hepatocellular carcinoma is associated with an increased risk of hepatitis b virus recurrence after...
TRANSCRIPT
CA
HH
LGD
*P
Bc(hgdigHhulHeOlm3rrv1Dmw3rHrHwtH
CLIN
ICA
LA
DV
AN
CES
INLIV
ER,PA
NCREA
S,A
ND
BILIA
RY
TRA
CT
GASTROENTEROLOGY 2008;134:1890–1899
LINICAL ADVANCES IN LIVER, PANCREAS,ND BILIARY TRACT
epatocellular Carcinoma Is Associated With an Increased Risk ofepatitis B Virus Recurrence After Liver Transplantation
UCIANA C. FARIA,*,‡,¶ MICHELLE GIGOU,*,§,¶ ANNE M. ROQUE–AFONSO,*,§,� MYLENE SEBAGH,*,§ BRUNO ROCHE,*,�,§
UILLAUME FALLOT,*,§ TERESA C. A. FERRARI,‡ CATHERINE GUETTIER,*,§ ELISABETH DUSSAIX,*,§
ENIS CASTAING,*,� CHRISTIAN BRECHOT,*,§ and DIDIER SAMUEL*,§,�
‡ §
INSERM, U 785, Villejuif, France; Federal University of Minas Gerais, Faculty of Medicine, Department of Internal Medicine, Belo Horizonte, MG, Brazil; Universitéaris-Sud, UMR-S 785, Villejuif, France; �AP-HP Hopital Paul Brousse, Centre Hepato-Billaire, Villejuif, France; and ¶ANRS HP EP 01 cccDNA, Paris, FrancecrHtHHrr
Hwp(dpblmgpldo
rfc
sahct
See Loomba R et al on page 696 in CGH; SeeCME quiz on page 2155.
ackground & Aims: Hepatitis B virus (HBV) re-urrence after orthotopic liver transplantationOLT) is significantly reduced by prophylaxis withyperimmune antibody to hepatitis B surface anti-en (anti-HBs) globulins (HBIG) and antiviralrugs. The role of hepatocellular carcinoma (HCC)
n HBV recurrence remains unclear. We investi-ated the association between HCC pre-OLT andBV recurrence post-OLT. Methods: We studied 99epatitis B surface antigen-positive patients whonderwent OLT for cirrhosis. The median fol-
ow-up period was 43 months. All patients receivedBIG, and 51 also received lamivudine and/or ad-
fovir. Of these 99 patients, 31 had HCC beforeLT. Total HBV DNA and covalently closed circu-
ar (ccc)-DNA were measured in tumor and nontu-or tissues from the explanted livers of 16 of these
1 HCC patients and, also, in a context of tumorecurrence, in 3 patients who developed HBV/HCCecurrence. Results: Fourteen patients (14.1%) de-eloped HBV recurrence within a median period of5 months post-OLT. HCC at OLT, a pre-OLT HBVNA viral load >100,000 copies/mL, and HBIGonoprophylaxis were independently associatedith HBV recurrence post-OLT. Eleven out of the1 patients with HCC at OLT presented with HBVecurrence and 3 out of the 68 patients withoutCC had HBV recurrence (P < .0001). HBV recur-
ence was more frequent in patients who developedCC recurrence (7/8 patients, 87.5%) than in thoseho did not (4/23 patients, 17.4%) (P < .0001). In
he 16 explanted livers, cccDNA was detectable in
CC cells in 11 and in nontumor cells in 12.ccDNA was detected in a context of HCC recur-ence in 2 of the 3 patients tested who developedBV/HCC recurrence. Conclusions: The associa-
ions of HCC pre-OLT, and HCC recurrence withBV recurrence post-OLT, and the detection ofBV DNA and cccDNA in HCC suggest that HBV
eplication in tumor cells may contribute to HBVecurrence post-OLT.
epatitis B is a leading cause of chronic hepatitis,cirrhosis, and hepatocellular carcinoma (HCC)
orldwide.1 In Europe and the United States, 4%–9% ofatients undergoing orthotopic liver transplantation
OLT) suffer from hepatitis B virus (HBV)-associated liveriseases.2,3 Long-term survival depends largely on therevention of allograft reinfection. Major advances haveeen achieved in recent years in the management of HBV
iver transplant recipients: the use of long-term hyperim-une antibody to hepatitis B surface antigen (anti-HBs)
lobulins (HBIG) post-OLT4 – 6 and then combined pro-hylaxis with highly effective anti-HBV agents, such as
amivudine and adefovir dipivoxil and HBIG, have pro-uced excellent results, ensuring a reduction in the riskf reinfection to approximately 10%.7–11
Chronic HBV infection is one of the most importantisk factors for HCC.12,13 Other known major risk factorsor the development of HCC are cirrhosis because ofhronic hepatitis C virus (HCV) infection or due to other
Abbreviations used in this paper: anti-HBs, antibody to hepatitis Burface antigen; cccDNA, covalently closed circular DNA; HBeAg, hep-titis B e antigen; HBIG, hyperimmune anti-HBs globulins; HBsAg,epatitis B surface antigen; HBV, hepatitis B virus; HCC, hepatocellulararcinoma; HCV, hepatitis C virus; HDV, hepatitis D virus; OLT, ortho-opic liver transplantation.
© 2008 by the AGA Institute0016-5085/08/$34.00
doi:10.1053/j.gastro.2008.02.064
cshaidinaccDcpwscv
piapafmpepntH
HPFwnc(cwdcltys(Oi
sa(p(
a1ftpi1boOclfvFfmat
T
MMHC
C
HHH
Htvac
CLI
NIC
AL
AD
VA
NCES
INLI
VER
,PA
NCREA
S,A
ND
BIL
IARY
TRA
CT
June 2008 HEPATOCELLULAR CARCINOMA AND HBV RECURRENCE 1891
auses such as alcohol use and hemochromatosis, maleex, and exposure to dietary aflatoxin.14 –17 Recent dataave demonstrated that high HBV viral loads are stronglyssociated with the development of HCC,18 –20 althought is commonly assumed that serum HBV DNA levelsecline by the time HCC develops. Through the use of
mmunohistochemical and in situ hybridization tech-iques, it has been suggested that tumor cells no longerllow viral DNA replication and that most tumor tissuesontain only integrated viral DNA.21,22 Covalently closedircular DNA (cccDNA) is a replicative form of HBVNA and is believed to be responsible for persistent
hronic HBV infection.23,24 Little is known about theresence of this episomal form of HBV DNA in HCC cellshen compared with adjacent nontumor tissue. Recent
tudies have shown the presence of HBV DNA andccDNA in HCC cells, thus suggesting the possibility ofiral replication in tumor tissue.25–27
The relationship between HCC and HBV recurrenceost-OLT remains unclear. The present study aimed to
nvestigate whether HCC before OLT and its recurrencefter OLT are associated with HBV recurrence posttrans-lantation. We studied a series of 99 hepatitis B surfacentigen (HBsAg)-positive patients who underwent OLTor cirrhosis, in the era of new antiviral agents. Further-
ore, with the purpose of searching for replication com-etent HBV DNA in HCC cells, we investigated the pres-nce of cccDNA and total HBV DNA using a real-timeolymerase chain reaction (PCR) assay28 in tumor andontumor liver tissues from the explanted livers of pa-ients who had undergone OLT for HCC and chronicBV infection.
