Ischemic myocardial protection

J THoRAc CARDIOVASC SURG 1988;95:239-46

Ischemic myocardial protection Comparison of nonoxygenated crystalloid, oxygenated crystalloid, and oxygenated fluorocarbon cardioplegic solutions This study was designed to compare myocardial protection with a nonoxygenated crystalloid solution, an oxygenated crystalloid solution, and an oxygenated fluorocarbon cardioplegic solution. Postischemic ventricular performance was studied in three equal (N = 7) groups of dogs subjected to 120 minutes of global ischemia induced at an average myocardial temperature of 18.5° ± 1.4° C (range 17.0° to 21.0° C). Left ventricular global and regional function was evaluated by sonomicrometry and micromanometers before ischemia and at 45 and 60 minutes after ischemia. Stroke volume index, left ventricular pressure-minor external diameter loop area, percent shortening, first derivative of left ventricular pressure, mean velocity of circumferential fiber shortening, and the slope of the end-systolic pressure were used to evaluate myocardial contractility. In vitro oxygen content of the three cardioplegic solutions was measured at a mean injection temperature of 8.3° ± 0.6° C: 0.8 ± 0.1 vol% (nonoxygenated crystalloid cardioplegia), 3.2 ± 0.2 vol% (oxygenated crystalloid cardioplegia), and 6.2 ± 0.2 vol% (oxygenated fluorocarbon cardioplegia). Recovery of global and regional function was significantly (p < 0.05) better with both oxygenated solutions than with the nonoxygenated solution. Differences between the oxygenated crystalloid and fluorocarbon groups were not significant. We conclude: (1) Compared to nonoxygenated crystalloid cardioplegia, oxygenated crystalloid and oxygenated fluorocarbon cardioplegic solutions gave superior myocardial protection during 2 hours of ischemic arrest; (2) no difference was found in protective effects between an oxygenated crystalloid and an oxygenated fluorocarbon solution.

Koichi Tabayashi, MD, Peter P. McKeown, MB, BS, Masaki Miyamoto, MD, Andrew E. Luedtke, BSEE, Robert Thomas, BA, Margaret D. Allen, MD, Gregory A. Misbach, MD, and Tom D. Ivey, MD, Seattle, Wash.

Rassium cardioplegic arrest and myocardial cooling are common adjuncts for myocardial protection during cardiac operations.':' However, complete recovery of ventricular performance and metabolism even in the arrested hypothermic heart is not always achieved because anaerobic metabolism is not completely adequate to satisfy metabolic demand during ischemia." The effects of oxygenated blood cardioplegia,' oxygen-

From the Division of Cardiothoracic Surgery. Department of Surgery, University of Washington School of Medicine, Seattle, Wash. Received for publication Aug. 12. 1986. Accepted for publication Dec. 4, 1986. Address for reprints: Peter P. Mckeown. MB, BS, FRCS(C). FRACS, Assistant Professor, Division of Cardiothoracic Surgery. Department of Surgery, RF-25, University of Washington School of Medicine, Seattle, W A 98195.

ated crystalloid cardioplegia (CC),8-10 and oxygenated fluorocarbon (Auosol-DA 20%*) cardioplegia (FC)11.12 on oxidative metabolism during induced arrest have been described in previous studies. Oxygenated cardioplegia solutions in these studies provided better metabolic and hemodynamic preservation than nonoxygenated CC solutions. Novick and associates 12 demonstrated that myocardial protection with an oxygenated FC solution was not only superior to nonoxygenated CC but was also superior to oxygenated blood cardioplegia. However, controlled laboratory studies comparing oxygenated CC with oxygenated FC solutions have not been reported. Oxygen content of an oxygenated FC solution under similar conditions is higher than that of an oxygenated *Fluosol-DA supplied by Alpha Therapeutic Corp., Los Angeles. Calif.

239

The Journal of Thoracic and Cardiovascular Surgery

240 Tabayashi et al.

Table I. Composition of cardioplegic solutions Determinations

K+ (mEq/L) Na+ (mEq/L) Ca"" (mEq/L) Mgt" (mEq/L) Glucose (rnmol/L) Hydroxyethyl starch (gm/L)

Perfluorodecalin (grn/L) Perfluorotripropylamine (grn/L)

