Assessment of mitral flow velocity with exercise by an index of stress-induced left ventricular ischemia in coronary artery disease

Assessmentof Mitral Flow Velocity with Exercise by an Index of Stress-Inducedleft Ventricular lschemia in Coronary Artery Disease GEORGE D. MITCHELL, MD, RICHARD C. BRUNKEN, MD, MARKUS SCHWAIGER, MD, BRYAN C. DONOHUE, MD, JANINE KRIVOKAPICH, MD, and JOHN S. CHILD, MD

Exercise-induced myocardial &hernia results in both diastolic and systolic left ventricular (LV) dysfunction. To investigate the utility of Doppler assessment of LV diastolic function with exercise, 28 consecutive patients underwent digital stress echocardiography, including measurement of mitral flow velocity by pulsed-wave Doppler and simultaneous stress thallium imaging. The mean mitral flow velocity was measured as the integrated area under the LV diastolic inflow Doppler spectral display. The change in mean mitral flow velocity from baseline to Immediate postexercise was compared among 3 patient groups: (1) patients with thallium redistribution or exercise-induced wall-motion abnormalities, or both, consistent with exercise-induced ischemia (n = 18), (2) patients with no evidence of stressinduced ischemia, with or without resting wall-motion abnormalities (n = IO), and (3) 10 control

subjects of similar age with normal resting la-lead electrocardiograms, normal resting and postexercise P-dimensional echocardiograms and normal electrocardiographic treadmill stress testing. The percent increase in mean mitral flow velocity was 101% (f59) for controls and 88% (f53) for patients without stress-induced ischemia versus 33% (f24) in patients with stress-induced ischemia (p 50% correctly identified 9 of IO nonischemic control patients. An increase in mean velocity of <50% predicted ischemia in 15 of 18 patients with evidence of stress-induced ischemia (p <0.005) Thus, Doppler assessment of LV diastolic function with exercise expressed as a change in the mean velocity of mitral flow is a useful indicator of stress-induced &hernia. (Am J Cardiol 1988;81:538-540)

S

tudies have shown that exercise-induced myocarphy is a noninvasive means of obtaining beat by beat dial ischemia can produce abnormalities in both dia- analysis of LV diastolic fillings6-lo The characteristics stolic and systolic left ventricular (LV) function.1-5 of the Doppler spectral display of diastolic transmitral Several methods have been used to assess diastolic flow velocity have been studied in patients with sysfunction, including frame by frame analysis of contrast temic hypertension, dilated cardiomyopathy, acute myocardial infarction and during percutaneous transventriculograms to determine rates of diastolic filling and analysis of time-activity curves obtained by gated luminal coronary angioplasty.11-16 Rassi et all7 have blood pool scintigraphy to obtain filling rates and fill-, shown in normal subjects undergoing supine bicycle ing fractions. These methods are not easily applied to exercise that the mean mitral flow velocity increased the study of large numbers of patients or serial assess- by more than 100% from baseline to peak exercise. Changes in transmitral flow occurring with exercise in ment of individual patients. Doppler echocardiograpatients with coronary artery disease have not been well described. This study investigates the usefulness From the Department of Medicine, Division of Cardiology, and of Doppler assessment of LV diastolic function during the Department of Radiology, Division of Nuclear Medicine, stress testing in patients with known or suspected coroUCLA School of Medicine, Los Angeles, California. Manuscript nary artery disease as a possible marker of exercisereceived July 24,1987; revised manuscript received and acceptinduced ischemia. ed October 29,1987. Dr. Mitchell’s present address: Baystate Medical Center, Springfield, Massachusetts. Address for reprints: George D. Mitchell, MD, Cardiology Service, Baystate Medical Center, 759 Chestnut Street, Springfield, Massachusetts 01199.

