pathophysiologic factors governing the variability of ischemic responses to treadmill and bicycle...

Pathophysiologic factors governing the variability of ischemic responses to treadmill and bicycle exercise

lschemic responses may vary considerably when patients with coronary artery disease (CAD) are tested serially, but the pathophysiologic mechanisms that govern this variability have not been well evaluated. We thus evaluated whether clinical, hemodynamic, physiologic, and anatomic factors influenced the variability in ischemic responses among 140 patients (mean age 54 + 11 years) subjected to both bicycle and treadmill exercise electrocardiography. Radionuclide ventriculography was obtained during bicycle exercise in each patient. The population included 77 patients with CAD, 21 patients with normal coronary arteriograms, and 42 patients with ~5% likelihood of CAD. Bicycle exercise evoked higher systolic blood pressure @ < 0.001) and double-product (p < 0.001) responses compared with treadmill exercise in the patients with CAD and in the normal subjects, and it evoked a lower frequency of chest pain (12% vs 41%, p < 0.001) in the 34 patients with CAD who had ST-segment depression during both exercise tests. There was a high frequency of variability in ischemic responses during treadmill versus bicycle exercise: 22 (39%) of the 56 CAD patients who had exercise-induced ST-segment depression manifested this response during one stress test only. This variability was strongly related to the functional and anatomic magnitude of disease. lschemic variability decreased progressively as the response of left ventricular ejection fraction (LVEF) to exercise worsened progressively (p - 0.003 by analysis of variance), from 63% in those with an LVEF increase of >lO% with exercise to only 13% in those with an LVEF fall of ~5% with exercise. Similarly, ischemic variability occurred in 6 (69%) of 9 patients with single-vessel CAD versus 14 (30%) of 47 patients with multivessel CAD @ < 0.005). In conclusion, bicycle and treadmill exercise are different stressors, evoking different hemodynamic and clinical responses in patients with CAD. lschemic ECG responses vary considerably when these patients undergo both stresses. This variability is governed by the functional and anatomic magnitude of ischemic heart disease. Variability in ischemic responses is reduced in the presence of multivessel coronary disease and in patients with abnormal LVEF responses to exercise. (AM HEART J 1994;128:946-55.)

Jacob Klein, MD,” Susan Cheo, MS,B Daniel S. Berman, MD,” and Alan Rozanski, MDb Los Angeles, Calif., and New York, N. Y.

A variety of physiologic “triggers” may stimulate the induction of myocardial ischemia in patients with coronary artery disease (CAD). Besides exercise, mental stress,1-4 cold stimulation,5 and smoking6 all

From the *Division of Cardiology and Department of Nuclear Medicine, Cedars-Sinai Medical Center, Los Angeles; the bDivision of Cardiology, De- partment of Medicine, St. Luke’s+Roosevelt Hospital Center, New York; and the Department of Medicine, Columbia University College of Physi- cians and Surgeons, New York.

Supported in part by the John D. and Catherine T. MacArthur Foundation and by the KROC Foundation. J.K. was formerly the Save-A-Heart Fellow in Preventive Cardiology. Received for publication Dec. 3, 1993; accepted Jan. 16, 1994.

Reprint requests: Alan Rozanski, MD, Section of Nuclear Cardiology and Cardiac Stress Testing, St. Luke’s-Roosevelt Medical Center, 114th Street and Amsterdam Avenue, New York, NY 10025. Copyright Q 1994 by Moaby-Year Book, Inc. 0002~8703/94/$3.00 + 0 4/I/5705%

have been shown to induce myocardial ischemia in the laboratory. However, the reproducibility of ischemic induction by these nonphysical stressors in the laboratory has not been evaluated. Ischemia during ambulatory electrocardiography (ECG) monitoring, which can occur in association with a wide variety of physical and mental triggers,7-10 has been shown to be highly variable during serial monitoring.rl So far, however, the underlying causes of ischemic variability-by any type of trigger, including exercise-have not been well studied. Under- standing the causes of ischemic variability could be important because both exercise-induced and ambu- lant ischemia are of prognostic significance in patients with CAD,12-i6 and elucidation of the causes of the variability could improve our understanding of pathophysiologic basis of CAD.

