- Email: [email protected]
Management of Hepatocellular Carcinoma Recurrence After Liver Transplantation
Management of Hepatocellular Carcinoma Recurrence After Liver Transplantation A. Valdivieso, J. Bustamante, M. Gastaca, J.G. Uriarte, A. Ventoso, P. Ruiz, J.R. Fernandez, I. Pijoan, M. Testillano, M.J. Suarez, M. Montejo, and J. Ortiz de Urbina ABSTRACT Management of patients with hepatocellular carcinoma (HCC) recurrence after liver transplantation (OLT) is not well established. We conducted a retrospective analysis of our results in the treatment of HCC recurrence after OLT Patients. The 23 HCC recurrences developed after 182 OLT performed for HCC within Milan criteria, had an average follow-up of 60 months. Results. The median time to recurrence was 23.4 months. Surgical resection of the recurrence was possible in 11 patients, but an R-0 resection was obtained in 8 patients. Four of these 8 patients developed another recurrence, with 3 succumbing due to tumor recurrence and 1 alive at 12 months with recurrence. The other 4 patients without recurrences, include 3 who are alive at 19, 31, and 86 months and 1 who died at 32.6 months due to hepatitis C recurrence. The 3 patients with palliative resections developed recurrences. Twelve patients were rejected for surgery: 8 were treated symptomatically, 2 with systemic chemotherapy, and 2 with everolimus and sorafenib. This last treatment was also prescribed for 2 patients after R-0 surgery who are alive at 19 and 31 months and for 1 patient after R-1 surgery who is alive at 19 months. Of 15 patients who died, 13 succumbed to HCC recurrence. The average survival from transplantation was 61.7 ⫾ 37.5 and 48 ⫾ 34.3 months for patients without and with recurrence, respectively (P ⬍ .001). The survival from the recurrence was significantly higher among patients with R-0 surgery: 32.3 ⫾ 21.5 versus 11.9 ⫾ 6.9 months (P ⫽ .006). Conclusions. HCC recurrence after OLT of patients within Milan criteria was low but had a great impact on survival. Few cases are amenable to R-0 resection, but when possible it was associated with a significantly increased survival, although with an high incidence of a new recurrence. There is a rationale for the use of sorafenib and mammalian target of rapamycin based immunosuppression, which warrants randomized studies. RTHOTOPIC liver transplantation (OLT) is the standard of care for many patients with early hepatocellular carcinoma (HCC). However, even when using the restrictive Milan criteria, the incidence of HCC recurrence ranges from 8% to 20%.1 The natural history and prognosis for posttransplant HCC recurrence are insufficiently known, but this complication has a significant impact on outcome; the median survival after this diagnosis is ⬍12 months.1,2 The optimal management of HCC recurrence is not established and the therapeutic options are limited. Most patients present with advanced disease not amenable to potentially curative treatments. A significant number of cases can be considered for resection of hepatic or extrahepatic recurrences. Few studies
O
0041-1345/10/$–see front matter doi:10.1016/j.transproceed.2010.02.014 660
have reported specifically the prognosis and treatment of posttransplant recurrent HCC, but the published data suggest that selected surgically treated patients seem to show a better prognosis.2– 4 Unfortunately, the use of tailored immunosupFrom the Hepatobiliary and Liver Transplantation Unit (A.Va., M.G., J.G.U., A.Ve., P.R., J.O.d.U.), the Digestive Service (J.B., J.R.F., M.T., M.J.S.), the Epidemiology Unit (I.P.), and the Infectios disease Unit (M.M.), Cruces University Hospital, Bilbao, Spain. Address reprint requests to Andrés Valdivieso, Unidad Hepatobiliar y Trasplante Hepático, planta 4a E, Hospital Universitario Cruces, Plaza de Cruces s/n, 48903, Baracaldo, Bilbao, Spain. E-mail: [email protected] © 2010 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710 Transplantation Proceedings, 42, 660 – 662 (2010)
MANAGING HCC RECURRENCE
661
pression and the role of new molecularly targeted therapies have not been addressed in prior studies. Therefore, we sought to retrospectively analyze our experience in the management of HCC recurrences posttransplantation. PATIENTS AND METHODS From 1996 to May 2008, we performed 719 liver transplantations in 684 patients. One hundred eighty-two patients were transplanted due to HCC within Milan criteria (26.6%). Incidental tumors were not included in the study. We retrospectively collected demographic data, etiology of underlying disease, alpha fctoprotein (AFP) values, waiting list time, and treatment on the waiting list. Twenty-three patients (12.6%) developed hepatic and/or extrahepatic recurrence. Patients with an enhancing hepatic or extrahepatic mass on dynamic studies or an increased AFP were considered to have a recurrence; otherwise, we obtained histologic confirmation. Surgery, when possible, was considered to be the first therapeutic option. At the beginning of our program, deoxyrubicin was used when surgery was contraindicated. Mammalian target of rapamycin (mTOR)-based immunosuppression was considered from 2007 onward. Recently, we have tested sorafenib together with everolimus. Conversion to everolimus was performed at 3– 4 weeks, and sorafenib was started after the liver function was stable under everolimus treatment. Mean follow-up was 60 months; minimum follow-up, 12 months.
