Back transport of bacterially derived products from dialysate fluid to the circulation has been correlated with both acute and long-term complications in hemodialysis (HD) patients. According to recent literature, acute pyrogenic reactions such as fever, chills, and hypotension occur in approximately 0.5–5 per 1,000 HD sessions. 1–3 In addition, nausea, myalgia, headache, and sleepiness have been attributed to the use of nonsterile dialysate. 4 It has been suggested that the repeated transmembrane passage of bacterial fragments activate peripheral blood mononuclear cells (PBMC), resulting in the intermittent secretion of a variety of proinflammatory cytokines, including interleukin-1α (IL-1α), interleukin-6 (IL-6), and tumor necrosis factor α (TNFα). 5 Due to this repetitive stimulation, a chronic microinflammatory state is induced, which may contribute to the development of long-term HD related complications such as dialysis related amyloidosis, 1,6,7 malnutrition, 8,9 accelerated atherosclerosis, 10 and increased mortality. 11 Several in vitro studies suggest that sterile dialysate is crucial to prevent back transport of these cytokine-inducing substances.
Therefore, it seems mandatory that water used for dialysis treatment meet stringent microbiological and nonpyrogenic criteria. However, these in vitro experiments were performed under highly nonphysiologic conditions. 12 In many of these studies lipopolysaccharide (LPS) challenge levels were 500 to 1,000-fold greater than the maximum accepted dose in clinical HD. 13 In fact, data from controlled clinical studies are lacking and evidence based guidelines are not available. Hence, it may not be surprising that official standards to prepare dialysis fluid and the final dialysate itself, vary markedly (Table 1). The strictest standard is met in the European Pharmacopoeia, with limits of 100 colony forming units per milliliter (CFU/ml) for bacteria and 0.25 endotoxin units (EU)/ml for LPS in dialysis water. 14 Higher bacterial counts are accepted by the American Association of Medical Instrumentation (AAMI), with levels up to 200 CFU/ml in the dialysis water and up to 2,000 CFU/ml in the final dialysate. 15 As for the maximal LPS concentration in dialysis fluid, no recommendations are given. In clinical practice, several studies have shown that many HD centers worldwide do not comply with the above prescriptions. 16–18 A survey of Greek centers shows that the total bacterial counts in dialysate exceeded the AAMI standards in 36.3%. 19 Furthermore, the situation is potentially aggravated by the widespread use of liquid bicarbonate concentrate, in conjunction with high flux membranes, predisposing to bacterial growth and back filtration of bacterial fragments. 20 Therefore, even more stringent criteria for the purity of the dialysis fluid have been recently formulated, e.g., < 0.1 CFU/ml for bacterial growth and 0.03–0.05 EU/ml for the presence of endotoxins. 21,22
Recently, dialyzers with superior permeability for middle and large molecular weight uremic toxins, 23 termed superflux dialyzers, were introduced in our center. As we moved to a new center, a new dialysate delivery system based on current recommendations, with a few modifications, 24–26 was constructed. The present report describes both the design and the effectiveness of this water treatment system to produce low to moderately contaminated dialysate without the need for a bacterial filter. In addition, bacterial counts and LPS concentrations in both the dialysate and blood were compared between groups of patients undergoing long-term dialysis treatment with two types of superflux dialyzers (polysulfon [PS] and cellulosic tri-acetate [CTA]) and a group of chronic HD patients undergoing HD treatment with a standard high flux dialyzer (PS). In the case of superflux CTA, standard dialysate was compared with ultra-pure dialysate.
Materials and Methods
Before participating in the study, all patients were dialyzed three times a week using a high flux PS dialyzer. Thirty-seven patients were randomized in two successive periods to HD with either a high flux PS (F 60), a superflux PS (F 500S), a superflux CTA (Tricea 150G) or a superflux CTA dialyzer with filtered dialysate (Tricea 150Gf) for 12 consecutive weeks, resulting in 74 periods in which measurements were obtained. To avoid carry-over effects, a washout period of 1 month was instituted. Blood samples were taken at the end of each study period from the afferent line before dialysis (t0) and from the efferent line 180 min (t180) afterward. Samples were analyzed for limulus amebocyte lysate (LAL) reactivity. Dialysate samples were taken at t180 from the outlet port of the dialysis machine and were cultured for bacterial growth and analyzed for LAL reactivity.
