Albumin dialysis: single pass vs. recirculation (MARS).

BackgroundThe aim of extracorporeal albumin dialysis (ECAD) is to reduce endogenous toxins accumulating in liver failure. To date, ECAD is conducted mainly with the Molecular Adsorbents Recirculating System (MARS). However, single-pass albumin dialysis (SPAD) has been proposed as an alternative. The aim of this study was to compare the two devices with a prospective, single-centre, non-inferiority crossover study design with particular focus on reduction of bilirubin levels (primary endpoint) and influence on paraclinical and clinical parameters (secondary endpoints) associated with liver failure.MethodsPatients presenting with liver failure were screened for eligibility and after inclusion were randomly assigned to be started on either conventional MARS or SPAD (with 4 % albumin and a dialysis flow rate of 700 ml/h). Statistical analyses were based on a linear mixed-effects model.ResultsSixty-nine crossover cycles of ECAD in 32 patients were completed. Both systems significantly reduced plasma bilirubin levels to a similar extent (MARS: median −68 μmol/L, interquartile range [IQR] −107.5 to −33.5, p = 0.001; SPAD: −59 μmol/L, −84.5 to +36.5, p = 0.001). However, bile acids (MARS: −39 μmol/L, −105.6 to −8.3, p < 0.001; SPAD: −9 μmol/L, −36.9 to +11.4, p = 0.131), creatinine (MARS: −24 μmol/L, −46.5 to −8.0, p < 0.001; SPAD: −2 μmol/L, −9.0 to +7.0/L, p = 0.314) and urea (MARS: −0.9 mmol/L, −1.93 to −0.10, p = 0.024; SPAD: −0.1 mmol/L, −1.0 to +0.68, p = 0.523) were reduced and albumin-binding capacity was increased (MARS: +10 %, −0.8 to +20.9 %, p < 0.001; SPAD: +7 %, −7.5 to +15.5 %, p = 0.137) only by MARS. Cytokine levels of interleukin (IL)-6 and IL-8 and hepatic encephalopathy were altered by neither MARS nor SPAD.ConclusionsBoth procedures were safe for temporary extracorporeal liver support. While in clinical practice routinely assessed plasma bilirubin levels were reduced by both systems, only MARS affected other paraclinical parameters (i.e., serum bile acids, albumin-binding capacity, and creatinine and urea levels). Caution should be taken with regard to metabolic derangements and electrolyte disturbances, particularly in SPAD using regional citrate anti-coagulation.Trial registrationGerman Clinical Trials Register (www.drks.de) DRKS00000371. Registered 8 April 2010.Electronic supplementary materialThe online version of this article (doi:10.1186/s13054-015-1159-3) contains supplementary material, which is available to authorized users.

