View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

lACC Vol. 6, No.2 275 August 1985:275-84

Silent Myocardial : Qemodynamic Changes During Dynamic Exercise in Patients With Proven Coronary Despite Absence of Pectoris

HEINZ O. HIRZEL, MD, RUTH LEUTWYLER, HANS P. KRAYENBUEHL, MD

Zurich, Switzerland

The hemodynamic changes during exercise occurring in fraction decreased (from 59 ± 7 to 50 ± 15 and from 36 patients with proven (10 with• 60 ± 4 to 52 ± 9% in the study and control groups out and 26 with previous myocardial ) who without prior infarction and from 54 ± 9 to 47 ± 10 tolerated the test without angina were analyzed and 55 ± 9 to 50 ± 4% in the respective groups with aud compared with changes observed in a control group prior infarction). Whereas the changes from rest to ex• of 36 carefully matched patients whose exercise was lim• ercise were highly significant within each group, no sig• ited by angina. All patients were exercised to the same nificant differences were noted between the correspond• extent, reaching a similar rate-pressure product at the ing groups. Regional de novo hypokinesia appeared in end of the stress test (19,508 ± 4,828 [SD] versus 19,247 all patients without prior infarction and in 25 and 22 ± 4,1l7 beats/min x mm ~g [NS] in the study and patients, respectively, of the groups with prior infarction. control groups without prior infarction, and 19,665 ± Thus, under similar physical stress conditions, com• 3,950 versus 17,701 ± 4,600 beats/min x mm Hg [NS] parable hemodynamic changes indicative of ischemia are in the respective groups with infarction). In all groups observed in patients with significant coronary artery le• left ventricular end-diastolic pressure increased from rest sions with or without previous ir• to exercise (from 18 ± 4 to 36 ± II and from 13 ± 5 respective of the occurrence of angina. Therefore, an• to 29 ± 9 mm Hg, respectively, in the study and control gina pectoris cannot be considered a prerequisite for groups without prior infarction and from 17 ± 7 to 32 hemodynamically significant ischemia during exertion. ± 13 and from 19 ± 7 to 36 ± 9 mm Hg in the respective (J Am Coil Cardiol 1985;6:275-84) groups with prior infarction). Left ventricular ejection

Angina pectoris is considered a sensitive marker of ischemic scure and little is known about the hemodynamic changes events. Indirect evidence, however, suggests that ischemia associated with such attacks. may also occur silently, without accompanying chest . Whereas global and regional changes of left ventricular Such evidence has been derived from exercise induced ST function in the anginal state induced by various stress tests segment changes (1-6), clear-cut defects at stress are well studied (15-25), only a few reports (25-27) have (7,8) or typical ischemic electrocardiographic dealt with alterations of function during dynamic exercise alterations during continuous ambulatory Holter monitoring tolerated without angina despite significant coronary artery (8-14) in asymptomatic patients with arteriographically proven lesions. Thus, we examined the hemodynamic changes dur• coronary artery lesions and patients who later developed ing exercise occurring in 36 of 258 consecutive patients symptomatic coronary disease. Yet, the underlying with proven coronary artery disease who tolerated the stress pathophysiologic mechanisms of silent ischemia remain ob- test without symptoms.

Methods From the Department of Medicine. Medical Policlinic. Division of StuJly patients (Groups I and II). Of 258 consecutive Cardiology. University Hospital. Zurich. Switzerland. This work was sup• ported in part by a grant of the EM DO-Foundation, Zurich. Switzerland. patients with arteriographically severe coronary artery le• Manuscript received December 5. 1984; revised manuscript received March sions who were evaluated by dynamic exercise during the 12. 1985. accepted March 29. 1985. invasive examination, 36 were angina-free on admission and Address for reprints: Heinz O. Hirzel, MD. Department of Medicine. Medical Policlinic. Division of Cardiology. University Hospital. CH-8091 had no anginal symptoms during a bicycle stress test con• Zurich. Switzerland. ducted in the upright position on the day before cardiac

© 1985 by the American College of Cardiology 0735-1097/85/$3.30 276 HIRZEL ET AL. JACC Vol. 6, No.2 HEMODYNAMIC CHANGES IN SILENT ISCHEMIA August 1985:275-84

