Hepatotrophic Factor Research Articles

Total blood exchange suppresses the early stage of liver regeneration following partial hepatectomy in rats.

Despite its obscure and short effect, plasma exchange (PE) remains a mainstay in the treatment of liver disease. However, the question still remains as to whether or not PE suppresses the regeneration of the liver because PE deprives patients of hepatotrophic factors. The effect of PE, which could be a total blood exchange (TBE) in a syngeneic setting, on liver regeneration following a 68% partial hepatectomy (PH) was investigated in rats. In Group 1, 20 ml of blood from normal rats was infused while native blood was removed at 6 and 12 h after PH. In Group 2, 20 ml of blood obtained from PH rats at the same time points was infused. The regeneration rate, labeling index of proliferating cell nuclear antigen (PCNA), and plasma hepatocyte growth factor (HGF) level were determined, and standard liver function tests performed at 24, 48, and 72 h. Although all liver function tests improved in Group 1 at 24 and 48 h, the regeneration rate was significantly impaired. Similarly, the PCNA labeling index was significantly lower in Group 1 than that in Group 2. The plasma HGF level was significantly reduced in Group 1 (6 h blood out versus blood in: 1.1+/-0.5 vs. 0.1+/-0.1 ng/ml, p < 0.05). TBE with normal blood following PH suppressed the early stage of liver regeneration, in part, because of the reduction of HGF even though the blood was purified.

Induction of mdr1b mRNA and P-glycoprotein expression by tumor necrosis factor alpha in primary rat hepatocyte cultures.

Mammalian liver exhibits expression of members of the family of multidrug resistance (mdr) transporters (P-glycoproteins). P-glycoprotein isoforms encoded by mdr1 genes participate in extrusion of an array of xenobiotics into the bile. Induction of mdr1b mRNA expression has been shown to occur in rat hepatocytes in response to hepatotrophic growth factors. As the cytokine tumor necrosis factor alpha (TNF-alpha) is known to exert a direct mitogenic effect on hepatocytes, its influence on mdr1b expression was investigated. In primary rat hepatocytes cultured in the absence of TNF-alpha, a time-dependent increase in basal expression of mdr1b mRNA and in immunodetectable P-glycoprotein was observed. In cells treated with TNF-alpha (4,000 U/ml) for 3 days, expression of mdr1b mRNA and of immunodetectable P-glycoprotein was induced approximately twofold. Moreover, intracellular steady-state levels of the mdr1 substrate rhodamine 123 were decreased in cells pretreated with TNF-alpha in comparison to controls, indicating an increase in functional transporter(s) mediating dye extrusion. Treatment of hepatocytes with antioxidants (1 mM ascorbic acid and 2% dimethyl sulfoxide) for 3 days markedly suppressed mdr1b mRNA and P-glycoprotein expression both in cells cultured in the presence of TNF-alpha and in the absence of the cytokine, but did not fully abolish mdr1b mRNA induction by TNF-alpha, supporting the notion that reactive oxygen species participate in regulation of basal mdr1b gene expression during hepatocyte culture. In conclusion, the present data indicate that by inducing mdr1b expression in hepatocytes, TNF-alpha may affect the capacity of the liver for extrusion or detoxification of endogenous or xenobiotic mdr1 substrates.

Aging and the liver: an update.

The issue of whether or not liver function is compromised in the elderly population remains unresolved. Numerous age-related changes in hepatic structure and function have been described, but many of these observations are qualitative, were made under suboptimal experimental conditions, or are simply contradictory. Changes in hepato-cellular structural parameters, e.g., increased hepatocyte size, increase in the number of binucleated cells, altered mitochondria, and endoplasmic reticulum, have been reported. However, quantitative morphological analyses have refuted many of these observations. There are few functional data that correlate with structural changes. Serum and biliary cholesterol appear to rise, predisposing elderly people to increased incidences of coronary disease and gallstones, respectively. The rate of liver regeneration declines in old animals, but the regenerative capacity remains unchanged, perhaps reflecting an age-associated reduction in the response to hepatotrophic factors. This senescent change has important clinical implications with regard to surgical intervention for liver disease, e.g., resection or transplantation. Nevertheless, most outcomes studies suggest that age alone should not be a determining factor in such clinical decisions. Geriatric patients exhibit a decline in the hepatic clearance of certain drugs and a marked increase in the frequency of adverse drug reactions, reflecting an increase in polypharmacy regimens and declines in liver volume and blood flow rather than reduced Phase I metabolism. Although the livers of elderly subjects are characterized by a decline in adaptive responsiveness and reduced reserve capacity, clinical tests suggest that liver function is well-maintained in this age group.