Patients and MethodsPatientsBetween December 1995 and December 2004, 99
BsAg-positive patients underwent OLT for cirrhosis ataul Brousse Hospital, Centre Hépato-Biliaire, Villejuif,rance, and were included in the present study. Patientsho underwent transplantation for HBV-related fulmi-ant hepatitis were not included. This patient populationomprised 84 males, and the median age was 49 yearsrange, 14 –71 years). Twenty-six patients (26.3%) wereoinfected with hepatitis delta virus (HDV), 17 (17.2%)ere HCV coinfected, and 6 (6.1%) had human immuno-eficiency virus (HIV) infection. HCC before OLT wasonfirmed by histologic examination of the explantediver in 31 patients (31.3%). In 14 out of these 31 patients,he HCC was outside the Milan criteria29 based on anal-sis of the explanted liver. Thirteen patients receivedystemic adjuvant antitumor chemotherapy after OLTapproximately 6 sessions during the first 6 months post-
LT). The general characteristics of the patients included
n the study are summarized in Table 1. mCombined immunosuppressive therapy post-OLT con-isted of corticosteroids associated with cyclosporine pluszathioprine (10 patients), tacrolimus plus azathioprine20 patients), tacrolimus (29 patients), cyclosporine (26atients), and tacrolimus plus mycophenolate mofetil
14 patients).Prophylactic HBIG was administered intravenously to
ll patients. During the anhepatic phase, they received0,000 IU of HBIG (IVHEBEX LFB, Les Ulis, France),ollowed by 10,000 IU/day during the first 6 postopera-ive days. Subsequently and throughout the follow-uperiod, 10,000 IU of HBIG were given whenever circulat-
ng anti-HBs antibody concentrations had fallen below50 IU/L (in HBV DNA-negative patients before OLT) orelow 500 IU/L (in HBV DNA-positive patients). Fifty-ne patients (51.5%) who were HBV DNA-positive pre-LT received combined prophylaxis with HBIG and nu-
leoside or nucleotide analogues: 36 patients receivedamivudine 100 mg/day (Zeffix; Glaxo Wellcome, Green-ord, United Kingdom), 8 received lamivudine and adefo-ir-dipivoxil 10 mg/day (Hepsera; Gilead Sciences, Inc.,oster City, CA), 2 received adefovir, and 5 HIV-coin-ected patients received lamivudine and tenofovir 300
g/day (Viread; Gilead Sciences, Inc). HBIG and thessociated antiviral drugs were prescribed indefinitely af-er OLT to all these 51 patients.
Patients were followed for a median period of 43
able 1. Characteristics of the 99 HBsAg-Positive PatientsWho Underwent OLT for Cirrhosis
Characteristics
Hepatic disease
HBV-C(n � 73)
HDV-C(n � 26)
Total(n � 99)
ale gender 66 18 84edian age at OLT (y) 51 41 49CC before OLT 25 6 31hemotherapy after
OLT11 2 13
ombined antiviralprophylaxis: HBIG� lamivudine,adefovir, ortenofovir
45 6 51
CV co-infection 10 7 17IV co-infection 4 2 6BV DNA levels at
OLT (copies/mL)�200 38 15 53200–9999 7 9 1610,000–99,999 9 1 0�100,000 11 10 11
(NA for 8 pts) (NA for 1 pt) (NA for 9 pts)
BsAg, hepatitis B surface antigen; OLT, orthotopic liver transplanta-ion; HBV-C, hepatitis B virus cirrhosis; HDV-C, hepatitis B and deltairus cirrhosis; HCC, hepatocellular carcinoma; HBIG, hyperimmunenti-HBs globulins; HCV, hepatitis C virus; HIV, human immunodefi-iency virus; NA, not available; pt, patient.
onths (range, 0.25–125 months) after OLT. HBV recur-
rsCp
wwaatbmbayii
mo(b(1(FttTwH
nattsbWtiOimn
fimtttCv
t5
T
H valentlyp ADV, a
CLIN
ICA
LA
DV
AN
CES
INLIV
ER,PA
NCREA
S,A
ND
BILIA
RY
TRA
CT
1892 FARIA ET AL GASTROENTEROLOGY Vol. 134, No. 7
ence was defined as the reappearance of HBsAg in theerum.4 The study was approved by the local Ethicsommittee, and all individuals gave informed consent toarticipate in the investigation.
Virologic MonitoringSerum HBsAg levels (Diasorin, Sallugia, Italy)
ere tested before and every 3 months post-OLT. Patientsere screened for hepatitis B e antigen (HBeAg) andnti-HBe antibodies (BioMérieux, Marcy l’Etoile, France)nd anti-delta antibodies (Diasorin, Sallugia, Italy) prioro OLT. Anti-HBs antibody titers (Dade Behring, Mar-urg, Germany) were tested every week during the firstonth posttransplantation. After that, anti-HBs anti-
ody titers were tested 1 month after the last HBIGdministration and then every 2 weeks. After the firstear, patients exhibiting stable levels underwent screen-ng for anti-HBs antibody titers every 2–3 months on anndividual basis.