948

Volume 12%. Number 5 American Heart Journal Klein et al. 949

To investigate this issue, we evaluated ECG results in patients undergoing both treadmill and bicycle exercise. The former test was performed alone or with myocardial perfusion scintigraphy, and the latter test was performed with exercise radionuclide ventriculography. Both treadmill and bicycle exercise are physical stressors, and documented differences in heart rate, blood pressure, and ischemic ECG re- sponses17-26 elicited by these two stressors make their comparison useful for assessing ischemic variability. We examined hemodynamic, clinical, functional (radionuclide ventriculographic), and angiographic factors associated with ischemic variability during serial treadmill and bicycle stress.

METHODS Patient population. We retrospectively analyzed ECG

results in 140 patients with possible or known CAD who underwent treadmill end upright bicycle exercise in sequential order at least 1 day apart. The latter test was performed with exercise radionuclide ventriculography. Pa- tients were excluded from analysis if their resting ECG was uninterpretable for ST-segment response to exercise (i.e., left bundle branch block, intraventricular conduction dis- turbance, left ventricular hypertrophy, or significant resting ST-segment abnormalities), if they were receiving digoxin, if they had any significant mitral or aortic valvu- lar heart disease or known cardiomyopathy, if the period between the two tests exceeded 3 weeks, or if there were any changes in clinical status during the 3-week period. The population of 140 patients was divided into 77 patients who had angiographically significant CAD (~50% stenosis in one or more coronary arteries) (mean age 57 f 9 years, 83% men), 21 patients who had normal coronary arteries (<50 % stenosis in all coronary arteries) (mean age 51 * 12 years, 48% men), and 42 patients with <5% pretest likelihood of CAD (mean age 48 ? 11 years, 86% men), CAD likelihood was based on CADENZA, a computer-based Bayesian program that incorporates patient age, gender, symptom classification, and Framingham risk factors as previously described.27 Thirty-two (42 %) of the 77 patients with CAD had a history of myocardial infarction. All patients were requested to discontinue P-blocking medica- tions for 48 hours, calcium-channel blockers for 24 hours, and nitrates for 6 hours before stress testing.

Bicycle and treadmill exercise protocol. Patients were requested to exercise maximally for both stress tests. Ex- ercise was terminated prematurely by the physician for severe chest pain, complex ventricular arrhythmias, or the development of exertional hypertension or hypotension. Treadmill exercise was performed by graded stress according to the standard Bruce protocol. Bicycle exercise was performed in the upright position, also by a graded-stress protocol. Patients began bicycling against a work load of 200 kilopond meters (kpm)/min, with the work load increased by 200 kpm/min every 3 minutes of exercise.26 For both tests, the exercise ECG was monitored continuously

in leads aVF, Vi, and Vs. Twelve-lead ECG was recorded at rest, at the end of each exercise stage, at peak exercise, and during each minute after exercise. Blood pressure was recorded by cuff sphygmomanometer at rest, during each stage of exercise, at peak exercise, and at Z-minute inter- vals after exercise. Patients were permitted to hold the treadmill handrails or bicycle handlebars during the tests. The ECG response to exercise was considered ischemic if horizontal or downsloping ST-segment depression ~1 mm or upsloping ST-segment depression 21.5 mm compared with the baseline PR segment developed and persisted for 80 msec after the J-point in at least three consecutive beats. In patients with ischemic responses to exercise, the magnitude of depression, heart rate at the onset of ischemia, and the postexercise duration of the ST-segment depression were recorded.

Bicycle exercise radionuclide ventriculography. After an intravenous injection of 25 mCi of autologous red blood cells labeled in vitro with technetium-99m, the patients were positioned on an upright, constant-load bicycle ergometer calibrated in kiloponds. A mobile gamma camera equipped with a 0.25-inch sodium iodide crystal and an all-purpose collimator was then positioned in a 40- to 45- degree left anterior oblique (LAO) projection; the degree of obliquity was determined by the position providing the best separation of the left and right ventricular blood pools. R-wave-synchronized, multiple-gated equilibrium cardiac blood pool scintigraphy was performed at rest and during exercise by acquisition of 20 frames distributed uniformly over the entire R-R interval. Scintigraphic images were ac- quired over a P-minute period, yielding approximately 100,000 counts per frame, and were collected with a 25% energy window centered on the 140 keV photopeak of technetium-99m. Scintigraphy was performed during the last 2 minutes of each 3-minute stage of exercise and during the first 2 minutes after exercise while the patients re- mained upright on the bicycle.28

