Superiority of the University of Wisconsin solution over simple crystalloid for extended heart preservation

Superiority of the University of Wisconsin solution over simple crystalloid for extended heart preservation A study of left ventricular pressure-volume relationship To compare the effects of the University of Wisconsin solution with those of an extracellular crystalloid solution, Krebs-Ringer bicarbonate, as cardiac preservation media, we studied 35 adult dogs in an isolated heart preparation. Four groups of seven hearts were preserved in University of Wisconsin solution for 6 or 12 hours or in Krebs-Ringer bicarbonate solution for 6 or 12 hours. An additional group of seven hearts with no ischemia was used for a control group. In the four preservation groups, hearts were arrested by electrolyte solution (Normosol with potassium chloride, 20 mEq/L, added, 4 0 C), flushed with 200 ml of the preservation solution, and then stored in the same solution at 10 to 2 0 C. The hearts were mounted on an isolated heart preparation equiped with a computer-controlled servo-pump system that used a mock arterial system to modulate the aortic input impedance presented to the left ventricle. Left ventricular pressure-volume loops were measured on-line for 2 hours of reperfusion with autologous warm oxygenated blood. Elastance was derived from the end-systolic pressure-volume relationship, and diastolic compliance was derived from the end-diastolic pressurevolume relationship. The total left ventricular performance was assessed by the preload recruitable stroke work area, the slope, and its x-intercept, all of which derived from the stroke work (pressurevolume area)-end-diastolic volume relationship. Extended global ischemia had more deleterious effects on the end-diastolic than the end-systolic pressure-volume relationship. In confirmation with other studies, elastance did not accurately reflect the level of ventricular contractile dysfunction because of the significant amount of diastolic dysfunction. The preservation of myocardial systolic and diastolic functions, as demonstrated by the preload recruitable stroke work area and diastolic compliance, was better in the University of Wisconsin solution groups than in the Krebs-Ringer bicarbonate solution groups after 6 and 12 hours of preservation. In addition, 6 hours of preservation with University of Wisconsin solution maintained normal systolic and diastolic functions as compared with those of the control group. Preservation with University of Wisconsin solution prevented any myocardial edema formation; by contrast, this was significantly increased after 12 hours in Krebs-Ringer bicarbonate solution. Groups preserved with University of Wisconsin solution had less reperfusion injury as evidenced by the release of coronary sinus creatine kinase during reperfusion; they also had improved oxygen use during reperfusion. This study demonstrated that the University of Wisconsin solution is excellent in preserving myocardial systolic and diastolic functions after 6 hours and appears to be acceptable for up to 12 hours of storage. It is superior to the extracellular type of solutions such as the Krebs-Ringer bicarbonate solution. (J THORAC CARDIOVASC SURG 1992;103:980-92)

Wilson Ko, MD, John A. Zelano, PhD, Richard Lazzaro, MD, W. Douglas Lazenby, MD, Thomas Hamilton, MD, O. Wayne Isom, MD, and Karl H. Krieger, MD, New York, N.Y.

From theCardiothoracic Surgery Research Laboratory, The New York Hospital-Cornell University Medical College, New York, N.Y. Read at the Seventeenth Annual Meeting of The Western Thoracic Surgical Association, Seattle, Wash., June 26-29,1991.

980

Address for reprints: Wilson Ko, MD, New York Hospital-Cornell University Medical College, Division of Cardiothoracic Surgery, Box-378, 525 East 68th St., New York, NY 10021. 12/6/35719

Volume 103 Number 5

Heart preservation

May 1992

The current cardiac preservation technique involving the use of static hypothermic hyperkalemic crystalloid solutions allows good clinical results in cardiac transplantation for up to 4 hours of storage.' New methods that will extend the safe period of storage have advantages, including more detailed tissue typing and matching of organs, an expansion of the donor pool by allowing more distant procurement, increased margin of safety in myocardial preservation, avoidance of the risks and logistic problems of rapid transport of organs, and reducing the inherent risks and economic costs of emergency surgery. The University of Wisconsin solution (UW) has been shown to be a superior preservation medium for kidney, pancreas, and liver transplantation.i" Its beneficial components include an intracellular type of electrolyte composition, two impermeable components-lactobionate and raffinose-and a colloid-pentastarch, all of which can minimize myocardial edema during storage. In addition, it contains adenosine, which may act like a potential high-energy phosphate precursor and a potent coronary vasodilator. Finally, allopurinol and glutathione are constituents that may reduce oxygen-derived free radical production during reperfusion. Survival studies with the heterotopic rat heart transplant model demonstrated that 12 to 18 hours of preservation with UW had yielded lower graft failure rates than the Stanford and Bretschneider solutions,": 8 Other small animal studies in which rat and rabbit isolated working heart models were used to measure cardiac outputs demonstrated that UW was superior to simple saline, St. Thomas' Hospital solution, Stanford solution, and Collins solution after 4 to 8 hours of storage. 9-IZ These results are encouraging; however, their methodology did not allow for the assessment of intrinsic myocardial contractility that may be derived from the measurement of the ventricular pressure-volume relationship. The current study involves an isolated adult canine heart preparation equiped with a computer-controlled mock arterial system that allows real-time measurements ofleft ventricular pressure-volume loops at different preset preloads and afterloads. The emphasis is on the analysis of left ventricular systolic and diastolic functions. To delineate the potential beneficial effects of the various components of UW, we also studied myocardial interstitial pH during storage, changes in myocardial water content, coronary hemodynamics, and myocardial oxygen consumption (MVO z). The effects of UW were compared with those of an extracellular crystalloid solution, Krebs-Ringer bicarbonate (KRB), for 6 and 12 hours of static cold storage.

Static Storage at 1- 2 Centigrade

I

Myocardial pH measurements

I

981

Analysis of LV Functions

I

I I

LV Biopsy

Normal Pressure Reperfusion LV Biopsy Coronary Flush with UW or KRB Arrest with Normosol + KCL

Low Pressure Reperfusion Flush with Normosol LV Biopsy

Fig. 1. Experimental protocol. LV, Left ventricular.

