F rent techniques of hypothermic cardioplegic arrest

Transcription

1 Does Hypothermic Fibrillatory Arrest Improve Myocardial Protection During Emerrrencv Revascularization? U J Peter S. Greene, MD, Duke E. Cameron, MD, Elaine M. Griffiths, MB, Joseph M. DiNatale, BS, and Timothy J. Gardner, MD Division of Cardiac Surgery, The Johns Hopkins Medical Institutions, Baltimore, Maryland Hypothermic fibrillatory arrest (HFA) was compared with conventional hypothermic cardioplegic arrest (HCA) in a model of acute regional ischemia. In 2 pigs, the left anterior descending coronary artery was occluded for 3 minutes before cardiopulmonary bypass. In the HCA group (n = lo), the heart was arrested with a hyperkalemic cold crystalloid solution, whereas in HFA animals (n =, the heart was vented and allowed to fibrillate spontaneously without cross-clamping. Miniature ph probes monitored intramyocardial ph during 45 minutes of arrest (HCA or HFA, both with systemic and topical myocardial cooling) and during two hours of coronary reperfusion. Hypothermic fibrillatory arrest did not ameliorate the acidosis in the ischemic (left anterior descending) region; indeed, after two hours of coronary reperfusion, there was a trend toward more acidosis in the postischemic left anterior descending territory in the HFA group. However, HFA did prevent acidosis in the nonischemic (left circumflex) territory. Infarct size expressed as percent of region at risk was 8.% 2 3.2% (mean f standard error of the mean) in the HCA animals and 8.8% 2 4.4% in the HFA animals. These results demonstrate that HFA offers no advantage over HCA in protection of regionally ischemic myocardium in a model with minimal collateral circulation. (Ann Thorac Surg 989;48:3842) or most patients undergoing cardiac operations, cur- F rent techniques of hypothermic cardioplegic arrest (HCA) preserve ventricular function well. Profound cardiac cooling and the immediate chemical induction of electromechanical arrest permit extended periods of cardiac arrest with excellent metabolic and mechanical recovery. However, in certain subsets of patients these techniques of myocardial protection remain inadequate. Patients with acute and ongoing myocardial ischemia, such as those with an evolving myocardial infarction or those with complications from failed percutaneous transluminal coronary angioplasty, are at high risk for low cardiac output after emergency revascularization. Recent clinical reports on urgent surgical revascularization for acute ischemia have documented higher operative mortality, higher perioperative myocardial infarction rate (even with rapid transit to the operating room), and greater need for postoperative inotropic or mechanical support [l, 2. Previous studies from this laboratory have demonstrated that when ischemic myocardium is subjected to hypothermic global ischemic arrest, myocardial salvage is adversely affected [3, 4. Therefore, any technique that minimizes or obviates the global ischemic insult of aortic cross-clamping might result in superior myocardial protection. Several clinical series using the technique of hypother- Presented at the Thirty-Fifth Annual Meeting of the Southern Thoracic Surgical Association, Marco Island, FL, Nov -2, 988. Address reprint requests to Dr Gardner, Johns Hopkins Hospital, 6 North Wolfe St, Blalock 68, Baltimore, MD 225. mic fibrillatory arrest (HFA) have suggested that HFA may be an effective alternative strategy for myocardial protection [5-7. Typically, the technique involves performance of proximal aortosaphenous anastomoses before cardiopulmonary bypass (CPB), followed by rapid systemic cooling on CPB. Systemic arterial pressure on CPB is kept high, usually above 8 to 9 mm Hg. Topical myocardial cooling and elective ventricular fibrillation provide a nearly still heart for the distal coronary anastomoses [5]. Advocates propose that HFA avoids the global ischemic insult of aortic cross-clamping and leads to earlier reperfusion of the most ischemic region. The excellent clinical results reported in high-risk patients undergoing emergency procedures [8] led us to examine whether this technique compares favorably with standard HCA in an animal model of acute regional ischemia. This model simulates emergency surgical revascularization for acute myocardial ischemia. Material and Methods Male and female farm pigs weighing 25 to 35 kg were premedicated with intramuscular ketamine hydrochloride (4 mg/kg) and anesthetized with intravenous sodium pentobarbital (3 mg/kg). Animals were intubated and maintained on a Harvard ventilator, with adjustments made to achieve a Po, of to 2 mm Hg, Pco, near 4 mm Hg, and ph of 7.4. Polyethylene catheters were inserted in the external jugular vein and common carotid artery to monitor central venous and arterial pressure. Intravenous bretylium tosylate (2 mg/kg) was adminis- 989 by The Society of Thoracic Surgeons /89/$3.5

