Fluid and electrolyte transport across the peritoneal membrane during CAPD according to the three-pore model

In the present review, we summarize the principles governing the transport of fluid and electrolytes across the peritoneum during continuous ambulatory peritoneal dialysis (CAPD) in "average" patients and during ultrafiltration failure (UFF), according to the three-pore model of peritoneal transport. The UF volume curves as a function of dwell time [V(t)] are determined in their early phase by the glucose osmotic conductance [product of the UF coefficient (LpS) and the glucose reflection coefficient (sigmag)] of the peritoneum; in their middle portion by intraperitoneal volume and glucose diffusivity; and in their late portion by the LpS, Starling forces, and lymph flow. The most common cause of UFF is increased transport of small solutes (glucose) across the peritoneum, whereas the LpS is only moderately affected. Concerning peritoneal ion transport, ions that are already more or less fully equilibrated across the membrane at the start of the dwell, such as Na+ (Cl-), Ca2+, and Mg2+, have a convection-dominated transport. The removal of these ions is proportional to UF volume (approximately 10 mmol/L Na+ and 0.12 mmol/L Ca2+ removed per deciliter UF in 4 hours). The present article examines the impact on fluid and solute transport of varying concentrations of Ca2+ and Na+ in peritoneal dialysis solutions. Particularly, the effect of "ultralow" sodium solutions on transport and UF is simulated and discussed. Ions with high initial concentration gradients across the peritoneum, such as K+, phosphate, and bicarbonate, display a diffusion-dominated transport. The transport of these ions can be adequately described by non-electrolyte equations. However, for ions that are in (or near) their diffusion equilibrium over the peritoneum (Na+, Ca2+, Mg2+), more complex ion transport equations need to be used. Due to the complexity of these equations, however, non-electrolyte transport formalism is commonly employed, which leads to a marked underestimation of mass transfer area coefficients (PS). This can be avoided by determining the PS when transperitoneal ion concentration gradients are steep.