- Email: [email protected]
Calcium and cardioplegia
J
THoRAc CARDIOVASC SURG
87:908-912, 1984
Calcium and cardioplegia The optimal calcium content for the St. Thomas' Hospital cardioplegic solution The relatiolLShip between the calcium content of the St, Thomas' Hospital cardioplegic solution and the degree of tissue protection it affords has been characterized by means of an isolated working rat heart preparation subjected to normothermic ischemic arrest With a 3 minute period of preischemic infusion of solutiolLS containing calcium cbloride coneeatrations of 0, 0.6, 1.0, 1.1, 1.2, 1.3, 1.4, and 2.4 mmol/L and 35 minutes of nonilothermic ischemic arrest, postischemic creatine kinase leakage was 562.3 ± 26.7, 64.8 ± 12.1, 63.3 ± 9.9, 37.5 ± 5.4, 33.4 ± 3.1, 41.2 ± 4.1, 58.6 ± 7.4, and 65.7 ± 9.1 1U/15 min/gm dry weight, respectively. The postischemic recovery of aortic flow was 0%, 29.2% ± 6.8%, 29.0% ± 3.3%, 45.8% ± 4.9%, 55.5% ± 1.4%, 38.5% ± 2.7%, 27.8% ± 8.6%, and 19.5 ± 7.1 %, respectively. The results indicate optimal protection with a calcium concentration of 1.2 mmol/L and a rapid decline in protection with even small changes in concentration either side of the optimum. The hazard of total absence of calcium was confirmed by the induction of the calcium paradox. This study shows that under normothermic conditiOILS a calcium concentration of 1.2 mmol/L is optimal for the St, Thomas' Hospital solution. This study reinforces the importance of undertaking dose-response studies for all components of all cardioplegic solutiolLS.
Furnio Yamamoto, M.D.,* Mark V. Braimbridge, F.R.C.S., and David J. Hearse, D.Sc., London, United Kingdom
One of the more variable components of the different cardioplegic solutions is the calcium ion. The Bretschneider' and Kirsch' solutions are essentially calcium free, whereas the Tyers' and St. Thomas' Hospital" solutions contain calcium in a concentration of up to 1.2 rnmol/L, Protagonists of low or zero calcium solutions argue that calcium reduction induces diastolic arrest rapidly and that extracellular calcium reduction should reduce tissue injury by avoiding intracellular calcium overload. Counteracting these advantages, however, is the danger of inducing a calcium paradox':" during reperfusion. As the effects of calcium are tightly linked From The Heart Research Unit, The Rayne Institute, St. Thomas' Hospital. London, United Kingdom. Supported by grants from the British Heart Foundation and St. Thomas' Hospital Research Endowment Fund. Received for publication May 31, 1983. Accepted for publication Aug. 30, 1983. Address for reprints: Dr. D. J. Hearse, The Rayne Institute, St. Thomas' Hospital, London, SEI 7EH. United Kingdom. *Kleinwort Fellow, The Rayne Institute, St. Thomas' Hospital, London, United Kingdom.
908
Table I. Composition of the basic St. Thomas' Hospital cardioplegic solution Concentration Compound
Sodium chloride Potassium chloride Magnesium chloride Calcium chloride Sodium bicarbonate pH adjusted to 7.8 Osmolarity = 324 Osrn/kg H 20
(mmolfl.)
110.0 16.0 16.0 1.2
10.0
to other ions, particularly sodium, it is not possible to make generalized statements about the ideal calcium concentration of an extracellular solution without consideration of all ions present and their interactions. Table I shows the formulation of the St. Thomas' Hospital solution. The sodium content is near normal and there is an excess of magnesium and potassium compared with their concentrations in extracellular fluid. With the exception of the calcium content, all the components of this solution have been optimized in relation one to another. The present study details the
Volume 87 Number 6
Calcium and cardioplegia
June 1984
Oxygen
+
Langendorff Reservoir
909
results of a normothermic isolated rat heart study in which the optimal calcium concentration of the St. Thomas' Hospital cardioplegic solution was determined on the basis of functional and enzymatic studies. Materials and methods
Aortict Flow
Cardioplegia Reservoir
••
··
Coronary Flow
Fig. 1. The isolated perfused working heart model was used in this study. The rat heart is cannulated via the left atrium and the aorta and is maintained in a thermostatically controlled heart chamber. In the Langendorff mode tap TI is open (taps T2. T3. and T4 are closed), and oxygenated perfused fluid flows from a thermostatically controlled reservoir to the aortic cannula and exits from the right heart. Conversion from the empty beating Langendorff preparation to the working (ejecting) mode is accomplished by closing tap T1 and opening taps T2 and T3. Under these conditions oxygenated perfusion fluid flows from the atrial reservoir, which delivers a constant perfusion pressure (18 em H 20) to the left atrium. The perfusate passes to the left ventricle, which spontaneously (electrical pacing was not used in this study) ejects the perfusion fluid via the aorta and an elasticity chamber (containing 2.5 ml air) against a hydrostatic pressure (100 em H 20) to the top of the lung. Coronary perfusate exits from the right heart and is mixed with the aortic outflow and reoxygenated. The oxygenated perfusion fluid is returned via a pump and a microfilter to the atrial reservoir. Aortic flow is measured in the aortic outflow tract, either manually or with an electromagnetic flowmeter; coronary flow is measured from the heart chamber; intracavity pressures are recorded with appropriately located pressure transducers, which also can be used for the measurement of heart rate. Total cardiopulmo-
