SEMINARS IN LIVER DISEASE-VOL. 19, NO.

3, 1999

Liver Transplantation for Hepatocellular Carcinoma

HENRl BISMUTH, M.D., F.A.C.S. (HON),
PIETRO E. MAJNO,* M.D., F.R.C.S. (ENG), and RENE ADAM, M.D., Ph.D.

ABSTRACT: Liver transplantation for hepatocellular carcinoma (HCC) in patients with cirrhosis is a radical
treatment of the tumor and associated precancerous state. It is potentially curative in a proportion of patients. The
outcomes of early studies of liver transplantation in this indication were initially unfavorable. Selection of transplant
candidates at an early stage, in the absence of extrahepatic spread, gives better survival than surgical resection and
alternative nonsurgical treatments. Transarterial chemoembolization can be used for preoperative control of the dis-
ease. Adjuvant chemotherapy may be indicated in the postoperative period for the prevention of recurrence in pa-
tients with histologic features of invasiveness in the surgical specimen. Liver transplantation as the treatment oj
choice for early HCC in screening programs in cirrhotic patients may become limited by graft availability as the
numbers of hepatitis C-related cases increase. Resection may be indicated if the waiting time is likely to be long.
KEY WORDS: chemoembolization, hepatectomy, hepatocellular carcinoma, liver cirrhosis, liver transplantation,

Downloaded by: Universite Laval. Copyrighted material.

patient selection, survival analysis

Hepatocellular carcinoma (HCC) is one of the most
common tumors worldwide. In the western world the
current epidemic of cirrhosis due to the hepatitis C virus
is increasing the number of new cases.1.' HCC used to
be discovered at an advanced stage, too late for curative
treatment because of the extent of the intrahepatic
spread, the associated degradation of liver function, or
the presence of extrahepatic metastases. Awareness of
the increased risk of HCC in patients with chronic hepa-
titis or cirrhosis is changing the spectrum of presenta-
tion of the disease. An increasing number of tumors are
now discovered at an earlier stage, and treatment has the
potential to be curative.' In addition to surgical treat-
ment by liver resection and liver transplantation (LT),
evidence is accumulating for a curative role for percuta-
neous destruction by ethanol injection. Percutaneous
ethanol injection (PEI) was formerly considered to be
palliative, as are other techniques such as cryotherapy
or radiofrequency thermal destruction. Traditionally,
liver resection has been preferred for small tumors and

Objectives
Upon completion of this article, the reader should be able to: I ) appreciate the role of liver transplantation in the management of
hepatocellular carcinoma (HCC); 2) list the indications for the procedure; and 3) recognize the limitations of liver resection in HCC patients.
Accreditation
The Indiana University School of Medicine is accredited by the Accreditation Council for Continuing Medical Education to sponsor
continuing medical education for physicians.
Credit
The Indiana University School of Medicine designates this educational activity for a maximum of 1.0 hours credit toward the AMA
Physicians Recognition Award in category one.

Disclosure
Statements have been obtained regarding the author's relationships with financial supporters of this activity. There is no apparent conflict
of interest related to the context of participation of the author of this article.

From the Centre Hipafobiliaire, Assi.stcmce Publique-H6pitaux de Paris, Universiti Paris-Sud, Hripital Paul Brousse, Villejug France.
*Current address: Transplantation Unit, Department of Surgery, University Hospital, Genevam Switzerland.
Reprint requests: Professor Henri Bismuth, Centre HCpatobiliaire. HBpital Paul Brousse, 94804 Villejuif, France. E-mail: henri.
bismuth@pbr.ap-hop-paris.fr

Copyright O 1999 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel.: +1(212) 760-0888 xl32.
0272-808711 99911098-897 1 (1999) 19:03:03 11-032 1 :SLD0001OX
31 1
 SEMINARS IN LIVER DISEASE-VOL. 19, NO. 3, 1999

TABLE 1. Characteristics of 125 Patients In recent years, the indications for LT in HCC have
Transplanted for HCC at Paul Brousse Hospital, evolved. Today, LT is the treatment of choice for early
1985-1 995 HCC. This article reviews the evolution of LT in the
N Percent treatment of HCC in two stages. The experience in our
center is described first. The key issues identified, espe-
Males
Females cially those related to criteria for the selection of pa-
Mean age (range), yr tients leading to long-term survivals, are then discussed
Cau.~eofcirrl7o.si.s in the context of other data published in the literature.
Hepatitis B
Hepatitis C
Hepatitis B + C
Alcohol THE PAUL BROUSSE EXPERIENCE
Primary biliary cirrhosis
Other
Between January 1984 and December 1998, 248
Degree of cirrhosi.\
Child's A patients with HCC associated with cirrhosis have been
Child's B treated by LT at the Hepatobiliary Centre of Paul
Child's C Brousse Hospital. At the end of December 1995, 125
Turnor charrrctrri.stics
patients had been transplanted and had at least 3 years
Number of nodules
1 of follow-up. The characteristics of the patients are
2-3 shown in Table 1.
>3

Downloaded by: Universite Laval. Copyrighted material.

