Calcium-induced ventricular contraction during cardioplegic arrest

J

THoRAc CARDIOVASC SURG

1987;94:606-13

Calcium-induced ventricular contraction during cardioplegic arrest Cardiac arrest induced by hyperkalemic perfusion is generally considered to represent a state of complete electromechanical arrest. However, high-energy phosphate concentrations and ventricular function decrease with increasing cardioplegic calcium concentrations, possibly because of elevated resting muscle tone produced by calcium influx. We examined isolated rat hearts containing an isovolumic intraventricular baUoon for the presence of contractile activity during the administration at 10° C of a cardioplegic solution containing potassium, 20 mEqjL. Significant left ventricular pressure was developed (35.6% ± 4.3 % of prearrest systolic pressure) during administration of a solution containing a calcium concentration of 1.0 mmoljL and far less (9.7% ± 1.6% of prearrest systolic pressure) with a calcium-free cardioplegic solution. The muscle contraction diminished with repeated doses, was increased by increasing cardioplegic calcium content, and was inihibited by magnesium. Adenosine triphosphate and creatine phosphate concentrations were 9.0 ± 1.4 and 7.0 ± 0.9 nmoljmg dry weight immediately after infusion of 15 mI of a hypoxic cardioplegic solution containing calcium, versus 13.3 ± 1.3 (p < 0.02) and 31.9 ± 3.5 nmoljmg dry weight (p < 0.0001) after a hypoxic acalcemic solution was given. When repeated doses of a hypoxic cardioplegic solution containing calcium in a concentration of 1.0 mmoljL were given at 15 minute intervals at 10° C, ischemic contracture (a sustained development of ventricular pressure, mean 51 % ± 4 % of prearrest systolic pressure) resulted within 1 hour. Coronary vascular resistance was increased during the muscle contractions induced by calcium-containing solutions, markedly so during contracture. Calcium-related mechanical activity was also observed during hypothermic cardioplegic arrest in five of six isolated isovolumic canine hearts. We conclude that hearts remain potentiaUy active mechanicaUy during cold hyperka1emic arrest and undergo energeticaUy wasteful contraction when stimulated with calcium-containing hyperkalemic cardioplegic solutions.

David F. Torchiana, MD, Tim R. Love, MD, William G. Hendren, MD, Gillian A. Geffin, MB, BS, James S. Titus, Brian E. Redonnett, Dennis D. O'Keefe, MD, and Willard M. Daggett, MD, Boston, Mass.

HyperkalemiC cardioplegic arrest is used to provide myocardial protection and a still, bloodless field for all types of cardiac surgical procedures. Perfusion of the coronary arteries with a cold hyperkalemic solution results in membrane depolarization and electrical inactivation of the heart, a state described as diastolic arrest. Prompt cessation of the heart's mechanical activity and the metabolic slowing provided by hypothermia are the key facets of this technique in maintaining myocardial From the Department of Surgery, Massachusetts General Hospital and Harvard Medical School, Boston, Mass. Supported in part by National Institutes of Health Grant HL 12777. Received for publication Sept. 25, 1986. Accepted for publication Nov. 17, 1986. Address for reprints: Willard M. Daggett, MD, Department of Surgery, Massachusetts General Hospital, Boston, MA 02114.

606

energy stores through a period of ischemia and optimizing functional recovery on reperfusion. The membrane action potential is the usual stimulus for the contractile elements of the heart through calcium release from the sarcoplasmic reticulum. However, electrical arrest by sustained membrane depolarization does not eliminate mechanical activity in all circumstances. Specifically, Niedergerke' observed in frog heart strips that high extracellular potassium concentrations (100 mEqjL) permitted increased influx of extracellular calcium resulting in sustained muscle contraction without electrical activity. This phenomenon has been called potassium contracture. Resting tone with hyperkalemic perfusion can be further elevated by extracellular sodium depletion. 2, 3 With typical hyperkalemic cardioplegic solutions (CS) in the isolated rat heart there is evidence that increasing CS calcium concentration can result in a

Volume 94 Number 4 October 1987

Calcium-induced contraction

607

40 36 32 28

~

24

-...I ~

20 16 12 8 4 0

2

3

4

5

CS DOSE

6

7

8

Fig. 2. Progressive decline in pressure development with successivedoses of CS containing calcium concentration of 1.0 mmol/L in a group of six hearts. Percent left ventricular pressure (LVP) = 100 X (peak pressure with each CS dose - prearrest end-diastolic pressure) / (prearrest peak systolic pressure). Bars represent ± SEM.

