of Calcium-Free and Calcium-Containing Solutions

Transcription

1 Cold Ischemic Arrest: Comparison of Calcium-Free and Calcium-Containing Solutions F. F. A. Hendriks, Ph.D., J. Jonas, M.D., A. van der Laarse, Ph.D., H. A. Huysmans, M.D., G. L. van Rijk-Zwikker, M.D., and J. J. Schipperheyn, M.D., Ph.D. ABSTRACT Isolated pumping rat hearts, perfused with reconstituted blood, were studied to compare the effects of 30 minutes of ischemic arrest following calcium-free or normal, calcium-containing cold cardioplegia on recovery of mechanical function, lactate production, myocardial adenosine triphosphate concentration, and release of creatine kinase (CK). As in clinical situations, the volume of the infusate was only three to four times the intracavitary blood volume. Hearts arrested with calcium-free solution showed incomplete recovery of mechanical function, whereas hearts arrested with calcium-containing solution recovered completely. After calcium-free arrest, stroke volume recovered to 76 t 29% (standard deviation [SDI) of its prearrest value. Enzyme release (CK) was significantly higher after calcium-free cardioplegia (7.7 t 4.6 units [SD]) than after cardioplegia with normal calcium (2.1? 1.6 units [SD]). Since the addition of only mmol calcium ions to a liter of calcium-free solution completely prevented its negative effect, it was concluded that calcium-free cardioplegia may cause limited but pronounced damage to myocardial cells, presumably because it removes calcium from the cellular membranes-the so-called calcium paradox. Probably due to residual calcium in blood and extracellular fluid, the damage is not so extensive after calcium-free cardioplegia as to be noticeable in clinical surgical situations. Residual calcium in the heart does not exclude the possibility, however, that a calcium paradox occurs in small scattered areas of the heart. Cardiac arrest with myocardial hypothermia induced by infusion of an appropriate cold cardioplegic solution into the coronary system is commonly accepted as a safe method to protect the heart from ischemic damage during surgical interventions lasting up to three hours [l- 41. Cardioplegic solutions contain potassium ions in a relatively high concentration of 10 to 20 mmol/l and magnesium ions in a concentration of 2 to 16 mmol/l [l]. This combination of ionic concentrations blocks the conduction of impulses and inhibits the influx of calcium From the Departments of Thoracic Surgery and Cardiology, University of Leiden, Leiden, The Netherlands. Accepted for publication June 7, Address reprint requests to Dr. Schipperheyn, Department of Cardiology, Academic Hospital, Rijnsburgerweg 10, 2333 AA Leiden, The Netherlands. ions, thereby reducing myocardial oxygen consumption. Further reduction is achieved by rapid cooling to a temperature between 15" and 20 C. In an earlier stage of development of the technique of cardioplegia, it was considered necessary to improve on the membrane-protective effect of the solutions by adding substances such as procaine, lidocaine, or calcium channel blockers. Amino acids or glucose was added, and the solutions were buffered. Lowering the concentrations of sodium and calcium ions and increasing the concentration of magnesium ions appeared to improve the protective action. Some of these refinements are now deemed unnecessary or even harmful, but some are widely used, although their possible merit is difficult to demonstrate [l]. Controversy still exists regarding the use of the so-called intracellular solution, which is completely devoid of calcium ions. Buckberg [l], in a review of the literature, concluded that lowering the calcium concentration indeed has a favorable effect, but he also quoted Tyers [5], who warned of the possible adverse effects of the use of completely calcium-free solutions. At the time of Tyers's observation, it had become known that perfusion of the isolated rat heart with a solution containing less than mmol/l of calcium ions damages the outer membranes of myocardial cells. When the heart is perfused with a solution containing a normal concentration of calcium ions, the cells are completely destroyed [6-91. Jynge and colleagues [lo-121 tested different types of cardioplegic solutions in an isolated pumping rat heart and found better recovery of function if the solutions contained calcium ions. Fairly large volumes of perfusate were infused in their studies, however, and the hearts were not perfused with blood. In clinical studies, cardioplegia with calcium-free solutions gave satisfactory results [13], and in experiments in dogs, no damage resulting from calcium paradox (the removal of calcium from cellular membranes) could be demonstrated [14]. It has been suggested that low temperature and the use of low concentrations of sodium prevented the calcium paradox in these studies [15]. Although low temperature and sodium-poor solutions indeed reduce and delay the calcium paradox [16], the residual calcium in the blood and in the extracellular space and calcium ions from noncoronary blood supply may be most important in preventing it. In clinical surgery, the volume of the cardioplegic solution is at most three to four times the volume of blood and extracellular fluid in the heart. Therefore, one cannot expect the calcium concentration to drop below the mmol/l required for the calcium paradox. 312

