After approval of the experimental protocol by the State and Institutional Animal Care Committee Rhineland Palatine (approval no. G16-1-042-E4), 18 male German landrace pigs (age range 16–20 weeks, weighing 30–35 kg) were acquired from a local farm after being screened by the breeder for any obvious medical conditions or diseases according to the German Animal Care Regulations. The animals were given pre-transport sedation via an intramuscular injection of azaperone (2 mg kg− 1) and ketamine (4 mg kg− 1) and were secured in a large box with hay bedding, in which they were transported to our facility (~ 30 minutes). After the pigs were in the Large Animal Research Facility, anaesthesia was induced via an intravenous catheter placed in the lateral/marginal auricular vein (22 gauge, B. Braun, Germany) by injecting into it fentanyl (4 µg kg− 1, Rotexmedica GmbH, Germany), propofol (4 mg kg− 1, Fresenius Kabi GmbH, Germany) and atracurium (0.5 mg kg− 1, Hikma Pharma GmbH, Germany). Then, a secure airway was established using a standard endotracheal tube (ID 6.0‒7.0 mm, Teleflex Medical, Ireland) under direct laryngoscopy. During the entire experiment, anaesthesia was maintained via continuous infusion of propofol (5‒10 mg kg− 1 hour− 1) and fentanyl (8‒12 µg kg− 1 hour− 1) as well as a balanced electrolyte infusion of 5 ml kg− 1 hour− 1 (Sterofundin ISO, B. Braun, Germany). Volume-controlled ventilation was provided and monitored using an intensive-care respirator (Engstroem Care Station, GE Healthcare, Germany) with tidal volumes (Vt) of 6‒7 mL kg− 1, peak inspiratory pressures of 40 cmH2O, positive end-expiratory pressure of 5 cmH2O and a respiratory rate adapted to end-expiratory carbon dioxide (CO2) levels below 6 kPa (45 mmHg), which usually resulted in 20 ‒ 30 breaths minute− 1.
After being anaesthetised, the animals were instrumented under the guidance of ultrasound with a central-venous catheter, pulse contour cardiac output system (PiCCO, Pulsion, Germany) and a Swan-Ganz catheter through introducer sheaths in the femoral veins and arteries as described before(13). When the instrumentation was completed, the animals were screened again for any previously inapparent cardiopulmonary pathologies during base line measurements (i.e. ventricular defects or severe oxygenation impairments due to infections). Any afflicted animals would have been excluded from the trial and euthanised.
Following the health assessment, an oscillation catheter (Osypka Medical GmbH, Rheinfelden-Herten, Germany) was placed intravenously. The fasting animals were given an initial fluid bolus of 30 ml/kg warm balanced electrolyte solution and left to stabilise for 30 minutes before base line measurements were taken.
Following base line measurements, ventricular fibrillation was induced via the oscillation catheter (13.8 V current according to manufacturer’s recommendation) and the ventilator was disconnected. Monitor-confirmed cardiac arrest was permitted for eight minutes, and the animals were randomised into two groups by blinded drawing of 1 of 18 envelopes containing the respective chest compression device (9 animals per group):
mCCD 1 - Continuous automated chest compressions via the LUCAS™ 2 device (Stryker® Corporation, Kalamazoo MI, USA) with a fixed rate of 100 min− 1 as described before(14).
mCCD 2 - Continuous automated chest compressions via the Corpuls™ cpr device, using a Recboard and a standard size stamp (GS Elektromedizinische Geraete, Kaufering, Germany) with a fixed rate of 100 min− 1 and a set compression depth of 5 cm and positioning as described before(12).
During chest compressions, both groups were ventilated with a guideline-based ventilation regimen (Vt 8–10 ml kg− 1, FiO2 1,0, RR 10 min− 1). After eight minutes of continuous CPR, resuscitation measures were continued according to the advanced life support algorithm: 2 minute compression cycles, rhythm analysis, defibrillation (200 J, bi-phasic), epinephrine (1 mg) and vasopressine (0.1 U kg− 1) administration as well as amiodarone (5 mg kg− 1) after the third and the sixth cycle. If ROSC was not achieved after the 10th cycle, the experiment was terminated. Animals achieving ROSC were switched back to standard ventilation and monitored for eight hours. During the monitoring period, mean arterial blood pressure was kept over 60 mmHg using a norepinephrine drip if necessary. The experiment was terminated with the animal being euthanised using high doses of propofol (200 mg) and potassium chloride (40 mmol).
Cardiopulmonary data were constantly measured and collected during the duration of the experiment using a Datex Ohmeda S5 monitor (GE Healthcare, Munich, Germany). These include respiratory rate, ventilation pressures, oxygen fractions, oxygen saturation, intra-arterial blood pressure, pulmonary artery pressure, heart rate and core temperature. Additionally, blood gas analyses were performed at base line (“BLH”) 5 minutes into chest compressions (“BLS”), after the fourth shock (“ALS 1”) and after the eighth shock (“ALS 2”).
Ventilation/perfusion (VA/Q) analyses were performed at base line and during CPR (at “BLS”) using the micropore membrane inlet mass spectrometry facilitated multiple inert gas elimination technique (MMIMS-MIGET, Oscillogy LLC, Philadelphia, USA) as described before(13). In short, subclinical, non-toxic doses of a saline solution containing six chemically inert gases with different elimination constants (sulphur hexafluoride, krypton, desflurane, enflurane, diethyl ether and acetone) were infused starting 20 minutes prior to measurements in order to reach an in vivo steady state. Blood samples from the pulmonary and femoral artery were taken and analysed via a mass spectrometer determining gas elimination during lung passage, thus allowing accurate VA/Q fraction determination for high, normal and low perfusion ratios as well as shunt volumes.
After termination, the thorax was examined using para- and substernal incisions and careful preparation in order not to inflict additional damage. Pericardium, pleura, rib cage and lung tissue were assessed using a scoring system developed by our group, consisting of 7 damage aspects (haematothorax, pneumothorax, rib fractures, sternal fractures, pericardial effusion/tamponade, blood in the gastric tube and blood in the tracheal tube, see Fig. 5). To support the clinical findings, sonographic analysis of the thorax was performed before the first incision, screening for pneumothorax or pleural effusion using a mobile device with a linear probe (Sonosite M, FUJIFILM Sonosite GmbH, Frankfurt, Germany). Exemplary pictures and videos as well as pictures of lung sonography and thoracic injuries are provided in the online supplement of this article.
Statistical analyses were performed using 2-way ANOVA inter-group tests for repeated measurements as well as students-t-tests for single measurements via GraphPad Prism 8 software (GraphPad Software Inc., La Jolla, CA, USA). Data in the text are presented as mean (standard deviation). p-values < 0.05 were considered significant.