Extracorporeal Techniques in Kidney Failure

Chronic kidney disease is a condition characterized by water retention as well as complex modification of plasma solutes, with deranged electrolyte and acid-base homeostasis [1, 2]. Over the years, various blood purification techniques have been developed [3] with different characteristics that are described in this article. Treatment effects are the result of the interaction of individual patient characteristics with device characteristics and treatment prescription. Historically, the nephrology community aimed to provide adequate treatment, along with the best possible quality of life and outcomes. Various strategies have been optimized over the years to enhance clearance of a variable spectrum of retained molecules, to ensure hemodynamic tolerance to fluid removal and improve long-term survival.

Treatment duration describes the amount of time per treatment episode and is preferably expressed in minutes (min), but for practical reasons, it can also be expressed in hours. Prescribed duration of treatment is fixed in most patients, but the actual duration may vary substantially from one patient to another, and across different centers. In general, treatment durations are shorter (≈200 min) in the USA than in other regions of the world [4], whereas treatment duration is longer than average (≈300 min) in Japan [5]. If necessary, it may be possible to extend the session duration to obtain adequate dialysis in order to achieve a sufficient correction of metabolic and electrolyte alterations [6, 7]. Treatment duration may even be longer than 300 min, e.g., in case of prolonged dialysis. Most treatments are delivered during daytime. Some patients are treated during nighttime, which are typically referred to as nocturnal treatments. Typically, these are much longer treatments (>360 min), but concurrently allows for patient recumbency and sleep.

Treatment frequency is the number of occurrences of a repeating treatment per unit of time, implicitly indicating intermittency. The vast majority of patients are treated with more than one session. Most patients receive treatments three times per week according to a standard regimen [8, 9]. Alternatively, sessions can be daily [10], six [11], five, or four times per week, every other day, twice a week [12] or once a week [13]. Frequency is dependent on the individual metabolic needs and amount of dialysis per treatment session. On one end of the spectrum, incremental dialysis is a regimen aimed at delivering a low (i.e., once or twice per week) treatment frequency to those patients having sufficient residual kidney function [13, 14]. The duration and/or frequency of dialysis is adjusted incrementally to reflect the loss of residual kidney function, vintage on dialysis, and/or advancing uremic solute burden. Any regimen exceeding thrice weekly sometimes is referred to as frequent or intensive treatment [11]. The combination of frequency and duration allows for a wide array of choice and a varied selection of hemodialysis (HD) prescriptions.

Treatment location describes not only where the treatment is delivered but also provides information about facilities and resources. In-center refers to a unit having all necessary facilities and resources for providing care for most if not all individuals having kidney failure. A satellite unit refers to a unit often located remote to a main nephrology center but having certain facilities and resources to provide HD for most patients. Home treatments are performed at the location of a (semi-) permanent residency [15]. It should be noted that location does not provide information on (in-)dependence of the patient as care for the patient can be self-provided or be assisted, both at home and in-center.

Based on treatment characteristics, we have defined a number of distinct extracorporeal treatment options, which are distinct from each other. For each alternative we provide a definition, a description of the technical set-up, as well as an overview of clinical indications. We however should keep in mind that there is a wide array of clinical applications dependent on the needs of individual patients, and physicians may modify schemes and prescriptions in terms of duration, frequency, efficiency, flows, and operating parameters. Nonetheless, dialysis equipment is increasingly safe and standardized and tends to maintain prescription parameters well within acceptable and predefined values to minimize excessive deviations from validated guidance.

Ultrafiltration (UF) is an extracorporeal treatment applying an exclusively convective process across a semipermeable membrane in the absence of dialysate, to remove excess plasma water without the use of replacement fluids. The device set-up needed for this treatment (Fig. 1) is basic and can be less sophisticated than that used for maintenance HD [16], although it is also possible to use the standard HD equipment for maintenance HD in UF mode [17, 18]. For this technique, filters with high water permeability (high ultrafiltration coefficient; K_UF) are generally preferred. The blood flow is generally between 100 and 250 mL/min, but can be higher. The process is not dependent on the blood flow. The transmembrane pressure will determine the UF rate, typically ranging from 2 to 10 mL/min, although in selected patients higher UF rates might be feasible. The fluid balance is obtained by a control system that regulates the flows in and out of the filter (Q_D^IN and Q_D^OUT) through peristaltic pumps. In modern equipment, management is extremely simple and straightforward and therapy is monitored by an integrated system.