Table I. Composition of solutions Fig. 1. Beating hearts were arrested with 15 ml of CS, and then 10 ml of additional CS was administered at 15 minute intervals. Three hearts receiving solutions with different calcium concentrations are illustrated. In each pair of ventricular pressure tracings, electrical signal is amplified in lower panel. Brackets represent time of CS administration. Pressure scales are in millimeters of mercury.

progressive reduction in postischemic high-energy phosphate stores and ventricular function.t' Calcium affects a variety of cellular functions concerning both energy production and consumption." Various authors'<' have stated that calcium in CS can produce an effect similar to potassium contracture with calcium influx into cells causing increased muscle tone and energy consumption. In this study we examine the mechanical action of isolated rat hearts during arrest with cold hyperkalemic CS and describe some aspects of such calcium-stimulated muscle activity. Canine hearts were examined for a similar phenomenon.

Materials and methods Adult male Sprague Dawley rats were given 1,000 units of heparin intraperitoneally and 20 minutes later anesthetized with 20 mg of phenobarbital. Beating hearts were excised, placed in 4° C Krebs-Henseleit

Na+ (mEq/L) K+ (mEq/L) Ca"' (mEq/L) Mg?" (mEq/L) Cl- (mEq/L) HCO, - (mfiq/L) H,PO, - (mEqjL) SO, -- (mliq/L) Glucose (mmol/L) Mannitol (rnmol/L)

Krebs-Henseleit

Cardioplegic

bicarbonate buffer

solution"

143.0 5.9 2.4 2.4 125.1 25.0 1.2 2.4 11.0

109.3

20.0

102.4 26.8

27.8 54.9

• Modified in certain protocols as described in the text.

buffer (Table I), and then perfused through the aortic root at 100 em H 20 pressure with the same solution oxygenated with 95% oxygen and 5% carbon dioxide at 37° C. An incision was made in the pulmonary artery to allow coronary effluent to escape. The left atrium was opened and the left ventricle cannulated via the mitral valve with a balloon-tipped, saline-filled catheter attached to a pressure transducer. The balloon was filled with saline to produce a left ventricular end-diastolic pressure of 5 to 10 mm Hg and a peak left ventricular pressure typically greater than 100 mm Hg. Measurements of left ventricular end-diastolic pressure, peak left ventricular pressure, coronary flow, and heart rate were

The Journal of

608

Thoracic and Cardiovascular Surgery

Torchiana et al.

Table II. Myocardial concentrations of adenine nucleotides and creatine phosphate (nmolfmg dry weight) Immediately after CS CS additives

No.

ATP

O 2, Ca, glucose O 2, no Ca, glucose N 2, Ca, no glucose N 2, no Ca, no glucose

8 8 9 10

18.1 ± 1.0'11 17.7 ± 0.6'11 9.0± ql 13.3 ± l.3t

ADP

I

7.1 ± 6.2 ± 10.4 ± 8.5 ±

0.3'11 0.3'11 0.411 O.4t

I

AMP

3.0 ± 2.5 ± 10.3 ± 5.4 ±

0.3'11 0.2'11 2.111 0.5t

I

TAN

28.1 ± 26.4 ± 29.8 ± 27.1 ±

1.0 0.7 l.3§ 1.1

I

CP

30.2 ± 40.4 ± 7.1 ± 31.9 ±

3.0'11 1.4t'll 0.9'11 3.5:111

Legend: Values are ±SEM. Dry weight values were calculated by correcting for the mean myocardial water for each group. Prearrest values were ATP = 12.1 ± 0.9, ADP = 8.9 ± 0.3, AMP = 6.3 ± 0.9, TAN = 27.3 ± 0.8, and CP = 18.8 ± 1.4 (n = 13). ATP, Adenoside triphosphate. ADP, Adenosine diphosphate. AMP, Adenosine monophosphate, TAN, Total adenine nucleotides. CP, Creatine phosphate. •p < 0.05 versus corresponding Ca ++ -containing solution. tp < 0.02 versus corresponding Cat-containing solution. :j:p < 0.001 versus corresponding Catt-containing solution. §p < 0.05 versus corresponding prearrest values. II'> < 0.02 versus corresponding prearrest values. ~ < 0.00 I versus corresponding prearrest values. #p < 0.05 versus corresponding values immediately after CS. "p < 0.02 versus corresponding values immediately after CS. ttp < 0.001 versus corresponding values immediately after CS.

