- Email: [email protected]
Alternans of ventricular gradient during percutaneous transluminal coronary angioplasty
Journal of Electrocardiology Vol. 34 No. 2 2001
Alternans of Ventricular Gradient During Percutaneous Transluminal Coronary Angioplasty
Shigeo Horinaka, MD, Suomi Hara, MD, Noriaki Tsuchiya, MD, Akihisa Yabe, MD, Hiroshi Asakawa, MD, Hiroshi Yagi, MD, Yousuke Mori, MD, and Hiroaki Matsuoka, MD
Abstract: We evaluated the influence of local myocardial ischemia induced by acute coronary occlusion during percutaneous transluminal coronary angioplasty (PTCA) on the ventricular gradients (VG) and investigated whether 2:1 alternans of VG occurs. Twenty-seven patients with angina pectoris, who had one-vessel coronary artery stenosis, were studied. The VG of each consecutive heartbeat before, during, and after PTCA over a 22-second interval was calculated using a microcomputer. The standard deviation and coefficient of variation of magnitude were used as indices of VG variability. Frequencydomain analysis of time series consisting of beat-to-beat VG magnitude for a 22-second interval was also performed by the maximum entropy method. The standard deviation and coefficient of variation of VG magnitude during PTCA were significantly greater than those before and after PTCA (P ⬍ .01, P ⬍ .01, respectively), and the indices before PTCA were also significantly greater than those after PTCA (P ⬍ .05). The maximum power spectrum peaks around 0.5 cycles/beat during PTCA were significantly greater than those after PTCA (P ⬍ .01); this suggests that the enhancement of VG alternans is reflected by 2:1 alternans of the action potential in the acute local ischemic myocardium during PTCA. Key words:Ventricular gradient variability, percutaneous transluminal coronary angioplasty, acute myocardial ischemia.
Recently, alternans of the ST segment has been observed in acute coronary artery occlusion in experimental animals (1,2), and clinically, in patients with variant angina (3,4) and during percu-
taneous transluminal coronary angioplasty (PTCA) (5,6). Nearing et al. (7), who used complex demodulation methods, found marked increases in the degree of T wave alternans that parallel the established time course of changes in vulnerability in coronary artery occlusion and reperfusion in dogs. Previously, we reported beat-to-beat variability of the parameters of the ventricular gradient (VG) of consecutive heart beats measured in Frank lead X, Y, Z scalar electrocardiograms at rest from patients with ischemic heart disease. The coefficient of variation of the VG magnitude in patients with
From the Department of Hypertension and Cardiorenal Medicine, Dokkyo University School of Medicine, Tochigi, Japan. Reprint requests: Shigeo Horinaka, MD, Department of Hypertension and Cardiorenal Medicine, Dokkyo University School of Medicine, 880 Kitakobayashi, Mibu, Tochigi, 321-0293, Japan; e-mail: [email protected]. Copyright © 2001 by Churchill Livingstone® 0022-0736/01/3402-0006$35.00/0 doi:10.1054/jelc.2001.23114
135
136 Journal of Electrocardiology Vol. 34 No. 2 April 2001 angina pectoris who had significant coronary stenosis was greater than that in healthy patients (8,9). Moreover, this index correlated with alternans of the monophasic action potentials in ischemic myocardium in experimental canine models (10). PTCA is an acute myocardial ischemic model in humans. Therefore, we investigated whether the VG variability increased and the beat-to-beat VG alternans was pronounced during PTCA.
Materials and Methods
the suitable view that best showed the lesion without overlap. The percentage of diameter stenosis of the luminal area was calculated by dividing the culprit luminal diameter by the normal luminal diameter for PTCA measured by videodiameter methods. Likewise, the percentage of stenosis of the densitometric luminal area was calculated by dividing the culprit lesion area by the normal luminal area measured by videodensitometric methods using a quantitative coronary angiographic analyzer (Cardio-500 system, Kontron, Germany) before and after PTCA.
Patient Selection
VG Alternans Analysis
Between January 1997 and April 1998, 247 patients underwent percutaneous revascularization procedures in our division. Our study comprised of 27 consecutive patients with de novo effort angina pectoris who underwent PTCA (mean age: 61 years; range, 41 to 73 years; sex: 21 men, 6 women). All patients gave written informed consent to participation in this study. Exclusion criteria were multivessel disease, hypertension, valvular heart disease, myocardial infarction, and arrhythmia.
