Systolic Time Intervals before and after Maximal Exercise Treadmill Testing for Evaluation of Chest Pein" Ronald E. Gillilan, M.D.; William D. Parnes, M.D.; Brian E. Mandell; Richard]. Bouchard, M.D.; and]. Richard Warbasse, M.D.
De cbange in systoUc time iatervaIs from before eJ:erclse to three to four minutes foDowing a mulmlll-eJ:erclse treadmill test was measured to evaluate chest pain in 110 fastiDg supinr subjects. Forty" (85 perceat) of 54 patlenD with chest pain and with abnormal fiDdiDp on c0ronary arteriograms were found to have at least a lo-msec prolongation in the left ventricular ejection time btdeJ: (LVETI), whereas only two (8 perceBt) of 25 subjects without heart disease and 5 (16 pettent) of 31 subjecfs with chest pain but with normal 8adiDgs on coronary
arteriograms had 10 nuec or more of prolongation of the LVETI after eJ:erdse. The cbanle III the other systolic: time intervals (total eIectromecllaDicJll systole, preejectloa pIaase [PEP], awd PEP/LVET) were less reliable III detediq tile presence or lIIMeBce of COI'OII8I'Y disease. We coadude ..... determblatioa of L VEn before ..... after maximal.eJ:erdse tradmDI ..... Is a dhlically UIefaI nonlDvasive diapoItIc: test fOl' obltrudh'e coroaary disease In patIeDts with chest pain.
limitations of the currently used clinical methods for evaluating chest pain make it apparent that there is a need to improve the diagnostic accuracy of noninvasive tests for coronary heart disease.' Since myocardial ischemia during attacks of angina pectoris often results in decreased contractility of the left ventricle," a relatively sensitive noninvasive test of left ventricular function, such as measurements of systolic time intervals," might be of value as a diagnostic test for obstructive coronary arterial disease. Although resting measurements of systolic time intervals have not provided a dependable means of detecting coronary heart disease,4,5 and these measurements during exercise have not proved useful," a previous report suggests that the change in measurements of systolic time intervals from the preexercise to the postexercise state may aid in the diagnosis of ischemic heart disease." The present investigation was designed to determine the diagnostic accuracy of the change in measurements of systolic time intervals before and three to four minutes after maximal-exercise treadmill testing in detecting obstructive coronary disease among patients with recurrent chest pains.
MATERIALS AND MEmODS
~e
-From the Cardiovascular Service and Laboratories, Public Health Service Hospital, Baltimore Md. Supported in part Iiy Base Researclt Program Support of the Federal Health Programs Service. Manuscript received June 11; revision accepted August 3.
Reprint requests: Cardiology Department, USPHS Hospital, 3100 Wyman Pllf'k Drive, Baltimore 21211
CHEST. 71: 4, APRIL, 1977
Study Population Following review and approval of the protocol by a medical ethics committee, fully informed verbal and written consent was obtained hom all 110 subjects comprising the two study groups participating in this investigation. The 6rst group, consisting of 56 sedentary subjects without evidence of significant coronary disease, is subdivided into 25 nonnaI "control" subjects and 31 subjects with chest pain and with normal findings on coronary arteriograms. The 25 normal subjects were 20 men and five women (mean age, 50 years; range, 35 to 63 years) who had DO evidence of heart disease by history, physical examination, chest x-ray film, and electrocardiogram. Coronary arteriographic studies were not performed on these asymptomatic, apparently healthy subjects. The 31 patients with nonnal coronary arteriograms were 17 women and 14 men (mean age, 50 years; range, 38 to 65 years) from among the 85 consecutive patients with a syndrome of recurrent chest pain. The second group consisted of 50 men and four women (mean age, 55 years; range, 35 to 64 years) with abnormal coronary arteriograms who were the remainder of the 85 consecutive patients with recurrent chest pain. Among these 54 patients with coronary disease, there were 28 who had previous (more than three months) myocardial infarctions. subjects with chest pain were considered by three cardiologists to have strictly clinical indications for coronary arteriographic studies. Patients with recent myocardial infarction, congestive heart failure, or coexisting valvular or hypertensive cardiovascular disease and subjects receiving tDerapy with digitalis and propranolol were excluded from the study.