Maximum size (mm)
530
30-50
>50
Portal vein tumor thrombus
Trunk
Main portal branch
Sectorial/segmental branch
Alpha-fetoprotein
Normal
Increased
Pretransplant TACE
Intraoperative blood transfusion (units)
0 during which the criteria for selection of transplant candi-
1-5
5-10
210
Postoperative chemotherapy
LT for large tumors not amenable to resection. Better
knowledge of the biology of HCC and of the results of
the different treatments, however, are challenging this
attitude. Liver resection, despite the progress of hepatic
surgery, can be performed only on 5 1 5 % of patients
presenting with HCC and is limited by the diminished
functional reserve of the diseased liver.4-7 Although per-
cutaneous treatments are less limited by liver function,
the long-term results of both liver resection and PEI are
worsened by the mortality due to recurrence and the nat-
ural history of the cirrhosis, with liver failure and rates
of new cancer development as high as 20% per year in
the cancer-prone cirrhotic tissue.8-11 In contrast, LT ap-
pears to be the most radical therapy. It removes the
tumor and the underlying cirrhosis, treating the risk of
hepatic recurrence at the same time as avoiding the
mortality of complications such as hepatocellular fail-
ure and portal hypertension. It needs to be performed in
patients before extrahepatic spread has occurred.'?
Selection of Transplant
Candidates with HCC
The indications for LT were either the known pres-
ence of the tumor (92 cases, 74%) or the severity of the
underlying cirrhosis. In the latter patients, the tumor was
an incidental finding during pretransplant assessment (1 3
cases, 10%) or in the hepatectomy specimen after trans-
plantation (20 cases, 16%). The data span two periods
dates with known tumors were different. From 1985 to
199 1 , 60 patients were transplanted for tumors that could
not be resected because of poor hepatic function, multiple
nodules, or excessive size. The only contraindication in
this first period was the presence of extrahepatic spread.
Analysis of the results obtained during this first period
suggested that, of patients with known tumors, transplan-
tation was indicated for those who had smaller and fewer
tumors and no vascular involvement.'? In the second pe-
riod, LT was considered as preferable to resection. Liver
resection for HCC was performed in patients with con-
traindications for LT or in patients with easily accessible
peripheral lesions in whom the disease could progress
during the waiting period to LT. During the period 1992
to 1995,45 patients were transplanted according to these
criteria. The tumor characteristics of the patients in the
two periods are compared in Table 2.
Preoperative Assessment
In the pretransplant assessment, the points specific to
the presence of the tumor were the staging of the intra-
hepatic disease and the search for extrahepatic spread. All
patients underwent Doppler-ultrasound studies of the
LIVER TRANSPLANTATION FOR HEPATOCELLULAR CARCINOMA-BISMUTH
TABLE 2. Tumor Characteristics and Outcomes
in Patients with Known Tumors (n = 105)
Nutnber of tutnors
5 3 nodules
>3 nodules
Size
530mm
>30mm
Number cn~dsize
5 3 0 mm + 5 3 nodules
>30 mm + >3 nodules
Rer~irrence,3-year survivals
Overall
Disease free
NS. not significant
liver, focusing on portal and hepatic venous involvement,
computed tomography (CT) o f the abdomen and chest,
and a radionuclide bone scan. Metastases to other sites
are sufficiently rare not to warrant screening other than
by clinical examination. The protocol for CT was system-
atized with the use of new generation spiral CT, with an
unenhanced study o f the liver and o f the chest, followed
by an arterial (30 seconds) and venous (120 seconds) spi-
ral on the liver. In 65 patients (2%) with preserved he-
patic function (Child's A), and in the absence o f other
contraindications such as hepatofugal portal flow or pe-
ripheral vascular disease, transarterial chemoemboliza-
tion (TACE) with lipiodol and cisplatin was performed
with the double aim o f diagnosing occult nodules and o f
achieving preoperative control o f the disease.'3 PEI had
been performed in four patients (3%). No biopsy o f the
tumor was performed in our unit to avoid needle tract dis-
semination, although in some cases biopsies had been
performed before referral. The general evaluation of the
patient in the pretransplant assessment was aimed at
excluding or treating the usual contraindications to trans-
plantation, such as hepatitis B viral replication, cardio-
pulmonary disorders, persistent alcohol abuse, psychiaric
disease, or anticipated noncompliance. Patients were put
on the routine waiting list, with no particular priority as-
signed to the indication o f LT for HCC.
Operative Technique
Laparotomy at the time o f transplantation was used
as the final confirmation that LT was not contraindi-
cated. A second recipient was kept on stand-by in case
extrahepatic tumor was found. This was preferred to the
alternative policy o f performing an exploratory laparo-
tomy during the pretransplant evaluation. The advan-
tage is to avoid a futile pretransplant assessment whose
negative findings become obsolete during the wait for
ET AL.
transplantation, or which results in adhesions that ren-
der the final operation more difficult.
Downloaded by: Universite Laval. Copyrighted material.
The transplant operation was started with a limited
incision only, and exploration o f the peritoneal cavity
was carried out first. Lymph nodes in the hepatic pedicle
were systematically resected and examined with frozen
sections. Tumoral thrombosis in the portal trunk discov-
ered preoperatively or intraoperatively was considered
as a contraindication to proceeding with transplantation
during the second part o f the study (1992-1995). In to-
tal, 9 o f 134 patients (7% o f the eligible patients) were
not transplanted for contraindications that emerged dur-
ing the laparotomy (peritoneal nodules in 3 cases, lymph
node invasion, neoplastic thrombosis in the portal vein,
and abdominal wall invasion in 2 cases each) and were
excluded from the analysis.
The cell saver was not used until after the hepatec-
tomy to decrease the risk o f disseminating tumor cells
that may be shed during manipulation o f the liver. Ex-
tracorporeal bypass was used only after clamping o f the
portal vein to avoid dislodging intrahepatic tumoral
cells during aspiration o f the portal trunk. To achieve
better tumor clearance, the retrohepatic vena cava was
resected. Transplantation preserving the inferior vena
cava was performed only in exceptional circumstances.
Postoperative Care
Postoperative care and immunosuppression were
identical to LT performed for other indications. Chemo-
therapy with doxo~ubicinand 5-fluorouracil was given
in the presence o f adverse features on histology (mi-
crovascular invasion, satellite nodules, absent or in-
vaded tumor capsule) as soon as the patient could toler-
ate it, for a total o f nine cycles, unless complications
delayed or impeded its administration. Chemotherapy
was given to 5 1 patients (41%). Alpha-fetoprotein blood

