Ultrasound in Med. & Biol., Vol. 36, No. 10, pp. 1756–1761, 2010Copyright � 2010 World Federation for Ultrasound in Medicine & Biology

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

asmedbio.2010.05.013

doi:10.1016/j.ultr

d Clinical Note

EVALUATION OF CORONARY FLOW VELOCITY RESERVE IN HOMOZYGOUSFAMILIAL HYPERCHOLESTEROLEMIA BY TRANSTHORACIC DOPPLERECHOCARDIOGRAPHY AND DUAL-SOURCE COMPUTED TOMOGRAPHY

YA YANG,* XIAOSHAN ZHANG,* RONGJUAN LI,* HONGYAN REN,* ZHENG WANG,* ZHIAN LI,* JIE LIN,y

LUYA WANG,y WEI YU,z and ZHAOQI ZHANGz

*Department of Ultrasound, Beijing Anzhen Hospital, Capital Medical University, Beijing, P. R. China; yBeijing Instituteof Heart Lung and Blood Vessel Diseases, Beijing, P. R.China; and zDepartment of Radiology, Beijing Anzhen Hospital,

Capital Medical University, Beijing, P. R. China

(Received 15 October 2009; revised 8 May 2010; in final form 14 May 2010)

Asound100029

Abstract—Homozygous familial hypercholesterolemia (HoFH) is a rare disorder characterized by the early onsetof atherosclerosis and usually occurs at the ostia of coronary arteries. In this study, we used transthoracic Dopplerechocardiography (TTDE) to evaluate the dynamic changes of coronary flow in HoFH patients and to detect aorticand coronary atherosclerosis by dual-source computed tomography (DSCT). We studied 20 HoFH patients(12 females, 8 males, mean age 13.1 ± 5.3 years, with a mean low density lipoprotein (LDL) cholesterol of 583 ±113 mg/dL) and 15 control patients (8 females, 7 males, mean age 15.2 ± 6.9 years, with a mean LDL cholesterol128 ± 71 mg/dL) using TTDE and DSCT. None of the patients showed evidence of ischemia with standard exercisetesting. Though the baseline coronary flow was similar between HoFH patients and normal controls, the hyperemicflow velocities and, thus, the coronary flow velocity reserve (CFVR) were significantly lower in those with HoFH. AllHoFH patients had aortic plaques, nine of them with the coronary artery ostia simultaneously, who had significantlyhigher LDL-cholesterol and lower CFVR than those without ostia plaques. Our data demonstrated that TTDEtogether with DSCT could be a useful noninvasive method for detection of coronary flow dynamics and atheroscle-rosis specifically in HoFH subjects with coronary ostia. (E-mail: lizhian_anzhen@yahoo.com.cn) � 2010 WorldFederation for Ultrasound in Medicine & Biology.

Key Words: Homozygous familial hypercholesterolemia, Coronary blood flow velocity reserve, TransthoracicDoppler echocardiography, Dual-source computed tomography.

INTRODUCTION

Familial hypercholesterolemia (FH) is a monogenic

disorder due to a mutation in the low density lipoprotein

receptor (LDL-R) gene (Kraft et al. 2000; Marais 2004).

It has been estimated to have a prevalence of 1/500

for the heterozygous form (HeHF) and, therefore,

a prevalence of about 1/1,000,000 for the homozygous

form (HoHF) (Marais 2004; Santos et al. 2008; Awan

et al. 2008). Individuals with this mutation have high

levels of low density lipoprotein (LDL) cholesterol in

plasma starting in childhood, early atherosclerosis and

a high incidence of coronary artery disease (Wiegman

et al. 2003; Civeira et al. 2005). Since atherosclerosis often

ddress correspondence to: Zhian Li, M.D., Department of Ultra-Beijing, Anzhen Hospital, Capital Medical University, Beijing,

P. R. China. E-mail: lizhian_anzhen@yahoo.com.cn

1756

happens at the ostia of coronary arteries, if untreated, the

prognosis is poor due to sudden death, which is oftscribed

to acute myocardial infarction before age 30.

