Vasodilation of epicardial coronary artery can be measured with transthoracic echocardiography

Ultrasound in Med. & Biol., Vol. 33, No. 3, pp. 362–370, 2007Copyright © 2007 World Federation for Ultrasound in Medicine & Biology

Printed in the USA. All rights reserved0301-5629/07/$–see front matter

doi:10.1016/j.ultrasmedbio.2006.08.012

● Original Contribution

VASODILATION OF EPICARDIAL CORONARY ARTERY CAN BEMEASURED WITH TRANSTHORACIC ECHOCARDIOGRAPHY

TUOMAS O. KIVINIEMI,* JYRI O. TOIKKA,*† JUHA W. KOSKENVUO,* ANTTI SARASTE,*‡

MARKKU SARASTE,* JUSSI P. PÄRKKÄ,* OLLI T. RAITAKARI,* and JAAKKO J. HARTIALA**Department of Clinical Physiology and Nuclear Medicine, Turku University Hospital, Turku, Finland; †Department

of Clinical Physiology, Tampere University Hospital, Tampere, Finland; and ‡Department of Medicine, TurkuUniversity Hospital, Turku, Finland

(Received 6 March 2006, revised 8 August 2006, in final form 17 August 2006)

Abstract—Transthoracic Doppler echocardiography (TTE) has been introduced as a noninvasive tool to measurecoronary flow velocity reserve (CFVR). Velocity measurement, however, fails to take into account epicardialcoronary artery vasodilation during hyperemia and this may cause underestimation of CFVR measurements.Therefore, we sought to determine whether the vasodilation of epicardial coronary artery can be measuredduring cold pressor test (CPT) and adenosine infusion simultaneously with the flow velocity measurement usingTTE. We studied 41 healthy nonsmoking men with a linear high-frequency 8.0-MHz transducer. The CPT andadenosine infusion dilated the diameter of the distal left anterior descending coronary artery (LAD) from 1.4 �0.4 mm to 1.5 � 0.4 mm (14 � 13%, p < 0.01) and from 1.4 � 0.4 mm to 1.8 � 0.5 mm (31 � 19%, p < 0.01),respectively. The CPT increased flow velocity and calculated coronary blood flow rate (velocity time integral �cross-sectional area) from 0.23 � 0.05 m/s to 0.36 � 0.13 m/s (31 � 34%, p < 0.01) and from 8.1 � 4.2 mL/minto 11.4 � 6.0 mL/min (47 � 51%, p < 0.01). CFVR and calculated coronary blood flow rate reserve were 3.9 �1.0 and 6.0 � 1.9, respectively. In Bland-Altman analysis, velocity measurements underestimated the vasodilationresponse of the CPT and adenosine compared with the measurements where epicardial diameter dilation wastaken into account. Intra- and interobserver variability of diameter measurements was low (coefficient ofvariation [CV] 2.6 to 6.5%). Day-to-day, within-day and intersonographer variabilities were of similar magnitude(CV 4.6 to 8.2%), suggesting good reproducibility. This study demonstrates that TTE can be used to assesschanges in both epicardial coronary artery diameter and flow velocity simultaneously in the distal LAD artery.(E-mail: [email protected]) © 2007 World Federation for Ultrasound in Medicine & Biology.

Key Words: Diameter, Velocity, Cold pressor test, Coronary flow reserve.

INTRODUCTION

Recent development in the ultrasound technology hasmade it possible to assess coronary flow velocity andvelocity-based coronary flow reserve (CFVR) duringadenosine infusion with transthoracic echocardiography(TTE) (Hozumi et al. 1998a; Caiati et al. 1999a, 1999b;Saraste et al. 2001; Hildick-Smith et al. 2002; Kosken-vuo et al. 2003). CFVR measured with TTE correlatesclosely with that measured with an intracoronary flowwire (Hozumi et al. 1998a; Caiati et al. 1999b), MRI(Koskenvuo et al. 2003) and PET (Saraste et al. 2001).All of the CFVR studies performed with TTE are based

Address correspondence to: Tuomas Kiviniemi, Department ofClinical Physiology and Nuclear Medicine, Turku University Hospital,

362

on the change in blood flow velocity. However, coronaryblood flow rate depends on both coronary flow velocityand artery diameter and, thus, epicardial coronary arteryvasodilation during cold pressor test (CPT) or adenosineinfusion may cause underestimation of the vasodilationresponse of CFVR if only flow velocities are measured.Coronary blood flow rate can be assessed by multiplyingthe cross-sectional area and flow velocity (velocity–timeintegral [VTI]) of a coronary artery (Doucette et al. 1992;Oskarsson and Pesonen 2002). If the diameter could bemeasured accurately, the extent of the epicardial coro-nary artery vasodilation could be taken into account inthe CFVR measurements.

