The evolution of dialyzers has been a key piece in the development of HD. In this evolution, they have highlighted a series of markers of its effectiveness in eliminating TU. Classically, to assess the elimination of low PM TU, the Kt/V or better the Kt corrected by body surface (Ktsc) is used.14 The transition from HD-LF to HD-HF was marked by the Kf titration of the dialyzers, greater than 20 mL/h/mmHg in those of HF.

Convective transport is assessed by the total ultrafiltered volume (VTU) and the Cc coefficient for the different molecules,15 being the Cc of β2 -mG the most used. In this sense, the EUDIAL16 group marked among the requirements to define an effective OL-HDF, to use a dialyzer with a Cc for β2-mG greater than 0.6 and a VTU per session greater than 23 L. The importance of the amount of convective volume administered is due to the fact that it has been directly related to the clearance of medium molecules,17 and as previously commented, with the mortality of HD patients.

To typify dialyzers with MCO membranes, two markers11 have been used, the retention point or “retention onset” (MWRO) and the cut-off point (MWCO): the first is determined by the MW of the molecule whose Cc is of 0.9 or expressed in another way, the MW of the molecules that begin to be retained by 10%, and the second, the MW that corresponds to a Cc of 0.1, close to the CO of the evaluated membrane (Fig. 2). These two markers have been used in the Cc curve for a membrane/dialyzer, because determining the MW with which a molecule begins to be retained and the MW of CO is very difficult. OLS dialyzers would have a Cc for β2-mG close to 1 and 0.9 for myoglobulin.

Fig. 2.

Comparison of the coefficients of screening of the membranes of high efficiency (high-flux) and those of the medium cut-off point (MCO).

(0.24MB).

Cc in dialyzers depend on the conditions under which they have been measured. The type of plasma, blood flow (Qb), ultrafiltration flow (Quf) and the time it took to obtain the sample are some of the factors that determine the result.15 The comparison of dialyzers, using their Cc is only reliable if the measurement conditions are similar. In the technical data sheet of the devices, it must be specified under which conditions the Cc has been measured.

The MCO membrane has been used in Theranova® (Ther) dialyzers, available in Spain for three years. Their design favors internal filtration/retrofiltration, as a form of purification by convective transport,18 which has been called internal hemodiafiltration (HDFi).19 One of the ways to increase HDFi is to decrease the internal diameter of the capillaries, up to 180 µ in diameter in these dialyzers, to increase internal resistances and achieve greater filtration/back-filtration. The internal radius of the capillary is listed to the fourth power in the Hagen-Poiseuille equation that calculates the resistance of the blood to the passage through the capillaries. A long, narrow dialyzer with a capillary diameter of 180 µ will cause a very large pressure drop within the capillary, of around 170 mmHg for 400 mL/min of Qb. Another way to increase internal filtration is to increase the hydraulic resistance in the dialysis fluid (LD) compartment, increasing the density of the capillaries inside the dialyzer, between 55 and 60%,20 creating a pressure drop of 30 mmHg to 500 mL/min of LD flow (Qd). In internal filtration, Qb and Qd directly influence.

Although internal convective transport has been enhanced in the new MCO dialyzers, recent studies suggest that diffusive transport is the main mechanism for elimination of medium molecules in HDx.21 Always remember the interference of these two types of transport on the same membrane.

A requirement in the development of MCO membrane dialyzers was that they did not lose a significant amount of albumin. The high permeability must be compatible with a minimal elimination of this protein, with a Cc less than 0.01. The loss of albumin is dependent on the size of the pores, its interaction with plasma and the transmembrane pressure (PTM). In OL-HDF, the prefilter pressure can reach 700 mmHg, being the main determinant of the loss of albumin. Dialyzers with OLS membranes should not be used at these pressures, so their use is contraindicated in OL-HDF and isolated ultrafiltration. With the HDF technique, OLS membranes can cause significant albumin losses. In any hemodialysis technique, the loss of this should be less than 4 g per session.6

To assess the effectiveness of OLS dialyzers, we must continue to use Kt/V or Ktsc. In the clinic we cannot measure HDFi, although it has been estimated that for an MCO dialyzer, Ther400, with a Qb of 400 mL/min it would be 41.6 mL/min and 53.1 mL/min with the Ther500,6,22 which would mean a four-hour HD session with an internal filtration of 9,984 mL and 12,744 mL, respectively, to which the programmed ultrafiltration would have to be added. Taking into account that the Cc of this dialyser for β2-mG is >0.9, a clearance of β2-mG similar to that of OL-HDF can be calculated. The Qb, therefore, would also be an important determinant of the purifying efficiency with these dialyzers. The preHD concentration of β2-mG could be useful, and should be kept between 20 and 30 mg/L, except in situations of hyperproduction of β2-mG. With OLS dialyzers, Kf no longer has the importance of HF dialyzers; the Kf and Cc relationship is not good.

