Comparison of the Protective Properties of Four

Comparison of the Protective Properties of Four Clinical Crystalloid Cardioplegic Solutions in the Rat Heart Lary A. Robinson, M.D., Mark V. Braimbridge, F.R.C.S., and David J. Hearse, D.Sc. ABSTRACT Although few surgeons dispute the benefits of high-potassium crystalloid cardioplegia, objective comparison of the efficacy of various formulations is difficult in clinical practice. We compared four commonly used cardioplegic solutions in the isolated rat heart (N = 6 for each solution) subjected to 180 minutes of hypothermic (20 C) ischemic arrest with multidose cardioplegia (3 minutes every half-hour). The clinical solutions studied were St. Thomas Hospital solution, Tyers solution, lactated Ringer s solution with added potassium, and a balanced saline solution with glucose and potassium. Postischemic recovery of function was expressed as a percentage of preischemic control values. Release of creatine kinase during reperfusion was measured as an additional index of protection. St. Thomas Hospital solution provided almost complete recovery of all indexes of cardiac function following ischemia including 88.1 +- 1.6% recovery of aortic flow,. compared with poor recovery for the Tyers, lactated Ringer s, and balanced saline solutions (20.6 f 6.5%, 12.5 f 6.4%, and 9.6 f 4.2%, respectively) ( p < 0.001). Spontaneous defibrillation was rapid (less than 1 minute) and complete (100%) in all hearts in the St. Thomas Hospital solution group, but much less satisfactory with the other formulations. Finally, St. Thomas Hospital solution had a low postischemic level of creatine kinase leakage, contrasting with significantly higher enzyme release in the other solutions tested ( p < 0.001). Although differences in composition are subtle, all potassium crystalloid cardioplegic solutions are not alike in the myocardial protection they provide. Comparative studies under controlled conditions are important to define which formulation is superior for clinical application. Since the description in 1955 by Melrose and colleagues [l] of the concept of elective cardiac arrest, the use of cold chemical cardioplegic solutions has evolved and received widespread acceptance: 99% of surgical units in From the Heart Research Unit, The Rayne Institute, St. Thomas Hospital, London, England, and the Department of Surgery, Duke University Medical Center, Durham, NC. Presented at the Thirtieth Annual Meeting of the Southern Thoracic Surgical Association, Marco Island, FL, Nov 3-5, 1983. Address reprint requests to Dr. Robinson, Section of Thoracic and Cardiovascular Surgery, University of Nebraska Medical Center, 42nd St and Dewey Ave, Omaha, NE 68105. the United States are now using this technique clinically [Z]. However, a bewildering variety of cardioplegic solutions and techniques are utilized in clinical practice in the United States and Europe, and objective comparison of these formulations has proven difficult. Numerous studies, summarized by Hearse and associates [3], have compared various solutions with other methods of tissue protection in both experimental and clinical protocols, but aside from one study comparing the German intracellular cardioplegic solutions with the extracellular St. Thomas Hospital formulation [4], there is a dearth of objective data evaluating the crystalloid solutions commonly used in clinical cardiac surgery in the United States. To avoid the limitations and inherent variability of clinical studies, we chose the isolated working rat heart model [3, 5, 61 to compare the efficacy of myocardial protection provided by four commonly used clinical cardioplegic solutions during prolonged (three hours) hypothermic ischemic arrest. We sought to test the validity of the widely held belief that all cold highpotassium crystalloid cardioplegic solutions are similar in effectiveness. Materials and Methods Hearts were obtained from male Sprague-Dawley rats ranging in weight from 200 to 250 gm. In all studies, the National Academy of Sciences Guide for the Care and Use of Laboratory Animals was followed (National Institutes of Health Publication 80-23, revised 1978). Experimental Model The isolated, perfused, working