Comparison of HTK-Custodiol and St-Thomas solution as cardiac preservation solutions on early and midterm outcomes following heart transplantation

Abstract

OBJECTIVES

The choice of the cardiac preservation solution for myocardial protection at time of heart procurement remains controversial and uncertainties persist regarding its effect on the early and midterm heart transplantation (HTx) outcomes. We retrospectively compared our adult HTx performed with 2 different solutions, in terms of hospital mortality, mid-term survival, inotropic score, primary graft dysfunction and rejection score.

METHODS

From January 2009 to December 2020, 154 consecutive HTx of adult patients, followed up in pre- and post-transplantation by 2 different tertiary centres, were performed at the University Hospital of Lausanne, Switzerland. From 2009 to 2015, the cardiac preservation solution used was exclusively St-Thomas, whereafter an institutional decision was made to use HTK-Custodiol only. Patients were classified in 2 groups accordingly.

RESULTS

There were 75 patients in the St-Thomas group and 79 patients in the HTK-Custodiol group. The 2 groups were comparable in terms of preoperative and intraoperative characteristics. Postoperatively, compared to the St-Thomas group, the Custodiol group patients showed significantly lower inotropic scores [median (interquartile range): 35.7 (17.5–60.2) vs 71.8 (31.8–127), P < 0.001], rejection scores [0.08 (0.0–0.25) vs 0.14 (0.05–0.5), P = 0.036] and 30-day mortality rate (2.5% vs 14.7%, P = 0.007) even after adjusting for potential confounders. Microscopic analysis of the endomyocardial biopsies also showed less specific histological features of subendothelial ischaemia (3.8% vs 17.3%, P = 0.006). There was no difference in primary graft dysfunction requiring postoperative extracorporeal membrane oxygenation. The use of HTK-Custodiol solution significantly improved midterm survival (Custodiol versus St-Thomas: hazard ratio = 0.20, 95% confidence interval: 0.069–0.60, P = 0.004).

CONCLUSIONS

This retrospective study comparing St-Thomas solution and HTK-Custodiol as myocardial protection during heart procurement showed that Custodiol improves outcomes after HTx, including postoperative inotropic score, rejection score, 30-day mortality and midterm survival.

Over the last decades, heart transplantation (HTx) has become the gold standard of care for well-selected end-stage heart disease patients [1].

Graphical Abstract

INTRODUCTION

Over the last decades, heart transplantation (HTx) has become the gold standard of care for well-selected end-stage heart disease patients [1]. Nowadays, HTx still remains the treatment of choice despite the increasing number of continuous-flow mechanical circulatory support devices and their favourable results in different clinical settings [1].

Successful organ preservation is a key element of transplantation since its goal is to maintain the viability of the organ until its implantation into the recipient. Two issues are important in this process: the type of preservation solution used to obtain the diastolic cardiac arrest and the duration of the cold ischaemic storage. The duration of the latter should be limited to 4–6 hours, and it is well known that longer preservation data alter outcomes [2], although ischaemic times as long as 13 hours have been reported [3]. In that perspective, the 2017 registry of the International Society for Heart and Lung Transplantation reported that allograft ischaemic time between 2 and 4 hours is associated with considerably higher survival and better early outcomes than allograft ischaemic time of >4 h [4]. More than 100 preservation solutions [2] have been developed and applied worldwide, but there is no consensus on the choice to use cardiac preservation solution (CPS), and uncertainties persist regarding its effect on early- and mid-term HTx outcomes, including a potential survival benefit.

The goal of this work was to report our two-centre (University Hospital of Lausanne and University Hospitals of Geneva, Switzerland) experience of HTx over a period of 12 years with 2 different CPS (St-Thomas and HTK-Custodiol). Based on unchanged patient profiles in the cohort of HTx recipients, we investigated the impact of these 2 CPSs on hospital mortality (30-day mortality) and mid-term mortality, inotropic score, primary graft dysfunction requiring extracorporeal membrane oxygenation (ECMO) and 1-year post-transplant rejection score.