The prescription will depend on the physician’s assessment of the patient’s short-term water balance, through calculating interdialytic weight gains and estimated dry weight assessment. The prescription requires frequent reassessment and adjustment. Net UF (UF^NET) corresponds to the volume of plasma water removed. Monitoring of treatment should include non-invasive blood pressure control and could also include monitoring on a weight scale bed when available, and/or blood volume monitoring as available on the treatment monitor.

The main indication for this technique is volume overload (often due to concomitant heart failure or excessive salt and water retention). Treatments can be delivered in emergency in case of acute pulmonary edema or can be scheduled in case of less emergent situations, sometimes also as additional sessions during the interdialytic gaps in schedule for hemodialysis sessions, especially where attainment of target dry weight has been difficult during normal dialysis sessions. In case of acute indications, pulmonary edema tends to resolve in course of a few min. after start of treatment (sometimes even in the blood priming phase of the circuit). In case of repeated sessions or sessions additional to normal dialysis, the patient’s target dry weight will be established on the basis of clinical criteria (nocturnal dyspnea in case of hyperhydration or cramps in case of excessive fluid subtraction) and, if needed, other methods to assess volume status (e.g., impedance measurements, pulmonary ultrasound, findings on transthoracic echocardiography). Adequate and stable UF is dependent on the rate of UF, and patient characteristics such as transcapillary refill and cardiovascular tolerance. UF rates exceeding 12 mL/kg/h have been associated with high mortality (both US and European data).

The working group discussed several synonyms of UF, including isolated UF, dry UF, and pure UF. We propose not to use dry UF as this is a technical reference, and ultimately both the blood side and dialysate side of the membrane are in contact with fluids. We propose not to use the descriptive term pure UF as pure connotes with water quality whereas dialysate is not a prerequisite for UF. The working group proposes to use isolated UF when no other technique is combined, but otherwise refer to UF.

Hemofiltration (HF) is an extracorporeal treatment applying an exclusively convective process across a semipermeable membrane in the absence of dialysate, to remove plasma water and plasma water constituents and substitute the plasma filtrate by an equal volume of replacement fluids. If needed, the technique is combined with UF to optimize volume status.

For this technique, filters with high water permeability (high K_UF) are used (Fig. 2). The transmembrane pressure is much higher than that during UF, resulting in filtration of large volumes of plasma water and its constituents. The total convective flow (Q_UF^TOT, usually from 40 to 80 mL/min = 2,400–4,800 mL/h) exceeds the volume needed to optimize the patient’s volume status (Q_UF^NET). The difference between total and net removal of plasma water is compensated by substitution with a replacement fluid. The infusion of fluid can occur at the filter outlet (post-filter), at the filter inlet (pre-filter), or a combination of both (mixed, i.e. pre- and post-filter). Originally, bags of sterile fluids were used [19], while today the preparation of replacement fluid is prepared “online” from an ultrapure dialysis solution. The UF rate is usually managed by a control system [20, 21] which also regulates the reinfusion rate to obtain an adequate fluid balance and the desired UF^NET. The blood flow is generally between 250 and 400 mL/min with session duration of at least 4 h for a total convective clearance ranging from about 10 to 20 L/session (post-filter).

HF was quite popular in the 1980s [22], at the time when especially removal of medium and high molecular weight molecules was pursued [23, 24]. However, this treatment approach has mostly been abandoned in favor of HD and hemodiafiltration (HDF). In some patients, however, HF is still practiced with benefit thrice weekly, despite lower achieved weekly Kt/V_urea in comparison with HD or HDF [25]. Given the lower urea clearance and to avoid excessive accumulation of urea, this technique is mostly restricted to individuals having a significant residual kidney function. The use of this exclusively convective technique has been related to better hemodynamic stability [26], perhaps linked to a negative heat balance and an isotonic composition of the ultrafiltrate, resulting in the absence of transcellular fluid shifts.