TABLE 3. Outcome of Transplanted Patients
as a Function of the Mode of Tumor Discovery
Mtrirz Irzrlictrtion
or1 R<+rrcrl
Number of cases
Mortality 1 2 mo
Recurrence
Dead. with recurrence
Dead. no recurrence
Alive, with recurrence
Alive, no recurrence
Mean follow-up (mo)
levels and liver ultrasound were performed every 3
months. Abdominal and chest CTs were performed
every 6 months during the initial 2 years of follow-up.
No patient was lost to follow-up, with a median length
of follow-up of 3.8 years (range, 0-1 1 years; interquar-
tile range, 1.4-6.2 years).
Outcomes
Perioperative Mortality
Four patients died within 2 months of LT (3%): one
from an intraoperative cardiac arrhythmia, one from
cardiac failure on day 1, one of bacterial sepsis after re-
transplantation for primary nonfunction on day 1 1 (day
3 of the retransplant), and one patient on day 1 1 with dif-
fuse aspergillosis.
Surgical Complications
Arterial complications occurred in seven patients
(6%), comprising two thromboses, two anastomotic
strictures, and three anastomotic aneurysms. Biliary
complications occurred in 17 patients (14%) (stenosis in
1 1 cases, biliary fistula in 2 cases, and biliary peritonitis
after removal of the T-tube in 4 cases).
Retransplantation
Eleven patients (9%) were retransplanted. The indi-
cations were in all cases reasons unrelated to the tu-
moral disease. The median delay was 5 months (range, 3
days to 95 months).
Recurrence of HCC
This was observed in 26 patients (21%) at a mean
of 18 months after transplantation (range, 2-94
months). Most recurrences were observed during the
first year (13126, 50%) or the second year (8126, 31%).
SEMINARS IN LIVER DISEASE-VOL. 19, NO. 3,1999
Incidental
Findirzg during Po.sttrun.splant
A.s.srs.srneizt Histolo~y
Only a few cases were observed after this time (5126,
19%). The initial sites of recurrence were, in decreasing
order of frequency, lung (8 cases, 31%), liver (6 cases,
23%), bone (5 cases, 19%), lymph nodes (4 cases, 15%)
and adrenals (3 cases, 12%).
Table 3 shows that recurrence was observed more
Downloaded by: Universite Laval. Copyrighted material.
frequently in cases where the tumor was the main indi-
cation for transplantation (25192, 27%). In fact, there
was only one recurrence among the 33 patients in whom
the tumor was not the main indication (3%, p = 0.001).
This was in 1 of 20 patients in whom HCC was discov-
ered on pathologic examination of the explanted liver.
There was no recurrence in the 13 cases in whom the tu-
mor was discovered during the pretransplant evaluation.
Survival
Of the initial 125 patients transplanted, with a mean
follow-up of 50 months (range, 2-138), 53 have died
(42%) and 72 are alive (58%), of whom 71 are without
recurrence. The actuarial survival for the whole group
was 80% at 1 year, 65% at 3 years, and 58% at 5 years
(Fig. l), with respective disease-free survivals of 74%,
62%, and 57%.
Factors Related to Survival After LT
Four factors were correlated to survival after trans-
plantation: the mode of presentation of the tumor, the
size of the nodules, the number of nodules, and the in-
volvement of the intrahepatic blood vessels (Table 4).
Mode of Presentation
Survival was similar in patients in whom the tumor
was discovered incidentally during the pretransplant as-
sessment and in patients in whom the tumor was discov-
ered on pathologic examination of the liver, and this
was greater than in cases where the tumor represented
the indication for transplantation (Fig. 2).
LIVER TRANSPLANTATION FOR HEPATOCELLULAR CARCINOMA-BISMUTH ET AL.

125 Patients
100
A % surviving

Downloaded by: Universite Laval. Copyrighted material.

01
0 1 2 Years 3 4 5
FIG. 1. Actuarial survival of patients with HCC and cirrhosis transplanted from November 1984 to December 1995.

Size of the Tumor mediate categories (with tumors larger than 30 mm

but with no more than three nodules or with more than
Patients whose tumors were smaller than 30 mm three nodules not larger than 30 mm), survival was
had a significantly better survival than patients with tu- intermediate.
mors between 30 and 50 mm or larger than 50 mm ( 5 -
year survivals, 70% vs. 41% and 35%, respectively; p
< 0.01).