obtained for 15 minutes at 5 minute intervals to establish the stability of each heart. The aortic line was clamped and the heart arrested with CS delivered at 4° C and at a pressure of 65 mm Hg while balloon pressure and aortic pressure were constantly recorded. The volume of CS used to produce arrest was 15 ml. In studies involving multidose CS all subsequent doses were 10 ml given at 15 minute intervals. The arresting CS dose cooled the hearts to 10° ± 2° C. Temperature was maintained at this level by a water-jacketed chamber surrounding the heart. CS composition is described in Table I. Unless otherwise noted CS used was oxygenated and contained glucose. When glucose was omitted an equiosmolar amount of mannitol was added to replace it. Added calcium chloride and magnesium chloride were not corrected for osmolarity. Cold CS was oxygenated with 98% oxygen and 2% carbon dioxide or made hypoxic with 98% nitrogen and 2% carbon dioxide, giving an oxygen tension of more than 500 or less than 50 mm Hg, respectively, when measured at 10° C and a pH of 7.35 to 7.50 at 37° C. Myocardial adenine nucleotides and creatine phosphate (CP) concentrations were assayed as follows: Hearts were removed from the apparatus and the ventricular apex was rapidly excised and frozen between metal paddles cooled to -70° C. The remaining part of the ventricles was assayed in duplicate for percentage heart water by dessication to constant weight at 90° C. The samples that had been frozen were divided into thirds, homogenized, and analyzed for adenosine triphosphate (ATP), adenosine diphosphate (ADP), adenosine monophosphate (AMP), and creatine phosphate

(CP) by high-pressure liquid chromotography as previously detailed.v" Total adenine nucleotides were calculated by adding total ATP, ADP, and AMP. Six isolated canine hearts were examined for calciuminduced mechanical activity during hypothermic arrest. The dogs were anesthetized with intravenous chloralose (150 mg/kg) and urethane (1.5 gm/kg), A water-filled balloon connected to a pressure transducer was inserted into the left ventricle either during coronary perfusion by a support dog'? or after cardioplegic arrest. All hearts were arrested with cold acalcemic CS (Table I) equilibrated with 98% oxygen and 2% carbon dioxide and delivered into the aortic root after cross-clamping of the aorta. Multidose cardioplegia maintained the hearts at 8° to 16° C. The effects on left ventricular pressure of doses of CS with and without calcium were compared. Data in Table II were subjected to two-way analysis of variance by the BMDP Statistical Software Program BMDP7D. 1I Data from the study of oxygenated CS were analyzed separately from data from the study of hypoxic CS. Time and CS calcium content were the between-subjects factors. For comparisons with prearrest data only, these data were entered as an additional level of the time factor at both levels of CS calcium content. When the analyses of variance rejected the hypothesis of equal means or there was a strong interaction between time and CS calcium content, specific pairs of means were compared by the unpaired t test. Values of p less than 0.05 were considered significant. All data are expressed as mean ± standard error of the mean (SEM). Animals were cared for in a humane way in compli-