Left ventricular ejection fraction (LVEF) measurements were determined by a computer operator, blind to study conditions, who used a light pen to designate end-diastolic, end-systolic, and left paraventricular background regions of interest. LVEF was calculated as stroke counts divided by background-corrected end-diastolic counts. Left ventricular wall motion was evaluated independently by two experienced physicians unaware of the clinical or test data. The rest images and all exercise images of each patient’s study were displayed simultaneously in a closed-loop for- mat on the computer’s video display. Before segmental wall motion analysis, images were processed by spatial and temporal smoothing. Each of the five 45-degree LAO segments and five anterior view segments was scored according to a five-point system: 3 = normal wall motion, 2 = mild hypokinesis, 1 = severe hypokinesis, 0 = akinesis, and -1 = dyskinesis.2g Exercise wall motion was considered abnormal if at least one left ventricular segment worsened by one or more points.

Coronary arteriography. Selective tine coronary arteriography was performed in multiple oblique projections by

950 Klein et al. November1994

AmericanHearlJournal

n = 77

rc = 0.46 p < 0.001

60 EC tee 126 140 160 160 260 60 ICO 120 f40 160 160 200 220 240

Treadmill Peak HR Treadmill Peak SBP

TM=67 j 12 i%=96jl5 r, = 0.37 p * 0.001 n = 62

%i = 6.4 f 3.4 m = 10.3 l 3.5

rc = 0.60 p < 0.001 n = 77

40 60 60 loo 120 140 0 2 6 0 12 16 16

Treadmill Peak DBP Treadmill Duration of Exercise

Fig. 1. Correlation of treadmill (TM) and bicycle (Bike, BK) values for peak heart rate (HR) (top left), peak systolic blood pressure (SBP) (top right), peak diastolic blood pressure (DBP) (bottom left), and exercise duration (bottom right) for 77 patients with coronary artery disease. Treadmill exercise values are on x axis for each group and bicycle exercise values are on y axis. Line of identity is shown for each graph, and mean measurements for each variable during treadmill and bike exercise, r,, Lin’s concordant coefficient and the p value for the mean differences, are listed on the bottom right corner of each graph. Most individual peak SBP and DBP measurements during bicycle exercise are to left of line of identity, indi- cating their higher values during bicycle vs treadmill exercise.

the Judkins technique. The results were interpreted by experienced angiographers. Significant stenosis was considered present when there was >50% narrowing of a major coronary vessel.

Statistical analysis. Comparison of clinical, hemodynamic, ECG, and scintigraphic responses to bicycle versus treadmill exercise was assessed for all patients with CAD and among and within each of four subgroups defined by ECG response (described later). Chi square analysis and Fisher’s exact test were used to compare categorical variables. Quantitative variables were analyzed by analysis of variance and the Kruskal-Wallis test. McNemar’s test and the paired t test were used for paired comparisons. The Lin30 concordance correlation coefficient was used as an index of agreement between bicycle and treadmill exercise. Test results were considered significant at the p < 0.05 level.

RESULTS Hemodynamic responses during bicycle and treadmill

exercise. The hemodynamic responses during bicycle versus treadmill exercise for the patients with CAD are shown in Fig. 1. The mean values for peak heart rate during treadmill and bicycle exercise were very similar. The mean peak systolic blood pressure, however, was significantly higher (by a mean of 25 mm Hg) during bicycle versus treadmill exercise (p < O.OOl), as was the mean peak diastolic pressure (p <O.OOl). Most of the individual peak systolic and diastolic blood pressure measurements were higher during bicycle exercise. The mean peak heart rate was slightly but insignificantly higher during treadmill exercise. Thus the double product was consis- tently greater during bicycle exercise (27,000) than

Vohmo 126, Number 5

American Hearl Journal

during treadmill exercise (23,900) Cp < 0.001). Mean exercise duration for our graded-stress bicycle protocol was longer than for treadmill exercise by a mean of approximately 2 minutes 0, < 0.001). Normal subjects and those with possible CAD also manifested substantially higher peak systolic and diastolic blood pressure responses during bicycle versus treadmill exercise (data not shown).