Maintained function of primary human hepatocytes by cellular interactions in coculture: implications for liver support systems.

The application of primary hepatocytes in hybrid artificial liver systems has been hampered by the gradual loss of differentiated morphology and function in vitro. Therefore, we have established a coculture model of autologous human hepatocytes and biliary epithelial cells (BEC) in collagen gel in the presence of hepatotrophic growth factors. Furthermore, we examined the effect of hepatocyte cell perfusion in a women multicompartment capillary membrane system. Normal hepatocytes isolated from human liver produced albumin for more than 2 weeks in serum-free media, and were further stimulated by conditioned medium. When cocultured with BEC, albumin secretion was greatly enhanced, suggesting that cellular interactions promote tissue-specific differentiation. When perfused in bioreactors, albumin levels were maintained at steady state for longer than 2 weeks. These data indicate that differentiation of primary hum hepatocytes can be maintained by coculture interactions and three-dimensional hybrid organ devices, providing appropriate growth factors and matrix for tissue regeneration.

Involvement of growth factor receptor‐bound protein‐2 in rat hepatocyte growth

Growth factor receptor-bound protein-2 (GRB-2) is a protein linking receptor tyrosine kinase and Sos (Son of Sevenless gene; Ras GDP/GTP exchange protein), leading to activation of the Ras-mitogen-activated protein kinase (MAPK) cascade. So far, it remains unclear how GRB-2 plays a role in signal transduction pathways evoked by hepatotrophic factors. This study was attempted to evaluate the involvement of GRB-2 in signalling in rat hepatocyte growth. Using rat cultured hepatocytes stimulated by hepatotrophic factors and regenerating livers after partial hepatectomy (PH) we examined GRB-2-mediated linkage of hepatotrophic factor receptors to signal transducing molecules such as Sos or dynamin-II by immunoprecipitation and western blot analysis. In primary cultured hepatocytes stimulated with hepatocyte growth factor (HGF) or epidermal growth factor (EGF), GRB-2 linked HGF receptor or EGF receptor, respectively, to Sos which activated the mitogen-activated protein kinase (MAPK) cascade. In contrast, in primary cultured hepatocytes stimulated with insulin, GRB-2 linked insulin receptor substrate-1 (IRS-1) to dynamin-II as well as Sos. In the early phase after PH, GRB-2 activated the Ras-MAPK cascade by linking HGF receptor, IRS-1, or EGF receptor to Sos. In the late phase after PH, a complex of IRS-1-GRB-2 associated with dynamin-II, indicating that GRB-2 may transduce signals from IRS-1 to dynamin-II. We conclude that GRB-2 may play a role in transmitting signals from hepatotrophic factors to not only MAPK but also to other signalling pathways in hepatocyte growth.

Maintained function of primary human hepatocytes by cellular interactions in coculture: implications for liver support systems

Abstract The application of primary hepatocytes in hybrid artificial liver systems has been hampered by the gradual loss of differentiated morphology and function in vitro. Therefore, we have established a coculture model of autologous human hepatocytes and biliary epithelial cells (BEC) in collagen gel in the presence of hepatotrophic growth factors. Furthermore, we examined the effect of hepatocyte cell perfusion in a woven multicompartment capillary membrane system. Normal hepatocytes isolated from human liver produced albumin for more than 2 weeks in serum-free media, and were further stimulated by conditioned medium. When cocultured with BEC, albumin secretion was greatly enhanced, suggesting that cellular interactions promote tissue-specific differentiation. When perfused in bioreactors, albumin levels were maintained at steady state for longer than 2 weeks. These data indicate that differentiation of primary hum hepatocytes can be maintained by coculture interactions and three-dimensional hybrid organ devices, providing appropriate growth factors and matrix for tissue regeneration.

