J A C C : C A R D I O V A S C U L A R I M A G I N G V O L . 1 , N O . 2 , 2 0 0 8

© 2 0 0 8 B Y T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N I S S N 1 9 3 6 - 8 7 8 X / 0 8 / $ 3 4 . 0 0

P U B L I S H E D B Y E L S E V I E R I N C . D O I : 1 0 . 1 0 1 6 / j . j c m g . 2 0 0 7 . 1 2 . 0 0 1

Impaired Coronary Vasodilation by MagneticResonance Angiography Is Associated WithAdvanced Coronary Artery Calcification

Masahiro Terashima, MD, PHD, FACC,* Patricia K. Nguyen, MD,* Geoffrey D. Rubin, MD,†Carlos Iribarren, MD, MPH, PHD,§ Brian K. Courtney, MD,* Alan S. Go, MD,§�Stephen P. Fortmann, MD,‡ Michael V. McConnell, MD, MSEE, FACC*

Stanford, Oakland, and San Francisco, California

O B J E C T I V E S This study evaluated the hypothesis that impaired nitroglycerin (NTG)-induced

coronary vasodilation is associated with advanced coronary atherosclerosis in asymptomatic older

patients.

B A C K G R O U N D Atherosclerosis is associated with both structural and functional abnormalities of

the vessel wall. Noninvasive functional measures of subclinical coronary atherosclerosis may help

characterize high-risk subjects and guide preventive therapy.

M E T H O D S A total of 236 older patients (age 60 to 72 years, 33% female) without a history of

cardiovascular disease were studied. Nitroglycerin-induced coronary vasodilation was measured by

magnetic resonance angiography (MRA). Cross-sectional images of the right coronary artery were

acquired before and 5 min after 0.4-mg sublingual NTG using a gated, breath-held spiral coronary MRA

sequence (0.7-mm resolution). Quantitative analysis of the increase in cross-sectional area was

performed in the 90% of patients (n � 212) with adequate image quality. Quantitation of coronary artery

calcification (CAC) was performed by multidetector computed tomography using the Agatston method.

R E S U L T S Forty patients (19%) had advanced CAC (�400). Coronary vasodilation to NTG was

significantly impaired (p � 0.02) in patients with advanced CAC (median [interquartile range] � 15.9%

[4.2% to 28.0%] vs. 21.5% [9.6% to 36.6%] for CAC �400). Importantly, NTG-induced coronary

vasodilation remained independently associated with advanced CAC after multivariate analysis incor-

porating risk factors (p � 0.02) and other potential confounders (p � 0.04). There was no significant

difference in coronary vasodilation between men and women, but few women (n � 3) had advanced

CAC.

C O N C L U S I O N S Impaired NTG-induced coronary vasodilation by MRA is associated with advanced

coronary atherosclerosis in a community-based cohort of older asymptomatic subjects. Coronary MRA

may provide a noninvasive functional assessment of subclinical coronary atherosclerosis. (J Am Coll

Cardiol Img 2008;1:167–73) © 2008 by the American College of Cardiology Foundation

From the *Division of Cardiovascular Medicine, †Department of Radiology, and ‡Stanford Prevention Research Center,Stanford University School of Medicine, Stanford, California; §Division of Research, Kaiser Permanente of NorthernCalifornia, Oakland, California; and the �Departments of Epidemiology, Biostatistics, and Medicine, University of California,San Francisco, San Francisco, California. This study was supported by grants from the Donald W. Reynolds Foundation (LasVegas, Nevada); National Heart, Lung, and Blood Institute (grant R01HL39297, Bethesda, Maryland); and General ElectricHealthcare (Waukesha, Wisconsin). Dr. Rubin is a consultant/advisory board member for TeraRecon, Inc., and SiemensMedical Solutions. Dr. McConnell receives research support from GE Healthcare, is on a scientific advisory board for Kowa,Inc., and has been an advisor for Philips Medical Systems.

Manuscript received October 18, 2007; revised manuscript received December 5, 2007, accepted December 8, 2007.

Ntarttcwnsac

c

md(coa

M

Pthmd

rgnsnAFTKotiwtvicl6CfipIaNNMmsem4vtcrm(sinb�stm

p(tC

A

A

C

c

C

E

H

M

c

M

a

M

i

N

R

R

c

c

SNR � signal-to-noise ratio

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 1 , N O . 2 , 2 0 0 8

M A R C H 2 0 0 8 : 1 6 7 – 7 3

Terashima et al.