An
smouc TIME INTERVAlS
479
Meawrements of Systolic Time Irrterool8 Systolic time intervals were measured according to the method of Weissler and Ganards on all 110 subjects prior to and three to four minutes following maximal~xercise treadmill testing. All subjects were studied between the hours of 8 AM and 12 noon, with the patients having fasted overnight for 10 to 16 hours. Patients abstained from smoking cigarettes and from the use of nitroglycerin on the morning of the test. Recordings were obtained at base line after the subjects had rested in the supine position fOl' at least ten minutes and also at the third to fonrth minute after stopping the exercise test. Measurements were made by recording simultaneously the ECG. the phonocardiogram, and the carotid arterial pulse at a paper speed of 200 mm/sec with 2O-msec time lines on an oscilloscopic recorder (Electmoia for Medicine DR-8). The ECG consisted of a single bipolar chest lead, CBs, as described herein. The phonocanliognun was obtained with a contact microphone (Electmoics for Medicine) placed at the point on the upper anterior portion of the chest wall showing the best delineatioo of the second heart sound. The carotid pulse recording was obtained by placing OYer the carotid pulsation a funnel that was air-coupled to a pressure transducer (Statham P23Db) . Total electromecbani.c systole (Q-S2) W9S measured from the onset of the QRS complex of the ECG to that highfrequency component of the second heart sound preceding the incisura of the carotid pulse by approximately 20 msee. Left ventricular ejection time (LVET) was measured from the beginning of the rapid upstroke of the carotid pulse to the nadir of the Incisura, The preejection phase (PEP) was derived by subtracting the LVET from Q-Sz. The measurements on ten consecutfve beats were determined to the nearest 5 msec and then were averaged, except when sinus anhythmia was present, in which case 15 consecutive beats were averaged. Indices of systolic time intervals (Q-S~, L VETI, and PEPI) were derived from the actual measured intervals by applying the regression equations of Weissler and Ganard. s Records of systolic time intervals were assigned coded numbers so that the technician making the measurements and the cardiologist rechecking the measurements were both unaware of whether a partiaJ1ar record was that of a subject with chest pain or a control subject. These data were subjected to statistical analysis, utilizing the t-test for difference of means and the x2 contingency test with Yates' correction.
110 subjects. Subjects walked or jogged on the treadmill to the point of chest pain or limiting symptoms, such as dyspnea or fatigue. A single bipolar chest lead (CBs), with the positive electrode at the V s position and the negative electrode over the right scapula, was used." Cuff blood pressures and ECGs were recorded at base line, during each stage of exercise, and during recovery. An ECG was considered to show an ischemic response to exercise if the ST segment developed 1 mm or more of depression (horizontal or downward sloping for at least O.08-second duration) below or elevation above the P-R segment. Selective Coronorv Arteriograms and Ventriculograms Eighty-five subjects underwent clinically indicated selec-
tive coronary arteriographic studies (percutaneous femoral approach of Judlcins10 ) for evaluation of chest pain. Multiple
projections of each coronary artery were 6lmed to assure adequate visualization of the entire coronary circulation of each patient. All arteriograms were "blindly" interpreted by two cardiologists experienced in interpreting coronary arteriograms who did not know the results of the measurements of systolic time intervals. The degree of coronary arterial obstruction was graded from zero to 100 percent. The main left coronary artery, the left anterior descending coronary artery. the major diagonal branches, the circumflex coronary artery, the major circumflex marginal branches, the right coronary artery, and the posterior descending coronary artery were considered major coronary arteries and were graded individually. Patients were classi6ed as having significant coronary obstruction (ie, abnormal coronary arteriogram) if there was 75-percent or greater obstruction of at least one of the previously listed major coronary arteries. Patients with less than 5O-percent obstruction of any major artery were considered to have no signi6cant coronary obstruction (ie, normal c0ronary arteriogram). Left ventricular volumes were measured in all 85 subjects, according to the method of Greene et al,11 from single-plane 35-mm cineangiograms 6lmed at 60 frames per second in the right anterior oblique projection. The anterior, apical, and inferior segments of the left ventricle were graded as normal, hypokinetic, alcinetic, or dyskinetic by two cardiologists with extensive experience with left ventricular ciDeangiograms. REsULTS
Maximal-Erercise Treadmill Testing Following determination of base-line resting systolic time graded-treadmiJl tests, as deintervals, maximal~xercise scribed by McDonough and Bmce,1I were performed by the
Selective Coronary Arteriograms
Abnormal coronary arteriograms, as defined previously, were found in 50 men and four women.
