in the SES group (7.9 � 10% vs. �2.3 � 10%,� 0.003) (Figs. 1 and 2). For the segment distal to the

tent, normal vasomotion (vasodilatation) was maintainedn the BES group, whereas vasoconstriction was observed inhe SES group (6.1 � 8% vs. �5.4 � 9%, p � 0.001) (Figs.and 2). In the SES group, the vasomotor response in the

eference segment was better than in both the proximal andhe distal segments (�2.3 � 10% for proximal and �5.4 �% for distal vs. 10 � 0.4% for reference, p � 0.001 foroth). In the BES group, reference segments had a higherasodilatory response from the distal segment (6.1 � 8% vs.3 � 0.6%, p � 0.003) but not from the proximal segment7.9 � 10% vs. 13 � 0.6%, p � NS). The stented vesselegments in both groups showed no diameter change (Table). The reference segment not related to the stentedegment showed pacing-induced vasodilatation in bothroups (13 � 0.6% in BES, 10 � 0.4% in SES, p � NS).ntracoronary nitrates were associated with maximal vaso-ilatation of all of the evaluated vessel segments (proximal,istal, and reference) in both groups (Table 3). The maxi-al vasodilatation observed after nitroglycerin administra-

ion was not statistically significantly different for any of theegments between the 2 groups.

iscussion

ur study is the first to show that a newer-generation DES isssociated with better endothelium-dependent coronary vaso-otor response than SES, whereas the nonendothelium-

ependent vasomotor response was preserved in all cases andid not differ significantly between the 2 groups. Thesendings pertain only to the 2 stents under study. Despitehe similarities between biolimus and sirolimus, it is un-nown whether our findings are related to stent, polymer, oro a drug class effect.ndothelial function after stent implantation. Implanta-

ion of balloon expandable stents is necessarily associatedith vessel wall injury and trauma to the coronary endothe-

emodynamic Data

Table 2 Hemodynamic Data

SES (n � 15) BES (n � 19) p Value

Mean arterial blood pressure atbaseline (mm Hg)

96.9 � 13.9 95.1 � 11.9 0.71

Mean arterial blood pressure atpeak pacing (mm Hg)

101 � 17 95.5 � 16 0.36

Heart rate at baseline (beats/min) 79.1 � 20.3 77.5 � 13.9 0.8

Maximal pacing rate (beats/min) 133.1 � 19.7 133.7 � 14.2 0.91

Rate-pressure product at baseline(mm Hg � beats/min)

10,202 � 4,174 10,039 � 2,436 0.89

Rate-pressure product at peakpacing (mm Hg � beats/min)

15,703 � 3,913 15,258 � 3,123 0.72

bbreviations as in Table 1.

ium. This trauma sets in motion the inflammatory reaction

aeocvrfgvpm

mao

Qua

ntit

ativ

eC

oron

ary

Ang

iogr

aphy

Mea

sure

men

tsin

Bot

hG

roup

sat

Bas

elin

e,M

axim

alP

acin

g,an

dA

fter

Nit

rogl

ycer

inA

dmin

istr

atio

n

able

3Q

uant

itat

ive

Cor

onar

yA

ngio

grap

hyM

easu

rem

ents

inB

oth

Gro

ups

atB

asel

ine,

Max

imal

Pac

ing,

and

Aft

erN

itro

glyc

erin

Adm

inis

trat

ion

Pro

xim

alD

ista

lR

efer

ence

Ste

nt

BES

(n�

15)

SES

(n�

13)

BES

(n�

19)

SES

(n�

15)

BES

(n�

19)

SES

(n�

15

)B

ES(n

�1

9)

SES

(n�

15

)

asel

ine

3.1

8�

1.0

3.1

8�

0.7

02

.49

�0

.43

2.4

3�

0.5

42

.47

�0

.56

2.5

4�

0.6

03

.27

�0

.37

3.0

3�

0.5

0

axim

umpa

cing

3.3

9�

0.9

9*

3.1

2�

0.7

42

.62

�0

.34

*2

.30

�0

.57

*2

.79

�0

.58

*2

.79

�0

.60

*3

.27

�0

.37

3.0

3�

0.4

8

(%ch

ange

from

base

line)

(�7

.9�

10

)†(�

2.3

�1

0)

(�6

.1�

8)†

(�5

.4�

9)

(�1

3�

0.6

)(�

10

�0

.4)

(0�

1)

(0�

2)

itrog

lyce

rine

3.6

4�

1.0

6*

3.5

0�

0.9

0*

2.8

8�

0.4

3*

2.7

7�

0.5

1*

3.0

5�

0.5

6*

3�

0.6

0*

3.2

8�

0.3

93

.06

�0

.51

esar

em

ean

ofdi

amet

ers

inm

m�

SD

.*p

�0

.05

from

base

line.

