Sorafenib delays recurrence and metastasis after liver transplantation in a rat model of hepatocellular carcinoma with high expression of phosphorylated extracellular signal-regulated kinase

Authors

  • Jun Yan,

    Corresponding author
    1. Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, Shanghai, People's Republic of China
    2. Department of Surgery, Fujian Provincial Tumor Hospital, Teaching Hospital of Fujian Medical University, Fuzhou, People's Republic of China
    • Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Changjun Tan,

    Corresponding author
    1. Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, Shanghai, People's Republic of China
    • Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Fangming Gu,

    Corresponding author
    1. Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, Shanghai, People's Republic of China
    • Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Jiahao Jiang,

    1. Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
    2. Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, Shanghai, People's Republic of China
    Search for more papers by this author
  • Min Xu,

    1. Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
    2. Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, Shanghai, People's Republic of China
    Search for more papers by this author
  • Xiuzhen Huang,

    1. Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
    2. Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, Shanghai, People's Republic of China
    Search for more papers by this author
  • Zhi Dai,

    1. Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
    2. Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, Shanghai, People's Republic of China
    Search for more papers by this author
  • Zheng Wang,

    1. Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
    2. Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, Shanghai, People's Republic of China
    Search for more papers by this author
  • Jia Fan,

    1. Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
    2. Shanghai Key Laboratory of Organ Transplantation, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
    3. Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, Shanghai, People's Republic of China
    Search for more papers by this author
  • Jian Zhou

    1. Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
    2. Shanghai Key Laboratory of Organ Transplantation, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
    3. Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, Shanghai, People's Republic of China
    Search for more papers by this author

  • This work was supported by the National Key Sci-Tech Special Project of China (2012ZX10002-016), the National Natural Science Foundation of China (81272574 and 81172277), the Natural Science Foundation of Fujian Province (2010J01136), Innovation Research of Fujian Health Bureau (2012-CXB-7), and the 2010 Excellent Ph.D. Research Foundation of Fudan University.

Address reprint requests to Jia Fan, M.D., Ph.D., Liver Cancer Institute, Zhongshan Hospital, Fudan University, 180 Feng Lin Road, Shanghai, People's Republic of China 200032. Telephone: 086-21-64041990; FAX: 086-21-64037181; E-mail: jiafan99@yahoo.com

Address reprint requests to Jian Zhou, M.D., Ph.D., Liver Cancer Institute, Zhongshan Hospital, Fudan University, 180 Feng Lin Road, Shanghai, People's Republic of China 200032. Telephone: 086-21-64041990; FAX: 086-21-64037181; E-mail: zhou.jian@zs-hospital.sh.cn

Abstract

Liver transplantation (LT) is one of the curative treatments for hepatocellular carcinoma (HCC). However, cancer recurrence and metastasis after LT are common in some HCC patients with high-risk factors (even in those within the Milan criteria). It remains unclear whether adjuvant therapy with sorafenib inhibits HCC recurrence and metastasis after LT. Therefore, we performed orthotopic LT in an August Irish Copenhagen (ACI) rat model of HCC. Because LT involves immune rejection and tolerance and it is unknown whether sorafenib influences the immune response, we also investigated the effects of sorafenib on immune balance. In this study, we established an allogeneic rat LT model in which liver grafts were taken from Lewis rats and transplanted into ACI rats with orthotopic HCC, and they were administered cyclosporine A to prevent acute allograft rejection. From day 7 after LT, sorafenib was administrated at 30 mg/kg/day for 3 weeks. Our results showed that the serum levels of vascular endothelial growth factor and hepatocyte growth factor significantly increased after LT, and the T helper 1 (Th1)/T helper 2 (Th2) immune balance was shifted toward a Th2 response after immunosuppressant administration. In comparison with controls, the rats in the sorafenib group showed significantly inhibited extracellular signal-regulated kinase phosphorylation and improved progression-free survival and overall survival. The tumor proliferation rate and angiogenesis in posttransplant recurrent tumor tissues decreased in the sorafenib group, and the tumor apoptosis rate increased. There was no significant difference in the Th1/Th2 immune balance between the sorafenib and control groups. In conclusion, adjuvant therapy with sorafenib is highly effective at inhibiting cancer recurrence and metastasis without influencing the immune balance after LT for HCC with high expression of phosphorylated extracellular signal-regulated kinase. This study suggests that sorafenib may have potential, particularly as part of a stratified medicine approach to HCC treatment after LT. Liver Transpl 19:507–520, 2013. © 2013 AASLD.

Abbreviations
ALT

alanine aminotransferase

AST

aspartate aminotransferase

CSA

cyclosporine A

CT

computed tomography

ELISA

enzyme-linked immunosorbent assay

ERK

extracellular signal-regulated kinase

HCC

hepatocellular carcinoma

H&E

hematoxylin and eosin

HGF

hepatocyte growth factor

IFN-γ

interferon-γ

IL-4

interleukin-4

I/R

ischemia/reperfusion

LT

liver transplantation

pERK

phosphorylated extracellular signal-regulated kinase

PET

positron emission tomography

Th1

T helper 1

Th2

T helper 2

TNM

tumor-node-metastasis

TUNEL

terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling

VEGF

vascular endothelial growth factor.