Vasodilation responses to the CPT and adenosineinfusion are widely used measures of coronary athero-sclerosis and cardiac microvascular function (Nitenberg

Epicardial coronary artery vasodilation ● T. O. KIVINIEMI et al. 363

causes vasodilation and increased blood flow in the epi-cardial coronary arteries via sympathetic activation (Ma-cho et al. 1981; Nabel et al. 1988). CFVR is defined as aratio of coronary blood flow during maximal pharmaco-logically-induced vasodilation to baseline blood flow(Baumgart et al. 1998). Adenosine infusion is used tomeasure CFVR because of its capability to induce themaximal dilation of small coronary arteries. Secondarily,adenosine increases flow velocity in the epicardial arter-ies, also leading to flow-mediated vasodilation. Impair-ment of coronary artery reactivity as shown by epicardialvasoconstriction in the CPT or decreased CFVR occursin various cardiovascular diseases such as hypertension(Antony et al. 1994; Nitenberg et al. 1995), diabetes(Pitkänen et al. 1998; Nitenberg et al. 2001, 2004),hypertrophic cardiomyopathy (Dimitrow et al. 2000),hypercholesterolemia (Pitkänen et al. 1996, 1999) andcoronary artery disease (Zeiher et al. 1991; Saraste et al.2000), and may be the result of impairment in vascularsmooth muscle or endothelial function, or both.

TTE has been used to measure coronary arterydiameters, with success rates varying from 63 to 98%(Weyman et al. 1976; Vered et al. 1986; Presti et al.1987; Ross et al. 1990; Faletra et al. 1995; Petrovic et al.1996; Hildick-Smith and Shapiro 1998; Gradus-Pizlo etal. 2001). A relatively good correlation (r � 0.83 to 0.93)has been shown between the measurements of the leftanterior descending coronary artery (LAD) diameter byTTE and quantitative coronary angiography (Hildick-Smith and Shapiro 1998; Kiviniemi et al. 2004), intra-coronary ultrasound (Hildick-Smith et al. 2002) and epi-cardial echocardiography (Gradus-Pizlo et al. 2003).Moreover, the variation of repeated off-line diametermeasurements by TTE has been very low (coefficient ofvariation [CV] 5.4 to 7.5%) (Kiviniemi et al. 2004).

We sought to determine whether the vasodilation ofthe distal LAD can be measured during the CPT andadenosine infusion simultaneously with flow velocitymeasurement using TTE.

MATERIALS AND METHODS

Forty-one healthy young (mean age 24 � 2.3 y)nonsmoking Finnish men with good echo window on thedistal LAD territory were included in the study. Thestudy was carried out in accordance with the Declarationof Helsinki (2000) of the World Medical Association andwas approved by the Ethics Committee of the SouthwestFinland Health Care District. All subjects gave theirwritten informed consent.

Echocardiography, coronary flow velocity and diametermeasurements

Subjects were instructed to avoid large meals for 4 h

studies. Variability assessment was performed after anovernight fast and a 12-h caffeine, tobacco and alcoholavoidance. TTE studies were performed with Sequoia C512 ultrasound apparatus (Acuson Inc., Mountain View,CA, USA) by four experienced sonographers with alinear 8.0-MHz transducer. B-mode and color-Dopplermapping were used to identify the distal LAD. Thecoronary artery appeared as a linear tubular structure,demonstrating mainly diastolic color-Doppler flow sig-nal. Diameters of the distal LAD were measured from themodified apical two-chamber, the modified apical shortaxis or the modified four-chamber view by aiming theultrasound beam superiorly in parallel with the epicardialsurface of the heart and scanning the interventricularsulcus when subjects were lying in the left lateral decu-bitus position. The part of the LAD that was apical to thepapillary muscle level was considered to be its distalpart. When performing echocardiography, the sonogra-pher marked the position of the transducer on the chestwall to visualize the same region-of-interest at baseline,during the CPT and adenosine infusion. All measure-ments were done at the end of expiration because thecoronary artery may move from the imaging plane dur-ing respiration. However, subjects were asked not to holdtheir breath during the measurements. The duration of asingle TTE study varied from 24 to 57 min.

B-mode images were stored in clips containing twoconsecutive cardiac cycles in the digital network (Ki-netDX WS3000, Acuson). In off-line analysis, clips con-taining 2-D images were first scrolled through to find thesame measurement position based on both intra- andextravascular structures. The same position was thenused at baseline, during the CPT and adenosine infusion.Diameters were measured from end-diastolic images us-ing ImageJ software (ImageJ 1.30N, National Institutesof Health, Bethesda, MD, USA). Usually, the coronaryartery was only partially visible on a single longitudinalimaging plane. Diameters were measured from the larg-est distance between the luminal edges from the black-white interface of the near and far walls and, whenpossible, the intima-to-intima distance was used for mea-surements. The diameter of each artery was calculated asan average of three measurements from three separateimages. The TTE images of the coronary artery diame-ters at baseline, during the CPT and adenosine infusionare shown in Fig. 1.

The coronary flow velocity was assessed with apulsed wave Doppler using a recently published imagingprotocol, at the same place as the diameter measurements(Saraste et al. 2000). The transducer was tilted laterallyto allow Doppler analysis of the distal part of the LADartery. The angle between the transducer and longitudi-nal axis of the coronary artery was smaller than 60° in

364 Ultrasound in Medicine and Biology Volume 33, Number 3, 2007

corrected. The length of the sample volume of thepulsed-wave Doppler was 2 mm. Mean diastolic flowvelocities (MDV) and velocity time integrals (VTI) were

measured. Coronary blood flow rate (mL min–1) wascalculated as

Flow rate � � � �D

2 �2

� VTI � HR � 0.5 (1)

where (D � coronary artery diameter [mm], HR � heartrate [1/min], VTI � velocity time integral —, VTI valuesare angle corrected) (Doucette et al. 1992; Oskarsson andPesonen 2002).