RESULTS
The average age of patients with HCC recurrence was 55.8 ⫾ 6.9 years. Most patients were males (91.3%). The main underlying diseases were hepatitis C virus cirrhosis (56.5%), alcohol cirrhosis (26%), and hepatitis B virus cirrhosis (8.7%). Median AFP was 8.7 (range, 2.7–5056). Median time on the waiting list was 129 days with 65% treated with TACE and/or PEI while a waiting OLT. No differences were observed in these variables when comparing transplanted patients with versus without HCC recurrence (Table 1). In the explanted liver, 134 patients fulfilled Milan criteria (73.6%); 25 fulfilled UCLA criteria (13.7%) and 23 exTable 1. Patient and Tumor Data
Age (mean ⫾ SD) Sex (M/F) % Etiology (%) Hepatitis C Alcohol Hepatitis B Others AFP (median) WL time (median) WL treatment (%) Criteria in explanted liver Milan UCLA ⬎UCLA Microvascular invasion
HCC Recurrence (n ⫽ 23)
No HCC Recurrence (n ⫽ 159)
55.8 ⫾ 6.9 91.3/8.7
58.1 ⫾ 7.6 67.6/32.4
56.5% 26% 8.7% 8.7% 8.7 (2.7–5056) 129 (6–323) 65
50% 38.5% 5% 6.5% 8.3 (0.4–1318) 120 (0–341) 60.5
10.4% 4% 35% 39%
73.6% 13.7% 12.6% 9%
AFP, alfa fetoprotein; WL, waiting list.
ceeded UCLA criteria (12.6%). Recurrence appeared in 10.4% of patients within Milan, 4% within UCLA and 35% exceeding UCLA criteria (Table 1). Microvascular invasion was observed in 39% of cases with HCC recurrence and only in 9% without recurrence (Table 1). Median time to recurrence was 23.4 months (range, 2–93). Recurrence was hepatic (n ⫽ 2), hepatic and extrahepatic (n ⫽ 5) and only extrahepatic (n ⫽ 16; 69.5%). The lung was affected in 56.5% of cases and bone in 21.7%. Surgical resection was performed in 11; chemotherapy with deoxyrubicin in 2, everolimus and sorafenib in 2, and best supportive care in 8 patients. Surgical resection was R-0 only in 8 patients involving the liver (n ⫽ 2), adrenal gland (n ⫽ 2), abdominal lymph nodes (n ⫽ 2), lung (n ⫽ 2). Four of these 8 patients developed a new recurrence, with 3 succumbing at 5.5, 35, and 52 months after surgery for HCC recurrence. The remaining patient, who was treated with everolimus and sorafenib as adjuvant treatment, was alive at 19.3 months after surgery. In the subgroup of 4 patients without recurrences, 1 died at 32.6 months due to hepatitis C recurrence and 3 are alive at 19, 31, and 86 months including 2 treated with everolimus and sorafenib. The median recurrence-free survival among the R-0 resection cohort was 21.2 months (range, 7–90). All patients with R-1 surgery developed new recurrences, with 1 succumbing at 19.2 months after surgery and 2 alive at 12 and 15.5 months. Their median recurrence-free survival after surgery was 4.5 months (range, 2–9). The 2 patients treated with deoxyrubicin died with recurrence at 16.3 and 21.7 months. The 2 patients treated solely with everolimus and sorafenib are alive at 18.5 months (without recurrence) and at 10 months (with a re-recurrence). Fifteen of 23 patients with recurrences died: 13 due to HCC recurrence and 2 to hepatitis C recurrence (1 with HCC recurrence). The average survival from transplantation was significantly different between patients with versus without a recurrence (48 ⫾ 34.3 vs 61.7 ⫾ 37.5 months P ⬍ .001) with 5-year survival rates of 48% and 83.5%, respectively (P ⬍ 0.001; Fig 1). The mean survival time from recurrence was also significantly different between patients with R-0 surgery versus the other patients (33.2 ⫾ 21.5 vs 11.9 ⫾ 6.9 months; P ⫽ .006). The estimated 5-year survival rate was 27% for patients with R-0 surgery and 0% at 3 years for the other subjects (Fig 2). DISCUSSION