rat heart model employed in this study has been described in detail elsewhere [3, 5, 61. In this left heart preparation, an ultrafiltered (pore size, 5 pm) perfusion medium is utilized (Krebs-Henseleit bicarbonate buffer [7, 81); the buffer has a ph of 7.4, contains 11.1 mmol/l of glucose, and is oxygenated with 95% oxygen and 5% carbon dioxide at 37 C. The perfusate enters the cannulated left atrium at a pressure equivalent to 17 cm H20 and passes into the left ventricle, from which it is spontaneously ejected through an aortic cannula against a hydrostatic pressure equivalent to 100 cm H20 (electrical pacing was not used in this study). Coronary effluent exiting from the right atrium and ventricle can be sampled for biochemical analysis or recirculated with the aortic outflow. Total cardiopulmonary bypass with coronary perfusion maintained may be simulated by clamping the left atrial cannula and introducing perfusion fluid at 37 C into the aorta from a reservoir located 100 cm above the 268

~ ~~~ ~ ~~~ 269 Robinson, Braimbridge, and Hearse: Crystalloid Cardioplegic Solutions in the Rat Heart Table 1. Composition of the Four Crystalloid Cardioplegic Solutions Testeda Heparin ph Solution Na++ C1++ K+ Ca++ Mg++ HC03- Acetate Gluconate Lactate Procaine Dextrose THAM (units) at 20Tb St. Thomas' 120 160 16 1.2 16 10..................... 8.10 ~f: 0.06 Hospital Tyers' 160 114 20 0.9 1.5 20 27 23............ 1,000 8.71 2 0.04 Lactated Ringer's 130 129 24 1.5......... 28............ 7.14 t 0.04 D2.5%/0.45%NaCl 114 170 17 2.4 17.,.......... 1.1 139 4.38... 7.80 3-0.01 "Where appropriate, values are expressed in millimoles per liter. bvalues shown are mean t standard error of the mean. D2.5%/0.45%NaCl = 2.5% dextrose in 0.45% normal saline; THAM = tris(hydroxymethy1)methylamine. heart. With this preparation, which is essentially that described by Langendorff [9], the heart continues to beat but does not perform any external work. Ischemic cardiac arrest may be induced by clamping the aortic cannula. During ischemia, the heart may be maintained at hypothermia (20 C) by means of cooling circuits supplying the water-jacketed heart chamber. Short periods of preischemic coronary infusion of a cardioplegic solution at 20 C may be achieved by use of a reservoir located 60 cm above the heart and attached to a sidearm of the aortic cannula. Card ioplegic Solutions The test solutions used in these studies were St. Thomas' Hospital solution [lo], Tyers' solution [ll], a modified lactated Ringer's solution with potassium as described by various authors [12-141, and a balanced saline solution (2.5% dextrose in 0.45% normal saline) containing glucose as suggested by Roe and colleagues [15] and Lolley and co-workers [16]. The comparative composition and measured ph (at 20 C) of the test solutions are outlined in Table 1. All tests were prepared from commercially available, clinical intravenous solutions in one-liter plastic or glass containers, and all additives were of pharmaceutical quality. All cardioplegic solutions were filtered through a cellulose acetate membrane (pore size, 0.8 Fm) just before use. None of the test solutions were oxygenated. Experimental Time Course Immediately after excision of the heart, the aorta was cannulated and Langendorff perfusion initiated for a 5- minute washout and equilibration period (Fig 1). Left atrial cannulation was accomplished during this time. The perfusion fluid during the 5-minute washout period and subsequent perfusion periods was Krebs-Henseleit bicarbonate buffer (37"C), as described earlier. The heart was converted to a working preparation by terminating retrograde aortic perfusion and initiating left atrial perfusion. During a 20-minute period, control values for aortic and coronary flow rates, peak aortic pressure, heart rate, and electronically differentiated first derivative of aortic pressure (dp/dt) were recorded. At the end of the control period, the atrial and aortic cannulas were clamped and the heart was subjected to a 3-minute pe- riod of hypothermic (20 C) coronary infusion with one of the four test cardioplegic solutions. Infusion was terminated, and the entire heart was maintained in a state