Liver transplantation for unresectable hepatocellular carcinoma in normal livers

Transcription

1 Liver transplantation for unresectable hepatocellular carcinoma in normal livers Hynek Mergental 1,39,, Rene Adam 2,, Bo-Goran Ericzon 3,, Piotr Kalicinski 4,, Ferninand Mühlbacher 5,, Krister Höckerstedt 6,, Jürgen L. Klempnauer 7,, Styrbjörn Friman 8,, Christoph E. Broelsch 9,, Georges Mantion 1,, Carlos Fernandez-Sellez 11,, Bart van Hoek 12,, Josef Fangmann 13,, Jacques Pirenne 14,, Paolo Muiesan 15,, Alfred Königsrainer 16,, Darius F. Mirza 1,, Jan Lerut 17,, Olivier Detry 18,, Yves-Ptrice Le Treut 19,, Vincenzo Mazzaferro,, Florian Löhe 21,, Marina Berenguer 22,, Pierre-Alain Clavien 23,, Xavier Rogiers 24,, Jacques Belghiti 25,, Laslo Kóbori 26,, Patrizia Burra 27,, Philippe Wolf 28,, Wolfgang Schareck 29,, Przemyslaw Pisarski 3,, Aksel Foss 31,, Franco Filipponi 32,, Marek Krawczyk 33,, Martin Wolff 34,, Jan M. Langrehr 35,, Keith Rolles 36,, Neville Jamieson 37,, Wim C.J. Hop 38,, Robert J. Porte 39,, 1 Queen Elizabeth Hospital, Birmingham, United Kingdom; 2 AP-HP, Hôpital Paul Brousse, University Paris Sud, Villejuif, France; 3 Karolinska Institutet, Stockholm, Sweden; 4 Children s Memorial Health Institute, Warsaw, Poland; 5 Medical University Vienna, Austria; 6 Helsinki University Central Hospital, Finland; 7 Medizinische Hochschule Hannover, Germany; 8 Sahlgrenska University Hospital, Goteborg, Sweden; 9 Universitatsklinikum Essen, Germany; 1 Hôpital Jean Minjoz, Besanscon, France; 11 University Juan Canalejo Medical Center, La Coruna, Spain; 12 Leiden University Medical Center, The Netherlands; 13 Universitatsklinikum Leipzig, Germany; 14 Universitaire Ziekenhuizen Leuven, Belgium; 15 King s College Hospital, London, United Kingdom; 16 Universitatsklinikum Tübingen, Germany; 17 Cliniques Universitaires Saint-Luc, Brussels, Belgium; 18 University Hospital, Liege, Belgium; 19 Hôpital de la Conception, Marseille, France; Istituto Nazionale Tumori, Milano, Italy; 21 Ludwig-Maximilians University Hospital Munich, Germany; 22 Hospital Universitario La Fe, Valencia, Spain; 23 Universitatspital Zurich, Switzerland; 24 Universitätsklinikum Hamburg-Eppendorf, Hamburg, Germany; 25 University Hôpital Beaujon, Clichy, France; 26 Semmelweis University Hospital, Budapest, Hungary; 27 Centro Trapianti di Fegato-Azienda Ospedaliera di Padova, Italy; 28 Hôpital Universitaire de Hautepierre, Strasbourg, France; 29 Universitätsklinikum Rostock, Germany; 3 Universitätsklinikum Freiburg, Germany; 31 Oslo University Hospital Rikshospitalet, Norway; 32 University of Pisa Medical School Hospital, Italy; 33 Medical University of Warsaw, Poland; 34 Universitätsklinikum, Bonn, Germany; 35 Charité Campus Virchow-Klinikum, Berlin, Germany; 36 Royal Free Hospital, London, United Kingdom; 37 Addenbrooke s Hospital, Cambridge, United Kingdom; 38 Erasmus Medical Center, Rotterdam, The Netherlands; 39 University Medical Center Groningen, Groningen, The Netherlands See Editorial, pages Background & Aims: The role of liver transplantation in the treatment of hepatocellular carcinoma in livers without fibrosis/cirrhosis (NC-HCC) is unclear. We aimed to determine selection criteria for liver transplantation in patients with NC-HCC. Keywords: Hepatocellular carcinoma; Liver transplantation; Non-cirrhotic; Nonfibrotic; Unresectable; Recurrent; Hepatoma. Received 7 October 11; received in revised form 21 March 12; accepted 22 March 12; available online 18 April 12 q DOI of original article: Corresponding author. Address: Department of Hepato-Pancreato-Biliary Surgery and Liver Transplantation, University Medical Center Groningen, P.O. Box 3.1, 97 RB Groningen, The Netherlands. Tel.: ; fax: address: r.j.porte@umcg.nl (R.J. Porte). For the European Liver and Intestine Transplant Association (ELITA). Abbreviations: HCC, hepatocellular carcinoma; NC-HCC, non-cirrhotic hepatocellular carcinoma; ELTR, European Liver Transplant Registry; ELITA, European Liver and Intestine Transplantation Association; LT, liver transplantation; primary-lt, primary liver transplantation; rescue-lt, rescue liver transplantation; CI, confidence interval; HR, hazard ration; AJCC, American Joint Committee on Cancer. Methods: Using the European Liver Transplant Registry, we identified 15 patients who underwent liver transplantation for unresectable NC-HCC. Detailed information about patient, tumor characteristics, and survival was obtained from the transplant centers. Variables associated with survival were identified using univariate and multivariate statistical analyses. Results: Liver transplantation was primary treatment in 62 patients and rescue therapy for intrahepatic recurrences after liver resection in 43. Median number of tumors was 3 (range 1 7) and median tumor size 8 cm (range.5 3). One- and 5-year overall and tumor-free survival rates were 84% and 49% and 76% and 43%, respectively. Macrovascular invasion (HR 2.55, 95% CI 1.34 to 4.86), lymph node involvement (HR 2., 95% CI 1.28 to 5.28), and time interval between liver resection and transplantation <12 months (HR 2.12, 95% CI.96 to 4.67) were independently associated with survival. Five-year survival in patients without macrovascular invasion or lymph node involvement was 59% (95% CI 47 7%). Tumor size was not associated with survival. Journal of Hepatology 12 vol. 57 j

