Cardioprotection with esmolol-based cardioplegia for non-infarcted and infarcted rat hearts

Abstract

 

Open in new tabDownload slide

OBJECTIVES

Esmolol-based cardioplegic arrest offers better cardioprotection than crystalloid cardioplegia but has been compared experimentally with blood cardioplegia only once. We investigated the influence of esmolol crystalloid cardioplegia (ECCP), esmolol blood cardioplegia (EBCP) and Calafiore blood cardioplegia (Cala) on cardiac function, metabolism and infarct size in non-infarcted and infarcted isolated rat hearts.

METHODS

Two studies were performed: (i) the hearts were subjected to a 90-min cardioplegic arrest with ECCP, EBCP or Cala and (ii) a regional myocardial infarction was created 30 min before a 90-min cardioplegic arrest. Left ventricular peak developed pressure (LVpdP), velocity of contractility (dLVP/dt_max), velocity of relaxation over time (dLVP/dt_min), heart rate and coronary flow were recorded. In addition, the metabolic parameters were analysed. The infarct size was determined by planimetry, and the myocardial damage was determined by electron microscopy.

RESULTS

In non-infarcted hearts, cardiac function was better preserved with ECCP than with EBCP or Cala relative to baseline values (LVpdP: 100 ± 28% vs 86 ± 11% vs 57 ± 7%; P = 0.002). Infarcted hearts showed similar haemodynamic recovery for ECCP, EBCP and Cala (LVpdP: 85 ± 46% vs 89 ± 55% vs 56 ± 26%; P = 0.30). The lactate production with EBCP was lower than with ECCP (0.6 ± 0.7 vs 1.4 ± 0.5 μmol/min; P = 0.017). The myocardial infarct size and (ECCP vs EBCP vs Cala: 16 ± 7% vs 15 ± 9% vs 24 ± 13%; P = 0.21) the ultrastructural preservation was similar in all groups.

CONCLUSIONS

In non-infarcted rat hearts, esmolol-based cardioplegia, particularly ECCP, offers better myocardial protection than Calafiore. After an acute myocardial infarction, cardioprotection with esmolol-based cardioplegia is similar to that with Calafiore.

INTRODUCTION

Myocardial ischaemia can be treated with either percutaneous coronary intervention or coronary artery bypass grafting [1]. Coronary artery bypass grafting is usually carried out using a heart-lung machine and cardioplegic arrest [2]. The most widely used cardioplegia solutions can be categorized into asanguinous crystalloid solutions and blood cardioplegia (BCP) solutions [3–5]. Particularly, Calafiore BCP (Cala), being a very simple formulation, has gained much popularity [6, 7]. Since depolarized arrest induced by hyperkalaemia also generates some detrimental effects on ischaemic myocardium, like ionic imbalance [8], maintained energy utilization [9] and persisting ischaemia–reperfusion injury [10], there is interest in alternative concepts, which maximize the functional recovery and minimize the extent of ischaemia–reperfusion injury.

Experimental studies [11–13] have demonstrated that high concentrations (∼1 mmol/l) of esmolol induce diastolic cardiac arrest [12]. When added to an oxygenated perfusate as a multi-dose infusion, esmolol provided superior cardioprotection to isolated rat hearts when compared with cross-clamp fibrillation [12] or St. Thomas' Hospital cardioplegia, a hyperkalaemic crystalloid cardioplegia (CCP) [11]. Therefore, our hypothesis was that esmolol-based cardioplegia offers better myocardial protection than regular hyperkalaemic solutions.

Experimental data related to the intermittent administration of esmolol-based cardioplegia with prolonged infusion intervals would be valuable because multi-dose esmolol infusion had shown to be more protective than continuous esmolol infusion [12], even for prolonged ischaemia periods [11] and because it is more applicable in clinical settings. Therefore, we compared the recovery from esmolol-based versus hyperkalaemic cardioplegia in blood-perfused rat hearts.

Cardiac surgery in acute myocardial infarction is a clinical setting where protection against ischaemia–reperfusion injury is of utmost importance. In particular, hearts with ischaemic damage benefit from the arresting, buffering and energy-balancing effects of cardioplegia. This can lead to better recovery from cardiac surgery in patients with acute myocardial infarction due to better cardiac function after surgery, if the most effective cardioplegia is used. Thus, since hearts with acute myocardial infarction may be more prone to ischaemia–reperfusion damage during cardiac surgery than stable hearts with intact coronary circulation, we also compared cardioplegia solutions in hearts with acute myocardial infarctions in an ischaemia–reperfusion setting.