of hypothermic 20 C ischemic arrest for 180 minutes. Cardioplegic solution at 20 C was reinfused for 3 minutes every half-hour during this three-hour ischemic period, after which the heart was reperfused initially in the Langendorff mode for 15 minutes with collection of coronary effluent for creatine kinase (CK) determination [17]. During this period, the time from initial postischemic reperfusion until the resumption of regular heart rhythm was noted. If regular rhythm did not resume by the end of the initial 15- minute reperfusion period, the heart was electrically defibrillated. Following the 15-minute reperfusion period, each heart was converted to the working mode for a further 20 minutes and recovery of cardiac function was recorded. At the end of each experiment, all hearts were heated to 110 C for 24 hours for the determination of dry weight. Analysis of Data During the preischemic working control period, the following variables were recorded: heart rate, coronary flow, aortic flow, aortic pressure, and instantaneous differentiation of aortic dp/dt. Cardiac output was derived from the sum of aortic and coronary flows, and stroke volume was obtained by dividing cardiac output by heart rate. Minute work was derived by multiplying cardiac output by systolic pressure. During the working recovery period, the absolute values for the various measured indexes of cardiac function in individual hearts were compared and expressed as a percentage of those values obtained during the preischemic control period. In addition to eliminating any inherent variability among individual hearts, this allowed the recovery of each index of function to be related to the nature and composition of the test cardioplegic solution. Creatine kinase leakage was expressed as international units per 15 minutes per gram of dry weight. At least 6 hearts were used for each condition studied, and all data were expressed as the mean f standard error of the mean. Statistical analysis of the results was made by unpaired Student t test, and statistical significance was assumed when p values were 0.05 or less.

270 The Annals of Thoracic Surgery Vol 38 No 3 September 1984-0.? c 0 z Pre-Ischemic Working Control 3 7 z.rd ~ c o -00 0- < a 55 $ : Hypothermic Non -Working Post- I sc hemi c Ischemia 20 C Reperfuston Working Table 2. Effect of Composition of Test Cardioplegic Solutions on Postischemic Recovery of Indexes of Cardiac Function and on Enzyme Leakage after 180 Minutes of Hypotherrnic Ischemia St. Thomas Hospital Tyers Lactated Ringer s D2.5%/0.45%NaCl Percent Percent Percent Percent Variable Control Recovewb Control Recovery Control Recover/ Control Recovery Aortic flow (mumin) 69.6 f 1.8 88.1 f 1.6 70.8 f 1.9 20.6 rt 6.5d 66.8 t 1.3 12.5 f 6.4d 69.3 rt 2.5 9.6 f 4.2d Cardiac output 94.8 f 2.4 88.1 f 1.7 94.0 f 2.0 29.6 f 6.2d 89.7 2 2.4 23.9 f 5.2d 90.8 f 3.2 18.5 f 3.p (mumin) Stroke volume 0.31 f 0.01 90.6 _r 3.7 0.32 f 0.01 32.5 _t 5.7d 0.30 2 0.01 25.8 rt 5Bd 0.30 f 0.01 20.4 f 4.4d (mumin) dp/dt (cmhzo/sec) 4,446.0 f 311.7 67.1 f 2.3 5,408.0 f 268.5 33.5 rt 2Ad 4,957.3 2 224.1 21.4 f 6.1d 4,264.0 f 382.5 29.8 c 6Ad Minute work 184.1 f 9.2 77.4 * 1.8 199.6 f 7.8 20.0 f 1.8d 188.4 2 10.3 15.3 f 1.6d 176.2 _t 10.6 12.0 * 1.6d (105dyne-cm/min) Coronary flow 23.9 f 1.3 88.1 f 3.3 23.2 -c 0.5 56.4 f 8.0 22.8 -r 1.4 58.8 -c 7.F 21.2 f 1.8 46.0 f 5.7d (mumin) Aortic pressure 200.6 c 7.8 87.8 f 0.7 216.7 f 7.4 71.4 f 3.4d 215.0 -?r 8.5 58.9 f 3.6d 197.5 f 8.2 59.7 f 6.9e (cm HzO) Creatine kinase 21.1 f 2.5 84.0 f 8.8d 70.4 rt 3.2d 123.5 f 22.0d leakage (IU/15 min/gm dry weight) Values shown are mean k standard error of the mean (of at least 6 hearts). bafter 35 minutes of reperfusion for all solutions. Only 5 of the 6 hearts in which this solution was used returned to regular rhythm after ischemia; therefore, the values indicated in this column are for 5 hearts. dsignjficance: p < 0.001 compared with St. Thomas Hospital solution values (Student t test). Sigmficance: p < 0.01 compared with St. Thomas Hospital solution values (Student t test). Results Functional Assessment Table 2 and Figure 2 show the postischemic recovery of various indexes of myocardial function after three hours of hypothermic ischemia for the four test cardioplegic solutions. As is apparent from these data, St. Thomas Hospital solution afforded good cardiac protection with almost complete return of function; aortic flow and cardiac output recovered to 88.1? 