Postransplant recurrence has been minimized by adopting the so called “Milan criteria.” Despite these strict selection criteria, recurrent HCC develops in 8%–20% of patients.1,2 Our recurrence rate of 12.5% was consistent with this result. As expected it produced a negative impact on survival (83.5% vs 43% at 5 years; P ⬍ .001). Well-known tumor factors associated with a high risk of HCC recurrence were also correlated with event in our population; size, number of lesions, and presence, of microvascular invasion in the explant. Nevertheless, in this study, 5-year survival
662
among patients with HCC recurrence was 43%, which favorably compared with results from other series.1 As previously reported,1–3 our data confirmed that the few patients who were amenable to resection with curative intention (hepatic resection ⫾ metastasectomy) showed significantly longer survival than those without this therapeutic option (27% at 5 years vs 0% at 3 years). There is an obvious selection bias because of the more favorable characteristics of the surgically treated group, but this reinforces the need for continued, prolonged surveillance, seeking to detect recurrence during an early, surgically treatable stage. Although our patients with “curative” surgery displayed significantly longer survival, most of them (50%) had developed a new recurrence, a finding consistent with the results reported by Roayaie et al1 and Regalia et al.4 Similar to nontransplant surgical HCC patients, there is no established adjuvant therapy. Some evidence favors conversion of immunosuppression from calcineurin to mTOR-inhibitors due to the antitumor effects of the latter drugs.5 Therefore, we decided in 2007 to switch patients with proven recurrences to everolimus as adjuvant treatment postoperatively, and as primary therapy for those with unresectable recurrences. Furthermore, it seems rational to use an mTOR-based regimen for patients whose pathologic examination of the explanted liver suggests a high risk of recurrence.5 Sorafenib, a tyrosine kinase inhibitor, has recently been approved for treatment of advanced HCC.6 We introduced it in 3 patients subsequent to surgical resection of aggressive recurrences and in 2 nonsurgical patients. In our experience with 5 patients, conversion to everolimus and sorafenib after surgery seemed to prolong survival (median survival after recurrence, 18.8 months) without significant postoperative, immunologic, or hematologic complications. No significant pharmacodynamic interactions were observed with everolimus and the adverse effects of sorafenib were mild, similar to those described among the nontransplant HCC population. Although this strategy needs to be validated in well-designed, randomized, controlled trials, it seems reasonable to speculate that everolimus could potentiate sorafenib to inhibit tumor recurrence. In our experi-
Fig 1. Survival from transplant based on HCC recurrence.
VALDIVIESO, BUSTAMANTE, GASTACA ET AL
Fig 2. Survival from HCC recurrence diagnosis based on treatment management.
