Bioreactors for liver assist tested on small animal models are generally scaled-up to treat humans by increasing their size to host a given liver cell mass. In this process, liver cell function in different culture devices is often established based on the metabolite concentration difference between the bioreactor inlet and outlet irrespective of how matter distributes in the bioreactor. In this paper, we report our investigation aimed at establishing whether bioreactor design and operating conditions influence the distribution of matter in two bioreactors proposed for liver assist. We investigated a clinical-scale bioreactor where liver cells are cultured around a three-dimensional network of hollow fiber membranes and a laboratory-scale bioreactor with cells adherent on collagen-coated flat substrata. The distribution of matter was characterized under different operating modes and conditions in terms of the bioreactor residence time distribution evaluated by means of tracer experiments and modeled as a cascade of N stirred tanks with the same volume. Under conditions recommended by the manufacturers, matter distributed uniformly in the clinical-scale bioreactor as a result of the intense backmixing (N=1) whereas axial mixing was negligible in the laboratory-scale bioreactor (N=8). Switching from recycle to single-pass operation definitely reduced axial mixing in the clinical-scale bioreactor (N=2). Increasing feed flow rate significantly enhanced axial mixing in the laboratory-scale bioreactor (N=4). The effects of design, operating mode and conditions on matter distribution in bioreactors for liver cell culture suggest that characterization of the distribution of matter is a necessary step in the scale-up of bioreactors for liver assist and when function of liver cells cultured in different bioreactors is evaluated and compared.
|Number of pages||7|
|Journal||International Journal of Artificial Organs|
|Publication status||Published - 2001|