METHODS

From January 2009 to December 2020, 165 consecutive HTx for end-stage heart failures from all aetiologies were performed in our institution (Lausanne University Hospital, Switzerland). The patients were followed up pre- and postoperatively by 2 different tertiary centres, respectively Lausanne University Hospital and Geneva University Hospitals, Switzerland. After excluding patients under 18 years of age, the study population included 154 adult patients. From 2009 to 2015, the CPS used was exclusively St-Thomas, whereafter the institution made a decisive switch to HTK-Custodiol only. Thus, patients were classified in 2 groups according to the solution used, St-Thomas or HTK-Custodiol.

The study was approved by the Ethic Committee of the Lausanne University Hospital (Switzerland) in March 2018 (CER-VD2019-704) after a thorough scrutiny of the study protocol as well as an analysis of a sample of patients from the study population. We requested and obtained a written informed consent for all patients.

Operative strategy

During organ procurement, CPS administration varied according to the type of CPS used. HTK-Custodiol was perfused at the dose of 30 ml/kg (of donor body weight) to achieve a total infusion time of 7 min. St-Thomas was administered at the dose of 20 ml/kg (of donor body weight). In both groups, topical cooling with ice-slush was also employed during harvest and transport. If allograft ischaemic time exceeded 150 min, 500 ml of CPS were re-administered upon graft arrival in the operating room (St-Thomas or HTK-Custodiol depending on the first solution administered).

Clinical evaluation and follow-up data

Patients’ demographic and clinical data recorded prior to surgery by the physician in charge were retrieved from electronic patient records without alteration. Operative, in-hospital postoperative and follow-up data were collected by the intensive care team and the heart failure cardiologists in charge of the patient from the time of the surgery. Primary outcomes of interest were hospital vasoactive inotropic score (VIS), rejection score, primary graft dysfunction requiring ECMO and 30-day mortality. Overall mid-term survival was considered as secondary outcome.

The International Society for Heart and Lung Transplantation histological rejection score [5] was obtained by endomyocardial biopsies every week for the first month, every 2 weeks for the next 6 weeks, monthly biopsies for 3–4 months and every 3 months until the end of the first year. The rejection score was calculated as the average of the scores obtained from the first 5 endomyocardial biopsies. The VIS was calculated according to Gaies et al. [6] formula, as a predictor of poor outcomes after cardiac surgery (death, cardiac arrest, need for mechanical circulatory support, renal replacement therapy and/or neurological injury) [6]. Hourly doses of all vasoactive medications were recorded and the maximum level of each medication through the first 48 h carefully noted. The first 3 post-transplantation endomyocardial biopsies were analysed in search of histological features of subendothelial ischaemia to evidence potential ischaemia reperfusion injuries. The cardiac allograft vasculopathy was also scrutinized 1 year after the surgery.

Statistical methods

Results were expressed as mean and standard deviation for quantitative variables or median and interquartile range for non-normally distributed variables. Frequency tables (numbers and %) were used for summarizing categorical data. Normality of distributions was assessed by the Shapiro–Wilk test. A log-transform was used to normalize non-normal distributions (VIS, waiting time). For quantitative variables, groups (St-Thomas versus HTK-Custodiol) were compared by the Kruskal–Wallis test while the chi-squared test for qualitative variables. Multiple linear regression was used to assess the relationship between a quantitative outcome and several covariates. For a binary (ordinal) outcome, (ordinal) logistic regression was applied to the data. Results were expressed as regression coefficient or odds ratio (OR) with 95% confidence interval (CI). The E-value was estimated to measure the effect of potential hidden biases on the association between the exposure (CPS) and outcome (30-day mortality). A high E-value suggests that uncontrolled confounders have to be strongly related to exposure and outcome to completely explain the association. The Kaplan–Meier method was used to estimate survival functions. The relationship between a survival outcome variable and covariates was assessed by Cox regression analysis. Results were then expressed by the hazard ratio and its 95% CI. Statistical calculations were always done on the maximum number of data available. Missing values were neither replaced nor imputed. Results were considered significant at the 5% critical level (P < 0.05). All calculations were done with SAS version 3.4 (SAS Institute, Cary, NC, USA) and R version 4.1.0 (R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Patient characteristics