Dialysis Procedure and Materials
The dialysis sessions lasted 3 to 5 hours, depending upon the individual prescription of the patient. Only first-use dialyzers were used. Characteristics of the three dialyzers used in this study (PS: F 60 and F 500S, Fresenius, Bad Homburg, Germany; CTA: Tricea 150G, Baxter, Osaka, Japan) are depicted in Table 2. As for the superflux dialyzers, these devices have been designed to maximize convective transport by increasing the pressure drop along the fibers of the membrane. Furthermore, the pore size and/or distribution of the pores influences the permeability. These characteristics result in a better clearance of middle and high molecular weight substances, as measured by β2-microglobulin (J. Vienken, personal communication). Filtered dialysate was obtained by the interposition of a bacterial PS filter (SPS 600, Fresenius, Bad Homburg, Germany). 27 According to the individual needs of the patients, blood flow and ultrafiltration (UF) rates were kept constant between 200 and 250 ml/min and between 300 and 1,000 ml/h, respectively. Isolated UF was not performed. Bicarbonate powder (BiBag) was used for preparation of the dialysate. Dialysate flow was 500 ml/min. Anticoagulation was achieved by dalteparin with an initial dose of 2,500 to 6,000 IU, followed by an extra dose of 500 to 1,000 IU during the dialysis treatment, if necessary. Individual conditions (blood flow, UF, dalteparin dose) were kept stable throughout the study period.
Water Treatment System
The water system (Figure 1) is supplied with drinking water from the community waterworks piping grid. Two consecutive coarse filters (100 μm and 10 μm) and two microfilters (5 and 1 μm) are installed to remove suspended particles. Next, water hardness is sharply reduced from 13° to 2° German Hardness by two parallel connected softeners. Downstream from the softener, a 100 L buffer tank separates the water from the main water supply and a high pressure pump delivers water from the tank to the reverse osmosis system (RO). The RO input pressure is 14.6 bar (212 PSI). The RO, represented in Figure 1 as one membrane, actually consists of four semipermeable membranes (molecular size exclusion, 90 Da) serially connected. The permeate is conducted to the distribution loop and the amount of water that is not used directly by the dialysis machines is recycled and passes once more through RO by means of the pressure pumping device. This repeated purification results in a water recovery rate of 75%, a rejection rate for bacteria of 99%, for organic substances of 99%, and for inorganic ions of 95–98% (Filmtec 4° Tapwater RO elements, USF Aquapur B.V., Zoetermeer, The Netherlands).
The distribution loop is built of PEX (polyethylene, crosslinked), is free of dead zones, and consists of stainless steel quick-couplings fitted directly to the loop without faucets. The RO water produced enters the distribution loop at a pressure of 2 bar (29 PSI) and flows with a speed of 1.2 m/s. A black sleeve shields the piping system from light.
To prevent bacterial contamination of the water treatment system, filters are replaced every 6 months or sooner when the pressure drop exceeds 0.5 bar. Alarms are triggered when the water hardness exceeds 0.5° German Hardness, the reject rate of the RO drops below 90%, and the conductivity pre RO and post RO exceeds 20 μS/cm. Lastly, samples for bacterial determination are taken monthly from the RO water distribution loop itself, as well as from the outlet of each dialysis machine.
Microbiologic Evaluation of the Dialysate.
