Routine Exercise Testing to Detect Coronary Artery Disease in Patients with Atherosclerotic Vascular Disease F.J.J. van der Watt, MBChB, P.J. Jordan, MMed, C.J.C. Nel, MMed, A. Travers, MBChB, Bloemfontein, South Africa
The ability of patients with peripheral vascular disease to perform exercise studies on a conventional treadmill is often hampered by claudication, amputation, ischemic ulceration or rest pain. This study was undertaken to investigate the use of the arm ergometer in these patients. Eighty-three patients admitted with peripheral vascular or carotid artery disease were subjected to electrocardiographic-monitoredexercise testing, using both the arm ergometer and conventional treadmill, where possible. Coronary arteriography was performed consecutively on 32 of these patients to establish a control group from which the sensitivity, specificity and predictive accuracy of both methods of exercise testing could be calculated. Nineteen of the 70 arm ergometry tests and 22 of the 48 treadmill tests were positive. Nineteen of the patients with a positive test using either of the methods were asymptomatic for cardiac disease. All five patients who developed cardiac events during surgery had positive exercise tests, preoperatively. The sensitivity of arm ergometry in detecting coronary artery disease was 45.5% and the specificity loo%, while the figures for treadmill testing were 82.4% and 83.3%, respectively. The combined sensitivity for the two tests was 81.8% and the specificity 87.5%. Using a combination of these two tests thus provided a highly specific and adequately sensitive means of detecting the presence of coronary artery disease in patients presenting for peripheral vascular surgery. (Ann Vasc Surg 1990;4:47!3-484). KEY WORDS: Asymptomatic coronary artery disease; exercise testing; atherosclerotic vascular disease and exercise testing.
Complications related to coronary artery disease (CAD) continue to be the most common cause of early and late postoperative morbidity and mortality among patients undergoing major vascular surgery [ 1-31. There is little difficulty in identifying the patient with symptomatic CAD. However, routine coronary arteriography From the Department of Surgery and Dr. Bill Venter Unit for Vascular Surgely, Universitas Hospital, and the Department of Cardiology, Universitas Hospital, Bloemfontein, South Africa. Reprint requests: Professor C.J.C. Nel, Department of Surgery (G72), Medical Faculty, University of the Orange Free State, P.O. Box 339, Bloemfontein 9301, South Africa. 479
in vascular patients has shown that the incidence of sigmficant but asymptomatic CAD may be as hgh as 30% [4]. While some authors suggest routine preoperative coronary arteriography in all patients before vascular surgery [5],others have emphasized noninvasive tests as the first step in identifying high-risk patients [6,7]. The ECG-monitored treadmill stress test has been shown to be cost-effective and relatively reliable screening technique to detect associated CAD in patients with peripheral vascular disease (PVD) [8,9]. However, the ability of patients with PVD to participate in exercise studies on a conventional treadmill is often hampered by claudication, amputation, ischemic ulceration or severe rest pain.
480
ANNALS OF VASCULAR SURGERY
ROUTINE EXERCISE TESTING TO DETECT CAD
To overcome these problems, arm ergometry has been suggested as a valuable alternative way of stress testing before vascular surgery 18,101. The aim of this study was to compare the applicability and diagnostic value of treadmill stress and arm ergometry in detecting associated CAD in patients with PVD, carotid artery disease and aneurysmal disease.
TABLE I.-Presenting diagnoses for patients in both groups Carotid artery disease Peripheral vascular disease Aneurysmal disease Total
Group A 13 57 13 83
Group B 4 22 6
32
RESULTS PATIENTS AND METHODS Group A
Eighty-three consecutive patients admitted to the Department of Surgery, Universitas Hospital, Bloemfontein, with either atherosclerotic occlusive disease of the extremities, carotid artery disease or aneurysms were included in the study. Patients were divided into two groups. Group A consisted of 83 patients that performed both methods of exercise where possible. Group B is a subgroup of Group A and consisted of 32 patients on whom coronary arteriography was performed to establish a control group from which the sensitivity, specificity and predictive accuracy of both exercise methods could be evaluated. All treadmill studies were performed on a programmed Burdick T5OO treadmill according to the Bruce protocol [ 1 I]. An adaptable Wurburg ergometer and the protocol suggested by Williams [I21 were used for upper extremity exercise testing. Patients with ischemic rest pain, ischemic ulceration or severe restricting claudication were exercised on the arm ergometer only. Continuous 12 lead electrocardiographic monitoring was used during the period of exercise. Resting ECGs were recorded in a supine and standing position before and after completion of exercise. Exercise was terminated in the event of claudication, dyspnea, fatigue, angina, ventricular dysrhythmia, hypotension or marked ischemic changes on electrocardiography. All ECGs were interpreted by a cardiologist familiar with exercise testing but without prior knowledge of the patient's clinical condition. Tests were taken as positive in the presence of new or additional horizontal or downsloping S-T segment depression of 1 mm or more, fall in systolic blood pressure of 10 mrnHg or more and exercise-induced angina or complex ventricular arrhythmias. Coronary arteriography was performed on patients in Group B employing the transcutaneous femoral Seldinger technique. Obstruction of more than SO% in at least one of the major coronary arteries was considered as being hemodynamically significant.
