Respiratory dialysis: Reduction in dependence on mechanical ventilation by venovenous extracorporeal CO2 removal*

OBJECTIVES: Mechanical ventilation is injurious to the lung. Use of lung-protective strategies may complicate patient management, motivating a search for better lung-replacement approaches. We investigated the ability of a novel extracorporeal venovenous CO2 removal device to reduce minute ventilation while maintaining normocarbia. DESIGN: Prospective animal study. SETTING: Government laboratory animal intensive care unit. SUBJECTS: Seven sedated swine. INTERVENTIONS: Tracheostomy, volume-controlled mechanical ventilation, and 72 hrs of round-the-clock intensive care unit care. A 15-F dual-lumen catheter was inserted in the external jugular vein and connected to the Hemolung, an extracorporeal pump-driven venovenous CO2 removal device. Minute ventilation was reduced, and normocarbia (Paco2 35-45 mm Hg) maintained. Heparinization was maintained at an activated clotting time of 150-180 secs. MEASUREMENTS AND MAIN RESULTS: Minute ventilation (L/min), CO2 removal by Hemolung (mL/min), Hemolung blood flow, O2 consumption (mL/min), CO2 production by the lung (mL/min), Paco2, and plasma-free hemoglobin (g/dL) were measured at baseline (where applicable), 2 hrs after device insertion, and every 6 hrs thereafter. Minute ventilation was reduced from 5.6 L/min at baseline to 2.6 L/min 2 hrs after device insertion and was maintained at 3 L/min until the end of the study. CO2 removal by Hemolung remained steady over 72 hrs, averaging 72 ± 1.2 mL/min at blood flows of 447 ± 5 mL/min. After insertion, O2 consumption did not change; CO2 production by the lung decreased by 50% and stayed at that level (p < .001). As the arterial PCO2 rose or fell, so did CO2 removal by Hemolung. Plasma-free hemoglobin did not change. CONCLUSIONS: Venovenous CO2 removal enabled a 50% reduction in minute ventilation while maintaining normocarbia and may be an effective lung-protective adjunct to mechanical ventilation.