catheterization and at catheterization. All 36 patients had weeks before the present admission. Thus, at hospital entry, experienced one or more episodes of chest pain 4 to 8 months none of the patients was receiving appropriate antianginal before admission and 26 patients had a history of myocardial therapy. infarction that dated back more than 6 months. The indi• Control patients (Groups III and IV). The 36 patients cation for catheterization was posed at the time when symp• without exercise-induced angina were matched with 36 pa• toms occurred or in the early recovery period after myo• tients who were limited by angina at stress. These control cardial infarction and was determined on the basis of objective patients were similar to the patients without stress-induced findings such as a positive exercise stress test or exercise• angina in age, number of diseased vessels, extent of ath• induced reversible perfusion defects in the -20l erosclerosis, left ventricular end-diastolic pressure at rest, scintigram, or both. All patients were male, ranging in age ejection fraction and absence of site of prior myocardial from 26 to 63 years (Table 1). Group I consisted of 10 infarction. The diagnosis of a previous myocardial infarction patients without prior myocardial infarction and Group II was based on the history of an appropriate event associated of 26 patients with prior myocardial infarction. with electrocardiographic or enzymatic changes, or both, Before hospital entry, 5 of the 10 patients without prior and circumscribed left ventricular asynergy in the angiogram infarction in Group I were treated intermittently with beta• at rest. Despite continued medical therapy, all control pa• receptor blocking agents, and long-acting nitrates, 3 patients tients continued to have effort-induced angina. To obtain with beta-blocking agents, long-acting nitrates and calcium the most precise information regarding the clinical relevance channel antagonists, 1 with beta-blocking agents and cal• and severity of the disease, all medication with the excep• cium antagonists and 1 with beta-blocking agents only. The tion of short-acting nitrates was discontinued in all patients 26 patients in Group II had received beta-blocking agents on admission. No withdrawal effects necessitated restoration and long-acting nitrates prophylactically for 3 months fol• of full antianginal therapy in any patient. lowing hospital discharge after infarction. Because symp• Clinical evaluation. All patients were examined clini• toms did not reappear when discontinuing medication the cally and by routine laboratory tests including a 12 lead majority of these patients had stopped medication several standard electrocardiogram at rest and a continuous muIti-

Table 1. Characteristics of the Four Patient Groups With Proven Coronary Artery Disease Patients Without MI Patients With MI

Group I Group III Group II Group IV (without AP) (with AP) (without AP) (with AP)

Patients (no.) 10 10 26 26 Age (yr) Mean 52 ± 7 50 ± 6 48 ± 9 49 ± 8 Range 38 to 63 39 to 63 26 to 61 33 to 60 Levine vascularization index -1.7 ± 0.9 -1.7 ± 0.8 -1.8 ± 0.9 -1.9 ± 0.8 Number of diseased vessels I vessel 2 2 8 6 2 vessels 6 6 5 7 3 vessels 2 2 13 13 Critically stenosed vessels (>70% narrowing) LAD 5 6 21 24 LCx 8 7 II 11 RCA 6 6 17 17 Site of infarction Anterior II II Posterior 15 15 Upright bicycle stress test 131 ± 21 98 ± 39* 129 ± 37 97 ± 32t work load attained (walls) ST segm~nt depression 3 6 12 13 (>0.1 mY) ST segment elevation o o (>0.1 mV) Angina pectoris o o 26t Values are expressed as mean values ± I SO. *p < 0.05 versus Group I; tp < 0.01 versus Group II; tp < 0.001 versus Groups I and II, respectively (unpaired t tests). AP = exercise-induced angina pectoris; LAD = left anterior descending coronary artery; LCx = left circumflex coronary artery; MI = myocardial infarction; RCA = right coronary artery. JACC Vol. 6, No.2 HIRZEL ET AL. 277 August 1985:275-84 HEMODYNAMIC CHANGES IN SILENT ISCHEMIA

