Preoperative tumour biopsy does not affect the oncologic course of patients with transplantable HCC

Preoperative tumour biopsy does not affect the oncologic course of patients with transplantable HCC

Accepted Manuscript Preoperative fine-needle aspiration tumour biopsy does not affect the oncologic course of patients with transplantable HCC Fuks Da...

474KB Sizes 0 Downloads 14 Views

Accepted Manuscript Preoperative fine-needle aspiration tumour biopsy does not affect the oncologic course of patients with transplantable HCC Fuks David, Cauchy Francois, Fusco Grazia, Paradis Valerie, Durand Francois, Belghiti Jacques PII: DOI: Reference:

S0168-8278(14)00310-9 http://dx.doi.org/10.1016/j.jhep.2014.04.046 JHEPAT 5152

To appear in:

Journal of Hepatology

Received Date: Revised Date: Accepted Date:

19 December 2013 17 April 2014 21 April 2014

Please cite this article as: David, F., Francois, C., Grazia, F., Valerie, P., Francois, D., Jacques, B., Preoperative fine-needle aspiration tumour biopsy does not affect the oncologic course of patients with transplantable HCC, Journal of Hepatology (2014), doi: http://dx.doi.org/10.1016/j.jhep.2014.04.046

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Preoperative fine-needle aspiration tumour biopsy does not affect the oncologic course of patients with transplantable HCC Running title: Preoperative HCC biopsy and liver transplantation Fuks David1, Cauchy Francois1, Fusco Grazia1, Paradis Valerie2, Durand Francois3, Belghiti Jacques1 Departments of (1) HPB surgery, (2) Pathology, (3) Hepatology Beaujon Hôpital, 100 boulevard Général Leclerc, 92110 Clichy, France

Corresponding author: Jacques Belghiti MD, Department of Surgery, Hospital Beaujon, HPB Surgery, Beaujon Hôpital, 100 boulevard Général Leclerc, 92110 Clichy, France Ph: (33)-1-40-87-58-95; Fax: (33)-1-40-87-17-24 E-mail: [email protected] Word count: 3672 Number of tables and figures: Three Tables and one Figure Abbreviations: Preoperative fine-needle aspiration biopsy (PFNAB); hepatocellular carcinoma (HCC); liver transplantation (LT); chronic liver disease (CLD); magnetic resonance imaging (MRI); computed tomography (CT); Fine needle aspiration (FNA); Radiofrequency ablation (RFA); Alpha-fetoprotein (AFP) level; Ultrasound (US)

Conflicts of interest: DF, FC, GF, VP, FD and JB have no conflict of interest No financial support

1

ABSTRACT (251 words) Background. Preoperative fine-needle aspiration biopsy (PFNAB) allows obtaining reliable hepatocellular carcinoma (HCC) diagnosis before liver transplantation (LT) in doubtful situations, but may result in higher recurrence rates following LT. This study aimed to evaluate whether PFNAB actually jeopardized the outcome of patients with transplantable HCC. Methods. From 2002 to 2012, among 309 HCC patients listed for LT, 80 (26%) underwent PFNAB (PFNAB+). Their characteristics, modalities of recurrence and survivals were retrospectively compared to those of the 229 (74%) patients without PFNAB (PFNAB-). Results. The two groups (PFNAB+ vs. PFNAB-) were similar in terms of demography, rates of lesions within the Milan criteria (81%vs.79%, p=0.676) and duration on the waiting list (7.0 vs. 6.9 months, p=0.891). Dropout following tumour progression was similar between both groups (6%vs.11%, p=0.424). Among the 278 (90%) transplanted patients, pathological analysis revealed that 11 (4%) patients had non-HCC lesions including 10 among PFNABpatients. Median follow-up was 34 months (12-135) and recurrence after LT was observed in 25 (9%) patients with no difference between both groups (9.3%vs.8.9%, p=0.904). Parietal recurrence was observed in one PFNAB+ patient and in 2 PFNAB- patients after radiofrequency ablation (p=0.797). On an intention to treat basis, 1-, 3- and 5-year overall survivals (89%, 69% and 60% vs. 85%, 67% and 61%, p=0.601) were not significantly different between PFNAB+ and PFNAB- patients. Conclusion: This study supports that preoperative tumour biopsy does not negatively influence the oncologic course of HCC patients eligible for LT. Hence, there is no argument to restrict biopsy in doubtful situations. Key-words: HCC; liver biopsy; liver transplantation; tumour recurrence