†p

�0

.05

from

SES

.bb

revi

atio

nsas

inTa

ble

1.

2126 Hamilos et al. JACC Vol. 51, No. 22, 2008Endothelium and DES June 3, 2008:2123–9

nd healing response that eventually leads to re-ndothelialization. In the first month after the implantationf a bare-metal stent, a new immature endothelial layerovers the stent struts, reestablishing a normal coronaryessel wall lining. Autopsy studies suggest that endotheliumecovery is complete 3 months after the implantation, butunctionality of the neo-endothelium has not been investi-ated (12). In a study by Maier et al. (13), coronaryasomotion after exercise testing during coronary angiogra-hy was normal in segments distal and proximal to bare-etal stent 6 months after implantation.After implantation of SES, a more pronounced inflam-atory response has been described that may occasionally be

ssociated with a local hypersensitivity reaction and eosin-philic infiltration. The synthetic polymer containing the

Figure 1 Distribution of DiameterChanges in Patients With BES and SES

Scatterplots showing the distribution of percent mean diameter changes frombaseline to maximal pacing rate in patients with biolimus A9-eluting stents(BES) and sirolimus-eluting stents (SES), in proximal (A) and distal (B)segments.

drug may be an important trigger of local coronary inflam-T B M N

Valu A

mpsvhpltcersaf

ialia(

cCsvtEiDdb(Aibgapmets

dsArnamesttePrwssttobhepo

afmsdv

2127JACC Vol. 51, No. 22, 2008 Hamilos et al.June 3, 2008:2123–9 Endothelium and DES

ation, even though the metal struts or the stent itself mayarticipate in this reaction. Coronary inflammation is re-ponsible for delayed re-endothelialization of stent andessel wall, which eventually results in a delayed vascularealing response. Sirolimus impairs the normal healingrocesses of the injured arterial wall by inhibiting endothe-

ial cell proliferation (14). In addition, sirolimus was showno enhance tissue factor expression in human endothelialells (15), and inhibits the migration and proliferation ofndothelium progenitor cells, which may play an importantole in endothelium regeneration (16,17). These findingsuggest that SES may impair both the endothelial regener-tion and the ability of endothelial cells to exert their normalunction.

Poor stent endothelialization means that stent struts aren direct contact with blood and its elements. Humanutopsies showed insufficient endothelial coverage and de-ayed arterial healing with SES, regardless of the delay frommplantation (18–23). These findings are complemented byngioscopy (24) and optical coherence tomographic imaging

Figure 2 Mean Diameter of the Proximal, Distal, andReference Segments in the BES and SES Groups

Mean diameter values at baseline, maximal pacing rate, and after nitroglycerinadministration in BES and SES groups, in proximal (A), distal (B), and refer-ence (A and B) segments. The p values for the comparison between BES andSES are indicated for both proximal and distal segments. Abbreviations as inFigure 1.

25), which showed incomplete endothelialization and un- p

overed stent struts 3 to 6 months after SES implantation.omplete or partial lack of re-endothelialization of stent

truts and vessel wall generates a long-lasting, unhealedessel wall surface favoring platelet adhesion and aggrega-ion, which may eventually cause thrombus formation.ndothelial recovery and vasomotor response after BES

mplantation. All of those concerns led to new targets inES development with use of bioerodable polymers and

rugs specifically developed for local application. The No-ori coronary stent system uses a bioresorbable polymerpolylactic acid) from which biolimus A9 is eluted. Biolimus9 is a new sirolimus derivative that binds to the cytosolic

mmunophilin FK binding protein 12. The formed complexinds to the mammalian target of rapamycin and inhibitsrowth factor-driven cell proliferation, including T-cellsnd vascular smooth muscle cells. It has the potential torevent restenosis in coronary stents by interrupting smoothuscle cell migration and proliferation. In addition, an

nhanced lipophilicity enables an increased uptake in localarget tissues and reduced presence in areas surrounding thetented segment or even in the systemic circulation.

