Removal of Metabolites, Cytokines and Hepatic Growth Factors by Extracorporeal Liver Support in Children

Auth, Marcus KH.*; Kim, Hyun Soo*; Beste, Mechthild†; Bonzel, Klaus E.‡; Baumann, Ulrich; Ballauff, Antje*; Wallot, Michael*; Borchers, Tanja*; Vester, Udo‡; Grasemann, Corinna*; Hauffa, Berthold§; Hoyer, Peter F.‡; Gerken, Guido†; Voit, Thomas*

Journal of Pediatric Gastroenterology & Nutrition:
Original Articles: Hepatology and Nutrition

Background: Molecular Adsorbents Recirculating System (MARS)-mini has recently been approved and applied in children with hepatic failure. However, its indication, efficacy and capability to induce liver regeneration remain unclear. The aim of our pilot study in children was to analyse the impact of MARS on markers of detoxification and regeneration.

Methods: In children with fulminant Wilson's disease and bridged with MARSmini for liver transplantation, we analyzed toxic metabolites (bile acids, bilirubin, lactate, ammonia, tryptophan and copper), regulators of the inflammatory cascade [nitrate, tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), methionine, cystine and hyaluronic acid] and hepatic growth factors [hepatocyte growth factor (HGF), epidermal growth factor (EGF), transforming growth factor-β1 (TGF-β1), cortisol, corticosteroid-binding globulin (CBG), insulin-like growth factor-1 (IGF-1), angiogenin, vascular endothelial growth factor (VEGF), IL-6 and TNF-α] from blood, albumin circuit and haemodialysate from four applications.

Results: In all four applications, transfer of toxic metabolites (6/6) and inflammatory mediators (6/6), but also of hepatic growth factors (9/10), into the albumin circuit of MARS was consistently detected. Corresponding blood levels were decreased for 3/6 metabolites, 3/6 inflammatory mediators and 1/10 growth factors and increased for 1/10 growth factors. Bridging for liver transplantation was successful with MARS.

Conclusions: In our prospective study, substantial extraction of albumin-bound and water-soluble candidate substances was detected with variable effect on respective blood levels. Notably, essential factors inducing liver regeneration were simultaneously removed. These data provide a basis for evaluation of liver restoration and efficacy of liver support in children with liver failure to devise a collaborative, multicentre trial.

Author Information

Department of Paediatrics, Division of *General Paediatrics; †Department of Internal Medicine, Division of Gastroenterology; Department of Paediatrics, Divisions of ‡Paediatric Nephrology and §Paediatric Haematology, Oncology and Endocrinology, University Hospital of Essen, Hufelandstr, Essen, Germany

Received February 10, 2004; accepted August 16, 2004.

Supported by DFG-grant AU 117/4-1 and 4-2.

Address correspondence and reprint requests to Dr. Marcus K.H. Auth, Universitäts-Kinderklinik, Abteilung für Allgemeine Pädiatrie, Hufelandstr. 55, D-45122 Essen, Germany (e-mail:

Article Outline
Back to Top | Article Outline


"Who told you that the evil is taken out and the good stays inside." - Johann Dryander, Professor of Medicine, Germany, 1585 (about purging).

Spontaneous recovery from fulminant hepatic failure (FHF) is rare and unpredictable, and mortality in this group of patients is high without emergency liver transplantation (1). Extracorporeal liver support systems (ELS) have been developed to facilitate recovery or bridging until liver transplantation. Ideally, an artificial liver should remove toxins, prevent brain damage and enable liver restoration. Although the exact pathophysiology is not resolved, it is established that ammonia and aromatic amino acids contribute to development of brain oedema, proinflammatory cytokines and nitric oxide to circulatory malfunction (2), bile acids and lactate to tissue damage and hyaluronic acid to liver damage (3). Residual liver cells are required to maintain homeostasis while being exposed to cytotoxins and simultaneously to induce cell proliferation triggered by hepatic growth factors. Following a priming signal by TNF-α and IL-6, hepatocyte proliferation requires major mitogenic stimulation by HGF, EGF and TGF-α (4). Co-stimulatory signals are exerted by insulin-like growth factor, cortisol, insulin, norepinephrine and extracellular matrix degradation (4,5). The tissue repair process also depends on the presence of TGF-β1, an inhibitor of hepatocyte replication, and on signals mediating vascularisation (e.g., VEGF and angiogenin) (4,5).