Of the 154 adult patients who underwent HTx for end-stage heart failure from all aetiologies, 75 (48.7%) received St-Thomas and 79 (52.3%) HTK-Custodiol as CPS. The overall percentage of missing data was 8.3%, respectively 6.3% (St-Thomas) and 10.1% (Custodiol). The mean number (range) of missing values per patient in St-Thomas group 1.3 (0–5) was significantly lower than in the Custodiol group 2.4 (0–7). However, for most variables, data were either complete or only barely missing in each group. Baseline patient (recipient and donor) characteristics are displayed in Table 1. Recipients did not differ by age, aetiology of the heart failure, presence of a ventricular assist device preoperatively, mean ejection fraction, cardiopulmonary bypass (CPB) time, gender-, height- or weight-mismatch, previous cardiac surgery and previous biventricular failure. By contrast, there were more women in the Custodiol group than in the St-Thomas group (29.2% vs 13.3%, P = 0.017) and the ischaemic time was shorter (172 ± 45.5 vs 144 ± 40.2 min, P < 0.001). As for donors, they were perfectly comparable with respect to cause of death (P = 0.39) and gender (P = 0.75) but were slightly older in Custodiol group than in St-Thomas group (43.5 ± 14.9 vs 49.2 ± 14.4 years, P = 0.038).

Table 1:

Baseline recipient and donor characteristics

Variable	St-Thomas	HTK-Custodiol	P-value
	N = 75	N = 79
Recipient
Age (years)	51.9 (12.1)	51.3 (12.9)	0.76
Female gender	10 (13.3)	23 (29.2)	0.017
Waiting time on list (days)a	170 (89–403)	209 (63–403)	0.87
Ischaemic aetiology	30 (40.0)	36 (45.6)	0.49
VAD	24 (32.0)	32 (40.5)	0.27
Diabetes	13 (17. 3)	21 (26.6)	0.17
RF (ml/min/1.73 m2)	50.8 (14.3)	56.7 (20.3)	0.079
Ejection fraction (%)	25.0 (19.1)	28.4 (14.4)	0.15
VO2 max (ml/min/kg)	14.1 (4.0)	18.3 (24.2)	0.17
PVR (WU)	2.4 (1.2)	2.3 (0.97)	0.55
Ischaemic time (min)b	172 (45.5)	144 (40.2)	<0.001
	(N = 68)	(N = 49)
CPB time (min)a	135 (110–188)	143 (103–180)	0.73
Gender mismatch	28 (37.3)	31 (39.2)	0.81
Height mismatch	1 (1.3)	1 (1.3)	1.0
Weight mismatch	18 (24.0)	23 (29.1)	0.47
Previous cardiac surgery	40 (53.3)	52 (65.8)	0.11
Emergency transplantation	18 (24.0)	19 (24.1)	0.99
Donor
Cause of death
Cerebral haemorrhage	33 (44.0)	35 (44.3)	0.39
Anoxia	12 (16.0)	10 (12.7)
Trauma	23 (30.7)	26 (32.9)
Cerebral oedema	7 (9.3)	8 (10.1)
Age (years)b	43.5 (14.9)	49.2 (14.4)	0.038
	(N = 72)	(N = 50)
Female genderb	25 (40.3)	11 (44.0)	0.75
	(N = 62)	(N = 25)

Summary statistics are presented as mean (SD) or number (%).

^aMedian (IQR).

^bActual sample sizes are given in parentheses.

CPB: cardiopulmonary bypass; IQR: interquartile range; PVR: pulmonary vascular resistance; RF: renal function; SD: standard deviation; VAD: ventricular assist device; VO₂ max: maximal oxygen consumption; WU: wood units.

