Abstract
Objective: Apoptotic cardiomyocyte death is induced during open heart surgery, but its determinants are poorly understood. Prolonged aortic clamping time is associated with adverse clinical outcomes. The purpose of this study was to determine whether occurrence of cardiomyocyte apoptosis is related to the duration of aortic clamping in experimental pig model of cardiac surgery with cardiopulmonary bypass. Methods: The pigs (mean weight 29 ± 1 kg) were randomly divided to undergo cardioplegic arrest for 60 (n = 4) or 90 (n = 4) min followed by reperfusion period of 120 min. Control group (n = 5) was connected to cardiopulmonary bypass for 120 min without cardioplegic arrest. Cardiomyocyte apoptosis was detected (TUNEL assay and immunohistochemical staining of active caspase-3) in left ventricular tissue samples obtained before ischemia and after the ischemia-reperfusion period. Results: Apoptotic cardiomyocytes were found in all samples obtained after cardioplegic arrest and cardiopulmonary bypass alone with the TUNEL assay. The amount of apoptosis after the 120 min of cardiopulmonary bypass alone in the control group was 0.006 ± 0.001%. Compared with this, cardiomyocyte apoptosis was increased after cardioplegic arrest. After 60 min of aortic cross-clamp the amount of apoptosis was 0.019 ± 0.004% (p = 0.031). After 90 min of aortic cross-clamp the amount was 0.042 ± 0.005% (p ≪ 0.001) being significantly higher than after 60 min (p = 0.001). Aortic cross-clamp of 90 min also resulted in a detectable increase in caspase-3 activation when compared with controls. Conclusions: The occurrence of cardiomyocyte apoptosis increases with prolonged aortic clamping time during open heart surgery.
1 Introduction
In clinical practice aortic clamping time can be extended up to several hours, although many problems occur much earlier than that. Prolonged aortic clamping time is associated with increased mortality [1], lowered cardiac contractility [1,2], frequent arrhythmias [3,4] and increased levels of circulating troponin-T (TnT) [5].
Apoptosis is morphologically distinct type of cell death, which is genetically controlled and requires energy [6,7]. Cardiomyocyte apoptosis is associated with several cardiac diseases such as acute myocardial infarction, heart failure and myocarditis [8–10]. In addition, it has been shown to take part in ischemia-reperfusion injury in open heart surgery in animal models and in human [11,12]. Whether the amount of cardiomyocyte apoptosis correlates with aortic cross-clamp time during open heart surgery, is not known. Thus, the purpose of this study was to determine whether occurrence of cardiomyocyte apoptosis is related to the duration of aortic clamping in experimental pig model of cardiac surgery with cardiopulmonary bypass (CPB).
2 Materials and methods
2.1 Study protocol
In this study we used experimental model of open heart surgery with CPB in pig. To compare the effect of different aortic cross-clamp times, the animals (Finnish landrace pigs, n = 13, age of 12 weeks, mean weight of 29 ± 1.0 kg) were randomly divided in three groups. The first group underwent cardioplegic arrest for 60 min (n = 4) and the second group for 90 min (n = 4). The cardiac arrest was followed by a reperfusion period of 120 min (n = 8). In the control group (n = 5) the animals underwent CPB without cardioplegic arrest for 120 min.