2

INTRODUCTION Liver transplantation (LT) is the most effective treatment for early hepatocellular carcinoma (HCC) in patients with chronic liver disease (CLD) [1]. Yet, the current existing imbalance between the increasing number of candidates and organ shortage [2–3] requires that a reliable diagnostic should be preoperatively obtained to avoid two major issues: first, to perform an unnecessary LT in case of benign dysplastic nodules and second, to transplant patients with HCC of poor histological prognosis [4-5]. Even though recent advances in imaging modalities have obviated the need for tissue confirmation in most hepatocellular carcinomas (HCCs), percutaneous biopsy of the tumour could solve these two concerns. Yet, both existence of false positive and risk of tumour cells seeding along the needle track following capsular rupture [6–8] currently discourages many authors [9-10] from performing preoperative tumour biopsy in patients with transplantable HCC. Nevertheless, while these arguments seem somehow intuitive, they have not been validated in large series. In this setting, it still remains to be elucidated whether tumour biopsy could actually lead to increased recurrence rates in a context of immunosuppression and negatively influence the oncologic course of patients qualifying for LT. The objective of the current study was therefore to assess whether PFNAB itself could jeopardize the long-term outcomes in large series of patients with HCC eligible for liver transplantation (LT) on an intention to treat basis.

3

MATERIALS AND METHODS Patient' selection Between January 2002 and December 2011, 309 patients with suspected HCC occurring on chronic liver disease qualifying and listed for LT at our institution were retrospectively analyzed. Diagnosis of HCC was retained either (i) morphologically for lesions >2 cm showing both early hyperenhancement and delayed hypoenhancement (washout) on at least two concordant imaging studies including triple-phase computed tomography (CT) and/or magnetic resonance imaging (MRI); or (ii) histologically for lesions <2 cm or in the absence of concordance between two coincidental imaging modalities according to the american association for the study of liver diseases (AASLD) practice guidelines for the management of HCC [3]. Preoperative tumour biopsy (PFNAB) Computed

tomography

(CT)

or

ultrasound

(US)

guided

percutaneous

(transabdominal) fine needle aspiration (FNA) biopsy, was performed using a 20 to 22 Gauge needle. PFNAB was repeated with a maximum number of three passes during the same procedure in case of insufficient material [6]. Types of tissue samples obtained with FNA technique included smears, needle rinse samples, cell-blocks, and microbiopsies from which core imprints could be made. Analyzed characteristics included existence and type of tumour cells as well as degree of differentiation. All biopsies were analysed in an on-site cytology service by a trained pathologist (VP). Management on the waiting list To limit patient dropout while on the waiting list and to achieve a good degree of necrosis of the tumour, TACE was performed whenever possible in Child A patients and

4

selectively in Child B patients. Radiofrequency ablation (RFA) or percutaneous ethanol injection (PEI) were undertaken in the presence of a maximum of 3 lesions <3 cm easily accessible using the transabdominal approach under US guidance, and without close contact to vascular or biliary structures. Follow-up after LT Patients with chronic hepatitis B were all treated using the appropriate antiviral therapy before and after LT. Treatment of hepatitis C virus (HCV) infection was considered on a case-by-case basis [11]. After LT, follow-up included liver function tests, alphafetoprotein (AFP) level, ultrasound (US), and triple phase CT scan or MRI every 3 months during the first year and every 6 months thereafter. The period of inclusion allowed for extension of follow-up for a minimum of 24 months in order to evaluate the rate of tumour recurrence. Recurrence was defined as the appearance of a new lesion with features of HCC on imaging. Recorded modalities of recurrence included number, site and time interval from liver transplantation. Recurrences were classified as hepatic, extra-hepatic and both hepatic and extra-hepatic [12]. As for the initial tumour, management of all recurrences were discussed in a multidisciplinary meeting and included surgery, ablation, embolization or systemic chemotherapy. Design of the study An intention to treat analysis was performed on all 309 patients listed for LT. Hence, both characteristics and outcomes of the 80 (26%) patients with preoperative pathologic confirmation of HCC (PFNAB+ group) were compared to that of 229 (74%) without PFNAB (PFNAB- group). Modalities of recurrence, overall and disease-free survivals were calculated in each group from the time of enrolment on the waiting list. Disease-free survival was

