Objectives: 

We designed a novel respiratory dialysis system to remove Co₂ from blood in the form of bicarbonate. We aimed to determine if our respiratory dialysis system removes Co₂ at rates comparable to low-flow extracorporeal Co₂ removal devices (blood flow < 500 mL/min) in a large animal model.

Design: 

Experimental study.

Setting: 

Animal research laboratory.

Subjects: 

Female Yorkshire pigs.

Interventions: 

Five bicarbonate dialysis experiments were performed. Hypercapnia (Pco₂ 90–100 mm Hg) was established in mechanically ventilated swine by adjusting the tidal volume. Dialysis was then performed with a novel low bicarbonate dialysate.

Measurements and Main Results: 

We measured electrolytes, blood gases, and plasma-free hemoglobin in arterial blood, as well as blood entering and exiting the dialyzer. We used a physical-chemical acid-base model to understand the factors influencing blood pH after bicarbonate removal. During dialysis, we removed 101 (±13) mL/min of Co₂ (59 mL/min when normalized to venous Pco₂ of 45 mm Hg), corresponding to a 29% reduction in Paco₂ (104.0 ± 8.1 vs 74.2 ± 8.4 mm Hg; p < 0.001). Minute ventilation and body temperature were unchanged during dialysis (1.2 ± 0.4 vs 1.1 ± 0.4 L/min; p = 1.0 and 35.3°C ± 0.9 vs 35.2°C ± 0.6; p = 1.0). Arterial pH increased after bicarbonate removal (7.13 ± 0.04 vs 7.21 ± 0.05; p < 0.001) despite no attempt to realkalinize the blood. Our modeling showed that dialysate electrolyte composition, plasma albumin, and plasma total Co₂ accurately predict the measured pH of blood exiting the dialyser. However, the final effluent dose exceeded conventional doses, depleting plasma glucose and electrolytes, such as potassium and phosphate.

Conclusions: 

Bicarbonate dialysis results in Co₂ removal at rates comparable with existing low-flow extracorporeal Co₂ removal in a large animal model, but the final dialysis dose delivered needs to be reduced before the technique can be used for prolonged periods.