Pluronic F-68 (gm/L) Glycerol (gm/L) Osmolarity (mOsm/L) pH

Crystalloid

Fluorocarbon

25 145 2

25 128

3

80

2 2 10

o

30

o

140 60

o o

27 8 410 7.4

o

390

7.4

ee solution." II, 12 The purpose of this study was to compare preservation of myocardial function with a nonoxygenated ee, an oxygenated ee, and an oxygenated Fe solution. Materials and methods Three equal groups of seven adult mongrel dogs whose weights ranged from 19,0 to 28,0 kg (mean 23,9 ± 0,7) were studied, Myocardial protection in Group I (control) consisted of a standard nonoxygenated CC solution, This group was compared to Group II (oxygenated CC) and Group III (oxygenated FC), Composition of each cardioplegic solution is shown in Table I. Cardioplegic solutions in Groups II and III were fully oxygenated by continuous bubbling of 100% oxygen through the recirculating solution. Oxygen tension was measured with a blood-gas analyzer (Model 113, Instrumentation Laboratory, Inc., Lexington, Mass.), Oxygen content was measured by a Model K LEX-0 2-CON analyzer (Lexington Instrument Co" Waltham, Mass.), All animals were permedicated with morphine sulfate (3 rug/kg, intramuscular), Anesthesia was induced with thiamylal sodium (Surital, 18 mg/kg) and maintained with a mixture of alpha chloralose (50 mg/kg) and urethane (400 rug/kg). Ventilation was controlled by a Harvard constant-volume ventilator (Harvard Apparatus Co" S. Natick, Mass.), The heart was exposed through a median sternotomy, Major vessels to the spleen were ligated to prevent splenic pooling, A 7F Mikro-Tip catheter transducer (Millar Instruments, Inc., Houston, Texas) was inserted into the ascending aorta via the right carotid artery for measurement of arterial pressure. Left ventricular (LV) pressure and its first derivative (L V dp/dt) were obtained from a 7F Mikro-Tip catheter transducer inserted via the LV apex, A fluid-filled polyethylene catheter was inserted into the body of the left atrium for measurement of left atrial pressure, Pressures obtained from fluid-filled catheters were measured with Statham P23b transducers (Gould Inc" Cardiovascular Products, Oxnard, Calif.), Three pairs of sonomicrometric piezoelectric crystals were implanted in the LV. One pair was sutured to the anterior and

posterior epicardial walls to obtain maximum transverse external diameter in the plane of the LV short axis, A second pair was implanted across the LV free wall to measure wall thickness, A third pair was implanted in the subendocardium perpendicular to the long axis of the ventricle near the left anterior descending artery for measurement of segmental shortening, Dimensional data were measured by a Triton sonomicrometer (Triton Technology, Inc" San Diego, Calif.), Orientation of crystal pairs was confirmed by postmortem dissection of the LV, After systemic heparinization (3 rng/kg), an arterial cannula was inserted into the left carotid artery and two venous perfusion cannulas were inserted into the right atrium, A Bentley BOS-5 pediatric bubble oxygenator (American Bentley, Irvine, Calif.) and Olson roller pump (Olson Medical Sales Corp" Ashland, Mass.) were used for extracorporeal support, The extracorporeal system was primed with Ringer's lactate (1,300 ml) and fresh homologous blood (700 ml), Baseline hematocrit after hemodilution averaged 31% (nonoxygenated CC), 33% (oxygenated CC), and 33% (oxygenated FC). Hemodynamic measurements were taken during atrial pacing at a rate of 135 beats/min, After baseline measurements, cardiopulmonary bypass was initiated, Vents were inserted into the right and left atria, During bypass, pH was maintained with sodium bicarbonate at a mean of7.4 ± 0,02; carbon dioxide tension at 21.0 ± 0,7 mm Hg; oxygen tension at 531 ± 15,6 mm Hg, and hematocrit at 24.6% ± 0,5%, Mean aortic pressure was maintained at 67 ± 2 mm Hg by adjusting systemic perfusion at flow rates between 70 and 100 ml/kgyrnin. After the preparation had stabilized, the animal was perfusion cooled to a mean rectal temperature of 29, I 0 ± 0.1 C C, the aorta cross-clamped, and chilled (8,3° ± 0.6° C) cardioplegic solution administered immediately, The initial dose of cardioplegic solution was 10 nil/kg with an additional 10 ml/kg infused every 30 minutes, The in-line injection pressure of the cardioplegic solution was maintained between 60 and 70 mm Hg by adjusting flow rate of a separate infusion pump, Cardioplegic solutions were infused over a period averaging approximately 3 minutes, Cardioplegic solution returning through the right atrial vent was diverted to a discard cylinder so that it would not accumulate in the cardiopulmonary circuit. A 24-gauge hypodermic thermistor (Model 524, Yellow Springs Instruments Co" Yellow Springs, Ohio) was placed in the anterior LV septum for continuous monitoring of myocardial temperature, Myocardial temperature was maintained at 18,5° ± 1.4° C (range 17,0° to 21,0° C), Cross-clamp time in all experiments was 120 minutes, During the last 10 minutes of aortic occlusion, systemic temperature was gradually increased to 36° C. After removal of the aortic cross-clamp, the heart was electrically defibrillated, and the animals were weaned from cardiopulmonary bypass after 30 minutes of bypass assist. Cardiotonic agents were not administered. Postbypass hemodynamics were measured at 45 and 60 minutes after removal of the cross-clamp, Hematocrit at 60 minutes after ischemia averaged 22% (nonoxygenated CC), 24% (oxygenated CC), and 23% (oxygenated FC). All experiments were performed in accordance with "The Principles of Laboratory Animal Care" formulated by the National Society for Medical Research and the "Guide for the