See “Effect of Molecular Adsorbents Recirculating System Treatment in Children With Acute Liver Failure Caused by Wilson Diseas” by Rustom et al on page 162. Molecular adsorbents recirculating system (MARS) is a form of dialysis using albumin-enriched dialysate to remove albumin-bound toxins via a specialised membrane. The dialysate is then regenerated through activated charcoal and anion exchange resin columns. Haemofiltration is required simultaneously to control the blood and dialysate circuits. Single-pass albumin dialysis is another form of albumin dialysis, without a perfusion pump. The accumulation of albumin-bound toxins (eg, bilirubin, bile salts) in liver failure has led to the curiosity behind the potential use of MARS in improving the overall clinical condition for successful bridging to transplantation, or even avoiding transplantation, in cases of acute liver failure (ALF) or acute-on-chronic liver failure (AoCLF). In this study, Rustom et al (1) have set out to show that MARS therapy may be useful as a bridge to transplantation in acute Wilson disease (WD) in children. The concept of MARS as a method for removing liver failure–associated albumin-bound toxins, as well as the increased fraction of toxic albumin-bound copper in acute WD, is the basis for their hypothesis. Data were retrospectively analysed (2004–2009) for all of the children who presented with acute WD and had MARS treatment in a single centre. All of the patients (n = 4) had encephalopathy, renal failure, and a Wilson prognostic index >11. Hence, all of the patients were listed for transplantation. Biochemical data were available for 3 patients and showed reduction in bilirubin, ammonia, aspartate aminotransferase, creatinine, and serum copper levels after MARS sessions. Data available for 2 patients showed an improvement in the Fischer index. All 4 patients were reported to have stable or improved encephalopathy after MARS treatment. Platelet count was significantly reduced in all of the patients and 1 patient had an adverse incident secondary to vascular catheter insertion. There was no improvement in hepatic synthetic function. The median time for transplantation was 5.5 days. Rustom et al suggest that the observed improvement in biological and clinical parameters render MARS a promising treatment modality in acute WD in children; however, the group recognises the shortcomings of such a small sample size and the incomplete dataset. They also acknowledge that the treatment effect is difficult to determine in the absence of a control group and the presence of simultaneous haemofiltration. The exact methodology of encephalopathy measurement in the 2 ventilated cases was unclear. Furthermore, although there is a postulated association with a higher Fischer index and improved encephalopathy, this is, as yet, not a known validated tool in liver failure. Most crucially, this article is unable to shed further light on the following clinically relevant question: Do the observed improved biological and clinical parameters correlate to improved survival and/or morbidity outcomes? Small randomised controlled trials (2–4) in adult studies of AoCLF concur with Rustom et al, namely, biological and clinical improvement is demonstrated in the MARS treatment group compared with the standard medical treatment group. Similarly, some small uncontrolled adult studies of ALF (5) and MARS therapy have observed such improvement. With respect to acute WD, case reports (using MARS or single-pass albumin dialysis) (6–8) have documented enhanced biochemical parameters with or without improved encephalopathy while being successfully bridged to transplantation. One common feature among all of the studies was the acceptable safety profile of MARS, with only complications of thrombocytopaenia and vascular catheter insertion being noted. Two recent large multicentre randomised controlled trials have shown that the survival benefit of MARS remains uncertain. Bañares et al (9) have shown no beneficial effect on survival at 28 days in MARS compared with a standard medical treatment group for AoCLF, despite improved biochemistry and encephalopathy in this group. Saliba et al (10) did show a statistically nonsignificant improved 6-month posttransplant and nontransplant survival rate in patients with ALF in the MARS therapy group, but no improvement in biochemical or clinical parameters. In paediatrics, Dhawan et al (11) revealed a median time of 3.5 days for transplantation of acute WD. This brief period to transplantation raises the question of the need for MARS as a bridging technique. Paediatric data for MARS are limited. Auth et al (12) used MARS as a bridging technique to transplantation for 2 paediatric cases of acute WD. They demonstrated that the MARS circuit could extract toxic metabolites; however, this was not consistently reflected by reduced serum levels of certain metabolites (copper, ammonia, bilirubin), suggesting saturation of the albumin-binding capacity. Essential growth factors and precursors to glutathione synthesis were also noted to be extracted by the albumin circuit, which could negatively affect mortality and liver regeneration. MARS does not prevent transplantation, and survival outcome posttransplantation is unclear. There is no robust evidence to justify the financial implications of this intervention in a clinical setting. The present role of MARS remains within the research setting. Alas, the quest for “life on MARS” continues.

In their article in Critical Care, Sponholz et al. [1] compared two devices: the molecular adsorbent recirculating system (MARS) and single pass albumin dialysis (SPAD). As Sponholz et al. claim, SPAD is not an effective method for achieving a significant decrease in the concentration of bile acids, creatinine or urea, because a single SPAD procedure induces a 7.62 % (median) reduction in the level of bile acids and an increase in the concentrations of creatinine and urea by respectively 5.04 % (median) and 4.69 % (median), compared with pre-dialysis levels [1]. In contrast, our as yet unpublished study demonstrates that SPAD can effectively reduce not only the level of bile acids but also the concentration of ammonia. Based on our study, the application of the SPAD method with a dialysate flow rate of 1000 ml/h for a total of around 10 h reduces the level of bile acids by 21.9 % (median) and ammonia by 16.25 % (median). A question arises as to the possible causes of such discrepancies. Since continuous venovenous haemodialysis was performed in both studies, the divergence cannot be attributed to the technique. In our view, this divergence is associated with several factors. One of these factors is the difference in dialysate flow rate. An albumin dialysate flow rate of 700 ml/h is highly insufficient. Taking into consideration the results obtained by Sponholz et al., an increase in the albumin dialysate flow rate to at least 1000 ml/h appears to be a necessary precondition to ensure effective elimination of bile acids. Also, the methodology adopted for the study is questionable. In our opinion, comparing the effectiveness of urea and creatinine elimination at radically different dialysate flow rates (SPAD 700 ml/h vs MARS 2000 ml/h) is unjustified and may lead to erroneous conclusions, which in fact occurred [1]. Differences in efficacy between the MARS and SPAD procedures reported in the article in question may also be related to inappropriately selected haemofilters in the extracorporeal blood purification systems under comparison. Furthermore, Sponholz et al. [1] pointed out metabolic derangements and electrolyte disturbances, particularly in SPAD using regional citrate anticoagulation. In our view, increasing the dialysate flow rate would significantly contribute to preventing the irregularities noted in the study. On the other hand, we believe that the risk of citrate overdose during liver dialysis, but also during continuous venovenous haemodialysis, is very low [2–4]. In our medical centre, we have been successfully performing continuous venovenous haemodialysis procedures with regional citrate anticoagulation in patients with severe liver dysfunction, sometimes for up to several weeks. We have never noted any symptoms of overdose apart from end-stage liver failure.

