Analysis of beta(2)-microglobulin kinetics in hemodialysis by a modified variable-volume one-compartment model

It has been reported that deposition of beta(2)-microglobulin (BMG) is associated with the occurrence of dialysis-related amyloidosis (DRA) in long-term hemodialysis (HD) patients. Though reduction of the BMG burden is essential in preventing DRA, simple BMG kinetic models applicable to clinical practice have not been established. We have reported a modified variable-volume one-compartment model (1CM) for analyzing urea nitrogen (UN) kinetics, in which no specific parameters other than a modification factor are necessary. If there is a constant relation between the BMG concentration of the dialyzer arterial line and the mean BMG concentration in the body during HD, the modified 1CM may also be applied to BMG. As such, in order to verify its validity, we analyzed UN and BMG kinetics by the modified 1CM in 28 HD patients in whom polysulfone dialyzers were used, and compared the calculated and measured solute rebound. In 3 of the patients, the spent dialysate was collected in a tank, and the BMG removal mass, calculated by the modified 1CM, was compared with that recovered in the tank. The BMG rebound ratio (%), calculated by the modified 1CM, was not different from the measured one (36.3 +/- 9.9 vs. 36.5 +/- 11.5, p = 0.954), as in the case of UN (15.8 +/- 4.5 vs. 16.3 +/- 4.6, p = 0.541). The solute dilution in the blood circuit by solute disequilibrium and blood recirculations for BMG was estimated to be 74% stronger than that for UN. The normalized solute generation rate (mg/min/l) and the time-averaged solute concentration over a 1-week period (mg/l) were 0.142 +/- 0.023 and 448 +/- 68 for UN and 0.00578 +/- 0.00125 and 22.4 +/- 4.6 for BMG, respectively. The differences between these two solutes resulted in a Dilution index for BMG that was 23% lower than that for UN (2.62 +/- 0.44 vs. 3.21 +/- 0.39, p < 0.0001). The modified 1CM overestimated the BMG removal mass (mg) by about 20% compared with that recovered in the tank (195 +/- 22 vs. 162 +/- 17, p < 0.0001). One of the causes of this discrepancy was speculated to be adsorption of BMG by polysulfone dialyzers. It was concluded that, despite several problems in quantitative analysis, the modified 1CM would be a useful model for estimating the dialysis efficiency for BMG removal in clinical practice.