Lactate Release During Reperfusion Predicts Low Cardiac Output Syndrome After Coronary Bypass Surgery

Transcription

1 Lactate Release During Reperfusion Predicts Low Cardiac Output Syndrome After Coronary Bypass Surgery Vivek Rao, MD, PhD, Joan Ivanov, RN, MSc, Richard D. Weisel, MD, Gideon Cohen, MD, Michael A. Borger, MD, and Donald A. G. Mickle, MD Division of Cardiovascular Surgery and Department of Clinical Biochemistry, Toronto General Hospital, Centre for Cardiovascular Research and University of Toronto, Toronto, Ontario, Canada Background. Cardioplegic arrest induces anaerobic myocardial metabolism with a net production of lactate from glycolysis. However, persistent lactate release during reperfusion suggests a delayed recovery of normal aerobic metabolism and may lead to depressed myocardial function necessitating inotropic or intraaortic balloon pump support (low output syndrome [LOS]). We examined the relation between perioperative myocardial metabolism and postoperative clinical outcomes in patients undergoing isolated coronary artery bypass surgery (CABG). Methods. We reviewed 623 patients who were enrolled in clinical studies evaluating perioperative myocardial metabolism between 1983 and Arterial and coronary sinus blood samples were obtained intraoperatively to assess myocardial metabolism. Clinical data regarding patient demographics and postoperative outcomes were prospectively collected and entered into our institutional database. Results. Low output syndrome developed in 36 patients (5.8%). Myocardial lactate release was higher in these patients compared with those who did not develop postoperative LOS. Advanced age and poor preoperative left ventricular function were independent predictors of lactate release during reperfusion. Persistent lactate release after 5 minutes of reperfusion was the only independent predictor of postoperative LOS in this low-risk population. Conclusions. Persistent lactate release during reperfusion occurs in a significant proportion of low-risk patients undergoing isolated CABG and is an independent predictor of postoperative low cardiac output syndrome. Persistent lactate release during reperfusion suggests a delayed recovery of aerobic myocardial metabolism and may be related to intraoperative misadventure or inadequate myocardial protection. Myocardial lactate release may be useful as an alternative end-point in clinical trials evaluating perioperative myocardial protection. (Ann Thorac Surg 2001;71: ) 2001 by The Society of Thoracic Surgeons Despite apparently adequate blood cardioplegic protection, sensitive indices reveal delayed myocardial metabolic and functional recovery after surgery [1 8]. In previous clinical trials at our institution [5 8], we have observed that aortic cross-clamping induces anaerobic myocardial metabolism with a net release of lactate. A return to normal aerobic lactate extraction may be delayed for as long as 4 hours after cross-clamp removal. Persistent anaerobic glycolysis during reperfusion may be associated with intraoperative ischemic injury. Kobayashi and Neely [9] demonstrated that the activity of mitochondrial pyruvate dehydrogenase (PDH) was inhibited within 2 minutes of reperfusion after global ischemia. PDH regulates the conversion of pyruvate to acetyl-coa and the inhibition of this metabolic pathway can lead to persistent lactate production with a resultant acidosis. In addition to the decreased efficiency of anaerobic ATP production, Weiss and colleagues [10] demonstrated that ATP produced by oxidative phosphorylation Accepted for publication March 13, Address reprint requests to Dr Weisel, Division of Cardiovascular Surgery, The Toronto Hospital, 200 Elizabeth St, E14215, Toronto, Ontario, Canada M5G 2C4; richard.weisel@uhn.on.ca. was preferentially utilized for myocardial contractile function. Therefore, it is possible that persisent anaerobic lactate release during reperfusion may lead to inadequate ventricular function in the early postoperative period. We reviewed 623 patients who were enrolled in clinical trials at our institution evaluating various techniques of myocardial protection. Metabolic and clinical data were collected prospectively on all patients. We have previously identified the clinical predictors of low output syndrome (LOS) in 4,558 patients undergoing isolated coronary artery bypass grafting [11]. Our objective definition of LOS involves the use of inotropic or intraaortic balloon pump support for more than 30 minutes in the intensive care unit in order to maintain adequate hemodynamics after appropriate adjustment of preload and afterload. The development of LOS is associated with a higher prevalence of perioperative myocardial infarction and operative mortality. The current low rates of perioperative infarction and mortality associated with isolated CABG [12] have required large sample sizes in order to demonstrate clinically significant differences between myocardial protective strategies. The use of LOS as an 2001 by The Society of Thoracic Surgeons /01/$20.00 Published by Elsevier Science Inc PII S (01)