1.6% and 88.1 * 1.7% of preischemic levels, respectively. In marked contrast, the other three solutions gave poor protection during this interval, with the best recovery of aortic flow reaching only 20.6 * 6.5% (Tyers solution) and the worst, 9.6 2 4.2% (D2.5%/0.45%NaCl). Although there were no significant differences in any of the functional indexes for these three solutions, there were highly significant differences in recovery when they were compared with St. Thomas Hospital solution. On average, St. Thomas Hospital solution gave three to five times greater recovery of hemodynamic indexes than any of the other solutions. The only difference between the four groups that was not statistically significant was the recovery of heart rate. Enzymatic Assessment The marked differences in functional recovery among the various solutions are paralleled by the results for postischemic CK leakage; Figure 3 displays these results. Again, St. Thomas Hospital solution demonstrated better myocardial preservation, allowing the release of significantly less CK (21.1 * 2.5 IU per 15 minutes per gram of dry weight) than the Tyers, lactated Ringer s,

271 Robinson, Braimbridge, and Hearse: Crystalloid Cardioplegic Solutions in the Rat Heart 1-2 90- c c u 80-.- u E 22 70- V v) - & 60- a 0 ST Tyer s LR D2 51 45NS.u 30 * I c g 10 a Aortic Stroke Minute dpldt Flow Volume Work CardiG output Coronary Flow Fig 2. Postischemic recovery of various functional indexes (expressed as a percentage of their preischemic baseline control value) in groups of hearts perfused with each of the following cardioplegic solutions: St. Thomas (ST), Tyers, lactated Ringer s (LR), and 2.5% dextrose in 0.45% normal saline (D2.51.45NS). Recovey was measured at the end of a 35-minute repelfusion period following 180 minutes of hypothermic (20 C) ischemia. Each column represents the mean of 6 hearts, and bars represent the standard error of the mean. (dp1dt = first derivative of aortic pressure; * = p < 0.001 compared with St. Thomas solution; ** = p < 0.01 compared with St. Thomas solution.) and 2.5% dextrose in saline soilutions (84.0 2 8.8 IU, 70.4? 3.2 IU, and 123.5? 22.0 IU per 15 minutes per gram of dry weight, respectively) (see Table 2). Arrhythmia Assessment Immediately following the three-hour ischemic period, measurements were made of the number of hearts that spontaneously defibrillated (number spontaneously converting to regular rhythm) during reperfusion and the duration of reperfusion before returning to normal rhythm. After 15 minutes of reperfusion, hearts not spontaneously converting to regular rhythm were electrically defibrillated. The results (Fig 4) revealed that perfusion with St. Thomas Hospital solution or Tyers solution led to spontaneous defibrillation in all hearts studied, whereas with the other solutions, the number of hearts that required electrical conversion ranged from 16.7% (D2.5%/0.45%NaCl) to 50% (lactated Ringer s solution). A similar trend was observed for the time to return to regular rhythm, with the shortest time for hearts perfused with St. Thomas Hospital solution (0.67 & 0.1 minutes). The other groups were significantly higher: 3.4 & 1.0 minutes (Tyers ), 10.0 & 1.5 minutes (lactated Ringer s), and 2.9 & 0.6 minutes (D2.5%/0.45% NaCl). If early spontaneous defibrillation is considered an indicator of good myocardial protection, St. Thomas Hospital solution again proved superior. Comment The description in 1955 by Melrose and colleagues [l] of a cardioplegic solution for elective arrest had a profound effect on the future conduct of cardiac surgery. However, the dose of potassium citrate chosen proved too large, resulting in myocardial necrosis; therefore, this technique was abandoned for a decade or more. In the early 1970s, the revival of cold chemical cardioplegia using more rational dosages of component ions resulted in the rapid adoption of this technique in clinical cardiac surgery throughout