METHODS JULIO A. PANZA, MD, ARSHED A. QUYYUMI, MD, JEAN G. DIODATI, MD, TIMOTHY S. CALLAHAN, MS, STEPHEN E. EPSTEIN, MD, FACC

Transcription

1 JACC Vol. 17. No.3 Marh : METHODS Predition of the Frequeny and Duration of Ambulatory Myoardial Ishemia in Patients With Stable Coronary Artery Disease by Determination of the Ishemi Threshold From Exerise Testing: Importane of the Exerise Protool JULIO A. PANZA, MD, ARSHED A. QUYYUMI, MD, JEAN G. DIODATI, MD, TIMOTHY S. CALLAHAN, MS, STEPHEN E. EPSTEIN, MD, FACC Bethesda. Maryland The relation between ambulatory myoardial ishemia and the results of exerise testing in patients with ishemi heart disease remains undefined, beause of the dissimilar results of previous reports. To further investigate this issue and, in partiular, to asertain the importane of the exerise protool in determining that relation, 70 patients with stable oronary artery disease underwent 48 h ambulatory eletroardiographi (ECG) monitoring and treadmill exerise tests after withdrawal of mediations. Patients exerised using two different protools with slow (National Institutes of Health [NIH] ombined protool) and brisk (Brue protool) work load inrements. Exerise duration was longer with the NIH ombined protool (14.1 ± 5 versus 6.8 ± 2 min; p < ), but the maximal work load and peak heart rate ahieved were greater with the Brue protool (9.8 ± 2 versus 6.5 ± 2 METs, and 142 ± 19 versus 133 ± 22 beats/min, respetively; p < ). A lose inverse orrelation between exerise testing and the results of ambulatory ECG monitoring was observed using the NIH ombined protool; the strongest orrelation was observed between time of exerise at 1 mm of ST segment depression and number of ishemi episodes (r = -0.86; P < ). With the Brue protool a signifiantly weaker inverse orrelation was found (r = -0.35). The mean heart rate at the onset of ST segment depression was similar during monitoring and during exerise testing with the NIH ombined protool (97.2 ± 13 versus ± 17 beats/min, respetively) but it was signifiantly higher (110.4 ± 13) when using the Brue protool (p < 0.001). These findings indiate that a relation does exist between ambulatory myoardial ishemia and the results of the exerise test but this relation is ritially determined by the exerise protool and is better observed with protools that produe slow work load inrements. These observations raise questions about the independent value and usefulness of ambulatory ECG monitoring in patients with stable ishemi heart disease. (J Am Coil CardioI1991;17:657-63) The episodes of myoardial ishemia that may our during daily life in patients with oronary artery disease are often symptomatially silent 0-3) and their presene during ambulatory eletroardiographi (ECG) monitoring is believed to provide important prognosti information in patients with stable forms of ishemi heart disease independent of that provided by the exerise stress test (4-6). Also, the lak of orrelation between the assessment of myoardial ishemia by exerise testing and ambulatory ECG monitoring reported in some studies (7-9) has supported the view that episodes of ambulatory ishemia are ommonly triggered by sudden redutions in oronary blood flow (2,10). However, the issues regarding the relation between the presene and magnitude of ambulatory myoardial ishemia From the Cardiology Branh. National Heart, Lung, and Blood Institute. National Institutes of Health. Bethesda. Maryland. Manusript reeived June ; revised manusript reeived August 29, aepted September Address for reprints: Julio A. Panza. MD. National Institutes of Health. Building 10, Room 7B-15, Bethesda, Maryland by the Amerian College of Cardiology and the results of the exerise stress test are largely unresolved beause previous studies have shown both a good 01-13) and a poor (7-9) orrelation. Resolution of these disparate results would be important beause, if no suh relation exists, identifiation of episodes of ambulatory ishemia would provide valuable independent information and. therefore, ambulatory ECG monitoring should be inluded in the routine evaluation of patients with oronary artery disease. Although there are several explanations that an aount for the dissimilar observations of previous reports, it is important to reognize that different exerise protools were used in these studies (7-9,11-13). The present investigation was therefore designed to further analyze the relation between the presene and magnitude of myoardial ishemia during daily life and the threshold at whih ECG evidene of myoardial ishemia ours during exerise testing (the ishemi threshold); in partiular, we wished to establish whether the exerise protool played a role in determining this relation /91/$3.50