Patients in this group had a mean age of 62 years (36-84 years) and included 67 men and 16 women. The presenting diagnoses are shown in Table I. Seven of the 83 patients included in the study could not perform any exercise, two because of previous strokes and five due to general debilitation. A further three patients presented with a left bundle branch block on resting ECG, making interpretation of the exercise electrocardiogram unreliable. This means that 12% of patients considered for vascular surgery in this study could not be considered for exercise testing. A further 25 patients (34.2%) could not be considered for treadmill testing, 24 due to severe claudication and one because of debilitating arthritis of the hip joints. The three patients that were unable to perform arm ergometry had severe arterial insufficiency of the upper extremities (Fig. 1). Exercise studies on the treadmill were terminated prematurely as a result of claudication in 21 patients, fatigue in l l patients, complex ventricular dysrhythmias in one and asthma in another. The main reason for premature cessation of arm ergometry was fatigue of the arms, 38 tests being terminated for this reason. Only two women (16.7%) participating in arm ergometry managed to reach
Patients
25
"
--
No exercise
A€
+
TM
Only AE
Only TM
LBBB
Method of exercise
AE Arm ergometry TM trsldmill leltinp L88S lell bundle branch block
Fig. 1. Applicability of treadmill testing and arm erga metry in vascular patients.
VOLUME4
48 1
ROUTINE EXERCISE TESTING TO DETECT CAD
NO 5 - 1990
Fig. 3. Clinical status of patients with positive exercise tests.
Fig. 2. Reasons for termlnation of exercise.
85% of their target heart rate, mainly due to fatigue of the arms. One patient developed angina and another asthma (Fig. 2). Fourteen (29.2%) of the 48 patients that performed treadmill testing managed to achieve 85% of their age-related maximum predicted heart rate, indicating that only 16.9% of patients under consideration for vascular surgery can reach 85% of their maximum predicted heart rate during treadmill testing. Thirty (42.9%) of the 70 patients that performed upper extremity stress testing reached 85% of their maximum predicted heart rate. Seen globally, 42.2% of all vascular patients thus reached 85% of their maximum predicted heart rate with one of the two methods (Table 11). Nineteen of the 70 arm ergometry tests were positive (27.1%), while 22 out of 48 treadmill tests were positive (45.8%). Thirty patients (41.1%) had at least one positive exercise test (Table 11). Thirteen patients on the treadmill and 10 with the arm ergometer had positive stress tests without reaching 85% of their target heart rate. A positive treadmill test with normal arm ergometry was seen in 11 patients. Nineteen (63.3%) of the patients with a positive stress test were totally asymptomatic for cardiac disease, while 12 of these 19 patients had signs of a previous myocardial infarction or ischeTABLE 11.-Summary
of exercise tests performed
mia on their resting ECG, showing that seven patients (23.3%) with no clinical or electrocardiographical stigmata of CAD presented with a positive exercise test (Fig. 3). Five perioperative cardiac events were documented in 49 patients eventually undergoing vascular surgery (Table 111). These included complex and resistant ventricular arrhythmias in two patients, ventricular fibrillation in one and lung edema in two patients. These five patients all had a positive stress test, while only two of them managed to reach 85% of their maximum predicted heart rate. Only one of these patients was able to exercise for longer than six minutes.
Group B
Patients in this group had a mean age of 59 (36-69) years and included 28 men and four women. The results of the two exercise tests for these 32 patients are presented in Table IV. Two patients showed signs of left bundle branch block on resting ECG and were not included in the final analysis. Coronary arteriographic findings are shown in Table V. Two of the patients classified as having normal coronary arteries had previous coronary angioplasties following myocardial infarctions. Twelve patients were referred for coronary bypass surgery before peripheral revascularization on grounds of coronary arteriographic findings, while another two patients underwent coronary bypass following peripheral vascular surgery. These two patients had bilateral high-grade carotid artery ste-
85% of
Tests performed Treadmill testing 48 Arm ergometry 70 Total tests 118 Total patients 73 '35183 = 42.2% of all vascular patients '30/83 = 36.1% of all vascular patients
target heart rate reached 14 (29.2%) 30 (42.9%) 44 (37.3%) 35* (47.9%)
Positive tests 22 (45.8%) 19 (27.1%) 41 (34.7%) 30t (41.lo/o)
TABLE Ill.-Perioperative
cardiac events
Patients operated Patients with cardiac complications Positive exercise test 85% of target heart rate reached Exercise time >6 minutes
49 5
5
2 1
482
ROUTINE EXERCISE TESTING TO DETECT CAD
TABLE 1V.-Accuracy
ANNALSOF VASCULAR SURGERY
of treadmill testing and arm ergometry to detect coronary artery &$ease in vascular patients compared with coronary arteriography TreadmiII THR* 10 8 7 1 1 8 2 5 2 3 0 17 7 82.4% 100% 83.3% 66.7% 93.3% 87.5% 62.5% 100% 82.6% 90.0% All 23 15 14
Total All positive True positive False positive All negative True negative False negative Coronary artery disease positive Sensitivity Specificity Positive predictive value Negative predictive value Predictive accuracy
Arm ergometry Al I TH R 29 15 10 6 10 6 0 0 19 9 7 4 5 12 22 11 45.5% 54.5% 100% 100% 100% 100% 36.9% 44.4% 58.6% 66.7%
Combination A1I THR 30 17 19 11 18 10 1 1 11 6 7 4 4 2 22 12 81.8% 83.3% 87.5% 80.0% 94.7% 90.9% 63.6% 66.7% 83.3% 82.4%
'(Target heart rate) 85% of maximum predicted heart rate reached
nosis and a symptomatic abdominal aortic aneurysm respectively. The postoperative course of the latter was complicated by ventricular fibrillation on the third postoperative day. The other two patients in this group who had postoperative cardiac complications were diagnosed preoperatively as having diffuse nonoperable CAD. Eleven of the 14 patients referred for coronary bypass surgery were totally asymptomatic (Table V), while six of the 14 had signs of a previous myocardial infarction or ischemia on resting ECG.