In our study, BES showed a better-preserved endothelium-ependent vasomotor response compared with SES. There areeveral possible explanations for this observation. Biolimus9 is a more lipophilic drug than sirolimus, and upon

elease would quickly bind to the target lipid-rich tissue. Ofote, drug is present only on the vessel side (abluminally),nd as such enters into peripheral circulation only ininimal quantities (26). This results in a more localized

ffect and less systemic drug exposure. Animal studieshowed that after BES implantation, the tissue concentra-ion of the drug in segments 5 mm proximal and distal tohe stent edges is almost nonmeasurable (26) and stentndothelialization is more complete than with SES andES (27). In addition, SES and BES have different drug

elease kinetics: total drug content is released from the SESithin 60 days with more than 60% released shortly after

tent implantation (28), versus a small initial burst andustained simultaneous drug release and polymer degrada-ion taking place over 6 months in the BES (26), exposinghe surrounding tissue at any given time to a lower amountf drug. Finally, the polymer used with BES is expected toe absorbed in a few months. Because durable polymersave been held responsible for some of the late adversevents related to SES, it is expected that the degradation ofolymer will improve arterial healing and long-term safetyf BES.The fact that the majority of our patients with BES had

vasodilatory response to pacing is a robust indication thatunctional endothelial re-growth is almost complete 9onths after implantation. The findings of our study give

ome promise that newer-generation DES with differentrug and polymer platforms may be more respectful of theessel healing while remaining at least as effective (29,30).

In our study, the response of the different patients to

acing was by and large homogenous. Some patients with

SpaatiuwaePpito6saf

SpvauaimioSsacdit

rtcpBoti

sswsMcN2e

teSgc

urddn(Acg

C

Emtams

ATpa

RdA

R

2128 Hamilos et al. JACC Vol. 51, No. 22, 2008Endothelium and DES June 3, 2008:2123–9

ES seemed to have a normal vasodilatory response toacing, albeit blunted, whereas some patients with BES hadvasoconstrictive response. Because the time course of

rterial healing after DES placement may vary from patiento patient, endothelialization was probably more completen some patients than in others, for reasons that are not wellnderstood. In addition, the time after stent implantationas longer than in previous studies, which may theoretically

llow for further healing. The time course of recovery ofndothelial function still needs to be studied systematically.revious studies. To our knowledge there is no otherublished study that examines coronary vasomotion aftermplantation of a newer-generation DES. On the contrary,here are 3 published studies that all showed agreement withur findings that SES cause local coronary vasoconstrictionmonths after their implantation compared with bare metal

tents (5–7). Two of those studies (5,7) used acetylcholinend the third used exercise (6) to assess coronary endothelialunction.

The more pronounced vasoconstrictive response afterES implantation reported in those studies could be ex-lained by the different time points at which coronaryasomotion was tested and the different methods used tossess vasomotion. In 2 of those studies, acetylcholine wassed to test the endothelial function. Despite the fact thatcetylcholine infusion has been extensively used for measur-ng coronary vasomotor response, we believe that flow-

ediated changes induced by pacing are a more physiolog-cal stimulus than a pharmacological test for the assessmentf endothelial function.tudy limitations. Baseline endothelial function beforetent implantation was not assessed in our study. Thessessment of endothelial function in patients with signifi-ant coronary lesions is not practicable, first because it isifficult to stop antianginal vasoactive medication before the

ntervention, and second because it is not feasible to avoidhe use of intracoronary nitrates during the procedure.

At the time of the intervention, stent assignment was notandomized. Patient enrollment was done in waves: allreated patients in each institution complying with inclusionriteria were treated simultaneously with BES or SES atre-defined study periods determined by the availability ofES. Because the assignment of stent type was dependentnly on the time point of study inclusion, bias with regardo stent-type decision was minimal and has probably notnfluenced the result.

