Journal Logo

Original Basic Science—General

Comparing Donor Heart Assessment Strategies During Ex Situ Heart Perfusion to Better Estimate Posttransplant Cardiac Function

Ribeiro, Roberto Vanin Pinto MD1,2; Alvarez, Juglans Souto MD, MSc1; Yu, Frank BSc1; Adamson, Mitchell Brady BSc, MSc1,2; Paradiso, Emanuela MD3; Hondjeu, Arnaud Romeo Mbadjeu MD3; Xin, Liming PhD1,4; Gellner, Bryan MSc, PhD4; Degen, Maja BSc1; Bissoondath, Ved BBA, MHSA1; Meineri, Massimiliano MD3,5; Rao, Vivek MD, PhD1,2,6; Badiwala, Mitesh Vallabh MD, PhD1,6

Author Information
doi: 10.1097/TP.0000000000003374



Heart transplantation remains the best treatment option for the eligible patient with advanced heart failure. Unfortunately, the number of donor hearts offered for transplantation has plateaued, while the number of adequate recipients continues to increase.1-3 The use of marginal donor organs4,5 and the resurface of donation after circulatory death (DCD)1,6-9 have been proposed to overcome organ shortage for heart transplantation, with similar early mortality rates when compared with standard donor hearts.10 However, this has also increased rates of primary graft failure and mechanical support requirements due to the inevitable warm ischemic damage DCD hearts suffer.2 The need for an accepted tool to evaluate these hearts is critical. Cold storage, the current standard preservation technique, does not enable the evaluation of donor hearts to determine the extent of graft damage.11

Ex situ heart perfusion (ESHP) platforms have been developed to limit ischemic periods and enable continuous metabolic monitoring of DCD hearts; however, a validated functional assessment method, capable of predicting cardiac performance following transplantation, has yet to be established.12-14 The current clinically available ESHP system relies solely on lactate levels to determine if the heart is suitable for transplant, limiting the ability to predict posttransplant functional recovery.1,7,12 Preclinical studies have evaluated left ventricular (LV) contractility parameters and demonstrated greater predictability of cardiac performance.13,15,16 However, these findings are limited to small-animal models and were not tested in full transplantation studies. Furthermore, right ventricular (RV) function in this setting remains unexplored.

To address this, our group developed a novel perfusion system capable of performing biventricular loading and evaluation.14 In this study, our objective was to compare biventricular contractility and metabolic parameters measured during ESHP to determine the best evaluation strategy to estimate cardiac function following transplantation.


Male Yorkshire pigs (40 ± 4 kg; Caughell Farms, CA) were used to perform 17 orthotopic heart transplants. The experimental protocol was approved by our institutional animal care committee, and animals were treated following the Guide for the Care and Use of Laboratory Animals. This study followed the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines.17 Full details regarding the methods used are given in Detailed Methods (SDC,

Donor Procedure

Anesthesia and Monitoring

Animals were premedicated with intramuscular Midazolam (0.3 mg/kg) and Ketamine (20 mg/kg). Anesthesia was induced and maintained using inhalational isoflurane through an endotracheal tube (end-tidal concentration: 1%–3%). An arterial line was inserted through a carotid artery, a central venous line was introduced into a jugular vein, and a pulmonary artery catheter was inserted via a jugular vein and directed beyond the pulmonary artery bifurcation. After the median sternotomy, the heart and great vessels were exposed. An umbilical tape was placed around the inferior vena cava, and a pressure-volume conductance catheter (Millar Instruments Inc) was inserted into the LV through an apical ventriculotomy. Systemic anticoagulation was achieved with an injection of 30 000 U heparin. Following a baseline assessment, donor animals were randomly assigned to undergo beating-heart donation (BHD; n = 9) or DCD (n = 8) (Figure 1) to avoid selection bias of animals and provide a wide range of posttransplant functional recovery, following the recommendations of the ARRIVE guidelines.17

Experimental design. Donor animals were randomly assigned to undergo beating-heart donation. BHD, beating-heart death (control hearts); DCD, donation after circulatory death; WLST, withdrawal of life sustaining therapies.

Beating-Heart Donation

Following baseline evaluation, a cardioplegia cannula was placed in the ascending aorta and an 18F venous cannula was placed in the right atrium (RA). We collected 1.5 L of whole blood into an autotransfusion system (Fresenius Kabi C.A.T.S., Terumo) to isolate the red blood cells. Simultaneously, the aorta was cross-clamped and hearts arrested with 1 L of histidine-ketoglutarate-tryptophan (HTK) at 4°C. The heart was excised and placed in ice-cold HTK for 1 hour while being cannulated for ESHP.