Bioartificial liver devices (using living hepatocytes) have been evaluated in several hundred adult patients, but there is no compelling evidence yet that these are effective (6). Hepatocytes in these devices were either porcine or derived from hepatoma cell lines. Controlled, randomised studies with MARS, a system of extracorporeal albumin dialysis recirculating over adsorption columns, have rekindelled hopes that there might be safe and effective ways for detoxification in FHF (7,8). In 2001, approval for a volume-adapted, paediatric system (MARSmini) was obtained in Europe, and more than 50 children have been treated in different centres for varying indications (9). At present, no protocol exists to evaluate the efficacy of the system or to compare the spectrum and proportions of removed tissue-damaging substances. More important, the possible transfer of hormones, peptides, and anti-oxidants (e.g., hepatic growth factors) has not yet been examined.

To address this issue as a starting point for a subsequent, collaborative study protocol, we have analysed the transfer of detrimental and protective factors in the clinical application of MARSmini. In the pilot study of our centre, only children with FHF according to the King's College criteria (10) and listing for high urgent liver transplantation are considered as candidates for temporal liver support.

Back to Top | Article Outline


Molecular Adsorbents Recirculating System (MARS) was used as bridging therapy for FHF in four applications in two children with Wilson's disease. Diagnosis was based on massively raised urine copper (basal urine copper 109 μmol/24h in patient 1, and urine copper after penicillamine challenge 404 μmol/24h in patient 2), low plasma caeruloplasmin (140 mg/L in patient 1, and 95 mg/L in patient 2), raised liver copper concentration and appearance of Kayser-Fleischer rings (in patient 2). Chelation therapy with D-penicillamine combined with zinc treatment was immediately started but failed, as indicated by a progression of the Nazer score from 9 to 11 in patient 1 and a sustained score of 8 in patient 2 (11). Acute accumulation of toxic-reactive copper results in FHF, renal failure and haemolytic anaemia. To prevent cerebral oedema and circulatory arrest before emergency liver transplantation could be performed, two sessions of 6 hours on consecutive days were performed in each child. In three sessions, the approved paediatric modification of MARSmini (filter volume 57 ml) was applied (patients 1 and 2), and in 1 session, the standard MARS filter (170-180 ml) was applied (patient 2). Heparin was administered in boluses (200 I.U.) as an anticoagulant to keep the activated clotting time between 120 to 180 seconds according to dialysis standards. MARS-mini (extracorporeal volume of 100 ml) combines high flux haemodialysis (molecular weight cut off 50 kD) against a solution of 20% human albumin, which is recirculated over a charcoal and an anion exchange resin column, and low flux dialysis of the albumin dialysate (8). Patient 1, a 6-year-old girl weighing 23 kg and progressing from grade III to grade IV hepatic encephalopathy (assessed clinically and subsequently by EEG) was treated with hyperventilation, mannitol and vasopressors. Patient 2, a 14-year-old girl weighing 71 kg admitted in multiorgan failure with grade III-IV encephalopathy (assessed by EEG), haemolytic anaemia, respiratory failure, anuria and acute pancreatitis, additionally required continuous haemofiltration. Both patients exhibited very poor prognostic scores (Child C, 14 points; pediatric end-stage liver disease (PELD) score 23 and 24 points) (12).

Parents gave informed consent and defined criteria for MARS applications in our pilot study were approved by the local ethics committee: MARS treatment required listing for high urgent liver transplantation, fulfilment of paediatric risk factors based upon King's College criteria for liver transplantation and no current availability of an donor organ. Analysis of blood, albumin dialysate and haemodialysate (substances listed in Tables 1-3) was performed from samples collected immediately before, 3 hours during and directly after MARS treatment (in Figure 1, absolute levels are depicted only as pre and post values), which had immediately been stored at 2 −70°C. Cytokine levels were measured by ELISA (R&D Systems, Wiesbaden, Germany), nitric oxide metabolites by chemiluminescent analyser (NOA 280; Sievers Instruments, Boulder, CO), cortisol (Adaltis, Freiburg, Germany) and corticosteroid-binding globulin (CBG) by RIA (EBL, Hamburg, Germany) and other clinical chemistry using standard laboratory protocols.