Methods Patients: A total of 28 consecutive patients referred to the UCLA noninvasive cardiology laboratory for treadmill stress testing underwent simultaneous digi536

March 1, 1988

tal stress echocardiography and stress thallium imaging. The “ischemic” group (group l] comprised 18 patients with exercise-induced wall:motion abnormalities, exercise-induced thallium perfusion defects, or both. The “nonischemic” group (group 2) included 10 patients without exercise-induced wall motion abnormalities and negative thallium scans. Since the time of initiation of stress echocardiography in this laboratory, a total of 10 patients undergoing digital stress echocardiography with measurement of transmitral flow velocity had normal results in their resting 12-lead electrocardiograms, resting and postexercise 2-dimensional echocardiograms and electrocardiographic treadmill stress testing. These patients, whose ages were similar to those of the study patients, constituted the control group (group 31. Echocardiography: Resting and immediate postexercise 2-dimensional echocardiographic imaging was performed on supine subjects with a HewlettPackard 77020A echocardiographic recorder with a 2.5 MHz electronically phased array transducer with pulsed Doppler (2.0 MHz) capability. Optimal windows for obtaining parasternal and apical views at rest were clearly marked on each patient’s chest wall. Immediate postexercise Z-dimensional echocardiographic images were obtained in all cases within 30 to 60 seconds of termination of upright treadmill exercise. Digitized tine loops of .&dimensional echocardiographic images obtained using the apical J-chamber, &chamber and long axis views and the parasternal short axis view were made on a Microsonics digitizing system. Each apical view was divided into 6 segments.18 Wall-motion analysis was performed independently by 2 observers unaware of other patient data using a grading system from -1 for dyskinesia to 4 for hyperkinesia. Discrepancies between observers were resolved by consensus review. The development of a new exercise-induced wall-motion abnormality was defined as >l grade deterioration in wall-motion score for any 1 segment from baseline to immediate postexercise. The development of dyskinesia in an area of prior akinesia was not considered a new wallmotion abnormality. Doppler: The time-velocity profile of transmitral flow by pulsed Doppler was obtained with the sample volume placed on the ventricular side of the mitral anulus near the tips of the mitral leaflets. Beats used for analysis were those that showed the greatest mitral peak flow velocity regardless of the phase of respiration. Four parameters of LV filling were obtained from the analysis of the time-velocity spectral display: (I] peak velocity in early diastole; (2) peak velocity during atria1 systole; (3) ratio of early diastolic peak flow velocity to peak flow velocity during atria1 systole; (4 mean mitral flow velocity-obtained by digitizing the contour of the darkest outline of the spectral display of the transmitral time-velocity profile (Figure 1). Digitization was performed using a standard, commercially available, digitizing pad from Trinity Systems. The integrated area under the curve was expressed as the mean mitral flow velocity. The integrated area under the spectral envelope (mean mitral flow velocity) normalized for the diastolic filling period was expressed

THE AMERICAN

JOURNAL

OF CARDIOLOGY

Volume 61

537

as the time-velocity integral. Pulsed Doppler measurements were obtained immediately after 2-dimensional echocardiographic imaging at baseline and upright treadmill exercise. Thallium imaging: Stress thallium imaging was performed after obtaining echocardiographic studies immediately after exercise. Thallium images (64 of 20) were obtained using a Siemens Orbiter-SPECT system CLC 7501X. Redistribution images were obtained 4 hours after exercise. Those thallium perfusion defects present immediately after exercise and exhibiting redistribution 4 hours later were considered to represent exercise-induced myocardial ischemia. Thallium studies were read by consensus of 2 observers blinded to other patient data. Statistical methods: Independent sample ,&tailed Wilcoxon rank sum tests for unpaired data were performed. Differences between means of more than 2 groups were evaluated by l-way analysis of variance. 2 X 2 contingency tables were analyzed using the 2tailed Fisher’s exact test. Values are reported as mean f standard deviation. A probability value
FIGURE 1. Sample spectral display of Doppler transmitrai flow velocity. E = peak velocity in early diastoie; A = peak velocity during atriai systoie; integrated area under the darkest outline of the spectral display = mean mitral flow velocity.