2 313 Hendriks et al: Cold Ischemic Arrest with Calcium and Noncalcium Solutions This study has been designed to compare, in an isolated blood-perfused heart, the effect on the recovery of myocardial function including lactate production, myocardial adenosine triphosphate (ATP) concentration, and release of creatine kinase (CK), of a cardioplegic solution without calcium ions and that of a solution containing a normal concentration of calcium ions. The volume of the infusate was small compared to the size of the heart, to mimic the clinical situation in that respect. In isolated preparations it is possible to detect minor cellular damage, which cannot be detected in whole animal studies or in clinical situations. Material and Methods Thirty-five male Wistar rats were used for the experiments. They ranged in weight from 232 to 320 gm. The animals were anesthetized with diethyl ether. The thorax was opened, 5 mg of heparin was injected into the caval vein, and the heart was taken out and immersed immediately in a cold saline solution at 4 C. The aorta was tied to a cannula, and perfusion of the coronary system was started within 1 minute after excision with a perfusion pressure of 60 mm Hg. The perfusion technique was adopted from Duvelleroy and associates [17] and has been modified and described in detail previously [18]. The hearts were perfused with reconstituted blood. To prepare this, freshly obtained bovine blood was treated with citric acid-citrate buffer to prevent coagulation. Erythrocytes were separated from plasma and leukocytes in a Haemonetics Cell-Saver and were washed repeatedly with a buffered Krebs-Henseleit solution to which 0.2% bovine serum albumin (fraction V) was added. The fluid was equilibrated to a 95% oxygen and 5% carbon dioxide gas mixture at 37 C. The ph was between 7.35 and 7.45, and the oxygen saturation was close to 100%. Hematocrit values ranged from 0.25 to After a few minutes of coronary perfusion, the left atrium was cannulated and connected to a reservoir. The fluid level in the reservoir was set at 10 cm above the level of the mitral valve. By opening the connection to the reservoir, the heart was filled and made to pump against a hemodynamic load. Pressure in the aortic cannula was set at 60 mm Hg by means of a Starling resistor. Blood from the coronary system, pumped out by the right ventricle, was collected in a beaker. Pressure in the left atrium and in the aorta was measured with two Statham pressure transducers connected to a Philips electrical manometer. Flows through the aorta and into the left atrium were measured with two flow probes connected to a Skalar electromagnetic flowmeter. Coronary flow was measured by subtracting aortic flow from flow into the left atrium. Heart rate was calculated continuously from the pressure signal with a Thompson heart rate meter. Samples of the arterial and the coronary venous blood were taken at regular intervals to determine ph, oxygen tension, and oxygen content. Coronary flow measurements were calibrated by measuring the volume of coronary venous blood collected Table 1. Corvipositioii of the Calcium-Free Bretschneider Solution and the Calcium-Coiztaitiing St. Thomas Hospital Solution Used in the Experirnents St. Thomas Bretschneider Hospital Variable Solution Solution Component (mmolil) Sodium chloride Sodium bicarbonate Potassium chloride Calcium chloride Magnesium chloride Magnesium sulfate Potassium hypophosphate Mannitol L-Histidine Histidine hydrochloride Tryptophan Equilibration 100% Or 95% 02,5% coz ph at 20 C Osmolarity 335 mosm/l 295 mosm/l over 2 minutes. From coronary flow and the arteriovenous difference of oxygen content, the oxygen consumption was calculated. After initial equilibration for 15 minutes, the cannulas to the left atrium and aorta were clamped and 3 ml of a cold (4 C) cardioplegic solution was gently infused into the coronary arteries through a side port just above the aortic valve. The infusion rate was set at 6 ml/min. The compositions of the calcium-free (Bretschneider) and the calcium-containing (St. Thomas Hospital) solutions are given in Table 1. Special care was taken to avoid contamination of the calcium-free solution with calcium ions. The solutions were prepared with Milli-Q water (Millipore), which contains ppb of calcium. To demonstrate that possible differences in the results obtained with the two types of solutions are caused by the difference in calcium concentration only and not by differences in the concentration of residual components, a low concentration of calcium ions (0.025 mmol/l) was added to the calcium-free Bretschneider solution in an additional series of experiments. The hearts were kept arrested for 30 minutes while submerged in saline solution at 20 C. Before and after cardioplegia, the temperature of the perfusate and of the humid air around the heart was maintained at 37 C. After 30 minutes, the hearts were reperfused, initially at a perfusion pressure of 40 mm Hg, which was gradually increased to 60 mm Hg in 10 minutes. During these 10 minutes, the cannula to the left atrium was kept closed. Then the cannula was opened, and the hearts were allowed to pump. After an additional 20 minutes, the experiment was terminated. The hearts were quickly frozen in liquid nitrogen and freeze-dried for 48 hours.