MATERIALS AND METHODS

Animals

The hearts of 42 adult (age: 3–4 months, mean body weight: 517 ± 77 g, mean heart weight: 2.3 ± 0.28 g) male Wistar rats (Janvier, St. Berthevin, France) were used for these experiments, which were registered with the regional authorities (Giessen, Germany) and conformed to the German Animal Protection Law.

Experimental model

A Langendorff apparatus (IH-SR Type 844, Hugo Sachs, Hugstetten, Germany) was filled with freshly prepared, filtered, heparinized bovine erythrocyte concentrates diluted in Krebs–Henseleit buffer (KHB) and warmed to 36°C (Fig. 1). The perfusate was gassed with a mixture containing 5% CO₂ and 95% O₂ (Carbogen, Linde, Pullach, Germany) to obtain a physiological pH (7.4) before utilization. To mimic the clinical situation in which warm (36°C) extracorporeal circulation is employed, the heart was continuously exposed to air in a heated (36°C) chamber. This model was previously described in more detail [14, 15].

During the experiment, functional parameters were recorded: heart rate, coronary flow (CF), left ventricular peak developed pressure (LVpdP), velocities of contractility (dLVP/dt_max) and relaxation (dLVP/dt_min). The isovolumetric measurement of left ventricular (LV) performance was carried out using a compliant latex balloon, inserted into the LV and filled with saline solution. The end-diastolic pressure was adjusted to 10–12 mmHg during the stabilization period and held constant afterwards. LV performance was assessed by measuring the LV systolic pressure (LVSP) and LV end-diastolic pressure (LVEDP; LVpdP = LVSP − LVEDP). First derivatives of LVSP (dLVP/dt_max, dLVP/dt_min) were recorded by a transducer amplifier module (TAM-A; Harvard Apparatus, Holliston, MA, USA) and calculated by the software ‘Isoheart’ (Harvard Apparatus; Fig. 1). All results were calculated as percentage of baseline (BL) to correct for initial differences. At BL, the haemodynamic performance and the CF in hearts without infarction and with later infarction were not different between the groups (Tables 1 and 2). The BL measurements were taken after the stabilization period and before ligature of the left anterior descending coronary artery (LAD) or aortic clamping. During reperfusion, the oxygen and lactate contents in the cardiac inflow (aortic block) and outflow (effluate collector) were measured. RapidLab 348 (Siemens, Eschborn, Germany) was used for oxygen saturation (sO₂) and oxygen partial pressure (pO₂) measurements and Radiometer ABL800 Flex (Radiometer, Krefeld, Germany) for lactate measurements. Myocardial oxygen consumption (MV_O2) was calculated according to Fick’s law [16]: MV_O2 = (c_artO₂ − c_venO₂) × (CF/heart weight) × 100 with c_xO₂ = s_xO₂ × cHb × 1.34 + p_xO₂ × 0.0031, where c_xO₂ = perfusate oxygen; s_xO₂ = perfusate oxygen saturation; cHb = perfusate haemoglobin concentration; and p_xO₂ = perfusate oxygen partial pressure. Myocardial lactate production (MV_Lac) was calculated using: MV_Lac = CF × (Lac_art − Lac_ven) × 100, where Lac_x = perfusate lactate content.

Perfusion medium, cardioplegic solutions

For the perfusate, with a haemoglobin concentration of 8.2 ± 1.5 g/l, 250 ml of KHB [17] was mixed with 5 g bovine serum albumin, 1 IU insulin and 250 ml bovine erythrocyte concentrate. Esmolol (Brevibloc; Baxter, Unterschleißheim, Germany) crystalloid cardioplegia (ECCP) was prepared by diluting with KHB to a concentration of 1.0 mmol/l esmolol. This solution was infused at 32°C for 3 min. Esmolol blood cardioplegia (EBCP) was prepared by diluting esmolol with a 1:1 mixture of blood perfusate and KHB to a concentration of 1.0 mmol/l esmolol and was infused at 36°C for 3 min. The initial hyperkalaemic BCP was induced with 14 ml of blood and 150 µl of Cala solution (2.9 × 10⁻³mol potassium chloride, 1.3 × 10⁻³mol magnesium sulphate; Cala) with a 150-µl Cala bolus for 2 min at 36°C. Subsequent doses were 10 ml of blood and 225 µl of Cala for 2 min. Each cardioplegia solution was oxygenated until the application. The flow rates of the cardioplegia solutions were adjusted to BL CF with the addition of a safety margin of 30% [total volume = x min × (BL CF + 30%)].

