Endotoxin adsorption of various dialysis membranes: in vitro study

Endotoxin (ET) in the dialysate is known to adsorb on dialysis membranes made of polyether polymer alloy (PEPA) and polymethylmethacrylate (PMMA). In the present study, we investigated the adsorption of ET on dialysis membranes with a focus on PEPA membranes and polysulfone (PS) membranes that are extensively used in artificial kidneys or as ET-removal filters. In the case of PEPA, the compounding of polyvinylpyrrolidone (PVP) was changed, and both a hydrophobic version and a hydrophilic version were used on the blood side. For the PS dialysis, commercial membranes (APS (Asahi), BSP (Toray), PSN (Fresenius), CLPS (Terumo)) were used. Adsorption was evaluated by exposing both sides of the membrane after it had been primed with physiological saline: the ET concentration on the blood side and dialysate side of the dialysis membrane was monitored during the 240 min from the start of the exposure. When the PEPA membrane was investigated, ET was significantly adsorbed on the hydrophobic version. For PS membranes, ET was adsorbed on the blood side or on both the blood and dialysate sides, depending on the membrane. PS dialysis membranes can adsorb ET but the power and site of adsorption are different even between membranes made of the same material. In addition to electrostatic action attributable to the compounding of hydrophilic agent PVP on the dialysis membrane, the distribution of PVP that was compounded and the potential of the membrane itself are considered to cause differences in adsorption.

Is steam sterilization really making any difference in dialysis-induced cytokine release?

Ethylene oxide (ETO) is presently the most commonly used sterilization method for medical devices. Although alternative sterilization modes such as steam sterilization have been suggested, the effect of steam on dialysis-induced cytokine release is unknown. We enrolled 9 patients on chronic hemodialysis and evaluated at different intervals IL-1beta production while treated with ETO (NC 1785-Bellco) and steam sterilized NC 1785S-Bellco) Synthetically Modified Cellulose (SMC). A basal test during treatment with NC 1785 was performed (A); the same test was set up 4 weeks after treatment with NC 1785S (B) and, lastly, 4 weeks after returning to NC 1785 (C). Peripheral blood mononuclear cells (PBMC) were purified before and after the dialysis session, were isolated on a Ficoll/Hypaque gradient and incubated for 24 h. Spontaneous IL-1beta release was evaluated in the supernatant and in the lysate. In A, IL-1beta levels were (in pg/ml/10(6) cells, in supematant and lysate, respectively): 5.8 +/- 4.8 and 7.6+/-5.2 in pre-HD and 4.68 +/- 3.6 and 9.7 +/- 6.65 in post-HD. These levels showed a clear reduction in B: 2.5 +/- 2.2 and 4.4 +/- 3.1 in pre-HD, and 4.35+/- 6.6 and 7.52 +/- 7.22 in post-HD. In the C test, 4 weeks after the return to the ETO membrane, IL-1beta levels remained unchanged: 2.9 +/- 1.8 and 4.5 +/- 3.1 in pre-HD; and 2.6 +/- 3 and 5.7 +/- 6.6 in post-HD. Statistical analysis showed significant changes in the pre-HD levels both in supematant (p < 0.04) and in lysate (p < 0.04). Steam sterilization of SMC induced a lower spontaneous IL-1beta release, but this effect was not statistically significant due to the large inter-individual variation. Hence, contrary to claims of better biocompatibility, steam sterilization does not result in a reduced production of pro-inflammatory IL-1beta.

Effects of ethylene oxide and steam sterilization on dialysis-induced cytokine release by cuprophan membrane

The effects of sterilization modalities on dialysis-induced cytokine release are still unknown. To investigate these effects, 8 patients on chronic hemodialysis were enrolled for evaluating at different intervals interleukin-1beta (IL-1beta) and tumor necrosis factor-alpha (TNF-alpha) production (pg/ml/106). They were using a 1.3 m2 ethylene oxide (E3) or steam (E3S) sterilized Cuprophan membrane. The patients underwent a basal test with E3 (A1) and 2 following tests after 1 (B1) and 2 (B2) months of E3S treatment, respectively. Finally, the last test was performed 1 month after the switch to E3 (A2). Il-1beta predialysis release by mononuclear cells was 162 +/- 114 pg/ml/106 in A1, 185 +/- 129 pg/ml/106 in B1, and 226 +/- 138 pg/ml/106 in B2, then decreased to 123 +/- 134 in A2 (p < 0.07). Il-1beta postdialysis levels were 234 +/- 238 pg/ml/106 in A1, 429 +/- 285 pg/ml/106 (B1), and 438 +/- 473 pg/ml/106 (B2) with the steam membrane, decreasing to 204 +/- 134 pg/ml/106 in A2 (p < 0.01). TNF-alpha predialysis basal release (A1) was 826 +/- 817 pg/ml/106, 720 +/- 496 in B1, and 1079 +/- 515 pg/ml/106 in B2, and finally 680 +/- 588 pg/ml/106 in A2 (p < 0.03). In postdialysis TNF-alpha levels were 963 +/- 542 pg/ml/106 in A1, 1,226 +/- 541 pg/ml/106, and 1,183 +/- 776 in B1 and B2 respectively, and 388 +/- 297 pg/ml/106 in A2 (p < 0.003). Steam sterilization seems to induce a higher cytokine release by mononuclear cells when a Cuprophan membrane is used. This finding may be related to a less physiologic action of the steam in the case of Cuprophan membranes. Further studies are needed to clarify this hypothesis.

In vitro and in vivo TNFalpha synthesis modulation by methylguanidine, an uremic catabolyte

This study was performed in order to examine whether the uraemic toxin, methylguanidine (MG), can modulate tumor necrosis factor alpha (TNF alpha) release by activated macrophages. In this study we have evaluated the ability of MG to influence TNF alpha release in vitro, in Escherichia coli lypopolysaccharide- (LPS)-stimulated J774 cells preincubated overnight with MG, and in vivo in rats treated with MG before and after LPS challenge. Parallel experiments employing N(G)-nitro-L-arginine methyl esther (L-NAME) were also carried out for comparison. The effect of LPS (6 x 10(3) u/ml) on TNF alpha release by J774, following overnight incubation with MG or L-NAME (1 mM), was examined 3 hours after LPS challenge. LPS-stimulated J774 released 287.83+/-88 u/ml TNF alpha into the culture medium. MG (1 mM) significantly inhibited TNF alpha release by 73% (P<0.05). L-NAME (1 mM) significantly inhibited TNF alpha release too by 72.88% (P<0.05). The effect of MG and L-NAME have been also studied in vivo. Serum TNF alpha levels in LPS treated rats 2 h after LPS challenge were 88.33+/-31.7 u/ml as compared to the serum TNF alpha levels of control rats (undetectable). Treatment of rats with MG (30 mg/kg, i.p.) strongly and significantly reduced TNF alpha release (98.71% inhibition; with P<0.001); in the same experimental setting L-NAME (10 mg/kg, i.p.) also significantly reduced TNF alpha serum levels (76.47% inhibition; with P<0.01). These results could indicate that immune disfunction related to uremia may be related to the inhibitory capability of uremic catabolyte, MG, on TNF alpha synthesis and release.