Impaired Coronary Vasodilation by MRA

168

oninvasive measures of subclinical coro-nary atherosclerosis have the potential tohelp identify patients at increased risk forfuture coronary events and guide therapy

hat may reduce the risks of death and disabilityttributable to coronary artery disease. Atheroscle-osis is associated with both structural and func-ional abnormalities of the vessel wall. For struc-ural assessment, coronary artery calcium (CAC) byomputed tomography (CT) has been the mostidely studied direct noninvasive measure of coro-ary atherosclerosis. Coronary artery calcium isupported by histological studies, with the presencend extent of CAC correlated to the magnitude oforonary plaque burden (1).

See page 174

For functional assessment, the standardhas been the measurement of epicardialcoronary artery vasodilation by invasiveX-ray angiography using intra-arterial stim-uli. The impairment of endothelium-dependent vasodilation is recognized as anearly step in the development of athero-sclerosis (2). Nitroglycerin (NTG) is anexogenous source of nitric oxide andcauses endothelium-independent, smoothmuscle cell– dependent vasodilation,which may also be affected by atheroscle-rotic changes in the vessel wall. Previousstudies have demonstrated that impairedcoronary vasodilation to NTG, as assessedby invasive X-ray angiography, is presentin patients who have coronary artery dis-ease and diabetes (3), and has been asso-ciated with an increased risk of future

ardiac events (4,5).We and others have developed a noninvasiveethod to measure NTG-induced coronary vaso-

ilation with magnetic resonance angiographyMRA) (6,7). In this study, we hypothesized thatoronary vasodilation by MRA may be impaired inlder patients with advanced subclinical coronarytherosclerosis.

E T H O D S

atient characteristics. Two hundred fifty-three pa-ients (age 60 to 72 years, 34% women) without aistory of cardiovascular disease, other major co-orbidities (including prior systemic malignancy,

ein

ht

ementia, cirrhosis, human immunodeficiency vi- a

us/acquired immunodeficiency syndrome, and or-an transplantation), or contraindication to mag-etic resonance imaging (MRI) were randomlyelected from the membership of Kaiser Perma-ente of Northern California as a substudy of theDVANCE (Atherosclerotic Disease, Vascularunction, and Genetic Epidemiology) study (8).he recruitment of these patients was performed byaiser Permanente of Northern California Divisionf Research, and all subjects underwent evaluationhrough the Stanford Prevention Research Center,ncluding CT for CAC. Patients were evaluatedithout any alteration of their medical regimen per

heir health care provider, including continuation ofasoactive medications. Magnetic resonance imag-ng was not completed in 17 subjects due tolaustrophobia, back pain, or technical problems,eaving an MRI cohort of 236 patients (mean age6.0 � 2.7 years, 33% women, and 37% non-aucasian). Magnetic resonance imaging was per-

ormed within 1 week of CT in all subjects. Writtennformed consent was obtained from all partici-ants. The study protocol was approved by thenstitutional Review Boards at Stanford Universitynd the Kaiser Foundation Research Institute.TG-induced coronary vasodilation with MRA.itroglycerin-induced coronary vasodilation byRI was performed using methods described inore detail previously (7). A 1.5-T Signa MRI

canner (GE Healthcare, Waukesha, Wisconsin)quipped with high-performance gradients (40T/m, 150 mT/m/ms) was used. A commercial

-channel cardiac phased-array surface coil pro-ided signal reception (GE Healthcare). A real-ime interactive MRI system (iDrive, GE Health-are) was used for coronary localization. High-esolution coronary MRA was performed using aultislice spiral sequence, as previously reported

9), with cardiac gating, breath-holding, and acqui-ition during diastole (field of view � 22 cm,n-plane spatial resolution � 0.7 mm, slice thick-ess � 5 mm, 3 slices, repetition time � 1 hearteat, echo time � 2.5 ms, 18 interleaves, flip angle

60°, acquisition window � 35 ms). A subject-pecific trigger delay time during diastole was de-ermined from reviewing a 4-chamber cardiac cineagnetic resonance acquisition (10).After real-time interactive MRI was used to

rescribe in-plane views of the right coronary arteryRCA), a cross-sectional view of a linear portion ofhe proximal- to mid-RCA was then prescribed.ross-sectional high-resolution coronary MRA im-

