Pyrogen transfer across high- and low-flux hemodialysis membranes

The extent to which bacterial products from contaminated dialysate enter a patient's blood depends upon the type and permeability of the hemodialysis membrane in use. This study was performed to assess the transfer of pyrogenic substances across both high- and low-flux membranes (DIAPES, Fresenius Polysulfone, Helixone, Polyamide S). All experiments were carried out in the saline-saline model. The dialysate pool was contaminated either with purified lipopolysaccharide (LPS) (250 and 500 EU/mL) or with sterile bacterial culture filtrates (20 EU/mL), and in vitro dialysis was performed under diffusive and convective conditions. A significant transfer of endotoxin was observed for both low- and high-flux DIAPES challenged with either LPS or with bacterial culture filtrates. Under identical conditions, no transfer of endotoxins was detectable across Fresenius Polysulfone and Helixone upon challenge with purified LPS. With bacterial culture filtrates, endotoxin concentrations for Polyamide S and Fresenius Polysulfone were about 10% and 1%, respectively, of those measured for DIAPES, whereas no transfer of endotoxin was detectable for Helixone. Using an alternative assay (induction of interleukin-1 receptor antagonist, IL-1Ra, in whole blood), only the DIAPES membrane showed the passage of cytokine-inducing substances. Thus, when saline is present in both the blood and dialysate compartments (i.e., the situation during predialysis priming procedures), dialysis membranes differ profoundly with respect to their permeability to endotoxins.

Differences in the permeability of high-flux dialyzer membranes for bacterial pyrogens

AIMS

The increasing use of high-flux membranes for hemodialysis has raised concerns that patients dialyzed with these membranes may be at higher risk of being exposed to cytokine-inducing bacterial substances in the dialysate than patients dialyzed with low-flux membranes. We investigated the permeability of various high-flux membranes for both purified E. coli lipopolysaccharide (LPS) as well as for LPS derived from Stenotrophomonas (Sten.) maltophilia.

MATERIALS AND METHODS

An in vitro dialysis circuit with saline in the blood compartment of 3 dialyzers containing different membranes (polysulfone, helixone and Diapes) was employed. The dialysate was challenged with increasing doses of sterile filtrates derived from Sten. maltophilia cultures or with purified LPS from E. coli. Samples from the blood compartment were tested for cytokine induction (IL-1beta, IL-6 and TNF) in mononuclear cells as well as for LPS by limulus amebocyte lysate test (LAL).

RESULTS

IL-6 induction above sterile controls (< 0.02 ng/ml IL-6) was observed by samples from the blood side of DIAPES dialyzers (1.2 +/- 0.7 ng/ml IL-6) after challenging the dialysate with 4.1 +/- 3.6 U/ml E. coli LPS (9.9 +/- 4.5 ng/ml IL-6). In contrast, at the same challenge dose no significant IL-6 induction above sterile controls was observed by blood side samples of polysulfone (0.15 +/- 0.07 ng/ml) and helixone (0.09 +/- 0.05 ng/ml) dialyzers. Increasing the amount of E. coli LPS in the dialysate further augmented IL-6 induction by blood side samples of Diapes but not of polysulfone and helixone dialyzers. Similar results were obtained for IL-1beta and TNF. After challenging the dialysate with E. coli LPS as well as with cultures of Sten. maltophilia, significantly more LAL reactivity was observed in the blood compartment of Diapes compared to polysulfone and helixone.

CONCLUSIONS

There are considerable differences between high-flux membranes regarding their permeability for cytokine-inducing substances from E. coli as well as for LPS derived from E. coli and Sten. maltophilia. Dialyzers that leak CIS under aqueous conditions in vivo should not be used unless the dialysate has passed through an ultrafilter.

Transfer of endogenous pyrogens across artificial membranes?

Synthetic high-flux dialyzer membranes used in continuous veno-venous hemofiltration are permeable to middle molecular size endogenous pyrogens, the pro-inflammatory cytokines IL-1 beta and TNF-alpha. The quantities removed by sieving are, however, negligible in vitro as well as in vivo. Adsorption of cytokines to the membrane polymer is the major mechanism of pyrogen removal. Adsorption seems to be semispecific for pro-inflammatory cytokines because levels of anti-inflammatory mediators were not changed or even increased during CVVH. Thus, CVVH may change cytokine profiles in septic patients supporting the predominance of anti-inflammatory over pro-inflammatory activity in plasma. It remains to be demonstrated whether modifications of extracorporeal blood purification systems (high-volume CVVH, plasma separation + adsorption) are able to amplify the change in cytokine profiles and whether this change influences outcome of septic patients.

Influence of perflubron emulsion particle size on blood half-life and febrile response in rats

Perfluorochemical (PFC) emulsions are particulate in nature and, as such, can cause delayed febrile reactions when injected intravenously. This study investigated the influence of emulsion particle size on intravascular retention and on body temperature changes in unrestrained conscious rats. Concentrated (60% to 90% w/v) emulsions based on perflubron (perfluorooctyl bromide [PFOB]) with mean particle sizes ranging from 0.05 microns to 0.63 microns were tested. Rats were fitted with a chronic jugular catheter and an abdominal body temperature telemetry unit. Fully recovered, conscious rats were monitored for 24 hours after infusion (dose = 2.7 g PFC/kg). Emulsion blood half-life (T1/2) was determined from blood perflubron levels measured by gas chromatography. Emulsions with a particle size of 0.2-0.3 microns caused fevers (6 to 8 hour duration) which peaked at 1-1.5 degrees C above normal (approximately 37.5 degrees C). Fevers could be blocked by i.v. treatment with either cyclooxygenase inhibitors (ibuprofen) or corticosteroids (dexamethasone). Both intensity and duration of the temperature response, quantified by area under the temperature curve, was decreased significantly for emulsions with a particle size < or = 0.12 micron. Blood T1/2 varied inversely with particle size, and was 3 to 4 fold longer for emulsions with a mean particle size < or = 0.2 micron. Thus, smaller emulsion particles more effectively evaded the reticuloendothelial system, which resulted in longer intravascular retention, less macrophage activity, and reduced febrile responses.

Measurement of back clearance

With improvement in dialysis membranes, back filtration becomes the focus of study. Because new membranes that remove low molecular weight proteins, such as beta 2-microglobulin by diffusion are now available for clinical use, both back filtration and back diffusion become problems. This paper gives the mass transfer rate of solutes from the dialysate to blood compartment. As markers of toxic or antigenic substances, inulin (MW 5,200), lysozyme (MW 14,300), and alpha-lactoalbumin (MW 14,400) were used. The dialyzers tested were the CLSU-12W, AM-FP10, BK-1.0P, and FB-110U. The UFR controller set the flow rate on the blood side at 200, dialysate at 500, and ultrafiltration (QF) as 0 to 50 ml/min. Under these conditions, the inulin back clearance is 16 (CLSU) to 26 ml/min (FB), even if the QF is 50 ml/min, and is 10 to 34 ml/min for lysozyme at QF = 0 ml/min. Although a substantial amount of solute moves into the blood compartment, back clearance does not drastically decrease with increase in QF. To avoid contamination of the blood side, clean-up techniques for the dialysis line or online protein adsorber should be followed.