Enhanced Preservation of Pig Cardiac Allografts by Combining Erythropoietin With Glyceryl Trinitrate and Zoniporide

Authors

  • A. J. Watson,

    1. Transplant Program, The Victor Chang Cardiac Research Institute, Sydney, NSW, Australia
    2. Heart & Lung Transplant Unit, and Department of Physiology and Pharmacology, University of New South Wales, Randwick, NSW, Australia
    3. Department of Cardiothoracic Surgery, University of New South Wales, Randwick, NSW, Australia
    Search for more papers by this author
  • L. Gao,

    1. Transplant Program, The Victor Chang Cardiac Research Institute, Sydney, NSW, Australia
    Search for more papers by this author
  • L. Sun,

    1. Transplant Program, The Victor Chang Cardiac Research Institute, Sydney, NSW, Australia
    Search for more papers by this author
  • J. Tsun,

    1. Transplant Program, The Victor Chang Cardiac Research Institute, Sydney, NSW, Australia
    Search for more papers by this author
  • A. Doyle,

    1. Transplant Program, The Victor Chang Cardiac Research Institute, Sydney, NSW, Australia
    Search for more papers by this author
  • S. C. Faddy,

    1. Department of Cardiology, St Vincent's Hospital, Darlihghurst NSW 2010, University of New South Wales, Randwick, NSW, Australia
    Search for more papers by this author
  • A. Jabbour,

    1. Transplant Program, The Victor Chang Cardiac Research Institute, Sydney, NSW, Australia
    2. Heart & Lung Transplant Unit, and Department of Physiology and Pharmacology, University of New South Wales, Randwick, NSW, Australia
    3. Department of Clinical Pharmacology, University of New South Wales, Randwick, NSW, Australia
    Search for more papers by this author
  • Y. Orr,

    1. Heart & Lung Transplant Unit, and Department of Physiology and Pharmacology, University of New South Wales, Randwick, NSW, Australia
    2. Department of Cardiothoracic Surgery, University of New South Wales, Randwick, NSW, Australia
    Search for more papers by this author
  • K. Dhital,

    1. Heart & Lung Transplant Unit, and Department of Physiology and Pharmacology, University of New South Wales, Randwick, NSW, Australia
    2. Department of Cardiothoracic Surgery, University of New South Wales, Randwick, NSW, Australia
    Search for more papers by this author
  • M. Hicks,

    1. Transplant Program, The Victor Chang Cardiac Research Institute, Sydney, NSW, Australia
    2. Department of Clinical Pharmacology, University of New South Wales, Randwick, NSW, Australia
    3. Department of Physiology and Pharmacology, University of New South Wales, Randwick, NSW, Australia
    Search for more papers by this author
  • P. C. Jansz,

    1. Heart & Lung Transplant Unit, and Department of Physiology and Pharmacology, University of New South Wales, Randwick, NSW, Australia
    2. Department of Cardiothoracic Surgery, University of New South Wales, Randwick, NSW, Australia
    Search for more papers by this author
  • P. S. Macdonald

    Corresponding author
    1. Transplant Program, The Victor Chang Cardiac Research Institute, Sydney, NSW, Australia
    2. Heart & Lung Transplant Unit, and Department of Physiology and Pharmacology, University of New South Wales, Randwick, NSW, Australia
    3. Department of Cardiology, St Vincent's Hospital, Darlihghurst NSW 2010, University of New South Wales, Randwick, NSW, Australia
    Search for more papers by this author

Abstract

Erythropoietin has a tissue-protective effect independent of its erythropoietic effect that may be enhanced by combining it with the nitric oxide donor glyceryl trinitrate (GTN) and the sodium–hydrogen exchange inhibitor zoniporide in rat hearts stored with an extracellular-based preservation solution (EBPS). We thus sought to test this combination of agents in a porcine model of orthotopic heart transplantation incorporating donor brain death and total ischaemic time of approximately 260 min. Pig hearts were stored in one of four storage solutions: unmodified EBPS (CON), EBPS supplemented with GTN and zoniporide (GZ), EBPS supplemented with erythropoietin and zoniporide (EZ), or EBPS supplemented with all three agents (EGZ). A total of 4/5 EGZ hearts were successfully weaned from cardiopulmonary bypass compared with only 2/5 GZ hearts, 0/5 CON hearts and 0/5 EG hearts (p = 0.017). Following weaning from bypass EGZ hearts demonstrated superior contractility and haemodynamics than GZ hearts. All weaned hearts displayed impaired diastolic function. Release of troponin I from EGZ hearts was lower than all other groups. In conclusion, supplementation of EBPS with erythropoietin, glyceryl trinitrate and zoniporide provided superior donor heart preservation than all other strategies tested.