538

MITRAL FLOW VELOCITY

IN CORONARY

ARTERY DISEASE

Results A summary of the overall results of simultaneous digital stress echocardiography and stress thallium imaging performed in the 28 consecutive patients with known or suspected coronary artery disease undergoing treadmill stress testing is shown in Figure 2. The 18 patients constituting the “ischemic group” (group 1) comprised 12 with positive SPECT thallium scan results (group 1A) and 6 with negative thallium scan findings but with exercise-induced wall-motion abnormalities (group 1B). Of the 10 patients with “nonischemic” studies (group 2), 7 had resting wall-motion abnormalities (Group 2A). For the 3 groups undergoing comparison, their mean ages (63 f lo,55 f 12 and 55 f 14 years for groups 1, 2 and 3, respectively], baseline heart rates (69 f 18, 67 f 12 and 61 f 9 beats/min), postexercise heart rates (88 f 23, 98 f 20 and 91 f beats/min), percentage of maximum predicted heart rates at peak exercise (81 f 15,87 f 13 and 88 f 12%), and double-products achieved (21,000 f 5,000, 25,000 f 7,000 and 25,000 f 6,000 mm Hg - beats/min] were comparable. The number of patients in each group with a history of hypertension (8 of 18 in group 1,3 of 10 in group 2 and 3 of 10 in group 3) was similar (p = 0.9). There was a trend toward greater use of calcium antagonists or P-blocking drugs in groups 1 and 2 (9 of 18 and 6 of 10, respectively), compared with group 3 (2 of lo] (p = 0.2). Mean mitral flow velocities at baseline and after exercise for each of the patient groups are listed in Table I. In controls, the mean mitral flow velocity doubled, whereas patients with exercise-induced ischemia had an increase in mean mitral flow velocity of only 33% (p
results and the group with exercise-induced wall-motion abnormalities and negative thallium scan results. An increase in mean mitral flow velocity of >50% with exercise correctly identified 9 of 10 controls as nonischemic. The absence of a >50% increase in mean mitral flow velocity predicted an ischemic response in 15 of 18 patients with exercise-induced ischemia (p
BASELINE

POST- EX.

A

29 PATIENTS

12 POSITIVE

/

THALLIUM

**

9 NEW WMA

3 NO NEW WMA

7 RESi

3 REST WMA

WMA

\

16 NEGATIVE

6 NEW WMA

6 REiT

WMA

**

B THALLIUM

* 10 NO NEW WMA

7 RESi

WMA

FIGURE 2. Overall results of simultaneous digital stress echocardiography and stress thallium imaging in 28 consecutive patients with known or suspected coronary after coronary artery disease. * nonlschemic group; * * ischemic group. WMA = wall-motion abnormality.

FIGURE 3. Preexercise and postexercise spectral display of Doppler lransmilral flow velocity in 2 representative patients. A, control: a 81.year-old female, 13.minute total exercise lime on Bruce prolocol, reached 88 % maxlmal predicted heart rate, double-product = 23,800, baseline mean mitral flow velocity = 33 cm/s, postexercise mean mitral flow velocity = 87 cm/s (103 % increase). S, thallium redistribution and exercise wall-motion abnormality developed in this 58-year-old male, 115minute total exercise lime on Bruce protocol, reached 100% maximal predicted heart rate, doubleproduct = 27,200, baseline mean mitral flow velocity = 42 cm/s, postexercise mean mitral flow velocity = 53 cm/s (28% increase). Each division = 20 cm/s.