Hypercholesterolemia is a major risk factor for coro-

nary artery disease. Deterioration of left ventricular (LV)

function has mainly been attributed to coronary insuffi-

ciency secondary to formation of atherosclerotic lesions

in epicardial coronary arteries that reduce blood flow at

rest or in response to increased demands. Hypercholester-

olemia also reduces coronary flow reserve and induces

microvascular dysfunction that may contribute to my-

ocardial ischemia and LV dysfunction. Patients with

syndrome X have myocardial ischemia without stegnotic

coronary artery lesions based on microvascular endothe-

lial dysfunction.

The objective of this study was to measure coronary

flow velocity reserve (CFVR) of HoFH patients by TTDE

and combine dual-source computed tomography (DSCT)

to evaluate the changes of coronary flow dynamics.

Evaluation of coronary flow velocity reserve d Y. YANG et al. 1757

MATERIALS AND METHOD

Study subjectsTwenty patients with HoFH but no history of ischemic

heart disease (8 men and 12 women; mean age 13.1 6 5.36

years) and 15 control subjects (7 men and 8 women; age

15.2 6 6.91 years) were studied. HoFH was diagnosed ac-

cording to the following criteria: (1) plasma cholesterol

concentration above 600 mg/dL or LDL cholesterol concen-

tration above 500 mg/dL; (2) occurrence of tendon xan-

thoma before age 20; and (3) having both parents who

were heterozygous familial hypercholesterolemia (Santos

et al. 2008; Faccenda et al. 1990). All patients satisfied

the aforementioned three criteria. Fifteen asymptomatic

subjects with normo-lipidemia, normo-glycemia, also the

siblings of HoFH patients, were selected as the control

group. Exclusion criteria were LV wall-motion abnor-

mality, previous myocardial infarction, LV hypertrophy,

previous cardiac surgery, atrial fibrillation, significant

valvular heart disease and unstable angina. All the study

subjects were prohibited to consume tea, coffee and other

caffeinated beverages during adenosine stress echocardi-

ography 12 hours prior to the study. The study was

approved by the Institutional Review Board of our institu-

tion and all study subjects gave informed consent prior to

the study. If the patients were children, appropriate parent or

guardian signatures were obtained on the informed

consents.

Dual-source computed tomographyAll of computed tomography (CT) scans were per-

formed on a DSCT (SOMATOM Definition; Siemens

Medical Solutions, Forchheim, Germany). The detector

collimation was 2 3 32 3 0.6 mm and acquisition colli-

mation was 2 3 32 3 0.6 mm by means of z-flying focal

spot. Gantry rotation time was 330 ms and pitch was 0.3–

0.45 adapted to heart rate (HR). DSCT angiography was

acquired using a retrospective gating with tube current

modulation scan protocol in the craniocaudal direction.

Full tube current was given during 30%–80% R-R

window. Bolus tracking was performed with a region-

of-interest placed in the root of ascending aorta and image

acquisition was automatically started 6 s after the signal

attenuation reached the predefined threshold of 100 HU.

The scanning range extended from the tracheal bifurcation

to just below the diaphragm.

The contrast medium injection used an 18-gauge

intravenous needle through right antecubital vein and

a dual-head power injector (Stellant D; Medrad, Indianola,

PA, USA). A two-phase protocol was used as a contrast

medium injection. According to scan time 50–70 mL,

contrast medium (Ultravist, 370 mgI/mL iopromide;

Bayer, Wayne, NJ, USA) was injected, followed by

30 mL saline (0.9% sodium chloride) as bolus chaser.

The injection rate for all phases was 4–4.5 mL/s.

The coronary tree was divided into segments accord-

ing to the modified American Heart Association classifica-

tion: left main and proximal mid and distal segments of the

left anterior descending artery, left circumflex, first and

second marginal branches and right coronary artery.

Transthoracic Doppler echocardiographyEchocardiographic examinations were performed

with a General Electronics Medical System (Vivid 7

dimension digital ultrasound system; General Electronics

Medical System, Waukesha, WI) with a harmonic probe at

a frequency of 4 MHz. Patients were examined in the left

lateral position using a modified left parasternal window.