Abbreviations
CON

control

EPBS

extracellular-based preservation solution

EGZ

erythropoietin glyceryl trinitrate and zoniporide

EZ

erythropoietin and zoniporide

GTN

glyceryl trinitrate

GZ

glyceryl trinitrate and zoniporide

Introduction

Primary allograft failure remains the leading cause of early mortality following heart transplantation, accounting for 36% of perioperative deaths in the most recent International Society for Heart & Lung Transplantation report [1]. Its multifactorial pathogenesis involves the complex interaction of the neurohormonal and inflammatory sequelae of brain death, ischaemia-reperfusion injury (IRI) and recipient factors [2]. Although most of these factors are non-modifiable, the effects of ischemia may be ameliorated by improved preservation strategies. Hypothermic cardioplegic arrest remains by far the most commonly employed means of donor heart preservation. Existing cardioplegic solutions may be modified by the addition of chemical adjuncts known to protect against IRI (“pharmacologic perconditioning”). This approach is a potentially simple and inexpensive means of improving donor heart preservation and thereby reducing early mortality.

We have previously demonstrated that myocardial protection afforded by Celsior, a commercially available cardioplegic solution, can be extended in both rodent and porcine models by the addition of glyceryl trinitrate (GTN) and the sodium-hydrogen exchange (NHE) inhibitor cariporide [3, 4]. Cariporide was withdrawn from clinical development following the publication of EXPEDITION trial. This trial demonstrated a reduction in perioperative myocardial infarction at the expense of an excess stroke risk in patients undergoing coronary artery bypass grafting and who were administered perioperative cariporide [5]. We have subsequently shown that zoniporide, an NHE inhibitor with greater selectivity for the NHE-1 isoform predominant in cardiomyocytes, confers equivalent cardioprotection in rats at a concentration tenfold less than cariporide [6].

Over the last decade several authors have demonstrated the cardioprotective efficacy of erythropoietin in a variety of rodent models of myocardial injury [7-10]. We subsequently evaluated erythropoietin as a supplement to Celsior in an ex vivo rodent model of prolonged, global cold ischaemia, demonstrating synergistic beneficial effects when combined with GTN and zoniporide [11]. The aim of the present study was thus to conduct further preclinical evaluation of this combination of pharmacological agents using a porcine model of orthotopic heart transplantation.

Methods

A porcine model of orthotopic heart transplantation incorporating donor brain death was used as previously described [4, 12]. All procedures were approved by the institutional animal ethics committee (Animal Research Authority #08/17) and complied with the Guidelines To Promote The Wellbeing Of Animals Used For Scientific Purposes (National Health & Medical Research Council, Australia) and the Guide For The Care And Use Of Laboratory Animals (National Institutes of Health, Bethesda, MD).

Experimental groups

Twenty orthotopic heart transplants were performed in four groups (each n = 5) that differed only with respect to the cardioplegic solution employed. The surgeons were blinded to treatment allocation and the randomization code was revealed only after collection of all data. In the control group (CON), unmodified Celsior was used for donor heart arrest and storage. Donor hearts in the remaining groups were arrested using Celsior modified by the addition of varying combinations of: recombinant human erythropoietin (5 U/mL; Eprex, Janssen-Cillag), GTN (100 mg/L; Hospira Australia Pty Ltd) and zoniporide (1 µM; provided by Pfizer under an independent external investigator agreement). The combinations tested were: GTN and zoniporide (GZ); erythropoietin and GTN (EG); and erythropoietin, GTN and zoniporide (EGZ).

Animals and anaesthesia

Forty juvenile Landrace pigs (40–60 kg) were obtained in pairs, the larger being used as the donor animal. Animals were premedicated with intramuscular ketamine (10 mg/kg), midazolam (1 mg/kg) and atropine (50 µg/kg) prior to tracheal intubation. General anesthesia was maintained with isofluorane (1–4%) and intermittent IV fentanyl (100–200 µg). Normal saline was infused to maintain central venous pressure (CVP) 0–5 mmHg. Continuous physiological monitoring included electrocardiogram, mean arterial blood pressure (MAP), CVP, pulse oximetry, end-tidal CO2 and core temperature. Arterial blood gases and blood glucose levels were analyzed hourly. Intravenous lignocaine (1 mg/kg) was administered prior to sternotomy for prophylaxis of ventricular arrhythmias. Ventricular fibrillation was cardioverted using internal DC countershock (10–20 J).