Volume 95 Number 2 February 1988

Ischemic myocardial protection

Care and Use of Laboratory Animals" (NIH Publication No. 80-23, revised 1978). Data analysis. The electrocardiogram, aortic pressure, LV pressure, LV dp/dt, and dimensional data were measured before ischemic and at 45 and 60 minutes after release of the cross-clamp. Each parameter was measured at rest, during vena caval occlusion, and during acute aortic constriction. Hemodynamic measurements at rest were taken at an LV end-diastolic pressure of 5 mm Hg. Analog data were digitized at 5 msec intervals with a Digital Equipment Corp. (Marlboro, Mass.) PDP 11/23 minicomputer. End-diastole was taken as the beginning of the initial upstroke in LV dp/dt. Onset of ejection was taken from the time of maximum positive LV dp/dt, End-systole was defined as the maximum ratio of high-fidelity LV pressure to volume." Hemodynamic indices were calculated from the following equations: Dint = (Dext - 2h)

(I)

LVV= 1.014 (Dintj3'55

(2)

SVI

(L VED V - LVESV)

(3)

BW TSR=

MAP (LVEDV- LVESV) X HR

C=

7r

MeanVCF

"vmax E

srnax =

(Dext - h) (EDC-ESC) ETxEDC LVESP

(LVESV- Vo) LVESP (LVESS -So)

x80

(4) (5) (6)

(7) (8)

where Dint = internal minor axis diameter, Dext = external minor axis diameter, h = wall thickness, LVV = LV intracavitary volume," LVEDV = LV end-diastolic volume, LVESV = LV end-systolic volume, SVI = stroke volume index, BW = body weight, TSR = total systemic resistance, HR = heart rate, MAP = mean aortic pressure, C = midwall minor axis circumference, meanVCF = mean velocity of circumferential fiber shortening, EDC = end-diastolic midwall circumference, ESC = end-systolic midwall circumference, ET = ejection time, LVESP = LV end-systolic pressure, 'vmax = slope of end-systolic pressure-volume relation, Vo = volume intercept, Esmax = slope of end-systolic pressuresegment length relation, LVESS = LV end-systolic segment length, and So = segment length intercept. Equations for the slope of the end-systolic pressure-volume relation and the slope of the end-systolic pressure-segment relation (Equations 7 and 8) were constructed from endsystolic pressure-volume points and segment length points that were obtained over a range of pressures, volumes, and lengths produced during vena caval occlusion, as described by Suga," Miller," and associates. Heart rate during vena caval occlusion was kept constant by atrial pacing. Instantaneous end-systolic pressure, volume, and segment length data from multiple systoles (10 to 20) were fitted by least-squared linear regression. To compare different dimensions, internal minor axis end-diastolic diameter and end-systolic diameter were normalized by dividing the measured length by prearrest end-diastolic

24 1

diameter and then multiplying by a constant of 30." Normalized systolic excursion was calculated as the difference between normalized end-diastolic and normalized end-systolic diameter. Percent shortening was determined by dividing normalized systolic excursion by the normalized end-diastolic diameter and multiplying by 100. Segment length change was normalized to an initial value of 10 mm and calculated similarly. The LV pressure-minor external diameter loop area was constructed by plotting LV pressure on the ordinate and minor external diameter on the abscissa. The area within these loops corresponds to total LV work. LV work function curves were constructed by plotting the relationship between LV end-diastolic pressure and LV pressure-minor external diameter loop area at rest and during afterload stress" induced by aortic constriction. Peak systolic LV pressure was increased 50% to 70% above control by constriction with an aortic snare. Aortic constriction was performed before ischemia and at 60 minutes after ischemia. Preischemic and postischemic heart rates were kept similar during aortic constriction so as to avoid the effects of reflex modulation on heart rate. Contractions analyzed when the aorta was undergoing constriction were measured during the initial 5 to 10 seconds of constriction and before secondary changes caused by reflex response. Water content. A section of myocardium was taken from the free wall of the LV for measurement of water content at 60 minutes after ischemia. Samples were desiccated over a period of 48 hours at 80 c C and water content expressed as the ratio of wet to dry weight. Statistical analysis. Results are expressed as a mean ± I standard error of the mean. Comparison of variables obtained at baseline with measurements taken during the postischemic period were made with the paired t test. Analysis of variance was used to determine the significance of differences between experimental groups. Differences were considered to be statistically significant at a probability value less than 0.05.

Results Volume validation. LV volume was calculated from sonomicrometrically determined dimensions where LV intravcavitary volume was taken as 1.014 (DintV 155• We compared LV stroke volume calculated by this method to simultaneously measured stroke volume derived from thermodilution cardiac output. The calculated and measured stroke volume in five dogs during control, volume loading with Ringer's lactate, and infusion of epinephrine gave a linear correlation coefficient of 0.94 (Y = 0.77 X +0.62; standard error of the estimate = 3.5 ml). The two methods were linearly related over a volume range of 4.0 to 45.0 ml. Cardioplegic oxygenation. Samples for oxygen content were taken from the cardioplegic delivery system in each animal at a mean temperature of the injected solution of 8.3° ± 0.6° C. Oxygen content of the three cardioplegic solutions averaged 0.8 ± 0.1 vol% (nonoxygenated CC); 3.2 ± 0.2 vol% (oxygenated CC), and 6.2 ± 0.2 vol% (oxygenated FC).

