Coronary Artery Bypass Grafting on the Beating Heart Evaluated With Integrated Backscatter Kenichi Imasaka, MD, Shigeki Morita, MD, Ichiro Nagano, MD, Munetaka Masuda, MD, Ryuji Tominaga, MD, and Hisataka Yasui, MD Department of Cardiovascular Surgery, Faculty of Medicine, Kyushu University, Fukuoka, Japan Background. In beating heart coronary artery bypass grafting (CABG) the effect of ischemic insult during coronary occlusion could not be evaluated immediately. Using transesophageal echocardiography, myocardial performance can be evaluated with analysis of integrated backscatter. Methods. In 15 beating heart CABGs, cyclic variation (CV) of integrated backscatter of the anterior wall before, during, and after the left internal thoracic artery to left anterior descending (LAD) branch anastomosis was measured with transesophageal echocardiography. The patients were divided into two groups according to collateral vessels status (good collateral group n 6, poor collateral group n 9). Results. In all patients, CV increased significantly after revascularization (8.56 2.50 to 11.47 3.32 db, p < 0.0001). During LAD occlusion, significant decrease in CV was found in patients who had poor collateral arteries. At 15 minutes of LAD occlusion, CV decreased from the preocclusion value of 7.51 2.21 to 3.23 4.03 db (p < 0.01). Conclusions. Measurement of CV can detect the ischemic insult during coronary occlusion and the effect of revascularization in beating heart CABG. (Ann Thorac Surg 2000;70:1049 53) 2000 by The Society of Thoracic Surgeons Coronary artery bypass grafting (CABG) without an aortic cross-clamp offers an effective alternative to the patients who have diseased aortas for preventing perioperative stroke [1]. Widespread application of beating heart CABG was made possible with the introduction of new stabilizing devices. In beating heart CABG, however, it is necessary to occlude the coronary artery while the heart is beating. Interruption of coronary flow may cause ischemic damage distal to coronary occlusion. No information is available regarding whether the myocardium is exposed to ischemic insult during occlusion or whether the procedure is safe. Because of the nature of the disease, coronary occlusion is always performed distal to the significant stenosis of the target coronary artery. To address this issue, we measured the cyclic variation (CV) of integrated backscatter before, during, and after coronary occlusion, using intraoperative transesophageal echocardiogram (TEE). Ultrasonic tissue characterization is emerging as a new technique capable of detecting myocardial structural changes. The CV in backscatter power levels observed in healthy hearts [2] is altered during ischemia [3, 4]. The magnitude of such variation has been shown to be related to the severity of ischemia in animal models [5]. Recovery of CV after reperfusion occurs more quickly Presented at the Sixth Annual Cardiothoracic Techniques and Technologies Meeting 2000, Ft Lauderdale, FL, Jan 27 29, 2000. Address reprint requests to Dr Morita, Department of Cardiovascular Surgery, Faculty of Medicine, Kyushu University, 3-1-1 Maidashi, Higashi-Ku, Fukuoka 812-8582, Japan; e-mail: morita@heart. med.kyushu-u.ac.jp. than does recovery of regional systolic wall thickening [6 8]. In this study, we characterized the temporal changes in CV to determine if we could detect ischemia during the coronary occlusion and if there was an immediate recovery in CV after revascularization. We also studied the relation between the above-mentioned changes and development of collateral vessels present on preoperative coronary angiograms. Patients and Methods Patients Fifteen patients (10 men, 5 women) with a mean age of 66 years (range 50 to 75 years) who underwent CABG without arresting the heart were studied. Of the 15 patients, 12 patients underwent CABG without cardiopulmonary bypass and 3 patients underwent CABG with cardiopulmonary bypass without arresting the heart. Indications for beating heart CABG were diseased ascending aorta or significant stenosis in carotid arteries. Additional indications for off-pump CABG were endorgan dysfunction, such as renal dysfunction, respiratory failure, or left ventricular dysfunction. Operative Procedure In 12 patients with off-pump CABG, patients were prepared similar to standard CABG, which includes placement of a pulmonary artery catheter, an arterial pressure line, and a urinary drainage catheter. The operation was performed either through a limited left anterior thoracotomy (n 2), or median strenotomy (n 13). Heparin 2000 by The Society of Thoracic Surgeons 0003-4975/00/$20.00 Published by Elsevier Science Inc PII S0003-4975(00)01793-8
