The host response to the presence of lipopolysaccharide (LPS) is complex and varied. Two closely related endogenous serum proteins, LPS-binding protein (LBP) and bactericidal/permeability-increasing factor (BPI), regulate delivery of LPS to CD14 antigen on effector cell surfaces and modulate the host response to LPS. In the study presented here, plasma levels of LBP and BPI were measured, predialysis, 15 min into dialysis and postdialysis in patients dialyzed with cellulose, cellulose-tri-acetate (CTA), and polysulfone dialyzers. Further, the association between LBP levels and BPI release during hemodialysis and clinical and laboratory characteristics of patients, complement activation represented by plasma C3a levels, and monocyte cytokine production represented by interleukin-1 receptor antagonist (IL-1Ra) synthesis was also studied. Predialysis plasma levels of LBP were 14,459 +/- 544, 13,889 +/- 1362 and 12,622 +/- 6305 ng/mL, respectively, with cellulose, CTA, and polysulfone dialyzers, and postdialysis levels were 17,834 +/- 861, 20,979 +/- 8485 and 18,177 +/- 1656 ng/mL, respectively. Postdialysis plasma levels of LBP were consistently higher than predialysis levels with all three dialyzers (P < 0.05). However, plasma LBP levels were not significantly different between the three dialyzers either predialysis (P = 0.28) or postdialysis (P = 2.8). There were no significant differences in predialysis BPI levels between the three dialyzers (P = 0.21). BPI levels at 15 min of dialysis with CTA (10.91 +/- 3.65 ng/mL) and polysulfone (10.73 +/- 2.24 ng/mL) dialyzers were significantly greater (P < 0.05) than that observed with cellulose (5.49 +/- 0.66 ng/mL). Similarly, postdialysis levels with CTA and polysulfone were significantly greater (P < 0.05) than that observed with cellulose dialyzers. The percentage change in BPI levels between predialysis and 15 min was 1341 +/- 243%, 2935 +/- 1033%, and 3790 +/- 1151% for cellulose, CTA, and polysulfone dialyzers, respectively. The changes in BPI levels from predialysis to 15 min and between pre- and postdialysis samples were statistically significant for all three dialyzers (P < 0.05). Postdialysis LBP:BPI ratios were 50 +/- 6%, 18 +/- 4%, and 22 +/- 6% of predialysis ratios for cellulose, CTA, and polysulfone dialyzers, respectively. These changes were statistically significant (P < 0.05) for all three dialyzers. There was no significant correlation between baseline clinical or laboratory characteristics and predialysis LBP levels. Similarly, the correlation between BPI levels at 15 min of dialysis with the clinical and laboratory characteristics was also poor, with the exception of serum albumin (r = 0.43, P = 0.008). The correlation between BPI levels at 15 min of dialysis with plasma LBP levels (r = -0.29; P = 0.08), plasma C3a levels (r = -0.1; P = 0.55), peripheral blood mononuclear cells (PBMC) content of IL-1Ra (r = 0.01; P = 0.94), and IL-1Ra production by unstimulated (r = 0.13; P = 0.45), and endotoxin-stimulated PBMC (r = 0.32; P = 0.06) was not statistically significant. The results of this study demonstrate that dialysis with cellulose, CTA, and polysulfone dialyzers results in a significant increase in LBP and BPI levels. BPI release is probably mediated by non-complement factors and may be related to the nutritional status of the patient. The release of BPI during HD and consequent lowering of the LBP:BPI ratio could potentially afford some protection against endotoxin in the dialysate.