Hearts. Hearts were obtained from male rats (280 to 320 gm body weight) of the Wistar strain. Experimental model (Fig. I). The isolated perfused working heart model, which has already been described in detail,' is a left heart preparation in which oxygenated perfusion medium at 37° C enters the cannulated left atrium at a pressure equivalent to 18 ern H 20 and is passed to the left ventricle. It is spontaneously ejected from the left ventricle (electrical pacing was not used in this study) at a rate of 60 to 70 mlJmin via an aortic cannula against a hydrostatic pressure equivalent to 100 em H 20. Coronary effluent can be sampled for biochemical analysis or pooled and recirculated with the aortic outflow. Total cardiopulmonary bypass with maintained coronary perfusion may be simulated by clamping the left atrial cannula and introducing perfusion fluid at 37° C into the aorta from a reservoir located 100 em above the heart. This preparation, which is essentially that described by Langendorff," will continue to beat but does not perform external work. Ischemic cardiac arrest may be induced in this preparation by clamping the aortic cannula. Short periods of preischemic coronary infusion (at 37° C) of cardioplegic solutions may be achieved by use of a reservoir (located 60 em above the heart) attached to a side arm of the aortic cannula. Experimental time course. Hearts were subjected to 5 minutes of aerobic Langendorff perfusion, 20 minutes of aerobic working perfusion, 3 minutes of cardioplegic infusion (37° C) with the St. Thomas' Hospital cardioplegic solution containing various concentrations of calcium, 35 minutes of ischemic arrest at 37° C, 15 minutes of aerobic Langendorff reperfusion, and 20 minutes of aerobic working reperfusion. nary bypass with maintained coronary perfusion is simulated by closing tap T2. tap T3. and tap T4. opening tap TI. and introducing perfusion fluid at 37° C into the aorta from a reservoir located 100 em above the heart. Ischemic arrest is simulated by closing tap T2 and tap T3 (tap TI and tap T4 are also closed). Single or multiple coronary infusion of cardioplegic solution is achieved by opening tap T4. which connects the aortic cannula to a reservoir containing cardioplegic fluid at the appropriate hydrostatic pressure (60 ern H 20). All reservoirs and chambers are surrounded by a water jacket to maintain the desired temperature.
The Journal of Thoracic and Cardiovascular Surgery
9 10 Yamamoto, Braimbridge, Hearse
60
ischemic recovery of function could be expressed as a percentage of initial functional performance and related to the calcium content of the cardioplegic solution. Six hearts were used for each condition studied and the results are expressed as the mean ± standard error of the mean.
A
i§ 50
u:
:Eo ""15
e-
40 30
~
lil
""
20
~
Results
10
Hearts were subjected to 3 minutes of cardioplegic infusion at 37° C with the St. Thomas' Hospital cardioplegic solution containing one of the following concentrations of calcium chloride: 0, 0.6, 1.0, 1.1, 1.2, 1.3, 1.4, and 2.4 mmol/L, After 35 minutes of normothermic ischemic arrest, a time which was deliberately selected so as to permit an approximate recovery of 50% in the control group (calcium concentration 1.2 mmol/ L) to allow the detection of modifications which either increase or reduce recovery, hearts were reperfused in the Langendorff mode for 15 minutes. During this time all coronary effluent was collected for the measurement of creatine kinase leakage. This was followed by a 20 minute perfusion in the working mode during which time functional indices recovered to plateau. Fig. 2 shows the results for the postischemic recovery of aortic flow and creatine kinase leakage in relation to the calcium content of the preischemic cardioplegic infusion solution. Table II shows the results for all other functional indices. In terms of cardiac function, optimal myocardial protection was observed with a calcium concentration of 1.2 mmol/L, Under these conditions the mean recovery of aortic flow was 55.5% ± 1.4%. As shown in Fig. 2, A, myocardial protection declined rapidly with increasing or decreasing calcium concentrations, and a range of only 0.2 mmol/L on either side of the optimum approximately halved recovery. Even a change in calcium content as small as 0.1 mmol/L resulted in a significant (p < 0.001) reduction in protection. Cardiac functional indices were accompanied by reciprocal changes in enzyme leakage (Fig. 2, B). Enzyme leakage (33.4 ± 3.1 IU/15 min/gm dry weight) was lowest at the calcium concentration of 1.2 mmoljL. Reduction of the calcium concentration to zero abolished postischemic recovery of function. Upon reperfusion, hearts failed to contract and massive enzyme leakage occurred (562.3 ± 26.7 IV j 15 mini gm dry weight). These changes were characteristic of those seen with the calcium paradox.
c: 0.