Currently, liver cancer is the fifth most frequently diagnosed cancer worldwide and the second most frequent cause of cancer-related death in men. In women, it is the seventh most commonly diagnosed cancer and the sixth leading cause of cancer-related death.[1] Hepatocellular carcinoma (HCC), the most common type of primary liver cancer, is a highly malignant tumor characterized by rapid progression, a poor prognosis, and frequent tumor recurrence and metastasis. Liver transplantation (LT) is one of the curative treatments for HCC.[2-4] However, cancer recurrence and metastasis after LT are common in some HCC patients with high-risk factors such as microvascular invasion, poor differentiation, allelic imbalance in microsatellites, and genetic variations of single-nucleotide polymorphisms (even in those within the Milan criteria, ie, a single tumor ≤ 5 cm or 3 or fewer tumors with none ≥ 3 cm without major vessel invasion or metastasis).[5-9] To date, no effective systemic treatment is available for inhibiting cancer recurrence and metastasis after LT.[10-13] Sorafenib, a multikinase inhibitor that blocks the vascular endothelial growth factor (VEGF) and proto-oncogene RAF/mitogen-activated protein kinase kinase (MEK)/extracellular signal-regulated kinase (ERK) pathways, has been approved for the treatment of advanced HCC by the US Food and Drug Administration since 2007, and this represents a milestone in HCC treatment.[14-16] Thus far, it remains unclear whether sorafenib inhibits cancer recurrence and metastasis after LT in HCC patients with high-risk factors. Therefore, we performed a study to assess the efficacy of sorafenib as an adjuvant therapy after orthotopic LT in an ACI rat model of HCC, which mimics LT for HCC patients and presents with posttransplant recurrence and metastasis at a rate of 100%. Because LT involves immune rejection and tolerance and it is unknown whether sorafenib has an influence on the immune response, we also investigated the effects of sorafenib on the immune balance.

At our institute, a previous in vitro study[17] found that basal phosphorylated extracellular signal-regulated kinase (pERK) levels increased stepwise in cell lines in accordance with their metastatic potential, and the effects of sorafenib on cell proliferation were significantly correlated with basal pERK levels. This previous in vitro study concluded that pERK was a potential predictor of sensitivity to sorafenib in the treatment of HCC. Therefore, in our present in vivo study, we tested the expression of pERK in our model and the changes in pERK before and after sorafenib treatment. Moreover, we also detected the expression of pERK in clinical HCC tissues by immunohistochemistry via tissue microarrays containing samples from a cohort of 323 HCC patients.

MATERIALS AND METHODS

Study Design and Sample Size Determination

There were 3 steps to this study. First, we established an allogeneic rat LT model in which liver grafts were taken from Lewis (haplotype of major histocompatibility complex) rats and transplanted into ACI (haplotype of major histocompatibility complex) rats with orthotopic HCC, and they were administered cyclosporine A (CSA) to prevent acute allograft rejection. Second, we assessed whether adjuvant therapy with sorafenib inhibited HCC recurrence and metastasis after allogeneic LT and whether overall survival was improved. Finally, if overall survival was improved, we planned to further explore the underlying mechanisms of this improvement by investigating whether sorafenib inhibited ERK phosphorylation, tumor proliferation, invasion, and metastasis under the conditions of ischemia/reperfusion (I/R) injury and immunosuppression after allogeneic LT. We also investigated whether sorafenib influenced the immune balance after allogeneic LT. The animal care and use committee of Fudan University and the institutional review board of Zhongshan Hospital approved this study.

In step 2, the ACI rats were randomized into sorafenib and control groups by body weight after allogeneic LT for HCC. Both groups received subcutaneous CSA injections at 3 mg/kg/day for 2 weeks after allogeneic LT to prevent liver rejection. The sorafenib group underwent sorafenib oral administration once daily at a dose level of 30 mg/kg of body weight for 3 weeks from day 7 after allogeneic LT, whereas the control group received oral administration of the solvent. HCC recurrence and metastasis were detected by positron emission tomography (PET)/computed tomography (CT). The endpoint was rat death, and the primary outcome was overall survival. The sample size was calculated as follows: the overall survival time from LT to death in the control group was estimated to be 34.00 ± 8.90 days on the basis of previous experience. We hypothesized that adjuvant therapy with sorafenib could improve the survival time by 50%, and then each group of 6 rats was determined. With the number of rats used, the study had 90% power to detect a 50% increase in survival time after LT with sorafenib treatment (2-sided type I error = 0.05).

In step 3, the same experiment was repeated to explore the underlying mechanisms, but the endpoint was changed from death to day 28 after LT. The rats were sacrificed, and blood and specimens, including livers and lungs, were harvested on day 28 after allogeneic LT. Tumor proliferation was determined with Ki-67 immunohistochemical staining, and tumor angiogenesis was determined with CD31 immunohistochemical staining. Tumor apoptosis was determined with the terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling (TUNEL) method. pERK was analyzed with immunohistochemistry and western blotting. Liver function, as determined by the serum activities of alanine aminotransferase (ALT) and aspartate aminotransferase (AST), was examined with a biochemical analysis. Serum levels of VEGF, hepatocyte growth factor (HGF), interferon-γ (IFN-γ), and interleukin-4 (IL-4) were measured with an enzyme-linked immunosorbent assay (ELISA). The T helper 1 (Th1)/T helper 2 (Th2) immune balance was determined with the IFN-γ/IL-4 ratio.