2 658 PANZA ET AL. JACC Vol. 17. No.3 Marh : Table 1. Desription of the Two Exerise Testing Protools Used in the Study Brue Protool NIH Combined Protool Speed Grade Time of Exerise Speed Grade (mph) (%) METs (min) (mph) (%) METs Methods Study patients. Seventy patients with stable oronary artery disease were inluded in this study, whih was approved by the National Institutes of Health Investigational Review Board. Mean age was 59.8 ± 8 years (range 41 to 75); there were 53 men and 17 women. All patients had undergone ardia atheterization as part of their evaluation for suspeted oronary artery disease, hest pain syndrome or previous myoardial infartion. Coronary angiograms were interpreted by an independent reader; a signifiant stenosis was onsidered as ::::70% narrowing in the internal diameter of at least one major oronary artery. Twenty-six patients (37%) had single vessel disease, 24 (34%) had two vessel disease and the remaining 20 (29%) had involvement of all three major oronary vessels. Fifty patients were asymptomati and the remaining 20 had stable symptoms. None had had a myoardial infartion in the 3 months before the study. Patients with left bundle branh blok, those who were reeiving digitalis and those who had any other form of ardia disease were exluded from the study. After informed onsent was obtained, tests were arried out after withdrawal of anti anginal mediations for at least 48 h. During ambulatory ECG monitoring, patients took sublingual nitroglyerin for hest pain but prophylati administration was not allowed. The ambulatory monitoring and exerise stress tests were performed within 72 h in 16 patients, within 1 week in 21 patients and within 1 month in the remaining 33 patients. Exerise testing protools (Table 1). Patients underwent two symptom-limited exerise tests, one using the standard Brue protool (14) and the seond using the National Institutes of Health (NIH) ombined protool (15). In 42 patients (60%) the exerise test with the NIH ombined protool preeded that with the Brue protool; in the remaining 28 patients (40%) the order was reversed. The NIH ombined protool utilizes a onstant treadmill speed with gradual inreases in slope during the first five stages (12.5 min) similar to the Balke-Ware (16) and Naughton (17) protools; after the fifth stage both the treadmill speed and slope are inreased at eah stage. In partiular, the NIH ombined protool employs slower inreases in work load than does the Brue protool. For instane, at 5 min of exerise patients ahieve 7 METs with the Brue protool but only half of that work load with the NIH ombined protool; similarly, subjets ahieve 10 METs after exerising for 20 min on the NIH ombined protool, although they do so after only 9 min on the Brue protool. The NIH ombined protool has proved to be useful for the serial noninvasive evaluation of patients with oronary artery disease (15). Twelve-lead ECGs were obtained at rest, at every minute during exerise and at peak exerise. On the ECG, lead avr was replaed by lead CMs. The exerise was terminated when hest pain, ST segment depression ::::4 mm, a derease in blood pressure ::::20 mm Hg, ventriular tahyardia, extreme fatigue or shortness of breath developed during the test. A test was onsidered positive when planar or downsloping ST segment depression:::: 1 mm, at 0.08 s after the J point, was observed. When the exerise test was onsidered positive, the ECG traings showing onset of ST segment depression and ahievement of 1 mm of ST segment depression were identified. For the overall patient group the onset of ST segment depression, both on exerise testing and ambulatory ECG monitoring, was defined as the earliest ECG manifestation of repolarization hanges harateristi of myoardial ishemia that preeded the development of more pronouned and definite hanges in the ST segment. This early manifestation of repolarization abnormalities was usually reognized when the ST segment shift was 0.5 mm, although in some patients it was possible to identify ECG traings showing even more minor hanges (suh as ST segment depression of 0.25 mm). Therefore, to ensure omparability of the readings, the ECG traings from ambulatory monitoring and exerise testing were mathed in retrospet for eah individual patient so that the extent of repolarization hanges, identified as the onset of ST segment depression, was similar for all tests. For the purpose of orrelation with the results of ambulatory ECG monitoring, the ishemi threshold was then assessed from the exerise test in four separate ways: 1) heart rate at onset of ST segment depression, 2) time of