o
o
0.6
1.0
l.l
1.3
1.2
1.4
2.4
Calcium Concentration Immoles/liter) B
i ~ = E" ~
=:
f
.s ~
c
;;;: ~
"
~
e
u
:l~ 100 80
60 40 20 0.6
1.0
l.l
1.2
1.3
1.4
2.4
Calcium Concentration (mmoles/liter>
Fig. 2. The effect of the concentration of calcium in the cardioplegic solution upon myocardial protection. The relationship between the calcium content of a preischemic cardioplegic solution and postischemic (A) recovery of aortic flow expressed as a percent of its preischemic control and (B) creatine kinase leakage. Hearts were subjected to 35 minutes of normothermic ischemic arrest. Six hearts were used for each group and the bars represent the standard error of the mean.
Cardioplegic solution. The basic St. Thomas' Hospital cardioplegic solution (without procaine) was used. The composition is shown in Table I. Creatine kinase assay. Creatine kinase leakage into the coronary effluent during the Langendorff reperfusion period was measured by the method of Urdal and
Stromme," Expression of results. During the preischemic working control period the following indices of cardiac function were recorded: heart rate, coronary flow, aortic flow, and aortic pressure. The cardiac output was derived from the sum of aortic and coronary flow and the stroke volume by dividing cardiac output by heart rate. During the recovery period these were again measured and expressed as a percentage of their individual preischemic control values. In this way the post-
Discussion The results of this study have demonstrated that under normothermic conditions a calcium concentration
Volume 87 Number 6
Calcium and cardioplegia 9 1 1
June 1984
Table II. Dose-response characteristics for the protective properties of calcium in the St. Thomas' Hospital
cardioplegic solution Calcium concentration
Percent recovery of cardiac function Coronary flow
Cardiac output
Aortic pressure
Heart rate
Stroke volume
Creatine kinase leakage (1U/15/min/gm dry wt]
(mmolfl.)
Aortic flow
0 0.6 1.0 1.1
0* 29.2 ± 6.8t 29.0 ± 3.3* 45.8 ± 4.9
1.2
55.5 ± 1.4
76.3 ± 5.6
61.3 ± 1.9
92.5 ± 2.1
89.3 ± 2.0
68.7 ± 2.9
33.4 ± 3.1
1.3 1.4 2.4
38.5 ± 2.7* 27.8 ± 8.6t 19.5 ± 7.1*
70.5 ± 2.1 68.2 ± 3.6 58.8 ± 12.4
47.3 ± 2.4t 39.3 ± 6.9:1: 30.8 ± 7.8t
87.3 ± 2.7 82.0 ± 2.4 70.2 ± 14.2
84.8 ± 2.5 95.7 ± 3.4 73.2 ± 15.2
62.2 ± 4.2 40.8 ± 6.6t 34.5 ± 8.2t
41.2 ± 4.1 58.6 ± 7.4t 65.7 ± 9.1t
I
25.0 ± 80.2 ± 79.3 ± 79.5 ±
16.4t 6.1 3.6 3.6
I
7.5 ± 43.8 ± 41.7 ± 55.2 ±
5.0* 5.7:1: 3.7* 4.0
I
24.0 ± 80.8 ± 82.7 ± 81.8 ±
15.2t 4.2 3.0 2.3
I
29.2 ± 87.2 ± 91.8 ± 84.3 ±
18.4t 3.1 2.8 2.7
I
8.7 ± 50.8 ± 47.7 ± 65.2 ±
5.8* 7.2§ 3.6t 4.0
562.3 ± 64.8 ± 63.3 ± 37.5 ±
26.7* 12.1§ 9.9t 5.4
*p < 0.001. tp < 0.01. :j:p < 0.02. §p < 0.05.