Human hepatopoietin augmenter of liver regeneration — A hepatotrophic factor for liver regeneration, and its potential antihepatitis effectin vivo

The effect of human hapatopoietin i. e. augmenter of liver regeneration (hALR) was determined on hepatocyte DNA synthesis, and on CCl4-induced hepatitis in animal invitro andin vivo. It is found that ALR could directly stimulate DNA synthesis of hepatocytes in primary culture in a dose-dependent manner. In 30% hepatectomied rats, significant DNA synthesis occurred in control rats; even so, exogenous hALR gene encoding protein increased DNA synthesis of regeneration liver by 2.3 fold compared with control rats. A lower dose of CCl4 was administrated in rats and the effect of ALR on DNA synthesis of regenerating liver 48 h after CCl4 administration was analyzed. Few Budr-labeled hepatocytes were visible in control rats. However, exogenous hALR markedly increased the number of labeled cells in a dose-dependent manner. Statistical analysis showed that 10 or 40 μg· kg-1 ALR protein stimulated DNA synthesis by 1.7 and 4.8 fold respectively.

Stimulation of rat hepatocyte proliferation in vitro and in vivo by factors derived from the bovine small intestinal mucosa.

A factor with a molecular weight of less than 1 kDa in the mucosa of the bovine small intestine (low molecular weight factor or LMW factor) stimulated DNA synthesis in rat hepatocytes in primary culture. This factor only showed its activity when it was added with a larger factor with a molecular weight of 30 kDa that was also found in the same tissue (high molecular weight factor or HMW factor). The LMW factor probably acts to enhance the action of a hepatotrophic growth factor, since EGF and HGF can substitute for the HMW factor. The action of the LMW factor was not due to the actions of low molecular weight substances such as norepinephrine, estradiol, triiodothyronine, and putrescine, which enhance the action of EGF or HGF, since substantial amounts of these substances were not found in the extract. When intraperitoneally administered into rats, after two-thirds hepatectomy, the LMW factor enhanced hepatocyte proliferation without the administration of the HMW factor. In the regenerating liver, a hepatotrophic growth factor(s), which acts synergistically with the LMW factor, might be properly provided, but the supply of the LMW factor might be below the level that maximally stimulates hepatocyte proliferation.

Stimulation by histamine of TPA-dependent hepatocyte growth factor production in promyelocytic leukemia HL-60 cells.

Hepatocyte growth factor (HGF) is most likely a physiological hepatotrophic factor that triggers regeneration of the injured liver. Histamine may also be important in the pathophysiology of the injured liver. Previously we showed that histamine production was increased in liver macrophages of mice injected with CCI4, a well-known hepatotoxin. Therefore, it is likely that the biological actions of histamine in repairing processes of the injured liver are mediated by HGF. This study was aimed at examining the effects of histamine on production of HGF using, as a model, the human promyelocytic leukemia cells, HL-60. 12-o-Tetradecanoylphorbol-13-acetate (TPA) markedly stimulated HGF production and release from the cells; the maximal amount of HGF was released at a concentration of 3 ng/ml of TPA. Histamine significantly stimulated the TPA-induced HGF production and release in these cells, depending on incubation time and its dose. These actions of histamine were abrogated by a H2 receptor antagonist, ranitidine.

Simultaneous or Delayed Administration of Hepatocyte Growth Factor Equally Represses the Fibrotic Changes in Murine Lung Injury Induced by Bleomycin

Hepatocyte growth factor (HGF) is a humoral mediator of epithelial-mesenchymal interactions, acting on a variety of epithelial cells as mitogen, motogen, and morphogen. Exogenous HGF acts as a hepatotrophic factor and a renotrophic factor during experimental injury. To investigate whether HGF has a pulmotrophic function, human recombinant HGF was administered to C57BL/6 mice with severe lung injury by bleomycin (BLM). Low dose simultaneous and continuous administration of HGF (50 micrograms/mouse/7 d) with BLM (100 mg/mouse/7 d) repressed fibrotic morphological changes at 2 and 4 wk. Ashcroft score showed a significant difference in lung fibrosis with and without HGF at 4 wk (3.7 +/- 0.4 versus 4.9 +/- 0.3, p < 0.05). Furthermore, either simultaneous or delayed administration of high dose HGF (280 micrograms/mouse/14 d) equally repressed fibrotic changes by BLM when examined at 4 wk (Ashcroft score: 2.6 +/- 0.4 and 2.4 +/- 0.2 versus 4.1 +/- 0.2, p < 0.01). Hydroxyproline content in the lungs was significantly lower in mice with either simultaneous or delayed administration of high dose HGF as compared to those administered BLM alone (121.8 +/- 8.1% and 113.2 +/- 6.2% versus 162.7 +/- 4.6%, p < 0.001). These findings indicate that exogenous HGF acts as a pulmotrophic factor in vivo and prevents the progression of BLM-induced lung injury when administered in either a simultaneous or delayed fashion. HGF may be a potent candidate to prevent or treat lung fibrosis.