Methods All animals received humane care in compliance with the "Principles of Laboratory Animal Care" formulated by the National Societyfor Medical Research and the "Guide for the Care and Useof LaboratoryAnimals" prepared by the National Academy of Sciencesand published by the National Institutes of Health (NIH Publication No. 80-23, revised 1978). Thirty-fiveheartworm-free mongrel dogs (30 to 40 kg) were randomly divided into five groups (n = 7 in each group) according to the type of preservationsolution and the duration of preservation: control; KRB-6, Krebs-Ringer bicarbonate for 6 hours; KRB-12, Krebs-Ringer bicarbonate for 12 hours; UW-6, University of Wisconsin solution for 6 hours; and UW-12, Universityof Wisconsin solution for 12 hours. Preservation solutions. The compositions of the two solutions in the study are depicted in Table I. UW was prepared accordingto the protocolfrom the Universityof Wisconsinwith a slight modification; calcium wasadded to prevent the calcium paradox during reperfusion. All solutions were freshly mixed in the same week of the experiments. Operative procedures. All animals were kept from food and water starting on the night before the experiments. Acepromazine(0.1 rug/kg intramuscularly) wasgivenfor preoperative sedation. General anesthesia was induced by pentobarbital (25 to 30 mg/kg intravenously) and maintained by isoflurane(0.5% to 1.25%). The animals' lungs were ventilated by a mechanical respiratorthrough an endotrachealtube. Bodytemperature was maintained at 370 C by a recirculatingwater blanket. The electrocardiogram was continuously monitored in lead I or II. The experimental protocol was as depicted in Fig. 1. Large-bore femoral arterial and venous lines were placed through a groin cutdown. The mediastinum was exposed through a sternotomy incision. The venae cavae, pulmonary artery, brachiocephalic artery, left subclavian artery, and the distal aortic arch were isolated and surrounded by sutures or tapes; the azygos vein was tied off. The animal was then fully heparinized (400 units/kg intravenously). Approximately half of the animal's blood volume was removed slowlythrough the femoralarterial linewith simultaneouscrystalloidresuscitation. A modified arterial tubing was placed through the left subclavianartery for cardioplegic infusion.In a rapid sequentialorder, the brachiocephalic artery and the cavae were tied to empty the heart, followed by crossclamping of the aortic arch and infusion of 750 ml of electrolyte solution (Normosol with potassium chloride, 20 mfiq/L added, at4 0 C through the subclavian arterial cannula under 150 mm Hg external pressure. This

982

The Journal of Thoracic and Cardiovascular Surgery

Ko et al.

oxygenator heat exchanger

carbogen EKG - Electrocardiogram Cpp - Coronary perfusion pressure LVP - Left ventricular pressure

Fig. 2. Isolated heart preparation.

Table I. Composition of cardiac preservation solutions

uw

KRB

Electrolytes: (rnmol/L)

Na K

CL HC03 P04 S04

25 125 I 25

138 20 128 15 2.2

5

5 Mg 0.5 Ca Osmotic agents: gm/L (mmol/L) Glucose 17.8 (300) Raffinose 35.8 (100) Lactobionate Pentastarch 50 Miscellaneous Adenosine (rnmol/L) 5 Glutathione (mrnol/L) 3 1 Allopurinol (rnmol/L) Kefzol (mg/L) 300 320 Osmolarity pH (25 0 C) 7.4 ± 0.05

I

1.4 1.8 (10)

300 275

7.4 ± 0.05

method produced an aortic root pressure of 60 to 80 mm Hg during cardioplegic infusion. Cardioplegic solution was vented through a venotomy made in the superior vena cava while cold saline (4 0 C) was poured over the heart. Simultaneously, the

animal was fully exsanguinated through the femoral cannulas, and the blood was stored at 4 0 C for the later priming of the isolated heart circuit. The heart was then excised, weighed, and flushed with 200 ml of preservation solution through the subclavian arterial cannula. The heart was then placed in a plastic container filled with 750 ml of the same precooled preservation solution and placed in an ice chest. Preparations were made for the isolated ventricular studies during the end of the storage period. The aorta was secured with an umbilical tape just distal to the left subclavian artery. The venae cavae and the main pulmonary arterial trunk were ligated. A 16F right-angle cannula was placed through the right ventricular free wall for the collection of coronary venous flow during reperfusion. A wire-reinforced 16F cannula was placed through the brachiocephalic artery for retrograde coronary perfusion. The subclavian arterial cannula was used to monitor coronary perfusion pressure during reperfusion. The chordae tendineae were released and a custom-made plastic anulus was sutured to the mitral anulus for mounting the heart on the servo-pump system. At the end of the preservation period, the heart was placed in a pan of room temperature saline solution. The aorta and the cannulas were cleared of air, and the coronary arteries were flushed with 100 ml of Normosol solution (no potassium added) through the subclavian cannula. The brachiocephalic cannula was then connected to the arterial side of the isolated heart circuit. Reperfusion was started while the heart was still under water. Once aortic valve competence was ascertained, the heart was then mounted on the servo-pump system. The left ventricular balloon was secured in the ventricular cavity with its tip placed through the apical stab wound. A small cannula was also placed through the apex to vent the left ventricle. Coronary flow

Volume 103 Number 5

Heart preservation

May 1992

End Systolic PV Relationship: Slope = Elastance

983

120

Pressure

End Diastolic PV Relationship: 1/ Slope = Diastolic Compliance

Pressure (mmHg)

Volume

o

Volume (ee)

40

120

Stroke Work

UW-12 PRSWA Pressure 50cc

(mmHg)

End Diastolic Volume

Fig. 3. Analysis of left ventricular mechanics. Pt/, Pressurevolume.

ol------~--,-------

Volume (ee)