ence, it did not affect graft function. Considering that there are no real therapeutic options for nonsurgical patients with posttransplant HCC recurrence, this combination therapy should also be evaluated in this population.7 In the near future, genomic markers may allow further refinement of candidate selection to achieve far lower recurrence rates. Until then, efforts should be directed to establish the most adequate immunosuppressive regimes for patients transplanted with HCC and to utilize surveillance protocols to identify early recurrences among at risk patients. In conclusion, HCC recurrence rate is low among patients transplanted within the Milan criteria (12.6%), but has a significant impact on survival. Therapeutic options for this complication are limited, but patients with recurrences amenable to surgery should be resected. Although there is an obvious selection bias, surgery seemed to prolong survival even when it was not curative (64% re-recurrence). There is a rationale for the use of sorafenib ⫹ mTOR-based immunosuppression, but this warrants further studies in randomized controlled trials. REFERENCES 1. Roayaie S, Schwartz JD, Sung MW, et al: Recurrence of hepatocellular carcinoma after liver transplant: patterns and prognosis. Liver Transpl 10:534, 2004 2. Hollebecque A, Decaens T, Boleslawski E, et al: Natural history and therapeutic management of recurrent hepatocellular carcinoma after liver transplantation. Gastroentrol Clin Biol 33: 361, 2009 3. Schwartz M, Roayaie S, Llovet J: How should patients with hepatocellular carcinoma after liver transplantation be treated?. J Hepatol 43:584, 2005 4. Regalia E, Fassati LR, Valente U, et al: Pattern and management of recurrent hepatocellular carcinoma alter liver transplantation. J Hepatobiliary Pancreat 5:29, 1998 5. Zimmerman MA, Trotter JF, Wachs M, et al: Sirolimus-based immunosuppression following liver transplantation for hepatocellular carcinoma. Liver Transpl 14:633, 2008 6. Llovet JM, Ricci S, Mazzaferro V, et al: Sorafenib in advanced hepatocellular carcinoma. N Engl J Med 359:378, 2008 7. Newell P, Toffanin S, Villanueva A, et al: Ras pathway activation in hepatocellular carcinoma and anti-tumoral effect of combined sorafenib and rapamycin in vivo. J Hepatol 51:725, 2009
O
0041-1345/10/$–see front matter doi:10.1016/j.transproceed.2010.02.014 660
have reported specifically the prognosis and treatment of posttransplant recurrent HCC, but the published data suggest that selected surgically treated patients seem to show a better prognosis.2– 4 Unfortunately, the use of tailored immunosupFrom the Hepatobiliary and Liver Transplantation Unit (A.Va., M.G., J.G.U., A.Ve., P.R., J.O.d.U.), the Digestive Service (J.B., J.R.F., M.T., M.J.S.), the Epidemiology Unit (I.P.), and the Infectios disease Unit (M.M.), Cruces University Hospital, Bilbao, Spain. Address reprint requests to Andrés Valdivieso, Unidad Hepatobiliar y Trasplante Hepático, planta 4a E, Hospital Universitario Cruces, Plaza de Cruces s/n, 48903, Baracaldo, Bilbao, Spain. E-mail: [email protected] © 2010 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710 Transplantation Proceedings, 42, 660 – 662 (2010)
MANAGING HCC RECURRENCE
661
pression and the role of new molecularly targeted therapies have not been addressed in prior studies. Therefore, we sought to retrospectively analyze our experience in the management of HCC recurrences posttransplantation. PATIENTS AND METHODS From 1996 to May 2008, we performed 719 liver transplantations in 684 patients. One hundred eighty-two patients were transplanted due to HCC within Milan criteria (26.6%). Incidental tumors were not included in the study. We retrospectively collected demographic data, etiology of underlying disease, alpha fctoprotein (AFP) values, waiting list time, and treatment on the waiting list. Twenty-three patients (12.6%) developed hepatic and/or extrahepatic recurrence. Patients with an enhancing hepatic or extrahepatic mass on dynamic studies or an increased AFP were considered to have a recurrence; otherwise, we obtained histologic confirmation. Surgery, when possible, was considered to be the first therapeutic option. At the beginning of our program, deoxyrubicin was used when surgery was contraindicated. Mammalian target of rapamycin (mTOR)-based immunosuppression was considered from 2007 onward. Recently, we have tested sorafenib together with everolimus. Conversion to everolimus was performed at 3– 4 weeks, and sorafenib was started after the liver function was stable under everolimus treatment. Mean follow-up was 60 months; minimum follow-up, 12 months.
