Computer simulations of ultrafiltration profiles for an icodextrin-based peritoneal fluid in CAPD

Background. The three-pore model of peritoneal transport has the ability to predict ultrafiltration (UF) profiles rather accurately, even when high molecular weight (MW) solutes are employed as osmotic agents in continuous ambulatory peritoneal dialysis (CAPD). In the present simulations, we wanted to assess, for various theoretical perturbations, the UF properties of a peritoneal dialysis (PD) solution with an osmotic agent having an average MW of 20 kD and a "number average MW" of 6.2 kD, which is similar to that of icodextrin (ICO). Methods. For a PD solution containing a completely monodispersed 20 kD MW osmotic agent, the degree of UF modeled is much higher than that reported for ICO. Hence, to model the behavior of ICO, we subdivided the ICO molecules into eight or more different MW size fractions. For simulations using six or eight subfractions, we obtained an excellent fit of simulated to reported UF data. More dispersed solutions produced UF profiles similar to that with eight fractions. Results. A 2.05 L 7.5% ICO PD solution, despite being slightly hypotonic, yielded a UF volume of nearly 600 mL in 12 hours, modeled for patients not previously exposed for ICO. After nine hours, the UF volume exceeded that produced by 3.86% glucose. The UF rate and volumes increased in proportion to (1) the ICO concentration, (2) the peritoneal surface area, and (3) the peritoneal UF coefficient, but was almost insensitive to increases in the instilled fluid volume. Simulated for patients previously exposed to ICO, having steady-state plasma concentrations of ICO degradation products, the predicted UF volume at 12 hours was reduced to approximately 400 mL. Conclusion. Employing the three-pore model of peritoneal transport and taking into account the polydispersed nature of ICO, it was possible to accurately computer simulate the UF profiles of ICO in accordance with reported data. The simulations suggest an advantage of using ICO in patients with type I UF failure, where UF with a high-MW osmotic agent will exceed that seen in patients not showing UF failure who are on glucose-based PD solutions.