Specific activation of AP-1 but not Stat3 in regenerating liver in mice.

Liver regeneration following partial hepatectomy is regulated by hepatotrophic factors whose precise roles are still elusive. In cell culture studies, some of them have been shown to activate members of the family of the signal transducers and activators of transcription (Stat). To test this contention in vivo, nuclear extracts were isolated from livers of partially hepatectomized and sham-operated mice killed at 30 different time points between zero and 108 h after surgery. Stat3 DNA binding is rapidly induced after surgery in both partially hepatectomized and sham-operated mice. Maximum activation of Stat3 is achieved 4-6 h after resection, and elevated Stat3 activation is detected as late as 60 h after surgery in both groups. Activated Stat5 is found sporadically in both sham-operated and resected mice but appears to be absent in the first 12 h after partial hepatectomy. Neither Stat1, Stat2, Stat4 nor Stat6 is induced during the time of observation. In contrast, AP-1 DNA binding activity is specifically induced in regenerating mouse liver.

Ethanolamine modulates the rate of rat hepatocyte proliferation in vitro and in vivo.

A low molecular weight, heat-resistant hepatotrophic factor in an extract from the bovine intestinal mucosa was purified and identified as ethanolamine by structural analyses. The mode of action of ethanolamine in vitro and in vivo coincided with that of the crude extract of the tissue, indicating that ethanolamine is the active component. Ethanolamine synergistically elevated the stimulation of DNA synthesis in hepatocytes in primary culture when added together with a growth factor, such as epidermal growth factor, with the ED50 being 20 microM, although it showed little stimulatory effect by itself. Contrary to these in vitro results, the intraperitoneal administration of ethanolamine hydrochloride (24 mg of ethanolamine per kg of body weight) enhanced hepatocyte proliferation in regenerating rat livers after two-thirds hepatectomy without the administration of any growth factors. In the regenerating liver, hepatocyte proliferation may be initiated by an endogenous growth factor, but the supply of ethanolamine in circulation may not be sufficient for optimal hepatocyte proliferation; thus, the exogenous administration of ethanolamine may further enhance hepatocyte proliferation. Ethanolamine in circulation may be a humoral hepatotrophic factor.

Serial Ultrasound-Guided Percutaneous Liver Biopsy in a Partial Hepatectomy Porcine Model: A New Technique in the Study of Liver Regeneration

Studies on liver regeneration are often carried out on small animal models such as the rat. The use of large animal models however has many advantages. Repeated biopsies of the liver may be carried out and the animal can thus serve as its own control. Sequential changes in the regenerative response to various putative hepatotrophic factors can therefore be studied. Ultrasound-guided percutaneous biopsy (USGPB) of the liver, a technique widely used in clinical practice to obtain liver samples, offers a means for repeated liver biopsies. This study was carried out to validate the use of serial USGPB of the liver as a research technique in the study of liver regeneration in a porcine hepatectomy model. Material and method: Left hepatectomy was carried out in nine Yorkshire pigs (20–30 kg). Bromodeoxyuridine (BrdU) at 50 mg/kg BW was administered daily for the first 5 days after surgery. USGPB was carried out 1 hr later using 16F biopsy needles in a Biopty device. Ultrasonography was carried out using an Acuson XP10 color doppler ultrasound machine. Immunoperoxidase staining of the biopsy specimens was performed using mouse monoclonal antibody to BrdU. Regenerating cells were identified and labeling indices constructed for each day after hepatectomy. All animals were sacrificed at the end of the study and postmortem was performed. Results and conclusion: USGPB was successful at every attempt and no animal suffered complications following the procedure. Median cumulative labeling indices increased rapidly after hepatectomy and the steepest rise occurred between the third and fourth day after surgery when it increased from 2.25 to 3.40% before leveling off. These results are consistent with those obtained by other reported techniques. USGPB of the liver overcomes the need for repeated laparotomies to obtain biopsy samples in a large animal model and has no morbidity and mortality that could lead to animal wastage.

