Determinants Of Coronary FlowAbnormalities in Obstructive Type

Hypertrophic Cardiomyopathy: NoninvasiveAssessment by Transthoracic Doppler

EchocardiographySeden Celik, MD, Bahadir Dagdeviren, MD, Aydin Yildirim, MD, Sevket Gorgulu, MD,Nevzat Uslu, MD, Mehmet Eren, MD, Tayfun Gurol, MD, Ersin Ozen, MD, and Tuna

Tezel, MD, Istanbul, Turkey

We aimed to visualize the coronary flow velocities(CFV) of patients with hypertrophic obstructive car-diomyopathy by using transthoracic Doppler echo-cardiography, and to determine the relationshipbetween abnormal CFV patterns and conventionalechocardiography indices. Guided by 2-dimensionalechocardiography and Doppler color flow mapping,CFV in the distal left anterior descending coronaryartery were measured in 21 patients with hypertro-phic obstructive cardiomyopathy using a 3.5-MHztransducer. The results were compared with those of18 control subjects. Abnormal systolic flow patternswere observed in 15 (71%) patients (11 systolic-reversal flow and 4 no systolic flow). For patientsand control subjects, peak diastolic velocity andvelocity-time integral obtained from distal left ante-

744

rior descending coronary artery were higher (63 �21 cm/s and 18.5 � 4 cm vs 41 � 11 cm/s and 14.2 �5 cm, respectively; P < .01 for both) whereas peaksystolic velocity and velocity-time integral were sig-nificantly lower (�17 � 10 cm/s and 4.5 � 6 cm vs24 � 9 cm/s and 9.5 � 4 cm, respectively; P < .001for both). Significant positive and negative correla-tions between diastolic CFV and septal thicknessindex (r � 0.79, P < .0001), and between systolicCFV and septal thickness index (r � �0.65, P <.005), have been observed. CFV abnormalities thatcould easily be recorded by a standard Dopplerechocardiographic study seem to be related to septalthickness rather than the degree of obstruction inhypertrophic obstructive cardiomyopathy. (J AmSoc Echocardiogr 2004;17:744-9.)

Hypertrophic cardiomyopathy (HCM) is a geneticcardiac disease characterized by left ventricular (LV)hypertrophy in the absence of another cause ofincreased cardiac mass.1 Patients with HCM com-monly have evidence of myocardial ischemia despiteangiographically normal coronary arteries.2,3 Thealterations of phasic coronary flow characteristicshave been reported and proposed as a mechanismfor the development of angina pectoris in thesepatients.4-6

Invasive or semi-invasive techniques assessing cor-onary flow such as Doppler flow wire and trans-esophageal echocardiography are not practical fordaily use. Recent advancements of Doppler transtho-racic echocardiography (TTE) provide noninvasivemeasurement of coronary flow velocities (CFV) in

From the Department of Cardiology, Siyami Ersek Cardiovascularand Thoracic Surgery Center.Reprint requests: Seden Celik, MD, Gardenya 4-1 Ka:2, D:9,Atasehir, Kucukbakkalkoy, 81120 Istanbul, Turkey (E-mail:sedenerten@hotmail.com).0894-7317/$30.00Copyright 2004 by the American Society of Echocardiography.doi:10.1016/j.echo.2004.03.028

the distal epicardial coronary arteries. However,studies evaluating coronary flow with a transtho-racic approach have mostly been performed byusing high-frequency transducers and conducted onlimited numbers of patients.7,8

The purposes of this study were: (1) to evaluatethe feasibility and accuracy of transthoracic coro-nary flow imaging by using an ultrasonic systemequipped with a standard adult transducer in pa-tients with obstructive HCM (HOCM); (2) to inves-tigate the relationship between the severity of LVoutflow tract obstruction and CFV; and (3) to assesswhether a relationship exists between abnormalCFV pattern and other conventional echocardiogra-phy parameters.

