Contribution of residual function to removal of protein-bound solutes in hemodialysis

Abstract

Background and objectives: This study evaluated the contribution of residual function to the removal of solutes for which protein binding limits clearance by hemdialysis.

Design, setting, participants, & measurements: Solute concentrations were measured in 25 hemodialysis patients with residual urea clearances ranging from 0.1 to 6.2 ml/min per 1.73 m2. Mathematical modeling assessed the effect of residual function on time-averaged solute concentrations.

Results: Dialytic clearances of the protein-bound solutes p-cresol sulfate, indoxyl sulfate, and hippurate were reduced in proportion to the avidity of binding and averaged 8±2, 10±3, and 44±13% of the dialytic urea clearance. For each bound solute, the residual clearance was larger in relation to the residual urea clearance. Residual kidney function therefore removed a larger portion of each of the bound solutes than of urea. Increasing residual function was associated with lower plasma levels of p-cresol sulfate and hippurate but not indoxyl sulfate. Wide variation in solute generation tended to obscure the dependence of plasma solute levels on residual function. Mathematical modeling that corrected for this variation indicated that increasing residual function will reduce the plasma level of each of the bound solutes more than the plasma level of urea.

Conclusions: In comparison to urea, solutes than bind to plasma proteins can be more effectively cleared by residual function than by hemodialysis. Levels of such solutes will be lower in patients with residual function than in patients without residual function even if the dialysis dose is reduced based on measurement of residual urea clearance in accord with current guidelines.

Figure 1.

Time-averaged plasma concentrations (left panel) and generation rates (right panel) for UreaN and the solutes plotted against values for residual urea clearance (K_ru) in the 25 study subjects. Plasma levels declined significantly (bold lines) with increasing K_ru for PCS (r² = 0.34, P < 0.01) and HIPP (r² = 0.25, P < 0.01) but not for urea nitrogen (UreaN) or IS (dashed lines). Solute generation rates were highly variable among individual solutes but did not exhibit any significant relation to K_ru.

Figure 2.

The predicted effect on time-averaged solute concentrations of varying the residual urea clearance (K_ru) while solute generation rate is held constant (UreaN, solid line; PCS, – - – - -; IS, · · · · ; HIPP, – – – –). Solute concentrations at each level of K_ru are depicted as a percent of the value in a patient with no residual function (K_ru = 0 ml/min). (A) The predicted solute concentrations if dialysis were prescribed to provide a spKt/V_urea of 1.4 at each of three weekly 180-minute treatments regardless of the level of residual function. (B and C) The predicted solute concentrations if dialysis dose is reduced as allowed by KDOQI Guidelines when the residual urea clearance is >2 ml/min. (B) The predicted solute concentrations if dialysis provided a spKt/V_urea of 1.4 three times weekly for K_ru < 2 ml/min, and treatment time was reduced to provide a spKt/V_urea of 1.05 three times weekly for K_ru ≥ 2 ml/min. (C) The predicted solute concentrations if dialysis provided a spKt/V_urea of 1.4 three times weekly for K_ru < 2 ml/min but dialysis time is adjusted to provide a spKt/V_urea of 2.33 twice weekly for K_ru ≥ 2 ml/min. Concentration profiles were modeled as described in the Materials and Methods section using the average observed ratios of bound solute to urea clearance for the native kidney and for hemodialysis as given in Table 2.

Figure 3.

The relation of generation rates for the bound solutes PCS and IS. There was no apparent correlation between the generation rates for these two solutes, which are both derived from the action of colon bacteria on amino acids that escape absorption in the small intestine.