Donation After Circulatory Death

Before cessation of mechanical ventilation, Propofol (1 mg/kg/min) and Remifentanil (1 µg/kg/min) infusions were commenced, ensuring adequate anesthesia. Mechanical ventilation was discontinued in the preheparinized animal resulting in progressive hypoxia and circulatory arrest. Asystole was determined as the loss of pulsatility on the arterial waveform. After the onset of asystole, a 15-minute warm ischemic period was observed. Functional warm ischemic time was defined as the time since systolic blood pressure fell below 50 mm Hg until the organ was flushed with cold HTK. Whole blood was then collected, and hearts flushed and retrieved as described for BHD.

Ex Situ Heart Perfusion

We used a custom ESHP system in these experiments as previously described.14 The system was primed with 500 mL of STEEN solution (XVIVO Perfusion), 500 mL of SOM-TRN-001 (Somahlution), Mannitol (1 mg/kg), Methylprednisolone (500 mg), Heparin (10 000 U), and Cefazolin (1 g). A sufficient volume of red blood cell was added to achieve a hematocrit of 15%. Calcium chloride 10%, sodium bicarbonate 8.4%, and dextrose 50% were added to correct calcium (1.1–1.3 mmol/L), glucose (5–10 mmol/L), and bicarbonate (24–30 mmol/L) concentrations, respectively.

Hearts were mounted on the ESHP system following 1 hour of cold storage. Once deairing was complete, retrograde aortic root perfusion was started at 40 mm Hg and hearts were rewarmed over 30 minutes to 37°C. All hearts were perfused in Langendorff mode, during which hourly perfusate samples were obtained for metabolic assessment. At 3 hours, hearts were transitioned into biventricular working mode by loading the left atrium (LA) and RA at a flow corresponding to a cardiac index (CI) of 1.8 L/min/m2.

Recipient Procedure

We performed orthotopic cardiac transplants as previously described.18 In brief, sedation, anesthesia, and monitoring were performed as in the donor protocol. After median sternotomy, we exposed the heart and great vessels and the superior and inferior vena cava were encircled with umbilical tapes. Systemic anticoagulation was achieved with an intravenous injection of 30 000 U of heparin. Ascending aortic and bicaval cannulations were used to initiate cardiopulmonary bypass (CPB). Normothermia was maintained and flow rates were adjusted to maintain a mean arterial pressure above 50 mm Hg. Norepinephrine (0.05–0.10 μg/kg/min) was administered if needed for pressure regulation.

After aortic cross-clamping, the recipient heart was excised and the anastomotic margins inspected and trimmed. At 4 hours of ESHP, the donor heart was flushed with an initial dose of cold blood cardioplegia and removed from the perfusion system. For implant, we used a standard biatrial anastomotic technique in sequence: LA, RA, pulmonary artery, and ascending aorta. We chose this technique as it avoids significant manipulation of the fragile porcine tissue,19 decreases implant time in a model particularly sensitive to ischemia,20 and has been well established within our group.11,18,21-24 Cardioplegic protection consisted of 300 mL of a 2:1 mixture of blood/crystalloid and was delivered at 10°C after the completion of each anastomosis. Before removal of the aortic cross-clamp, an additional 300 mL of warm blood cardioplegia and 500 mg of methylprednisolone were administered.

Once the aortic cross-clamp was removed, hearts were reperfused for 60 minutes, after which they were weaned from CPB. Weaning was deemed successful if the animals maintained a systolic arterial pressure of 60 mm Hg for 30 minutes after the discontinuation of CPB. A vasoactive infusion of Dobutamine (5 μg/kg/min) and Norepinephrine (0.1 μg/kg/min) was used to assist in weaning from CPB. Hemodynamic and biventricular functional assessments were performed at 3 hours postreperfusion, following which, the experiment was terminated.

Biochemical Parameters

Blood Gases, Electrolytes, and Oncotic Pressure

Arterial and venous samples were collected at baseline, following DCD induction, hourly during ESHP, and hourly following transplantation. Electrolyte, lactate and hemoglobin concentrations, pH, Po2, Pco2, and oxygen saturation were measured using a blood gas analyzer (RAPIDPoint 500 Blood Gas Systems; Siemens). All metabolic parameters were assessed in Langendorff mode.

Myocardial Viability Assessment

Measurement of metabolic parameters during ESHP has been used to gauge myocardial viability before transplantation.25 Myocardial lactate extraction, coronary vascular resistance (CVR), myocardial oxygen consumption (MVO2), and oxygen extraction rate were derived as previously described.13,14

Contractility Assessment

In Vivo

Global cardiac function was evaluated by cardiac output measurements using the pulmonary artery catheter and the traditional thermodilution technique. CI was derived as the ratio between cardiac output and body surface area. RV function was assessed with the same catheter using RV Stroke Work Index (RVSWI). LV function was assessed using a conductance catheter (Ventri-Cath 507S; Millar Inc, Houston, TX). Data were collected and analyzed using IOX v1.8.9.13 software (EMKA Technologies Inc.). All assessments were performed at baseline and at 3 hours of reperfusion posttransplant. Percent recovery (%) was then calculated relative to baseline values.

