Version May 2024
INTRODUCTION — Hemodiafiltration (HDF) is a form of kidney replacement therapy that utilizes convective in combination with diffusive clearance, which is used in standard hemodialysis. Compared with standard hemodialysis, HDF removes more middle-molecular-weight solutes. Some, though not all, studies have suggested that HDF is associated with improved clinical outcomes, providing adequate convection volumes are achieved.
However, HDF is more complex than standard hemodialysis and places increased demands on the user and outpatient dialysis center. HDF is not available in the United States. In Europe, Japan, and some other countries where HDF is available, most clinicians use a specific type of HDF termed online HDF [1]. In online HDF, the substitution fluid is produced by the dialysis machine, which enables large convection volumes.
The use of chronic maintenance HDF requires changes in infrastructure in outpatient dialysis units. Decisions to implement such changes are made on a corporate level, although, when HDF is available in a given center, the practicing clinician can decide to perform either HDF or standard hemodialysis in individual patients. Indications for HDF in individual patients have not been identified. The possible beneficial effect of online HDF on survival is consistent in different patient groups (including age, sex, dialysis vintage, absence or presence of residual kidney function, diabetes) [2].
This topic reviews dosing recommendations and clinical outcomes for HDF.
The technical aspects of HDF, including established guidelines for its implementation, are discussed elsewhere. (See "Technical aspects of hemodiafiltration".)
BASIC PRINCIPLES — The basic principles underlying solute clearance differ between hemodialysis and HDF and are discussed elsewhere. (See "Technical aspects of hemodiafiltration", section on 'Principles of hemodiafiltration'.)
OUTCOMES — Some, though not all, studies have suggested that HDF is associated with improved clinical outcomes, providing adequate convection volumes are achieved.
All-cause and cardiovascular mortality — Accumulating data suggest that, if adequate convection volumes are achieved and compared with standard in-center hemodialysis, online HDF may reduce the risk of all-cause and cardiovascular mortality [2-11]. These data are discussed in detail elsewhere. (See "Alternative kidney replacement therapies in end-stage kidney disease", section on 'Mortality'.)
The mechanisms by which online HDF may reduce mortality, particularly cardiovascular mortality, are not known. Suggested mechanisms include improved clearance of middle-molecular-weight molecules, improved hemodynamic stability with fewer periods of cardiac ischemia, and reduced inflammation [2,11-14]:
●Clearance of middle-molecular-weight molecules and phosphate – The removal of middle-molecular-weight solutes, such as beta-2-microglobulin, myoglobin, leptin, and advanced glycation end products (AGEs), is better with online HDF when compared with low-flux hemodialysis [3,15-17], though, for some solutes, such as beta-2-microglobulin, not when compared with high-flux hemodialysis [4,5]. In addition, although improved clearance has been suggested as a mechanism contributing to improved survival of HDF [15-17], in one randomized trial, better clearance of beta-2-microglobulin was not related to outcome [3].
HDF provided better clearance of phosphate in several crossover studies and one large, randomized, controlled trial [3,18-20]. HDF resulted in lower serum phosphate concentrations and the prescription of fewer phosphate binders in some [18,21], but not all [4,5,22], studies. Hyperphosphatemia and its treatment with phosphate binders have been implicated in increased cardiovascular risk among end-stage kidney disease (ESKD) patients. (See "Vascular calcification in chronic kidney disease", section on 'Hyperphosphatemia and hypercalcemia'.)
●Hemodynamic stability – HDF has been associated with better intradialytic hemodynamic stability, which may contribute to improved survival [4,23,24]. As an example, among 146 patients, treatment with convective techniques (hemofiltration and predilution HDF) resulted in a decreased risk of symptomatic hypotensive intradialytic episodes (for online predilution HDF: odds ratio [OR] 0.46, 95% CI 0.33-0.63) [23]. In another trial, high-volume postdilution HDF reduced the incidence of intradialytic hypotensive episodes as compared with both hemodialysis (HD) and low-volume HDF [25].
The mechanism by which HDF may improve hemodynamic stability is not known. Possible mechanisms include a cooling effect on the patient and better preservation of left ventricular function induced by HDF compared with hemodialysis.