HBV DNA was determined in patient sera every 3– 4onths before and post-OLT using the following meth-
ds: Digene Hybrid Capture System HBV DNA assayMurex Diagnostics, Chatillon, France) between Decem-er 1995 and August 1997; branched DNA methodQuantiplex; Chiron, Emeryville, CA) between September997 and December 2000; and Cobas HBV Monitor assayRoche Diagnostics, Meylan, France) from January 2001.or patients who underwent transplantation before 2001,he HBV DNA load at the time of OLT was retrospec-ively analyzed on sera stored at �80°C using the Cobasaqman HBV Test (Roche Diagnostics, Meylan, France),hich presents an excellent correlation with the Cobas
able 2. Characteristics of the 16 Patients With HepatocelluSamples of Explanted Liver Were Available, and ResNon-Tumor Tissues
PtAntiviral therapy
before OLT
Serum HBV-DNAbefore antiviral
therapy(copies/mL)
Serum HBV-DNAat OLT
(copies/mL)
H
Total(copies/cell)
T NT
1 �200 65.5 2.142 LAM � ADV 5.78 � 108 6.07 � 103 61.3 266.03 ADV NA �200 39.0 4.264 �200 0.02 0.025 �200 1.06 0.086 LAM 2.69 � 103 1.96 � 103 27.2 16.57 �200 0.16 0.138 �200 59.6 1.299 LAM 2.65 � 104 6.37 � 103 86.4 46.9
10 �200 0.006 0.1711 1.42 � 105 530.0 1576.012 �200 0.06 137.013 LAM NA �200 0.43 7.2314 2.7 � 104 99.0 5.5715 LAM � Tenofovir NA �200 545.0 66.216 LAM� ADV 1.47 � 109 6.3 � 105 106.0 1007.0
BV, hepatitis B virus; Pt, patient number; OLT, orthotopic liver transplantation; cccDNA, core-S1 Ag, pre-S1 hepatitis B antigen; T, tumor cells; NT, non-tumor cells; LAM, lamivudine;
BV Monitor assay.30 G
Quantification of HBV cccDNA and TotalDNA in Liver TissueSixteen HCC patients, for whom frozen tumor and
ontumor specimens from their explanted livers were avail-ble, were selected for the measurement of cccDNA andotal HBV DNA levels. The characteristics of these 16 pa-ients are shown in Table 2. A real-time PCR assay for thepecific detection of cccDNA and total HBV DNA in liveriopsy specimens was performed as previously described.28
e also measured cccDNA and total HBV DNA in tumorissue available in the context of post-OLT HCC recurrencen patients who presented with HBV/HCC recurrence post-
LT. Tissue samples were stored at �80°C until the exper-ments were carried out. The diagnoses of the conditions
entioned above were confirmed by the histologic exami-ation of frozen specimens.
Immunostaining of Liver Tumor andNontumor SamplesImmunostaining was performed on sections of
xed liver tumor and nontumor samples using a com-ercial 3-step streptavidin-biotin technique, according
o the manufacturer’s instructions (Ventana Medical Sys-ems, Strasbourg, France). The primary monoclonal an-ibodies employed were anti-HBsAg, anti-HBcAg (Dako,openhagen, Denmark), and anti-pre-S1 (kindly pro-ided by Dr M. A. Petit).31
Sequence AnalysisS gene sequences were obtained from sera and liver
issue using the following forward and reverse primers:=-TGGYTATCGCTGGATGTGTC-3= and 5=-CCCAAAA-
arcinoma Transplanted for HBV-Related Disorders for Whomof HBV-DNA and HBV Antigen Determinations in Tumor and
A Immunostaining
cccDNA(copies/cell) HBsAg HBcAg Pre-S1 Ag
RecurrenceHBV/HCCT NT T NT T NT T NT
0 0.0077 � � � � � � N/N0.053 7.33 �� � � � ��� �� N/N0.68 0.03 � � � � �� �� N/N0.0086 0.07 � � � � �� � N/N0.037 0 � � � � �� �� N/N
26.0 1.33 NA � NA � NA � N/Y0 0.02 � � � � � � Y/Y0 0 � � � � ��� �� Y/Y0.0023 20.2 � � � � � � N/N0 0 � � � � � � N/Y3.78 8.5 � �� � � � ��� Y/Y0 0 �� � � � � �� N/N9.8 5.01 � � � � � �� N/N0.042 0.01 � � � � � � N/Y0.00007 0.0016 NA NA NA NA NA NA N/N0.41 4.87 NA � NA �� NA ��� N/N
closed circular DNA; HBsAg, hepatitis B surface antigen; HBcAg, hepatitis B core antigen;defovir; �, negative; �, positive; NA, not available; N, no; Y, yes.
lar Cults
BV-DN
21
2
ACCCACAATTC-3=. The cycle sequencing reaction was
pRta
SCtldvc
HlOHuaadTeHcWCSc
tmf
fatpsy
(0dd�
T
A
MHHC
H
Htha
T
AMHHHHHHCG
Hha*
CLI
NIC
AL
AD
VA
NCES
INLI
VER
,PA
NCREA
S,A
ND
BIL
IARY
TRA
CT
June 2008 HEPATOCELLULAR CARCINOMA AND HBV RECURRENCE 1893
erformed with the BigDye Terminator Version 3.1eady Reaction Cycle Sequencing Kit (Applied Biosys-
ems, Foster City, CA). Sequences were read bidirection-lly on an ABI377 (Applied Biosystems) instrument.
Statistical AnalysisStatistical analyses were performed using the
PSS 10.0 statistical software (SPSS Inc and Microsoftorp, Chicago, IL). The Kaplan–Meier method was used
o estimate rates of HBV recurrence-free survival, and theog-rank test was used to test hypotheses concerningifferences in survival between the groups selected. Theariables analyzed were age, sex, hepatitis D virus (HDV)oinfection, HCV coinfection, HCC pre-OLT, HBeAg,
able 3. Clinical and Virological Characteristics ofTransplanted HBsAg-Positive Cirrhotic PatientsWith Hepatocellular Carcinoma Compared to ThoseWithout Hepatocellular Carcinoma Before OLT
HCC patients(n � 31)
Non-HCC patients(n � 68) P value
ge (median, rangein y)
53.5 (23�71) 48 (14�63) .06
ale gender 29 55 .18CV co-infection 2 15 .10DV co-infection 6 20 .42ombined prophylaxis
(HBIG� antiviral)12 39 .13
BV DNA levels atOLT (copies/mL)
�200 15 38200�9999 3 13 .210,000�99,999 3 7�100,000 6 5
(NA for 4 pts) (NA for 5 pts)
BsAg, hepatitis B surface antigen; OLT, orthotopic liver transplanta-ion; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HDV,epatitis delta virus; HBIG, hyperimmune anti-HBs globulins; NA, notvailable; pt, patient.
able 4. Actuarial Risk of HBsAg Recurrence Among PatientsVariables Analyzed
Variable P valu
ge �50 y .065ale gender .42*BV-C .23*DV-C .23*CV co-infection .3*BeAg positive .65*CC before OLT �.000BV DNA levels in serum at OLT � 100,000 copies/mL .002ombined prophylaxis (HBIG� antiviral) .014lobal
BsAg, hepatitis B surface antigen; HBV, hepatitis B virus; OLT, orepatitis B and delta virus cirrhosis; HCV, hepatitis C virus; HBeAg, hnti-HBs globulins.