Ex Situ

Biventricular functional parameters were measured using a conductance catheter. Volume-dependent measurements were collected under steady loading conditions (CI = 1.8 L/min/m2), while volume-independent measurements were collected by decreasing the RA and LA loading. Mean aortic pressure and systolic aortic pressure (SBP), mean RV pressure, systolic pulmonary artery pressure, LA pressure and RA pressure, and stroke volume (SV = cardiac output per heart rate) were recorded to derive noninvasive (NI) estimates of preload recruitable stroke work (PRSW) and maximum elastance (Emax) as previously described.13,26 In brief, PRSW is the slope of the linear regression between LV or RV stroke work (SW) and atrial pressures, and Emax is the slope of the linear regression between SBP or systolic pulmonary artery pressure and atrial pressures, where LV SW = mean aortic pressure (mm Hg) × SV (mL) and RV SW = RV pressure (mm Hg) × SV (mL).


Surface echocardiography was performed using a standard 3-dimensional transesophageal echocardiographic probe (Z6Ms; ACUSON SC2000; Siemens). The scan protocol includes 3 views: LV short axis, apical 4, and 2 chamber. Systolic function is evaluated by ejection fraction, fractional area change (FAC), global longitudinal strain, and global circumferential strain.

Statistical Analysis

Statistical analysis was performed using the SPSS 24 (IBM) and Prism 8 (GraphPad). Data are reported as median values (interquartile range). Comparisons between BHD and DCD hearts were performed using nonparametric tests. Spearman correlation analyses were used to determine the correlation between ESHP parameters and posttransplant function. Correlations were classified as weak (r < 0.5), moderate (r = 0.5–0.7), or strong (r > 0.7) as previously described.27 Significance was set at α = 0.05.


Posttransplant Outcomes

We performed 17 transplants; 14 successfully weaned from CPB after reperfusion (BHD: 8 of 9, DCD: 6 of 8; P = 0.580). Table 1 illustrates the procedural time points in each group. Median functional warm ischemic time of DCD hearts was 25 minutes (23–30 min). There were no differences in cold storage, ESHP, implant, or CPB times. As expected, total ischemic time was significantly greater in DCD hearts (BHD = 427 [419–433], DCD = 455 [446–469]; P = 0.002).

Protocol timeline comparison between BHD and DCD hearts

We initially correlated biventricular contractility parameters with CI measured following transplantation to determine the relative contribution of each variable to global cardiac function. Tau, end-systolic elastance, end-diastolic pressure-volume relationship (EDPVR), maximum rate of developed pressure (dP/dtmax) versus end-diastolic volume, Emax, and RV-FAC showed weak or no correlation with global myocardial function and were excluded from further analyses. Developed pressure (dP), minimum rate of dP (dP/dtmin), and RVSWI showed moderate correlation; dP/dtmax, SW, and PRSW had a strong correlation (Table S1, SDC, As such, posttransplant measurements of dP, dP/dtmin, dP/dtmax, SW, PRSW, and RVSWI were chosen as end points for correlation with ESHP data.

Metabolic Assessment

CVR and MVO2 were similar between BHD and DCD hearts throughout ESHP (Figure 2A and B). BHD hearts showed increasing oxygen extraction during the first hour of perfusion and remained stable afterward, while DCD hearts demonstrated a decrease throughout the same period (P < 0.001) (Figure 2C). DCD hearts demonstrated significantly higher total lactate concentration throughout ESHP (BHD = 0.9 mmol/L [0.4–2.2 mmol/L], DCD = 3.6 mmol/L [1.2–5.1 mmol/L] at 4 h; P < 0.001), but no difference was seen in lactate extraction between groups (P = 0.110) (Figure 2D and E).

Metabolic assessment during perfusion. A, Coronary vascular resistance (CVR). B, Myocardial oxygen consumption (MVO2). C, Oxygen extraction. D, Total venous lactate level. E, Presence of arteriovenous lactate extraction. CVR and MVO2 were similar between beating-heart donation (BHD) and donation after circulatory death (DCD) hearts throughout perfusion. Regarding oxygen extraction, BHD hearts increased extraction during the first h of perfusion and remained stable afterward, while DCD hearts demonstrated a decrease throughout the same period (P < 0.001). DCD hearts demonstrated significantly higher total lactate concentration throughout ESHP (P < 0.001), but no differences were seen regarding lactate extraction. *P < 0.05, ***P < 0.001. ESHP, ex situ heart perfusion.