Table l--a-ll. o/IIa.dJruJl-Esen:iN TretIdmUl Tab (IIETT) and 11--..--.. 0/ SratoIie TillU11lllenHlla ill 110 Su6}ee.. • METT
Group
No. of Subjectll
Subjectll without coronary disease Normal control Chest pain and normal CA··
56 25 31
Subjectll with coronary disease Chest pain and abnormal CA··
54
Chest Pain Only 8 (14)
o
8 (26)
10 (18)
Ischemic ECG Only 4 (7)
o
4 (13)
4 (7)
Chest Pain and Ischemic ECG
Ischemic STl ReBpoJ18e
1 (3)
7 (12) 2 (8) 5 (16)
38 (70)
46 (85)
1 (2)
o
*Table values are numbers of patientll; numbers within parentheses are percents. ECG is lead CBa. ··CA, Coronary arteriogram.s.
480 GIWWI ET At
CHEST. 71: 4, APRIL, 1977
Normal coronary arteriograms, defined as less than 5O-percent obstruction in all major coronary arteries, were found in 17 women and 14 men. There were no subjects with an obstruction of 50 to 74 percent of only one major coronary artery. The 31 subjects with normal coronary arteriograms had normal heart size, normal right and left ventricular pressures, normal indices of cardiac output, normal measurements of left ventricular volume, and no evidence of localized disorders of myocardial contraction on the left ventricular cineangiograms. Maximal-Exercise Treadmill Testing Table 1 summarizes the results of the maximalexercise treadmill tests . All 25 control subjects attained 100 percent of the maximal predicted heart rate for their age. 1Z None of these subjects developed chest pain or evidence of ischemia on the ECG. Although the 31 subjects with recurrent chest pain but with normal coronary arteriograms showed a high incidence of equivocally positive findings on maximal-exercise treadmill tests by virtue of developing chest pain (eight subjects or 26 percent) or ischemic ST-segment depression (four subjects or 13 percent), only one subject (3 percent) developed both chest; pain and ST-segment depression (unequivocally positive findings) with exercise testing. Thirty-eight of the 54 subjects with recurrent chest pain who had abnormal coronary arteriograms stopped exercise with chest pain consistent with angina pectoris and evidence of ischemia as defined by either 1 mm or more of ischemic ST-
segment depression (34 patients) or ischemic STsegment elevation (four patients) on the exercise ECG. Thus, the maximal-exercise treadmill test gave unequivocally positive findings in 70 percent (38) of 54 patients with significant coronary obstruction. An additional 14 patients with coronary disease developed only chest pain (ten patients) or ischemic ST-segment depression (four patients) and may therefore be classified as having equivocally positive findings on exercise tests. Only two of the 54 patients had unequivocally negative findings on maximalexercise treadmill tests (no chest pain or ischemic ST-segment depression). Measurements of Systolic Time Intervals Table 2 summarizes the results of the determinations of systolic time intervals before and three to four minutes following maximal-exercise treadmill testing in all subjects. Comparison of resting baseline measurements of systolic time intervals in the groups shows that although the mean LVETI is significantly reduced (P < 0.001) and the mean PEPI, as well as the PEP/LVET, are significantly prolonged (P < 0.001) in patients with coronary disease, a considerable number (18) of these 54 patients fail to demonstrate abnormality in these base-line measurements of cardiac function. The changes in the indices of systolic time intervals from before to three to four minutes after exercise show more significant differences between the groups. The patients with coronary disease have significant prolongation of the mean LVETI but no change in
Table 2-Mean Sy.wlie Time ',lIel'llflla before find Three 10 ' _ r ltIlnaIea c/fer ltIasi'""'-E_ra.e Treadmill Tad... (110 Sdjee..)·
Measurement
No Coronary Disease
Normal Controls
Chest Pain and NormalCA"
Coronary Disease (Chest Pain and Abnormal CA..)
