Scientific Articles

Research Proposal: Modular Extracorporeal Liver Therapy

Ryder Damen, MEng


Recent advances in extracorporeal liver support devices have been trending towards biological solutions. Known as Bioartificial Livers (BAL), these devices employ live hepatocytes to process the plasma of a patient with acute liver failure (ALF).8 Older, but not as efficacious solutions rely on albumin dialysis (AD) to relieve an ALF patient of built up toxic metabolites such as bilirubin and ammonia.11 Individually, outcomes of both these devices are pre-clinically, and clinically significant in the treatment of ALF respectively.8 There exists however, a gap in knowledge on treatment efficacy when combining BAL and AD. Therefore, the proposed research program aims to evaluate the benefits of combining therapies not only on patient outcome, but on device hepatocyte cytotoxicity. By such evaluation, we gain knowledge in potential complications and benefits of combining these therapies in a pre-clinical setting, and useful data for assisting in BAL scale-up efforts.



Affecting millions worldwide, Acute Liver Failure (ALF) is characterized by the sudden onset of severe hepatic dysfunction, often as the result of viral infection or drug overdose.9 In patients with severe ALF, the current gold standard of treatment is complete orthotopic liver transplant (OLT).9 Many patients die on the waiting list, and mortality rates post-transplant can be as high as 50% in the first year, making this a medical problem without a definitive solution.10 An important tool in combating ALF is the Extracorporeal Liver Support Device (ELSD) – a recent innovation in the field of hepatology and tissue engineering. Known simply as an artificial liver, or external liver, these devices can be subdivided into non-biological and biological systems.

Non-biological systems function on the basis of albumin dialysis, dialyzing toxic albumin-bound built-up metabolites from a patient’s blood into an albumin dialysate.11 One example of such technology is molecular adsorbent recirculating system (MARS) therapy. Once metabolites have dialyzed from a patient’s blood across a semi-permeable membrane, they bind exogenous albumin molecules in a secondary circuit.11 This secondary circuit then undergoes low flux dialysis, stripping metabolites from exogenous albumin into a third waste circuit, and recycling exogenous albumin for continued use.11 While effective in removing protein-bound and water-soluble metabolites, non-biological systems like MARS are ineffective in replacing full hepatic function, simply because they cannot address the wide range of chemical reactions performed by in vivo hepatocytes. 9, 11

To address this challenge, the concept of Bioartificial Livers (BAL) were introduced; employing hepatocytes of an allogeneic or xenogeneic origin as a temporary replacement for a patient’s damaged organ.12 One such example of this technology is the Academic Medical Center of Amsterdam Bioartificial Liver (AMC-BAL), employing in its most recent iteration an immortalized hepatocyte line (HepaRG progenitor line) within a bioreactor module. Affluent media to the AMC-BAL is fed from a plasmaphoresis device connected to a patient’s circulation, allowing the HepaRG progenitor hepatocytes to perform reactions upon it.12 While not yet proven clinically efficacious, this device demonstrates considerably significant improvements in biochemical profiles of ALF animals.12 A major setback in clinical implementation however, is the ability to scale up the device. The AMC-BAL is only currently able to support 40 g in a 537mL bioreactor, or 5% of the functional liver mass of a 70kg human.12 The need remains for additional technology to bridge the gap, and bring the benefits of BAL technology to the clinical stage at the earliest possible opportunity. An additional setback is the response of the device hepatocytes to a patient’s plasma. Compared to standard hepatocyte media, when introduced to healthy, pooled human plasma, hepatic gene transcription rates indicative of function fall to 20%, followed shortly by cytotoxic events.4 This loss of function is only further compounded with the use of ALF plasma.4 It is hypothesized that a protein-bound toxic metabolite is the culprit. Studies have eluded to possible positive effects of combining AD with BAL treatment, but efficacies of combinations have not yet been established experimentally.4 By pre-treating plasma with an albumin dialysis therapy like MARS, toxic-metabolite exposure to BAL hepatocytes is greatly reduced. This in turn allows BAL housed hepatocytes to function to offset the rest of the hepatic spectrum; while greatly reducing their exposure to (and need to reduce) toxic metabolites.4,12 With this reduction in toxic-metabolite exposure, there is the potential for decreased cytotoxicity upon serum exposure. Thus, there exists the potential for increased device longevity and efficacy, potentially requiring a lower cell mass for an effective treatment in humans.