stage bicycle exercise test in the upright posture with re• Coronary . The angiographic severity of cording of precordial leads V2 , V4 , and Vo. Exercise testing coronary artery disease was assessed by the numerical score was performed according to a modified Bruce protocol (28). proposed by Levine et al. (32). Thus, fixed values were The target work load was determined from a nomogram that assigned to each obstructive lesion located in any of the adjusted for age, sex and height (29). Exercise was started three coronary vessels and their major side branches. Le• at about one third of the target work load; the work load sions with less than 70% reduction in luminal diameter were was then increased stepwise every 3 minutes until chest considered hemodynamically insignificant and were as• pain, dyspnea, fatigue or muscle weakness forced interrup• signed the value O. A 70 to 90% was assigned the tion of exertion. value - 0.5, and a more than 90% stenosis or occlusion Cardiac catheterization and angiography. Premed• of the vessel was given the value - 1.0. By summing the ication consisted of chlordiazepoxide (Librium), 10 mg, values calculated for each individual coronary artery system given perorally I hour before the study. After left heart the vascularization index was established. The more neg• catheterization by the transfemoral arterial route, the feet ative this vascularization index was, the more extensive the of the supine patient were attached to the pedals of the coronary artery alterations. bicycle ergometer, whose axis was 42 cm above the cath• Ventricular hypokinesia. The appearance of de novo hy• eterization table. The left ventricular pressure was then re• pokinesia during exercise was evaluated in two ways: I) corded at rest on an oscillograph (VR 12, Electronics for visual inspection of the biplane left ventricular angiograms Medicine). Thereafter, biplane cineangiography at rest was by two independent observers; in all cases there was agree• performed in the right and left anterior oblique projections ment; and 2) comparison of the fractional shortening of the (Cardioskop-U, Siemens) by injecting an average volume six hemiaxes (23,33) in the right and left anterior oblique of 44 ml of contrast medium (Urografin, 76%, Schering) projections. Thereby, de novo hypokinesia was defined as by means of an electrocardiographically triggered power a greater than 10% decrease in fractional shortening of one injector (Contrac, Siemens) at a flow rate of 13 mils (Fig. hemiaxis from rest to exercise without a concomitant in• I). A 15 minute pause was allowed to restore hemodynamic crease in shortening of the corresponding opposite hemiaxis. baseline levels. Exercise was started at about half the work Hypokinesia was considered present only if both methods load tolerated during the upright stress test performed the yielded a positive result. day before. The work load was then increased stepwise Statistics. The results were expressed as mean values ± every 2 to 3 minutes. At the highest work load tolerated or I standard deviation. To evaluate statistical differences be• at the onset of angina and with the patient still pedaling, tween the data, paired and unpaired Student's t tests, anal• the left ventricular pressure was recorded again and cine• ysis of variance and the chi-square test with the continuity angiography repeated using an average volume of 72 ml of correction described by Yates for small sample sizes were contrast agent injected at a flow rate of 16 mils (Fig. I). used. Selective coronary arteriography using the Judkins tech• nique was performed after cineangiography. During the en• tire examination the patients received 10,000 units of . Results Ventriculographic analysis. The left ventriculograms were The results are summarized in Tables I to 3 and Figures recorded on 35 mm cinefilm exposed at 50 frames/so The 2 and 3. left ventricular cavity silhouettes were drawn from end• Clinical and coronary status of the study and control diastolic and end-systolic frames in both the left and right groups. Because the study and control groups were care• anterior oblique projections (Vanguard projector). The cor• fully matched, they did not differ in mean age, coronary responding frames in end-diastole and end-systole were status (including the number of diseased vessels), mean identified using a numerical code displayed on both the cine vascularization index or site of previous myocardial infarc• frames and the recording paper. The end-diastolic frame tion. However, at the stress test in the upright position the was chosen at the peak of the R wave in the electrocardio• work load tolerated by the patients without exercise-induced gram and the end-systolic frame when the left ventricular angina (Groups I and II) was significantly higher than the area was smallest. The volumetric analysis was confined to work load at which the stress test had to be terminated contractions occurring during normal electrical depolari• because of angina in the respective control groups (Groups zation; extrasystoles or postextrasystolic beats were ex• III and IV). In the groups without prior myocardial infarc• cluded from analysis. After correction for image magnifi• tion, the exercise electrocardiogram revealed an ST segment cation, volume estimates were made using the area-length depression of greater than 0.1 mV in three patients without method (30,31), and biplane ejection fraction was com• stress-induced angina (Group I) and six with angina (Group puted. In addition, rate-pressure product, as an estimate of III) (NS). Among the patients with prior infarction, ST myocardial consumption, and the increase in the segment depression was found in 12 patients without angina rate-pressure product from rest to exercise were calculated. (Group II) and 13 with angina (Group IV) (NS) and sig- 278 HIRZEL ET AL. lACC Vol. 6, No.2 HEMODYNAMIC CHANGES IN SILENT ISCHEMIA August 1985:275-84

Figure 1. Cineangiograms obtained in the right anterior oblique projec• tion in end-diastole (above) and end• systole (below) at rest (left panels) and during exercise (right panels) at similar work loads in two patients with severe three vessel coronary ar• tery disease. A, Case I, a 63 year old man who tolerated the stress test without anginal symptoms. During exercise (for 3 minutes, to a work load of 80 watts), left ventricular ejection fraction decreased from 52 to 35% and end-diastolic pressure in• creased from 28 to 40 mm Hg. Fur• thermore, severe hypokinesia of the anterolateral wall developed (ar• rows). B, Case 2, a 47 year old man who was limited by severe angina at exercise. During exercise (for 2 min• utes to a work load of75 watts), left ventricular ejection fraction de• creased from 69 to 44% and end• diastolic pressure increased from 16 to 40 mm Hg. Again, marked de novo hypokinesia of the anterolateral, ap• ical and inferior portions of the left ventricle became apparent (arrows). 280 HIRZEL ET AL. JACC Vol. 6, No.2 HEMODYNAMIC CHANGES IN SILENT ISCHEMIA August 1985:275-84

Table 3. Hemodynamic Changes During Exercise in the 52 Patients With Myocardial Infarction With and Without Angina Pectoris

Patients With Myocardial Infarction

p Group II Group IV (Group II versus (without AP) (with AP) Group IV)