1050 MINIMALLY INVASIVE IMASAKA ET AL Ann Thorac Surg ECHO BACKSCATTER AND BEATING HEART CABG 2000;70:1049 53 Fig 1. Short axis-view of the left ventricle with transesophageal echocardiography in a patient who underwent coronary bypass grafting on the beating heart. The short axis view was zoomed in the area of the anterior wall of the left ventricle (the right panel). A region of interest was drawn that excluded endocardial and pericardial borders. The measurements of integrated backscatter in this area were performed throughout cardiac cycle. (LV left ventricle; ROI region of interest.) sodium (1 mg/kg) was administered intravenously before dividing arterial conduits. For patients with a limited anterior thoracotomy, stabilization of the target segment was achieved with the aid of commercially available stabilizer (US Surgical, Norwalk, CT). For those with median sternotomy, Octopus II (Medtronic, Minneapolis, MN) or US-Surgical s stabilizers were used to stabilize the target segment. A 5-minute period of coronary artery occlusion followed by a 5-minute reperfusion was performed to induce ischemic preconditioning. After the anastomosis, protamine sulfate (1 mg/kg) was administered intravenously. Three patients who underwent beating heart CABG with cardiopulmonary bypass were expected to have hemodynamic instability that made it impossible to perform off-pump CABG. Heparin sodium (3 mg/kg) was administered intravenously and cardiopulmonary bypass was instituted. The operative procedure for CABG was the same as that for off-pump CABG. Protamine sulfate (3 mg/kg) was administered after coming off bypass. In cases of multiple graft CABG, the first anastomosis was the left internal thoracic artery to left anterior descending branch (LAD) anastomosis. Measurement of Cyclic Variation of Integrated Backscatter After the induction of anesthesia and endotracheal intubation, a multiplane TEE probe was inserted into esophagus and connected to an echocardiographic system (SONOS 5500, Hewlett-Packard). The left ventricle was imaged along the short-axis plane at the level of the papillary muscle. The system is capable of providing either conventional echocardiographic images or twodimensional images in which the gray level is displayed proportional to integrated backscatter amplitude. Sixty consecutive frames of images (30 frames/second) were displayed on the monitor and were stored on an optical disk for off-line analysis. We analyzed the digitally acquired images with a software package (Acoustic Densitometry, Hewlett-Packard) incorporated into the system to construct time intensity waveforms of the integrated backscatter. The perfused areas of LAD were chosen as the region of interest, and the size of the region of interest was made as large as possible, which excluded endocardial and epicardial reflectors (Fig 1). A single observer (K.I.) adjusted the location of the site on a frame-byframe basis to keep it well within the myocardial midwall throughout the cardiac cycle. Then a time intensity waveform of the integrated backscatter was reconstructed. The transmission power, compression setting, and individual values of the time gain compensation were kept constant throughout the experimental protocol. Adjustments of the above-mentioned settings were made during the harvesting of the internal thoracic artery. We determined the magnitude (in db) of the CV of integrated backscatter as the differences between the minimal and maximal values in one cardiac cycle. Because the regional contraction in the infarct zone may not be synchronized with global contractile events, we calculated a time delay for regional CV with respect to global ventricular mechanical systole. For this purpose, we calculated normalized delay, which is defined as the interval from the upstroke of QRS complex to the nadir of the CV divided by QT interval. If the normalized delay value was larger than 1.2, we considered it indicative of asynchronous contraction or passive stretch, and multiplied the magnitude of CV by 1.0 [4, 5, 8]. If the normalized delay value was less than or equal to 1.2, the CV value was considered to be the same as the measured magnitude value. All patients underwent a TEE examination before occlusion, 5 minutes, 10 minutes, and 15 minutes after LAD occlusion, and immediately after revascularization.
