*Written as a layman review for general scientists.
Acute Liver Failure (ALF) is a disease without a definitive cure.4 Characterized by the sudden onset of liver dysfunction, ALF is most commonly caused by hepatitis (viral infection of the liver) in the developing world, and paracetamol (Tylenol) toxicity in the developed world.4 Orthotopic liver transplant (OLT) remains the gold standard for treating ALF to this day with many patients on waiting lists across the globe.5 With the scarcity of donor organs, surgical complications, and a post-procedure mortality rate ranging from 20-50%, the field of hepatology has been quick to explore other options, namely the concept of the artificial liver.
The concept of an External Liver Support Device (or artificial liver) was first introduced with non-cellular dialysis-based technologies.6 While effective in some aspects of treatment, true hepatic functionality can only be replicated by living hepatocytes – the functional cell type of the liver. Thus the concept of the bioartificial liver (BAL) was introduced. Using a bioreactor filled with hepatocytes, a patient’s blood is filtered into plasma, passed through several membranes, and exposed to a bioreactor (an external device containing live hepatocytes).1 This allows the external device to do the work of a patient’s liver. The goals of using such devices are to either bridge the patient to transplant, or take stress off the in vivo liver and allow it to regenerate.
Building a functional bioartificial liver is tough; currently there exist two methods for populating the bioreactor with hepatocytes.1 The first uses immortalized hepatocyte lines, which are subject to the same difficulties as all immortalized cell lines: loss of full-function and cellular dedifferentiation. The second uses hepatocytes extracted from non-human animals – most commonly pigs. While pig hepatocytes have a functional profile quite similar to human hepatocytes, there exist big immunological risks with the possible transmission of porcine-endogenous retroviruses.7 These viruses present in the genome of pigs can reactivate, cross the human-device barrier and infect patients. This in addition to the necessary animal sacrifice and preparation time make non-human hepatocytes a flawed model at best.
Recently, studies have been exploring the concept of hiHep cells: fibroblasts (connective tissue cells) transformed into hepatocytes.2,3 Using a viral vector to up-regulate three specific genes, fibroblasts of both mice and humans can be transformed into functional hepatocytes with a similar profile to the gold standard primary human hepatocyte found in your liver. This genetic transformation does however carry inherent risks, most notably carcinogenesis (cancer). Previous pre-clinical trials have directly implanted these transformed hiHep cells into test animals with good results.1 In vivo transplantation of hiHeps in humans however carries far too much risk of carcinogenesis to ever be approved in clinical trials or standard practices. If hiHep cells can be housed in an external (extracorporeal) environment with a barrier, they can bring their full range of benefits to a patient without the inherent risk. Such a concept already exists in standard practice for BAL devices, with bioreactors and patients separated through a plasma filter, allowing only the exchange of proteins and small molecules.1,6
This recent study by Xiao-lei et al has demonstrated efficacy in using hiHep cells in a BAL device in a pre-clinical trial, representing a potential new frontier in the treatment of human ALF, and a step towards a mass-produced, functioning artificial liver.1
Cells were first converted from purchased human fetal fibroblasts into hiHeps.1 In a method similar to induced pluripotent stem cell induction, changes in the regulation of 3 genes (FOXA3, HNF1A, HNF4A) were induced by infecting cultures with a lentiviral vector.1,2,3 With increased expression of these genes, fetal fibroblasts underwent a phenotypic transformation to a hepatocyte-like cell: hiHeps.
An important innovation of this trial was the development of a hiHeps optimization technique.1 Induced cells themselves can not yet compare functionally or morphologically to primary human cells, but with optimization the gap can be lessened.2 By washing with collagenase during passage, testing for adhesive proteins (CDH1 and ZO-1), and discarding cells with immature hepatocyte genetic profiles, researchers were able to generate a better functioning pool of hepatocytes than without optimization.1
Once induced, hiHeps were expanded to approximately 3 billion cells in 10 days using 8 1x720cm2 hyper-flasks.1 A major concern during the expansion and optimization process was dedifferentiation and subsequent loss of function, a problem all too common with expansion of cells in vitro. Researchers monitored morphology, testosterone clearance, ammonia clearance, and other markers of hepatic functionality before, during, and after expansion. Data indicated a slight, but not significant decline in functionality and genetic expression of certain factors. This bodes well for future clinical use, as it demonstrates hiHeps can be expanded quickly without the common loss of function seen all too often in tissue engineering.