of 1.2 mmol/L (approximately 1.0 mmoljL ionized calcium) in the St. Thomas' Hospital formulation is associated with maximal myocardial protection. Second, a bell-shaped dose-response curve exists, but this exhibits an exceptionally narrow optimal range, with relatively small increases or decreases in calcium content resulting in a large fall in protective capability. Third, under conditions less than 25 ~moljL, massive injury results probably through the induction of the calcium paradox. It is perhaps no coincidence that the optimal calcium concentration as defined by these studies matched our previously used concentration because all of the other ionic components of the solution were optimized with a calcium concentration of 1.2 mmoljL. As the concentration of one ion can greatly influence the effect of another, it is necessary to stress that the calcium concentration of 1.2 mmol/L may apply only to this particular cardioplegic formulation and may even apply only under normothermic conditions. A small change in formulation, such as a reduction of sodium, may well influence the optimal calcium concentration and the extent of protection afforded at that concentration. As stressed earlier, one limitation of these studies is that they were carried out under conditions of normothermic ischemic arrest and therefore cannot be automatically extrapolated to hypothermic ischemic arrest. It is possible that our observed bell-shaped dose-response curve may differ or be shifted by cooling. Although this possibility cannot be excluded, in an earlier study of magnesium dose-response characteristics, we" observed a bell-shaped dose-response curve with normothermic experiments. When these experiments were repeated under hypothermic conditions, an identical dose-
response profile was observed. One aspect of the present normothermic investigation which we would expect to change with hypothermia is the severe calcium paradox-like injury observed with zero calcium cardioplegia. As we6 have discussed previously, a number of factors, including hypothermia, protect against the calcium paradox; thus, if the present experiments were repeated at 20° C, considerably less damage would be expected in the zero calcium group. The observation of a sharp peak in the dose-response curve was unexpected. It emphasizes that optimal protection depends upon fme ionic balance and shows how sensitive this balance is to even small changes of concentration. Whatever the mechanism, the results clearly argue for very careful quantity control in the preparation of cardioplegic solutions. This point is particularly relevant when hygroscopic compounds like calcium chloride are used, as variable degrees of hydration can lead to substantial errors of composition. The third point to emerge from these studies is the confirmation that, with the St. Thomas' Hospital cardioplegic solution, the elimination of calcium is hazardous and may result in lethal calcium paradox. This danger exists for any cardioplegic solution that contains nearly normal physiological levelsof sodium and that is used in the physiological pH range. As we12• 13 have stressed before, certain solutions such as Bretschneider's are protected by their low sodium content (and probably also by contaminant calcium) and by the fact that the solutions are normally used under hypothermic conditions, as both of these factors are known to protect against induction of the calcium paradox. 10, II However, other solutions, some of which are in clinical use, do not have such margins of safety, and a strong argument
The Journal of Thoracic and Cardiovascular Surgery
912 Yamamoto. Braimbridge, Hearse
exists for considering increasing the calcium content of such solutions. Discussion of the preceding points serves to stress how a small change in formulation can result in major changes in the efficacy of cardioplegic solutions. This work reinforces the argument that each component of any cardioplegic solution in clinical use should be carefully characterized in terms of dose-response. As previously shown with magnesium," procaine and lignocaine," and calcium antagonists such as verapamil," excessively high or low concentrations may fail to exploit the protective properties of the additive and, more seriously, may cause toxic effects that negate the protective properties of the entire solution. The assistance of Dr. D. Chambers is gratefully acknowledged. REFERENCES
2
3
4
5
Bretschneider HJ, Hubner G, Knoll D, Lohr B, Nordbeck H, Spieckermann PG: Myocardial resistance and tolerance to ischemia. Physiological and biochemical basis. J Cardiovasc Surg (Torino) 16:241-260, 1975 Kirsch Y, Rodewald G, Kalmar P: Induced ischemic arrest. Clinical experience with cardioplegia in open-heart surgery. J THORAC CARDIOVASC SURG 63:121-130, 1972 Tyers GFO, Manley NJ, Williams EH, Shaffer CW, Williams DR, Kurusz M: Preliminary clinical experience with isotonic hypothermic potassium-induced arrest. J THORAC CARDIOVASC SURG 74:674-681, 1977 Hearse DJ, Braimbridge MY, Jynge P: Protection of the ischemic myocardium, Cardioplegia, ed I, New York, 1981, Raven Press, p, 346-348 Zimmerman Ane, Daems W, Hulsmann W, Snyder J,
6
7
8 9
10
II
12
13
14