To assess the frequency of chest pain in association with ischemia, we evaluated chest pain frequency among the 34 patients who had ST-segment depression during both treadmill and bicycle exercise. In these patients, it was possible to compare the association between exercise-induced ST-segment depression during treadmill and bicycle exercise and exercise-induced angina. Exercise-induced chest pain was more than three times more frequent during treadmill exercise than during bicycle exercise in this subgroup (41% vs 12 % , respectively, p < 0.02), de- spite a similar magnitude of exercise-induced ST- segment depression during treadmill and bicycle exercise (2.3 + 0.9 vs 2.0 + 0.0 mm, respectively, p = not significant [NS]) and a similar heart rate threshold for the onset of ST-segment depression (117 + 21 vs 117 + 22 beats/min, respectively, p = NS) in these patients.

Electrocardiographic responses during bicycle and treadmill exercise. The sensitivity of bicycle and treadmill exercise for CAD detection was evaluated first by excluding the 32 patients with previous myocardial infarction. In the remaining 42 patients with CAD, the frequency of an ischemic ECG response to treadmill exercise (69%) was slightly and insigniti- cantly higher than the response bicycle exercise (62 % ). The specificity value for the two tests was the same (57%) among the 21 patients with normal coronary arteriograms. The normalcy rate in the group with a low CAD likelihood, however, was significantly higher during bicycle versus treadmill exercise (98% vs 67 % , respectively, p = 0.001). Fig. 2 shows the frequency of a positive ECG response to exercise according to the magnitude of angiographic CAD. For this analysis, the 32 patients with previous myocardial infarction were included. Overall, the frequency of an ischemic response was 11% higher during treadmill than during bicycle exercise (p = NS), but the difference tended to be larger for patients with single-vessel CAD and smaller for the patients with triple-vessel CAD.

Overall, 56 (73 % ) of the 77 patients with CAD had a positive ECG response to treadmill or bicycle exercise or both. However, the ECG responses to the two stressors were variable, with 22 (39%) of these 56

Klein et al. 951

100

1

60

il - 64

60

%

40

20

n ”

All CAD

(n=77)

1 VD 2VD

(tl=i 3) (n=22)

67

(2, m Treadmill ECG @?i Bicycle ECG

Fig. 2. Frequency of abnormal (ischemia) electrocardiographic (ECG) responses to treadmill and bicycle exercise in 77 patients with coronary artery disease (CAD) documented by angiography and in these patients further divided according to number of diseased coronary vessels 0’0).

patients having an ischemic response during one form of exercise stress but not the other. Fifteen (27 % ) of the 56 patients had a positive ECG response to treadmill exercise only, and 7 patients (13 % ) had a positive response to bicycle exercise only.

Based on the ECG response to treadmill and bicycle exercise, we divided the patients into four subgroups for subsequent analyses: group 1, an ischemic ECG response to both exercise tests (34 patients); group 2, an ischemic ECG response to bicycle exercise only (7 patients); group 3, an ischemic response to treadmill exercise only (15 patients); and group 4, nonischemic responses to both exercise tests (21 patients). The hemodynamic, ECG, and radionuclide ventriculographic responses to exercise in these four groups are summarized in Table I. The comparative heart rate and blood pressure responses to exercise were similar among the four subgroups. Each subgroup manifested a markedly higher peak systolic blood pressure response during bicycle versus treadmill exercise 0, < 0.001 for each subgroup), which resulted in a significantly higher double product during bicycle exercise for each subgroup as well. When the four subgroups were compared with each other, however, there were significant differences in the magnitude of induced ischemic abnormalities.

Compared to the other subgroups, group 1 patients manifested a markedly lower threshold to the onset of ST-segment depression (p = 0.002), a higher frequency of abnormal LVEF responses (p = 0.002), and a more abnormal mean LVEF response to exercise (JJ = 0.0003). Group 1 patients also tended to have a greater frequency and magnitude of exercise-

952 Klein et al. November 1994

American Heart Journal

Table I. Comparison of hemodynamic responses among patients with coronary artery disease

Group 1 Group 2 Group 3 Group 4 +TM/+BK ECG -TM/+BK ECG +TM/-BK ECG -TM/-BK ECG

(n = 34) (n = 7) (n = 15) (n =21) P

Hemodynamic responses Peak HR

TM 136 + 19 BK 139 + 20

Peak SBP TM 168 + 28 BK 188 + 28t

Peak DBP TM 87 -c 12 BK 97 + 17$

Peak DP TM 23.0 k 5.3 BK 26.4 + 6.7$

HR threshold for onset of ST-segment depression TM 117 & 21 BK 118 t 21

Scintigraphy responses Abnormal EX LVEF response (ischemic) 88% (30 Mean change in LVEF with EX -3.7% + 9% >l WMA EX 85% (29)