Extracorporeal liver support procedures based on albumin dialysis require the use of pharmaceutical-grade human serum albumin (HSA). Those preparations contain octanoate, which is added as stabilizer during the production process. For octanoate, a direct involvement in the pathogenesis of liver failure complications as well as an indirect influence by competitive displacement effects at the albumin molecule have been described. During five Single Pass Albumin Dialysis (SPAD) and three Molecular Adsorbent Recirculating System (MARS) treatments the changes of octanoate concentrations in blood and dialysate were investigated. An octanoate increase in patient blood was observed during passage of the filter for both SPAD (585 micromol/L [338-1022 micromol/L]) (median [range]) and MARS (182 micromol/L [71-437 micromol/L]) during the first three hours of treatment. The molar ratio of octanoate/albumin at the blood outflow was significantly higher during SPAD treatments (1.73 [0.86-2.64] vs. 0.54 [0.31-1.1]; P = 0.001) during MARS. Changes of octanoate blood levels during SPAD were significantly higher than during MARS (P < 0.001). The shift of octanoate from the dialysate to the patient was persistent during SPAD (median 67.6 micromol/min), whereas during MARS a decrease over time was observed (from 25.5 to 7.5 micromol/min). During albumin dialysis procedures a transfer of octanoate into patient blood occurs. The time-course and extent are different between both albumin dialysis procedures. Given the positive clinical effects reported mainly for MARS, the clinical impact of albumin dialysis-associated transfer of octanoate during extracorporeal liver support needs to be evaluated further.

Artificial liver systems are used to bridge between transplantation or to allow a patient's liver to recover. They are used in patients with acute liver failure (ALF) and acute-on-chronic liver failure. There are five artificial systems currently in use: molecular adsorbent recirculating system (MARS), single-pass albumin dialysis (SPAD), Prometheus, selective plasma filtration therapy, and hemodiafiltration. The aim is to compare existing data on the efficiency of these devices. A literature search was conducted using online libraries. Inclusion criteria included randomized control trials or comparative human studies published after the year 2000. A systematic review was conducted for the five individual devices with a more detailed comparison of the biochemistry for the SPAD and MARS systems. Eighty-nine patients were involved in the review comparing SPAD and MARS. Results showed that there was an average reduction in bilirubin (-53μmol/L in MARS and -50μmol/L in SPAD), creatinine (-19.5μmol/L in MARS and -7.5μmol/L in SPAD), urea (-0.9mmol/L in MARS and -0.75mmol/L in SPAD), and gamma-glutamyl transferase (-0.215μmol/L·s in MARS and -0.295μmol/L·s in SPAD) in both SPAD and MARS. However, there was no significant difference between the changes in the two systems. This review demonstrated that both MARS and SPAD aid recovery of ALF. There is no difference between the efficiency of MARS and SPAD. Because of the limited data, there is a need for more randomized control trials. Evaluating cost and patient preference would aid in differentiating the systems.

Systematic review of artificial liver support systems: Current clinical challenges