B B R E V I A T I O N S

N D A C R O N YM S

AC � coronary artery

alcium/calcification

T � computed tomography

CG � electrocardiography

DL � high-density lipoprot

DCT � multidetector

omputed tomography

RA � magnetic resonance

ngiography

RI � magnetic resonance

maging

TG � nitroglycerin

CA � right coronary artery

CAC � coronary artery

alcification score for the rig

oronary artery

ges were then acquired both before and 5 min after

0sapimawtntb(csvitcCssSAt

heagitstAtwCwssapSpdbuoW

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 1 , N O . 2 , 2 0 0 8

M A R C H 2 0 0 8 : 1 6 7 – 7 3

Terashima et al.

Impaired Coronary Vasodilation by MRA

169

.4-mg sublingual NTG was administered to theubject while in the magnet (Fig. 1). The time pointfter NTG (5 min) was chosen based on ourrevious data (7) on the time course of NTG-nduced coronary vasodilation, demonstrating the

ost consistent coronary vasodilation at 3 to 7 minfter NTG. Matching pre- and post-NTG imagesere identified, as described previously (7), and

hen all images were pooled and randomized, witheither patient nor NTG information provided onhe images. Prior to analysis, images were judged toe good, fair, or poor based on signal-to-noise ratioSNR), vessel border sharpness, and artifacts. Theross-sectional area of the RCA was traced by aingle investigator, with a similar low intraobserverariability as found previously (7) (mean differencen cross-sectional area, expressed as a percentage ofhe overall mean, was 6.3 � 5.0% with correlationoefficient � 0.97).AC. The presence and severity of CAC was as-essed by 4- or 16-row multidetector CT (MDCT)canners (GE Healthcare and Siemens Medicalolutions, Erlangen, Germany, respectively) (8).fter a scout image had been acquired, a prospec-

Figure 1. Coronary MRA of NTG-Induced Coronary Vasodilation

Coronary magnetic resonance angiography (MRA) images are showtion. Coronary MRA of the right coronary artery (RCA) was performesequence. Cross-sectional area of the RCA was traced manually (redcalcium (CAC) score � 0 show a greater degree of NTG coronary vawith a CAC score � 443 (coronary vasodilation � 17.0%).

ive electrocardiography (ECG)-triggered breath- C

eld CT scan through the heart was performed withither a 4 � 2.5-mm or a 16 � 1.25-mm nonhelicalcquisition, gated to mid-diastole based upon araphical prescription from the ECG tracing. Twodentical scans were acquired. Contiguous 2.5-mmransverse sections were reconstructed using a half-can reconstruction algorithm to give an effectiveemporal resolution of 200 to 250 ms per section.n experienced technologist analyzed the CT sec-

ions using an AccuImage coronary calcium scoringorkstation (Merge/eFilm, South San Francisco,alifornia) to determine the Agatston score (11),ith the final score equal to the average of the 2

cans (providing both total CAC score and CACcore for the RCA). The total CAC was considereddvanced if the score was �400, as indicated byrevious reports (12,13).tatistical analysis. Continuous variables are ex-ressed as means with standard deviations or me-ians with interquartile ranges. The differencesetween CAC groups were compared using thenpaired t test if data were normally distributed; iftherwise distributed, the nonparametric Mann-

hitney U test was used. Differences among the 4

monstrating the analysis of nitroglycerin (NTG)-induced vasodila-re- and post-NTG using a 0.7-mm gated, breath-held spiral MRAcles). The MRA images (A) from the patient with a coronary arteryilation (28.6%) compared with MRA images (B) from the patient

n ded pcirsod

AC groups, 0 � CAC � 10, 10 � CAC � 100,

1adawvaiism�clmcv

R

PwiMMFag�mmNtitc[miv

cfbrtpiac

lrv0fl(33q�piNaiw21lnvwmNaotsdTp�ipp(30

D

C

CAC � coronary artery ca

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 1 , N O . 2 , 2 0 0 8

M A R C H 2 0 0 8 : 1 6 7 – 7 3

Terashima et al.