Back to Top | Article Outline


Fifty-six hours and 108 hours after listing and subsequent bridging with MARSmini, liver transplantation was successfully performed. During two MARS cycles in each child, patient 1 remained haemodynamically stable, did not require haemofiltration between MARS sessions, and her PELD score dropped from 23 to 10. Patient 2 required markedly less vasopressor support (norepinephrine was decreased from 9 to 1 μg/kg/min and dobutamine from 10 to 5 μg/kg/min) and parenteral fluid (from 1000 ml to 100 ml per hour), accompanied by a drop in her PELD score from 24 to 19. Patient 2 was off haemofiltration 1 week after LTx, and both patients could be extubated after 1 to 3 days. Neurologic recovery was excellent on discharge after 3 weeks. Liver histology demonstrated end stage liver cirrhosis in both patients and raised liver copper (720 μg/g dry weight in patient 1).

Laboratory data revealed several trends that deserve clinical attention. Although the capacity of the albumin circuit to remove different substances varied intra- and inter-individually, the qualitative analysis of the removed factors was consistent (Table 1). As anticipated, considerable amounts of toxic metabolites, tissue-damaging nitrate and proinflammatory cytokines in the recirculating albumin circuit were transferred, with exception of hyaluronic acid, an extracellular marker associated with liver damage. On the other hand, albumin circuit and dialysate were also removing 9 of 10 growth factors (only excluding CBG) and essential precursors to glutathione synthesis. Binding to the absorbers was detected for amino acids, lactate and 6 of 10 examined hepatic growth factors.

Relative transfer from blood to albumin compartment was high for all metabolites, amino acids, copper and nitrate (median, 34-167%), and lower for cytokines (median, 3.7-10%) (Table 2). Among hepatic growth factors, 4 of 10 factors were consistently transferred at a proportion of greater than 20% (median): cortisol, TGF-β1, IGF-1 and angiogenin. However, transfer of major liver regeneration factors HGF and IL-6 was less than 10% (median), and albumin circuit levels of EGF and VEGF did not increase (Tables 2 and 3).

Notably, considerable removal of a substance was not necessarily accompanied by a decrease of the corresponding blood level (Fig. 1A, B). Whereas reduced blood copper levels of patient 1 were associated with an only moderate increase of albumin copper levels (Fig. 1A), substantially enhanced albumin copper levels (reaching a plateau after 3 hours MARS treatment) yielded no decline of corresponding blood levels in patient 2 (Fig. 1B), who required continuous haemofiltration. However, diuresis of patient 1 was maintained at 2400 mL/d, which contained 41-52 μmol/L urine copper. Consistently decreased blood levels after each MARS treatment were measured for 3 of 6 metabolites, 3 of 6 inflammatory mediators and 1 of 10 growth factors (Table 1).

Data on individual blood levels (Table 3) showed that MARS treatment can involve constantly high bilirubin levels (patient 1) or increasing ammonia and copper levels (patient 2). With regard to side-effects of extracorporeal membrane treatment, no general activation of the inflammatory or complement cascade was observed, and drop of platelets associated with extracorporeal treatment did not require substitution (patient 1: 67/nl to 46/nl platelets; patient 2: 124/nl to 106/nl platelets; the numbers were assessed before the first and after the second treatment with MARS in each patient). No additional substitution of fresh frozen plasma was required under ACT-directed heparinisation, and clotting factors II, V, VII, protein C and S remained stable under MARS therapy (data not shown). Composition of amino-acids changed, reflected by improvement of the Fischer ratio, a marker associated with hepatic encephalopathy (13). Interestingly, during MARS bridging, an increase of hepatic growth factors EGF, TGF-β1 and IGF-1 was detected in serum of both patients.

Back to Top | Article Outline


We have examined the capacity of an extracorporeal adsorption system (MARSmini) to remove candidate substances in two paediatric patients with fulminant hepatic failure of nearly identical aetiology and severity. Taking in consideration the limitations of only four procedures (i.e., dependence on variations in production, equilibration, and adsorbing capacity of the circuit), we believe that indices for important trends can be derived from this pilot study. The novel concept of using albumin as a circulating dialysate purified by adsorbents has been claimed to selectively remove protein-bound-toxins and water-soluble metabolites, being impermeable to valuable proteins, hormones and clotting factors (7). However, although MARSmini application as a bridging device appeared safe and neurologic outcome was excellent in these patients, we clearly demonstrated that not only metabolites and cytokines, but also essential hormones and growth factors are removed by MARS, the most selective form of ELS.