Perfusion protocols

We applied an experimental model simulating the clinical setting of a patient having heart surgery using extracorporeal circulation with aortic clamping and cardioplegic arrest. First, we simulated a setting with a patient having a cardiac procedure (Fig. 2A). Afterwards, we simulated a patient with an acute myocardial infarction having cardiac bypass surgery (Fig. 2B).

Step 1: myocardial protection during cardiac arrest without myocardial infarction

After excision, the hearts were mounted onto the Langendorff apparatus and the perfusion was initiated. After the hearts were haemodynamically stable for up to 30 min, in the cardioplegic arrest group, the aorta (18 hearts, Fig. 2A) was ‘clamped’ by blocking the arterial blood flow in the aortic block. Afterwards, either ECCP, EBCP or Cala was administered in an antegrade fashion into the coronary arteries (Fig. 1). The cardioplegia application was repeated every 20 min until a clamping time of 90 min was reached (Fig. 2A). Then, the hearts were reperfused for 90 min further.

Step 2: myocardial protection during cardiac arrest after previous myocardial infarction

In the infarction group (24 hearts, Fig. 2B), a myocardial infarction was created by the ligation of the LAD for 30 min, as previously described [17, 18]. After 30 min of regional ischaemia of the LAD territory, the aorta was ‘clamped’ and the cardioplegic solutions were administered similarly (Fig. 2B). After 90 min of aortic clamping, the ligature was removed and the hearts were reperfused similarly.

Infarct size planimetry

After 90 min of reperfusion, the hearts of the infarction group were frozen at −20°C for 30 min. Subsequently, the hearts were cut into 16 ± 2 slices and incubated in 1.2% triphenyl tetrazolium chloride for 30 min at 38°C. Then, the slices were fixed in 7% formalin at room temperature overnight. Digital images were taken from both sides of the slices with a M60 microscope (Leica, Wetzlar, Germany) at 1.25-fold magnification. The Infarct size was determined by planimetry as the mean proportion of the infarcted area of the whole slice using Leica Application Suite LAS version 4.6 (Leica).

Electron microscopy

Two hearts in each of the cardioplegic arrest groups were fixed by vascular perfusion via the aortic cannula with 1.5% glutaraldehyde and 1.5% paraformaldehyde in 0.15 M HEPES buffer. After the storage at 4°C for at least 24 h, the LV including septum was sampled by systematic uniform random sampling [19]. The myocardial samples were osmicated, stained en-bloc with half-saturated uranyl acetate, dehydrated in an ascending acetone series and embedded in epoxy resin. Semi- and ultrathin sections were cut using an ultramicrotome and then stained with toluidine blue or lead citrate and uranyl acetate, respectively. Using morphometry, we estimated a cellular oedema index, which rises, if cardiomyocytes swell [20], and volume-to-surface ratio of mitochondria, which provides a suitable parameter for ischaemia-induced swelling of mitochondria [21].

Statistical analysis

All parameters were analysed using GraphPad Prism version 8 (GraphPad Software Inc., San Diego, CA, USA). Intergroup differences at different points in time were analysed using one-way analysis of variance. Tukey’s post hoc test was used, if any difference was observed. We did not correct for multiple testing for all of our tests and models. The data are shown as mean ± standard deviation. A statistical significance was assumed at a level of P < 0.05.

RESULTS

Step 1: myocardial protection during cardiac arrest without myocardial infarction

During aortic clamping, LVEDP, as a measure of ischaemic contracture, increased in Cala (289%), whereas with ECCP (61%) and EBCP (58%), it fell to <65% without showing intermittent or increasing contraction (Fig. 3C).