Log-rank test.BV DNA level at the time of OLT, HBIG monoprophy-axis (vs combined prophylaxis), chemotherapy after
LT, HCC stage according to the Milan criteria, andCC recurrence. Multivariate analyses were performedsing the Cox regression model. Variables associated withP value � .25 under univariate analysis were included instepwise Cox regression analysis to identify indepen-
ent factors associated with HBV recurrence after OLT.he Fisher exact and �2 tests were used to test for differ-nces between subgroups of patients. To compare totalBV DNA and cccDNA levels in tumor and nontumor
ells, the Mann–Whitney U test (unpaired samples) orilcoxon signed ranks test (paired samples) were used.
orrelations between variables were calculated usingpearman correlation coefficient test. P values � .05 wereonsidered to be statistically significant.
ResultsThe HBV viral load at transplantation was quan-
ified in 90 patients and was undetectable (�200 copies/L) in 53 (58.9%) of them. Patient distribution as a
unction of HBV DNA levels is shown in Table 1.The 31 patients with HCC did not differ significantly
rom the 68 without HCC in relation to sex, age, HDVnd HCV coinfection, antiviral prophylaxis, HBeAg sta-us, and HBV DNA levels. However, the median age ofatients with HCC (53.5 years; range, 23–71 years) waslightly higher than that of patients without HCC (48ears; range, 14 – 63 years) (P � .06) (Table 3).
Overall HBV RecurrenceHBsAg reappeared in the serum of 14 patients
14.1%) within a median period of 15 months (range,.5– 60 months) after OLT. At that time, HBV DNA wasetectable in the serum in 10 of these 14 patients (me-ian HBV DNA, 6.21 � 106 copies/mL; range, 929 to6.14 � 108 copies/mL) Overall actuarial rates of HBV
splanted for HBV-Related Cirrhosis According to the
HBVrecurrence
Actuarial rate of HBV recurrence after OLT (%)
1 year 2 years 5 years 8 years
10/49 8.7 13.3 27.0 27.013/84 7.5 12.8 17.4 17.412/73 8.6 13.4 20.4 20.42/26 0 4.0 8.8 8.81/17 0 7.1 7.1 7.12/15 0 7.1 18.7 18.7
11/31 16.7 27.8 43.1 43.15/11 30.0 41.7 61.1 61.13/51 2.0 4.0 7.0 7.0
14/99 6.3 10.9 17.3 17.3
pic liver transplantation; HBV-C, hepatitis B virus cirrhosis; HDV-C,tis B e antigen; HCC, hepatocellular carcinoma; HBIG, hyperimmune
Tran
e
*
1***
thotoepati
ra
ulpmdHi�P1
D(r01P
pr
wit
CLIN
ICA
LA
DV
AN
CES
INLIV
ER,PA
NCREA
S,A
ND
BILIA
RY
TRA
CT
1894 FARIA ET AL GASTROENTEROLOGY Vol. 134, No. 7
ecurrence at 1, 2, 5, and 8 years were 6.3%, 10.9%, 17.3%,nd 17.3%, respectively (Table 4).
The variables associated with HBV recurrence undernivariate analysis were HCC (P � .0001), pre-OLT viral
oad �100,000 copies/mL (P � .002), and HBIG mono-rophylaxis (P � .014) (Table 4, Figure 1A–C). Underultivariate analysis, the same factors were indepen-
ently associated with a higher risk of HBV recurrence:CC pre-OLT (relative risk [RR], 6.1; 95% confidence
nterval [95% CI]: 1.6 –23.3; P � .008), HBV DNA100,000 copies/mL at OLT (RR, 5.9; 95% CI: 1.5–23.4;� .01), and HBIG monoprophylaxis (RR, 3.8; 95% CI:
.04 –14.3; P � .04). y
By comparison with patients with undetectable HBVNA levels at OLT, the RR of HBV recurrence was 1.7
95% CI: 0.3–9.3; P � .55) for those with viral loadsanging from 200 to 9999 copies/mL, 3.02 (95% CI:.6 –16.8; P � .2) for those with viral loads ranging from0,000 to 99,999 copies/mL, and 5.9 (95% CI: 1.5–23.4;� .01) for those with viral loads �100,000 copies/mL.
HBV Recurrence in Patients With HCCEleven (35.5%) out of the 31 patients with HCC
resented with HBV recurrence. Actuarial rates of HBVecurrence in this group of patients at 1, 2, 5, and 8
ure 1. Actuarial risk of hepatitis B recurrence afterT in (A) patients who underwent transplantation forV-related cirrhosis who did and did not suffer fromC; (B) in terms of serum HBV DNA levels at the timeLT; (C) in patients treated with HBIG monoprophy-
is and in those receiving combined prophylaxisIG � antiviral agent); (D) in patients without HCC at
T, in patients with HCC but no recurrence of HCCr OLT, and in those presenting with HCC recurrencet-OLT; (E) according to the Milan criteria in patients
h pre-OLT HCC.