Volume 94 Number 4 October 1987

Calcium-induced contraction

609

After CS and 15 min ischemia ATP

No. 10

10 10 9

11.1 14.2 10.3 12.7

± ± ± ±

I.3tt 1.3# 1.0 0.6

AMP

ADP 8.1 7.8 8.7 8.4

± ± ± ±

0.4 0.4** 0.6** 0.4

ance with the "Guiding Principles In the Care and Use of Animals" approved by the Council of the American Physiologic Society (revised 1980) and the "Guide for the Care and Use of Laboratory Animals" (NIH Publication No. 78-83, revised 1978). Results 1. In preliminary experiments, infusion of acalcemic CS resulted in an increase in left ventricular pressure that was of low magnitude and fell away to baseline upon cessation of CS administration (Fig. 1, a). With calcium added to the CS, left ventricular pressure rose markedly, and slowly returned to baseline once perfusion ended (Fig. 1, b). In six hearts given acalcemic CS the contractions averaged 9.7% ± 1.6% of prearrest developed pressure. With calcium added to a concentration of 1.0 mmol/L, the hearts produced 35.6% ± 4.3% of prearrest developed pressure. When the electrocardiogram was monitored during this period no electrical activity was observed with either solution. 2. A series of CS doses containing an ionized calcium concentration of 1.0 mmoljL were given to a group of six hearts. The contractions observed during CS infusion progressively decreased in magnitude (Fig. 2). 3. When a sequence of calcium-containing CS doses was interrupted with a CS dose of higher calcium concentration, the ensuing contraction was increased over the previous contraction size, which suggests a relationship between peak pressure height and calcium concentration (Fig. 1, c). 4. To establish that the pressure increases measured by the balloon were actively generated by the myocardium and not simply a hydrostatic effect of CS delivery, we performed experiments on hearts containing both an intraventricular balloon and ultrasonic crystals. The 2 mm piezoelectric crystals were placed in the ventricular wall approximately parallel to the left ventricular equator to measure myocardial chord length. The crystals were connected to a sonomicrometer (Norland NI-202, Fort Atkinson, Wis) and the output recorded on a strip chart. There was systolic shortening both during the prearrest perfusion period and during reperfusion after ischemia, which indicated appropriate crystal position-

6.0 5.0 7.1 4.9

± ± ± ±

1.0** 1.0# 1.2 0.6

TAN 25.2 27.0 26.1 25.9

± ± ± ±

0.8** 0.6 0.8** 0.8

CP 10.0 24.4 5.0 12.0

± i.ztt ± 2.3Ht ± 0.5 ± 1.2:j:tt

ing. As shown in Fig. 3, the rise in balloon pressure observed with CS infusion is mirrored by chord shortening and confirms that the contraction was an active event. 5. Magnesium can inhibit calcium influx at the cell membrane" and is included in the disparate CS formulations of Hearse, Stewart, and Braimbridge!' and Bretschneider." When added to calcium-containing CS at a concentration of 16 mmol/L, magnesium inhibited the contractions seen with a calcium-containing solution (Fig. 4). 6. High-energy phosphate concentrations were determined immediately after a single arresting dose of CS containing calcium, 1.0 mmoljL, or acalcemic CS or after arrest and 15 minutes of ischemia (Table 11). When the solutions were oxygenated and contained glucose, ATP concentration was greater after arrest than beforehand and did not differ significantly with or without calcium. CP levels were greater with the oxygenated acalcemic solution: Both groups had levels greater than control. After arrest and the ensuing period of ischemia both ATP and CP fell, more so with the calcium-containing group. Similar experiments were done subsequently using hypoxic, glucose-free CS to minimize aerobic energy production during CS administration and thus emphasize differences in energy consumption. Immediately after arrest ATP concentrations were 13.3 ± 1.3 nmol/ mg dry weight with acalcemic CS and 9.0 ± 1.4 nmol/mg with calcium added (p = 0.Q1l4). The calcium effect on CP concentrations was even greater, with values of 31.9 ± 3.5 and 7.1 ± 0.9 nmol/rng dry weight, respectively. The corresponding values after 15 minutes of ischemia failed to show the trend toward further depletion seen with calcium-containing oxygenated CS. There were no major changes in total adenine nucleotides with varied CS calcium content. When ATP values diminished, AMP tended to show a reciprocal rise in all groups. We confirmed in these hearts that the left ventricular pressure developed during CS administration was much higher with calcium-containing solutions. Expressed as a percentage of prearrest systolic pressure, a single dose of

6 10

The Journal of Thoracic and Cardiovascular Surgery

Torchiana et al.

Fig. 3. Simultaneous tracings recorded by intraventricular balloon and intramyocardial ultrasonic crystals are displayed in upper and lower panels. respectively. Before arrest, crystals demonstrate myocardial chord shortening during ventricular systole and lengthening during diastole. With infusion of CS containing calcium, 1.0 mmol/L, myocardial chord shortening occurs when ventricular pressure rises. LVDP, Left ventricular diastolic pressure.

50f-

0 {;

•

No Co '" Ca++ Ca++ + Mg++

,1

40

~

-....I

I I I I I I I

30

~

20

I

Fig. 5. Successive doses of hypoxic CS containing calcium, 1.0 mmol/L, were given. Baseline pressure started to rise before fifth CS dose with a subsequent sustained increase in ventricular pressure representing contracture. Brackets represent time of CS administration. Pressure scales in millimeters of mercury.