METHODS

Patients

We studied 21 consecutive patients (mean age: 38 � 13years; 8 women) with the clinical and echocardiographicdiagnosis of HOCM on the basis of the demonstration of ahypertrophied and nondilated LV in the absence of a second-

Journal of the American Society of EchocardiographyVolume 17 Number 7 Celik et al 745

ary cause of LV hypertrophy. Inclusion criteria were: (1)interventricular septum thickness � 15 mm and septum/posterior wall ratio � 1.3; and (2) a basal systolic gradient �30 mm Hg in the LV outflow tract. In all study patients, drugtreatments were withheld 48 hours before the echocardio-graphic examination. A control group consisted of 18healthy age- and sex-matched volunteers. All patients wereinformed about the investigative nature of the study, and thelocal ethical committee approved the study protocol.

Echocardiographic Examination

TTE recordings were obtained from parasternal, apical,and subcostal windows by using ultrasound systems (GE-Vingmed System V and Vivid 7, GE-Vingmed Ultrasound,Horten, Norway) equipped with 1.5- to 3.7-MHz broad-band electronic transducers. Conventional M-mode, 2-di-mensional, pulsed wave, and color Doppler images wereacquired with simultaneous electrocardiographic tracings.LV systolic and diastolic dimensions, and wall thickness,were measured from the M-mode traces according to therecommendations of The American Society of Echocardi-ography.9 LV mass was calculated by the formula ofDevereux and Reichek.10 LV ejection fraction was calcu-lated with the Teichholz formula. Transmitral inflowvelocities were recorded at the level of mitral leaflet tips.The early and late diastolic velocities, deceleration time ofearly diastolic velocities, and isovolumic relaxation timewere also measured with previously described methods.11

Coronary Flow Imaging

For the optimal coronary flow imaging, the system presetswere adjusted as follows: Nyquist limit was 18 to 20 cm/s;color pulse frequency and pulse Doppler frequency were3.5 MHz; lateral and radial averaging was lowest; colorsample volume was minimum; and Doppler sample vol-ume was 1 to 2 mm. The patients were examined in theleft lateral position using a modified left parasternal win-dow. The transducer position was around the left midcla-vicular line in the fourth and fifth intercostal spaces. Afterthe anterior interventricular sulcus was imaged by anglingthe ultrasonic beam laterally and superiorly, coronaryblood flow of distal left anterior descending coronaryartery (LAD) was identified by Doppler color flow map-ping. Doppler spectral trace was recorded with a samplevolume positioned on the color signal within the artery.The long-axis sections were carefully adjusted to minimizethe angle (which was kept �20 degrees) between theDoppler beam and the long axis of the artery. Samplevolume was located within the vessel lumen, and was keptthere as long as possible during the cardiac cycle toacquire a typically diastolic predominant phasic CFVDoppler pattern (Figure 1). All echocardiographic imageswere stored in 5.5-in MO (magneto optic) disks, andmeasurements were made by using the internal analyzingsoftware package of the system.

By tracing the coronary blood flow Doppler spectrum,the peak systolic and diastolic velocities, and their veloc-ity-time integrals, were measured. For each of those 4

variables, the average value of the measurements from 3 to5 cardiac cycles was used for statistical calculations.

Doppler Flow Wire Study

A total of 10 patients for whom a coronary angiographywas planned by their physicians to role out the possibilityof coronary artery disease underwent Doppler flow wirestudy after the procedure was completed. Informed con-sent was obtained from those patients. Coronary angiog-raphy revealed normal findings in all. After the completionof the angiographic study, a 0.014-inch Doppler flow wire(FloWire, Cardiometrics, Mountain View, Calif) wasplaced in the center of the vessel coaxial to the lumen inthe distal portion of the LAD. A high quality of blood flowDoppler spectrum was attempted by carefully manipulat-ing the flow wire (Figure 1). Particular attention was paidto Doppler wire position, located between the last diago-nal branch and mustache of LAD, which was approxi-mately the same location viewed by Doppler TTE. Thepeak values and the velocity-time integrals of systolic anddiastolic CFV were measured in these 10 patients tocompare corresponding values recorded by Doppler TTE.All studies were continuously recorded on super-VHSvideotape.

