Uremic toxicity: present state of the art.

The uremic syndrome is a complex mixture of organ dysfunctions, which is attributed to the retention of a myriad of compounds that under normal condition are excreted by the healthy kidneys (uremic toxins). In the area of identification and characterization of uremic toxins and in the knowledge of their pathophysiologic importance, major steps forward have been made during recent years. The present article is a review of several of these steps, especially in the area of information about the compounds that could play a role in the development of cardiovascular complications. It is written by those members of the Uremic Toxins Group, which has been created by the European Society for Artificial Organs (ESAO). Each of the 16 authors has written a state of the art in his/her major area of interest.

Preparation of and optimal module housings for hollow fibre membrane ion exchangers.

Macroporous polyamide 6 hollow fibres can be polymer coated by a three-step procedure: first, reaction of the amino end groups with a bifunctional, double-bond-containing reagent; second, block polymerization with different monomers; and third, polymer analogue reactions with amines or sulphite salts to produce ion exchanger groups. The densities of double bonds are dependent on the amino densities and are in the range of 20-30 mumol/g polyamide 6. The ion exchanger fibres were packed in different types of module housings to get an optimal separation unit. The best housing seems to be a so-called single-dead-end arrangement of fibres. Three types of ion exchanger hollow fibres have been produced: a weak and a strong anion exchanger and a strong cation exchanger. The dynamic protein-binding capacities are in the range of 40 mg/ml membrane. Using these membrane modules, it is possible to separate proteins in the same way as with particle-based ion exchangers. Fast protein separations with low pressure drop are possible.

Membranes and polymer structures--biocompatibility aspects with respect to production limits.

Plasmapheresis can be performed by centrifugation and by use of membrane technology. With the latter technique we receive a plasma which is absolutely free from platelets. This is why membranes are gaining market shares in this particular field of medical application. Today plasmapheresis membranes are mostly fabricated from synthetic polymers, such as polypropylene (e.g. PLASMAPHAN), polysulfone, polyacrylonitrile, polymethylmethacrylate, polyvinylalcohol and others, the only exception being cellulose acetate. Parameters determining the biocompatibility of plasmapheresis membranes are generation of complement C3a or C5a, hemolysis and possible thrombus formation. These parameters depend on various properties of the membrane polymer: e.g. the nature of the molecular end/side-groups, the distribution of electrical charges on the polymer surface and the different chemical structures and conformation of the polymer. In addition, membrane properties like pore distribution and geometry or the flow characteristics of a particular device-design may trigger cell activation or influence biocompatibility through the adsorption of various plasmacomponents. Most of the polymers which are used today for manufacturing plasmapheresis membranes have not been developed for this purpose. They were originally selected to be used as textile fibers. Further, no present membrane polymer has been specifically developed to achieve high biocompatibility. The membrane profile was designed in such a way that pheresis properties were met rather than optimizing biochemical blood/polymer interactions. One reason for this decision may be that the market volume of plasmapheresis technology is too small in order to justify specific and high-cost developments of polymers for this purpose. Polymer selection to achieve excellent biocompatibility profiles is determined by polymer-availability, costs, membrane-forming processes and environmental aspects related to possible pollution during the manufacturing process. The production of PLASMAPHAN by the unique Accurel-process combines several of these parameters. The main membrane production processes and especially the Accurel-process are described here. The influence of polymer-surface properties, membrane structure and module-design on the biocompatibility of plasmapheresis treatments are discussed and explained by appropriate examples.

Biocompatibility and membrane development.

Development of new biocompatible membranes for clinical application needs the expertise of various specialists, chemists, chemical engineers, and clinicians. As the biocompatibility of a membrane cannot be considered in terms of a single biochemical pathway, due to the interrelation between the complement, coagulation, and inflammatory systems, it is difficult to achieve optimal results with one single polymer. A compromise between different approaches has therefore to be found. Membrane development also needs sophisticated test systems, which simulate the clinical situation as closely as possible. Adequate results are achieved with the 'ex vivo' model, which represents open-loop haemodialysis. The ex vivo model gives good results which are close to those from clinical experiments.

Comparison of polyamide and polypropylene membranes for plasma separation.