As a pre-clinical trial, the study was performed on 20 adult Bama miniature pigs.1 In order to best replicate human acute liver failure, animals were injected with 0.40 g/kg dosages of D-Galactosamine, a hepatotoxic drug. A dose of 0.40 g/kg in previous trials had shown animal survival times extending to 1 week, providing a broad treatment window. Animals were allowed to progress without treatment for 24 hours, allowing adequate time for them to clear the D-Galactosamine from their systems. Animals were stratified into three groups for adequate controls: no-treatment (6 pigs), empty-BAL treatment (6 pigs), and hiHeps-BAL treatment (8 pigs).
BAL devices were perfused with hiHeps cells and 24 hours was allowed for the cells to adhere to the polycarbonate plates.1 In a 65-layer stack, these cells were circulated with media that was in contact with the ALF plasma of test animals, allowing for the free exchange of small molecules and proteins – but not cells. Treatment consisted of exposing the animals to the BAL device for a period of 3 hours at 24 hours post D-Galactosamine injection. Results were then observed for the remainder of the study (until t=7 days).
The main outcome this trial sought to measure was survivorship.1 Out of the 8 pigs that received hiHeps treatment, 7 survived until trial conclusion (1-week post D-Galactosamine injection). In the empty-BAL treatment group, none of the 6 animals survived until trial conclusion. In the no-treatment group, only one animal survived until trial conclusion.
In addition to survivorship, standard blood biochemical profiles were continuously monitored.1 In the hiHeps treatment group, alanine aminotransferase (ALT), aspartate aminotransferase (AST), serum ammonia and total bilirubin levels all declined and eventually normalized by end of trial. This normalization was not seen in other groups.
While catalysis is commonly measured in BAL studies, anabolism is equally as important. This study also profiled the production of human-liver proteins – albumin and alpha-1 antitrypsin (AAT) – in pig serum and in bioreactor media.1 These proteins in the trial could only be produced by the hiHeps cells.
With the survival of all but one of the hiHeps-BAL treatment animals, and early death of all but 1 of the control animals, there is a strong therapeutic correlation between hiHeps BAL treatment and survivorship in an ALF event.1 Survivorship correlations have often demonstrated in animal models with other BAL treatments, but the use of hiHeps presents a promising opportunity for this model.
The declines in ALT, AST, serum ammonia, and total bilirubin values are often accompanied by the restoration of normal daily activity in animals and humans that have undergone acute liver failure.1 In the case of the trial, treated pigs with these declined values were up and moving after treatment, indicating an at least partial recovery from ALT-associated conditions like hepatic encephalopathy.
The presence of human albumin and AAT was found in serum, indicating the ability of hiHeps synthesized proteins to cross the semi-permeable membrane.1 The membrane exists to protect the patient from the device, but it often limits the exchange of small molecules and proteins. These findings indicate that hiHeps-based bioreactors can bring the full range of metabolism (both anabolic and catalytic) to a patient.
With demonstrated efficacy in an animal model, clinical trials for hiHeps based BAL systems have been scheduled and are currently preparing for recruitment.8 Before these devices can be used in a clinical setting, scientists need to overcome some challenges.
To adequately treat a 70kg human, approximately 10 billion hepatocytes are required.1 This in itself is a formidable challenge, as the production of this much functional tissue requires an immense amount of time and materials. Combining hyperstack technology and cryopreservation (shown to be effective with hiHeps) is however a method that could grow the necessary tissue mass. The first step of these clinical trials is to establish safety and efficacy in humans. While these devices are not yet as effective as a full healthy liver, with continued research and optimization, they could one day exceed that efficacy. It is entirely likely that in the near future these devices will be as ubiquitous in hospitals as renal dialysis machines, allowing patients with ALF to resume a normalized life, and paving the way for a true cure.