the world [3] so that by 1981, 99% of clinical surgical units in the United States were using it [2]. A large number of cardioplegic solutions had been developed and evaluated. Nearly all were based on potassium in the range of 15 to 35 mmoy liter [MI, and potassium frequently was the only cardioplegic agent. Other components varied widely in terms of the specific additive chosen and its dosage. In most instances, the inclusion of such additives and their dosages were based on theory, rather than definitive and controlled experimental studies. The recent literature abounds with studies in which various cardioplegic solutions have been compared with regard to their efficacy of protection; these have been summarized by Hearse and colleagues [3]. In general, however, comparisons have been with other methods of tissue protection. Jynge and associates [4] performed the only controlled experimental evaluation of cardio-

272 The Annals of Thoracic Surgery Vol 38 No 3 September 1984 T*,* I* 100 80 F Z 60 W 0 40 n 20 0 st. Tyers Lactated D2.5/ Thomorl Ringerr 0.45NaCI Tver s Lactated Ringers DZ 51 045NS Fig 3. Creatine kinase release following ischemia in the four test cardioplegic solution groups. Coronary effuent for the first 25 minutes of reperfusion was collected for measurement of enzyme release. Each column represents the mean of 6 hearts, and bars represent the standard error of the mean. (Abbreviations and symbols same as in Fig 2.) 12 L A plegic solutions in which they compared St. Thomas Hospital solution with the German intracellular solutions (Bretschneider and Kirsch). The present study examined four potassium-based clinical cardioplegic solutions commonly used in the United States. Three markers of myocardial protection (functional recovery, CK leakage, and spontaneous defibrillation) were used to assess the relative efficacy of these solutions. Despite the generally high level of protection afforded, not all solutions provided the same extent of protection, and the St. Thomas Hospital solution proved superior in all aspects to the other three. The reasons for the differences among the test solutions include variations in specific components and differences in dosage of those components (see Table 1). Because of complex interactions and dose-response relationships among several of the important basic cations (sodium, potassium, and calcium), variations in dosage of any of these ions in solutions may lead to deleterious effects on postischemic recovery. Tyers solution [ll] differs in several major respects from St. Thomas Hospital solution. The low dose of magnesium used (one-tenth that of St. Thomas ) does not take full advantage of the increased protection afforded by higher concentrations (e.g., 16 mmol/l), as has been demonstrated by controlled dose-response studies [19]. The acetate (27 mmom) and gluconate (23 mmol/l) included in Tyers solution may not be inert anions substituting for chloride, but they may have potential negative effects. Christoffersen and Skibsted [20] found acetate and gluconate to have marked calciumbinding actions, and inclusion of these additives may Fig 4. (Top) Percentage of hearts spontaneously defibrillating to regular rhythm upon reperfusion after three hours of hypothemzic (20 C) ischemia in each cardioplegic solution test group. (Bottom) Time (mean 5 standard error of the mean) to resumption of regular rhythm following ischemia in hearts that spontaneously defibrillated. Each test group contained 6 hearts. (* = p < 0.02; ** = p < 0.02; *** = p < 0.001; all comparisons between St. Thomas solution and other test groups.) affect the level of free ionized calcium. A recent study [21] found the optimal dose of calcium in cardioplegic solution to be 1.2 mmol/l. The calcium level in Tyers solution of 0.9 mmol/l may have been effectively lowered by the addition of acetate and gluconate, and as a consequence may have an adverse effect on the efficacy of this solution as a cardioplegic agent. The solution based on lactated Ringer s prime [12-141 also differs markedly from the St. Thomas Hospital solution, which may account for its inferior cardioprotective effects. Again, the absence of magnesium may be an important deficiency [19]. Equally important is the potential hazard of lactate in infusates [22]. Although the optimal ph for cardioplegic solutions is controversial, many studies [3] suggest that with cooling of the heart, a mildly alkaline solution is advisable, and any buffer used should have a pk value closest to the desired ph to be effective. The pk (at 25 C) for lactate is 3.86; lactate therefore is an inappropriate buffer because its action is outside of the range normally experienced in cardiac tissue. The measured ph of the test lactated Ringer s solution is 7.14 2 0.04, and because lactate is the buffer, the