  1. Pre-MARS treated plasma decreases cytotoxicity in AMC-BAL HepaRG progenitor hepatocytes.
  2. Pre-MARS treated plasma in concert with AMC-BAL treatment results in a significantly improved ALF biochemical profile for induced-ALF pigs.


Objective & Specific Aims

The long term objective of this research is to evaluate ways of decreasing the required mass of hepatocytes in BAL devices, and to better understand modular ALF therapies. The immediate objective of this proposal is to evaluate patient and device outcomes of combining AMC-BAL and MARS therapy. The specific aims of this proposal are:

  1. To determine if MARS therapy has an impact on AMC-BAL HepaRG progenitor hepatocyte cytotoxicity.
  2. To determine if MARS and AMC-BAL combined therapy can significantly alter ALF biochemical profiles in a porcine model.


Research Plan

Aim I: Determine if MARS therapy has an impact on AMC-BAL HepaRG progenitor hepatocyte cytotoxicity.

With knowledge that human plasma, especially ALF human plasma, contributes highly to hepatocyte cytotoxicity the aim is to evaluate the impact of MARS on the hepatocytes housed within the AMC-BAL device.3 In order to best simulate physiological human conditions, 350 mL samples of plasma will be collected from 20 ALF patients in Toronto hospitals with standard plasmaphoresis under the supervision of a clinician.3,4 Inclusion criteria for patients include the onset of ALF induced by acetaminophen / paracetamol, 48-96 hours post overdose.3,4 Exclusion criteria for plasma include detectable amounts of acetaminophen / paracetamol in serum, as detected by a standard high performance liquid chromatography (HPLC) assay.4 Upon plasma collection, samples will be pooled into a 7L pool, divided into 12 500mL samples, and one 10 100mL samples for top up, frozen, and at trial commencement thawed. Pooled plasma 500mL samples will be evenly split amongst two categories: experimental and control, the latter receiving only AMC-BAL treatment. The experimental arm will be first subjected to a MARS therapy circuit before being processed with the AMC-BAL device. This order is important to preserve the optimal health of the device hepatocytes. If performed in the opposite direction, hepatocytes would first receive the full force of the plasma’s toxic effects, as opposed to the benefits of MARS beforehand. Twelve replicates of 9mL AMC-BAL reactors will be constructed and prepared with a 2mL cellular pellet comprising of HepaRG progenitor hepatocytes.16 Hepatocytes will be cultured in advance from HepaRG progenitor seed stock, with recommended HepaRG media (without DMSO) in monolayer, with passage once every two weeks at a 1/6 ratio.16 Seed stock will be obtained from ThermoFischer Scientific (HPRGC10 – HepaRG Progenitor Line 10×106 cells) as will HepaRG culture media (HPRG620 16 mL HepaRG media supplement per 100mL Williams Media E A1217601 with 1mL GlutaMAX 35050061).17 Cultures will be maintained according to specification in an incubator with 95% humidity, 20% O2 and 5% CO2.16 After passaging and culturing to a sufficient quantity (2mL cellular pellet), pellets will be resuspended in standard HepaRG progenitor media, and seeded directly into 9mL AMC-BAL devices. Cells will then be allowed to affix for a period of 3 hours, during which time a gas mixture will be passed through device capillaries comprising of 5% CO2, 40% O2 and 55% N2.16 After hepatocyte fixation the device will be perfused with 500mL of HepaRG media, changed every 3 days until the next stage of the experiment, for a maximum of 9 days.16 With all plasma thawed and all devices seeded and ready, culture media will be drained from six control reactors and replaced with the control category of plasma (500mL/device/3 days) (t=0). The remaining six reactors in the experimental arm will have plasma replaced by the same methods, with the addition of an affluent circuit diversion into a MARS circuit, between the peristaltic pump and the reactor input (t=0). MARS circuits will be prepared by purchasing a MARS system with the addition of a MARS Treatment Kit, from GAMBRO (Baxter Intl.).  Plasma will be sampled every 6 hours from the circuit (10mL samples) and replaced with thawed plasma pre-warmed to 37oC. Samples will then be sent for analysis of free ammonia (methods described later). Plasma will be perfused through devices for a total of 72 hours. On experimental completion (t=72 hours), plasma circuits will be disconnected and the devices will be perfused with HepaRG culture media for 20 minutes before hepatotoxicity testing, to remove endogenous plasma artifacts or resident cells. The primary outcome of this study is hepatocyte toxicity; this will be measured by means of flow cytometry (FACS). After HepaRG media flushing, HepaRG cells will be dissociated from devices with 0.25% trypsin-EDTA, washed in PBS and centrifuged.5 Re-suspension of the pellet in PBS will then be followed by staining with the LIVE/DEAD™ Fixable Far Red Dead Cell Stain Kit optimized for red laser flow cytometry, followed by a second centrifugation, wash, and suspension in PBS, followed by immediate flow cytometry interrogation at 650nm after which samples will be discarded.13 Concurrently, residual circuit plasma will be measured for free ammonia, as will all other intermittent samples. Free ammonia will be measured with the AA0100 Ammonia Assay kit as purchased from Sigma Aldrich, an enzymatic kinetic assay standard with serum ammonia measurement.6,7 Live/Dead cellular data and ammonia levels will be analyzed with respective student’s t tests, and any necessary post-hoc analysis will be performed to determine if differences in observed cytotoxicity as observed from flow cytometry have statistical significance.


