Exercise response assessed by continuous monitoring of ventricular function in patients with coronary bypass operations

Exercise response assessed by continuous monitoring of ventricular function in patients with coronary bypass operations The response of left ventricular function during exercise and recovery after exercise was assessed in 35 patients with coronary artery bypass grafting before and after the operation by means of a continuous ventricular function monitor, which records serial beat-to-beat radionuclide data and calculates left ventricular ejection fractions every 20 seconds. The mean ejection fraction decreased with graded bicycle exercise from 48 % ± 9% to 41 % ± 11 % (p < 0.001) before operation but increased with exercise from 50% ± 9% to 55% ± 11 % (p < 0.001) after operation. Cardiac response was divided into four types with respect to the profiles of the ejection fractions during exercise. Type A continued to increase; type B initially increased but then decreased in late exercise stages; type C did not change significantly; type D continued to decrease. Most patients had type C or D responses before operation but type A after operation. Seven patients with occluded grafts or ungrafted coronary arteries had type B or D responses. Three patients with complete revascularization, including an internal thoracic artery and saphenous vein grafts, had type B responses. Three patients with extensive infarction and poor left ventricular function showed type C. In the early recovery period after exercise, most patients had an "overshoot" elevation of ejection fraction. The mean value increased from 59% ± 10% before operation to .64% ± 11 % after operation (p < 0.01). The recovery time after exercise was reduced from 2.8 minutes before operation to 1.8 minutes after operation (p < 0.001). The continuous ventricular function monitor elucidated changes in left ventricular function both during exercise and recovery after exercise, as well as unmasking abnormalities in left ventricular function after coronary bypass operation. (J THoRAe CARDIOVASC SURG 1992;103:849-54)

Michio Kawasuji, MD, Hirobumi Takemura, MD, Takeo Tedoriya, MD, Shigeharu Sawa, MD, Junichi Taki, MD, and Takashi Iwa, MD, Kanazawa, Japan

Radionuclide angiocardiography has been used frequently to define left ventricular function.l' It is well established that coronary artery bypass grafting (CABG) can reverse exercise-induced wall motion abnormalities and improve ventricular function during exercise.v? Equilibrium radionuclide angiocardiography requires 2 to 3 minutes for acquisition of data and permits only From the Department of Surgery (I) and the Department of Nuclear Medicine, Kanazawa University School of Medicine, Kanazawa, Japan. Received for publication March 28, 1990. Accepted for publication Dec. 31, 1990. Address for reprints: Michio Kawasuji, MD, Department of Surgery (I), Kanazawa University School of Medicine, Takaramachi 13-1, Kanazawa 920, Japan.

12/1/28348

intermittent measurements of left ventricular function at rest, at maximal exercise, and additionally at the graded exercise stage. In 1979 Strauss and associates" reported a wearable instrument, consisting of a radionuclide detector/recorder and an electrocardiographic monitor capable of continuously showing cardiac function. The ambulatory ventricular function monitor enabled us to assess sequential changes of left ventricular function during exercise and recovery from exercise, which elucidated the mechanism of the exercise response in patients with coronary artery disease." Continuous monitoring ofventricular function during exercise and recovery after exercise may unmask abnormalities in ventricular function but has not been studied in patients with CABG. The purpose of this study was to investigate, by means of a continuous ventricular function monitor, the response of left ventricular function during exercise and recovery 849

850

The Journal of Thoracic and Cardiovascular Surgery

Kawasuji et al.

1.0 E 2

2.0 E 2

2.0 E 2

2.0 E 2

EF

HR

EDV

ESV

II

0

50

100

o

o

•

A

100

100

o +--+----l.-,--'---\-----'--.--.,.-----,r---..... -'-..,..-"-+--'--+ 0 30:00 0:00

Fig. 1. Changes of left ventricular function at rest, during exercise, and during recovery after exercise before operation in a patient with triple-vessel disease. EF, Ejection fraction; HR, heart rate; EDV, relative end-diastolic volume; ESV, relative end-systolic volume. after exercise in patients with coronary artery disease and to examine the effect of CABG on cardiac response to exercise.