HD is an extracorporeal treatment applying a mainly diffusive process across a semipermeable membrane in the presence of dialysate and without the use of replacement fluids, to remove plasma constituents. The technique is often combined with UF to optimize volume status. It is a suitable treatment for the vast majority of patients having kidney failure.

During treatment, blood flows through an extracorporeal circuit and a filter (Fig. 3) at a flow rate typically between 250 and 450 mL/min. Blood flow can be (much) lower in specific settings (small body size including children, continuous treatments [see article 3], and at the beginning of treatment). The semipermeable membrane separates the blood from the dialysate. In the early days of dialysis, dialysate was not continuously circulated [27‒29]. Instead, the dialysate bath was intermittently replaced by fresh dialysate. With modern filter designs, dialysate flows countercurrent to blood flow [3]. This set-up optimizes the concentration gradient between blood and dialysate across the length of the filter [30]. In most treatments, dialysate flow is between 400 and 600 mL/min, although both lower and higher dialysate flows are used in specific settings [31]. In individuals tolerating high blood flows (>300 mL/min) and high dialyzer surfaces, the dialysate flow can be increased up to 800 mL/min, obtaining significantly better urea clearance values. In this case, we refer to high efficiency HD, and the treatment time can be reduced to below 4 h/session if permitted by hemodynamic tolerance, volume status, and metabolic control.

Historically, filter characteristics have been used pars pro toto to describe and classify clinical HD treatments. The working group recognizes the conflation of filter characteristics and the use of this terminology as a classifier of HD treatments [32]. Historically, this conflation can be traced back to the introduction of synthetic polymer-based membranes with membrane characteristics that differed from the original cellulose (-based) membranes [33]. Further innovations, resulting in different dialysis performance characteristics, contribute to confusion of terminology.

The K_UF of a filter (DK_UF) describes the permeability to water in response to a hydrostatic pressure gradient and per time unit. Factors that determine the K_UF include pore density (the number of pores per surface area) and the average pore size. The DK_UF is the product of K_UF and membrane area [33, 34]. As the Stokes-Einstein radius of water is small, both a higher number of relatively smaller pores and a lower number of relatively larger pores can result in a similar K_UF. Some synthetic polymers have a large average pore size without a low pore density [33]. These filters have higher K_UF values, leading to a larger transfer from blood to dialysate of plasma water per unit of hydrostatic pressure gradient.

The increase of average pore size has ramifications beyond water permeability. Membranes with larger average pore sizes are permeable to molecules with larger Stokes-Einstein radiuses [34]. Such membranes thus have a higher molecular weight cut-off. The correlation between higher K_UF and permeability to larger molecules contributed to the conflation of concepts. Over time, different regulatory bodies, handbooks, and study-steering committees opted for various definitions of low- versus high-flux [9, 35]. Some definitions were based solely on K_UF, others solely on permeability (or even clearance) of molecules, and some definitions combined both. We propose to restrict the use of the K_UF to describe the water permeability of membranes and develop separate standardized measures of clearance of molecules of different sizes.

Flux can be bidirectional, i.e., both plasma water from the blood compartment into the dialysate compartment as well as dialysate into the blood compartment [36, 37]. The hydrostatic gradient can be such that the water flux reverses over the length of the dialyzer. Determinants of this include the K_UF and the transmembrane pressure, which itself is determined by the membrane geometry (inner diameter and length of the hollow fibers), blood and dialysate flows, and viscosity of the patients’ blood. The resultant convective transport within the filter is balanced and therefore not dependent on a separate substitution fluid or additional pumps, which discriminates it from HDF. The internal convective transport adds to the diffusive clearance during HD.

Given these considerations, the working group discussed whether the use of “flux,” which originally described a physical characteristic of individual filters, should be abandoned as classifier terminology to describe different types of dialysis treatments. On the one hand, it was recognized that apart from K_UF, additional filter characteristics including filter geometry, fluid dynamics, and patient characteristics (e.g., blood rheology) all influence solute and plasma water transport across the filter during HD. On the other hand, the filter characteristics are used for regulatory and marketing purposes, and in some countries for differential reimbursement of dialysis treatments. For these reasons, disentanglement proved difficult.

