The institutional review board of the Academic Medical Center, Amsterdam, The Netherlands, approved the trial registered in the Dutch trial register (number NTR1226). Inclusion into this prospective, placebo controlled, investigator blinded, parallel arm single-center study took place from 8/7/2008 to 3/7/2011 at the Academic Medical Center, Amsterdam, The Netherlands, in accordance with the International Conference on Harmonization on Good Clinical Practice Guidelines and the Declaration of Helsinki. Patients were randomized to one of five parallel groups in a 1:1:1:1:1 allocation ratio using web-based randomization software (ALEA; NKI; Amsterdam, The Netherlands) with a fixed block scheme, a block size of 5 patients and stratification on sex (see Fig. 1). While the anesthesiologist and the investigator in the operating room were not blinded, the patients as well as the investigators performing laboratory data analysis (troponin T values, Western Blot experiments) were blinded to the randomization strategy. Patients were recruited by self-selecting, and all subjects gave written informed consent. Exclusion criteria were age <18 years, legal incapacity, emergency operations, combined coronary artery and heart valve procedures, off-pump procedures, diabetes mellitus, severe chronic obstructive pulmonary disease (COPD), and left ventricular ejection fraction <30 %. The last two criteria were added after publication on http://www.trialregister.nl but before start of the study.
Study protocol
At least two cycles inhalation of sevoflurane were necessary to induce preconditioning in humans [18, 19]. We wanted to extend the preconditioning stimulus and decided to use three cycles of conditioning, as this protocol was also used in most experimental studies [20]. The first group received helium preconditioning (He-Pre) by inhalation of three cycles of helium for 5 min, followed by 5 min inhalation of oxygen-enriched air (30 % oxygen). Helium was obtained as a mixture with oxygen (Heliox: 79 % helium and 21 % oxygen, BOC, Mordon, United Kingdom) and administered using a non-invasive helium delivery system (Helontix Vent, Linde Therapeutics, Eindhoven, The Netherlands) modified to allow manual ventilation in a Maplesons A configuration. All patients were ventilated the same way by the same investigator. Extra oxygen was added and the final concentration of the gas-mixture was 70 % helium and 30 % oxygen. He-Pre was administered shortly before start of cardiopulmonary bypass (CPB). A graphical presentation of our study protocol is represented in Fig. 1. The postconditioning group (He-Post) received at least 15 min of helium at the end of aortic cross-clamping, lasting up to 5 min after release of the clamp. The third group received helium as pre- and postconditioning stimulus (He-PP). Patients receiving helium pre- and postconditioning thus received two conditioning stimuli of helium with double time of helium ventilation. To compare the effects of helium with the known effects of anesthetic preconditioning (APC), the fourth group received three cycles of 5 min sevoflurane inhalation with a minimal alveolar concentration (MAC) of 1.0 MAC, and the auto-flow function of the anesthesia machine (Zeus, Dräger Medical, Lübeck, Germany) was used to ensure rapid wash in and wash out of sevoflurane. The 5th group was an untreated control group.
Anesthesia
Patients received premedication with temazepam 10 mg per os. Induction of anesthesia was performed with intravenous administration of midazolam 0.1–0.2 mg kg−1 and target controlled infusion of propofol (dosage was 1–2 mg/kg for induction), sufentanil 1.0–1.5 μg kg−1, and rocuronium 0.6 mg kg−1 for muscle relaxation. Target controlled infusion of propofol was continued to maintain anesthesia in combination with either continuously or intermittently sufentanil.
Surgery
All patients received routine monitoring during operation and routine surgical techniques were used. A pulmonary artery catheter was used for cardiac output monitoring. The left internal mammary artery was used to graft the left anterior descending artery. As additional grafts, harvested veins from the leg, the right internal mammary artery or one of the radial arteries were used. Both, cold crystalloid and cold blood cardioplegia were administered antegrade via the aortic root, and management of the cardiopulmonary bypass (CPB) was according to standard procedure.
Median sternotomy was performed, followed by pericardiotomy after which the first sample of the right atrial appendage was obtained. Then the left internal thoracic artery was prepared, during which time systemic heparinization was started (300 IU/kg goal: coagulation time >450 s). After venous and arterial cannulas for CPB were inserted and secured, the second sample of the right atrium was obtained which was directly after preconditioning in the applicable groups. Then CPB was started, and the aorta was cross-clamped and cardioplegia solution was infused. All distal anastomoses were performed during aortic cross-clamping. Additional cardioplegic solution was administered at intervals to maintain a flat electrocardiogram. Fifteen minutes before expected release of the aortic cross clamp, we started helium postconditioning in the designated groups, and continued helium ventilation until 5 min after the start of reperfusion. After completion of coronary artery bypass grafting, CPB was discontinued and the third sample of the right atrium was obtained. After surgery, patients were transferred to the intensive care unit (ICU), received routine therapy and were weaned from the ventilator. ICU and ward staffs were blinded to the treatment allocation.