CVR at 30 minutes (r = −0.525; P = 0.031) and 3 hours (r = −0.503; P = 0.047) and MVO2 at 3 hours (r = 0.650; P = 0.006) showed moderate correlation with posttransplant RVSWI (Table 2). Neither CVR nor MVO2 correlated with CI following transplantation. Also, no correlations were seen between oxygen extraction and any posttransplant measurements. Lactate at 1 and 3 hours demonstrated weak correlation with %RVSWI (r = −0.484; P = 0.049) and CI (r = −0.495; P = 0.043), respectively. The presence of lactate extraction was associated with a decreased RVSWI following transplant (r = −0.608; P = 0.012) (Table 2). Metabolic variables did not correlate with posttransplant LV function. Table S2 (SDC, illustrates all metabolic correlation coefficients.

Spearman correlation coefficients of significant metabolic predictive parameters

LV Assessment

Regarding LV functional assessment during ESHP, dP was significantly higher in BHD hearts (BHD = 138 [114–159], DCD = 101 [81–115]; P = 0.004). No differences were noted between BHD and DCD hearts in other invasive measurements of LV function. NI measurements of LV Emax (BHD = 8.4 [4.0–12.2], DCD = 3.0 [1.1–5.5]; P = 0.015) and PRSW (BHD = 98 [46–140], DCD = 17 [6–42]; P = 0.008) were significantly greater in BHD hearts (Figure 3A).

Biventricular contractility assessment during perfusion. A, Left ventricular (LV) functional assessment. Developed pressure was significantly higher in beating-heart donation (BHD) hearts (BHD = 138 [114–159], donation after circulatory death [DCD] = 101 [81–115]; P = 0.004). No major differences were noted between BHD and DCD hearts in other invasive measurements of LV function. Noninvasive (NI) measurements of LV maximum elastance (Emax) (BHD = 8.4 [4.0–12.2], DCD = 3.0 [1.1–5.5]; P = 0.015) and preload recruitable stroke work (PRSW) (BHD = 98 [46–140], DCD = 17 [6–42]; P = 0.008) were significantly greater in the BHD group. B, Right ventricular (RV) functional assessment. Systolic (developed pressure, stroke work, PRSW, and end-systolic elastance [Ees]) and diastolic (Tau) invasive measurements of RV function during perfusion were markedly improved in BHD hearts compared with the DCD group. NI measurements of RV Emax (BHD = 6.4 [3.3–12.9], DCD = 2.1 [0.9–3.9]; P = 0.008) and PRSW (BHD = 30 [21–66], DCD = 8 [6–12]; P = 0.001) were also significantly greater in the BHD group. *P < 0.05, **P < 0.01. dP/dt, rate of developed pressure; EDPVR, end-diastolic pressure-volume relationship.

Significant correlations with posttransplant myocardial performance were found for several LV contractility parameters (Table 3). There was significant variation within measured parameters in the ability to predict LV and global function following transplantation. LV predictive parameters dP, dP/dtmax, dP/dtmin, tau, EDPVR, PRSW, NI Emax, and NI PRSW showed moderate to strong correlations with LV performance following transplant. Of note, PRSW (r = 0.770; P = 0.009), NI Emax (r = 0.706; P = 0.002), and NI PRSW (r = 0.730; P = 0.001) also showed strong correlations with %CI after reperfusion. Table S3 (SDC, illustrates all LV function correlations.

Spearman correlation coefficients of significant left ventricular predictive parameters

RV Assessment

dP, SW, PRSW, end-systolic elastance, and tau during ESHP were markedly improved in BHD compared with DCD hearts (Figure 3B). NI RV Emax (BHD = 6.4 [3.3–12.9], DCD = 2.1 [0.9–3.9]; P = 0.008) and PRSW (BHD = 30 [21–66], DCD = 8 [6–12]; P = 0.001) were also significantly greater in BHD hearts. RV predictive parameters demonstrated fewer significant correlations with posttransplant performance (Table 4). dP (r = 0.560; P = 0.046), SW (r = 0.571; P = 0.041), EDPVR (r = 0.629; P = 0.028) demonstrated moderate correlation with %RVSWI following reperfusion. NI PRSW showed moderate correlation with CI (r = 0.688; P = 0.003). Table S4 (SDC, illustrates all RV function correlations.