Compared with dialysis, HDF is characterized by a lower energy transfer rate that results in a cooling effect that is comparable with hemodialysis with low-temperature dialysate [26-29]. Dialysis with low-temperature dialysate is associated with improved vascular reactivity and is a commonly used method of treating intradialytic hemodynamic instability. (See "Intradialytic hypotension in an otherwise stable patient", section on 'Prevention of recurrent episodes'.)
HDF may decrease or prevent the development of left ventricular hypertrophy, which is associated with standard hemodialysis (and is of itself a major risk factor for cardiovascular morbidity and mortality in ESKD patients) (see "Overview of screening and diagnosis of heart disease in patients on dialysis"). One small study, however, showed no differences in myocardial perfusion and cardiovascular response during HD or HDF [30,31].
Two small, randomized studies (n = 21 and n = 20) found a decrease in left ventricular mass (LVM) index in HDF patients [32,33]. In the CONTRAST trial cited above, there was no difference in LVM in patients treated with online HDF compared with hemodialysis [34]. However, whereas LVM remained stable in patients treated with HDF, it tended to increase in patients treated with hemodialysis.
●Inflammation – HDF may cause less inflammation compared with standard hemodialysis, as indicated by lower levels of inflammatory markers, C-reactive protein (CRP), and interleukin-6 [4,5,18,35]. Chronic inflammation may increase overall mortality, although this has not been conclusively shown [12]. In one study, CRP was inversely related to the convection volume and survival [36].
Erythropoiesis-stimulating agent resistance — Although several observational and crossover trials [18,20,37,38] and one of the large, randomized, controlled trials [5] suggested that HDF reduced erythropoiesis-stimulating agent (ESA) resistance, this could not be confirmed in most large, prospective, randomized trials [4,39,40].
Nutrition — Data on nutritional status are limited and conflicting. Although randomized trials showed no effect of HDF on albumin levels [4,5,35] or body weight [4,5,41], in one study, online HDF induced an increase in normalized protein catabolic rate (nPCR), which is an indicator of protein intake [4].
Quality of life and symptoms — Some studies found better physical wellbeing during treatment with HDF [19,42,43], whereas others did not [44]. In the largest trial (714 patients, followed for two years), there was no significant effect of HDF on quality of life [45]. However, one trial of 381 older adult patients showed a lower incidence of adverse events during treatment with online HDF compared with hemodialysis (OR 0.86, 95% CI 0.79-0.94) [46].
A randomized control trial did not show any effect of HDF on the progression of neuropathy [47].
RISKS — HDF requires the infusion of large volumes of fluid. There is a potential risk of transmitting infection or inducing inflammatory reactions if the infusion fluid is not ultrapure. Additionally, the high rate of ultrafiltration could result in loss of nutritional and other relevant substances and endothelial or blood cell activation during the treatment.
Studies of HDF have not suggested any safety concerns providing HDF is performed according to recommendations, although there are no studies specifically designed to address safety. Inflammatory mechanisms have not been shown to be activated, and more infections do not appear to occur [36,48]. In fact, some studies suggest that inflammatory markers are lower during HDF [35,41]. Some data suggest that activation of platelets and coagulation with HDF is more pronounced than with standard hemodialysis [49-51]. The clinical relevance of these effects is uncertain.
DIALYSIS PRESCRIPTION — The prescribed dose of HDF should define both a minimum urea clearance (which is referred to by the formula "Kt/V") and a minimum effective convection volume. We target a minimum Kt/V of 1.2 to 1.4 per session, which is the same as for standard hemodialysis. Studies that demonstrated improved survival with HDF all achieved this minimum Kt/V in addition to a minimum convection volume [3-9,52-54]. (See "Prescribing and assessing adequate hemodialysis", section on 'The optimal amount of dialysis'.)
We target an effective convection volume of 23 L/session or 70 L/week when HDF is delivered in postdilution mode. The effective convection volume may be adjusted to body size (figure 1). When performed with this convection volume in postdilution mode, HDF may confer a survival benefit compared with conventional hemodialysis (HD) [2,3,5,11,36,55]. In one study using the predilution mode, HDF with a volume of 50 L/session was associated with a survival benefit compared with conventional HD [56]. (See 'All-cause and cardiovascular mortality' above.)