3 314 The Annals of Thoracic Surgery Vol 39 No 4 April 1985 I -' -160 mm Ha. 10 sec c c Fig 1. Chart recording of (A) left atr<aal pressure (PI,),(B) aortic pressure (Pa<,), and (C) aortic flow (QaJ measured in an isolated rat heart during 30 minutes' ischemic arrest at 20 C after cardioplegia with St. Thomas' Hospital solution. Pieces of dry myocardium were trimmed off, weighed, and homogenized in 10% trichloroacetic acid. After neutralization of the supernatant with potassium hydroxide (5 M), the ATP concentration was determined with a bioluminescent technique (Lumac). The coronary effluent was collected during the 30-minute reperfusion period after ischemic arrest, and the total activity of released CK was determined. The CK activity was measured spectrophotometrically at 25 C using a commercially available kit (Merck). Additional samples were taken to determine lactate concentrations (Automatic Laboratory Analyzer). Results Thirty-five rat hearts were used for the study. Mean heart weight was 1.05? 0.09 gm (standard deviation [SD]). Thirteen hearts were arrested with calciumcontaining St. Thomas' Hospital solution, 14 with the calcium-free Bretschneider solution, and 8 with the modified Bretschneider solution containing some calcium. Mean cardiac output in the prearrest period was 45.7? 6.6 mlimin, and mean heart rate was 245? 46 bpm). Figure 1 is a chart recording of the left atrial and aortic pressure and aortic flow before, during, and after cardioplegia with St. Thomas' Hospital solution. Figure 2 shows the same signals obtained before, during, and after cardioplegia with calcium-free Bretschneider solution. After calcium-free cardioplegia, recovery is slow and incomplete, stroke volume is reduced, and initially heart rate is low. Left atrial pressure is high, although the fluid level in the reservoir was set at 10 cm above the mitral valve. The presence of high V waves suggests mitral insufficiency, possibly caused by papillary muscle dysfunction. Table 2 summarizes the values of prearrest and postarrest stroke volume, oxygen consumption, and lactate production. After cardioplegia with the calcium-free solution, stroke volume was 76% of its prearrest value ( p < 0.01), oxygen consumption was 91% (not significant), and lactate concentration in the effluent was 0.38 mmol, which was significantly higher than the mean value after cardioplegia with St. Thomas' Hospital solution. After cardioplegia with St. Thomas' Hospital solution, stroke volume, oxygen consumption, and lactate production were not significantly different from prearrest values. Adding only mmol/l of calcium ions improved the results obtained with Bretschneider solution: Stroke volume, oxygen consumption, and lactate production showed complete recovery after the ischemic arrest. Measurements of enzyme (CK) release and myocardial ATP content after ischemic arrest point in the same direction (Table 3). After calcium-free cardioplegia, the total activity of CK released from the preparation (7.7? 4.6 units) is more than three times higher than that after normal calcium cardioplegia (2.1 & 1.6 units). The presence of only a small quantity of calcium ions reduces the CK release to values found after cardioplegia with St. Thomas' Hospital solution. The ATP concentrations measured 30 minutes after reperfusion were not significantly different for either type of solution. Comment The results clearly show that if the infusate is completely calcium-free, functional recovery from ischemic arrest is delayed and occasionally incomplete. Release of CK is higher after calcium-free cardioplegia, indicating that significantly more cells are damaged by that kind of solution than by calcium-containing solutions. The ATP concentration remains normal after calcium-free cardioplegia, which is consistent with damage caused by calcium paradox. The variation of the CK release values

4 315 Hendriks et al: Cold Ischemic Arrest with Calcium and Noncalcium Solutions B. i tzo mllrnin Fig 2. Chart recording of (A) left atrial pressure (PI.), (B) aortic pressure (Pa<,), and (C) aorticfim (QaJ of an isolated rat heart during ischemic arrest at 20 C after cardioplegia with calcium-free Bretschneider solution. after calcium-free cardioplegia is great, indicating that the effect of the low calcium concentration is somewhat erratic. Some hearts recover completely, whereas others incur considerable damage. The extent of the damage can be estimated from the total activity values of CK released after reperfusion. The hearts of the laboratory rats used in this study contained between 750 and 760 units of releasable CK per gram of wet weight. If completely destroyed, the hearts in