FigOLHBHCof Olax(HBOLaftepos
ears were 16.7%, 27.8%, 43.1%, and 43.1%, respectively
(pr8a(HcpttMu
mHAprt.1rc
niombic(
l
clPtnwtlDcb(
nt3(
rracw(i
strea2
T
11112
HcN
CLI
NIC
AL
AD
VA
NCES
INLI
VER
,PA
NCREA
S,A
ND
BIL
IARY
TRA
CT
June 2008 HEPATOCELLULAR CARCINOMA AND HBV RECURRENCE 1895
Table 4). HCC recurrence occurred in 8 out of these 31atients. Seven patients presented with HBV/HCC recur-ence (Table 5). HBV recurrence occurred in 7 out of the
patients (87.5%) who presented with HCC recurrencend in 4 out of the 23 (17.4%) who did not (P � .0001)Figure 1D). In 6 out of the 8 patients who presented with
CC recurrence, the tumor did not meet the Milanriteria29 at OLT. The risk of HBV recurrence in HCCatients was higher in those whose tumors were beyondhe Milan criteria (RR, 18.1; 95% CI: 4.6 –71.9; P � .0001)han in those who presented with a tumor within the
ilan criteria (RR, 4.2; 95% CI: 0.93–18.9; P � .06) (Fig-re 1E).Seven of the 13 patients who received antitumor che-otherapy after OLT presented with a recurrence ofCC, and 6 of these 7 experienced a recurrence of HBV.mong the other 6 without HCC recurrence, only 1resented with HBV recurrence. The risk of HBV recur-ence was higher in patients who received post-OLT an-itumor chemotherapy (RR, 23.2; 95% CI: 5.4 –99.3; P �0001) than in those who did not (RR, 3.9; 95% CI:.04 –17.5; P � .05). No patients without HCC whoeceived combined prophylaxis presented with HBV re-urrence.
Total HBV DNA and cccDNA Levels andHBV Immunostaining in Tumor andNontumor Tissues From Native LiverTotal HBV DNA was detected in all 16 tumor and
ontumor specimens. In 10 cases, cccDNA was detectedn both tumor and nontumor samples; in 1 case, it wasnly found in the tumor and, in 2 cases, only in nontu-or specimens. There were no significant differences
etween paired tumor and nontumor specimens regard-ng both median cccDNA levels (0.023 vs 0.024 copies/ell, respectively, P � .7) and median total HBV DNA49.3 vs 6.4 copies/cell, respectively, P � .9).
Total HBV DNA in tumor and nontumor cells corre-
able 5. HBV-DNA and HBV Antigens in a Context of Tumor RHepatocellular Carcinoma and HBV
PtSerum HBV DNA
at OLT (copies/mL)
Time ofRecurrence(months) Serum HBV
at HBV recurr(copies/mHCC HBV
7 �200 12 6 �2008 �200 13.5 15 �2001 �200 1.5 2.5 6307 �200 9.5 11.5 �1.1 � 108 �200 22.5 24 �2009 �200 With lamivudine 36 27 17600 21,600 With adefovir 31 27 607
BV, hepatitis B virus; OLT, orthotopic liver transplantation; HCC, hircular DNA; HBsAg, hepatitis B surface antigen; HBcAg, hepatitis B coA, not available.
ated positively with each other (r � 0.63; P � .009). b
ccDNA levels in tumor cells were also positively corre-ated with cccDNA levels in nontumor cells (r � 0.66;
� .006). There was also a significant positive correla-ion between total HBV DNA and cccDNA levels inontumor cells (r � 0.5; P � .04); however, no correlationas found between total HBV DNA and cccDNA levels in
umor specimens (r � 0.26; P � .33). Serum HBV DNAevels at OLT were positively correlated with total HBV
NA levels in tumor (r � 0.627; P � .009) and nontumorells (r � 0.669; P � .005) and also with cccDNA levels inoth tumor (r � 0.508; P � .045) and nontumor cells
r � 0.652; P � .006).Pre-S1 antigen staining was positive in 12 out of 15
ontumor tissues and in 8 out of the 13 tumor tissuesested. HBsAg staining was positive in 9 out of 15 and in
out of 13 nontumor and tumor tissues, respectivelyTable 2).
Total HBV DNA and cccDNA Levels andHBV Antigen Immunostaining in TumorTissue in a Context of HCC RecurrenceSeven patients presented with HBV/HCC recur-
ence post-OLT. Tissue specimens from 6 patients with aecurrence of HCC were examined for total HBV DNAnd HBV antigens. Total HBV DNA was detectable in 5ases, and pre-S1 staining was positive in 3 cases. cccDNAas examined in 3 patients and was detectable in 2
hepatic recurrence in one patient and adrenal recurrencen the other) (Table 5).
Sequence AnalysisAmong the 14 patients with a recurrence of HBV,
erum HBV DNA levels were undetectable in 4 of them athe time of HBV recurrence: 3 presented with HBV/HCCecurrence (patients 7, 8, and 18 in Tables 2 and 5) and 1xperienced an HBV recurrence without any known HCCt OLT. However, HBV DNA levels became detectable inof these 4 patients, 5 and 24 months after HBsAg values
rence After OLT in 7 Patients With Recurrences of Both
Tumor HBV DNA(copies/cell) Tumor immunostaining
Total cccDNA HBsAg HBcAg Pre-S1 Ag
5326.0 NA � � �13.1 NA � � �NA NA NA NA NA
147.6 7 � 10�4 � � ��0 NA � � �
1970.0 0 � � ��0.27 0.52 � � �
ocellular carcinoma; Pt, patient number; cccDNA, covalently closedtigen; pre-S1 Ag, hepatitis B pre-S1 antigen; �, negative; �, positive;
ecur
DNAenceL)
8
epatre an
ecame positive (34,700 and 146,000 copies/mL, in pa-
tawaHatwDcardot
umaldbsbartjdt
luoaiisat2dtH
rgwctcOtpDmpoHar
FcchHrrHpH
CLIN
ICA
LA
DV
AN
CES
INLIV
ER,PA
NCREA
S,A
ND
BILIA
RY
TRA
CT
1896 FARIA ET AL GASTROENTEROLOGY Vol. 134, No. 7
ients 7 and 18, respectively). S gene sequences werevailable from serum HBV DNA from 7 of the 10 patientsith detectable HBV DNA after HBV recurrence. Aminocid variations within the major hydrophilic region ofBsAg were observed in 5 individuals with HBV recurrence
lone, but HCC was diagnosed at OLT in 4. The substitu-ions observed were, respectively, as follows: G145A (patientithout HCC); D144E�G145R; T125M�D144G; P120Q�144G; S177I�G119R�T123N�C124R. HBsAg sequen-
es were also available from 2 patients with both HBVnd HCC recurrences (patients 19 and 20 in Table 5). Noelevant mutations were observed within the major hy-rophilic region of HBsAg in patient 19, but an insertionf 5 amino acids, S117SSTTTS, and F/Y134L substitu-ion were observed in patient 20 (Figure 2).