Cold pressor testBaseline values were detected as an average of three

separate diameter and three flow velocity measurements.After the baseline measurements, the subjects’ righthands were placed into ice-cold water for 120 s. Duringthe hand immersion, both diameter and flow velocitywere measured in turn continuously. After 120 s of handimmersion, the measurements were continued for a fol-low-up period of 5 min to assess the recovery of the flowvelocity and diameter to baseline.

Epicardial coronary artery diameter during the CPTis expressed as an average of the measurements between30 and 180 s, and flow velocity as an average of themeasurements between 10 and 150 s after placing thehand into the cold water. The minimum of three mea-surements was used.

Adenosine infusionIntravenous (IV) adenosine (adenosine item 5 mg/

mL) was infused at a rate of 0.14 mg/kg/min for 5 min,and diameters and flow velocities were measured in turncontinuously during the infusion. Blood pressure wasmeasured at baseline and during the stimulation, andelectrocardiography (ECG) was recorded continuously.Adenosine infusion was started 22 � 7 min after theCPT.

The results of the measurements are expressed as anaverage of three separate maximum diameter and flowvelocity measurements. Coronary blood flow rate reservewas measured as a ratio of coronary blood flow rateduring maximally induced vasodilation to baseline, and

Fig. 1. The transthoracic echocardiography images of the lu-men of the LAD artery diameter (a) at baseline, (b) during theCPT and (c) adenosine-induced maximal vasodilation. Thecoronary artery appears as two linear echogenic lines. Thelumen is between these lines (*). Measurements were donefrom one luminal edge of vessel wall to the other (arrow heads).Coronary artery diameter vasodilation during the CPT was 14%and during adenosine infusion was 31%. The imaging proper-ties are on the right side of the image. (d) MDV in the distalLAD artery at baseline (on left), during the CPT (middle) and

adenosine infusion (right).

Epicardial coronary artery vasodilation ● T. O. KIVINIEMI et al. 365

CFVR was measured with the corresponding MDV val-ues. Study protocol for the CPT and adenosine infusionmeasurements is shown in Fig. 2.

Data analysisData are presented as mean � standard deviation

unless stated otherwise. Paired t-tests were used whenappropriate. Two independent observers analyzed thecoronary flow velocity and diameter measurements,blinded to each other, to assess the intra- and interob-server variability of the off-line measurements. Within-day variability was analyzed by performing the CPT andadenosine infusion again 2 h after the first test by thesame sonographer. Day-to-day variability was analyzedby performing the CPT and adenosine infusion on sep-arate days by the same sonographer, and intersonogra-pher variability was analyzed by comparing the testsperformed by two sonographers on separate days. Allvariability measurements were carried out blinded. Vari-ability was assessed in 10 subjects.

RESULTS

The measurements were performed successfully inall 41 subjects during adenosine infusion. The CPT datacould not be collected in three of them because of poorimage quality. The CPT and adenosine infusion weregenerally tolerated well. The time courses of the percentchange of coronary artery diameter and MDV flow ve-locity during the CPT and adenosine infusion are pre-sented in Fig. 3. The complete data of hemodynamicresponse of the subjects to the CPT and adenosine infu-sion is presented in Table 1.

CPT measurementsThe maximal vasodilation occurred between 30 and

120 s after the onset of hand immersion in the CPT (seeFig. 3). The coronary artery dilated from 1.4 � 0.4 mmto 1.5 � 0.4 mm (14 � 13%, p � 0.01). MDV increasedfrom 0.23 � 0.05 m/s to 0.36 � 0.13 m/s (31 � 34%, p� 0.01) and VTI from 0.12 � 0.04 m to 0.14 � 0.05 m

-diameter -velocity-ECG -BP

-diameter -velocity-ECG -BP

AT REST

-diameter -velocity-ECG -BP

CFRCPT

adenosine infusion 0.14 mg/kg/min 5 minutes

hand immersion in cold water 2 minutes

Fig. 2. The study protocol for CPT and coronary flow reservemeasurements. ECG � electrocardiography; BP � blood

pressure.

(19 � 33 %, p � 0.01), respectively. As a result, coro-

nary blood flow rate increased from 8.1 � 4.2 mL/min to11.4 � 6.0 mL/min (47 � 51%, p � 0.01). Flow velocitydecreased markedly 30 s after the end of the hand im-mersion. Artery diameter remained dilated for the wholefollow-up period after the CPT.

Adenosine infusionThe maximal vasodilation occurred 60 s after the

onset of adenosine infusion (see Fig. 3). During adeno-sine infusion, the coronary artery dilated from 1.4 � 0.4mm to 1.8 � 0.5 mm (31 � 19%, p � 0.01); MDVincreased from 0.23 � 0.06 m/s to 0.96 � 0.28 m/s (288� 100%, p � 0.01) and VTI from 0.12 � 0.04 m to 0.34� 0.12 m (202 � 80%, p � 0.01), respectively. Coro-nary blood flow rate increased from 7.8 � 3.6 mL/min to45.0 � 18.7 mL/min (498 � 190%, p � 0.01). Coronary

a)

b)

0 1 2 3 4 5 6 7-20

-10

0

10

20

30

40

50

60

70

80

90

100

Time (min)

Per

cen

t ch

ang

e (%

) Diameter

Cold pressor test

MDV

0 1 2 3 4 50

10

20

30

40

50

200

300

Time (min)

Per

cen

t ch

ang

e (%

)

MDV

Adenosine infusion

Diameter

Fig. 3. The time courses (mean � SEM) of the coronary arterydiameter and MDV during (a) CPT and (b) adenosine

infusion.