Plasma separation experiments were made with polyamide experimental prototype hollow-fiber plasma filters with surface areas between 0.025 m2 and 0.1090 m2 using bovine blood collected in acid citrate dextrose (ACD). The maximum filtration velocity rose with the wall shear rate gamma w as gamma w 0.72 +/- 0.02 and decreased with the length of fiber L as L-0.41 with a correlation coefficient of 0.97 +/- 0.02. The results were similar to those with polypropylene fibers. We also investigated the occurrence of hemolysis as a function of shear rate and transmembrane pressure. The free hemoglobin concentration of filtered plasma was checked using a U.V. spectrophotometer. It was concluded that polyamide membrane filters can be safely used for plasma separation from blood.

Dialysate contamination and back filtration may limit the use of high-flux dialysis membranes.

Endotoxins, or fragments thereof, can reach the blood stream of dialysis patients, transported by diffusion and connection across the intact high-flux membrane. This transfer depends upon the phenomenon of back filtration. Back filtration generally occurs under conventional high-flux dialysis conditions with membranes having an ultrafiltration coefficient in blood (UF-C) above 20 ml/hr/m2/mmHg. The clinical consequences of back filtration vary from center to center depending primarily on the quality of dialysate. We therefore surveyed the bacterial and endotoxin levels of purified water and effluent dialysate in a cross section of dialysis centers in the central United States. Using a high recovery medium, we found that 53% of the centers had bacterial counts above the Association for the Advancement of Medical Instruments standard in water (20% cfu/ml) and 35% above the standard in dialysate (2,100 cfu/ml). Endotoxin concentrations higher than 5.0 EU/ml in both water and dialysate were found in 4% and 11.8% of the centers, respectively. Since high-flux membranes are believed to be of benefit for long-term dialysis patients, manufacturers will have to offer dialysate preparation systems with additional safety features. The proper membrane design will be a key to the success of such systems.

Cellulosic versus synthetic membranes: a reasonable comparison?

Of the two main classes of dialysis membranes, cellulosic and synthetic, the former represents the standard membrane used in dialysis therapy. The disputed properties of cellulosic membranes related to biocompatibility have inspired a number of authors to compare these two membrane classes from a variety of perspectives. However, such a strict categorization as "synthetic" or "cellulosic" is of doubtful value from the point of view of polymer chemistry. Here, biocompatibility and performance properties of these two membrane classes are compared with the aim of investigating the validity of this categorization. The biocompatibility parameters studied are complement and leukocyte activation together with activation of the coagulation cascade. Analysis of a variety of both cellulosic and synthetic membranes with different degrees of biocompatibility showed that biocompatibility can be achieved by both classes of membranes and is, therefore, not a particular property of one class only. Furthermore, performance and beta 2-microglobulin removal properties of the two classes of membranes do not particularly favor one of these classes. Therefore, differences between cellulosic and synthetic membranes are not manifested in parameters like biocompatibility, hydraulic permeability, and overall performance.

Backfiltration: does it occur in membrane plasma separation?

Backfiltration in hemodialysis refers to local filtration of dialysate into the blood compartment. This investigation was performed to examine whether there is a similar backfiltration phenomenon in membrane plasma separation. It was shown experimentally in hollow fiber devices that there is a flow of plasma through the filtrate side parallel to the blood flow inside the fiber. This bypass flow includes a backfiltration of the plasma from filtrate to the blood compartment near the end of the module. The following experimental results show the existence of bypass flow and backfiltration. In the case of no net filtration, these findings were made: 1) the existence of an offset of mean transmembrane pressure and 2) the blood side pressure drop in ACD-blood was less in impermeable fibers than in permeable ones. Filtrate pressure is higher than module outlet pressure at a wide range of filtration rates. In summary, backfiltration is not restricted to zero net filtration but occurs under the conditions of clinical use. Therefore, backfiltration is a crucial issue for device safety, because any contamination or wash-out from the membrane can reach the blood side.

Successful long-term use of a miniaturized plasmapheresis circuit in rabbits.

Dual lumen silicon rubber right atrial catheters were implanted into the jugular of 8 rabbits and tunneled subcutaneously to exit sites between the ears. Using a miniaturized tubing-pump system, blood flow rates of 25 ml/min could be achieved for up to 3 1/2 hours without sign of hemolysis in an extracorporeal blood circuit. Seven catheters functioned an average of 75 +/- SE 19 days (range 24-117). One catheter remains functional after 176 days. Infection and thrombosis were the main reasons for failure. 20 plasmaphoresis experiments were carried out in four heparinized rabbits (blood flow rates 15 ml/min, plasmaflux 1.5-2.0 ml/min) using polypropylene minifilters (average pore size, 0.55 micron) with the plasma recirculated back into the animal. No hemolysis was detectable throughout the 4 hr experiment. Plasma proteins with a MW of 69 X 10(3) to 3 X 10(6) (Albumin, LDH, SGOT, SGPT, CPK, fibrinogen, LDL) showed a sieving coefficient close to 1.0. The good filtration performance and the absence of side effects make this system a possible use for plasmaphoresis in neonates.