Donor surgical procedures

A right fronto-parietal burr-hole was drilled and the tip of a Foley catheter placed in the subdural space. The lumen of the catheter was primed with saline and connected to a pressure transducer to monitor intracranial pressure. Median sternotomy was then performed and a micromanometer-tipped catheter (Millar Instruments, Houston, TX) was placed in the left ventricle via the right carotid artery. Two pairs of ultrasonic dimension transducers (Sonometrics Corporation, Ontario, Canada) were sutured to the epicardium to measure the base-apex long axis and anterior–posterior short axis dimensions of the left ventricle [12, 13].

After acquisition of baseline data, brain death was induced by injecting 24 mL saline into the Foley catheter balloon in 3 mL increments every 30 s. This produced a reproducible “autonomic storm” characterized by extreme tachycardia and hypertension, followed by mild hypotension. Anaesthesia was terminated 15 min after induction of brain death, which was confirmed by: 1) typical haemodynamic changes as described, 2) intracranial pressure in excess of mean arterial pressure and 3) absence of brainstem reflexes following cessation of anaesthesia.

One hour after brain death intravenous heparin (15 000 IU) was administered, then the ascending aorta was cross-clamped and 900 mL chilled Celsior (Genzyme, Naarden, The Netherlands) with or without supplements was administered, titrating the rate of infusion to achieve an aortic root pressure of 60 mmHg. The heart was then excised and stored in 100 mL of the same Celsior used to arrest the heart in a sealed plastic bag, which was placed in a second bag containing chilled saline, then packed in ice slurry in an insulated plastic container.

Recipient procedures

The recipient animal was anesthetized as described above. Methyl prednisolone 500 mg was administered on induction of anesthesia and just prior to myocardial reperfusion. A median sternotomy was performed, heparin 300 IU/kg administered then the pig placed on aorto-bicaval cardiopulmonary bypass (CPB). A further 10 000 IU heparin was added to the pump prime and the animal actively cooled to 32°C. Orthotopic transplantation of the donor heart was performed using the biatrial method of Lower and Shumway [14] with a cold storage time of approximately 210 min and warm ischaemic time of approximately 50 min. Following completion of the left atrial anastomosis an LV vent was placed to prevent distension and 300 mL unmodified Celsior was administered via the aortic root. Rewarming was commenced whilst performing the aortic anastomosis. After removal of the aortic cross-clamp the heart was defibrillated if necessary. After 45 min of reperfusion, a dobutamine infusion was commenced at 10 µg/kg/min. In addition, a noradrenaline infusion was initiated if needed to maintain a mean arterial pressure 40–50 mmHg, and the heart was paced (VVI 120 bpm).

One hour after reperfusion an attempt was made to wean the animal from CPB. Successful weaning was defined as the ability to maintain a self-sustaining aortic root pressure greater than 40 mmHg. If this attempt was unsuccessful the recipient animals was placed back on bypass and two further attempts at weaning were made after 2 and 3 h of reperfusion, the latter after dobutamine was increased to 20 µg/kg/min. Once off bypass the animal was monitored for a further 3 h with haemodynamic data acquired hourly. During this time the rate of dobutamine infusion was kept constant; noradrenaline was gradually weaned if possible. After data acquisition at 3 h noradrenaline infusion was ceased. Data acquisition was repeated 15 min later, then the experiment was terminated.

Outcome measures

The primary outcome measure of the study was the ability to wean from CPB and maintain cardiac output for the duration of the study. Myocardial contractile function was assessed using the stroke work index (SWI), derived from the preload recruitable stroke work (PRSW) relationship using left ventricular pressure–volume loops as previously described in detail [13, 15]. Myocardial diastolic function was assessed using the slope of the end-diastolic pressure volume relationship (EDPVR) from the same pressure volume loops. Other haemodynamic parameters including heart rate, MAP, CVP and cardiac output (derived from pressure–volume loop data) were recorded. Data were collected in the donor just prior to induction of brain death and prior to cardioplegic arrest, and in the recipient at hourly intervals after weaning from CPB. Cardiac troponin I released into blood was assessed in the donor before surgery, immediately prior to induction of brain death and cardioplegic arrest, then in all recipients at 15 min, 1, 2 and 3 h following reperfusion.