5

computed considering patients who developed HCC recurrence and patients who died as censored. Analysis of recurrence, overall and disease-free survival covered only the subgroup of patients reaching 24 months unless they experienced earlier recurrences. Patients who were excluded from the waiting list were analysed and the reasons was reported. Thereafter, similar analysis was performed considering only the 278 (90%) patients who actually underwent LT. Statistical analysis Patients’ data were collected from a prospectively maintained and computerized database. Nonparametric data are presented as median (range), and categorical data as both frequency and proportion (%). Univariate analysis was performed to assess for any marked differences on various outcome variables among patients in this study. Continuous variables were compared using the Mann-Whitney U test. Categorical data were analyzed using the Pearson’s chi square test. A multivariate analysis was performed by Cox regression (step-wise forward model) for variables reaching p <0.1 significance on univariate analysis or those known to strongly affect overall and disease-free survival. All statistical analyses were performed using the SPSS for Windows version 18.0 (SPSS Inc), and statistical significance was taken at the 5% level.

6

RESULTS Intention-to-treat analysis (n=309) The 80 PFNAB+ and 229 PFNAB- patients shared similar demographic characteristics, tumour characteristics on imaging including rates of lesions within the Milan criteria (81% vs. 79%, p=0.676), types of underlying liver disease, α-fetoprotein level at the time of listing (9 vs. 9 ng/ml, p=0.211), duration on the waiting list (7 vs. 6.9 months, p=0.891), and treatment while on the waiting list (80% vs. 66%, p=0.097). However, PFNAB- patients had significantly higher median baseline MELD score at the time of inclusion on the waiting than PFNAB+ patients (13.2 vs. 11.9, p=0.019). Among the 309 patients listed for LT, dropout related to tumour progression was observed in 31 (10%) and was not significantly different between PFNAB+ and PFNABpatients (6% vs. 11%, p=0.424). No PFNAB-related complications led to dropout. On an intention-to-treat basis, 1-, 3- and 5-year overall (89%, 69% and 60% vs. 85%, 67% and 61%, p=0.601) and disease-free survivals (84%, 67% and 59% vs. 81%, 65% and 57%, p=0.446) were not significantly different between PFNAB+ and PFNAB- patients. In 4 (5%) PFNAB+ patients, histological analysis was negative and these patients were initially not listed for LT. Among them, 3 (3.7%) developed tumour size increase during follow-up and 2 were secondarily listed and successfully transplanted. The remaining patient was listed but could not be transplanted due to tumour progression while on the waiting list. Transplanted patients (n=278) As presented in Table 1, the 75 PFNAB+ and the 203 PFNAB- transplanted patients shared similar demographic and tumour characteristics on imaging as well as types of underlying liver disease. PFNAB+ patients displayed significantly less frequently single lesions compared to PFNAB- patients. However, pathological analysis of the explants showed

7

comparable tumour characteristics (detailed in Table 2) and similar cirrhosis rate (97 vs. 99%, p=0.92) between the two groups. In the PFNAB+ group, 3 patients (4%) experienced pain after the procedure including 2 (2%) with Child C cirrhosis and low platelet count who showed a sub-capsular hematoma on CT scan. None of them experienced intra-peritoneal bleeding. On final postoperative pathological examination 41 (14.7%) patients displayed no feature of HCC including 11 (4%) with other types of lesions. Complete tumour response rates were similar between the 60 PFNAB+ and 135 PFNAB- transplanted patients treated before LT (15% vs. 16.3%, p=0.819). Among the 11 patients with other types of lesions, 10 were in the PFNAB- group compared to one in the PFNAB+ group (p=0.172). These included dysplastic nodules (DN) in 5 patients (3 high-grade DN and 2 low-grade DN), regenerative nodules in 2 patients, absence of lesion in 1 patient, cholangiocellular carcinoma in 2 patients, and hemangioma in one patient. Of these 11 patients with non-HCC lesions, 8 (72.8%) had solitary nodules measuring 2 to 3 cm. After a median follow-up of 34 months (24-135), 25 (9%) patients experienced tumour recurrence after LT. The rate of tumour recurrence was similar between PFNAB+ and PFNAB- patients (8.9% vs. 8.8%, p=0.780). Median time to recurrence was 28 months (range: 3-75) and was similar in PFNAB+ and PFNAB- patients (26 vs. 24 months, p=0.721). Locations of recurrences are detailed in Table 2. Recurrences were solitary in 11 (44%) patients, involved more than one organ in 12 (48%) patients including the lungs in 14 (56%) cases, the bones in 9 (36%) cases and the liver 8 (32%) cases. None of the patients had exclusive intrahepatic recurrence. Three (1%) patients experienced exclusive parietal recurrence including 2 PFNAB- patients who had undergone preoperative radiofrequency ablation and 1 PFNAB+ patient (p=0.797). All these 3 patients with solitary parietal recurrence were treated with