Patients and methods This study was conducted on 35 patients who had undergone elective CABG at Kanazawa University Medical Center. The ages of the patients ranged from 35 to 69, with a mean age of 59 years. Coronary arteries with a 70% or greater reduction in luminal diameter were considered significantly obstructed. Two of the patients had single-vesselcoronary disease, nine had double-vessel disease, 18 had triple-vessel disease, and six had disease of the left main coronary artery. Twenty-three patients had a history of remote myocardial infarction documented by electrocardiographic changes, enzymatic changes, or both. Eleven patients had inferior infarction, 10 patients, anterior infarction, and two patients, inferolateral infarction. CABG was conducted with myocardial preservation accomplished by the administration of a cold crystalloid potassium cardioplegic solution and topical cooling with saline slush. Thirty-one patients received an internal thoracic artery graft to the left anterior descending coronary artery and saphenous vein grafts to the right and circumflex coronary artery systems. Four patients received only saphenous vein grafts. Each patient received an average of 2.6 grafts. There were no instances of perioperative myocardial infarction or significant complications in any of these patients. Coronary angiography or digital subtraction angiography was performed in all of the patients I month after operation. Seven of the patients had occluded grafts, including four grafts to the right coronary artery and three grafts to the left circumflex artery. Left ventricular function was evaluated by the radionuclide angiocardiographic method before and I month after operation. Each patient received the same medications, such as nitrates, calcium-channel blockers, and l3-adrenergic blockade, during the preoperative and postoperative studies. At first, multiple-

gated equilibrium blood pool imaging was performed after the equilibra tion of 20 mCi technetium 99m-Iabeled autologous red blood cells in the intravascular space. Patients were examined by means of an Anger camera (GE Medical Systems Inc., Milwaukee, Wis.) equipped with a dedicated computer in the 30to 40-degree left anterior oblique projection and 35-degree caudal angulation.f After completion of conventional radionuclide angiocardiography, patients wore a continuous ventricular function monitor, (RRG-607, Aloka Incorporated, Tokyo, Japan): This system consists of two radionuclide detectors, recorders, and a computer. The radiation detectors consist of cadmium telluride crystals placed behind straight-bore collimators." One radionuelide detector was used to monitor activity from the left ventricle. The position of the left ventricular detector was confirmed by acquiring static images with a gamma camera. The second detector was used to monitor activity in the lung. After initial amplification, the output from the left ventricle and lung background detectors was subject to threshold analysis, the composite signals summated for 50 msec and recorded in the computer memory. The subject was fitted with an elastic, vestlike garment to hold the detectors in place over the left ventricle and the lung. A resting control study for 4 minutes was performed in all patients, followed by a graded exercise protocol that used a supine bicycle ergometer. The work load began at 25 watts and increased by 25-watt increments at 2-minute intervals until fatigue, dyspnea, or chest pain occurred. Data acquisition continued until 10 minutes after exercise. A modified V 5 electrocardiogram was used for continuous electrocardiographic monitoring. The changes of loading were marked on the tape, to correlate exercise grades with left ventricular response. Blood pressures were recorded by cuff sphygmomanometry. Ejection fraction was calculated from the stroke counts divided by the scatter-corrected end-diastolic counts. A scatter correction of 70% of the end-diastolic counts produced ejection fraction values by the continuous ventricular function monitor that correlated well with those measured by the gamma cam-

Volume 103 Number 5 May 1992

Continuous ventricular function monitor

Table I. Changes of left ventricular function at rest, at maximal exercise, and during recovery after exercise, before and after CABG Factor

Before CABG

Heart rate (beats/min) Rest 70 ± 7 ] t Exercise 109 ± 19 Pressure-rate product (x I 02 mm Hg . beats/min) Rest 76±16J Exercise 159 ± 37 :j: Relative end-diastolic volume (ml) Rest 109 ± 21] ] Exercise 113 ± 22 t Recovery' 109 ± 21 Relative end-systolic volume (ml) Rest 59 ± 21 Exercise 68 ± 21 :j: Recovery' 47 ± 19 Ejection fraction (%) Rest 48 ± 9 ] J~ Exercise 41 ± II t Recovery' 59 ± 10

J ]

85I

EF(O/O)

®

After CABG 80 ± 9 ] 118 ± 17

82 ± 14] 185 ± 53

:j:

:j:

104 ± 20l NS 106 ± 21 103 ± 19

:j:

52 ± 21 J~ 50 ± 21 39 ± 19

:j:

50 ± 9 ] 55 ± 11 64 ± 11

o -j----,--J'---+---L-,-'----+---.-.,.....L--.-~~+_4

© 50

CABG, Coronary artery bypass grafting; NS, not significant. 'The point of peak ejection fraction during recovery after exercise. tp
era.' Relative end-diastolic and end-systolic volumes werecal-

culated assuming that strokevolume was 50 ml. The radionuelide andelectrocardiographic data were summed for20-second intervals tocalculate ejection fraction, relative end-diastolic and end-systolic volumes, and heart rate, which were displayed graphically foranalysis (Fig. I). Changes in respiratory pattern did not affect the data by the continuous ventricular function monitor, because it provided averaged data for a 20-second interval. Changes in ejection fraction of 5% or morewereconsidered significant. Cumulative data wereexpressed as mean ± standard deviation. Continuous variables were analyzed by Student's t test to detectsignificant (p < 0.05)differences between the measured variables. Results

