Phosphate in dialysis patients.

: A single pool extracellular compartment (10-15 L) model was demonstrated to fit changes in plasma Pi (at steady state or rebound during HD) on the assumption that its G increased linearly or exponentially during HD. G at the end of HD was estimated to be 30-300 times that at the start of HD, which suggested rapid outflow of Pi from its reservoirs. When bovine or human RBC as a cellular model of probable Pi reservoirs were incubated with solute-free NSS or dialyzed plasma, BUN instantaneously reached a new equilibrium, but Pi and CR gradually flowed out from RBC. Pi efflux from RBC was observed even after the Pi concentration in the incubation media exceeded its initial plasma concentration. With regard to BUN and Cr, the concentration in RBC estimated from that in the incubation media decreased exponentially during HD corresponding to the change in their plasma concentrations. Estimated concentration of Pi in RBC, however, remained unchanged during HD in spite of significant decreases in its plasma concentration. These results suggest that Pi does not leave RBC by simple diffusion and that RBC contains Pi precursor(s). 31P NMR spectra obtained from packed RBC showed that uremic RBC contained more ATP and unidentified compounds (possibly sugar phosphates) than nonuremic RBC and that the peak amplitude of the latter unidentified compounds significantly decreased in uremic RBC leaving the dialyzer or at the end of HD. Furthermore, it was demonstrated that during incubation at 37 degrees C with NSS or dialyzed plasma, the peak amplitude of 2, 3-DPG decreased in contrast with an increased Pi peak.(ABSTRACT TRUNCATED AT 250 WORDS)