Before, during the last 22 seconds of, and 30 minutes after PTCA. Frank lead X, Y, Z scalar ECGs were recorded with the patients in the angiographic room and simultaneously stored on the data recorder (Sony FC-14, Tokyo, Japan) with a vectorcardiograph (Fukuda Denshi, Cardio-Pro FX 3211, Tokyo, Japan). Simultaneously, standard 12-lead ECGs were also recorded to confirm the location of myocardial ischemia. During PTCA, Frank lead X, Y, Z scalar ECGs were analyzed during the last 22 seconds after balloon inflation when the ST-segment shift was stable. The mean and standard deviation (SD) of the VG magnitude for all heartbeats in the 22-second interval were obtained by computer (NEC-9801BX, Tokyo, Japan). The coefficient of variation (SD/mean ⫻ 100) of this value was used as index of beat-to-beat VG variability as previously described in detail (9,10). Furthermore, spectral analysis of VG variability was performed. Time series of the VG magnitude for all heartbeats in the 22-second interval were analyzed by the MemCalc system (Suwa Trust Co. Ltd, Tokyo, Japan) (11), which is a combination of the maximum entropy method for spectral analysis and nonlinear least squares method for fitting analysis. This new method was appropriated to the analysis of VG alternans (ie, 1:2) over a maximum interval of 22 seconds.
Cardiac Catheterization and PTCA All patients underwent diagnostic coronary angiography with a transfemoral approach with the Judkins technique and exhibited at least 75% organic stenosis of a major coronary artery on coronary angiography after intracoronary injection of a nitrate drug. The single ischemia-related artery stenosis (culprit lesion) was also confirmed by the localization of 12-lead electrocardiographic changes with spontaneous chest pain or exercise treadmill test. A few days after diagnostic coronary angiography, PTCA was performed with the conventional balloon technique with the Judkins technique as a primary treatment. Balloon inflation was repeated to dilate the culprit lesion sufficiently. The balloon was inflated each time for 60 to 120 seconds. Angiographic success was defined as less than 30% residual diameter stenosis after PTCA. Quantitative Coronary Angiography Quantitative coronary angiography (QCA) of the culprit lesion was performed in patients with patent arteries by a single operator who had no knowledge of the patients’ characteristic data and coronary morphology. The QCA analysis was performed in
Statistical Analysis All calculated data are expressed as the mean ⫾ SD. The one-way analysis of variance, which was subsequently subjected to Fisher’s protected least significant difference test for multiple comparisons, was used to determine the statistical significance of differences. The statistical analysis was conducted with a commercially available statistical software program (STSTVIEW Version 4.58, Abacus Concepatients, Berkeley, CA). Statistical significance was accepted at P ⬍ .05.
VG Alternans During PTCA •
Horinaka et al. 137
Table 1. Percentage of Culprit Lesion Stenosis, HR, Mean, Standard Deviation, Coefficient of Variation and Maximum Power Around 0.5 Cycles of the Ventricular Gradients Before, During, and After PTCA (n ⫽ 26) Before PTCA L-Area % diameter stenosis (%) Densito L-Area % stenosis (%) HR (bpm) Mean of VG magnitude (Vsec) SD of VG magnitude (%) CV of VG magnitude (%) Power spectrum density around 0.5 cycles of VG
91.5 ⫾ 6.0 86.5 ⫾ 13.0 72.4 ⫾ 12.3 74.68 ⫾ 20.72 6.76 ⫾ 1.76 9.51 ⫾ 3.15 0.43 ⫾ 0.53†
During PTCA
After PTCA
75.9 ⫾ 13.0 78.38 ⫾ 31.49 8.32 ⫾ 2.64† 12.11 ⫾ 6.18† 1
43.0 ⫾ 16.1† 25.7 ⫾ 22.6† 75.4 ⫾ 10.2 76.52 ⫾ 20.94 5.81 ⫾ 1.19*‡ 8.10 ⫾ 2.61*‡ 0.33 ⫾ 0.40‡
* P ⬍ .05 vs. before PTCA, † P ⬍ .01 vs. before PTCA, ‡ P ⬍ .01 vs. during PTCA. Data are the mean ⫾ standard deviation. Mean: mean value of all the original values, SD: standard deviation of all the original values, CV: coefficient of variation (SD/Mean ⫻ 100), Densito: densitometric, L-Area: Lumen-Area. Power spectrum density (Vsec2) around 0.5 cycles of VG: relative values were presented when the value of integration of the power spectrum density from 0.45 to 0.5 cycles/beat of VG magnitude during PTCA was 1.