8

surgical resection and survived 29, 34 and 45 months respectively after diagnosis of the recurrence (Table 3). Patients without RFA had similar recurrence rate than patients treated with RFA (10 vs. 9%, p=0.92). On multivariate analysis, PFNAB was not significantly associated with recurrence (HR: 2.73 (0.21-8.44), p=0.77). Overall survival rates at 1, 3 and 5 years were neither significantly different between overall PFNAB+ and PFNAB- transplanted patients (94%, 72% and 63% vs. 91%, 73% and 63%, p=0.895, Figure 1).

9

DISCUSSION While the exact influence of PFNAB on the risk of tumour recurrence following LT has to date remained unclear, the present study highlights for the first time that PFNAB did not impair the oncologic prognosis of patients eligible for LT on an intention-to-treat basis. In a context of organ shortage, the primary goal of performing PFNAB in patients with suspected transplantable HCC would be to prevent unnecessary LT in patients with wellcompensated cirrhosis [6]. In the present study, only 4% of the patients who did not undergo PFNAB displayed other lesions than pure-HCC on liver explants (mainly dysplastic or regenerative nodules) and especially those with isolated lesions of 2-3 cm. This low rate could be explained by excellent radiological assessment. While PFNAB could be of use in such context, several authors have emphasized that its use was also limited by high false negative rates, reaching up to 30% [13-14]. Even, if the current study was not designed to assess the accuracy of PFNAB in the assessment of HCC, failure of PFNAB was observed in only 4 (5%) patients and negatively influenced the outcome of two (2.5%). This finding might be related to the repetition of the procedures during a strict follow-up in case of limited type and size of tumour sampling or inconclusive pathological results as previously reported [6]. Indeed, the main result of the present study was that PFNAB neither negatively influenced the short nor the long-term outcomes of patients qualifying for LT. Firstly, PFNABrelated complications rarely occurred (2.5%) and never discarded patients from LT in the present series. Accordingly even-though PFNAB can be particularly difficult, especially for small lesions, the present study confirms that PFNAB can be safely carried out until reaching adequate diagnostic accuracy. Secondly, while several authors argue that PFNAB could favour tract seeding or hematogenous dissemination of tumour cells [6-8], we observed similar dropout rates as well as identical survivals in PFNAB+ and PFNAB- patients on an

10

intention to treat basis. This finding suggests that PFNAB neither affected the oncologic prognosis of patients qualifying for LT including those dropping out while on the waiting list and those who actually underwent LT. Likewise, we observed both identical local and systemic recurrence rates following LT regardless of the existence of PFNAB and none of these recurrences occurred exclusively in the liver. These results are in line with previous studies reporting that recurrences following LT are a more a consequence of preoperative microinvasive tumour cell dissemination [15-16]. Altogether, results of the current study support that avoiding PFNAB is neither technically nor oncologically relevant and that the balance between the risks of the procedure, of needle-tract recurrence and systemic disseminations as well as the risk of a pointless and demanding surgery associated with a mortality of approximately 10% [17] should clearly be in favour of PFNAB. In addition, recent studies have emphasized that the assessment of both tumoral and peri-tumoural gene-expression signatures could improve the diagnosis of vascular invasion beyond clinical variable-based prediction [18]. Even though the purpose of this study was not to address these issues, it is likely that obtaining tumour signature using PFNAB in order to perform relevant candidate selection before LT is a path worth exploring. The current study has some limitations. First, the sample size was quite small but nevertheless represents the largest collection of patients with PFNAB prior LT reported to date. Secondly, the major limitation of our study is that the two groups did not present completely comparable tumour stage. This is observation, is likely a consequence of the challenging morphological characterization of small well-differentiated HCC (<2cm) leading to considering PFNAB in patients with earlier stage tumours. Nevertheless, matching the two groups for lesions of identical stage would have precluded relevant analysis on an intention to treat basis, which best represents the true clinical situation faced in daily practice. Finally,

11

we did not analyze the use of coaxial technique of biopsy allowing multiple samples without repeated placement of the needle; thus, potentially reducing the risk of needle tract seeding. This approach helps to reduce the number of inadequate biopsies. Additionally, the track left after biopsy with these larger needles would be embolized with fibrin sealant as a plug. In conclusion, this series shows for the first time that PFNAB does not affect the oncologic outcomes of patients with transplantable HCC. In this setting, there is currently no argument to restrict PFNAB in doubtful situation. Nevertheless, when no reasonable doubt about HCC diagnosis is raised by imaging, the risk of seeding should not be neglected and should not prime the collection of tissue biopsies without proper information to patients.