Spearman correlation coefficients of significant right ventricular predictive parameters

Echocardiographic Assessment

LV and RV function, as noted by LV global circumferential strain (BHD = −12.1% [−16.4% to −10.3%], DCD = −4.8% [−10.8% to −3.3%]; P = 0.010) and RV-FAC (BHD = 41.7% [30.1%–50.1%], DCD = 23.6% [17.1%–33.5%]; P = 0.021; Figure 4), were significantly better in BHD hearts during ESHP. There were no differences in LV ejection fraction, LV FAC, or LV global longitudinal strain between groups. There were no significant correlations between echocardiographic data obtained during ESHP and posttransplant performance (Table S5, SDC,

Echocardiographic assessment during perfusion. Left ventricular (LV) and right ventricular (RV) systolic function, as noted by LV global circumferential strain (GCS) (beating-heart donation [BHD] = −12.1% [−16.4% to −10.3%], donation after circulatory death [DCD] = −4.8% [−10.8% to −3.3%]; P = 0.010) and RV fractional area change (FAC) (BHD = 41.7% [30.1%–50.1%], DCD = 23.6% [17.1%–33.5%]; P = 0.021), was significantly better in BHD hearts during ex situ heart perfusion (ESHP). There were no differences in LV ejection fraction (EF), LV FAC, or LV global longitudinal strain (GLS) between groups. *P < 0.05, **P < 0.01.


The need to determine the extent of cardiac damage before transplantation is critical to increase the use of donor hearts procured from DCD or extended-criteria donors. We compared contractile and metabolic parameters measured during ESHP and identified highly predictive measures of myocardial functional recovery following transplantation. The main findings of this study were (1) ESHP measures of LV contractility, especially LV work, correlated better with posttransplant cardiac performance compared with RV and metabolic measurements; and (2) simple NI estimations of LV and RV work proved useful to predict function following transplant.

ESHP can improve resuscitation and preservation of marginal and DCD hearts previously deemed unsuitable for transplantation and thus expand the donor pool.1,7 However, previous studies report acute graft dysfunction rates of up to 30%.8,10 Thus, it is imperative to evaluate the viability of these donor hearts to reduce the risk of posttransplant complications. Various methods can be used to assess organ viability during ESHP, including biomarkers of tissue injury (ie, troponin I), metabolic measurements (ie, MVO2), and hemodynamic and contractility parameters (ie, pressure-volume loops and echocardiography).16,28-30 While these assessment tools are invaluable, reliable, easy to use, and reproducible, parameters are needed to assess myocardial function before transplant.

In the clinical setting, allograft assessment using the Organ Care System (Transmedics) uses lactate levels and arteriovenous lactate extraction as markers of heart viability and suitability for transplantation.12,28,31 This is supported by Hamed et al,32 who revealed that a serum lactate above 4.96 mmol/L was a strong predictor of graft dysfunction at 30 days, and has been used to identify suitable DCD hearts for clinical transplantation.7 In our study, we evaluated several metabolic markers repeatedly throughout ESHP. However, these only demonstrated a modest number of weak correlations. Dornbierer et al,16 using a rat model of DCD heart procurement and ex situ reperfusion and assessment, and White et al,13 using a similar porcine model, have previously reported on the limited applicability of metabolic measurements. However, their work was exclusively performed with assessment during ESHP and a transplantation model was required before translating their findings to a clinical setting. Other biomarkers of myocardial damage, such as troponin I, have also proven to be of limited value in determining organ viability due to their natural elevation during the warm ischemic and preservation insult.28,33

MVO2 measures the oxygen consumption of the myocardium while performing work and represents a functioning myocardium capable of adequately consuming oxygen.28,31 CVR represents the capacity of the vascular endothelium to adjust blood flow and oxygen delivery.34 Finally, lactate concentration has been used as a surrogate marker of anaerobic metabolism due to insufficient oxygen delivery or a compromised oxidative metabolism.13 These metabolic parameters are readily measurable during ESHP; however, they are also highly dependent on oxygen delivery, which can be easily controlled on perfusion platforms. For example, coronary blood flow can be increased in response to rising lactate levels to increase oxygen delivery and induce aerobic metabolism.12 Thus, the easy manipulation and high variability of these markers makes their use limited to predict posttransplant performance or graft dysfunction. In contrast, we believe that these markers may have a role to play in providing an initial assessment of DCD hearts to determine the extent of ischemic damage when measured following cardiac arrest or during initial reperfusion. In line with this are our previous findings demonstrating that lactate levels measured in DCD donors immediately before in situ reperfusion showed a strong correlation with myocardial function following transplantation.35 This is also seen in this study, where DCD hearts demonstrated significantly higher lactate levels throughout perfusion.

