The extracorporeal bioartificial liver (BAL) assist system represents a promising methodology for providing temporary support for patients with liver failure either until an organ becomes available, or until the patient's native liver regenerates. Although much work has already been carried out in the development of such devices, many unsolved critical issues remain which hamper the adoption of this treatment modality in the clinic.

The foremost limiting issue involves identification of a reliable source of large numbers of highly functional liver cells. Normal hepatocytes are in limited supply because they have not yet been shown to proliferate to any significant extent in vitro. To address this issue, we are investigating the use of embryonic stem (ES) cells as a potential source for hepatocytes. We have focused on a number of scalable manipulations to differentiate and separate pure populations of hepatocyte-like cells. These approaches include the use of: 1) microencapsulation, 2) techniques for inducing the metabolic machinery that accompanies hepatocyte differentiation, and 3) co-culture with adult hepatocytes and mesenchymal cells. In the latter case for example, ES cells are grown on top of collagen sandwiched hepatocytes to yield a relatively homogeneous population of early endodermal lineage cells. Subsequent co-culture with 3T3 fibroblast feeder layers induces gene and protein expression patterns consistent with hepatocyte-like cells.

A second important problem involves BAL bioreactor design. Due to the absence of oxygen carriers in the fluid that nourishes hepatocytes in the BAL, high flow rates (with accompanying high shear rates) are needed to deliver sufficient oxygen to the large cell mass contained within the BAL. Because hepatocytes are highly sensitive to fluid shear stress, we have developed reactors that are shear protective. In these studies, we have used photolithographic techniques to generate microgrooves onto the underlying glass substrate of a microchannel flat plate bioreactor in an effort to protect the hepatocytes from the shear stress. Hepatocyte viability and function remained stable over several days of perfusion in the grooved-substrate bioreactor, whereas viability and function in the flat-substrate bioreactor decreased over the same time period. We have also scaled-up our microgrooved substrate bioreactor in a radial flow design, consisting of 20 stacked glass substrates with microgrooves on the top surface of each substrate, creating a channel height of 100 ƒÝm between each substrate, with a total bioreactor surface area of 400 cm2 available for cell seeding. Cell density in the bioreactor was as high as 40°Ñ106 hepatocytes in 5 mL fluid volume. In this radial configuration, liver-specific functions also remained stable over 5 days of perfusion.

A third area of investigation involves the modulation of the inflammatory response in hepatic failure. In this regard, we have focused on interleukin-1 (IL-1) blockade as an adjunct therapeutic modality for the treatment of fulminant hepatic failure (FHF). We transfected primary hepatocytes with an adenoviral vector encoding human IL-1 receptor antagonist (AdIL-1Ra), incorporated these cells into our flat-plate BAL device, and evaluated their efficacy in treating D-galactosamine (GalN)-induced FHF in a rat model. After 10 hours of extracorporeal perfusion with the BAL device containing the transfected hepatocytes, there were significant reductions in the plasma levels of hepatic enzymes (aspartate aminotransferase and alanine aminotransferase) and cytokines (IL-1 and IL-6), indicating a beneficial effect. Furthermore, animal survival was significantly improved in the treated group compared to the IL-1Ra alone or hepatocyte alone control groups. These experiments demonstrate that combining inflammatory cytokine blockade with a functional BAL device may provide added therapeutic benefit in the treatment of FHF.