2 Ann Thorac Surg 989;48:3842 GREENE ET AL 39 tered to prevent ventricular fibrillation during median sternotomy and exposure of the heart. Regional myocardial protection was assessed by on-line monitoring of myocardial tissue ph. Subminiature ph electrodes (Ingold Electrodes, Inc, Andover, MA) were placed at two locations near the center of the region at risk (left anterior descending distribution) and in one location in the posterolateral left ventricular wall (left circumflex distribution). The Khuri Regional Tissue Ischemia Monitor (Vascular Technology, Inc, North Chelmsford, MA) [9, calculated actual tissue ph based on prior calibration and simultaneous regional myocardial temperature. Thirty minutes of equilibration time were allowed after placement of the ph probes to minimize effects of local tissue damage produced by ph probe insertion. To create reversible regional ischemia, the left anterior descending coronary artery was occluded with an atraumatic vascular clip just distal to the first large diagonal branch. Hearts that fibrillated during ischemia were defibrillated quickly with a direct current countershock. After 3 minutes of normothermic regional ischemia, all animals were heparinized and connected for CPB using a 28F venous cannula in the right atrium and a 4F arterial cannula in the left femoral artery. A Harvey H-5 bubble oxygenator primed with L of homologous blood completed the CPB circuit. In the HCA animals (n = lo), the aorta was crossclamped and the heart arrested with an aortic root injection of 25 ml of 4 C hyperkalemic crystalloid cardioplegic solution. In the HFA group (n = lo), the heart was vented via the left ventricular apex and allowed to fibrillate spontaneously without cross-clamping. In both groups, topical iced saline solution maintained myocardial temperature below 5"C, as measured by needle thermistor probes. All animals were cooled to a systemic temperature of 28 C. In both groups, after the 45-minute arrest period the coronary artery occluder was removed, simulating surgical revascularization. Animals were weaned from CPB over the next 45 minutes without the use of inotropic drugs. Frequent measurements of arterial blood gases were made to insure normal values; sodium bicarbonate was given to correct metabolic acidosis if present. After two hours of reperfusion, the ph probes were removed and the region at risk for ischemic injury was determined by intracoronary injection of monastryl blue dye (B.W. 372-P phthalo-blue B, DuPont De Nemours Company, Wilmington, DE). The right ventricle and atria were removed and the left ventricle was cut into -cm slices in a plane parallel to the atrioventricular groove. Slices were incubated in a % solution of nitro blue tetrazolium chloride (Polysciences, Warrington, PA) for 2 minutes at 37 C to determine if infarction had occurred. Cross-sectional areas of the region at risk and infarction were determined by planimetry. Values are reported as mean * the standard error of the mean. The sizes of the region at risk and the infarction were compared by the nonparametric Wilcoxon rank sum test. All regional ph values were corrected to 37 C by the following formula: ph(corrected) = ph(measured) +.5 (T - 37), where T is myocardial temperature in Ibpemulon Z Time (min) Fig. Myocardial temperature in the nonischemic region (left circumflex territory) during cardioplegic versus fibrillatory arrest. degrees centigrade. The ph trends were compared by an analysis of variance for repeated measures. Arrest and reperfusion phases were examined separately. All animals received humane care according to the "Guidelines for the care and use of laboratory animals" prepared by the National Academy of Science and published by the National Institutes of Health (NIH publication No , Revised 985). Results There were no significant differences between HFA and HCA groups in physiological variables, including hematocrit, arterial blood gases, and mean perfusion pressure on CPB. All but two animals (one in each group) had at least one episode of ventricular fibrillation during the 3-minute period of regional ischemia. In all cases, it was possible to restore sinus rhythm by defibrillation before the myocardial arrest period. No animals were excluded from the study. Figure depicts the time course of myocardial temperature for the two groups recorded at 5-minute intervals during the study. In the nonischemic area of the left ventricle there were no statistically significant differences in temperature between HFA and HCA groups during the prebypass, arrest, or reperfusion periods. However, there