Summed wall motion worsening score 3.6 k 3.5

151 f 24 153 + 19 144 + 22 146 + 22

175 2 13 209 2 21t

90 + 11 102 -+ 12

26.6 f 5.0 30.3 + 5.5*

NA 136 + 20

71% (n=5) -0.9% + 9% 71% (n=6)

2.7 + 2.0

167 jI 22 193 f 30t

86 t 11 102 -t 21

25.8 -t 5.4 28.5 + 7.4*

147 -c 20 NA

53% (n=8) 5.1% + 8% 53% (n=8)

1.5 + 2.6

145 k 23 137 + 24*

157 + 29 184 f 27f

85 + 14 96 _+ lO*

23.0 k 6.2 25.9 r 6.93

NA NA 0.0002f

43%(n = 9) 0.002 6.3% f 9% 0.003 62% (n = 13) 0.08

2.0 + 3.6 0.13

BK, Bicycle; DBP, diastolic blood pressure; DP, HR X SBP double product x10 3; ECG, electrocardiogram; EX, exercise; HR, heart rate; LVEF, left ventricular ejection fraction; NA, not applicable; SBP, systolic blood pressure; TM, treadmill; WMA, wall motion abnormality. Abnormal EX LVEF response defined as failure of LVEF to increase 0.05 units with EX. *p < 0.01. yp < 0.0001. fp < 0.001. 8By analysis of variance for groups 1 to 3 (with use of bicycle threshold for group 1).

induced wall motion abnormalities. Thus, compared with the other ECG subgroups, group 1 patients manifested more ischemic abnormalities by exercise ECG and radionuclide ventriculography.

Reproducibility of electrocardiographic responses to bicycle and treadmill exercise: Scintigraphic and angiographic correlates. To assess the variability of ECG responses to bicycle and treadmill exercise as it may relate to the functional severity of CAD, we divided the 77 CAD patients according to the magnitude of change in LVEF during bicycle exercise into those whose LVEF increased by 210% (14 patients), increased by 0 % to 9 % (27 patients), decreased from 1% to 4% (17 patients), or decreased by 25% (19 patients). As the LVEF response to exercise become more abnormal, the frequency of exercise test posi- tivity to treadmill or bicycle exercise increased progressively in these four groups (by 43 % , 78%, 76 % , and 84 % , respectively, p = 0.02 by analysis of variance). Furthermore, as indicated in Fig. 3, the frequency of variability in ischemic responses during bicycle versus treadmill exercise diminished substantially as the LVEF response to exercise worsened among the 56 patients who manifested an ischemic

response to bicycle and or treadmill exercise 0, = 0.002). Th us whereas 5 (83%) of the 6 patients with CAD who had a positive exercise ECG and a 110% increase in exercise LVEF manifested an ischemic ECG response on one of the two exercise stress tests but not both, only two (13 % ) of the 16 patients with a positive ECG and ~5% decrease in LVEF during exercise manifested similar variability in ischemic responses.

A similar pattern was noted when test results of the 56 patients with ischemic ECG responses were assessed according to the results of angiography. Whereas the frequency of the variability in ischemic responses during bicycle versus treadmill exercise was very high in patients with single-vessel disease, the frequency of variability was significantly lower in the patients with multivessel CAD (Fig. 4).

DISCUSSION

Studies have demonstrated that bicycle and treadmill exercise are not necessarily the “same” stressors. These studies, based primarily on hemodynamic and ECG comparisons, have demonstrated significant differences between the two forms of exercise in heart

vol~mo 128, Number 5

Amoflcan Howl Journal Klein et al. 953

% of +Ex ECG patients with Re-test variability 100

80

80

40

20

iEF 110%? (n=6)

0 - 9%,t -4 to 0%5 s-5%5 (n=21) (n-13) (n=l6)

p=o.o02

Fig. 3. Seventy-seven patients with coronary artery disease divided into four groups based on the change in left ventricular ejection fraction (EF) from rest to peak exercise Ix axis). y axis, Frequency of patients who had an ischemic exercise electrocardiographic (Ex ECG) response during only one of two exercise tests. As EF response to exercise decreased progressively, variability in ischemic responses during bicycle and treadmill exercise also decreased.