DISCUSSION DeBakey and coworkers [I31 were among the first to point out the frequent association of PVD and CAD. In an analysis of 5,000 patients with peripheral atherosclerosis, 25% of patients with occlusive disease and 29% with abdominal aortic aneurysms (AAA) were found to have significant CAD. This was followed by numerous other studies documenting the high incidence of myocardial complications in patients undergoing vascular reconstructive procedures 114-171.
TABLE V.-Classification of coronary arteriographic findings in relation to coronary artery disease symptoms NO.
Three-vessel disease Two-vessel disease Single-vessel disease Normal Total
8 10 6 8 32
Asym tomatPc 3 8 4
Acce ted tor C ~ G 6 6 2
6 15+
0 14
'Coronary artery bypass grafting surge?, +I5124 = 62.5% of patients with significant coronary artery disease were asymptomatic
Certainly one of the most important findings that emerged from our study is the high percentage of asymptomatic patients with significant and in many cases life-threatening CAD (Table V). In a series of 1,000 consecutive patients studied by Hertzer and colleagues [ 5 ] , CAD was clinically suspected in 52% of 263 patients presenting with AAA, 56% of 381 with lower limb ischemia and 57% of 169 patients with cerebrovascular disease. However, when coronary arteriography was performed on all 1,000 patients, only 8% were found to have normal coronary arteries. Severe but correctable CAD representing a high risk of myocardial infarction was found in 25% of patients. Following a policy of routine preoperative coronary arteriography before vascular surgery, early encouraging reports showed low mortality rates in patients having elective aortic reconstruction following myocardial revascularization [18]. This highlights the importance of an effective preoperative diagnostic algorithm for the detection of asymptomatic CAD in vascular patients. Due to inherent risks and high cost of the procedure, the policy of routine preoperative coronary arteriography has been questioned [7,10]. The selective application of coronary arteriography in patients with clinically suspected CAD or positive noninvasive screening tests may be a safer and more cost-effective means of lowering peri- and postoperative mortality. ECG-monitored exercise testing has been proposed as a relatively inexpen* sive, easily applicable means of screening for asymptomatic CAD in patients presenting for major vascular surgery [8,9,19]. The results of this study confirm the fact that in excess of 10% of patients considered for vascular surgery cannot perform any exercise (Fig. 1). As a result of their high general operative risk, only one of these patients was eventually operated upon. Less than 20% (14/83) of
VOLUME 4 SO 5 - 1990
ROUTINE EXERCISE TESTING TO DETECT C A D
vascular patients reached 85% of their predicted heart rate on a treadmill, while 40% (35/83) were able to exercise satisfactorily using at least one of the described niethods (Table 11). Despite the frequent limitation of exercise by claudication (Fig. 2), 45.8% of patients participating in treadmill tests in the present study developed electrocardiographic evidence of myocardial ischemia (Table 11). This represents 26.5% of all patients included in the study and compares well with figures found in other studies (29% [ 191-38% [8,9]). The frequent inability of vascular patients to participate in treadmill exercise has led to some investigators proposing arm ergometry as an alternative to treadmill testing [8,10]. Although arm ergometry is less sensitive than leg ergometry in eliciting exercise-induced ischemic and arrhythmic abnormalities [22], the addition of arm ergometry to a preoperative algorithm, using treadmill testing and exercise radionuclide ventriculography increased the predictive capacity of the algorithm from 64% to 89% [lo]. The combined treadmill stress plus arm ergometry alone had a predictive accuracy of 74%. The sensitivity of arm ergometry to elicit ischemic ECG changes in the present study was disappointingly low (45.57eTable IV) and the value of a negative stress test during arm ergometry in a patient who cannot exercise properly on the treadmill will have to be investigated again. Despite the fact that only 16% of vascular patients can reach 85% of their target heart rate on the treadmill, it must be regarded as the method of choice when evaluating vascular patients for associated CAD (Table IV). Of particular interest is the fact that the three patients with false negative treadmill tests did not reach 85% of their predicted heart rate. The patient with a false positive treadmill test on the other hand, was using a diuretic at the time of the test. Arm ergometry can be used in conjunction with or as a valuable alternative to treadmill testing in patients with a normal treadmill test who cannot reach a satisfactory heart rate. Proponents of exercise testing claim that an ischemic response during exercise is a useful predictor of perioperative cardiac complications [19,20] and that patients able to achieve more than 85% of their predicted maximum heart rate may represent a low risk group for major vascular surgery [21]. This was supported in the present study as all five patients that suffered perioperative cardiac events had a positive exercise test and only two of them managed to reach 85% of their maximum predicted heart rate (Table 111).The ability of exercise testing to identify patients with significant but asymptomatic CAD may be of even greater importance for their long-term prognosis. Using routine exercise testing enabled us to identify 18 of the 24 patients with asymptomatic but hemodynamically significant CAD (75%). Fourteen patients
483
were subsequently referred for coronary artery bypass surgery and in 12 cases this was done before the initially planned peripheral vascular procedure (Table V).