RESULTS
The average age of patients with HCC recurrence was 55.8 ⫾ 6.9 years. Most patients were males (91.3%). The main underlying diseases were hepatitis C virus cirrhosis (56.5%), alcohol cirrhosis (26%), and hepatitis B virus cirrhosis (8.7%). Median AFP was 8.7 (range, 2.7–5056). Median time on the waiting list was 129 days with 65% treated with TACE and/or PEI while a waiting OLT. No differences were observed in these variables when comparing transplanted patients with versus without HCC recurrence (Table 1). In the explanted liver, 134 patients fulfilled Milan criteria (73.6%); 25 fulfilled UCLA criteria (13.7%) and 23 exTable 1. Patient and Tumor Data
Age (mean ⫾ SD) Sex (M/F) % Etiology (%) Hepatitis C Alcohol Hepatitis B Others AFP (median) WL time (median) WL treatment (%) Criteria in explanted liver Milan UCLA ⬎UCLA Microvascular invasion
HCC Recurrence (n ⫽ 23)
No HCC Recurrence (n ⫽ 159)
55.8 ⫾ 6.9 91.3/8.7
58.1 ⫾ 7.6 67.6/32.4
56.5% 26% 8.7% 8.7% 8.7 (2.7–5056) 129 (6–323) 65
50% 38.5% 5% 6.5% 8.3 (0.4–1318) 120 (0–341) 60.5
10.4% 4% 35% 39%
73.6% 13.7% 12.6% 9%
AFP, alfa fetoprotein; WL, waiting list.
ceeded UCLA criteria (12.6%). Recurrence appeared in 10.4% of patients within Milan, 4% within UCLA and 35% exceeding UCLA criteria (Table 1). Microvascular invasion was observed in 39% of cases with HCC recurrence and only in 9% without recurrence (Table 1). Median time to recurrence was 23.4 months (range, 2–93). Recurrence was hepatic (n ⫽ 2), hepatic and extrahepatic (n ⫽ 5) and only extrahepatic (n ⫽ 16; 69.5%). The lung was affected in 56.5% of cases and bone in 21.7%. Surgical resection was performed in 11; chemotherapy with deoxyrubicin in 2, everolimus and sorafenib in 2, and best supportive care in 8 patients. Surgical resection was R-0 only in 8 patients involving the liver (n ⫽ 2), adrenal gland (n ⫽ 2), abdominal lymph nodes (n ⫽ 2), lung (n ⫽ 2). Four of these 8 patients developed a new recurrence, with 3 succumbing at 5.5, 35, and 52 months after surgery for HCC recurrence. The remaining patient, who was treated with everolimus and sorafenib as adjuvant treatment, was alive at 19.3 months after surgery. In the subgroup of 4 patients without recurrences, 1 died at 32.6 months due to hepatitis C recurrence and 3 are alive at 19, 31, and 86 months including 2 treated with everolimus and sorafenib. The median recurrence-free survival among the R-0 resection cohort was 21.2 months (range, 7–90). All patients with R-1 surgery developed new recurrences, with 1 succumbing at 19.2 months after surgery and 2 alive at 12 and 15.5 months. Their median recurrence-free survival after surgery was 4.5 months (range, 2–9). The 2 patients treated with deoxyrubicin died with recurrence at 16.3 and 21.7 months. The 2 patients treated solely with everolimus and sorafenib are alive at 18.5 months (without recurrence) and at 10 months (with a re-recurrence). Fifteen of 23 patients with recurrences died: 13 due to HCC recurrence and 2 to hepatitis C recurrence (1 with HCC recurrence). The average survival from transplantation was significantly different between patients with versus without a recurrence (48 ⫾ 34.3 vs 61.7 ⫾ 37.5 months P ⬍ .001) with 5-year survival rates of 48% and 83.5%, respectively (P ⬍ 0.001; Fig 1). The mean survival time from recurrence was also significantly different between patients with R-0 surgery versus the other patients (33.2 ⫾ 21.5 vs 11.9 ⫾ 6.9 months; P ⫽ .006). The estimated 5-year survival rate was 27% for patients with R-0 surgery and 0% at 3 years for the other subjects (Fig 2). DISCUSSION
Postransplant recurrence has been minimized by adopting the so called “Milan criteria.” Despite these strict selection criteria, recurrent HCC develops in 8%–20% of patients.1,2 Our recurrence rate of 12.5% was consistent with this result. As expected it produced a negative impact on survival (83.5% vs 43% at 5 years; P ⬍ .001). Well-known tumor factors associated with a high risk of HCC recurrence were also correlated with event in our population; size, number of lesions, and presence, of microvascular invasion in the explant. Nevertheless, in this study, 5-year survival
662