Work load attained (watts) 107 :+: 35 99 :+: 33 NS HR (beats/min) Rest 78 :+: 13 71 :+: II <0.05 Exercise 127 :+: 18* 118 :+: 22* NS LVEOP (mm Hg) Rest 17 :+: 7 19 :+: 7 NS Exercise 32 :+: 13* 36 :+: 9* NS LVSP (mm Hg) Rest 130 :+: 22 135 :+: 15 NS Exercise 154 :+: 22* 149 :+: 23* NS Rate-pressure product (beats/min x mm Hg) Rest 10,185 :+: 2,326 9,954 :+: 2,182 NS Exercise 19,665 :+: 3,950* 17,701 :+: 4,600* NS % increase 97 :+: 37 80 :+: 40 NS EOVI (mllm2) Rest 121 :+: 37 III:+: 26 NS Exercise 130 :+: 39 125 :+: 25 NS ESVI (ml/m2) Rest 58 :+: 30 50 :+: 18 NS Exercise 71 :+: 32* 64 :+: 24* NS EF(%) Rest 54 :+: 9 55 :+: 9 NS Exercise 47 :+: 10* 50 :+: 14* NS Regional de novo hypokinesia 25/26 22/26 NS

Values expressed as mean:+: SO. Symbols and abbreviations as in Table 2. significantly between the groups, averaging 82 ± 30 and and 3). In none of the patients was global hypokinesia noted. 98 ± 55%, respectively, of the value at rest for Groups I The de novo contraction abnormalities during exercise as and III (without infarction), and 97 ± 37 and 80 ± 40% assessed by analysis of the fractional shortening of the six of the value at rest for Groups II and IV (with infarction). hemiaxes in either of the two projections (see Methods) Correspondingly, the final values for heart rate and systolic involved I to 3 of the 12 analyzed segments in 3 and 5 left ventricular pressure achieved during exercise and their patients, respectively, in Groups I and III (without infarc• increase from rest to exercise were nearly the same in the tion) and in 17 and 8 patients in Groups II and IV (with respective study and control groups. The end-diastolic left infarction). Four to six segments were hypokinetic in 5 and ventricular pressure increased significantly in all four groups, 3 patients, respectively, in Groups I and III and in 6 and reaching 36 ± 11 and 29 ± 9 mm Hg, respectively, in II patients, respectively, in Groups II and IV. Finally, seven Groups I and III (without infarction) and 32 ± 13 and to nine hypokinetic segments were observed in nine patients; 36 ± 9 mm Hg in Groups II and IV (with infarction). two each in Groups I and III and two and three, respectively, Because end-diastolic and especially end-systolic left ven• in Groups II and IV. tricular volumes also increased similarly, left ventricular With re~pect to the appearance of regional contraction ejection fraction decreased to the same extent in the cor• abnormalities. the decrease in ejection fraction and the responding groups, that is, from 59 ± 7 to 50 ± 15% and increase in left ventricular end-diastolic pressure during 60 ± 4 to 52 ± 9%, respectively, in Groups I and III and exercise as indicators of ischemia, findings were similar in from 54 ± 9 to 47 ± 10 and 55 ± 9 to 50 ± 14% in individual patients whether or not these hemodynamic changes Groups II and IV. Although the values were significantly were accompanied by angina (Fig. 2 and 3). Thus, in the different within each group, they did not differ between the two groups without myocardial infarction all three findings respective study and control groups. were present in seven patients without exercise-induced an• Contraction abnormalities occurring during exercise, such gina (Group I) and eight patients with angina (Group III), as regional de novo hypokinesia, were detected in all patients two findings were present in two patients of each group and without myocardial infarction (Groups I and III; Fig. 1) and only de novo hypokinesia developed in one patient in Group 25 of 26 patients without angina (Group II) and in 22 of 26 I. In the two groups with myocardial infarction, all three patients with angina (Group IV) with infarction (Tables 2 findings were observed in 16 patients in Group II (without JACC Vol. 6, No.2 HIRZEL ET AL. 279 August 1985:275-84 HEMODYNAMIC CHANGES IN SILENT ISCHEMIA

Table 2. Hemodynamic Changes During Exercise in the 20 Patients Without Myocardial Infarction With and Without Exercise-Induced Angina Pectoris (AP) p (Group I Group I versus (without AP) Group III (with AP) Group III)

Work load attained (watts) 115 ± 31 94 ± 28 NS HR (beats/min) Rest 77± 12 72± 13 NS Exercise 126 ± 19* 115 ± 16* NS L VEOP (mm Hg) Rest 18 ± 4 13 ± 5 <0.02 Exercise 36 ± 11* 29 ± 9* NS LVSP (mm Hg) Rest 138 ± 21 141 ± 18 NS Exercise 152 ± 19* 168 ± 25* NS Rate-pressure product (beats/min x mm Hg) Rest 10,499 ± 1,410 10,208 ± 2,879 NS Exercise 19,508 ± 4,828* 19,247 ± 4,117* NS % increase 82 ± 30 98 ± 55 NS EOVI (mllm2) Rest 100 ± 23 100 ± 18 NS Exercise 116 ± 18* 103 ± 14 NS ESVI (mllm2) Rest 44 ± 14 39 ± 7 NS Exercise 59 ± 25* 50 ± 13* NS EF(%) Rest 59 ± 7 60 ± 4 NS Exercise 50 ± 15* 52 ± 9* NS Regional de novo hypokinesia 10/10 10/10 NS Values are expressed as mean values ± SO. * P < 0.001 different from value at rest (paired and unpaired t tests, one-way analysis of variance). EOVI = left ventricular end-diastolic volume index; EF = left ventricular ejection fraction; ESVI = left ventricular end-systolic volume index; HR = heart rate; LVEOP = left ventricular end-diastolic pressure; LVSP = left ventricular systolic pressure.