I

I

10

0

~ I

1

I

2

3

CS DOSE Fig. 4. After two acalcemic doses of CS, hearts received either CS containing calcium in a concentration of 1.0 mmol/L (n = 6) or CS containing calcium, 1.0 mmol/L, and magnesium, 16 mmoljL (n = 6). Added magnesium almost totally inhibited augmented pressure development seen with addition of calcium only. Percent LVP as in Fig. 2.

calcium-containing CS produced contractions of 29.6% ± 1.7% (hypoxic, n = 19) and 30.4% ± 2% (oxygenated, n = 18), whereas the pressures produced by acalcemic CS were 6.5% ± 0.7% (hypoxic, n = 17) and 9.9% ± 0.9% (oxygenated, n = 18). 7. Successive doses of hypoxic CS containing a calcium concentration of 1.0 mmoljL were given to 10 hearts..After three to five doses of CS, left ventricular pressure began to rise, plateauing at a high level of

resting tone (51% ± 4% of prearrest developed pressure in the group of 10 hearts) (Fig. 5). Morphologically, this pressure recording is analogous to those observed in published descriptions of ischemic contracture," a state of tetanic contraction resulting from ATP depletion. When a single dose of hypoxic, calcium-containing CS was given to 10 hearts followed by unmodified ischemia at 10° C, no contracture was observed over a similar period of I hour. 8. Oxygenated CS was given in multidose fashion to groups of hearts for a total of 2 hours. Three groups were compared: CS containing a calcium concentration of 1.0 mmoljL with and without glucose and acalcemic CS without glucose (Fig. 6). CS with calcium and glucosecaused contractions of progressively diminishing magnitude, as described earlier in Fig. 2. Acalcemic CS produced contractions of smaller magnitude with little change after the third dose. When calcium-containing CS was given without glucose, most of the hearts had evidence of contracture within the 2 hour period, as evidenced by increasing left ventricular pressure between CS doses typically starting late during a IS

Volume 94 Number 4 October 1987

Calcium-induced contraction 6 I 1

N2,1.0mM Co", Glucose o 02,1.0mM Co", No Glucose • 02 ,1.0mM"Co*, Glucose o 02, No Co ,No Glucose

l:>

40

20

01.0 mM Co*,No Glucose e1.0 mM Co*, GkJcase o No Co* ,No Glucose

36

32 28

~ ..... ~

24 20

10

16

8

12

o'------:------=---~--'------!

CS DOSE

8 4 0

6

7

8

CS DOSE

Fig. 6. Pressure developed with multidose oxygenated CS of varying calcium and glucose contents is illustrated. With glucose and calcium (1.0 mmoljL) present (n = 6), contractions progressively declined (data the same as in Fig. 2). In absence of glucose (n = 6), calcium-containing CS produced contractions of similar magnitude, but four of six hearts developed contracture (arrows) after the fifth dose and one more heart after seventh dose of CS. Glucose-free acalcemic CS (n = 6) produced contractions of low magnitude without causing contracture. Percent LVP as in Fig. 2.

minute ischemic interval (in four of six hearts after the fifth dose; in one more heart after the seventh). Before development of contracture in these hearts there was no difference in the magnitude of the contractions with the calcium-containing CS in the presence or absence of glucose, as seen from the near superimposition of the data points in Fig. 6. 9. Coronary vascular resistance was higher during arrest with calcium-containing CS than with acalcemic CS. This seemed to be a specific effect of the muscle contraction itself rather than calcium-induced vasoconstriction. The larger contractions associated with the first few infusions of calcium-containing CS were associated with a moderate elevation of coronary vascular resistance (Fig. 7). With later doses, producing smaller contractions, coronary vascular resistance fell to a level similar to that found with acalcemic CS. When contracture occurred with hypoxic calcium-containing CS, coronary vascular resistance rose markedly. The increased coronary vascular resistance found with con-

Fig. 7. Coronary vascular resistance (CVR) is shown for same three groups as in Fig. 6 and for additional hearts receiving hypoxic calcium-containing CS (n = 6). First two doses of calcium-containing CS produced large contractions (Fig. 6) associated with CVR greater than with acalcemic CS. As contractions diminish, difference in CVR decreases. With hypoxic calcium-containing CS, onset of contracture between third and fifth CS doseis associated with elevated CVR. Bars represent ± SEM. CVR is calculated by dividing time required to deliver CS dose by volume of dose.