273 Robinson, Braimbridge, and Hearse: Crystalloid Cardioplegic Solutions in the Rat Heart ph will tend to stabilize or become lower after use. In a study of the efficacy of various buffers in cardioplegic solution, bicarbonate or phosphate afforded considerable myocardial protection, whereas lactate was detrimental [22]. Thus, a lactate buffer is also not recommended because it adds to intracellular lactate, which accumulates during ischemia and inhibits glycolysis and anaerobic adenosine triphosphate production. The final test solution using D2.5%/0.45%NaCl as its base [15, 161 also differs in composition from the St. Thomas Hospital solution. The level of calcium in this solution (2.4 mmol/l) is twice that found optimal in the only dose-response study in the literature [21], and the efficacy of this solution may suffer as a result. Tris(hydroxymethy1)methylamine buffer is used instead of bicarbonate as the buffering agent. Although it has a favorable pk of 8.08, Tris is less likely to be advantageous during tissue acidosis as, unlike the other buffers, its ability to act as a hydrogen acceptor declines at lower ph values and is likely to be relatively ineffective in severely ischemic tissue. Additionally, Tris penetrates the cell membrane, and has been shown to be toxic in a number of tissues and to exert a negative inotropic effect [23, 241. The other and perhaps most important difference between St. Thomas Hospital solution and D2.5%/0.45% NaCl is the presence of glucose (139 mmol/l) in the formulation. There is an extensive literature suggesting that glucose in combination with insulin and potassium is beneficial to the ischemic myocardium during acute myocardial infarction [25-271, and its use has been similarly advocated in cardioplegic solutions [15, 161 for elective surgical ischemia. However, studies of glucose-containing cardioplegic solutions compared results with former methods of myocardial preservation and not with similar cardioplegic solutions that did not contain glucose. This may explain why Hearse and co-workers [28] found glucose to be deleterious in the isolated rat heart model with a dose-dependent reduction in myocardial protection. The mechanism to explain these detrimental effects of glucose is unknown but may involve stimulation of glycolysis and accumulation of lactate and protons; or, exogenous glucose utilization may replace the energetically favorable glycogen utilization. The solutions used in this study are representative of the formulations in frequent clinical use, but they do not encompass the entire range of crystalloid cardioplegic solutions or additives. In this comparative controlled study, St. Thomas Hospital solution proved superior to the other test solutions for myocardial protection in the isolated rat heart. If the results for this model prove valid in clinical trials, as has been the case in the past [3], our findings have important implications for the practicing surgeon. This work was supported in part by grants from the British Heart Foundation and the St. Thomas Hospital Research Endowments Fund. References 1. Melrose DG, Dreyer B, Bentall HH, Baker JBE: Elective cardiac arrest. Lancet 221, 1955 2. Miller DW, Binford JM, Hessel EA: Results of a survey of the professional activities of 811 cardiopulmonary perfusionists. J Thorac Cardiovasc Surg 83:385, 1982 3. Hearse DJ, Braimbridge MV, Jynge P: Protection of the Ischemic Myocardium: Cardioplegia. New York, Raven, 1981 4. Jynge P, Hearse DJ, de Leiris J, et al: Protection of the ischemic myocardium: ultrastructural, enzymatic, and