No. of subjeets
56
25
31
54
LVETI, msee Before After P valuet Q-S,I, msee Before After P valuet
415 ± 12 412 ± 13 NS
±9 ± 13
NS
414 412
± 15 ± 13
401 ± 15 420 ± 16
De cbange in systoUc time iatervaIs from before eJ:erclse to three to four minutes foDowing a mulmlll-eJ:erclse treadmill test was measured to evaluate chest pain in 110 fastiDg supinr subjects. Forty" (85 perceat) of 54 patlenD with chest pain and with abnormal fiDdiDp on c0ronary arteriograms were found to have at least a lo-msec prolongation in the left ventricular ejection time btdeJ: (LVETI), whereas only two (8 perceBt) of 25 subjects without heart disease and 5 (16 pettent) of 31 subjecfs with chest pain but with normal 8adiDgs on coronary
arteriograms had 10 nuec or more of prolongation of the LVETI after eJ:erdse. The cbanle III the other systolic: time intervals (total eIectromecllaDicJll systole, preejectloa pIaase [PEP], awd PEP/LVET) were less reliable III detediq tile presence or lIIMeBce of COI'OII8I'Y disease. We coadude ..... determblatioa of L VEn before ..... after maximal.eJ:erdse tradmDI ..... Is a dhlically UIefaI nonlDvasive diapoItIc: test fOl' obltrudh'e coroaary disease In patIeDts with chest pain.
limitations of the currently used clinical methods for evaluating chest pain make it apparent that there is a need to improve the diagnostic accuracy of noninvasive tests for coronary heart disease.' Since myocardial ischemia during attacks of angina pectoris often results in decreased contractility of the left ventricle," a relatively sensitive noninvasive test of left ventricular function, such as measurements of systolic time intervals," might be of value as a diagnostic test for obstructive coronary arterial disease. Although resting measurements of systolic time intervals have not provided a dependable means of detecting coronary heart disease,4,5 and these measurements during exercise have not proved useful," a previous report suggests that the change in measurements of systolic time intervals from the preexercise to the postexercise state may aid in the diagnosis of ischemic heart disease." The present investigation was designed to determine the diagnostic accuracy of the change in measurements of systolic time intervals before and three to four minutes after maximal-exercise treadmill testing in detecting obstructive coronary disease among patients with recurrent chest pains.
MATERIALS AND MEmODS
~e
-From the Cardiovascular Service and Laboratories, Public Health Service Hospital, Baltimore Md. Supported in part Iiy Base Researclt Program Support of the Federal Health Programs Service. Manuscript received June 11; revision accepted August 3.
Reprint requests: Cardiology Department, USPHS Hospital, 3100 Wyman Pllf'k Drive, Baltimore 21211
CHEST. 71: 4, APRIL, 1977
Study Population Following review and approval of the protocol by a medical ethics committee, fully informed verbal and written consent was obtained hom all 110 subjects comprising the two study groups participating in this investigation. The 6rst group, consisting of 56 sedentary subjects without evidence of significant coronary disease, is subdivided into 25 nonnaI "control" subjects and 31 subjects with chest pain and with normal findings on coronary arteriograms. The 25 normal subjects were 20 men and five women (mean age, 50 years; range, 35 to 63 years) who had DO evidence of heart disease by history, physical examination, chest x-ray film, and electrocardiogram. Coronary arteriographic studies were not performed on these asymptomatic, apparently healthy subjects. The 31 patients with nonnal coronary arteriograms were 17 women and 14 men (mean age, 50 years; range, 38 to 65 years) from among the 85 consecutive patients with a syndrome of recurrent chest pain. The second group consisted of 50 men and four women (mean age, 55 years; range, 35 to 64 years) with abnormal coronary arteriograms who were the remainder of the 85 consecutive patients with recurrent chest pain. Among these 54 patients with coronary disease, there were 28 who had previous (more than three months) myocardial infarctions. subjects with chest pain were considered by three cardiologists to have strictly clinical indications for coronary arteriographic studies. Patients with recent myocardial infarction, congestive heart failure, or coexisting valvular or hypertensive cardiovascular disease and subjects receiving tDerapy with digitalis and propranolol were excluded from the study.