tracture and with the larger contractions appears to result from compression of the coronary circulation by the contracting muscle. 10. In five of six canine hearts, calcium-containing CS produced greater increases in left ventricular pressure than acalcemic CS. Traces from one representative experiment are displayed in Fig. 8. Discussion During cold hyperkalemic arrest, the rat myocardium remains potentially active mechanically and contracts when stimulated with calcium-containing solutions despite continuing electrical arrest. The magnitude of this contraction is increased with increasing calcium concentration and is extinguished with repetition. Magnesium markedly reduces the calcium-induced contraction. A single hypoxic, glucose-free arresting dose of calcium-containing CS produces significantly lower ATP and CP concentrations than the same CS without calcium. If produced repetitively the contractions are associated with energy depletion of such severity that a contracture state develops more rapidly with multiple CS infusions than with a single CS infusion followed by hypothermic ischemia. Contracture induced in this manner can be avoided by including oxygen and glucose

6 I 2 Torchiana et al.

The Journal of Thoracic and Cardiovascular Surgery

Fig. 8. Sequential doses of acalcemic and calcium-containing (1.1 mmoljL) CS at 4° C were administered to isolated supported isovolumic canine heart at approximately 10 minuteintervals. Calcium-containing CS produced larger and more sustained elevations of ventricular pressure than acalcemic CS. Brackets represent time of CS administration. Pressure scale in millimeters of mercury.

in the CS. Similar calcium-dependent contractions have been observed in our laboratory during cold hyperkalemic arrest in the dog heart (Fig. 8). The increase in resting muscle tone produced by calcium-containing cardioplegic perfusates can be modified in a number of ways. Alkalosis," extracellular sodium depletion, z.a and high potassium concentrations I all increase calcium influx. Acidosis" and magnesium" are protective. In addition to the substantial cost of tension development, calcium influx during CS administration can also expend energy by stimulating the sarcoplasmic reticulum calcium pump" and inhibiting mitochondrial ATP production." The increased wall tension produced by calcium-containing CS may also indirectly affect energy balance by raising coronary resistance and thereby slowing the rate of myocardial cooling and delaying the onset of cardiac arrest. In a previous report we" noted a threefold increase in coronary vascular resistance when hypoxic calciumcontaining CS was repeatedly administered to isolated rat hearts. In our present study the increase in coronary vascular resistance accompanies the development of

contracture and appears to result from extrinsic compression of the coronary circulation that occurs when contracture develops. Coronary vascular resistance is also elevated to a lesser degree in association with the muscle contractions produced by calcium-containing CS. The elevation of coronary vascular resistance from increased muscle tone presumably is most pronounced in the subendocardium, where wall tension is highest, and could reduce CS delivery to this region. Analysis of coronary resistance data from a previously reported canine study? showed elevated coronary vascular resistance in some animals after a prolonged cross-clamp period with multidose hypoxic calcium-containing CS, which suggests that contracture may develop in the dog in the same fashion. In our previous rat study' coronary vascular resistance returned to normal, and functional recovery, although depressed, was seen in these hearts after reperfusion with warm oxygenated Krebs-Henseleit buffer. Whether to include calcium in CS and how much to include remain controversial problems. Calcium is present in plasma and extracellular fluid in the physio-