among patients with HCC recurrence was 43%, which favorably compared with results from other series.1 As previously reported,1–3 our data confirmed that the few patients who were amenable to resection with curative intention (hepatic resection ⫾ metastasectomy) showed significantly longer survival than those without this therapeutic option (27% at 5 years vs 0% at 3 years). There is an obvious selection bias because of the more favorable characteristics of the surgically treated group, but this reinforces the need for continued, prolonged surveillance, seeking to detect recurrence during an early, surgically treatable stage. Although our patients with “curative” surgery displayed significantly longer survival, most of them (50%) had developed a new recurrence, a finding consistent with the results reported by Roayaie et al1 and Regalia et al.4 Similar to nontransplant surgical HCC patients, there is no established adjuvant therapy. Some evidence favors conversion of immunosuppression from calcineurin to mTOR-inhibitors due to the antitumor effects of the latter drugs.5 Therefore, we decided in 2007 to switch patients with proven recurrences to everolimus as adjuvant treatment postoperatively, and as primary therapy for those with unresectable recurrences. Furthermore, it seems rational to use an mTOR-based regimen for patients whose pathologic examination of the explanted liver suggests a high risk of recurrence.5 Sorafenib, a tyrosine kinase inhibitor, has recently been approved for treatment of advanced HCC.6 We introduced it in 3 patients subsequent to surgical resection of aggressive recurrences and in 2 nonsurgical patients. In our experience with 5 patients, conversion to everolimus and sorafenib after surgery seemed to prolong survival (median survival after recurrence, 18.8 months) without significant postoperative, immunologic, or hematologic complications. No significant pharmacodynamic interactions were observed with everolimus and the adverse effects of sorafenib were mild, similar to those described among the nontransplant HCC population. Although this strategy needs to be validated in well-designed, randomized, controlled trials, it seems reasonable to speculate that everolimus could potentiate sorafenib to inhibit tumor recurrence. In our experi-
Fig 1. Survival from transplant based on HCC recurrence.
VALDIVIESO, BUSTAMANTE, GASTACA ET AL
Fig 2. Survival from HCC recurrence diagnosis based on treatment management.
ence, it did not affect graft function. Considering that there are no real therapeutic options for nonsurgical patients with posttransplant HCC recurrence, this combination therapy should also be evaluated in this population.7 In the near future, genomic markers may allow further refinement of candidate selection to achieve far lower recurrence rates. Until then, efforts should be directed to establish the most adequate immunosuppressive regimes for patients transplanted with HCC and to utilize surveillance protocols to identify early recurrences among at risk patients. In conclusion, HCC recurrence rate is low among patients transplanted within the Milan criteria (12.6%), but has a significant impact on survival. Therapeutic options for this complication are limited, but patients with recurrences amenable to surgery should be resected. Although there is an obvious selection bias, surgery seemed to prolong survival even when it was not curative (64% re-recurrence). There is a rationale for the use of sorafenib ⫹ mTOR-based immunosuppression, but this warrants further studies in randomized controlled trials. REFERENCES 1. Roayaie S, Schwartz JD, Sung MW, et al: Recurrence of hepatocellular carcinoma after liver transplant: patterns and prognosis. Liver Transpl 10:534, 2004 2. Hollebecque A, Decaens T, Boleslawski E, et al: Natural history and therapeutic management of recurrent hepatocellular carcinoma after liver transplantation. Gastroentrol Clin Biol 33: 361, 2009 3. Schwartz M, Roayaie S, Llovet J: How should patients with hepatocellular carcinoma after liver transplantation be treated?. J Hepatol 43:584, 2005 4. Regalia E, Fassati LR, Valente U, et al: Pattern and management of recurrent hepatocellular carcinoma alter liver transplantation. J Hepatobiliary Pancreat 5:29, 1998 5. Zimmerman MA, Trotter JF, Wachs M, et al: Sirolimus-based immunosuppression following liver transplantation for hepatocellular carcinoma. Liver Transpl 14:633, 2008 6. Llovet JM, Ricci S, Mazzaferro V, et al: Sorafenib in advanced hepatocellular carcinoma. N Engl J Med 359:378, 2008 7. Newell P, Toffanin S, Villanueva A, et al: Ras pathway activation in hepatocellular carcinoma and anti-tumoral effect of combined sorafenib and rapamycin in vivo. J Hepatol 51:725, 2009