nificant ST segment elevation in 1 patient in each of these the somewhat higher mean left ventricular end-diastolic two groups (Table 1). pressure in the angina-free patients without myocardial in• In further analysis of the risk profiles of the four groups, farction (Group I) and the higher mean heart rate in the no differences were detected. In the patients without myo• asymptomatic patients with infarction (Group II). The rate• cardial infarction (Groups I and III) no patient was diabetic. pressure product, as an estimate of basal oxygen consump• Moderate hypertension was present in two patients without tion, however, was almost identical in all four groups (Ta• exercise-induced angina and one patient with angina. Hy• bles 2 and 3). perlipidemia was found in five patients without angina and Hemodynamic and angiographic changes during ex• four with angina; serum cholesterol levels ranged from 6.7 ercise. At the stress test during catheterization, the external to 8.6 mmollliter. Seven patients in Group I and seven in work load tolerated was slightly higher in Groups I and II Group III were smokers ( > 20 cigarettes/day for more than (patients without exercise-induced angina) than in Groups 2 years). In Groups II and IV (patients with myocardial III and IV (patients with angina), averaging 115 ± 31 and infarction), two patients without and three with exercise• 94 ± 28 watts (NS), respectively, in Groups I and III induced angina had a modest elevation of glucose (patients without infarction) and 107 ± 35 and 99 ± 33 levels in the fasting state. In three patients without and four watts (NS) in Groups II and IV (patients with infarction). with angina, serum cholesterol levels were above 6.5 The rate-pressure products achieved during exercise, how• mmollliter. Two patients with and two without angina had ever, were almost identical (19,508 ± 4,828 and 19,247 ± hypertension. Thirteen patients without angina and 12 with 4,117 beats/min X mm Hg [NS] for the groups without angina were heavy smokers before myocardial infarction prior infarction and 19,665 ± 3,950 and 17,701 ± 4,600 had occurred. beats/min x mm Hg [NS] for the groups with prior in• Baseline at rest. Baseline hemody• farction, respectively). Moreover, the relative increases in namic values at rest were also nearly identical except for the rate-pressure product from rest to exercise did not differ lACC Vol. 6, No.2 HIRZEL ET AL. 281 August 1985:275-84 HEMODYNAMIC CHANGES IN SILENT ISCHEMIA

20 Figure 2. Clinical, hemodynamic and angio• graphic characteristics during exercise in Groups I ANGINA and III, 20 patients with proven coronary artery II ~ disease but without prior myocardial infarction. Ten I ------I patients (Group III) were limited by angina, 10 (Group DE NOVO HYPOKINESIA /~ ~ I) tolerated the exercise test without anginal symp• toms. Both patient groups were carefully matched FALL OF EF for extent and severity of coronary artery lesions. ~. ~ ~ No differences in prevalence of either regional de RISE OF LVEDP " novo hypokinesia, decrease of left ventricular ejec• .. b .b II CD tion fraction (EF) or increase in end-diastolic pres• .'0 sure (L VEDP) were noted between the two groups Group III Group I during exercise at comparable external work loads (with AP) (without AP) and rate-pressure products. l1li FINDING PRESENT angina) and 16 patients in Group IV (with angina), whereas The prevalence of significant coronary artery lesions in two of these indicators were present in 8 patients in Group asymptomatic patients has been estimated by various in• II and 6 patients in Group IV; only one indicator was present vestigators (34-36) to be roughly 4.5% on the basis of in 2 patients in Group II and in 4 patients in Group IV. arteriographic and postmortem findings in large groups of These differences between subgroups were not statistically patients who were either asymptomatic or not known to significant. have had coronary heart disease before death. However, it remains an unresolved issue whether this group of asymp• tomatic patients exhibited attacks of so-called silent ischemia. Even though the degree of coronary luminal obstruction Discussion that may lead to compromised myocardial perfusion is the• Markers and determinants of ischemia. The existence oretically well established (37), several additional factors of silent myocardial ischemia has been postulated for several influence the significance of a stenosis. Aside from the driv• years to explain asymptomatic episodes of ST segment changes ing pressure, the length (38), shape and dynamic properties in patients with coronary artery disease (1-8). Our knowl• of the stenosis, the magnitude of the supplied muscle mass, edge remains sparse, however, about the patients prone to the autoregulatory capacity of the vascular bed and the num• development of such attacks and the underlying pathophy• ber and size of functioning collateral vessels (39) play an siologic mechanisms. Moreover, it is still unclear whether important role. Thus, in a given anatomic setting the an• ischemia without angina is related to the degree or special giographic appearance of a stenosis cannot be viewed as the dynamic properties of coronary artery lesions, or both, and only indicator of its significance in relation to myocardial to what extent it is accompanied by hemodynamic alterations ischemia. More precise information is derived from isch• characteristic of ischemic myocardium. emia-provoking tests such as dynamic exercise.