was a trend toward warmer myocardial temperatures in the HFA group during the arrest procedure. In Figure 2 are the average temperatures of the two probes placed in the ischemic region. No significant temperature differences were seen between the HFA and HCA groups during any phase of the study, as determined by analysis of variance for repeated measures. Displayed in Figure 3 are the ph measurements made in the nonischemic area, corrected to 37 C. Although the curves appear similar during the period of regional ischemia, there was significantly greater myocardial acidosis during the arrest and reperfusion periods in the HCA group. At the conclusion of the myocardial arrest period, mean ph in the HCA group was 6.92 &.4 compared with 7.2 *.6 in the HFA group. This trend of more acidosis in HCA animals persisted in the reperfusion period, although by the end of reperfusion, mean ph in both groups converged to I

3 4 GREENE ET AL HYI'OTHERMIC FIBRILLATORY ARREST Ann Thorac Surg 989;48:38-Q d i - Cardioplegic h a t A-A flmtoqhd n - o 26 w ni 26 iw im m -8 (h) Fig 2. Myocardial temperature in the ischemic region (left anterior descending territory) during cardioplegic versus fibrillatory arrest. Figure 4 shows the time course of tissue ph in the ischemic region. At the conclusion of myocardial arrest, myocardial ph in the two groups was similar: 5.87 f. for HCA animals and 5.88 *.5 for the HFA animals. However, during coronary reperfusion there was a trend toward more myocardial acidosis in the HFA group, with a final ph of 6.92 f.9 compared with 7.9 k.6 in HCA animals. The two experimental groups had comparable areas of the left ventricle at risk for ischemic injury: 3.5% +- 2.% in HCA and 3.9% -I- 2.4% in HFA animals (Fig 5). Infarct sizes measured by vital staining and expressed as percent of the region at risk were also similar: 8.% f 3.2% in the HCA and 8.8% f 4.4% in the HFA animals. Myocardial necrosis was not seen in any area outside the region at risk. To evaluate the usefulness of regional intramyocardial ph measurements, the minimum ph reached during ischemia and a mean ph during reperfusion were correlated with the eventual infarct size. As shown in Figures 6 and 7, these correlations were both quite poor (I = -.3 and Y = -.25, respectively). Comment Surgical revascularization for rapidly progressive unstable angina, evolving myocardial infarction, and persistent - Cardioplegic Arrest.-. Fibrillatory Arrest 7.6 E on Time Fig 3. Myocardial ph in the nonischemic region (left circuniflex territory) during cardioplegic versus fibrillatory arrest Cardioplegic Arrest 7.8 Fibrillatory Arrest g# W Time (min) Fig 4. Myocardial ph in the ischemic region (left anterior descending territory) during cardioplegic uersus fibrillatory arrest. ischemia after infarction has a higher overall mortality rate than routine coronary revascularization [ -5]. A growing body of experimental and clinical evidence has helped to explain the increased risk of perioperative injury. With ongoing ischemia before operation, the myocardium has less metabolic reserve and, therefore, increased susceptibility to further injury. Although standard intraoperative myocardial protective techniques have been refined to minimize gobal ischemic injury, the added insult of aortic cross-clamping may allow the evolving infarction to continue to enlarge, creating greater irreversible injury. In addition, reperfusion of the ischemic myocardium likely contributes to further necrosis due to lipid peroxidation of cellular membranes by oxygen radicals, the influx of calcium into impaired cells, and the immediate functional demands placed on the injured myocardium during reperfusion. Studies carried out in our laboratory using a porcine model of reversible regional ischemia have allowed us to examine the role of various interventions that might limit both infarct extension and postischemic ventricular dysfunction. In hearts subjected to mid-left anterior descending coronary artery occlusion followed by global ischemia and reperfusion, the administration of superoxide dismutase was effective in reducing infarct size [6] and in z LVZO : I ' 6 6 Cardioplegic Arrest m Fibrillatory Arrest 36 so 26 ~ % RR Region At Risk Infarct Size Fig 5. Region at risk (RR) for infarction and size of infarction after regional ischemia: cardioplegic uersus fibrillatory arrest. (Left) Region at risk expressed as percent of left ventricular (LV) mass, cardioplegic versus fibrillatory arrest. (Right) Infarct size expressed as percent of region at risk, cardioplegic uersus fibrillatory arrest. 6 6