rate, blood pressure, oxygen consumption, or exercise ECG responses.17-26 Our study confirms these marked hemodynamic differences. Most notably, systolic blood pressure responses were substantially higher during bicycle than during treadmill exercise in our study, by a mean of 25 mm Hg. It is possible that a tighter gripping of the handlebars during bicycle exercise, compared with the holding of the handrails during treadmill exercise, may have contributed to this blood pressure difference. Diastolic blood pressure elevations also were significantly greater during bicycle versus treadmill exercise. Similar differences also were noted in normal subjects.

The goal of our study, however, was not to confirm differences between bicycle and treadmill exercise. Rather we sought to use the different responses elicited by these stressors to evaluate the pathophysiologic cause(s) of ischemic variability during exposure to sequential physical stressors,

First, whereas the overall frequency of ST-segment depression was not significantly different between bicycle and treadmill exercise among our patients with CAD, there were considerable differences in ischemic ECG responses among individual patients. Thus, among the 56 patients with a positive ECG response to either exercise test, only 34 (61%) had concordantly positive responses during both stress tests. For the others (39%), results were positive during one stress or the other but not both. Only a few other studies have assessed variability in ECG re-

SINGLE VESSEL MULTIVESSEL DISEASE DISEASE

(rk9) (n=47)

Fig. 4. Frequency of patients who had ischemic exercise electrocardiographic response during only one of two exercise tests (y axis) for patients with single-vessel disease and multivessel disease.

sponses during bicycle and treadmill exercise.23-26 In a comparably sized study, Wetherbee et a1.25 found variability in ischemic ST-segment depression during bicycle and treadmill exercise to be similar. A similar discordance was noted also in 4 (36 %) of 11 patients in a small study by Wicks et aL2* However, these studies did not examine potential factors governing this variability in ischemic responses.

Second, we observed a higher frequency of silent ischemia during bicycle exercise than during treadmill exercise. Fewer than half of patients with ST- segment depression had associated chest pain with treadmill exercise, but among patients with ST-segment depression during both bicycle and treadmill exercise (group l), the frequency of exercise-induced chest pain during treadmill stress was more than threefold higher than during bicycle stress. The cause for this difference is not readily apparent. The heart rate threshold for ST-segment depression and the magnitude of induced ST-segment depression were similar in the group 1 patients, and the mean peak heart rates were nearly identical. Only peak systolic blood pressure responses differed: they were higher during bicycle exercise, as for the total patient population. Further study is needed to evaluate differences in silent ischemia during different exertional stressors.

Third, we observed a strong association between the pattern of ECG response to the two exercise stressors and indices of ischemic magnitude. For ex- ample, group 1 patients (those with ischemic ECG responses to both bicycle and treadmill exercise) had

November 1994

954 Klein et al. American Heart Journal

a significantly lower heart rate threshold for onset of ST-segment depression during exercise compared with that of patients with a variable ischemic ECG response to bicycle and treadmill exercise (groups 2 and 3). Similarly, group 1 patients had the most abnormal LVEF responses to stress, including both more frequent and more abnormal LVEF responses to stress. Conversely, group 4 patients (those with no ST-segment depression during bicycle and treadmill exercise) had the least abnormal radionuclide ventriculographic responses to stress. These results im- ply that patients who manifest reproducible ischemic ST-segment responses to bicycle and treadmill exercise are “functionally sicker” compared with patients in whom ischemic ST-segment responses are not re- produced during serial bicycle and treadmill exercise.