The Journal of Thoracic and Cardiovascular Surgery

242 Tabayashi et al.

Table Il, LV global function during baseline and at 45 and 60 minutes after ischemia at LVEDP = 5 mm Hg Oxygenated CC

Nonoxygenated CC Variables HR (beats/min) LVSP (mm Hg) (% recovery of control, %) SVI (rnl/beat/kg) (% recovery) LVPDLA (mm-mrn Hg) (% recovery) % Shortening (%) (minor diameter) +LV dp/dt (mm Hg/sec) (% recovery) Evrnax (% recovery) Vo(mm) Mean VCF (eire/sec) (% recovery)

Baseline 141 ± 4 112 ± 5 0.9 ± 0.2 207 ± 27 22.9 ± 4.1 2,419 ± 307 9.1 ± 2.2 6.2 ± 2.7 0.67 ± 0.09

45 min 137 64 (61 0.5 (54 97 (48 15.5 (55 1,188 (55 3.1 (44 7.4 0.42 (55

± ± ± ± ± ± ± ± ± ± ±

± ±

± ±

±

3 6t 7) 0.1 * 9)

zrt

10) 5.3* 13) 122t 6) 0.7t 10) 2.2 0.12* 12)

60 min

Baseline

± ± ± ± ± ± ± ± ± ± ± ± ± ±

137 ± 10 98 ± 6

134 76 (70 0.5 (67 150 (75 17.8 (72 1,515 (71 3.1 (43 9.7 0.46 (67

2 7 9) 0.1 * 9) 26* 9) 4.6* 10) 192* 9) 0.7t 8) 2.6 ± 0.10* ± 10)

0.7 ± 0.1 178 ± 29 19.4 ± 2.6 1,939 ± 144 9.7 ± 1.7 9.2 ± 2.5 0.59 ± 0.05

45 min 136 75 (78 0.6 (97 166 (109 18.0 (94 1,623 (82 8.0 (81 8.4 0.50 (85

± ± ± ± ± ± ± ±

± ± ± ± ± ± ± ±

5 4* 5) 0.1 14)+ 22 24)+ 2.5 9):1: 134* 5)§ 1.8 11):1: 2.8 0.06 9):1:

60 min 140 80 (83 0.7 (99 197 (129 20.1 (103 1,905 (100 7.7 (77 7.5 0.54 (93

± ± ± ± ± ± ±

3 5 7) 0.1 7) 17 18)11 ± 2.8 ± 3~1 ± 157 ± 7)11 ± 1.6 ± 9)11 ± 2.6 ± 0.05 ± 6~1

Legend: Values are expressed as a mean ± I standard error of the mean. LVEDP. Left ventricular end-diastolic pressure. HR, Heart rate. LVSP, Left ventricular systolic pressure. SVI, Stroke volume index. LVPDLA, Left ventricular pressure-minor external diameter loop area. +LV dpjdt, Maximum positive first derivative of left ventricular pressure. 'vmax, Slope of the end-systolic pressure-volume relation. Yo, Volume intercept of the 'vmax slope line. Mean VCF, Mean velocity of circumferential fiber shortening. CC, Crystalloid cardioplegia. FC, Fluorocarbon cardioplegia.

'p < 0.05 versus baseline; tp < 0.01 versus baseline. :j:p< 0.05 versus nonoxygenated CC at 45 minutes; §p < 0.01 versus nonoxygenated CC at 45 minutes.

lip < 0.05

versus nonoxygenated CC at 60 minutes.

LV global function. LV global function parameters at baseline and 45 and 60 minutes after ischemia are presented in Table II. Each parameter is represented by a mean value averaged over 5 cardiac cycles and is expressed as percent recovery from baseline. There were no significant differences in LV global function among the three groups at baseline. There were no significant differences in heart rate before and after arrest. Postischemic LV hemodynamics were significantly worse than baseline in the nonoxygenated CC group, In the oxygenated CC and oxygenated FC groups no significant differences were found between determinations taken at baseline and at 45 and 60 minutes after ischemia. Percent recovery of LV global function as evaluated by LV pressure-minor external diameter loop area, percent shortening of diameter, positive LV dp/dt, and mean velocity of circumferential fiber shortening were significantly better in the oxygenated CC and FC groups than in the nonoxygenated CC group. However, there were no significant differences in these parameters between the oxygenated CC and FC groups. The slope of the end-systolic pressure-volume relation was used as an index of myocardial contractility independent of preload and afterload. Postischemic recovery of the slope of the end-systolic pressure-volume relation in the oxygenated CC and FC groups was better than in the