Impaired Coronary Vasodilation by MRA

170

00 � CAC � 400, and CAC � 400, werenalyzed by the Kruskal-Wallis test. Differences inistribution of CAC scores �400 and �400 weressessed by the chi-square test. Regression analysisas performed comparing NTG-induced coronaryasodilation to log10 (CAC score � 1). Multivari-ble logistic regression analysis was performed todentify independent predictors of advanced CAC,ncorporating conventional coronary risk factors,uch as age, gender, body mass index, diabetesellitus (history of treatment or fasting glucose126 mg/dl), systolic blood pressure, total

holesterol/high-density lipoprotein (HDL) cho-esterol ratio, and smoking into the model. For

ultivariate analysis, tertiles of the percentages oforonary vasodilation were also used. A 2-tailed palue �0.05 was considered statistically significant.

E S U L T S

atient characteristics. Of the 236 patients, 10%ere excluded prior to analysis because of poor

mage quality on either the pre- or post-NTGRA, leaving 212 subjects for quantitative analysis.edian CAC was 37 (interquartile range 1 to 269).

orty patients (19%) had advanced CAC, defineds CAC �400. Clinical characteristics of the 2roups are shown in Table 1. In subjects with CAC400, there was a significantly higher proportion ofen (p � 0.001) and of subjects with diabetesellitus (p � 0.04).TG-induced coronary vasodilation and coronary ar-ery calcium. Patients with CAC �400 had signif-cantly impaired coronary vasodilation (median [in-erquartile range] � 15.9% [4.2% to 28.0%])ompared with patients with CAC �400 (21.5%9.6% to 36.6%], p � 0.02) (Fig. 2). The results ofultivariable logistic regression analysis are shown

n Table 2. The degree of NTG-induced coronary

cs of Study Patients Stratified by CAC Score

sticsCAC <400(n � 172)

CAC >400(n � 40) p Value

65.9 � 2.7 66.4 � 2.7 0.30

106/66 37/3 <0.0012) 26.9 � 4.1 27.9 � 4.7 0.19

15.7% 30% 0.04

(mm Hg) 132 � 18 135 � 19 0.32

cholesterol ratio 3.9 � 1.1 3.9 � 0.9 0.68

8.1% 10% 0.70

n � standard deviation or proportions. Values in bold indicate statistical

lcium; HDL � high-density lipoprotein.

asodilation was independently and inversely asso- N

iated with CAC �400 (p � 0.02) after adjustmentor age, gender, body mass index, diabetes, systoliclood pressure, total cholesterol/HDL cholesterolatio, and current cigarette smoking. The associa-ion of NTG-induced coronary vasodilation inatients with advanced CAC also remained signif-cant (p � 0.04) in multivariable logistic regressionnalysis after vasoactive medication use was in-luded as a potential confounder.

Further analysis of NTG coronary vasodilation iness advanced CAC did not reveal a “dose-response”elationship, as regression analysis of vasodilationersus log10 (CAC � 1) was not significant (p �.15), and the degree of vasodilation was relativelyat for CAC subgroups: 0 � CAC � 10: 21.4%10.7% to 36.2%); 10 � CAC � 100: 25.1% (8.8 to6.8%); and 100 � CAC � 400: 21.4% (9.7% to7.3%) (p � 0.16). High CAC did affect imageuality (CAC �400: 68% fair, 32% good vs. CAC400: 45% fair, 55% good, p � 0.01), but the

ercent vasodilation was similar for fair and goodmage quality (p � 0.87).TG-induced coronary vasodilation by gender. Over-ll there was not a significant difference in NTG-nduced coronary vasodilation between men andomen (men: 20.8% [8.0% to 36.3%] vs. women:1.1% [6.9% to 34.0%], p � 0.9). In men only (n �43), the impaired NTG-induced coronary vasodi-ation in patients with CAC �400 remained sig-ificant by both univariate (p � 0.02) and multi-ariate analysis (p � 0.01). The sample size foromen (n � 3 for CAC �400) was too small foreaningful statistical evaluation.T-induced coronary vasodilation and RCA coronaryrtery calcium. As the MRI measurement was donen the RCA, we also compared the MRI vasodila-ion measures to the coronary artery calcificationcore for the right coronary artery (RCAC). Me-ian RCAC was 0 (interquartile range 0 to 47).here was �RCAC (RCAC score �0) in 86atients (41%, 71 male). Only 2 subjects with CAC400 did not have �RCAC. The degree of NTG-

nduced coronary vasodilation was significantly im-aired in patients with �RCAC compared toatients with �RCAC by univariate analysis17.4% [5.0% to 33.7%] vs. 22.6% [11.4% to6.5%], p � 0.049) and multivariate analysis (p �.03).

