The Removal of Uremic Solutes by Peritoneal Dialysis

Abstract

Peritoneal dialysis (PD) is now commonly prescribed to achieve target clearances for urea or creatinine. The International Society for Peritoneal Dialysis has proposed however that such targets should no longer be imposed. The Society's new guidelines suggest rather that the PD prescription should be adjusted to achieve well-being in individual patients. The relaxation of treatment targets could allow increased use of PD. Measurement of solute levels in patients receiving dialysis individualized to relieve uremic symptoms could also help us identify the solutes responsible for those symptoms and then devise new means to limit their accumulation. This possibility has prompted us to review the extent to which different uremic solutes are removed by PD.

Figure 1

The concentration of urea in the peritoneal dialysate rises to approach the concentration of urea in the plasma with time. Values were recorded after the infusion of 2 L of dialysate in a single patient by Popovich et al. Calculation of the mass transfer area coefficient for urea from the initial points along the curve would here yield a value of approximately 20 ml/min. A notable difference between peritoneal dialysis and hemodialysis is that clearance declines during the course of a peritoneal exchange as the concentration of solute in the dialysate rises and the gradient driving solute diffusion into the dialysate declines. Peritoneal solute clearances are therefore usually expressed as effective clearances, obtained by dividing the total daily solute removal by the plasma solute concentration which remains nearly stable through the day. Redrawn from Popovich et al.

Figure 2

The mass transfer area coefficient (MTAC – logarithmic scale) declines with increasing solute molecular mass (logarithmic scale). The MTAC values obtained by Kredeit et al. after infusion of 2 L of dialysate into eight patients have been adjusted downward for an average body surface area of 1.73 m² as compared with the average body surface area of 2.02 in the original study. The solutes studied were ß₂M, ß₂ microglobulin; A, albumin; C, creatinine; G, glucose; I, inulin; IgG, immunoglobulin G; K, kanamycin; L, lactate; MTAC, mass transfer area coefficient; U, urea. Redrawn from Kredeit et al.

Figure 3

The predicted relation of plasma solute concentrations in patients maintained on three times weekly HD and patients maintained on PD has changed greatly over 40 years. The solid lines represent the predicted ratios of the time-averaged concentrations achieved with HD to the more stable concentrations achieved with PD, and the dashed lines represent the higher predicted ratios of the average peak concentrations achieved with HD relative to the concentrations achieved with PD. Predicted relative concentrations are shown for HD and PD prescriptions commonly in use in 1980 (left) and modern HD and PD prescriptions (right) as further detailed in Supplemental Table 1. In 1980, HD was predicted to yield the lower clearances and higher plasma levels relative to PD for solutes with molecular mass above approximately 200 Da. Since that time, however, changes in hemodialysis membrane materials along with increases in membrane surface areas and blood flow rates have indeed resulted in HD patients having lower predicted plasma solute concentrations than PD patients for all solutes across the size range depicted if the solutes' volumes of distribution are similar to that of urea. Intermittent HD is predicted to provide inferior control only of the predialysis or peak plasma concentrations of small solutes with small volumes of distribution as further shown under Supplemental Table 1. Solute concentrations have been modeled assuming that solute production is the same in HD and PD patients and that there is no residual function. HD, hemodialysis; PD, peritoneal dialysis. The left panel is redrawn from the original analysis of Hiatt et al.

Figure 4

Superior control by PD of the plasma levels of small solutes which are reabsorbed in the native kidney. The bottom panel shows the GFR declining from 10 to 0 ml/min over time. The upper panel shows the plasma levels of three small uremic solutes: one of which is secreted and has a clearance twice that of the GFR (red line), one of which has a clearance equal to the GFR (green line), and one of which is partially reabsorbed in the kidney and has a clearance half the GFR (blue line). Before dialysis is initiated, the plasma levels of the solutes rise together to reach an arbitrarily assigned value of 1.0 when the GFR reaches 6 ml/min. PD is then initiated using a prescription (described in Supplemental Table 1) which would provide a stdKt/V of 1.7 in an anuric patient. The initiation of PD reduces the level of each solute with the reduction being relatively largest for the solute which is partially reabsorbed by the native kidney. If solute generation remains constant, the plasma level of each solute will then rise again as residual function declines. Full-dose PD can maintain the plasma level lower than it was at the time of dialysis initiation only for the solute which is reabsorbed by the native kidney. PD, peritoneal dialysis.

Figure 5

The predicted effect of initiating low-dose PD is greater for solutes which have low residual kidney clearances. The figure shows the predicted effect of initiating PD using a prescription which provides a solute clearance of only 1.5 ml/min, such as might be obtained using a single long-duration exchange. If dialysis is initiated when the GFR is 6 ml/min, this low-dose PD would reduce plasma levels by more than 30% for a solute which is reabsorbed in the native kidney and has a residual clearance of 3 ml/min. The levels of solutes with higher residual clearances including solutes cleared by secretion would be reduced less. Successful initiation of dialysis with a single 2 L nocturnal icodextrin exchange has been reported by Jeloka et al.; however, the extent to which this relieved uremic symptoms was not described. PD, peritoneal dialysis.