Wisse E, Durrer D: Morphological changes of heart muscle caused by successive perfusion with calcium-free and calcium-containing solutions (calcium paradox). Cardiovasc Res 1:201-209, 1967 Jynge P, Hearse DJ, Braimbridge MY: Myocardial protection during ischemic cardiac arrest. A possible hazard with calcium-free cardioplegic infusates. J THORAC CARDIOVASC SURG 73:846-855, 1977 Hearse DJ, Braimbridge MY, Jynge P: Protection of the ischemic myocardium, Cardioplegia, ed 1, New York, 1981, Raven Press, p 59-63 Langendorff 0: Untersuchungen am iiberlebenden Saugertierherzen. Pflugers Arch 61:291, 1895 Urdal P, Stromme JH: Effects of Ca, Mg and EDTA on creatine activity in cerebrospinal fluid. Clin Chern 25:147150, 1979 Hearse DJ, Braimbridge MY, Jynge P: Protection of the ischemic myocardium, Cardioplegia, ed 1, New York, 1981, Raven Press, pp 219-224 Jynge P: Protection of the ischemic myocardium. An experimental evaluation of cardioplegic infusates, Doctoral Thesis, University of Tromso, Norway Hearse DJ, Stewart DA, Braimbridge MY: Myocardial protection during ischemic cardiac arrest. The importance of magnesium in cardioplegic infusates. J THORAC CARDIOVASC SURG 75:877-885, 1978 Hearse DJ, O'Brien K, Braimbridge MY: Protection of the myocardium during ischemic arrest. Dose-response curves for procaine and lignocaine in cardioplegic solutions. J THORAC CARDIOVASC SURG 81:873-879, 1981 Yamamoto F, Manning AS, Braimbridge MY, Hearse DJ: Cardioplegia and slow calcium-ehannel blockers. Studies with verapamil. J THORAC CARDI0VASC SURG 86:252-261, 1983
THoRAc CARDIOVASC SURG
87:908-912, 1984
Calcium and cardioplegia The optimal calcium content for the St. Thomas' Hospital cardioplegic solution The relatiolLShip between the calcium content of the St, Thomas' Hospital cardioplegic solution and the degree of tissue protection it affords has been characterized by means of an isolated working rat heart preparation subjected to normothermic ischemic arrest With a 3 minute period of preischemic infusion of solutiolLS containing calcium cbloride coneeatrations of 0, 0.6, 1.0, 1.1, 1.2, 1.3, 1.4, and 2.4 mmol/L and 35 minutes of nonilothermic ischemic arrest, postischemic creatine kinase leakage was 562.3 ± 26.7, 64.8 ± 12.1, 63.3 ± 9.9, 37.5 ± 5.4, 33.4 ± 3.1, 41.2 ± 4.1, 58.6 ± 7.4, and 65.7 ± 9.1 1U/15 min/gm dry weight, respectively. The postischemic recovery of aortic flow was 0%, 29.2% ± 6.8%, 29.0% ± 3.3%, 45.8% ± 4.9%, 55.5% ± 1.4%, 38.5% ± 2.7%, 27.8% ± 8.6%, and 19.5 ± 7.1 %, respectively. The results indicate optimal protection with a calcium concentration of 1.2 mmol/L and a rapid decline in protection with even small changes in concentration either side of the optimum. The hazard of total absence of calcium was confirmed by the induction of the calcium paradox. This study shows that under normothermic conditiOILS a calcium concentration of 1.2 mmol/L is optimal for the St, Thomas' Hospital solution. This study reinforces the importance of undertaking dose-response studies for all components of all cardioplegic solutiolLS.
Furnio Yamamoto, M.D.,* Mark V. Braimbridge, F.R.C.S., and David J. Hearse, D.Sc., London, United Kingdom
One of the more variable components of the different cardioplegic solutions is the calcium ion. The Bretschneider' and Kirsch' solutions are essentially calcium free, whereas the Tyers' and St. Thomas' Hospital" solutions contain calcium in a concentration of up to 1.2 rnmol/L, Protagonists of low or zero calcium solutions argue that calcium reduction induces diastolic arrest rapidly and that extracellular calcium reduction should reduce tissue injury by avoiding intracellular calcium overload. Counteracting these advantages, however, is the danger of inducing a calcium paradox':" during reperfusion. As the effects of calcium are tightly linked From The Heart Research Unit, The Rayne Institute, St. Thomas' Hospital. London, United Kingdom. Supported by grants from the British Heart Foundation and St. Thomas' Hospital Research Endowment Fund. Received for publication May 31, 1983. Accepted for publication Aug. 30, 1983. Address for reprints: Dr. D. J. Hearse, The Rayne Institute, St. Thomas' Hospital, London, SEI 7EH. United Kingdom. *Kleinwort Fellow, The Rayne Institute, St. Thomas' Hospital, London, United Kingdom.
908
Table I. Composition of the basic St. Thomas' Hospital cardioplegic solution Concentration Compound
Sodium chloride Potassium chloride Magnesium chloride Calcium chloride Sodium bicarbonate pH adjusted to 7.8 Osmolarity = 324 Osrn/kg H 20
(mmolfl.)
110.0 16.0 16.0 1.2
10.0
to other ions, particularly sodium, it is not possible to make generalized statements about the ideal calcium concentration of an extracellular solution without consideration of all ions present and their interactions. Table I shows the formulation of the St. Thomas' Hospital solution. The sodium content is near normal and there is an excess of magnesium and potassium compared with their concentrations in extracellular fluid. With the exception of the calcium content, all the components of this solution have been optimized in relation one to another. The present study details the
Volume 87 Number 6
Calcium and cardioplegia
June 1984
Oxygen
+
Langendorff Reservoir
909
results of a normothermic isolated rat heart study in which the optimal calcium concentration of the St. Thomas' Hospital cardioplegic solution was determined on the basis of functional and enzymatic studies. Materials and methods
Aortict Flow
Cardioplegia Reservoir
••
··
Coronary Flow
Fig. 1. The isolated perfused working heart model was used in this study. The rat heart is cannulated via the left atrium and the aorta and is maintained in a thermostatically controlled heart chamber. In the Langendorff mode tap TI is open (taps T2. T3. and T4 are closed), and oxygenated perfused fluid flows from a thermostatically controlled reservoir to the aortic cannula and exits from the right heart. Conversion from the empty beating Langendorff preparation to the working (ejecting) mode is accomplished by closing tap T1 and opening taps T2 and T3. Under these conditions oxygenated perfusion fluid flows from the atrial reservoir, which delivers a constant perfusion pressure (18 em H 20) to the left atrium. The perfusate passes to the left ventricle, which spontaneously (electrical pacing was not used in this study) ejects the perfusion fluid via the aorta and an elasticity chamber (containing 2.5 ml air) against a hydrostatic pressure (100 em H 20) to the top of the lung. Coronary perfusate exits from the right heart and is mixed with the aortic outflow and reoxygenated. The oxygenated perfusion fluid is returned via a pump and a microfilter to the atrial reservoir. Aortic flow is measured in the aortic outflow tract, either manually or with an electromagnetic flowmeter; coronary flow is measured from the heart chamber; intracavity pressures are recorded with appropriately located pressure transducers, which also can be used for the measurement of heart rate. Total cardiopulmo-