Outcomes

As seen in Table 2, the 2 groups differed for inotropic score [median (interquartile range): 71.8 (31.8–127) vs 35.7 (17.5–60.2), P < 0.001] (Fig. 1), rejection score [0.14 (0.05–0.25) vs 0.08 (0.0–0.25), P = 0.036] (Fig. 2), and for 30-day mortality rate (14.7% vs 2.5%, P = 0.0068). The groups were similar for primary graft dysfunction requiring postoperative ECMO, immediately at the end of the surgery or within the first 24 h (16.0% vs 16.5%, P = 0.94). The microscopic analysis of the first 3 endomyocardial biopsies revealed specific histological features of subendothelial ischaemia in 13 (17.3%) patients of the St-Thomas group and 3 (3.8%) in the Custodiol group (P = 0.006). One year after HTx, there was no significant difference between groups regarding the cardiac allograft vasculopathy.

Table 2:

Comparison of outcomes according to preservation solution

Outcome	St-Thomas	HTK-Custodiol	P-value
	N = 75	N = 79
Inotropic score	71.8 (31.8–127)	35.7 (17.5–60.2)	<0.001
Intra/postoperative ECMO	12 (16)	13 (16.5)	0.94
Rejection score	0.14 (0.05–0.25)	0.08 (0.0–0.25)	0.036
30-Day mortality	11 (14.7)	2 (2.5)	0.007

Summary statistics are presented as median (IQR) or number (%).

ECMO: extracorporeal membrane oxygenation; IQR: interquartile range.

Figure 1:

Distribution of the inotropic score in St-Thomas (N = 66) and HTK-Custodiol (N = 74) solution groups.

Figure 2:

Distribution of the rejection score in St-Thomas (N = 62) and HTK-Custodiol (N = 69) solution groups.

Cardiac preservation solution and inotropic score

Linear regression of log-transformed inotropic scores on CPS confirmed that scores were lower for HTK-Custodiol compared to St-Thomas solution (regression coefficient: −0.69, 95% CI: −1.0 to −0.38, P < 0.001). The significant relationship between inotropic score and CPS remained unchanged after adjusting for any of the patient characteristics, even for ischaemic time and CPB time both positively associated with the inotropic score (data not shown). Multiple linear regression confirmed that, when combined with ischaemic time (0.0052, 95% CI: 0.0011–0.0093, P = 0.017) and log-transformed CPB time (0.71, 95% CI: 0.16–1.3, P = 0.015), the preservation solution remained significantly related to the inotropic score (−0.60, 95% CI: −0.99 to −0.21, P = 0.003) (Table 3).

Table 3:

Relationship between cardiac preservation solution and inotropic score^a adjusted for ischaemic time and cardiopulmonary bypass time as derived by multiple linear regression

Covariate	Regression (95% CI)	P-value
CPS (Custodiol versus St-Thomas)	−0.60 (−0.99 to −0.21)	0.003
Ischaemic time (min)	0.0052 (0.0011 to 0.0093)	0.017
CPB time (min)a	0.71 (0.16 to 1.3)	0.015

^aLog-transform.

CI: confidence interval; CPB: cardiopulmonary bypass; CPS: cardiac preservation solution; SE: standard error.

Cardiac preservation solution and rejection score

The overall distribution of the rejection score could not be normalized, therefore the 131 patients with a rejection score were classified into 3 categories as follows: 41 (31.3%) had a score equal to 0, 49 (37.4%) had a score between 0 and 0.2, and 41 (31.3%) has a rejection score >0.2. Ordinal logistic regression confirmed that the rejection score was significantly impacted by CPS in favour of Custodiol (OR = 0.41, 95% CI 0.24–0.86, P = 0.016). No patient-related characteristics was associated with the rejection score, except renal glomerular function (N = 109 patients; OR = 0.979, 95% CI 0.961–0.999, P = 0.036). The effect of CPS on the rejection score remained significant after adjusting for any of the patient-related characteristics but only a tendency remained for renal glomerular function (P = 0.098) (data not shown).

Cardiac preservation solution and 30-day mortality

Overall, 13 (8.4%) died within 30 days in the patient series, significantly more in the St-Thomas group than in the Custodiol group as mentioned above (OR = 6.62, 95% CI: 1.41–30.9, P = 0.016). None of the other recipient-related preoperative or intraoperative characteristics was related to 30-day mortality rate (Table 4). CPS remained associated with 30-day mortality after adjusting for any of these covariates. The E-value to assess the potential effect of non-controlled confounders was 12.7 (lower limit 2.18) confirming the strong association of CPS and 30-day mortality. Of note, however, the association between CPS and 30-day mortality vanished (P = 0.86) when inserting the outcome variable log (VIS) in the logistic regression, emphasizing the strong relationship between CPS and VIS.