Although previous studies have suggested the advantage of contractility measurements over metabolic parameters during ESHP to determine early graft function following transplantation, these were limited to LV function.13,16,28,36 The impact of RV evaluation has remained unexplored in this setting. This is of critical importance as RV failure has been described in the majority of patients developing primary graft failure following transplant.37 Several factors may contribute to this, such as the thinner RV walls which may lead to faster temperature changes and greater warm ischemic times during exposure to room temperature during implantation.37 This has been further highlighted by Ou et al,38 who demonstrated the deleterious effect in the myocardium associated with unprotected rewarming during the vulnerable implantation phase. As such, there has been a belief that RV function may prove to be a better predictor of cardiac function following transplant.13

Here, RV function during ESHP showed limited ability to correlate with posttransplant performance. Although the lack of an afterload device for the RV on our system14 may be considered a limitation for the assessment of this ventricle, our evaluation of RV function during ESHP clearly demonstrated differences between BHD and DCD hearts as would be expected. It is well known that RV function has prognostic value in acute, chronic, and perioperative disease.37 However, the complex RV contractile pattern makes assessment difficult, and the evaluation of RV function is still a clinical challenge which merits further investigation.39

In comparison, LV functional assessment has been thoroughly investigated in this setting.13,16,28,36 The lack of posttransplant evaluations, however, does not allow correlation between ESHP findings and outcomes. In this study, we found that various measures of LV contractility correlated strongly with LV and global myocardial performance following transplant. Ideal parameters of contractility should be sensitive to the heart’s inotropic state but relatively insensitive to heart size, rate, and loading conditions to ensure a reliable and reproducible measurement.31 Interestingly, PRSW, a marker of LV work and reserve, meets these criteria and we found it to be strongly correlated with posttransplantation performance.

Conductance catheters to assess LV function have been well established in the research and clinical setting.28,31 However, these are invasive and can be cumbersome to utilize, decreasing result reproducibility. Ideally, methods to evaluate donor hearts should be easily and rapidly performed. With this in mind, we tested an NI method in this study. We calculated NI estimations of ventricular work as previously described by White et al.13,26 In our study, these simple NI estimates of LV function demonstrated once again a strong correlation with cardiac function following transplant. In contrast, the RV-PRSW estimate only showed a moderate correlation with posttransplant RVSWI. Again, assessment of the complex contractile pattern of the RV remains a significant challenge.

Finally, we evaluated whether echocardiographic measurements could predict posttransplant myocardial performance. Although during ESHP these measurements demonstrated significant differences between BHD and DCD hearts, they did not correlate with posttransplant measurements. We previously demonstrated that obtaining high-quality images during ESHP is feasible.40 However, the assessment was only performed with steady preload, without the evaluation of different preload and afterload conditions as done with pressure-volume loops. This may, at least in part, explain the lack of significant correlations between echocardiographic parameters and posttransplant function. Nevertheless, the ability to perform echocardiographic assessments in this setting is important and should be further explored in future studies.

Overall, such NI approaches could eliminate the need for conductance catheters and facilitate clinical adoption of functional evaluation during ESHP. Our group and others have been working on the development of novel NI assessment methods to predict myocardial performance following transplantation. Specifically, we have previously demonstrated the ability to derive LV dP/dtmax from aortic pressure curves during ESHP while perfusing the heart in a nonworking mode.41 This not only facilitates data acquisition and increases reproducibility but can also eliminate the need to transition donor hearts into working mode during perfusion. A feature which is currently not available in the Organ Care System.1,7,12

Strengths and Limitations

To our knowledge, this is the first study to directly compare the ability of ESHP contractility and metabolic markers to estimate posttransplant function in a large-animal transplant model. Furthermore, we repeatedly measured metabolic parameters throughout perfusion, in contrast with single measurements performed previously.13,15,16 Finally, our experimental model closely mimics the clinical scenario of standard and DCD heart transplantation, ensuring readily translatable findings. We sought to build upon previous work and investigate multiple markers which are readily available for clinical use, making our findings easily translatable to practice. Further research and development of new clinical ESHP systems should be capable of evaluating donor hearts in a working mode and providing real-time estimates of myocardial function. Preferentially, these will be NI and avoid injury to the organ. Also, these will then need to be applied in a larger clinical scale to determine cutoff values that accurately determine organ suitability for transplantation.

Nevertheless, this work has important limitations. Our sample size was small, donor and recipient animals were of young age—potentially increasing ischemic tolerance, and follow-up was limited to the early posttransplant period. Conclusions regarding the ability of our measurements to estimate myocardial performance should be considered hypothesis generating. However, we still detected relevant and consistent information which is in line with previous studies. Although we provided precise preload, afterload, and inotropic conditions during functional assessment, we did not control heart rate which increases variability in some parameters (ie, dP/dtmax). It is important that these parameters be acquired under equivalent preload, afterload, chronotropic, and inotropic conditions to ensure reproducibility and reliability. Finally, we did not use supplemental postconditioning agents which our cardioplegia which improve functional recovery of DCD hearts and are currently standard practice7,25 because we wished to achieve a wide range of functional recovery to detect significant correlations.


