Reduced-size orthotopic liver transplantation

Transcription

1 RAPID COMMUNICATION Arterialized Partial Orthotopic Liver Transplantation in the Mouse: A New Model and Evaluation of the Critical Liver Mass Yinghua Tian, * Rolf Graf, * Wolfram Jochum, and Pierre-Alain Clavien * The availability of a model of partial orthotopic liver transplantation (OLT) in the mouse would be an important tool for studying injuries associated with transplantation. The goals of this study were three-fold: (1) to develop a model of partial OLT in the mouse, (2) to determine the minimal graft volume in this model, and (3) to define the injury associated with small volume incompatible with animal survival. Putative grafts of 30% and 50% were prepared. Their weight was 30 5% and 45 10%, respectively. Subsequently, 30% and 45% syngeneic partial liver grafts were orthotopically transplanted into C57BL/6 mice. Each recipient receiving a 45% graft survived permanently, whereas those receiving only a 30% graft volume died within 2 to 4 days of surgery. Serum transaminase levels normalized in the 45% graft group within 14 days after surgery. In this group, small foci of necrosis and mild steatosis were noted on histology at postoperative day 2, but no abnormalities were noted after 14 days and 100 days. In contrast, recipients who underwent transplantation with a 30% graft volume showed a comparable amount of necrosis and significant microvesicular steatosis in most hepatocytes 2 days after surgery. Hepatocyte proliferation was reduced in this group when compared with animals who underwent transplantation with a 45% graft volume. In conclusion, partial liver transplantation is feasible in the mouse with a critical graft volume ranging between 30% and 45%. Small liver grafts develop massive microvesicular steatosis and impaired regeneration rapidly leading to animal death. (Liver Transpl 2003;9: ) Reduced-size orthotopic liver transplantation (OLT) is a surgical technique that reduces a whole liver to a smaller functional size by means of an anatomic hepatectomy. It includes reduced-size and split OLT and living-related OLT. The first two surgical procedures use livers from cadaveric donors, and the last involves a hepatectomy in a healthy donor. Because the shortage of organs available for transplantation is currently the single most important factor limiting the success of the procedure, these new techniques of partial OLT have triggered much enthusiasm to expand the donor pool, and become an effective therapy in saving many patients with end-stage liver disease. 1,2 The liver is the only solid organ that can restore major tissue loss by regeneration. 3 Although this growth response is critical for the success of partial OLT, the reduction of the liver graft cannot go below a critical mass to ensure normal function and patient survival in the recipient, as well as in the donor in case of living donation. The minimal volume of the graft is determined by its ability to safely repair the injury and regenerate after partial liver transplantation. This size is usually given as the ratio of the graft to recipient body weight (GRBW). In people, it is around 1%; a 75-kg recipient needs a 750-g graft for successful transplantation. Another way of reporting graft sizes is the ratio of the graft volume to standard liver volume (GV/SLV). A GRBW of 1% corresponds to about a 40% GV/SLV. 4 The critical volume necessary to sustain life after resection is significantly less than after transplantation. The reason that more volume is required after OLT is unclear. The availability of a model of partial OLT in mice may help elucidate the mechanisms of graft injury and failure of regeneration after transplantation, and eventually may lead to successful implantation of small grafts. For example, the need for less liver volume would enable safer operation in living related donors, thereby increasing the attractiveness for this unlimited source of organs. Therefore, the purpose of this study was three-fold: first to develop a model of partial OLT in the mouse, second to determine the critical mass necessary for successful OLT, and third to identify the injury associated with a small-forsize graft incompatible with animal survival. From the Laboratory of Hepato-Biliary and Transplantation Surgery, *Division of Visceral and Transplantation Surgery and Department of Pathology, University Hospital of Zurich, Switzerland. Supported by a grant from the National Institutes of Health (DK A1) and from the Swiss National Science Foundation (SNF ) (P.A.C.). Address reprint requests to Pierre-Alain Clavien, MD, PhD, FACS, University Hospital Zürich, Rämistrasse 100, 8091 Zürich, Switzerland. Telephone: ; FAX: ; clavien@chir.unizh.ch Copyright 2003 by the American Association for the Study of Liver Diseases /03/ $30.00/0 doi: /jlts Liver Transplantation, Vol 9, No 8 (August), 2003: pp