We propose, despite conflated terminology, to retain three major types of HD: low-flux and high-flux HD, and expanded HD (HDx). These are not based upon categories of K_UF and internal convective transport as these are a continuum. However, from a clinical perspective, these classifiers represent different HD treatment strategies (Table 1).

Low-flux HD describes a treatment using a filter with low DK_UF. The filter has a low molecular weight retention onset and a low molecular weight cut-off. Plasma water constituents are removed almost exclusively using diffusion, and internal convective transport is very low to non-existent. This treatment is considered most suitable for patients without high-risk characteristics. In case of reduced water quality, this treatment is preferred over the other types of dialysis due to negligible convective transport.

High-flux HD describes a treatment using a filter with a high DK_UF. The filter has a variable molecular weight retention onset and a molecular weight cut-off exceeding that of low-flux HD. Dependent on patient characteristics and fluid dynamics, a small amount of convective transport is combined with diffusive transport. This treatment is suitable for most patients, but requires a more rigorous standard of water quality.

HDx describes a treatment using a filter with a high molecular weight retention onset and a molecular weight cut-off similar to or exceeding high-flux HD. Due to changes in membrane geometry with a smaller inner diameter to promote internal filtration and backfiltration, the volume of convective transport exceeds that of high-flux HD. This is combined with diffusive transport, without the use of replacement fluids. This allows augmented clearances of molecules in a very broad spectrum of molecular weight from small water-soluble molecules to middle molecules with a molecular weight/size cutoff just below albumin [38].

Recently, new classes of medium cut-off and high cut-off filters have been developed. The high cut-off filter was developed to be used for specific pathologies (myeloma, rhabdomyolysis) given the significant loss of albumin. The design of the medium cut-off filter makes these filters suitable to deliver HDx [39]. The larger pores permit passage of medium-middle and large-middle weight molecules, and the change in geometry with a smaller inner diameter of the filter fibers promotes internal filtration and backfiltration (Fig. 4). This allows augmented clearances of molecules in a very broad spectrum [38, 40], without significant losses of albumin [41]. Several studies suggest an improved erythropoietin responsiveness [42, 43], improved muscle strength, improved nutritional status in parallel to a reduction of chronic inflammation [44].

HDF is an extracorporeal treatment applying the combination of convective and diffusive processes across a semipermeable membrane in the presence of dialysate to remove large volumes of plasma water and blood constituents and substitute the plasma water by an equal amount of replacement fluids. If needed, the technique is combined with UF to optimize volume status. It is a suitable treatment for the vast majority of patients having kidney failure.

HDF, a combination of HF and HD (Fig. 5), is performed using filters with a high K_UF (high-flux filters, see above) with the aim of controlling solute levels in a broad spectrum of molecular weights as well as the fluid balance with possibly improved hemodynamic tolerance [45, 46]. The system combines diffusion (presence of dialysis solution countercurrent to the blood flow) with convection (the UF rate produced is significantly in excess of the amount that should be subtracted from the patient and thus requires substitution volume). Blood flow is generally between 300 and 500 mL/min, while dialysate flow is between 400 and 800 mL/min.

The UF control system, consisting of volumetric chambers, regulates the dialysis fluid pumps (D_I and D_o in Fig. 5) controlling the flow of dialysate, while a set of pumps and scales manages the the ultrafiltration flow (Q_UF) and replacement flow (Q_R). Reinfusion can take place at the filter inlet (pre-filter), filter outlet (post-filter), via mixed pre- and post-filter, and mid-filter using a dedicated filter (see below) [47, 48]. The convective exchange (total UF) generally ranges from approximately 9 L up to more than 23 L in high-volume post-filter HDF and up to more than 40 L in pre-filter HDF.

There are a number of requirements to perform HDF. The dialysis monitor should be equipped with adequate software and hardware (including at least two pumps). Two different systems can be used for UF: (1) a system of equalization of the dialysis solution flow in and out of the dialyzer (in this case a pump for the production of Q_UF^NET that drains the liquid from the line leaving the dialyzer, in front of the flow equalizer is required) and (2) a differential flow meter that directly manages the difference between the dialysis solution flow at the outlet (Q_D^OUT) and the one of the solution at the inlet (Q_D^IN) by adjusting the speed of the peristaltic pumps.