This may have important clinical implications for wider application in children. In a recent, systematic review of randomised trials with 483 patients, the Cochrane group concluded that ELS had no significant effect on mortality compared with standard medical therapy in patients with acute liver failure (6). Lack of efficacy on neurologic outcome or liver restoration was also reported in a series of 49 children with FHF treated with plasmapheresis, an unselective method of detoxification, concomitantly removing essential growth factors and hormones (14). Transfer of growth factors could affect mortality and restrain liver regeneration, as indicated by the vital role of HGF, IL-6, or TNF-α in animal models (5) and significance of adrenal insufficiency in patients with acute hepatic failure (15). The value of protective factors present in plasma from patients with FHF has been attributed to proliferation, cytochrome P450-activity, and reversal of oxidative stress in human liver cells (16). Candidate factors include hormones (4), hepatic growth factors (5), and anti-oxidant precursors (16) (i.e., all factors, which are removed by different ELS to various degree).

In our patients, however, reduction of growth factor blood levels appeared moderate, and some factors (CBG, EGF, TGF-β1, IGF-1) increased over treatment time. Our data agree with recent reports from adults that MARS is capable of removing substantial amounts of metabolites, toxins, cytokines and nitric oxide in FHF. Such removal has been correlated with haemodynamic and neurologic improvement (2,7,17). Safety of MARS has been demonstrated in more than 1500 adult patients (7,9), and mortality could be significantly reduced in patients with acute-on-chronic liver failure, hepatorenal syndrome and liver cirrhosis with progressive hyperbilirubinaemia in randomised, controlled trials (7,8). On the other hand, detection of persisting, high or even increasing levels of ammonia, bilirubin, copper and cytokines following individual courses of a 6-hour treatment in our children could indicate persisting liver damage and also a limitation in detoxification capacity. Saturation of albumin binding capacity and differences in renal elimination may account for divergent pattern of copper distribution in our patients and those reported by other investigators during MARS therapy (18).

The clinical importance of our findings can only be explored in larger series examining patterns of candidate parameters in blood and dialysate under controlled conditions. Collaboration among institutions using ELS on a regular basis would greatly advance the critical evaluation of this promising therapy in paediatric hepatology. The possibility of improving the outcome of liver transplantation and facilitating the induction of liver regeneration are the key issues which should foster application of ELS in children.

Acknowledgments: The kits for the treatment of the patients were kindly donated by Teraklin AG. The authors thank the contributing laboratories of the University of Essen: hygienics (Dr. H. Diederichs), coagulation (Dr. S. Schimanski), clinical chemistry (Dr. L. Volbracht); our paediatric gastroenterologists (A. Hinsberger), nephrologists (A. Wingen, R. Büscher, B. Kranz), and transplant surgeons (PD Dr. M. Malago, Dr. S. Nadalin), Ms. B.K. Gunson for kind review of the manuscript; and Prof. M. Sachs for review of medical history.

Back to Top | Article Outline


1. Shneider B, Alonso EM, Narkewicz MR. Research agenda for pediatric gastroenterology, hepatology and nutrition: hepatobiliary disorders. Report of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition for the Children's Digestive Health and Nutrition Foundation. J Pediatr Gastroenterol Nutr 2002;35(Suppl 3):S268-S274.
2. Sen S, Williams R, Jalan R. The pathophysiological basis of acute-on-chronic liver failure. Liver 2002;22:5-13.
3. Williams AM, Langley PG, Osei-Hwediah J, et al. Hyaluronic acid and endothelial damage due to paracetamol-induced hepatotoxicity. Liver Int 2003:23:110-15.
4. Michalopoulos GM, DeFrances MC. Liver regeneration. Science 1997;276:60-6.
5. Fausto N, Laird AD, Webber EM. Role of growth factors and cytokines in hepatic regeneration. FASEB J 1995;9:1527-36.
6. Kjaergard LL, Liu J, Als-Nielsen B, et al. Artificial and bioartificial support systems for acute and acute-on-chronic liver failure. JAMA 2003;289:217-22.
7. Stange J, Hassanein TI, Mehta R, et al. The molecular adsorbents recycling system as a liver support system based on albumin dialysis: a summary of preclinical investigations, prospective, randomized, controlled clinical trial, and clinical experience from 19 centres. Artif Organs 2002;26:103-10.
8. Heemann U, Treichel U, Philipp T, et al. Albumin dialysis in cirrhosis with superimposed acute liver injury: a prospective, controlled study. Hepatology 2002;36:949-58.
9. Steiner C. International MARS registry [database of the University of Rostock, advised by the MARS research group]. Available at Accessed January 4, 2004.
10. O'Grady JG, Alexander GJ, Hayllar KM, et al. Early indicators of prognosis in fulminant hepatic failure. Gastroenterology 1989;97:439-45.
11. Nazer H, Ede RJ, Mowat AP, et al. Wilson's disease: clinical presentation and use of prognostic index. Gut 1986;27:1377-81.
12. McDiarmid SV, Anand R, Lindblad AS, et al. Development of a pediatric end-stage liver disease score to predict poor outcome in children awaiting liver transplantation. Transplantation 2002;74:173-81.
13. Fischer JE, Rosen HM, Ebeid AM, et al. The effect of normalization of plasma amino acids on hepatic encephalopathy in man. Surgery 1976;80:77-91.
14. Singer AL, Olthoff KM, Kim H, et al. Role of plasmapheresis in the management of acute hepatic failure in children. Ann Surg 2001;234:418-24.
15. Harry R, Auzinger G, Wendon J. The clinical importance of adrenal insufficiency in acute hepatic dysfunction. Hepatology 2002;36:395-402.
16. McCloskey P, Tootle R, Seldon C, et al. Modulation of hepatocyte function in an immortalized human hepatocyte cell line following exposure to liver-failure plasma. Artif Organs 2002;26:340-8.
17. Guo LM, Liu JY, Xu DZ, et al. Application of molecular adsorbents recirculating system to remove NO and cytokines in severe liver failure patients with multiple organ dysfunction syndrome. Liver Int 2003;23:16-20.
18. Sen S, Felldin M, Steiner C, et al. Albumin dialysis and Molecular Adsorbents Recirculating System (MARS) for acute Wilson's disease. Liver Transpl 2002;8:962-7.

‡Johann Dryander (1500-1560): Vom Eymsser Bade/was natures in jmhab. Mainz: Peter Jordan M.D. XXXV. [1585], fol. CIV. Cited Here...

Cited By:

This article has been cited 5 time(s).

Journal of Pediatric Gastroenterology and Nutrition
Serum Hepatocyte Growth Factor and Vascular Endothelial Growth Factor in Children with Acute Liver Failure
Aw, MM; Mitry, RR; Hughes, RD; Dhawan, A
Journal of Pediatric Gastroenterology and Nutrition, 44(2): 224-227.
PDF (110) | CrossRef
Journal of Pediatric Gastroenterology and Nutrition
Are Hepatic Growth Factors Predictors of Clinical Outcome in Fulminant Hepatic Failure?
Auth, MK
Journal of Pediatric Gastroenterology and Nutrition, 44(2): 168-170.
PDF (59) | CrossRef
Current Opinion in Nephrology and Hypertension
Albumin dialysis: an update
Mitzner, SR
Current Opinion in Nephrology and Hypertension, 16(6): 589-595.
PDF (143) | CrossRef
ASAIO Journal
Albumin Dialysis MARS: Knowledge from 10 Years of Clinical Investigation
Mitzner, SR; Stange, J; Klammt, S; Koball, S; Hickstein, H; Reisinger, EC
ASAIO Journal, 55(5): 498-502.
PDF (193) | CrossRef
Current Opinion in Gastroenterology
Inherited metabolic liver disease
Schilsky, ML; Fink, S
Current Opinion in Gastroenterology, 22(3): 215-222.
PDF (156) | CrossRef
Back to Top | Article Outline

Liver failure; Cytokines; Hepatic growth factors; Artificial liver; Children

© 2005 Lippincott Williams & Wilkins, Inc.