Figure 1 Distal left anterior descending coronary arteryDoppler spectra obtained by transthoracic echocardiogra-phy (A) and flow wire (B) methods demonstrate typicalsystolic reversal flow for patient with hypertrophic obstruc-tive cardiomyopathy.

Journal of the American Society of Echocardiography746 Celik et al July 2004

Reproducibility and Agreement of CFVMeasurements by Doppler TTE and Flow Wire

To evaluate the effect of interobserver variability on CFVmeasurements by Doppler TTE, two independent, blindedobservers (A. Y. and N. U.) analyzed 21 randomly selectedDoppler velocity recordings. To evaluate intraobservervariability, 21 randomly selected Doppler velocities wereremeasured by the same operator (S. C.) 1 week apart. Toevaluate the agreement of CFV measurements obtained byDoppler TTE and Doppler flow wire methods, 10 patientswere analyzed. Interobserver and intraobserver variabili-ties, and the agreement of coronary flow measurementsby two methods, were calculated as the SD of the differ-ences between the two measurements or assessments,expressed as a percentage of the average value.

Statistical Analysis

Continuous variables were expressed as mean � 1 SD. Anunpaired 2-tailed t test was used for comparing valuesbetween groups. Linear regression analysis was used forthe correlations between CFV measurements and theother echocardiographic variables. Significance was set ata P value � .05. Software (SPSS, Version 10.0, Chicago, Ill)was used for computations.

RESULTS

Clinical, hemodynamic, and echocardiographic dataof the patients with HOCM and control subjects aregiven in Table. The groups were comparable interms of age, sex, body surface area, blood pressure,and heart rate. Patients with HOCM had a signifi-cantly thickened interventricular septum and, con-sequently, a higher LV mass index and ventricularseptum/posterior wall ratio than control subjects.

Table. Demographic, hemodynamic, and echocardiographcardiomyopathy and control subjects.

Variable Patients

Age (y) 38.7 � 13Sex (F/M) 8/13BSA (m2) 1.78 � 0.26HR (bpm) 72 � 8BP-sys (mm Hg) 121 � 29BP-dias (mm Hg) 70 � 11LV mass index (g/m2) 257 � 35LVOT max-grd (mm Hg) 60 � 33EF (%) 76 � 11E vel (cm/s) 87 � 32A vel (cm/s) 66 � 20IVRT (msec) 90 � 17

A vel, Atrial filling velocity; BP-dias, diyastolic blood pressure; BP-sys, systovelocity; F, female; HR, heart rate; IVR T, isovolumic relaxation time; LV,not significant.

The average level of intraventricular pressure gradi-ent was 60 � 33 mm Hg in study patients.

CFV Analysis

Technically adequate Doppler flow velocity tracingsof the distal LAD could be obtained from all 21consecutive study patients, but only from 18 of 24control subjects (75%).

The mean diastolic and systolic CFV of patientswith HOCM and control subjects are displayed inFigure 2. The peak diastolic velocity and velocity-time integral were significantly higher in patientswith HOCM than in control subjects (63 � 21 cm/sand 18.5 � 4 cm vs 41 � 11 cm/s and 14.2 � 5,respectively; P � .01 for both).

Because CFV were in normal antegrade form in allcontrol subjects but there was a reversal flow in 11and no systolic flow in 4 patients with HOCM (P �.0001), the peak systolic velocity and velocity-time

of patients with hypertrophic obstructive

Control

subjects P

35 � 11 NS7/11 NS

1.80 � 0.30 NS68 � 10 NS

115 � 25 NS65 � 10 NS

112 � 15 0.001

67 � 10 NS89 � 24 NS62 � 15 NS78 � 20 NS

d pressure; BSA, body surface area; EF, ejection fraction; E vel, early fillingricular; LVOT max-grad, LV outflow tract maximal gradient; M, male; NS,

Figure 2 Mean values of peak diastolic and systolic leftanterior descending coronary artery (LAD) velocities forpatients with hypertrophic obstructive cardiomyopathy(HOCM) and control subjects.

ic data

lic blooleft vent

tion ti

Journal of the American Society of EchocardiographyVolume 17 Number 7 Celik et al 747

integral were significantly lower in HOCM group (24� 9 cm/s and 9.5 � 4 cm vs �17 � 10 cm/s and 4.5� 6 cm, respectively; P � .001 for both).