Volume 94 Number 4 October 1987

logic state at an ionized concentration of approximately 1.0 mmol/L, Its absence from coronary perfusates can result in devastating consequences with cellular rupture and myocardial necrosis on readmission of calcium." However, this sort of picture-the calcium paradox-is not seen clinically with acalcemic CS because of the protective influences of hypothermia." relatively low CS volumes," and noncoronary collateral flOW 22 with resultant calcium contamination. Rapid arrest is facilitated by acalcemic CS, and addition of calcium can cause energetically wasteful mechanical activity and elevate coronary vascular resistance. Clearly it is counterproductive to stimulate muscle activity and energy consumption with a solution intended to minimize both. Our current practice is to accept the low level of calcium (0.2 to 0.3 mmol/L, ionized) obligated by our use of oxygenated dilute blood CS and to try to minimize its detrimental aspects by the addition of magnesium at a concentration of 16 mmol/L, as employed by Braimbridge and Hearse in the St. Thomas' Hospital solution. We wish to thank Alvin Denenberg, Douglas Malnati, James Vath, Carmello Bondi, and Richard Wawrzynski for performing the high-pressure liquid chromatographic and chemical analyses. We thank Matthew Parker and Michael Geffin for technical assistance and help with the data analysis. The manuscript was typed by Mary Chasse and Judy Feiner, to whom we are also grateful. REFERENCES 1. Niedergerke R. The potassium chloride contracture of the heart and its modification by calcium. J Physiol 1956;134:584-99. 2. Luttgau HC, Niedergerke R. The antagonism between Ca and Na ions on the frog's heart. J Physiol 1958;143:486-505. 3. Hearse DJ, Braimbridge MV, Jynge P. Components of cardioplegic solution. In: Protection of the ischemic myocardium: cardioplegia, New York: Raven Press, 1981: 230-1. 4. Boggs BR, Torchiana DF, Geffin GA, et al. Optimal myocardial preservation with an acalcemic crystalloid cardioplegic solution. J THoRAc CARDIOVASC SURG 1987;93:838-46. 5. Pernot AC, Ingwall JS, Menasche P, et al. Evaluation of high-energy phosphate metabolism during cardioplegic arrest and reperfusion: a phosphorus-31 nuclear magnetic resonance study. Circulation 1983;67:1296-1303. 6. Weber A, Murray JM. Molecular control mechanisms in muscle contraction. Physiol Rev 1973;53:612-73. 7. Gebhard MM, Bretschneider HJ, Gersing E, Preusse CJ, Schnabel PA, Ulbricht LJ. Calcium-free cardioplegiapro. Eur Heart J 1983;4(Suppl H):151-60.

Calcium-induced contraction 6 1 3

8. DeBoer LWV, Ingwall JS, Kloner RA, Braunwald E. Prolonged derangements of canine myocardial purine metabolism after a brief coronary occlusion not associated with anatomic evidence of necrosis. Proc Nat! Acad Sci USA 1980;77:5471-5. 9. Randolph JD, Toal KW, Geffin GA, et al. Improved myocardial preservation with oxygenated cardioplegic solutions as reflected by on-line monitoring of intramyocardial pH during arrest. J Vase Surg 1986;3:216-25. 10. Teplick R, Haas GS, Trautman E, Titus J, Geffin G, Daggett WM. Time dependence of the oxygen cost of force development during systole in the canine left ventricle. Circ Res 1986;59:27-38. 11. Brown MB, Engelman L, Frane JW, Hill MA, Jennrich RI, Toporek JD: BMDP Statistical Software 1981, In Dixon WJ, ed. Berkeley, California: The University of California Press, 1981:105-15. 12. Shine KI, Douglas AM. Magnesium effects on ionic exchange and mechanical function in rat ventricle. Am J Physiol 1974;227:317-24. 13. Hearse DJ, Stewart DA, Braimbridge MV. Myocardial protection during ischemic cardiac arrest: the importance of magnesium in cardioplegic infusates. J THORAC CARDIOVASC SURG 1978;75:877-85. 14. Bretschneider HJ. Myocardial protection. Thorac Cardiovase Surg 1980;28:295-302. 15. Hearse DJ, Garlick PB, Humphrey SM. Ischemic contracture of the myocardium: mechanisms and prevention. Am J Cardiol 1977;39:986-93. 16. Langer GA. The effect of pH on cellular and membrane calcium binding and contraction of myocardium. Circ Res 1985;57:374-82. 17. Rich TL, Brady AJ. Potassium contracture and utilization of high energy phosphates in rabbit heart. Am J Physiol 1974;226:105-13. 18. Lehninger AL. Ca 2+ transport by mitochondria and its possible role in the cardiac contraction-relaxation cycle. Circ Res 1974;34,35(Suppl III):III83-8. 19. Zimmerman ANE, Hulsmann we. Paradoxical influence of calcium ions on the permeability of the cell membranes of the isolated rat heart. Nature 1966; 211:646-7. 20. Holland CE, Olson RE. Prevention by hypothermia of paradoxical calcium necrosis in cardiac muscle. J Mol Cell Cardiol 1975;7:917-28. 21. Jynge P. Protection of the ischemic myocardium: calciumfree cardioplegic infusates and additive effects of coronary infusion and ischemia in the induction of the calcium paradox. Thorac Cardiovasc Surg 1980;28:303-9. 22. Brazier J, Hottenrott C, Buckberg G. Noncoronary collateral myocardial blood flow. Ann Thorac Surg 1975;19:426-35.