nonoxygenated CC group, There were no significant differences in slope of the end-systolic pressure-volume relation between the oxygenated CC and FC groups, Volume intercept of the slope line of the end-systolic pressure-volume relation showed no significant differences between baseline and postischemic values in any of the three groups, Mean slope of LV work-function curves relating LV pressure-minor external diameter loop area to LV end-diastolic pressure during baseline and at 60 minutes after ischemia are shown in Fig. 1. The mean difference in slope between nonoxygenated CC (34.6 ± 17.5) and oxygenated CC (71.4 ± 17.5) was significant (p < 0.05). When the nonoxygenated CC group was compared to the oxygenated FC group (83.3 ± 11.8), there was also a significant difference, but no difference was found between the oxygenated CC and FC groups. Total systemic resistance at baseline averaged 3,175 ± 697 dyne-sec-em." (nonoxygenated CC), 3,187 ± 280 dyne-sec-em:" (oxygenated CC), and 2,922 ± 490 dyne-sec-em:" (oxygenated FC). Total systemic resistance at 60 minutes after ischemia averaged 2,359 ± 584 dyne-see-ems (nonoxygenated CC), 2,231 ± 176 dyne-sec-ern" (oxygenated CC), and 2,487 ± 700 dyne-sec-em:" (oxygenated FC). There

Volume 95 Number 2

Ischemic myocardial protection 2 4 3

February 1988

CCS02

CCC02

FCC02

Oxygenated Fe Baseline 136 ± 2 97 ± 4 0.7 ± 0.2 212 ± 24 22.4 ± 2.7 2,309 ± 218 9.2 ± 1.9 10.6 ± 5.7 0.66 ± 0.08

45 min 140 ± 77 ± (79 ± 0.6 ± (77 ± 204 ± (100 ± 20.1 ± (86 ± 1,712 ± (74 ± 7.1 ± (76 ± 13.7 ± 0.59 ± (87 ±

2 4* 2):1: 0.1 II) 25 14):1: 3.8 7):1: 167* 3):1: 1.6* 3):1: 7.0 0.09 6):1:

60 min 138 ± 85 ± (87 ± 0.7 ± (90 ± 244 ± (117 ± 22.2 ± (99 ± 2,068 ± (90 ± 8.3 ± (89 ± 14.5 ± 0.64 ± (101 ±

2 5 3) 0.2 10) 25 II~I

100 LVPDLA LVEDP

80 60

3.5 5~1

187 3)11 2.3 8)11 7.5 0.08 6)11

40 20 I-P
were significant differences (p < 0.05) in total systemic resistance before and after arrest, and no significant difference was found among the three groups. LV regional function. Table III presents LV regional function parameters measured at baseline and 45 and 60 minutes after ischemia. Changes in LV regional function before and after arrest were similar to changes in global function. Postischemic regional function was better in the oxygenated CC and FC groups than in the nonoxygenated CC group, and no significant difference between the two oxygenaged groups was found. Segment length intercept of the slope line of the end-systolic pressure-volume relation showed no significant differences between baseline and postischemic values in any of the three groups. Water content. Water content in the nonxoygenated CC, the oxygenated CC, and the oxygenated FC groups at 60 minutes after ischemia was 82.2% ± 0.7%, 79.6% ± 0.4%, and 80.1% ± 1.1%, respectively. Water content in the nonoxygenated CC group was significantly higher (p < 0.05) than in the oxygenated CC and oxygenated FC groups. Discussion Cardioplegic arrest and myocardial cooling are commonly used to enhance myocardial protection during ischemic arrest.'? However, complete myocardial protection is not always achieved and a consensus as to which modality affords the greatest degree of myocardial protection has not been reached. J-6 The efficacy of numerous cardioplegic agents has been studied clinically and experimentally.i': 7,19 yet investigation for the ideal

I'

P<0.05--l

Fig. 1. Slope (mean ± standard error of the mean) of LV function curves relating LVPDLA to LVEDP at 60 minutes (solid bars) after 2 hours of cardioplegic arrest with ee s O 2, ee c O 2 or Fe c O 2 compared with baseline (open bars). LVPDLA, Left ventricular pressure-left ventricular minor external diameter loop area. LVEDP, Left ventricular enddiastolic pressure. CCs 0 20 Nonoxygenated crystalloid cardioplegia. CCc 0 20 Oxygenated crystalloid cardioplegia. FCc O2, Oxygenated fluorocarbon cardioplegia. N'S; Not significant.

agent continues. The optimal solution should be capable of providing for the metabolic demands of the arrested ischemic heart so that preischemic energy stores are preserved. Furthermore, the ideal agent should be easy to prepare, inexpensive, and should not cause adverse systemic effects. The present study was designed to compare the myocardial protective effects of oxygenated CC and oxygenated FC solutions with a nonoxygenated CC solution. Cardiac oxygenation. Oxygen content of the three cardioplegic solutions averaged 0.8 vol% (nonoxygenated CC), 3.2 vol% (oxygenated CC), and 6.2 vol% (oxygenated FC). The directly measured oxygen contents in our study were different from values reported by others."!' Guyton,'? Kanter, I I and their associates used in vitro oxygen tensions of the cardioplegic solution to calculate oxygen contents. This method does not take into account the fact that oxygen solubility is appreciably altered by the addition of dissolved solutes in fluids at lower temperatures." Dissimilarities in measuring

244

Tabayashi et af.

Table

m.