Figure 3. Clinical, hemodynamic and angiographic characteristics during ex• ercise in Groups II and IV, 52 patients with proven coronary artery disease and previous myocardial infarction. Twenty• ANGINA six patients (Group IV) developed an• gina during the stress test and 26 pa• DE NOVO tients (Group II) tolerated the exercise HYPOKINESIA test without anginal symptoms. Both pa• tient groups were carefully matched for FALL OF EF extent and severity of coronary artery lesions and location of infarction. No RISE OF LVEDP differences in prevalence of either re• gional de novo hypokinesia, decrease of Group I V Group I I left ventricular ejection fraction (EF) or (with AP) (without AP) increase in end-diastolic pressure (L VEDP) were noted between the two groups during exercise at comparable • FINDING PRESENT external work loads and rate-pressure products. 282 HIRZEL ET AL. lACC Vol. 6. No.2 HEMODYNAMIC CHANGES IN SILENT ISCHEMIA August 1985:275-84

Thus, the development of contraction abnormalities lead• the different groups suggest that the internal work load of ing to a decrease in ejection fraction and the increase in left the heart may have been much the same. Thus, our study ventricular end-diastolic pressure during exercise clearly in• further demonstrates that in the absence of angina the hemo• dicate left ventricular dysfunction. Such changes, however, dynamic alterations during exertion may be of a magnitude are observed in a variety of pathologic conditions and do similar to that of changes accompanied by angina. There• not necessarily imply that ischemia is the underlying path• fore, angina pectoris cannot be considered a prerequisite for ophysiologic cause. In patients who have valvular insuffi• hemodynamically significant ischemia during exertion in ciency or of various origins with normal patients with coronary artery disease. coronary , it is tempting to attribute this phenomenon Similar conclusions were reached by Cecchi et al. (40) to myocardial failure due to either increased fibrosis and Chierchia and coworkers (41). Using long-term am• or disease-related biochemical changes at the cellular level bulatory Holter monitoring alone (40) or combined with left than to inadequate blood flow, especially because the con• ventricular or pUlmonary and systemic arterial pressure mea• traction abnormalities are uniform. The exercise-induced surements (41), these investigator~ observed spontaneous appearance of regional wall motion abnormalities in patients asymptomatic ischemic episodes similar in duration and se• with significant coronary artery lesions, however, may in• verity in patients with either effort-induced angina or angina dicate more specifically localized impairment of myocardial at rest. This finding again favors the view that neither du• perfusion. ration nor severity of ischemia are the only factors respon• Periods without angina in the course of coronary ar• sible for the genesis of anginal pain. In addition, these tery disease. The present study focuses on 36 (14%) of studies do~ument that ischemic attacks with angina may 258 consecutive patients with severe coronary artery lesions occur side by side with attacks without anginq in the same who tolerated a dynamic exercise test during cardiac cath• patient. eterization without angina pectoris despite pronounced Possible pathophysiologic mechanisms of painless hemodynamic changes indicative of exertion-induced isch• ischemic episodes. There is uncertainty about the patho• emia. Group I comprised J0 (7.9%) of 127 patients in whom physiologic mechanisms underlying this lack of sensitivity neither anamnestic, clinical nor angiographic data suggested to pain, and it is tempting to attribute the phenomenon of the presence of previous myocardial infarction and Group silent ischemia to individual differences in pain threshold, II comprised 26 (19.8%) of 131 patients who had experi• as suggested by experimental pain measurements (42). In enced such an event as proven by the existence of an akinetic patients with prior myocardial infarction, the morphologic region in the left ventriculogram. All patients, however, destruction of nociceptive pathways could render an alter• were limited by angina at the time the decision for invasive native explanation for lack of pain sensations in the ischemic evaluation was made but became spontaneously angina-free state. However, observations made after aortocoronary by• thereafter. This observation suggests that: I) the incidence pass surgery indicate that only subsequent to extremely rad• of ischemic attacks without angina may not be that small ical denervation procedures may measurable changes in ex• because 14% of our patient group exhibited such an event, perienced pain occur (43). In addition, such effects would and 2) that such attacks seem to occur significantly more be expected to be present in the control group as well. frequently (p < 0.0 I) in patients after myocardial infarction. Finally, it might be argued that differences in risk profile Moreover, it documents that the course of clinically manifest account for variations in the perceptibility of ischemic pain. coronary heart disease is by no means uniform and may be In our study no differences were detectable between patients interrupted at times by periods without angina despite con• with and without exercise-induced angina with regard to the tinued silent ischemic attacks during exertion. Therefore, prevalence of hypertension, high serum cholesterol levels the spontaneous disappearance of anginal pain in a patient and smoking habits. Among patients without angina, dia• with coronary heart disease has to be interpreted with caution betes was excluded in all 10 patients without prior myo• because it may not necessarily reflect true improvement in cardial infarction, whereas modest elevations of blood glu• terms of disappearance of myocardial ischemia. cose levels in the fasting state were found in only 2 of the In comparing the hemodynamic changes that occurred in 26 patients with prior infarction. On the basis of anamnestic these two groups without angina withthe changes observed data, none of the 36 patients without angina had evidence in two carefully matched control groups of patients who of excessive consumption thus ruling out polyneu• developed angina at nearly identical exercise conditions ropathy as a possible cause of lowered pain threshold in this (Groups III and IV), no significant differences were de• patient cohort. tected. Although a similar external work load is not nec• Thus, even though the present study offers no explanation essarily indicative of similar myocardial oxygen demand, for the altered pathophysiologic mechanisms operating in the comparable rate-pressure product as an estimate of myo• silent forms of myocardial ischemia, it confirms the exis• cardial oxygen consumption achieved during exercise by tence of such events in an appreciable proportion of patients JACC Vol. 6, No.2 HIRZEL ET AL. 283 August 1985:275-84 HEMODYNAMIC CHANGES IN SILENT ISCHEMIA