Experimental Animals and Drugs

Inbred male ACI (RT1av1) rats (Harlan, Indianapolis, IN) and Lewis (RT1l) rats (Vitalriver, Beijing, China), each 12 to 16 weeks old and weighing 250 to 280 g, were maintained in laminar flow cages in a specific pathogen-free animal facility with a standard diet and water. All procedures involving rat experiments were performed according to the recommendations of Guide for the Care and Use of Laboratory Animals (National Institutes of Health). CSA (Sandimmune, Novartis, NJ) was administered via subcutaneous injections at 3 mg/kg/day for 14 days to prevent liver rejection after allogeneic LT in accordance with a previous study.[18] Sorafenib was purchased from Bayer Pharmaceutical Corp. and was formulated as previously described.[19] Sorafenib was administered orally once daily at a dose level of 30 mg/kg of body weight for 3 weeks in accordance with a previous study.[14]

Allogeneic Rat LT Model of Acute Allograft Rejection in Which Grafts From Lewis Rats Were Transplanted Into ACI Rats With Orthotopic HCC

To imitate clinical LT with acute rejection, we established an allogeneic rat LT model of acute allograft rejection in which grafts from Lewis rats were transplanted into ACI rats with orthotopic HCC. A Morris hepatoma line (MH-3924A, Institute for Applied Cell Culture, Munich, Germany) was maintained by intrahepatic passage in male ACI rats.[20] Intrahepatic tumor implantation with Morris hepatoma was performed as described in a previous study.[18] The implanted tumor grew to approximately 1 cm without metastasis (Supporting Fig. 1A-C) by day 14, and all rats with Morris hepatoma in the liver exhibited 100% transplantability and metastatic ability. Allogeneic orthotopic LT from Lewis rats to ACI rats was performed on day 14 after intrahepatic tumor implantation with arterial reconstruction, as previously described.[21, 22] There was no metastasis in the ACI rats before transplantation. The recipient rats in both the sorafenib and control groups received subcutaneous CSA injections at 3 mg/kg/day for 14 days to prevent rejection after LT. Without any posttransplant immunosuppressant agents, rats died within 2 weeks after LT because of rejection. This allogeneic LT model of acute allograft rejection in ACI rats that received grafts from Lewis rats is shown in Fig. 1. In this study, the CSA levels were the same in the 2 groups. The rats were followed until death or were sacrificed on day 28 after LT according to the study design.

Figure 1.

Allogeneic rat LT model of acute allograft rejection in which a graft from a Lewis rat is transplanted into an ACI rat with orthotopic HCC. (A) A donor Lewis rat is shown. (B) The donor liver is procured. (C) The donor liver is excised and stored in an ice-cold histidine-tryptophan-ketoglutarate preservative solution. (D) A recipient ACI rat with orthotopic Morris hepatoma is shown. (E) The recipient liver is removed (the white arrow shows the Morris hepatoma). (F) An anastomosis is hand-sewn for the suprahepatic vena cava. (G) Portal vein and intrahepatic vena cava anastomoses (black and white arrows, respectively) with 2 cuffs are shown. (H) The recipient ACI rat is awake after LT. (I) H&E staining shows acute rejection, which is characterized by portal inflammation, bile duct damage, and venous endothelial inflammation on day 7 after allogeneic LT without any posttransplant immunosuppressant agents. (J) H&E staining shows no acute rejection on day 7 after allogeneic LT with CSA administration. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com

PET/CT Imaging

A high-resolution, small-animal PET/CT scan (Inveon micro-PET/CT, Siemens Preclinical Solution, Knoxville, TN) was used to detect HCC recurrence and metastasis each week after LT. Each rat was anesthetized with 5% isoflurane in an induction chamber and injected intravenously with 0.2 to 0.3 mCi of 2-deoxy-2-[F-18]fluoro-d-glucose. During image acquisition, the rats were anesthetized with 1% to 2% isoflurane gas delivered through a custom face mask. All micro-PET images were reconstructed with the standard ordered-subset expectation maximization method.

Liver Function Examination

Blood samples were collected on day 28 after LT. Serum was obtained by centrifugation. ALT and AST levels were determined with a biochemical analysis.

ELISA

ELISA was performed to measure the serum levels of VEGF, HGF, IFN-γ, and IL-4 according to the manufacturer's guide. Blood samples were taken from each rat in the anesthesia induction period before surgery and on postoperative days 7, 14, 21, and 28. All samples were assayed in duplicate with commercially available ELISA kits for VEGF, IFN-γ, and IL-4 (Quantikine kit, R&D Systems, Abingdon, United Kingdom) and HGF (Rapid Bio Lab, Calabasas, CA).