March 1, 1988

TABLE I

Pt Group 1 IA IB 2 2A 3

Mitral Flow Velocity Baseline Mitral Flow Velocity (cm/s) 39f 42f 32 f 33 Lt 35 f 31 f

THE AMERICAN

JOURNAL

OF CARDIOLOGY

Volume 61

539

Data Postexercise Mitral Flow Velocity (cm/s)

13 15 6 7 7 7

Patients with X0% Increase in Mitral Flow Velocity* (cm/s)

Percent Change in Mitral Flow Velocity* (cm/s)

51 f 18 56 f 20 41 f 9 59f 12 63f 10 60f 12

33 f 24 31 f29 34& 24 > 86 f 53 90 f 64 101 f 59 >

.

All values are mean f standard deviation. l p = <0.005 between groups.

TABLE II

Diastolic

Pt Group

Baseline DFP

1 2 3

0.48 f. 0.16 0.48 3z 0.11 0.59 f 0.13 NS

Filling Period and Time-Velocity Postexercise DFP 0.37 f 0.15 0.29 f 0.06 0.34 f 0.09 NS

Integral Data

Percent Change in DFP 23f 375 40f p =

16 14 18 0.02

All values are in mean rt standard deviation. DFP = diastolic filling period (in seconds of diastolelbeat);

groups are listed in Table II. The percent change in time velocity integral from baseline to postexercise was no different among the 3 groups. The percent decrease in diastolic filling period was less for the ischemic group than either controls or patients with nonischemic studies (p = 0.02) despite similar heart rates at baseline and postexercise. The ratios of peak velocity in early diastole to peak velocity during atria1 systole (E/A ratios), were no different among the patient groups at baseline or postexercise. The percent change in this parameter from baseline to postexercise did not differ among the 3 groups.

Discussion Recent interest has focused on the use of Doppler velocity measurements to assess global LV function. Aortic and transmitral flow velocity determinations combined with anular cross-sectional area measurements have been used to attempt to estimate stroke volume and cardiac output.lgJo The response of aortic flow velocity to exercise has been extensively studied.21-23Little is known of the response of transmitral flow to upright treadmill exercise in normal subjects or patients with coronary artery disease. This study corroborates the increase in transmitral flow velocity seen by Rassi et all7 in normal subjects, by showing a similar increase of at least 100% in controls after treadmill exercise. It also demonstrates an abnormal response of the velocity of transmitral flow with exercise in patients with noninvasive evidence of stress-induced myocardial ischemia. The failure in ischemic patients to increase transmitral flow velocity with exercise to the extent seen in controls or nonischemic patients

Baseline TVI

Postexercise TVI

18f5 16 f 4 17f4 NS

17 f 4 l7f4 20 f 4 NS

NS = not significant;

Percent Change in TVI

TVI = time-velocity

-1 P 18 13 f 23 11 f 17 NS

integral.

may be a reflection of ischemia-induced compromise of stroke volume, LV filling, or both. The change from baseline to postexercise in the time-velocity integral, which relates more directly to total transmitral flow than mean velocity alone, was not significantly different among the 3 study groups. Although given the small number of patients the likelihood of beta error must be considered, this would suggest that changes in stroke volume were not significantly different and that an abnormal response of mean mitral flow velocity with exercise in ischemic patients more likely represents an ischemia-induced change in LV compliance and, therefore, resistance to filling. Likewise, the smaller percent decrease in diastolic filling period from baseline to postexercise in ischemic patients may reflect a relative prolongation of the diastolic filling period due to ischemia-induced decrease in LV compliance. The changes in the mean velocity of transmitral flow with exercise were similar in controls and patients without evidence of exercise-induced ischemia, even though 7 of 10 patients in this nonischemic group had resting wall-motion abnormalities on 2-dimensional echocardiography. When considered separately, these 7 patients with presumably nonischemic but infarcted hearts had quite similar increases in transmitral flow with exercise compared with controls. This would suggest that the abnormal response of transmitral flow velocity with exercise (i.e., failure to increase by at least 50%) may be a marker of exercise-induced ischemia rather than merely a marker of resting LV compliance abnormalities related to prior ischemic injury.