Results Sixteen, 7, and 4 patients had left anterior descending coronary artery, right coronary artery, and left circumflex coronary artery stenosis, respectively. The reference diameter of the culprit lesion for PTCA was 2.51 ⫾ 0.61 mm. The percentages of luminal diameter and area stenosis of the culprit lesion for PTCA reduced from 91.5 ⫾ 6.0 and 86.5 ⫾ 13% to 43.0 ⫾ 16.1 and 25.7 ⫾ 22.6%, respectively, as shown in Table 1. There was no difference in heart rate (HR) before, during, and after PTCA (Table 1). A localized STsegment shift greater than 1 mm in 12-lead electrocardiogram induced by acute myocardial ischemia during PTCA was observed in all patients. The ST-segment or T-wave alternans (ie, 2:1 or 3:1) on 12-lead ECG were not visually observed in any patients. The VG of the QRST complexes recorded before, during, and after PTCA in one patient with coronary artery disease are shown in Figure 1. Before, during the last 22 seconds of, and 30 minutes after balloon inflation, there were no significant differences in mean VG magnitude among the three periods. The SD of VG magnitude was significantly
Fig. 1. Superimposed ventricular gradients of each heartbeat in the 22-second interval before, during, and after PTCA in one patient with coronary artery disease. Left, before PTCA; middle, during PTCA; right, after PTCA.
greater during PTCA than before or after PTCA (P ⬍ .01, P ⬍ .01, respectively) and significantly smaller after PTCA than before PTCA (P ⬍ .05). The coefficient of variation of the VG was significantly greater during PTCA than before or after PTCA (P ⬍ .01) and significantly smaller after PTCA than before PTCA (P ⬍ .05) (Table 1). The frequency power of the spectral analysis of the VG magnitude recorded before, during, and after PTCA in one patient with coronary artery disease is shown in Figure 2. For spectral analysis, the number of patients who had maximum power around 0.5 cycles/beat was 4 (15%) before, 14 (52%) during, and 2 (7%) after PTCA. The power spectral density around 0.5 cycles/beat (integration of the power spectrum from 0.45 to 0.5 cycles/beat) in all patients during PTCA was significantly greater than that before and after PTCA (P ⬍ .01, respectively) (Table 1).
Discussion This study showed that 2:1 alternans of the VG magnitudes were enhanced in acute local myocardial ischemia during PTCA even in humans. Inter-
138 Journal of Electrocardiology Vol. 34 No. 2 April 2001
Fig. 2. Frequency power spectrum of the magnitude of ventricular gradients in the 22-second interval before, during, and after PTCA in one patient with coronary artery disease. left, before PTCA; middle, during PTCA; right, after PTCA. power sp: relative value when the maximum power spectral density (Vsec2) was 1.
estingly, this phenomenon was observed at rest in 4 patients (15%). The coronary blood flow was reduced in severe stenosis because the mean stenotic rate was about 90% in our patients (12). Moreover, 2:1 alternans almost disappeared after PTCA in these cases. Thus, it is likely that myocardial electrical instability exists in severe stenosis even at rest. The VG, which was first described by Wilson et al. (13), reflects nonuniformity in ventricular action potentials. Theoretical and clinical reports have documented that the VG is independent of the activation sequence (14,15). Geselowitz (16) showed mathematically that the VG is the difference in area of the individual action potentials of the myocardium. Recently, we reported that alternans of the monophasic action potentials appeared in the lowflow myocardial area in an experimental canine coronary stenosis model, and the magnitude of these alternans significantly correlated with that of variability of the VG magnitude calculated by recording body surface leads (10). Therefore, it is likely that 2:1 alternans of the action potential of the acute local ischemic myocardium occur in acute coronary occlusion, even in humans. Although 2:1 alternans cannot be confirmed visually, as shown in Fig. 1, the alternans may be obscured by the fluctuation of the distance and angle between the heart and body surface leads by breathing. Only 50% of our patients had maximum power around 0.5 Hz in the frequency analysis. The other patients might have had weaker ischemia, or smaller areas, as previously described (17), or relatively larger fluctuation of distance and angle between heart and body surface leads. In this study, the maximum entropy method was used to prove the appearance of the 2:1 alternans of VG in myocardial ischemia. However, time series analysis comprises 2 kinds of analysis; one is the frequency-domain analysis known as spectral anal-
ysis, and the other is the time-domain analysis. Fast fourier transformation (FFT) is one of the most popular methods of spectral analysis widely used up to date. However, FFT has such disadvantage as poor resolution because of the effect of window functions and unrealistic assumptions about extending the data, needing at least 128 samples (18). Thus, FFT is inadequate for estimating the precious power spectral density from short time series data such as the small number of beats during the 22-second interval in our study. On the other hand, with respect to the time-domain series analysis in the time-domain, the complex demodulation method has been used (19). This technique calculates the instantaneous amplitude and phase by using moving-average filtered data. However, complex demodulation is inappropriate for analysis of rapidly varying fluctuations because the technique is applicable to only highly smoothed time series data. Therefore, we used the time series of the VG magnitude for all heartbeats in the 22-second interval analyzed by the MemCalc method, which is a combination of the maximum entropy method for spectral analysis and nonlinear least squares method for fitting analysis. With this new method, we achieved a reliable analysis of the relatively low frequency component (0.5 Hz) of the VG alteration on a maximum interval of 22 seconds. Recently, a relationship has been reported between T-wave alternans and ventricular tachyarrhythmia (20,21). However, no patients had ventricular arrhythmia during or after PTCA in our study. Therefore, further studies are necessary to evaluate the relation between the alternans of VG and ventricular arrhythmia. It may be necessary to elucidate the influence of voltage change in the body surface leads by the depth and frequency of the breathing, and the effect of chest pain and intracoronary nitrate drug
VG Alternans During PTCA •
administration during PTCA, and autonomic nervous system function in further study. In summary, we observed not only enhanced alternans of the VG magnitudes, but also a maximum power spectrum for this alternans around 0.5 cycles/beat in acute myocardial ischemia during PTCA. These data suggest that the 2:1 alternans of the action potentials is enhanced in acute ischemic myocardium.