12

REFERENCES 1 Clavien PA, Lesurtel M, Bossuyt PM, Gores GJ, Langer B, Perrier A, et al. Recommendations for liver transplantation for hepatocellular carcinoma: an international consensus conference report. Lancet Oncol 2012;13:e11-22. 2 Llovet JM, Burroughs A, Bruix J. Hepatocellular carcinoma. Lancet 2003;362:1907-17. 3 Bruix J, Sherman M, Practice Guidelines Committee, American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma. Hepatology 2005; 42: 1208-1236. 4 Compagnon P, Grandadam S, Lorho R, Turlin B, Camus C, Jianrong Y, et al. Liver transplantation for hepatocellular carcinoma without preoperative tumour biopsy. Transplantation 2008;86:1068-76. 5 Kim H, Choi GH, Na DC, Ahn EY, Kim GI, Lee JE, et al. Human hepatocellular carcinomas with "Stemness"-related marker expression: keratin 19 expression and a poor prognosis. Hepatology 2011;54:1707-17. 6 Durand F, Regimbeau JM, Belghiti J, Sauvanet A, Vilgrain V, Terris B, et al. Assessment of the benefits and risks of percutaneous biopsy before surgical resection of hepatocellular carcinoma. J Hepatol 2001;35:254-8. 7 Silva MA, Hegab B, Hyde C, Guo B, Buckels JA, Mirza DF. Needle track seeding following biopsy of liver lesions in the diagnosis of hepatocellular cancer: a systematic review and meta-analysis. Gut 2008;57:1592-6. 8 Stigliano R, Marelli L, Yu D, Davies N, Patch D, Burroughs AK. Seeding following percutaneous diagnostic and therapeutic approaches for hepatocellular carcinoma. What is the risk and the outcome? Seeding risk for percutaneous approach of HCC. Cancer Treat Rev 2007 ;33:437-47. 9 Hemming AW, Cattral MS, Reed AI, Van Der Werf WJ, Greig PD, Howard RJ. Liver transplantation for hepatocellular carcinoma. Ann Surg 2001;233:652-9. 10 Stigliano R, Burroughs AK. Should we biopsy each liver mass suspicious for HCC before liver transplantation?--no, please don't. J Hepatol 2005;43:563-8. 11 Roche B, Samuel D. Liver transplantation in viral hepatitis: prevention of recurrence. Best Pract Res Clin Gastroenterol 2008;22:1153-69. 12 Fuks D, Dokmak S, Paradis V, Diouf M, Durand F, Belghiti J. Benefit of initial resection of hepatocellular carcinoma followed by transplantation in case of recurrence: an intention-to-treat analysis. Hepatology 2012;55:132-40. 13 Durand F, Belghiti J, Paradis V. Liver transplantation for hepatocellular carcinoma: role of biopsy. Liver Transpl 2007;13:S17-23. 14 Gilmore IT, Burroughs A, Murray-Lyon IM, Williams R, Jenkins D, Hopkins A. Indications, methods, and outcomes of percutaneous liver biopsy in England and Wales: an audit by the British Society of Gastroenterology and the Royal College of Physicians of London. Gut 1995;36:437-41. 15 Sharma P, Welch K, Hussain H, Pelletier SJ, Fontana RJ, Marrero J, et al. Incidence and risk factors of hepatocellular carcinoma recurrence after liver transplantation in the MELD era. Dig Dis Sci 2012;57:806-12. 16 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 2009;10:35-43.

13

17 Lesurtel M, Raptis DA, Melloul E, Schlegel A, Oberkofler C, El-Badry AM, et al. Low Platelet Counts after Liver Transplantation Predict Early Post-Transplant Survival: The "60-5 Criterion". Liver Transpl 2013 Oct 3. 18 Rodríguez-Perálvarez M, Luong TV, Andreana L, Meyer T, Dhillon AP, Burroughs AK. A systematic review of microvascular invasion in hepatocellular carcinoma: diagnostic and prognostic variability. Ann Surg Oncol 2013;20:325-39.