4 Ann Thorac Surg 989;48:38-42 GREENE ET AL 4 so Ll 9- El K 2- - I r = Minimum ph during Arrest Period Fig 6. Relationship of minimum myocardial ph during myocardial arrest to eventual infarction size (data from both cardioplegic and fibrillatory arrest groups). limiting postischemic ventricular dysfunction [7]. Even more dramatic improvement was observed in this porcine model when a superoxide dismutase-enriched solution was administered via the coronary sinus during the global ischemic period. This porcine model of regional ischemia has also enabled definition of the time course of infarction and the reliability of vital staining techniques for measurement of infarct size [9]. Although the frequency of ventricular tachyarrhythmias makes the pig a challenging experimental animal, we prefer the porcine model because the end-artery coronary pattern more closely resembles the circulation of the human heart than does the more extensively collateralized canine heart. This experimental model of emergency revascularization is analogous to failure of percutaneous transluminal coronary angioplasty (sudden and total coronary occlusion) in a patient lacking well-developed collateral myocardial blood flow. The present study was undertaken to examine the hypothesis that hypothermic fibrillatory arrest is peferable to cardioplegic arrest in hearts with preexisting regional ischemia. Theoretically, the operative strategy of HFA avoids the additive insult of global ischemia as the aorta is not cross-clamped and coronary artery perfusion (except in the region at risk) is maintained. If there are zones of myocardium within the region at risk that are on the threshold of necrosis but are still salvageable, maintenance of myocardial perfusion during bypass grafting might result in ultimate infarct reduction. The safety and clinical efficacy of HFA has been well documented in clinical reports by Akins and associates [5-8, In a recent review of, consecutive patients undergoing nonemergency coronary artery bypass grafting with HFA between 979 and 984, Akins and Casoll [22] reported a hospital mortality of only.4% and a perioperative myocardial infarction rate of.8%. Survival at 5 years in this patient group was 9.6% and actuarial event-free rates at 5 years were 97.7% for myocardial infarction, 99.4% for percutaneous transluminal coronary angioplasty, 99.5% for reoperative revascularization, and 88.6% for all combined morbidity and mortality. Even more remarkable were results reported in high-risk emergency patients managed with HFA [8]. In 27 consecutive patients who required emergency myocardial revascularization, there was only one hospital death (.8%) and a single perioperative myocardial infarction (.8%). Actuarial survival at 45 months was 9.8%, and the mean New York Heart Association classification for the 7 long-term survivors was.2. The definition of an emergency operation in this study was broad. Included as emergencies were patients whose operations were performed outside normal operating room hours and patients whose deteriorating clinical condition required cancellation of an elective procedure. Nonetheless, this report and other successes with patients having left ventricular aneurysms [2] demonstrate that HFA can be used safely in a broad range of clinical circumstances. As Akins and Casoll [22] note, the efficacy of HFA challenges the view that refinements in myocardial preservation with HCA have been the reason for improved results with coronary bypass grafting. In addition to avoiding the global ischemia that accompanies cardioplegic techniques, HFA can be expected to have other possible benefits in special circumstances. Avoidance of aortic cross-clamping is especially critical in patients with severe calcification of the ascending aorta. This problem has become increasingly frequent as older patients become surgical candidates. Also, in patients at risk of postoperative right ventricular dysfunction, HFA would seem ideal because coronary perfusion can be maintained during revascularization. However, there are theoretical problems with the clinical use of HFA. As described by Akins, in contrast to the usual technique of HCA, with HFA proximal aortosaphenous anastomoses must be performed before CPB, requiring the placement of a partial occlusion clamp on the aorta. Potentially, this may increase afterload and oxygen demand on a heart already stressed by inadequate coronary blood flow [23]. These partial aortic occlusions also introduce risk of embolization of arteriosclerotic debris to the coronary and cerebral circulations. Also, with HFA the surgeon loses the bloodless, quiet surgical field achieved with cardioplegic arrest. Few experimental data have been available to compare so- 3 tr 2- - r = Mean ph during Reperfusion Fig 7. Relationship of mean ph during coronary reperfusion to eventual infarction size (data from both cardioplegic and fibrillatory arrest groups).