References 1. Kleber AG, Janse MJ, van Capelle FJL, et al: Mechanism and time course of S-T and T-Q segment changes during acute regional myocardial ischemia in the pig heart determined by extracellular recordings. Circ Res 42:603, 1978 2. Hashimoto H, Asano M, Nakashima M: Potentiating effects of a ventricular premature beat on the alternation of ST-T complex of epicardial electrocardiograms and the incidence of ventricular arrhythmias during acute coronary artery occlusion in dogs. J Electrocardiol 17:229, 1984 3. Kleinfeld MJ, Rozanski JJ: Alternans of the ST segment in Prinzmetal’s angina. Circulation 55:574, 1977 4. Turitto G, El-Sherif N: Alternans of the ST segment in variant angina. Chest 93:587, 1988 5. Okino H, Arima S, Yamaguchi H, et al: Marked alternans of the elevated ST segment during occlusion of the left anterior descending coronary artery in percutaneous transluminal coronary angioplasty: Clinical background and electrocardiographic features. Intern J Cardiol 37:23, 1992 6. Sochanski M, Feldman T, Chua KG, et al: ST segment alternans during coronary angioplasty. Catheter Cardiovasc Diag 27:45, 1992 7. Nearing BD, Huang AL, Verrier RL: Dynamic tracking of cardiac vulnerability by complex demodulation of the T wave. Science 252:437, 1991 8. Horinaka S, Yamamoto H, Tabuchi T, et al: New electrocardiographic method for detection of isch-
9.
10.
11.
12.
13.
14. 15.
16.
17.
18.
19.
20.
21.
Horinaka et al. 139
emic heart disease: Alterations in ventricular gradient. Jpn Heart J 35:183, 1994 (suppl) Horinaka S, Yamamoto H, Tabuchi T, et al: Ventricular gradient variability: New ECG method for detection of ischemic heart disease. J Electrocardiol 28: 177, 1995 Horinaka S, Yamamoto H: Enhancement of ventricular gradient variability during acute myocardial ischemia. Int J Cardiol 65:173, 1998 Takusagawa M, Komori S, Umetani T, et al: Alteration of autonomic nervous activity in recurrence of variant angina. Heart 82:75, 1999 Gould KL, Lipscomb K: Effect of coronary stenosis on coronary flow reserve and resistance. Am J Cardiol 34:48, 1974 Wilson FN, Macleod AG, Barker PS, et al: The determination and the significance of the areas of the ventricular deflections of the electrocardiogram. Am Heart J 10:46, 1934 Burger H: A theoretical elucidation of the motion “ventricular gradient.” Am Heart J 53:240, 1957 Berkun MA, Kesselman RH, Donoso E, et al: The spatial ventricular gradient: Intermittent Wolff-Parkinson-White syndrome, intermittent left bundle branch block and ventricular premature contractions. Circulation 13:562, 1956 Geselowitz DB: The ventricular gradient revisited: Relation to the area under the action potential. IEEE Trans Eng 30:76, 1983 Kwan T, Feit A, Alam M, et al: ST-T alternans and myocardial ischemia. Catheter Cardiovasc Diag 35: 121, 1995 Smith JM, Clancy EA, Valeri CR, et al: Electrical alternans and cardiac electrical instability. Circulation 77:110, 1988 Nearing BD, Huang AL, Verrier RL: Dynamic tracking of cardiac vulnerability by complex demodulation of the T wave. Science 252:437, 1991 Nearing BD, Oesterle SN, Verrier RL: Quantification of ischaemia induced vulnerability by precordial T wave alternans analysis in dog and human. Cardiovasc Res 28:1440, 1994 Rosenbaum DS, Jackson LE, Smith JM, et al: Electrical alternans and vulnerability to ventricular arrhythmias. N Engl J Med 330:235, 1994
Alternans of Ventricular Gradient During Percutaneous Transluminal Coronary Angioplasty
Shigeo Horinaka, MD, Suomi Hara, MD, Noriaki Tsuchiya, MD, Akihisa Yabe, MD, Hiroshi Asakawa, MD, Hiroshi Yagi, MD, Yousuke Mori, MD, and Hiroaki Matsuoka, MD