Echocardiographic images were obtained from the

acoustic window around the midclavicular line in the

fourth and fifth intercostal spaces. After the lower portion

of the interventricular sulcus had been located in the long-

axis cross-section, the ultrasound beam was rotated

laterally, visualizing the distal portion of the left anterior

descending artery (LAD) under color flow-mapping guid-

ance. Color flow imaging was visualized using a high

frequency color Doppler technique. Blood flow velocity

was measured by pulsed-wave Doppler echocardiog-

raphy, using a sample volume (1.5–2.0 mm) placed on

the color signal in the distal LAD. After obtaining coro-

nary flow velocity at the baseline, adenosine was infused

into the peripheral arm vein at a rate of 140 mg/kg per min

for a total of 3 min. During the infusion, heart rate, blood

pressure and ECG were recorded and monitored at base-

line, every minute during adenosine infusion and for the

first 5 min after the infusion. An experienced operator

who was blinded to the patient’s data measured coronary

flow velocities. Both mean and peak diastolic flow veloc-

ities at baseline and at peak hyperemia were measured

manually by tracing the contour of the spectral Doppler

signals using analysis software incorporated into the

ultrasound system. Averages of the measurements were

obtained in three cardiac cycles. Peak diastolic velocity

(PDV) and mean diastolic velocity (MDV) were measured

off-line. The coronary flow velocity reserve (CFVR) was

calculated as the ratio of hyperemic to basal mean diastolic

flow velocities.

Intraobserver and interobserver variabilityIntraobserver and interobserver coefficient of varia-

tion for MDV and PDV measurements were analyzed, first

by one operator on two different occasions for validation

of intraobserver variability, and then by a second operator

for evaluation of interobserver variability.

Statistical analysisData are presented as mean 6 standard deviation and

compared with the two sample t-test. Data was analyzed

using SAS 9.0 (SAS Institute Inc., Cary, NC, USA)

1758 Ultrasound in Medicine and Biology Volume 36, Number 10, 2010

and p , 0.05 was considered statistically significant.

Intra- and interobserver variability was calculated using

intraclass correlation coefficients (ICC).

Fig. 1. An oblique coronal view in two-dimensional (2-D) mul-tiplanar reformatted image of hypercholesterolemia (HoFH)patient atherosclerotic plaques compromising the coronary

artery ostia.

RESULTS

Table 1 shows age, plasma lipid concentrations and

coronary flow parameters for the two study groups.

HoFH patients were similar to control subjects in age

and plasma triglyceride concentrations but had plasma

LDL concentrations higher than fourfold. Mean and

peak diastolic velocities in the LAD were not significantly

different in the two groups under basal conditions. When

vasodilator of the coronary arteries was induced with

adenosine, the increase in mean and peak diastolic veloc-

ities in the HoFH group was significantly lower than those

of the control group. As consequence of this inability to

achieve a normal increase in velocity during vasodilator

infusion, coronary flow velocity reserve was also signifi-

cantly lower in HoFH than in control subjects (3.36 vs.

1.92 respectively, p , 0.0001).

Aortic and coronary plaquesAll HoFH patients had aortic plaques and nine

patients had aortic plaques compromising the coronary

ostia (Fig. 1). In four patients, both the left and right coro-

nary artery ostia were involved. Five patients presented

coronary artery calcification. All patients had no signifi-

cant coronary artery stenosis (Table 2).

Table 3 listed lipid and coronary flow parameters of

HoFH subjects with and without plaques compromising

the coronary artery ostia. LDL cholesterol was signifi-

cantly higher in coronary artery ostia with plaques

subjects than those without. Coronary flow velocity

reserve was significantly lower.

Table 1. Blood lipid and coronary flow dynamics data inpatients with HoFH and control group

Control (n 5 15) HoFH (n 5 20) p

Age (years) 15.2 6 6.91 13.1 6 5.36 0.5772TC (mg/dL) 201.59 6 54.71 679.00 6 87.82 ,0.0001*LDL-C (mg/dL) 128.01 6 71.33 583.59 6 113.74 ,0.0001*TG (mg/dL) 105.12 6 48.96 106.22 6 43.43 0.9522

MDVb 23.60 6 5.58 27.82 6 7.72 0.1438PDVb 36.00 6 6.93 35.65 6 7.38 0.9034MDVh 77.96 6 18.12 51.88 6 11.78 ,0.0001*PDVh 91.60 6 15.09 69.76 6 13.91 0.0008*CFVR 3.36 6 0.61 1.92 6 0.40 ,0.0001*

Data are presented by mean 6 standard deviation.TC 5 total serum cholesterol; LDL-C 5 low-density lipoprotein

cholesterol; TG 5 triglyceride; MDVb 5 mean diastolic velocity at base-line; PDVb 5 peak diastolic velocity at baseline; MDVh 5 mean dia-stolic velocity in hyperemia; PDVh 5 peak diastolic velocity inhyperemia; CFVR 5 coronary flow velocity reserve.