2 790 Tian et al Materials and Methods Experimental Design Syngeneic partial OLTs were performed in mice using a 30% and 50% GV/SLV. Recipients were killed at 2 hours, 2 days, 14 days, and 100 days after OLT. Recipient survival, serum aspartate aminotransferase (AST) levels, histology of the graft, and markers of hepatocyte proliferation were examined at each time point. Animals Male inbred C57BL/6, 10 to 12 weeks old with a body weight 23 to 30 g, were used as isograft transplant donors and recipients. Animal care and handling were performed according to Swiss federal guidelines. Operations and treatment protocols were approved by the local ethics committee. All animals were kept in a temperature-controlled environment with a 12-hour light dark cycle; food and water were given ad libitum. Surgical Materials and General Preparations Nylon sutures (5-0, 8-0, 10-0, 11-0) were purchased from Aesculap, Tuttlingen, Germany. Isofluran was from Baxter AG, Volketswil, Switzerland; the polyethylene tubing was from SIMS Portex, Kent, UK. The analgesic, Temgesic, was bought from Essex Chemie AG, Luzern, Switzerland. Surgical instruments were bought from Aesculap, Tuttlingen, Germany. Each operation was performed under general anesthesia using an automatic delivery system (Provet, Lyssach, Switzerland) that provides a mixture of isofluran and oxygen. Surgical procedures were performed under clean but not sterile conditions; a surgical microscope (SZH10, Olympus, Zurich, Switzerland) was used during the entire surgical procedure. The donor and recipient mice were prepared by shaving the abdominal area and disinfecting with Betadine (B. Braun Medical AG, Emmenbrücke, Switzerland). The abdominal cavity of the donor was exposed with retractor assistance. The intestines were wrapped in saline-moistened gauze and gently pulled out of the abdominal cavity. All of the ligaments of the liver were dissected. In the donor, the bile duct was prepared and a tube was inserted and secured with 8-0 suture. The bile duct was cannulated with a 3-mm polyethylene tube with an inner diameter of 0.28 mm and an outer diameter 0.61 mm. A detailed description of the development of the surgical procedure preparing the liver graft and the reconstruction in the recipient is given in the Results section. After reconstruction of the hepatic artery and the bile duct connection in the recipient, the abdomen was closed with a double-layer running suture (5-0). The animal received a subcutaneous injection of 1 ml lactated Ringer solution to replace the fluid lost during the operation and 0.1 mg/kg body weight of Temgesic. During recovery, the recipient mice were placed on a warming pad for 1 hour and were allowed free access to food and water. Biochemical Analysis Serum levels of AST and bilirubin were measured in serum after centrifugation of heparinized blood (3000g for 5 min at 4 C). A multiple biochemical analyzer for serum was used (Ektachem DTSCII, Johnson & Johnson, Rochester, NY). Histology and Immunohistochemistry Tissue specimens were fixed in 4% phosphate-buffered saline buffered formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin or sirius red using standard histologic techniques. Immunohistochemical staining for proliferation cell nuclear antigen (PCNA) was performed as previously described. 5 Statistics Values are expressed as means SD. The data were analyzed using SPSS software version (SPSS Inc, Chicago, IL). Differences of continuous parameters between groups were compared using the Student s t-test. Animal survival was evaluated using the Kaplan-Meyer plot. Results To establish a model of partial liver transplantation, several steps during the operation were carefully monitored and evaluated. Among these, key issues were the control of the anhepatic time, the exact determination of the liver volume, and the reconstruction of the hepatic artery. Which Lobes Can Be Transplanted With an Approximate Graft Size of 50% and 30%? First, we aimed to identify which combination of lobes would yield approximately 30% and 50% graft weight. The actual size was determined by weighing each lobe individually. Before dissection the weight of the total livers (n 5) was also determined (Table 1). The approximate weights of the lobes and the need to select a continuous graft led to the following graft composition: the 30% GV/SLV was obtained by removal of the left lateral lobe, the caudate lobe, and the right lobe. The remaining liver to be transplanted included the median lobe only (Fig. 1A). The 50% GV/SLV was obtained by resecting the left lateral lobe, the caudate lobe, and the left portion of the median lobe. The reduced graft was composed of the right portion of the median lobe and the right lobe (Fig. 1B). Thus, the combined weight of the lobes used for transplantation was calculated to be 45% for the 50% graft and 30% for the 30% graft (Table 1). Therefore, the 50% graft will be referred to as the 45% graft from now on.