Abstract: We evaluated the influence of local myocardial ischemia induced by acute coronary occlusion during percutaneous transluminal coronary angioplasty (PTCA) on the ventricular gradients (VG) and investigated whether 2:1 alternans of VG occurs. Twenty-seven patients with angina pectoris, who had one-vessel coronary artery stenosis, were studied. The VG of each consecutive heartbeat before, during, and after PTCA over a 22-second interval was calculated using a microcomputer. The standard deviation and coefficient of variation of magnitude were used as indices of VG variability. Frequencydomain analysis of time series consisting of beat-to-beat VG magnitude for a 22-second interval was also performed by the maximum entropy method. The standard deviation and coefficient of variation of VG magnitude during PTCA were significantly greater than those before and after PTCA (P ⬍ .01, P ⬍ .01, respectively), and the indices before PTCA were also significantly greater than those after PTCA (P ⬍ .05). The maximum power spectrum peaks around 0.5 cycles/beat during PTCA were significantly greater than those after PTCA (P ⬍ .01); this suggests that the enhancement of VG alternans is reflected by 2:1 alternans of the action potential in the acute local ischemic myocardium during PTCA. Key words:Ventricular gradient variability, percutaneous transluminal coronary angioplasty, acute myocardial ischemia.
Recently, alternans of the ST segment has been observed in acute coronary artery occlusion in experimental animals (1,2), and clinically, in patients with variant angina (3,4) and during percu-
taneous transluminal coronary angioplasty (PTCA) (5,6). Nearing et al. (7), who used complex demodulation methods, found marked increases in the degree of T wave alternans that parallel the established time course of changes in vulnerability in coronary artery occlusion and reperfusion in dogs. Previously, we reported beat-to-beat variability of the parameters of the ventricular gradient (VG) of consecutive heart beats measured in Frank lead X, Y, Z scalar electrocardiograms at rest from patients with ischemic heart disease. The coefficient of variation of the VG magnitude in patients with
From the Department of Hypertension and Cardiorenal Medicine, Dokkyo University School of Medicine, Tochigi, Japan. Reprint requests: Shigeo Horinaka, MD, Department of Hypertension and Cardiorenal Medicine, Dokkyo University School of Medicine, 880 Kitakobayashi, Mibu, Tochigi, 321-0293, Japan; e-mail: [email protected]. Copyright © 2001 by Churchill Livingstone® 0022-0736/01/3402-0006$35.00/0 doi:10.1054/jelc.2001.23114
135
136 Journal of Electrocardiology Vol. 34 No. 2 April 2001 angina pectoris who had significant coronary stenosis was greater than that in healthy patients (8,9). Moreover, this index correlated with alternans of the monophasic action potentials in ischemic myocardium in experimental canine models (10). PTCA is an acute myocardial ischemic model in humans. Therefore, we investigated whether the VG variability increased and the beat-to-beat VG alternans was pronounced during PTCA.
Materials and Methods
the suitable view that best showed the lesion without overlap. The percentage of diameter stenosis of the luminal area was calculated by dividing the culprit luminal diameter by the normal luminal diameter for PTCA measured by videodiameter methods. Likewise, the percentage of stenosis of the densitometric luminal area was calculated by dividing the culprit lesion area by the normal luminal area measured by videodensitometric methods using a quantitative coronary angiographic analyzer (Cardio-500 system, Kontron, Germany) before and after PTCA.
Patient Selection
VG Alternans Analysis
Between January 1997 and April 1998, 247 patients underwent percutaneous revascularization procedures in our division. Our study comprised of 27 consecutive patients with de novo effort angina pectoris who underwent PTCA (mean age: 61 years; range, 41 to 73 years; sex: 21 men, 6 women). All patients gave written informed consent to participation in this study. Exclusion criteria were multivessel disease, hypertension, valvular heart disease, myocardial infarction, and arrhythmia.