Patient 19 had undetectable HBV DNA levels at OLTnder lamivudine. HBV recurred 27 months after OLT, 9onths before the diagnosis of HCC recurrence in the
drenal gland. At HBV recurrence, the serum HBV viraload was low (1760 copies/mL). Total HBV DNA wasetected in the adrenal metastasis but in neither the liveriopsy specimen nor serum collected at the same time (aturgical resection of the adrenal metastasis). HBsAg levelsecame negative after resection of the adrenal metastasisnd then became positive again after a second tumorecurrence (Figure 3). The HBV sequence obtained duringumor recurrence exhibited no mutations within the ma-or hydrophilic loop of HBsAg (Figure 2) and no lamivu-ine resistance mutations, despite the fact that this pa-ient was receiving lamivudine and HBIG.
Patient 20 underwent transplantation with HBV DNAevels at 21,600 copies/mL under adefovir. HBV DNA wasndetectable in the serum after OLT and until the timef HBV recurrence, which occurred 27 months after OLTnd 4 months before the diagnosis of HCC recurrencen the adrenal gland. The HBV sequences obtained dur-ng tumor recurrence and from serum sampled at theame time were identical and revealed a marked alter-tion of the major hydrophilic loop when compared withhe sequence determined in serum prior to OLT (Figure). No mutations responsible for adefovir resistance wereetected in the polymerase open reading frame, despitehe fact that this patient was receiving adefovir and
BIG. rDiscussionDuring this study, we analyzed the risk of HBV
ecurrence in HBsAg-positive patients who had under-one OLT for cirrhosis, the purpose being to investigatehether HCC prior to OLT is associated with HBV re-
urrence posttransplantation. Our results add strengtho the argument put forward since the late 1990s that thearriage of HBsAg constitutes a valuable indication forLT. We showed that, in a modern context marked by
he use of effective prophylaxis for anti-HBV recurrenceost-OLT and the availability of sensitive assays for HBVNA quantification, viral load at transplantation re-ains an independent risk factor for HBV recurrence, as
reviously described.4,6,10,11,32–34 Our results also corrob-rate the beneficial effects of combined prophylaxis withBIG and a nucleoside or nucleotide analogue, which is
n independent factor associated with lower HBV recur-ence.
Figure 2. Progression of themajor hydrophilic loop of HBsAgin 2 patients with a recurrence ofboth HBV and HCC.
igure 3. This patient (patient 19) underwent OLT for HBV-relatedirrhosis and HCC. Prior to OLT, the HBV DNA viral load was 4.76 � 105
opies/mL and became negative under lamivudine therapy. After OLT,e received HBIG and lamivudine. Twenty-seven months after OLT,BV DNA and HBsAg levels became positive. A few months later, a
ecurrence of HCC was diagnosed in the adrenal gland. After surgicalesection of the adrenal metastasis, HBsAg in the serum disappeared.BV DNA was detected in the adrenal metastasis but not in the trans-lantation liver at the same time. After a second HCC recurrence,BsAg and HBV DNA became positive again. The patient died of HCC
ecurrence, 6 years after OLT.
ftrrsfr5htsh
otHprsmppeHppttpta
swaorphsmttHntIrrimqad
PlspcdaslsmticHbtreiptb
HHHcntrluicPnotnctrcchtawsIr
t
CLI
NIC
AL
AD
VA
NCES
INLI
VER
,PA
NCREA
S,A
ND
BIL
IARY
TRA
CT
June 2008 HEPATOCELLULAR CARCINOMA AND HBV RECURRENCE 1897
The original finding of this study was to demonstrate,or the first time, that the presence of HCC at transplan-ation is an independent risk factor associated with HBVecurrence. Furthermore, in patients with HCC, HBVecurrence was associated with HCC recurrence. This wasupported by the detection of cccDNA in tumor cellsrom the explanted liver and also in HCC metastasesecurring after transplantation, as shown in Tables 2 and, respectively. Except for one study that suggested aigher incidence of HBV recurrence in transplanted pa-ients with HCC,35 previous studies had not demon-trated any association between HCC prior to OLT and aigher risk of HBV recurrence.10,34,36 –38
Several mechanisms may be involved in the higher ratef HBV recurrence in HCC patients. Antitumor chemo-herapy may possibly play some role. A reactivation ofBV replication is a frequent and well-recognized com-lication in patients with chronic HBV infection whoeceive chemotherapy.39,40 However, according to our re-ults, most of the patients who received post-OLT che-
otherapy and presented with HBV recurrence also ex-erienced a recurrence of HCC. Only one individualresented with recurrent HBV infection alone and novidence of tumor recurrence. This finding suggests thatCC recurrence itself, and not chemotherapy, was therincipal factor responsible for HBV recurrence in theseatients. Although it is not possible to rule out a con-ributory role for chemotherapy, our data do emphasizehe fact that the higher rate of HBV recurrence in HCCatients who received such treatment was probably dueo the selection of patients with more advanced tumorsnd, therefore, at a higher risk of HCC recurrence.
The association between HCC and HBV recurrenceuggests the possibility of viral replication in HCC cells,hich would then act as a viral reservoir. The disappear-nce of HBsAg after resection of an adrenal metastasis inne patient who presented with both HBV and HCCecurrence, and its reappearance after a second extrahe-atic tumor recurrence (Figure 3), points toward thisypothesis. This idea is reinforced by the fact that, in thisame patient, HBV DNA was detected in the adrenal
etastasis but not in the transplanted liver at the sameime. It is possible that the immunosuppression relatedo HCC recurrence itself contributes to HBV recurrence.owever, the direct temporal relationship between adre-al gland resection and the disappearance of HBsAg fromhe serum suggests HBV replication in HCC metastasis.n patient 20, who also presented with HBV/HCC recur-ence, the HBV sequences obtained from both the recur-ing tumor and serum sampled at the same time weredentical and exhibited a marked alteration within the
ajor hydrophilic loop when compared with the se-uence observed prior to OLT (Figure 2). This findinglso constitutes powerful evidence in favor of virus pro-
uction by tumor cells. lIn this study, we used a recently developed real-timeCR assay that enables the quantification of cccDNA in
iver tissue.28 Its specificity is based on using a plasmid-afe DNAse treatment to digest non-cccDNA forms andrimers located on both sides of the gap of relaxedircular (rc) DNA to preferentially amplify cccDNA. Theetection of cccDNA in HCC cells from explanted livernd from HCC recurring post-OLT in patients who pre-ented with HBV/HCC recurrence demonstrated the rep-ication competence of HBV DNA in HCC cells andtrongly supports the hypothesis that HBV replication
ay occur in HCC cells. Extrahepatic HCC micrometas-ases could act as potential reservoirs for HBV graftnfection. After HCC recurrence, the tumor burden in-reases, and viral production in the tumor increasesBIG consumption, which ultimately exceeds the capa-
ility to neutralize all the virions produced. With respecto patients who received combined prophylaxis, no HBVecurrence occurred in those without HCC. In the currentra of highly effective prophylaxis against HBV, this find-ng suggests that HCC itself is the main factor of HBVrophylaxis failure. Furthermore, patients with bulkierumors are at higher risk, as shown by the comparisonetween patients beyond and within the Milan criteria.