2 1926 RAO ET AL Ann Thorac Surg LACTATE RELEASE AFTER BYPASS SURGERY 2001;71: alternative clinical outcome may improve the statistical power to observe differences between treatment groups. We believe that myocardial lactate release during reperfusion is a marker of perioperative ischemic injury, and may predict the development of postoperative LOS. The purpose of this study was to examine the relationship between lactate release during reperfusion and postoperative LOS. We reviewed a predominantly low risk patient population in which significant morbidity and mortality was not anticipated in order to determine if myocardial lactate release can be used as a sensitive predictor of inadequate ventricular function in the early postoperative period. Material and Methods Six hundred and twenty-three patients agreed to participate in prospective clinical trials at our institution evaluating alternate myocardial protective strategies between 1983 and All patients signed an informed consent form approved by our institutional ethics committee. Our conduct of operation and techniques of myocardial protection have been previously described in detail [5 8]. Cardioplegic arrest was achieved using a mixture of systemic blood and crystalloid cardioplegia in a 2:1 or 4:1 ratio. After initial antegrade arrest, cardioplegia was delivered either retrograde into the coronary sinus or antegrade into completed vein grafts. All distal and proximal coronary anastomoses were performed during a single aortic cross-clamp period. Arterial and coronary sinus blood samples were obtained during the cross-clamp period and at prespecified intervals during reperfusion. Specifically, samples were obtained immediately after aortic cross-clamp removal, 5 and 10 minutes after cross-clamp removal, and 5 and 10 minutes after discontinuation of cardiopulmonary bypass. Low Cardiac Output Syndrome Low output syndrome was diagnosed if the patient required an intraaortic balloon pump either in the operating room or in the intensive care unit for hemodynamic compromise. Low output syndrome was also diagnosed if the patient required inotropic medication to maintain the systolic blood pressure greater than 90 mm Hg and the cardiac output greater than 2.2 L min 1 m 2 for at least 30 minutes in the intensive care unit after correction of all electrolyte and blood gas abnormalities, and after adjusting the preload to its optimal value [13]. Afterload reduction was also attempted when possible. Patients received either dopamine, dobutamine, milrinone or epinephrine. Patients who received less than 4 g/kg of dopamine to increase renal perfusion were not considered to have LOS. Patients who received vasoconstricting medication because of a high cardiac output ( 2.5 L min 1 m 2 ) and low peripheral resistance were also not considered to have LOS. Biochemical Measurements Arterial and coronary sinus blood samples were assayed to estimate the myocardial consumption of oxygen and production of lactate or acid. Oxygen content (O 2 Con) was calculated from the formula O 2 Con 1.39Hgb SaO po 2, where Hgb is the hemoglobin concentration, SaO 2 is the oxygen saturation and po 2 is the partial pressure of oxygen. Myocardial oxygen extraction (O 2 Ex) was calculated as the arterial or cardioplegic oxygen content minus the coronary venous oxygen content and myocardial oxygen consumption (MVO 2 ) determined after correcting for coronary flow. Measurements were made at 37 C and then corrected to the cardioplegic temperature at the time of sampling. Blood lactate concentration was determined using a commercially available assay (Rapid Lactate Stat Pack kit; Calbiochem-Behring, La Jolla, CA). Lactate extraction (LEx) was calculated as the difference between arterial and coronary sinus lactate content. Negative lactate extraction is expressed as net lactate release. Lactate consumption or production (MVL) was determined after correcting lactate extraction for coronary blood flow. The concentration of hydrogen ion (H ) in blood was determined by converting the measured ph value to [H ] by the formula: [H ] antilog (-ph). Measurements were made at 37 C and corrected to the myocardial temperature at the time of sampling. Myocardial acid production was calculated