Modified cellulosic dialyzer membranes: an investigative tool in thrombogenicity studies.

We have previously demonstrated that chemical modification of cellulosic membranes with dimethyl-amino-ethyl (DEAE) groups significantly improves membrane properties in terms of biocompatibility. Here, we show that DEAE substitution also alters the membrane's thrombogenic properties, and cellulosic membranes with various amounts of DEAE substitution were produced. Clinical dialyzers were constructed using two experimental membrane materials: modified cellulose-low (MC-low) and MC-high; standard unsubstituted cellulose was used as a control. Six patients were treated for a period of 3 weeks with each type of dialyzer and a heparin dose of less than 6000 IU/treatment. MC-low exhibited less extracorporeal beta-thromboglobulin and thromboxane B2 release than MC-high or Cuprophan. In addition, residual blood volume after clinical use was lower in the MC-low type. MC-low and MC-high induced less complement activation than Cuprophan, as characterized by extracorporeal C5a and C3a plasma concentrations (75% less C5a generation and 50 to 70% less C3a generation than unsubstituted cellulose).

Optimal dimensions of capillary membranes for plasma separation.

The work presented in this article examines the relationship between the efficacy of the design (membrane consumption) and the design parameters for plasma separation modules. A computer simulation program for the design of hollow fiber modules was developed. It is based on a formula for filtrate flux prediction by Jaffrin. The limiting conditions set by red and white blood cell lysis are also taken into account. The results show that membrane area consumption is strictly related to the internal hollow fiber diameter. Low-efficiency devices (low infiltrate flux) can be designed nearly optimally, using 330-micron fibers. On the other hand, for high-efficiency devices, our model predicts lowest membrane consumption and, therefore, lowest costs using 220-micron fiber diameter. Furthermore, the results demonstrate that shear rates in commercially available plasma filters are too low.

Ethylene oxide in dialyzer rinsing fluid: effect of rinsing technique, dialyzer storage time, and potting compound.

Ethylene oxide (ETO) is recognized as one of the main causes of dialyzer-associated hypersensitivity reactions. We studied the amount of ETO in the rinsing fluid of ETO-sterilized hollow-fiber dialyzers as a function of rinsing technique, dialyzer storage time, and the amount of potting compound (known to be an ETO reservoir) in the dialyzer. The results suggested that the initial 500 ml of rinsing fluid removes much of the residual ETO in the dialyzer. Ethylene oxide extraction was enhanced substantially by rinsing at 37 degrees C versus 5 degrees C. However, considerable amounts of ETO remained in the dialyzer after an initial 500 ml rinse, some of which could be removed by rinsing with an additional 1,500 ml. High concentrations of ETO were measured in fluid that had been recirculated through the dialyzer for 10 min or longer and in fluid that had been allowed to remain in the dialyzer for 10 min under zero-flow conditions. The amount of ETO in the rinsing fluid decreased markedly as the dialyzer storage time was increased from 4 to 8 weeks and in dialyzers in which a portion of the potting compound had been replaced with a polycarbonate ring. Our results suggest that the dose of ETO administered to the patient at the outset of dialysis can be minimized by rinsing the dialyzer with 2 L of fluid at 37 degrees C and by avoiding administration of rinsing fluid that has been allowed to remain in contact with the dialyzer for more than several minutes. Use of a long storage interval and use of dialyzers containing reduced amounts of potting material will also reduce the ETO load.

Antibody responses to hemodialysis-related antigens in chronic hemodialysis patients.

Allergic-type reactions during hemodialysis are sometimes due to sensitization to ethylene oxide. To examine the possibility that additional antigens might be a basis for unexplained reactions, antibodies to formaldehyde and phthalate-related antigens and to dialyzer extracts were measured. Unselected sera from 113 chronic hemodialysis patients (CHP) and 200 control subjects were tested for IgG antibodies to formaldehyde-treated human serum albumin (HSA). The IgG antibody activity was confirmed in sera of five CHP who had used formaldehyde-treated dialyzers. These antibodies also reacted with formaldehyde-treated red blood cells. Sera from 71 CHP and 80 controls were tested for IgE antibodies to diethylphthalate-treated HSA; antibody was detected in two CHP sera. With extracts from hollow-fiber dialyzers, IgG antibody was detected in approximately 1/3 and IgM antibodies in approximately 1/2 of CHP sera. This antibody was found in comparable numbers of control sera. It was concluded that these additional substances are immunogenic and could be involved in allergic-type reactions.