Linear regression revealed that for patients withHOCM, no correlation existed between intraventric-ular pressure gradient and either systolic or diastolicLAD flow velocities. On the other hand, a significantnegative correlation was observed between inter-ventricular septum thickness index (interventricularseptum thickness/body surface area) and LAD peaksystolic velocity (r � �0.65; P � .005) (Figure 2). Inthe analysis of diastolic CFV, significant positivecorrelations were observed between peak diastolicLAD velocities and both interventricular septumthickness and its indices (r � 0.79; P � .0001).Among the conventional diastolic Doppler parame-ters, the peak early inflow velocities showed a goodnegative correlation (r � �0.79; P � .005), whereasisovolumic relaxation time showed moderately pos-itive correlation with peak systolic CFV (r � 0.60; P� .05) (Figure 3).

Figure 3 Correlation between peak diastolic left ainterventricular septum thickness index for patients(A). Relation between peak systolic LAD velocitiesearly filling (E) velocities (C), and isovolumic relaxa

Reproducibility

Interobserver variability of Doppler velocity mea-surements was 7.2% for systolic and 4.4% for dia-stolic velocities. Intraindividual variability was excel-lent and never exceeded 2 cm/s for both diastolicand systolic peak velocity, which provided a maxi-mal �5.6% difference in relative terms. The agree-ment of CFV measurements obtained by DopplerTTE and flow wire methods was also in acceptablelimits. The mean variability of CFV measurements bytwo methods was 4.2% and 6.1% for diastolic andsystolic velocities, respectively.

DISCUSSION

In patients with HOCM, the noninvasive DopplerTTE technique allows acquisition of the flow pat-terns of distal LAD, measuring the peak velocity andvelocity-time integrals, and demonstrating the flow

descending coronary artery (LAD) velocities andypertrophic obstructive cardiomyopathy (HOCM)interventricular septum thickness index (B), mitralme (IVRT) (D) for patients with HOCM.

nteriorwith hwith

Journal of the American Society of Echocardiography748 Celik et al July 2004

pattern abnormalities involving the systolic phasesuch as its absence and paradoxical reversal. Fur-thermore, patients with HOCM showed significantlyincreased diastolic CFV in distal LAD when com-pared with control subjects. Both systolic and dia-stolic CFV were observed to significantly correlatewith septal thickness and diastolic function param-eters, but not with intraventricular pressure gradi-ent.

Coronary Flow Abnormalities in HOCM

Some previous studies using invasive and noninva-sive methods demonstrated decreased or reversedsystolic and augmented diastolic coronary flow inpatients with HOCM.5-7,12 Possible mechanisms thathave been proposed to explain these alterations arethe abnormally increased coronary vascular resis-tance as a result of systolic compression of largeintramural arteries, throttling effect of small intramu-ral coronary arteries, and reduced compliance ofepicardial coronary arteries because of maximalvasodilatator state even in resting condition.

Consistent with those mechanisms, CFV alter-ations seem to have a close relationship with thedegree of septal thickening, as our study demon-strated. This scenario emphasizes that as the myo-cardial wall thickens, intramural compression on thecoronary artery increases, and both the throttlingeffect and the reduced compliance because of max-imal-basal vasodilatation of LAD become more prom-inent.

In the literature, there are some observationssupporting our findings. Tomochika et al5 measuredCFV in the proximal LAD of 7 patients with nonob-structive HCM by TEE; systolic flow was reducedand reversal flow occurred in two patients withventricular septal thickness � 2 cm, and a significantcorrelation was found between septum thicknessand peak systolic velocities. Memmola et al13 alsoreported a significantly greater diastolic and lowersystolic component of CFV in the proximal LAD of10 patients with HOCM by TEE. In our study, anentirely noninvasive approach performed to a largerpatient group with HOCM demonstrated consistentfindings with those observations.