3 Partial Liver Transplantation in the Mouse 791 Table 1. Weight of Liver Weight (g) Percent of Total 50% graft 30% graft Left lateral lobe Right lobe Left median lobe Right median lobe Caudate lobe Total Is Arterialized Orthotopic Partial Liver Transplantation Feasible in Mice? After preparing the mouse for the donor operation, the pyloric vein, right adrenal veins, and renal arteries and veins were ligated and severed. The gastroduodenal, splenic, and left gastric arteries were similarly prepared. The branches of the aorta from the celiac artery to the iliac arteries were ligated and severed. The liver lobes were then prepared as described above. After clamping the inferior hepatic vena cava and the portal vein, the remnant liver was perfused through the portal vein with 1 to 2 ml of 0 to 4 C cold saline solution mixture containing 50 U heparin. The suprahepatic vena cava was cut near the diaphragm to allow the outflow of the flush solution. The aorta was transected above the celiac axis and near the iliac artery levels. Thus, the hepatic celiac axis aorta artery segment was carefully preserved. The liver was excised from the surrounding tissue and placed in a container with 0 to 4 C saline solution. In preliminary experiments we found no difference in the outcome between University of Wisconsin (UW) solution or saline solution as long as the preservation time did not exceed 1 hour. The saline solution, probably because of the lower viscosity, is less sticky than the UW solution, we decided to use saline throughout the whole experiment. After the excision of the graft, the gallbladder was removed after ligature of the cystic duct. Polyethylene cuffs were placed on the portal vein and on the infrahepatic vena cava as previously described, 6 and the graft liver was stored in 4 C saline solution until implantation. In the recipient, all ligaments and connections of the liver were dissected. The small branch of the esophageal artery, which supplies the left lobe of the liver, was ligated and severed. The proper hepatic artery was double-ligated and severed. The bile duct was dissected from surrounding tissues and severed at the level of hepatic hilum. A single ligature was placed around the right adrenal vessels. After clamping the portal vein and the infrahepatic vena cava with microvascular clamps, the suprahepatic vena cava was cross-clamped using a custom-made Satinsky clamp to remove the host liver of the recipient. The donor liver was then placed orthotopically into the abdominal cavity of the recipient. The suprahepatic vena cava was anastomosed with a 10-0 running suture. The portal vein was reconnected with a cuff anastomosis. The clamps on the portal and suprahepatic vena cava were released, and the portal venous blood flow was then re-established immediately. Initially, the anhepatic times were close to 20 minutes, a time that is known to be critical for survival of the mice. Figure 1. Schematic representation of the liver indicating the method of hepatic graft reduction. (A) The 30% graft liver was prepared by removal of the left lateral lobe, caudate lobe, and right lobe (darkly hatched areas). The remnant hepatic graft consists of the median lobe only. (B) The 50% graft reduction was prepared by removal of the left lateral lobe, the caudate lobes, and the left median lobe (darkly hatched areas). The reduced graft consisted of the right portion of the median lobe and the right lobe.