Lower volumes may have insufficient benefit to justify the increased complexity and cost of HDF. When substitution fluid is infused in predilution mode, the convection volume should be doubled in order to obtain a comparable clearance.
An adequate convection volume is achieved by prescribing a sufficiently long treatment time and a high blood flow rate [3-5]. We prescribe a treatment time of four hours, three times weekly, and a minimum blood flow rate of 350 to 400 mL/min. More intense schedules (daily, nightly, longer sessions, etc) have not been extensively studied in randomized trials, although limited data are available [57].
We initially prescribe postdilution HDF rather than predilution, mixed dilution, or mid-dilution [58]. In postdilution HDF, the infusate is delivered into the tubing downstream of the dialyzer (figure 2). The majority of clinical trials have used postdilution HDF, and postdilution HDF is the most effective modality for middle-molecule clearance [59]. (See "Technical aspects of hemodiafiltration", section on 'Modes of infusion of replacement fluid'.)
Pre- or mixed dilution might be preferable in selected cases (eg, with very low blood flow rates or high filter pressures). Pre- and postdilution can be performed with all HDF machines and dialyzers. Special equipment is required to perform mid- or mixed dilution.
ACHIEVED Kt/V — In contrast to standard hemodialysis, the achieved Kt/V among patients undergoing HDF depends on whether replacement fluid is delivered into the tubing upstream of the dialyzer (predilution), downstream of the dialyzer (postdilution), both upstream and downstream (mixed dilution), or into the middle of the dialyzer blood pathway (mid-dilution) (figure 2). (See "Technical aspects of hemodiafiltration", section on 'Principles of hemodiafiltration'.)
The achieved Kt/V is reduced by predilution replacement due to a decrease in the concentration difference of the solute with respect to blood and dialysate. If predilution HDF is used, the dialysis time or blood flow may need to be increased in order to achieve the Kt/V target. With postdilution HDF, Kt/V will be slightly higher than standard hemodialysis when the same blood flow, dialyzer, and time are used due to the addition of convection to the diffusive clearance.
EFFECTIVE CONVECTION VOLUME — The effective convection volume is equal to the substitution volume plus the net ultrafiltration volume (ie, the difference between patient weight before and after dialysis).
The effective convection volume is determined by three parameters, which are blood flow, treatment time, and the filtration fraction. The relationship between these parameters is demonstrated in the table (table 1).
The blood flow and treatment time make up the total processed blood volume, which is the total amount of blood that goes through the dialyzer. Thus, for a patient with blood flow of 350 mL/min who undergoes HDF for four hours (ie, 240 minutes), the total blood processed is 350 mL/min x 240 minutes, or 84 L.
The effective convection volume is calculated differently for postdilution HDF and for pre-, mid-, and mixed-dilution HDF.
For postdilution HDF, the effective convection volume is equal to the volume ultrafiltered during the treatment session.
For pre-, mid-, or mixed-dilution HDF, the effective convection volume is the ultrafiltration volume, as defined above, multiplied by the dilution fraction (DF). The DF is the plasma water concentration of any solute at the dialyzer inlet relative to its concentration leaving the patient. DF is reduced below 1 due to infusion of replacement fluid into the blood upstream of the dialyzer (Qiu). DF is calculated from the plasma water flow leaving the patient (Qpw), divided by plasma water flow at the dialyzer inlet (Qpw + Qiu). The Qpw is estimated using the equation below:
Qpw = Qb (1 - blood hematocrit) x 0.93
Where Qb is the blood flow rate leaving the patient, and 0.93 represents the fraction of plasma that is water since 7 percent of the plasma volume consists of nonwater components such as protein and lipid. A more accurate approximation for the fraction of plasma that is water can be calculated from total plasma protein concentration [60].
To achieve a target effective convection flow rate (Qfeff), the ultrafiltration rate in predilution (Qfpre) must be increased by a factor of 1/DF so that Qfpre = Qfeff/DF. The table shows the various Qfpre required to achieve an equivalent adequate Qfeff (table 1) [61].
In general, if predilution HDF is used, the ultrafiltration rate should be at least 30 to 50 percent of blood flow entering the dialyzer in order to achieve the target of an effective convection rate of at least 20 percent of the undiluted blood flow rate (table 1) [61].
Blood flow and treatment duration — As noted above, an adequate blood flow rate and treatment time are required to achieve the minimum effective convection volume. (See 'Effective convection volume' above.)