Ann Thorac Surg MINIMALLY INVASIVE IMASAKA ET AL 2000;70:1049 53 ECHO BACKSCATTER AND BEATING HEART CABG 1051 Determination of Coronary Collaterals From preoperative coronary angiography, development of collateral vessels in the LAD region was classified according to Rentrop and colleagues [9]: grade 0 no filling of collateral vessels; grade 1 filling of collateral vessels without any epicardial filling of the artery to be dilated; grade 2 partial epicardial filling of the artery to be dilated by collateral vessels; and grade 3 complete epicardial filling of the artery to be dilated by collateral vessels. Patients were divided into two groups: a poor collateral group, which included 9 patients with grade 0 or 1, and a good collateral group, which included 6 patients with grade 2 or 3. Statistical Analysis Data were presented as mean and standard deviation. A paired t test was used to compare the CV values before occlusion and after revascularization. Two-way analysis of variance (ANOVA) with repeated measures was used to evaluate the effect of ischemic duration on CV. The differences in CV between two groups were evaluated by Student s t test. A p value less than 0.05 was considered to be statistically significant. Results Table 1. Patient Characteristics and Surgical Data Characteristic Poor collateral group (n 9) Good collateral group (n 6) p Age (years) a 67.9 9.3 63.3 9.2 NS Gender (male/female) 5/4 5/1 NS Previous myocardial 2/7 5/1 0.05 infarction (Y/N) (The area of LAD) NYHA a 2.3 0.9 2.3 0.5 NS LVEF (%) a 58.4 16.1 57.3 8.3 NS Medication before the operation (Y/N) Digitalis 0/9 0/6 NS blocker 5/4 3/3 NS ACE inhibitor 2/7 1/5 NS Nitrate 7/2 5/1 NS Ca channel blocker 8/1 4/2 NS Type of operation NS Off-pump CABG 6 6 On-pump beating bypass 3 0 LAD occlusion time (min) a 22.9 9.7 32.0 8.6 NS No. of distal anastomosis a 2.0 1.0 1.7 0.5 NS a Data are given as the mean the standard deviation of the mean. ACE angiotensin-converting enzyme; CABG coronary artery bypass grafting; LAD left anterior descending artery; LVEF left ventricular ejection fraction; NS not significant; NYHA New York Heart Association. Fig 2. The change in the magnitude of the cyclic variation of integrated backscatter in all patients (n 15) who underwent coronary artery bypass grafting on the beating heart. There was a significant improvement in the cyclic variation magnitude after revascularization. Error bars represent standard deviation. Patients characteristics and surgical data are shown in Table 1. Between the good and poor collateral groups, there were no differences except for the history of previous myocardial infarction. The incidence of previous myocardial infarction was higher in the good collateral group than the poor collateral group. In all patients, CV magnitude increased significantly from the preocclusion value of 8.56 2.50 db to 11.47 3.32 db immediately after revascularization (Fig 2). The preocclusion value of the good collateral group (10.08 2.26 db) was significantly higher than that of the poor collateral group (7.54 2.21 db). In the poor collateral group, serial decrease in CV magnitude was observed (7.54 2.21 db before occlusion, 5.89 1.97 db at 5 minutes of occlusion, 4.49 3.65 db at 10 minutes of occlusion, 3.23 4.03 db at 15 minutes of occlusion), whereas no serial change in CV magnitude was observed in the good collateral group (Figs 3, 4). Eight of 9 patients in the poor collateral group developed ST-T changes during LAD occlusion, whereas all patients in the good collateral group showed