Dialysate samples were drawn into sterile tubes. Total plate counts were performed on Reasoners 2A-agar (R2A) (Difco Laboratories, Detroit, MI) after 24 hours of incubation at 37°C and 7 days at 21°C. Isolation of pathogens was performed on McConkey agar (Becton-Dickinson, Germany) after 48 hours of incubation at 37°C. Identification of microorganisms was achieved using Analytical Profile Index identification (Analytical Profile Index system S.A., Montalieu Verceu, France).
Blood samples for LPS determinations were collected at t0 and t180 in heparinized and pyrogen-sterile Endo Tubes (Chromogenix Instrumentation Laboratory Spa, Milano, Italy). After centrifugation (10 min, 190 g), samples were stored at −20°C until determination. LPS concentrations in platelet rich plasma (PRP) were quantified by an end-point chromogenic method based on the activation of clotting factors in LAL (Bio Whittaker, Boehringer Ingelheim Bioproducts, Verviers, France) and subsequent conversion of chromogenic substrate by the activated clotting factors. The chromogenic end-product was measured bichromatically at 405–490 nm. As LPS in plasma is partly bound to lipoprotein-binding protein (LBP), phospholipid transfer protein (PLTP), or various lipoprotein fractions, including LDL and VLDL, 28–30 all plasma samples were diluted 10 times and heated for 15 minutes at 75°C to overcome inhibition/enhancement by these proteins. The highest concentration of the standard curve was 1.2 EU/ml PRP. Aliquots of all plasma samples were spiked with 0.024 EU/ml LPS. Mean ± SD recovery at t0 was 72.7% ± 14.3%, whereas the mean ± SD recovery at t180 was 67.2% ± 15.6%. Sample recoveries (50–100%) were below the value of the recoveries of healthy volunteers (100%). The latter observation may be due to a rise in acute phase proteins in HD patients, associated with an increased LPS binding, which is then not completed inhibited by the dilution/heating procedure as described above. Data at t180 were corrected for changes in hematocrit (Ht): corrected valuet180 = (Htt0/Htt180) × valuet180.
Dialysate samples for LPS determinations were collected at t180 in pyrogen-sterile FALCON 2063 polypropylene tubes (Becton Dickinson, Franklin Lakes, USA). LPS activity in dialysate was quantified by a kinetic chromogenic method based on the LAL assay (Bio Whittaker, Wakersville, USA). Standard series of purified Escherichia coli 055:B5 LPS were made in the range of 0.005 to 50.0 EU/ml. Inhibition and interference testing was performed on each sample by an endotoxin spike. To overcome inhibition/enhancement, all dialysate samples were diluted 10 times. All determinations were performed in duplicate, and recoveries of spikes between 50 and 150% were accepted; limit of determination was 0.05 EU/ml.
Data are expressed as mean (± SD) or median and range when appropriate. Analysis was performed with the Statistical Package for Social Sciences/PC + software system using analysis of variance, and paired and unpaired t-tests to study the differences between groups. Differences were considered statistically significant at p < 0.05.
Bacterial culture results from the dialysate are shown in Table 3. Gram positive organisms were cultured in 8 of 74 cases, whereas 7 of 8 also showed gram negative organisms. When comparing filtered with nonfiltered dialysate, a significant difference was observed (filtered dialysate 0 [0–3] CFU/ml; nonfiltered dialysate 26 [0–310] CFU/ml, p < 0.001). Pseudomonas species were not isolated.
LPS Content of Dialysate
The LPS concentrations in dialysate during HD with filtered and nonfiltered dialysate are shown in Table 3. Marked differences were not found between filtered and nonfiltered dialysate (filtered dialysate 0.051 ± 0.0052 EU/ml, nonfiltered dialysate 0.051 ± 0.0048 EU/ml, p = not significant [ns]).
LPS Content of Plasma
At baseline, all groups showed comparable LPS concentrations (F 60, 0.032 ± 0.005 EU/ml; F 500S, 0.031 ± 0.004 EU/ml; Tricea 150G, 0.032 ± 0.004 EU/ml; and Tricea 150G, 0.034 ± 0.007 EU/ml, p = ns). Both at the beginning and end of the dialysis session, mean LPS concentrations in plasma were significantly below the values in the dialysate (plasma, t0 0.032 ± 0.005 and t180 0.028 ± 0.006 EU/ml, respectively; dialysate, 0.051 ± 0.005 EU/ml, p < 0.001).
During HD, a significant decrease was observed in all four modalities (F 60: t0 mean, 0.032 ± 0.005, t180 mean, 0.026 ± 0.009 EU/ml, p = 0.001; F 500S: t0 mean, 0.031 ± 0.004, t180 mean, 0.027 ± 0.005 EU/ml, p = 0.001; Tricea 150G: t0 mean, 0.032 ± 0.004, t180 mean, 0.025 ± 0.005 EU/ml, p < 0.001; and Tricea 150Gf: t0 mean, 0.034 ± 0.007, t180 mean, 0.025 ± 0.006 EU/ml, p < 0.001). A “borderline” significant difference was observed between PS and CTA dialysis (PS −18.8%, CTA −24.2%, p = 0.06) (Figure 2).
Correlation Between Dialysate Cultures, LPS Content of Dialysate, and Plasma
No correlation was found between the dialysate cultures and the LPS concentrations in the dialysate (r = −0.035, p = ns). Furthermore, marked correlations were not found between dialysate cultures and plasma LPS concentrations, nor between the dialysate and plasma LPS concentrations (r = −0.06, p = ns, respectively, r = −0.02, p = ns).
The data presented in this report clearly show that the water treatment system used in our center produces low to moderately contaminated dialysate without the need for expensive, disposable bacterial filters. 31,32 With respect to the water delivery system, several factors may be responsible for the excellent bacterial quality of the dialysate. 21,24,33–35 To prevent bacterial growth and bio-film formation, dead spaces in the distribution loop are completely avoided, whereas a water flow with a speed of 1.2 m/s is provided. In addition, the distribution loop itself is built of PEX, which is an attractive alternative to stainless steel, having a smooth inert surface resistant to biologic erosion. 21,26 Furthermore, to provide optimal purification of the incoming water, the RO system consists of four serially connected membranes, while in contrast to many other reports, 19,36 the storage tank is placed before the RO system. This arrangement is aimed at preventing bacterial growth due to stagnancy of the purified permeate downstream of the RO system. Finally, the use of dry powder bicarbonate cartridges improved the microbiological quality of the dialysate. 24,26
Plasma LPS concentrations were significantly below the value in the dialysate, both before and after HD. Remarkably, during HD with all four modalities, plasma LPS concentrations decreased significantly. As we did not measure the rate of back transport, these findings suggest that either back transport of LPS did not occur at all or plasma clearance of LPS was greater than back transport and/or clearance of LPS at the dialyzer.
The mechanism that may account for the decrease of LPS during HD has not been elucidated. As the concentration of LPS in the plasma was below the level of the dialysate, diffusive removal of LPS seems highly unlikely. A bactericidal/permeability increasing protein (BPI) dependent mechanism may influence the biologic response to LPS, as BPI increases during HD. 37,38 However, the stimulation of BPI during HD in a recently published study showed comparable data between CTA and PS devices. 38 Finally, reliable data about the adsorptive capacity and/or the convective transport of membranes for LPS are lacking.
Our findings are interesting for several reasons. First, dialyzers with superior permeability for harmful middle and large molecular weight substances, such as leptin, 39,40 β2-microglobulin, 41 uremic toxins involved in extrarenal homocysteine metabolism, 42 and advanced glycation end-products (AGEs), 43 appear to be safe and do not permit the transfer of LAL positive material from the dialysate to the blood, at least under the conditions found in our center. Therefore, our data shed some doubt on the clinical relevance of the induction of a microinflammatory state due to the transfer of LAL positive bacterial fragments during both high flux and superflux bicarbonate HD and low to moderate levels of dialysate contamination. Second, surprisingly, all dialyzers used in this study reduced the LPS concentrations in the plasma. In this respect, it is of note that mild endotoxemia is not confined to dialysis patients but has been described recently in apparently healthy persons 44 or may originate from chronic infections, smoking, gut barrier dysfunction, and chronic heart failure. 45,46 It has been well documented that most chronic HD patients exhibit multiorgan diseases. It is intriguing to speculate that, HD with biocompatible high flux and superflux dialyzers, such as PS and CTA, instead of worsening, may improve the microinflammatory state that is induced by the endotoxemia accompanying these comorbid conditions.
Of note, the higher levels of cultured micro-organisms in the nonfiltered modalities did not result in higher LPS concentrations. However, it has been suggested that not only LAL positive substances but also other bacterially derived fragments, such as exotoxins and outer membrane proteins (peptidoglycans, muramyl-peptides), may permeate the membrane of the dialyzer and activate mononuclear cells. 13 Like LPS, these cytokine-inducing substances may intermittently elicit the secretion of a variety of proinflammatory cytokines, resulting in a chronic microinflammatory state. However, preliminary data on plasma levels of IL-1β, IL-6, TNF-α, and various acute phase proteins (C-reactive protein, prealbumin, and lipoprotein [a]) showed no marked changes during either a single dialysis session or during long-term HD. 47,48
To summarize, the construction of a new water delivery system resulted in the production of moderately contaminated dialysis fluid, without the need for a bacterial filter. Almost all LPS measurements were within the reference range, whereas during HD with both high flux and superflux dialyzers these values were lower in the plasma than in the dialysate, before as well as after HD. Remarkably, as demonstrated in this study, concomitant use of highly permeable dialyzers and moderately contaminated dialysate resulted in a significant decrease of LPS concentrations during HD, irrespective of the material used (PS or CTA), the flux characteristics of the devices (high flux or superflux), or the presence of a bacterial filter.
We wish to thank Professor A. Sturk and J. Duits for their critical reading of the manuscript and the staff and patients of the dialysis department for their indispensable support and enthusiasm. Finally, we thank Stichting Diafoon, Baxter B.V., and Fresenius Medical Care B.V. for financial support.
1. Baz M, Durand C, Ragon A, et al: Using ultrapure water in hemodialysis delays carpal tunnel syndrome. Int J Artif Organs 4: 681–685, 1991.
2. Lonneman G, Koch KM: Pyrogenic reactions, in Jacobs C (ed), Replacement of Renal Function by Dialysis
. Dordrecht The Netherlands: Kluwer, 1996, pp. 726–733.
3. Oettinger CW, Arduino MJ, Oliver JC, Bland LA: The clinical relevance of dialysis sterility. Semin Dial 7: 263–266, 1994.
4. Grooteman MC, Nubé MJ: Hemodialysis-related bioincompatibility: Fundamental aspects and clinical relevance. Neth J Med 52: 169–178, 1998.
5. Dinarello CA, Lonneman G, Maxwell R, Shaldon S: Ultrafiltration to reject human interleukin-1-inducing substances derived from bacterial cultures. J Clin Microbiol 25: 1233–1238, 1987.
6. Schoels M, Jahn B, Hug F, Deppisch R, Rits E, Hansch GM: Stimulation of mononuclear cells by contact with cuprophan membranes: Further increase of β2
- microglobulin synthesis by activated late complement components. Am J Kidney Dis 21: 394–399, 1993.
7. De Broe ME, Nouwen J, Waeleghem JP: On the mechanism and site of production of β2
-microglobulin during hemodialysis. Nephrol Dial Transplant 2: 124–126, 1987.
8. Hakim RM, Levin N: Malnutrition in hemodialysis patients. Am J Kidney Dis 21: 125–137, 1993.
9. Parker TF, Wingard RL, Husni L, Ikizler TA, Parker RA, Hakim RM: Effect of membrane biocompatibility on nutritional parameters in chronic hemodialysis patients. Kidney Int 49: 551–556, 1996.
10. Memoli B: Cytokine production in hemodialysis. Blood Purif 17: 148–158, 1999.
11. Zimmermann J, Herrlinger S, Pruy A, Metzger T, Wanner C: Inflammation enhances cardiovascular risk and mortality in hemodialysis patients. Kidney Int 55: 648–658, 1999.
12. Nubé MJ, Grooteman MPC: Impact of contaminated dialysate on long-term hemodialysis-related complications: Is it really that important? Nephrol Dial Transplant 16: 1986–1981, 2001.
13. Lonneman G: The quality of dialysate: An integrated approach. Kidney Int 58: S112–S119, 2000.
14. European Pharmacopoeia: Hemodialysis Solutions, Concentrates, Water for Dilution
, 3rd ed. Strasbourg: Council of Europe, 1997.
15. Association of the Advancement of Medical Instrumentation:AAMI Standards and Recommended Practices
. Arlington Virginia: AAMI, 1996.
16. Bambauer R, Schauer M, Jung WK, Daum V, Vienken J: Contamination of dialysis water and dialysate: A survey of 30 centers. ASAIO J 40: 1012–1016, 1994.
17. Laurence RA, Lapierre ST: Quality of hemodialysis: A 7-year multicenter study. Am J Kidney Dis 5: 738–750, 1995.
18. Klein E, Pass T, Harding GB, Wright R, Million C: Microbial and endotoxin contamination in water and dialysate in the Central United States. Artif Organs 14: 85–94, 1990.
19. Arvanitodou M, Spaia S, Katsinas C, et al: Microbiological quality of water and dialysate in all hemodialysis centers of Greece. Nephrol Dial Transplant 13: 949–954, 1998.
20. Hakim RM: Endotoxin transfer during dialysis: Less than meets the eye? J Lab Clin Med 124: 742–743, 1994.
21. Lebedo I, Nystrand R: Defining the microbiological quality of dialysis fluid. Artif Organs 23: 37–43, 1999.
22. Brunet P, Berland Y: Water quality and complications of hemodialysis. Nephrol Dial Transplant 15: 578–580, 2000.
23. Van Tellingen A, Grooteman MPC, Bartels PCM, et al: Long-term reduction of plasma homocysteine levels by superflux dialyzers in hemodialysis patients. Kidney Int 59: 342–347, 2001.
24. Ismail N, Becker BN, Hakim RM: Water treatment for dialysis. Am J Kidney Dis 16: 60–72, 1996.
25. Chang PC: Waterbereiding voor hemodialyse in Academisch Ziekenhuis Leiden. Med Technol 3: 8–10, 1996.
26. Protocol: Waterbereiding voor hemodialyse. Kommissie Kwaliteitsbewaking Dialyse Groep Nederland, 1996.
27. Lonneman G, Schindler R: Ultrafiltration using the polysulphon membrane to reduce cytokine-inducing activity of contaminated dialysate. Clin Nephrol 42: 37–43, 1994.
28. Mattsby-Baltzer I, Lindgren K, Edebo L: Endotoxin shedding by enterobacteria: Free and cell-bound endotoxin differ in Limulus activity. Infect Immun 59: 689–695, 1991.
29. Vesy CJ, Kitchens RL, Wolfbauer G, Albers JJ, Munford RS: Lipopolysaccharide-binding protein and phospholipid transfer protein release lipopolysaccharides from gram-negative bacterial membranes. Infect Immun 68: 2410–2417, 2000.
30. Levels JH, Abraham PR, van den Ende A, van Deventer SJ: Distribution and kinetics of lipoprotein-bound endotoxin. Infect Immun 69: 2821–2828, 2001.
31. Schindler R, Lonneman G, Schäffer J, Shaldon S, Koch KM, Krautzig S: The effect of ultrafiltered dialysate on the cellular content of interleukin-1 receptor antagonist in patients on chronic hemodialysis. Nephron 68: 61–68, 1994.
32. Tielemans C, Husson C, Schurmans T, et al: Effects of ultrapure and non-sterile dialysate on the inflammatory response during in vitro hemodialysis. Kidney Int 49: 236–243, 1996.
33. Van der Linde K, Lim BT, Rondeel JMM, Antonissen LPMT, de Jong GMT: Improved bacteriological surveillance of hemodialysis fluids: A comparison between Tryptic soy agar and Reasonor’s 2A media. Nephrol Dial Transplant 14: 2433–2437, 1999.
34. Lonneman G: Assessment of quality of dialysate. Nephrol Dial Transplant 13: 17–20, 1998.
35. Perez-Garcia R, Rodriques-Benitez PO: Why and how to monitor bacterial contamination of dialysate? Nephrol Dialysis Transplant 15: 760–764, 2000.
36. Grassmann A, Uhlenbusch-Körwer I, Bonnie-Schorn E, Vienken J: Composition and management of hemodialysis fluids, in Vienken J (ed), Good Dialysis Practice. Lengerich, Pabst Science, 2000, pp. 181–198.
37. Schindler R, Marra MN, McKelligon BM, et al: Plasma levels of bactericidal/permeability-increasing protein (BPI) and lipopolysaccharide-binding protein (LBP) during hemodialysis. Clin Nephrol 40: 346–351, 1993.
38. Sundaram S, King AJ, Pereira BJ: Lipopolysaccharide-binding protein and bactericidal/permeability-increasing factor during hemodialysis: Clinical determinants and the role of different membranes. J Am Soc Nephrol 8: 463–470, 1997.
39. Wiesholzer M, Harm F, Hauser A, Pribasnig A, Balcke P: Inappropriately high plasma leptin levels in obese haemodialysis patients can be reduced by high flux haemodialysis and hemodiafiltration. Clin Sci 94: 431–435, 1998.
40. Widjaja A, Kielstein JT, Horn R, von zur Mühlen A, Kliem V, Brabant G: Free serum leptin but not bound leptin concentrations are elevated in patients with end-stage renal disease. Nephrol Dial Transplant 15: 846–850, 2000.
41. Floege J, Ehlerding G: β2
-microglobulin-associated amyloidosis. Nephron 72: 9–26, 1996.
42. Van Tellingen A, Grooteman MPC, Bartels PCM, et al: Long-term reduction of plasma homocysteine levels by super-flux dialyzers in hemodialysis patients. Kidney Int 59: 342–347, 2001.
43. Van Tellingen A, Schalkwijk CG, Teerlink T, Grooteman MPC, ter Wee PM, Nubé MJ: AGE-peptides are cleared during HD with high permeable dialyzers (Abstract). J Am Soc Nephrol 10: 308A, 1999.
44. Schouten WEM, Grooteman MPC, van Houte AJ, Schoorl M, van Limbeek J, Nubé MJ: Effects of dialyzer and dialysate on the acute phase reaction in clinical bicarbonate dialysis. Nephrol Dial Transplant 15: 379–384, 2000.
45. Wiedermann CJ, Kiechl S, Dunzendorfer S, et al: Association of endotoxemia with carotid atherosclerosis and cardiovascular disease: Prospective results from the Bruneck Study. J Am Coll Cardiol 34: 1975–1981, 1999.
46. Goto T, Eden S, Nordenstam G, Sundh V, Svanborg-Eden C, Mattsby-Baltzer I: Endotoxin levels in sera of elderly individuals. Clin Diagn Lab Immunol 1: 684–688, 1994.
47. Van Tellingen A, Schoorl M, Bartels PC, et al: No signs of cytokine release during hemodialysis with superflux dialyzers and standard bicarbonate dialysate: Preliminary data (Abstract). Neth J Med 54: A23, 1999.
48. Van Tellingen A, Grooteman MPC, Bartels PC, et al: Monocyte activation and subsequent acute phase response (APR) during superflux hemodialysis: Long-term results (Abstract). Neth J Med 58: A12, 2001.