An
smouc TIME INTERVAlS
479
Meawrements of Systolic Time Irrterool8 Systolic time intervals were measured according to the method of Weissler and Ganards on all 110 subjects prior to and three to four minutes following maximal~xercise treadmill testing. All subjects were studied between the hours of 8 AM and 12 noon, with the patients having fasted overnight for 10 to 16 hours. Patients abstained from smoking cigarettes and from the use of nitroglycerin on the morning of the test. Recordings were obtained at base line after the subjects had rested in the supine position fOl' at least ten minutes and also at the third to fonrth minute after stopping the exercise test. Measurements were made by recording simultaneously the ECG. the phonocardiogram, and the carotid arterial pulse at a paper speed of 200 mm/sec with 2O-msec time lines on an oscilloscopic recorder (Electmoia for Medicine DR-8). The ECG consisted of a single bipolar chest lead, CBs, as described herein. The phonocanliognun was obtained with a contact microphone (Electmoics for Medicine) placed at the point on the upper anterior portion of the chest wall showing the best delineatioo of the second heart sound. The carotid pulse recording was obtained by placing OYer the carotid pulsation a funnel that was air-coupled to a pressure transducer (Statham P23Db) . Total electromecbani.c systole (Q-S2) W9S measured from the onset of the QRS complex of the ECG to that highfrequency component of the second heart sound preceding the incisura of the carotid pulse by approximately 20 msee. Left ventricular ejection time (LVET) was measured from the beginning of the rapid upstroke of the carotid pulse to the nadir of the Incisura, The preejection phase (PEP) was derived by subtracting the LVET from Q-Sz. The measurements on ten consecutfve beats were determined to the nearest 5 msec and then were averaged, except when sinus anhythmia was present, in which case 15 consecutive beats were averaged. Indices of systolic time intervals (Q-S~, L VETI, and PEPI) were derived from the actual measured intervals by applying the regression equations of Weissler and Ganard. s Records of systolic time intervals were assigned coded numbers so that the technician making the measurements and the cardiologist rechecking the measurements were both unaware of whether a partiaJ1ar record was that of a subject with chest pain or a control subject. These data were subjected to statistical analysis, utilizing the t-test for difference of means and the x2 contingency test with Yates' correction.
110 subjects. Subjects walked or jogged on the treadmill to the point of chest pain or limiting symptoms, such as dyspnea or fatigue. A single bipolar chest lead (CBs), with the positive electrode at the V s position and the negative electrode over the right scapula, was used." Cuff blood pressures and ECGs were recorded at base line, during each stage of exercise, and during recovery. An ECG was considered to show an ischemic response to exercise if the ST segment developed 1 mm or more of depression (horizontal or downward sloping for at least O.08-second duration) below or elevation above the P-R segment. Selective Coronorv Arteriograms and Ventriculograms Eighty-five subjects underwent clinically indicated selec-
tive coronary arteriographic studies (percutaneous femoral approach of Judlcins10 ) for evaluation of chest pain. Multiple
projections of each coronary artery were 6lmed to assure adequate visualization of the entire coronary circulation of each patient. All arteriograms were "blindly" interpreted by two cardiologists experienced in interpreting coronary arteriograms who did not know the results of the measurements of systolic time intervals. The degree of coronary arterial obstruction was graded from zero to 100 percent. The main left coronary artery, the left anterior descending coronary artery. the major diagonal branches, the circumflex coronary artery, the major circumflex marginal branches, the right coronary artery, and the posterior descending coronary artery were considered major coronary arteries and were graded individually. Patients were classi6ed as having significant coronary obstruction (ie, abnormal coronary arteriogram) if there was 75-percent or greater obstruction of at least one of the previously listed major coronary arteries. Patients with less than 5O-percent obstruction of any major artery were considered to have no signi6cant coronary obstruction (ie, normal c0ronary arteriogram). Left ventricular volumes were measured in all 85 subjects, according to the method of Greene et al,11 from single-plane 35-mm cineangiograms 6lmed at 60 frames per second in the right anterior oblique projection. The anterior, apical, and inferior segments of the left ventricle were graded as normal, hypokinetic, alcinetic, or dyskinetic by two cardiologists with extensive experience with left ventricular ciDeangiograms. REsULTS
Maximal-Erercise Treadmill Testing Following determination of base-line resting systolic time graded-treadmiJl tests, as deintervals, maximal~xercise scribed by McDonough and Bmce,1I were performed by the
Selective Coronary Arteriograms
Abnormal coronary arteriograms, as defined previously, were found in 50 men and four women.
