Correcting acidosis in hemodialysis

Abstract

Control of uremic acidosis by hemodialysis carries the potential risks of reducing phosphate clearance and worsening metastatic calcification; modeling bicarbonate delivery has been proposed to adequately correct acidosis without impairing phosphate removal. To test the efficacy and safety of different methods for controlling acidosis, nine stable adults received in random order standard (S; dialysate HCO3- 30 to 34 mmol/L), high (H; 40 mmol/L) or modeled (M; 28 mmol/L, rising exponentially to 35 mmol/L at 3 h, 40 mmol/L at 4 h) bicarbonate dialysis for 4 wk each, and were tested during the last two dialyses of each treatment. More oral bicarbonate capsules were required with M than H (2.8 +/- 0.4 versus 1.4 +/- 0.4/day, P = 0.04) to maintain predialysis HCO3- at 24 to 26 mmol/L. Plasma HCO3- was significantly higher with H than M during dialysis, and than S before, during, and after dialysis. Plasma inorganic phosphate, phosphate rebound, clearance of phosphate from plasma (80 to 90 mL/min) and mass transfer of phosphate into dialysate (12 to 13 mmol/4 h dialysis) were no different among the three treatments. Similarly, there were no differences in plasma concentration of urea, total calcium, estimated ionized calcium, lipids, and potassium, clearance and mass transfer of urea, blood pressure, and symptoms with the three treatments. Estimated levels of tribasic inorganic phosphate, the phosphate component of hydroxyapatite, were very similar before and after each treatment. Plasma calcium x phosphate product was less than 3.5 mmol2/L2 at all times with each treatment. A risk factor for metastatic calcification was calculated from the relative saturation ratio of its principle component, hydroxyapatite (Ca5 (PO4)3 OH); this was no different among each of the treatments, and was not altered significantly by dialysis. Uremic acidosis can be fully corrected by high or modeled bicarbonate dialysis without any reduction of phosphate clearance or increased risk of metastatic calcification. The added cost of modeling technology is not justified by the criterion of phosphate clearance alone.