Pathological Examination

After the livers and the lungs were harvested, the specimens were processed with routine histopathological procedures, which included 10% buffered formalin processing, paraffin embedding, sectioning at 5 μm, and hematoxylin and eosin (H&E) staining.

Immunohistochemical Staining and TUNEL Staining

Immunohistochemical staining and TUNEL staining were conducted and analyzed according to a previous study.[14] A rabbit polyclonal anti–Ki-67 antibody (Abcam, Cambridge, MA), a monoclonal anti-CD31 antibody (Abcam), and a rabbit monoclonal anti–phospho-ERK1/2 antibody (phospho-p44/42 mitogen-activated protein kinase, Thr202/Tyr204, Cell Signaling Technology, Beverly, MA) were used for immunohistochemical staining according to the manufacturer's protocol. The ApopTag Plus peroxidase in situ apoptosis detection kit (Oncor, Gaithersburg, MD) was used for TUNEL staining according to the manufacturer's protocol.

Western Blot Analysis

Western blot analysis was performed as previously described.[14] Briefly, total proteins were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and this was followed by a transfer to polyvinylidene difluoride membranes. The membranes were washed, blocked, and incubated with specific primary antibodies against pERK1/2 (1:1000), ERK1/2 (1:1000), or β-actin (1:2000), and this was followed by incubation with horseradish peroxidase–conjugated secondary antibodies. Proteins were detected with an enhanced chemiluminescence assay (Pierce-Thermo Scientific).

Tissue Microarrays and Immunohistochemistry

Tissue microarrays were constructed by Shanghai Biochip Co., Ltd., as previously described.[23] Three tissue microarrays containing samples from a cohort of 323 HCC patients were used in this study. pERK was detected by immunohistochemistry. A rabbit monoclonal anti–phospho-ERK1/2 antibody (phospho-p44/42 mitogen-activated protein kinase, Thr202/Tyr204, Cell Signaling Technology) was used for immunohistochemical staining according to the manufacturer's protocol.

Statistical Analysis

The Student t test (means and standard deviations) and the Mann-Whitney test (medians and ranges) were used for continuous data to compare the sorafenib and control groups. Comparisons between groups with respect to categorical data were analyzed with the chi-square test. Kaplan-Meier curves for overall survival and progression-free survival were compared with a log-rank test. All data were analyzed with SPSS 16.0 (SPSS, Inc., Chicago, IL), and significance for these differences was defined as P < 0.05.

RESULTS

Adjuvant Therapy With Sorafenib Significantly Inhibited HCC Recurrence and Metastasis After LT and Improved Progression-Free Survival and Overall Survival

After allogeneic LT, small-animal PET/CT was used to detect HCC recurrence and metastasis every week. In the first week after LT, no recurrence or metastasis was detected by PET/CT in either group. In the second week after LT, PET/CT detected recurrence and metastasis in 33.3% of the rats (2/6) in the control group, but no rat (0/6) in the sorafenib group had recurrence or metastasis (P = 0.12). During the 2 weeks after LT, there was no significant difference in body weight between the sorafenib and control groups. However, in the third week after LT, the first notable differences were observed in HCC recurrence and metastasis between the sorafenib and control groups. PET/CT showed that 100% of the rats (6/6) in the control group had obvious intrahepatic recurrence, and 83% of the control rats (5/6) had lung metastases (Fig. 2A), whereas no rats in the sorafenib group had detectable intrahepatic recurrence or lung metastases (Fig. 2B). There was a significant difference between the 2 groups in intrahepatic recurrence (P = 0.001) and lung metastases (P = 0.003). With disease progression in the fourth week after LT, rats in the control group became cachectic and were characterized by weight loss, anorexia, ascites, and abdominal distension. PET/CT showed that all rats in the control group had larger intrahepatic recurrent tumors with widespread extrahepatic metastases (Fig. 3A). In contrast, the rats in the sorafenib group appeared stable during this period, and their body weight continued to increase, although small intrahepatic recurrent lesions without extrahepatic metastasis were detected by PET/CT (Fig. 3B). The differences between the 2 groups became greater over the following 2 weeks. In the control group, the rats began to die, with no rat surviving to day 48 after LT, whereas in the sorafenib group, all rats remained stable, and there were no rat deaths at that time (P = 0.001). From the seventh week to the ninth week after LT, the rats in the sorafenib group gradually became cachectic and began to die. The changes in body weight are shown in Fig. 4A. There was a significant difference in body weight between the 2 groups by the third week after LT. Progression-free survival from LT to HCC recurrence and/or metastasis is shown in Fig. 4B. The progression-free survival times for the control and sorafenib groups were 18.67 ± 3.61 and 29.17 ± 2.86 days, respectively, and the difference was significant (P = 0.001). Overall survival from LT to death is shown in Fig. 4C. The overall survival times for the control and sorafenib groups were 35.50 ± 6.60 and 59.67 ± 7.00 days, respectively; this difference was significant (P = 0.001). The median survival times for the control and sorafenib groups were 32 days (95% confidence interval = 26-38 days) and 58 days (95% confidence interval = 52-64 days), respectively. Autopsies confirmed that all of the rats in both groups died of HCC recurrence and metastasis.