2 Conclusions: This is the largest reported series of patients transplanted for NC-HCC. Selection of patients without macrovascular invasion or lymph node involvement, or patients P12 months after previous liver resection, can result in 5-year survival rates of 59%. In contrast to HCC in cirrhosis, tumor size is not a predictor of post-transplant survival in NC-HCC. Ó 12 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Introduction Hepatocelullar carcinoma (HCC) represents the third most frequent cause of cancer-related death worldwide and the incidence in Western countries is rising [1]. More than 9% of all HCCs occur in diseased livers, and chronic inflammation, liver fibrosis, and cirrhosis are well-known risk factors [2,3]. In less than 1% of all cases, HCC occurs in patients without any evidence for an underlying liver disease [2,4,5]. Those patients have no signs of liver fibrosis or cirrhosis and no evidence of a current or previous inflammatory disease, such as hepatitis B or C virus infection. The exact pathogenesis of HCC in a non-fibrotic and non-cirrhotic liver (NC-HCC) remains unclear, although studies suggest that some may have developed in a pre-existing liver adenoma [6]. The peak incidence of NC-HCC is in the fourth decade, which is much earlier than that for HCC in diseased livers. Because of the relatively young age, the absence of an underlying liver disease, and the lack of early symptoms, most patients with NC-HCC present at a relatively late stage when the tumor has reached a considerable size or (intrahepatic) metastases have developed already [7 9]. The treatment of first choice in patients with NC-HCC is a partial liver resection [1 12]. Because of the late presentation and often a large tumor size, a partial liver resection may not always be possible, despite the otherwise normal liver parenchyma. Although the development of innovative strategies in hepatic surgery have extended the possibilities of tumor resection during the past decade, partial resection will be always limited by anatomical factors and the need of preserving a sufficient amount of liver function to avoid death from post-operative liver failure [13]. In case of primarily unresectable NC-HCC, liver transplantation may be the only alternative with a chance of cure. The same is true for patients who develop intrahepatic recurrence after one or more previous partial liver resection(s), and in whom another partial liver resection is no longer possible. The role of primary or rescue liver transplantation in the treatment of NC-HCC, however, is poorly defined and criteria for patient selection are lacking. While liver transplantation is a well-established therapy for patients with HCC in a fibrotic or cirrhotic liver that remains within the Milan criteria (i.e. one tumor with a diameter 65cm or maximum of three tumors with a diameter 63 cm), it is unclear whether these criteria are also applicable to patients with NC-HCC [14]. In fact, the vast majority of patients with a NC-HCC that meet the Milan criteria will be able to undergo a partial liver resection and do not need transplantation. Previous reports based on small single center series or case reports have suggested that outcome after liver transplantation for NC-HCC is poor [15,16]. Due to the lack of larger clinical experiences with detailed long-term follow-up, the role of liver transplantation for NC-HCC, therefore, remained controversial. We aimed to determine outcome after liver transplantation for NC-HCC and to identify variables that are associated with survival. We, therefore, performed a detailed analysis of a large cohort of patients transplanted for NC-HCC and with a long-term follow-up. Materials and methods Study patients To identify patients who underwent liver transplantation for NC-HCC, we used the European Liver Transplant Registry (ELTR), a regularly audited registry of patients who underwent liver transplantation in one of the 137 contributing European centers [17,18]. All patients who were registered in the ELTR with the diagnosis HCC in a non-cirrhotic liver, and who underwent transplantation between January 1, 1994 and December 31, 5, were selected. In addition, we identified the following two subgroups: patients who underwent liver transplantation as a primary therapy for HCC in an otherwise completely normal liver (primary-lt group), and patients who underwent liver transplantation as rescue treatment for recurrent NC-HCC after a previous partial liver resection (rescue- LT group). This study was approved by and performed under the auspices of the Board of European Liver and Intestine Transplant Association (ELITA), the governing society of the ELTR. Data collection Upon identification of patients within the ELTR, the individual transplant centers were contacted to provide detailed patient data. A standardized case record form containing 89 items was used to record clinical variables, including information about patient demographics, pre-transplant HCC management, staging methods, details about the transplant procedure, pathology report, post-transplant treatment, methods of recurrence surveillance, and survival data (the case record form is provided as Supplementary material and available online). Case record forms were collected at the co-coordinating center, the University Medical Center Groningen, reviewed for internal consistency and completeness and transferred into a SPSS database. Detailed tumor characteristics, including tumor size, number of nodules, and histological characteristics were obtained from pathology reports of the explanted livers. All presented data describing the tumor characteristics are based on the post-transplant histopathological findings. In seven patients who had a histologically confirmed NC-HCC prior to treatment with transarterial chemoembolization, the final pathology after transplantation revealed completely necrotic tumor. Only patients with histologically confirmed absence of an underlying liver disease (i.e. no histological signs of inflammation, fibrosis, or cirrhosis) and negative serology testing for hepatitis B and C virus infection with a post-transplant follow-up of at least 12 months were included in the analysis. Overall, there were less than 5% missing data values in all variables except AFP levels (27 cases missing), information concerning adjuvant chemotherapy (7 cases missing) intraoperative blood transfusion requirements ( cases missing) and details of immunosuppressive medication (1 cases missing). Statistical analysis Categorical variables were presented as number and percentages. Continuous variables were presented as median and range (unless indicated otherwise). When clinically relevant, median values of continuous variables were used as cut-off value to create subgroups for comparison. Differences in categorical variables between groups were analyzed using the Chi-square test or Fisher s exact test. Continuous data were compared using the Mann Whitney test. Survival rates and 95% confidence intervals (95% CI) were determined using the Kaplan Meier method and groups were compared using the log-rank test. In case the effect of a variable was not significantly different between the two transplantation groups (as assessed by Cox-regression using appropriate interaction terms), the prognostic effect of that variable was evaluated in the total population using the log-rank test stratified for the transplantation group. For the entire study population, all variables tested in the univariate survival analysis with a p.1 were included in a multivariate Cox regression analysis with backward elimination, and hazard ratios (HR) and 95% CI were calculated. p <.5 (two-sided) were considered significant in all analyses. SPSS software, 298 Journal of Hepatology 12 vol. 57 j

3 Patients identified in ELTR n = 147 (time period 1/1/1995 to 31/12/5) version 15. for Windows (SPSS Inc. Chicago, IL, USA) was used for all analyses. For presenting the results, percentages were rounded to whole numbers and p values to three decimals. Results Patients included into analysis n = 15 Primary transplantation n = 62 Patients excluded upon histology review n = 28 Liver fibrosis, n = Mixed cholangiocellular carcinoma, n = 4 Underlying liver disease, n = 4 Patients excluded upon serology review n = 8 Hepatitis C infection, n = 5 Hepatitis B infection, n = 3 Patients excluded for insufficient data n = 6 Rescue transplantation n = 43 Fig. 1. Selection of patients from the European Liver Transplant Registry (ELTR). Patient characteristics and pre-transplant treatment Using the ELTR, we identified 147 patients who underwent liver transplantation for HCC in a non-cirrhotic liver. Twenty-eight patients were subsequently excluded from the analyses because more detailed information about the pathological examination of the explanted liver demonstrated presence of an underlying liver disease, liver fibrosis, or presence of mixed type HCC/ cholangiocarcinoma. Eight patients were excluded because they were seropositive for hepatitis B or C, and 6 patients were excluded because we were unable to collect sufficient information about tumor staging or patient follow-up was incomplete (Fig. 1). The remaining 15 patients who were included in the study were transplanted in 38 European centers, with a range of 1 13 patients/center (list of centers is provided as Addendum). In 62 (59%) patients, liver transplantation was used as primary therapy (primary-lt) and in 43 (41%) patients transplantation was performed as a rescue treatment for intrahepatic tumor recurrence after previous partial liver resection (rescue-lt). In the rescue-lt group, the median time interval between partial liver resection and transplantation was 23 (2 127) months. Fifteen patients in this subgroup (34%) underwent repeated liver resections prior to transplantation. Details about patient characteristics, tumor histology, as well as surgical and post-transplant variables for the entire group and for the two subgroups are provided in Table 1. Noteworthy were the female preponderance, relatively young age, and the large tumor diameter. The median (range) diameter of the largest tumor was 8 (.5 3) cm for the entire patient population, and 11 (3 3) cm and 3 (.5 12) cm for the primary-lt and rescue- LT subgroups, respectively. The majority of all 15 patients (63%) had more than one tumor nodule and only 12 patients (11%) had tumor characteristics that were within the Milan criteria. We classified all patients according to the TNM staging system of the American Joint Committee on Cancer (AJCC) Manual, 7th Edition. However, we did not use other staging systems, such as the Barcelona Clinic Liver Cancer Staging (BCLCS), the Okuda system, or the Cancer of Liver Italian Program (CLIP), as these systems have been developed for patients with fibrosis or cirrhosis and severity of underlying liver disease is an important discriminating factor in these scoring systems. Survival analysis JOURNAL OF HEPATOLOGY The median (range) length of follow-up for the entire study population was 75 (17 146) months. Overall 1-, 3-, and 5-year patient survival rates were 84%, 66%, and 49%, respectively. Tumor-free survival rates at 1-, 3-, and 5-year were 76%, 57%, and 43%, respectively (Table 2). In the primary-lt group, overall 1-, 3-, and 5-year survival rates were 8%, 62%, and 43%, respectively. In the rescue-lt group, overall survival rates at the same time intervals were 86%, 68%, and 58%, respectively. Kaplan Meier survival curves are presented in Fig. 2A. There were no major differences between overall patient survival and tumor-free survival in the entire study population or the two subgroups (Table 2). The results of a univariate analysis of variables associated with 5-year patient survival in the entire population are presented in Table 3. Variables significantly associated with 5-year survival were macrovascular invasion, hilar lymph node involvement and number of tumors (Fig. 2B, D and Supplementary Fig. 1F). Tumor differentiation (e.g. fibrolamellar versus non-fibrolamellar type NC-HCC) was not associated with survival (Supplementary Fig. 1B). There was a significant interaction between the largest tumor size and the indication for transplantation (primary-lt versus rescue-lt). Therefore, tumor size as a risk factor was analyzed only separately for each subgroup. In the primary-lt group, no additional variables were identified that were significantly associated with 5-year survival (Table 3). In the rescue-lt group, the time interval between partial liver resection and transplantation was identified as an additional factor that was strongly associated with 5-year survival (Table 3). Patients with a time interval of P12 months between liver resection and transplantation had a 5-year post-transplant survival rate of 71% (95% CI 54 87%), compared to 24% (95% CI 47%) in patients with an interval <12 months (Fig. 2F). Interestingly, there were no significant differences in 5-year survival rates in patients with tumor characteristics within or outside the Milan criteria (Supplementary Fig. 1C and D). In the primary-lt group, 5-year survival in patients with the largest tumor size up to the median value of 11 cm was 55% (95% CI 38 73%, Fig. 2E). Kaplan Meier survival curves for subgroups based on different maximum tumor diameters in patients in the primary-lt group are presented in Supplementary Fig. 2. Journal of Hepatology 12 vol. 57 j