no change in ST-T segment (Table 2). All patients underwent postoperative coronary angiogram, which showed patent left internal thoracic arteryto-lad anastomoses. No perioperative myocardial infarction occurred in this series of patients. Comments In this study, we were able to show that in beating heart CABG the CV magnitude significantly increased after revascularization. Also base line CV magnitude before coronary occlusion in patients with poor collateral vessels was significantly smaller than those in patients with good collateral vessels. Lastly, there was a serial timedependent decline in the CV magnitude during coronary occlusion in the patients with poor collateral vessels, whereas those with good collateral vessels showed no decline in the CV magnitude during coronary occlusion. Analysis of the magnitude of CV of integrated back-
1052 MINIMALLY INVASIVE IMASAKA ET AL Ann Thorac Surg ECHO BACKSCATTER AND BEATING HEART CABG 2000;70:1049 53 Fig 3. Representative integrated backscatter wave forms in patients with poor collateral vessels (A) and with good collateral vessels (B), before occlusion (base line), at 15 minutes of left anterior descending artery (LAD) occlusion, and after revascularization. Note that there was a decline in the magnitude of cyclic variation in the patient with poor collateral vessels during LAD occlusion (A), whereas no significant change in the cyclic variation magnitude was observed in patients with good collateral vessels (B). scatter is one of the clinical applications of the ultrasonic tissue characterization. The factors that influence CV magnitude are changes in acoustic impedance, changes in fiber orientation or shape from diastole to systole, and changes in the elastic modulus during sarcomere shortening [10 12]. Several studies have shown that the change in CV magnitude reflects an ischemic insult to the myocardium. Wickline and associates [5] examined the relation between the duration of coronary occlusion and CV magnitude in an animal experiment. A rapid and complete recovery of CV magnitude was observed when the duration of coronary occlusion was 5 minutes. After 20 minutes of coronary occlusion, the recovery of CV magnitude was delayed but the recovery was complete. When the duration of coronary occlusion was 60 minutes, the recovery of the CV magnitude was incomplete (50%) even after 3 hours of reperfusion. Takiuchi and colleagues [8] showed in their clinical study that after acute myocardial infarction recovery of the CV magnitude preceded recovery of regional wall motion. These studies indicate that changes in the CV magnitude of integrated backscatter promptly detects ischemic insults, and may delineate potential beneficial effects of coronary reperfusion in the presence of wall motion abnormalities. Our results are in accordance with previous reports studying the effect of ischemia or reperfusion on CV magnitude. When myocardium was well perfused, such as after reperfusion or base line status in patients with good collateral vessels, larger CV magnitudes were observed. A unique finding in our study is a time-dependent serial decline in the CV magnitude in patients who had poor collaterals. It should be mentioned that in these patients there was no serial decline in blood pressure or cardiac output that produced hemodynamic deterioration. The only conventional parameter that might have detected an ischemic insult was a ST-T change in the chest lead electrocardiogram. However, there was no time-dependent change in the amount of ST-T change.