After 90 min of reperfusion, the haemodynamic function of hearts with both ECCP and EBCP recovered better than that of hearts with Cala: LVpdP, dLVP/dt_max and dLVP/dt_min showed better results (Table 1 and Fig. 3). CF recovery during reperfusion was higher in the ECCP (228%, P < 0.001) and EBCP (132%, P = 0.17) groups compared with Cala (65%) (Fig 3D). At 10-min reperfusion, the oxygen consumption of the ECCP (148%; P = 0.013) and EBCP groups (127%; P = 0.062; Table 2) was higher than within the Cala group (61%), whereas after 90-min reperfusion there was no statistically significant difference in this parameter (ECCP 100%, EBCP 87%, Cala 55%; P = 0.085; Fig. 3E). After 90-min reperfusion the lactate production in the ECCP group was nominally, but not statistically significant lower [0.67 µmol/(ml min)] than in the EBCP [1.72 µmol/(ml min)] and Cala groups [0.97 µmol/(ml min); Fig. 3F].

The electron microscopic analyses showed similar cellular oedema index in the Cala (0.40), EBCP (0.39) and ECCP (0.40) groups. Similarly, the mitochondrial swelling was not different (volume-to-surface ratio of mitochondria: Cala 0.35, EBCP 0.43, ECCP 0.40). Also, there was not a qualitative ultrastructural difference of cardiomyocytes between the analysed groups (Fig. 4).

Step 2: myocardial protection during cardiac arrest after previous myocardial infarction

During aortic clamping LVEDP increased in the Cala group (221%) while with ECCP (57%) and EBCP (73%), it fell to <75% without showing intermittent or increasing contraction (Fig. 5C).

After 90 min of reperfusion, the haemodynamic function of hearts treated with ECCP and EBCP recovered similar to the ones with Cala: LVpdP, dLVP/dt_max, dLVP/dt_min recovered slightly better with EBCP and ECCP, although there was no statistically significant difference (Table 2 and Fig. 5). The CF during reperfusion was higher in the ECCP (137%, P = 0.027) and EBCP (147%, P = 0.011) groups compared with Cala (62%) respectively (Fig 5D). The oxygen consumption of the ECCP (76%; P = 0.64) and EBCP groups (81%; P = 0.29) were higher compared with Cala (54%; Fig. 5E), although not statistically significant different. The lactate production in the ECCP [1.43 µmol/(ml min); P = 0.017] and Cala [1.04 µmol/(ml min); P = 0.26] groups was higher than with EBCP [0.58 µmol/(ml min); Fig. 5F].

The creation of a myocardial infarction was successful in each heart as documented by the size of infarction, which was evaluated after reperfusion. The myocardial infarct size (Fig. 6) was somewhat smaller in hearts with EBCP (15%) and ECCP (16%) than in the Cala (24%) group, although not statistically significant different (P = 0.21).

DISCUSSION

This study investigated the recovery of hearts in an in vitro rat heart model. Our primary results are that esmolol-based cardioplegia solutions are more protective than the widely used Calafiore. Also after acute myocardial infarction esmolol-based cardioplegia solutions provide the same myocardial protection as Calafiore.

In non-infarcted rat hearts, esmolol-based cardioplegia solutions, particularly ECCP, offered better myocardial protection. ECCP improved the haemodynamic function and stabilized the CF of the treated hearts during and after reperfusion better than Cala, although oxygen consumption and lactate production during and after reperfusion were similar. In contrast to our results, Fujii and Chambers [22] previously showed a similar recovery in ischaemic hearts treated with ECCP and with EBCP.

In our acute myocardial infarction model, esmolol-based cardioplegia solutions offered myocardial protection as good as that of Calafiore. The infarct size tended to be smaller in hearts treated with esmolol-based cardioplegia solutions, but this was not statistically significant. Bell et al. [23] assumed that 30–35 min of no-flow normothermic ischaemia is optimal for a myocardial infarction model utilizing the Langendorff apparatus because it results in ∼50% infarction within the myocardium at risk. We observed that all cardioplegia solutions showed equally good protection from myocardial injury (15–24% of the infarcted area). But the lactate production in ECCP-treated hearts was higher than in Cala-treated hearts during early reperfusion and higher than in EBCP-treated hearts during late reperfusion. Although haemodynamic measurements and myocardial oxygen consumption were similar between all cardioplegia types, the higher lactate levels suggest that ECCP might offer inferior cardioprotection in infarcted hearts.

Differences between the results of all groups could be explained by the different mechanisms of causing cardiac arrest. While Cala is a potassium-based solution, causing a depolarized diastolic arrest (membrane potential around −40 mV), esmolol is a short-acting β-blocker causing a diastolic arrest with nearly normal resting membrane potential (around −70 mV) [24]. At an extracellular potassium level of around 30 mmol/l, Ca²⁺ channels are activated and there is Ca²⁺ uptake into the myocytes, with a chance of Ca²⁺ overload, known to result in an increased myocardial oxygen consumption and ischaemia–reperfusion injury [3]. Contrarily, esmolol decreases myocardial oxygen consumption and improves oxygen usage in human patients [25].

