Effect of dialysate osmolarity on the transport of low-molecular weight solutes and proteins during CAPD

Abstract

Osmotic-induced fluid and solute transport was studied in ten stable CAPD patients, who were examined twice within one week, using dialysate with 1.36% glucose on the first and 3.86% glucose on the second day. Peritoneal fluid kinetics were determined using intraperitoneally administered dextran 70 as a volume marker. After a four-hour dwell period, an increase in mean transcapillary ultrafiltration rate (TCUFR) with 3.86% glucose compared to 1.36% glucose was found (3.40 ± 0.62 ml/min vs. 1.20 ± 0.57, P < 0.001), but the lymphatic absorption was unchanged (1.32 ± 0.10 ml/min vs. 1.42 ± 0.15). The increased TCUFR resulted in a higher clearance of β2-microglobulin, but no differences were present in the clearances of albumin, transferrin, IgG, IgA and α2-macroglobulin. This is consistent with the two pore theory for transcapillary transport with a small pore size of less than 40 Å. The contribution of osmotic induced convection to the total transport of β2-microglobulln was small (6% during 1.36% glucose, 16% during 3.86% glucose), suggesting that macromolecules are mainly transported by diffusion or hydrostatic convection. The peritoneal restriction coefficient was 2.37 ± 0.04, indicating restricted diffusion for macromolecules. In contrast, the restriction coefficient for low-molecular weight solutes was 1.24 ± 0.03, in accordance with a process of mainly unrestricted diffusion for solutes smaller than 16 Å. Higher values of protein clearances were found during the first hour of dialysis compared with the subsequent hours. This was probably caused by a vasoactive effect of commercially available dialysis fluid resulting in an increase in effective peritoneal surface area, and to a lesser extent by an increase in intrinsic peritoneal permeability. Various models for the calculation of the mass transfer area coefficient (MTC) of urea, creatinine and urate were used to evaluate possible interterence by convection on the results of the calculations. With the exception of MTC urea, MTC values of the other solutes were similar, irrespective of the relative weight given to convective transport. It implies that the simplified Garred model can also be used during hypertonic dwells for the calculation of MTC creatinine and urate.