Figure 2.

PET/CT showed significant differences in HCC recurrence and metastasis between (A) the control group and (B) the sorafenib group in the third week after LT. (A) PET/CT showed that 100% of the rats (6/6) in the control group had obvious intrahepatic recurrence (white arrows), and 83% (5/6) had lung metastases (red arrows). (B) No rat in the sorafenib group had detectable intrahepatic recurrence or lung metastases.

Figure 3.

PET/CT detection in the fourth week after LT. (A) PET/CT showed that all rats in the control group had obvious recurrent tumors (white arrows) with lung metastases (red arrows). (B) Small intrahepatic recurrent lesions (white arrows) without extrahepatic metastases in the sorafenib group were detected by PET/CT.

Figure 4.

Sorafenib delayed disease progression and improved survival. (A) By the third week after LT, there was a significant difference in body weight between the control group and the sorafenib group. (B) Adjuvant therapy with sorafenib significantly improved progression-free survival in comparison with the control group (29.17 ± 2.86 versus 18.67 ± 3.61 days, P = 0.001) after LT. (C) Adjuvant therapy with sorafenib significantly improved overall survival in comparison with the control group (59.67 ± 7.00 versus 35.50 ± 6.60 days, P = 0.001) after LT.

In short, adjuvant therapy with sorafenib after LT significantly inhibited HCC recurrence and metastasis and improved progression-free survival and overall survival.

Adjuvant Therapy With Sorafenib After LT Significantly Suppressed Tumor Proliferation and Tumor Angiogenesis, Induced Tumor Apoptosis, and Inhibited ERK Phosphorylation

In step 3, the same experiment was repeated to explore the underlying mechanisms of the effects of sorafenib, but the endpoint was changed from rat death to day 28 after LT. Blood samples were taken from rats on days 0, 7, 14, 21, and 28 after LT. The dynamics of the serum levels of VEGF and HGF are shown in Fig. 5A,B. The serum levels of both VEGF and HGF increased dramatically, reached a peak on day 7 after LT, and decreased gradually thereafter. There were significant differences in the serum levels of VEGF and HGF before and after the operation (P < 0.001). The Th1/Th2 immune balance, marked by the IFN-γ/IL-4 ratio, shifted toward Th2, and this indicated immunosuppression after CSA administration (Fig. 5C). The IFN-γ/IL-4 ratio decreased from 1.76 ± 0.24 to 1.25 ± 0.19 in the control group (P = 0.002) and from 1.80 ± 0.26 to 1.17 ± 0.23 in the sorafenib group (P = 0.001) after the administration of CSA to prevent liver rejection. Rat liver function, as determined by the serum activities of ALT and AST, is shown in Fig. 5D. There was no significant difference in liver function between the sorafenib and control groups on day 28 after LT (for ALT, 179.17 ± 68.84 versus 111.67 ± 60.16 IU/L, P = 0.10; for AST, 364.17 ± 125.40 versus 248.50 ± 131.59 IU/L, P = 0.15). The rats were sacrificed, and specimens were harvested on day 28 after LT. Intrahepatic recurrent tumors in the control group were much larger than those in the sorafenib group (Fig. 6A). There were significant differences between the sorafenib group and the control group with respect to the tumor volume (1.96 ± 0.45 versus 30.27 ± 6.90 cm2, P < 0.001) and the tumor weight (2.30 ± 0.37 versus 25.83 ± 3.19 g, P < 0.001). All rats in the control group had lung metastases, whereas no rat in the sorafenib group had lung metastases (Fig. 6A). The median number of lung metastases was 7 (range = 3-11) in the control group. H&E staining confirmed the presence of lung metastases in the control group and the absence of lung metastases in the sorafenib group (Fig. 6B). Immunohistochemical staining demonstrated that Ki-67–positive tumor cells and CD31-positive microvessels were significantly decreased in the sorafenib group versus the control group (Fig. 7A,B). The average Ki-67–positive rate was 68.67% ± 7.29% in the control group and 14.50% ± 3.83% in the sorafenib group (P < 0.001). The microvessel density identified by CD31 was 189.33 ± 19.58 vessels/mm2 in the control group and 46.67 ± 8.16 vessels/mm2 in the sorafenib group (P < 0.001). TUNEL staining showed more tumor cell apoptosis in the sorafenib group versus the control group (Fig. 7C). The average tumor apoptosis rate was 12.40% ± 3.65% in the sorafenib group and 4.80% ± 2.59% in the control group (P = 0.005). More importantly, sorafenib treatment significantly inhibited ERK phosphorylation. Immunohistochemical staining showed that there was a significant difference between the control group and the sorafenib group with respect to the pERK-positive rate (47.33% ± 9.09% versus 15.32% ± 6.92%, P < 0.001; Fig. 8A-C). Western blot analysis confirmed that sorafenib treatment significantly inhibited ERK phosphorylation. As Fig. 8D shows, total ERK was unchanged, and sorafenib decreased pERK expression. Collectively, adjuvant therapy with sorafenib after LT significantly suppressed tumor proliferation and tumor angiogenesis, induced tumor apoptosis, and inhibited ERK phosphorylation.