4 Table 1. Characteristics of all 15 patients as well as primary and rescue transplant subgroups. All patients (n = 15) Primary-LT (n = 62) Rescue-LT (n = 43) p value* Situation before transplantation Gender, male, No. (%) 45 (43) 3 (48) 15 (35).229 Age (yr), median (range) (4-68) 39 (4-66) 41 (12-68).691 Repeated liver resection, No. (%) n.a. n.a. 15 (35) n.a. Time resection to LT (mo), median (range) n.a. n.a. 23 (2-127) n.a. TACE, No. (%) 37 (35) 15 (27) 22 (51).7 Neoadjuvant chemotherapy 27 (26) 18 (3) 9 (21).37 Tumor characteristics Largest tumor size (cm), median (range) 8 (.5-3) 11 (3-3) 3 (.5-12) <.1 Number of lesions, median (range)** 3. (1-7) 3. (1-7) 4. (1-7).279 Solitary tumor, No. (%) 37 (35) 28 (45) 9 (21) tumors, No. (%) (57) 36 (58) 24 (56).843 Within Milan criteria, No. (%) 12 (11) 2 (3) 1 (23).7 Tumor differentiation, No. (%).57 Fibrolamellar 11 (11) 7 (11) 4 (9) Well 45 (43) 33 (53) 12 (28) Moderate 33 (31) 15 (24) 18 (42) Poor 9 (9) 5 (8) 4 (9) Not assessable (only necrosis) 7 (7) 2 (3) 5 (12) Microvascular vasoinvasion positive, No. (%) 48 (46) 3 (48) 18 (43).554 Macrovascular vasoinvasion positive, No. (%) 15 (14) 11 (18) 4 (9).268 Alpha fetoprotein (ng/ml), median (range) 12 (-29,25) 49 (1-29,25) 8 (-9213).67 Alpha fetoprotein above 1 ng/ml, No. (%) 24 (31) 19 () 5 (16).26 Lymph node involvement positive, No. (%) 12 (11) 8 (13) 4 (9).757 TNM tumor stage***, No. (%).2 Stage I 19 (18) 14 (23) 5 (12) Stage II 37 (35) 12 (19) 25 (58) Stage III A 25 (24) 18 (29) 7 (16) Stage III B 12 (11) 1 (16) 2 (5) Stage IV A 12 (11) 8 (13) 4 (9) Surgical variables and post-operative management Full size graft, No. (%) 82 (78) 49 (79) 33 (77).814 Living donor, No. (%) 18 (17) 9 (15) 9 (21).437 Conventional transplantation, No. (%) 66 (64) (66) 26 (61).68 RBC units given, median (range) 4 (-35) 3 (-35) 5 (-28).59 Adjuvant chemotherapy, No. (%) 38 (39) 27 (47) 11 (28).62 Maintain immunosuppressive treatment, No. (%) 1. Calcineurin inhibitors 91 (95) 53 (95) 38 (95) Rapamycin 9 (1) 3 (5) 6 (15) TACE, transarterial chemoembolization; RBC, red blood cells; LT, liver transplantation; n.a., not applicable. Missing values: overall, there were less than 5% missing data values in all variables except AFP levels (27 cases missing), information concerning adjuvant chemotherapy (7 cases missing) peroperative blood transfusion requirements ( cases missing) and details of immunosuppressive medication (1 cases missing). p value for comparison between primary-lt and rescue-lt group. Highest number limited to more than 6 on the case report form. According to the 7th edition of the Tumor, Node, Metastasis staging system of the American Joint Committee on Cancer (there were no patients with TNM stage IIIC or IVB). After multivariate Cox regression analysis, only the presence of macrovascular invasion, involvement of hilar lymph nodes, and rescue-lt for intrahepatic recurrence within 12 months after partial liver resection were identified as independent risk factors for 5-year patient survival (Table 4). Overall 5-year survival in patients without macrovascular invasion and hilar lymph node involvement was 59% (95% CI 47 7%), 3 Journal of Hepatology 12 vol. 57 j

5 JOURNAL OF HEPATOLOGY Table 2. Outcome after transplantation. Outcome Survival rate No. (%) (95% CI) All patients (n = 15) 1-year survival 84 (77-91) 1-year tumor-free survival 76 (68-85) 5-year survival 49 (39-59) 5-year tumor-free survival 43 (33-53) Patient status Alive, No. (%) 47 (45) Alive with recurrence, No. (%) 6 (6) Death 58 (55) HCC unrelated death, No. (%) 11 (11) Primary transplantation subgroup (n = 62) 1-year survival 8 (7-9) 1-year tumor-free survival 74 (62-85) 5-year survival 43 (3-55) 5-year tumor-free survival (26-53) Patient status Alive, No. (%) 24 (39) Alive with recurrence, No. (%) 3 (5) Death 38 (61) HCC unrelated death, No. (%) 8 (13) Rescue transplantation subgroup (n = 43) 1-year survival 86 (76-97) 1-year tumor-free survival 8 (68-93) 5-year survival 58 (43-73) 5-year tumor-free survival 48 (31-64) Patient status Alive, No. (%) 23 (54) Alive with recurrence, No. (%) 3 (7) Death (47) HCC unrelated death, No. (%) 3 (7) compared to 16% (95% CI 36%) in patients with these two main risk factors (Fig. 2D). Absence of these two major risks factors, in patients undergoing rescue-lt for intrahepatic recurrences P12 months after a previous partial liver resection, was associated with an even higher 5-year survival rate of 83% (95% CI 68 98%). When patients were categorized based on the AJCC TNM staging system also a clear difference in patient survival was noted with significantly worse survival in patients with TNM stage IIIB or higher. Patients with TNM stage I to IIIA (any tumor size, but no macrovascular or lymph node involvement) had a 5-year survival rate of 59% (95% CI 47 7%), compared to 17% (95% CI 1 34%) in patients with stage IIIB or higher. Discussion We here present the largest series of patients transplanted for HCC in an otherwise normal liver (NC-HCC). Patients with HCC occurring in a liver without any proven underlying pathology, such as (chronic) inflammation, fibrosis, or cirrhosis and negative testing for hepatitis B or C infection, represent a small