24-hour performance of albumin dialysis assessed by a new two-compartment in vitro model.

Extracorporeal albumin dialysis as a measure to remove water soluble and protein bound toxins simultaneously has been shown to improve complications of liver failure. However, recent research suggests that only treatments associated with a measurable improvement of patient's albumin binding function by effective removal of albumin bound toxins leads to better survival. The aim of the present work was to develop a test platform for upcoming devices to evaluate long term effectiveness on toxin removal and improvement of patients' albumin binding capacity. The classical one compartment model consisting of a closed pool of toxin spiked plasma was combined with continuous infusion of water soluble and protein bound toxins mimicking physiological rebound rates reflected in the literature. The model was used to demonstrate the effect of stabilizer contamination of dialysate albumin on toxin clearance and albumin binding function. In comparison to the classical one compartment model, the two compartment model allows for long term effectiveness tests of liver assist devices not only for strongly albumin-bound, but also water-soluble molecules. The limitations of commercial albumin overloaded with caprylate ligands (5:1 molar ratio) were demonstrated by presenting a significant improvement of albumin binding function using 80 g deligandized albumin compared to no significant improvement using the standard 120 g albumin as dialysate. The new two compartment model allows for pre-clinical evaluation of new upcoming devices aiming for improvement of patients' albumin binding function as a measure for clinically meaningful extracorporeal detoxification of albumin-bound toxins.