In addition, it is interesting that the CFV variableswere not related to the degree of obstruction sever-ity in our study patients. Altered coronary flowcharacteristics—regardless of the severity of systolicpressure increase in LV—in relation with the degreeof thickening of surrounding myocardium, couldsuggest the mechanisms involving the local func-tional and structural changes regulating coronaryflow that might probably contribute to the genera-tion of abnormal CFV patterns in patients withHOCM. The findings of a previous study by Kyriaki-dis et al6 suggested regional difference of coronaryflow between the coronary arteries responsible

from the perfusion of hypertrophic and normal LVsegments; they reported an increased coronary flowin LAD compared with left circumflex coronaryartery by using flow wire method in 18 patients withHOCM. Thus, one might argue that the regionaldistribution of hypertrophy closely relates with theregional impairment of coronary flow, and the localregulatory mechanisms largely influence the coro-nary flow dynamics rather than increased intraven-tricular pressure. Furthermore, previous observa-tions, which reported similar CFV abnormalities,also for patients with nonobstructive HCM,5,14 maylet us think that the intraventricular pressure gradi-ent is not the predominant factor responsible fromthe CFV alterations for patients with HOCM. Incontrast to our findings, transvalvular pressure gra-dient has been shown to correlate with the degreeof systolic reversal coronary flow in patients withvalvular aortic stenosis, which may be considered asanother form of LV outflow tract obstruction.15,16 Inour opinion, these controversial results might berelated to the difference between hemodynamiceffects of fix-valvular stenosis in which peak gradi-ent was measured at midsystole and dynamic ob-struction of HCM, which has a less effective gradientat end-systole.

This study also demonstrates that the alterationsin CFV significantly relate to the degree of LVdiastolic functional impairment. The correlation ofabnormally decreased systolic coronary flows withhigher early diastolic velocities and shorter isovolu-mic relaxation time may be caused by an interactionbetween more advanced form of diastolic dysfunc-tion or high LV filling pressures with coronary flowdynamics.

Clinical Implications

Noninvasive imaging of distal LAD flow velocities byDoppler TTE without requiring a high-frequencytransducer would stimulate further research to ex-plore the prognostic implication of coronary flowabnormalities, and to assess whether the abnormalcoronary flow pattern, when established, couldguide patient selection for different medical andinterventional treatment strategies.

Study Limitations

Although our comparative Doppler flow wire studydemonstrated a satisfactory correlation of the resultsin 10 patients with HOCM, the study fails to validatethe Doppler echocardiography method of assessingdistal LAD flow in a larger group of individualsbecause such a study of an invasive nature could nothave been justified because of ethical reasons. Incontrast to patients with hypertrophied ventricles,echocardiographic assessment of LAD flow was notpossible in all control subjects. This difficulty toobtain CFV in control subjects in our study may be

Journal of the American Society of EchocardiographyVolume 17 Number 7 Celik et al 749

related to the use of an ultrasonic system equippedwith a standard adult transducer, which is probablysuboptimal for control subjects without hypertro-phied ventricles.

Conclusion

In conclusion, CFV measurements of distal LAD byDoppler TTE revealed a significant increase in dia-stolic velocities whereas a significant reduction ofsystolic velocities was shown for patients withHOCM compared with control subjects. These ab-normalities of CFV closely correlate with septalthickness and severity of diastolic dysfunction ratherthan the degree of intraventricular pressure gradient

REFERENCES

1. Maron BJ, Epstein SE. Hypertrophic cardiomyopathy: a dis-cussion of nomenclature. Am J Cardiol 1979;43:1242-4.

2. O’Gara PT, Bonow RO, Maron BJ, Damske BA, van LingenA, Bachrach SL. Myocardial perfusion abnormalities in pa-tients with hypertrophic cardiomyopathy: assessment withthallium-201 emission-computed tomography. Circulation1987;6:1214-23.