CONCLUSIONS Although many previous studies related the results of noninvasive tests to perioperative complications only [7-9,19-21], the long-term prognosis of patients also has to be taken into account [18]. ECG-monitored stress testing in general proved to be a fairly sensitive and very specific way of detecting associated asymptomatic CAD in patients for vascular surgery. We suggest that every patient under consideration for major vascular surgery should at least be subjected to an ECG-monitored exercise test. More expensive methods such as dipyridamole and exercise thallium scanning may have a place in patients who cannot reach a satisfactory heart rate with any of the described methods. Bearing in mind the systemic nature of atherosclerosis, the admission of a patient for vascular surgery must be seen as a golden opportunity to diagnose associated asymptomatic CAD.
ACKNOWLEDGMENTS This study was supported by the Medical Research Council of South Africa.
REFERENCES 1. HERTZER NR. Fatal myocardial infarction following abdominal aortic aneurysm resection. Ann Surg 1980;192:667673. 2. HERTZER NR, LEES CD. Fatal myocardial infarction following carotid endarterectomy Ann Surg 1981;194:212218. 3. HERTZER NR. Fatal myocardial infarction following lower extremity revascularization. Ann Surg 1981;193:492-498. 4. TOMATIS LA, FIERENS EE, VERBRUGGE GP. Evaluation of surgical risk in vascular disease by coronary arteriography: series of 100 cases. Surgery 1972;71:429-435. 5. HERTZER NR, BEVEN EG, YOUNG JR, et al. Coronary artery disease in peripheral vascular patients. A classification of 1000 angiograms and results of surgical management. Ann Surg 1984;199:223-233. 6. NICOLAIDES AN, SALMASI A, SONECHA TN. How should we investigate the arteriopath for coexisting lesions? J Cardiovasc Surg 1986;27515-533. 7. EAGLE KA, SINGER DE, BREWSTER DC, et al. Dipyridamole- thallium scanning in patients undergoing vascular surgery. Optimizing preoperative evaluation of cardiac risk. JA MA I987;2572 185-2 189. 8. CUTLER BS, WHEELER HB, PARASKOS JA, et al. Applicability and interpretation of electrocardiographic stress testing in patients with peripheral vascular disease. A m J Surg 1981;141:501-506. 9. CUTLER BS, WHEELER HB, PARASKOS JA, et al. Assessment of operative risk with electrocardiographic ex-
484
10.
11. 12.
13. 14.
IS. 16.
ROUTINE EXERCISE TESTING TO DETECT CAD
ercise testing in patients with peripheral vascular disease. A m J Surg 1979;137:484-490. YOUNGMAN DJ, DOVE T, BOCCUZZI SJ, et al. Noninvasive and angiographic evaluation of coronary artery disease in patients with peripheral vascular disease. Am J Cardiol 1989;63:14461449, BRUCE RA. Exercise testing of patients with coronary heart disease. Principles and normal standards for evaluation. Ann Clin Rest 1971;3:323-332. WILLIAMS J , COTTRELL E , POWERS SK, et al. Arm ergometry: a review of published protocols and the introduction of a new weight adjusted protocol. J Sports Med 198333:107-1 12. D E BAKEY ME, CRAWFORD ES, MORRIS GC Jr, et al. Late results of vascular surgery in the treatment of arteriosclerosis. J Cardiovasc Surg 1964;5:473480. CRAWFORD ES, BOMBERGER RA, GLAESER DH, et al. Aortoiliac occlusive disease: factors influencing survival and function following reconstructive operation over a twenty- five year period. Surgery 1981;90:10SS-1067. HOLLIER L H , PLATE G , O’BRIEN PC, et al. Late survival after abdominal aortic aneurysm repair: influence of coronary artery disease. J Vusc Surg 1984;1:29&299. JAMIESON WRE, JANCESZ TM, MIYAGISHIMA RT, et
17.
18.
19. 20. 21.
22.
ANNALSOF VASCULAR SURGERY
al. Influence of ischemic heart disease on early and late mortality after surgery for peripheral occlusive disease. Circulation 1982;66:92-97. THOMPSON JE, AUSTIN DJ, PATMAN RD. Carotid endarterectomy for cerebrovascular insufficiency: long term results in 592 patients followed up to thirteen years. Ann Surg 1970;172:663-679. HERTZER NR, YOUNG JR, BEVEN EG, et al. Late results of coronary bypass in patients with infrarenal aortic aneurysms. The Cleveland Clinic study. Ann Surg 1987305: 360-367. MC CABE CJ, REIDY NC, ABBOT WM, et al. The value of electrocardiogram monitoring during treadmill testing for peripheral vascular disease. Surgery 1981$9: 183-186. AROUS EJ, BAUM PL, CUTLER BS. The ischemic exercise test in patients with peripheral vascular disease. Arch Surg 1984;119:78&783. MC PHAIL N , CALVIN JE, SHERIATMANDAR A, et al. The use of preoperative exercise testing to predict cardiac complications after arterial reconstruction. J Vusc Surg 1986;7:60-68. SCHWADE J, BLOMQVIST CG, SHAPIRO W. A comparison of the response to arm and leg work in patients with ischemic heart disease. A m Heurt J 1977;94:203-208.