540

MITRAL FLOW VELOCITY

IN CORONARY

ARTERY DISEASE

Because baseline, postexercise and percent of maximum predicted heart rates were all similar, the differences noted between groups in the response of mean mitral flow velocity to exercise are not a result of differences in heart rate. Likewise, the degree of stress induced was similar, as there was no difference between groups in double-product achieved. Changes in the pattern of LV filling by pulsed Doppler may occur with aging.24-26The control group used in this study was of an age distribution similar to the ischemic and nonischemic groups. Hypertension and therapy with calcium antagonists or P-blocking drugs can affect LV compliance. There was no significant difference between groups in the incidence of hypertension. Although there was a trend toward more frequent use of ,&blockers or calcium antagonists in the ischemic and nonischemic groups, as compared with controls, this difference did not reach statistical significance. The effect of these agents on diastolic LV function would be to increase myocardial compliance, which would theoretically improve transmitral flow and, therefore, would not account for the differences shown. Regardless, each patient served as his or her own control for the comparison of rest versus immediate postexercise mitral flow velocities. The peak velocity in early diastole (E) and peak velocity during atria1 systole (A], obtained by pulsed Doppler, and their ratio [E/A ratio], have been studied in normal subjects and in various forms of acquired heart disease. The absolute values for the E/A ratio and the percent changes in E/A ratio from baseline to postexercise were no different between patient groups. With increased heart rates after exercise, it was noted that the E and A velocities tended to fuse, making separate determination of E and A velocity integrals technically difficult. The mean velocity of transmitral flow, which takes into account the velocities in early diastole and during atria1 systole, would appear to be a more reliable and reproducible measure of Doppler-derived changes in transmitral flow with exercise.

References 1. Cove11JW, Ross J Jr. Nature and significance of alterations in myocardial compliance. Am J Cardiol 1973:32:449-455. 2. Bristow JD, Van Zee BE, Judkins MP. Systolic and diastolic abnormalities of the left ventricle in coronary artery disease. Circulation 1970;42:219-228. 3. Bonow RO, Bacharach SL, Green MV, Kent KM, Rosing DR, Lipson LC, Leon MB, Epstein SE. Impaired left ventricular diastolic filling in patients with coronary artery disease: assessmentwith radionuclide angiography. Circulation 1981;64:315-323. 4. Reduto LA, Wickemeyer WJ, Young JB, Del Ventura LA, Reid JW, Glaser DH, Quinones MA, Miller RN. Left ventricular diastolic performance at rest and during exercise in patients with coronary artery disease. Circulation 198X63:1228-1237. 5. Bourdillon PD, Lore11BH, Mirsky I, Paulus WJ, Wynne J, Grossman W.

Increased myocardial stiffness of the left ventricle during pacing-induced angina in man. Circulation 1983;67:316-322. 6. Hammermeister KE, Warbasse JR. The rate of change of left ventricular volume in man. II. Diastolic events in health and disease. Circulation 1974;49:739-747. 7. Rokey R, Kuo LC, Zoghbi WA, Limacher MC, Quinones MA. Determinotion of parameters of left ventricular diastolic filling with pulsed Doppler echocardiography: comparison with cineangiography. Circulation 1985; 71:543-550.

8. Spirit0 P, Maron BJ, Bonow RO. Noninvasive assessmentof left ventricular diastolic function: comparative analysis of Doppler echocardiographic and radionuclide angiographic techniques. JACC 1986;7:518-526. 9. Friedman BJ, Drinkovic N, Miles H, Shih W, Mazzoleni A, DeMaria AN. Assessment of left ventricular diastolic function: comparison of Doppler echocardiography and gated blood poool scintigraphy. JACC 1986;8:23481354.

10. Kitabatake A, Inoue M, Asao M. Transmitral blood flow reflecting diastolic behavior of the left ventricle in health and disease-a study by pulsed Doppler technique. Jpn,Circ J 1982;46:92-102. 11. Snider AR, Gidding SS, Rocchini AP. Doppler evaluation of left ventricular diastolic filling in children with systemic hypertension. Am J Cardiol 1985;56:921-926.