Table 4:

Relationship between 30-day mortality rate and each recipient-related characteristics adjusted for cardiac preservation solution as derived by logistic regression analysis

Risk factor	OR (95% CI)	P-value
Recipient-related preoperative
Age (years)	1.03 (0.97–1.08)	0.32
Gender (male versus female)	1.55 (0.33–7.37)	0.58
Aetiology (ischaemic versus other)	1.16 (0.37–3.62)	0.80
VAD	1.10 (0.34–3.55)	0.87
Diabetes	0.27 (0.034–2.18)	0.22
RF (ml/min/1.73 m2)	0.99 (0.95–1.03)	0.65
Ejection fraction (%)	1.002 (0.96–1.04)	0.94
VO2 max (ml/min/kg)	0 0.90 (0.75–1.09)	0.28
PVR (WU)	1.10 (0.65–1.88)	0.72
Waiting time on list (days)a	1.03 (0.67–1.59)	0.89
Recipient-related intraoperative
Ischaemic time (min)	1.00 (0.99–1.02)	0.61
CPB time (min)a	4.06 (0.85–19.3)	0.078
CPS (Custodiol versus St-Thomas)	0.15 (0.032–0.71)	0.016

^aLog-transform.

CI: confidence interval; CPB: cardiopulmonary bypass; CPS: cardiac preservation solution; OR: odds ratio; PVR: pulmonary vascular resistance; RF: renal function; VAD: ventricular assist device; VO₂ Max: maximal oxygen consumption; WU: Wood units.

Cardiac preservation solution and midterm survival

The follow-up for HTK-Custodiol patients was necessarily shorter than for St-Thomas patients (3.1 ± 1.5 vs 7.0 ± 3.9 years). Globally, 26 (16.9%) patients died, 21 in the St-Thomas group and 4 in the Custodiol group. The Kaplan–Meier survival functions of both groups (Fig. 3) differed significantly (log-rank test, P = 0.001). Cox regression analysis applied to each patient-related characteristic showed than CPS was the only significant factor affecting overall survival (HTK-Custodiol versus St-Thomas: hazard ratio = 0.20, 95% CI 0.069—0.60, P = 0.004) (Table 5). The impact of CPS on midterm survival remained unchanged after adjusting for any of the other patient-related factors.

Figure 3:

Kaplan–Meier post-transplant survival curves with censoring marks and 95% confidence limits in St-Thomas and HTK-Custodiol solution groups.

Table 5:

Relationship between overall survival and each recipient-related characteristics adjusted for cardiac preservation solution as derived by Cox regression analysis

Risk factor	HR (95% CI)	P-value
Recipient-related preoperative
Age (years)	1.03 (0.99–1.07)	0.11
Gender (male versus female)	1.38 (0.47–4.00)	0.56
Aetiology (ischaemic versus other)	0.95 (0.43–2.06)	0.89
VAD	0.97 (0.43–2.18)	0.94
Diabetes	0.63 (0.22–1.82)	0.39
RF (ml/min/1.73 m2)	0.998 (0.971–1.03)	0.90
Ejection fraction (%)	1.01 (0.98–1.04)	0.46
VO2 max (ml/min/kg)	0 0.95 (0.86–1.06)	0.36
PVR (WU)	1.07 (0.77–1.50)	0.68
Waiting time on list (days)a	0.92 (0.71–1.20)	0.55
Recipient-related intraoperative
Ischaemic time (min)	1.00 (0.99–1.01)	0.94
CPB time (min)a	2.16 (0.75–6.26)	0.16
Preservation solution (Custodiol versus St-Thomas)	0.20 (0.069–0.6)	0.004

CI: confidence interval; CPB: cardiopulmonary bypass; HR: hazard ratio; PVR: pulmonary vascular resistance; RF: renal function; VAD: ventricular assist device; VO₂ max: maximal oxygen consumption; WU: wood units.

^aLog-transform.