Table I shows the summarized data of left ventricular function at rest, during exercise, and during recovery after exercise for all patients. Preoperative ventricular function. Before CABG, patients exercised to a mean of 71 watts, reaching a heart rate of 109 beats/min and rate-pressure product of 159 X 102 mm Hg . beats/min. Both relative end-diastolicand end-systolicvolumes increased significantly with exercise (p < 0.01; P < 0.001). The mean ejection fraction decreased with exercise from 48% ± 9% to 41% ± II % (p < 0.001), while the heart rate continued to increase throughout exercise. The profiles of ejection

50

24 aomln Fig. 2. Profiles of left ventricular ejection fraction during exercise. Four types (A, B, C, and D) are shown. EF, Ejection fraction; Ex, exercise. 6

12

18

fraction during exercise were divided into four types (Fig. 2). In four patients, ejection fraction continued to increase until maximal exercise (type A). In four patients, ejection fraction initially increased and then decreased in late exercise stages (type B). In three patients with type B responses, ejection fraction at maximal exercise was below the preexercise baseline. In eight patients, ejection fraction did not change significantly during exercise (type C). In 19 patients, ejection fraction decreased throughout exercise (type D). Chest pain developed during exercise in IS patients. Thirteen of them showed a decrease in both ejection fraction and ST segment depression in the electrocardiogram during exercise. In five of them, decrease in ejection fraction began earlier than the onset of chest pain, while in eight patients they started at the same time. In the other two patients in whom chest pain developed,

The Journal of Thoracic and Cardiovascular Surgery

8 5 2 Kawasuji et al.

before

CABG

after

CABG

Fig. 3. A-D,Changesin the types of profiles of leftventricular ejection fraction during exercise beforeand after CABG. ejection fraction did not change. In nine patients ejection fraction decreased without chest pain, and five of them showed ST segment depression in the electrocardiogram. In the early recovery period after exercise, the mean relative end-diastolic volume decreased to the mean baseline value at rest. The mean relative end-systolic volume decreased significantly (p < 0.001), and the mean ejection fraction increased significantly (p < 0.001). An "overshoot" elevation of ejection fraction above preexercise baseline was seen during early recovery in 32 patients. In three patients with type D responses, ejection fraction rose gradually but not above baseline values (Fig. 2, D). The mean value of peak ejection fraction during recovery after exercise was 59% ± 10%.The time interval between the end of exercise and the point of maximal ejection fraction during recovery was 2.8 minutes. In the late recovery period after exercise, ejection fraction returned gradually to the preexercise baseline value. Postoperative ventricular function. After CABG, patients exercised to a mean of73 watts, reaching a heart rate of 118 beats/min and rate-pressure product of 185 X 1<>2 mm Hg . beats/min. Relative end-diastolic volume increased and relative end-systolic volume decreased with exercise, but neither change reached statistical significance. The mean ejection fraction increased with exercise from 50% ± 9% to 55% ± II % (p < 0.001). Twenty-one patients showed a type A exercise response, seven showed type B, four showed type C, and three showed type Dafter CABG (Fig. 3). In 10

patients who showed type B or D responses after operation, ejection fraction began to decrease at a mean work load of 68 ± II watts. Chest pain did not develop in any of them, and only two showed ST segment depression in the electrocardiogram. Four patients with type A responses, three with type B, six with type C, and eight with type D before CABG had type A responses after CABG. One patient with a type B response and six patients with type D responses before operation showed type B responses after operation, but ejection fractions at maximal exercise were not below the baseline in any of the patients with type B responses after operation. Two patients who had had preoperative ejection fractions less than 35%, with a history of extensive infarction, showed type C responses both before and after CABG. In three patients having type D responses after operation, ejection fractions at maximal exercise were above preoperative levels. In the early recovery period after exercise, the mean relative end-diastolic volume decreased to the baseline value at rest. The mean relative end-systolic volume decreased significantly (p < 0.001), and the mean ejection fraction increased significantly (p < 0.001). An "overshoot" elevation bf ejection fraction above baseline values was seen during early recovery in all patients. The mean value of peak ejection fraction during recovery after exercise was 64% ± II %, which was significantly greater than the preoperative peak value during recovery after exercise (p < 0.01). The time interval between the cessation of exercise and the point of peak ejectionfraction during recovery was 1.8 minutes. This was significantly shorter than that before operation (p < 0.001). In the late recovery period after exercise, ejection fractions returned gradually to preexercise baseline values.

