Effect of acetate, bicarbonate dialysis, and acetate-free biofiltration on nitric oxide synthesis: Implications for dialysis hypotension

The effect of acetate dialysis (AD), bicarbonate dialysis (BD), and acetate-free biofiltration (AFB) on nitric oxide (NO) synthesis and the implications for dialysis hypotension was studied. The finding that uremic plasma is a potent inducer of NO synthesis by endothelial cells in vitro suggested that the cardiovascular instability of dialysis patients might result from excessive NO formation. Cardiovascular instability is more frequent in patients undergoing AD than BD. To see whether these differences were attributable to NO, we studied the NO synthetic pathway ex vivo in patients undergoing different dialysis procedures. Five patients were treated, in a random order, with AD, BD, and AFB, a technique using a buffer-free dialysate and postdilution of a sterile bicarbonate solution. Each type of dialysis was used for 1 week, comprising three dialysis sessions. A polyacrylonitrile dialyzer was used for all three methods. Before and after the third dialysis, plasma was collected, added to [3H]L-arginine, and incubated with human umbilical vein endothelial cells (HUVECs) for 24 hours. NO synthesis was evaluated as [3H]L-citrulline formation. Plasma concentrations of interleukin-1beta (IL-1beta), a potent inducer of inducible NO synthase (iNOS) in endothelial cells, were also measured. Plasma collected from patients after AD stimulated endothelial NO synthesis more than plasma from the same patients before the dialysis session (pre-AD, 0.173+/-0.028 nmol/10(5) cells v post-AD, 0.280+/-0.093 nmol/10(5) cells; P < 0.05). A slight, although not significant, increase was also observed when HUVECs were incubated with plasma drawn after BD (pre-BD, 0.151+/-0.014 nmol/10(5) cells; post-BD, 0.230+/-0.055 nmol/10(5) cells). AFB did not aggravate the stimulatory effect of uremic plasma on endothelial NO synthesis (pre-AFB, 0.184+/-0.038 nmol/10(5) cells; post-AFB, 0.189+/-0.040 nmol/10(5) cells). Plasma IL-1beta was greater (P < 0.01) after AD than after BD and AFB (post-AD, 0.234+/-0.028 pg/mL; post-BD, 0.124+/-0.019 pg/mL; post-AFB, 0.120+/-0.013 pg/mL). With AD, there was a greater intradialytic decrease in systolic blood pressure than with BD or AFB. Weight and blood volume loss and sodium balance were similar in AD, BD, and AFB. These data were consistent with the possibility that NO and cytokines, released in excessive amounts during AD, may contribute to hemodynamic instability.