I S C U S S I O N

oronary MRA was successfully applied to assess

Table 1. Characteristi

Characteri

Age (yrs)

Gender (men/women)

Body mass index (kg/m

Diabetes mellitus (%)

Systolic blood pressure

Total cholesterol/HDL

Current smoker (%)

Results are given as measignificance.

TG vasodilation in a large cohort of asymptom-

atpmisMcCfpdetbea

dpmselasciiii-wNcecciTtemcou

pbtrvewbo

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 1 , N O . 2 , 2 0 0 8

M A R C H 2 0 0 8 : 1 6 7 – 7 3

Terashima et al.

Impaired Coronary Vasodilation by MRA

171

tic, older community-based patients. We foundhat NTG coronary vasodilation was impaired inatients with advanced CAC, which were primarilyen in this cohort. In addition, this association was

ndependent of other major coronary risk factors,uggesting that NTG coronary vasodilation by

RA may be a functional measure of subclinicaloronary atherosclerosis.oronary vasodilation and CAC. Coronary vasomotorunction plays an important role in the onset androgression of coronary atherosclerosis. Endothelialysfunction, as measured by the impairment ofndothelium-dependent vasodilation, is consideredo be an early marker of atherosclerosis, which cane found without overt angiographic or ultrasoundvidence of coronary atherosclerosis (14,15). The

Figure 2. Comparison of NTG-Induced Coronary Vasodilationand CAC

The degree of coronary vasodilation is shown for patients withand without advanced CAC. Coronary MRA images were ana-lyzed in 212 patients and coronary vasodilation to NTG wasquantified. Coronary artery calcium scoring was performed bycomputed tomography. Box and whisker plots show themedian, interquartile range, and full range of percent coronaryvasodilation between groups with and without advanced CAC(�400). Patients with CAC �400 had significantly impaired NTG-induced coronary vasodilation compared to patients with CAC�400 by univariate and multivariate analysis (p � 0.02). Abbre-viations as in Figure 1.

ssessment of epicardial coronary vasodilation is

istinct from the assessment of coronary flow orerfusion reserve, which are primarily indicators oficrovascular dysfunction or flow-limiting steno-

es. Several studies have found a predictive value ofndothelium-dependent epicardial coronary vasodi-ation in response to intracoronary infusion ofcetylcholine in patients with mild coronary athero-clerosis or normal coronary angiograms by invasiveoronary angiography (4,15,16). Although NTG-nduced (endothelium-independent) vasodilation has typ-cally been used as a “positive control” in these patients,mpairment has been found (17). Schachinger et al. (4)nvestigated both endothelium-dependent andindependent coronary vasodilation in 147 patientsith coronary artery disease and reported thatTG-induced coronary vasodilation was a signifi-

ant predictor of prognosis. Similarly, von Meringt al. (5) studied 163 women with suspected myo-ardial ischemia and reported that women with aonfirmed acute ischemic event had significantlympaired NTG-induced coronary vasodilation.hese data suggest that atherosclerotic changes to

he vessel wall can impair vasodilation beyond theffects on the endothelium, although the preciseechanisms underlying the impairment remain un-

lear. Our data suggest that there may be a thresh-ld effect, with NTG vasodilation remaining intactntil atherosclerosis is advanced.Interestingly, smooth muscle cells are not only the

rimary mediators of NTG-induced vasodilation (18),ut are also thought to be responsible for the forma-ion of vascular calcification (19,20). Kullo et al. (21)ecently reported that a decreased brachial arteryasodilator response to NTG, but not flow-mediatedndothelium-dependent vasodilation, is associatedith subclinical coronary atherosclerosis as measuredy CAC in 441 asymptomatic patients. A recent studyf subclinical atherosclerosis in 292 participants from

Table 2. Multivariate Analysis of NTG-Induced Coronary VasodiCAC (>400)