4 792 Tian et al Figure 2. Schematic drawing of the arterialization and bile duct anastomosis in the recipient mouse during the 45% partial OLT operation. The cuff technique is used to anastomose the vena cava (a). The donor hepatic-celiac-aorta segment (b) is end-to-side anastomosed with the recipient abdominal aorta (c). The portal vein is anastomosed using a cuff (d). The bile duct is anastomosed with a stent (e). The gall bladder of the donor is removed, and the cystic duct is ligated (f). The learning curve for this part of the procedure was very short because similar operations have already been performed in whole liver transplantation. 6 Eventually, the anhepatic times could be kept between 14 and 17 minutes. The infrahepatic vena cava was then reconnected using the cuff technique. The hepatic celiac axis aorta artery segment of donor liver was connected using an end-to-side anastomosis to the aorta of the recipient below the renal artery level with a 11-0 suture. The arterialization procedure is critical to the success of the operation. The rearterialization using the hepatic celiac axis aorta segment was developed as an alternative to the procedure in the whole organ transplantation. Although this procedure takes 10 to 15 minutes longer for donor preparation than the method described previously, 6 it improves access and facilitates hepatic artery reconstruction in the recipient operation. Finally, the operation was concluded by reconstruction of the bile duct with a polyethylene stent, which was secured with a circular 8-0 suture (Fig. 2). We conclude that this newly developed surgical procedure is a feasible method for establishing a model of partial liver transplantation in mice. What is the Recipient Survival After Partial OLT Using 30% and 45% GV/SLV Grafts? Clinical experience with partial liver transplantation indicates that transplantation of a 50% GV/SLV graft is associated with excellent outcome. Thus, we decided to perform a 50% partial OLT, effectively representing a 45% graft, as a test for this model in mice. After establishment of the surgical procedure, we explored whether there was mortality associated with this operation. All mice that received a 45% liver graft survived until they were killed for analysis. To be sure that long-term survival was possible in this model, we kept six animals for over 100 days without mortality (Fig. 3). To test whether a reduced graft size of 30% GV/SLV was associated with increased mortality, we planned a similar survival study. However, none of the animals survived more than 4 days: all animals showed a high morbidity on day 2 after the operation and died between days 2 and 4 (Fig. 3). Thus, similar to the human situation, a 30% liver graft is not sufficient for achieving recipient survival. This suggests that the minimal partial liver graft required for recipient survival is between 30% and 45% donor liver weight.

5 Partial Liver Transplantation in the Mouse 793 What Type of Liver Injury Is Associated With 30% and 45% GV/SLV Partial OLT? To determine the extent of liver injury in 45% and 30% partial OLT grafts, serum AST levels were measured. In the 45% partial OLT group, AST levels of the recipient mice were initially elevated and decreased in the course of time. They were U/L, U/L, U/L, and U/L at 2 hours, 2 days, 2 weeks, and 100 days after transplantation, respectively, in comparison with baseline values of 80 U/L in untreated livers (Fig. 4). Sera of 30% partial OLT recipients were only available from a few animals at 2 days after the operation, and their AST levels were very high (the average AST level was over 2000 U/L, data not shown). Histologic analysis of 45% partial OLT grafts at 2 days posttransplantation showed small foci of necrosis and mild microvesicular steatosis in less than 10% of the hepatocytes (Fig. 5A). Few neutrophils were present around interlobular bile ducts and within bile duct epithelia. Differences in dilatation of interlobular veins, sinusoids, and central veins were not observed in comparison with age-matched normal liver tissue and 100% OLT grafts. Analysis of 50% partial OLT grafts from 14 days and 100 days posttransplantation showed Figure 3. Kaplan-Meyer survival plot of mice receiving a partial graft. 45% partial OLT: the dashed line represents six animals each that were kept for predefined time points at 14 and 100 days after the operation. None of these animals showed any apparent morbidity. 