The Journal of Thoracic and Cardiovascular Surgery

LV regional function during baseline and at 45 and 60 minutes after ischemia at LVEDP = 5 mm Hg Nonoxygenated CC Variables

Oxygenated CC

Baseline

45 min

60 min

Baseline

45 min

60 min

% Shortening (%) (segment

18.6 ± 2.7

10.0 ± 3.0t

14.9 ± 2.3'

17.5 ± 3.5

16.5 ± 3.5

17.2±3.8

length) (% recovery of baseline, %) E smax (% recovery of baseline, %) So (mm)

52.6 ± 8.5

(45 29.5 (59 7.7

(70 29.1 (54 7.5

(92 45.3 (73 8.1

(96 51.3 (78 8.2

7.6 ± 1.6

± II) ± 5.0t ± 7) ± 1.5

± 10) ± 6.4t ± 6) ± 1.7

65.4 ± 7.3 7.8 ± 1.0

± 4):1: ± 4.8 ± 9) ± 0.9

± 4)§

± 7.7' ± 6)§

± 0.8

Legend: All values represent a mean ± I standard error of the mean. "smax, Slope of the end-systolic pressure-segment length relation. SO' Segment length intercept of the "smax line. LVEDP, Left ventricular end-diastolic pressure. CC, Crystalloid cardioplegia. FC, Fluorocarbon cardioplegia.

*p < 0.05 versus baseline. tp < 0.01 versus baseline. :j:p< 0.01 versus nonoxygenated CC at 45 minutes. §p < 0.05 versus nonoxygenated CC at 60 minutes.

techniques between our study and methods used by

others'"" may have contributed to these discrepancies. Cardiac metabolism during cardioplegic arrest is significantly lower than that in the normothermic beating heart (1.1 to 2.0 versus 3.4 to 8.0 ml oxygen/lOO gm/min)Z1.22 and is further reduced by cooling. Metabolic demand of the heart has been reported to be 0.31 ml oxygen/miri/Itx) gm LV at 22° C,22 0.27 at 15° C,23 and 0.13 at 5° C. 23Nonetheless, some myocardial energy demand during ischemia is still present, as evidenced by gradual depletion of creatine phosphate and adenosine triphosphate stores." 24 In view of these findings, the concept of providing oxygen to the heart during cardioplegia has been proposed and studied by Follette,' Engelman," Bretschneider," and their colleagues. Several investigators" have used oxygenated blood potassium cardioplegic solutions as a means to augment oxidative metabolism during ischemic arrest, but their results have been inconsistent. Kanter and associates" compared oxygenated blood and fluorocarbon (Fluosol43) cardioplegic solutions with the effectiveness of a standard hyperkalemic CC solution. They reported that an FC solution produced better preservation of ventricular performance than oxygenated blood and nonoxygenated CC solutions and that there were no significant hemodynamic differences between oxygenated blood cardioplegia and nonoxygenated CC solutions. Their results may have been affected by the decreased ability of hemoglobin to release oxygen under hypothermic conditions, which leads to a marked leftward shift in the oxygen-hemoglobin dissociation curve." Oxygenation of a cold CC solution is characterized by a linear relation and an increase in oxygen solubility as temperature is lowered. 20. 26 Digerness, Vanini, and Wideman" showed that at 10° C an oxygenated CC solution is capable of releasing all of its oxygen and that

the amount of oxygen available is higher than the amount of oxygen available from arterial whole blood, oxygenated pump perfusate, or an oxygenated blood cardioplegic solution at the same low temperature. The volume of oxygen dissolved in fluorocarbon (Fluosol-DA 20%) also changes linearly with the partial pressure of oxygen, according to Henry's law." The uptake and release of oxygen is completely reversible and the rate is very fast. 27 The solubility of oxygen increases as temperature is lowered and its solubility in fluorocarbon suspension (1.32 X 10- 4 ml of oxygen/nil solution/rum Hg oxygen tension at 37° C) is higher than in a CC solution (3.22 X 10- 5 ml of oxygen/nil solution/rum Hg oxygen tension at 35° C).20.n Therefore, an FC solution can carry more oxygen than a CC solution. In this study, oxygen content of oxygenated FC (6.2 vol%) was higher than oxygen content in oxygenated CC (3.2 vol%), but recovery of ventricular performance in the oxygenated CC group was similar to that of the oxygenated FC group. Hemodynamic recovery in both the oxygenated CC and oxygenated FC groups approximated preischemic conditions of each group, These results suggest that cardiac metabolic demand during ischemic insult appeared to be satisfied with the use of these oxygenated cardioplegic solutions. In addition, this study showed that an oxygenated CC solution appears to deliver enough oxygen to the ischemic arrested heart at a myocardial temperature of 18.5° C to get excellent preservation of LV function comparable to that afforded by an oxygenated FC solution. An oxygenated CC solution has advantages in that this agent is compatible with most cardiopulmonary bypass systems, is inexpensive, and is easy to prepare. The major disadvantage of an oxygenated CC solution is related to the possibility that gas embolization may occur if temperature of the solution is not adequately