The ability of patients with peripheral vascular disease to perform exercise studies on a conventional treadmill is often hampered by claudication, amputation, ischemic ulceration or rest pain. This study was undertaken to investigate the use of the arm ergometer in these patients. Eighty-three patients admitted with peripheral vascular or carotid artery disease were subjected to electrocardiographic-monitoredexercise testing, using both the arm ergometer and conventional treadmill, where possible. Coronary arteriography was performed consecutively on 32 of these patients to establish a control group from which the sensitivity, specificity and predictive accuracy of both methods of exercise testing could be calculated. Nineteen of the 70 arm ergometry tests and 22 of the 48 treadmill tests were positive. Nineteen of the patients with a positive test using either of the methods were asymptomatic for cardiac disease. All five patients who developed cardiac events during surgery had positive exercise tests, preoperatively. The sensitivity of arm ergometry in detecting coronary artery disease was 45.5% and the specificity loo%, while the figures for treadmill testing were 82.4% and 83.3%, respectively. The combined sensitivity for the two tests was 81.8% and the specificity 87.5%. Using a combination of these two tests thus provided a highly specific and adequately sensitive means of detecting the presence of coronary artery disease in patients presenting for peripheral vascular surgery. (Ann Vasc Surg 1990;4:47!3-484). KEY WORDS: Asymptomatic coronary artery disease; exercise testing; atherosclerotic vascular disease and exercise testing.
Complications related to coronary artery disease (CAD) continue to be the most common cause of early and late postoperative morbidity and mortality among patients undergoing major vascular surgery [ 1-31. There is little difficulty in identifying the patient with symptomatic CAD. However, routine coronary arteriography From the Department of Surgery and Dr. Bill Venter Unit for Vascular Surgely, Universitas Hospital, and the Department of Cardiology, Universitas Hospital, Bloemfontein, South Africa. Reprint requests: Professor C.J.C. Nel, Department of Surgery (G72), Medical Faculty, University of the Orange Free State, P.O. Box 339, Bloemfontein 9301, South Africa. 479
in vascular patients has shown that the incidence of sigmficant but asymptomatic CAD may be as hgh as 30% [4]. While some authors suggest routine preoperative coronary arteriography in all patients before vascular surgery [5],others have emphasized noninvasive tests as the first step in identifying high-risk patients [6,7]. The ECG-monitored treadmill stress test has been shown to be cost-effective and relatively reliable screening technique to detect associated CAD in patients with peripheral vascular disease (PVD) [8,9]. However, the ability of patients with PVD to participate in exercise studies on a conventional treadmill is often hampered by claudication, amputation, ischemic ulceration or severe rest pain.
480
ANNALS OF VASCULAR SURGERY
ROUTINE EXERCISE TESTING TO DETECT CAD
To overcome these problems, arm ergometry has been suggested as a valuable alternative way of stress testing before vascular surgery 18,101. The aim of this study was to compare the applicability and diagnostic value of treadmill stress and arm ergometry in detecting associated CAD in patients with PVD, carotid artery disease and aneurysmal disease.
TABLE I.-Presenting diagnoses for patients in both groups Carotid artery disease Peripheral vascular disease Aneurysmal disease Total
Group A 13 57 13 83
Group B 4 22 6
32
RESULTS PATIENTS AND METHODS Group A
Eighty-three consecutive patients admitted to the Department of Surgery, Universitas Hospital, Bloemfontein, with either atherosclerotic occlusive disease of the extremities, carotid artery disease or aneurysms were included in the study. Patients were divided into two groups. Group A consisted of 83 patients that performed both methods of exercise where possible. Group B is a subgroup of Group A and consisted of 32 patients on whom coronary arteriography was performed to establish a control group from which the sensitivity, specificity and predictive accuracy of both exercise methods could be evaluated. All treadmill studies were performed on a programmed Burdick T5OO treadmill according to the Bruce protocol [ 1 I]. An adaptable Wurburg ergometer and the protocol suggested by Williams [I21 were used for upper extremity exercise testing. Patients with ischemic rest pain, ischemic ulceration or severe restricting claudication were exercised on the arm ergometer only. Continuous 12 lead electrocardiographic monitoring was used during the period of exercise. Resting ECGs were recorded in a supine and standing position before and after completion of exercise. Exercise was terminated in the event of claudication, dyspnea, fatigue, angina, ventricular dysrhythmia, hypotension or marked ischemic changes on electrocardiography. All ECGs were interpreted by a cardiologist familiar with exercise testing but without prior knowledge of the patient's clinical condition. Tests were taken as positive in the presence of new or additional horizontal or downsloping S-T segment depression of 1 mm or more, fall in systolic blood pressure of 10 mrnHg or more and exercise-induced angina or complex ventricular arrhythmias. Coronary arteriography was performed on patients in Group B employing the transcutaneous femoral Seldinger technique. Obstruction of more than SO% in at least one of the major coronary arteries was considered as being hemodynamically significant.