It is important to note the achieved blood flow rate in individual patients. In clinical practice, the effective or achieved blood flow rate is on average 5 percent lower than the set blood flow rate [62-64]; the difference may be as high as 50 percent in individual patients, especially in patients with catheters [65].
An adequate vascular access is required to accomplish acceptable blood flow rates. The achieved blood flow rate tends to be lower in central venous catheters as compared with fistulae [64-66]. However, central venous catheters have been successfully used for HDF [67]. The size of the needle used for puncture of the vascular access helps to determine the blood flow rate [62] and varies greatly between dialysis centers and patients [68,69]. We generally use a needle size of 14G or 15G.
We do not use single-needle dialysis systems for HDF, as this method cannot provide adequate convection volumes. In single-needle dialysis, only one needle is inserted into the fistula and utilizes two pumps for the arterial and venous line. During single-needle hemodialysis, effective blood flow rates are approximately half of the rates during double-needle treatment [70], and the blood flow varies, resulting in variations in transmembrane pressure (TMP) and filtration fraction.
If a blood flow of ≥350 mL/min cannot be achieved, the treatment time should be increased (table 1).
Filtration fraction — The filtration fraction is the convection volume divided by the processed blood volume (or convective flow rate / blood flow rate). The filtration fraction is not set as a separate treatment parameter but is adjusted by altering the substitution ratio, the substitution flow rate, or the target substitution volume, depending on the equipment used.
The filtration fraction is most precisely calculated from the plasma water flow rate or volume rather than the blood flow rate or volume. However, clinicians use the blood flow rate since the blood flow rate is readily available at the bedside [61].
When the blood flow rate is used to calculate the filtration fraction, hemoconcentration can vary markedly with variations in hematocrit and plasma protein content. In addition, the actual blood flow may be lower than the set blood flow, resulting in a higher actual filtration fraction as the recorded filtration (or substitution) fraction keeps pace with set blood flow.
At high filtration fractions, more plasma water is extracted from the blood, and convection volume increases. This results in increased hemoconcentration inside the dialyzer. In postdilution HDF, the maximal filtration fraction is limited by an increase in TMP. The increased TMP is caused by hemoconcentration, which may result in clotting, altered membrane performance, and increased filter entrance pressure [71]. These parameters change during the treatment and may be enhanced by changes in net ultrafiltration rate (eg, by individualized ultrafiltration profiles) [72]. Hence, the tolerance for a certain filtration fraction may change during the treatment.
A high TMP causes alarms and interrupts dialysis, which decreases the total dialysis time. The total convection volume is associated with the frequency of TMP alarms during online postdilution HDF [73]. This problem may be mitigated by automated machine configurations that utilize TMP-controlled treatments [73,74]. Another approach is the automatic adaptation of the substitution flow rate to changes in blood viscosity, based on TMP assessment and pressure transmitted by the peristaltic blood pump [66,75] or on assessment of the global hydraulic permeability coefficient of the dialysis system (ultrafiltration flow ÷ TMP) [76].
DIALYZERS — A membrane with a high hydraulic permeability (ie, a high-flux dialyzer) is required for HDF. However, specific information on the performance of different dialyzer membranes in postdilution online HDF is scarce. In the multicenter Dutch Convective Transport Study (CONTRAST), dialyzers were used with membrane surface areas between 1.7 and 2.2 m2, ultrafiltration coefficients between 56 and 85 mL/hour/mmHg/m2, capillary lumen diameters between 185 and 215 microm, and capillary lengths between 225 and 280 mm [3]. Despite these differences in dialyzer characteristics, the convection volume was determined by treatment time and blood flow and not the type of membrane and its surface area.
By contrast, a crossover study that tested four different dialyzers showed that the highest convection volumes and filtration fraction were achieved by the dialyzer membrane with the largest surface area, a high ultrafiltration coefficient, a wide capillary lumen diameter (≥200 micrometer), and a capillary length of 200 mm [77]. Increasing the dialyzer surface area from 1.4 to 1.8 m2 in a crossover study resulted in higher convection volumes [78]. Hence, a dialyzer surface area of at least 1.7 to 1.8 m2 seems advisable. Further research on this subject is warranted.