subgroup of around 1% of all patients with HCC [2,4,5]. Although the exact pathogenesis of this type of HCC remains elusive, some studies have suggested that some NC-HCC may develop in a preexisting liver cell adenoma [6]. The absence of any of the known risk factors for HCC, in combination with the female preponderance and relatively young age at presentation, suggests a different pathogenesis and possibly also different tumor biology in NC- HCC, compared to HCC occurring in a diseased liver [19]. In the current study, patient characteristics were very similar to those reported in other series of patients with NC-HCC [23]. While partial liver resection is the treatment of first choice in patients with NC-HCC, this may not always be possible in individual cases because of late discovery and large tumor diameter or multiplicity of tumor nodules [2,1,15,]. The role of liver transplantation as a primary therapy in these patients has remained controversial and specific selection criteria are lacking [15,21]. In addition, the indication for liver transplantation in patients who present with intrahepatic recurrence(s) after a previous partial liver resection for NC-HCC remains to be determined. In the current study, we have analyzed outcome in a large series of 15 patients who underwent liver transplantation for NC- HCC. The study was based on a collaborative analysis of 38 European transplant centers and the number of patients per center varied between 1 and 13, emphasizing the unique character of this type of HCC. Conclusions on the role of liver transplantation for this specific indication can, therefore, only be drawn from a multicenter study. It is important to emphasize that all patients included in this study had no histological signs of liver fibrosis or cirrhosis, no evidence of any other underlying liver disease, and were all seronegative for hepatitis B and C. The most important finding of this study is that a 5-year survival rate of 59% can be obtained in patients without macrovascular invasion or lymph node involvement, independent of tumor size. This survival rate is well above the cut-off generally required to justify liver transplantation for patients with a malignancy [22]. Many patients who present with HCC in an otherwise normal liver are previously healthy and relatively young (median age in this series was 45 years) and symptoms may not occur until tumors have reached a considerable diameter. This is different from patients who are known with a liver disease and who are generally screened at regular intervals for the development of HCC, resulting in an earlier detection of tumors with a smaller diameter. In the current series, the median largest tumor size was 8 cm for the entire group and 11 cm for the primary-lt group. Despite these large tumors, size was not identified as a predictor of survival after transplantation. In fact, the Milan criteria, generally used to select patients for transplantation with HCC in a cirrhotic liver, were not associated with post-transplant survival in patients with NC-HCC. In patients with NC-HCC, tumor size alone should therefore not be used to discriminate patients in whom transplantation may or may not be justified. Macrovascular tumor invasion and lymph node involvement were identified as the two most important variables associated with poor survival. These findings are in line with data from patients transplanted for HCC in cirrhotic livers [15,24 26]. Based on these findings, we can conclude that the presence of macrovascular invasion or the detection of tumor metastases in the hilar lymph nodes should be considered a contraindication for Journal of Hepatology 12 vol. 57 j

6 A B C Overall Primary Rescue No invasion Microvascular Macrovascular Lymph node negative Lymph node positive p =.231 p =.11 p = D E F Without involvement Vaso- and LN-invasion Up to 11 cm Larger than 11 cm Up to 12 mo Longer than 12 mo p = p = p = Fig. 2. Impact of selected variables on 5-year patient survival. (A) Kaplan Meier estimates of patient survival after liver transplantation in 15 patients with hepatocellular carcinoma in an otherwise normal liver (NC-HCC). There was no statistically significant difference in overall survival between patients who underwent liver transplantation as primary therapy for NC-HCC (n = 62) and those who underwent liver transplantation as rescue therapy for intrahepatic recurrence after a previous partial liver resection (n = 43). Probability of survival in patients with or without evidence of micro- or macrovascular tumor invasion in the entire cohort is shown in (B), and in patients with or without hilar lymph node involvement in (C). Kaplan Meier estimates of patient survival after liver transplantation for all patients with NC-HCC, according to the combined presence or absence of macrovascular invasion and lymph node (LN) involvement is shown in (D). Kaplan Meier estimates of patient survival after liver transplantation as primary therapy for NC-HCC, according to the largest tumor size (n = 62). Patient with a largest tumor size less than the median value of 11 cm had a 5- year survival rate of 55% (95% CI 38 73%), compared to 34% (95% CI 17 52%) in patients with a largest tumor size >11 cm as is shown in (E). The effect of a time interval between a previous partial liver resection and liver transplantation of more or less than 12 months in the rescue therapy group (n = 43) is shown in (F). liver transplantation in patients with NC-HCC. Nowadays, crosssectional imaging in combination with endoscopic ultrasoundguided fine needle aspiration cytology can yield more than 9% accuracy of the preoperative HCC staging [27 29]. However, we recommend routine lymph node dissection and frozen section examination as a first step in the transplant procedure and the case should be aborted if these lymph nodes are positive. This will require the need for a back-up patient on stand by in order to avoid loss of a donor liver. In contrast to macrovascular invasion, microvascular tumor invasion was not associated with post-transplant survival. This is in contrast with a well described association between microvascular invasion and post-transplant survival in patients transplanted for HCC in a diseased liver [3]. On the other hand, the current findings are in accordance with studies in patients undergoing partial liver resection for NC-HCC, which have also not shown an association between microvascular invasion and post-operative survival [31,32]. In a recent study in patients undergoing liver resection for NC-HCC, the investigators noted that the frequency of vascular invasion was only 21%, which is unexpectedly low for the mean tumor size of 1.3 cm in this series [32]. Altogether, data suggest that microvascular invasion 32 Journal of Hepatology 12 vol. 57 j