rate was adjusted to maintain reperfusion pressure at 40 mm Hg for the first 15 minutes. Subsequently, it was maintained between 65·and 85 mm Hg. Electrical defibrillation attempts were started once the perfusion pressure was increased at 15 minutes of reperfusion. Lidocaine (I 0 mg/L) was rarely given if electrical defibrillation was unsuccessful or if the severity of multifocal premature ventricular contractions prevented the ventricular function studies. Once defibrillated, the ventricles were paced at a rate of 120 beats/min via epicardial silver wires. The left ventricular balloon was fully collapsed during the first 30 minutes of reperfusion. The procedures for the control group were slightly modified to prevent any period of myocardial ischemia. The brachiocephalic cannula was placed in situ. Extracorporeal retrograde coronary perfusion was initiated through this cannula immediately after the great vessels and the aorta were tied and before the heart was excised from the animal. Isolated heart circuit. The circuit consists of a pediatric venous reservoir (Terumo Corp., Tokyo, Japan), a hollow-fiber membrane oxygenator with an integral heart exchanger (Terumo), on-line pH, carbon dioxide tension, and oxygen tension electrodes (Cardiomet, Shiley Inc., Irvine, Calif.), ultrasonic aortic inflow and coronary sinus outflow probes (Transonics Systems, Inc., Ithica, N.Y.), a 20 /Lm filter/bubble trap (Shiley), a 40 /Lm blood filter (Pall Corp., Biomedical Products Division, Glen Cove, N.Y.), and a roller pump (Masterflex, Cole-Parmer Instrument Co., Chicago, Ill.), as depicted in Fig. 2. These components are connected by polyvinylchoride tubings (% inch and \4 inch). The circuit was primed with 800 to 1000 ml of autologous heparinized blood (hemoglobin 8 to 12 gm/dl), warmed to 38° C, and oxygenated with carbogen (95% oxygen and 5% carbon dioxide). Sodium bicarbonate was added to correct to a blood pH of 7.35 to 7.45. Dextrose 50 (50 ml) and mannitol (12.5 gm) were also added. Blood gases and oxygen saturation of hemoglobin were measured by a blood pH analyzer and a Co-Oximeter device (Instrumentation Laboratory, Lexington, Mass.). Gas exchange with carbogen in this

40

Fig. 4. Representative samples of pressure-volume loops. set-up produced oxygen tension and carbon dioxide tension in the ranges of 550 to 650 mm Hg and 35 to 45 mm Hg, respectively. Left ventricular function studies. We used a mock circulatory system consisting of a servo-controlled volume pump to assess left ventricular function. The pump uses an electronically controlled piston of fixed cross-sectional area to alter the volume of the balloon placed in the left ventricular cavity. The left ventricular pressure measured by a micromanometer (Millar Instruments, Inc., Houston, Tex.) cannulated in the left ventricular balloon, is sampled in real-time by a personal computer at 250 Hz. During systole, the measured left ventricular pressure together with the preset aortic impendance are used to compute aortic root flow, and the computer then positions the piston (it is moved back to withdraw fluid within the balloon) so that the change in volume of the left ventricular cavity or balloon will correspond to the calculated flow in the sampling interval. During diastole, the piston is pushed forward by the computer according to the preset diastolic filling pressure until the isovolumetric contraction is sensed. Both diastolic filling pressure and total peripheral resistance are user selectable, which enables complete variation of ventricular filling (preload) and aortic input impedance (afterload). For each data collection point, a set of pressure-volume loops is obtained with varying ventricular diastolic filling pressures using a constant afterload. Data are recorded for IS seconds at 250 Hz for each diastolic filling pressure setting. The measured parameters that are recorded simultaneously on a separate computer with a custom data acquisition software at 250 Hz are the left ventricular pressure, left ventricular volume, aortic root or coronary perfusion pressure, retrograde aortic flow, and coronary sinus flow. Pressure-volume loops are generated and recorded at 30, 60, and 90 minutes of reperfusion. The left ventricular balloon is kept fully collapsed until simulation. Data were pooled and analyzed on separate occasions. Two

984

The Journal of Thoracic and Cardiovascular Surgery

Ko et al.

Table II. Left ventricular function studies ESPVR

SW-EDVR Slope (erg· cm! . j()3)

x-Intercept (ml)

Control UW-6 KRB-6 UW-12 KRB-12

9 11 10 12 II

± ± ± ± ±

53 56 32 47 4

1 2 2 2 3

± ± ± ± ±

PRSWA (erg. ml . jll)

4 4 9* 10 2t

55 57 31 36 3

± ± ± ± ±

8 9 9* 9 2t

x-Intercept (ml)

1 2 7 II II

± ± ± ± ±

3 2 2 3:1: 2:1:

EDPVR

Elastance (mm Hgfml]

4.2 4.4 4.4 5.4 5.3

± ± ± ± ±

0.6 0.3 0.7 1.0 1.0

Diastolic compliance (mlfmm Hg)

1.92 1.84 0.94 0.75 0.22

± ± ± ± ±

0.33 0.42 0.19* 0.19 0.03§

SW-EDVR, Stroke work-end-diastolic volume relationship; ESPVR, end-systolic pressure-volume relationship; EDPVR, end-diastolic pressure-volume relationship; PRSWA, preload recruitable stroke work area; UW, University of Wisconsin solution; KRB, Krebs-Ringer bicarbonate solution.

*p < 0.05 versus control

and UW -6.

tp < 0.005 versus control and UW-l2. :j:p < 0.05 versus control and UW -6. §p < 0.02 versus control and UW -12.

Table III. Percent myocardial water content

KRB 6 hr 12 hr UW

6 hr 12 hr

Baseline

End of preservation

End of reperfusion

78.0 ± 2.5 77.2 ± 2.2

78.1 ± 1.9 81.5 ± 1.5*

82.5 ± 0.9 80.0 ± l.l

75.7 ± 0.6 74.5 ± 2.2

75.8 ± 0.7 74.3 ± 1.8*

79.7 ± 1.2 80.5 ± 1.4

*p < 0.05, KRB-12 versus UW-12 by unpaired t test.

sets of pressure-volume loops were used for analysis for each time point. Elastance (Emax) and dead volume (Vo) were derived from the slope and the x-intercept of the left ventricular end-systolic pressure-volume relationship (ESPVR), respectively (Fig. 3). The left ventricular stroke work (SW) was calculated from the area of the pressure-volume loop for each measured end-diastolic volume (EDV). Linear regression of the left ventricular SW versus the EDV yielded the x-intercept (Vsw) and the slope (M sw) in the following equation: SW

= M sw • (EDV -

Vsw)

Preload recruitable stroke work area (PRSWA), defined as the area under this curve extrapolated to an end-diastolic volume of 50 ml (Fig. 3), was calculated based on this equation: PRSWA =

1/2

M sw ' (50 - Vsw)2

The left ventricular diastolic compliance was defined as the inverse of the slope of the left ventricular end-diastolic pressurevolume relationship (EDPVR) as illustrated in Fig. 3. Mean coronary resistance was calculated by average coronary pressure divided by average coronary flow. Myocardial interstitial pH. Interstitial pH as an index of tissue ischemia was measured by the Khuri pH monitor (Vascular Technology, Inc., Chelmsford, Mass.) during the preservation period. A glass pH microelectrode and a plunge needle temperature probe were placed into the interventricular septum after harvest until the end of the storage period. The temperature and the pH automatically corrected to the measured temperature were plotted continuously on a chart recorder. Reported values were read from specific time points for comparison.