366 Ultrasound in Medicine and Biology Volume 33, Number 3, 2007

2 3 4 5 6 7 8 9 10-2

0

2

4

6

Mean - 1.96 SD

Mean

Mean + 1.96 SD

Mean (CFVR and Flow rate reserve)

2 3 4 5 6

2

4

6

8

10n = 41r = 0.64p < 0.001

CFVR

0.5 1.0 1.5 2.0

1

2

3

4 n = 36

r = 0.54

p < 0.001

CPT Velocity ratio

1 2 3-1

0

1

2

Mean - 1.96 SD

Mean

Mean + 1.96 SD

Mean of CPT (Flow rate and Velocity)

CP

T f

low

rat

e ra

tio

Dif

fere

nce

of

CP

T (

Flo

w r

ate

- V

elo

city

)

Dif

fere

nce

(F

low

rat

e re

serv

e-C

FV

R)

Flo

w r

ate

rese

rve

a) b)Fig. 4. (a) Correlation graph and Bland-Altman analysis of CPT hyperemia/baseline measurements performed withvelocity and blood flow rate values, and (b) corresponding graphs of coronary blood flow rate reserve and CFVR

Table 1. Hemodynamic data of study population (adenosine infusion, n � 41; CPT, n � 38)

CPT Adenosine infusion

Baseline Hyperemia Ratio Baseline Hyperemia Ratio

Diameter (mm) 1.4 � 0.4 1.5 � 0.4 1.14 � 0.13 1.4 � 0.4 1.8 � 0.5 1.31 � 0.19MDV (m s�1) 0.23 � 0.05 0.36 � 0.13 1.31 � 0.34 0.23 � 0.06 0.96 � 0.28 3.88 � 1.0VTI (m) 0.12 � 0.04 0.14 � 0.05 1.19 � 0.33 0.12 � 0.04 0.34 � 0.12 3.02 � 0.80Blood flow rate (mL min�1) 8.1 � 4.2 11.4 � 6.0 1.47 � 0.51 7.8 � 3.6 45.0 � 18.7 5.98 � 1.9Heart rate (beats min�1) 61 � 8 69 � 10 61 � 8 82 � 15Rate pressure product (mm Hg s�1) 7210 � 1450 8060 � 1790 6440 � 2200 8530 � 1950

MDV � mean diastolic velocity; VTI � velocity time integral; blood flow rate � � · (diameter/2)2 · VTI · HR � 0.5; ratio � hyperemia/baseline.

measurements. Differences are presented as absolute values.

Epicardial coronary artery vasodilation ● T. O. KIVINIEMI et al. 367

flow reserves using MDV, VTI and blood flow rate were3.9 � 0.8, 3.0 � 0.8 and 6.0 � 1.9, respectively.

Comparisons and repeatabilityVelocity-based hyperemia-to-baseline ratios under-

estimated the vasodilation effect of the CPT and adeno-sine infusion in Bland-Altman analysis (Fig. 4) (Blandand Altman 1986). There was a significant correlationbetween hyperemia-to-baseline ratios obtained with flowvelocity and blood flow rate in the CPT (r � 0.54, p �0.001); however, the Bland-Altman analysis showed aclear trend upwards in those subjects, with greater meanin the measurements carried out with the two ways.CFVR and blood flow rate reserve correlated signifi-cantly (r � 0.64, p � 0.001), but Bland-Altman analysisshowed even greater upward trend in those with greatermean.

Intraobserver, interobserver, within-day, day-to-dayand intersonographer variabilities for the change of di-ameter, flow velocity and blood flow rate measurementsare presented in Table 2.

DISCUSSION

This study demonstrates that TTE can be used tomonitor changes in both the epicardial coronary arterydiameter and flow velocity simultaneously during the

Table 2. Intraobserver, interobserver, within-day, day-adenosine infusion mea

Diameter ratio

CV (mean � SD)(%)

Mean difference(range)

CV (m

Intraobserver 2.9 � 2.5 0.0 (�0.1 to 0.1) 3.0Interobserver 3.7 � 4.0 �0.0 (�0.1 to 0.2) 3.7Within-day 5.6 � 4.3 0.0 (�0.2 to 0.2) 12.6Day-to-day 8.0 � 5.4 0.0 (�0.3 to 0.3) 13.4Intersonographer 6.6 � 6.4 0.0 (�0.3 to 0.3) 12.4

Diameter ratio

CV (mean � SD)(%)

Mean difference(range)

CV (m

Intraobserver 5.1 � 3.8 0.0 (�0.2 to 0.3) 2.6Interobserver 6.5 � 3.2 �0.1 (�0.3 to 0.1) 8.6Within-day 8.2 � 6.0 0.0 (�0.4 to 0.4) 11.5Day-to-day 6.8 � 6.1 0.0 (�0.4 to 0.3) 11.4Intersonographer 4.6 � 3.3 0.0 (�0.2 to 0.2) 16.8

CV � Coefficient of variation: SD/mean � 100; mean difference �diameter hyperemia/diameter rest.

CPT and adenosine infusion. Flow velocity measure-

ments alone appeared to underestimate the vasodilationresponse in the CPT and CFVR when compared withcalculated blood flow rate measurements. This findingmay have important applications in the assessment ofcoronary atherosclerosis and microvascular reactivitynoninvasively by TTE.