* p , 0.05 marks statistical significance.

Intra- and interobserver variabilityTable 4 showed intra- and interobserver repeated

measure analysis using intraclass correlation coefficients

in the MDVb and MDVh.

DISCUSSION

Although HoFH is a rare disease, the mortality is

high due to the high prevalence of cardiovascular events.

It has been previously shown that sudden death and heart

failure are frequent causes of death in HoFH. Angina is

also a feature of HoFH since atherosclerosis may occur

on both aortic plaques and the coronary ostia.

A number of imaging techniques can be used to assess

coronary function and each has both advantages and disad-

vantages. Angiography is able to observe aortic and coro-

nary artery lumens and the degree of stenosis can be judged

directly. However, this method is invasive and cannot

accurately identify information about the vessel wall.

Magnetic resonance imaging (MRI) and CT, as noninva-

sive examinations, can detect early coronary atheroscle-

rosis in patients with HoFH and visualize plaque in both

the aorta and coronary artery but it is unable to evaluate

coronary hemodynamics (Santos et al. 2008; Summers

et al. 1998). Several tools have already been used to

measure coronary flow and coronary flow reserve (CFR)

for decades, including coronary sinus thermodilution,

cardiac nuclear imaging, cardiac magnetic resonance and

intracoronary Doppler flow wire. However, these

methods, mainly used research, have major limitations

on clinical analysis because of being complex, time-

consuming, expensive, invasive and not easily available.

TTDE can measure coronary flow and CFR and it is nonin-

vasive, easily available at bedside, not expensive and

Table 2. Aortic and coronary artery atherosclerosis

Patients

Aortic plaquesadjacent to

coronary ostia Aortic plaques compromising the coronary ostiaCoronary artery

calcification Affected artery

1 Yes compromising the left coronary ostia No None2 Yes Compromising the left coronary ostia Yes Mid LAD3 Yes compromising the left and right coronary ostia No None4 Yes compromising the left coronary ostia No None5 Yes compromising the left coronary ostia Yes Left main6 Yes compromising the left and right coronary ostia Yes Mid RCA7 Yes compromising the left coronary ostia Yes Left main8 Yes compromising the left and right coronary ostia Yes Left main and mid RCA9 Yes compromising the left and right coronary ostia No None

Evaluation of coronary flow velocity reserve d Y. YANG et al. 1759

without radiation exposure with multiple potential clinical

applications.

Using the noninvasive technique, TTDE, we found

a significant decrease in coronary flow velocity reserve

in the HoFH patients. In this study, TTDE and DSCT

were used to detect coronary flow dynamics and athero-

sclerosis in HoFH subjects. The most striking finding

was that CFVR was significantly lower in coronary artery

ostia with plaques subjects than those without. These

results suggest that TTDE assessment of CFVR might

be used in HoFH patients to assess their coronary artery

hemodynamics.

We used the ratio of hyperemic to rest coronary flow

velocity to calculate the CFVR, which is an important

index of coronary hemodynamics. CFR is defined as the

ratio of coronary flow under maximal coronary vasodila-

tion to coronary flow in resting condition. Intracoronary

Doppler and simultaneous angiography were used to deter-

mine CFR itself. However, this method is invasive and

costly and its clinical use is restricted. If coronary artery

cross-sectional area is similar during basal and hyperemic

conditions, CFVR can be used to represent CFR (Erbel

et al. 1996). TTDE can measure CFVR as a noninvasive

way. Average peak velocity, measured by TTDE, had

been found to correlate well with that measured invasively

Table 3. Lipid and hemodynamic parameters inhomozygous familial hypercholesterolemia patients withand without aortic plaques combined with the coronary

ostia

Aortic plaquescompromising the

coronary ostia

No aortic plaquescompromising the

coronary ostia p

CFVR 1.72 6 0.32 2.15 6 0.38 0.0029*TC (mg/dL) 714.67 6 105.57 638.88 6 37.60 0.0710LDL-C (mg/dL) 634.63 6 119.05 526.18 6 79.08 0.0455*TG (mg/dL) 95.45 6 37.62 118.33 6 48.75 0.2925