2.2 Experimental model
The animals were premedicated with 34 mg/kg (1000 mg) of S-ketamine (Ketanest-S®, Pfizer AB, Taby, Sweden) intramusculary (i.m.). A peripheral vein in the ear was cannulated and 10 mg of diazepam (Stesolid Novum®, A/S Dumex, Denmark) was given intravenously (i.v.). For intubation, the trachea was surgically exposed and the intubation was performed openly. Before intubation, the animals received 4 mg i.v. bolus of pancuronium (Pavulon®, Organon, The Netherlands). The animals were connected to respirator (Respiration Pump model 607, Harvard Apparatus, Millis, MA, USA) and were ventilated with room air. The tidal volume (450 ml/min) and the frequency (16–18 min−1) were set according to blood gas analysis (ABL 50, Radiometer A/S, Copenhagen, Denmark). The anaesthesia was maintained with continuous intravenous infusion of S-ketamine (0.27 mg/kg min) and pancuronium (0.007 mg/kg min). In addition, the animals received solution containing succinylated gelatin (Gelofusine®, B. Braun Melsungen AG, Melsungen, Germany, 4%, 1.8 mg/kg min) throughout the experiment. ECG, heart rate and blood pressures were monitored during the experiment (Uniflow TM pressure monitoring kit 43-600F, Baxter, Uden, Holland and Olli 530, Kone Oy, Espoo, Finland). Internal carotid artery and both external jugular veins were cannulated for monitoring and blood sampling. The heart was exposed by medial sternotomy and the pericardium was opened and lifted. Before manipulating the heart, the animals received a 100 mg slowly infused bolus of lidocain hydrocloride (Xylocard® Hässle Läkemedel AB, Mölndal, Sweden) to reduce the amount of cardiac arrhythmias. The animals were surgically prepared for the CPB by placing purse string sutures on ascending aorta and right atrium. Before the cannulation, the animals received 3.4 mg/kg heparin (Heparin®, Lövens, Ballerup, Denmark). On the venous side we used a two-stage cannula. The animals were connected to CPB and the flow in the aortic line was adjusted to 2.5–3 l/min (85–100 ml/kg min) according to blood gas analysis and venous blood oxygen saturation percentage (Oxysat® Meter SM-0200, Baxter, Bentley, Irvine, CA, USA). A pediatric membrane oxygenator (Midiflo Pediatric D705, Dideco, Mirandola, Italy) was used during the CPB. The oxygenator was primed with 1500 ml of fresh pig blood containing 3800 mg of sodium citrate and 50 mg of heparin. The left hemiazygos vein draining to coronary sinus was ligated. After the aorta was clamped, initial 500 ml of cold (+5 °C) crystalloid modified St. Thomas Hospital N:o II cardioplegic solution (concentrate including electrolytes: Ca2+ 1.12 mmol, K+ 9.99 mmol, Mg2+ 7.99 mmol, Cl− 28.23 mmol, 10 ml of concentrate blended in 500 ml of 0.9% NaCl-solution) was infused in 18 G cannula (Venlon®2, Viggo AB, Helsingborg, Sweden) placed in the proximal ascending aorta. An additional dose of cardioplegia (250 ml) was given every 30 min. After the declamping of the aorta, cardioversion was performed in case of ventricular fibrillation. The animals were weaned from the CPB after 17 ± 3 min (p = 0.725, 60 min vs 90 min) after the cardiac arrest and the effect of heparin was antagonized with protamine (Protaminsulfat Leo®, LEO Pharma A/S, Ballerup, Denmark). After the cardiac arrest the animals underwent a reperfusion period of 120 min. The animals were sacrificed with a potassium chloride injection given directly to the left atrium after the experiment. The study protocol was approved by the Ethical Committee for Animal Research at the University of Turku.
2.3 Myocardial samples
In order to detect apoptosis, transmyocardial samples were collected during the experiment from the left ventricle. A needle biopsy was taken before ischemia (tru-cut needle, Pharmaseal®, Allegiance Healtcare Corporation, McGaw Park, IL, USA). After the ischemia-reperfusion time at the end of the experiment averagely 1 cm3 of the ventricular wall was obtained. Samples were fixed in neutral buffered formalin overnight, embedded in paraffin and cut at 4 μm sections for analysis of apoptosis.
2.4 Detection of apoptosis
Cardiomyocyte apoptosis was detected using TUNEL (terminal transferase mediated ddUTP nick end labelling) assay as previously described [7,9]. In brief, paraffine embedded myocardial sections were heated in sodium citrate solution and digested with proteinase-K to expose DNA. The DNA strand breaks were then labelled using terminal transferase with digoxigenin-conjugated ddUTP and visualized using alkaline phosphatase immunohistochemistry. The assay was standardized with the use of serial sections treated with DNaseI as positive control for apoptosis.