Figure 5.

Increases in growth factors and decreased recipient immunity after allogeneic LT. (A) The serum levels of VEGF increased dramatically, reached a peak on day 7 after LT, and decreased gradually thereafter. There was a significant difference between preoperative and postoperative levels (P < 0.001). (B) The dynamics of HGF serum levels were similar to those for VEGF. (C) The Th1/Th2 immune balance, marked by the IFN-γ/IL-4 ratio, shifted toward Th2 after immunosuppression. The IFN-γ/IL-4 ratio decreased from 1.76 ± 0.24 to 1.25 ± 0.19 in the control group (P = 0.002) and from 1.80 ± 0.26 to 1.17 ± 0.23 in the sorafenib group (P = 0.001) after CSA administration. Sorafenib did not influence the immune balance. There were no significant differences between the sorafenib and control groups on days 0 and 28 after LT. (D) There was no significant difference between the sorafenib group and control group in liver function on day 28 after LT (for ALT, 179.17 ± 68.84 versus 111.67 ± 60.16 IU/L, P = 0.10; for AST, 364.17 ± 125.40 versus 248.50 ± 131.59 IU/L, P = 0.15).

Figure 6.

Comparison of liver recurrence and lung metastasis in the sorafenib and control groups on day 28 after LT. (A) The recurrent tumors in the control group were much larger than the tumors in the sorafenib group. There were lung metastases (white arrows) in the control group, but there were no lung metastases in the sorafenib group. (B) H&E staining confirmed the presence of lung metastases (white arrows) in the control group and the absence of lung metastases in the sorafenib group.

Figure 7.

Adjuvant therapy with sorafenib after LT significantly suppressed tumor proliferation, inhibited tumor angiogenesis, and induced tumor apoptosis. (A) Immunohistochemical staining demonstrated that Ki-67–positive tumor cells were significantly decreased in the sorafenib group versus the control group. (B) Immunohistochemical staining showed that the density of CD31-positive microvessels was significantly lower in the sorafenib group versus the control group. (C) TUNEL staining showed more tumor cell apoptosis in the sorafenib group versus the control group.

Figure 8.

Sorafenib treatment significantly inhibited ERK phosphorylation. (A) Immunohistochemical staining showed that pERK expression was strongly positive in the control group. (B) Immunohistochemical staining showed that pERK expression significantly decreased in the sorafenib group. (C) There was a significant difference between the control group and the sorafenib group with respect to the pERK-positive rate (47.33% ± 9.09% versus 15.32% ± 6.92%, P < 0.001). (D) Western blot analysis showed that total ERK was unchanged, and sorafenib decreased pERK expression.

pERK Expression in Tissue Microarrays

Three tissue microarrays containing samples from 323 Chinese HCC patients were used in this study. These patients, who did not have distant metastases or any prior treatment, underwent curative resection for primary tumors consecutively between January 2003 and March 2004 at the Liver Cancer Institute (Zhongshan Hospital, Fudan University). Their preoperative liver function was defined as Child-Pugh A. The diagnosis of HCC was pathologically proven. The tumor stage was determined according to the International Union Against Cancer tumor-node-metastasis (TNM) classification system. Tumor differentiation was graded with the Edmondson grading system. Detailed clinicopathological characteristics are listed in Table 1. The expression of pERK in the tissue microarrays is shown in Fig. 9. The rate of positive pERK expression was 57.89% (187/323), and the rate of negative pERK expression was 42.11% (136/323).

Table 1. Clinicopathological Characteristics of a Tissue Microarray Containing Samples From 323 Chinese Patients With HCC
Variablen (%)
Sex 
Male277 (85.8)
Female46 (14.2)
Age 
≤50 years166 (51.4)
>50 years157 (48.6)
Hepatitis B surface antigen 
Negative45 (13.9)
Positive278 (86.1)
Liver cirrhosis 
Absent36 (11.1)
Present287 (88.9)
Alpha-fetoprotein 
≤20 μg/L96 (29.7)
>20 μg/L227 (70.3)
Tumors 
1276 (85.4)
≥247 (14.6)
Tumor size 
≤5 cm164 (50.8)
>5 cm159 (49.2)
Tumor encapsulation 
Complete164 (50.8)
None159 (49.2)
Venous invasion 
Absent175 (54.2)
Present148 (45.8)
Edmondson grade 
I or II255 (78.9)
III or IV68 (21.1)
TNM stage 
I154 (47.7)
II or III169 (52.3)
pERK expression 
Positive187 (57.89)
Negative136 (42.11)
Figure 9.

pERK expression in tissue microarrays. (A) Three tissue microarrays containing samples from a cohort of 323 HCC patients were used in this study. pERK was detected by immunohistochemistry. (B) The rate of positive pERK expression in the tissue microarrays was 57.89% (187/323), and the rate of negative pERK expression was 42.11% (136/323). (C) Positive pERK expression is shown (×100). (D) Negative pERK expression is shown (×100).