Variable

Adjusted Odd(95% Confid

Interval

NTG coronary vasodilation (per SD increment) 0.60 (0.34–0

Age (per yr) 1.08 (0.94–1

Gender (men vs. women) 8.45 (2.37–3

Body mass index (per kg/m2) 1.03 (0.94–1

Diabetes mellitus (yes/no) 1.66 (0.7–3.9

Systolic blood pressure (per mm Hg) 1.01 (0.99–1

Total cholesterol/HDL cholesterol ratio 0.87 (0.53–1

Current smoker (yes/no) 1.25 (0.32–4

Values in bold indicate statistical significance.

lation and Advanced

s Ratioence)

pValue

.91) 0.02

.24) 0.28

0.09) 0.001

.13) 0.54

5) 0.25

.03) 0.54

.42) 0.57

.94) 0.75

NTG � nitroglycerin; SD � standard deviation; other abbreviations as in Table 1.

tccTtanlacRgoetbgoIchsId�aicserH1s(SriPded

epMstcmNwh(op

hatsawtia

C

IiMwot

AToSt

RMSD

R

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 1 , N O . 2 , 2 0 0 8

M A R C H 2 0 0 8 : 1 6 7 – 7 3

Terashima et al.

Impaired Coronary Vasodilation by MRA

172

he Framingham Heart Study included aortic MRI,oronary and thoracic aortic calcification by CT, andarotid intima-media thickness by ultrasound (22).he results indicated that a combination of imaging

ests may be needed to fully identify higher-risksymptomatic patients. Further studies of the prog-ostic significance of NTG-induced coronary vasodi-

ation by coronary MRA and its value relative to CACnd other cardiovascular tests and risk factors arelearly needed.isk factors. According to previous reports, age andender are strong risk factors for advanced burdenf CAC (23–26). In our study, the age range ofnrolled subjects was narrow (60 to 72 years), andherefore we did not observe a significant associationetween age and advanced CAC. However, maleender was a strong predictor of advanced CAC inur study, which is consistent with prior studies (26).n other studies of asymptomatic patients, manyonventional coronary risk factors, such as obesity,ypertension, hypercholesterolemia, diabetes, andmoking, were strongly associated with CAC (23,27).nterestingly, in our study, only male gender andiabetes were significantly associated with CAC400. Besides the sample size and the relatively

dvanced age range, this may have been attributed to,n part, the relatively low prevalence of the otheroronary risk factors in our sample compared withamples from previous studies (Table 1) (23,28). Forxample, the mean total cholesterol/HDL cholesterolatio and systolic blood pressure were 3.9 and 133 mm

g in our study compared with 4.0 to 5.3 and 139 to44 mm Hg in other studies (23,28), and currentmokers were 8.5% compared with 15% to 19%23,28).tudy limitations. The main limitation of the cur-ent study is that we assessed only endothelium-ndependent coronary vasodilation using NTG.rior studies indicate that an endothelium-ependent method may detect abnormalities at anarlier stage of disease (2,29) and produce moreivergent responses (vasoconstriction vs. vasodila-

2157– 62. measurements of lu

ndothelium-dependent stimuli, such as the cold-ressor test and mental stress, are challenging in theRI environment, and pharmacological agents,

uch as acetylcholine, cannot safely be given sys-emically. A preliminary MRI report using theold-pressor test shows promise (30). The otherain limitation is the interpatient variability ofTG vasodilation in this study. Although the RCAas used for higher SNR and spatial resolution, andas been shown previously to be very reproducible7,31), further improvements (32,33) in SNR, res-lution, and breath-hold reproducibility may im-rove image quality and reduce variability.Multidetector CT angiography has emerged as a

igh-resolution noninvasive approach for coronaryrtery imaging. A recent study did look retrospec-ively at patients who had more than 1 coronary CTcan, where NTG was used in 1 and not in another,nd did show significantly larger coronary diameterith NTG (34). However, the radiation and con-

rast involved make it suboptimal for repeat imag-ng to measure coronary vasodilation, particularly insymptomatic subjects.

O N C L U S I O N S

n an asymptomatic older community patient cohort,mpaired NTG-induced coronary vasodilation by

RA was significantly and independently associatedith advanced CAC. Noninvasive assessment of cor-nary vasodilation may provide an additional func-ional measure of subclinical coronary atherosclerosis.

cknowledgmentshe authors thank the staff of the Kaiser Permanentef Northern California Division of Research and thetanford Prevention Research Center for their assis-ance in patient recruitment and evaluation.

eprint requests and correspondence: Dr. Michael V.cConnell, Division of Cardiovascular Medicine,

tanford University School of Medicine, 300 Pasteurrive, Room H-2157, Stanford, California 94305-5233.

tion) to improve discrimination. Nonpharmacologic E-mail: mcconnell@stanford.edu.