7 JOURNAL OF HEPATOLOGY Table 3. Univariate analysis of variable effect on 5-year survival. Variable 5-year survival p value rate (95% CI) All patients (n = 15) Pretransplant situation Indication for transplantation.231 Primary 45 (32-57) Rescue 57 (41-72) Transarterial.942 chemoembolization Yes 46 (29-62) No 51 (39-64) Neoadjuvant chemotherapy.121 Yes 62 (43-8) No 46 (34-57) Tumor characteristics Number of tumors.772 Solitary 47 (31-64) Multiple 51 (38-63) Number of tumors.33 Up to 4 nodules 57 (44-7) More than 4 nodules 39 (24-54) HCC differentiation.435 Fibrolamellar Non-fibrolamellar 55 (25-84) 49 (38-59) HCC differentiation.4 Well (including fibrolamellar) 5 (36-63) Moderate 46 (29-64) Poor 39 (5-73) Not assessable 71 (38-1) (only necrosis) Microvascular invasion.25 Negative 57 (44-7) Positive (25-54) Macrovascular invasion.3 Negative 55 (44-65) Positive 16 (-36) Lymph node involvement.7 Negative 54 (43-64) Positive 17 (-38) Macrovascular invasion and.2 lymph node involvement Negative 59 (47-7) Positive 16 (-36) Variable 5-year survival p value rate (95% CI) Primary transplant subgroup (n = 62) Tumor size.461 Up to 11 cm 55 (38-73) Larger than 11 cm 34 (17-52) Number of tumors.833 Solitary 43 (27-62) Multiple 46 (29-63) Milan criteria.153 Within 33 (-86) Outside 46 (31-56) Rescue transplant subgroup (n = 43) Previous partial liver resection.986 One time 58 (32-84) Multiple 58 (-76) Time interval after partial liver.5 resection to transplantation Up to 12 months 24 (-47) Longer than 12 months 71 (54-87) Tumor size.96 Up to 5 cm 59 (48-81) Larger than 5 cm 5 (3-64) Number of tumors.929 Solitary 53 (19-87) Multiple 58 (41-75) Milan criteria.65 Within 79 (57-1) Outside 5 (32-67) may be a delayed feature in the growth of NC-HCC and, therefore, a less important risk factor in patients with NC-HCC, compared to patients with HCC in a cirrhotic liver. More research in this area will be needed. In the current study, tumor differentiation was not associated with post-transplant survival. This is in contrast with previous case reports and small single center series suggesting an unfavorable outcome in patients with a non-fibrolamellar type of NC- HCC and suggesting that liver transplantation should not be offered to those patients [9,11,16,33,34]. In the current series, post-transplant survival was not different in patients with a fibrolamellar or non-fibrolamellar type of NC-HCC [35]. Therefore, we conclude that non-fibrolamellar NC-HCC alone should not be considered a contraindication for transplantation. There is still controversy about whether or not fibrolamellar HCC is a different tumor compared to non-fibrolamellar HCC in an otherwise normal liver. The results from this study did not show any difference in outcome between these two patients groups, which Journal of Hepatology 12 vol. 57 j

8 Table 4. Multivariate survival analysis (Cox-regression) of various risk factors. Factor Hazard 95% CI p value ratio Macrovascular invasion Negative 1 # - - Positive 2.55 ( ).5 Lymph nodes Negative 1 # - - Positive 2. ( ).8 Indication Primary transplantation 1 # - - Rescue transplantation 2.12 ( ) mo* Rescue transplantation >12 mo*.47 ( ).41 # Reference category. p =.2 for the difference. is in line with the assumption that fibrolamellar HCC is not a biologically more favorable type of HCC in a non-cirrhotic liver [36]. A literature review by Kakar et al. has shown that outcomes of patients with fibrolamellar and non-fibrolamellar HCC are similar when same-stage diseases are considered [36]. The question whether or not fibrolamellar HCC is a variant of NC-HCC or a different entity has not been completely settled in the literature and our study cannot address this either. One of the key aspects in selecting patients with HCC for liver transplantation remains the identification of reliable surrogate markers of tumor biology. Liver transplantation will be futile in those patients with aggressive tumor biology whereas transplantation may be justified in patients with milder tumor behavior. In patients in whom liver transplantation is considered as rescue therapy for intrahepatic recurrences after a previous partial liver resection, two important pieces of information will be available that can help physicians in differentiating relatively favorable tumor biology from aggressive and unfavorable tumor biology. These factors are: (a) the possibility of histological assessment of the previously resected tumor and (b) the time interval between partial liver resection and tumor recurrence. Unfortunately, we were unable to collect detailed information on the pathological examination of the previously resected liver tumors as most of these resections were performed in another center than the transplant center. It appeared impossible to retrospectively identify the referring centers for all 43 patients who underwent partial liver resection prior to their transplant and to collect detailed information on the pathology reports. However, it seems reasonable to assume that presence of either macrovascular invasion or hilar lymph node involvement at the time of partial liver resection is associated with worse outcome after a subsequent liver transplantation. The