with coronary heart disease and, more important, it docu• 20. Dwyer EM. Left ventricular pressure-volume alterations and regional ments that silent ischemic attacks may be accompanied by disorders of contraction during myocardial ischemia induced by atrial pacing. Circulation 1970;42: 1111-22. hemodynamic changes equal in severity to those observed 21. Amende I. Coltart DJ, Krayenbuehl HP. Rutishauser W. Left ven• in patients who experience angina. tricular contraction and relaxation in patients with coronary heart dis• ease. Eur J Cardiol 1975;3:37-45. We are grateful to Joerg Grimm. PhD for his help in performing the 22. Hess OM, Goebel NH, Gruentzig AR, Krayenbuehl HP. Linksven• statistical evaluation of the data. trikuliire Funktion bei Patienten mit koronarer Herzkrankheit vor und wahrend Ergometrie. Schweiz Med Wochenschr 1978;108:1726-8. 23. Carroll 10, Hess OM. Studer NP. Hirzel HO. Krayenbuehl HP. Sys• tolic function during exercise in patients with coronary artery disease. References J Am Coli Cardiol 1983;2:206-16. I. Froelicher VF. Yanowitz FG. Thompson AJ. The correlation of coro• 24. Tebbe U, Hoffmeister N. Sauer G, Neuhaus KL. Kreuzer H. Changes nary angiography and the electrocardiographic response to maximal in left ventricular diastolic function in coronary artery disease with treadmill testing in 76 asymptomatic men. Circulation 1973;48:597-604. and without angina pectoris assessed from exercise ventriculography. 2. Gettes LS. Painless myocardial ischemia. Chest 1974;66:612-3. Clin Cardiol 1980;3:19-25. 3. Froelicher VF. Thompson AJ. Longo MR Jr. Triebwasser JH. Lan• 25. Iskandrian AS. Segal BL. Anderson GS. Asymptomatic myocardial caster Me. Value of exercise testing for screening asymptomatic men ischemia. Arch Intern Med 1981; 141 :95-7. for latent coronary artery disease. Prog Cardiovasc Dis 1976;18:265-76. 26. Uhl GS. Kay TN, Hickman JR Jr. Comparison of exerciseradionuclide 4. Cohn PF. Severe asymptomatic coronary artery disease. A diagnostic. angiography and thallium perfusion imaging in detecting coronary prognostic and therapeutic puzzle. Am J Med 1977;62:565-8. disease in asymptomatic men. J Cardiac Rehab 1982;2: 118-24. 5. Weiner DA. Ryan TJ. McCabe CH. et al. Exercise stress testing. 27. Cohn PF. Brown EJ Jr. Wynne J. Holman L, Atkins HL. Global and Correlation among history of angina. ST segment response and prev• regional left ventricular ejection fraction abnormalities during exercise alence of coronary artery disease in the Coronary Artery Surgery Study in patients with silent myocardial ischemia. J Am Coli Cardiol (CASS). N Engl J Med 1979;301:230-5. 1983;1:931-3. 6. Langou RA. Huang EK. Kelley MJ. Cohen LS. Predictive accuracy of coronary artery calcification and abnormal exercise test for coronary 28. Bruce RA. Exercise testing of patients with coronary heart disease. artery disease in asymptomatic men. Circulation 1980;62:1196-203. Ann Clin Res 1971;3:323-32. 7. Berman DS. Amsterdam EA. Joye JA. et al. Thallium-201 stress 29. Krayenbuehl HP. Dyspnoe infolge Erkrankungen des Herzens. In: myocardial scintigraphy: application in asymptomatic patients with Siegenthaler W. ed. Differentialdiagnose innerer Krankheiten. Stutt• positive exercise electrocardiograms (abstr). Am J Cardiol 1978;41 :380. gart. New York: Georg Thieme. 1984:10-12. 8. Deanfield JE. Maseri A. Selwyn AP. et al. Myocardial ischemia during 30. Dodge HT, Sandler H. Ballew DW, Lord 10 Jr. The use of biplane daily life in patients with stable angina: its relation to symptoms and angiocardiography for the measurement of left ventricular volume in heart rate changes. Lancet 1983;2:753-8. man. Am Heart J 1960;60:762-76. 