TABLE 4. Risk of Recurrence

Number of Nodules as a Function of Tumor Characteristics
Number of Rec~trrences
Patients with single tumors had a better survival p r CNSCS Observerl I'
than patients with two or three nodules and than patients
Size
with more than three nodules (5-year survival, 68%, <30 mm
58%, and 42%, respectively; not significant). The dif- 3 1-50 mm
ference was statistically significant when patients with >50 mm
one to three nodules were compared with patients with Number of nodules
1
more than three nodules. 2-3
>3
Size and number
Combination of Size and Number 5 3 0 mm and 5 3
>30 mm and 5 3
5 3 0 mm and >3
Patients with small tumors (<30 mm) and no more >30 mm and >3
than three nodules had a markedly better survival than Portal tumor thrombus
patients with larger (>30 mm) and multinodular tumors Absent
Sectorial-segmental
(more than three nodules) (5-year survival, 74% vs.
Trunk-main branch
38%; p < 0.001). For patients with disease in the inter-
3 16
% surviving disease-free
100
80
60
0 1 2 Years
FIG. 2. Actuarial survival after LT for HCC according to the mode of tumor discovery. Discovery on pathologic examination of
the hepatectomy specimen after transplantation (Histological); discovery during the pretransplant assessment in a patient with de-
compensated cirrhosis (Incidental); cases in whom the tumor was the main indication for transplantation (Main).
Portal Tumor Thrombus
Survival was significantly worse when the tumor
was associated with portal vein tumor thrombus, con-
firmed on histology ( p = 0.002), whether distal (seg-
mental or sectorial branch, 5-year survival, 28%) or
proximal (portal trunk or main branch, 5-year survival,
0%) as compared with patients without portal vein in-
volvement (5-year survival, 68%).
Impact of the Restrictive Selection Criteria
The results observed in the second period were bet-
ter than the results of the first period (Table 2, Fig. 3) in
terms of prevention of recurrence (11% vs. 33%), of
overall survival (76% vs. 53%), and of disease-free sur-
vival (74% vs. 43%).
LT FOR HCC: BACKGROUND
AND LITERATURE ANALYSIS
Unresectable hepatic malignancy was one of the
first indications for LT. The initial results were poor due
SEMINARS IN LIVER DISEASE-VOL. 19, NO. 3, 1999
p=NS
Histological finding (n = 20)
Incidental finding (n = 13)
53 %
Main indication (n = 92)
Downloaded by: Universite Laval. Copyrighted material.
3 4 5
to the high mortality associated with the procedure,
ranging between 24% and 70%, and recurrence of the
tumors (Table 5).'4-16.19 At that time, transplantation
was performed for tumors that were unresectable be-
cause of their large size or because they were multiple.
Both are factors that are correlated with the incidence of
distant spread.20 Patients with lymph node invasion
were included in some series, up to 25% of the cases in
one of the largest centers." Immunosuppression, favor-
ing proliferation of occult malignant cells already pres-
ent in extrahepatic sites before LT, was considered in
part responsible for the very high incidence of recur-
r e n ~ e . ~ ?A. 'period
~ of disillusion followed, in which the
role of LT for malignant disease was seriously ques-
tioned. LT for cancer in Europe fell from 29% of all
indications in the period 1983 to 1987 to 15% in the
period 1988-1992.24 Some large centers, however, had
drawn attention to the good outcome of patients trans-
planted with tumors discovered only on the hepatec-
tomy specimen, a factor probably related to their small
size and to the low occurrence of vascular invasion and
distant spread in the early stages.16
The first report suggesting that better patient selec-
tion was the key to improving the outcome of LT for
HCC came from the retrospective analysis of the results
LIVER TRANSPLANTATION FOR HEPATOCELLULAR CARCINOMA-BISMUTH
% surviving disease-free
100
80
60
40
0
0 1
FIG. 3. Actuarial survival after LT for HCC in periods with different patient selection criteria. In 1992 a policy was imple-
mented favoring LT in patients with up to three tumor nodules, up to 30 mm in size, and no tumour involvement of the portal vein.
in 60 patients transplanted in our unit between 1985 and
1991.12 These patients underwent LT, rather than liver
resection, because of poor liver function or because of
multifocal disease. In all patients, extrahepatic disease
had been excluded according to the protocol detailed
above. The results of LT were compared with the results
of liver resection in 60 patients who were operated dur-
ing the same period. This study identified tumor size,
number of foci, and vascular invasion as the main deter-
minants for recurrence. It showed that patients with
small HCC (<3 cm) and one or two tumors, traditionally
considered the best candidates for liver resection, had a
significantly better disease-free survival with LT (83%
at 3 years for LT vs. 18% at 3 years for liver resection)
because of the high occurrence of de novo tumors in the
residual cirrhotic liver after resection. The importance
of tumor size, tumor number, and vascular invasion was
confirmed prospectively in the second period of the ex-
perience of LT for HCC at Paul Brousse, where a dis-
ease-free survival of 73% at 5 years was obtained in 45
patients selected, taking into account the abovemen-
tioned criteria.
Results comparable with ours have been reported
by other teams applying the same, or a similar, restric-
tive poiicy to the selection of patients with HCC for LT
(Table 4) and obtaining a similarly low intraoperative
ET AL.
1992-1995 (n = 45)
Downloaded by: Universite Laval. Copyrighted material.
I
2 Years 3 4 5
mortality. The group from the National Cancer Institute
in Milan reported a 4-year survival of 85% in a group of
patients transplanted for solitary tumors less than 5 cm
in size or with less than three tumor nodules, none larger
than 30 mm. A report from the International Registry of
Hepatic Tumors in Liver Transplantation confirmed the
importance of tumor size and vascular invasion and
drew attention to the possible impact of tumor grade,
with patients with grade 3 or 4 having a significantly
worse outcome.40 However, the number of patients with
poorly differentiated tumors was small (261410). Histo-
logic grade was not reported for 55% of patients in the
study, and the slides were not reviewed by a centralized
pathology service. The suggestion that tumor biopsy
should be performed to select patients with better out-
come needs to be confirmed with more stringent data
collection. We remain concerned by this practice be-
cause of the risk of tumor dissemination along the
biopsy tract.
Resection versus Transplantation in HCC
The good results of LT for HCC put the role of liver
resection, traditionally the preferred treatment in pa-
tients with smaller tumors, into a different perspective.
 SEMINARS IN LIVER DISEASE-VOL. 19, NO. 3, 1999