other known risk factors for NC-HCC recurrence after partial liver resection are incomplete tumor resection, multiple tumors, poor tumor differentiation, absence of tumor capsule and perioperative administration of blood transfusion [1,32,35,37 39]. Interestingly, several reports have suggested that outcome after partial liver resection for NC-HCC is not influenced by tumor size [38,,41]. This is in line with the findings in the current study. The time interval between previous partial liver resection and transplantation, however, was known for all patients in the rescue-lt group in the current study. It appeared that a time interval of P12 months was associated with significantly better posttransplant survival than transplantation for intrahepatic recurrences within 12 months after a partial liver resection. In the former group, 5-year survival after transplantation was 71%. Absence of macrovascular invasion and lymph node involvement, in patients undergoing rescue-lt for intrahepatic recurrences P12 months after a previous partial liver resection, was associated with an even higher 5-year survival rate of 83%. These findings are clinically very relevant as 75% of the recurrences after liver resection for NC-HCC occur exclusively in the liver [32]. This study has some limitations due to its retrospective, multicenter design. First of all, it is likely that criteria to decide whether a tumor was resectable or not were not uniform in all centers. Threshold for resectability may also have changed over time. However, the majority of patients included in this study were transplanted in high volume centers with a large experience in both liver surgery and liver transplantation and it is likely that the vast majority of patients included in this series had truly unresectable tumors. Secondly, pre-transplant and post-transplant therapy has been different in the centers. Although we have collected information on additional therapies, such as transarterial chemoembolization or (neo-) adjuvant chemotherapy for all patients, we could not demonstrate any effect of these therapies on post-transplant survival. However, the numbers in these subgroups were small. In conclusion, our findings indicate that liver transplantation may be justified as a curative treatment option in a selected group of patients with unresectable HCC in an otherwise normal liver. Partial liver resection, however, should always be the treatment of first choice. In contrast to patients with HCC in a diseased (i.e. fibrotic or cirrhotic) liver, tumor size is not an important risk factor for post-transplant survival in patients with NC-HCC. In fact, many of these patients have a tumor diameter that is already outside the Milan criteria at the time of first presentation. Selection of patients without macrovascular invasion or hilar lymph node involvement is associated with a 5-year post-transplant survival rate of 59%. Presence of any of these two risk factors, however, should be considered a contraindication for transplantation. Favorable results with a 5-year survival rate of 71% can also be obtained after liver transplantation in patients with an intrahepatic tumor recurrence P12 months after a previous partial liver resection for NC-HCC. Conflict of interest The authors who have taken part in this study declared that they do not have anything to disclose regarding funding or conflict of interest with respect to this manuscript. Authors contribution R.J.P. initiated this study and was the principal investigator. H.M. and R.J.P. prepared the study design, were responsible for the study management, data collection and interpretation, manuscript preparation and submission. H.M. and W.H. were responsible for the statistical data analysis. All the other co-authors were responsible for data collection and completion of the case report forms for patients included from their center. All co-authors reviewed, amended the manuscript and approved the final ver- 34 Journal of Hepatology 12 vol. 57 j

9 sion of it (the order of co-authors is based on number of patients included from each center). Acknowledgement The authors thank Vincent Karam, data manager of the ELTR, for his support in identifying patients for this study. Supplementary data Supplementary data associated with this article can be found, in the online version, at References [1] Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2. CA Cancer J Clin 5;55: [2] Llovet JM, Burroughs A, Bruix J. Hepatocellular carcinoma. Lancet 3;362: [3] Sherman M. Hepatocellular carcinoma: epidemiology, risk factors, and screening. Semin Liver Dis 5;25: [4] Bruix J, Sherman M, Llovet JM, Beaugrand M, Lencioni R, Burroughs AK, et al. Clinical management of hepatocellular carcinoma. Conclusions of the Barcelona- EASL conference. European Association for the Study of the Liver. J Hepatol 1;35: [5] Bosch FX, Ribes J, Borras J. Epidemiology of primary liver cancer. Semin Liver Dis 1999;19: [6] Stoot JH, Coelen RJ, De Jong MC, Dejong CH. Malignant transformation of hepatocellular adenomas into hepatocellular carcinomas: a systematic review including more than 1 adenoma cases. HPB [Oxford] 1;12: [7] Smalley SR, Moertel CG, Hilton JF, Weiland LH, Weiand HS, Adson MA, et al. Hepatoma in the noncirrhotic liver. Cancer 1988;62: [8] Bralet MP, Regimbeau JM, Pineau P, Dubois S, Loas G, Degos F, et al. Hepatocellular carcinoma occurring in nonfibrotic liver: epidemiologic and histopathologic analysis of 8 French cases. Hepatology ;32: 4. [9] El-Serag HB, Davila JA. Is fibrolamellar carcinoma different from hepatocellular carcinoma? A US population-based study. Hepatology 4;39: [1] Bismuth H, Chiche L, Castaing D. Surgical treatment of hepatocellular carcinomas in noncirrhotic liver: experience with 68 liver resections. World J Surg 1995;19: [11] Ringe B, Pichlmayr R, Wittekind C, Tusch G. Surgical treatment of hepatocellular carcinoma: experience with liver resection and transplantation in 198 patients. World J Surg 1991;15: [12] Llovet JM, Bruix J, Gores GJ. Surgical resection versus transplantation for early hepatocellular carcinoma: clues for the best strategy. Hepatology ;31: [13] Clavien PA, Petrowsky H, DeOliveira ML, Graf R. Strategies for safer liver surgery and partial liver transplantation. N Engl J Med 7;356: [14] Mazzaferro V, Regalia E, Doci R, Andreola S, Pulvirenti A, Bozzetti F, et al. Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N Engl J Med 1996;334: [15] Pichlmayr R, Weimann A, Oldhafer KJ, Schlitt HJ, Klempnauer J, Bornscheuer A, et al. Role of liver transplantation in the treatment of unresectable liver cancer. World J Surg 1995;19: [16] Houben KW, McCall JL. Liver transplantation for hepatocellular carcinoma in patients without underlying liver disease: a systematic review. Liver Transpl Surg 1999;5: [17] Adam R, Cailliez V, Majno P, Karam V, McMaster P, Caine RY, et al. Normalised intrinsic mortality risk in liver transplantation: European Liver Transplant Registry study. Lancet ;356: [18] Karam V, Gunson B, Roggen F, Grande L, Wannoff W, Janssen M, et al. Quality control of the European Liver Transplant Registry: results of audit visits to the contributing centers. Transplantation 3;75: [19] Wong N, Lai P, Lee SW, Fan S, Pang E, Liew CT, et al. Assessment of genetic changes in hepatocellular carcinoma by comparative genomic hybridization analysis: relationship to disease stage, tumor size, and cirrhosis. Am J Pathol 1999;154: JOURNAL OF HEPATOLOGY [] Llovet JM, Schwartz M, Mazzaferro V. Resection and liver transplantation for hepatocellular carcinoma. Semin Liver Dis 5;25:181. [21] El-Gazzaz G, Wong W, El-Hadary MK, Gunson BK, Mirza DF, Mayer AD, et al. Outcome of liver resection and transplantation for fibrolamellar hepatocellular carcinoma. Transpl Int ;13:S6 9. [22] O Grady JG. Treatment options for other hepatic malignancies. Liver Transpl ;6:S [23] Trevisani F, Frigerio M, Santi V, Grignaschi A, Bernardi M. Hepatocellular carcinoma in non-cirrhotic liver: a reappraisal. Dig Liver Dis 1;42: [24] Iwatsuki S, Starzl TE, Sheahan DG, Yokoyama I, Demetris AJ, Todo S, et al. Hepatic resection versus transplantation for hepatocellular carcinoma. Ann Surg 1991;214: , [discussion ]. [25] Sotiropoulos GC, Radtke A, Schmitz KJ, Molmenti EP, Schroeder T, Saner FH, et al. Liver transplantation in the setting of hepatocellular carcinoma and portal vein thrombosis: a challenging dilemma? Dig Dis Sci 8;53: [26] Sotiropoulos GC, Malago M, Molmenti EP, Losch C, Lang H, Frilling A, et al. Hilar lymph nodes sampling at the time of liver transplantation for hepatocellular carcinoma: to do or not to do? Meta-analysis to determine the impact of hilar lymph nodes metastases on tumor recurrence and survival in patients with hepatocellular carcinoma undergoing liver transplantation. Transpl Int 7;: [27] Choi SH, Lee JM, Yu NC, Suh KS, Jang JJ, Kim SH, et al. Hepatocellular carcinoma in liver transplantation candidates: detection with gadobenate dimeglumine-enhanced MRI. AJR 8;191: [28] Murakami T, Imai Y, Okada M, Hyodo T, Lee WJ, Kim MJ, et al. Ultrasonography, computed tomography and magnetic resonance imaging of hepatocellular carcinoma: toward improved treatment decisions. Oncology 11;81: [29] Crowe A, Knight CS, Jhala D, Bynon SJ, Jhala NC. Diagnosis of metastatic fibrolamellar hepatocellular carcinoma by endoscopic ultrasound-guided fine needle aspiration. Cyto J 11;8:2. [3] Mazzaferro V, Llovet JM, Miceli R, Bhoori S, Schiavo M, Mariani L, et al. Predicting survival after liver transplantation in patients with hepatocellular carcinoma beyond the Milan criteria: a retrospective, exploratory analysis. Lancet Oncol 9;1: [31] Lang H, Sotiropoulos GC, Domland M, Fruhauf NR, Paul A, Husing J, et al. Liver resection for hepatocellular carcinoma in non-cirrhotic liver without underlying viral hepatitis. Br J Surg 5;92: [32] Smoot RL, Nagorney DM, Chandan VS, Que FG, Schleck CD, Harmsen WS, et al. Resection of hepatocellular carcinoma in patients without cirrhosis. Br J Surg 11;98: [33] Yokoyama I, Todo S, Iwatsuki S, Starzl TE. Liver transplantation in the treatment of primary liver cancer. Hepatogastroenterology 199;37: [34] Schlitt HJ, Neipp M, Weimann A, Oldhafer KJ, Schmoll E, Boeker K, et al. Recurrence patterns of hepatocellular and fibrolamellar carcinoma after liver transplantation. J Clin Oncol 1999;17: [35] Lerut J, Mergental H, Kahn D, Albuquerque L, Marrero J, Vauthey JN, et al. Place of liver transplantation in the treatment of hepatocellular carcinoma in the normal liver. Liver Transpl 11;17:S9 97. [36] Kakar S, Burgart LJ, Batts KP, Garcia J, Jain D, Ferrell LD. Clinicopathologic features and survival in fibrolamellar carcinoma: comparison with conventional hepatocellular carcinoma with and without cirrhosis. Mod Pathol 5;18: [37] Bege T, Le Treut YP, Hardwigsen J, Ananian P, Richa H, Campan P, et al. Prognostic factors after resection for hepatocellular carcinoma in nonfibrotic or moderately fibrotic liver. A 116-case European series. J Gastrointest Surg 7;11: [38] Dupont-Bierre E, Compagnon P, Raoul JL, Fayet G, de Lajarte-Thirouard AS, Boudjema K. Resection of hepatocellular carcinoma in noncirrhotic liver: analysis of risk factors for survival. J Am Coll Surg 5;1: [39] Grazi GL, Cescon M, Ravaioli M, Ercolani G, Gardini A, Del Gaudio M, et al. Liver resection for hepatocellular carcinoma in cirrhotics and noncirrhotics. Evaluation of clinicopathologic features and comparison of risk factors for long-term survival and tumour recurrence in a single centre. Aliment Pharmacol Ther 3;17: [] Verhoef C, de Man RA, Zondervan PE, Eijkemans MJ, Tilanus HW, Ijzermans JN. Good outcomes after resection of large hepatocellular carcinoma in the non-cirrhotic liver. Dig Surg 4;21: [41] Laurent C, Blanc JF, Nobili S, Sa Cunha A, le Bail B, Bioulac-Sage P, et al. Prognostic factors and longterm survival after hepatic resection for hepatocellular carcinoma originating from noncirrhotic liver. J Am Coll Surg 5;1: Journal of Hepatology 12 vol. 57 j