The activation of apoptosis-specific caspase-3 with an antibody specific for large (17–20 kDa) fragments of cleaved caspace-3 (Cell Signaling Technology, Beverly, MA, USA) was also analysed to detect myocardial apoptosis. Deparaffined hydrated sections were treated in microwave oven for 10 min in sodium citrate buffer (pH 6.0) to expose antigens, followed by inhibition of endogenous peroxidase activity by 1% H2O2. The primary antibody (1:100) was visualized using Vectastain ABCElite Kit (Vector Laboratories, Burlingame, CA, USA) according to the manufacture’s instructions using the avidine–biotine immunoperoxidase technique with diaminobenzine as the chromogen. Sections of inflamed human tonsil showing positive staining in some lymphocytes served as positive controls. Negative control sections incubated without primary antibody showed no staining.
The numbers of TUNEL-positive cardiomyocytes and cardiomyocytes containing cleaved caspase-3 were calculated using light microscopy (250×) with an ocular grid. The amounts of TUNEL-positive cardiomyocytes and cardiomyocytes containing cleaved caspase-3 were expressed as percentages of positively stained cardiomyocyte nuclei or cardiomyocytes from the total number of cardiomyocyte nuclei or cardiomyocytes, respectively. Only cells containing both fragmented DNA and typical cardiomyocyte myofilaments were calculated to maximize the probability, that only the apoptosis of cardiomyocytes were detected. In some cases consecutive histological sections or TUNEL-stained sections stained by immunohistochemistry for cardiac myosin were studied.
An average of 17,500 and 78,000 cardiomyocyte nuclei were analysed in TUNEL-stained tissue sections of each tissue sample obtained before or after operation, respectively.
2.5 Data analysis
The differences between the groups were tested with two-tailed Student’s t-test and with analysis of variance (ANOVA) followed by post hoc test (SPSS Statistical Program, SPSS Inc., Chicago, IL, USA). P-value ≪0.05 was considered statistically significant.
3 Results
3.1 Hemodynamics
Before ischemia, mean arterial pressures and heart rates were comparable among groups. Compared with 60 min of aortic clamping, mean arterial pressure was lower after 90 min of cardiac arrest immediately after the ischemia period (79.5 ± 9.1 mmHg vs 47 ± 121 mmHg, p = 0.021) and it stayed low trough the reperfusion period (75 ± 12.4 mmHg vs 50 ± 11.5 mmHg, p = 0.044). There were no statistical differences in heart rate before or immediately after the ischemia period, but after the 120 min reperfusion time the heart rate was lower in the 90 min aortic clamping group when compared with the 60 min group (152 ± 2 min−1 vs 111 ± 5.9 min−1, p ≪ 0.001). There were no statistical differences in central venous pressure at any point between the groups.
3.2 Cardiomyocyte apoptosis
In pre-operative biopsies very few apoptotic cardiomyocytes were found. In biopsies of six out of eight pigs there were no TUNEL-positive cardiomyocytes. However, apoptotic cardiomyocytes were found in all post-operative samples as detected by either TUNEL-method or immunohistochemistry for active caspase-3 (Figs. 1 and 2 ).
Percentages of TUNEL-positive cardiomyocytes in controls (120 min of cardiopulmonary bypass without ischemia) and after 60 or 90 min of aortic cross-clamp followed by 120 min of reperfusion. The amount of cardiomyocyte apoptosis is significantly higher after 90 than 60 min of cardioplegic arrest. The dots are indicating individual values and the short lines the median values.