Statistics

Statistical analysis was performed using SPSS Statistics 20 (SPSS Inc, Chicago, IL). Data were expressed as mean ± standard error of the mean; differences between groups were compared by one-way analysis of variance (ANOVA) followed by Fisher's LSD post-test. Categorical variables were reported as a percentage and compared using Fisher's exact test for a 2 × 4 contingency table. Regression coefficients were reported as mean ± standard error of the mean. Derived regression estimates (e.g. SWI and nVw) were reported as absolute values based on the mean value of the regression coefficient. Two-tailed p values of less than 0.05 were considered statistically significant.

Results

Selected baseline characteristics are presented in Table 1. Donor and recipient animals were well matched across groups with respect to gender and body mass. Discrepancy in body mass for animal pairs was no greater than 15% of the recipient animal mass. There were no significant differences between groups with respect to duration of brain death or myocardial ischaemic times (see Table 2).

Table 1. Baseline characteristics of the control and treatment groups
CharacteristicCONGZEGEGZp value
  1. Fisher's exact test performed for categorical variables. One-way ANOVA performed for continuous variables, values expressed as mean ± standard error of the mean. D:R, donor: recipient ratio.
Gender (% male)
Donor40%80%60%80%0.742
Recipient80%80%60%80%1.000
Weight (kg)
Donor51.4 ± 3.751.4 ± 2.552.4 ± 0.951.8 ± 1.20.989
Recipient50.0 ± 3.548.2 ± 2.449.6 ± 0.548.6 ± 0.90.929
D:R weight ratio1.03 ± 0.041.07 ± 0.011.06 ± 0.021.07 ± 0.030.783
Table 2. Ischaemic times and weaning from cardiopulmonary bypass
CharacteristicCONGZEGEGZp value
  1. Fisher's exact test performed for categorical variables. One-way ANOVA performed for continuous variables, values expressed as mean ± standard error of the mean. CPB, cardiopulmonary bypass. Reperfusion, interval between release of cross-clamp and separation from CPB. A single animal in the CON group initially weaned from CPB after 65 min; however, the heart subsequently succumbed to delayed graft failure, thus not considered a successful wean.
Times (min)
Brain death72.2 ± 2.678.2 ± 3.978.8 ± 1.277.6 ± 1.30.256
Warm ischemia53.0 ± 8.156.2 ± 2.652.8 ± 3.749.0 ± 2.20.767
Cold ischemia206.8 ± 7.4227.6 ± 8.0211.6 ± 2.4212.4 ± 6.00.151
Total ischemia259.8 ± 7.7283.8 ± 9.8264.4 ± 6.0261.4 ± 6.90.143
Reperfusion6577.0 ± 2.077.5 ± 8.80.772
Weaned from CPB (%)0%40%0%80%0.017

Weaning from cardiopulmonary bypass

One of five animals in the CON group was weaned from CPB; however, the heart fibrillated after 90 min and was unable to be resuscitated despite adrenaline boluses and defibrillation. It was placed back on CPB support and regained an organized rhythm, however, could not be weaned again despite a further 90 min of myocardial rest. Thus, none of the 5 CON group, nor any of the EG animals, completed the reperfusion protocol. In contrast, 2 of 5 GZ animals and 4 of 5 EGZ animals were weaned from CPB (p = 0.017, Fisher's exact test). All of these animals completed the experimental protocol, with a trend for improved haemodynamics over time. Reperfusion time prior to separation from CPB was similar among those animals that were successfully weaned (Table 2).

All animals that were weaned from CPB did so at the first attempt and remained on dobutamine at 10 µg/kg/min for the duration of the experiment. The dose of noradrenaline required varied among animals but tended to decrease with time from a mean of 0.051 µg/kg/min at 1 h postweaning to 0.037 µg/kg/min at 3 h. Noradrenaline was then withdrawn altogether for the final data acquisition at 3.25 h to standardize the degree of hemodynamic support provided.

Myocardial contractility, diastolic function and hemodynamics pre- and posttransplant

Posttransplant contractility, diastolic function and haemodynamic data were available for a total of only six animals (two from the GZ group and four from the EGZ group), precluding meaningful statistical comparison of groups. A number of generalizations can be made, however, regarding the posttransplant data that are in keeping with the proportion of animals weaned from CPB in these two groups. All measures of contractile and hemodynamic performance were similar between groups at pre-brain death baseline.

Representative pressure–volume loops from GZ and EGZ animals are presented in Figure 1 and demonstrate a decrease in stroke work (area within the loop) at 3.25 h postwean from CPB in the GZ animal compared with the EGZ animal. They also demonstrate a decrease in end-diastolic volume (x-axis) in both animals posttransplantation, likely representing a degree of diastolic dysfunction. The corresponding PRSW relationships in these animals (Figure 2) show a decrease in contractility (slope of the curve) compared with baseline in the GZ animal, in contrast to an increase in contractility in the EGZ animal. Results of multiple linear regression (MLR) analysis of pooled PRSW relationships are presented in Table 3. At each time-point following transplantation the group mean stroke work index (SWI) was superior for EGZ animals compared with GZ animals.