Ex vivo biocompatibility evaluation of a new modified cellulose membrane.

To evaluate membrane biocompatibility, an open loop ex vivo model was designed simulating the hemodialysis procedure. Blood was withdrawn continuously from healthy nonuremic donors, heparinized, and pumped through a module containing the membrane to be studied. C3a generation in the module was determined at various time points comparing the cuprammonium cellulose (CC) membrane and four types of modified cellulose (MC) membrane, each with a different degree of hydroxyl (OH-) group substitution. In other studies, C3a generation in the ex vivo mode was compared with that during in vivo dialysis. In the ex vivo model, C3a generation with MC membranes was reduced by 70% compared with CC. However, within the MC group, the degree of C3a generation did not correlate with the degree of OH-group substitution. In vivo studies confirmed the reduced degree of C3a generation with the MC membrane compared with CC. Additionally, validation studies using the CC membrane showed excellent agreement between C3a generation during ex vivo perfusion and in vivo dialysis. The results suggest that a group of new MC membranes causes substantially less complement activation than the CC membrane but that the degree of complement activation with various subtypes of MC membranes is not related to the degree of OH-group substitution.

Dialyzer membranes: effect of surface area and chemical modification of cellulose on complement and platelet activation.

Using an ex vivo model, the effects of membrane composition and surface area on both the complement system (as reflected by plasma C3a levels) and platelets [as indicated by plasma concentrations of thromboxane B2 (TXB2) and platelet factor 4 (PF4)] were studied. In this model, polyacrylonitrile (PAN) was associated with less complement activation than cuprammonium cellulose (CC). A new "modified cellulose" (MC) membrane, in which a small number of the free hydroxyl groups on cellulose are substituted with a tertiary amino compound, was also associated with a low degree of complement activation, similar to that with PAN. However, the extent of hydroxyl group substitution in four MC membrane subtypes did not correlate with the reduction in complement activation. In studies using CC, the amount of generated C3a correlated with the membrane surface area, although the relationship was curvilinear. Plasma concentrations at the "dialyzer" outlet of TXB2 and PF4 were similar with CC, PAN, and MC. In studies with the MC subtypes, increasing the extent of hydroxyl group substitution paradoxically increased, albeit slightly, the amount of TXB2 generation. In studies with CC, a linear relationship between membrane surface area and TXB2 generation was found. The results suggest a dissociation between platelet and complement effects among different dialyzer membranes, and underline the importance of membrane surface area.

A clinical study on different cellulosic dialysis membranes.

A controlled clinical study was performed over a period of 8 weeks in two dialysis centres (Rostock, GDR, and Munich, FRG). The aim was to compare a dialysis membrane made of modified cellulose (Hemophan) with classical regenerated cellulose (Cuprophan). Dialysers containing these membranes, together with a cellulose acetate dialyser, were therefore incorporated in a cross-over programme and clinical and biochemical investigations undertaken. The efficacy of the modified cellulosic membrane with respect to urea and creatinine clearance was shown to be comparable to that of regenerated cellulose and cellulose acetate. However, modified cellulose showed an increased clearance for inorganic phosphate, significantly different from that demonstrated by both regenerated cellulose and cellulose acetate. Blood compatibility studies, which included the assessment of C3a activation and the reduction of white blood cell (WBC) and platelet count, clearly demonstrated that in comparison to regenerated cellulose, modified cellulose resulted in significantly less complement activation and WBC reduction. Similarly in comparison to cellulose acetate, modified cellulose showed reduced complement-activating and WBC-reducing properties. The reason for the improved blood compatibility of modified cellulose is not, as was originally assumed, related to binding of complement-inhibiting heparin, but appears instead to be due to the substitution of hydroxyl groups of regenerated cellulose.

Ex vivo model for pre-clinical evaluation of dialyzers containing new membranes.

The ex vivo model which reflects hemodialysis modulating factors during the first twenty minutes of blood membrane interaction, is applicable as a pre-clinical test for new membranes. The biocompatibility of a new cellulosic membrane (MC) proved to be superior to regenerated cellulose and comparable to synthetic membranes such as PAN regarding complement activation.