DISCUSSION

More than 50 years after the first human HTx by Christian Barnaard, HTx remains the preferred surgical option for selected patients with end-stage heart disease. The fact that the number of patients on waiting list and the duration of their HTx candidacy are continuously growing in Europe and the USA is an indirect sign of this trend [1, 7, 8].

Despite major recent progresses in the field of HTx, organ preservation remains imperfect and still impacts patients’ survival and outcomes [3]. The ex vivo period is the vulnerable stage during which the organ can undergo cellular damage that is further compounded by reperfusion injury after the implantation. The goal during procurement and preservation is to minimize these injuries and maintain the viability of the organ until its implantation in the recipient. Rapid diastolic cardiac arrest and subsequent cold ischaemic storage (at 4°C) are the 2 cornerstones of the cardiac procurement technique. Diastolic cardiac arrest preserves adenosine triphosphate (ATP) levels by comparison to ischaemic myocardial contracture [9], and cooling down the organ to 4°C results in a 10- to 12-fold decrease in metabolic demand. However, the persistence of a level of metabolism at 5–10% of normal values explains why cooling alone does not prevent all cellular damages [10]. During cold ischaemic storage, the only source of energy for the graft is anaerobic glycolysis, but the enzymes involved in this process are inhibited by the acidosis resulting from the ischaemia. Therefore, it is necessary to use a CPS containing buffers to maintain the cellular pH stable and allow a minimal ATP production [11].

CPSs are classified as intracellular or extracellular according to their concentration in sodium and potassium. Intracellular CPSs contain high potassium and low sodium and tend to be like the intracellular milieu. As a result, they limit the movement of ions and water across the cell membrane. Extracellular CPSs contain low potassium and were initially developed to prevent hyperkalemia related to the infusion of intracellular CPSs [12]. However, this classification remains rather artificial and subjective given that each CPS is best defined by its own ionic concentration and mostly by the residual osmotic space for the addition of other substances. These other substances could reduce intra- and extracellular oedema, limit intracellular acidosis, reduce reactive oxygen species (ROS) generation, and increase ATP production. All these factors tend to decrease the myocardial injury and thus improve the outcomes after HTx [13].

HTK-Custodiol is a hyperpolarizing solution with low sodium concentration that allows a large osmotic space as well as the addition of numerous other highly concentrated substances [14]. Among these substances, there is a high concentration of histidine/histidine hydrochloride intracellular buffering system, which enhances buffering capacity during ischaemic induced acidosis; amino-acid tryptophan alpha ketoglutarate, which protects cell membrane as a substrate for anaerobic metabolism; and mannitol, which is an osmotic agent that helps reducing cellular and tissue oedema. It is also an excellent scavenger of ROS [13, 15]. HTK-Custodiol has also been shown to maintain high levels of intra-cellular ATP after reperfusion and this is known to be directly correlated with low output syndrome, which usually develops a few hours after surgery and is the result of myocardial oedema during the ischaemic phase. The latter decreases coronary blood flow and thus intra-cellular ATP levels [14].

St-Thomas solution is an extracellular solution, which provides a rapid diastolic cardiac arrest by high potassium and magnesium concentration as well as by the membrane’s stabilizing effect of procaine hydrochloride. Cellular oedema is reduced by the extracellular sodium concentration, procaine, and a variable concentration of bicarbonates [16]. The increase in extracellular potassium concentration causes a progressive depolarization of the membrane potential for each level of potassium concentration. Solutions with a high concentration of potassium, such as St-Thomas, are however known to cause toxicity to the vascular endothelium. Carpentier was the first to demonstrate reduced viability and function of endothelial cells after exposure to high potassium concentration [17, 18]. The endothelium is however important as it locally regulates coronary perfusion and cardiac function through the secretion of nitric oxide and vasoactive peptides. Therefore, after administration of a high potassium concentration solution, endothelial dysfunction occurs, which could lead to myocardial dysfunction [19].