Despite improvement in critical care, liver failure is still associated with high mortality. Therapeutic concepts are aimed at restoring endogenous liver function or to bridge the time to liver transplantation. In addition to standard medical treatment, extracorporeal liver support with albumin dialysis is used for this purpose. The aim of this study was to analyze the efficacy of single pass albumin dialysis (SPAD) in comparison to the molecular adsorbent recirculating system (MARS) in patients treated at our university hospital intensive care unit between July 2004 and August 2008. In this retrospective analysis we studied patients presenting with liver failure who were treated with albumin dialysis. Laboratory parameters, daily health scoring, the number of transfusions, and mortality were recorded. The (paired) t-test, Mann-Whitney U-test, and Wilcoxon test were used for statistical analysis. In all, 163 albumin dialysis treatments, 126 with MARS and 37 with SPAD, in 57 patients were performed. MARS resulted in a significant decrease in bilirubin (-38 +/- 66.5 micromol/L from a baseline of 301 +/- 154.6 micromol/L), gamma-glutamyltransferase (gamma-GT), alanine aminotransferase, creatinine, and urea. SPAD resulted in a significant decrease in bilirubin (-41 +/- 111.2 micromol/L from a baseline of 354 +/- 189.4 micromol/L) and gamma-GT, while lactate levels increased. No differences in the need for blood transfusion, health scoring, or mortality between the two treatment modalities were detected. This retrospective analysis suggests equal efficacy of MARS and SPAD; however, prospective assessment to further define the role of SPAD in the treatment of acute or acute-on-chronic liver failure is needed.

Albumin dialysis with the molecular adsorbent recirculating system (MARS) or single pass albumin dialysis (SPAD) uses human serum albumin (HSA) as an addendum of the dialysate fluid. The purpose of this in vitro study was to evaluate the impact of the dialysate albumin concentration on removal efficacy. Heparinized human plasma (3 L/test) was spiked with creatinine (1000 mg/L), unconjugated bilirubin (100 mg/L), chenodeoxycholic acid (CDCA) (100 mg/L), and diazepam (3 mg/L). The MARS albumin circuit was primed with different amounts of HSA (150, 100, 60, and 40 g). The plasma, albumin, and dialysate flow rates were 200, 200, and 40 mL/min, respectively. Clearances were calculated based on repeated sampling during the experiments, which lasted 480 min. The effective HSA concentrations in the dialysate were 175, 115, 77, and 46 g/L, respectively. They decreased over treatment time to 147, 99, 63, and 41 g/L, respectively, due to surface adsorption. The plasma-HSA concentration remained unchanged over time in all experiments (average 39 g/L). The creatinine clearance was not impacted by dialysate HSA concentration. For the albumin-bound markers a clear correlation between HSA-concentration and clearance was demonstrated with the highest clearances for the 100 and 150 g HSA experiments. The 100 g HSA setup appeared to be the one with best cost-benefit ratio, resulting in clearances (after 1 h of treatment) of 31 mL/min creatinine, 0.3 mL/min unconjugated bilirubin, 11 mL/min CDCA, and 35 mL/min diazepam. Low albumin concentrations, such as in SPAD, result in low clearance rates for albumin-bound substances. The optimal clearances in these experiments were reached with a priming dose of 100 g HSA.

Albumin dialysis is widely accepted as a liver-support technique for patients with liver failure. The Molecular Adsorbent Recirculating System, the widely accepted albumin dialysis technique, has limited use in developing countries because of its technical difficulties and high cost. Therefore, we assessed the efficacy of the more practical modality, the single-pass albumin dialysis (SPAD), in terms of bilirubin reduction, as a marker of albumin-bound toxins removal, as well as the patient outcomes. Twelve acute or acute-on-chronic patients with liver failure who had hyperbilirubinemia (total bilirubin > 20 mg/dL) were treated with SPAD by using 2% human serum albumin dialysate for 6 h. SPAD treatment significantly improved the levels of total bilirubin, conjugated bilirubin, urea, and creatinine (P < 0.001 for all parameters). The reduction ratios of these four parameters were 22.9 +/- 3.8%, 20.9 +/- 5%, 19.0 +/- 4.1%, and 27.7 +/- 3.2%, respectively. No significant difference was observed between serum ammonia before and after treatment. No significant changes in mean arterial pressures were noted during the maneuver, representing cardiovascular tolerability. No treatment-related complications were found. The 15-day in-hospital survival was 16.7%. However, a subgroup of the patients who had moderate severity showed 100% 15-day-survival rate (2 of 2 patients). In conclusion, SPAD is salutarily effective in reducing bilirubin in patients with liver failure. The procedure is safe and simply set up.