Data are presented by mean 6 standard deviation.CFVR 5 coronary flow velocity reserve; TC 5 total serum choles-

terol; LDL-C 5 low-density lipoprotein cholesterol; TG 5 triglyceride.* p , 0.05 marks statistical significance.

by intracoronary Doppler (Bartel et al. 1999; Murata et al.

2006). However, coronary artery flow velocity was

lower in the systolic, especially during the baseline; flow

spectrum can not be displayed completely in some

patients at the systolic. Nevertheless, flow spectrum can

be displayed thoroughly in the systolic either in the

baseline or in hyperemic. As a result, MDV was used to

calculate CFVR in this study.

All the patients were successfully measured with

CFVR in current study since all the patients were

children. Although no significant difference between

HoHF patients and controls was seen in MDV and PDV

at baseline, MDV and PDV were significantly decreased

in HoHF patients compared with controls during hyper-

emia (Figs. 2 and 3). All HoFH patients had aortic

plaques by DSCT and nine patients had aortic plaques

combined with the coronary ostia. In four patients,

both the left and right coronary artery ostia were

compromised while all patients had no significant

coronary artery stenosis. CFVR was impaired in HoFH

patients probably because the high serum cholesterol led

to an increase in coronary resistance (Kaufmann et al.

2000; Yokoyama et al. 1996; Awan et al. 2008). A

typical characteristic of HoFH is aortic plaques. CFVR

was lower and LDL-C concentration was increased signif-

icantly in the aortic plaques compromising the coronary

ostia compared with the group without compromising

the coronary ostia. Decrease of coronary artery flow perfu-

sion due to plaques compromising the coronary artery ostia

and coronary artery calcification would explain the signif-

icant decrease in CFVR seen in HoHF patients with aortic

Table 4. Intra- and interobserver ICC

MDVb MDVh

intraobserver interobserver intraobserver interobserver

r 0.987 0.989 0.992 0.990

MDVb 5 mean diastolic velocity at baseline; MDVh 5 mean diastolicvelocity in hyperemia.

Fig. 2. Transthoracic color Doppler echocardiography demon-strating coronary blood flow in the distal left anterior descending

artery (LAD) when basal.

1760 Ultrasound in Medicine and Biology Volume 36, Number 10, 2010

plaques compromising the coronary ostia (Beppu et al.

1983).

CONCLUSION

TTDE together with DSCT could be used to detect

both coronary and aortic atherosclerosis in HoFH subjects.

Noninvasive CFVR measurement using TTDE may be

useful to evaluate coronary hemodynamics. CFVR is

impaired in patients with HoFH due to existence of

coronary atherosclerosis. It is further impaired in HoFH

patients with aortic plaques compromising the coronary

ostia.

Study limitationsMurata et al. measured CFVR using TTDE in three

major coronary arteries. They defined CFVR , 2.0 in at

least one vessel as being positive for myocardial ischemia

and, then, compared TTDE results with single photon

Fig. 3. Transthoracic color Doppler echocardiography demon-strating coronary blood flow in the distal left anterior descending

artery (LAD) when hyperemic.

emission computed tomography (SPECT). The sensitivity

and specificity of CFVR , 2.0 in at least one coronary

vessel were 86% and 89%, respectively. However, it

was more difficult to measure CFVR in the posterior de-

scending artery or the lateral circumflex artery than in

the LAD (Murata et al. 2006). In the present study,

CFVR was only calculated in the LAD. DSCT angiog-

raphy can identify coronary plaques but the hemodynamic

significance of these is uncertain as most operators are

only able to report .50% stenosis and unable to distin-

guish hemodynamically significant disease.

Acknowledgments—This work was supported by the research grants fromNational Natural Science Foundation of China Research Grants30772356, the Natural Science Foundation of Beijing, China (Grant7062010).