Second, the production of the substitution fluid should be rigorous and standardized. This takes place through several filtration steps, until an ultrapure solution is reached, suitable for direct infusion into the blood circuit (online HDF). The concept of sterility is distinct from that of an ultrapure solution [47, 49]. Sterility in the online production of liquids can only be certified if there is a continuous analysis of the liquid produced. For the production of online reinfusion fluids by HD equipment, the concept of redundancy is applied. The water produced by the normal treatment systems of dialysis centers today has a high degree of both chemical and microbiological purity. Then, to this are added, at the level of equipment, sterile concentrates, and the resulting solution is filtered in successive passages through membranes with high bacterial retention capacity. Therefore, it is calculated that the probability of the presence of a single bacterium in 1 L of online substitution fluids is very low and these fluids are referred to as sterile and non-pyrogenic. In these circumstances, safety is considered adequate and there have been no reports of adverse reactions despite the high number of treatments performed worldwide. Guidelines review this concept and are based on the ISO13959, European Pharmacopoeia, the AAMI Standards and Recommended Practices, European Best Practice Guidelines for hemodialysis and literature reviews [47, 50].

Proposed in 1970s [51], HDF was considered experimental until the 1990s when it began to be used extensively in many countries except for the USA. Over the years, the practical application of HDF has evolved and different device setups have been used (Fig. 6). In the pioneering phase, HDF sessions were delivered using 9 L of sterile solution (three three-liter pre-packaged bags) that were hung or deposited on the reinfusion scale above the machine (HDF or standard HDF). This increased cost significantly, and the complexity of the system made this technique unattractive for many centers. To overcome this, soft HDF was introduced. Also known as “biofiltration,” only 3 L of substitution volume were used, subsequently increased to 6 L. As the substitution fluid used was acetate-free, this modality has been referred to as “acetate-free biofiltration.” In some Italian centers, an early version of high-volume HDF was practiced during some time, exchanging 15–18 L per session. Under these conditions, the pre-dialysis levels of β2 microglobulin tended to be much lower, demonstrating efficiency of the increased convection volumes in terms of β2 microglobulin clearance. The development of online production of substitution fluid from circulating dialysate solution proved a turning point. Originally, this was referred to as online HDF [52]. Nowadays, this is simply named HDF and has become standard-of-care, recommended by the National Institute of Clinical Excellence (NICE). In Table 2, we summarize different types of HDF treatment according to convective volume, retention onset, and albumin loss.

Several variants have been developed. Each of these has advantages and disadvantages with variable technical and economical limitations. They are reviewed not only to provide historical insight but also to illustrate the potential for future developments. One variant (Fig. 6) was the creation of an HDF treatment with separate chambers, originally defined as paired filtration dialysis [52]. The method consisted of a circuit in which two filters were placed in series. In the first filter, convection was achieved by means of pure UF, while diffusive clearance was achieved in the other filter. The reinfusion of sterile liquid from bags occurred in between the two devices. This variant theoretically avoided backfiltration, i.e., the movement of liquid from the dialysate compartment into the blood compartment once the pressure gradient along the hollow fibers was reversed. Subsequently, the dialysis community moved on to a variant of this technique was introduced called HFR (online HDF with regeneration and reinfusion of the ultrafiltrate). The ultrafiltrate produced by the UF filter was processed through a sorbent cartridge and was reinfused once purified as an endogenous reinfusion liquid. The advantage of the technique was that it avoided losses of important plasmatic components such as vitamins and amino acids. Another variant was represented by mid-dilution HDF (Mid Dil HDF) characterized by a device with two internal bundles of fibers in which the blood initially runs against current and subsequently co-current, after reinfusion has taken place at one end of the filter (Fig. 6). The advantage of this technique should be an increased shear rate of the blood at the interface with the membrane and less packing due to high UF rates.