LV function, especially ventricular work and reserve, measured during ESHP provided the best estimation of myocardial performance following transplantation. Furthermore, simple NI estimates of ventricular function proved to be useful in this setting, potentially precluding the need for invasive and cumbersome techniques. RV and metabolic measurements were limited in their ability to correlate with myocardial recovery. This emphasizes the need for an ESHP platform capable of assessing myocardial function in a physiological working mode and suggests that metabolic parameters alone do not provide a reliable evaluation of organ viability. Future studies are needed to validate these findings in the clinical scenario and to investigate potential cutoff values to accurately determine organ suitability for transplantation.


The authors would like to acknowledge Professor Yun Sun, Professor Craig Simmons, Professor Jean Zu, and Drs Marcelo Cypel, Lisa Robinson, and Markus Selzner for their important advice regarding the conductance of this project. We thank Somahlution for generously providing SOM-TRN-001 for these experiments.


1. Messer SJ, Axell RG, Colah S, et al. Functional assessment and transplantation of the donor heart after circulatory death. J Heart Lung Transplant. 2016; 35:1443–1452
2. Messer S, Page A, Axell R, et al. Outcome after heart transplantation from donation after circulatory-determined death donors. J Heart Lung Transplant. 2017; 36:1311–1318
3. Lund LH, Khush KK, Cherikh WS, et al.; International Society for Heart and Lung Transplantation. The Registry of the International Society for Heart and Lung Transplantation: Thirty-fourth Adult Heart Transplantation Report-2017; Focus Theme: allograft ischemic time. J Heart Lung Transplant. 2017; 36:1037–1046
4. Popov AF, García Sáez D, Zych B, et al. The Organ Care System heart in high risk transplantation with an adverse donor/recipient profile. Ready for daily practice? Thorac Cardiovasc Surg. 2015; 63S 01OP162
5. García Sáez D, Zych B, Sabashnikov A, et al. Evaluation of the organ care system in heart transplantation with an adverse donor/recipient profile. Ann Thorac Surg. 2014; 98:2099–2105. discussion 2105
6. Hesse K, Aitken E, Clancy M, et al. Expanded criteria donor and donation after circulatory death renal allografts in the west of Scotland: Their place in the kidney allocation process Surgeon. 2014; 14:136–141
7. Dhital KK, Iyer A, Connellan M, et al. Adult heart transplantation with distant procurement and ex-vivo preservation of donor hearts after circulatory death: a case series. Lancet. 2015; 385:2585–2591
8. Messer S, Ardehali A, Tsui S. Normothermic donor heart perfusion: current clinical experience and the future. Transpl Int. 2015; 28:634–642
9. Messer S, Lannon J, Wong E, et al. The potential of transplanting hearts from donation after circulatory determined death (DCD) donors within the United Kingdom J Heart Lung Transplant. 2015; 34:S275
10. Dhital KK, Chew HC, Macdonald PS. Donation after circulatory death heart transplantation. Curr Opin Organ Transplant. 2017; 22:189–197
11. Ribeiro RVP, Alvarez JS, Yu F, et al. Hearts donated after circulatory death and reconditioned using normothermic regional perfusion can be successfully transplanted following an extended period of static storage. Circ Heart Fail. 2019; 12:e005364
12. Ardehali A, Esmailian F, Deng M, et al.; PROCEED II trial investigators. Ex-vivo perfusion of donor hearts for human heart transplantation (PROCEED II): a prospective, open-label, multicentre, randomised non-inferiority trial. Lancet. 2015; 385:2577–2584
13. White CW, Ambrose E, Müller A, et al. Assessment of donor heart viability during ex vivo heart perfusion. Can J Physiol Pharmacol. 2015; 93:893–901
14. Xin L, Gellner B, Ribeiro RVP, et al. A new multi-mode perfusion system for ex vivo heart perfusion study. J Med Syst. 2017; 42:25
15. Sourdon J, Dornbierer M, Huber S, et al. Cardiac transplantation with hearts from donors after circulatory declaration of death: haemodynamic and biochemical parameters at procurement predict recovery following cardioplegic storage in a rat model. Eur J Cardiothorac Surg. 2013; 44:e87–e96
16. Dornbierer M, Stadelmann M, Sourdon J, et al. Early reperfusion hemodynamics predict recovery in rat hearts: a potential approach towards evaluating cardiac grafts from non-heart-beating donors. PLoS One. 2012; 7:e43642
17. Kilkenny C, Browne W, Cuthill IC, et al.; NC3Rs Reporting Guidelines Working Group. Animal research: reporting in vivo experiments: the ARRIVE guidelines. Br J Pharmacol. 2010; 160:1577–1579