Acute liver failure (ALF) and acute on chronic liver failure (AoCLF) carry a high mortality. The rationale for extracorporeal systems is to provide an environment facilitating recovery or a window of opportunity for liver transplantation. Recent technologies have used albumin as a scavenging molecule. Two different albumin dialysis systems have been developed using this principle: MARS (Molecular Adsorbent Recirculation System) and SPAD (Single-Pass Albumin Dialysis). A third system, Prometheus (Fractionated Plasma Separation and Adsorption), differs from the others in that the patient's albumin is separated across a membrane and then is run over adsorptive columns. Although several trials have been published (particularly with MARS), currently there is a lack of controlled studies with homogenous patient populations. Many studies have combined patients with ALF and AoCLF. Others have included patients with different etiologies. Although MARS and Prometheus have shown biochemical improvements in AoCLF and ALF, additional studies are required to show conclusive benefit in short- and long-term survival. The appropriate comparator is standard medical therapy rather than head-to-head comparisons of different forms of albumin dialysis.

The molecular adsorbent recirculating system (M.A.R.S.) is widely used as liver support therapy in patients with hepatic dysfunction. The goal of this study was to measure changes in dialysate albumin and bilirubin concentrations during clinical MARS treatments. Eight patients with acute liver dysfunction and hyperbilirubinemia were enrolled in this study. Five of them received a total of 10 treatments with MARS, in which 600 mL of 20% human albumin was used as dialysate, continuously regenerated by two adsorbent columns in the circuit. Three patients received 4 treatments of a modified MARS, in which the two adsorbent columns were bypassed in the first course for 4 h, and then connected to the circuit in the second course for another 4 h. The total, conjugated and unconjugated bilirubin (TB, CB, UCB) and albumin concentrations in serum and albumin dialysate were dynamically measured, and the adsorbent column inlet pressures were recorded during each session. In one session, dialysate albumin levels were measured during the priming process, at the time points prior to the priming process, immediately after priming, and at the end of the treatment. During MARS therapies, the reduction ratio of serum TB, CB and UCB was 26.6+/-9.0%, 29.5+/-9.6% and 14.8+/-12.3%, respectively. The molar ratio of TB/albumin in serum was approximately 20-fold higher than dialysate at all time points. A significant albumin concentration decrease from baseline in the dialysate was found (mean+/-SD, 34.6+/-16.6%). For the first four hours of modified treatments, in which only albumin dialysis without albumin regeneration by adsorbent columns was performed, the dialysate albumin decrease was substantially smaller (mean, 8.3+/-1.5%). After switching to standard MARS, there was a further decrease in the dialysate albumin concentration of 35.1+/-14.5%. In one session, dialysate albumin concentrations were measured during the priming process, and levels decreased from 196.9 g/L to 144.4 g/L. Adsorber inlet pressure increased from 40+/-10 mmHg at the start of priming to 150+/-50 mmHg at the end of priming, and further increased to 340+/-100 mmHg at the end of treatment. There is a significant reduction in dialysate albumin concentration during MARS therapy. Binding of albumin to the adsorbent columns used for albumin regeneration is largely responsible for this decrease.

Background:Besides standard medical therapy and critical care monitoring, extracorporeal liver support may provide a therapeutic option in patients with liver failure. However, little is known about detoxification capabilities, efficacy, and efficiency among different devices.Methods:Retrospective single-center analysis of patients treated with extracorporeal albumin dialysis. Generalized Estimating Equations with robust variance estimator were used to account for repeated measurements of several cycles and devices per patient.Results:Between 2015 and 2021 n = 341 cycles in n = 96 patients were eligible for evaluation, thereof n = 54 (15.8%) treatments with Molecular Adsorbent Recirculating System, n = 64 (18.7%) with OpenAlbumin, n = 167 (48.8%) Advanced Organ Support treatments, and n = 56 (16.4%) using Single Pass Albumin Dialysis. Albumin dialysis resulted in significant bilirubin reduction without differences between the devices. However, ammonia levels only declined significantly in ADVOS and OPAL. First ECAD cycle was associated with highest percentage reduction in serum bilirubin. With the exception of SPAD all devices were able to remove the water-soluble substances creatinine and urea and stabilized metabolic dysfunction by increasing pH and negative base excess values. Platelets and fibrinogen levels frequently declined during treatment. Periprocedural bleeding and transfusion of red blood cells were common findings in these patients.Conclusions:From this clinical perspective ADVOS and OPAL may provide higher reduction capabilities of liver solutes (i.e. bilirubin and ammonia) in comparison to MARS and SPAD. However, further prospective studies comparing the effectiveness of the devices to support liver impairment (i.e. bile acid clearance or albumin binding capacity) as well as markers of renal recovery are warranted.

There are various Extra Blood Purification Therapies (EBPTs) used in the context of critical care, including but not limited to Acute Kidney Injury (AKI). These therapies aim to remove toxins, inflammatory mediators, and excess fluids from the bloodstream. While some blood purification therapies were initially developed for renal support, they have been explored for use in other medical conditions as well, including liver pathologies and sepsis. Here is a brief explanation of some therapies such as MARS (Molecular Adsorbents Recirculating System), Prometheus, CPFA (Coupled Plasma Filtration Adsorption), PAP (Plasma Adsorption), and SPAD (Single-Pass Albumin Dialysis). Some of these therapies have entered clinical use, while others have faced challenges, such as negative evidence, poor purifying efficacy, or difficulties in practical use. The field of extracorporeal liver support is dynamic, with ongoing developments aimed at improving the effectiveness and practicality of these therapies. Sorbents mark the latest frontiers in blood purification to remove various toxic molecules, with specific emphasis on the modulation of bilirubin and other substances in critically ill patients suffering from liver failure. In the above-mentioned pathologies, substances may be continuously generated within the body, and Mass Balance is the only valuable tool for distinguishing between generation and removal processes. The effectiveness of sorbents in removing bilirubin and bile acids, as demonstrated in both in vitro and in vivo studies, distinguishes them and shows their superiority over traditional liver cleansing methods, such as CPFA, PAP, SPAD, Prometheus, and MARS.

Albumin dialysis by the molecular adsorbents recirculating system (MARS) and by fractionated plasma separation, adsorption, and dialysis (Prometheus[PROM]) represent novel nonbiological liver support systems specifically designed to remove albumin-bound substances. Preliminary evidence suggests a favorable impact of MARS on the course and outcome of liver failure. This study aimed at comparing the detoxification capacity of both devices. For this purpose, we performed a retrospective analysis on data prospectively collected in patients with acute-on-chronic liver failure treated with either the MARS (n = 9) or the PROM (n = 9) device on 2-5 consecutive days. Each treatment was performed for at least 5 h at identical blood and dialysate flows. Blood clearances were calculated during the first treatment session for urea nitrogen, creatinine, total bilirubin, and bile acids from paired arterial and venous line samples after 1, 4, and 6 h of treatment. Reduction ratios for all single-treatment sessions, and the overall treatment phase, were calculated from pretreatment and post-treatment values. For all markers but bile acids, the single-treatment as well as the overall treatment phase reduction ratios obtained with PROM were significantly higher compared with those obtained with MARS. PROM led at all time points to higher clearances for all evaluated solutes. Blood clearances of protein-bound substances declined over time with MARS, but not with PROM. In conclusion, a significant decline in the serum level of water-soluble and protein-bound toxins was achieved with both devices. PROM produces higher blood clearances for most toxins, which results in higher delivered treatment doses compared with a matching treatment with MARS.

Acute-on-chronic liver failure (ACLF) is a frequent cause of death in cirrhosis. Albumin dialysis with the molecular adsorbent recirculating system (MARS) decreases retained substances and improves hemodynamics and hepatic encephalopathy (HE). However, its survival impact is unknown. In all, 189 patients with ACLF were randomized either to MARS (n=95) or to standard therapy (SMT) (n=94). Ten patients (five per group) were excluded due to protocol violations. In addition, 23 patients (MARS: 19; SMT: 4) were excluded from per-protocol (PP) analysis (PP population n=156). Up to 10 6-8-hour MARS sessions were scheduled. The main endpoint was 28-day ITT and PP survival. There were no significant differences at inclusion, although the proportion of patients with Model for Endstage Liver Disease (MELD) score over 20 points and with spontaneous bacterial peritonitis (SBP) as a precipitating event was almost significantly greater in the MARS group. The 28-day survival was similar in the two groups in the ITT and PP populations (60.7% versus 58.9%; 60% versus 59.2% respectively). After adjusting for confounders, a significant beneficial effect of MARS on survival was not observed (odds ratio [OR]: 0.87, 95% confidence interval [CI] 0.44-1.72). MELD score and HE at admission and the increase in serum bilirubin at day 4 were independent predictors of death. At day 4, a greater decrease in serum creatinine (P=0.02) and bilirubin (P=0.001) and a more frequent improvement in HE (from grade II-IV to grade 0-I; 62.5% versus 38.2%; P=0.07) was observed in the MARS group. Severe adverse events were similar. At scheduled doses, a beneficial effect on survival of MARS therapy in patients with ACLF could not be demonstrated. However, MARS has an acceptable safety profile, has significant dialysis effect, and nonsignificantly improves severe HE.