Patients in this group had a mean age of 62 years (36-84 years) and included 67 men and 16 women. The presenting diagnoses are shown in Table I. Seven of the 83 patients included in the study could not perform any exercise, two because of previous strokes and five due to general debilitation. A further three patients presented with a left bundle branch block on resting ECG, making interpretation of the exercise electrocardiogram unreliable. This means that 12% of patients considered for vascular surgery in this study could not be considered for exercise testing. A further 25 patients (34.2%) could not be considered for treadmill testing, 24 due to severe claudication and one because of debilitating arthritis of the hip joints. The three patients that were unable to perform arm ergometry had severe arterial insufficiency of the upper extremities (Fig. 1). Exercise studies on the treadmill were terminated prematurely as a result of claudication in 21 patients, fatigue in l l patients, complex ventricular dysrhythmias in one and asthma in another. The main reason for premature cessation of arm ergometry was fatigue of the arms, 38 tests being terminated for this reason. Only two women (16.7%) participating in arm ergometry managed to reach
Patients
25
"
--
No exercise
A€
+
TM
Only AE
Only TM
LBBB
Method of exercise
AE Arm ergometry TM trsldmill leltinp L88S lell bundle branch block
Fig. 1. Applicability of treadmill testing and arm erga metry in vascular patients.
VOLUME4
48 1
ROUTINE EXERCISE TESTING TO DETECT CAD
NO 5 - 1990
Fig. 3. Clinical status of patients with positive exercise tests.
Fig. 2. Reasons for termlnation of exercise.
85% of their target heart rate, mainly due to fatigue of the arms. One patient developed angina and another asthma (Fig. 2). Fourteen (29.2%) of the 48 patients that performed treadmill testing managed to achieve 85% of their age-related maximum predicted heart rate, indicating that only 16.9% of patients under consideration for vascular surgery can reach 85% of their maximum predicted heart rate during treadmill testing. Thirty (42.9%) of the 70 patients that performed upper extremity stress testing reached 85% of their maximum predicted heart rate. Seen globally, 42.2% of all vascular patients thus reached 85% of their maximum predicted heart rate with one of the two methods (Table 11). Nineteen of the 70 arm ergometry tests were positive (27.1%), while 22 out of 48 treadmill tests were positive (45.8%). Thirty patients (41.1%) had at least one positive exercise test (Table 11). Thirteen patients on the treadmill and 10 with the arm ergometer had positive stress tests without reaching 85% of their target heart rate. A positive treadmill test with normal arm ergometry was seen in 11 patients. Nineteen (63.3%) of the patients with a positive stress test were totally asymptomatic for cardiac disease, while 12 of these 19 patients had signs of a previous myocardial infarction or ischeTABLE 11.-Summary
of exercise tests performed
mia on their resting ECG, showing that seven patients (23.3%) with no clinical or electrocardiographical stigmata of CAD presented with a positive exercise test (Fig. 3). Five perioperative cardiac events were documented in 49 patients eventually undergoing vascular surgery (Table 111). These included complex and resistant ventricular arrhythmias in two patients, ventricular fibrillation in one and lung edema in two patients. These five patients all had a positive stress test, while only two of them managed to reach 85% of their maximum predicted heart rate. Only one of these patients was able to exercise for longer than six minutes.
Group B
Patients in this group had a mean age of 59 (36-69) years and included 28 men and four women. The results of the two exercise tests for these 32 patients are presented in Table IV. Two patients showed signs of left bundle branch block on resting ECG and were not included in the final analysis. Coronary arteriographic findings are shown in Table V. Two of the patients classified as having normal coronary arteries had previous coronary angioplasties following myocardial infarctions. Twelve patients were referred for coronary bypass surgery before peripheral revascularization on grounds of coronary arteriographic findings, while another two patients underwent coronary bypass following peripheral vascular surgery. These two patients had bilateral high-grade carotid artery ste-
85% of
Tests performed Treadmill testing 48 Arm ergometry 70 Total tests 118 Total patients 73 '35183 = 42.2% of all vascular patients '30/83 = 36.1% of all vascular patients
target heart rate reached 14 (29.2%) 30 (42.9%) 44 (37.3%) 35* (47.9%)
Positive tests 22 (45.8%) 19 (27.1%) 41 (34.7%) 30t (41.lo/o)
TABLE Ill.-Perioperative
cardiac events
Patients operated Patients with cardiac complications Positive exercise test 85% of target heart rate reached Exercise time >6 minutes
49 5
5
2 1
482
ROUTINE EXERCISE TESTING TO DETECT CAD
TABLE 1V.-Accuracy
ANNALSOF VASCULAR SURGERY
of treadmill testing and arm ergometry to detect coronary artery &$ease in vascular patients compared with coronary arteriography TreadmiII THR* 10 8 7 1 1 8 2 5 2 3 0 17 7 82.4% 100% 83.3% 66.7% 93.3% 87.5% 62.5% 100% 82.6% 90.0% All 23 15 14
Total All positive True positive False positive All negative True negative False negative Coronary artery disease positive Sensitivity Specificity Positive predictive value Negative predictive value Predictive accuracy
Arm ergometry Al I TH R 29 15 10 6 10 6 0 0 19 9 7 4 5 12 22 11 45.5% 54.5% 100% 100% 100% 100% 36.9% 44.4% 58.6% 66.7%
Combination A1I THR 30 17 19 11 18 10 1 1 11 6 7 4 4 2 22 12 81.8% 83.3% 87.5% 80.0% 94.7% 90.9% 63.6% 66.7% 83.3% 82.4%
'(Target heart rate) 85% of maximum predicted heart rate reached
nosis and a symptomatic abdominal aortic aneurysm respectively. The postoperative course of the latter was complicated by ventricular fibrillation on the third postoperative day. The other two patients in this group who had postoperative cardiac complications were diagnosed preoperatively as having diffuse nonoperable CAD. Eleven of the 14 patients referred for coronary bypass surgery were totally asymptomatic (Table V), while six of the 14 had signs of a previous myocardial infarction or ischemia on resting ECG.