Few quantitative data are available on intrahepaticBV DNA levels in tumor and nontumor tissues fromCC patients. Zanella et al25 quantified intrahepaticBV DNA using a real-time PCR method and found
omparable levels of total HBV DNA in the tumor andontumor tissues. However, the method used duringheir study was not able to distinguish between ccc andcDNA forms. Wong et al27 quantified total and cccDNAevels in tumor and nontumor tissues from HCC patientssing the Invader assay (Third Wave Technologies, Mad-
son, WI) and demonstrated significantly higher levels ofccDNA in HCC cells. Pollicino et al26 adopted a nestedCR approach and detected cccDNA in both tumor andontumor specimens from 46.7% of 30 patients withccult hepatitis B. They also performed reverse-transcrip-ion PCR analysis in 10 cccDNA-positive tumor andontumor liver tissues and detected HBV RNA in allases, thus demonstrating a persistence of HBV transcrip-ional activity. Taken together, these data show thateplication competent HBV DNA may be present in HCCells. The previous concept of an absence of HBV repli-ation in HCC cells was based on immunostaining orybridization techniques, which are much less sensitivehan PCR assays.21,22 In this context, the levels of totalnd cccDNA found in tumor tissues from our patientsere not very high, and this was supported by the ab-
ence of HBcAg from tumor tissue in most patients.ndeed, HBcAg is not a very sensitive marker of HBVeplication, unlike preS1 antigen staining.31
Although our data point toward a relationship be-ween HCC and HBV recurrence, it is important to high-
ight the limitations of our sample regarding its size,
hsh
twHsHcrrpdlrur
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
CLIN
ICA
LA
DV
AN
CES
INLIV
ER,PA
NCREA
S,A
ND
BILIA
RY
TRA
CT
1898 FARIA ET AL GASTROENTEROLOGY Vol. 134, No. 7
eterogeneity, and possible sampling errors. Furthertudies are necessary to confirm the findings presentederewith.In conclusion, HCC pre-OLT, viral load at transplan-
ation �100,000 copies/mL, and HBIG monoprophylaxisere independent factors associated with a higher risk ofBV recurrence. HCC recurrence post-OLT was also as-
ociated with HBV recurrence. The association betweenCC and HBV recurrence after OLT and the detection of
ccDNA in HCC cells point toward the possibility of HBVeplication in tumor cells, which could act as potentialeservoirs for HBV recurrence, especially in patients whoresent with a recurrence of HCC. Thus, despite theramatic improvements achieved in HBV prophylaxis fol-
owing OLT, we should remain cautious concerning theisk of HBV recurrence in patients with HCC and partic-larly in those at a high risk of HCC recurrence. Theseesults require confirmation by further investigations.
References
1. Lee WM. Hepatitis B virus infection. N Engl J Med 1997;337:1733–1745.
2. Adam R, McMaster P, O’Grady JG, et al, and the European LiverTransplantation Association. Evolution of liver transplantation inEurope: report of the European liver transplant registry. LiverTranspl 2003;9:1231–1243.
3. Camci C, Gurakar A, Rose J, et al. Liver transplantation forhepatitis B in the United States. Transplant Proc 2005;37:4350–4353.
4. Samuel D, Muller R, Alexander G, et al. Liver transplantation inEuropean patients with the hepatitis B surface antigen. N EnglJ Med 1993;329:1842–1847.
5. Terrault NA, Zhou S, Combs C, et al. Prophylaxis in liver trans-plant recipients using a fixed dosing schedule of hepatitis Bimmunoglobulin. Hepatology 1996;24:1327–1333.
6. Shouval D, Samuel D. Hepatitis B immune globulin to preventhepatitis B virus graft reinfection following liver transplantation: aconcise review. Hepatology 2000;32:1189–1195.
7. Markowitz JS, Martin P, Conrad AJ, et al. Prophylaxis againsthepatitis B recurrence following liver transplantation using com-bination lamivudine and hepatitis B immune globulin. Hepatology1998;28:585–589.
8. Han SHB, Ofman J, Holt C, et al. An efficacy and cost-effective-ness analysis of combination hepatitis B immune globulin andlamivudine to prevent recurrent hepatitis B after orthotopic livertransplantation compared with hepatitis B immune globulinmonotherapy. Liver Transpl 2000;6:741–748.
9. Marzanno A, Salizzoni M, Debernardi-Venon W, et al. Preven-tion of hepatitis B virus recurrence after liver transplantation incirrhotic patients treated with lamivudine and passive immuno-prophylaxis. J Hepatol 2001;34:903–910.
0. Roche B, Feray C, Gigou M, et al. HBV-DNA persistence 10 yearsafter liver transplantation despite successful anti-HBs passiveimmunoprophylaxis. Hepatology 2003;38:86–95.
1. Terrault N, Roche B, Samuel D. Management of the hepatitis Bvirus in the liver transplantation setting: a European and anAmerican perspective. Liver Transpl 2005;11:716–732.
2. Szmuness W. Hepatocellular carcinoma and the hepatitis B virus:evidence for a causal association. Prog Med Virol 1978;24:40–69.
3. Beasley RP, Hwang LY, Lin CC, et al. Hepatocellular carcinomaand hepatitis B virus. A prospective study of 22,707 men in
Taiwan. Lancet 1981;2:1129–1133.4. Brechot C. Hepatitis B and C viruses and primary liver cancer.Baillieres Clin Gastroenterol 1996;10:335–373.
5. Idilman R, De Maria N, Colantoni A, et al. Pathogenesis ofhepatitis B and C-induced hepatocellular carcinoma. J Viral Hepat1998;5:285–299.
