**6. Overview of clinical trials on HDF**

Hemodiaflitraion refers to a combination of convective and diffusive therapies. An example is the use of continuous venovenous hemodiafiltration (CVVHDF) in critically ill patients. It constitutes lower blood flow rate and slower fluid removal which may cause less hemody‐

The setup for HDF requires a double-lumen central venous catheter, a peristaltic pump, and a filter inserted into a venovenous extracorporeal circuit [52] (**Figure 1**). Negative pressure generated by the pump allows blood circulation in the circuit, starting from the central vein and returning to the patient through the filter. Pore fibers of the filter keep water and small particles separated from blood through convective transport. Plasma water flow makes the blood concentration transient, eventually causing intravascular refilling while transferring liquids from the outer to the inner vascular space, thus safeguarding the circulating volume.

After the dispatched intravascular fluid is replaced by extravascular fluid refilling, hypovo‐ lemia can be prevented, thus preventing hemodynamic worsening. According to previous reported studies, fluid removal through HDF has affirmative hemodynamic effects. A study was done on 24 patients with resistant CHF undergoing UF; blood gas analysis (in both systemic and pulmonary arteries), plasma volume changes, and plasma refilling rate were measured after every liter of plasma water removed; UF was performed safely without side effects or hemodynamic instability; and mean right atrial, pulmonary artery and wedge pressures progressively reduced during the procedure [18]. Heart rate and systemic vascular resistance did not increase, and other peripheral biochemical parameters did not worsen during UF. Cardiac output increased at the end of the procedure and, to a greater extent, 24 h later, in relation to the increase of stroke volume. Intravascular volume remained stable throughout the entire duration of the procedure, indicating that a proportional volume of fluid was refilled from the congested parenchyma. Thus, UF is associated with hemodynamic improvement. Fluid refilling from the over-hydrated interstitium is the major compensatory mechanism for intravascular fluid removal, and hypotension does not occur when plasma

When a patient in decompensated heart failure is treated with HDF, it is important to ensure that the amount of removed intravascular water is below the amount of the capillary refilling rate; otherwise, neurohumoral activation can lead the patient to hypotension and abnormal renal function [18]. Thus, the distinctive feature of HDF is its effectiveness in removing extravascular fluids without increasing or decreasing the circulating volume and in turn avoiding neurohormonal activation and its consequences [52]. Respiratory symptoms such as dyspnea and orthopnea can be improved not only by total fluid removal but also by improve‐ ment in lung mechanics, pulmonary gas exchange, interstitial edema, vascular congestion, etc. [53, 54]. The two strategies (mechanical and pharmacological) of fluid withdrawal can have different effects on intravascular volume, hemodynamics, neurohumoral activity, and other regulatory mechanisms, leading to differing clinical outcomes. The tonicity of the fluid removed by the two strategies is different (isotonic with ultrafiltration [UF] and hypotonic with loop diuretics) [52, 55, 56]. With HDF, intravascular volume can be preserved, whereas it can be reduced with loop diuretics. After HDF, a better water-salt balance can be achieved and maintained in the body, whereas rapid worsening of this balance to the baseline abnormal

namic instability in a hypotensive critically ill patient [51].

64 Advances in Hemodiafiltration

refilling rate is adequate to prevent hypovolemia.

An Israeli study aimed to determine the appropriateness and safety of UF therapy in patients with CHF who were admitted to outpatient clinics between April and September 2013 [58]. HDF was used to treat several patients with chronic CHF with NYHA III-IV and diuretic resistance. A total of 38 courses of treatment were administered. The results showed that about 1982 mL of fluid was removed per course, and patients reported having a subjective feeling of improvement in symptoms. Another Singaporean retrospective study was conducted among 44 patients who were admitted to the hospital from October 2011 to July 2013 [59] for decom‐ pensated heart failure and diuretic resistance. The study authors concluded that HDF is an effective and safe treatment strategy that can ameliorate the health of diuretic-resistant Asian patients with decompensated CHF.

Therefore, HDF has proven to be an effective method for fluid removal. It is adjustable in terms of both fluid volumes and rates with no changes in the levels of serum electrolytes. However, despite the effectiveness of HDF in fluid removal, clinical studies have not discovered any clinical benefit of this method over diuretic therapy. Furthermore, compared with diuresis, HDF does not preserve renal function.