Previous investigations have indicated that patients who manifest ischemia during both exercise and nonexertional ambulatory activity are “functionally sicker” than patients who manifest ischemia during exercise testing alone31: they manifest more hemodynamic and exercise ECG abnormalities and more abnormalities by thallium-201 stress-redistribution single-photon-emission computed tomogra- phy.l” In the laboratory, patients in whom ischemia is induced by a nonexertional stimulus such as mental stress are also functionally sicker compared with patients in whom ischemia is provoked by exercise only.lm4 These observations suggested the following question: could variability in ischemic responses to frank physical stressors also be related to the underlying functional or anatomic magnitude of coronary artery disease? To evaluate this possibility, we divided patients in terms of a marker of functional disease severity: the magnitude of LVEF change to exercise. Patients with the most normal LVEF responses to exercise manifested the greatest variability in exercise ECG responses. For instance, of 6 patients with a positive exercise ECG response and an exercise LVEF increase 110 % ,5 (83 % ) had ST-segment depression during bicycle or treadmill exercise but not during both tests. But as the LVEF response to exercise worsened, the frequency of this variability diminished markedly. Thus, among the 16 patients with ECG ischemia and a frank decrease in exercise LVEF (~5% ), the variability in ischemic responses was only 13 % . Indirect support for this type of functional relation can be provided by the bicycle-treadmill comparison reported by Wetherbee et al.25: 50 % of 45 patients with concordantly positive ECG responses to treadmill and bicycle exercise had New York Heart Association class 3 or 4 angina, whereas

this degree of angina was present in only 14 % of the 21 patients with discordant ECG responses.

Similar results were observed if patients were stratified on the basis of an anatomic standard: among the 56 patients with an ischemic ECG response to exercise, the ischemic response lacked reproducibility in 92% of patients with single-vessel disease, compared with a lack of reproducibility in only 30% of the patients with multivessel disease. Thus the functional and anatomic severity of CAD may be important determinants of the reproducibility in ischemic responses during serial stress.

Conclusions. Compared with treadmill exercise, bicycle exercise produces more hypertensive blood pressure responses to stress and a higher frequency of silent myocardial ischemia. Patients who show reproducible ECG responses to both bicycle and treadmill exercise appear to have greater functional and anatomic disease compared with that of patients with ischemic responses that are not reproducible during serial stress. Further study of the factors that govern this phenomenon may provide new insights into the pathophysiologic mechanisms of myocardial ischemia.

REFERENCES

1. Rozanski A, Bairey CN, Drantz DS, Friedman J, Resser K, Morel1 M, Hilton-Chalfen S, Hestrin L, Bietendorf J, Berman DS. Mental stress and the induction of silent myocardial ischemia in patients with coronary artery disease. New Engl J Med 1988;318:1005-12.

2. LaVeau PJ, Rozanski A, Krantz D, Cornell CE, Cattanech L, Zaret BL, Wockers FJT. Transient left ventricular dysfunction during provoca- tive mental stress in patients with coronary artery disease. AM HEART J 1989;118:1-8.

3. Deanfield JE, Kensett M, Wilson RA, et al. Silent myocardial ischemia due to mental stress. Lancet 1984;2:1001-4.

4. Bairey CN, Krantz DS, DeQuattro V, Berman DS, Rozanski A. Effect of beta-blockade on low heart rate-related ischemia during mental stress. J Am Co11 Cardiol 1991;1388-95.

5. Deanfield JE, Shea MJ, Wilson RA, Horlock P, de Landsheere CM, Selwyn AP. Direct effects of smoking on the heart: silent ischemic dis- turbances of coronary flow. Am J Cardiol 1986; 5’7:1005-9.

6. Shea MJ, Deanfield SE, DeLandstere CM, Wilson RA, Kensett M, Sel- wyn AP. Asymptomatic myocardial ischemia following cold provoca- tion. Aru HJL4RT J 1987;114:469-76.

7. Allen RD, Gettes LS, Phalan C, Avington MD. Painless ST-segment depression in patients with angina pectoris: correlation with daily activities and cigarette smoking. Chest 1976;69:467-73.

8. Schang SJ Jr, Pepine CJ. Transient asymptomatic ST-segment depression during daily activity. Am J Cardioll977; 39396402.

9. Barry J, Selwyn AP, Nable EG, Rocco MB, Meod K. Frequency of ST- segment depression produced by mental stress in stable angina pectoris from coronary disease. Am J Cardiol 1988;61:979-83.

10. Klein J, Gabbay FH, Howell RH, Krantz DS, Hedges SM, Berman DS, Rozanski A. The relationship between myocerdial hypoperfusion and the daily life triggers of myocardial ischemia [Abstract]. J Am Co11 Car- dial 1991;17:331.

v~lme 128, Number 5 American Heart Journal Klein et al. 955

11. Nabel EG, Roceo BJ, Rocco MB, Campbell S, Meod K, Fenton T, Orav ET, Selwyn AP. Variability of transient myocardial ischemia in ambulatory patients with coronary artery disease. Circulation 1988;78:60-7.