3 lacc Vol. 17, No.3 Marh I, 1991: PANZA ET AL. 659 Table 2. Results of Exerise Testing Using Two Different Exerise Protools in Patients with Stable Coronary Artery Disease Brue Protool NIH Combined Protool Exerise duration (min) 6.8 ± ± 5 Maximal work load (METs) 9.8 ± ± 2 Peak heart rate (beats/min) 142 ± ± 22 Time at onset of ST * (min) 3.2 ± ± 3 Heart rate at onset of ST " 113 ± ± 15 (beats/min) Time at I mm ST " (min) 4.9 ± ± 5 Heart rate at I mm ST " 126 ± ± 9 (beats/min) p Value <0.001 "Inludes only 47 patients with abnormal exerise tests using both protools. ST = ST segment depression. exerise at onset of ST segment depression, 3) heart rate at 1 mm of ST segment depression, and 4) time of exerise at 1 mm of ST segment depression, Ambulatory ECG monitoring. Eah patient underwent a 48 h period of ambulatory ECG monitoring using lead CM 5 and modified lead II (11). Patients were instruted to arry out their normal daily ativities and to keep a detailed diary of their ativities, symptoms and onsumption of sublingual nitroglyerin. Tapes were analyzed by an independent reader at 60 to 120 times normal speed, utilizing a Del Mar Avionis model 750A system. Printouts at a paper speed of 25 mmls were obtained before, during and after any hange in the ST segment level. An ishemi episode was defined as :2: 1 mm ST segment depression, at 0.08 s after the J point, lasting for :2: 1 min. Return of ST segment level to baseline for :2:3 min was required between two episodes. The magnitude of ambulatory myoardial ishemia was assessed in two separate ways: 1) the number and 2) the total duration of ishemi episodes during the period of monitoring. For eah ishemi episode the heart rate at the onset of ST segment depression was determined. In eah patient with more than one episode during monitoring the values of heart rate at the onset of repolarization hanges for eah episode were averaged for the purpose of orrelation with the heart rate at the onset of ST segment hanges during exerise testing. Statistial analysis. Comparison of mean values on exerise testing and ambulatory monitoring was performed using the paired t test. Inreases in heart rate during the two exerise protools were ompared by using one-way analysis of variane for repeated measurements. Proportions were ompared by the hi-square test. Relation between different variables were assessed by means of Pearson's orrelation oeffiient. Different r values were then ompared using Fisher's z transformation. Data are presented as mean values ± SD. A p value <0.05 was onsidered signifiant. Table 3. Reasons for Termination of Exerise Treadmill Tests Using Two Different Exerise Protools in 70 Patients With Stable Coronary Artery Disease* Brue NIH Combined Protool Protool p Chest pain 32 (46%) 26 (37%) NS ST t 2: 4 mm 2 (3%) 6(9%) NS Fatigue or breathlessness 34 (48%) 38 (54%) NS Ventriular tahyardia 2 (3%) 0(0%) NS Derease in blood 0(0%) 0(0%) NS pressure 2:20 mm Hg "Figures denote number of patients. ST t = ST segment depression. Results Exerise testing (Tables 2 and 3). As expeted, patients exerised longer with the NIH ombined protool than with the Brue protool; however, the maximal work load and peak heart rate ahieved were greater with the Brue protool (Table 2). No signifiant differenes were found between the two exerise protools with regard to the reasons for stopping the tests (Table 3). Inreases in heart rate in relation to exerise time were signifiantly steeper during the Brue protool than during the NIH ombined protool (Fig. 1). Ishemia developed during exerise testing in 47 (67%) of the 70 patients when using the NIH ombined protool and in 58 patients (82%) when using the Brue protool. None of the patients with a normal Brue exerise test had an abnormal test when using the NIH ombined protool. In the 47 patients in whom both exerise tests were abnormal, the time at onset and at 1 mm of ST segment depression were greater using the NIH ombined protool. Conversely, the heart rate at onset and at 1 mm ST segment depression was higher when using the Brue protool (Table 2). Ambulatory ECG monitoring. A total of 3,297 reorded hours were suitable for ST segment analysis. Forty-five patients (64%) had at least one episode of ST segment Figure 1. Inrease in heart rate during treadmill exerise testing using the standard Brue protool (solid irles) and the NIH ombined protool (open irles) in 70 patients with stable oronary artery disease. As a onsequene of faster inrements in work load, the Brue protool produes more rapid inreases in heart rate. oo i :960! OJ II: t::40 :r:.5:20 Ip<O.ll n=70 0L-----L----L----.: Exerise Time (min)