Myocardial water content. Transmural biopsy specimens (200 to 400 gm) were taken from the left ventricular apex immediately after harvest, at the end of the preservation period, and at the end ofthe 2 hours of reperfusion. Dry weightswere obtained after the specimens were dried to constant weights at 60° C. Percent water content was calculated as follows: % H 20 = ([Wet weight - Dry weightj/Wet weight) X 100

Creatine kinase release. Blood samples were collected from the isolated heart circuit right before the heart was mounted and from the coronary sinus cannula after 5 minutes of reperfusion. Serum creatine kinase was measured by an enzymatic method dependent on nicotinamide-adenine dinucleotide, reduced (Fisher Scientific Co., Pittsburgh, Pa.). MV0 2• Aortic inflow and coronary sinus blood oxygen tension, hemoglobin, and oxygen saturation of hemoglobin were measured by a pH/blood gas analyzer and a Co-Oximeter device (Instrumentation Laboratory). MV0 2was calculated as follows: MV0 2 = ([Ca02 - Cv021 X Coronary flow)/Heart weight 02 content = (1.38 X Hgb X O 2 saturation)

+ (Po 2 X 0.003)

Ca02 and CV02 denote arterial oxygen content and venous oxygen content, respectively. These were measured and calculated for 5, 30, and 60 minutes of reperfusion. Statistical methods. All numerical values are expressed as means ± standard error of the mean. Unpaired t test was used for comparisons between groups. Two-way analysis of variance was used for the analysis of the pH data. A p value lessthan 0.05 was considered to be statistically significant. Results Heart weights. The heart sizes were similar in all five groups. The wet weights of the control, UW -6, UW -12, KRB-6, and KRB-12 groups were 231 ± 15, 259 ± 13, 217 ± 10, 256 ± 14, and 235 ± 10 gm, respectively. The measurements included the four chambers of the heart and the great vessels used for the preparation. Left ventricular function studies. The data of the left ventricular function studies at 60 minutes of reperfusion are presented. Elastance or Emax calculated as the slope

Volume 103

Heart preservation

Number 5 May 1992

985

3000

:ll:

a::

0

0" 0

:ll:

-"

a::

~

== W

0

ICIl

2000

--tr--e--

--.--

0

!:!'

-···-0····-_···-c··-··

1000

CONT UW-6 UW-12 KRB-6 KRB-12

_ _.. - --2

...................10

20

30

40

50

60

LEFT VENTRICULAR END·DIASTOLIC VOLUME (CC)

Fig. 5. Left ventricular SW-EDVR. 100

w a::

;:)

CIl CIl W

80

a::

D.

0

Ci ::J :I:

0

l-

CIl

< Ci

60

E

S. 40

····c···

Q

····0···

Z

W

>

20

--.--

--e---tr-

..I

0 0

10

20

KRB-12 KRB-6 UW-12 UW-6

control

30

DIASTOLIC VOLUME (CCl

Fig. 6. Left ventricular EDPVR.

of the ESPVR did not show any significant changes among the five groups (Table II). The x-intercepts or Vo of the ESPVR significantly shifted to the right on the x-axis for the preserved groups as compared with the control group; this shift was more pronounced for the 12-hour groups than for the 6-hour groups. The analysis of elastance was not useful in comparing the levels of ventricular performance among groups of hearts with varying periods of global ischemia, partly because of the rightward shifts of the x-intercepts. This problem is illustrated in the representative samples in Fig. 4. Although the slopes of the ESPVR of the two sets of pressure-volume loops are similar, the x-intercept of the UW-12 group is shifted more to the right and its slope of the EDPVR

is steeper than that of the UW-6 group. Furthermore, the sizes of the pressure-volume loops or SWs in the UW-12 group are much smaller than those of the UW-6 group. The levels of left ventricular performance were well demonstrated in the left ventricular SW-EDV relationship (SW-EDVR), which takes into account of the areas of the pressure-volume loops. The slope of this relationship determines the ability of the left ventricle to increase its output of kinetic energy as diastolic filling is increased. Glower and his associates! 3 have shown this to be a linear relationship over a wide range of EDVs, and this was confirmed in our study. The? values of the linear regressions for the control, UW-6, UW-12, KRB-6, and KRB12 groups were 0.95 ± 0.02,0.97 ± 0.01,0.94 ± 0.05,

986

The Journal of Thoracic and Cardiovascular Surgery

Ko et al.

8.0

-

UN

KR8

---0-

7.5 J:

a.

...J


C a:

7.0


o

o

>

:E

6.5

6.0

+--O-,.-......,--.-T-r-r-"r""T~T""T-,-T"""T-.-........-r-................,,...,....,

o

2

3

4

5

6

7

8

9

10

11

12

13

TIME (hr)

Fig. 7. MyocardialpH during preservation. 2500

2000

:::J

1500

ll:

1000

o

500

Baseline

Reperfusion

Fig. 8. Creatine kinase (CPK) release during reperfusion. = 0.025 KRB-6 versus UW-6. tp = 0.007 KRB-12 versus UW-12. *p

0.96 ± 0.02, and 0.87 ± 0.06, respectively. The same group also introduced the PRSWA as the most reliable index of intrinsic myocardial contractility; this is calculated as the area under the SW-EDVR curve extrapolated to a predefined EDV, which was 50 ml in this study. The x-intercepts of the SW-ED VR were similar in all five groups of hearts (Table 11). The slope of this curve remained normal after 6 hours of preservation in the UW groups as compared with the control group and was slightly lower after 12 hours, not reaching statistical significance (Fig. 5). In contrast, there was significant deterioration in the slope of the SW-EDVR after 6 hours of preservation in KRB as compared with the control or