Volume 95 Number 2 February 1988

Ischemic myocardial protection

Oxygenated

Fe

Baseline

45 min

18.4 ± 2.4

15.6 ± 2.0*

64.2 ± 8.3 8.3 ± 1.1

(85 48.4 (78 8.3

± ± ± ±

5lt 5.6 9) 1.2

60 min 17.0 ± 2.6 (92 56.7 (98 8.9

± ± ± ±

5)§ 5.8 15)§ 1.0

controlled. However, this disadvantage can be corrected with a thermostatically controlled recirculating delivery system that incorporates an air filter. An oxygenated Fe solution under hypothermic conditions has a greater affinity for oxygen than an oxygenated ee or oxygenated blood cardioplegic solution.v" However, this agent is relatively expensive, and adverse reactions in the cardiovascular and respiratory systems have been reported.28 Left ventricular function. In this study, LV global function as evaluated by LV pressure-minor external diameter loop area, percent shortening of minor axis diameter, LV dp/dt, and mean velocity of circumferential fiber shortening recovered to more than 100% of baseline. Functional recovery greater than baseline may be due to the effect of intrinsic catecholamines and to the postischemic decrease in afterload after extracorporeal hemodilution." Faris and associates" showed that cardiac output and systemic vascular conductance were increased by hemodilution. Percent shortening, LV dp/dt, and mean velocity of circumferential fiber shortening are also affected by changes in loading conditions." Total systemic resistance decreased significantly after ischemia, but there was no significant difference in total systemic resistance among the three groups. Therefore, changes in total systemic resistance are not considered to interfere with comparison of LV function among the three groups. LV regional function was evaluated by percent shortening of segment length and slope of the end-systolic pressure-segment length relation. LV regional parameters showed a more consistent pattern of improved recovery in oxygenated ee and oxygenated Fe groups than in the nonoxygenated ee group. Water content. Water content in the nonoxygenated ce group was significantly higher than in the oxygenaged ee and oxygenated Fe groups. Myocardial edema depresses ventricular performance and limits coronary perfusion during reoxygenation." The impaired LV functional recovery in the nonoxygenated ee group

245

may have been due in part to the increase in myocardial water content. Myocardial edema may result from capillary membrane changes and alterations in hydrostatic and colloid osmotic pressure on either side of the capillary.9.31 Preservation of high-energy phosphates during arrest is instrumental in minimizing capillary membrane changes." Osmolarity of ee solution was lower than osmolarity of the Fe solution. However, there was no significant difference in water content between the oxygenated ee and oxygenated Fe groups, and water content in the nonoxygenated ee group was higher than in the oxygenated ee group (p < 0.05). These results suggest that oxygenation of cardioplegic solution may preserve high-energy phosphate stores and reduce the possibility of acute changes in myocardial capillary membrane integrity. Summary

This study confirms the advantages of oxygenated cardioplegic solutions. Oxygenated ee and oxygenated Fe solutions demonstrated better myocardial preservation than a nonoxygenated ee solution. The fact that we were unable to demonstrate any significant difference between oxygenated ee and oxygenated Fe groups suggests that oxygenated Fe should be further evaluated to identify any possible advantages of this agent as a cardioplegic solution before its clinical use is considered. From a practical point, oxygenated ee is inexpensive, readily available, and easy to deliver. REFERENCES I. Gay WA Jr, Ebert PA. Functional, metabolic, and morphologic effects of potassium-induced cardioplegia. Surgery 1973;74:284-90. 2. Braimbridge MV, Chayen J, Bitensky L, Hearse OJ, Junge P, Cankovic-Darracott S. Cold cardioplegia or continuous perfusion? Report on preliminary clinical experience as assessed cytochemically. J THORAC CARDIOVASC SURG 1977;74:900-6. 3. Follette 0, Fey K, Mulder JV, Buckberg GO. Prolonged safe aortic clamping by combining membrane stabilization, multidose cardioplegia, and appropriate pH reperfusion, J THORAC CARDIOVASC SURG 1977;74:683-94. 4. Nelson RL, Goldstein SM, McConnell DH, Maloney JV, Buckberg GO. Improved myocardial performance after aortic cross-clamping by combining pharmacologic arrest with topical hypothermia. Circulation 1976;54(Pt 2):IIIlI-16. 5. Engelman RM, Rousou JH, Longo F, Auvil J, Vertrees RA The time course of myocardial high-energy phosphate degradation during potassium cardioplegic arrest. Surgery 1979;86: 138-47. 6. Grover FL, Fewel JG, Schrank KP, Ghidoni JJ, Arom KV, Trinkle JK. Effects of various periods of cold