Hearts. Hearts were obtained from male rats (280 to 320 gm body weight) of the Wistar strain. Experimental model (Fig. I). The isolated perfused working heart model, which has already been described in detail,' is a left heart preparation in which oxygenated perfusion medium at 37° C enters the cannulated left atrium at a pressure equivalent to 18 ern H 20 and is passed to the left ventricle. It is spontaneously ejected from the left ventricle (electrical pacing was not used in this study) at a rate of 60 to 70 mlJmin via an aortic cannula against a hydrostatic pressure equivalent to 100 em H 20. Coronary effluent can be sampled for biochemical analysis or pooled and recirculated with the aortic outflow. Total cardiopulmonary bypass with maintained coronary perfusion may be simulated by clamping the left atrial cannula and introducing perfusion fluid at 37° C into the aorta from a reservoir located 100 em above the heart. This preparation, which is essentially that described by Langendorff," will continue to beat but does not perform external work. Ischemic cardiac arrest may be induced in this preparation by clamping the aortic cannula. Short periods of preischemic coronary infusion (at 37° C) of cardioplegic solutions may be achieved by use of a reservoir (located 60 em above the heart) attached to a side arm of the aortic cannula. Experimental time course. Hearts were subjected to 5 minutes of aerobic Langendorff perfusion, 20 minutes of aerobic working perfusion, 3 minutes of cardioplegic infusion (37° C) with the St. Thomas' Hospital cardioplegic solution containing various concentrations of calcium, 35 minutes of ischemic arrest at 37° C, 15 minutes of aerobic Langendorff reperfusion, and 20 minutes of aerobic working reperfusion. nary bypass with maintained coronary perfusion is simulated by closing tap T2. tap T3. and tap T4. opening tap TI. and introducing perfusion fluid at 37° C into the aorta from a reservoir located 100 em above the heart. Ischemic arrest is simulated by closing tap T2 and tap T3 (tap TI and tap T4 are also closed). Single or multiple coronary infusion of cardioplegic solution is achieved by opening tap T4. which connects the aortic cannula to a reservoir containing cardioplegic fluid at the appropriate hydrostatic pressure (60 ern H 20). All reservoirs and chambers are surrounded by a water jacket to maintain the desired temperature.
The Journal of Thoracic and Cardiovascular Surgery
9 10 Yamamoto, Braimbridge, Hearse
60
ischemic recovery of function could be expressed as a percentage of initial functional performance and related to the calcium content of the cardioplegic solution. Six hearts were used for each condition studied and the results are expressed as the mean ± standard error of the mean.
A
i§ 50
u:
:Eo ""15
e-
40 30
~
lil
""
20
~
Results
10
Hearts were subjected to 3 minutes of cardioplegic infusion at 37° C with the St. Thomas' Hospital cardioplegic solution containing one of the following concentrations of calcium chloride: 0, 0.6, 1.0, 1.1, 1.2, 1.3, 1.4, and 2.4 mmol/L, After 35 minutes of normothermic ischemic arrest, a time which was deliberately selected so as to permit an approximate recovery of 50% in the control group (calcium concentration 1.2 mmol/ L) to allow the detection of modifications which either increase or reduce recovery, hearts were reperfused in the Langendorff mode for 15 minutes. During this time all coronary effluent was collected for the measurement of creatine kinase leakage. This was followed by a 20 minute perfusion in the working mode during which time functional indices recovered to plateau. Fig. 2 shows the results for the postischemic recovery of aortic flow and creatine kinase leakage in relation to the calcium content of the preischemic cardioplegic infusion solution. Table II shows the results for all other functional indices. In terms of cardiac function, optimal myocardial protection was observed with a calcium concentration of 1.2 mmol/L, Under these conditions the mean recovery of aortic flow was 55.5% ± 1.4%. As shown in Fig. 2, A, myocardial protection declined rapidly with increasing or decreasing calcium concentrations, and a range of only 0.2 mmol/L on either side of the optimum approximately halved recovery. Even a change in calcium content as small as 0.1 mmol/L resulted in a significant (p < 0.001) reduction in protection. Cardiac functional indices were accompanied by reciprocal changes in enzyme leakage (Fig. 2, B). Enzyme leakage (33.4 ± 3.1 IU/15 min/gm dry weight) was lowest at the calcium concentration of 1.2 mmoljL. Reduction of the calcium concentration to zero abolished postischemic recovery of function. Upon reperfusion, hearts failed to contract and massive enzyme leakage occurred (562.3 ± 26.7 IV j 15 mini gm dry weight). These changes were characteristic of those seen with the calcium paradox.
c: 0.
o
o
0.6
1.0
l.l
1.3
1.2
1.4
2.4
Calcium Concentration Immoles/liter) B
i ~ = E" ~
=:
f
.s ~
c
;;;: ~
"
~
e
u
:l~ 100 80
60 40 20 0.6
1.0
l.l
1.2
1.3
1.4
2.4
Calcium Concentration (mmoles/liter>
Fig. 2. The effect of the concentration of calcium in the cardioplegic solution upon myocardial protection. The relationship between the calcium content of a preischemic cardioplegic solution and postischemic (A) recovery of aortic flow expressed as a percent of its preischemic control and (B) creatine kinase leakage. Hearts were subjected to 35 minutes of normothermic ischemic arrest. Six hearts were used for each group and the bars represent the standard error of the mean.