as the difference in H concentration between arterial and coronary sinus blood corrected for coronary flow. Creatine Kinase Measurement An antibody inhibition technique was employed to measure the MB isozyme of creatine kinase (CK-MB). Sequential CK-MB measurements were performed at 2, 4, 8, 16, 24, and 48 hours after the removal of the aortic cross-clamp. Integration of the area under the concentration-time curve for CK-MB within the first 48 hours postoperatively allowed calculation of the total CK-MB release, expressed as units (IU) hours. Statistical Analysis Statistical analysis was performed using the SAS program (SAS Institute, Cary, NC). Categorical variables were analyzed with 2 or Fisher s exact test as appropriate. Continuous variables were analyzed using analysis of variance (ANOVA). Stepwise multiple logistic regression was employed to determine the independent predictors of LOS and lactate release during reperfusion. Variables were included in the multivariable model if their univariate p value was less than 0.15 or if they were of known clinical importance. Multivariable models were evaluated by comparing the Hosmer-Lemeshow goodness of fit statistic and the area under the receiver operating characteristic curve [11]. Results Thirty-six patients (5.8%) developed postoperative LOS. Figure 1 illustrates the relationship between myocardial lactate release, oxygen extraction, and acid release and the development of postoperative LOS. Myocardial lactate release at cross-clamp removal was higher in pa-

3 Ann Thorac Surg RAO ET AL 2001;71: LACTATE RELEASE AFTER BYPASS SURGERY 1927 than 0.4 mmol/l of lactate after 5 minutes of reperfusion (n 540). Patients who displayed persistent lactate release during reperfusion were older and suffered from preoperative left ventricular dysfunction. Left main coronary artery disease was uncommon in both groups (14% versus 9%, p 0.2), a significant minority of patients were in New York Heart Association class IV (37% versus 29%, p 0.3) and an overwhelming majority of patients had triple vessel coronary artery disease (72% versus 77%, p 0.6). Table 2 compares the operative and postoperative data in these two groups. The release of creatine kinase and the MB isozyme was similar in both groups. The prevalence of postoperative LOS and operative mortality was significantly higher (p and p 0.02, respectively, by Fisher s exact test) in patients who released more than 0.4 mmol/l of lactate after 5 minutes of reperfusion. Fig 1. The relationship between myocardial lactate release, oxygen extraction, and acid release during reperfusion and the development of postoperative low output syndrome (LOS). Patients in whom LOS developed had significantly higher lactate release immediately after aortic cross-clamp removal (XCL OFF) and at 5 minutes of reperfusion (5 ). However, net myocardial lactate release at 10 minutes of reperfusion (10 ) was not different between groups. There were no differences in myocardial oxygen extraction or acid release at any time point. tients who developed LOS ( mmol/l versus , p 0.01). Similarly, myocardial lactate release after 5 minutes of reperfusion was higher in patients who developed LOS ( mmol/l versus mmol/l, p 0.05). Myocardial oxygen extraction or acid release were not different in patients who developed postoperative LOS. There were no differences between groups in any metabolic measurements during the crossclamp period or beyond 10 minutes of reperfusion. Stepwise logistic regression identified persistent lactate release after 5 minutes of reperfusion to be the only predictor of postoperative LOS (odds ratio 5.85, 95% confidence interval 2.1 to 16.3). Figure 2 displays the proportion of patients who developed LOS at arbitrary cut-points of myocardial lactate release at 5 minutes of reperfusion. The ROC curve constructed in panel B of Figure 2 demonstrates that the optimal cut-point is a myocardial lactate release of 0.4 mmol/l. Table 1 compares the preoperative demographics of these patients (n 83) to those patients who released less Fig 2. (A) The relationship between myocardial lactate release at 5 minutes of reperfusion and the development of postoperative low output syndrome (LOS). Arbitrary cut-points (POS positive; NEG negative) reveal an increased prevalence of LOS in those patients who released more than 0.4, 0.6 or 0.8 mmol/l of lactate. (B) A receiver operating characteristics (ROC) curve was constructed from the arbitrary cut-points used in panel A. The optimal balance between sensitivity and specificity for predicting postoperative LOS is determined at a cut-point of 0.4 mmol/l of lactate (the point closest to the upper left quadrant). The area under the ROC curve, which estimates the precision of the diagnostic test, is