TABLE 5. Main Series in Liver Transplantation for HCC

lwatsuki ( 16) 25 -
O'Grady (17) 32 32
Ringe ( 18) 20 20
lsrnail (19) 0 0
Ringe (2 1 ) 15 15
Iwatsuki (25) 52 49
Haug (26) 32 -

Mc Peake (27) - -
Bismuth ( 12) 49 -
Chung (28) 46 -
Romani (29) 71 -

Schwartr (30) 60 -
Selby (3 1 ) 39 36
Tan (32) 63 -
Olthoff (33) 46 -
Mazraferro (34) 83;(4 yr) -
Figueras (35) 75 63
Colella (36) 72 68
Bechstein (37) - 6077
Otto (38) 53'3 -

Downloaded by: Universite Laval. Copyrighted material.

Llovet (39) 74 74
Prcsent series 68 59
7htT 74:'

"Includes patients with and without cirrhosis.

'Mortality at 30 days
?Mortality at 90 days.
Tumors discovered after hepatecto~nyexcluded.
Mortality at 60 days.
'llncludes only patients with tumors 4 0 mm in sire
*'4'Hospitalmortality.
,iDisease free.
-:Thirty-six percent of patients with turno1.s >5 cm.
"Patients transplanted 1992- 1995.

Liver resection, even when performed only on patients
with preserved hepatic function, is followed by a yearly
incidence of recurrence in the residual liver of 20 to
25% in most series in the literature.8-1' Although it is
possible that this may be diminished by the use of drugs
favoring cellular differentiation or by the selection of a
subgroup of patients at a lower risk of recurrence, there
is the long-term mortality due to the complications of
the underlying liver cirrhosis."-1'
Most studies performing a direct comparison
between liver resection and LT suffer from the meth-
odologic bias of a nonrandomized allocation of pa-
tients with early HCC to the different treatment
groups.l'.".38.11.Q Despite this, the difference in disease-
free survival for patients with small tumors in our expe-
rience is so striking (3-year disease-free survivals of
83% for transplantation vs. 18% for resection in patients
with one or two tumors 1 3 cm) that it makes a random-
ized trial protocol ethically questionable. The superior-
ity of LT in the treatment of small HCC has tended to
confine liver resection in our unit to the treatment of pa-
tients with contraindications to transplantation, such as
advanced age, hepatitis B viral replication, associated
extrahepatic diseases, or psychosocial factors, or to pa-
tients with easily accessible solitary tumors and good
hepatic function in whom the risk of liver resection is
low. In this group, the shortage of grafts is a factor that
favors resection. Some patients may become candidates
for LT in the event of a recurrence.
Preoperative and Postoperative
Antitumoral Treatments
Whether or not patients with HCC who are consid-
ered for LT should undergo some form of pretransplan-
tation treatment of their cancer is still debated. Because
the recurrence of HCC after LT depends mainly on ex-
trahepatic spread, there is little reason to believe that
preoperative control of the hepatic disease would affect
the outcome. Furthermore, because patients normally
considered for LT have small tumors in which extrahe-
patic spread is unlikely, a survival benefit is difficult to
demonstrate. Some authors report a survival advantage
in patients treated with TACE and PEI before transplan-
tation." Others do not confirm a benefit of preoperative
LIVER TRANSPLANTATION FOR HEPATOCELLULAR CARCINOMA-BISMUTH
% surviving
100
80
60 No TACE (n = 22)
40
TACE Downstaging - (n = 16)
Years
FIG. 4. Actuarial disease-free survival in patients with and without chernoembolization. Patients had an initial tumor size >31
mm. TACE Downstaging + indicates those in whom transarterial lipiodol chemoembolization was followed by a significant reduction
in tumor size. TACE downstaging-indicates those without significant response in size.
TACE.13J4,44Some authors conclude that this treatment
is unwarranted.44 Deeper analysis of the reports, how-
ever, leads to a more balanced view. For patients with a
tumor larger than 3 cm, a survival advantage was seen
after response to TACE in our center.lWf 19 of such
patients, 5 patients died, only 3 of recurrence (2 with
portal vein thrombosis, the other with extensive multi-
nodular disease), for a cumulative disease-free survival
at 5 years of 70%, similar to the disease-free survival of
patients with tumors <3 cm in size (Fig. 4). Oldhafer et
al. did not show a survival advantage in 21 patients un-
dergoing LT after TACE compared with 21 historical
control subjects. These patients were not matched for
tumor size but to tumor TNM stage, which is poorly
correlated to outcome of LT for HCC.39.45 Recurrence
was reported in only 10% of the patients in both groups,
and survival in the TACE group was penalized by three
cases of pneumonia that were related to TACE, a com-
plication that has not been observed by other groups.
Also, TACE was carried out with doxorubicin rather
than with cisplatin, a less effective regimen in previous
studies and in our experience." Factors other than pa-
tient survival after LT need to be taken into account to
decide whether preoperative treatment is warranted.
The delay before an organ is available may be impor-
tant. Patients with blood group 0 or B may wait more
ET AL.
TACE Downstaging + (n = 19)
29 %
Downloaded by: Universite Laval. Copyrighted material.
than a year before an organ is available. The proportion
of patients excluded from transplantation because of tu-
mor progression should be considered in evaluating the
impact of pretransplant treatment, as discussed by
Sarasin et al.46 In practice, it is reasonable to offer some
form of tumor control in patients within the traditionally
recognized inclusion criteria (less than three tumors <3
cm or a single tumor <5 cm) who are unlikely to
undergo LT for HCC within 6 months. Whether tumor
control should be obtained by liver resection, radiofre-
quency thermal destruction, PEI, TACE, or a combina-
tion of these methods should be further investigated.
Also, in patients with preserved liver function, the feasi-
bility of a policy of primary liver resection or percuta-
neous destruction and salvage LT for recurrence may be
worth investigating with regard to both patient survival
and graft saving. Liver resection did not appear to jeop-
ardize the results of a subsequent LT for recurrence in a
small series of patients in our unit.
Chemotherapy is used to complement LT for HCC in
most centers reviewed by Cherqui?' mostly postopera-
tively and less frequently preoperatively. The results of
pre- and postoperative chemotherapy with doxorubicin
were reported in series of 20 patients transplanted for
large HCC by Stone et al.48 The disease-free survival of
the 17 patients with tumors larger than 5 cm was 70% at 1