this series of experiments would have yielded 790 units of CK. The release of 2.1 units after cardioplegia with St. Thomas' Hospital solution, therefore, corresponds to cellular damage of 0.4% of total cardiac mass. The highest CK value found after calcium-free cardioplegia was 16.8 units, corresponding to more than 2% of mean cardiac mass. It is known that residual components of calcium-free solutions may influence the tendency of such solutions to induce a calcium paradox. Low temperature, a low concentration of sodium, and a moderately high magnesium concentration are known to delay the development of the calcium paradox and to limit the extent of cellular damage. In this study, they did not completely prevent the damage, however. In our view, there is little doubt that myocardial cellular damage is caused by the low calcium concentration, since the addition of only Table 2. Stroke Volume, Oxygen Consumption, and Lactate Production before and after Cardioplegia and lschernic Arrest" St. Thomas' Hospital Bretschneider Solution Solution Bretschneider Solution with Calcium" Variable (N = 13) (N = 14) (N = 8) Stroke volume (ml) Before CP After CP Percent' i: ? i: f ? t 29.2* f ? f 9.9 Myocardial oxygen consumption (ml/min) Before CP After CP Percentc Lactate concentration (mmol/l) Before CP After CP f ? f f f ? f f f f 0.27* f f f ? f 0.12 "Values shown are mean? standard deviation. bsolution with rnmol/l of calcium ions. 'Mean and standard deviation of postarrest values expressed as percentages of prearrest values. "Significantly different (p < 0.01; paired t test) compared with values obtained after cardioplegia with St. Thomas' solution. CP = cardioplegia.

5 ~ ~~~~ ~~~ ~~~~ 316 The Annals of Thoracic Surgery Vol 39 No 4 April 1985 Table 3. Myocardial Adenosine Triphospkate and Creatine Kinase Releasearb St. Thomas' Bretschneider Hospital Solution Bretschneider Solution Solution with Calcium' Variable (N = 13) (N = 14) (N = 8) ATP concentration ? 4.2 (pmol/gm of dry weight) CK release (units) 2.1? I~I 4.6" 2.7 I~I 1.5 "Measured after a 30-minute reperfusion period following 30 minutes of ischemic arrest under cardioplegic protection. bvalues shown are mean r standard deviation. 'Solution with mniolil of calcium ions. dsignificantly different (p < 0.001; paired L test) compared with results after cardioplegia with St. Thomas' solution. ATP = adenosine triphosphate; CK = creatine kinase mmolll of calcium ions prevents it. This is in keeping with observations made by Jynge and colleagues [lo, 111. The mmol/l is known to be the lower limit of safety above which the calcium paradox does not occur It is believed that the main reason that calcium paradox is not seen after calcium-free cardioplegia in clinical operations is that residual calcium in blood and extracellular fluid prevents this. In the present study, we used approximately the same volume of calcium-free infusate relative to heart weight as is used in clinical surgical procedures. The results suggest that although the mean calcium concentration must have been well above mmovl, concentrations below the lower limit of safety must have occurred regionally, causing scattered damage. We believe that something similar might happen in cardiac operations in human beings if calciumfree solutions are used to obtain cardiac arrest. The extent of the damage will not be dramatic. Loss of 2% of total muscle mass, if scattered over the whole muscle, will not be detected easily in clinical situations, especially if it does not occur in all instances. The prevention of such damage is not unimportant, however. Release of CK in the same order of magnitude was found to be associated with postoperative complications that necessitated treatment with dopamine or prolonged assistance by partial cardiopulmonary bypass, for example, and also with an increased incidence of ventricular fibrillation and repolarization abnormalities [20]. Since most of the damage to and ultimate destruction of myocardial cells after ischemia is induced by calcium ions entering the cell, it seems advantageous to reduce the calcium concentration of the extracellular fluid to protect the heart during ischemic arrest. By exposing the tissue to a fluid with the same composition as the intracellular fluid, one expects to prevent all unwanted uptake of calcium. The intracellular concentration of free calcium is extremely low (less than mmovl), so even the addition of mmol of calcium ions per liter of calcium-free solution violates the principle of protection by equilibration of intracellular and extracellular fluid. Apart from the great practical problem of rapidly changing the composition of the extracellular