UW-6 groups. The results after 12 hours of preservation in KRB were poor. Identical trends were demonstrated in the calculated PRSWA among the fivegroups (Table 11). The diastolic compliance (Table II) was calculated as the inverse of the slope of the EDPVR (Fig. 6). The diastolic compliance was well preserved in the normal range after 6 hours of storage in UW. Unlike the systolic function, as demonstrated by the PRSWA, however, the diastolic compliance was significantly reduced after 12 hours of storage in UW as compared with control values. The diastolic compliance after 6 hours of storage in KRB was significantly less than in the control and the UW-6 groups; it was unacceptably poor after 12 hours of storage in KRB. Percent water content. There was no significant water gain after 6 or 12 hours of preservation in UW. Six hours of storage in KRB also did not produce any water gain; however, the hearts stored in KRB for 12 hours had significantly higher water content than did the UW-12 group (Table III). This difference was not maintained at the end of the 2 hours of reperfusion. Myocardial interstitial pH. On-line myocardial pH measurements were used as an index of local lactic acidosis or tissue ischemia during the storage period. The initial pH was in the neighborhood of 7.60 for both of the preservation media; the relative alkalosis was due to the effect of hypothermia (Fig. 7). The most dramatic reduction of pH occurred within the first hour of storage, with a reduction of 0.59 and 0.54 in the UW and KRB groups, respectively. Thereafter, the pH declined more gradually. Although the average pH of the UW hearts was consistently higher than that of the KRB hearts after the first hour of storage, there was no statistical difference between the UW and KRB groups at any of the time points. Release of creatine kinase. Serum creatine kinase levels were significantly elevated at 5 minutes of reperfusion for all groups (p < 0.05) as compared with the baseline samples collected from the isolated heart circuit immediately before the hearts were mounted. The amount of creatine kinase released in the KRB-6 group was 2.79-fold higher than that released in the UW-6 group; it was 3.98-fold higher in the KRB-12 groups than in the UW-12 group (Fig. 8). Coronary resistance. The resting coronary resistance of the nonischemic ex vivo hearts in the control group after 60 minutes of extracorporeal perfusion was 27.5 ± 1.3 mm Hg . sec/ml. The coronary resistance of all the preserved groups at 5 minutes of reperfusion was about twofold lower than that of the control group (Fig. 9). It then increased toward the normal range at 60 min-

Volume 103 Number 5 May 1992

Heart preservation

987

w

0

z

Cl: I-

C/)

U

enw U" II:

J:

z

§.

0

II:

I

•"! '" E

>

II: Cl:

20

10

KRB-6 KRB-12 UW-6 UW-12

0 0

5 min

30 min

60 min

REPERFUSION

Fig. 9. Coronary resistance. utes of reperfusion. There was no statistical difference between the UW and the KRB groups. MVOz. The MVO z was 4.11 ± 0.53 ml/rnin - 100 gm for the nonischemic control hearts. The UW-6 group had immediate recovery of this level of oxygen use during reperfusion (Fig. 10). The KRB-6 group had significantly lower level of MV02 than that of the UW-6 group during the first 30 minutes of reperfusion; it then increased to the level of the UW-6 group at 60 minutes of reperfusion. The levels of MVO z were lower for both the UW-12 and KRB-12 groups than for the UW-6 and control groups during the entire reperfusion period. The MV0 2 measured at 5 minutes of reperfusion represented the unloaded fibrillating state of the hearts in these experiments (Fig. 10). It is a state that requires less than 40% of the oxygen requirement of the fully loaded working state. This factor needs to be included in comparing the MVO z at 5 minutes of reperfusion with the subsequent time points when the hearts were in the working state. The oxygen use at 5 minutes of reperfusion was, therefore, more than twofold higher than the subsequent time points considering the two different states of ventricular activities. The remarkably elevated amount of oxygen use during the very early phase of reperfusion would be expected from prolonged ischemia. The MVOz measured in the subsequent time points (30 and 60 minutes of reperfusion) represented minimally loaded but working hearts in a regularly paced rhythm. These blood gas samples were collected during the first and the lowest loading condition of each set of pressure-volume loop simulation. The same loading condition was used to allow for comparisons among the different groups. The lower than expected values of MV02 in the control group and

during the 30 and 60 minutes of reperfusion for the study groups were due to the minimum loading conditions when the measurements were made.

Discussion The slope of the ESPVR or elastance (Emax) was introduced by Sagawa'" and Suga and Sagawa'" as an index of myocardial contractility. Its advantages include the linearity of the relationship within the physiologic range of ventricular volumes and its independence from preload and afterload conditions, as opposed to other indexes such as rate of pressure rise, ejection fraction, or cardiac output. Although it is a good index of systolic elasticity, it may not represent the ventricular performance in its entirety. The total energy output of the ventricle is limited by the area between the ESPVR and the EDPVR. 16 Therefore, the ventricular performance may be reduced by a reduction in the slope of the ESPVR, or an elevation in the slope of the EDPVR, or both. This study demonstrated that extended global ischemia had a more profoundly deleterious effect on the EDPVR than the ESPVR. Although a decrease in the systolic elasticity was reflected in the rightward shifts in the x-intercept of the ESPVR, the slope or elastance calculated from the ESPVR was unchanged among the five groups of hearts. There was an obvious subjective difference in the levels of contractile activities between the KRB-6 and UW-6 groups, and there was clear and dramatic deterioration in the levels of contractile activities of the 12-hour groups based on visual inspection by the investigators of this study. This wide range of contractile function among the five groups of hearts was not reflected in the calculated elastance from the ESPVR. A recent report closely

988

The Journal of Thoracic and Cardiovascular Surgery

Ko et al.

7

*

6

E Cl

5

0 0

,...

c 'E 0

~

4

D

KRB-6 UW-6

D

KRB·12 UW-12

~

3

N

0

> ==

2

0

5 min

60 min

30 min

REPERFUSION 7 6

E Cl

5

0 0

,...

C

4

E 0

~

E::I

3

N

0

>

2

==

5 min

30 min

60 min

REPERFUSION

Fig. 10. MV02. *p = 0.006, UW-6 versus KRB-6.

confirmed our findings in that the elastance (Emax) did not reflect left ventricular contractile dysfunction (was unchanged) when there was significant decrease in the diastolic compliance induced experimentally.l? Glower and his associates" have demonstrated the unreliability of using elastance to compare the normal state and the ischemic state of the ventricle. They further described the superiority of the SW-EDVR and its derived slope and PRSWA. 13, 18 The left ventricular SW is equal to the area of each of the corresponding pressure-volume loops, which is limited by the ESPVR and EDPVR. The analysis of the SW-EDVR in essence takes both the ESPVR and EDPVR into account and truly represents the total ventricular performance.

