Elimination of gaseous microemboli from cardiopulmonary bypass using hypobaric oxygenation

Abstract

Background: Numerous gaseous microemboli (GME) are delivered into the arterial circulation during cardiopulmonary bypass (CPB). These emboli damage end organs through multiple mechanisms that are thought to contribute to neurocognitive deficits after cardiac surgery. Here, we use hypobaric oxygenation to reduce dissolved gases in blood and greatly reduce GME delivery during CPB.

Methods: Variable subatmospheric pressures were applied to 100% oxygen sweep gas in standard hollow fiber microporous membrane oxygenators to oxygenate and denitrogenate blood. GME were quantified using ultrasound while air embolism from the surgical field was simulated experimentally. We assessed end-organ tissues in swine postoperatively using light microscopy.

Results: Variable sweep gas pressures allowed reliable oxygenation independent of carbon dioxide removal while denitrogenating arterial blood. Hypobaric oxygenation produced dose-dependent reductions of Doppler signals produced by bolus and continuous GME loads in vitro. Swine were maintained using hypobaric oxygenation for 4 hours on CPB with no apparent adverse events. Compared with current practice standards of oxygen/air sweep gas, hypobaric oxygenation reduced GME volumes exiting the oxygenator (by 80%), exiting the arterial filter (95%), and arriving at the aortic cannula (∼100%), indicating progressive reabsorption of emboli throughout the CPB circuit in vivo. Analysis of brain tissue suggested decreased microvascular injury under hypobaric conditions.

Conclusions: Hypobaric oxygenation is an effective, low-cost, common sense approach that capitalizes on the simple physical makeup of GME to achieve their near-total elimination during CPB. This technique holds great potential for limiting end-organ damage and improving outcomes in a variety of patients undergoing extracorporeal circulation.

Appendix Figure 1

Microscopic evaluation of peripheral blood, brain, and kidney from animals managed with hypobaric oxygenation revealed normal cytoarchitecture. (A) Peripheral blood smears were made from samples taken before initiation of CPB (baseline), then after two and four hours of CPB with hypobaric oxygenation. At 1000X magnification, smears display red blood cell echinocytes with a crenated appearance, which is typical for swine, but no evidence of cellular injury or hemolysis. (B–D) Hematoxylin-eosin stained, paraffin-embedded, 4 μm thick sections from cerebral cortex (B), hippocampal region CA1 (C), and renal cortex (D) are without apparent abnormality. Scale bars = 100 μm

Figure 1. Hypobaric oxygenation reduces dissolved gases in vitro

(A) Hypobaric oxygenation apparatus. Pure oxygen is supplied to the sweep gas inlet of a standard hollow fiber microporous membrane oxygenator with a sealed housing. A regulated vacuum source at the sweep gas outlet applies user-determined variable subatmospheric pressure to the sweep gas compartment to regulate the partial pressure gradient for blood oxygenation. A vacuum gauge measures the pressure applied, while a positive-pressure relief valve (PPR) insures against creation of positive pressures. A needle-valve flowmeter at the sweep gas inlet regulates sweep gas flow rate and thus CO₂ removal, while allowing a pressure drop from ambient to subatmospheric. (B) In vitro gas exchange circuit. A mixture of human RBCs, FFP, and minimal crystalloid (Hct∼30%) from a CPB reservoir is pumped (3.5 liters/minute) to the CPB oxygenator (37°C), where oxygenation occurs with pure oxygen sweep gas at variable subatmospheric pressure. The blood then passes into a Patient Simulator consisting of a reservoir, pump, and oxygenator that removes O₂ and adds CO₂ using pure CO₂ sweep gas at very low pressure (1 liter/minute, 0.1 ata). Blood gases were sampled downstream of the CPB oxygenator (arterial) and the Patient Simulator (venous, n=3 samples per condition). The simulated patient produced normal venous blood gas values. (C, D) Application of subatmospheric sweep gas pressure in the CPB oxygenator reduced arterial oxygenation in the expected linear manner independent of CO₂ removal.

Figure 2. Hypobaric oxygenation greatly enhances GME removal in vitro

(A) Single oxygenator CPB circuit configuration with arterial filter and purge line returning to reservoir. Locations of air introduction and Doppler GME monitoring are shown. Arrow indicates direction of blood flow (5 liters/minute, 37°C). (B) A tiny (∼20 μl) air bubble agitated by hand in 10 ml blood (using two 12 ml syringes and a 3-way stopcock) was injected upstream of the oxygenator. The emboli traversed the oxygenator and were detected proximal to the arterial filter. Application of slightly subatmospheric sweep gas pressures strongly attenuated the GME Doppler signal. Data are averages of 9–10 trials per condition. (C, D) Continuous entrainment of air (500 ml/minute, via luer connector) into the venous line at the reservoir entrance simulated a large, continuous embolic challenge. With ambient sweep gas pressures, strong Doppler signals were observed upstream and downstream of the arterial filter. Hypobaric oxygenation produced a robust dose-dependent reduction of the Doppler signal in both locations. Data are averages of 3 continuous trials that employed 2-minute steps to each listed pressure level.

Figure 3. Hypobaric oxygenation nearly eliminates GME delivery in vivo

(A) CPB circuit with EDAC and Doppler monitors and venous air entrainment site (200 ml/minute). (B) Histograms showing EDAC GME counts and sizes at the four monitoring sites for both control (O₂/air sweep gas at ambient pressure) and hypobaric (O₂ sweep gas at subatmospheric pressure) conditions. Under control conditions ∼4500 GME/minute were delivered to the patient. Under hypobaric conditions, GME counts and volumes were similar to control at the preoxygenator location, but were progressively eliminated as they traversed the CPB circuit, reducing GME delivery to only 2/minute during this large embolic load.

Figure 4. Animals managed with hypobaric oxygenation display reduced microvascular injury in cerebral white matter

(A) 10X photomicrographs of 4 μm-thick hematoxylin-eosin stained sections of periventricular white matter. Dilated capillaries appear as voids (white) surrounded by a single layer of endothelial cells. Scale bars = 100 μm. (B, C) Dilated capillaries were fewer in number and in area in animals managed using hypobaric oxygenation (control n=30 10X-fields, N=3 animals; hypobaric n=51 fields, N=5 animals; *=p<0.001). (D) The difference in numbers of dilated capillaries between conditions was present at all sizes studied, reassuring us that the result is not an artifact of the measurement criteria.