functional assessment of the efficacy of various cardioplegic infusates. J Thorac Cardiovasc Surg 76:2, 1978 5. Neely JR, Liebermeister H, Battersby EJ, Morgan HE: Effect 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. of pressure development on oxygen consumption by isolated rat hearts. Am J Physiol 2i5804, 1967 Hearse DJ, Stewart DA, Braimbridge MV Hypothermic arrest and potassium arrest: metabolic and myocardial protection during elective cardiac arrest. Circ Res 36:481, 1975 Krebs HA, Henseleit K Untersuchungen uber die Harnstoffbildung im Tierkorper. Hoppe Seylers Z Physiol Chem 210:33, 1932 Umbreit WW, Burns RH, Stauffer JF: Preparation of Krebs- Ringer phosphate and bicarbonate solution. In Manometric Techniques. Minneapolis, Burgess, 1972, p 146 Langendorff 0: Untersuchungen am uberlebeden Saugethierherzen. Pfluegers Arch 61:291, 1895 Jynge P, Hearse JD, Feuvray D, et al: The St. Thomas Hospital cardioplegic solution: a characterization in two species. Scan J Thorac Cardiovasc Surg [Suppl] 30:5, 1981 Tyers GFO, Manley NJ, Williams EH, et al: Preliminary clinical experience with isotonic hypothermic potassiuminduced arrest. J Thorac Cardiovasc Surg 74:674, 1977 Tyers GFO, Hughes HC Jr, Todd GJ, et al: Protection from ischemic cardiac arrest by coronary perfusion with cold Ringer s lactate solution. J Thorac Cardiovasc Surg 67:411, 1974 Fisk RL, Gelfand ET, Callahan JC: Hypothermic coronary perfusion for intraoperative cardioplegia. Ann Thorac Surg 23:58, 1977 Adams PX, Cunningham JN Jr, Trehan NK, et al: Clinical experience using potassium-induced cardioplegia with hypothermia in aortic valve replacement. J Thorac Cardiovasc Surg 75:564, 1978 Roe BB, Hutchinson JC, Fishman NH, et al: Myocardial protection with cold, ischemic, potassium-induced cardioplegia. J Thorac Cardiovasc Surg 73:366, 1977 Lolley DM, Ray JF 111, Meyers WO, et al: Reduction of intraoperative myocardial infarction by means of exogenous anaerobic substrate enhancement: prospective randomized study. Ann Thorac Surg 26515, 1978 Szoaz GS, Gruber W, Bernt E: Creatine kinase in serum: I. Determination of optimal reaction conditions. Clin Chem 22:50, 1976 Mitchell BA, Litwak RS: Myocardial protection with cold ischemic potassium-induced cardioplegia: an overview. Proc/Am SOC Extracorp Tech 6:127, 1978 Hearse DJ, Stewart DA, Braimbridge MV Myocardial protection during ischemic cardiac arrest: the importance of magnesium in cardioplegic infusates. J Thorac Cardiovasc Surg 75:877, 1978 Christoffersen GRJ, Skibsted LH: Calcium ion activity in physiological salt solutions: influence of anions substituted for chloride. Comp Biochem Physiol52A:317, 1975

274 The Annals of Thoracic Surgery Vol 38 No 3 September 1984 21. Yamamoto F, Braimbridge MV, Hearse DJ: Myocardial protection during global ischemia: optimization of the calcium content of the St. Thomas cardioplegic solution (abstract). J Mol Cell Cardiol 15:Suppl 1:272, 1983 22. Hearse DJ, Stewart DA, Braimbridge MV: Myocardial protection during bypass and arrest: a possible hazard with lactate-containing infusates. J Thorac Cardiovasc Surg 72: 880, 1976 23. GiUespie JS, McKnight AT Adverse effects of Tris hydrochloride, a commonly used buffer in physiologic media. J Physiol259561, 1976 24. Turlapaty PDMV, Altura BT, Altura BM: Influence of Tris on contractile responses of isolated rat aorta and portal vein. Am J Physiol235:H208, 1978 25. de Leiris J, Opie LH: Effects of substrates and coronary artery ligation on mechanical performance and on release of lactate dehydrogenase and creatine phosphate in isolated working rat hearts. Cardiovasc Res 12:585, 1978 26. Majid PA, Sharma 8, Meeran MKM, Taylor SH Insulin and glucose in the treatment of heart failure. Lancet 2:937, 1972 27. Opie LH, Owen P: The effect of glucose-insulin-potassium infusions on arteriovenous differences of glucose and of free fatty acids and on tissue metabolic changes in dogs with developing myocardial infarction. Am Heart J 38:310, 1976 28. Hearse DJ, Stewart DA, Braimbridge MV: Myocardial protection during ischemic cardiac arrest possible deleterious effects of glucose and mannitol in coronary infusates. J Thorac Cardiovasc Surg 7616, 1978