The left ventricular studies described here clearly demonstrated that UW was able to maintain normal levels of systolic function in terms of the SW-EDVR and diastolic function in terms of compliance after 6 hours of static hypothermic storage. After 12 hours of storage in UW, there was a significant deterioration of diastolic compliance, while the systolic function was only slightly depressed and not reaching statistical significance.More important, preservation with UW was far superior to that with KRB after 6 and 12 hours of storage in this model. Progressive myocardial edema has been shown to be a major problem in extended periods of hypothermic organ preservation. 19 The accumulation of interstitial fluid may be attributed to the disparity between the oncotic pres-

Volume 103 Number 5 May 1992

sures in the traditional crystalloid type of preservation solutions and the interstitial space. Furthermore, intracellular edema occurs because of an influx of sodium and chloride ions caused by a lossof the sodium-potassium ion pump function during ischemic and hypothermic storage. 20,21 This influx of ions further increases the passive diffusion of water into the cytoplasm and intracellular organelles. Interstitial edema may impede capillary flow and decrease ventricular wall compliance.F and intracellular edema may distort intracellular cytoskeleton and impair mitochondrial and other cellular functions.P This study demonstrated that UW prevented any fluid accumulation in the myocardium up to 12 hours of hypothermic preservation. UW's intracellular composition of sodium and potassium greatly reduced the electrochemical gradients that may promote fluid shifts in the extracellular type of solutions such as KRB. The balance of negative ions across the cytoplasmic membrane in homeostasis is largely based on the negatively charged intracellular proteins and the extracellular chloride ion. The loss of energy-dependent membrane ion carriers during storage allows for the passive diffusion of the chloride ions into the cells. The replacement of the chloride anion by the large lactobionate anion, which is impermeable to cellular membranes, prevents this problem in the case of UW.24 Furthermore, the additional osmotic pressure generated by the other impermeable component, raffinose, and the oncotic pressure exerted by the pentastarch prevented any edema formation, as shown in this study. This benefit was substantiated by studies of kidney, pancreas, and liver slices by other investigators.P: 26 This advantage may explain in part the superiority of UW over KRB after 12 hours of preservation. However, there was no difference in the water content between the two preservation media after 6 hours of storage, despite differences in the ventricular function studies. The level of postischemic recovery closely correlated with the amount of adenosine triphosphate remaining in the myocardium at the end of the preservation period ranging from 2 to 12 hours of hypothermic static storage in rat hearts." A theoretical advantage of UW over simple crystalloid solutions such as KRB has been attributed to the ability of UW's adenosine and phosphate to act as precursors for adenosine triphosphate synthesis during the storage period. The availability of these precursors, therefore, may reduce the amount of energy deprivation and anaerobic metabolism during storage. However, the levelsof tissue acidosis or ischemia as measured by tissue pH were similar for the two solutions. Myocardial pH measured by the method used in this study has been shown to be a reliable indicator of the severity of ischem-

Heart preservation

989

ic damage during anoxic arrest under normothermic and hypothermic conditions." It is likely that these metabolic precursors need to be actively and continually perfused into the coronary circulation to be used by the endothelium and myocytes. There is much evidence that tissue acidosis causes cellular damage by inducing lysosomal instability, by activating lysosomal enzymes, and by impairing mitochondrial electron transport.i? A remarkable amount of acidosis developed in all the preserved hearts during the storage period in this experimental protocol despite the presence of the buffering capacity of both of these solutions. Hypothermia reduces but does not eliminate cellular metabolism. Therefore anaerobic metabolism seems unavoidable shortly after cardiac arrest, as evidenced by the pH data presented here. Aims to prevent the development of myocardial acidosis necessitate some type of coronary perfusion with buffer agents during the storage period, be it intermittent or continuous. Adenosine is the most potent coronary vasodilator available." Therefore storage of hearts in UW with adenosine may facilitate coronary flow during reperfusion. This study showed that the coronary resistance of the ischemic hearts was much reduced during the early phase of reperfusion and it normalized at 60 minutes of reperfusion. There was no statistical difference in the coronary resistance between the two solutions at any time points of the reperfusion phase. It appears that the length of ischemia in this study already induced maximal coronary vasodilatation; therefore adenosine did not exert any advantage in this regard. The amount of myocardial damage was reflected in the release of creatine kinase from the coronary sinus during reperfusion. Significant amounts of creatine kinase were released from all the preserved hearts, as expected. However, hearts preserved in KRB had several-fold higher levels of creatine kinase released than did the hearts preserved in UW. This differential amount of myocardial damage could not be explained by the levels of ischemia experienced between the two solutions during the storage period; they were similar, as indicated by the tissue pH data. Indeed, UW could not be expected to be any more protective against ischemia and its consequential acidosis than KRB, since neither group was perfused or oxygenated. However, the pretreatment with allopurinol and glutathione contained in UW may have reduced the amount of myocardial reperfusion injury caused by oxygen-derived free radical formation during reflow and reoxygenation. There is certainly an enormous amount of literature supporting the efficacy of oxygen free radical scavenging in reducing myocardial reperfusion injury.31-33 Furthermore, adenosine recently has been

The Journal of Thoracic and Cardiovascular Surgery

990 Ko et al.

shown to reduce myocardial reperfusion injury. Two recent reports demonstrated the ability of intravenous and intracoronary adenosine to reduce reperfusion-related myocardial infarcts and to improve regional ventricular function in canine models of regional ischemia-reperfusion.34, 35 The biochemical mechanism of this benefit will require further investigation. The level of ventricular functional recovery depends largely on the cellular integrity of the energy production apparatus-the glycolytic pathway and mitochondrial oxidative phosphorylation-since 80% of oxygen use is devoted to the contractile activities in the heart.P In a sheep model of reversiblewarm global ischemia, Furukawa and his associates'? showed a significant increase in left ventricular MV02 during recovery. In the present study of extended hypothermic ischemia, the level of oxygen use was well maintained throughout the reperfusion period in the UW-6 hearts compared with control values;it was not found to be elevated, as expected.On the contrary, the MV02 was transiently depressed for the KRB-6 group as compared with the control and UW-6 groups; it was depressed for both preservation media after 12hours of storage. This suggeststhat either hypothermia or the prolonged period of ischemia ameliorated the ability of the cells to increase their oxygen use during recovery. The level of MV0 2 reflects the amount of chemical bond energy generated from available substrates; the chemical bond energy in the forms of adenosine triphosphate and nicotinamide-adenine dinucleotide, reduced, are converted into kinetic energy during myocardial fiber shortening. Therefore the depressed levels of MV02 would in part account for the depressed ventricular functions in the KRB-6 and the 12-hour groups. The elucidation of the mechanisms responsible for this depressed oxygen use after prolonged hypothermic arrest in the absence of myocardial necrosis will be fruitful if these mechanisms are reversible or preventable. The biochemical basis for this uncoupling of energy requirement for recovery and oxidative phosphorylation (aerobic metabolism) after extended hypothermic storage requires much basic scientificinvestigation. The possible mechanisms may include a lossof nucleotide metabolites for production of adenosine triphosphate and nicotinamide-adenine dinucleotide, reduced, a loss of adenine nucleotide translocase activity in the mitochondrial membrane, an impairment of the proton gradient across the mitochondrial membrane, a loss of cytosolic calcium, or an accumulation mitochondrial calcium, all of which are affected by hypothermic and ischemic storage. Although the tissue pH data did not show any advantage of adenosine in UW in terms of the degree of ischemia