4 660 PANZA ET AL. JACC Vol. 17. No.3 Marh : NIH 228 Combined Neg. 23 pts 1 Brue Neg. 12 pts NIH Combined Protool Pos. 47 pts Ishemi episodes during monitoring Pos. 58 pts Figure 2. Pie harts relating the results of ambulatory ECG monitoring and exerise testing with the NIH ombined protool (left panel) and the Brue protool (right panel). Most patients with a negative (Neg.) exerise test did not have episodes of myoardial ishemia during monitoring, whih were frequently observed in patients with ST segment depression during exerise. Pos. = positive; pts = patients. 0).;:,g 0 0) 'g 8! 'B.!Il. w u 'E Q).l: u!!l '5 i z Brue Protool depression during monitoring. A total of 321 ishemi episodes were reorded, 84% of whih were silent. Average number of ishemi episodes per patient over the 48 h of reording was 4.4 (range, 0 to 18). The total duration of the ishemi episodes was 7,957 min (72% were silent) with a mean of 113 min per patient (range 0 to 955). Relation between exerise testing and ambulatory ECG monitoring (Fig. 2 and 3, Table 4). Only 1 (4%) of the 23 patients with a normal exerise test using the NIH ombined protool and none of the 12 patients with a normal test using the Brue protool had ishemi episodes during ambulatory ECG monitoring (Fig. 2). Thus, 316 of the total 321 episodes of myoardial ishemia during monitoring (98%) were observed in the group of patients with ST segment depression on exerise testing (p < when ompared with patients with a normal exerise test using either protool). With the NIH ombined protool, patients who had an abnormal exerise test showed a lose inverse orrelation between the ishemi threshold and the number and duration of ishemi episodes during ambulatory monitoring (Table 4). The strongest orrelation was observed when the ishemi threshold was taken as the time of exerise at 1 mm ST segment depression and the magnitude of ambulatory myoardial ishemia was assessed as the number of ishemi episodes (r = -0.86; p < ) (Fig. 3). Thus, patients in whom 1 mm ST segment depression developed early during the Time at 1 mm sn (min) Figure 3. Correlation between the number of ishemi episodes during ambulatory ECG monitoring and the time at 1 mm of ST segment depression (ST ) during exerise with the NIH ombined protool (top panel) and the Brue protool (bottom panel). Although both exerise protools showed an inverse relation with the results of ambulatory monitoring. the orrelation with the NIH ombined protool was signifiantly stronger. exerise test (low ishemi threshold) had several ishemi episodes during ECG monitoring. Conversely, patients in whom I mm ST segment depression ourred during the later stages of the test (high ishemi threshold) had few or no episodes. When using the Brue protool, a similar but weaker inverse orrelation was found (Fig. 3). When the orrelations between the two exerise protools were ompared, the NIH ombined protool had a signifiantly stronger relation with the results of ambulatory ECG monitoring than did the Brue protool for most of the variables seleted (Table 4). Heart rate during exerise testing and ECG monitoring (Fig. 4 and 5). Analysis of the frequeny distribution of heart rate at onset of ECG hanges during ambulatory monitoring and exerise tests using both protools showed Table 4. Correlation Between Myoardial Ishemia During Ambulatory ECG Monitoring and Ishemi Threshold Assessed From Exerise Testing Using Two Different Protools in Patients with Stable Coronary Artery Disease Time of Exerise at Onset of Heart Rate at Onset of ST 1 SIt Heart Rate at 1 mm ST 1 Time of Exerise at 1 mm ST 1 NIH NIH NIH NIH Brue Combined p' Brue Combined p* Brue Combined p' Brue Combined p' Ishemi episodes Number r = r = <0.05 r = r = <0.002 r = r = <0.05 r = r = Duration r = r = r = r = r = r = r = r = <0.02 'omparison between orrelation oeffiients obtained with the two exerise protools. ST t = ST segment depression.