year and 56% at 3 years, leading the authors to recom-
mend such treatment for high-risk patients. A similar pos-
itive effect was reported by Olthoff et al.," with a sur-
vival of 46% at 3 years for 25 patients transplanted for
unresectable HCC, a recurrence rate of 20%, and 3 pa-
tients with tumors >5 cm having long-term disease-free
survival. Although no controlled series is available con-
firming the advantages of postoperative chemotherapy, at
Paul Brousse Hospital doxorubicin and 5-fluorouracil are
given to patients with unfavorable histologic characteris-
tics. Chemotherapy is generally well tolerated, but in our
series it contributed to one death from neutropenic sepsis.
LT in Patients with Borderline
or Advanced Tumors
The commonly accepted criteria for transplantation
for HCC exclude a large number of patients (most in
some centers) in whom LT still represents the best treat-
ment possible, but it cannot be offered because the long-
term results are inferior to the results of patients with be-
nign disease. They compete for the same limited pool of
donor organs. From among the patients with apparently
unfavorable tumor characteristics, it is important to iden-
tify the subgroup in whom the tumor has a low malignant
potential and the results of LT are acceptable. It is in this
group that preoperative or postoperative treatments will
have a greater impact and that further study is needed.
Response to TACE was a favorable factor in our
unit. In cases where preoperative treatment offers pro-
longed control of the tumor, such treatment may help in
the selection of patients outside the commonly accepted
criteria by allowing a period of observation during
which the biology of the tumor may become more evi-
dent. Other parameters may offer further clues. The vi-
ral status of the host does not appear correlated to the in-
cidence of recurrence or survival." The absence of
estrogen receptors in the tumor is associated with a
more favorable o ~ t c o m e . " ~
To minimize the inappropriate use of grafts, it is ad-
visable that patients outside the accepted criteria for LT are
treated in strict protocols, possibly with the use of marginal
organs such as from hepatitis C virus-positive donors or
from donors with neurologic malignancies. More recently,
we have used the livers explanted from amyloid neuropa-
thy patients in domino transplantation pr0cedures.5~
Treatment of Recurrence
Even with stringent selection criteria, recurrence
will be observed in 10 to 15% of the patients trans-
planted for HCC. In published series, this will be ob-
served in the transplanted liver or in the lung in approxi-
mately one fourth to one fifth of the cases and less
frequently in the lymph nodes, bone, and adrenals. In a
SEMINARS IN LIVER DISEASE-VOL. 19, NO. 3, 1999
recent publication collecting the recurrences of three
major centers in Northern Italy, 21 patients had a recur-
rence after an average of 7.8 months ( 1 to 25 months).
Although overall the outcome was poor, seven patients
with resectable recurrences had a 4-year survival of
57%, and one case is still alive without evidence of dis-
ease 6 years after resection of a recurrence in the liver."
CONCLUSIONS
The recent experience in LT for HCC at Paul
Brousse Hospital and in other centers allows the defini-
tion of a subgroup of patients with small tumors (up to
30 mm or 50 mm if solitary), no more than three nod-
ules, and the absence of portal vein tumor thrombus. In
these patients LT offers probabilities of survival similar
to patients with benign liver disease. LT is indicated in
these patients, even in the current climate of graft short-
age, because the results of LT are better than those of
Downloaded by: Universite Laval. Copyrighted material.
liver resection. Resection has a role restricted to the
treatment of patients in whom LT is otherwise con-
traindicated or as a first-line treatment in patients with
easily accessible tumors and good liver function for
whom a long wait for a graft is predictable.
Unfortunately, LT is possible only for a small propor-
tion of patients with HCC because in most cases the dis-
ease is discovered at an advanced stage. Further progress
can be expected by more widespread diffusion of screen-
ing programs for the detection of tumors at an earlier stage
in patients with cirrhosis, by further refining the criteria
for the selection of patients at low risk of recurrence to
widen the indications for transplantation, and by the use of
effective methods of tumor control before and after the
procedure. We anticipate that as this progress is made,
graft shortage will represent the main limit to the success-
ful treatment of HCC with LT. New means to increase the
number of available grafts, such as an expansion in the use
of split or "domino" LT, living donor LT, or treatment
strategies that reserve LT for recurrence after local treat-
ment of the tumors by resection or percutaneous tumor de-
struction, will need to be explored.
ABBREVIATIONS
LT liver transplantation
HCC hepatocellular carcinoma
PEI percutaneous ethanol injection
TACE transarterial chemoembolization
REFERENCES
I. El-Serag HB, Andrew C. The increasing incidence of hepato-
cellular carcinoma in the United States. N Engl J Med 1999;
340:745-750
LIVER TRANSPLANTATION FOR HEPATOCELLULAR CARCINOMA-BISMUTH
Ganne-CariC N, Chastang C, Chapel F, et al. Predictive score for
the development of hepatocellular carcinoma and additional
value of liver large cell dysplasia in western patients with cirrho-
sis. Hepatology 1996;23: 11 12-1 118
Llovet JM, Bustamante J, Castells A, et al. Natural history of un-
treated hepatocellular carcinoma: Rationale for design and eval-
uation of therapeutic trials. Hepatology 1999;29:62-67
Makuuchi M, Takayama T, Kubota K, et al. Hepatic resection for
hepatocellular carcinoma. Japanese experience. Hepatogastroen-
terology 1998;45: 1267-1 274
Liver Cancer Study Group of Japan. Predictive factors for long
term prognosis after partial hepatectomy for patients with hepa-
tocellular carcinoma in Japan. Cancer 1994;74:2272-2280
Castells A, Bruix J, Bru C. Treatment of small hepatocellular
carcinoma in cirrhotic patients: A cohort study comparing surgi-
cal resection and percutaneous ethanol injection. Hepatology
1993;18:1121-1126
Lai ECS, Fan ST, Lo CM, Chu KM, Liu CL, Wong J. Hepatic rc-