The in vivo effect of hepatotrophic factors augmenter of liver regeneration, hepatocyte growth factor, and insulin-like growth factor-II on liver natural killer cell functions.

Fine balanced sequential changes of the levels of circulating hepatotrophic factors are essential for normal liver regeneration. Our recent studies have indicated that liver-resident natural killer (NK) cells are important regulators of liver regeneration and have raised the possibility that hepatotrophic factors might mediate their activities through NK cells. In the present study, we assessed the effects of in vivo administration of three hepatotrophic factors (augmenter of liver regeneration [ALR], insulin-like growth factor-II [IGF-II], and hepatocyte growth factor [HGF]) on NK cells in normal rats. Each of the three, given over a 1-day period in doses known to produce hepatotrophic activity, induced inhibition of NK cell cytotoxic activities in the population of mononuclear leukocytes (MNL) in the liver, but not in MNL from the spleen or peripheral blood. In contrast to these results obtained by the whole animal treatment, the three molecules had no effect on NK cell functions when added to cultures of MNL from the livers, spleens, or blood of untreated rats. These data support and extend our previously advanced hypothesis that ALR and other hepatotrophic factors play an important role in liver regeneration by regional regulation of NK cells through some as-yet-unknown intermediary mechanism.

The in vivo effect of hepatotrophic factors augmenter of liver regeneration, hepatocyte growth factor, and insulin-like growth factor-II on liver natural killer cell functions

Fine balanced sequential changes of the levels of circulating hepatotrophic factors are essential for normal liver regeneration. Our recent studies have indicated that liver-resident natural killer (NK) cells are important regulators of liver regeneration and have raised the possibility that hepatotrophic factors might mediate their activities through NK cells. In the present study, we assessed the effects of in vivo administration of three hepatotrophic factors (augmenter of liver regeneration [ALR], insulin-like growth factor-II [IGF-II], and hepatocyte growth factor [HGF]) on NK cells in normal rats. Each of the three, given over a 1-day period in doses known to produce hepatotrophic activity, induced inhibition of NK cell cytotoxic activities in the population of mononuclear leukocytes (MNL) in the liver, but not in MNL from the spleen or peripheral blood. In contrast to these results obtained by the whole animal treatment, the three molecules had no effect on NK cell functions when added to cultures of MNL from the livers, spleens, or blood of untreated rats. These data support and extend our previously advanced hypothesis that ALR and other hepatotrophic factors play an important role in liver regeneration by regional regulation of NK cells through some as-yet-unknown intermediary mechanism.(Hepatology 1997 Feb;25(2):411-5)

Contribution of primary cultures of adult human hepatocytes to the pathophysiology of hepatocellular carcinoma

Background/Aims: The mechanisms of hepatocarcinogenesis are still poorly understood. The development of hepatocellular carcinoma has recently been shown to be associated with increased DNA synthesis in cirrhosis. The aim of this work was to determine whether the high rate of hepatocyte regeneration observed in cirrhotic liver with hepatocellular carcinoma is associated with the presence of a growth factor that could be detectable in the serum. Methods: Adult human hepatocytes in primary culture, allowing the evaluation of the release of circulating hepatotrophic factors, were used. These cultures were treated for 48 h with serum from patients with cirrhosis with and without hepatocellular carcinoma, from patients with liver metastasis, and from healthy subjects. The rate of DNA synthesis in these cultures was assessed by measuring the amount of [ 3H]-thymidine incorporation into genomic DNA. Results: On average, the synthesis of DNA was increased 2.5-, 2.2-, 2.1-, and 2.3-fold, respectively, in response to serum from patients with cirrhosis with hepatocellular carcinoma, from patients with cirrhosis without hepatocellular carcinoma, from patients with liver metastasis, and from healthy subjects. Conclusions: We conclude that the hepatotrophic activity of the serum is not significantly different in patients with cirrhosis with or without hepatocellular carcinoma. These results suggest that the increased DNA synthesis in hepatocytes of cirrhotic liver with hepatocellular carcinoma might be due to prolfierative factor(s) acting by paracrine or autocrine pathways.