We assessed endothelial function 9 months after bothtent implantations, and accordingly our results refer to thispecific time point. Further studies are needed to seehether the differences observed between the 2 stents are

till present more than 9 months after their implantation.oreover this study cannot indicate differences in the

linical outcomes after both stent implantations, althoughOBORI CORE data at 9 months show equivalence of thestents in terms of safety, angiographic results, and clinical

fficacy (30).

Calcium channel blockers were more frequently used inhe SES group. The 24-h period of interruption is longnough for a complete washout. Moreover, the fact that theES group showed more vasoconstriction than the BESroup indicates that the difference in the use of these drugsould not have influenced our results.

Finally, the clinical relevance of our findings remainsnknown. Local coronary vasoconstriction is only an indi-ect sign of endothelial dysfunction, probably caused byelayed endothelialization in the segments proximal andistal to the stent. Delayed endothelialization in the coro-ary vessels has been closely related to several adverse events31), the most severe of which is thrombus generation.ccordingly this finding could partially explain the in-

reased late thrombotic events observed with the first-eneration DES.

onclusions

ndothelium-dependent vasomotion at adjacent stent seg-ents seems to be better preserved after BES implantation

han SES implantation. This finding is a strong indicator offunctionally complete endothelial regeneration, a fact thatay hold promise for reducing thrombotic events with this

tent platform.

cknowledgmentshe authors thank all of the patients who agreed toarticipate in this study. Special thanks to the techniciansnd support staff in all participating hospitals.

eprint requests and correspondence: Dr. William Wijns, Car-iovascular Center Aalst, OLV Clinic, Moorselbaan, 164, B-9300alst, Belgium. E-mail: William.Wijns@village.uunet.be.

EFERENCES

1. Camenzind E, Steg PG, Wijns W. Stent thrombosis late afterimplantation of first-generation drug-eluting stents: a cause for con-cern. Circulation 2007;115:1440–55.

2. Stone GW, Moses JW, Ellis SG, et al. Safety and efficacy of sirolimus-and paclitaxel-eluting coronary stents. N Engl J Med 2007;356:998–1008.

3. Lagerqvist B, James SK, Stenestrand U, SCAAR Study Group.Long-term outcomes with drug-eluting stents versus bare-metal stentsin Sweden. N Engl J Med 2007;356:1009–19.

4. Togni M, Raber L, Cocchia R, et al. Local vascular dysfunction aftercoronary paclitaxel-eluting stent implantation. Int J Cardiol 2007;120:212–20.

5. Hofma S, Van der Giessen W, Van Dalen B, et al. Indication oflong-term endothelial dysfunction after sirolimus-eluting stent im-plantation. Eur Heart J 2006;27:166–70.

6. Togni M, Windecker S, Cocchia R, et al. Sirolimus-eluting stentsassociated with paradoxic coronary vasoconstriction. J Am Coll Car-diol 2005;46:231–6.

7. Fuke S, Maekawa K, Kawamoto K, et al. Impaired endothelialvasomotor function after sirolimus-eluting stent implantation. Circ J2007;71:220–5.

8. Nabel EG, Selwyn AP, Ganz P. Paradoxical narrowing of atheroscle-rotic coronary arteries induced by increases in heart rate. Circulation1990;81;850–9.

9. Prasad A, Husain S, Quyyumi A. Abnormal flow-mediated epicardialvasomotion in human coronary arteries is improved by angiotensin-

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

F

2129JACC Vol. 51, No. 22, 2008 Hamilos et al.June 3, 2008:2123–9 Endothelium and DES

converting enzyme inhibition. A potential role of bradykinin. J AmColl Cardiol 33;1999:796–804.

0. Hanet C, Schroeder E, Michel X, et al. Flow-induced vasomotorresponse to tachycardia of the human internal mammary artery andsaphenous vein grafts late following bypass surgery. Circulation1991;84 Suppl:III268–74.

1. Hanet C, Evrard P, Jacquet L, Goenen M, Robert A. Flow-mediatedvasodilator response to tachycardia of epicardial coronary arteries ispreserved in heart transplant recipients. Circulation 1993;88:II257–62.