246 Tabayashi et al.

potassium cardioplegic arrest upon myocardial contractility and metabolism. 1 Surg Res 1980;28:328-37. 7. Follette DM, Mulder DG, Maloney rv, Buckberg GD. Advantages of blood cardioplegia over continuous coronary perfusion or intermittent ischemia: experimental and clinical study. 1 THORAC CARDIOVASC SURG 1978;76:60419. 8. Engelman, RM, Rousou lH, Dobbs W, Pels MA, Longo F. The superiority of blood cardioplegia in myocardial preservation. Circulation 1980;62(Pt 2):162-6. 9. Bedenhamer RM, DeBoer LWV, Geffin GA, et al. Enhanced myocardial protection during ischemic arrest: oxygenation of a crystalloid cardioplegic solution. 1 THORAC CARDIOVASC SURG 1983;85:769-80. 10. Guyton RA, Dorsey LMA, Craver 1M, et al. Improved myocardial recovery after cardioplegic arrest with an oxygenated crystalloid solution. 1 THORAC CARDIOVASC SURG 1985;89:877-87. II. Kanter KR, laffin lH, Ehrlichman Rl, Flaherty J'T, Gott VL, Gardner TJ. Superiority of perfluorocarbon cardioplegia over blood or crystalloid cardioplgia. Circulation 1981;64(Pt 2):1175-80. 12. Novick Rl, Stefaniszyn Hl, Michel RP, Burdon FD, Salerno T A. Protection of the hypertrophied pig myocardium: a comparison of crystalloid, blood, and Fluosol-DA cardioplegia during prolonged aortic clamping. 1 THORAC CARDIOVASC SURG 1985;89:547-56. 13. Sagawa K. The end-systolic pressure-volume relation of the ventricle: definition modifications and clinical use [Editorial]. Circulation 1981;63:1223-7. 14. Suga H, Sagawa K. Assessment of absolute volume from diameter of the intact canine left ventricular cavity. 1 Appl Physiol 1974;36:496-9. 15. Suga H, Sagawa K, Shoukas AA. Load independence of the instantaneous pressure-volume ratio of the canine left ventricle and effects of epinephrine and heart rate on the ratio. Circ Res 1973;32:314-22. 16. Miller WP, Liedtke Al, Nellis SH. Regional end-systolic pressure-length relationships using a volume loading technique in the intact pig heart. Circ Res 1984;55:326-35. 17. Theroux P, Ross 1 Fr, Franklin D, Kemper WS, Sasayama S. Regional myocardial function in the conscious dog during acute coronary occlusion and responses to morphine, propranol, nitroglycerin and lidocaine. Circulation 1976;53:302-14. 18. Sasayama S, Franklin D, Ross 1 lr. Hyperfunction with normal inotropic state of the hypertrophied left ventricle. Am 1 Physiol 1977;232:H418-25. 19. Johnson RG, lacocks MA, Aretz TH, et al. Comparison

The Journal of Thoracic and Cardiovascular Surgery

of myocardial preservation with hypothermic potassium and nifedipine arrest. Circulation 1982;65(Pt 2):173-80. 20. Umbreit WW, Burris RH, Stauffer IF. Manometric techniques. Minneapolis: Burgess Publishing Co, 1947:5. 21. McKeever WP, Gregg DE, Canney Pc. Oxygen uptake of the nonworking left ventricle. Circ Res 1958;6:61223. 22. Buckberg GD, Brazier lR, Nelson RL, Goldstein SM, McConnell DH, Cooper N. Studies of the effects of hypothermia on regional myocardial blood flow and metabolism during cardiopulmonary bypass. I. The adequately perfused beating, fibrillating, and arrested heart. 1 THORAC CARDIOVASC SURG 1977;73:87-94. 23. Bretschneider Hl, Hubner G, Knoll D, Lohr B, Nordbeck H, Spieckerrnann PG. Myocardial resistance and tolerance to ischemia: physiological and biochemical basis. 1 Cardiovasc Surg 1975;16:241-60. 24. Flaherty rr, lacobus WE, Weisfeldt ML, Hollis DL, Gardner T1, Gott VL. Use of phosphorus 31 nuclear magnetic resonance to assess myocardial protection during global ischemia [Abstract]. Am 1 Cardiol 1979; 43:362. 25. Severinghaus lW. Oxyhemoglobin dissociation curve correction for temperature and pH variation in human blood. 1 Appl Physiol 1958;12:485-6. 26. Digerness SB, Vanini V, Wideman FE. In vitro comparison of oxygen availability from asanguineous and sanguineous cardioplegic media. Circulation 1981;64(Pt 2):11803. 27. Naito R, Yokoyama K. Perfluorochemical blood substitutes: "FC-43 emulsion" "Fluosol-DA, 20% and 35%". Technical Information Series No.5, Osaka; The Green Cross, 1978:10-14,73-78. 28. Vercellotti GM, Hammerschmidt DE, Craddock PR, lacob HS. Activation of plasma complement by perfluorocarbon artificial blood: probable mechanism of adverse pulmonary reactions in treated patients and rationale for corticosteroid prophylaxis. Blood 1982;59: 1299-1304. 29. Faris IB, lannos 1, lamieson G, Ludbrook 1. The circulatory effects of acute hypervolemia and hemodilution in conscious rabbits. Circ Res 1981;48:825-34. 30. Mahler F, Ross 1 lr, O'Rourke RA, Covell lW. Effects of changes in preload, afterload, and inotropic state on ejection and isovolumic phase measures of contractility in the conscious dog. Am 1 Cardiol 1975;35:626-34. 31. Foglia RP, Steed DL, Follete DM, DeLand E, Buckberg GD. Iatrogenic myocardial edema with potassium cardioplegia. 1 THORAC CARDIOVASC SURG 1979;78:217-22.