Cardioplegic solution. The basic St. Thomas' Hospital cardioplegic solution (without procaine) was used. The composition is shown in Table I. Creatine kinase assay. Creatine kinase leakage into the coronary effluent during the Langendorff reperfusion period was measured by the method of Urdal and
Stromme," Expression of results. During the preischemic working control period the following indices of cardiac function were recorded: heart rate, coronary flow, aortic flow, and aortic pressure. The cardiac output was derived from the sum of aortic and coronary flow and the stroke volume by dividing cardiac output by heart rate. During the recovery period these were again measured and expressed as a percentage of their individual preischemic control values. In this way the post-
Discussion The results of this study have demonstrated that under normothermic conditions a calcium concentration
Volume 87 Number 6
Calcium and cardioplegia 9 1 1
June 1984
Table II. Dose-response characteristics for the protective properties of calcium in the St. Thomas' Hospital
cardioplegic solution Calcium concentration
Percent recovery of cardiac function Coronary flow
Cardiac output
Aortic pressure
Heart rate
Stroke volume
Creatine kinase leakage (1U/15/min/gm dry wt]
(mmolfl.)
Aortic flow
0 0.6 1.0 1.1
0* 29.2 ± 6.8t 29.0 ± 3.3* 45.8 ± 4.9
1.2
55.5 ± 1.4
76.3 ± 5.6
61.3 ± 1.9
92.5 ± 2.1
89.3 ± 2.0
68.7 ± 2.9
33.4 ± 3.1
1.3 1.4 2.4
38.5 ± 2.7* 27.8 ± 8.6t 19.5 ± 7.1*
70.5 ± 2.1 68.2 ± 3.6 58.8 ± 12.4
47.3 ± 2.4t 39.3 ± 6.9:1: 30.8 ± 7.8t
87.3 ± 2.7 82.0 ± 2.4 70.2 ± 14.2
84.8 ± 2.5 95.7 ± 3.4 73.2 ± 15.2
62.2 ± 4.2 40.8 ± 6.6t 34.5 ± 8.2t
41.2 ± 4.1 58.6 ± 7.4t 65.7 ± 9.1t
I
25.0 ± 80.2 ± 79.3 ± 79.5 ±
16.4t 6.1 3.6 3.6
I
7.5 ± 43.8 ± 41.7 ± 55.2 ±
5.0* 5.7:1: 3.7* 4.0
I
24.0 ± 80.8 ± 82.7 ± 81.8 ±
15.2t 4.2 3.0 2.3
I
29.2 ± 87.2 ± 91.8 ± 84.3 ±
18.4t 3.1 2.8 2.7
I
8.7 ± 50.8 ± 47.7 ± 65.2 ±
5.8* 7.2§ 3.6t 4.0
562.3 ± 64.8 ± 63.3 ± 37.5 ±
26.7* 12.1§ 9.9t 5.4
*p < 0.001. tp < 0.01. :j:p < 0.02. §p < 0.05.
