Open chest and pericardium facilitate transpulmonary passage of venous air emboli

Transpulmonary passage of air emboli can lead to fatal brain‐ and myocardial infarctions. We studied whether pigs with open chest and pericardium had a greater transpulmonary passage of venous air emboli than pigs with closed thorax.


| INTRODUC TI ON
Venous air emboli frequently complicate surgery, interventional procedures, and trauma. 1 Symptoms range from a transient drop in end-tidal CO 2 concentration, decreased lung compliance, pulmonary edema, cerebral-or myocardial infarction, hemodynamic collapse to death. 2,3 The lungs usually filter emboli, but this filtering capacity is limited and can be overwhelmed, allowing air to traverse the lungs into the systemic circulation, potentially causing embolic infarctions .4,5 The thorax and pericardium provide a rigid framework for the lungs and heart, limiting overexpansion of the organs. 6 Opening the thorax, including pleura and pericardium, has been shown to reduce air emboli tolerance. 7 However, the effect of open chest and pericardium, including the pleura and pericardial sac, on the lungs' ability to filter venous air embolism has not, to our knowledge, been studied.
Cognitive impairment and micro-infarctions are frequent complications in open-heart surgery. Studies suggest cerebral microemboli of particulate or gaseous matter as a potential mechanism. 8,9 Studies have suggested that systemic air emboli can occur during open-heart surgery or if air enters a cardiopulmonary bypass circuit. 8 Several studies have proposed echocardiography as a sensitive monitor of venous air emboli. [10][11][12][13] We hypothesized that opening the chest, including the pleura and pericardial sac, reduces the lung filtering capacity of venous air emboli, potentially leading to systemic egress of air. Thus, we studied the transthoracic passage of venous air emboli in pigs with an open chest and pericardium compared to pigs with closed thorax.

| MATERIAL S AND ME THODS
The Norwegian Animal Research Authority approved the studies (FOTS ID9466), and we performed the experiments under the Norwegian Laboratory Animal Regulations and the EU directive 2010/63/EU.

| Experimental animals
In a two-center non-randomized, experimental study using a convenience sample, we allocated Norwegian landrace pigs from two suppliers to undergo an intravenous ambient air infusion either with open chest and pericardium, including open pleura and pericardium or closed thorax, or to serve as sham animals not receiving air. We excluded sick pigs and pigs with open foramen ovale. Finally, 30 pigs were included (Figure 1). Due to laboratory logistics, we retrieved animals allocated to sternotomy from one supplier. We summarize the animal characteristics in Table 1.

| Instrumentation, anesthesia, and surveillance
We anesthetized all pigs with azaperone 40 mg, ketamine 500 mg, and atropine 0.5 mg intramuscularly and maintained the anesthesia with an intravenous infusion of morphine 2 mg/kg/h, midazolam 0.15 mg/kg/h and pentobarbital 4 mg/kg/h. We endotracheally intubated the pigs with a 6 mm outer diameter tube. We mechanically ventilated the pigs with a tidal volume of 10-15 mL/kg, a rate of 20/ min, and zero positive end-expiratory pressure. Tidal volume and respiratory rate were adjusted to maintain a pH of 7.35-7.45. Inspiratory oxygen fraction (FiO 2 ) was adjusted to maintain an arterial pulse oximetry saturation (SpO 2 ) above 90%. We infused Ringer's acetate to We recorded cardiac output by hourly thermodilution (average of three injections of 5 mL ice-cold Ringer's Acetate) by the Edwards Vigilance II (Edwards) and Pulsion PICCO2 (Pulsion/Getinge).

| Air infusion and data collection
All pigs, except sham animals, received an infusion of ambient air through an ear vein, initially at a rate of 4-6 mL/kg/h, and increased by 2 mL/kg/h every hour until the pigs died or until the end of the experiment after 300 minutes. The air infusion was titrated based on previous studies, 4,7,13 aiming to cause hemodynamic instability, but F I G U R E 1 Inclusion of animals. We allocated forty-one pigs to receive infusion with either open chest and pericardium or closed thorax or serve as sham animals. We excluded animals with PFO, pneumonia, perioperative lung injury, or ventricular fibrillation related to wedging of the PA catheter. TEE, transesophageal echocardiogram; PFO, patent foramen ovale; VAE, venous air emboli; VF, ventricle fibrillation; PA, pulmonary artery not death. We euthanized pigs still alive 300 minutes after start of the air infusion by injection of potassium chloride.

| Power calculation
We expected to find systemic egress of intravenously infused air in

| Statistics
We used Prism for Mac 8.4.2 (Graphpad Software) for statistical calculations and used the Mantel-Cox log-rank test to compare Kaplan-Mayer survival curves and Spearman's test for correlations. We considered a P <.05 significant.  TA B L E 1 Animal characteristics after instrumentation, before air infusion between infused air and time to death ( Figure 3D).

| Cardiovascular effects of infused air
Air infusion triggered an immediate rise in the mean pulmonary artery pressure and reduced systemic arterial pressure resulting in an increase in the mean pulmonary artery to mean arterial pressure ratio (MPAP/MAP) in all pigs receiving air infusion. The MPAP/ MAP ratio remained constant in the sham animals ( Figure 4A,B).
Air infusion also led to reduced gas exchange and increased airway compliance, and triggered a tachycardia, as described in Supporting Information 2.
In most of the pigs receiving air infusion, we observed by echo-  (Table 2). that the lungs' filtering capacity is limited, in accordance with other studies. 4,5 We found that opening the thorax, pleura, and pericardium greatly reduced the threshold for when air reached the systemic circulation through the lungs. We had not designed our study to elaborate on the anatomical mechanisms underlying these findings. However, we suggest that recruitment of extra-alveolar shunt vessels by the following mechanism played an important role: Opening of the chest and pericardium resulted in reduced trans-