Aim II: Determine if MARS and AMC-BAL combined therapy significantly alters ALF biochemical profiles in a porcine model

With HepaRG progenitor hepatocytes being of human origin, the optimal testing of this device would be in a human ALF population. Due to standard preclinical and clinical trial procedures however, the combinations of these two therapies must first be evaluated in an animal model with a similar hepatic profile to humans: a porcine model.15 18 female pigs will be obtained with a healthy weight range (43-51kg), housed in a vivarium with ad libitum food (until hepatectomy), and subjected to an initial procedure to establish central access and baseline histology.15 During this procedure, a standard open liver biopsy will be performed to assess baseline histology via a midline laparotomy.15 In addition, a subcutaneous dual-lumen catheter will be placed into the animal’s jugular vein, allowing for access for blood draws, administration of drugs and fluids, and administration of plasmaphoresis and treatment.15 Basic histology will be established by paraffin sectioning and subsequent staining with H&E, and Masson’s Trichrome for the visualization of hepatic structure and pathology.18 Animals will then be given 1 week’s recovery time, at which point blood will be drawn to establish a baseline biochemical profile with the following variables: Serum Ammonia, Blood Urea Nitrogen (BUN), Creatinine, Alanine aminotransferase (ALT), and Aspartate transaminase (AST).15 Variables will be measured offsite at the St. Michael’s Hospital blood laboratory with standard blood analysis procedures: Ammonia (AA0100 Ammonia Assay kit as purchased from Sigma Aldrich), BUN (Sigma Aldrich MAK006 kinetic UV blood urea assay), Creatinine (Sigma Aldrich MAK080 Creatinine colrimetric assay kit), ALT (Alanine Aminotransferase Activity Assay Kit – MAK052), and AST (Aspartate Aminotransferase Activity Assay MAK055).15 Healthy animals, as measured by the absence of sepsis or infection from the previous procedure, will then be randomly assigned into one of two arms: control (AMC-BAL only), and experimental (MARS + AMC-BAL) equally. Animals unable to be included in the study at any time, or at trial end will be euthanized by sodium pentobarbital overdose (with inhaled isofluorane 2-3% as an anesthetic).15 Upon recovery from initial procedure and establishment of a biochemical baseline profile, animals will be sedated with 0.1-0.2mg/kg/min IV propofol, and administered a hepatotoxic dose of 0.75g/kg D-galactosamine (a standard drug for the initiation of drug-overdose induced acute liver failure) via the subcutaneous port.15 At this point, all feeding will be ceased. Upon initiation of the hepatotoxic event, animals will be allowed to progress 48 hours to establish complete hepatic failure confirmed with blood draws every 4 hours (analyzed for aforementioned biochemical profile), at which point a 24-hour intervention will be started.15 In order to minimize complications, overdoses and all other procedures will be administered on a staggered basis, so that no procedure will be performed simultaneously. During