This study showed that the epicardial coronary ar-tery diameter dilated by 14% and 31% during the CPTand adenosine infusion, respectively. Moreover, the ex-tent of the CPT-induced vasodilation response and cor-onary flow reserves were higher with blood flow ratemeasurements when compared with flow velocity mea-surements. Blood flow rate and velocity-based hyper-emia/baseline ratios during the CPT and adenosine infu-sion seemed to correlate moderately, but Bland-Altmangraphs showed a clear trend upward in subjects withgreater mean value of the two measurements. The find-ings are consistent with the hypothesis that flow velocityalone may underestimate CFVR. For example, subjectshaving CFVR around 4 may have blood flow rate–basedreserve from 4 to 9. This is most likely explained by thevasodilation of the epicardial coronary artery. Namely,the smooth muscle relaxation of the small coronary ar-teries leads to an increase of flow velocity in the epicar-dial coronary arteries as well, and the increase inducesflow-mediated vasodilation of the epicardial coronary

and intersonographer variabilities of (a) CPT and (b)ents in 10 subjects (a)

velocity ratio Blood flow rate ratio

SD) Mean difference(range)

CV (mean � SD)(%)

Mean difference(range)

0.0 (�0.1 to 0.1) 10.4 � 8.4 �0.1 (�0.6 to 0.4)0.0 (�0.2 to 0.2) 19.5 � 5.5 �0.3 (�1.0 to 0.5)

�0.1 (�0.6 to 0.5) 16.7 � 17.0 �0.2 (�1.1 to 0.7)4 0.0 (�0.6 to 0.6) 28.1 � 21.6 �0.1 (�1.6 to 1.7)2 �0.2 (�0.6 to 2.0) 24.1 � 18.3 �0.3 (�1.6 to 1.0)

CFVR Blood flow rate reserve

SD) Mean difference(range)

CV (mean � SD)(%)

Mean difference(range)

�0.1 (�0.4 to 0.2) 7.3 � 6.1 �0.5 (�1.7 to 0.7)0.3 (�0.8 to 1.4) 15.6 � 8.6 �1.0 (�3.4 to 1.5)

�0.5 (�2.1 to 1.1) 24.0 � 20.2 �0.5 (�6.7 to 5.7)�0.1 (�1.7 to 1.6) 25.4 � 13.4 �0.2 (�5.8 to 5.4)

6 0.5 (�1.6 to 2.6) 22.8 � 18.0 �0.4 (�6.2 to 4.1)

f (measurement 1 – measurement 2); range �2 SD; Diameter ratio �

to-daysurem

(a)

Flow

ean �(%)

� 2.0� 3.7� 9.4� 10.� 10.

(b)

ean �(%)

� 4.0� 9.8� 8.7� 8.9� 13.

mean o

artery. Because adenosine infusion is known to induce

368 Ultrasound in Medicine and Biology Volume 33, Number 3, 2007

the maximal dilation of the small coronary arteries,CFVR alone is likely to be a measure of the microcircu-latory function. Blood flow rate reserve, conversely, is acombined measure of flow-mediated epicardial arteryvasodilation and microvascular function.

Intra- and interobserver variabilities of the changeof diameter, velocity and calculated blood flow rate werelow. Low intra- and interobserver variabilities have alsobeen reported in previous studies for off-line diameter(Kiviniemi et al. 2004) and flow velocity measurements(Saraste et al. 2001). In addition, moderate within-day,day-to-day and intersonographer variabilities suggestgood reproducibility of the measurements using TTE.This information is relevant in studies where repeatedcoronary flow reserve measurements are performed.

The time courses demonstrate that velocity and di-ameter should be measured 30 to 120 s after the onset ofhand immersion in the CPT, to see the maximal vasodi-lation. Moreover, the artery diameter did not return tobaseline during the follow-up period of 5 min after theend of the CPT. Therefore, the baseline values of theCPT should be used as baseline values for adenosineinfusion, when assessing coronary flow reserve shortlyafter the CPT measurements. The time course of adeno-sine infusion shows that flow velocity response is notsteady. It seems, however, that the first maximal flowvelocity response assessed after 1 min from the onset ofthe infusion yielded the highest values. Consequently,5-min adenosine infusion is not needed if a good re-sponse is detected already after 2-min infusion.

Our study is the first to demonstrate that the simul-taneous assessment of the changes in coronary arterydiameter and velocity by TTE in the LAD artery isfeasible in humans. We were able to measure simulta-neously both the diameter and flow velocity in the LADby using TTE and a linear 8-MHz transducer. The high-frequency transducer was chosen for this study to obtainoptimal axial and lateral resolution. This transducer hasan axial resolution of 0.25 mm at a depth of 50 mm(Gradus-Pizlo et al. 2001). The fact that the LAD arterycan be visualized with TTE has been recognized forseveral years (Weyman et al. 1976; Vered et al. 1986;Presti et al. 1987; Ross et al. 1990; Faletra et al. 1995;Petrovic et al. 1996; Hildick-Smith and Shapiro 1998;Gradus-Pizlo et al. 2001; Kiviniemi et al. 2004). Inprevious studies, a relatively good correlation was shownbetween measurements of the LAD artery diameter byTTE and quantitative coronary angiography (Hildick-Smith and Shapiro 1998; Kiviniemi et al. 2004), TTEand intracoronary ultrasound (Hildick-Smith et al. 2002)and TTE and epicardial echocardiography (Gradus-Pizloet al. 2003). In a recent study, TTE revealed differencesin the coronary artery vasodilation response during the