In the era of the 1980s and 1990s, double-high-volume HDF (Double HV HDF) was tested in Los Angeles (Fig. 6). The goal was to significantly reduce dialysis time, thanks to the increase of the clearances. The technique used blood flows between 400 and 500 mL/min and dialysate flows of around 1,000 mL/min and required two filters aligned in series. While in the first filter, filtration took place, in the second one, reinfusion to the blood compartment occurred by backfiltration. Convective transport is regulated by a valve placed in the connection between the first and second filter in the dialysis liquid circuit. The use of two filters is sporadic today and is used mainly for research purposes.

A further variant of HDF, proposed in Japan was the push-pull HDF (Fig. 6). This circuit involves the use of two pumps. The first, on the inlet line determines the stroke volume, while the second on the outlet line is firm and occluding. At this stage, the pressure in the blood circuit is positive and UF occurs. In the second phase of the cycle, an equivalent stroke volume is produced by the second pump while the first is stationary and occluding thereby generating a negative pressure in the filter which causes internal reinfusion by backfiltration.

In clinical practice today, the most widely used variant is HDF (Fig. 7), and to a far lesser extent HFR. The European Dialysis working group (EUDIAL), established in 2010 by the European Renal Association revised the definition of HDF as blood clearance treatment that combines diffusive and convective transport using a high-flux filter with a DK_UF exceeding 20 mL/mm Hg/h/m², a sieving coefficient for β₂-microglobulin greater than 0.6 and a percentage of effective convective transport greater than 20% of the total processed blood. Convection volume was defined as the total UF volume obtained over the entire HDF session, the sum of the replacement volume and the intradialytic weight loss achieved. In post-filter HDF, the effective convection volume will be equal to the total ultrafiltered volume but, in pre-, mid-, or mixed replacement, the UF volume must be adjusted for a dilution factor that is calculated as the total plasma water volume processed divided by the plasma water plus upstream-infused fluid. A comparative review of the global removal score of uremic toxins of various techniques described above demonstrates the broadest spectrum of clearance is achieved with post-filter HDF and expanded HD (HDx) [38]. These data are summarized in Table 3.

Plasmapheresis is an extracorporeal treatment to remove plasma and to substitute removed plasma by an equal volume of replacement solution containing plasma proteins. The separation of plasma from the blood can be done using density centrifugation or by means of a semipermeable membrane (plasmafiltration) applying an exclusively convective transport across a semipermeable membrane to remove plasma.

Plasmafiltration is performed using a plasma filter, typically with a membrane surface area ranging from 0.3 to 0.8 m² (Fig. 8). The molecular weight cut-off is much higher than that in a conventional dialysis filter, allowing passage of high molecular weight solutes including plasma proteins. Dedicated equipment is required. The blood flow generally varies between 100 and 250 mL/min with the flow of filtered plasma varies from 10 to 20 mL/min for a total PE ranging from 1,500 to more than 4,000 mL. Sessions last from 2 to 4 h. In current equipment, the speed of plasmafiltration is managed by a balancing system which also regulates the speed of reinfusion to obtain an adequate plasma balance and prevent generation of negative pressures within the filtered plasma compartment.

The technique is applied in acute and chronic care settings and the number of sessions delivered would vary according to the disease indication and its severity. Plasmapheresis is indicated for a whole series of antibody-mediated syndromes, in transplantation and other pathologies that require removal of solutes bound to proteins or specific protein fractions.

There are a number of variants of plasmafiltration in which the filtered plasma is not solely discarded, but subject to additional manipulation. These include plasma filtration-adsorption, double-plasma filtration and molecular adsorption system, and cascade plasmapheresis (Fig. 9). The latter represents a variant of PE in which the plasma filtrate is subjected to further filtration by defined cut-off membranes, in order to separate IgG, IgM, HDL, and other plasma components. The purified plasma from those components is reinfused into the original circuit (generally no reinfusion of albumin is required, unless the amount of subtracted plasma is not significant).

Recently the use of adsorption combined with diffusion and convection has been proposed in patients with symptoms due to accumulation of large or protein-bound solutes. The sorbent cartridge is placed in series before the hemodialysis and the session remains unchanged compared to the classic hemodialysis prescription. Excellent results have been reported in two large trials [53‒55] and more research is undergoing.