during storage, the exposure of the cells to a bath ofadenosine may have prevented the loss of intracellular adenosine, a crucial salvageable metabolite for adenosine triphosphate production that is needed during recovery. This may in fact explain the improved oxygen use of the UW-6 hearts compared with KRB-6 hearts during the initial phase of reperfusion. Indeed, supplementation of exogenous adenosine or augmentation of intracellular adenosine by inhibiting adenosine deaminase via cardioplegic solutions has been shown to improve postischemic ventricular function in canine and rabbit hearts. 37, 38 Finally, adenosine added in cardioplegic solution induced more rapid cardiac arrest than when potassium was used alone.l" 40 This may provide additional myocardial protection, although it was not tested in thisstudy. In summary, the ventricular function studies of this report unequivocallydemonstrated the superiorityofUW over the simple extracellular type of crystalloids such as KRB in maintaining left ventricular systolicand diastolic functions after 6 and 12 hours of static hypothermic storage. UW was also found to be superior in the prevention of edema formation during storage, the prevention of reperfusion injury during reflow and reoxygenation, and the maintenance of a normal level of oxygen use during the early phase of reperfusion. On the basis of UW deserves serious consideration in clinical cardiac transplantation. REFERENCES I. Baumgartner WA. Evaluation and management of the heart donor. Baumgartner WA, Reitz RA, Aschuff Sc. Heart and heart-lung transplantation. Philadelphia: WB Saunders, 1990:86-102. 2. Ploeg RJ, Goossens D,McAnulty JF, Southard JH, Belzer FO. Successful 72-hour cold storage of dog kidneys with UW solution. Transplantation 1988;46: 191-6. 3. Wahlberg JA, Love R, Landegaard L, Southard JH, Belzer FO. 72 Hour preservation of canine pancreas. Transplantation 1987;43:5-8. 4. Jamieson NV,Sundberg R, Lindell S, et al.Preservation of the canine liver for 24-48 hours using simple cold storage with UW solution. Transplantation 1988;46:517-22. 5. Liu T, Walsh TR, Nalesnik M, Makowka L. Improved preservation of the rate liver fororthotopic liver transplantation: useof University of Wisconsin-Iactobionate solution and retrograde reflushing. Surgery 1990;I08:890-7. 6. Ploeg RJ. Preliminary results of the European multicenter study on UW solution in liver transplantation. Transplant Proc 1990;22:2185-8.

7. Okouchi Y,Shimizu K, Yamaguchi A, Kamada N. Effectiveness of modified University of Wisconsin solution for heart preservation as assessed in heterotopic rat heart

Volume 103 Number 5 May 1992

transplant model. J THORAC CARDIOVASC SURG 1990; 99:1104-8. 8. Makowka L, Zerbe TR, Chapman F, et al. Prolonged rat cardiac preservation with UW lactobionate solution. Transplant Proc 1989;21:1350-2. 9. Yeh T,Hanan SA, Johnson DE, et al. Superior myocardial preservation with modified UW solution after prolonged ischemia in the rat heart. Ann Thorac Surg 1990;49:932-9. 10. Maurer EJ, Swanson DK, DeBoer LWV. Comparison of UW and Collins solution for preservation of the rat heart. Transplant Proc 1990;22:548-50. 11. Wicomb WN, Collins GM, Wood J, Hill JD. Improved cardioplegia using new perfusates. Transplant Proc 1989; 21:1357-8. 12. Ledingham SJM, Katayama 0, Lachno DR, Yacoub M. Prolonged cardiac preservation: evaluation of the University of Wisconsin preservation solution by comparison with the St. Thomas' Hospital cardioplegic solutions in the rat. Circulation 1990;82:IV351-8. 13. Glower DD, Spratt JA, Snow ND, et al. Linearity of the Frank-Starling relationship in the intact heart: the concept of preload recruitable stroke work. Circulation 1985; 71:994-1009. 14. Sagawa K. The end-systolic pressure-volume relation of the ventricle: definition, modification, and clinical use. Circulation 1981;63:1223-7. 15. Suga H, Sagawa K. Instantaneous P-V relationships and their ratio in the excised, supported canine left ventricle. Circ Res 1974;35:117-26. 16. Sagawa K, Maughan L, Suga H, Sunagawa K. Energetics of the heart. In: Sagawa H, Maughan L, Suga H, Sunagawa K, eds. Cardiac contraction and the pressure-volume relationship. New York: Oxford University Press, 1988: 171-231. 17. Zile MR, Izzi G, Gaasch WHo Left ventricular diastolic dysfunction limits use of maximum systolic elastance as an index of contractile function. Circulation 1991;83:674-80. 18. Glower DD, Spratt JA, Kabas S, Davis JW, Rankin JS. Quantification of regional myocardial dysfunction after acute ischemic injury. Am J Physiol 1988;255:H85-93. 19. Hearse DJ, Braimbridge MV, Jynge P. Protection of the ischemic myocardium: cardioplegia. New York: Raven Press, 1981:167-208. 20. MacKnight ADC, Leaf A. Regulation of cellular volume. Physiol Rev 1977;57:510-6. 21. Pegg DE. The principles of organ storage procedures. In: Pegg DE, Jacobsen lA, Halasz NA, eds. Organ preservation: basic and applied aspects. Lancaster, England: MTP Press, 1982;55-66. 22. Bethencourt DM, Laks H. Importance of edema and compliance changes during 24 hours of preservation of the dog heart. J THORAC CARDIOVASC SURG 1981;81:440-9. 23. Feinberg H. Energetics and mitochondria. In: Pegg DE, Jacobsen lA, Halasz NA, eds. Organ preservation: basic and applied aspects. Lancaster, England: MTP Press, 1982;3-16.

