HIGHLIGHTS
- After ex vivo machine perfusion of porcine donation after circulatory death hearts with normothermic blood but not with the hypothermic histidine-tryptophane-ketoglutarate-N solution, left-ventricular pressure increment and decrement are impaired during early reperfusion.
- An increased left-ventricular preload can reverse this effect.
INTRODUCTION
The organ shortage is a limiting factor for heart transplantation (HTX). Thus, many transplant centers have started to accept hearts donated after circulatory death (DCD) [1]. Unlike donation after brain death, in DCD donors, the life-sustaining therapy of a potential DCD donor is withdrawn. As a consequence of the subsequent breathing arrest, the whole body of the donor becomes ischemic [2]. Due to global ischemia, the heart cannot maintain circulation in the body and circulatory death occurs [3]. According to local regulations, a specific period has to elapse before organ procurement. Therefore, DCD hearts are exposed to warm ischemia, known as warm ischemic time (WIT), before harvesting [4]. DCD hearts are mainly maintained by ex vivo machine perfusion (EVMP) with normothermic donor blood [1]. Nevertheless, we demonstrated in a porcine DCD model that EVMP with the hypothermic, crystalloid Histidine-Tryptophane-Ketoglutarate-N solution instead of blood is superior to preserving left-ventricular contractility, relaxation and coronary microcirculation during early reperfusion [5]. However, what remains unknown is whether left-ventricular contractility and relaxation change by reperfusion time after EVMP of DCD hearts.
MATERIAL AND METHODS
Animals and Anesthesia
The animals received humane care. We sedated healthy pigs (40-50 kg bodyweight) with intramuscular injection of Ketamine (22.5 mg/kg; Bremer Pharma, Warburg, Germany) and Midazolam (0.375 mg/kg; Hameln pharma plus, Hameln, Germany). Anesthesia was maintained intravenously with Pentobarbital-Sodium (15 mg/kg/h; Boehringer Ingelheim Vetmedica, Ingelheim, Germany).
Donation after Circulatory Death model and organ procurement
According to a previously published model, we induced circulatory death by the termination of mechanical ventilation [5]. Within the subsequent period of 30 min, we collected blood and harvested the heart. After a total WIT of 30 min, we flushed the DCD-hearts with 2 l of cold (4°C) Custodiol solution (Köhler Chemie GmbH, Bensheim, Germany), followed by mounting on the perfusion system (fig. 1).
Study groups and machine perfusion
We maintained DCD hearts either by EVMP with normothermic blood of the organ donor (BP group, n=6) or 4 °C, oxygenated HTK-N (HTK-N group, n=6; Köhler Chemie GmbH, Bensheim, Germany) for 4 hours in Langendorff fashion through the ascending aorta (fig. 2). The perfusion system consisted of an open venous reservoir, a hollow fiber membrane oxygenator with an integrated heat exchanger (Affinity Pixy, Medtronic, Minneapolis, USA) and 1/4” x 3/32” tubing. During reperfusion in both groups and during 4 h of EVMP in the BP group, the system was primed with 800 mL of non-diluted donor blood, added with 5000 i.E. Heparin, 250 mg Sodium-Prednisolone, and 25 mL Mannitol 15%. Hearts were perfused with a perfusion pressure of 50 mmHg in the BP group based on experimental pre-trials, and only 20 mmHg in the HTK-N group, concerning the increased risk of edema formation during crystalloid perfusion. Arterial and venous blood gas, including lactate, were measured every 30 min with a point-of-care analyzer (RAPID Point 500, Siemens) to maintain pH of 7.35-7.45, paO2 of 200 mmHg and paCO2 of 40-50 mmHg. If necessary, hearts were paced at 80 beats per minute in BP and during reperfusion. During reperfusion, perfusion pressure was adjusted to 60 mmHg.
Functional assessment during reperfusion
After maintenance perfusion, we reperfused the hearts of both groups for 120 min with fresh blood to mimic transplantation. During reperfusion, we evaluated left-ventricular contractility and relaxation using a balloon catheter inserted through the mitral valve. To determine a potential time-dependent effect of reperfusion, we measured the end-systolic pressure (ESP), maximal pressure increment (dp/dtmax) and maximal pressure decrement (dp/dtmin) after 30, 60 and 120 min of reperfusion at 10 and 20 mL of left-ventricular filling volume (LVV). We also analyzed the relative time-dependent change of contractility markers using the measurement after 30 min as a baseline value, thereby creating relative ESP, dp/dtmax, and dp/dtmin.
Statistical analysis was performed using IBM SPSS Statistics for Windows (Version 25.0, IBM Corp. Armonk, NY, USA). Results are expressed as mean ± standard error. We tested for homogeneity of variances by the Levene test. Data were analyzed using a one-way analysis of variance for multiple comparisons with Tukey adjustment in case of variance homogeneity and Games-Howell adjustment in case of variance inhomogeneity. For analysis of relative contractility, we applied a two-tailed unpaired classical t-test in case of variance homogeneity and a Welch t-test in case of variance inhomogeneity. We compared the data using 95% confidence intervals (CI) of the difference in means (DIM) and p-values. A p-value < 0.05 was considered statistically significant. Relative contractile parameters were calculated by normalizing contractility at 60 and 120 min or reperfusion to 30 min of reperfusion. Therefore, they allow a comparison of the shift of ESP, dp/dtmax, and dp/dtmin by time.