Figure 1.

Left ventricular pressure–volume (PV) loops Representative PV loops obtained during transient inferior vena caval occlusion taken at pre-brain death baseline (upper panels) and 3.25 h after weaning form CPB (lower panels). An example is shown from an animal in the GZ group (left panels) and the EGZ group (right panels). Note the marked decrease in PV loop area (representing stroke work) in the GZ animal following transplantation, compared with both the baseline loop and the EGZ animal at the corresponding time point. Also note the decrease in end-diastolic volume (x-axis) in both hearts posttransplantation.

Figure 2.

Preload recruitable stroke work (PRSW) relationships End-diastolic volume (x-axis) and stroke work (y-axis) plotted from the PV loops demonstrated in Figure 1. Data normalised to the baseline steady state values (i.e. prior to vena caval occlusion) obtained prior to brain death. Note the increase in contractility (slope of the curve; i.e. stroke work greater for any given end-diastolic volume) in the EGZ heart at 3.25 h following weaning from CPB, contrasted with the decrease in the GZ heart. Note again the decrease in posttransplant end-diastolic volume in both hearts.

Table 3. Preload recruitable stroke work (PRSW) relationship; MLR analysis of pooled group data
Time pointSlope (nMw)y-Axis intercept (nSW)x-Axis intercept (nVw)Stroke work index
  1. y-axis, normalized stroke work; x-axis, normalized epicardial end-diastolic volume. Regression coefficients (nMw and y-axis intercept) are group mean ± standard error of the normalized PRSW; nVw and stroke work index are calculated from the mean value of these regression coefficients. For donor values n = 5 each group; for posttransplant values n = 2 for GZ, n = 4 for EGZ.
Donor
Baseline (CON)2.534 ± 0.049−1.516 ± 0.0420.5981.018
Baseline (GZ)2.881 ± 0.046−1.841 ± 0.0400.6391.040
Baseline (EG)2.642 ± 0.055−1.630 ± 0.0460.6171.012
Baseline (EGZ)2.672 ± 0.029−1.648 ± 0.0250.6171.024
Posttransplant
1 h (GZ)1.681 ± 0.161−1.141 ± 0.1190.6790.540
1 h (EGZ)3.350 ± 0.258−2.267 ± 0.2000.6771.083
2 h (GZ)2.176 ± 0.169−1.448 ± 0.1220.6650.728
2 h (EGZ)5.243 ± 0.160−3.606 ± 0.1180.6881.637
3 h (GZ)2.952 ± 0.124−1.983 ± 0.0910.6720.969
3 h (EGZ)5.326 ± 0.173−3.715 ± 0.1310.6981.611
3.25 h (GZ)2.759 ± 0.151−1.874 ± 0.1110.6790.885
3.25 h (EGZ)4.239 ± 0.200−2.949 ± 0.1530.6961.290

Left ventricular minute work (LVMW = stroke work × heart rate) for the different groups is shown in Figure 3. LVMW did not differ between groups at baseline or after 1 h of brain death, immediately prior to cardioplegic arrest of the donor heart. Posttransplant LVMW was similar to the pre-brain death baseline value for EGZ hearts that were weaned from CPB. In contrast, in the GZ group mean LVMW was less than half the baseline value at all posttransplant time points.

Figure 3.

Left ventricular minute work (LVMW) at baseline and after 1 h of brain death, and at 1, 2, 3 and 3.25 h after weaning the transplanted heart from CPB Order of histograms from left to right is: CON (diagonal bars, n = 5 and n = 0 in donor and recipient respectively); GZ (grey bars, n = 5 and n = 2); EG (checked bars, n = 5 and n = 0); EGZ (black bars, n = 5 and n = 4). There were no significant differences between groups in the donor at baseline or after 1 h of brain death. Posttransplant LVMW was similar to the baseline value in EGZ hearts that could be weaned from CPB, however tended to be less than half the baseline value in GZ hearts.

EDPVR slopes for the different groups are shown in Figure 4. EDPVR slopes did not differ between groups either at baseline or after 1 h of brain death. Posttransplant, EDPVR slope increased two- to threefold in the six hearts that could be weaned from CPB. The slope remained markedly elevated in the GZ group but tended to fall in the EGZ group over the first 3 h after weaning.

Figure 4.

