Osmotic agents and buffers in peritoneal dialysis solution: monocyte cytokine release and in vitro cytotoxicity

Peritonitis remains a major problem in peritoneal dialysis. The incidence of peritonitis may be reduced by the use of more "biocompatible" peritoneal dialysis solutions that do not impair local host defense mechanisms, such as occurs with conventional lactate-buffered glucose solutions. In the present study, we investigated the use of bicarbonate and lactate as buffer systems and glucose, amino acids, and glucose polymer as osmotic agents on specific cellular functions of isolated fresh blood monocytes in vitro. The bicarbonate-buffered solutions had a physiologic pH (7.0 to 7.6). Lactate-buffered solutions were tested with a pH between 5.5 and 7.3. RPMI 1640 (Roswell Park Memorial Institute, supplied by Biochrom, Berlin, Germany) and phosphate-buffered saline were used as control mediums. The test solutions were incubated with 200,000 monocytes/mL for 45 minutes followed by a 1:1 mix with RPMI 1640 (with supplements) during a 24- or 4-hour tetrazolium bromide test (MTT test) recovery period. Constitutive and lipopolysaccharide (LPS)-stimulated release of interleukin-1beta (IL-1beta) and IL-6 in the supernatants as parameters of cellular host defense and lactate dehydrogenase concentrations and MTT-formazan production as parameters for cell cytotoxicity were measured. Significantly higher IL-6 and IL-1beta release was found in the bicarbonate-buffered solutions, both under basal conditions and after LPS stimulation, compared with the lactate-buffered solutions (LPS stimulation: 1% amino acids/34 mmol/L bicarbonate, IL-1beta: 1,166 +/- 192 pg/mL; 1.5% glucose/34 mmol/L bicarbonate, IL-1beta: 752 +/- 107 pg/mL; 1.5% glucose/35 mmol/L lactate/pH 5.5, IL-1beta: 174 +/- 51 pg/mL). Some of these differences could even be detected in spent dialysate after a 6-hour dwell in continuous ambulatory peritoneal dialysis patients (n = 10). A lower degree of cellular cytotoxicity (lactate dehydrogenase activity) and better-preserved metabolic activity (MTT test) also were found for the bicarbonate-buffered solutions. Amino acids (1%) proved to be comparable to glucose (1.5%) as an osmotic agent at a neutral pH with regard to LPS-stimulated cytokine release and cytotoxicity. The incubation with a glucose polymer solution (7.5% glucose polymer in phosphate-buffered saline, pH 7.3) resulted in a significantly lowered cytokine release (LPS stimulation: IL-1beta, 69 +/- 19 pg/mL) compared with the other solutions with neutral pH (P < 0.01). These results suggest that bicarbonate as a buffer provided better biocompatibility with regard to mononuclear cytokine release and viability compared with lactate. Amino acids and glucose were equivalent to these parameters at a physiologic pH. The glucose polymer solution, however, was associated with a marked depression of cytokine release.