3. Cannon RO, Rosing DR, Maron BJ, Leon MB, Bonow RO,Watson RM, et al. Myocardial ischemia in hypertrophic car-diomyopathy: contribution of inadequate vasodilator reserveand elevated left ventricular filling pressures. Circulation1985;71:234-43.

4. Maron BJ, Wolfson JK, Epstein SE, Roberts WC. Intramural(“small vessel”) coronary artery disease in hypertrophic cardio-myopathy. J Am Coll Cardiol 1986;8:545-57.

5. Tomochika Y, Tanaka N, Wasaki Y, Shimizu H, Hiro J,Takahashi T, et al. Assessment of flow profile of left anteriordescending coronary artery in hypertrophic cardiomyopathyby transesophageal pulsed Doppler echocardiography. Am JCardiol 1993;72:1425-30.

6. Kyriakidis MK, Dernellis JM, Androulakis AE, Kelepeshis GA,Barbetseas J, Anastasakis AN, et al. Changes in phasic coronaryblood flow velocity profile and relative coronary flow reserve inpatients with hypertrophic obstructive cardiomyopathy. Cir-culation 1997;96:834-41.

7. Crowley JJ, Dardas PS, Harcombe AA, Shapiro LM. Trans-thoracic Doppler echocardiographic analysis of phasic coro-nary blood flow velocity in hypertrophic cardiomyopathy.Heart 1997;77:558-63.

8. Dimitrow PP, Krzanowski M, Nizankowski R, Szczeklik A,Dubiel JS. Effect of verapamil on systolic and diastolic coro-nary blood flow velocity in asymptomatic and mildly symp-tomatic patients with hypertrophic cardiomyopathy. Heart2000;83:262-6.

9. Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R,Feigenbaum H, et al. Recommendations for quantitation ofthe left ventricle by two-dimensional echocardiography:American Society of Echocardiography committee on stan-dards, subcommittee on quantitation of two-dimensionalechocardiograms. J Am Soc Echocardiogr 1989;2:358-67.

10. Devereux RB, Reichek N. Echocardiographic determinationof left ventricular mass in man: anatomic validation of themethod. Circulation 1977;55:613-8.

11. Oh JK, Appleton CP, Hatle LK, Nisgimura RA, Seward JB,Tajik AJ. The noninvasive assessment of left ventricular dia-stolic function with two-dimensional and Doppler echocardi-ography. J Am Soc Echocardiogr 1997;10:246-70.

12. Akasaka T, Yoshikawa J, Yoshida K, Maeda K, Takagi T,Miyake S. Phasic coronary flow characteristics in patients withhypertrophic cardiomyopathy: a study by coronary Dopplercatheter. J Am Soc Echocardiogr 1994;7:9-19.

13. Memmola C, Iliceto S, Venanzio F, Napoli V, Cavallari D,Santoro G, et al. Coronary flow dynamics and reserve assessedby transesophageal echocardiography in obstructive hypertro-phic cardiomyopathy. Am J Cardiol 1994;74:1147-51.

14. Petkow Dimitrow P, Krzanowski M, Nizankowski R, Szczek-lik A, Dubiel JS. Effect of verapamil on systolic and diastoliccoronary blood flow velocity in asymptomatic and mildlysymptomatic patients with hypertrophic cardiomyopathy.Heart 2000;83:262-6.

15. Petropoulakis PN, Kyriakidis MK, Tentolouris CA, Kourou-clis CV, Toutouzas PK. Changes in phasic coronary bloodflow velocity profile in relation to changes in hemodynamicparameters during stress in patients with aortic valve stenosis.Circulation 1995;92:1437-47.

16. Tamborini G, Barbier P, Doria E, Galli C, Maltagliati A,Ossoli D, et al. Influences of aortic pressure gradient andventricular septal thickness with systolic coronary flow inaortic valve stenosis. Am J Cardiol 1996;78:1303-6.