E F E R E N C E S

1. Rumberger JA, Simons DB, Fitz-patrick LA, Sheedy PF, Schwartz RS.Coronary artery calcium area byelectron-beam computed tomogra-phy and coronary atheroscleroticplaque area. A histopathologic cor-relative study. Circulation 1995;92:

2. Bonetti PO, Lerman LO, Lerman A.Endothelial dysfunction: a marker ofatherosclerotic risk. ArteriosclerThromb Vasc Biol 2003;23:168–75.

3. Vavuranakis M, Stefanadis C, Tri-andaphyllidi E, Toutouzas K, Toutou-zas P. Coronary artery distensibility indiabetic patients with simultaneous

minal area and in-

tracoronary pressure: evidence of im-paired reactivity to nitroglycerin. J AmColl Cardiol 1999;34:1075–81.

4. Schachinger V, Britten MB, ZeiherAM. Prognostic impact of coronaryvasodilator dysfunction on adverselong-term outcome of coronary heartdisease. Circulation 2000;101:1899–

906.

1

1

1

1

1

2

2

2

2

3

3

3

3

3

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 1 , N O . 2 , 2 0 0 8

M A R C H 2 0 0 8 : 1 6 7 – 7 3

Terashima et al.

Impaired Coronary Vasodilation by MRA

173

5. von Mering GO, Arant CB, WesselTR, et al. Abnormal coronary vasomo-tion as a prognostic indicator of cardio-vascular events in women: results fromthe National Heart, Lung, and BloodInstitute-Sponsored Women’s IschemiaSyndrome Evaluation (WISE). Circu-lation 2004;109:722–5.

6. Pepe A, Lombardi M, Takacs I, Posi-tano V, Panzarella G, Picano E.Nitrate-induced coronary vasodilationby stress-magnetic resonance imaging:a novel noninvasive test of coronaryvasomotion. J Magn Reson Imaging2004;20:390–4.

7. Terashima M, Meyer CH, Keeffe BG,et al. Noninvasive assessment of cor-onary vasodilation using magnetic res-onance angiography. J Am Coll Car-diol 2005;45:104–10.

8. Fair JM, Kiazand A, Varady A, et al.Ethnic differences in coronary arterycalcium in a healthy cohort aged 60 to69 years. Am J Cardiol 2007;100:981–5.

9. Yang PC, Meyer CH, Terashima M,et al. Spiral magnetic resonance coro-nary angiography with rapid real-timelocalization. J Am Coll Cardiol 2003;41:1134–41.

0. Kim WY, Stuber M, Kissinger KV,Andersen NT, Manning WJ, BotnarRM. Impact of bulk cardiac motionon right coronary MR angiographyand vessel wall imaging. J Magn Re-son Imaging 2001;14:383–90.

1. Agatston AS, Janowitz WR, HildnerFJ, Zusmer NR, Viamonte M Jr.,Detrano R. Quantification of coronaryartery calcium using ultrafast com-puted tomography. J Am Coll Cardiol1990;15:827–32.

2. He ZX, Hedrick TD, Pratt CM, et al.Severity of coronary artery calcificationby electron beam computed tomogra-phy predicts silent myocardial ischemia.Circulation 2000;101:244–51.

3. Newman AB, Naydeck BL, Sutton-Tyrrell K, et al. Relationship betweencoronary artery calcification and othermeasures of subclinical cardiovasculardisease in older adults. ArteriosclerThromb Vasc Biol 2002;22:1674–9.

4. Nishimura RA, Lerman A, ChesebroJH, et al. Epicardial vasomotor re-sponses to acetylcholine are not pre-dicted by coronary atherosclerosis asassessed by intracoronary ultrasound.

J Am Coll Cardiol 1995;26:41–9.

15. Suwaidi JA, Hamasaki S, Higano ST,Nishimura RA, Holmes DR Jr., Ler-man A. Long-term follow-up of pa-tients with mild coronary artery dis-ease and endothelial dysfunction.Circulation 2000;101:948–54.