Left ventricular end-diastolic pressure–volume relationships (EDPVR) at baseline and after 1 h of brain death, and at 1, 2, 3 and 3.25 h after weaning the transplanted heart from CPB Order of histograms and posttransplant n as for Figure 3. Left ventricular end-diastolic pressure (y-axis) was plotted against end-diastolic volume (x-axis) during transient vena caval occlusion and the slope of the relationship calculated. Note the increase in the slope of the EDPVR following transplantation (indicating an increase in pressure change for a given change in volume, thus a loss of ventricular compliance). This loss of compliance appeared to be attenuated in EGZ hearts compared with GZ hearts 3.25 h following weaning from CPB.

Cardiac output (CO) data are presented in Figure 5. Donor CO did not differ between groups either at baseline or after 1 h of brain death. CO was similar to the baseline value at all posttransplant time points in EGZ hearts that were weaned from CPB. In contrast, mean CO in the GZ group tended to increase with time following weaning, reaching only 50% of the baseline value at 3.25 h.

Figure 5.

Cardiac output (CO) at baseline and after 1 h of brain death, and at 1, 2, 3 and 3.25 h after weaning the transplanted heart from CPB Order of histograms and posttransplant n as for Figure 3. There were no significant differences between groups in the donor at baseline or after 1 h of brain death. Posttransplant CO was similar to the baseline value in EGZ hearts that could be weaned from CPB, however tended to be roughly 50% of the baseline value in GZ hearts.

Mean arterial blood pressure (MAP) data are presented in Figure 6. MAP did not differ between groups either at baseline or after 1 h of brain death. MAP was similar to the baseline value at all posttransplant time points in EGZ animals that were weaned from CPB. In contrast, MAP in the GZ groups was less than the baseline value, and less than MAP in the EGZ group at all posttransplant time points.

Figure 6.

Mean arterial blood pressure (MAP) at baseline and after 1 h of brain death, and at 1, 2, 3 and 3.25 h after weaning the transplanted heart from CPB Order of histograms and posttransplant n as for Figure 3. There were no significant differences between groups in the donor at baseline or after 1 h of brain death. Posttransplant MAP was similar to the baseline value in EGZ hearts that could be weaned from CPB. Posttransplant MAP improved with time in GZ hearts however remained depressed compared with baseline and with EGZ at all time points.

Troponin release pre- and posttransplant

Cardiac troponin I release was measured in donors before commencement of surgery (baseline) then 1 h after induction of brain death then in recipients at 15 min, 1, 2 and 3 h postreperfusion of the heart, regardless of whether or not the animal was maintained on CPB at the time (Figure 7). Thus, a complete data set for troponin release was obtained for each group (i.e. n= 5 each group). Plasma troponin levels progressively increased following reperfusion in all groups (repeated measures ANOVA, p < 0.001), with the lowest values being recorded in the EGZ group at all time points. Significant differences between groups began to emerge following 1 h of reperfusion, with troponin levels in the EGZ group being lower than GZ at 1 h (60.3 ± 7.9 vs. 95.5 ± 8.3 µg/L; p = 0.004), 2 h (128.8 ± 7.9 vs. 251.4 ± 34.5; p = 0.033), and 3 h (206.2 ± 27.6 vs. 392.5 ± 53.7; p = 0.035). EGZ was also lower than EG at 3 h (206.2 ± 27.6 vs. 489.5 ± 96.5; p = 0.005), and GZ higher than CON at 1 h (95.5 ± 8.3 vs. 72.6 ± 8.2; p = 0.045).

Figure 7.

Plasma troponin I levels at baseline and after 1 h of brain death, and at 0.25, 1, 2 and 3 h following reperfusion of the transplanted heart Plasma samples were obtained for all animals pre- and posttransplant. Animals that could not be weaned from bypass were maintained on bypass support to allow blood sampling to be completed for a total of 3 h. Troponin I rose in all groups over the reperfusion period, with significant differences at 1, 2 and 3 h postreperfusion (see text for details of post hoc tests). Histograms appear in same order as Figure 3, n = 5 in all groups at all time points.

Discussion

We have previously demonstrated, using both rodent and porcine models, that cardioprotection afforded by Celsior may be enhanced by the addition of the nitric oxide donor GTN and the NHE inhibitor cariporide [3, 4]. The key finding of the current study was that the cardioprotective effect of providing exogenous nitric oxide combined with NHE inhibition (in this case by zoniporide) may be extended further by the addition of erythropoietin. The dosage and timing of administration of all three agents was based on our previous studies in isolated rat hearts [3, 6, 11] and in pigs [4]. In an earlier study [11], erythropoietin supplementation of Celsior significantly improved functional recovery of isolated rat hearts subjected to 6 h of cold ischaemia followed by 45 min reperfusion. Erythropoietin alone failed to protect against I-R injury when the storage time was extended to 10 h. At this more prolonged storage time either erythropoietin or zoniporide used in combination with GTN provided a modest protective effect; however, robust protection against IRI was afforded by the combination of all three agents.