of 1.2 mmol/L (approximately 1.0 mmoljL ionized calcium) in the St. Thomas' Hospital formulation is associated with maximal myocardial protection. Second, a bell-shaped dose-response curve exists, but this exhibits an exceptionally narrow optimal range, with relatively small increases or decreases in calcium content resulting in a large fall in protective capability. Third, under conditions less than 25 ~moljL, massive injury results probably through the induction of the calcium paradox. It is perhaps no coincidence that the optimal calcium concentration as defined by these studies matched our previously used concentration because all of the other ionic components of the solution were optimized with a calcium concentration of 1.2 mmoljL. As the concentration of one ion can greatly influence the effect of another, it is necessary to stress that the calcium concentration of 1.2 mmol/L may apply only to this particular cardioplegic formulation and may even apply only under normothermic conditions. A small change in formulation, such as a reduction of sodium, may well influence the optimal calcium concentration and the extent of protection afforded at that concentration. As stressed earlier, one limitation of these studies is that they were carried out under conditions of normothermic ischemic arrest and therefore cannot be automatically extrapolated to hypothermic ischemic arrest. It is possible that our observed bell-shaped dose-response curve may differ or be shifted by cooling. Although this possibility cannot be excluded, in an earlier study of magnesium dose-response characteristics, we" observed a bell-shaped dose-response curve with normothermic experiments. When these experiments were repeated under hypothermic conditions, an identical dose-
response profile was observed. One aspect of the present normothermic investigation which we would expect to change with hypothermia is the severe calcium paradox-like injury observed with zero calcium cardioplegia. As we6 have discussed previously, a number of factors, including hypothermia, protect against the calcium paradox; thus, if the present experiments were repeated at 20° C, considerably less damage would be expected in the zero calcium group. The observation of a sharp peak in the dose-response curve was unexpected. It emphasizes that optimal protection depends upon fme ionic balance and shows how sensitive this balance is to even small changes of concentration. Whatever the mechanism, the results clearly argue for very careful quantity control in the preparation of cardioplegic solutions. This point is particularly relevant when hygroscopic compounds like calcium chloride are used, as variable degrees of hydration can lead to substantial errors of composition. The third point to emerge from these studies is the confirmation that, with the St. Thomas' Hospital cardioplegic solution, the elimination of calcium is hazardous and may result in lethal calcium paradox. This danger exists for any cardioplegic solution that contains nearly normal physiological levelsof sodium and that is used in the physiological pH range. As we12• 13 have stressed before, certain solutions such as Bretschneider's are protected by their low sodium content (and probably also by contaminant calcium) and by the fact that the solutions are normally used under hypothermic conditions, as both of these factors are known to protect against induction of the calcium paradox. 10, II However, other solutions, some of which are in clinical use, do not have such margins of safety, and a strong argument
The Journal of Thoracic and Cardiovascular Surgery
912 Yamamoto. Braimbridge, Hearse
exists for considering increasing the calcium content of such solutions. Discussion of the preceding points serves to stress how a small change in formulation can result in major changes in the efficacy of cardioplegic solutions. This work reinforces the argument that each component of any cardioplegic solution in clinical use should be carefully characterized in terms of dose-response. As previously shown with magnesium," procaine and lignocaine," and calcium antagonists such as verapamil," excessively high or low concentrations may fail to exploit the protective properties of the additive and, more seriously, may cause toxic effects that negate the protective properties of the entire solution. The assistance of Dr. D. Chambers is gratefully acknowledged. REFERENCES
2
3
4
5
Bretschneider HJ, Hubner G, Knoll D, Lohr B, Nordbeck H, Spieckermann PG: Myocardial resistance and tolerance to ischemia. Physiological and biochemical basis. J Cardiovasc Surg (Torino) 16:241-260, 1975 Kirsch Y, Rodewald G, Kalmar P: Induced ischemic arrest. Clinical experience with cardioplegia in open-heart surgery. J THORAC CARDIOVASC SURG 63:121-130, 1972 Tyers GFO, Manley NJ, Williams EH, Shaffer CW, Williams DR, Kurusz M: Preliminary clinical experience with isotonic hypothermic potassium-induced arrest. J THORAC CARDIOVASC SURG 74:674-681, 1977 Hearse DJ, Braimbridge MY, Jynge P: Protection of the ischemic myocardium, Cardioplegia, ed I, New York, 1981, Raven Press, p, 346-348 Zimmerman Ane, Daems W, Hulsmann W, Snyder J,
6
7
8 9
10
II
12
13
14
Wisse E, Durrer D: Morphological changes of heart muscle caused by successive perfusion with calcium-free and calcium-containing solutions (calcium paradox). Cardiovasc Res 1:201-209, 1967 Jynge P, Hearse DJ, Braimbridge MY: Myocardial protection during ischemic cardiac arrest. A possible hazard with calcium-free cardioplegic infusates. J THORAC CARDIOVASC SURG 73:846-855, 1977 Hearse DJ, Braimbridge MY, Jynge P: Protection of the ischemic myocardium, Cardioplegia, ed 1, New York, 1981, Raven Press, p 59-63 Langendorff 0: Untersuchungen am iiberlebenden Saugertierherzen. Pflugers Arch 61:291, 1895 Urdal P, Stromme JH: Effects of Ca, Mg and EDTA on creatine activity in cerebrospinal fluid. Clin Chern 25:147150, 1979 Hearse DJ, Braimbridge MY, Jynge P: Protection of the ischemic myocardium, Cardioplegia, ed 1, New York, 1981, Raven Press, pp 219-224 Jynge P: Protection of the ischemic myocardium. An experimental evaluation of cardioplegic infusates, Doctoral Thesis, University of Tromso, Norway Hearse DJ, Stewart DA, Braimbridge MY: Myocardial protection during ischemic cardiac arrest. The importance of magnesium in cardioplegic infusates. J THORAC CARDIOVASC SURG 75:877-885, 1978 Hearse DJ, O'Brien K, Braimbridge MY: Protection of the myocardium during ischemic arrest. Dose-response curves for procaine and lignocaine in cardioplegic solutions. J THORAC CARDIOVASC SURG 81:873-879, 1981 Yamamoto F, Manning AS, Braimbridge MY, Hearse DJ: Cardioplegia and slow calcium-ehannel blockers. Studies with verapamil. J THORAC CARDI0VASC SURG 86:252-261, 1983