DISCUSSION DeBakey and coworkers [I31 were among the first to point out the frequent association of PVD and CAD. In an analysis of 5,000 patients with peripheral atherosclerosis, 25% of patients with occlusive disease and 29% with abdominal aortic aneurysms (AAA) were found to have significant CAD. This was followed by numerous other studies documenting the high incidence of myocardial complications in patients undergoing vascular reconstructive procedures 114-171.
TABLE V.-Classification of coronary arteriographic findings in relation to coronary artery disease symptoms NO.
Three-vessel disease Two-vessel disease Single-vessel disease Normal Total
8 10 6 8 32
Asym tomatPc 3 8 4
Acce ted tor C ~ G 6 6 2
6 15+
0 14
'Coronary artery bypass grafting surge?, +I5124 = 62.5% of patients with significant coronary artery disease were asymptomatic
Certainly one of the most important findings that emerged from our study is the high percentage of asymptomatic patients with significant and in many cases life-threatening CAD (Table V). In a series of 1,000 consecutive patients studied by Hertzer and colleagues [ 5 ] , CAD was clinically suspected in 52% of 263 patients presenting with AAA, 56% of 381 with lower limb ischemia and 57% of 169 patients with cerebrovascular disease. However, when coronary arteriography was performed on all 1,000 patients, only 8% were found to have normal coronary arteries. Severe but correctable CAD representing a high risk of myocardial infarction was found in 25% of patients. Following a policy of routine preoperative coronary arteriography before vascular surgery, early encouraging reports showed low mortality rates in patients having elective aortic reconstruction following myocardial revascularization [18]. This highlights the importance of an effective preoperative diagnostic algorithm for the detection of asymptomatic CAD in vascular patients. Due to inherent risks and high cost of the procedure, the policy of routine preoperative coronary arteriography has been questioned [7,10]. The selective application of coronary arteriography in patients with clinically suspected CAD or positive noninvasive screening tests may be a safer and more cost-effective means of lowering peri- and postoperative mortality. ECG-monitored exercise testing has been proposed as a relatively inexpen* sive, easily applicable means of screening for asymptomatic CAD in patients presenting for major vascular surgery [8,9,19]. The results of this study confirm the fact that in excess of 10% of patients considered for vascular surgery cannot perform any exercise (Fig. 1). As a result of their high general operative risk, only one of these patients was eventually operated upon. Less than 20% (14/83) of
VOLUME 4 SO 5 - 1990
ROUTINE EXERCISE TESTING TO DETECT C A D
vascular patients reached 85% of their predicted heart rate on a treadmill, while 40% (35/83) were able to exercise satisfactorily using at least one of the described niethods (Table 11). Despite the frequent limitation of exercise by claudication (Fig. 2), 45.8% of patients participating in treadmill tests in the present study developed electrocardiographic evidence of myocardial ischemia (Table 11). This represents 26.5% of all patients included in the study and compares well with figures found in other studies (29% [ 191-38% [8,9]). The frequent inability of vascular patients to participate in treadmill exercise has led to some investigators proposing arm ergometry as an alternative to treadmill testing [8,10]. Although arm ergometry is less sensitive than leg ergometry in eliciting exercise-induced ischemic and arrhythmic abnormalities [22], the addition of arm ergometry to a preoperative algorithm, using treadmill testing and exercise radionuclide ventriculography increased the predictive capacity of the algorithm from 64% to 89% [lo]. The combined treadmill stress plus arm ergometry alone had a predictive accuracy of 74%. The sensitivity of arm ergometry to elicit ischemic ECG changes in the present study was disappointingly low (45.57eTable IV) and the value of a negative stress test during arm ergometry in a patient who cannot exercise properly on the treadmill will have to be investigated again. Despite the fact that only 16% of vascular patients can reach 85% of their target heart rate on the treadmill, it must be regarded as the method of choice when evaluating vascular patients for associated CAD (Table IV). Of particular interest is the fact that the three patients with false negative treadmill tests did not reach 85% of their predicted heart rate. The patient with a false positive treadmill test on the other hand, was using a diuretic at the time of the test. Arm ergometry can be used in conjunction with or as a valuable alternative to treadmill testing in patients with a normal treadmill test who cannot reach a satisfactory heart rate. Proponents of exercise testing claim that an ischemic response during exercise is a useful predictor of perioperative cardiac complications [19,20] and that patients able to achieve more than 85% of their predicted maximum heart rate may represent a low risk group for major vascular surgery [21]. This was supported in the present study as all five patients that suffered perioperative cardiac events had a positive exercise test and only two of them managed to reach 85% of their maximum predicted heart rate (Table 111).The ability of exercise testing to identify patients with significant but asymptomatic CAD may be of even greater importance for their long-term prognosis. Using routine exercise testing enabled us to identify 18 of the 24 patients with asymptomatic but hemodynamically significant CAD (75%). Fourteen patients
483
were subsequently referred for coronary artery bypass surgery and in 12 cases this was done before the initially planned peripheral vascular procedure (Table V).