intervention, animals will be subjected to either AMC-BAL or MARS + AMC-BAL therapies depending on their assignment in the study. All animals will be lightly sedated during the 24-hour procedure with 0.1mg/kg/min IV propofol. During this procedure standard plasmaphoresis (0.47mm pore size) will be performed via the subcutaneous dual-lumen shunt.14,15 Concurrently, a transcutaneous hepatic biopsy will be performed to establish a comparative histological profile (analyzed by aforementioned histological sectioning and staining). Within the plasmaphoresis circuit, plasma from animals will either be subjected to AMC-BAL therapy, or MARS therapy, and then AMC-BAL therapy, before being returned to the plasmaphoresis circuit, and the patient. Devices will be obtained and set up as per previous methods in Aim 1. Flow will be maintained continuously for the complete duration of the treatment. AMC-BAL devices will be prepared first by culturing hepatocytes from the HepaRG progenitor line with a similar procedure as described in Aim 1.16 HepaRG populations will be first expanded with HepaRG media without DMSO, cultured at 37oC in monolayer, with passage every two weeks at a 1/6 ratio.16 Upon sufficient expansion (15g cells/reactor), cells will be loaded into 150mL AMC-BAL devices and allowed to adhere for 3 hours.16 During this time a standard AMC-BAL gas mixture will be passed through device capillaries (40% oxygen, 55% nitrogen, 5% CO2). After fixation, the device will be perfused with 2.5L of HepaRG media, until ready for trial inclusion. During the 24-hour intervention, blood draws will be drawn every four hours from the affluent lumen before plasmaphoresis, and subjected to the same biochemical profile testing as previously mentioned. Post treatment, animals will be allowed to progress until trial endpoint (t=72 hours) with no further interventions, with blood draws every 4 hours. At the endpoint, animals will be anesthetized with inhaled isofluorane 2-3% and 0.1mg/kg/min IV propofol, subjected to a transcutaneous hepatic biopsy (prepared and stained with previously described methods), and euthanized with sodium pentobarbital overdose.15 Upon trial conclusion, individual biochemical variables will be analyzed with the student’s unpaired t-test, and and post-hoc analysis required. Histological sectioning will be observed for pathology between time points.  


Milestones & Timeline

For both major projects, the following timeline is proposed:


Anticipated Significance

By demonstrating an improved biochemical profile and a reduced cytotoxicity the amount of live hepatic tissue required by a BAL can be potentially reduced. This has far reaching effects when considering the current scope of AMC-BAL and other BAL technologies – scale-up. By furthering our knowledge in modular and combinatory therapies, this research has the potential to bring bioartificial liver devices to functional clinical applications more rapidly. Furthermore, we open up new treatment options for patients with acute, chronic, and acute-on-chronic liver failure, the lattermost of which have fewer treatments available.