al. 2001). The extent of the vasodilation during the CPTwas similar to previous studies using TTE (13.2%)(Deng et al. 2001) or quantitative coronary angiography(8.9 to 17.2%) (Nabel et al. 1988; Zeiher et al. 1989;Antony et al. 1994; Nitenberg et al. 1995; Benvenuti etal. 1995; Schindler et al. 2003; Nitenberg et al. 2004).Furthermore, TTE-derived coronary flow velocity mea-surements have been validated with an intracoronaryDoppler ultrasound (Hozumi et al. 1998a), magneticresonance imaging (MRI) (Koskenvuo et al. 2003) andpositron emission tomography (PET) (Saraste et al.2001). TTE has been used to measure coronary flowduring adenosine infusion (Hozumi et al. 1998a, 1998b;Caiati et al. 1999a, 1999b; Saraste et al. 2001) and alsoduring the CPT (Dimitrow et al. 2000). Moreover, TEEhas been used to measure flow velocity in the CPT(Chandraratna et al. 1999). Coronary flow velocity re-sponse to the CPT in earlier studies (46.7 to 51%)(Antony et al. 1994; Nitenberg et al. 1995) is also in linewith our results.

The combined diameter and flow velocity measure-ments provided us with data to calculate the actual cor-onary blood flow rate, i.e., flow volume per unit of time.Previously, LAD artery blood flow rate at baseline hasbeen measured with TTE in children in whom the bloodflow rate is increased in dilated cardiomyopathy (Oskars-son and Pesonen 2002) and ventricular septal defect(Yasuoka et al. 2002) compared with healthy children.Moreover, coronary blood flow rate has been measuredwith TTE during adenosine infusion in healthy volun-teers and endurance athletes, but the coronary arterydiameter was dilated and fixed with nitroglycerine beforeadenosine infusion to prevent further flow-induced vaso-dilation (Hildick-Smith and Shapiro 1999; Hildick-Smithet al. 2000). The accuracy of the blood flow rate mea-surements by TTE remains to be studied using a directcomparison with other methods such as intracoronaryflow wire. Previously, TEE appeared to overestimatecoronary blood flow rate at rest when compared withepicardial Doppler echocardiography and electromag-netic flow assessment (Chaudhry et al. 2001) in dogs. Inour study, the vasodilation response in the CPT wascomparable with previous studies with healthy subjectsusing intracoronary-derived flow velocity and quantita-tive coronary angiography-derived diameter measure-ments (Nitenberg et al. 1995, 2001). Moreover, our re-sults of blood flow rate are rational compared with thoseobtained with MRI in the proximal LAD artery (46.6 mLmin–1) (Clarke et al. 1995) and an intracoronary ultra-sound study of the distal LAD (18.4 mL min–1 at rest)(Hildick-Smith et al. 2002). Compared with the MRIstudy, the smaller blood flow rate values in our study arelikely the result of more distal measuring position. In a

Epicardial coronary artery vasodilation ● T. O. KIVINIEMI et al. 369

state (13.4 mL min–1) (Hildick-Smith et al. 2000) com-pared with our results (8.1 mL min–1). This is presum-ably explained with the nitroglycerine administrationthat is likely to cause increased flow at baseline becauseof artery dilation.

There are several potential applications for our re-sults. In the future, noninvasive blood flow rate measure-ments during the CPT and adenosine infusion will beuseful in the detection of dysfunction of the microcircu-lation and epicardial vasodilation. Although intravascu-lar ultrasound and PET are accurate methods to assessblood flow in the heart, they have limitations that pre-clude their applicability for repeated measurements ofcoronary flow reserve and the CPT. Namely, IVUS isinvasive and PET causes radiation exposure for subjects.TTE, however, is especially suitable for comparing dif-ferent cardiovascular drugs and their dose-dependent ef-fects. For instance, TTE has been used in the evaluationof immediate effects of IV drugs on coronary flow ve-locity (Snapir et al. 2003).

Study limitationsSome limitations should be pointed out. First, we

studied only healthy subjects. Second, we did not vali-date this new method against other methods measuringcoronary blood flow rate. Nevertheless, a relatively goodcorrelation has been reported in studies comparing eitherTTE-derived diameter change (Deng et al. 2001) or flowvelocity change with intracoronary ultrasound (Hozumiet al. 1998a; Caiati et al. 1999b) and flow velocity withPET (Saraste et al. 2001) and MRI (Koskenvuo et al.2003), as discussed before. Finally, we measured dia-stolic flow velocities only. Flow in the LAD is mainlydiastolic because the systolic-to-diastolic ratio is 0.22 formean flow (Marcus et al. 1999). Although diastolic timeshortens when the heart rate increases and this is likely tochange systolic-to-diastolic ratio, the heart rate increasewas moderate during both the provocations (8 beatsmin–1 during the CPT and 21 beats min–1 during aden-osine infusion). Therefore, it has only a minor effect onour blood flow rate measurements.

CONCLUSIONS

We conclude that TTE can be used to assess thevasodilation of the epicardial coronary artery simulta-neously with flow velocity measurement during the CPTand adenosine infusion.