30% partial OLT: the solid line represents 10 animals. None of the 30% partial OLT recipient mice survived for more than 4 days. Figure 4. AST levels in animals receiving a 45% partial liver graft. Before killing the animals for histology, blood was collected for the determination of serum aspartate aminotransferase levels. Means SD of six animals each are given (baseline is 80U/L, indicated by dashed line). normal liver tissue without evidence of necrosis, bile duct abnormalities, or fibrosis (Figs. 5C and 5D). In the 30% partial OLT group, grafts were analyzed at 2 days posttransplantation because of the high mortality of recipients between 2 and 4 days posttransplantation. Grafts showed few small foci of necrosis and massive microvesicular steatosis (Fig. 5B). Does Liver Regeneration Occur After Partial OLT Using 30% and 50% GV/SLV Grafts? To test the ability of the partial liver graft to regenerate, we stained tissue sections for the proliferating cellular nuclear antigen, a protein upregulated in dividing cells. In the 45% partial OLT group, PCNA immunohistochemistry labeled numerous hepatocyte nuclei compared with untreated controls (Fig. 6A), indicating ongoing liver regeneration. In contrast, few PCNApositive hepatocytes were detected in the 30% partial OLT grafts (Fig. 6B). An alternative protocol for the detection of mitotically active cells measuring the incorporation of 5-Bromo-2 -deoxyuridine was not applied because of the morbidity and mortality of the 30% recipient mice. We conclude that regeneration in the 45% graft is induced, whereas the 30% partial OLT does not show signs of proliferative activity. Discussion This study establishes a new model of partial arterialized OLT in the mouse. Although a GV/SLV of 45%

6 794 Tian et al Figure 5. Histology of the graft after partial OLT. (A) Hematoxylin-eosin stained liver section from a 45% graft 2 days after transplantation with small foci of necrosis. (B, C) Liver sections from a 45% graft 14 and 100 days after transplantation. Note the normal histology of liver acini (B) and portal tracts (C). (D) A 30% graft 2 days after the operation; note massive microvesicular steatosis. was associated with consistent long-term recipient survival, minimal graft injury, and liver graft regeneration, smaller grafts with a GV/SLV of 30% resulted in recipient death within a few days after OLT, impaired regeneration, and massive microvesicular steatosis. The mouse offers unique opportunities for studying the biology of various diseases because its genome is very similar to that of humans. 7 More than 1300 mutant mouse strains are available, offering novel models for studying many human diseases. 8 In particular, transgenic and knockout mice can be used to dissect more specific aspects of cellular injuries. Our recent description of a whole organ OLT model in mice 6 and this article validating a partial OLT model in this species may provide breakthrough tools for further investigating many aspects related to liver transplantation. A central feature for success in our recently reported model of OLT in mice was the need for full rearterialization of the graft. 6 The absence of a patent hepatic artery was associated with poor long-term survival and the development of extensive liver damage. This model has served as the basis for the development of the current model of partial OLT. Because it was clear that the graft would significantly benefit from rearterialization, no attempt was made to perform partial OLT without rearterialization. We modified our previous technique for reconstruction of the hepatic artery 6 by using a hepatic celiac axis aorta segment that was anastomosed to the recipient aorta. With this approach we found that the recipient operation seems easier in the partial than the whole organ OLT because the smaller graft occupies less space in the abdominal cavity. The critical graft size in our model ranges between 30% and 45% GV/SLV, because all recipients receiving a 30% graft died within a few days, whereas all animals receiving a 45% graft survived permanently. This result is similar to the human situation, in which the critical mass seems to be about 40% (or 1% GRBW) 4 and graft Figure 6. Regeneration in liver grafts 2 days after transplantation using proliferation cell nuclear antigen (PCNA) immuohistochemistry. (A) Liver section of 45% graft showing numerous PCNA-positive nuclei. (B) Section from a 30% graft with very few PCNA-positive nuclei.