2. Grewe PH, Deneke T, Machraoui A, Barmeyer J, Muller KM. Acuteand chronic tissue response to coronary stent implantation: pathologicfindings in human specimen. J Am Coll Cardiol 2006;35:157–63.

3. Maier W, Windecker S, Küng A, et al. Exercise-induced coronaryartery vasodilation is not impaired by stent placement. Circulation2002;105;2373–7.

4. Parry TJ, Brosius R, Thyagarajan R, et al. Drug-eluting stents:sirolimus and paclitaxel differentially affect cultured cells and injuredarteries. Eur J Pharmacol 2005;524:19–29.

5. Steffel J, Latini RA, Akhmedov A, et al. Rapamycin, but not FK-506,increases endothelial tissue factor expression: implications for drug-eluting stent design. Circulation 2005;112:2002–11.

6. Fukuda D, Sata M, Tanaka K, Nagai R. Potent inhibitory effect ofsirolimus on circulating vascular progenitor cells. Circulation 2005;111:926–31.

7. Inoue T, Sata M, Hikichi Y, et al. Mobilization of CD34-positivebone marrow-derived cells after coronary stent implantation: impacton restenosis. Circulation 2007;115:553–61.

8. Finn AV, Kolodgie FD, Harnek J, et al. Differential response ofdelayed healing and persistent inflammation at sites of overlappingsirolimus- or paclitaxel-eluting stents. Circulation 2005;112:270 – 8.

9. Finn AV, Nakazawa G, Joner M, et al. Vascular responses todrug-eluting stents: importance of delayed healing. ArteriosclerThromb Vasc Biol 2007;27:1500–10.

0. Finn AV, Joner M, Nakazawa G, et al. Pathological correlates of latedrug-eluting stent thrombosis. Strut coverage as a marker of endothe-lialization. Circulation 2007;115:2435–41.

1. Virmani R, Guagliumi G, Farb A, et al. Localized hypersensitivity andlate coronary thrombosis secondary to a sirolimus-eluting stent: should

we be cautious? Circulation 2004;109:701–5. p

2. Joner M, Finn A, Farb A, et al. Pathology of drug-eluting stents inhumans. Delayed healing and late thrombotic risk. J Am Coll Cardiol2006;48:193–202.

3. Luscher TF, Steffel J, Eberli FR, et al. Drug-eluting stent andcoronary thrombosis: biological mechanisms and clinical implications.Circulation 2007;115:1051–8.

4. Kotani J, Awata M, Nanto S, et al. Incomplete neointimal coverage ofsirolimus-eluting stents: angioscopic findings. J Am Coll Cardiol2006;47:2108–11.

5. Matsumoto D, Shite J, Shinke T, et al. Neointimal coverage ofsirolimus-eluting stents at 6-month follow-up: evaluated by opticalcoherence tomography. Eur Heart J 2007;28:961–7.

6. Ostojic M. Nobori Pharmacokinetics study. Available at: http://www.europcronline.com/fo/lecture. Accessed November 19, 2007.

7. Virmani R. Coronary drug-eluting stents: from animal studies toclinical practice. Available at: http://www.europcronline.com/fo/lecture. Accessed November 19, 2007.

8. Vetrovec G, Rizik D, Williard C, et al. Sirolimus PK trial: apharmacokinetic study of the sirolimus-eluting Bx velocity stent inpatients with de novo coronary lesions. Catheter Cardiovasc Interv2006;67:32–7.

9. Chevalier B, Serruys PW, Silber S, et al. Randomized comparison ofNobori, Biolimus A9-eluting coronary stent with a Taxus, paclitaxel-eluting coronary stent in patients with stenosis in native coronaryarteries: The Nobori 1 trial. Eurointerv 2007;2:426–34.

0. Ostojic M, Sagic D, Beleslin B, et al. First clinical comparison ofNobori-Biolimus A9 drug eluting stent with Cypher-Sirolimus elutingstent: NOBORI CORE nine months and one year clinical outcomes.Eurointerv 2008;3:574–9.

1. Schachinger V, Britten MB, Zeiher AM. Prognostic impact ofcoronary vasodilator dysfunction on adverse long-term outcome ofcoronary heart disease. Circulation 2000;101:1899–906.

APPENDIX

or a complete list of investigators,

lease see the online version of this article.