Discussion The continuous ventricular function monitor is anextension of a device concept "nuclear stethoscope" described by Wagner and associates''' in 1976 for repetitive measurement of left ventricular function at the bedside. Measurements of left ventricular ejection fraction calculated by the continuous ventricular function monitor have been shown to correlate well with gamma camera measurements of left ventricular ejection fraction.?: II, 12 Continuous measurements of left ventricular function elucidate the cardiac response to exercise and unmask abnormalities during exercise and recovery after exercise in patients with coronary artery disease. The result of the present study elucidated changes in left ventricular function during exercise and recovery after exercise in patients with CABG and demonstrated the effect of CABG on left ventricular function. In a study of normal subjects.P continuous monitoring of left ven-

Volume 103 Number 5 May 1992

tricular function during exercise showed an increase in end-diastolic count and decrease in end-systolic count, with a resultant increase in stroke count and an increase in ejection fraction during exercise. In the present study, continuous ventricular function monitoring before CABG showed an increase in end-diastolic count and an increase in end-systolic count, with a resultant increase in stroke count but with a decrease in ejection fraction during exercise. These findings were consistent with the report of exercise response in patients with coronary artery disease. 7 The result of the present study showed four types of exercise response with respect to the profile of left ventricular ejection fraction during exercise in patients with CABG. Type A was considered to be the normal response of the left ventricular ejection fraction during exercise. Four patients (11%) showed type A responses before CABG and 21 patients (60%) after CABG. Three patients with type A responses postoperatively had occluded grafts to the right coronary artery. Type B, in which ejection fraction increased initially and then decreased in the late exercise stages, was considered abnormal but is also considered to contain a normal response to exercise. Tamaki and associates'! reported that end-diastolic count increased at peak exercise, which caused a slight decrease in ejection fraction in some normal subjects. However, ejection fraction below the baseline at maximal exercise is considered abnormal. In the present study, three patients with preoperative type B responses had ejection fractions below the baseline at maximal exercise, but none of seven with postoperative type B responses showed such decrease in ejection fraction. Three patients with postoperative type B responses had occluded grafts to the left circumflex artery, and one had an ungraftable left circumflex coronary artery. Three other patients with type B responses postoperatively had complete revascularization, with a patent internal thoracic artery graft to the left anterior descending coronary artery and patent saphenous vein grafts to the right and left circumflex coronary artery systems. Multiple gated equilibrium blood pool imaging showed an exerciseinduced decrease in the anteroseptal ejection fraction below resting values. It was suspected that flow of the internal thoracic artery graft to the left anterior descending artery was adequate at moderate exercise but inadequate at maximal exercise.' Types C and D are considered to be abnormal ejection fraction responses during exercise. Three patients having type C responses after operation all had patent grafts but had had poor ventricular function preoperatively with extensive myocardial infarction and possibly little viable myocardium. Two patients with postoperative typeD responses had ungraft-

Continuous ventricular function monitor

853

ed left circumflex coronary arteries, and another had an occluded graft to the dominant right coronary artery. Five patients who had residual lesions in the left circumflex coronary arteries showed type B or D responses after operation. These data suggest that ischemia of the left circumflex coronary system leads to left ventricular dysfunction during exercise. The entire profiles of left ventricular ejection fraction during exercise are important because a possible rise during exercise could be missed or a rise during submaximal exercise could be mistaken for a normal ejection fraction response. Taki and associates'? reported that left ventricular dysfunction manifested by a decrease in ejection fraction is an earlier indicator of myocardial ischemia than angina or electrocardiographic evidence of ischemia. It is important to detect a fall in ejection fraction at late exercise stages. In our postoperative study, 10 patients showed a decrease in ejection fraction at late exercise stages but only two of them showed ST segment depression and no patients had chest pain during exercise. The continuous ventricular function monitor will help to identify silent ischemia and define appropriate exercise limits for patients with residual coronary artery lesions. Continuous ventricular function monitoring also demonstrated changes in left ventricular function during recovery after exercise. In the early recovery period, a decrease in both end-diastolic and end-systolic volume was observed, but the decrease in end-systolic count was greater, with a resultant further increase in ejection fraction. Pfisterer and associates'" described this "overshoot" elevation of ejection fraction above resting levels after exercise in most patients with coronary artery disease. Plotnick, Becker, and Fisher!" postulated that decreased afterload, along with continuing sympathetic tones, allowed ejection fraction to increase in the early recovery period after exercise. Schneider and associates 17 analyzed the rate of recovery of left ventricular function with respect to duration and degree of exercise-induced myocardial ischemia and noted that delayed functional recovery was associated with extensive exercise-induced regional asynergy as a result of severe coronary artery disease. In the present study the recovery time from exercise that is needed became significantly shorter, and an "overshoot" elevation of ejection fraction was higher after CABG. These results also indicate improvement of exercise-induced ischemia after CABG. The continuous ventricular function monitor provides the means for noninvasively assessing pathophysiologic changes in left ventricular function in response to exercise. It also provides a new aspect of assessing the effects of CABG. CABG is generally recognized as reversing exercise-induced wall motion abnormalities and improving