16. Halcox JP, Schenke WH, Zalos G, etal. Prognostic value of coronary vascu-lar endothelial dysfunction. Circula-tion 2002;106:653–8.

17. Adams MR, Robinson J, McCredieR, et al. Smooth muscle dysfunctionoccurs independently of impairedendothelium-dependent dilation inadults at risk of atherosclerosis. J AmColl Cardiol 1998;32:123–7.

18. Mehta JL. Endothelium, coronary va-sodilation, and organic nitrates. AmHeart J 1995;129:382–91.

19. Iyemere VP, Proudfoot D, WeissbergPL, Shanahan CM. Vascular smoothmuscle cell phenotypic plasticity andthe regulation of vascular calcification.J Intern Med 2006;260:192–210.

20. Trion A, van der Laarse A. Vascularsmooth muscle cells and calcificationin atherosclerosis. Am Heart J 2004;147:808–14.

21. Kullo IJ, Malik AR, Bielak LF,Sheedy PF 2nd, Turner ST, PeyserPA. Brachial artery diameter and va-sodilator response to nitroglycerine,but not flow-mediated dilatation, areassociated with the presence andquantity of coronary artery calcium inasymptomatic adults. Clin Sci (Lond)2007;112:175–82.

22. Kathiresan S, Larson MG, Keyes MJ,et al. Assessment by cardiovascularmagnetic resonance, electron beamcomputed tomography, and carotidultrasonography of the distribution ofsubclinical atherosclerosis across Fra-mingham risk strata. Am J Cardiol2007;99:310–4.

23. Oei HH, Vliegenthart R, Hofman A,Oudkerk M, Witteman JC. Risk fac-tors for coronary calcification in oldersubjects. The Rotterdam CoronaryCalcification Study. Eur Heart J 2004;25:48–55.

24. Hoff JA, Chomka EV, Krainik AJ,Daviglus M, Rich S, Kondos GT. Ageand gender distributions of coronaryartery calcium detected by electronbeam tomography in 35,246 adults.Am J Cardiol 2001;87:1335–9.

25. Raggi P, Shaw LJ, Berman DS, Cal-

lister TQ. Gender-based differences in

the prognostic value of coronary calci-fication. J Womens Health (Larchmt)2004;13:273–83.

6. Janowitz WR, Agatston AS, KaplanG, Viamonte M Jr. Differences inprevalence and extent of coronary ar-tery calcium detected by ultrafastcomputed tomography in asymptom-atic men and women. Am J Cardiol1993;72:247–54.

7. Hoff JA, Daviglus ML, ChomkaEV, Krainik AJ, Sevrukov A, Kon-dos GT. Conventional coronary ar-tery disease risk factors and coronaryartery calcium detected by electronbeam tomography in 30,908 healthyindividuals. Ann Epidemiol 2003;13:163–9.

8. Greenland P, LaBree L, Azen SP,Doherty TM, Detrano RC. Coronaryartery calcium score combined withFramingham score for risk predictionin asymptomatic individuals. JAMA2004;291:210–5.

9. Ludmer PL, Selwyn AP, Shook TL,et al. Paradoxical vasoconstriction in-duced by acetylcholine in atheroscle-rotic coronary arteries. N Engl J Med1986;315:1046–51.

0. Jain H, Borges A, Iwuchukwu C, et al.Noninvasive assessment of the rela-tionship of coronary vasomotion toplaque burden (abstr). Circulation2006;114 Suppl II:II541.

1. Keegan J, Horkaew P, Buchanan TJ,Smart TS, Yang GZ, Firmin DN.Intra- and interstudy reproducibilityof coronary artery diameter measure-ments in magnetic resonance coronaryangiography. J Magn Reson Imaging2004;20:160–6.

2. Yang PC, Nguyen P, Shimakawa A,et al. Spiral magnetic resonance coro-nary angiography—direct comparisonof 1.5 Tesla vs. 3 Tesla. J CardiovascMagn Reson 2004;6:877–84.

3. Santos JM, Cunningham CH, LustigM, et al. Single breath-hold whole-heart MRA using variable-densityspirals at 3T. Magn Reson Med 2006;55:371–9.

4. Dewey M, Hoffmann H, Hamm B.Multislice CT coronary angiography:effect of sublingual nitroglycerine onthe diameter of coronary arteries. Rofo

2006;178:600–4.