Regarding the buffering system, St-Thomas solution contains only extracellular buffers, which are less effective than the intracellular buffers used in HTK-Custodiol and other CPS in preventing intracellular oedema [20, 21]. Although St-Thomas is beneficial and still widely used in non-transplant cardiac surgery, our study, like others [13, 20], demonstrates that using St-Thomas solution leads to worse immediate outcomes after HTx, which likely explains its overall decreasing use. Concerning current trends in CPS use, most European centres moved from St-Thomas solution to HTK-Custodiol after 2010, and in the USA, in the past years, nearly half of the grafts were stored in the University of Wisconsin solution, one-fourth in Celsior and one-fourth in HTK-Custodiol [13].

Another salient element arising from our study is the difference in rejection score in favour of the Custodiol group, which to our knowledge, has not been described before. This could be interpreted as a reflection of improvement of the overall HTx patient care [21], given that the same trend has been observed in several other European countries during the last decades and seems to be related to the improvement of the immunosuppression monitoring [21, 22]. However, over the whole duration of our study, no changes in the immunosuppression protocol or its monitoring occurred. We therefore suggested that the integrity of the endothelial cells of the graft could be compromised by the different preservation and storage techniques and in particular by the type of CPS used. Indeed, it is now well known that endothelial cell damage leads to increased capillary permeability, cellular and tissue oedema, vasospasm and microvascular hypoperfusion [23, 24]. As endothelial cell function is directly correlated to cardiomyocyte function, all these elements can lead to primary graft dysfunction [25, 26]. Moreover, different studies confirm that preservation related injuries in HTx can be the cause of early complications but also of late events such as graft rejection and chronic transplant arteriopathy [27, 28].

To confirm our hypothesis, we reviewed the anatomopathological reports of the first 3 endomyocardial biopsies for each patient, in both groups. This time, we were interested not only in the overall rejection score but also in the microscopic analysis when it showed typical lesions of ischaemia reperfusion phenomena. Specifically, the lesions found are infiltrates of mononuclear cells and granulocytes located in the endothelial layer and associated with interstitial oedema. These lesions are specifically different from rejection lesions and are interpreted as typical of ischaemia-reperfusion phenomena by our pathologists.

Interestingly, we found that there were significantly more of these specific histological features in the St-Thomas group than in the Custodiol group.

Several factors can explain these endothelial lesions during the graft harvesting and storage process. At first, the duration of ischaemia can directly affect the viability of endothelial cells through different pathways. These include reduced protein synthesis and ATP levels [28], increased anaerobic metabolism, and both intracellular and extracellular acidosis [25]. Under these conditions, the endothelium releases large quantities of proinflammatory chemoattractant cytokines (IL-1α, IL-8) and the availability of antioxidants is reduced [26]. All these elements lead to potassium efflux with membrane depolarization, cellular swelling, alteration of the endothelial barrier, and tissue oedema. This in turn leads to abnormalities in the distribution of CPS but also in blood flow at reperfusion, which aggravates the phenomenon [13].

Second, reperfusion is accompanied by a real burst of ROS which occurs only 15 s after the onset of the reperfusion [15]. This increases the endothelial lesions and the previously mentioned inflammatory reaction. Usually 2–3 h after reperfusion, activated neutrophils adhere to the endothelium, release large amounts of free radicals resulting in loss of endothelial barrier function, tissue oedema and a functional impairment of both endothelial cell and cardiomyocytes [16].

It is likely that the difference in outcomes obtained, especially regarding the rejection score in favour of the Custodiol group, is explained by the response of the 2 CPSs to various lesional factors affecting the endothelium and consequently the cardiomyocytes, during graft harvesting and preservation.

As mentioned above, the St-Thomas solution is a high concentration of potassium solution, and it has been known since the 1980s and Carpentier [17] that solutions of this type induce vasoconstriction and an impairment of the endothelial function with a decrease in nitric oxide release and other factors including prostacyclin, endothelium derived hyperpolarization factor and adenosine [13]. In addition, potassium-induced depolarization is known to promote platelet adhesion, neutrophil activation, inflammation, and ROS generation, which could explain our results. On the contrary, HTK-Custodiol is a low concentration potassium solution that contains different substances such as histidine, ketoglutarate, tryptophan and mannitol, whose role is to counteract the deleterious effects on the endothelium and the myocardium. Those differences in chemical composition may explain our results.