12. Phillips RA, Coplan NL, Krakoff LR, Yeaner K. Ross RS, Gorlin R, Goldman ME: Dopplerkchocardiographic anaIy& of left ventricular filling in treated hypertensive patients. JACC 2987;9:317-322. 13. Takenaka K, Dabestani A, Gardin JM, Russell D, Clark S, Allfie A, Henry WL. Left ventricular filling in hypertrophic cardiomyopathy: a pulsed DoppIer echocardiographic study. JACC 1986;6:1263-1271. 14. Takenaka K, Dabestani A, Gardin JM, Russell D, Clark S, Allfie A, Henry WL. Pulsed Doppler echocardiographic study of left ventricular filling in dilated cardiomyopathy. Am J Cardiol 1986;58:143-147. 15. Fujii J,Yazaki Y, Sawada H, Aizawa T, Watanabe H, Kato K. Noninvasive assessmentof left and right ventricular filling in myocardial infarction with a two-dimensional Doppler echocardiographic method. JACC 1985;5:11551160.

16. Wind BE, Snider AR, Buela AJ, O’Neill WW, Top01 EJ, Dilworth LR. Pulsed Doppler assessment of left ventricular diastolic filling in coronary artery disease before and immediately after coronary angioplasty. Am J Cardial 1987;59:1040-1046. 17. Rassi A Jr, Richards K, Miller J,Crawford M. Echo/Doppler assessmentof mitral blood flow during supine exercise (abstr). Circulation 1986;74:11-47. 18. Edwards WD. Tajik AJ, Seward JB. Standardized nomenclature and anatomic basis for regional tomographic analysis of the heart. Mayo CIin Proc 1981;56:479-497. 19. Huntsman LL, Stewart DK, Barnes SR, Franklin FB, Colocousis BR, Hesel EA. Noninvasive Doppler determination of cardiac output in mon. Clinical validation. Circulation 1983;67:593-601. 20. Lewis JF, Kuo LC, Nelson JG, Limacher MC, Quinones MA. Pulsed Doppler echocardiographic determination of stroke volume and cardiac output: clinical validation of two new methods using the apical window. Circulation i984;70:425-431.

21. Mehdirad AA, Williams GA, Labovitz AJ, Bryg RJ, Chaitman BR. Evaluation of left ventricular function’during upright exercise: correlation of exercise Doppler with post exercise two-dimensional echocardiographic results. Circulation 1987;75:413-419. 22. Gardin JM, Kozlowski J, Dabestani A, Murphy M. Kusnick C. Allfie A, Russell D, Henry WL. Studies of Doppler aortic flow velocjty during supine bicycle exercise. Am J Cardiol 1986;57:327-332. 23. Bryg RJ, Labovitz AJ, Mehdirad AA, Williams GA, Chaitman BR. Effect of coronary artery disease on Doppler-derived parameters of aortic flow during upright exercise. Am J Cardiol 1986;58:14-19. 24. Miyatake K, Okamoto M, Kinoshita N, Owa M, Nakasone I. Sakakibara H, Nimura Y. Augmentation of atria1 contribution to left ventricular inflow with aging as assessedby intracardiac Doppler flowmetry. Am J CardioI1984;53: 586-589.

25. Kuo LC, Quinones MA, Rokey R, Sartori M, Abinader EG, Zoghbi WA. Quantification of atria1 contribqtion to left ventricular filling by pulsed Doppler echocardionraphv and the effect of age in normal and diseased hearts. Am J Cardiol 1987>9%?‘4-1178. .. 1 26. Sartori MP, Quinones MA, Kuo LC. Relation of Doppler-derived left ventricular fiiling parameters to age and radius/thickness ratio in normal and pathoIogic states. Am J Cardiol 1987:59:1179-1182.