Percentages of TUNEL-positive cardiomyocytes in controls (120 min of cardiopulmonary bypass without ischemia) and after 60 or 90 min of aortic cross-clamp followed by 120 min of reperfusion. The amount of cardiomyocyte apoptosis is significantly higher after 90 than 60 min of cardioplegic arrest. The dots are indicating individual values and the short lines the median values.
Percentages of cardiomyocytes stained with antibody to active form of caspase-3 were significantly increased after 90 min of cardioplegic arrest and 120 min when compared with controls (120 min of CBP without ischemia). The dots are indicating individual values and the short lines the median values.
Percentages of cardiomyocytes stained with antibody to active form of caspase-3 were significantly increased after 90 min of cardioplegic arrest and 120 min when compared with controls (120 min of CBP without ischemia). The dots are indicating individual values and the short lines the median values.
In the control group, the amount of TUNEL-positive apoptotic cardiomyocytes after the 120 min of CPB without ischemia was 0.006 ± 0.001% (n = 5). Compared with this, the amounts of TUNEL-positive cardiomyocytes were significantly higher after ischemia-reperfusion.
After 60 min aortic cross-clamp, followed by 120 min of reperfusion, the amount of TUNEL-positive cardiomyocytes was 0.019 ± 0.004% (n = 4, p = 0.031 vs control). After 90 min of aortic cross-clamp and 120 of reperfusion, the amount was 0.042 ± 0.005% (n = 4, p ≪ 0.001 vs control). Thus, the amount of apoptotic cardiomyocytes as shown by TUNEL-staining was significantly higher after 90 min than 60 min of aortic clamping.
In the control group (n = 5), the amount of containing active caspase-3 cardiomyocytes after the 120 min of CPB was 0.064 ± 0.048% (p = 0.014 vs pre-ischemia). The amounts of containing active caspase-3 cardiomyocytes were increased after ischemia-reperfusion. After 60 min of ischemia, the amount of cardiomyocytes containing active caspase-3 was already over five times higher than in controls being 0.362 ± 0.231% (n = 4, p = 0.362). After 90 min of ischemia, the amount was 0.772 ± 0.448% (n = 4), being statistically significantly higher than in controls (p = 0.019). Thus, these results were concordant with the results of TUNEL-staining.
4 Discussion
In this study we have analysed the influence of aortic clamping time on apoptosis in experimental open heart model. Apoptosis is a highly regulated type of cell death, deposed from the uncontrolled manner called necrosis. It is a programmed form of cell death and it can be characterised by a series of typical morphological events, such as shrinkage of the cell, fragmentation into membrane-bound apoptotic bodies and instant phagocytosis by neighbouring cells [6,7]. Apoptosis can be detected by some of its biochemical hallmarks, such as internucleosomal fragmentation of genomic DNA and activation of caspase enzymes [7,9].
According to a novel knowledge, cardiomyocyte apoptosis is an important mechanism in several cardiac diseases leading to heart failure, since functional cardiomyocytes are lost by apoptosis [8,13,14]. Examples of these kinds of diseases are acute myocardial infarct [9], myocarditis [10,15] and hypertension [16]. Programmed cell death of cardiomyocytes has been associated with open heart surgery and ischemia-reperfusion injury in experiments done with animals and as well as humans [9,11,12,17,18]. The determinants of cardiomyocyte apoptosis during cardiac surgery remain poorly understood. There is evidence that retrograde cardioplegia induces more apoptotic cardiomyocytes compared with antegrade [19] and few experiments suggest that longer aortic clamping time might correlate with higher amount of apoptotic cardiomyocytes [2,20] and declining cardiac contractility [2]. These last two studies were carried out in atrial biopsies obtained from patients undergoing elective coronary artery bypass grafting (CABG) and also here the cardiomyocyte apoptosis was detected by TUNEL-method. In the study done by Schmitt [2], the amount of cardiomyocyte apoptosis was compared with changes in cardiac index, pulmonary capillary wedge pressure and mean pulmonary artery pressure. Schmitt and co-workers also found a correlation between release of cytochrome-c from mitochondria, a mediator of apoptosis, and aortic clamping time. Longer aortic cross-clamp time is associated also with higher amount of cardiomyocyte necrosis in Langendorff-perfused rabbit hearts [21] and increased levels of TnT in humans during CABG [5].