5 lacc Vol. 17. No.3 Marh I : PANZA ET AL l!l 10 0::: '". 8.. '0 Holter - NIH Combined Brue Heart Rate at Onset sn (beats/min) Figure 4. Frequeny distribution urves of the heart rate at the onset of ST segment depression (ST t ) during ambulatory (Holter) monitoring (dotted line) and exerise testing with the NIH ombined protool (solid line) and the Brue protool (dashed line). Note the rightward shift of the Brue protool urve in relation to both the Holter monitoring and NIH ombined protool urves that share a similar distribution. that ambulatory monitoring and the NIH ombined protool shared a similar distribution (Fig. 4). In fat, the mean heart rate at onset of ishemi episodes during monitoring did not differ signifiantly from the mean heart rate at onset of ST segment depression during exerise testing using the NIH ombined protool (97.2 ± 13 versus ± 17 beats/min, respetively). However, the frequeny distribution urve for the Brue protool was displaed to the right (Fig. 4), so that the heart rate at whih ST segment depression developed was signifiantly higher with the Brue protool (110.4 ± 13 beats/min) than either with the NIH ombined protool or Figure 5. Correlation between the heart rate at onset of ishemi episodes during ambulatory monitoring and the heart rate at onset of ST segment depression (ST t ) during exerise testing with the NIH ombined protool (top panel) and the Brue protool (bottom panel). Both exerise protools orrelated with ambulatory ECG monitoring; however. the orrelation for the NIH ombined protool was signifiantly stronger l!l ill I w (J 'E. '" M '0 i 0::: o 120 NIH Combined Protool Brue Protool Heart Rate at Onset of sn (beats/min) during ambulatory ECG monitoring (p < in both ases). Moreover, the average heart rate at onset of episodes of ambulatory myoardial ishemia for eah patient signifiantly orrelated with the heart rate at onset of ST segment hanges during exerise testing (Fig. 5); this orrelation was lose (r = 0.69) for the NIH ombined protool and signifiantly weaker (r = 0.39) for the Brue protool (p < when both r values were ompared). Disussion The relation between the presene and magnitude of episodes of myoardial ishemia during ambulatory monitoring and the results of exerise testing in patients with oronary artery disease has been the subjet of inreasing interest (18). espeially in the light of reent studies (4-6) suggesting that silent episodes of myoardial ishemia during daily life may onvey important information independent of that provided by exerise testing. However, beause previous studies have shown disparate results (7-9,11-13), this issue still remains largely unanswered. Relation between ambulatory ECG monitoring and exer ise testing. The results of the present study indiate that a strong orrelation between the frequeny and duration of ambulatory myoardial ishemia and the ishemi threshold (the heart rate or the time of exerise at whih ECG evidene of ishemia develops) assessed from exerise testing does exist, but suh a orrelation is ritially determined by the exerise protool. Thus, when the NIH ombined protool, whih employs slow work load inrements, is used: 1) patients with a normal exerise test usually do not have episodes of myoardial ishemia during monitoring; 2) patients with a low ishemi threshold (that is, ST segment depression that develops early during the test at a relatively low heart rate) have many ishemi episodes during ambulatory monitoring; and 3) patients with a high ishemi threshold have few or no episodes. These findings are reinfored by the observation that, when the NIH ombined exerise protool was used, the mean heart rate at the onset of ST segment depression did not differ between ambulatory monitoring and exerise testing. Although a similar inverse orrelation was found with the results of ambulatory ECG monitoring when using the Brue protool (Fig. 3), this was signifiantly weaker and did not ahieve statistial signifiane for some of the variables seleted (Table 4). Analysis of different exerise protools. The different results obtained by the two exerise protools in the orrelation with the magnitude of myoardial ishemia as assessed by ambulatory ECG monitoring an be best explained by analysis of the effets of the diverse rate in work load inrements during exerise. The brisk inrements in work load employed by the Brue protool produe rapid inreases in the oxygen demands of the myoardium (Fig. 1). Beause there must be some delay in the appearane of ECG hanges one ishemia ours, this rapid inrease in myoardial oxygen demands would allow the heart rate to