The primary outcome measure in the present study was the ability to wean the transplanted heart from CPB. Not only were EGZ hearts more likely to wean from CPB, but they also tended to perform better than GZ hearts in all measures of contractile function and haemodynamics assessed. In contrast, no CON or EG hearts were able to support the recipient animal's circulation following transplantation.

Analysis of posttransplant pressure–volume loop data provides some interesting insights. Contractility, assessed using SWI, was increased by 30% compared with baseline in the EGZ group at 3.25 h postweaning. This increase is attributable to the use of dobutamine; however, the contractile response to exogenous catecholamine was blunted in these hearts; we have previously demonstrated that 10 µg/kg/min dobutamine results in a doubling of SWI in the normal porcine heart [13]. In contrast to the increase in SWI seen in EGZ hearts there was a 12% decrease in GZ hearts at 3.25 h despite the use of dobutamine at the same dose. Furthermore, hearts in both groups were operating at lower left ventricular end-diastolic volumes and had marked elevations in the slope of the EDPVR following transplantation compared with baseline reflecting reduced left ventricular compliance. As a result, cardiac output was approximately equivalent to the baseline value in EGZ hearts despite the dobutamine-mediated increase in contractility. Thus, these data provide evidence of both systolic and diastolic dysfunction in the transplanted hearts, the former being significantly attenuated in hearts receiving all three pharmacologic adjuncts.

Troponin release was measured as a biochemical marker of cardiomyocyte injury. These biochemical data corroborated the contractile and hemodynamic data described above, insofar as plasma troponin was lower at all posttransplant time points in hearts from the EGZ group compared with hearts from the GZ group. Surprisingly, troponin values in the CON and EG groups were somewhat discordant, not being significantly higher than the GZ group (values in the CON group in fact being lower). Nonetheless, serum troponin values at all posttransplant time points were lower in the EGZ group than in each of the other three groups tested.

The cardioprotective actions of erythropoietin have been extensively tested in small (and to a lesser extent large) animal models. The majority of these studies have employed models of coronary ligation, with or without reperfusion, simulating acute coronary syndromes [8, 16-21]. To our knowledge this is the first study examining erythropoietin's cardioprotective effects in the context of cardiac transplantation. Cardiac allograft preservation is an attractive clinical application for pharmacologic conditioning strategies as onset of both ischaemia and reperfusion may be anticipated, and the administration of pharmacologic agents may be timed accordingly. The mechanism of erythropoietin's cardioprotective effect remains a matter of some debate. It is best known for its role in erythropoiesis, whereby it acts on cell surface receptors on bone marrow erythroid precursor cells to activate the intracellular PI3K-Akt and JAK-STAT signaling cascades. It is thought that these signalling molecules act on downstream elements to rescue the erythroblast from apoptosis, thus allowing it to differentiate and ultimately to form mature erythrocytes [22].

It has been known for some time that the signalling elements Akt and ERK1/2 are activated by cardioprotective stimuli such as ischaemic preconditioning. These agents, their upstream precursors, and their downstream targets have been collectively named the Reperfusion Injury Salvage Kinase (RISK) pathway [23], and have been extensively studied in models of erythropoietin-mediated cardioprotection. There has not been universal consensus regarding the importance of RISK pathway elements, with several investigators reporting significant increases in Akt activity [9, 17, 20, 21, 24-26] while others have not [11, 27]. ERK activity has received less attention; however, there has again been similar disagreement in the literature regarding its importance in erythropoietin-mediated cardioprotection, with some authors reporting an increase in activity [17, 21, 28, 29] while others do not [11, 27].

The importance of the JAK2-STAT3 signaling pathway in mediating ischemic pre- and postconditioning has been appreciated more recently, earning the epithet Survivor Activating Factor Enhancement (SAFE) pathway [30, 31]. Recently, mitochondrial STAT3 phosphorylation has been observed in postconditioning-mediated protection in pigs with regional cardiac ischemia/reperfusion [32]. The SAFE pathway has also been investigated in erythropoietin-mediated cardioprotection with multiple authors reporting an activation of STAT3 signalling during normothermic and hypothermic myocardial ischemia/reperfusion [11, 20, 21, 27, 29].