9. Stem S. Tzivoni D. Dynamic changeS in the ST-T segment during 31. Dodge HT, Sandler H. Baxley WA. Hawley RR. Usefulness and sleep in ischemic heart disease. Am J Cardiol 1973;32: 17-20. limitations of radiographic methods for determining left ventricular 10. Schang SJ Jr. Pepine CJ. Transient asymptomatic ST-segment depres• volume. Am J Cardiol 1966;18:10-24. sion during 4aily activity. Am J Cardiol 1977;39:396-402. 32. Levine JA, Bechtel OJ. Cohn PF. et al. Ventricular function before II. Selwyn AP. Fox K. Eves M. Oakley D. Dargie H. Shillingford JP. and after direct surgery. Circulation 1975;51: 1071-8. Myocardial ischemia in patients with frequent angina pectoris. Br Med 33. Herman MV, Heinle RA, Klein MD, Gorlin R. localized disorders J 1978;2: 1594-6. in myocardial contraction. Asynergy and its role in congestive heart 12. Lynch P. Dargie H, Krikler S, Krikler D. Objective assessment of failure. N Engl J Med 1967;277:222-32. antianginal treatment: a double-bli~d comparison of propranolol. ni• 34. Gensini GG. Kelly AE. Incidence and progression of coronary artery fedipine. and their combination. Br Med J 1980;281: 184-7. disease: an angiographic correlation in 1,263 patients. Arch Intern 13. Balasubramanian V. Lahiri A. Green HL. Stott FD. Raftery EB. Med 1972; 129:814-27. Ambulatory ST segment monitoring. Problems. pitfalls. solutions and clinical applications. Br Heart J 1980;44:419-25. 35. Erikssen J. Enge I. Forfang K, Storstein O. False positive diagnostic tests and coronary angiographic findings in 105 presumably healthy 14. Biagini A. Mazzei MG. Carpeggiani C. et al. Vasospastic ischemic males. Circulation 1976;54:371-6. mechanism of frequent asymptomatic transient ST-T changes during continuous electrocardiographic monitoring in selected 36. Diamond GA. Forrester JS. Analysis of probability as an aid in the patients. Am Heart J 1983;103:13-20. clinical diagnosis of coronary artery disease. N Engl J Med 1979;300: 1350-8. 15. Miiller 0, Rorvik K. Hemodynamic consequences of coronary heart disease. Br Heart J 1958;20:302-10. 37. Santamore WP. Bove AA. Alterations in the severity of coronary stenosis: effects of intraluminal pressure and proximal coronary artery 16. Cohen LS. Elliott WS. Rolett El. Gorlin R. Hemodynamic studies . In: Santamore WP, Bove AA, eds. Coronary Artery during angina pectoris. Circulation 1965;31 :409-16. Disease. Baltimore, Munich: Urban & Schwarzenberg, 1982:157-72. l7. Parker JO. DiGiorgi S. West RO. Hemodynamic study of coro• nary insufficiency precipitated by exercise. Am J Cardiol 38. Feldman RL, Nichols WW, Pepine CJ. Conti CR. Hemodynamic 1966;l7:470-83. significance of the length of a coronary arterial narrowing. Am J Cardiol 1978;41 :865-71. 18. Wiener L, Dwyer EM Jr. Cox JW. Left ventricular hemodynamics in exercise-induced angina pectoris. Circulation 1968;38:240-9. 39. Cohn PF, Maddox DE, Holman BL. See JR. Effect of coronary col• lateral vessels on regional myocardial blood flow in patients with 19. Lichtlen PR, Baumann PC. Albert H. The role of left ventricular coronary artery disease. Am J Cardiol 1980;46:359-64. abnormalities in exercise-induced performance in patients with severe coronary artery disease. Cardiologia 1969;54:295-319. 40. Cecchi AC. Dovellini EV, Marchi F. Pucci P. Santoro GM, Fazzini 284 HIRZEL ET AL. JACC Vol. 6, No.2 HEMODYNAMIC CHANGES IN SILENT ISCHEMIA August 1985:275-84

PF. Silent myocardial ischemia during ambulatory electrocardio• 42. Droste C, Roskamm H. Experimental pain measurements in patients graphic monitoring in patients with effort angina. J Am Coli Cardiol with asymptomatic myocardial ischemia. J Am Coli Cardiol 1983;1 :934-9. 1983;1:940-5. 41. Chierchia S, Lazzari M, Freedman MB, Brunelli C, Maseri A. Im• 43. Benrand ME, Lablanche JM, Rousseau MF, Warembourg HH, Stan• pairment of myocardial perfusion and function during painless myo• kowtak C, Soots G. Surgical treatment of variant angina: use of pIe x• cardial ischemia. J Am Coli Cardiol 1983; I :924-30. ectomy with aortocoronary bypass. Circulation 1980;61 :877-82.