7 Partial Liver Transplantation in the Mouse 795 function strongly depends on the patency of the hepatic artery. The results also emphasize previous reports 6 suggesting that mice are more representative than rats for human conditions. Rats are more tolerant to small grafts, and partial OLT can be successfully performed with small-for-size liver graft of 30%, even without hepatic artery reconstruction. 5,9 Other similarities between mice and humans, but not rats, include (1) the mouse major histocompatibility complex immune system in mice and the human major histocompatibility complex system in humans; (2) the bile duct anatomy, including the presence of a gallbladder; and (3) the occurrence of bile duct sclerosis after nonarterialized whole organ OLT. 6 It is clear that animal models have some limitations and do not completely mimic the human clinical situation. The advantage of the mouse model, however, is its versatility with respect to genetic strains available: it is conceivable that mice bearing diseases resembling human diseases (e.g., steatosis, cancer, cirrhosis) are used in experiments to better understand their effect on transplantation. Although the precise mechanisms responsible for liver failure after small-for-size graft OLT were not evaluated in depth in this model, the data indicate a consistent pattern of high AST release after reperfusion, impaired regeneration, and diffuse microvesicular steatosis, eventually leading to animal death, typically occurring 2 to 4 days after surgery. The underlying mechanisms responsible for these changes remain speculative. The microvesicular steatosis very likely is not the reason for the animal death, but surely it impairs the metabolic and regenerative capacity of the hepatocytes. Although microsteatosis has not been observed in the clinic, it is not clear whether it occurs after transplantation because of the lack of biopsies performed during the first few days after surgery. Therefore, an unbiased diagnosis of posttransplantation histopathologic changes is not available. In the rat model of small-forsize liver transplantation, vacuolar changes in the cytoplasm of hepatocytes early after transplantation 10 were reported. They seemed similar to the microvesicular steatosis in mice in our report. The metabolic workload of the remnant hepatocytes is increased to maintain homeostasis. The compensatory function of the liver, however, is limited by a certain metabolic rate. If the demand on the liver is too high because of a reduced graft volume, the remnant liver cannot function properly. 3 With significant loss of or damage to the hepatic parenchymal tissue, the need to replace or repair injured cells is induced. 11 Each recipient of 30% partial OLT died despite evidence of a patent hepatic artery. This emphasizes that adequate liver graft volume is critical for survival in this model. Small-for-size grafts lead to lower graft survival. Reduced metabolic capacity 3 may lead to the accumulation of toxic catabolites, which may enhance the injury to the parenchymal cells. This in turn may compromise the ability of hepatocytes to regenerate. In conclusion, partial OLT is feasible in the mouse. The success of the procedure relies on the availability of an experienced and talented microsurgeon and proper reconstruction of the hepatic artery. The model simulates closely the human situation, with a critical graft volume ranging between 30% and 45% GV/SLV. Small-for-size liver grafts cause a consistent pattern of injury resulting in animal death within 2 to 4 days of surgery. This novel model opens the avenue to identifying mechanisms responsible for the poor outcome after small-for-size OLT, and thereby may enable identification of novel protective strategies with positive impact on the shortage of organs. References 1. Shiffman ML, Brown RS Jr, Olthoff KM, Everson G, Miller C, Siegler M, et al. Living donor liver transplantation: Summary of a conference at The National Institutes of Health. Liver Transpl 2002;8: Busuttil RW, Goss JA. Split liver transplantation. Ann Surg 1999;229: Fausto N. Liver regeneration. J Hepatol 2000;32(suppl 1): Kiuchi T, Kasahara M, Uryuhara K, Inomata Y, Uemoto S, Asonuma K, et al. Impact of graft size mismatching on graft prognosis in liver transplantation from living donors. Transplantation 1999;67: Selzner N, Selzner M, Tian Y, Kadry Z, Clavien PA. Cold ischemia decreases liver regeneration after partial liver transplantation in the rat: A TNF-alpha/IL-6 dependent mechanism. Hepatology 2002;36: Tian Y, Rudiger HA, Jochum W, Clavien PA. Comparison of arterialized and nonarterialized orthotopic liver transplantation in mice: Prowess or relevant model? Transplantation 2002;74: Waterston RH, Lindblad-Toh K, Birney E, Rogers J, Abril JF, Agarwal P, et al. Initial sequencing and comparative analysis of the mouse genome. Nature 2002;420: Erickson RP. Mouse models of human genetic disease: Which mouse is more like a man? Bioessays 1996;18: Sun J, Qin G, Wu L, Wang C, Sheil AG. Antigenic load and peripheral chimeric levels in entire and partial liver allograft recipients. Microsurgery 2001;21: Man K, Lo CM, Ng IO, Wong YC, Qin LF, Fan ST, et al. Liver transplantation in rats using small-for-size grafts: A study of hemodynamic and morphological changes. Arch Surg 2001;136: Olthoff KM. Molecular pathways of regeneration and repair after liver transplantation. World J Surg 2002;26: