Discriminative impact of ultrafiltration on peritoneal protein transport.

OBJECTIVE The dialysate concentration of large proteins increases, on average, linearly during the whole peritoneal dialysis dwell, and this linear pattern seems to be independent of the rate of ultrafiltration induced by dialysis fluid. However, we observed a high variability of protein kinetics in individual dwell studies. Therefore, we studied the details of the kinetic pattern of peritoneal transport. DESIGN AND METHODS Kinetics of beta2-microglobulin, albumin, and total protein was examined in 23 clinically stable continuous ambulatory peritoneal dialysis patients using Dianeal 3.86% (15 dwell studies) or Dianeal 1.36% (9 dwell studies) dialysis fluid. Dialysate volume was measured using radioisotopically labeled albumin as a volume marker, with corrections for sample volume and absorption of fluid and marker from the peritoneal cavity. The generalized version of the Babb-Randerson-Farrell model was applied to estimate diffusive mass transport coefficient (K(BD)) and sieving coefficient (S) for proteins and small solutes (urea, creatinine, glucose, sodium, potassium). To quantify deviations from the linear pattern of protein dialysate concentration increase, the ratio (SR) of the slope of the linear regression line for the initial 3-30 minutes, divided by the slope for the next 60 - 360 minutes, was evaluated for albumin. RESULTS In 5 dwell studies with Dianeal 3.86% fluid, SR was lower than 1 [low albumin transport (LAT) group, median SR = 0.49, range -4.39 - 0.71], while in the other 10 dwell studies with this solution, SR was higher than 1 [high albumin transport (HAT) group, median SR = 2.77, range 1.32 - 7.56]. Clearances of albumin up to 120 minutes were higher in the HAT group than in the LAT group. The transport of fluid, beta2-microglobulin, and small solutes did not differ between the LAT and the HAT groups. K(BD) values for proteins did not differ between the groups, but S values for albumin and total protein were lower for the LAT group than for the HAT group. A similar diversity was found in the dwell studies with Dianeal 1.36%: In three dwell studies, SR for albumin was lower than 1 (median SR = 0.95, range 0.70 - 0.97), and in six dwells it was higher than 1 (median SR = 1.55, range 1.23 - 1.98). In general, the SR values observed with Dianeal 1.36% were closer to 1 than those for Dianeal 3.86%. CONCLUSIONS Ultrafiltration may affect the initial kinetic patterns of large protein (such as albumin) transport in two opposing ways: (1) by slowing the increase of protein concentration in dialysate (due to a low sieving coefficient, LAT group), and (2) by speeding up the increase of protein concentration in dialysate (due to a high sieving coefficient, HAT group). The average pattern in a non-selected group of studies is, however, close to a steady (linear) increase.