section for hepatocellular carcinoma. An audit of 343 patients.
Ann Surg 1995;22 1 :29 1-298
Belghiti J, Panis Y, Farges 0, et al. Intrahepatic recurrence after
resection for hepatocellular carcinoma complicating cirrhosis.
Ann Surg 1991;214:114-117
Nagasue N, Uchida M, Makino Y, et al. Incidence of factors as-
sociated with intrahepatic recurrence following resection of he-
patocellular carcinoma. Gastroenterology 1993; 105:488494
Muto Y, Moriwaki H, Ninomiya M, et al. Prevention of second
primary tumors by an acyclic retinoid, polyprenoic acid, in pa-
tients with hepatocellular carcinoma. N Engl J Med 1996:
334: 1561-1567
Ko S, Nakajima Y, Kanehiro H, et al. Significant influence of
accompanying chronic hepatitis status on recurrence of hepa-
tocellular carcinoma after hepatectomy. Ann Surg 1996;214
:591-595
Bismuth H, Chiche L, Adam R, et al. Liver resection versus
transplantation for hepatocellular carcinoma in cirrhotic pa-
tients. Ann Surg 1993;218:145-151
Majno P, Adam R, Bismuth H, et al. Influence of preoperative
transarterial lipiodol chemoembolization on resection and trans-
plantation for hepatocellular carcinoma in patients with cirrho-
sis. Ann Surg 1997;226:688-703
Starzl TE, Putnam CW, eds. Experience in Hepatic Transplanta-
tion. Philadelphia, PA: W.B. Saunders, 1969, pp 3-8
Mac Dougal BUD, Williams R. Indications and assessmcnt
for orthotopic liver transplantation. In: RY Calne, ed. Liver
Transplantation. London, England: Grune & Stratton, 1983,
pp 59-66
Iwatsuki S, Gordon RD, Shaw BW, Starzl TE. Role of liver
transplantation in cancer therapy. Ann Surg 1985;202:401407
O'Grady, JG. Polsen RJ, Rolles K, et al. Liver transplantation
for malignant disease: Results in 93 consecutive patients. Ann
Surg 1988;207:373-379
Ringe B, Wittekind C, Bechstein H, et al. The role of liver trans-
plantation in hepatobiliary malignancy: A retrospective analysis
of 95 patients with particular regard to tumor stage and recur-
rence. Ann Surg 1989;209:88-98
Ismail T, Angrisani L, Gunson BK, et al. Primary hepatic malig-
nancy: The role of liver transplantation. Br J Surg 1990;77:983-
987
Yuki H, Hirohashi S, Sakamoto M, et al. Growth and spread of
hepatocellular carcinoma. A review of 240 consecutive autopsy
cases. Cancer 1990;66:2174-2 179
Ringe B, Pichlmayer R, Wittekind C, et al. Surgical treatment of
hepatocellular carcinoma: Experience with liver resection and
transplantation. World J Surg 1991 ; 15:270-285
Kar S, Carr B. Detection of liver cells in peripheral blood of pa-
tients with advanced stage hepatocellular carcinoma. Hepatol-
ogy 1995;21:403407
ET AL. 32 1
Yokoyama I, Carr B, Saitsu H, Iwatsuki S, Starzl TE. Acceler-
ated growth rates of recurrent hepatocellular carcinoma after
liver transplantation. Cancer 199 1 ;68:2095-2 100
Pichlmayr R, Weimann A, Ringe B. Indications for liver trans-
plantation for hepatobiliary malignancy. Hepatology 1994;
20:33s40s
lwatzuki S, Starzl TE, Sheahan DG, et al. Hepatic resection ver-
s u ~transplantation for hepatocellular carcinoma. Ann Surg
199 1;214:221-229
Haug CE, Jenkins RL, Rohrer RJ, et al. Liver tranplantation for
primary hepatic cancer. Transplantation 1992;53:376-382
McPeake JR, O'Grady JG, Zaman S, et al. Liver transplantation
for primary hepatocellular carcinoma: Tumor size and number
determine outcome. J Hepatol 1993; 18:226-234
Chung SW, Toth JL, Rezieg M, et al. Liver transplantation for
hepatocellular carcinoma. Am J Surg 1994;167:3 17-321
Romani F, Belli LS, Rondinara GF, et al. The role of transplanta-
tion in small hepatocellular carcinoma complicating cirrhosis of
the liver. J Am Coll Surg 1994;178:379-384
Schwartz ME, Sung M, Mor E, et al. A multidisciplinary ap-
proach to hepatocellular carcinoma in patients with cirrhosis. J
Am Coll Surg 1995; 180596-603
Selby R, Kadry Z, Carr B, et al. Liver transplantation for hepato-
Downloaded by: Universite Laval. Copyrighted material.
cellular carcinoma. World J Surg 1995; 19:53-58
Tan KC, Ryder SD, Rizzi PM, et al. Experience of orthotopic
liver transplantation and hepatic resection for hepatocellular car-
cinoma of less than 8 cm in patients with cirrhosis. Br J Surg
1995;82:253-256
Olthoff KM, Rosove MH, Shackleton CR, et al. Adjuvant
chemotherapy improves survival after liver transplantation. Ann
Surg 1995;221:734-743
Mazzaferro V, Regalia E, Doci R, et al. Liver transplantation for
the treatment of small hepatocellular carcinomas in patients with
cirrhosis. N Engl J Med 1996;334:693-699
Figueras J, Jaurrieta E, Valls C, et al. Survival after liver trans-
plantation in cirrhotic patients with and without hepatocellular
carcinoma: A comparative study. Hepatology 1997;25:1485-
1489
Colella G, Bottelli R, De Carlis, et al. Hepatocellular carcinoma:
Comparison between liver transplantation, resective surgery,
ethanol injection and chemoembolization. Transplant Int 1998;
11:s193-196
Bechstein WO, Guckelberger 0, Kling N, et al. Recurrence-free
survival after liver transplantation for small hepatocellular carci-
noma. Transplant Int 1998; l l :s 1 8 9 192
Otto G, Heuschen U, Hofman W, Krumm G, Hinz U, Herfarth C.
Survival and recurrence after liver transplantation versus liver
resection for hepatocellular carcinoma. Ann Surg 1998; 227:
424432
Llovet JM, Bruix J, Fuster J, et al. Liver transplantation for
small hepatocellular carcinoma: The tumor-node-metastasis
classification does not have prognostic power. Hepatology 1998;
27: 1572-1577
Klintmalm GB. Liver transplantation for hepatocellular carci-
noma. A registry report of the impact of tumor characteristics on
outcome. Ann Surg 1998;228:479-490
Bronowicki JP, Boudjema K, Chone L, et al. Comparison of resec-
tion, liver transplantation and oily chemoembolization in the treat-
ment of hepatocellular carcinoma. J Hepatol 1996;24:293- 300
Michel J, Suc B, Montpeyroux F, et al. Liver resection or trans-
plantation for hepatocellular carcinoma? Retrospective analysis
of 2 15 patients with cirrhosis. J Hepatol 1997;26: 1274-1 280
Troisi R, Defreyne L, Hesse UJ, Praet M, Ducruyenaere J, De
Hemptinne B. Multimodal treatment for hepatocellular carcinoma
on cirrhosis: The role of chemoembolization and alcoolization be-
fore liver transplantation. Clin Transplant 1998; 12;313-3 19
Oldhafer KJ, Chavan A, Friihaus NU, et al. Arterial chemoem-
bolization before liver transplantation in patients with hepato-