The use of HF-HD and OL-HDF has necessitated other advances in dialysis. Among them, the machines with precise ultrafiltration (UF) control, the LD with bicarbonate and ultrapure quality stand out. HDx also requires a dialysis monitor with these advances, although, unlike OL-HDF, it does not require a LD infusion system tied to UF monitoring. The HDx can be performed with any modern HD machine with endotoxin filters.

The two dialyzers available in Spain with an MCO membrane, composed of polyarylether sulfone and polyvinylpyrrolidone, free of bisphenol A (BPA), are the Ther400 and the Ther500. Both have an internal diameter of the capillaries 180 µm 35 µm thick wall capillary. The Ther400 has a surface area of 1.7 m2 and a Kf of 48 mL/h/mmHg and the Ther500 2.0 m2 and a Kf of 59 mL/h/mmHg (in vitro with bovine blood, Ht. 32 % Pt 6 g/dL and 37 °C). The Cc is 1 for β2 -mG, 0.9 for myoglobin and 0.008 for albumin (measured with human plasma and Qb 300 mL/min and UF 60 mL/min).

The clearance of TU with these dialyzers has been compared with that of the HF-dialyzers and with OL-HDF (Table 2).23–34 HD-HF has similar or slightly lower results regarding the elimination of molecules of low PM in contrast to HDF-OL and HDx, measured by the proportion of the decrease in the concentration of the molecules before and after HD (RR) or by clearing it. Regarding the medium molecules, post-dilution OL-HDF and HDx are superior to HF-HD and if OL-HDF is performed with high convective transport, OL-HDF is superior to HDx. Regarding large molecules, as the MW increases, HDx surpasses OL-HDF, as for example with λ light chains.

The characteristics of dialysis that make OL-HDF superior to HDx have been evaluated.34 With a Qb of 350 mL/min, OL-HDF exceeds HDx in terms of the percentage reduction (RR) of the molecules, evaluated by means of a « Global Removal Score » (GRS), from 80.6 mL/min of Quf, with a Qb of 400 mL/min it would be from 74.1 mL/min. The main problem with this comparison is in the composition of the GRS, which includes, among other molecules, urea. If the RR of large molecules, such as λ light chains and interleukin-6 (IL-6), is analyzed, HDx would be superior to OL-HDF.24–26 The influence of the different large TUs on morbidity and mortality remains to be weighed.

With OLS dialyzers, albumin loss in LD is usually greater than in HF-HD and OL-HDF, in any case less than 3.5 g/session, between 0.03 and 3.15 g/session.25,29,30,32,35,36 Albumin loss is dependent on the type of membrane and transmembrane pressures used and can exceed HDF-OL 10 g/4 h with some dialyzers.37 In HDx patients, serum albumin is maintained after an initial drop.24,38 At 12 weeks and 12 months with OLS membranes, no significant changes in albuminemia have been detected.31,33 In the work of Bunch et al.39 performed in 638 patients, after one year, found a decrease of 3.5%.

One question that arises with MCO membranes is whether, while they cleanse more TU of medium and high MW than HD-HF, they remove more clotting factors, nutrients, drugs, and other molecules that are beneficial to the body. An in vitro study suggests that the change from HD-HF to Ther500 does not require variations in anticoagulation or in the adjustment of a drug such as vancomycin.40 The changes in insulin and erythropoietin concentrations would be similar with Ther500, HD-HF with Polyflux210H™ and HDF with Fx CorDiax 800™.40

Although the ability of MCO dialyzers to retain endotoxins and other pyrogenic substances has been reported, 41,42 given the significant back-filtration of these dialyzers, it is reasonable to use ultra-pure LD.