18. Badiwala MV, Ramzy D, Tumiati LC, et al. Donor pretreatment with hypertonic saline attenuates primary allograft dysfunction: a pilot study in a porcine model Circulation. 2009; 120:S206–S214
19. Robinson N, Souslian L, Gallegos RP, et al. Iaizzo PA. Animal models for cardiac research. In: Handbook of Cardiac Anatomy, Physiology, and Devices. 2015, Cham, Switzerland: Springer International Publishing, 469–491
20. Bianco RW, Gallegos RP, Rivard AL, et al. Iaizzo PA. Animal models for cardiac research. In: Handbook of Cardiac Anatomy, Physiology, and Devices. 2009, Totowa, NJ: Humana Press, 393–410
21. Ramzy D, Rao V, Mallidi H, et al. Cardiac allograft preservation using donor-shed blood supplemented with L-arginine. J Heart Lung Transplant. 2005; 24:1665–1672
22. Rao V, Feindel CM, Weisel RD, et al. Donor blood perfusion improves myocardial recovery after heart transplantation. J Heart Lung Transplant. 1997; 16:667–673
23. Ribeiro RVP, Alvarez JS, Yu F, et al. A pre-clinical porcine model of orthotopic heart transplantation J Vis Exp. 2019146:e59197
24. Ribeiro RVP, Badiwala MV, Ramzy D, et al. Recipient hypertonic saline infusion prevents cardiac allograft dysfunction. J Thorac Cardiovasc Surg. 2019; 157:615–625.e1
25. Iyer A, Gao L, Doyle A, et al. Increasing the tolerance of DCD hearts to warm ischemia by pharmacological postconditioning. Am J Transplant. 2014; 14:1744–1752
26. White CW, Nagendran J, Freed DH. Assessment of preload-recruitable stroke work during biventricular ex vivo heart perfusion: a novel approach eliminating the pressure-volume loop catheter Can J Cardiol. 2016; 32:S173–S174
27. Mukaka MM. Statistics corner: a guide to appropriate use of correlation coefficient in medical research. Malawi Med J. 2012; 24:69–71
28. Beuth J, Falter F, Pinto Ribeiro RV, et al. New strategies to expand and optimize heart donor pool: ex vivo heart perfusion and donation after circulatory death: a review of current research and future trends. Anesth Analg. 2019; 128:406–413
29. White CW, Ali A, Hasanally D, et al. A cardioprotective preservation strategy employing ex vivo heart perfusion facilitates successful transplant of donor hearts after cardiocirculatory death. J Heart Lung Transplant. 2013; 32:734–743
30. Ribeiro R, Ghashghai A, Yu F, et al. Comparison between Steen and Somah solutions as primary perfusate components for ex vivo heart perfusion Can J Cardiol. 2017; 33:S69
31. White CW, Messer SJ, Large SR, et al. Transplantation of hearts donated after circulatory death. Front Cardiovasc Med. 2018; 5:8
32. Hamed A, Tsui S, Huber J, et al. 19: Serum lactate is a highly sensitive and specific predictor of post cardiac transplant outcomes using the Organ Care System J Heart Lung Transpl. 2009; 28:S71
33. Collins MJ, Moainie SL, Griffith BP, et al. Preserving and evaluating hearts with ex vivo machine perfusion: an avenue to improve early graft performance and expand the donor pool. Eur J Cardiothorac Surg. 2008; 34:318–325
34. Goodwill AG, Dick GM, Kiel AM, et al. Regulation of coronary blood flow. Compr Physiol. 2017; 7:321–382
35. Ribeiro R, Alvarez JS, Paradiso E, et al. Functional assessment of DCD hearts following normothermic regional perfusion can predict post-transplantation contractility J Heart Lung Transpl. 2019; 38:S182
36. Suehiro K, Mohri M, Yamaguchi H, et al. Posttransplant function of a nonbeating heart is predictable by an ex vivo perfusion method. Ann Thorac Surg. 2001; 71:278–283
37. Cosío Carmena MDG, Gómez Bueno M, Almenar L, et al. Primary graft failure after heart transplantation: characteristics in a contemporary cohort and performance of the RADIAL risk score J Heart Lung Transplant. 2013; 32:1187–1195
38. Ou R, Gavin JB, Esmore DS, et al. Increased temperature reduces the protective effect of University of Wisconsin solution in the heart. Ann Thorac Surg. 1999; 68:1628–1634. discussion 1634
39. Dawes TJW, de Marvao A, Shi W, et al. Identifying the optimal regional predictor of right ventricular global function: a high-resolution three-dimensional cardiac magnetic resonance study. Anaesthesia. 2019; 74:312–320
40. Ruggeri GM, Qua Hiansen J, Pivetta E, et al. Description of a novel set-up for functional echocardiographic assessment of left ventricular performance during ex vivo heart perfusion Anesth Analg. 2018; 127:e36–e39
41. Xin L, Yao W, Peng Y, et al. Primed left ventricle heart perfusion creates physiological aortic pressure in porcine hearts ASAIO J. 2020; 66:55–63

Supplemental Digital Content

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.