Table l--a-ll. o/IIa.dJruJl-Esen:iN TretIdmUl Tab (IIETT) and 11--..--.. 0/ SratoIie TillU11lllenHlla ill 110 Su6}ee.. • METT
Group
No. of Subjectll
Subjectll without coronary disease Normal control Chest pain and normal CA··
56 25 31
Subjectll with coronary disease Chest pain and abnormal CA··
54
Chest Pain Only 8 (14)
o
8 (26)
10 (18)
Ischemic ECG Only 4 (7)
o
4 (13)
4 (7)
Chest Pain and Ischemic ECG
Ischemic STl ReBpoJ18e
1 (3)
7 (12) 2 (8) 5 (16)
38 (70)
46 (85)
1 (2)
o
*Table values are numbers of patientll; numbers within parentheses are percents. ECG is lead CBa. ··CA, Coronary arteriogram.s.
480 GIWWI ET At
CHEST. 71: 4, APRIL, 1977
Normal coronary arteriograms, defined as less than 5O-percent obstruction in all major coronary arteries, were found in 17 women and 14 men. There were no subjects with an obstruction of 50 to 74 percent of only one major coronary artery. The 31 subjects with normal coronary arteriograms had normal heart size, normal right and left ventricular pressures, normal indices of cardiac output, normal measurements of left ventricular volume, and no evidence of localized disorders of myocardial contraction on the left ventricular cineangiograms. Maximal-Exercise Treadmill Testing Table 1 summarizes the results of the maximalexercise treadmill tests . All 25 control subjects attained 100 percent of the maximal predicted heart rate for their age. 1Z None of these subjects developed chest pain or evidence of ischemia on the ECG. Although the 31 subjects with recurrent chest pain but with normal coronary arteriograms showed a high incidence of equivocally positive findings on maximal-exercise treadmill tests by virtue of developing chest pain (eight subjects or 26 percent) or ischemic ST-segment depression (four subjects or 13 percent), only one subject (3 percent) developed both chest; pain and ST-segment depression (unequivocally positive findings) with exercise testing. Thirty-eight of the 54 subjects with recurrent chest pain who had abnormal coronary arteriograms stopped exercise with chest pain consistent with angina pectoris and evidence of ischemia as defined by either 1 mm or more of ischemic ST-
segment depression (34 patients) or ischemic STsegment elevation (four patients) on the exercise ECG. Thus, the maximal-exercise treadmill test gave unequivocally positive findings in 70 percent (38) of 54 patients with significant coronary obstruction. An additional 14 patients with coronary disease developed only chest pain (ten patients) or ischemic ST-segment depression (four patients) and may therefore be classified as having equivocally positive findings on exercise tests. Only two of the 54 patients had unequivocally negative findings on maximalexercise treadmill tests (no chest pain or ischemic ST-segment depression). Measurements of Systolic Time Intervals Table 2 summarizes the results of the determinations of systolic time intervals before and three to four minutes following maximal-exercise treadmill testing in all subjects. Comparison of resting baseline measurements of systolic time intervals in the groups shows that although the mean LVETI is significantly reduced (P < 0.001) and the mean PEPI, as well as the PEP/LVET, are significantly prolonged (P < 0.001) in patients with coronary disease, a considerable number (18) of these 54 patients fail to demonstrate abnormality in these base-line measurements of cardiac function. The changes in the indices of systolic time intervals from before to three to four minutes after exercise show more significant differences between the groups. The patients with coronary disease have significant prolongation of the mean LVETI but no change in
Table 2-Mean Sy.wlie Time ',lIel'llflla before find Three 10 ' _ r ltIlnaIea c/fer ltIasi'""'-E_ra.e Treadmill Tad... (110 Sdjee..)·
Measurement
No Coronary Disease
Normal Controls
Chest Pain and NormalCA"
Coronary Disease (Chest Pain and Abnormal CA..)
No. of subjeets
56
25
31
54
LVETI, msee Before After P valuet Q-S,I, msee Before After P valuet
415 ± 12 412 ± 13 NS
±9 ± 13
NS
414 412
± 15 ± 13
401 ± 15 420 ± 16