Heparin Functions as a Hepatotrophic Factor by Inducing Production of Hepatocyte Growth Factor

Hepatocyte growth factor (HGF), a potent mitogen for hepatocytes, is a hepatotrophic factor for liver regeneration. Addition of heparin to cultures of human fibroblasts, leukemic cells, and umbilical vein endothelial cells stimulated HGF production to 3-6-fold higher levels than seen in the absence of heparin. Administration of heparin to rats increased blood HGF levels to a 2.5-5-fold higher level than that in control rats given saline alone, and consequently induced a remarkable enhancement of liver regenerationin vivoafter a 30% partial hepatectomy. However, heparin had no stimulatory effect on DNA synthesis of hepatocytes in primary culture. These results indicate that heparin has hepatotrophic function, an action possibly mediated through the stimulation of production of HGF.

Hepatocyte growth factor and c-met expression in Long-Evans Cinnamon rats with spontaneous hepatitis and hepatoma

The Long-Evans Cinnamon (LEC) rat is characterized by the spontaneous onset of acute and chronic hepatitis, followed by occurrence of liver cancer, and is thus able to provide a unique experimental model for human genetical liver disease, Wilson's disease. Hepatocyte growth factor (HGF) is a potent hepatotrophic factor in liver regeneration, and its expression is up-regulated in response to liver injuries. We found that the plasma HGF level in LEC rats rose markedly during the fulminant hepatitis phase, fell during the phase of chronic/cholangiofibrosis, and fluctuated during the hepatoma phase. Immunohistological staining of the liver revealed that the number of HGF-positive cells increased remarkably during the fulminant hepatitis phase, and that many of these cells were localized at the portal triads. Fewer HGF-positive cells were observed during the phase of chronic hepatitis. The surface of the hepatocellular carcinoma (HCC) cells and the cytoplasm of the nonepithelial cells in cancerous liver tissues were HGF-positive. The HGF-messenger RNA (mRNA) level in the liver rose in the fulminant hepatitis phase, fell in the chronic hepatitis phase, and was intermediate or high during the hepatoma phase. The expression of c-met mRNA was strong in the tissues of LEC rats with fulminant hepatitis and, especially, in the cholangiofibrosis tissues. c-met mRNA was also detected in HCCs. These results suggest that the HGF-c-met system may play an important role in the regeneration of hepatocytes as well as in the development of HCC in paracrine or autocrine mechanisms. (Hepatology 1996 Sep;24(3):596-602)

Hepatocyte Growth Factor and C– Met Expression in Long–Evans Cinnamon Rats With Spontaneous Hepatitis and Hepatoma

The Long-Evans Cinnamon (LEC) rat is characterized by the spontaneous onset of acute and chronic hepatitis, followed by occurrence of liver cancer, and is thus able to provide a unique experimental model for human genetical liver disease, Wilson's disease. Hepatocyte growth factor (HGF) is a potent hepatotrophic factor in liver regeneration, and its expression is up-regulated in response to liver injuries. We found that the plasma HGF level in LEC rats rose markedly during the fulminant hepatitis phase, fell during the phase of chronic/cholangiofibrosis, and fluctuated during the hepatoma phase. Immunohistological staining of the liver revealed that the number of HGF-positive cells increased remarkably during the fulminant hepatitis phase, and that many of these cells were localized at the portal triads. Fewer HGF-positive cells were observed during the phase of chronic hepatitis. The surface of the hepatocellular carcinoma (HCC) cells and the cytoplasm of the nonepithelial cells in cancerous liver tissues were HGF-positive. The HGF-messenger RNA (mRNA) level in the liver rose in the fulminant hepatitis phase, fell in the chronic hepatitis phase, and was intermediate or high during the hepatoma phase. The expression of c-met mRNA was strong in the tissues of LEC rats with fulminant hepatitis and, especially, in the cholangiofibrosis tissues. c-met mRNA was also detected in HCCs. These results suggest that the HGF-c-met system may play an important role in the regeneration of hepatocytes as well as in the development of HCC in paracrine or autocrine mechanisms.

Original extrahilar approach for hepatic portal revascularization and relief of extrahepatic portal hypertension related to later portal vein thrombosis after pediatric liver transplantation. Long term results.

Original extrahilar approach for hepatic portal revascularization and relief of extrahepatic portal hypertension related to later portal vein thrombosis after pediatric liver transplantation. Long term results.