  1. Van Wenum, M., Chamuleau, R. A. F. M., Hendriks, E. J., van Gulik, T. M., & Hoekstra, R. (2015). P1303: Human plasma toxicity in differentiated HepaRG progenitor cells in the context of the bioartificial liver. Journal of Hepatology, 62, S328.
  2. Li, J., for Diagnosis, S. K. L., Zhang, Y., Zhou, N., Chen, E., Lu, J., … & Li, L. (2017). LBP-520-Comparison between Li-ALS and Molecular Adsorbents Recirculating System (MARS) for the treatment of acute liver failure in pigs: a randomized controlled study. Journal of Hepatology66(1), S102.
  3. Hughes, R. D., Cochrane, A. M., Thomson, A. D., Murray-Lyon, I. M., & Williams, R. (1976). The cytotoxicity of plasma from patients with acute hepatic failure to isolated rabbit hepatocytes. British journal of experimental pathology, 57(3), 348.
  4. Shi, Q., Gaylor, J. D. S., Cousins, R., Plevris, J., Hayes, P. C., & Grant, M. H. (1998). The effects of serum from patients with acute liver failure on the growth and metabolism of Hep G2 cells. Artificial Organs-Cambridge, 22(12), 1023-1030.
  5. Higuchi, Y., Kawai, K., Yamazaki, H., Nakamura, M., Bree, F., Guguen-Guillouzo, C., & Suemizu, H. (2014). The human hepatic cell line HepaRG as a possible cell source for the generation of humanized liver TK-NOG mice. Xenobiotica; the Fate of Foreign Compounds in Biological Systems, 44(2), 146–153.
  6. Humphries, B. A., Melnychuk, M., Donegan, E. J., & Snee, R. D. (1979). Automated enzymatic assay for plasma ammonia. Clinical chemistry, 25(1), 26-30.
  7. Pesh-Imam, M., Kumar, S., & Willis, C. E. (1978). Enzymatic determination of plasma ammonia: evaluation of Sigma and BMC Kits. Clinical chemistry, 24(11), 2044-2046.
  8. van de Kerkhove, M. P., Chamuleau, R. A. F. M., & Van Gulik, T. M. (2003). Clinical application of bioartificial liver support systems. In Encephalopathy and nitrogen metabolism in liver failure (pp. 389-406). Springer Netherlands.
  9. Bernal, W., & Wendon, J. (2013). Acute liver failure. New England Journal of Medicine, 369(26), 2525-2534.
  10. Plevris, J. N., Schina, M., & Hayes, P. C. (1998). the management of acute liver failure. Alimentary pharmacology & therapeutics, 12(5), 405-418.
  11. Hanish, S. I., Stein, D. M., Scalea, J. R., Essien, E. O., Thurman, P., Hutson, W. R., … & Scalea, T. M. (2017). Molecular adsorbent recirculating system effectively replaces hepatic function in severe acute liver failure. Annals of Surgery, 266(4), 677-684.
  12. van Wenum, M., Treskes, P., Tang, C. Y., Coppens, E. J., Jansen, K., Hendriks, E. J., … & Hoekstra, R. (2017). Scaling-up of a HepaRG progenitor cell based bioartificial liver: optimization for clinical application and transport. Biofabrication, 9(3), 035001.
  13. Noto, A., Ngauv, P., & Trautmann, L. (2013). Cell-based Flow Cytometry Assay to Measure Cytotoxic Activity. Journal of Visualized Experiments : JoVE, (82), 51105. Advance online publication.
  14. Di Nicuolo, G., Kerkhove, M. P., Hoekstra, R., Beld, M. G., Amoroso, P., Battisti, S., … & Bracco, A. (2005). No evidence of in vitro and in vivo porcine endogenous retrovirus infection after plasmapheresis through the AMC‐bioartificial liver. Xenotransplantation, 12(4), 286-292.
  15. Glorioso, J. M., Mao, S. A., Rodysill, B., Mounajjed, T., Kremers, W. K., Elgilani, F., … & Nyberg, S. L. (2015). Pivotal preclinical trial of the spheroid reservoir bioartificial liver. Journal of hepatology, 63(2), 388-398.
  16. Nibourg, G. A., Chamuleau, R. A., Van der Hoeven, T. V., Maas, M. A., Ruiter, A. F., Lamers, W. H., … & Hoekstra, R. (2012). Liver progenitor cell line HepaRG differentiated in a bioartificial liver effectively supplies liver support to rats with acute liver failure. PloS one, 7(6), e38778.
  17. ThermoFischer Scientific. (2017). HepaRG™ Maintenance/Metabolism Medium Supplement Retrieved from
  18. Krishna, M. (2013). Role of special stains in diagnostic liver pathology. Clinical Liver Disease, 2(S1).