Blood sampling and tissue preparation
Blood samples were taken before cardiopulmonary bypass, 10 min after cardiopulmonary bypass and at the end of operation, as well as at 4, 12, 24 and 48 h after cardiopulmonary bypass. We measured troponin-T, creatinine kinase and its myocardial specific isoform Creatine Kinase-Muscle/Brain as markers of cellular injury. All samples were analysed in the Laboratory of Clinical Chemistry of the Academic Medical Centre, Amsterdam, The Netherlands.
Atrial samples were immediately flash frozen in liquid nitrogen and stored at −80 ◦C until further processing. Tissue fractionation was performed as described by Weber et al. [6]. Cytosolic, membrane, and the particulate fraction were immunoblotted using the Criterion Western Blotting system (Biorad, Hercules, CA).
After protein determination by the Lowry method, samples were thawed and diluted 1:5 with Sample Buffer 5 times containing Tris–HCl, glycerol and bromophenol blue. Samples were vortexed and boiled at 95 °C before being subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis using Criterion™ XT precast gels (Biorad, Hercules, CA). The proteins were separated by electrophoresis and transferred to a polyvinylidenfluorid membrane by tank blotting (Voltage 200 V for 50–55 min). Non-specific binding of the antibody was blocked by incubation with 5 % fat dry milk powder or bovine serum albumin solution in tris-buffered saline containing tween (TBS-T) for 2 h. Subsequently, the membrane was incubated overnight at 4 °C with the respective primary antibody at indicated concentrations. After washing in fresh, cold TBS-T, the blot was subjected to the appropriate horseradish peroxidase-conjugated secondary antibody for 2 h at room temperature. Immunoreactive bands were seen by chemiluminescence detected on X-ray film (Hyperfilm ECL, Amersham) using the enhanced chemiluminescence system Santa Cruz. The blots were quantified using a Kodak Image station® (Eastman Kodak Co., Rochester, NY, USA) and the results are presented as the ratio of phosphorylated to total protein. Values are expressed as x-fold average light intensity (AVI) compared with control. Equal loading of protein on the gel was additionally confirmed by detection of actin/α-tubulin and Coomassie staining of the gels.
Antibodies
We used anti-phospho PKC-ε, antibody (1:10.000) and total PKC-ε, both from Upstate (Lake Placid, NY). Phospho-ERK1/2 (1:10.000), total ERK 1/2 (1:10.000), Phospho p38 MAPK (1:5.000) and total p38 MAPK (1:5.000) were obtained from Cell Signalling (Danvers, MA), HSP 27 (1:5.000) from Abcam (Cambridge, UK). Both actin (1:10.000) and α-tubulin (1:40.000) were obtained from Sigma (St. Louis, MO). Peroxidase-conjugated goat anti-rabbit and donkey anti-mouse antibodies were from Jackson Immunoresearch (Suffolk, UK). The enhanced chemiluminescence protein detection kit was purchased from Santa Cruz (Heidelberg, Germany).
Endpoints and data collection
Primary endpoints of this study are phosphorylation of ERK1/2, p38MAPK and expression of HSP27 and PKC-ε in the particulate fraction. Secondary endpoints include post-operative troponin T release.
Data were collected on age, sex, race, length and weight, co-morbidities and risk factors for cardiovascular disease, Euroscore, medication usage, duration of bypass, and aortic clamping, number and type of grafts. Because of technical difficulties establishing reproducible results for the western blot we lost n = 5 patients per group for these targets (p38 MAPK, ERK1/2, HSP27 and PKC-ε).
Sample size calculation and statistics
Regarding our primary endpoint, no data on the effect of noble gas preconditioning on protein expression in human myocardial tissue was available while setting up the study. A proper sample size calculation was therefore not possible at start of the study. However, based on previous experimental research and a similar clinical study [19], we expected to find any—also clinically relevant differences—with a sample size of 25 patients per group.
Numerical data are presented as mean ± SD or median with interquartile range, as appropriate. Categorical data are presented as numbers and percentages. Statistical analyses were done using SPSS version 22 (IBM, Armonk, New York, USA).
We considered a p value of <0.05 to be statistically significant. Statistical testing of the western blot data was done using Shapiro–Wilk test for normality and Friedmann test for non-parametric data followed by Bonferroni correction for multiple testing (GraphPad Prism version 5.0, GraphPad, La Jolla, CA). We chose to graphically represent our data as mean + SD, however detailed information regarding the mean differences of the timepoints from our primary endpoints is available in Additional file 1.
To compare post-operative troponin T release the area-under-the-curve was calculated (mentioned as arbitrary units) and compared in a one-way-ANOVA.