12. Stern S, Tzivoni D. Early detection of silent ischemic heart disease by 24-hour electrocardiographic monit,oring of active subjects. Br Heart J 1974;36:481-6.

13. Rocco MB, Nabel EG, Campbell S, Goldman L, Barry 3, Mead K, Sel- wyn AP. Prognostic importance of myocardial ischemia detected by ambulatory monitoring in patients with stable coronary artery disease. Circulation 1988;78:877-84.

14. Deedwania PC, Carabajal EV. Silent ischemia during daily life is an in- dependent predictor of mortality in stable angina. Circulation 1990; 81:148-56.

15. Rozanski A, Berman DS. The efficacy of cardiovascular nuclear medicine studies. Semin Nucl Med 1987;27:104-20.

16. Klein J, Rodriques EA, Berman DS, Prigent F, Chao SY, Maryon T, Rozanski A. The prevalence and functional significance of transient ST-segment depression during daily life activity: with comparison of ambulatory ECG to stress-redistribution thallium-201 SPECT. AM HEART J 1993;125:1247-57.

17. Hermanser L, Saltin B. Oxygen uptake during maximal treadmill and bicycle exercise. J Appl Physiol 1969;26:31-7.

18. McKay GA, Bannister EW. A comparison of maximum oxygen uptake determination by bicycle ergometry at various pedalling frequencies and by treadmill running at various speeds. Em J Appl Physiol 1976; 35:191-200.

19. Niederberger M, Bruce RA, Kusumi F, Whitkanack S. Disparities in ventilatory and circulatory responses to bicycle and treadmill exercise. Br Heart J 1974;36:377-82.

20. Mama C, Bianchi G, Effendy FN, Bolognesi, Cucchini F, Visiolo 0. Comparison of 5 different stress testing methods in the ECG diagnosis of coronary artery disease. Cardiology 1979; 64325-32.

21. Kaltenbach M, Bischofs W, Hopi R, Bohmer D. Physical and physio- logical work in treadmill testing compared with other types of ergometry. Curr Heart J 1982;3:9:%9.

22. Hermansen L, Saltin B. Oxygen uptake during maximal treadmill and bicycle exercise. J Appl Physiol 1969;26:31-‘7.

23. Calvert AF, Pater G, Pye D, Mann J, Chalmers D, Ayres B. A matched pairs comparison of cycle ergometry and treadmill exercise testing in the evaluation of coronary heart disease. Aust N Z J Med 1987$7:472-S.

24. Wicks JR, John MB, Sutton JR, Oldridge NB, Jones NL. Comparison of the electrocardiographic changes induced by maximum exercise testing with treadmill and cycle ergometer. Circulation 1978;57:1066-70.

25. Wetherbee JN, Bamrah VS, Ptacin MJ, Kalbfleish JH. Comparison of ST segment depression in upright treadmill and supine bicycle exercise testing. J Am Co11 Cardiol 1988;11: 330-7.

26. Hambrecht R, Schuler GC, Muth T, Grunza MF, Marburger CT, Nie- bauer J, Methfessel SM, Kubler W. Greater diagnostic sensitivity of treadmill versus cycle exercise testing of asymptomatic men with coronary artery disease. Am J Cardiol 1992;70:141-6.

27. Diamond GA, Forrester JS, Hirsch M. Application of conditional prob- ability analysis to the clinical diagnosis of coronary artery disease. J Clin Invest 1980;65:1210-21.

28. Rozanski A, Berman D, Gray R, Diamond G, Raymond M, Prause J, Maddahi J, Swan HJC, Matloff J. Preoperative prediction of reversible myocardial asynergy by postexercise radionuclide ventriculography. N Engl J Med 1982;307:212-6.

29. Rosanski A, Elkayam U, Berman D, Diamond G, Prause J, Swan HJC. Improvement of resting myocardial asynergy with cessation of upright bicycle exercise. Circulation 1983;67:529-35.

30. Lin LI. Concordance correlation coefficient to evaluate reproducibility. Biometrics 1989;45:255-68.

31. Rozanski A, Berman DS. Silent myocardial ischemia: II. Prognosis and implications for the clinical assessment of patients with coronary artery disease. AM HEAHT J 1987;114:627-38.

pathophysiologic factors governing the variability of ischemic responses to treadmill and bicycle...

Documents