DISCUSSION

In this study, to imitate clinical LT for HCC patients with acute rejection and posttransplant HCC recurrence and metastasis, we established an allogeneic rat LT model in which liver grafts were taken from Lewis rats and transplanted into ACI rats with orthotopic HCC, and we administrated CSA to prevent acute allograft rejection. In a previous study by Freise et al.,[18] ACI-to-ACI syngeneic LT for orthotopic HCC was performed to assess the role of CSA in tumor recurrence, but a lack of allograft rejection impaired the validity of the LT model. Likewise, Matsuzaki et al.[24] used a rat model of HCC induced by diethylnitrosamine to assess the effects of immunosuppression with tacrolimus or CSA on survival after isogeneic Fischer F344 Inbred Rat (F344)-to-F344 LT, and survival was significantly enhanced in the absence of immunosuppression versus the presence of tacrolimus or CSA. The drawbacks of this model without rejection led to the modification reported by Ceriello et al.[25], who used the allogeneic combination of Brown Norway (BN) rats and Lewis rats with HCC induced by diethylnitrosamine. However, liver allografts from BN rats are accepted in Lewis rats without immunosuppression, as reported by other authors.[26] To address the limitations associated with a lack of alloimmunity in these models, we transplanted graft livers from Lewis rats into ACI rats with orthotopic HCC. Without any posttransplant immunosuppressant agents, the rats died within 2 weeks after LT because of rejection. According to multidose pilot data, the combination of these allogeneic strains required immunosuppression (CSA at 3 mg/kg/day for 14 days) to prevent allograft rejection. Moreover, we performed arterial reconstruction in our LT model because Morris hepatoma is a hypervascular tumor and has angiographic features similar to those of human HCC with an increase in the tumoral arterial blood flow.[27, 28] The findings of previous studies performed with nonarterialized grafts do not reflect the clinical setting of orthotopic HCC because hepatic artery reconstruction contributes to intrahepatic recurrence after LT for HCC.[18, 24, 25] As shown in Figs. 2A and 6A, 100% intrahepatic recurrence after LT was observed in our study. To the best of our knowledge, the allogeneic rat LT model established in this study—a liver graft from a Lewis rat transplanted into an ACI rat with orthotopic HCC—is the first LT rat model to mimic LT in HCC patients with acute rejection and present with posttransplant intrahepatic recurrence and extrahepatic metastasis at a rate of 100%.

HCC recurrence and metastasis after LT are common in some patients with high-risk factors. According to our study, there are 2 factors accelerating HCC recurrence and metastasis after LT. One factor is the release of growth factors, and the other is immunosuppression. Our data demonstrated that VEGF and HGF significantly increased after LT. Previous studies have suggested that the removal of the primary tumor could be associated with enhanced tumor growth of residual micrometastases.[29, 30] Recent clinical data have confirmed that VEGF and HGF are released in response to liver resection, and these 2 growth factors are necessary for wound healing but promote tumor growth.[31] As Tamagawa et al.[32] reported, hepatic I/R significantly increased VEGF and cancer growth in rats. They found that I/R-induced VEGF expression could enhance the growth of microscopic tumors via VEGF receptors on tumor cells and thus promote cancer metastasis. In addition to the release of growth factors, immunosuppression also plays an important role in determining HCC recurrence and metastasis after LT. Freise et al.[18] evaluated the effect of immunosuppression on tumor recurrence after LT in a rat model of HCC. They found that the overall survival rate was significantly reduced after treatment with the immunosuppressant CSA, and the incidence of pulmonary nodules and extrapulmonary sites was significantly increased. Results similar to those of animal models have also been found in the clinical setting.[33] Our study used CSA to prevent rejection in allogeneic LT. The Th1/Th2 immune balance, determined by the IFN-γ/IL-4 ratio, shifted toward Th2. The increased growth factors and decreased recipient immunity contributed to the posttransplant tumor proliferation, angiogenesis, invasion, and metastasis.

To date, no effective systemic treatment has been reported to inhibit HCC recurrence and metastasis after LT. Chemotherapy and chemoembolization remain disappointing as adjuvant or neoadjuvant therapies for improving overall survival and progression-free survival.[10-13] Sorafenib, an oral, multitargeted, tyrosine kinase inhibitor of VEGF receptors, RAF kinase, and platelet-derived growth factor receptors, has been shown to improve survival for patients with advanced HCC in 2 randomized phase 3 studies.[15, 16] Because of the survival benefit in patients treated with sorafenib for advanced HCC and the lack of established treatments to prevent or inhibit posttransplant HCC recurrence and/or metastasis, we assessed whether sorafenib could be useful as an adjuvant therapy via an orthotopic rat model of HCC. Our study found that sorafenib adjuvant therapy after LT significantly inhibited HCC recurrence and metastasis and improved overall survival. Overall survival was improved more than 50% by sorafenib treatment after LT. At our institute, a previous in vitro study[17] found that basal pERK levels increased stepwise in cell lines in accordance with their metastatic potential, and the effects of sorafenib on cell proliferation were significantly correlated with basal pERK levels. This previous in vitro study concluded that pERK was a potential predictor of sensitivity to sorafenib in the treatment of HCC. In our present in vivo study, we used a highly metastatic HCC model in which the pERK test was strongly positive, and we found that adjuvant therapy with sorafenib was very effective at delaying this highly metastatic HCC recurrence and metastasis after LT by inhibiting ERK phosphorylation. In comparison with controls, rats in the sorafenib treatment group had significantly inhibited ERK phosphorylation, as immunohistochemistry and western blotting showed. In short, the underlying mechanisms of sorafenib inhibition of HCC recurrence and metastasis after LT are as follows. First, hepatic I/R and wound healing promote the release of growth factors such as VEGF and HGF, which stimulate the growth of residual micrometastases. Sorafenib blocks the VEGF and RAF/MEK/ERK pathways and, therefore, inhibits residual tumor angiogenesis and proliferation. Second, immunosuppression after LT accelerates tumor invasion and metastasis. Sorafenib induces the apoptosis of residual micrometastases and recurrent cancer, inhibits tumor growth, and reduces the tumor burden. Altogether, adjuvant therapy with sorafenib provides an early intervention for inhibiting tumor proliferation, angiogenesis, invasion, and metastasis after LT (Fig. 10).