5 42 GREENE ET AL Ann Thorac Surg 989;48: HFA and HCA [23]. This study was undertaken because we believed that if HFA provided better salvage of myocardial tissue with evolving ischemia, disadvantages of the technique could be tolerated. In a porcine model of regional ischemia and simulated emergency revascularization, HFA was no better than HCA at preventing necrosis in jeopardized myocardial tissue. Furthermore, although there was evidence that HFA prevented intramyocardial acidosis, as assessed by continuous ph monitoring, this benefit was observed only in nonischemic regions of the heart and not in areas of preexisting ischemia which were at risk of infarct extension. Interestingly, there was no marked difference between groups in the temperatures recorded in the nonischemic region, despite the HFA group having no cross-clamp. This likely reflects the remarkable efficiency of topical cooling, especially in the thin-walled porcine myocardium. These data confirm clinical reports that HFA is a safe method of myocardial protection: postoperative infarcts were no larger in animals subjected to HFA. However, HFA was not superior to hypothermic cardioplegic techniques in this model of acute regional ischemia in which there is little or no collateral coronary blood flow. The most useful role for HFA may be in situations in which the aorta cannot be safely cross-clamped because of severe aortic calcification and the attendant high risk of embolic injury. Supported in part by the National Heart, Lung, and Blood Institute (grant R 944) and the Anselm Talalay Cardiovascular Research Fund. References. Connor AR, Vliestra RE, Schaff HV, et al. Early and late results of coronary artery bypass after failed angioplasty: actuarial analysis of late cardiac events and comparison with initially successful angioplasty. J Thorac Cardiovasc Surg 988; Parsonnet V, Fisch D, Gielchinsky, et al. Emergency operation after failed angioplasty. J Thorac Cardiovasc Surg 988; 96:9% Horneffer PJ, Gott VL, Gardner TJ. The deleterious effects of global ischemia on an evolving myocardial infarction. Surg Forum 984;35: Horneffer PJ, Gott VL, Gardner TJ. Reperfusion before global ischemic arrest improves the salvage of infarcting myocardium. Ann Thorac Surg 985;4: Akins CW. Noncardioplegic myocardial preservation for coronary revascularization. J Thorac Cardiovasc Surg 984;88: Akins CW. Early and late results following elective isolated myocardial revascularization during hypothermic fibrillatory arrest. Circulation 985;72(Suppl 3): Akins CW, Block PC. Surgical intervention for failed percutaneous transluminal coronary angioplasty. Am J Cardiol 984;53: 8C- C. 8. Akins CW. Early and late results following emergency isolated myocardial revascularization during hypothermic fibrillatory arrest. Ann Thorac Surg 987;43: Khuri SF, Marston WA. On-line metabolic monitoring of the heart during cardiac surgery. Surg Clin North Am 985; 65:43%53.. Khuri SF, Josa M, Marston WA, et al. First report of intramyocardial ph in man:. Assessment of adequacy of myocardial preservation. J Thorac Cardiovasc Surg 983; 86: Cosgrove DM, Loop FD, Lytle BW, et al. Primary myocardial revascularization: trends in surgical mortality. J Thorac Cardiovasc Surg 984;88: Berg R, Selinger SL, Leonard JJ, Grunwald RP, OGrady WP. Immediate coronary artery bypass for acute evolving myocardial infarction. J Thorac Cardiovasc Surg 98;8: DeWood MA, Spores J, Berg R, et al. Acute myocardial infarction; a decade of experience with surgical reperfusion in 7 patients. Circulation 983;68(Suppl 2): Phillips SJ, Kongtahworn C, Skinner JR, Zeff RH. Emergency coronary artery reperfusion: a choice therapy for evolving myocardial infarction. Results in 339 patients. J Thorac Cardiovasc Surg 983;86: DiSesa VJ, ONeil AC, Bitran D, et al. Aggressive surgical management of post-infarction angina: results of myocardial revascularization early after transmural infarction. Tex Heart Inst J 985;2: Horneffer PJ, Gott VL, Gardner TJ. Intraoperative use of superoxide dismutase improves salvage of infarcting myocardium. Surg Forum 986;37: Johnson DL, Horneffer PJ, DiNatale JM, Gott VL, Gardner TJ. Free radical scavengers improve functional recovery of stunned myocardium in a model of surgical coronary revascularization. Surgery 987;2: Horneffer PJ, Gott VL, Gardner TJ. Retrograde coronary sinus perfusion prevents infarct extension during intraoperative global ischemic arrest. Ann Thorac Surg 986;42: Horneffer PJ, Gott VL, Gardner TJ. The rapid evolution of myocardial infarction in an end-artery model. Circulation 987;76V Akins CW. Reoperation for stenotic saphenous vein bypass grafts without cardiopulmonary bypass. Ann Thorac Surg 983;35: Akins CW. Resection of left ventricular aneurysm during hypothermic fibrillatory arrest without aortic occlusion. J Thorac Cardiovasc Surg 986;9: Akins CW, Casoll DL. Event free survival following nonemergency myocardial revascularization during hypothermic fibrillatory arrest. Ann Thorac Surg 987;43: Levitsky S. Is hypothermic fibrillatory arrest a rational alternative to cardioplegic arrest? Ann Thorac Surg 987;43:27-8.