Heart preservation

99 1

24. Jamieson NV, Lindell S, Sundberg R, Southard JH, Belzer FO. An analysis of the components in UW solution using the isolated perfused rabbit liver. Transplantation 1988; 46:512-6. 25. Wahlberg JA, Southard JH, Belzer FO. Development of a cold storage solution for pancreas preservation. Cryobiology 1986;23:477-82. 26. Ar'Rajab A, Sundberg R, Ahren B. The functional effects of a colloid in liver cold storage preservation. Transplant Proc 1990;22:2191-3. 27. Connery CP, Hicks GL, Wang T. Positive correlation of functional recovery and tissue ATP levels in the hypothermically stored cardiac explant. Surg Forum 1990;41:2824. 28. Lange R, Kloner RA, Zierler M, Carlson N, Seiler M, Khuri SF. Time course of ischemic alterations during normothermic and hypothermic arrest and its reflection by on-line monitoring of tissue pH. J THORAC CARDIOVASC SURG 1983;86:418-34. 29. Belzer FO, Southard JH. Principles of solid-organ preservation by cold storage. Transplantation 1988;45:673-6. 30. Christensen CW, Rosen LB, Gal RA, Haseeb M, Lassar TA, Port SC. Coronary vasodilator reserve: comparison of the effects of papaverine and adenosine on coronary flow, ventricular function, and myocardial metabolism. Circulation 1991;83:294-303. 31. Bando K, Tago M, Teramoto S. Prevention of free radicalinduced myocardial injury by allopurinol: experimental study in cardiac preservation and transplantation. J THORAC CARDIOVASC SURG 1988;95:465-73. 32. Gharagozloo F, Melendez FJ, Hein RA, Shemin RJ, DiSesa VJ, Cohn LH. The effect of superoxide dismutase and catalase on the extended preservation of the ex vivo heart for transplantation. J THORAC CARDIOVASC SURG 1988; 95:1008-13. 33. Stewart JR, Gerhardt EB, Wehr CJ, et al. Free radical scavengers and myocardial preservation during transplantation. Ann Thorac Surg 1986;42:390-3. 34. Pitarys CJ II, Virmani R, Vildibill HD Jr, Jackson EK, Forman MB. Reduction of myocardial reperfusion injury by intravenous adenosine administered during the early reperfusion period. Circulation 1991;83:237-47. 35. Homeister JW, Hoff PT, Fletcher DD, Lucchesi BR. Combined adenosine and lidocaine administration limits myocardial reperfusion injury. Circulation 1990;82:595608. 36. Furukawa S, Bavaria JE, Kreiner G, Edmunds LH. Relationship between total mechanical energy and oxygen consumption in the stunned myocardium. Ann Thorac Surg 1990;49:543-9. 37. Wyatt DA, Ely SW, Lasley RD, et al. Purine-enriched asanguineous cardioplegia retards adenosine triphosphate degradation during ischemia and improves postischemic ventricular function. J THORAC CARDIOVASC SURG 1989; 97:771-8. 38. Bolling SF, Bies LE, Bove EL, Gallagher KP. Augmenting

992

Ko et al.

intracellular adenosine improves myocardial recovery. J THORAC CARDIOVASC SURG 1990;99:469-74. 39. de Jong JW, van der Meer P, van Loon H, Owen P, Opie LH. Adenosine as adjunct to potassium cardioplegia: effect on function, energy metabolism, and electrophysiology. J THORAC CARDIOVASC SURG 1990;100:44554. 40. Schubert T, Vetter H, Owen P, Reichart B, Opie LH. Adenosine cardioplegia. Adenosine versus potassium cardioplegia: effects on cardiac arrest and postischemic recovery in the isolated rat heart. J THORAC CARDIOVASC SURG 1989;98: I057-65.

Discussion Dr. Jack G. Copeland (Tucson, Ariz.). On the basis of what the authors have reported, I am convinced that UW merits use in human transplantation, and I believe studies are now underway in some centers. However, this paper evokes several memories and provokes several questions. Twenty years ago, using continuous low-flow perfusion with oxygenated modified buffered Krebs solution, which was the extracellular solution in this study, Dr. Stinson and I preserved hearts in vitro for 24 to 48 hours, then orthotopically transplanted them, and had survivors. The question that this brings up is this: How valid do you think this type of study is for clinical transplantation? Should you next evaluate an actual transplantation model? Dr. Ko. I agree that we need to be careful in projecting data of in vitro studies such as ours to the clinical situation. However, the preparation used in this study allowed us to study intrinsic myocardial functions under strictly controlled conditions that are not possible in intact animals or human beings. Long-

The Journal of Thoracic and Cardiovascular Surgery

term animal studies will be required to validate the long-term benefits of UW in extended heart preservation. In regard to your comment about continuous perfusion for preservation, we attempted this in our laboratory and found substantial edema formation by 24 hours. We were not able to achieve any meaningful results after 24 hours of preservation with continuous coronary perfusion. Dr. Copeland. I think your experiment is highly sophisticated but on the other hand perhaps treacherous. I wonder if there is a learning curve in performing the experiment, Did you randomize your animals in this experiment and, if so, how? Did you do all the control studies first and then all of the KRB studies and then all of the UW studies? If that is so, then I would expect you to get better with time. Dr. Ko. That is a valid and insightful question regarding the methodology in this type of study. The preparation used in this study was developed in collaboration with the investigators in Dr. Sagawa's laboratory at Johns Hopkins University. Dr. Sagawa and his associates were the first to describe this method of measuring ventricular contractility and had contributed much to this field. Approximately 40 pilot experiments were performed to ensure reproducibility of this preparation in our laboratory. In regard to the randomization process for this study, the control group was done first. The animals were randomized either into the KRB group or into the UW group. The 6-hour experiments were performed before the 12-hour experiments. The randomization was obviously not blinded, but we alternated between KRB and UW groups. Dr. Copeland. UW is the intuitive product of some very wise minds in Wisconsin, but we know it isjust a "best guess" recipe. Are there any ingredients that should be changed? Dr. Ko. After completing this study, we began the second arm of the project. I have sequentially and individually deleted the five important constituents in UW that I have shown earlier. The data are currently being analyzed.