6 662 PANZA ET AL. JACC Vol. 17, No.3 Marh I, 1991 : ahieve relatively higher levels before ST segment depression develops and would result in the assessment of a falsely high ishemi threshold in some patients. This and the fat that the heart rate hanges are ompressed into a relatively brief time interval probably prevents the adequate separation of patients with diverse ishemi thresholds into appropriate subgroups, resulting in a poorer orrelation with the findings of ambulatory monitoring. This would also explain why the heart rate at the onset of ST segment depression was signifiantly higher on exerise testing with the Brue protool than during ambulatory monitoring (Fig. 4), as also shown in several other studies (7-9). In ontrast, the NIH ombined protool employs slower work load inrements and therefore auses more gradual inreases in myoardial oxygen demands (Fig. 1). This permits a more aurate approximation of the true ishemi threshold in eah individual patient beause the ourrene of onset of ST segment depression is distributed over a wider range of heart rate and time of exerise. Suh an analysis may aount for the disrepanies in previous studies, as suggested by the fat that those reports failing to show a orrelation between the results of exerise testing and ambulatory monitoring used the standard Brue protool (8,9) and those reporting a good orrelation between the two tests (11-13) used the modified Brue protool (a slower work load protool). This observation also suggests that the strong orrelation we found between the results of exerise testing and ambulatory ECG monitoring an probably be obtained with any exerise protool that employs slow work load inrements. Although it was not the purpose of this study to ompare the sensitivity of different exerise protools, it should be noted that the Brue protool proved to be more sensitive than the NIH ombined protool for detetion of ishemia in our group of patients with known oronary artery disease; that is, a positive exerise test was more frequently observed with the Brue protool (Fig. 2). This differene is probably due to the higher work loads ahieved by patients with this protool (Table 2), whih inreased the possibility of deteting ishemia during exerise. This observation suggests that the Brue protool is potentially more useful for the purpose of sreening patients with suspeted oronary disease. Clinial impliations. Our findings may have important linial impliations relative to the value of ambulatory ECG monitoring in the assessment of patients with ishemi heart disease. First, aording to the results of reent studies (4-6), the presene of episodes of ST segment depression on ambulatory ECG monitoring is believed to onvey important prognosti information in patients with stable oronary artery disease, independent of the results of other noninvasive tests. However, the present study indiates that the results of ambulatory ECG monitoring an be largely predited by analysis of the exerise test if the exerise protool employs gradual work load inrements; therefore, ambulatory ECG monitoring may not add substantial information to that obtained by exerise stress testing in most patients with stable ishemi heart disease. Of note, studies showing that episodes of ambulatory myoardial ishemia provide independent prognosti information have used the Brue exerise protool (4-6), a fat that ould have prevented adequate stratifiation of patients aording to their exerise ishemi threshold, therefore limiting the usefulness of the exerise test. Seond, the disrepanies between the results of exerise testing and ambulatory ECG monitoring reported in previous studies (7-9) have supported the view that episodes of myoardial ishemia during daily life are largely a onsequene of signifiant redutions in oronary blood flow (2, to). Although the design of this investigation does not allow us to asertain the mehanisms responsible for the ourrene of episodes of ambulatory myoardial ishemia, the observation that the heart rate at the onset of ST segment depression was similar during ambulatory ishemi episodes and during exerise testing when using the NIH ombined protool suggests that both forms of ishemia (on exerise testing and during monitoring) usually share a ommon trigger, that is, an inrease in myoardial oxygen demand. This is further supported by the strong orrelation observed between the ishemi threshold obtained by exerise testing and the frequeny and duration of ambulatory episodes. This view does not exlude the ourrene of superimposed vasomotor hanges that might also influene the ishemi threshold of spontaneous episodes of ishemia during daily life, as suggested by the observation that the heart rate at the onset of ishemi episodes during monitoring shows a wide variation in eah individual patient (11). Limitations. Certain limitations in our study design must be aknowledged. First, our population omprised patients with angiographially doumented oronary artery disease who were linially stable and in most ases asymptomati. Therefore, the results of our study annot be extrapolated to other subsets of patients with ishemi heart disease, suh as patients with angina at rest, those who have reently had a myoardial infartion or an exaerbation in their anginal pattern or those without signifiant narrowing