cellular carcinoma: Marked tumor necrosis but no survival bene-
fit '? J Hepatol 1998;29:953-959
45. Marsh JW, Dvorchick I, Subotin M, et al. The prediction of risk
of recurrence and time to recurrence of hepatocellular carcinoma
after orthotopic liver transplantation: A pilot study. Hepatology
1997;26:444450
46. Sarasin FP, Mentha G, Giostra E, Hadengue A. Partial hepatec-
tomy or liver transplantation for the treatment of resectable he-
patocellular carcinoma. A cost-effectiveness perspective. Hepa-
tology 1998;28:436442
47. Cherqui D. Role of adjuvant treatment in liver transplantation
for advanced hepatocellular carcinoma. J Hepatobiliary Pancre-
atic Surg 1998;5:3540
48. Stone NJ, Klintmalrn GBG, Polter D, et al. Neoadjuvant
chemotherapy and liver transplantation for hepatocellular carci-
SEMINARS IN LIVER DISEASE-VOL. 19, NO. 3, 1999
noma: A pilot study in 20 patients. Gastroenterology 1993;
104: 196-202
49. Adam R, Arenas J, Feray C, Gigou M, Samuel D, Bismuth H.
Impact of B and C viral infection on hepatocellular carcinoma
treated by transplantation. XVI International Congress of the
Transplantation Society, Barcelona, 1996
50. Jonas S, Bechstein WO, Heinze T, et al. Female sex hormone re-
ceptor status in advanced hepatocellular carcinoma and outcome
after surgical resection. Surgery 1997; 12 1 :456-461
5 1. Monteiro E, Perdigoto R, Furtado AL. Liver transplantation for
familial arnyloid polyneuropathy. Hepatogastroenterology
1998;23:1375-1 380
52. Regalia E, Fassati LR, Valente U. Pattern and management of re-
current hepatocellular carcinoma after liver transplantation. J
Hepatal Bil Pancr Surg 1998;5:29-34
Downloaded by: Universite Laval. Copyrighted material.