JOURNAL OF SURGICAL RESEARCH 66, 57–63 (1996) ARTICLE NO. 0372 Extracorporeal Plasma Perfusion of Cultured Hepatocytes: Effect of Intermittent Perfusion on Hepatocyte Function and Morphology PETER STEFANOVICH, M.D., HOWARD W. T. MATTHEW,PH.D., MEHMET TONER,PH.D., RONALD G. TOMPKINS, M.D., SC.D., AND MARTIN L. YARMUSH, M.D., PH.D. Center for Engineering in Medicine and Surgical Services, Massachusetts General Hospital and the Shriners Burns Institute, Boston, Massachusetts 02114 Submitted for publication February 8, 1996 support system capable of providing a bridge to either The most promising approaches to developing a tem- liver regeneration or liver transplantation [1, 2]. Since porary bioartificial liver support system involve incor- the metabolic abnormalities arising from AHF are not porating cultured primary hepatocytes into an extra- completely understood, there is some uncertainty as to corporeal perfusion device. As a result, it is important which hepatocyte functions are most critical for main- to characterize both the phenotypic response of these taining homeostasis. Therefore, the most promising ap- cells during extracorporeal perfusion and the critical proach to a solution involves developing bioartificial factors involved in maintaining differentiated cell systems which incorporate living hepatocytes. function over extended periods of perfusion. In this In vitro function of hepatocytes in bioreactors per- study, hepatocytes cultured in a collagen sandwich fused with medium has been demonstrated by several configuration were connected to a rat via a hollow fi- investigators [3 – 5]. However, it has been difficult to ber plasma separator and perfused with plasma on quantify their function during extracorporeal perfusion line. Perfusions were either continuous for 48 hr or in animal models of AHF, and in some cases results intermittent for up to 174 hr with 6 hr per day of extra- have conflicted. Increases in animal survival and im- corporeal plasma perfusion alternating with 18 hr of provement of key metabolic parameters have been re- culture medium perfusion. During perfusion cell mor- ported after treatment of animal models of AHF with phology was continuously monitored by time-lapse hepatocyte based bioreactors [6, 7]. In other studies [8 – video microscopy. After the procedure, hepatocytes 10], specific hepatocyte functions have been demon- were returned to static culture and function was eval- strated during short (4–6 hr) intervals of perfusion. In uated by measuring the rates of urea synthesis daily for 7 days. During plasma perfusion all hepatocytes these cases, assessment of the effects of extracorporeal accumulated cytoplasmic lipid droplets in a time de- perfusion on the cultured cells was limited to cell viabil- pendent manner. Urea synthesis was maintained at ity tests after perfusion. Kasai et al. reported a 30% initial levels for up to 20 hr of continuous plasma per- decrease in the viability of isolated dog hepatocytes fusion. However, urea synthesis rates were reduced by after 6 hr perfusion [11]. Arnout et al. reported a 55% 31 and 52% after 30 and 48 hr of continuous plasma decrease in the viability of microcarrier attached rat exposure, respectively. With intermittent perfusions, hepatocytes in a hollow fiber system after 8 hr of perfu- as well as with control cells perfused with culture me- sion [8]. In contrast, Rozga et al. reported that micro- dium only, urea synthesis rates did not decrease for at carrier attached dog hepatocytes remained 90% viable least 78 hr of total perfusion. There was no difference after 6 hr of extracorporeal perfusion [6]. Since it is between the urea synthesis rates after 48 hr of cumula- anticipated that patients will require liver support for tive plasma exposure time between cells subjected to several days or even weeks, the hepatocyte response to continuous and intermittent plasma perfusion. These prolonged plasma perfusion needs to be characterized. results suggest that cultured hepatocytes may be ex- Such studies would be facilitated by a simplified perfu- posed to plasma for at least 20 hr with no significant sion system where all cells are subjected to identical, reduction in liver-specific function. Furthermore, an well-controlled conditions. The information obtained intermittent plasma perfusion schedule can be used to should be useful in developing optimal perfusion proto- divide the useful plasma perfusion time over several cols which maintain the long-term functional capacity days with no adverse effects on cell function.  1996 of hepatocyte-based liver assist devices. Academic Press, Inc. Recently, we showed that hepatocytes cultured be- tween two layers of collagen gel maintain differentiated INTRODUCTION function at physiological levels for at least 6 weeks [12]. In a previous study we described a simple technique to The metabolic derangements associated with acute hepatic failure (AHF) have created a need for a liver subject cultured hepatocytes to plasma perfusion and 57 0022-4804/96 $18.00 Copyright  1996 by Academic Press, Inc. All rights of reproduction in any form reserved. AID JSR 4930 / 6n15$$$201 11-18-96 14:49:23 srga AP: Surg Res