Before, during the last 22 seconds of, and 30 minutes after PTCA. Frank lead X, Y, Z scalar ECGs were recorded with the patients in the angiographic room and simultaneously stored on the data recorder (Sony FC-14, Tokyo, Japan) with a vectorcardiograph (Fukuda Denshi, Cardio-Pro FX 3211, Tokyo, Japan). Simultaneously, standard 12-lead ECGs were also recorded to confirm the location of myocardial ischemia. During PTCA, Frank lead X, Y, Z scalar ECGs were analyzed during the last 22 seconds after balloon inflation when the ST-segment shift was stable. The mean and standard deviation (SD) of the VG magnitude for all heartbeats in the 22-second interval were obtained by computer (NEC-9801BX, Tokyo, Japan). The coefficient of variation (SD/mean ⫻ 100) of this value was used as index of beat-to-beat VG variability as previously described in detail (9,10). Furthermore, spectral analysis of VG variability was performed. Time series of the VG magnitude for all heartbeats in the 22-second interval were analyzed by the MemCalc system (Suwa Trust Co. Ltd, Tokyo, Japan) (11), which is a combination of the maximum entropy method for spectral analysis and nonlinear least squares method for fitting analysis. This new method was appropriated to the analysis of VG alternans (ie, 1:2) over a maximum interval of 22 seconds.
Cardiac Catheterization and PTCA All patients underwent diagnostic coronary angiography with a transfemoral approach with the Judkins technique and exhibited at least 75% organic stenosis of a major coronary artery on coronary angiography after intracoronary injection of a nitrate drug. The single ischemia-related artery stenosis (culprit lesion) was also confirmed by the localization of 12-lead electrocardiographic changes with spontaneous chest pain or exercise treadmill test. A few days after diagnostic coronary angiography, PTCA was performed with the conventional balloon technique with the Judkins technique as a primary treatment. Balloon inflation was repeated to dilate the culprit lesion sufficiently. The balloon was inflated each time for 60 to 120 seconds. Angiographic success was defined as less than 30% residual diameter stenosis after PTCA. Quantitative Coronary Angiography Quantitative coronary angiography (QCA) of the culprit lesion was performed in patients with patent arteries by a single operator who had no knowledge of the patients’ characteristic data and coronary morphology. The QCA analysis was performed in
Statistical Analysis All calculated data are expressed as the mean ⫾ SD. The one-way analysis of variance, which was subsequently subjected to Fisher’s protected least significant difference test for multiple comparisons, was used to determine the statistical significance of differences. The statistical analysis was conducted with a commercially available statistical software program (STSTVIEW Version 4.58, Abacus Concepatients, Berkeley, CA). Statistical significance was accepted at P ⬍ .05.
VG Alternans During PTCA •
Horinaka et al. 137
Table 1. Percentage of Culprit Lesion Stenosis, HR, Mean, Standard Deviation, Coefficient of Variation and Maximum Power Around 0.5 Cycles of the Ventricular Gradients Before, During, and After PTCA (n ⫽ 26) Before PTCA L-Area % diameter stenosis (%) Densito L-Area % stenosis (%) HR (bpm) Mean of VG magnitude (Vsec) SD of VG magnitude (%) CV of VG magnitude (%) Power spectrum density around 0.5 cycles of VG
91.5 ⫾ 6.0 86.5 ⫾ 13.0 72.4 ⫾ 12.3 74.68 ⫾ 20.72 6.76 ⫾ 1.76 9.51 ⫾ 3.15 0.43 ⫾ 0.53†
During PTCA
After PTCA
75.9 ⫾ 13.0 78.38 ⫾ 31.49 8.32 ⫾ 2.64† 12.11 ⫾ 6.18† 1
43.0 ⫾ 16.1† 25.7 ⫾ 22.6† 75.4 ⫾ 10.2 76.52 ⫾ 20.94 5.81 ⫾ 1.19*‡ 8.10 ⫾ 2.61*‡ 0.33 ⫾ 0.40‡
* P ⬍ .05 vs. before PTCA, † P ⬍ .01 vs. before PTCA, ‡ P ⬍ .01 vs. during PTCA. Data are the mean ⫾ standard deviation. Mean: mean value of all the original values, SD: standard deviation of all the original values, CV: coefficient of variation (SD/Mean ⫻ 100), Densito: densitometric, L-Area: Lumen-Area. Power spectrum density (Vsec2) around 0.5 cycles of VG: relative values were presented when the value of integration of the power spectrum density from 0.45 to 0.5 cycles/beat of VG magnitude during PTCA was 1.