CONCLUSIONS Although many previous studies related the results of noninvasive tests to perioperative complications only [7-9,19-21], the long-term prognosis of patients also has to be taken into account [18]. ECG-monitored stress testing in general proved to be a fairly sensitive and very specific way of detecting associated asymptomatic CAD in patients for vascular surgery. We suggest that every patient under consideration for major vascular surgery should at least be subjected to an ECG-monitored exercise test. More expensive methods such as dipyridamole and exercise thallium scanning may have a place in patients who cannot reach a satisfactory heart rate with any of the described methods. Bearing in mind the systemic nature of atherosclerosis, the admission of a patient for vascular surgery must be seen as a golden opportunity to diagnose associated asymptomatic CAD.
ACKNOWLEDGMENTS This study was supported by the Medical Research Council of South Africa.
REFERENCES 1. HERTZER NR. Fatal myocardial infarction following abdominal aortic aneurysm resection. Ann Surg 1980;192:667673. 2. HERTZER NR, LEES CD. Fatal myocardial infarction following carotid endarterectomy Ann Surg 1981;194:212218. 3. HERTZER NR. Fatal myocardial infarction following lower extremity revascularization. Ann Surg 1981;193:492-498. 4. TOMATIS LA, FIERENS EE, VERBRUGGE GP. Evaluation of surgical risk in vascular disease by coronary arteriography: series of 100 cases. Surgery 1972;71:429-435. 5. HERTZER NR, BEVEN EG, YOUNG JR, et al. Coronary artery disease in peripheral vascular patients. A classification of 1000 angiograms and results of surgical management. Ann Surg 1984;199:223-233. 6. NICOLAIDES AN, SALMASI A, SONECHA TN. How should we investigate the arteriopath for coexisting lesions? J Cardiovasc Surg 1986;27515-533. 7. EAGLE KA, SINGER DE, BREWSTER DC, et al. Dipyridamole- thallium scanning in patients undergoing vascular surgery. Optimizing preoperative evaluation of cardiac risk. JA MA I987;2572 185-2 189. 8. CUTLER BS, WHEELER HB, PARASKOS JA, et al. Applicability and interpretation of electrocardiographic stress testing in patients with peripheral vascular disease. A m J Surg 1981;141:501-506. 9. CUTLER BS, WHEELER HB, PARASKOS JA, et al. Assessment of operative risk with electrocardiographic ex-
484
10.
11. 12.
13. 14.
IS. 16.
ROUTINE EXERCISE TESTING TO DETECT CAD
ercise testing in patients with peripheral vascular disease. A m J Surg 1979;137:484-490. YOUNGMAN DJ, DOVE T, BOCCUZZI SJ, et al. Noninvasive and angiographic evaluation of coronary artery disease in patients with peripheral vascular disease. Am J Cardiol 1989;63:14461449, BRUCE RA. Exercise testing of patients with coronary heart disease. Principles and normal standards for evaluation. Ann Clin Rest 1971;3:323-332. WILLIAMS J , COTTRELL E , POWERS SK, et al. Arm ergometry: a review of published protocols and the introduction of a new weight adjusted protocol. J Sports Med 198333:107-1 12. D E BAKEY ME, CRAWFORD ES, MORRIS GC Jr, et al. Late results of vascular surgery in the treatment of arteriosclerosis. J Cardiovasc Surg 1964;5:473480. CRAWFORD ES, BOMBERGER RA, GLAESER DH, et al. Aortoiliac occlusive disease: factors influencing survival and function following reconstructive operation over a twenty- five year period. Surgery 1981;90:10SS-1067. HOLLIER L H , PLATE G , O’BRIEN PC, et al. Late survival after abdominal aortic aneurysm repair: influence of coronary artery disease. J Vusc Surg 1984;1:29&299. JAMIESON WRE, JANCESZ TM, MIYAGISHIMA RT, et
17.
18.
19. 20. 21.
22.
ANNALSOF VASCULAR SURGERY
al. Influence of ischemic heart disease on early and late mortality after surgery for peripheral occlusive disease. Circulation 1982;66:92-97. THOMPSON JE, AUSTIN DJ, PATMAN RD. Carotid endarterectomy for cerebrovascular insufficiency: long term results in 592 patients followed up to thirteen years. Ann Surg 1970;172:663-679. HERTZER NR, YOUNG JR, BEVEN EG, et al. Late results of coronary bypass in patients with infrarenal aortic aneurysms. The Cleveland Clinic study. Ann Surg 1987305: 360-367. MC CABE CJ, REIDY NC, ABBOT WM, et al. The value of electrocardiogram monitoring during treadmill testing for peripheral vascular disease. Surgery 1981$9: 183-186. AROUS EJ, BAUM PL, CUTLER BS. The ischemic exercise test in patients with peripheral vascular disease. Arch Surg 1984;119:78&783. MC PHAIL N , CALVIN JE, SHERIATMANDAR A, et al. The use of preoperative exercise testing to predict cardiac complications after arterial reconstruction. J Vusc Surg 1986;7:60-68. SCHWADE J, BLOMQVIST CG, SHAPIRO W. A comparison of the response to arm and leg work in patients with ischemic heart disease. A m Heurt J 1977;94:203-208.