4 1928 RAO ET AL Ann Thorac Surg LACTATE RELEASE AFTER BYPASS SURGERY 2001;71: Table 1. Preoperative Demographics Variable A multiple regression analysis for myocardial lactate release after 5 minutes of reperfusion demonstrated a significant effect of age (slope , p ) and preoperative left ventricular function (slope , p ). Comment Lactate Release 0.4 mmol/l n (%) Lactate Release 0.4 mmol/l n (%) p Value n 83 (13) 540 (87) Age (years) Male 75 (90) 510 (94) Female 8 (10) 30 (6) Timing of surgery Elective 61 (74) 430 (80) Nonelective 22 (26) 110 (20) LVEF % 20 (24) 243 (45) 41 60% 38 (46) 222 (42) 20 40% 23 (28) 66 (12) 20% 2 (2) 4 (1) LVEF left ventricular ejection fraction. Advances in myocardial protection have led to lower risk-adjusted rates of morbidity and mortality after coronary bypass surgery [12]. However, the prevalence of postoperative LOS remains relatively high at approximately 9% [11]. In this series of 623 predominantly low risk patients, the overall incidence of LOS was 5.4%, despite an operative mortality of only 2.1% and a perioperative infarction rate of only 2.9%. In our institution, LOS is only diagnosed after an active intervention on the part of the surgeon or attending intensivist. The development of postoperative LOS, in the absence of an intraoperative misadventure, represents a failure of myocardial protection. We have previously shown that LOS, which may be a transient event in some patients, is associated with an increased length of intensive care unit and hospital stay and a higher mortality [11]. There are several preoperative clinical predictors of LOS in a diversified surgical population. However, in the low-risk homogeneous populations often employed for clinical studies, these risk factors are no longer predictive. Furthermore, the low rates of perioperative myocardial infarction and operative mortality necessitate the recruitment of large numbers of patients in order to discriminate between treatment strategies. Alternative outcome measures are required to increase the feasibility of clinical trials in myocardial protection. In previous clinical trials at our institution [5 8], we have observed a delayed recovery of myocardial metabolism after coronary bypass surgery. In particular, we have reported that anaerobic metabolism is induced by aortic crossclamping and cardioplegic arrest. However, in some patients, anaerobic metabolism persists even after cross-clamp removal and reperfusion. We hypothesized that persistent anaerobic metabolism, which we detect as net myocardial lactate release, is a marker of inadequate intraoperative myocardial protection and may predict early postoperative left ventricular dysfunction. Predictors of LOS Persistent lactate release during reperfusion was found to be the only multivariable predictor of LOS (odds ratio 5.85, 95% confidence interval 2.1 to 16.3). In a diversified surgical population, we have previously found that poor preoperative left ventricular function, repeat operation, urgent surgery for unstable angina, female gender, diabetes mellitus, age more than 70 years, left main disease, preoperative myocardial infarction, and triple vessel coronary artery disease to be independent predictors of postoperative LOS. None of these clinical risk factors predicted the development of LOS in this homogeneous study population. Although myocardial lactate release at 5 minutes of reperfusion had a poor predictive capability (area under ROC ), it was the only significant risk factor for the development of postoperative LOS. The poor predictive capability may be due to the fact that in some patients with adequate myocardial protection, postoperative LOS may be a result of intraoperative technical problems. Lactate release during reperfusion may be highly predictive of LOS in patients who had an uneventful intraoperative course. Unfortunately, it is difficult to distinguish between the effects of inadequate myocardial protection and intraoperative misadventure Table 2. Perioperative Data Variable Lactate Release 0.4 mmol/l n (%) Lactate Release 0.4 mmol/l n (%) p Value Aortic cross-clamp 70 2 minutes 66 2 minutes time Cardiopulmonary minutes minutes bypass time Bypass grafts 3 grafts 5 (6) 57 (10) grafts 78 (94) 482 (90) Creatine kinase release Total (IU/L) CK-MB (IU/L) %MB fraction Lactate release (mmol/l) Cross-clamp removal Five minutes reperfusion Perioperative MI 4 (5) 12 (2) Low output 11 (13) 25 (5) syndrome Stroke 2 (2) 7 (1) Mortality 5 (6) 8 (1) MI myocardial infarction (new Q wave or left bundle branch block on postoperative electrocardiogram).