Ann Thorac Surg MINIMALLY INVASIVE IMASAKA ET AL 2000;70:1049 53 ECHO BACKSCATTER AND BEATING HEART CABG 1053 in which myocardial fibers in the area were oriented nearly perpendicular to the ultrasound beam when TEE was used. For the analysis of the posterior wall, the distance between the TEE probe and the posterior wall is too small for satisfactory analysis of the integrated backscatter. We therefore did not perform the analysis for the posterior wall of the left ventricle. In conclusion, we were able to demonstrate that the change in the CV magnitude of integrated backscatter reflects ischemia or reperfusion of the myocardium during the CABG on the beating heart. The analysis of integrated backscatter is a promising tool to predict ischemic damage to the myocardium, and to prevent hemodynamic deterioration. Fig 4. The magnitude of the cyclic variation of integrated backscatter at base line, 5, 10, 15 minutes during the left anterior descending artery occlusion, and after revascularization. Two-way analysis of variance with repeated measures on one factor showed that the value of the magnitude of the cyclic variation of integrated backscatter was significantly different between the two groups (p 0.01). At base line, the cyclic variation of integrated backscatter of the good collateral group was significantly higher than that of poor collateral group. Error bars represent standard deviation. The result of serial decline in CV magnitude indicates that further study is warranted to determine the threshold of CV magnitude needed to predict hemodynamic deterioration during the coronary occlusion. Another future application of this technique is to determine the effectiveness of ischemic preconditioning. The effect of ischemic preconditioning could be evaluated by the rate of decline in the CV magnitude. Our preliminary study showed that in some patients the CV magnitude after 5 minutes of ischemic preconditioning was smaller than the CV magnitude after 5 minutes of coronary occlusion during the anastomosis, which indicates enhanced ischemic tolerance. Further study is warranted to evaluate these issues. Technical limitations should be mentioned. The magnitude of the CV of integrated backscatter is dependent on the angle between fiber orientation and the ultrasonic beam. Data in this study were obtained exclusively from the anterior wall of the left ventricle in a short-axis view, Table 2. Cases of ST Changes During LAD Occlusion Group ST changes ( ) ST changes ( ) Poor collateral group (n 9) 8 (88.9%) 1 (11.1%) Good collateral group (n 6) 0 (0%) 6 (100%) LAD left anterior descending artery. References 1. Borst C, Grundeman PF. Minimally invasive coronary artery bypass grafting: an experimental perspective. Circulation 1999;99:1400 3. 2. Madaras EI, Barzilai B, Perez JE, Sobel BE, Miller JG. Changes in myocardial backscatter through the cardiac cycle. Ultrason Imaging 1983;5:229 39. 3. Barzilai B, Madaras EI, Sobel BE, Miller JG, Perez JE. Effects of myocardial contraction on ultrasonic backscatter before and after ischemia. Am J Physiol 1984;247:H478 84. 4. Vitale DF, Bonow RO, Gerundo G, et al. Alterations in ultrasonic backscatter during exercise-induced myocardial ischemia in humans. Circulation 1995;92:1452 7. 5. Wickline SA, Thomas LJ III, Miller JG, Sobel BE, Perez JE. Sensitive detection of the effects of reperfusion on myocardium by ultrasonic tissue characterization with integrated backscatter. Circulation 1986;74:389 400. 6. Milunski MR, Mohr GA, Wear KA, et al. Early identification with ultrasonic integrated backscatter of viable but stunned myocardium in dogs. J Am Coll Cardiol 1989;14:462 71. 7. Milunski MR, Mohr GA, Gessler CJ, et al. Ultrasonic tissue characterization with integrated backscatter: acute myocardial ischemia, reperfusion, and stunned myocardium in patients. Circulation 1989;80:491 503. 8. Takiuchi S, Ito H, Iwakura K, et al. Ultrasonic tissue characterization predicts myocardial viability in early stage of reperfused acute myocardial infarction. Circulation 1998;97: 356 62. 9. Rentrop KP, Cohen M, Blanke H, Phillips RA. Changes in collateral channel filling immediately after controlled coronary artery occlusion by an angioplasty balloon in human subjects. J Am Coll Cardiol 1985;5:587 92. 10. Wickline SA, Thomas LJ III, Miller JG, Sobel BE, Perez JE. A relationship between ultrasonic integrated backscatter and myocardial contractile function. J Clin Invest 1985;76: 2151 60. 11. Aygen M, Popp RL. Influence of the orientation of myocardial fibers on echocardiographic images. Am J Cardiol 1987; 60:147 52. 12. O Brien PD, O Brien WD, Rhyne TL, Warltier DC, Sagar KB. Relation of ultrasonic backscatter and acoustic propagation properties to myofibrillar length and myocardial thickness. Circulation 1995;91:171 5.