It is important to note that other studies have not found results similar to ours. For example, the study by Cannata et al. [29] reported retrospectively133 HTx with 3 different CPSs (Custodiol versus St-Thomas versus Celsior). Custodiol was mainly used. Outcomes included intraoperative biventricular dysfunction requiring ECMO and in-hospital mortality. There was no difference between groups. In comparison, our study confirms that there is no difference in biventricular dysfunction, but our mortality differs between the groups. However, our study was designed differently, the aim being to determine the patient's postoperative condition other than only by mortality (inotropic score, biventricular dysfunction) and to see whether the advantage of Custodiol based on its chemical composition is confirmed at histological level (rejection score, ischaemia-reperfusion lesions).

Another interesting study written by Karduz et al. [30] aimed at evaluating the effect of HTK-Custodiol, St-Thomas and del Nido solutions functionally and biochemically in a rat model of donor heart. Custodiol administration led to reduced myocardial contraction, decreased ATP level, increased TNF-α and increased troponin-I levels. The results of this observational study run counter to several other studies on humans [14–16], especially regarding the ATP levels. However, the study is well conducted, and the results are very interesting.

It is likely that in the future, further studies, especially randomized control trials, could be necessary to confirm our data.

Limitations

This retrospective longitudinal study longitudinal of HTx patients suffers from the shortcomings of all retrospective observational studies, including selection biases, reliability, quality, and completeness of data collected from patient electronic records, even though a special effort was made in this study to eliminate erroneous data entry and avoid as much as possible missing data. In this respect, the data collection was complete, and the outcome measures were confirmed in our local database as well as in the Swiss Death registry and the Swiss Cohort Study.

CONCLUSION

In our regional cohort of consecutive HTx recipients in pre- and post-transplant follow-up by 2 different tertiary centres, we observed that the use of HTK-Custodiol as myocardial protection during heart procurement leads to improved outcomes after HTx, including postoperative inotropic score, 30-day mortality, mid-term survival, rejection score and presence of specific ischaemia–reperfusion lesions.

Even though the present study is not a head-to-head comparison, our results suggest the superiority of HTK-Custodiol over the St-Thomas solution, in the context of very few differences in the baseline patient’s characteristics, an unchanged pre- and post-transplant follow-up and an unchanged national donor heart allocation system during the study period. Further studies, especially randomized control trials, are necessary to confirm these data.

Conflict of interest: none declared.

Glossary

ABBREVIATIONS

ATP

Adenosine triphosphate

CI

Confidence interval

CPB

Cardiopulmonary bypass

CPS

Cardiac preservation solution

ECMO

Extracorporeal membrane oxygenation

HTx

Heart transplantation

OR

Odds ratio

ROS

Reactive oxygen species

VIS

Vasoactive inotropic score

DATA AVAILABILITY

The data that support the findings of this study are available on request from the corresponding author.

Author contributions

Filip Dulguerov: Conceptualization; Data curation; Investigation; Methodology; Validation; Visualization; Writing—original draft; Writing—review & editing. Tamila Abdurashidowa: Data curation. Emeline Christophel-Plathier: Data curation; Software; Validation. Lucian Ion: Conceptualization; Data curation; Validation. Ziyad Gunga: Validation. Valentina Rancati: Conceptualization; Validation. Patrick Yerly: Validation. Piergiorgio Tozzi: Supervision; Validation. Adelin Albert: Conceptualization; Formal analysis; Validation; Writing—review & editing; Statistics. Zied Ltaief: Validation; Writing—review & editing. Samuel Rotman: Data curation; Formal analysis; Methodology. Philippe Meyer: Validation; Writing—review & editing. Karl Lefol: Conceptualization; Data curation; Validation; Visualization. Roger Hullin: Conceptualization; Validation; Writing—review & editing. Matthias Kirsch: Conceptualization; Methodology; Supervision; Validation; Writing—review & editing.

Reviewer information

Interdisciplinary CardioVascular and Thoracic Surgery thanks Kartik Patel, Bilal Haneef Kirmani, Roman Gottardi and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.