Extracorporeal therapies for blood purification have become standard-of-care in the treatment of chronic diseases. Different device set-ups allow therapy to be tailored to individual needs. Nursing and technician care is of paramount importance to successfully apply these techniques. A thorough understanding by the whole caring team is also necessary to ensure adequate patient care. This requires the correct use of the equipment. In fact, multiple techniques can be performed with a single device. Some techniques however may be proprietary and specifically executable with only one type of device. It is essential that the physician has a thorough knowledge of the techniques available to be able to perform personalization of the treatment and a prescription appropriate to the needs of each patient.

We wish to thank all experts who attended the Extracorporeal Blood Purification meeting in November 2022 for their invaluable input and suggestions. In addition to the authors of this paper, the Nomenclature Standardization Faculty included:

Ghada Ankawi, King Abdulaziz University, Arabia Saudity; Vincenzo Cantaluppi, University of Piemonte Orientale, Italy; Rajasekara Chakravarthi Madarasu, Renown Clinical Services, India; William Clark, Purdue University, School of Chemical Engineering, West Lafayette, USA; Silvia De Rosa, University of Trento, Centre for Medical Sciences, Trento, Italy; Kristin Dolan, Mercy Children's Hospital Kansas City, USA; Lui G Forni, University of Surrey, Guilford, UK; Faeq Husain-Syed, University Hospital Giessen and Marburg, Justus-Liebig-University Giessen, Germany; Kianoush Kashani, Mayo Clinic, USA; John A. Kellum, University of Pittsburgh, USA; Anna Lorenzin, International Renal Research Institute of Vicenza, Vicenza, Italy; Francisco Maduell, Hospital Clínic, Barcelona, Spain; Ravindra Mehta, University of California San Diego, USA; Marlies Ostermann, King's College London, UK; John R. Prowle, Queen Mary University of London, UK; Thiago Reis, University of Brasilia, Brazil; Zaccaria Ricci, University of Florence, Florence, Italy; Thomas Rimmelé, Edouard Herriot Hospital, University of Lyon, France; Mitchell Rosner, University of Virginia, Charlottesville, USA; Danielle Soranno, University of Colorado, Aurora, USA; Gianluca Villa, University of Florence, Florence, Italy; Alexander Zarbock, University Hospital Münster, Germany.

Bjorn Meijers received speaker fees from Baxter and Nipro and received restricted grants from Fresenius and Nipro. Vega Almudena reports lecture fees from Baxter and B.Braun. Laurent Juilard Kirsch reports speaker fees and grants from Baxter and FMCHideki Kawanishi: none. Alexander H. Kirsch reports speaker fees and grants from Baxter and FMC. Francisco Maduell has received consultancy fees and lecture fees from Baxter, Fresenius Medical Care, Medtronic, Nipro, Toray, and Vifor. Ziad A. Massy declares having support from Amgen, Baxter, Fresenius Medical Care, GlaxoSmithKline, Merck Sharp and Dohme-Chibret, Genzyme/Sanofi, Lilly, Otsuka, Boehringer and Government support for CKD REIN project and experimental projects., and to charities from AstraZeneca, GSK, and Boehringer. Sandip Mitra is supported by NIHR D4D MedTech and In Vitro Diagnostics Co-operatives (MICs). He holds competitively awarded research grants from Fresenius, Invizius, and Baxter. Raymond Vanholder has received consultancy fees and lecture fees from AstraZeneca, Glaxo Smith Kline, Fresenius Kabi, Novartis, Kibow, Baxter, Nipro, Fresenius Medical Care, and NextKidney. Claudio Ronco had in the last 3 years the role of member of advisory board, speaker bureau, or consultation for ASAHI, Aferetica, bioMerieux, CytoSorbents, Jafron, Baxter, GE, Nipro, B.Braun, AstraZeneca. Mario Cozzolino reports grants from FMC and Baxter.

No funding has been received for this article.

Bjorn Meijers, Vega Almudena, Laurent Juilard, Hideki Kawanishi, Alexander H. Kirsch, Francisco Maduell, Ziad A. Massy, Sandip Mitra, Raymond Vanholder, Claudio Ronco, and Mario Cozzolino contributed in writing the manuscript.

A list of all the other faculty members and their affiliations can be found at the end of the paper.