5 Ann Thorac Surg RAO ET AL 2001;71: LACTATE RELEASE AFTER BYPASS SURGERY 1929 as these two events may be highly correlated. For example, diffuse atherosclerotic coronary artery disease not easily amenable to bypass grafting would also preclude adequate distribution of cardioplegia. Nevertheless, the development of postoperative LOS in a patient who had an uneventful intraoperative course may be due to a delayed recovery of normal aerobic myocardial metabolism. Relationship Between Lactate Release and Postoperative LOS The relationship between contractile function and myocardial metabolism is not clear. Weiss and associates [10] in an isolated rabbit heart model demonstrated that aerobically produced ATP was preferentially used for contractile function while ATP produced by anaerobic glycolysis was preferentially used for sarcolemnal ion channel function. Our findings in this study support Weiss observation in that persistent anaerobic lactate release during reperfusion predicted early postoperative left ventricular dysfunction. There was no evidence of irreversible myocardial injury as assessed by CK-MB enzyme release, thus the metabolic abnormalities are likely transient and may be amenable to treatment. Ando and colleagues [14] found similar results in a smaller study involving 66 patients undergoing isolated CABG with or without left main coronary artery stenosis. In patients with left main disease, they observed net lactate release in the empty beating state prior to cardioplegic arrest. These patients displayed a depression in postoperative left ventricular stroke work index, suggesting that subtle changes in normal myocardial metabolism may affect postoperative functional recovery [15]. The effect of left ventricular mass on myocardial lactate release was not evaluated in this study. One would expect myocardial oxygen and lactate consumption to increase as mass increases. However, the conversion from anaerobic to aerobic metabolism (lactate consumption to production) is independent of left ventricular mass. The exact mechanism of the derangement in normal myocardial metabolism remains unknown. Kobayashi and Neely [9] demonstrated that mitochondrial pyruvate dehydrogenase (PDH) activity was inhibited during early reperfusion after ischemia. As PDH regulates the conversion from pyruvate to acetyl-coa, its inhibition would result in greater lactate production. Attempts to improve myocardial functional recovery after ischemia and reperfusion by stimulating mitochondrial PDH activity have been inconsistent [16 19]. Mazer and colleagues [16] employed dichloroacetate (DCA), a potent stimulator of PDH activity, in a porcine model and found improved oxidative metabolism but no effect on functional recovery. In contrast, Wahr and associates [17] demonstrated in an isolated rabbit heart model that DCA resulted in a significant improvement in left ventricular functional recovery after ischemia and reperfusion. We have previously evaluated the effect of insulin on mitochondrial pyruvate dehydrogenase activity after simulated ischemia and reperfusion [20]. Inhibition of these enzymes would also impair normal oxidative phosphorylation and may result in inadequate functional recovery after surgery. Our results suggest that restoring normal aerobic metabolism during early reperfusion is crucial to preventing left ventricular dysfunction after surgery. We found that the metabolic differences that occurred in patients in whom LOS developed were attenuated within 10 minutes of reperfusion. In Kobayashi s report [9], mitochondrial pyruvate dehydrogenase activity was inhibited within 2 minutes of reperfusion and remained depressed for 45 minutes. Preventing this early inhibition of mitochondrial PDH activity may prevent left ventricular dysfunction after cardioplegic arrest. Summary This study evaluated the relationship between myocardial metabolism and postoperative low cardiac output syndrome in patients undergoing coronary bypass surgery. We believe that this is the largest invasive study of intraoperative myocardial metabolism in patients undergoing coronary bypass surgery. The results of this study suggest that delayed recovery of normal aerobic myocardial metabolism can adversely affect the clinical results of