RESULTS
Left-ventricular contractility and relaxation: Effects of reperfusion time and left-ventricular filling
Over time, the end-systolic pressure (ESP) remained constant at 10 mL of LVV in the BP and HTK-N group (fig. 3A). At an LVV of 20 mL, the ESP improved slightly between 30 and 60 min (155±9 vs. 176±7 mmHg; 95%CI DIM: -8 – 50) of reperfusion (fig. 3B) in the HTK-N group but not in the BP group. The maximal slope of pressure increment (dp/dt- max) at 10 mL of LVV decreased stepwise from 1442±152 mmHg/s at 30 min to 872±180 mmHg/s at 120 min (95%CI DIM: 1242 – -103) of reperfusion and remained constant in the HTK-N group (fig. 3C). At 20 mL of LVV, dp/dtmax did not show major differences by reperfusion time in both groups (fig. 3D). The maximal slope of pressure decrement (dp/ dtmin) impaired from -935±236 mmHg/s at 30 min to -522±67 mmHg/s at 120 min (95%CI DIM: -347 – 1172) of reperfusion time in the BP group at 10 mL of LVV. In the HTK-N group, dp/dtmin improved from 30 to 60 min of reperfusion (-1032±151 vs. -1416±384 mmHg/s; 95%CI DIM: -1460 – 689) followed by a slight impairment to 120 min (-1416±384 vs. -1180±293 mmHg; 95%CI DIM: -838 – 1311) at 10 mL of LVV. At 20 mL of LVV, dp/dtmin did not majorly change within 120 min of reperfusion in BP. In the HTK-N group, dp/ dtmin minorly improved from 30 to 60 min of reperfusion and then at 120 min decreased to the level shown at 30 min.
Shift of normalized left-ventricular contractility and relaxation: blood vs. HTK-N perfusion
Relative ESP at 20 mL of LVV after 60 min (98±3% vs. 116±9%; 95%CI DIM: -38 – 3) and 120 min (99±7% vs. 113±9%; 95%CI DIM: -39 – 13) of reperfusion was increased in the HTK-N group compared to the BP group (fig. 4B). At an LVV of 10 mL, relative dp/dtmax was higher in HTK-N than in BP, especially at 120 min of reperfusion (79±17% vs. 131±21%; 95%CI DIM: -48 – 9; fig. 4C). At 20 mL, the relative dp/dtmax did not majorly differ between both groups (fig. 4D). Relative dp/dtmin highly increased in HTK-N compared to BP with 10 mL of LVV at 60 min but especially after 120 min (79±17% vs. 131±22%; 95%CI DIM: -114 – 9) of reperfusion (fig. 4E).
DISCUSSION
We could already show in previous experiments that after BP, contractility and relaxation are impaired compared to perfusion with hypothermic, oxygenated HTK-N due to increased ischemia/reperfusion (I/R) injury [5]. I/R injury develops over time during the early reperfusion period [6]. Thus, the reason for a further decrease of dp/dtmax and dp/dtmin by reperfusion time after BP could be a further increasing reperfusion injury that does not occur after HTK-N perfusion. Unlike BP, after HTK-N perfusion, dp/dtmax, and dp/dtmin remain stable, presumably because I/R injury is not only less profound but is also not further increasing – at least not during the observed reperfusion time of 2 h.
HTK-N contains several pharmacological agents to protect the myocardium from I/R injury, such as membrane-permeable and non-permeable iron chelators LK-614 and deferoxamine, which are known to reduce the production of reactive oxygen species [7]. Furthermore, it contains endothelial protective amino acids, such as L-Alanine, to reduce Hydrogen-Peroxide mediated oxidative stress and to stimulate the antioxidant defense enzyme Heme Oxygenase-1 and the vasodilatory amino acid L-Glycine [8]. Compared to the traditional HTK solution, the chloride concentration in HTK-N was reduced because it was shown in an in vitro model of vascular I/R injury that it mediates cellular injury during reperfusion [9]. The buffer substance Histidine was also partially substituted by Nα-Acetyl-Histidine to retain the buffer capacity but reduce cytotoxicity [7].
The finding that an impaired dp/dtmax and dp/dtmin after ischemia can be partially reversed by elevated left-ventricular filling volumes is consistent with the results from Takano et al. in an experimental series on myocardial ischemia in dogs and from Tayama et al. in pigs [10,11]. Another experimental series is needed to perform molecular biological analysis from tissue samples at multiple points in time during reperfusion to determine the potential progress of I/R injury during reperfusion of DCD hearts. Regarding the progress of functional parameters during reperfusion it is known from the literature that beyond 2 hours of reperfusion, the impact of I/R injury most likely reaches a plateau and does not lead to a further functional decrease of the left ventricle [6].
LIMITATIONS
A limitation of the present work is that we did not perform an orthotopic HTX. Nevertheless, this is also beneficial because in an isolated heart perfusion model, confounding factors that would occur in an in vivo experiment can be minimized, thus reducing the number of animals and making small differences visible simultaneously.
CONCLUSION
Left-ventricular contraction and relaxation of porcine DCD hearts decrease over time during early reperfusion after EVMP with normothermic blood and remain stable after EVMP with hypothermic, oxygenated HTK-N solution. Nevertheless, higher left-ventricular filling volumes can reverse the impaired tensioning and relaxation.
ETHICAL APPROVAL
The animal study was approved by the Ethics Committee “Regierungspräsidium Karlsruhe” (G150/19; 4 July 2019).
CONFLICT OF INTERESTS
Nothing to declare.
ACKNOWLEDGMENTS
Lars Saemann is a fellow of the Josef Güttler Scholarship by the German Society for Cardiovascular Perfusion (DGfK).