AVCO2R PERIOPERATIVE MANAGEMENT

Abstract

Percutaneous arteriovenous CO2 removal (AVCO2R) is a simple A-V shunt for near-total CO2 removal that allows significant reductions in minute ventilation, decreased ventilator days, and increased survival in our LD40 model of ARDS. We critically reviewed our perioperative anesthetic management in this ovine smoke-burn injury model of ARDS. Sheep (n=16) require general anesthesia for the 3 steps of our protocol allowing AVCO2R outcomes studies. (1) Vascular access: 1–3 days prior to injury, 4.5–11.5 mg/kg IM ketamine and 4–5% halothane by mask for induction. Intubation, then maintenance by 1.5–2.5% halothane in 100% O2 titrated for MAP 70–95 mm Hg. Peak airway pressures were < 25 mmHg and end tidal CO2 30–40 mm Hg. (2) RDS model: Tracheostomy, then smoke insufflation (36 breaths) plus a 40% TBSA III° burn followed by mechanical ventilation. Intraop, Lactated Ringer's (LR) titrated for CVP 4–6 mm Hg. Postop, LR by Parkland formula then CVP titration. (3) AVCO2R/SHAM: When PaO2/FiO2 < 200 (48–52 hours), animals randomized to AVCO2R (n=8) or SHAM (n=8). Induction, 35% N2O and 5% isofluorane in 100% O2; maintenance 1.5–3.2% isofluorane in 100% O2 allowed exposure of the carotid artery and jugular vein. AVCO2R were cannulated (10F arterial; 14F venous) and connected to a membrane gas exchanger. SHAM received identical operative exposure without cannulation. Postop, both underwent algorithmic reduction in vent settings. All received identical heparin (ACT >300 sec) and fluid management. All sheep met criteria for RDS, survived both preps, and were standing in 0.5–5 hrs. Anesthetic time was equivalent for both groups. There were no complications attributable to anesthesia (hypotension requiring pressor support, arrythmias, or arrest). Algorithmic management protocol was followed 90% of the time. 8/8 AVCO2R and 3/8 SHAM survived the study. The absence of critical care and anesthetic related complications allows model development for independent evaluation of AVCO2R in outcomes studies.