14

Table 1. Preoperative characteristics of 75 transplanted HCC patients with pre-LT tumour biopsy (PFNAB group) and 203 patients without (No PFNAB group)

Gender (Male) Age (year)* Baseline MELD score* Tumour on imaging Milan criteria on imaging Diameter** Diameter >3 cm Number of nodules* Single nodule Macroscopic vascular invasion Median α-fetoprotein level** α-fetoprotein >400 Underlying liver disease HCV HBV Chronic alcohol consumption NASH Hemochromatosis Cirrhosis (F4) Time on waiting list (months)** Treatment on waiting list RFA TACE Resection

PFNAB+ group (n=75) 66 (88) 55±8 11.5±5

PFNAB- group (n=203) 172 (85) 54±7 13.0±5

p-value

61 (81) 22 (2-70) 29 (44) 1.7±0.8 35 (47) 0 (0) 9 (1-170340) 6 (8)

161 (79) 27 (5-70) 87 (43) 1.5±0.9 122 (60) 1 (1) 9 (1-12494) 12 (6)

0.676 0.040 0.529 0.123 0.045 0.787 0.211 0.729

30 (40) 18 (24) 26 (35) 4 (5) 3 (4) 73 (97) 7.0 (3-13) 60 (80) 39 (52) 34 (45) 8 (11)

78 (38) 33 (16) 88 (43) 11 (5) 4 (2) 201 (99) 6.9 (4-22) 135 (66) 65 (32) 75 (37) 40 (20)

0.455 0.098 0.121 0.623 0.287 0.921 0.891 0.097 0.072 0.117 0.096

0.312 0.214 0.019

PFNAB: preoperative fine-needle aspiration biopsy; *mean± SD; **mean (range); MELD : model for end stage liver disease ; HCV: chronic hepatitis C infection; HBV: chronic hepatitis B infection; NASH: non-alcoholic steatohepatitis; RFA : radiofrequency ablation ; TACE : trans arterial chemoembolization

15

Table 2. Pathological characteristics and recurrence pattern of 75 transplanted HCC patients with pre-LT fine-needle aspiration biopsy (PFNAB group) and 183 patients without (No PFNAB group)

HCC absent Tumour on pathology Milan criteria on explant Diameter (mm)* Diameter >3 cm Number of nodules* Single nodule Poor differentiation Satellites nodules Microvascular invasion Recurrence Delay (months)** Location Liver Lung Bone Peritoneal Parietal

PFNAB+ group (n=75)

PFNAB- group (n=183)

p-value

54 (72) 28 (2-73) 32 (43) 2.7±1.8 24 (32) 5 (7) 10 (13) 18 (24) 7 (9) 23 (3-56)

142 (77) 29(4-72) 81 (44) 2.6±1.9 80 (44) 6 (3) 36 (20) 43 (23) 18 (9) 21 (1-75)

.223 .747 .814 .638 .265 .089 .094 .542 .531 .563

3 (4) 4 (5) 3 (4) 2 (3) 1 (1)

5 (3) 10 (5) 6 (3) 2 (1) 2 (1)

.580 .504 .382 .423 .627

PFNAB: preoperative fine-needle aspiration biopsy; * mean± SD; ** mean (range)

16

Table 3. Details of the 3 patients experiencing exclusive parietal recurrence

Gender / age (years) PTB Radiofrequency ablation Tumour Within Milan criteria on imaging Largest diameter (mm)* Number of nodules* Subcapsular location Poor differentiation Satellites nodules Microvascular invasion AFP level (ng/ml) Fine needle aspiration / radiofrequency ablation Rim of non tumoral liver parenchyma Presence of ascites Number of procedures Needle Size (gauge) Recurrence Delay of recurrence (months) Survival after recurrence (months) Latest news

Patient 1 Male / 64 Yes No

Patient 2 Male / 51 No Yes

Patient 3 Male / 49 No Yes

Yes 22 3 Yes No No Yes 5

Yes 30 1 No No Yes No 9

Yes 36 1 No Yes No Yes 13

No No 1 22

Yes Yes 1 -

Yes No 1 -

19.1 45 Alive without recurrence

73.9 29 Died of recurrence

11.4 34 Alive without recurrence

17

Figure 1. Overall survival in the 278 transplanted patients, as a function of the preoperative fine-needle aspiration biopsy (PFNAB+: green, PFNAB-: blue) (Kaplan Meier, log-rank p=0.895)

PFNAB+ (n=75) Number at risk PFNAB- (n=203) Number at risk

12 months 94% 65 91% 182

36 months 72% 45 73% 123

60 months 63% 31 63% 90

18