REFERENCES

Awan Z, Alrasadi K, Francis GA, Hegele RA, McPherson R, Frohlich J,Valenti D, Varennes BD, Marcil M, Gagne C, Genest J, Couture P.Vascular calcifications in homozygote familial hypercholesterol-emia. Arterioscler Thromb Vasc Biol 2008;28:777–785.

Bartel T, Muller S, Baumgart D, Mathew B, Haude M, Erbel R.Improved high frequency transthoracic flow velocity measurementin the left anterior descending coronary artery after intravenousperipheral injection of Levovist. J Am Soc Echocardiogr 1999;12:252–256.

Beppu S, Minura Y, Sakakibara H, Nagata S, Park YD, Nambu S,Yamamoto A. Supravalvular aortic stenosis and coronary ostialstenosis in familial hypercholesterolemia: Two-dimensional echocar-diographic assessment. Circulation 1983;67:878–884.

Civeira F, Castillo S, Alonso R, Merino Ibarra E, Cenarro A,Artied M, Martin Fuentes P, Ros E, Pocovi M, Mata P. Tendonxanthomas in familial hypercholesterolemia are associated withcardiovascular risk independently of the low-density lipoproteinreceptor gene mutation. Arterioscler Thromb Vasc Biol 2005;25:1960–1965.

Erbel R, Ge J, Bockisch A, Kearney P, GOrge G, Haude M, Schumann D,Zamorano J, Rupprecht HJ, Meyer J. Value of intracoronary ultra-sound and Doppler in the differentiation of angiographically normalcoronary arteries: A prospective study in patients with angina pecto-ris. Eur Heart J 1996;17:880–889.

Faccenda F, Rubba P, Gnasso A, Pauciullo P, Postiglione A, Cortese C,Mancini M. Noninvasive ultrasound evaluation of pressure gradientsin aortic root of homozygotes for familial hypercholesterolemia.Arteriosclerosis 1990;10:710–713.

Kaufmann PA, Gnecchi-Ruscone T, Schafers KP, Luscher TF,Camici PG. Low density lipoprotein cholesterol and coronary micro-vascular dysfunction in hypercholesterolemia. J Am Coll Cardiol2000;36:103–109.

Kraft HG, Lingenhel A, Raal FJ, Hohenegger M, Utermann G. Lipopro-tein(a) in homozygous familial hypercholesterolemia. ArteriosclerThromb Vasc Biol 2000;20:522–528.

Marais AD. Familial hypercholesterolemia. Clin Biochem Rev 2004;25:49–68.

Murata E, Hozumi T, Matsumura Y, Fujimoto K, Sugioka K,Takemoto Y, Watanabe H, Yamagishi H, Yoshiyama M, Iwao H,Yoshikawa J. Coronary flow velocity reserve measurement in threemajor coronary arteries using transthoracic Doppler echocardiog-raphy. Echocardiography 2006;23:279–286.

Santos RD, Miname MH, Martinez LC, Rochitte CE, Chacra AM,Nakandakare ER, Chen D, Schaefer EJ. Noninvasive detection ofaortic and coronary atherosclerosis in homozygous familial hyper-cholesterolemia by 64-slice multidetector row computed tomographyangiography. Atherosclerosis 2008;197:910–915.

Summers RM, Andrasko-Bourgeois J, Feuerstein IM, Hill SC, Jones EC,Busse MK, Wise B, Bove KE, Rishforth BA, Tucker E, Spray TL,

Evaluation of coronary flow velocity reserve d Y. YANG et al. 1761

Hoeg JM. Evaluation of the aortic root by MRI: Insights frompatients with homozygous familial hypercholesterolemia. Circula-tion 1998;98:509–518.

Wiegman A, Rodenburg J, de Jongh S, Defesche J, Bakker H,Kastelein JP, Sijbrands EG. Family history and cardiovascular risk

in familial hypercholesterolemia: Data in more than 1000 children.Circulation 2003;107:1473–1478.

Yokoyama I, Murakami T, Ohtake T, Momomura S, Nishikawa J,Sasaki Y, Omata M. Reduced coronary flow reserve in familialhypercholesterolemia. J Nucl Med 1996;37:1937–1942.