Acknowledgements—This study was supported by The Turku Univer-sity Hospital Research Foundation, the Turku University Foundation

REFERENCES

Antony I, Aptecar E, Lerebours G, Nitenberg A. Coronary arteryconstriction caused by the cold pressor test in human hypertension.Hypertension 1994;24:212–219.

Baumgart D, Haude M, Liu F, Ge J, Goerge G, Erbel R. Currentconcepts of coronary flow reserve for clinical decision makingduring cardiac catheterization. Am Heart J 1998;136:136–149.

Benvenuti C, Aptecar E, Mazzucotelli JP, Jouannot P, Loisance D,Nitenberg A. Coronary-artery response to cold-pressor test is im-paired early after operation in heart-transplant recipients. J Am CollCardiol 1995;26:446–451.

Bland JM, Altman DG. Statistical methods for assessing agreementbetween two methods of clinical measurement. Lancet 1986;1:307–310.

Caiati C, Montaldo C, Zedda N, Bina A, Iliceto S. New noninvasivemethod for coronary flow reserve assessment: Contrast-enhancedtransthoracic second harmonic echo Doppler. Circulation 1999a;99:771–778.

Caiati C, Montaldo C, Zedda N, et al. Validation of a new noninvasivemethod (contrast-enhanced transthoracic second harmonic echoDoppler) for the evaluation of coronary flow reserve: Comparisonwith intracoronary Doppler flow wire. J Am Coll Cardiol 1999b;34:1193–1200.

Chandraratna PA, Nimalasuriya AR, Vlachonassios KD, et al. Useful-ness of the response of flow velocity in the left anterior descendingcoronary artery to the cold pressor test for evaluating endothelium-dependent vascular relaxation in the coronary microvasculature bytransesophageal echocardiography in subjects with angiographi-cally normal coronary arteries. Am J Cardiol 1999;84:1362–1365,A8.

Chaudhry FA, Ren JF, Ramani K, et al. Validation of transesophagealechocardiography to determine physiologic coronary flow. Echo-cardiograph Am J Cardiovasc Ultrasound Allied Technique 2001;18:553–557.

Clarke GD, Eckels R, Chaney C, et al. Measurement of absoluteepicardial coronary artery flow and flow reserve with breath-holdcine phase-contrast magnetic resonance imaging. Circulation 1995;91:2627–2634.

Deng YB, Wang XF, Li CL. A new noninvasive method for evaluationof coronary endothelial function in hypertensive patients based onchange in diameter of the left main coronary artery induced by coldpressor test using echocardiography. Clin Cardiol 2001;24:291–296.

Dimitrow PP, Krzanowski M, Nizankowski R, Szczeklik A, Dubiel JS.Comparison of the effect of verapamil and propranolol on responseof coronary vasomotion to cold pressor test in symptomatic patientswith hypertrophic cardiomyopathy. Cardiovasc Drugs Ther 2000;14:643–650.

Doucette JW, Corl PD, Payne HM, et al. Validation of a Doppler guidewire for intravascular measurement of coronary artery flow veloc-ity. Circulation 1992;85:1899–1911.

Faletra F, Cipriani M, Dechiara F, et al. Imaging the left anteriordescending coronary-artery by high-frequency transthoracic echo-cardiography in heart-transplant patients. Am J Cardiol 1995;75:855–858.

Gradus-Pizlo I, Bigelow B, Mahomed Y, Sawada SG, Rieger K,Feigenbaum H. Left anterior descending coronary artery wall thick-ness measured by high-frequency transthoracic and epicardial echo-cardiography includes adventitia. Am J Cardiol 2003;91:27–32.

Gradus-Pizlo I, Sawada SG, Wright D, Segar DS, Feigenbaum H.Detection of subclinical coronary atherosclerosis using two-dimen-sional, high-resolution transthoracic echocardiography. J Am CollCardiol 2001;37:1422–1429.

Hildick-Smith DJ, Johnson PJ, Wisbey CR, Winter EM, Shapiro LM.Coronary flow reserve is supranormal in endurance athletes: Anadenosine transthoracic echocardiographic study. Heart 2000;84:383–389.

Hildick-Smith DJ, Maryan R, Shapiro LM. Assessment of coronary

370 Ultrasound in Medicine and Biology Volume 33, Number 3, 2007

tion with intracoronary Doppler. J Am Soc Echocardiogr2002;15:984–990.

Hildick-Smith DJ, Shapiro LM. Transthoracic echocardiographic mea-surement of coronary artery diameter: Validation against quantita-tive coronary angiography. J Am Soc Echocardiogr 1998;11:893–897.

Hildick-Smith DJ, Shapiro LM. Potential use of transthoracic echocar-diography in the assessment of coronary flow reserve. J Am SocEchocardiogr 1999;12:590–595.

Hozumi T, Yoshida K, Akasaka T, et al. Noninvasive assessment ofcoronary flow velocity and coronary flow velocity reserve in the leftanterior descending coronary artery by Doppler echocardiography:Comparison with invasive technique. J Am Coll Cardiol 1998a;32:1251–1259.

Hozumi T, Yoshida K, Ogata Y, et al. Noninvasive assessment ofsignificant left anterior descending coronary artery stenosis bycoronary flow velocity reserve with transthoracic color Dopplerechocardiography. Circulation 1998b;97:1557–1562.

Kiviniemi TO, Saraste M, Koskenvuo JW, et al. Coronary arterydiameter can be assessed reliably with transthoracic echocardiog-raphy. Am J Physiol Heart Circ Physiol 2004;286:H1515–H1520.