In February 2016, a published study compared the effectiveness of HDF and loop diuresis in patients with heart failure (HF) [60]. This study comprised 105 patients with HF who were receiving UF and 108 patients with HF who were receiving adjustable loop diuretics. The aim of the study was to compare the recurrence of HF in the two groups at 90-day follow-up. However, the small sample size provided inconclusive data, which could not be used to identify a significant difference between the two patient groups. The authors of this study reported that the first HF event within 90 days of hospital discharge was seen in 25% of patients who received adjustable UF and in 35% of patients who received adjustable loop diuretics. Furthermore, there were approximately 62 days leading to the first HF event in the 25% patient group and 34 days leading to this even in the 35% patient group. Despite the trend toward better results with HDF, this difference did not reach statistical significance (log rank p 1∕4 0.106). The hazard ratio, which was found to be 0.663 (95% confidence interval [CI] 0.402 to 1.092), suggested a 37% lower risk of an HF event with adjustable UF compared with adjustable loop diuretic therapy, but this result was statistically insignificant. The results should be interpreted with caution, as the study was unilateral and was prematurely terminated by the sponsor.

Another study that included 188 patients investigated UF in CRS [37]. Half of the patients were given stepped pharmacological therapy, and the other half were treated with UF. The bivariate change of the serum creatinine level and body weight from baseline, calculated after 96 h, was the primary end point of the study. The patients were then followed for 60 days. After 96 h, patients undergoing stepped pharmacological therapy had superior renal function compared with that in patients receiving UF. A similar amount of weight loss was noted in both groups of patients. In addition, UF was found to be associated with a higher rate of adverse events.

For patients with CHF, UF is usually reserved for cases of renal failure or unresponsiveness to pharmacological therapy. To study the safety and effectiveness of UF in patients hospitalized with decompensated CHF, investigators conducted the relief for acutely fluidoverloaded patients with decompensated congestive heart failure (RAPID-CHF) trial. This study, the first randomized controlled trial to compare UF with medical therapy, comprised 40 subjects [61]. A single 8-h course of UF with peripherally inserted catheters, along with supportive care and treatment, was compared with supportive care only. Patients who received UF had a greater volume of fluid removal (median 4650 vs. 2838 mL, *p* = 0.001) and improved signs and symptoms of congestion. Treatment with UF was well tolerated by the patients and did not result in any major adverse effects such as hemodynamic and renal complications. Thus, as can be deduced from the study, early treatment with UF for patients with CHF is a feasible treatment option that is well tolerated and can result in significant weight loss and fluid removal.

Another study supporting the findings of the RAPID-CHF trial was the ultrafiltration versus intravenous diuretics for patients hospitalized for acute decompensated heart failure (UN‐ LOAD) trial [62]. The purpose of this randomized controlled trial was to study the safety and effectiveness of early treatment with primary UF in patients with ADHF. The study comprised 200 subjects with two or more signs of hypervolemia. At 48 h after the onset of treatment, weight reduction (5.0 ± 3.1 vs. 3.1 ± 3.5 kg, *p* = 0.001) and net fluid loss (4.6 vs. 3.3 L. *p* = 0.001) were higher in patients who were receiving UF. Furthermore, compared with the patients who were receiving diuretic therapy, the UF group showed lower rates of rehospitalization (0.22 ± 0.54 vs. 0.46 ± 0.76) with fewer days of in-hospital stay (1.4 ± 4.2 days vs. 3.8 ± 8.5 days) and fewer unscheduled medical visits (21% vs. 44%) within 90 days of hospital discharge. Through‐ out the study, changes in serum creatinine levels were similar in both groups. Episodes of hypotension during the first 48 h after hospitalization were similar (4% vs. 3%) in both groups. Thus, the UNLOAD trial demonstrated that, in patients with ADHF, UF is a safe treatment choice that produces higher weight and fluid loss compared with treatment with intravenous diuretics. In addition, UF also reduces a 90-day resource utilization in patients with ADHF. UF may thus prove to be an effective alternative to high doses of diuretic therapy in patients with ADHF, particularly in the presence of renal insufficiency.

Another small study designed for patients with ADHF investigated the effects of UF and standard intravenous diuretic (furosemide) therapy on the GFR and renal plasma flow [63]. The study comprised 19 patients who were randomized to receive UF (*n* = 9) or intravenous diuretics (*n* = 10). After treatment, no significant difference was seen in the GFR, renal plasma flow, and filtration fraction in the two groups. The difference in the net 48-h fluid removal between the two groups was also found to be insignificant. However, urine output during the first 48 h was significantly higher in the furosemide group than in the UF group.

The results of these randomized trials demonstrate that UF is associated with no major safety concerns, and compared with diuretics, it improves total fluid volume removal with decreased hospitalizations for CHF at 90 days in selected patients. However, the clinical trials have not evaluated theoretical concerns such as predisposition to infections and bleeding complications, which can be caused by the systemic heparinization and large-bore intravenous access needed during the UF procedure.