BCP has been shown to improve the myocardial protection compared with CCP [26]. However, experimental or clinical data addressing specific erythrocytes effects and their interaction with infarcted myocardium are not available.

The cardioplegia temperature application might be another critical aspect: In a previous study of our group, we found warm Cala application to result in a higher LVdP and less cellular oedema [14]. Therefore, we used Cala and EBCP at 36°C. For ECCP, we chose a temperature of 32°C because Nishina and Chambers [27] showed this temperature as ideal for esmolol CCP and improved the cardioprotection compared to hyperkalaemic CCP. Also, Melendez et al. [28] examined esmolol in human whole blood at temperatures between 37 and 4°C and observed an increase in half-life of esmolol from 19 min at 37°C to 227 min at 4°C. It was assumed that the extended half-life could lead to persistent β-blockade that would either provide additional cardiomyocytes protection or lead to a difficulty in weaning after cardiopulmonary bypass [28].

Esmolol cardioplegia provides complete protection for extended periods of hypothermic [27] and normothermic [11] ischaemia. However, it does not appear that prolonged infusion intervals we used compared to previous studies [11–13] and higher esmolol concentration of 1 mmol/l caused any harm because there was good cardiac recovery. Even in the infarction setting where nearly 20% of the viable myocardium was lost, we observed a return of parameters to near-BL values.

Limitations

Our results were obtained in isolated healthy rat hearts. Thus, it is unclear whether the results can be transferred to humans with pre-existing heart disease.

The context of an organism is lacking, although we have tried to mimic the clinical setting regarding temperature, application and composition of all cardioplegia solutions. Also, the myocardial infarction was created in a perfused heart for only 120 min and not a working heart over a longer period of time.

We made every effort to occlude the LAD at the same level; however, length, size and distribution area of the LAD and its collaterals can vary, but we were not able to take this into account.

Clinically used approaches to deliver cardioplegia to the infarcted area (retrograde application, application via bypass grafts) could not be incorporated in this model.

This study is limited by the small sample size. Therefore, the results do not allow for definitive conclusions but should rather be interpreted as exploratory and hypothesis generating.

CONCLUSION

In non-infarcted isolated rat hearts, esmolol-based cardioplegia solutions, particularly ECCP, offer better myocardial protection than widely used Calafiore cardioplegia. After acute myocardial infarction, ECCP and EBCP solutions led to effective myocardial protection and provided myocardial protection which was as good as Calafiore.

Presented at the 49th meeting of German Society for Thoracic, Vascular and Cardiac Surgery, Wiesbaden, Germany, 3 March 2020.

ACKNOWLEDGEMENTS

We acknowledge the editorial assistance, formatting and language editing services of Elizabeth Martinson.

Conflict of interest: none declared.

Author contributions

Alexander B. Veitinger: Data curation; Formal analysis; Investigation; Visualization; Writing—original draft; Writing—review & editing. Audrey Komguem: Investigation. Lena Assling-Simon: Investigation. Martina Heep: Data curation; Investigation; Methodology. Julia Schipke: Data curation; Investigation; Methodology. Christian Mühlfeld: Data curation; Investigation; Methodology; Writing—review & editing. Bernd Niemann: Writing—review & editing. Philippe Grieshaber: Formal analysis; Writing—review & editing. Kerstin Boengler: Methodology; Writing—review & editing. Andreas Böning: Conceptualization; Methodology; Project administration; Resources; Supervision; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Wolfgang A. Goetz, Paulo Roberto B. Evora and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

 
• BL

  Baseline

 
• BCP

  Blood cardioplegia

 
• Cala

  Calafiore blood cardioplegia

 
• CF

  Coronary flow

 
• CCP

  Crystalloid cardioplegia

 
• EBCP

  Esmolol blood cardioplegia

 
• ECCP

  Esmolol crystalloid cardioplegia

 
• KHB

  Krebs–Henseleit buffer

 
• LAD

  Left anterior descending coronary artery

 
• LV

  Left ventricular

 
• LVpdP

  Left ventricular peak developed pressure

 
• LVEDP

  LV end-diastolic pressure

 
• LVSP

  LV systolic pressure