There is little clinical evidence in the medium and long term with HDx (Table 3).23,28,31,33,38,43–45 The study that includes the most patients is the Colombian COREXH Registry,39 there they recruit 992 patients dialysed with HDx and 638 complete one year of follow-up (Table 3). They have a mortality of 8.54 per 100 patient-years, low compared to other similar studies with other HD techniques.46,47 The admission rate would be 0.79/patient-year and 6.91 days/patient-year.39 No adverse effects have been described with HDx,25,39,43 nor hypersensitivity reactions to synthetic membranes/polysulfones.48 However, because it is a synthetic membrane, these hypersensitivity reactions may occur.

The perceived quality of life has been assessed in a prospective study33 in which 49 patients were randomized, 24 were dialyzed with Ther400 and 25 with HD-HF with the FX CorDiax 80-60™. Basically and at 12 weeks, the Kidney Disease Quality of Life Short Form-36 (KQDOL-36) test with 36 items was performed. Ther400 patients expressed a better quality of life, mainly in the physical components/domains, as well as less itching and sleep disturbances.

Cho et al.31 did not find significant differences in the preHD concentration of TUs between patients dialysed with Ther400 and those with FX CorDiax 80 after 12 months. The low serum concentrations of β2-mG, 25.6 mg/L, which rise to 28.4 mg/L per year are noteworthy; these concentrations are common in OL-HDF or when there is significant residual renal function (RRF). The work does not provide this last data, which could mask the effect of the Ther400 through its decrease over time.

With HDx, an improvement in pro-inflammatory parameters has been reported.38 One of the factors that explain the loss of RRF in dialysis is inflammation.49 Some UT with high PM would be harmful to the renal tubules and to RRF2,3,27; their greater clearance by the MCO dialyzers27 could better preserve RRF, which will have to be investigated.

The induction of vascular calcification is reduced with plasma dialyzed with MCO and HCO,34 so a possible beneficial role of HDx to avoid vascular calcifications should be studied.

At the close of this review, there are 16 ongoing HDx studies, registered in the ClinicalTrials.50 Six of these are complete and eight are in recruitment. In six HDx was compared with HD-HF, in four with HDF and in one with both techniques. Some focus on specific aspects such as: anticoagulation, preservation of RRF, calcifications and mineral metabolism or symptoms. Among them is the Spanish multicenter, open, prospective, randomized study MoTHER to explore morbidity and mortality in dialysis patients with HDx compared with OL-HDF.

What makes HDx different from other forms of HD? Its ability to remove large molecules, high PM TU (Table 1). The HDx can be defined as a high Cc HD, tuning further, with a MWRO in the range of high PM TUs. Therefore, to conceptualize it and differentiate it from conventional HD-HF, the Cc of myoglobin, ≥0.9, should be used. The HDx is therefore a high-screening HD. Myoglobin is a 17,000 Da molecule that is easy to measure and belongs to the low range of high MW molecules.

There is little clinical evidence to determine what types of patients may benefit the most from HDx. Based on previous experience with other high-level TU debugging techniques, such as OL-HDF, we could propose some aspects to check in future studies. 1. Patients without a significant RRF; 2. Patients with the prospect of staying on hemodialysis for years, for example, not candidates for kidney transplantation; 3. Patients with a sufficient intake of nutrients; 4. Patients with a lot of comorbidity; 5. As an alternative to OL-HDF in cases where high convective transport cannot be guaranteed (elevated Hb, suboptimal Qb).51 The cases with greater comorbidity that could benefit from HDx would be those that have a clear relationship with the retention of high-PM TU: chronic inflammation; resistance to EEE; Restless leg syndrome, secondary immunodeficiency, and cardiovascular disease.52

Regarding the competition with OL-HDF, HDx would be useful in patients in whom it is not possible to achieve an adequate convective volume per session (23 L) or when OL-HDF is suspended for safety reasons.

Good results have been reported in people with certain pathologies 43,53: pruritus,33 post-HD asthenia, anorexia, restless legs syndrome, light chain myeloma,54 rhabdomyolysis, severe inflammation. Some of these indications coincide with our own experience.

The superiority of HDx over HD-HF in eliminating high PM TU, its easy implementation and its safety suggest creating a new section in the classification of HD techniques. The HDx is not a “conventional” HD in the sense of being within established standards. With the existing data, the HDx should be in a new category of HD. Morbidity and mortality clinical studies are needed to demonstrate the non-inferiority of HDx over OL-HDF.50