6. Sun Z, Lu P, Gail MH, et al. Increased risk of hepatocellularcarcinoma in male hepatitis B surface antigen carriers withchronic hepatitis who have detectable urinary aflatoxin metabo-lite M1. Hepatology 1999;30:379–383.
7. Llovet JM, Burroughs A, Bruix J. Hepatocellular carcinoma. Lan-cet 2003;362:1907–1917.
8. Yu MW, Yeh SH, Chen PJ, et al. Hepatitis B virus genotype andDNA level and hepatocellular carcinoma: a prospective study inmen. J Natl Cancer Inst 2005;97:265–272.
9. Chen CJ, Yang HI, Su J, et al, for the REVEAL-HBV Study Group.Risk of hepatocellular carcinoma across a biological gradient ofserum hepatitis B virus DNA level. JAMA 2006;295:65–73.
0. Liu CJ, Chen BF, Chen PJ, et al. Role of hepatitis B virus precore/core promoter mutations and serum viral load on noncirrhotichepatocellular carcinoma: a case-control study. J Infect Dis2006;194:594–599.
1. Raimondo G, Burk RD, Lieberman HM, et al. Interrupted replica-tion of hepatitis B virus in liver tissue of HBsAg carriers withhepatocellular carcinoma. Virology 1988;166:103–112.
2. Wang WL, London WT, Feitelson MA. Hepatitis B x antigen inhepatitis B virus carrier patients with liver cancer. Cancer Res1991;51:4971–4977.
3. Tuttleman JS, Pourcel C, Summers J. Formation of the pool ofcovalently closed circular viral DNA in hepadnavirus infectedcells. Cell 1986;47:451–460.
4. Newbold JE, Xin H, Tencza M, et al. The covalently closed duplexform of the hepadnavirus genome exists in situ as a heteroge-neous population of viral minichromossomes. J Virol 1995;69:3350–3357.
5. Zanella I, Rossini A, Domenighini D, et al. Real-time quantitationof hepatitis B virus (HBV) DNA in tumorous and surroundingtissue from patients with hepatocellular carcinoma. J Med Virol2002;68:494–499.
6. Pollicino T, Squadrito G, Cerenzia G, et al. Hepatitis B virusmaintains its pro-oncogenic properties in the case of occult HBVinfection. Gastroenterology 2004;126:102–110.
7. Wong DK, Yuen MF, Poon RT, et al. Quantification of hepatitis Bvirus covalently closed circular DNA in patients with hepatocellu-lar carcinoma. J Hepatol 2006;45:553–559.
8. Werle-Lapostolle B, Bowden S, Locarnini S, et al. Persistence ofcccDNA during the natural history of chronic hepatitis B anddecline during adefovir dipivoxil therapy. Gastroenterology 2004;126:1750–1758.
9. Mazzaferro V, Regalia E, Doci R, et al. Liver transplantation forthe treatment of small hepatocellular carcinomas in patients withcirrhosis. N Engl J Med 1996;334:728–729.
0. Hochberger S, Althof D, Gallegos de Schrott R, et al. Fully auto-mated quantitation of hepatitis B virus (HBV) DNA in humanplasma by the COBAS AmpliPrep/COBAS Taqman System. J ClinVirol 2006;35:373–380.
1. Petit MA, Dubanchet S, Capel F. A monoclonal antibody specificfor the hepatocyte receptor binding site on hepatitis B virus. MolImmunol 1989;26:531–537.
2. Samuel D, Bismuth A, Mathieu D, et al. Passive immunoprophy-laxis after liver transíplantation in HBsAg-positive patients. Lan-cet 1991;337:813–815.
3. Neff GW, O’Brien CB, Nery J, et al. Outcomes in liver transplantrecipients with hepatitis B virus: resistance and recurrence pat-terns from a large transplant center over the last decade. Liver
Transpl 2004;10:1372–1378.
3
3
3
3
3
3
4
Hd
VtS
DV
AN
CES
AN
CREA
S,RY
TRA
CT
June 2008 HEPATOCELLULAR CARCINOMA AND HBV RECURRENCE 1899
4. Marzanno A, Gaia S, Ghisetti V, et al. Viral load at the time of livertransplantation and risk of hepatitis B recurrence. Liver Transpl2005;11:402–409.
5. Wong PYN, McPeake JR, Portmann B, et al. Clinical course andsurvival after liver transplantation for hepatitis B virus infectioncomplicated by hepatocellular carcinoma. Am J Gastroenterol1995;90:29–34.
6. Mazzaferro V, Regalia E, Montalto F, et al. Risk of HBV reinfectionafter liver transplantation in HBsAg-positive cirrhosis. Primaryhepatocellular carcinoma is not a predictor for HBV recurrence.The European Cooperative Study Group on Liver Cancer andTransplantation. Liver 1996;16:117–122.
7. Steinmüller T, Seehofer D, Rayes N, et al. Increasing applicabilityof liver transplantation for patients with hepatitis B-related liverdisease. Hepatology 2002;35:1528–1535.
8. Wong SN, Reddy KR, Keeffe EB, et al. Comparison of clinicaloutcomes in chronic hepatitis B liver transplant candidates withand without hepatocellular carcinoma. Liver Transpl 2007;13:
334–342. c9. Lok AS, Liang RH, Chiu EK, et al. Reactivation of hepatitis B virusreplication in patients receiving cytotoxic therapy. Report of aprospective study. Gastroenterology 1991;100:182–188.
0. Markovic S, Drozina G, Vovk M, et al. Reactivation of hepatitis Bbut not hepatitis C in patients with malignant lymphoma andimmunosuppressive therapy. A prospective study in 305 pa-tients. Hepatogastroenterology 1999;46:2925–2930.
Received August 02, 2007. Accepted February 21, 2008.Address requests for reprints to: Didier Samuel, MD, Prof, Centre
epato-Biliaire, Hôpital Paul Brousse, 94800 Villejuif, France. e-mail:[email protected]; fax: (33) 1 45 59 38 57.Supported by the French National Agency for Research on AIDS and
iral Hepatitis (ANRS), ANRS HB EP 01 cccDNA, who is the sponsor ofhis trial, and by Coordenação de Aperfeiçoamento de Pessoal de Níveluperior (CAPES; to L.C.F.), Brazil.Conflicts of interest: None of the authors have any potential financial
onflicts of interest related to this paper.
CLI
NIC
AL
AIN
LIV
ER,P
AN
DBIL
IA