ANTICOAGULATION — As for routine hemodialysis, anticoagulation consists of a standard dose of unfractionated heparin given as a bolus at the start of the dialysis treatment with an additional mid-treatment dose, or a continuous infusion during treatment, or a standard dose of low-molecular-weight heparin (LMWH). HDF often requires a higher dose of anticoagulant than is commonly used for hemodialysis. We start with the same dose of anticoagulation for HDF as for standard hemodialysis. However, if clotting or high filter entrance pressure occurs, the dose of LMWH is increased by 25 to 50 percent.
In our experience, the dose of the prescribed anticoagulant, either unfractionated heparin or LMWH, is approximately 10 to 20 percent higher in patients treated with online postdilution HDF than in hemodialysis patients [5,79]. Injection of LMWH in the outlet blood line seems most effective [80].
Both platelet activation and coagulation activity are increased during online postdilution HDF as compared with hemodialysis [49-51]. The increase in coagulation results from a combination of greater hemoconcentration and shear stress within the dialyzer and possibly removal of anticoagulant medication [49-51].
DRUG DOSING — Little data are available on drug dosing in online postdilution HDF. Drug molecular weight and protein binding are reliable predictors of dialyzability, which affects dosing [81,82]. Drugs larger than 1 kD may be cleared more extensively with HDF than with standard hemodialysis since the removal of such drugs is dependent on convection rather than diffusion. HDF does not effectively remove molecules >60 kD and protein-bound drugs [83].
Drug dosing tables are available [83]. However, the majority of the data on drug dosing in convective techniques is derived from studies in intensive care patients. In these patients, protein binding and volume of distribution can be quite different from stable individuals [84].
COSTS — Starting up online HDF requires investments in equipment, organization, and training. However, once online HDF has become a routine treatment in a particular center, its costs appear comparable with those of hemodialysis with high-flux dialyzers.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Dialysis".)
SUMMARY AND RECOMMENDATIONS
●Overview – Hemodiafiltration (HDF) is a form of kidney replacement therapy that utilizes convective in combination with diffusive clearance, which is used in standard hemodialysis. HDF allows increased removal of middle-molecular-weight solutes. (See 'Introduction' above.)
●Outcomes of HDF – If adequate convection volumes are achieved and compared with standard in-center hemodialysis, maintenance online HDF may reduce the risk of all-cause and cardiovascular mortality. The mechanisms by which mortality, particularly cardiovascular mortality, is reduced are not known. (See 'All-cause and cardiovascular mortality' above.)
●Potential risks – Increased risks of HDF compared with standard hemodialysis have not been demonstrated. Potential risks include transmission of infection or induction of inflammatory reactions related to the large volume of replacement fluid. The high rate of ultrafiltration may also result in loss of nutritional and other relevant substances and endothelial or blood cell activation during the treatment. (See 'Risks' above.)
●Dose of HDF – The prescribed dose of HDF should define both a minimum urea clearance (which is referred to by the formula "Kt/V") and a minimum effective convection volume.
•The minimum Kt/V is the same as for standard hemodialysis, which is 1.2 to 1.4 per session. (See "Prescribing and assessing adequate hemodialysis", section on 'The optimal amount of dialysis'.)
•For patients selected for treatment with HDF, we suggest targeting an effective convection volume >23 L/treatment rather than lower volumes (Grade 2C). Studies suggest that a minimum of 23 L/session or 70 L/week convection volume is required to confer the survival benefit of HDF. Lower volumes may have insufficient benefit to justify the increased complexity and cost of HDF. The effective convection volume may be adjusted to body size (figure 1). (See 'Effective convection volume' above.)
●Treatment duration – An adequate convection volume is achieved by prescribing a sufficiently long treatment time and a high blood flow rate. We initially prescribe a treatment time of four hours and a minimum blood flow rate of 350 to 400 mL/min. We perform HDF three times weekly. If a blood flow of ≥350 mL/min cannot be achieved, the treatment time should be increased. (See 'Blood flow and treatment duration' above.)
●Dialyzer selection – A membrane with a high hydraulic permeability (ie, a high-flux dialyzer) is required for HDF. (See 'Dialyzers' above.)
آیا می خواهید مدیلیب را به صفحه اصلی خود اضافه کنید؟