Figure 10.

Mechanisms of sorafenib inhibition of HCC recurrence and metastasis after LT. Hepatic I/R and wound healing promote the release of growth factors such as VEGF and HGF, which stimulate the growth of residual micrometastases. Sorafenib treatment blocks the VEGF and RAF/MEK/ERK pathways and thereby inhibits residual tumor angiogenesis and proliferation. Meanwhile, immunosuppression after LT accelerates tumor invasion and metastasis. Sorafenib treatment induces residual micrometastases and recurrent cancer apoptosis and, therefore, reduces the tumor burden and inhibits tumor growth. Collectively, adjuvant therapy with sorafenib after LT provides an early intervention for inhibiting tumor proliferation, angiogenesis, invasion, and metastasis. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com

Because LT is involved in immune rejection and tolerance and it is unknown whether sorafenib influences the immune response, we also investigated the effects of sorafenib on the immune balance in this study. The results showed that the Th1/Th2 immune balance shifted toward Th2 after immunosuppressant administration, and sorafenib did not influence the immune balance after LT. Sorafenib is Food and Drug Administration–approved only for metastatic or unresectable HCC; thus, potential applications in the adjuvant setting are limited to patients with high-risk HCC undergoing curative resection (see the ongoing Sorafenib as Adjuvant Treatment in the Prevention of Recurrence of Hepatocellular Carcinoma trial).[34] Our study has demonstrated that sorafenib is highly effective at inhibiting HCC recurrence and metastasis after LT. Moreover, sorafenib does not influence the immune balance after LT.

As a matter of fact, HCCs constitute a very heterogeneous population of tumors, and responses to sorafenib vary widely between subjects in clinics. Our work has focused on adjuvant therapy with sorafenib after LT for HCC. In terms of the inhibition of recurrence and tumor growth, there are 2 reasons that sorafenib is very impressive here and that the results are quite different from the clinical experience with advanced HCC, in which the impact of sorafenib has been certain but at the same time not as impressive in comparison with this study. First, our model is a highly metastatic HCC model in which pERK expression is strongly positive, and adjuvant therapy with sorafenib is very effective at inhibiting ERK phosphorylation. Second, unlike the clinical situation in which sorafenib is usually used in patients with advanced HCC and most patients have lost the opportunity to undergo surgery, adjuvant therapy with sorafenib after LT in our study is an early intervention for inhibiting tumor recurrence and metastasis. Once again, because the pERK protein is known to be a key downstream component of the RAF/MEK/ERK pathway and both the previous in vitro study and this in vivo study have shown that the RAF/MEK/ERK signaling pathway is essential for sorafenib-mediated growth inhibition and that the sensitivity to sorafenib is directly related to the activation of this pathway and the basal pERK level, we believe that pERK may be a potential predictor of sensitivity to sorafenib when it is used as an adjuvant therapy to treat high-risk HCC after LT. Because an ACI rat LT model bearing HCC with low expression of pERK is not available yet, we could not repeat the experiment in a model that did not have overactive expression of pERK as a comparative group. This is the limitation of this study. As a substitute, we tested the expression of pERK in clinical HCC tissues. Our results from the tissue microarrays showed that 57.89% of 323 HCC patients had positive pERK expression, and 42.11% of the patients had negative pERK expression. Our animal study has shown that sorafenib is highly effective at inhibiting the recurrence and metastasis of HCC with high expression of pERK. Because more than 50% of HCC patients highly express pERK in the clinic, these patients might benefit from sorafenib treatment for recurrence and metastasis after surgery. Further investigation is needed to confirm this in a clinical trial.

In conclusion, we have established the first rat LT model (from a Lewis rat to an ACI rat) that mimics orthotopic LT in HCC patients and presents with both graft rejection and posttransplant recurrence and metastasis at a rate of 100%. Adjuvant therapy with sorafenib is highly effective at inhibiting the recurrence and metastasis of HCC with high expression of pERK, and it does not influence the immune balance after LT. This study suggests that sorafenib may have potential, particularly as part of a stratified medicine approach to HCC treatment after LT.

Ancillary