Results Sixteen, 7, and 4 patients had left anterior descending coronary artery, right coronary artery, and left circumflex coronary artery stenosis, respectively. The reference diameter of the culprit lesion for PTCA was 2.51 ⫾ 0.61 mm. The percentages of luminal diameter and area stenosis of the culprit lesion for PTCA reduced from 91.5 ⫾ 6.0 and 86.5 ⫾ 13% to 43.0 ⫾ 16.1 and 25.7 ⫾ 22.6%, respectively, as shown in Table 1. There was no difference in heart rate (HR) before, during, and after PTCA (Table 1). A localized STsegment shift greater than 1 mm in 12-lead electrocardiogram induced by acute myocardial ischemia during PTCA was observed in all patients. The ST-segment or T-wave alternans (ie, 2:1 or 3:1) on 12-lead ECG were not visually observed in any patients. The VG of the QRST complexes recorded before, during, and after PTCA in one patient with coronary artery disease are shown in Figure 1. Before, during the last 22 seconds of, and 30 minutes after balloon inflation, there were no significant differences in mean VG magnitude among the three periods. The SD of VG magnitude was significantly
Fig. 1. Superimposed ventricular gradients of each heartbeat in the 22-second interval before, during, and after PTCA in one patient with coronary artery disease. Left, before PTCA; middle, during PTCA; right, after PTCA.
greater during PTCA than before or after PTCA (P ⬍ .01, P ⬍ .01, respectively) and significantly smaller after PTCA than before PTCA (P ⬍ .05). The coefficient of variation of the VG was significantly greater during PTCA than before or after PTCA (P ⬍ .01) and significantly smaller after PTCA than before PTCA (P ⬍ .05) (Table 1). The frequency power of the spectral analysis of the VG magnitude recorded before, during, and after PTCA in one patient with coronary artery disease is shown in Figure 2. For spectral analysis, the number of patients who had maximum power around 0.5 cycles/beat was 4 (15%) before, 14 (52%) during, and 2 (7%) after PTCA. The power spectral density around 0.5 cycles/beat (integration of the power spectrum from 0.45 to 0.5 cycles/beat) in all patients during PTCA was significantly greater than that before and after PTCA (P ⬍ .01, respectively) (Table 1).
Discussion This study showed that 2:1 alternans of the VG magnitudes were enhanced in acute local myocardial ischemia during PTCA even in humans. Inter-
138 Journal of Electrocardiology Vol. 34 No. 2 April 2001
Fig. 2. Frequency power spectrum of the magnitude of ventricular gradients in the 22-second interval before, during, and after PTCA in one patient with coronary artery disease. left, before PTCA; middle, during PTCA; right, after PTCA. power sp: relative value when the maximum power spectral density (Vsec2) was 1.
estingly, this phenomenon was observed at rest in 4 patients (15%). The coronary blood flow was reduced in severe stenosis because the mean stenotic rate was about 90% in our patients (12). Moreover, 2:1 alternans almost disappeared after PTCA in these cases. Thus, it is likely that myocardial electrical instability exists in severe stenosis even at rest. The VG, which was first described by Wilson et al. (13), reflects nonuniformity in ventricular action potentials. Theoretical and clinical reports have documented that the VG is independent of the activation sequence (14,15). Geselowitz (16) showed mathematically that the VG is the difference in area of the individual action potentials of the myocardium. Recently, we reported that alternans of the monophasic action potentials appeared in the lowflow myocardial area in an experimental canine coronary stenosis model, and the magnitude of these alternans significantly correlated with that of variability of the VG magnitude calculated by recording body surface leads (10). Therefore, it is likely that 2:1 alternans of the action potential of the acute local ischemic myocardium occur in acute coronary occlusion, even in humans. Although 2:1 alternans cannot be confirmed visually, as shown in Fig. 1, the alternans may be obscured by the fluctuation of the distance and angle between the heart and body surface leads by breathing. Only 50% of our patients had maximum power around 0.5 Hz in the frequency analysis. The other patients might have had weaker ischemia, or smaller areas, as previously described (17), or relatively larger fluctuation of distance and angle between heart and body surface leads. In this study, the maximum entropy method was used to prove the appearance of the 2:1 alternans of VG in myocardial ischemia. However, time series analysis comprises 2 kinds of analysis; one is the frequency-domain analysis known as spectral anal-
ysis, and the other is the time-domain analysis. Fast fourier transformation (FFT) is one of the most popular methods of spectral analysis widely used up to date. However, FFT has such disadvantage as poor resolution because of the effect of window functions and unrealistic assumptions about extending the data, needing at least 128 samples (18). Thus, FFT is inadequate for estimating the precious power spectral density from short time series data such as the small number of beats during the 22-second interval in our study. On the other hand, with respect to the time-domain series analysis in the time-domain, the complex demodulation method has been used (19). This technique calculates the instantaneous amplitude and phase by using moving-average filtered data. However, complex demodulation is inappropriate for analysis of rapidly varying fluctuations because the technique is applicable to only highly smoothed time series data. Therefore, we used the time series of the VG magnitude for all heartbeats in the 22-second interval analyzed by the MemCalc method, which is a combination of the maximum entropy method for spectral analysis and nonlinear least squares method for fitting analysis. With this new method, we achieved a reliable analysis of the relatively low frequency component (0.5 Hz) of the VG alteration on a maximum interval of 22 seconds. Recently, a relationship has been reported between T-wave alternans and ventricular tachyarrhythmia (20,21). However, no patients had ventricular arrhythmia during or after PTCA in our study. Therefore, further studies are necessary to evaluate the relation between the alternans of VG and ventricular arrhythmia. It may be necessary to elucidate the influence of voltage change in the body surface leads by the depth and frequency of the breathing, and the effect of chest pain and intracoronary nitrate drug
VG Alternans During PTCA •
administration during PTCA, and autonomic nervous system function in further study. In summary, we observed not only enhanced alternans of the VG magnitudes, but also a maximum power spectrum for this alternans around 0.5 cycles/beat in acute myocardial ischemia during PTCA. These data suggest that the 2:1 alternans of the action potentials is enhanced in acute ischemic myocardium.
References 1. Kleber AG, Janse MJ, van Capelle FJL, et al: Mechanism and time course of S-T and T-Q segment changes during acute regional myocardial ischemia in the pig heart determined by extracellular recordings. Circ Res 42:603, 1978 2. Hashimoto H, Asano M, Nakashima M: Potentiating effects of a ventricular premature beat on the alternation of ST-T complex of epicardial electrocardiograms and the incidence of ventricular arrhythmias during acute coronary artery occlusion in dogs. J Electrocardiol 17:229, 1984 3. Kleinfeld MJ, Rozanski JJ: Alternans of the ST segment in Prinzmetal’s angina. Circulation 55:574, 1977 4. Turitto G, El-Sherif N: Alternans of the ST segment in variant angina. Chest 93:587, 1988 5. Okino H, Arima S, Yamaguchi H, et al: Marked alternans of the elevated ST segment during occlusion of the left anterior descending coronary artery in percutaneous transluminal coronary angioplasty: Clinical background and electrocardiographic features. Intern J Cardiol 37:23, 1992 6. Sochanski M, Feldman T, Chua KG, et al: ST segment alternans during coronary angioplasty. Catheter Cardiovasc Diag 27:45, 1992 7. Nearing BD, Huang AL, Verrier RL: Dynamic tracking of cardiac vulnerability by complex demodulation of the T wave. Science 252:437, 1991 8. Horinaka S, Yamamoto H, Tabuchi T, et al: New electrocardiographic method for detection of isch-
9.
10.
11.
12.
13.
14. 15.
16.
17.
18.
19.
20.
21.
Horinaka et al. 139
emic heart disease: Alterations in ventricular gradient. Jpn Heart J 35:183, 1994 (suppl) Horinaka S, Yamamoto H, Tabuchi T, et al: Ventricular gradient variability: New ECG method for detection of ischemic heart disease. J Electrocardiol 28: 177, 1995 Horinaka S, Yamamoto H: Enhancement of ventricular gradient variability during acute myocardial ischemia. Int J Cardiol 65:173, 1998 Takusagawa M, Komori S, Umetani T, et al: Alteration of autonomic nervous activity in recurrence of variant angina. Heart 82:75, 1999 Gould KL, Lipscomb K: Effect of coronary stenosis on coronary flow reserve and resistance. Am J Cardiol 34:48, 1974 Wilson FN, Macleod AG, Barker PS, et al: The determination and the significance of the areas of the ventricular deflections of the electrocardiogram. Am Heart J 10:46, 1934 Burger H: A theoretical elucidation of the motion “ventricular gradient.” Am Heart J 53:240, 1957 Berkun MA, Kesselman RH, Donoso E, et al: The spatial ventricular gradient: Intermittent Wolff-Parkinson-White syndrome, intermittent left bundle branch block and ventricular premature contractions. Circulation 13:562, 1956 Geselowitz DB: The ventricular gradient revisited: Relation to the area under the action potential. IEEE Trans Eng 30:76, 1983 Kwan T, Feit A, Alam M, et al: ST-T alternans and myocardial ischemia. Catheter Cardiovasc Diag 35: 121, 1995 Smith JM, Clancy EA, Valeri CR, et al: Electrical alternans and cardiac electrical instability. Circulation 77:110, 1988 Nearing BD, Huang AL, Verrier RL: Dynamic tracking of cardiac vulnerability by complex demodulation of the T wave. Science 252:437, 1991 Nearing BD, Oesterle SN, Verrier RL: Quantification of ischaemia induced vulnerability by precordial T wave alternans analysis in dog and human. Cardiovasc Res 28:1440, 1994 Rosenbaum DS, Jackson LE, Smith JM, et al: Electrical alternans and vulnerability to ventricular arrhythmias. N Engl J Med 330:235, 1994