The Journal of Thoracic and Cardiovascular Surgery

8 5 4 Kawasuji et al.

ventricular function during exercise.v ' Coronary artery disease initially results in an alteration of segmental ventricular function. The information provided by the continuous ventricular function monitor would be even more sensitive and possibly more specific for coronary artery disease if segmental function were also analyzed.P REFERENCES 1. Berger HJ, Reduto LA, Johnstone DE, et al. Global and regional left ventricular response to bicycle exercise in coronary artery disease: assessment by quantitative radionuelide angiocardiography. Am J Med 1979;66:13-21. 2. Jones RH, McEwan P, Newman GE, et al. Accuracy of diagnosis of coronary artery disease by radionuclide measurement of left ventricular function during rest and exercise. Circulation 1981;64:586-601. 3. Borer JS, Kent KM, Bacharach SL, et al. Sensitivity, specificity and predictive accuracy of radionuclide cineangiography during exercise in patients with coronary artery disease: comparison with exercise electrocardiography. Circulation 1979;60:572-80. 4. Lim YL, Kalff V, Kelly MJ, et al. Radionuclide angiographic assessment of global and segmental left ventricular function at rest and during exercise after coronary bypass graft surgery. Circulation 1982;66:972-9. 5. Taylor NC, Barber RW, Crossland P, English TAH, Petch MC. Effects of coronary artery bypass grafting on left ventricular function by multiple gated ventricular scintigraphy. Br Heart J 1983;50:149-56. 6. Strauss HW, Lazewatsky J, Moore RH, et al. The VEST: a device for the continuous monitoring of cardiac function in ambulatory patients [Abstract]. Circulation 1979;59(Pt 2):11246. 7. Tamaki N, Yasuda T, Moore RH, et al. Continuous monitoring of left ventricular function by an ambulatory radionuclide detector in patients with coronary artery disease. J Am Coli CardioI1988;12:669-79. 8. Kawasuji M, Tsujiguchi H, Tedoriya T, Taki J, Iwa T.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

Evaluation of postoperative flowcapacity of internal mammary artery. J THORAC CARDIOVASC SURG 1990;99:696702. Ide M, Kanemoto N, Goto Y, Suzuki Y. Evaluation ofleft ventricular function using a cadmium telluride probe. Jpn Nucl Med 1986;23:1675-82. Wagner HJ Jr, Wake R, NickoloffE, Natarajan TK. The nuclear stethoscope: a simple device for the generation of left ventricular volume curves. Am J CardioI1976;38:74750. Wilson RA, Sullivan PJ, Moore RH, et al. An ambulatory ventricular monitor: validation and preliminary clinical results. Am J Cardiol 1983;52:601-6. Harrison KS, Liu X, Han ST, et al. Evaluation of a miniature CdTe detector for monitoring left ventricular function. Eur J Nucl Med 1982;7:204-6. Tamaki N, Gill JB, Moore RH, Yasuda T, Boucher CA, Strauss HW. Cardiac response to daily activities and exercise in normal subjects assessed by an ambulatory ventricular function monitor. Am J Cardiol 1987;59:1164-9. Taki J, Yasuda T, Tamaki N, et al. Temporal relationship between left ventricular dysfunction and chest pain in patients with coronary artery disease during activities of daily living. Am J Cardiol 1990;66:1455-8. Pfisterer ME, Slutsky RA, Schuler G, et al. Profiles of radionuclide left ventricular ejection fraction changes induced by supine bicycle exercise in normals and patients with coronary heart disease. Cathet Cardiovasc Diagn 1979;5:305-17. Plotnick GO, Becker LC, Fisher ML. Changes in left ventricular recovery from upright bicycle exercise in normal persons and patients with coronary artery disease. Am J CardioI1986;58:247-51. Schneider RM, Weintraub WS, Klein LW, et al. Rate of left ventricular functional recovery by radionuclide angiography after exercise in coronary artery disease. Am J Cardiol 1986;57:927-32. Willerson JT. Ambulatory monitoring of left ventricular function. J Am Coli Cardiol 1988;12:680-1.