fluid into one approaching the intracellular fluids composition, it is now clear that protection by equilibration of intracellular and extracellular fluid is not practical, because low extracellular concentrations of calcium are damaging to the glycocalix (i.e., the outer layer of the cell membrane). This effect, called the calcium paradox, occurs in several types of tissue and in several species including human beings [21]. It is possible that low sodium and high magnesium levels and low temperature delay the calcium paradox, but if calcium-free solutions are used in clinical operations, damage is most likely prevented by residual calcium in the heart. Apart from the fact that true protection by intracellular and extracellular fluid equilibration is not achieved in this way, we have demonstrated in a rat heart model that despite residual calcium, myocardial cell damage develops. The damage is slight and scattered, varies from heart to heart, and, if it occurs the same way in human beings, is likely to pass unnoticed postoperatively in clinical situations. Therefore, we conclude that calcium-free solutions should not be used for cardioplegia in clinical surgery. Because of the damaging effect of low extracellular calcium, cardioprotection by equilibration of the fluid composition inside and outside the cell (if it can be achieved in clinical operations) is not a practical solution to the problem of cardioprotection during ischemic arrest. References 1. Buckberg GD: A proposed solution to the cardioplegic controversy. J Thorac Cardiovasc Surg 77:803, Conti VR, Bertrand EG, Blackstone EH: Cold cardioplegia versus hypothermia for myocardial protection. J Thorac Cardiovasc Surg 76:577, Roberts AJ, Abel Rh4, Alonso DR Advantages of hypothermic potassium cardioplegia and superiority of continuous versus intermittent aortic cross-clamping. J Thorac Cardiovasc Surg 79:44, Alfieri 0, Vermeulen FEF, Knaepen JJ, et al: Extensive myocardial revascularization: influence of cardioplegia on operative results. Thorac Cardiovasc Surg 28:343, Tyers GFO: Metabolic arrest of the ischemic heart. Ann Thorac Surg 20:91, Zimmerman ANE, Hulsmann WC: Paradoxical influences of calcium ions on the permeability of the cell membranes of the isolated rat heart. Nature 211:646, Zimmerman ANE, Daems W, Hulsmann WC, et al: Mor-

6 317 Hendriks et al: Cold Ischemic Arrest with Calcium and Noncalcium Solutions phological changes of heart muscle caused by successive perfusion with calcium-free and calcium-containing solutions (calcium paradox). Cardiovasc Res 1:201, Yates JC, Dhalla NS: Structural and functional changes associated with failure and recovery of hearts after perfusion with Ca-free medium. J Mol Cell Cardiol 791, Frank IS, Lange GA, Nudd LM, Seraydarian K The myocardial cell surface, its histochemistry, and the effect of sialic acid and calcium removal on its structure and cellular exchange. Circ Res 41:702, Jynge P, Hearse DJ, Braimbridge MV: Myocardial protection during ischemic arrest: a possible hazard with calciumfree cardioplegic infusates. J Thorac Cardiovasc Surg 73:848, Jynge P: Protection of the ischemic myocardium. Calciumfree cardioplegic infusates and the additive effects of coronary infusion and ischemia in the induction of the calcium paradox. Thorac Cardiovasc Surg 28:303, 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, Nordbeck H, Bretschneider HJ, Fuchs C, et al: Methode und Ergebnisse einer neuen Form des kunstlichen Herzstillstandes im Tierexperiment und unter klinischen Bedingungen. Thoraxchirurgie 22:582, Stapenhorst K Prolonged safe ischemic arrest using hypotherinic Bretschneider cardioplegia combined with topical cardiac cooling. Thorac Cardiovasc Surg 29:272, Bretschneider HJ: Myocardial protection. Thorac Cardiovasc Surg 28:295, Boink ABTJ, Ruigrok TJC, de Moes D, et al: The effect of hypothermia on the occurrence of the calcium paradox. Pflugers Arch 385:105, Duvelleroy MA, Duruble M, Martin JL, et al: Bloodperfused working isolated rat heart. J Appl Physiol41:603, Hendriks FFA, Jonas J, van der Laarse A, Huysmans HA: Cardioplegic arrest in isolated blood-perfused working rat hearts. J Surg Res 35:41, Ruigrok TJC, Burgersdijk FJA, Zimmerman ANE: The calcium paradox: a reaffirmation. Eur J Cardiol 3:59, Brower RW, de Jong JW, Haalebos M, et al: Evaluation of cardioplegia in coronary artery bypass graft surgery. In Just H, Tschirkov A, Schlosser V (eds): Kalziumantagonisten zur Kardioplegie und Myokardprotektion in der offenen Herzchirurgie. Stuttgart, Thieme, 1982, pp Lomsky M, Ekroth R, Poupa 0: The calcium paradox and its protection by hypothermia in human myocardium. Eur Heart J 4H:139, 1983