Cardiomyocyte apoptosis is induced during myocardial ischemia, but its execution requires energy provided during reperfusion [22]. Thus, we hypothesized that exposure to longer period of ischemia during aortic cross-clamp and cardioplegic arrest might induce more cardiomyocytes than shorter period. In line with this hypothesis we found more cardiomyocyte apoptosis after 90 than 60 min of ischemia despites the same reperfusion period of 120 min. At the end of the experiment the amounts of apoptotic cardiomyocytes, detected by either TUNEL-method or staining of active caspase-3, were roughly two times greater in the 90 min group than in the 60 min group. This is the first time to show that apoptosis in ventricular myocytes is increased with prolonged durations of aortic clamping.
In our study the CPB alone seemed to induce cardiomyocyte apoptosis. The duration of CPB was necessarily different among groups, although the reperfusion time was the same between the ischemic groups. The control group was connected to the CPB comparable time to the ischemic groups, being almost the same as the reperfusion time. CPB itself induces inflammatory responses that may contribute to induction of apoptosis [23]. However, as shown by the small amount of apoptosis after CPB alone when compared with cardiac arrest, this effect is likely to be very small.
With our experimental set up we wanted to imitate the clinical practise as much as possible. The duration of aortic clamping time of 60–90 mins seems to be the average time needed in open heart surgery. On the other hand, we also wanted to keep the model as simple as possible and therefore we used cold crystalloid cardioplegia. Although today there are much more sophisticated forms of cardioplegia, cold crystalloid cardioplegia has been a corner stone of myocardial protection for years and is still used by many prominent cardiac surgeons. There is recent data demonstrated by Feng et al. [24], showing that the use of cold blood cardioplegia is associated with less apoptosis than warm blood cardioplegia during cardioplegic ischemia. This studies were carried out with Langendorff-perfused rabbit hearts and they also indicate that cold crystalloid cardioplegia might induce less cardiomyocyte apoptosis than warm crystalloid cardioplegia.
It has been previously shown that the duration of cardioplegic arrest seems to correlate with higher patient mortality [1] and lower cardiac contractility (stunning) [1,2]. Some experiments indicate a correlation between increasing duration of cardioplegic arrest and the development of atrial fibrillation [3,4]. Few experiments suggest that apoptosis is involved in myocardial stunning [2,25].
In this study we have shown that a longer aortic clamping time induces more myocardial apoptosis. For further research of myocardial apoptosis and stunning in open heart surgery this is an important knowledge. Our results correlate with the results of similar kinds of experiments done with humans [2,20]. To have stronger evidence and to understand the clinical significance of increased cardiomyocyte apoptosis in open heart surgery, it is necessary to do more research on this subject in future.
4.1 Limitations of the study
The number of test animals in this study was small and the statistical significance is therefore weak with caspase-3 enzyme. However, since the difference in the amount of cardiomyocyte apoptosis between 60 and 90 min of aortic cross-clamp became evident in this material we considered it unnecessary to increase the sample size. Although cardiac anatomy and physiology of pig are very similar to humans, the aortic cross-clamp times in pigs and humans cannot be directly compared. In humans up to 4–6 h of cross-clamp has been used. During the experiment we tried to expand the aortic clamping time up to 3 h, but with our model of healthy pigs, we learned, that maximal clamping time can only be 90 min. Longer aortic clamping time leads to irreversible cardiac dysfunction and to a stone heart. To keep the protocol as simple as possible, we did not measure cardiac output and pulmonary wedge pressure. No doubt, these measurements would have brought additional valuable information to this study.
Acknowledgement
This study was supported by a grant from the Finnish Cultural Foundation and by the Aarno Koskelo Foundation.