Nitric oxide (NO) homeostasis is perturbed in ischemia-reperfusion and activation of endothelial nitric oxide synthase (eNOS) may in fact contribute to IRI. Under such conditions eNOS is activated by intracellular calcium influx, and following an initial burst the production of NO declines as both substrate (L-arginine) and cofactor (tetrahydrobiopterin) are consumed [33-35]. Activated eNOS then produces progressively increasing amounts of the reactive oxygen species superoxide (O2), particularly upon reperfusion [34, 35], a process referred to as “uncoupling” of eNOS [36, 37]. Provision of exogenous NO to ischaemic tissues has been shown to abrogate IRI in a variety of experimental scenarios. NO has numerous beneficial effects mediated by cGMP, including vasodilatation, inhibition of platelet aggregation and leucocyte adhesion [36]. In addition, exogenous NO has been shown to have cGMP-independent effects, possibly mediated by opening mitochondrial ATP-sensitive K+-channels [3, 38]. Exogenous NO also exerts a negative-feedback effect on eNOS (product-inhibition) thus diminishing eNOS-derived superoxide production [35, 39]. Interestingly, exogenous erythropoietin has also been shown to suppress eNOS activity and nitrosative stress in rat hearts subjected to IRI [29].

The sodium-hydrogen exchanger (NHE) is another potential therapeutic target. Activated in conditions of ischemia-reperfusion, it acts to maintain intracellular pH by extruding H+ in exchange for Na+. Coupled with inactivity of the Na+–K+ ATPase, this results in intracellular Na+ accumulation secondarily activating the Na+–Ca2+ antiporter with a consequent influx of Ca2+ [40]. Pharmacological NHE inhibition has been shown to protect against myocardial IRI in a variety of animal models, using agents such as amiloride [41, 42], cariporide [43-45], and more recently zoniporide [6, 46, 47]. These agents have been thought to act primarily by attenuating intracellular Ca2+ overload. We have recently shown however that zoniporide supplementation of Celsior in isolated rat hearts subjected to prolonged ischaemia results in an increase in phosphorylation of both ERK and STAT3 [6].

We have employed the current model in an attempt to mimic as closely as possible the conditions at play during human heart transplantation. We have previously demonstrated the protective efficacy of combination treatment with erythropoietin, GTN and zoniporide in isolated rat hearts subjected to prolonged hypothermic ischemia [11]. Certain important elements were missing in this screening model that we have incorporated in the current study, namely: 1) donor brain death, 2) orthotopic transplantation into a separate recipient animal and 3) reperfusion in whole blood. Prior studies from our group have utilized inbred Westran pigs. Due to unavailability of this strain we were obliged instead to use outbred Landrace pigs. We have previously successfully transplanted Westran hearts arrested with Celsior supplemented with cariporide and GTN after a period of 6 h donor brain death and 14 h ischaemia [4]. After a series of preliminary experiments it became apparent that the Landrace heart would not tolerate such extremes of ischemia using an identical preservation strategy (unpublished data). We thus modified our protocol to restrict ischemic time to a more clinically relevant 4½ h. Even at this ischemic time only a minority of hearts could be weaned from bypass. It would appear that hearts from different pig strains vary markedly in their susceptibility to IRI with Landrace hearts being more susceptible to IRI than human hearts. Nonetheless, we feel this model provided a rigorous test of the cardioprotective efficacy of EGZ supplemented Celsior and that our observations of superior myocardial performance in this group of animals remain valid.

In summary, we have demonstrated that the cardioprotective potential of erythropoietin may be exploited in a combined cardioplegic strategy incorporating NHE inhibition and nitric oxide donation, in a large animal model of orthotopic heart transplantation utilising grafts from brain dead donors. We believe it is likely that such a combined modality approach will be required if pharmacologic preconditioning strategies are to be successful in reducing the incidence and mortality of early graft failure in human heart transplantation.

Acknowledgments

The authors gratefully acknowledge the generous support of St Vincent's Hospital's clinical perfusion service; Andrew Dinale, Jonathan Cropper and Claudio Soto, led by Dr Frank Junius. Without their technical and material support this study would not have been possible.

Funding source: This project was given by an Australian National Health & Medical Research Council Program Grant (N° 573732). Dr Watson was supp orted by a Francis & Phyllis Thornell Shore Memorial Scholarship and a Sir Roy McCaughey Surgical Research Fellowship provided by the Royal Australasian College of Surgeons.

Disclosure

The authors of the manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.

Ancillary