surgery. We believe that myocardial lactate release during reperfusion is a marker of inadequate myocardial protection during cardioplegic arrest. Future clinical trials evaluating novel myoprotective strategies should aim to convert anaerobic lactate release during reperfusion to myocardial lactate extraction and oxidation. Our results suggest that improving the transition from anaerobic to aerobic metabolism after cardioplegic arrest may have a significant impact on clinical outcomes after coronary bypass surgery. This study was supported by the Medical Research Council of Canada (Grant MT1989) and the Heart and Stroke Foundation of Canada (HSFC). Joan Ivanov is a research fellow of the HSFC; Richard D. Weisel is a career investigator of the HSFC. The authors wish to thank the nursing staff of the cardiovascular intensive care unit and operating room for their cooperation with this study. The authors are indebted to the clinical perfusionists of The Toronto Hospital for their valuable assistance with the intraoperative measurements of myocardial metabolism. Laura Tumiati and Molly Mohabeer are acknowledged for their assistance in analyzing perioperative blood samples. References 1. Weisel RD, Mickle DAG, Finkle CD, Tumiati LC, Madonik MM, Ivanov J. Delayed myocardial metabolic recovery after blood cardioplegia. Ann Thorac Surg 1989;48: Krukenkamp IB, Silverman NA, Sorlie D, Pridjian A, Feinberg H, Levitsky S. Characterization of post-ischemic myocardial oxygen utilization. Circulation 1986;74(Suppl III):III Christakis GT, Weisel RD, Mickle DAG, et al. Right ventricular function and metabolism. Circulation 1990;82(Suppl IV):IV Teoh KH, Mickle DAG, Weisel RD, et al. Improving myocardial metabolic and functional recovery after cardioplegic arrest. J Thorac Cardiovasc Surg 1998;95: Yau TM, Ikonomidis JS, Weisel RD, et al. Ventricular function after normothermic versus hypothermic cardioplegia. J Thorac Cardiovasc Surg 1993;105:833.

6 1930 RAO ET AL Ann Thorac Surg LACTATE RELEASE AFTER BYPASS SURGERY 2001;71: Hayashida N, Weisel RD, Shirai T, et al. Tepid antegrade and retrograde cardioplegia. Ann Thorac Surg 1995;59: Hayashida N, Ikonomidis JS, Weisel RD, et al. Adequate distribution of warm cardioplegic solution. J Thorac Cardiovasc Surg 1995;110: Shirai T, Rao V, Weisel RD, et al. Antegrade and retrograde cardioplegia: Simultaneous or alternate? J Thorac Cardiovasc Surg 1996;112: Kobayashi K, Neely JR. Effects of ischemia and reperfusion on pyruvate dehydrogenase activity in isolated rat hearts. J Mol Cell Cardiol 1983;15: Weiss J, Hiltbrand B. Functional compartmentation of glycolytic versus oxidative metabolism in isolated rabbit heart. J Clin Invest 1985;75: Rao V, Ivanov J, Weisel RD, Ikonomidis JS, Christakis GT, David TE. Predictors of low output syndrome following coronary bypass surgery. J Thorac Cardiovasc Surg 1996;112: Christakis GT, Birnbaum PL, Weisel RD, Ivanov J, David TE, Salerno TA. The changing pattern of coronary bypass surgery. Circulation 1989;80(Suppl I):I Weisel RD, Burns RJ, Baird RJ, et al. Optimal postoperative volume loading. J Thorac Cardiovasc Surg 1983;85: Ando H, Tanaka J, Hisahara M, Nagano I, Shimizu I. Implication of myocardial lactate metabolism during coronary artery bypass grafting. Cardiovasc Surg 1997;5: Wechsler AS. Implication of myocardial lactate metabolism during CABG. Cardiovasc Surg 1997;5: Mazer CD, Cason BA, Stanley WC, Schnier CB, Wisneski JA, Hickey RF. Dichloroacetate stimulates carbohydrate metabolism but does not improve systolic function in the pig heart. Am J Physiol 1995;268:H Wahr JA, Olszanski D, Childs KF, Bolling SF. Dichloroacetate enhanced myocardial functional recovery postischemia: ATP and NADH recovery. J Surg Res 1996;63: Vanoverschelde JL, Janier MF, Bakke JE, Marshall DR, Bergmann SR. Rate of glycolysis during ischemia determines extent of ischemic injury and functional recovery after reperfusion. Am J Physiol 1994;267:H Delehanty JM, Imai N, Liang CS. Effects of dichloroacetate on hemodynamic response to dynamic exercise in dogs. J Appl Physiol 1992;72: Rao V, Merante F, Weisel RD, et al. Insulin stimulates pyruvate dehydrogenase and protects human ventricular cardiomyocytes from simulated ischemia. J Thorac Cardiovasc Surg 1998;116: