Ultrapure dialysis fluid

INTRODUCTION — Dialysis fluid, produced by combining specially treated water with a concentrated electrolyte solution, is generally not sterile and may contain a range of microorganisms [1-5] and bacterial products [6-10]. Historically, some level of microbial contamination of the dialysis fluid was considered acceptable. This was based upon the false assumption that the dialyzer membrane provided an effective barrier against bacterial contamination of the blood during dialysis. However, subsequent studies demonstrated that bacterial products from the dialysis fluid were able to cross the dialyzer membrane [11-13]. Such bacterial products were also capable of stimulating the immune system, possibly contributing to the heightened inflammatory state among patients receiving hemodialysis [11-17]. Consequently, efforts to create a more stringent standard for the microbial quality of dialysis fluid led to the development of "ultrapure" dialysis fluid [18].

The vast majority of conventional dialysis machines in the world are equipped to filter bacteria and endotoxins and, thereby, generate ultrapure dialysis fluid. However, regulatory agencies generally do not require dialysis units to comply with the standards for production of ultrapure fluid, except for fluid used to provide hemodiafiltration. (See "Technical aspects of hemodiafiltration", section on 'Components and technology of online HDF'.)

This topic will review the difference between ultrapure and standard dialysis fluid, the methods for production and monitoring of ultrapure dialysis fluid, and the clinical benefits of its use.

Other aspects pertaining to dialysis fluid are covered elsewhere:

●A detailed discussion of contaminants and clinical manifestations of toxicity from such contaminants. (See "Contaminants in water used for hemodialysis".)

●A detailed discussion of the water treatment system that produces standard dialysis fluid and monitoring of such a system. (See "Assuring water quality for hemodialysis".)

PRODUCTION OF ULTRAPURE DIALYSIS FLUID — The difference between ultrapure and standard dialysis fluid is only based upon its microbiologic quality expressed in terms of maximum allowable levels for bacteria and endotoxins (table 1). The maximum allowable levels of chemical contaminants do not differ between ultrapure and standard dialysis fluid. The maximum allowable limits for various contaminants are presented elsewhere. (See "Contaminants in water used for hemodialysis", section on 'Microbial contaminants' and "Contaminants in water used for hemodialysis", section on 'Clinically relevant contaminants'.)

Ultrapure dialysis fluid is produced by filtering standard dialysis fluid through one or more bacterial- and endotoxin-retentive ultrafilters immediately prior to the dialysis fluid entering the dialyzer (a process known as point-of-use ultrafiltration) [19-22]. The production of standard dialysis fluid is discussed at length elsewhere. (See "Assuring water quality for hemodialysis".)

Contemporary dialysis machines are designed with the bacterial- and endotoxin-retentive ultrafilters as integral components of the machine's dialysis fluid pathway [23,24]. This design enables the ultrafilters to be disinfected whenever the dialysis machines are disinfected. Ultrafilters can also be retrofitted into some older dialysis machines. Ultrafilters should be used in accordance with the manufacturer's instructions for use [23-26].

Point-of-use ultrafiltration can reduce but not eliminate the level of bacterial endotoxin in the dialysis fluid. Some small bacterial products, such as fragments of bacterial DNA, are not filtered out [6,27,28].

ASSURING THE QUALITY OF ULTRAPURE DIALYSIS FLUID — Dialysis water can contain bacterial products from a variety of water-borne organisms, such as Pseudomonas aeruginosa, Stenotrophomonas maltophilia, Brevundimonas vesicularis, Alcaligenes species, Moraxella, Corynebacterium, fungi, and yeast [1-5]. These organisms can enter from the incoming water supply, be housed within the water storage and distribution system, or be introduced when liquid, rather than powdered, bicarbonate concentrate is used for preparation of dialysis fluid. The microbiologic quality of ultrapure dialysis fluid is assured by maintaining a routine disinfection schedule of fluid pathways. In some cases, monitoring the quality of the ultrapure dialysis fluid is required.

Maintenance of a routine disinfection schedule — The routine disinfection process limits bacterial contamination and the formation of biofilm. The details regarding the schedule and process for routine disinfection are presented elsewhere. (See "Assuring water quality for hemodialysis", section on 'Disinfection'.)

Determining the need for quality assurance testing — The routine quality assurance and monitoring of the ultrapure dialysis fluid produced by a conventional dialysis machine is not required if the dialysis machine and the ultrafilter are operated according to the manufacturer's instructions, specifically with regards to:

●Replacement of the ultrafilter at the recommended interval (eg, the Diaclear filter should be replaced after use in 50 dialysis sessions or 100 heat disinfections)

AND

●The dialysis water used to prepare the standard dialysis fluid, which then enters the ultrafilter, contains acceptable levels of bacteria (<100 colony forming units [CFU]/mL) and endotoxin (<0.25 endotoxin units [EU]/mL)

Quality assurance of the ultrapure dialysis fluid should be performed if there is a deviation from any of the above conditions. Even in the absence of deviations, local regulatory bodies may require routine testing to ensure the quality of the ultrapure dialysis fluid. The process of quality assurance and monitoring is discussed below. (See 'Process for quality assurance testing, if required' below.)

In addition to testing the product (ie, ultrapure dialysis fluid), some manufacturers may require routine testing of the dialysis water used for preparation of the dialysis fluid. The details regarding testing of the dialysis water are presented in detail elsewhere. (See "Assuring water quality for hemodialysis", section on 'Sampling for microbiologic contaminants' and "Assuring water quality for hemodialysis", section on 'Testing for endotoxins'.)

Process for quality assurance testing, if required — The monitoring of the microbiologic quality of the ultrapure dialysis fluid is performed by culture of the dialysis fluid produced and by direct measurement of endotoxin levels in that fluid.

Culture — Samples used for testing should be obtained before a disinfection cycle to capture bacterial growth at the highest possible levels. The samples should be drawn from a dialysis fluid sample port of the dialysis machine according to the machine manufacturer's instructions for use. The media that should be used for culture of ultrapure dialysis fluid are tryptone glucose extract agar (TGEA) or Reasoner's agar number 2 (R2A), both of which provide superior recovery of bacteria compared with other media such as tryptic soy agar (TSA) [18,29-32]. Regardless of which medium is used, samples should be incubated at 17 to 23°C for seven days.  

Compared with sampling methods for standard dialysis fluid (pour plate or spread plate methods) (see "Assuring water quality for hemodialysis", section on 'Sampling for microbiologic contaminants'), those for ultrapure dialysis fluid need to have additional sensitivity. This is because, by design, fewer bacteria are present in ultrapure dialysis fluid compared with standard dialysis fluid. Instead, the sampling method suitable for ultrapure dialysis fluid is the membrane filtration method, which entails filtration of the dialysis fluid through a membrane filter with a pore diameter of ≤0.45 micrometer. The membrane filter (through which dialysis fluid has passed) is then applied directly to the surface of the incubation medium. The volume of fluid filtered should be sufficient to detect bacteria when they are present in that fluid at a concentration <0.1 CFU/mL, the maximum allowable level for ultrapure dialysis fluid. This level of detection can be obtained by filtering 200 to 1000 mL. Historical data from prior testing of the dialysis fluid from the same source can be used as a guide to determine the amount of fluid to be filtered.

In cases where historical data are unavailable, several membrane filters with different sample volumes filtered will need to be prepared for plating.

Measurement of endotoxin levels — Endotoxin levels are measured using the limulus amebocyte lysate (LAL) assay, a test based on the humoral coagulation cascade of the Atlantic horseshoe crab, Limulus polyphemus.  

There are three types of LAL assay: gel clot, turbidimetric, and chromogenic assay. The gel-clot assay, available in a range of sensitivities (0.015 to 0.5 EU/mL), is a semiquantitative assay that can be performed in a dialysis unit without the need for specialized laboratory equipment. While it is adequate for measurement of endotoxin levels in the ultrapure dialysis fluid, the gel-clot assay is the least sensitive of the three assays. The kinetic turbidimetric and kinetic chromogenic assays are both quantitative and capable of detecting much lower concentrations of endotoxin (as low as 0.001 EU/mL and 0.0002 EU/mL respectively). However, they both require specialized laboratory equipment and are generally not suited for use in a dialysis unit.

The LAL cascade is activated by the binding of endotoxin to Factor C of the coagulation cascade. Factor C has been produced as a recombinant protein (rFC) and assays for endotoxin based on its binding to rFC are now available. In these assays, activated rFC cleaves a synthetic fluorogenic substrate causing the test solution to fluoresce. This approach provides a quantitative range of 0.005 to 5 EU/mL in a single step, with improved resolution compared with conventional kinetic LAL assays [33,34].

While these assays are sensitive for detection of endotoxin, they do not detect other proinflammatory bacterial products, such as oligonucleotides, peptidoglycans, or fragments of bacterial DNA [6,27,28,35]. Other methods are available for detection of such substances but their use is not required as part of monitoring of ultrapure dialysis fluid [6,7,28,35].

POTENTIAL CLINICAL BENEFITS OF ULTRAPURE DIALYSIS FLUID — The use of ultrapure dialysis fluid is associated with a decline in the chronic inflammatory response that is common among hemodialysis patients [6,14,36-51] (see "Inflammation in patients with kidney function impairment"); such a decline in inflammatory response may lower cardiovascular risk and provide other clinical benefits [52-55]:

●Unclear impact on mortality – In an observational study of over 130,000 patients receiving hemodialysis, compared with those exposed to dialysis fluid containing endotoxin levels of <0.001 EU/mL (as in an ultrapure dialysate), patients exposed ≥0.1 EU/mL (as in a standard dialysate) had a 28 percent higher relative rate of all-cause mortality [56]. However, in a trial of over 700 patients who were randomly assigned to receive either ultrapure or standard dialysate and either low-flux or high-flux dialysis (total of four groups), there was no meaningful difference in mortality between the two dialysate groups (24.7 versus 24.1 percent in ultrapure and standard dialysate groups, respectively) over a mean of 35 months [57].

●Improvement in some surrogate endpoints – The use of ultrapure dialysis fluid is associated with improvement in the following:

•Rate of loss of residual kidney function (RKF)

•Anemia (decreased need for erythropoiesis-stimulating agents)

•Serum albumin

•Inflammatory markers

•Markers of oxidative stress

•Beta-2 microglobulin

Ultrapure dialysis fluid may slow the rate of decline in RKF. This was examined in one small trial of 30 patients who were randomly assigned to receive hemodialysis using either conventional or ultrapure dialysis fluid [58]. At 24 months, compared with patients assigned to conventional dialysis fluid, those assigned to ultrapure dialysis fluid retained more RKF as defined by a higher measured 24-hour creatinine clearance (mean 4.3 versus 2.5 mL/min) and a higher 24-hour urine volume (mean 1 versus 0.5 L). In addition, patients treated with ultrapure dialysis fluid had lower C-reactive protein (CRP; mean 0.3 versus 0.8 mg/dL) and interleukin (IL)-6 (mean 11 versus 38 pg/mL).

In another study of 475 patients followed for two years, the rate of decline in RKF was comparable between patients who initiated on conventional hemodialysis with ultrapure dialysis fluid and those initiated on peritoneal dialysis [59]. By contrast, prior studies have reported a faster rate of decline in RKF among patients receiving conventional hemodialysis compared with patients receiving peritoneal dialysis [60-64]. (See "Residual kidney function in kidney failure", section on 'Hemodialysis versus peritoneal dialysis'.)

Changes in the remainder of the parameters were summarized in a meta-analysis of 31 studies (16 observational studies, 15 trials) and 1580 patients that compared outcomes among patients who converted to treatment with ultrapure dialysis fluid with those who remained on standard dialysis fluid [49].

In the analysis of anemia parameters limited to trial data (206 patients), ultrapure dialysis fluid led to a small mean increase in hemoglobin (0.13 g/dL, 95% CI 0.0-0.26) and a decline in erythropoietin dose (-1188 units per week, 95% CI -2371 to -4), albeit with significant heterogeneity in the finding among included studies. No meaningful change was observed in serum transferrin saturation or ferritin levels.

In the analysis of albumin limited to trial data (176 patients), compared with patients who remained on standard dialysis fluid, those converted to ultrapure dialysis fluid had a relative increase in serum albumin (mean 0.25 g/dL, 95% CI 0.02-0.48).

Some inflammatory markers improved upon switching from standard to ultrapure dialysis fluid. As examples, CRP and IL-6 levels decreased. In the analysis limited to controlled trials (414 patients), the mean change in CRP was -4.4 mg/L (95% CI -7.5 to -1.3) and that in IL-6 was -11.6 pg/mL (95% CI -17.7 to -5.5). However, other markers such as tumor necrosis factor-alpha and IL-1 receptor antagonist remained without meaningful change.

There was also a decline in markers of oxidative stress such as pentosidine, myeloperoxidase, oxidized low-density lipoprotein, and beta-2 microglobulin.

SUMMARY AND RECOMMENDATIONS

●Introduction to ultrapure dialysis fluid – Efforts to create a dialysis fluid meeting more stringent microbial quality standards led to the development of "ultrapure" dialysis fluid. The vast majority of dialysis units in the world that use conventional dialysis machines are capable of producing ultrapure dialysis fluid, but the production of such fluid is not always required by regulatory agencies. (See 'Introduction' above.)

●Production – Ultrapure dialysis fluid is produced by filtering standard dialysis fluid through one or more bacterial- and endotoxin-retentive ultrafilters immediately before its entry into the dialyzer (a process known as point-of-use filtration). (See 'Production of ultrapure dialysis fluid' above.)

●Assuring quality – Unless mandated by local regulatory bodies, routine quality assurance testing of ultrapure dialysis fluid is not required if the ultrafilters are used according to the manufacturers' instructions for use. The quality assurance of ultrapure dialysis fluid, when required, is performed by culturing the fluid and by directly measuring the endotoxin levels in the fluid (see 'Assuring the quality of ultrapure dialysis fluid' above and 'Determining the need for quality assurance testing' above):

•Culture – Dialysis fluid should be sampled from the appropriate port on the dialysis machine before a disinfection cycle. For the assay to have sufficient sensitivity, the dialysis fluid needs to be filtered through a membrane filter with a pore diameter of ≤0.45 micrometer. The membrane filter is then plated on tryptone glucose extract agar or Reasoner's agar number 2 and incubated at 17 to 23°C for seven days. (See 'Culture' above.)

•Measurement of endotoxin levels – The measurement of endotoxin levels should be performed using the limulus amebocyte lysate (LAL) assay. The gel-clot type of the LAL assay is generally adequate and can be used at the point-of-care. Additional, more sensitive methods are also available. (See 'Measurement of endotoxin levels' above.)

●Potential clinical benefits – The use of ultrapure dialysis fluid may lead to a decline in the chronic inflammatory response that is common among hemodialysis patients; such a decline may lower cardiovascular risk and provide other clinical benefits such as improvement in anemia.