Koskenvuo JW, Saraste M, Niemi P, et al. Correlation of transthoracicDoppler echocardiography and magnetic resonance imaging inmeasuring left anterior descending artery flow velocity and time-course of dipyridamole-induced coronary flow increase. ScandJ Clin Lab Invest 2003;63:65–72.

Macho P, Hintze TH, Vatner SF. Regulation of large coronary-arteriesby increases in myocardial metabolic demands in conscious dogs.Circ Res 1981;49:594–599.

Marcus JT, Smeenk HG, Kuijer JPA. Flow profiles in the left anteriordescending and the right coronary artery assessed by MR velocityquantification: Effects of through-plane and in-plane motion of theheart. J Comput Assist Tomogr 1999;23:567–576.

Nabel EG, Ganz P, Gordon JB, Alexander RW, Selwyn AP. Dilation ofnormal and constriction of atherosclerotic coronary arteries causedby the cold pressor test. Circulation 1988;77:43–52.

Nitenberg A, Antony I, Aptecar E, Arnoult F, Lerebours G. Impairmentof flow-dependent coronary dilation in hypertensive patients. Dem-onstration by cold pressor test induced flow velocity increase. Am JHypertens 1995;8:13S–18S.

Nitenberg A, Ledoux S, Valensi P, Sachs R, Attali JR, Antony I.Impairment of coronary microvascular dilation in response to coldpressor–induced sympathetic stimulation in type 2 diabetic patientswith abnormal stress thallium imaging. Diabetes 2001;50:1180–1185.

Nitenberg A, Valensi P, Sachs R, Cosson E, Attali JR, Antony I.Prognostic value of epicardial coronary artery constriction to thecold pressor test in type 2 diabetic patients with angiographicallynormal coronary arteries and no other major coronary risk factors.Diabetes Care 2004;27:208–215.

Oskarsson G, Pesonen E. Flow dynamics in the left anterior descendingcoronary artery in infants with idiopathic dilated cardiomyopathy.

Petrovic O, Elsner GB, Wilensky RL, Swanson ST, Feigenbaum H.Transthoracic echocardiographic detection of coronary atheroscle-rosis. Am J Cardiol 1996;77:569–574.

Pitkänen OP, Nuutila P, Raitakari OT, et al. Coronary flow reserve inyoung men with familial combined hyperlipidemia. Circulation1999;99:1678–1684.

Pitkänen OP, Nuutila P, Raitakari OT, et al. Coronary flow reserve isreduced in young men with IDDM. Diabetes 1998;47:248–254.

Pitkänen OP, Raitakari OT, Niinikoski H, et al. Coronary flow reserveis impaired in young men with familial hypercholesterolemia. J AmColl Cardiol 1996;28:1705–1711.

Presti CF, Feigenbaum H, Armstrong WF, Ryan T, Dillon JC. Digitaltwo-dimensional echocardiographic imaging of the proximal leftanterior descending coronary-artery. Am J Cardiol 1987;60:1254–1259.

Ross JJ, Mintz GS, Chandrasekaran K. Transthoracic 2-dimensionalhigh-frequency (7.5 MHz) ultrasonic visualization of the distal leftanterior descending coronary-artery. J Am Coll Cardiol 1990;15:373–377.

Saraste M, Koskenvuo JW, Knuuti J, et al. Coronary flow reserve:Measurement with transthoracic Doppler echocardiography is re-producible and comparable with positron emission tomography.Clin Physiol 2001;21:114–122.

Saraste M, Koskenvuo JW, Mikkola J, Pelttari L, Toikka JO, HartialaJJ. Technical achievement: Transthoracic Doppler echocardiogra-phy can be used to detect LAD restenosis after coronary angio-plasty. Clin Physiol 2000;20:428–433.

Schindler TH, Hornig B, Buser PT, et al. Prognostic value of abnormalvasoreactivity of epicardial coronary arteries to sympathetic stim-ulation in patients with normal coronary angiograms. ArteriosclerThromb Vasc Biol 2003;23:495–501.

Snapir A, Koskenvuo J, Toikka J, et al. Effects of common polymor-phisms in the alpha1A-, alpha2B-, beta1- and beta2-adrenorecep-tors on haemodynamic responses to adrenaline. Clin Sci (Lond)2003;104:509–520.

Vered Z, Katz M, Rath S, et al. Two-dimensional echocardiographicanalysis of proximal left main coronary-artery in humans. AmHeart J 1986;112:972–976.

Weyman AE, Feigenbaum H, Dillon JC, Johnston KW, Eggleton RC.Noninvasive visualization of left main coronary-artery by cross-sectional echocardiography. Circulation 1976;54:169–174.

Yasuoka K, Harada K, Tamura M, Toyono M, Takada G. Blood flowin the left anterior descending coronary artery in children withventricular septal defect. J Am Soc Echocardiogr 2002;15:807–813.

Zeiher AM, Drexler H, Wollschlaeger H, Saurbier B, Just H. Coronaryvasomotion in response to sympathetic-stimulation in humans—Importance of the functional integrity of the endothelium. J AmColl Cardiol 1989;14:1181–1190.

Zeiher AM, Drexler H, Wollschlager H, Just H. Modulation of coro-nary vasomotor tone in humans—Progressive endothelial dysfunc-tion with different early stages of coronary atherosclerosis. Circu-