of the epiardial vessels. Seond, beause we ould not measure the blood pressure by sphygmomanometry reliably every minute during exerise, suh measurements were not taken into aount for assessment of the ishemi threshold. This may have affeted the interpretation of our results in ertain patients beause episodes of myoardial ishemia deteted by radionulide tehniques an our as a onsequene of inreases in blood pressure without signifiant hanges in heart rate (19). However, the strong orrelation we found between the heart rate at whih myoardial ishemia ours during exerise testing and ambulatory monitoring suggests that suh episodes of ishemia are unommonly deteted by eletroardiography in patients with stable oronary artery disease. Finally, sine all our study patients were withdrawn from their antianginal mediations during the tests, we annot be ertain whether a similar orrelation exists when patients are reeiving medial therapy. However, Mulahy

7 JACC Vol. 17, No.3 Marh I, 1991 : PANZA ET AL. 663 et al. (13), who inluded in their study patients who were reeiving onventional anti-ishemi therapy, found that this treatment did not affet the relation between the results of ambulatory monitoring and exerise testing. Conlusions. Our study demonstrates that a strong orrelation exists between the frequeny and duration of myoardial ishemia during daily life and the ishemi threshold as assessed by exerise testing, but only when an exerise protool is used that employs slow work load inrements. The existene of suh a relation suggests that episodes of myoardial ishemia during daily life in patients with stable ishemi heart disease an be largely predited by exerise testing and therefore raises questions about the utility of ambulatory ECG monitoring as part of the routine evaluation of these patients. Referenes I. Stern S, Tzivoni D. Early detetion of silent ishemi heart disease by 24-hour ECG monitoring of ative subjets. Br Heart J 1974;36: Shang SJ, Pepine CJ. Transient asymptomati ST segment depression during daily ativity. Am J Cardiol 1977;39: Mulahy D, Keegan J, Crean P, et al. Silent myoardial ishemia in hroni stable angina: a study of frequeny and harateristis in 150 patients. Br Heart J 1988:60: Roo MB, Nabel EG, Campbell S, et al. Prognosti importane of myoardial ishemia deteted by ambulatory monitoring in patients with stable oronary artery disease. Cirulation 1988:78: Tzivoni D, Weisz G. Gavish A, Zin D, Keren A, Stern S. Comparison of mortality and myoardial infartion rates in stable angina petoris with and without ishemi episodes during daily ativities. Am J Cardiol 1989:63: Deedwania PC, Carbajal EV. Silent ishemia during daily life is an independent preditor of mortality in stable angina. Cirulation 1990:81: Deanfield JE, Selwyn AP, Chierhia S, et al. Myoardial ishemia during daily life in patients with stable angina: its relation to symptoms and heart rate hanges. Lanet 1983:2: Mody FV, Nademanee K, Intarahot V, Josephson MA, Robertson HA, Singh BN. Severity of silent myoardial ishemia on ambulatory eletroardiographi monitoring in patients with stable angina petoris: relation to prognosti determinants during exerise stress testing and oronary angiography. J Am Coll CardioI1988:12: Benhorin J, Tzivoni D, Moriel M, et al. Can exerise test parameters predit severity of ishemi hanges during daily ativities? (abstr). Cirulation 1989:80(suppl 11): Cohn PF, Lawson WE. Charateristis of silent myoardial ishemia during out-of-hospital ativities in asymptomati angiographially doumented oronary artery disease. Am J CardioI1987:59: II. Quyyumi AA. Mokus L, Wright C, Fox KM. Morphology of ambulatory ST segment hanges in patients with varying severity of oronary artery disease: investigation of the frequeny of noturnal ishemia and oronary spasm. Br Heart J 1985 :53: Campbell S, Barry J, Roo MB, et al. Features of the exerise test that reflet the ativity of ishemi heart disease out of hospital. Cirulation 1986:74: Mulahy D, Keegan J, Sparrow J, Park A, Wright C, Fox K. Ishemia in the ambulatory setting-the total ishemi burden: relation to exerise testing and investigative and therapeuti impliations. J Am Coll Cardiol 1989:14: Brue RA. Exerise testing of patients with oronary artery disease. Ann Clin Res 1971:3: Kent KM, Bonow RO, Rosing DR, et al. Improved myoardial funtion during exerise after suessful perutaneous transluminal oronary angioplasty. N Engl J Med 1982:306: Balke B, Ware RW. An experimental study of "physial fitness" of Air Fore personnel. US Armed Fores Med J 1959:10: Naughton J. Balke B. Poarh A. Modified work apaity studies in individuals with and without oronary artery disease. J Sports Med Phys Fitness 1964:4: Epstein SE, Quyyumi AA, Bonow RO. Myoardial ishemia-silent or symptomati. N Engl J Med 1988;318: Rozanski A, Bairey CN, Krantz DS, et al. Mental stress and the indution of silent myoardial ishemia in patients with oronary artery disease. N Engl J Med 1988;318: