-
V O L U M E 1 1 • S U P P L E M E N T 1 • M A R C H 2 0 0 9
www.hpbjournal.com
THE OFFICIAL JOURNAL OF THE
International Hepato-Pancreato-Biliary Association
American Hepato-Pancreato-Biliary Association
European Hepato-Pancreato-Biliary Association
IN THIS ISSUE
Abstracts of the Ninth Americas Hepatopancreatobiliary
Congress
12–15 March 2009
Miami Beach, FL, USA
HPB
VO
LUM
E 11 • SUPPLEM
ENT 1 • M
ARC
H 2009
HPB_v11_is1_cover_B_5.0mm.indd 1 2/19/2009 12:08:43 PM
-
Editor-in-ChiefO. J. Garden, UK
Associate EditorsM. Callery, USAS. Connor, New Zealand
Honorary Regional EditorsY. Nimura, JapanH. Obertop, the
NetherlandsS. Strasberg, USA
Editorial BoardR. Andersson, SwedenC. Bassi, ItalyJ. Belghiti,
FranceG. Belli, ItalyM. Buechler, GermanyR. Chari, USAM.-F. Chen,
TaiwanM. Choti, USAC. Christophi, AustraliaP.-A. Clavien,
SwitzerlandJ. E. Cunha, BrazilE. de Santibanes, Argentina
C. Dervenis, GreeceJ. Espat, USAL. Fernández-Cruz, SpainY. Fong,
USAA. Frilling, SwitzerlandS. Gallinger, CanadaJ.-F. Gigot,
BelgiumH. Gooszen, the NetherlandsD. Gouma, the NetherlandsS.
Helton, USAM. Henderson, USAJ. Izbicki, Germany
P. Jagannath, IndiaB. Jeppsson, SwedenL. Kow, AustraliaM.
Krawczyk, PolandJ. Krige, South AfricaP. Lai, ChinaD. Mahvi, USAM.
Makuuchi, JapanD. Nagorney, USAR. Padbury, AustraliaT. Pappas,
USAR. Parks, UK
W. Pinson, USAH. Pitt, USAA. Principe, ItalyM. Rees, UKM. Ryska,
Czech RepublicM. Selzner, CanadaF. Sutherland, CanadaT. Takada,
JapanY. Tekant, TurkeyT. van Gulik, the NetherlandsJ. Windsor, New
Zealand
HPB Editors and Editorial Board
Aims and Scope. HPB is an international forum for clinical,
scientifi c and edu-cational communication. Eight issues a year
bring the reader leading articles, expert reviews, original
articles, images, editorials, and reader correspondence
encompass-ing all aspects of benign and malignant hepatobiliary
disease and its management. HPB features relevant aspects of
clinical and translational research and practice. Specifi c areas
of interest include HPB diseases encountered globally by clinical
practitioners in this specialist fi eld of gastrointestinal
surgery. The journal addresses the challenges faced in the
management of cancer involving the liver, biliary system and
pancreas. While surgical oncology represents a large part of HPB
practice, submission of manuscripts relating to liver and pancreas
transplantation, the treat-ment of benign conditions such as acute
and chronic pancreatitis, and those relating to hepatobiliary
infection and infl ammation are also welcomed. There will be a
focus on developing a multidisciplinary approach to diagnosis and
treatment with endoscopic and laparoscopic approaches, radiological
interventions and surgical techniques being strongly represented.
HPB aims to help its readers – surgeons, physicians, radiologists
and basic scientists – to develop their knowledge and prac-tice.
HPB will be of interest to specialists involved in the management
of hepat-obiliary and pancreatic disease however will also inform
those working in related fi elds.
ISSN (Print) 1365-182X; ISSN (Online) 1477-2574
Disclaimer. The Publisher, the International
Hepato-Pancreato-Biliary Associa-tion and Editors cannot be held
responsible for errors or any consequences arising from the use of
information contained in this journal; the views and opinions
expressed do not necessarily refl ect those of the Publisher, the
International Hepato-Pancreato-Biliary Association and Editors,
neither does the publication of advertisements constitute any
endorsement by the Publisher, the International
Hepato-Pancreato-Biliary Association and Editors of the products
advertised.
Journal compilation © 2009 International
Hepato-Pancreato-Biliary Association. All rights reserved. No part
of this publication may be reproduced, stored or transmitted in any
form or by any means without the prior permission in writing from
the copyright holder. Authorization to photocopy items for internal
and personal use is granted by the copyright holder for libraries
and other users registered with their local Reproduction Rights
Organisation (RRO), e.g. Copyright Clearance Center (CCC), 222
Rosewood Drive, Danvers, MA 01923, USA (www.copyright.com),
pro-vided the appropriate fee is paid directly to the RRO. This
consent does not extend to other kinds of copying such as copying
for general distribution, for advertising and promotional purposes,
for creating new collective works or for resale. Special requests
should be addressed to: [email protected].
Information for subscribers. HPB is published in 8 issues per
year. Subscrip-tion prices for 2009 are:Print + Online: US$920 (The
Americas), #593 (Europe), $920 (Rest of World). Prices are
exclusive of tax. Australian GST, Canadian GST and European VAT
will be applied at the appropriate rates. For more information on
current tax rates, please
go to www.wiley.com, click on Help and follow the link through
to Journal subscrip-tions. The institutional price includes online
access to the current and all online back fi les to January 1st
1997, where available. For other pricing options, including access
information and terms and conditions, please visit
www.interscience.wiley.com/journals.
Delivery Terms and Legal Title. Prices include delivery of print
journals to the recipient’s address. Delivery terms are Delivered
Duty Unpaid (DDU); the recipient is responsible for paying any
import duty or taxes. Legal title passes to the customer on
despatch by our distributors.
Back issues. Single issues from current and recent volumes are
available at the current single issue price from
[email protected]. Earlier issues may be obtained from
Periodicals Service Company, 11 Main Street, Germantown, NY 12526,
USA. Tel: +1 518 537 4700, Fax: +1 518 537 5899, Email:
[email protected].
HPB (ISSN 1365-182X), is published eight times a year. US
mailing agent: Mercury Airfreight International Inc., 365 Blair
Road, Avenel, NJ 07001, USA. Periodical postage paid at Rahway, NJ,
USA.
Postmaster: Send all address changes to HPB, Journal Customer
Services, John Wiley & Sons Inc., 350 Main St., Malden, MA
02148-5020, USA.
HPB is published by Blackwell Publishing Ltd, 9600 Garsington
Road, Oxford OX4 2DQ, UK. Tel: +44 1865 776868, Fax: +44 1865
714591.
Journal Customer Services. For ordering information, claims and
any enquiry concerning your journal subscription please go
tointerscience.wiley.com/support or contact your nearest offi
ce:Americas: Email: [email protected]; Tel: +1 781 388 8598 or
1 800 835 6770 (Toll free in the USA & Canada)Europe, Middle
East and Africa: Email: [email protected]; Tel: +44(0) 1865 778
315Asia Pacifi c: Email: [email protected]; Tel: +65 6511
8000
Production Editor: David Christie (email: [email protected]).
For submission instructions, subscription and all other
information visit: www.hpbjournal.com
This journal is available online at Wiley InterScience. Visit
www3.interscience.wiley.com to search the articles and register for
table of contents and e-mail alerts.
Printed in the UK by the Charlesworth Group.
Abstracting and Indexing Services. The Journal is indexed by
EMBASE/Excerpta Media, SIIC database.
HPB_v11_is1_ifc_B_5.0mm.indd 1 2/19/2009 12:23:12 PM
-
HPB
CONTENTS Volume 11, Supplement 1, March 2009
Abstracts of the Ninth Americas Hepatopancreatobiliary
Congress
Annual Scientifi c Session and Postgraduate Program
March 12–15 2009
Eden Roc Renaissance Resort
Miami Beach, FL, USA
Disclaimer
This abstract book has been produced using authors-supplied
copy. Editing has been restricted to some corrections of spelling
and style where appropriate. No responsibility is assumed for any
claims, instructions, methods or drug dosages contained in the
abstracts: it is recommended that these are verifi ed
independently.
(fm_i_ii)HPB_v11_is1_toc.indd fm_i 2/19/2009 12:24:23 PM
-
(fm_i_ii)HPB_v11_is1_toc.indd fm_ii 2/19/2009 12:24:23 PM
-
Abstracts 1 to 8President’s Plenary Session
Friday, March 13, 2009 7:45–9:30 am
1
Second hepatic resection for recurrenthepatocellular
carcinoma
S. ROAYAIE, MD, D. BASSI, S. HIOTIS, MD,D. LABOW, MD and M.
SCHWARTZ, MDMount Sinai Hospital, New York, NY, USA
While several Asian studies have shown promising resultsfor
second resection for recurrent hepatocellular carci-noma (HCC),
there are no Western series on this topic.The purpose of this study
was to determine the outcomeof patients undergoing a second hepatic
resection forrecurrent HCC at a Western center and to
identifyprognostic variables.Methods: A retrospective review of all
patients under-going hepatic resection for HCC from 1/1990 to
1/2008was conducted. Patients underwent second resection ifthey had
a single tumor on imaging, Child’s A liverfunction, platelets >
100 000 and no extrahepatic dis-ease.Results: During this period,
487 patients underwentresection with 221 having documented
recurrence. Ofthese, 30 underwent second resection. Underlying
liverdisease included HBV (n = 18), HCV (n = 6), none(n = 4), and
other (n = 2). Mean tumor size was 9.6 cmat first resection and 4.0
cm at second resection. Medianinterval between first resection and
recurrence was15.5 months. There were no perioperative mortalities
and1, 3, and 5 years survivals were 88%, 65%, and 65%.Recurrence
rate at 3 year was 80%. Univariate predic-tors of survival included
< 1year interval from firstresection to recurrence, gross
vascular invasion at secondresection, tumor > 5cm at second
resection, and bloodtransfusion at second resection. Multivariate
analysisfound gross vascular invasion at second resection as
theonly independent predictor of mortality (HR 76.9,P = 0.002).
Patients without gross vascular invasion at
2nd resection had median survival of 76.8 months and5 year
survival of 74%.Conclusions: Second resection for recurrent HCC
hasexcellent outcomes in well selected patients. Grossvascular
invasion is the only independent predictor ofoutcome after second
resection.
2
Incorporating an HPB fellowship does notdiminish surgical
residents’ HPB experience in ahigh-volume training program
N. J. ZYROMSKI, MD, L. TORBECK, PHD,D. F. CANAL, MD, K. D.
LILLEMOE, MD andH. A. PITT, MDIndiana University, Indianapolis, IN,
USA
Background: A major paradigm shift in surgical educa-tion has
recently been instituted by the American Boardof Surgery and the
Surgical Council on Resident Edu-cation (SCORE). Specific surgical
procedures have beendefined as ESSENTIAL (Common/Uncommon –
specificprocedural competency required by the end of training)and
COMPLEX (generic competence required, but notcompetence in
individual procedures). Importantly, vir-tually all elective HPB
procedures fall into the SCORE‘COMPLEX’ category. Trainees who wish
to practiceHPB surgery will therefore be required to obtainadvanced
training. As this training paradigm evolves, it isequally crucial
that incorporation of an HPB fellowshipinto an established surgical
residency program does notdiminish surgical residents’ exposure to
complex HPBprocedures. We hypothesized that incorporation of anHPB
fellowship into a high volume clinical trainingprogram would not
detract from residents’ HPB experi-ence.Methods: Our institution
incorporated an HPB trainingprogram in 2005–2006. Resident
operative case logs(provided by the American Council of Graduate
MedicalEducation) and HPB fellow case logs were reviewed.Resident
exposure to complex HPB procedures for3 years prior to and 3 years
after fellowship incorpo-ration were compared. Student’s t-test was
appliedwhere appropriate; P < 0.05 was accepted as
statisticallysignificant.Results: The ACGME requires graduating
residents’exposure to a minimum of three pancreas and fourliver
cases (complex biliary cases are not subdividedfrom the category
‘alimentary’). In 2007, the nationalaverage exposure of graduating
chief residents was:pancreas – 11; liver – 9; biliary – 5. The
InternationalHepatopancreatobiliary Association (IHPBA)
guidelinesfor HPB fellowship training call for exposure to:pancreas
– 30; liver – 25; and biliary – 20. Our institu-tional resident and
fellow HPB experience is shown inthe Table 1.
0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0 P
-
Table 1. Resident and fellow HPB experience 2002–2008.
Pancreas Liver Complex BiliaryResident Fellow Resident Fellow
Resident Fellow
2002–2003 20 – 13 – 8 –2003–2004 25 – 22 – 11 –2004–2005 27 – 15
– 9 –2005–2006 23 100 14 40 4 382006–2007 23 109 14 42 8
442007–2008 19 114 13 40 ** 56P value 0.39 0.34 0.17
**Data will be available 12/2008.
Conclusions: These data show that an HPB fellowshipprogram can
be incorporated into a high volume clinicaltraining program without
detracting from resident HPBexperience. Prior to initiating an HPB
fellowship pro-gram, individual training programs must carefully
assesstheir capability to provide an adequate clinical
experiencefor fellows without diminishing resident exposure
tocomplex HPB procedures.
3
Improved survival with aggressive resection ofhilar
cholangiocarcinoma
Y. L. CHEAH, MD, M. AKOAD, MD, K. VAKILI,MD, J. J. POMPOSELLI,
MD, PHD, E. A. POMFRET,MD, PHD, W. D. LEWIS, MD and R. L.
JENKINS,MDLahey Clinic, Burlington, MA, USA
Background: Complete surgical resection with negativehistologic
margins is a major determinant of long-termsurvival for patients
with hilar cholangiocarcinoma. Asour experience with surgical
resection for hilar cholan-giocarcinoma evolved, we adopted a more
radical surgi-cal approach with the addition of hepatectomy
andcomplex vascular reconstruction to achieve negativesurgical
margins. In this study we sought to examine theresults of this
aggressive surgical approach and its impacton patient
survival.Methods: All patients with the diagnosis of hilar
cho-langiocarcinoma evaluated and treated by a single teamof
surgeons between 1986 and 2007 were identified fromour computerized
database. A total of 120 patientsunderwent resection with curative
intent; 65 patientsbetween 1986 and 1998 at the former New
EnglandDeaconess Hospital (period 1), and 55 patients between1999
and 2007 at the Lahey Clinic (period 2). Patientdemographics,
extent of surgical resection and outcomeswere retrospectively
analyzed and compared betweenperiods 1 and 2. Survival is
calculated using the Kaplan–Meir method and curves were compared
using the logrank test.Results: The overall 1, 3, and 5-year
survival rates ofpatients with R0 resection were 85.5%, 63.7%
and35.5% compared to 56.6%, 8.2% and 0% for R1 resec-tion (P <
0.001). Patients in period 2 received moreextensive surgical
resections; hilar combined with hepaticresection was performed in
69.1% of patients in period 2compared to 48.6% in period 1.
Concomitant vascularresection was performed in 17 patients (three
in period1 and 14 in period 2), with portal vein resection in
13
patients, hepatic artery resection in two patients, andboth
hepatic artery and portal vein resection in twopatients. Negative
margins were achieved in 81.8%in period 2 compared to 50.0% in
period 1 (P < 0.05).The perioperative mortality rate improved
from 6% inperiod 1 to 1% in period 2. The 1, 3 and 5 year
survivalin period 1 was 67.1%, 39.4%, and 18.5% compared to79.6%,
48.4%, and 31.1% in period 2 (P < 0.05). Post-operative
complications occurred in 22 (40%) patientsin period 2; the most
common complication was bile leakin eight patients.Conclusion:
Aggressive surgery with the addition ofpartial hepatectomy and
vascular resection in treatinghilar cholangiocarcinoma improves
patient survival withacceptable morbidity and mortality.
4
Validation of a predictive algorithm to maxi-mize resectability
of pancreatic adenocarcinoma
P. BAO, MD, D. POTTER, PHD, J. YOUNG,D. EISENBERG, MD, D.
LENZNER, MS, K. LEE,MD, H. ZEH, MD, M. SANDERS, MD, S. HUGHES,MD
and A. J. MOSER, MDUniversity of Pittsburgh, Pittsburgh, PA, USA;
UPCIBiostatistics Facility, Pittsburgh, PA, USA
Introduction: The surgeon’s major contribution topatients with
localized pancreatic cancer is a marginnegative resection. We
hypothesized that a predictionalgorithm based on preoperative
computed tomography(CT) and endoscopic ultrasound (EUS) could
maximizethe rate of R0 resection while reducing the risk of
non-therapeutic surgery.Methods: 197 patients with biopsy-proven
pancreaticadenocarcinoma (157 head; 40 body/tail)
underwentexploration with intent to resect from 2002 to 2007.
Allpatients had staging helical CT and 143 had EUS. Aprediction
model was developed from the imaging data of65 patients during
2002–2005. This algorithm classifiedpatients as high or low risk
for noncurative surgery. Itwas validated in a subsequent cohort of
78 patientsbetween 2005 and 2007. Model performance was eva-luated
using contingency table and survival analysis.
0.0 0.0
0.2
0.4
0.6
0.8
1.0
1.0 2.0 3.0 Years
Cu
m S
urv
ival
P
-
Results: Predictors of resectability in the developmentcohort
were: any evidence of vascular involvement on CTscan, EUS stage IIB
by AJCC criteria, and EUS tumorsize greater than 2.6 cm. The
resectability rate in thevalidation cohort was 77% with a 58% R0
resection rate.Compared to these outcomes, selecting operative
patientsclassified by the model as favorable for curative
surgerywould have increased the resectability rate from 77% to92%
(P = 0.03) and the R0 resection rate from 58% to73% (P = 0.08). The
model was 71% accurate with 78%sensitivity and 61% specificity for
R0 resection. Therewas a 40% difference in R0 resectability between
highrisk and low risk patients (P = 0.001). High risk
classi-fication was associated with unresectable locally-ad-vanced
disease (P = 0.007), metastasis (P = 0.012), andthe need for
mesenteric vein resection (P = 0.012).Compared to patients at low
risk, those at high predictedrisk had a 67% rate of noncurative
surgery and experi-enced significantly shorter median survival
(12.3 vs.20.6 months, P = 0.022).
Conclusion: This validated prediction algorithm usesstandard CT
and EUS criteria to identify patients mostlikely to benefit from
resection for pancreatic adenocar-cinoma. The model significantly
increased the R0 resec-tion rate and reduced the rate of
nontherapeutic surgery.Prediction models may be used to classify
patientsentering neoadjuvant therapy trials.
5
Serial molecular analysis of EUS guided FNAaspiration of
pancreatic cysts: quantitativemolecular evidence detects neoplastic
cystprogression, stability or regression
D. MALLAT, MD, J. P. TAZELAAR, C. SPENCE,PHD, E. M. ELDER, SCD
and S. D. FINKELSTEIN,MDBaylor University, Dallas, TX, USA; RedPath
IntegratedPathology, Pittsburgh, PA, USA
Introduction: The biological behavior of pancreaticcystic
neoplasms, particularly those of mucinous origin,is difficult to
predict. Progression rates of 30% andhigher have been reported in
mucinous cystic processesleading to recommended surgical excision
in all cases.Regression of mucinous cysts over time is suspectedto
occur, however objective support has been lacking.
Molecular analysis of aspirated pancreatic cyst fluid(point
mutation, loss of heterozygosity) is used here toevaluate the
presence and rate of that mutational changeover time.Design: The
fluid from 52 patients with EUS-FNApancreatic cysts underwent
molecular analysis con-sisting of measurement of 1) DNA quantity,
2) DNAquality (extent of degradation), 3) Kras point mutation,4)
allelic imbalance (loss of heterozygosity determina-tion [LOH]) for
a panel of 16 markers and 5) degree ofclonal expansion of DNA
alterations when present(PathFinderTG�). In each patient, serial
aspirationswere available (40-two serial, 8-three, 4-four) totaling
120separate cyst fluid genotyping reactions. The intervalbetween
serial analyses ranged from 3 to 36 months.Mucinous cysts were
defined as cysts containing gross ormicroscopically visible mucin,
elevated CEA, Kras pointmutation and/or multiple LOH alterations.
Cysts notmeeting these criteria were classified as
serous/reactiveprocesses.Results: Forty-two of 52 patients had
cysts meeting thecriteria for mucinous etiology. 20 (48%)
manifestedneoplastic regression by reduction in DNA amount,
shiftfrom good to poor quality, elimination of some or all ofthe
point mutations/LOH change present previously and/or lowering in
the degree of clonality. The mucinous cystsof 19 patients (45%)
remained essentially stable whilethree patients (7%) manifested
molecular evidence ofprogression at 6, 8 and 10 months. All 10
non-mucinouscyst fluids showed stable indolent molecular features
(lowDNA, poor quality, 0–2 low clonality alterations).Mucinous
cysts with indolent molecular features did notmanifest neoplastic
progression.Conclusions: Molecular analysis provides a useful
ancil-lary tool with which to characterize aspirated
pancreaticfluid. The vast majority of mucinous cysts did
notmanifest neoplastic progression for up to 36-monthfollow-up.
Only 7% of mucinous cysts showed mole-cular progression which was
predicted in the initialsample by aggressive molecular changes. The
resultssupport integrated molecular pathology risk stratificationof
patients with pancreatic cysts including mucinouscystic
lesions.
6
A simple risk score predicts inpatient mortalityafter liver
resection for hepatocellularcarcinoma
J. P. SIMONS, MD, S. C. NG, MS, J. S. HILL, MD,S. A. SHAH, MD,
Z. ZHOU, MD, PHD andJ. F. TSENG, MD, MPHUniversity of Massachusetts
Medical School, Worcester,MA, USA
Objective: To develop an integer-based risk score basedon
national data to estimate the risk of in-hospitalmortality in
patients undergoing procedures for hepato-cellular
carcinoma.Background: There is a wide spectrum of disease burdenin
hepatocellular carcinoma accompanied by severaloptions for surgical
management. However, the associ-ated mortality of such procedures
is not well-defined.Accurate predictions of patients’ perioperative
riskwould be helpful to guide decision-making.
Low Risk73% R0
(+) (–)
(–) (+)
(–) (+)
EUS Stage 1A
EUS Stage 2B/3
(–) (+) Size>2.6 cm
Prediction algorithm for Risk of Nontherapeutic Laparotomy
CT vessel
High-Risk67% Not R0
Figure 1.
Abstracts 3
� 2009 The AuthorsJournal Compilation � 2009
Hepato-Pancreato-Biliary Association, HPB, 11 (Suppl. 1), 1–104
-
Methods: The Nationwide Inpatient Sample, the largestall-payer
discharge database in the U.S., was queriedfrom 1998 to 2005. A
cohort of patient-discharges forhepatic procedures with a diagnosis
of primary liverneoplasm was assembled. Procedures were categorized
ashepatic lobectomy, wedge resection, or enucleation/ablation.
Logistic regression and bootstrap methods wereused to create an
integer risk score for estimating the riskof in-hospital mortality
using procedure type, patientdemographics, comorbidities, and
hospital type. A ran-domly selected sample of 80% (n = 2274) of the
cohortwas used to create the risk score with validation of thescore
conducted in the remaining 20% (n = 560).Results: A total of 2834
patient-dischargeswere identified.Overall in-hospital mortality was
6.53%. Factors includedin the final model were age, gender,
Charlson comorbidityscore, procedure type, and teaching hospital
status. Inte-ger values were assigned to these characteristics, and
thenused for calculating an additive score (Figure, right
panel).Four clinically relevant score groups were then assembledto
stratify risk of in-hospital mortality, with a 13-foldgradient of
mortality ranging from 1.6 to 21.5% (Figure,left panel; P <
0.0001). In the derivation set, as inthe validation set, the score
discriminated well, with ac-statistic of 0.74 and 0.72,
respectively.Conclusion: An integer-based risk score can be used
topredict in-hospital mortality after surgical procedure
forhepatocellular carcinoma, and may be useful for preop-erative
risk stratification and patient counseling.
7
SNPs in RecQL predict survival in pancreaticadenocarcinoma
M.-C. GINGRAS, PHD, D. LI, PHD, S. E. HODGES,R. A. GIBBS, PHD,
F. C. BRUNICARDI, MD andW. E. FISHER, MDBaylor College of Medicine,
Houston, TX, USA; M. D.Anderson Cancer Center, Houston, TX, USA
Introduction: RecQL is a DNA helicase involved in DNAmismatch
repair. The RecQL 3’UTR A159C genotypehas previously been
associated with overall survival ofpatients with pancreatic cancer.
However, the functionalsignificance of this SNP remains
unknown.Hypothesis: The RecQL A159C SNP is in linkagedisequilibrium
with other functional SNPs of the gene.Somatic mutations of the
RecQL gene in tumors mayinfluence the clinical outcome of
pancreatic cancer.Methods: We sequenced the entire coding regions
of theRecQL gene in paired blood and tumor DNA of 35patients with
resectable pancreatic cancer treated withsurgery and adjuvant
chemoradiation. DNA was isolatedfrom blood using the PAXgene Blood
DNA kit (Pre-AnalytiX) and from matched tumors using the QIAamp
DNA Mini kit (Qiagen). Primer sets were designed tocover the 15
RecQL exons with their surroundingintronic regions. Sequencing was
performed on ABI 3700DNA Sequencers. SNPs and somatic mutations
werevalidated with Biotage pyrosequencing. SNPs thatshowed
significant association with overall survival werefurthered tested
in blood DNA samples of 120 patientswith resectable pancreatic
adenocarcinoma who receivedneoadjuvant chemoradiation. Univariate
analysis of theeffect of genotype on time to recurrence and
overallsurvival was performed using the Cox proportionalhazards
models.Results: Several previously reported and newly
identifiedSNPs but no nonsynonymous SNPs were found inthis study.
In addition to the RecQL A159C SNP(rs13035), 2 SNPs, located in
introns 2 and 11(rs10841834, rs2159943) showed a significant
associationwith overall survival. The three SNPs are part of
thesame haplotype block (linkage disequilibrium). The var-iant
allele of each SNP had a similar effect on overallsurvival of
patients receiving either adjuvant or neoad-juvant therapy. No
mutations were detected in thetumors in the exonic regions of the
RecQL gene.
Conclusion: The SNPs from the 3’UTR and intronic 2and 11 regions
(CC, TT, and AA genotypes) of RecQLare associated with improved
overall survival of patientswith resectable pancreatic cancer. The
functional signi-ficance of these SNPs warrants further
investigation.
8
Biliary complications including a single donormortality:
experience of 207 adult-to-adultliving donor liver
transplantation
M. A. EL-METEINI, MD, A. F. HAMZA, MD,A. A. ABDALAAL, MD, M. F.
FATHY, MD,M. BAHAA, MD, A. MOKHTAR, MD,F. ABOUELFETOUH, MD, I.
MOSTAFA, MD andA. EL-DORRY, MDAin-Shams University, Cairo, Egypt;
Wady ElNeelHospital, Cairo, Egypt
Introduction: Donor safety is crucial in Living donorliver
transplantation (LDLT) with ‘do no harm’aphorism printed in the
transplant team mind. Biliaryanomalies are more common in right
liver compared toleft liver grafts. Biliary complication is the
main cause ofmorbidity following right lobe donation.
Mortalityfollowing right lobe donation has been estimated to beless
than 0.5%.Patients and methods: Between November 2001 to date,207
adult-to-adult LDLT has been done using right lobegrafts. The
donors included 173 men and 34 women withmean age 28.4 ± 5.2 years.
Siblings were144 (69.6%)cases while unrelated donors were 63
(30.4%) with amean body mass index 25.2 ± 2.4. Liver biopsy is
Score Calculation
Age groups 0 ≤553 56-65 3 66-75 6 >75 0 0 Charlson score
Procedure type
Gender
Hospital type N(%): 554(24.4) 859(37.8) 624(27.4) Score
Est
imat
ed m
orta
lity
(%)
0–5 6–10 11–15 16–26
21.47
8.77
3.78 1.65
0
5
10
15
20
25
237(10.4)
2 1 5 2 10 3 0 Enucleation 2 Wedge 7 Lobectomy 0 Female 1 Male 0
Teaching 2 Non-teaching
Figure 1.
rs10541534 (C/T)
Log-rank P value: 0.005 Log-rank P value: 0.01 Log-rank P value:
0.001
rs2159943 (T/C) rs13035 (A/C)
Figure 1.
Abstracts4
� 2009 The AuthorsJournal Compilation � 2009
Hepato-Pancreato-Biliary Association, HPB, 11 (Suppl. 1), 1–104
-
routinely done and steatosis less than 15% is accepted.Single
and multiple right bile duct (RBD) were present in82 (39.6%) and
125 (60.3%) donor, respectively. Multi-ple RBD included 2 or 3 RBD
in 116 (56%) and 9 (4.3%)cases, respectively. The mean operative
time was360 ± 50 min. with an estimated blood loss around950 ± 450
mL and returned cell-saver amount of450 ± 334 mL. Two donors
(0.9%), each received twoblood bank units. Donor remnant liver
volumes (RLV)were 33.5 ± 3.2%. Mean ICU stay was 3 ± 0.7 daysand
mean hospital stay was 14 ± 3.5 days.Results: The overall biliary
complications occurred in 27(13.04%) cases. Following modified
Clavien classifica-tion, biliary complications were graded as grade
I(n = 10), grade II (n = 2), grade III (n = 14) and gradeV (n = 1).
Grade I and II (n = 12) biliary complicationswere successfully
managed conservatively. Grade IIIcases were treated by ultrasound
guided aspiration,
ERCP and surgery in 10, 2 and 2 donors, respectively.The later 2
donors were treated by t-tube insertion in onecase and duct-to-duct
anastomosis following transactionof the CBD in the other case. The
later case neededultrasound guided dilatation and 10F stent
insertiontwice with normal liver profile thereafter. Single
donormortality (Grade V) (0.4%) occurred following biliaryleakage
from RBD stump followed by peritonitis thatnecessitated exploration
on day 11, re-exploration onday 31 and ERCP with stent insertion on
day 38 andthe donor succumbed on day 43 due to
uncontrolledsepsis.Conclusion: Donor biliary complications accounts
for upto 43% of all complications. The majority of
biliarycomplications is minor and could be managed conser-vatively.
However, uncontrolled biliary leakage is aserious morbidity that
should be avoided as it could leadto a mortality.
Abstracts 5
� 2009 The AuthorsJournal Compilation � 2009
Hepato-Pancreato-Biliary Association, HPB, 11 (Suppl. 1), 1–104
-
Abstracts 9 to 16Free Papers – Liver I
Saturday, March 14, 2009 8:10–10:10 am
9
Surgical downstaging and neo-adjuvant therapyin metastatic
colorectal carcinoma with irino-tecan drug eluting beads: a
multi-institutionalstudy
M. BOWER, MD, K. ROBBINS, MD, D. TOMALTY,MD, C. R. SCOGGINS, MD,
K. M. MCMASTERS,MD, PHD and R. C. MARTIN, MDUniversity of
Louisville, Louisville, KY, USA; BaptistHealth, Little Rock, AR,
USA; Huntsville Hospital,Huntsville, AL, USA
Background: Neoadjuvant chemotherapy for potentiallyresectable
metastatic colorectal cancer (MCC) is becom-ing a more common
treatment algorithm. However, withthe ever increasing hepatic
toxicity with systemic che-motherapy a more target approach with
maximumresponse and minimal hepatic toxicity is needed. The aimof
this study was to evaluate the efficacy of precisionhepatic
arterial Irinotecan therapy in potentially resect-able MCC.Methods:
An open-label, multi-center, multi-nationalsingle arm study of MCC
patients, who receivedhepatic arterial Irinotecan. Primary
endpoints weresafety, tolerance and metastatic tumour
resection.Results: Fifty-five patients with metastatic colorectal
tothe liver underwent a total of 90 hepatic arterialIrinotecan
treatments. Thirty (55%) Women, 25 (45%)Men, 39 (71%) Caucasian,
with a median age of52 years (range 42–75) were treated. The extent
ofliver involvement was 41 (75%) of patients had< 25% Tumor
Replacement, 15% (< 26–50% tumorreplacement), 10% (> 50%
replacement), with mediannumber of hepatic lesions being 4 (range
1–20), totalsize of all target lesions being 9 cm (range 5.5–28
cm),and 50% of patients having bilobar tumor distribution.Median
number of irinotecan treatments were two(range: 1–5), median
treatment dose was 100 mg (range100–200), with total hepatic
treatment of 200 mg(range 200–650), with 86% of treatments being
per-formed in a lobar infusion treatment, and 30% ofpatients
treated with concurrent simultaneouschemotherapy. Eleven (20%)
patients demonstratedsignificant response and downstage of their
disease ordemonstrated stable disease without extra-hepaticdisease
progression that they under went resection,ablation, or resection
and ablation. There were nodeaths with morbidity occurring in 20%
ofpatients, with none of them being hepatic related.Non-tumorous
liver resected demonstrated no evi-dence of steatohepatitis from
the Irinotecan arterialinfusion.Conclusions: Hepatic arterial
infusion with Irinotecanwas safe and effective in the treatment of
MCC asdemonstrated by a minimal infusion complication
rate,acceptable tumor response and sustained diseasestabilization.
Hepatic arterial infusion is an acceptabletherapy for evaluating
the biology of metastatic colo-rectal to the live prior to planned
hepatic resection.
10
Exsanguination protocol improves survival aftermajor hepatic
trauma
V. ZAYDFUDIM, MD, W. D. DUTTON, MD, I. D.FEURER, PHD, B. K. AU,
B.SC., C. W. PINSON,MD, MBA and B. A. COTTON, MDVanderbilt
University Medical Center, Nashville, TN,USA
Background: Hepatic injury remains an important causeof
exsanguination after major trauma. Recent studieshave noted a
dramatic reduction in mortality amongseverely injured patients when
trauma exsanguinationsprotocols (TEP) are employed. We hypothesized
thatutilization of our institution’s TEP at the initiation
ofhospital resuscitation would improve survival in patientswith
significant hepatic trauma.Methods: TEP, which involves immediate
and sustainedrelease of blood products to the
hemodynamicallyunstable trauma patients in the operating room
inpre-defined ratios, was initiated in February 2006. Allpatients
who (1) underwent immediate operative inter-vention, (2) sustained
intra-abdominal hemorrhage withGrade 3–5 hepatic injury, and (3)
received the TEPbetween February 2006 and January 2008 were
pro-spectively identified. A pre-TEP cohort was retrospec-tively
identified from among all trauma patients admittedbetween February
2004 and January 2006 who (1)received immediate operative
intervention, (2) weretreated for major intra-abdominal hemorrhage
with aGrade 3-5 hepatic injury, and (3) received more than 10units
packed red blood cells in the first 24 h. Univariateand
multivariate analyses were used to examine the ef-fects of TEP on
blood product use during the first 24 h ofresuscitation and
evaluated the effects of demographicand clinical covariates on
survival.Results: Seventy-five patients were included in the
anal-ysis: 39 constituted the pre-TEP cohort (31% 30-daysurvival)
and 36 were treated with the TEP protocol(53% 30-day survival).
There were no differences in age,gender, mechanism, grade of
injury, lobar involvement,or injury to major hepatic vasculature
between the twocohorts (all P > 0.25). However, patients treated
withTEP had higher injury severity scores (P < 0.01).
While24-hour use of blood products did not differ betweencohorts,
TEP patients received more FFP and plateletsduring operative
intervention and significantly less crys-talloid (all P < 0.01).
After adjusting for age, gender,mechanism, and injury severity
score; Grade 5 injury andinvolvement of major hepatic vasculature
had significantnegative effects on survival (both P £ 0.02), while
utili-zation of TEP improved the odds of 30-day survival by78% (OR
= 0.22, CI: 0.06–0.81, P = 0.02).Conclusions: An exsanguination
protocol allows formore effective utilization of FFP and platelets
duringintra-operative management of major hepatic injury.TEP was
associated with a significant improvement in30-day survival among
patients treated for intra-abdominal hemorrhage associated with
significanthepatic trauma.
� 2009 The AuthorsJournal Compilation � 2009
Hepato-Pancreato-Biliary Association, HPB, 11 (Suppl. 1), 1–104
-
11
Evaluation of perioperative chemotherapy usinga prognostic
nomogram for survival followingresection of colorectal liver
metastases
S. K. REDDY, MD, M. W. KATTAN, PHD, C. YU,PHD, E. P. CEPPA, MD,
S. G. DE LA FUENTE, MD,Y. FONG, MD, B. M. CLARY, MD and R. R.
WHITE,MDDuke University Medical Center, Durham, NC, USA;Cleveland
Clinic, Cleveland, OH, USA; Memorial SloanKettering Cancer Center,
New York, NY, USA
Introduction: Nomograms are statistical tools designedto predict
outcomes and risk stratify patients for clinicaltrials and
multi-modality therapy. The aim of this studywas to evaluate the
effects of perioperative chemotherapyon disease-specific survival
(DSS) after resection ofcolorectal liver metastases (CLM) using a
prognosticnomogram incorporating demographic and clinicopath-ologic
variables established at Memorial-Sloan KetteringCancer Center
(MSKCC).Methods: An external cohort comprised of 203 consec-utive
patients who underwent resection of CLM between1996 and 2006 at
Duke University Medical Center wasused to validate the nomogram and
to evaluate the effectsof perioperative chemotherapy on DSS after
resection.Results: Similar to the MSKCC population, the
externalcohort included patients with node-positive primarydisease
(61.1%), rectal primary tumors (22.2%), disease-free interval
-
13
Resection of colorectal cancer (CRC) livermetastases: What is an
adequate margin?
R. T. PADBURY, MD, PHD, D. VANDEWEYER,MBBS, G. J. MADDERN, MD,
PHD and J. W. CHEN,MD, PHDFlinders Medical Centre, Adelaide,
Australia; The QueenElizabeth Hospital, Adelaide, Australia.
Traditionally a 1cm margin has been accepted as the goalwhen
resecting CRC liver metastases. Evidence is emerg-ing that a lesser
margin may provide good outcomes, buta critical margin, below which
recurrence is higher andsurvival poorer, has not been universally
agreed. In arecent publication1 we reported peri-operative
morbidityand clear margin as the two independent prognosticfactors.
In this study we defined a clear margin as anabsence of tumor cells
within 1 mm of the transectedsurface. The aim of the current study
is to further analysethe effect of the width of the surgical margin
on patientsurvival to determine whether a margin of 1 mm
isadequate.Methods: Two hundred and sixty-one consecutive pri-mary
liver resections for CRC mets from 1992 to 2007were analysed
(including 197 patients from1). 163(62.5%) were male. The median
age was 64 (22–92) years.The 30 day (inhospital) mortality was
1.5%. The primaryCRC Duke staging include 11 As, 70 Bs, 110 Cs and
70Ds. Initial analysis was performed on five groupsaccording to the
resection margins; involved, minimalmargin (0–1mm), > 1–4mm,
> 4 to < 10mm and‡10mm. Subsequent analysis was based on two
groups:margin £ 1mm and > 1mm.Results: With a median follow-up
of 4.7 years, theoverall 5 year patient & disease free survival
(DFS) were38% & 22% respectively. The 5 year patient and DFSwas
significantly better in patients with ‡ 10mm resectionmargin
compared with those of involved margin (43.4%vs. 19.4% P < 0.03;
28.7% vs. 16.9% P < 0.02 respec-tively). There was no
significant difference in patient orDFS between the three groups
with margin > 1mm. The5-year patient survivals in these groups
were similar.When a comparison is made between patient with
eitherinvolved or £ 1mm margin with patients with > 1mmmargin,
there is a significant 5 year patient survival dif-ference of 25%
versus 43% (P < 0.04). The DFS dif-ference however did not reach
statistical significance(P = 0.14)
Conclusions: In this cohort with a medium follow-up of5 years,
we can demonstrate that a margin of > 1 mm isassociated with
significant better 5 year survival. Thepossible beneficial effect
of greater margin beyond 1mmcould not be clearly demonstrated in
this cohort.Reference:1. Scheisser M, Chen JW, Maddern GJ, Padbury
RT.
Perioperative morbidity affects long term survival inpatients
following liver resection for colorectal metas-tases. J
Gastrointestinal Surg 2008; 12: 1054–60.
14
Locally advanced intrahepatic cholangiocarci-noma: survival
benefit with aggressive hepaticresection
S. H. TEH, MD, D. CUSATI, MD, J. GRAMS, MD,E. ABOIAN, MD, L.
CHANG, MD, S. CHA, PHD,L. BURGART, MD and D. M. NAGORNEY, MDSacred
Heart Medical Center, Eugene, OR, USA; MayoClinic, Rochester, MN,
USA; Mayo Clinic, Rochester,MN, USA
Introduction: Intrahepatic cholangiocarcinoma (ICC) israre and
often presents as locally advanced disease. Wehypothesized that
aggressive major hepatic resection mayimprove the survival of
patients with ICC.Methods: All consecutive patients who
underwenthepatic resection for ICC from Jan 1993 to Jan 2003
wereretrospectively reviewed. All pathological specimens
werere-examined by two independent pathologists.Results: There were
100 patients (male = 40, female =60) with a mean age of 63 years
old. Eighty-six patientshad major hepatic resections, of which 30
were extendedhepatic resections and eleven involved resection
andreconstruction of an extrahepatic structure (vena cava =7, main
portal vein = 2, and common bile duct = 2).The mean tumor size was
7.7 cm; multifocal disease waspresent in 27 and regional invasion
in 17. There were 89R0 resections and eight R1 resections despite
an extendedhepatic resection. Perioperative morbidity and
mortalitywere 37% and 3%, respectively. Mean hospitalizationwas 9
days. Multivariate analysis demonstrated that thefactors
significantly predicting survival were R0 resection(P < 0.0001,
HR 3.92), tumor size (P = 0.0001, HR3.74), metastasis to lymph
node(s) (P = 0.0001, HR3.241), older age (P = 0.002, HR 3.11), and
tumormultiplicity (P = 0.02, HR 1.78). The median survivalrate for
R0 and R1 resections was 4.97 years vs.8.9 months, respectively.
Overall 1-, 3-, and 5- year sur-vival rates for R0 and R1 resection
were 84% vs. 45%,57% vs. 18% and 49% vs. 0%, respectively.
Diseaserecurred in 59% of patients after a mean follow up of4 years
(hepatic recurrence = 53%, extrahepatic =47%).Conclusions:
Multivariate analysis demonstrated bothtumor and patient factors as
critical predictors of longterm outcome in the management of
locally advancedICC. However, the most significant factor
predictingsurvival was ability to achieve an R0 resection.
Effec-tive preoperative multimodality therapy is needed todownstage
tumors, and patients should undergoaggressive major hepatic
resection with the goal of R0resection.
0 1 2 3Patient Survival (years)
Resection margins
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
P1 mm
4 5 6
Figure 1.
Abstracts8
� 2009 The AuthorsJournal Compilation � 2009
Hepato-Pancreato-Biliary Association, HPB, 11 (Suppl. 1), 1–104
-
15
Long-term outcome following sequentialresections of liver and
lung metastases fromcolorectal carcinoma
R. MARUDANAYAGAM, MS,R. KRISHNAMOORTHY, MS, V. SHANMUGAM,MS, C.
COLDHAM, NONE, S. BRAMHALL, MD,MBBS, D. MAYER, MD, MBBS, J.
BUCKELS, MD,MBBS and D. MIRZA, MSQueen Elizabeth Hospital,
Birmingham, UK
Background: Surgical resection of colorectal livermetastases
(CLM) is an established form of treatment.Limited data exists on
the value of sequential hepatic andpulmonary metastatectomy. We
analysed patients whounderwent sequential liver and lung resections
for CLM.Methods: A total of 910 patients who underwent
liverresection for CLM between January 2000 and December2007, were
analysed to identify patients with resectablepulmonary metastases
(n = 43; 4.7%). Patient demo-graphics, overall survival and
survival difference betweensynchronous and metachronous pulmonary
metastatec-tomy group and between ‘liver and lung resection’
groupand matched ‘liver resection only’ group without pul-monary
metastases (matched for age, primary diseasestage, interval to
liver resection, and liver disease stage)were analysed.Results:
Forty-three patients underwent sequential liverand lung resections.
The median age was 62 years. Themedian number of liver lesion
detected was three and themost commonly performed procedure was
right hemi-hepatectomy (41.9%). Right upper lobe was the
pre-dominant site of lung metastasis (51%). Ten patients
hadsynchronous lung metastasis. The median interval be-tween liver
and lung metastatectomy was 25 months. The1-, 3- and 5-year overall
survival rates after first metas-tatectomy were 100%, 87.1% and
53.9% respectivelywith a median survival of 42 months.
Metachronouspulmonary metastatectomy group had better 1-, 3-
and5-year survival rates than the synchronous group (100%,88.9% and
60.9% vs. 100%, 75% and 0% respec-tively).There was no significant
survival differencebetween the ‘liver and lung resection’ and the
‘liverresection only’ groups.Conclusion: Sequential liver and lung
resections formetastases from colorectal carcinoma have good
long-term survival for selected patients. Presence of synchro-nous
lung and liver metastases was not associated withlong term
survival.
16
Predictors of blood transfusion requirement inelective liver
resection
A. J. COCKBAIN, MBBS, T. MASUDI, MBBS,G. J. TOOGOOD, MD, J. P.
LODGE, MD andK. RAJ PRASAD, MDSt James’ University Hospital, Leeds,
UK
Background: Liver resection for primary and secondarymalignancy
remains major surgery frequently requiringintra-operative blood
transfusion. The balance betweensafer surgical techniques and more
extensive resections
has direct implications on blood transfusion require-ments. In
addition to increasing shortage of blood stocks,and the cost and
side effects of blood transfusion, there isconcern that
perioperative blood transfusion has animmunomodulatory effect which
adversely affectstumour recurrence and prognosis.Aim: To identify
predictors of perioperative bloodtransfusion.Methods: A
retrospective review of a prospectively col-lected database of all
elective hepatic resections under-taken in a tertiary referral
centre over a 4-year period wasperformed. Data analysed included
patient demograph-ics, comorbidities, underlying liver disease,
haematolog-ical parameters, preoperative radiological
intervention,chemoradiotherapy, previous liver resection, number
oftumours, extent and method of resection, use of hae-mostatic
agents and histological diagnosis and grade.Number of units of
blood crossmatched and transfusedwere collected from the hospital’s
blood bank database.Multivariate regression analysis was performed
on sig-nificant factors on univariate analysis to
determineindependent predictors of blood transfusion in
theimmediate perioperative period (48 h).Results: 599 patients were
identified with a median age of64 years and a male: female ratio of
8:5. In the periop-erative period patients were crossmatched a
median of10 units blood. Ratio of units crossmatched:
unitstransfused was 13:1. Fifteen percent of patients received
ablood transfusion with a median transfusion of two
units.Transfusion requirement varied by operation from 12%of
patients undergoing hemihepatectomy or metastecto-my to 48% of
patients undergoing trisectionectomy.Multivariate regression
analysis identified seven inde-pendent factors predictive of
transfusion requirement(Table 1).
Table 1. Independent predictors of blood
transfusionrequirement.
Predictor P-valueOddsratio 95% CI
Coronary artery disease 0.009 2.769 1.287–5.960Preoperative
biliary drainage < 0.001 6.120 2.276–16.452Previous liver
resection 0.001 4.450 1.830–10.824Preoperative platelet count 0.002
1.005 1.002–1.008No. of segmentsresected: 3–4
0.507 1.300a 0.599–2.824
a) compared to 1–2 segments resectedNo. of segments resected: 5+
< 0.001 4.182a 1.874–9.328Hepatocellular carcinoma 0.03 2.443
1.092–5.464Preoperative haemoglobin< 10 g/dL
0.021 6.556b 1.325–32.440
b) compared to Hb > 12.5 g/dLPreoperative haemoglobin10–12.5
g/dL
0.041 2.019b 1.030–3.958
Conclusions: Results from this single centre study suggestthat
major liver surgery may be safely performed withfewer crossmatched
units of blood. Here we have iden-tified seven independent
predictors of transfusionrequirement. These factors could be used
as criteria forcross matching of blood, with group and save being
asafe and more cost effective measure in certain groups
ofpatients.
Abstracts 9
� 2009 The AuthorsJournal Compilation � 2009
Hepato-Pancreato-Biliary Association, HPB, 11 (Suppl. 1), 1–104
-
Abstracts 17 to 24Free Papers – Basic Science
Saturday, March 14, 2009 8:10–10:10 am
17
Targeting focal adhesion kinase inhibitspancreatic cancer growth
and metastasis
J. B. STOKES, MD, R. W. TILGHMAN, PHD,E. DAN HERSHEY, MS, J. K.
SLACK-DAVIS, PHD,J. THOMAS PARSONS, PHD and T. W. BAUER,
MDUniversity of Virginia, Charlottesville, VA, USA
Background: Focal adhesion kinase (FAK) is a cyto-plasmic
protein tyrosine kinase involved in the regulationof cellular
signaling, migration, apoptosis, and cell cycleprogression. FAK has
been shown to activate prolifera-tion and inhibit apoptosis in
cancer; however, its role inpancreatic cancer is not well
understood. We analyzedthe effects of PF-562,271, an inhibitor of
FAK, on pan-creatic cancer growth, invasion and metastasis.Methods:
The human pancreatic adenocarcinoma celllines L3.6pl and MPanc 96
were investigated in vitro,while MPanc 96 was used for in vivo
investigation. FAKphosphorylation was assessed by Western blot
analysis.Transwell migration assays were used to evaluate the
roleof FAK inhibition in pancreatic cancer cell migrationand
invasion after stimulation with growth factors andextracellular
matrix proteins. To evaluate the effects ofFAK inhibition in vivo,
we used an orthotopic mousemodel in which 33 mg/kg of PF-562,271
was adminis-tered twice daily by gastric lavage. Tumor volume
wasevaluated by magnetic resonance imaging (MRI) andtumor size,
retroperitoneal invasion, and metastasis wereevaluated at necropsy.
PF-562,271 was generously pro-vided by Pfizer Inc.Results: In
pancreatic cancer cell lines FAK phosphor-ylation was significantly
reduced by PF-562 271 admin-istration exhibiting an IC50 of 0.1 uM.
Stimulation byIGF-I resulted in a 7-fold increase (P < 0.05) in
cellmigration which was inhibited 94% by pretreatment withPF-562
271. Coating of transwell migration chambermembranes with
collagen-I increased migration 6-foldand this was inhibited 32% by
PF-562 271. In micebearing orthotopic pancreatic tumors, PF-562,271
ther-apy (vs. control) significantly inhibited tumor growth byMRI
volumetric analysis (58.75 mm3 vs. 127 mm3,P < 0.05) and by
pathologic examination (325 mm3 vs.816 mm3, P < 0.05). Treatment
with PF-562,271 alsoyielded significantly fewer abdominal
metastases (29%
vs.100%, P < 0.05), less retroperitoneal invasion(0% vs. 83%,
P < 0.05), and showed a trend towardsfewer liver metastasis (0%
vs. 50%, P = 0.06), comparedto control.Conclusions: Our study
demonstrates the important roleof FAK in pancreatic cancer growth,
invasion andmetastasis. These findings support FAK as a
potentialtarget for therapy in patients with pancreatic cancer.
18
The molecular mechanism of HIF-1independentVEGF expression in
hepatocellular carcinomacell line
S. B. CHOI, MD, J. B. PARK, PHD, K. S. KIM, MD,PHD and T. J.
SONG, MD, PHDYonsei University College of Medicine, Seoul, Republic
ofKorea;Research Institute and Hospital, National CancerCenter,
Ilsan, Republic of Korea; Korea University Collegeof Medicine,
Ansan, Republic of Korea
Purpose: Hypoxia-inducible factor-1 (HIF-1) is a
mastertranscription factor that plays a central role in
hypoxicexpression of various genes. The aim of this study was
toprovide molecular pathway of vascular endothelialgrowth factor
(VEGF) expression of HIF-1 independentpathway in hepatocellular
carcinoma cell line (Hep3B).Methods: HIF-1a, HIF-2a dominant
negative lentiviralvector was introduced to decrease the expression
of HIFin Hep3B cell line. Cells were incubated at 37�C
undernormoxic and hypoxic condition. We performed VEGFELISA using
supernatant, and Western Blotting todemonstrate the difference of
protein expression innormoxic or hypoxic condition. To validate the
HIF-1dependent or HIF-1 independent pathway, we treated thecells
with PI3K inhibitor and Erk kinase inhibitor.Results: VEGF level
was increased under the hypoxiccondition. HIF-1a protein expression
was induced inHep3B cell line after 24 h of exposure of hypoxia.
Weused siHIF-1a and siHIF-2a transfected Hep3B cell linewhich was
incubated in the normoxic or hypoxic condi-tion. The production of
the VEGF was done by theHIF-1 pathway. However significant portion
of theVEGF was produced by HIF-1 independent pathway.We treated the
vector, siHIF-1a and siHIF-2a transfectedcells with ERK inhibitor
before incubating normoxic orhypoxic condition. The VEGF expression
was not dif-ferent in each cell lines in hypoxic condition.
Thereforethe regulation of VEGF expression was not influenced byERK
pathway. We treated the vector, siHIF-1a andsiHIF-2a transfected
cells with PI3K inhibitor beforeincubating normoxic or hypoxic
condition. The PI3Kinhibitor decreased VEGF expression in the
siHIF-1atransfected cells in the hypoxic condition. We treated
thevector, siHIF-1a and siHIF-2a transfected cells withsiSP1
transient transfection before incubating normoxicor hypoxic
condition. The siSP1 transient transfectiondecreased VEGF
expression in the siHIF-1a transfectedcells in the hypoxic
condition. Therefore the VEGFregulation of Hep3B cell line was
mainly controlled by
160
140
120
100
80
60
40
20
0Week 0 Week 1 Week 2
Control
PF-562,271
Time
MRI Growth Curves
Tum
or
Volu
me
(mm
3 )
Week 3
Figure 1.
� 2009 The AuthorsJournal Compilation � 2009
Hepato-Pancreato-Biliary Association, HPB, 11 (Suppl. 1), 1–104
-
Akt/PI3K and SP1 pathway which was independent onHIF-1 in
hypoxic condition.Conclusions: Under the hypoxic condition,
VEGFexpression was controlled by HIF-1. However the VEGFexpression
by Akt/PI3K and SP1 pathway is not medi-ated by HIF-1 in the Hep3B
cell line.
19
Proteasome inhibition in combination therapy ofexperimental
pancreatic cancer: in vitro andin vivo evaluation
N. AWASTHI, PHD, M. A. SCHWARZ, MD andR. E. SCHWARZ, MDUT
Southwestern Medical Center, Dallas, TX, USA
Background: Use of targeted therapy to enhance clinicalbenefits
of cytotoxic chemotherapy has failed to improveresponses in
pancreatic ductal adenocarcinoma (PDAC).Proteasome inhibition (PI)
with agents such as bortezo-mib (B, Velcade) has shown anticancer
benefits throughincrease in proapoptotic mechanisms and cell
cycle-re-lated antiproliferative effects. This and the possibility
foran increased biologic activity of other agents after PIprovide
the rationale for PI combination therapy. Wehave previously shown
that endothelial monocyte acti-vating polypeptide II (EMAP II, E),
an antiangiogeniccytokine, enhances combination treatment effects
inexperimental PDAC through VEGF and integrin-relatedmechanisms.
Testing E and B for combination benefitsin vivo and in vitro was
the purpose of this study.Methods: In vitro studies used WST-1
proliferationassays with three human PDAC lines and
HUVECsendothelial cells (ECs) in a dose escalation matrix.Human
ASPC PDAC cells were used in a murine xeno-graft survival model.
Twelve days after i.p. or s.c. injec-tion of 7.5x10E5 tumor cells,
animals underwenttreatment with various combinations of E (80 ug/kg
i.p.daily), B (0.8 mg/kg i.p. twice weekly), or gemcitabine(G, 100
mg/kg i.p. twice weekly) for maximally 30 days.The resulting group
survival (n = 6/group) wascompared via logrank statistic.Results:
In 48-h. assays, B showed strong activity againstvarious PDAC cells
(IC50 range: 0.5-50 uM) and ECs(25 nM). In a 72-h. combination
assay, IC50s againstASPC (in uM) were: B = 0.1, G = 10, E=not
reached at> 50; addition of G and/or E to B at various doses did
notaffect proliferation differently. In a 48-hr. combinationassay,
IC50s against ECs (in uM) were: B = 0.025,G = 5, E = 10; addition
of G or E (at their IC20)increased B toxicity by 20%, and the
triple combination by42%, suggesting someadditive effects. After
initiation of invivo PDAC therapy, median group survival (d) was:
con-trol = 19, B = 18, E = 20, G = 28; of all three mono-therapy
groups, only G had a significant survival impact(P = 0.02). All
combination groups except B+E led toextended survival compared to
monotherapies or control(P = 0.003). However, addition of B to E,G,
or E+Gdidnot enhance survival (p=NS for all three
comparisons).Conclusions: PI with bortezomib mediates strong
anti-proliferative effects against PDACs and ECs in vitro, butfails
to enhance survival. Although combinations aremore effective than
monotherapies, the in vivo synergisticpotential is limited. The
findings highlight both benefitsand challenges of combination
therapy approaches, whilesuggesting limitations of cancer cell in
vitro assay efficacypredictions.
20
Anti-inflammatory effects of the nigella sativaseed extract,
thymoquinone, in pancreaticcancer cells
H. A. ARAFAT, MD, PHD, N. CHEHL, B.SC.,G. CHIPITSYNA, PHD, Q.
GONG, MD andC. J. YEO, MDThomas Jefferson University, Philadelphia,
PA, USA
Background: Both hereditary and sporadic forms ofchronic
pancreatitis are associated with an increased riskof developing
pancreatic ductal adenocarcinoma (PDA).Inflammation has been
identified as a significant factor inthe development of solid tumor
malignancies. Thymo-quinone (Tq), the major constituent of the
Nigella sativaoil extract induces apoptosis and inhibits PDA cell
pro-liferation. Tq also increases p21WAF1 expression,
inhibitshistone deacetylase (HDAC) activity, and induces his-tone
hyperacetylation. HDAC inhibitors have beenshown to ameliorate
inflammation-associated cancer inseveral animal models.Objective:
To evaluate the anti-inflammatory potentialof Tq in PDA cells in
comparison to a specific HDACinhibitor, trichostatin A
(TSA).Methods: PDA cells (AsPC-1, HS766T; MiaPaca) werecultured and
treated with or without Tq (25–75 lM),with or without pre-treatment
of TNF- a (30 nM). Theeffect of Tq on the expression of different
proinflamma-tory cytokines and chemokines was analyzed by real
timePCR. Luciferase-labeled promoter studies evaluated theeffect of
Tq on the transcription of monocyte chemo-attractant protein-1
(MCP-1) and nuclear factor-jB(NF-jB). The effect of Tq on the
endogenous and TNF-a-induced activation and nuclear translocation
of NF-jBwas examined by ELISA and immunohistochemistry.Results:
Within 6 h, Tq significantly and dose-depen-dently reduced PDA cell
production of TNF-a (P <0.02), interleukin (IL-1b) (P <
0.02), IL-8 (P < 0.05),Cox-2 (P < 0.002), and MCP-1 (P <
0.005). There wasno reduction in interferon-c (IFN-c) in the same
cultures.Within the same time period, TSA reduced the produc-tion
of Cox-2 (P < 0.02) and MCP-1 (P < 0.05), buthad no effect on
TNF-a, IL-8, or IL-1b. Tq, but notTSA, significantly and
dose-dependently reduced theintrinsic activity of the MCP-1
promoter. Tq alsoinhibited the intrinsic and the TNF-a-mediated
activa-tion of NF-jB in PDA cells and reduced the transport ofNF-jB
from the cytosol to the nucleus.Conclusions: Our data demonstrate
previously unde-scribed anti-inflammatory activities of Tq in PDA
cells,which are paralleled by inhibition of NF-jB. Tq as anovel
inhibitor of proinflammatory pathways provides apromising strategy
that combines anti-inflammatory andproapoptotic modes of
action.
21
A proteomic based platform for markerdiscovery in
cholangiocarcinoma
G. K. BONNEY, R. CRAVEN, A. MELCHER,P. SELBY, R. BANKS and R.
PRASADSt James’ University Hospital, Leeds, UK
Cholangiocarcinoma (CCA) is the second commonestprimary
malignancy of the liver with a rising incidenceworldwide and a
dismal prognosis. There remains a lack
Abstracts 11
� 2009 The AuthorsJournal Compilation � 2009
Hepato-Pancreato-Biliary Association, HPB, 11 (Suppl. 1), 1–104
-
of sensitive and specific biomarkers for this
diseaseparticularly in distinguishing it from patients with
benignand predisposing diseases such as Primary
SclerosingCholangitis.Bile, directly draining the liver, may
contain higherconcentrations of biomarkers than those found in
thegeneral circulation, as these may be tumour derived, shedor
secreted proteins. The proteomic analysis of bile indisease
pathogenesis (i.e. malignancy and gallstone dis-ease), drug
metabolism and biomarker discovery hasreceived recent interest. We
have developed a reproduc-ible and accurate method of quantifying,
desalting anddelipidating bile prior to proteomic analysis. Using
thismethod we ran two parallel proteomic based approaches,with the
aim of marker discover in CCA.The first, using Differential In Gel
Electrophoresis(DIGE), was a study comparing bile from 4 groups
ofpatients: healthy liver donors (n = 5), benign obstruc-tion (n =
5), Primary Sclerosing Cholangitis (n = 3)and CCA (n = 5). This
study resulted in six proteins thatwere significantly
differentially expressed between groups(downregulated n = 5,
upregulated n = 1) which werethen identified by mass spectrometry.
Downstreamvalidation of the upregulated protein was performed
byWestern blotting, which confirmed a significant differ-ence in
abundance of this protein between the fourgroups.The second method
used a Shotgun based approach,namely GeLC-MS/MS. For this, a sample
of bile from apatient with CCA was analysed and over 1200
proteinsidentified. In validations studies, by Western blotting,two
of these biliary proteins were differentially upregu-lated in CCA
when compared to other disease groups.Using a proteomic based
approach we have identified 3differentially expressed proteins in
the bile of patientswith CCA compared to normal, benign and
predisposedpatient groups.
22
Microenvironment-induced gene expressionchanges in breast cancer
liver metastases
S. TABARIÈS, PHD, Z. DONG, MD, F. PÉPIN,M. HALLETT, PHD, A.
OMERGLU, MD,M. HASSANAIN, MD, P. METRAKOS, MD andP. M. SIEGEL,
PHDMcGill University, Montreal, QC, Canada
Introduction: Breast cancer is the most common canceraffecting
Canadian women and is the second leadingcause of cancer related
deaths in these patients. Theacquisition of metastatic abilities by
breast cancer cells isthe most deadly aspect of the disease. Upon
dissemina-tion from the primary tumor, breast cancer cells
displaypreferences for specific metastatic sites. The liver
repre-sents the third most frequent site for breast
cancermetastasis, following bones and lungs. Despite the evi-dence
that hepatic metastases are associated with poorclinical outcome in
breast cancer patients, little is knownabout the molecular
mechanisms governing the spreadand growth of breast cancer cells
within the liver.Methods: We have employed 4T1 murine
mammarycarcinoma cells that were subjected to three rounds
ofsplenic injection, which permitted the isolation of breastcancer
cells that aggressively grow in the liver. LaserCapture
Microdissection (LCM) was used to identifygene expression changes
within breast cancer liver
metastatic cells that occur in situ in response to the
livermicroenvironment.4T1 breast cancer cells located at themargin
and the centre of several metastatic lesions havebeen individually
sampled by LCM and the recoveredRNA subjected to linear
amplification, labeling andhybridization to Agilent microarrays. To
date, we haveanalyzed four matched sets of margin vs. core
livermetastases samples. Moreover, we have complementedthe analyses
of 4T1-derived lesions with LCM experi-ments performed on liver
metastases isolated from breastcancer patients.Results: Among
several candidate genes that were dif-ferentially expressed between
the margin and core of theliver metastases, the gene encoding for
the leptin receptorappeared to be up-regulated in the margin. The
Leptin/LEPR axis has gained considerable attention as animportant
regulator of breast cancer progression. Indeed,Leptin and LEPR are
over-expressed in breast cancerscompared to normal breast
epithelium and Leptin hasbeen shown to stimulate the proliferation
of breast cancercells. We have validated LEPR expression in
the4T1-derived liver lesions, and more importantly, observeLEPR
positivity in multiple liver metastases obtainedfrom breast cancer
patients.Conclusion: The identification and functional validationof
candidate genes important for the ability of breastcancer cells
will provide basic insights into the pathwaysrequired for breast
cancer cells to metastasize to the liver.Our results suggest that
the LEPR may play an impor-tant role in enabling breast cancer
cells to metastasize tothe liver.
23
Validation of a novel, physiologic model ofexperimental acute
pancreatitis in the mouse
N. J. ZYROMSKI, MD, T. E. WADE, MD,H. A. PITT, MD, S. WANG, MD
andD. A. SWARTZ-BASILE, PHDIndiana University, Indianapolis, IN,
USA
Background: Acute pancreatitis is a devastating diseasethat
affects 240 000 Americans each year. No specifictreatment for acute
pancreatitis currently exists, largelybecause its precise
pathophysiology is poorly understood.Murine experimental models are
attractive, as completeknowledge of the mouse genome permits
precise geneticmanipulation. Unfortunately, current methods used
toinduce acute pancreatitis in the mouse (cerulein
hyper-stimulation, intravascular bile salt infusion,
supraphysi-ologic arginine administration) are of
questionableclinical relevance. Therefore, the aim of the current
studywas to validate a recently reported murine model of
acutepancreatitis that is more representative of the humandisease
process.Methods: Twenty C57BL/6J and 11 CF-1 mice werestudied.
Under general anesthesia, transduodenal cann-ulation of the
pancreatic duct was accomplished with a30-gauge catheter, and 50 lL
of 5% Sodium Taurocho-late (NaT) or 0.9 normal Sodium Chloride
(NaCl) wasinfused. Mice were euthanized 24 h later. Three
observersrated pancreatitis severity by light microscopic
evalua-tion of H&E sections. A validated scale
incorporatingdegree of edema, vacuolization, and inflammatory
cellinfiltrate comprise the total pancreatitis score.
Pancreatictissue concentration of the chemoattractant molecule
Abstracts12
� 2009 The AuthorsJournal Compilation � 2009
Hepato-Pancreato-Biliary Association, HPB, 11 (Suppl. 1), 1–104
-
monocyte chemoattractant protein-1 (MCP-1) and
theproinflammatory cytokine interleukin-6 (IL-6) weredetermined by
ELISA. ANOVA and Student’s t-test wereapplied where appropriate; P
value < 0.05 was acceptedas statistically significant.Results:
Thirteen mice (NaCl – 6; NaT – 7) survived for24 h. The total
pancreatitis score was significantly greaterin mice undergoing
retrograde pancreatic duct infusionof NaT (Table 1). Pancreata of
mice infused with NaTdemonstrated significant necrosis, consistent
with severeacute pancreatitis. Pancreatic concentrations of
MCP-1and IL-6 are shown in the Table 1.
Table 1. Pancreatic concentrations of MCP-1 and IL-6.
Pancreatitisscore
MCP-1(pg/mg)
IL-6(pg/mg)
NaCl(n = 6)
1.2 ± 0.4 2350 ± 1386 452 ± 269
NaT(n = 7)
6.3 ± 1.2* 3633 ± 1853 2028 ± 1612*
*P < 0.05 vs. NaCl
Conclusions: Retrograde pancreatic duct infusion ofSodium
Taurocholate induces severe acute pancreatitis inthe mouse. Though
associated with a discrete learningcurve, this model is likely more
representative humanpancreatitis pathophysiology, and therefore
provides apowerful tool with which to elucidate clinically
importantbasic mechanisms underlying the pathogenesis of
acutepancreatitis.
24
The expression of interferon receptor alpha/betain human
pancreatic cancer in nude mice isessential for tumor response to
interferon alphatreatment
R. F. SAIDI, MD, A. W. AHAD, MD,I. NALBANTOGLU, MD and M. J.
JACOBS, MDMassachusetts General Hospital, Boston, MA,USA;Providence
Hospital, Southfield, MI,USA;Department of Pathology, St John
Hospital, Detroit,MI, USA; Department of Surgery, Providence
Hospital,Southfield, MI, USA
Introduction: Adjuvant interferon (IFN) therapy
andchemoradiation status post pancreaticoduodenectomy
for patients with pancreatic cancer have renderedpromising
results.The aim of this study was to evaluatethe in vivo effect of
interferon alpha on human pan-creatic carcinoma implanted
orthotopically into nudemice.Material and methods: Human pancreatic
cancer celllines MiaPaCa-2 and Panc-1 were used. MiaPaCa-2 isknown
to express the interferon alpha/beta receptor andPanc-1 cells do
not. The cells were implanted into thepancreas of nude mice and
treatment was initiated sevendays later. Regimen I consisted of
intraperitoneal single-agent gemcitabine (125-mg/kg biweekly) and
Regimen IIconsisted of IFN-alpha (10 000-units daily,
subcutane-ously) and gemcitabine biweekly for 30 days. Animalswere
sacrificed after 30 days or if they became moribund.Body weight was
determined and the primary tumors inthe pancreas were excised,
measured, and weighed. Visi-ble metastases or adjacent organ
invasion were countedand processed for H&E staining. All
macroscopicallyenlarged regional (celiac and para-aortal) lymph
nodeswere harvested and the presence of metastatic disease
wasconfirmed by histology.Results: The mice that were implanted
with MiaPaCa-2cells showed a more dramatic response to Regimen
IIwhen compared to Panc-1 implanted mice. The Mia-PaCa-2 group that
was treated with Regimen II showedan 87% reduction in tumor volume
compared to 50% inthe Panc-1 group treated with the same regimen(P
< 0.001). Mice implanted with MiaPaCa-2 andtreated with Regimen
II showed less metastasis, less localinvasion, and a longer
survival compared to miceimplanted with Panc-1 treated with same
regimen.Regimen II was more effective on MiaPaCa-2 comparedto
Regimen I (p
-
Abstracts 25 to 31Free Papers – Transplant
Saturday, March 14, 2009 2:00–4:00 pm
25
Is the use of donors with high donor risk indicescost-effective
in liver transplantation?
D. E. MOORE, MD, MPH, I. FEURER, PHD andC. W. PINSON, MD,
MBAVanderbilt University, Nashville, TN, USA
Background: The ever widening gap between the avail-ability of
donor organs and the number of liver trans-plant candidates
requires the maximal use of allavailable grafts, including those
with high donor riskindices (DRI). High DRI organs are classified
as thosehaving DRIs greater than 2.0, while low DRI organshave DRIs
less than 2.0. Organs with a DRI above 2.0account for about 10% of
the donor pool. Recipientsreceiving high DRI organs have been shown
to havetwice the length of stay and a two fold increase inhospital
costs when compared to recipients of lowDRI organs. Unless these
grafts are used judiciously,outcomes and transplant resource
utilization will benegatively affected. The aim of this study was
to eval-uate the cost-effectiveness of an organ allocation
schemeusing high and low DRI organs compared to an allo-cation
scheme using only low DRI organs. Additionally,the impact of these
schemes on wait list mortality will beevaluated.Methods: A
Markov-based decision analytic model wascreated to simulate
outcomes for liver transplantation inrecipients of organs from low
and high DRI donorsversus recipients of organs from low DRI donors.
Asecond model was created to simulate outcomes that re-flect
alterations in wait list mortality under both organallocation
schemes. Baseline values and ranges weredetermined from the UNOS
database (19 000 recipientstransplanted between 2002 and 2007) and
Medicare costdata. Sensitivity analyses were conducted to test
modelstrength and parameter variability.Results: Recipients of
organs from low DRI donors hada three year survival of 75% versus
65% for recipients oforgans from high DRI donors. After
transplantation,recipients of high DRI organs gained 5.1
QALYs(Quality Adjusted Life Years) at a cost of $75,000/QALYversus
5.9 QALYs at a cost of $52 000/QALY forrecipients of low DRI
organs. However, the allocationscheme in which both low and high
DRI organs wereused together proved more cost-effective when
comparedto the scheme in which only low DRI organs were usedand
there was a 25% increase in wait list mortality; 5.8QALYs at a cost
of $54 000/QALY versus 5.6 QALYs ata cost of $59
000/QALY.Conclusions: High DRI donors provide a significantnumber
of grafts. While these grafts may represent anincreased cost per
individual transplant, the overallcontribution of these grafts to
the donor pool and asubsequent reduction in wait list mortality
make themcost-effective.
26
The development of a classification system forbiliary
complications following orthotopic livertransplantation
A. NEVILLE, MD, M. BOUTROS, MD andJ. BARKUN, MDRoyal Victoria
Hospital, McGill University, Montreal,QC, Canada
Introduction: Biliary tract complications remain a sig-nificant
source of morbidity and mortality followingorthotopic liver
transplantation (OLTx). Even thoughthese occur with an estimated
incidence of 10–40% andmay lead to graft failure, there exists no
universallyaccepted classification system. As such, descriptions
ofbiliary complications in the literature often lack consis-tency
and clarity which limits evaluation and comparisonof
post-transplantation outcomes. We propose a struc-tured
classification system for biliary complications fol-lowing
choledocho-choledochal anastomosis (CCA) atOLTx. The classification
is based on four major com-ponents: leaks (anastomotic or cystic
duct, anastomoticdisruption), filling defects (stones/sludge/casts
in theintrahepatic, recipient or donor common ducts), stric-tures
(unilateral intrahepatic, common hepatic or anas-tomotic) and
development of sclerosing cholangitis.Complications are identified
and classified based on theirappearance on a contrast
cholangiogram, mostly endo-scopic cholangiopancreatography (ERCP).
The goal ofthis initial report is to ensure that the proposed
classifi-cation is consistent and reliable.Methods: Patients having
undergone OLTx with CCA atthe McGill University Health Center from
2004 to 2007and who had at least one contrast cholangiogram
post-operatively were candidates for the study. ERCP filmswere
assessed by two independent reviewers to determineinter-rater
reliability and the kappa coefficient of agree-ment of
classification of the complication.Results: One hundred and eight
ERCP films werereviewed. Overall inter-rater reliability (percent
agree-ment) among the reviewers was 88.5%. Agreement onspecific
elements of the classification scheme (leak, stric-ture, filling
defect) was similarly high. Agreement amongstudy reviewers was
greater than agreement betweenstudy reviewers and the official
radiographic report(77.9%), although this was not statistically
significant.The kappa coefficient was 0.87 (95% confidence
interval0.79-0.94).Conclusion: The proposed classification system
demon-strates high reliability and would allow for
consistentinterpretation and reporting of biliary
complicationsfollowing OLTx. Validation of this classification
systemis ongoing through correlation of complications with
theend-points of patient survival, graft survival, number
ofadditional interventions required, length of hospital stayand
number of readmissions.
� 2009 The AuthorsJournal Compilation � 2009
Hepato-Pancreato-Biliary Association, HPB, 11 (Suppl. 1), 1–104
-
27
Regionalization of liver transplantation to highvolume centers
is associated with diminishedvolume-outcomes relationship and
underutiliza-tion of high volume centers by minority patients
E. T. TRACY, MD, E. M. AVIKI, T. N. PAPPAS, MD,B. H. COLLINS,
MD, J. E. TUTTLE-NEWHALL,MD, C. E. MARROQUIN, MD, P. C. KUO, MD
andJ. E. SCARBOROUGH, MDDuke University, Durham, NC, USA
Introduction: Our purpose was to determine the temporaltrends in
liver transplant center volume as well as theeffect of those trends
on both the volume-outcomesrelationship and the utilization of high
volume centers byminority patients.Methods: We performed a
retrospective analysis of35 374 adult orthotopic liver transplant
proceduresincluded in the Scientific Registry of Transplant
Recipi-ents (SRTR) for three consecutive 30-month time
periodsbetween 1999 and 2006. Transplant centers were dividedinto
five categories based on annual volume (very low-volume < 36
procedures per year, low volume = 37–55,moderate-volume 56–79,
high-volume group = 80–129,very high volume > 130). One-month
and 1-yearobserved-to-expected patient death ratios (provided bythe
SRTR), were calculated for each time period.Comparisons between
volume groups in each time periodwere made using chi square
analysis. Trends in thedemographic characteristics were also
assessed.Results: The percentage of liver transplants performed
atlower volume centers decreased significantly from 51% inPeriod 1
to 33% in Period 3 (P < 0.0001), while thoseperformed at higher
volume centers increased signifi-cantly from 27% in Period 1 to 44%
in Period 3(P < 0.0001). In Period 1, the 1-month and 1-year
O:Epatient death ratios at very low-volume centers (1-monthratio
1.33, 1-year ratio 1.18) were significantly higherthan the ratios
at very high volume centers (1-monthratio 0.89, 1-year 0.93, P <
0.0001 at both 1-month and1-year). By Period 3, the disparity
between very low andvery high volume centers was no longer
statistically sig-nificant at either 1 month or 1 year. Meanwhile,
inPeriod 1 the percentage of Black or Hispanic patients atvery
high-volume centers (28.4%) was significantlygreater than at very
low-volume centers (20.6%,P < 0.0001). By Period 3, the
percentage of Black orHispanic patients at very high volume centers
hadbecome significantly lower (18.1%) than at low volumecenters
(22.7%, P = 0.0004).Conclusions: As the percentage of liver
transplants per-formed at higher volume centers has increased over
time,the volume-outcomes relationship has become lessprominent.
However, regionalization of liver transplan-tation does appear to
be associated with a growingunderutilization of high volume centers
by minoritypatients.
28
Is magnetic resonance imaging superior tocomputed tomography for
preoperativeevaluation of cystic lesions of the pancreas?
C. K. CHU, MD, A. E. MAZO, B.SC., C. A. STALEY,MD, J. M.
SARMIENTO, MD, D. R. MARTIN, MD,W. E. TORRES, MD, N. V. ADSAY, MD
andD. A. KOOBY, MDEmory University School of Medicine, Atlanta, GA,
USA
Purpose: Cystic lesions of the pancreas (CLP) represent adiverse
range of pathologies. Appropriate managementdepends on accurate
diagnostic imaging. Althoughmagnetic resonance imaging (MRI)
provides a theoreti-cal advantage in improved soft-tissue contrast,
nodefinitive superiority in its diagnostic capability has
beenestablished over computed tomography (CT) in evalua-tion of
CLP.Methods: Database review identified patients undergoingsurgical
management at our institution between 1/1/00and 12/31/07 for CLP as
confirmed by postoperativepathologic analysis. Records of patients
who alsounderwent same-institution preoperative CT and/or MRIwere
selected for study. Diagnostic accuracy and confi-dence were rated
on a 5-point scale (Table 1) for eachradiologic study as compared
with histologic review.Mann-Whitney U and Chi-square tests were
used tocompare scores for CT and MRI.Results: One hundred and
forty-eight patients receivedoperative interventions for CLP. Of
these, 96 (65%)underwent preoperative CT and/or MRI (43 CT only,41
MRI only, 12 both) at our institution with a total of108 radiologic
studies. Histologically, 24 patients (25%)had serous cystadenomas;
23 (24%), intraductal papil-lary mucinous neoplasms (IPMN); 20
(21%), mucinouscystadenoma/adenocarcinomas; 19 (20%), pseudocysts;
7(7%), solid pseudopapillary tumors; 2 (2%), simplescysts; and 1
(1%), lymphoepithelial cyst. No significantdifference between
diagnostic scores of CT and MRI wasdetected (median 3 vs. 2, P =
0.14). However, only43.6% of CT studies received scores of 1 or 2,
while62.3% of MRI studies achieved the same (P = 0.05).There was a
trend towards improved preoperative iden-tification of IPMN using
MRI as evidenced by percent ofscore 1 (50% by MRI vs. 13% by CT, P
= 0.075). Nosuch trend was observed among cystadenomas.Conclusions:
MRI may provide greater diagnostic accu-racy and confidence in
preoperative evaluation of cysticlesions of the pancreas. MRI is
prudent when uncertaintyregarding diagnosis and management of
CT-visualizedlesions arise. Additionally, cystic lesions suspicious
forIPMN may be better evaluated for ductal involvementwith MRI than
CT.
Abstracts 15
� 2009 The AuthorsJournal Compilation � 2009
Hepato-Pancreato-Biliary Association, HPB, 11 (Suppl. 1), 1–104
-
Table 1. Diagnostic scoring system used for CT andMRI.
Score
Degree of accuracy and confidence of radiologicreport results in
comparison to surgicalpathology results
1 Definitive: correct answer proposed as single ortop
diagnosis
2 Suggestive: mentions correct answer within anarrow range (3 or
less) possible diagnoses ofequal likelihood
3 Somewhat suggestive: mentions correct answerwithin a wide
range (greater than 3) of possiblediagnoses
4 Vague: does not mention possibility of correctdiagnosis
anywhere within report
5 Contradictory: correct diagnosis not proposedas possibility,
and top proposed diagnosisinaccurate or contradictory to correct
diagnosis
29
Comparative performance of staging systemsfor early
hepatocellular carcinoma
H. NATHAN, G. MENTHA, H. P. MARQUES,L. CAPUSSOTTI, P. MAJNO, L.
ALDRIGHETTI,C. PULITANO, L. RUBBIA-BRANDT,N. RUSSOLILLO, B.
PHILOSOPHE, E. BARROSO,A. FERRERO, M. A. CHOTI and T. M. PAWLIKThe
Johns Hopkins University, Baltimore, MD, USA
Introduction: While several staging systems for hepato-cellular
carcinoma (HCC) have been proposed for pa-tients (pts) with a wide
spectrum of HCC disease, mostsurgical candidates in fact have early
HCC. The ability ofstaging systems to discriminate prognosis after
surgeryfor early HCC is unknown. We sought to assess theperformance
of major HCC staging systems in a largecohort of pts with early
HCC.Methods: Data on pts undergoing liver resection (LR)
ortransplantation (LT) for early HCC (£ 5 cm, N0M0, nomajor
vascular invasion) with well-compensated cirrhosiswere collected
from an international group of sevenhepatobiliary centers. We
evaluated five staging systems:American Joint Committee on Cancer
(AJCC), Interna-tional Hepato-Pancreato-Biliary Assoc. (IHPBA),
Japa-nese Integrated Staging (JIS), Cancer of the Liver
ItalianProgram (CLIP), and Okuda. A previously proposedscoring
system for early HCC was also evaluated; thissystem allots 1 point
each for size > 2 cm, multifocality,and microscopic vascular
invasion (mVI). The discrimi-native abilities of the systems were
quantified via boot-strap-corrected concordance indices
(c).Results: Of 366 eligible pts, 236 underwentLRand 130LT.Overall
survival was 74% at 3 years and 53% at 5 years(median 63 months).
While most pts had size > 2 cm(n = 278, 76%), a minority had
multifocality (n = 47,13%)ormVI (n = 76, 21%).Thedistributionofpts
amongstages was: AJCC: T1 (n = 250), T2 (n = 116); IHPBA:T1 (n =
67), T2 (n = 273), T3 (n = 26); JIS 0 (n = 51),1 (n = 249), 2 (n =
54), 3 (n = 12); CLIP: 0 (n = 206),1 (n = 81), 2 (n = 19), 3 (n =
2); Okuda: A (n = 69), B(n = 219), C (n = 3); early HCC scoring
system: 0
(n = 58), 1 (n = 219), ‡ 2 (n = 89). Except for the AJCCsystem,
all established staging systems performed poorly asjudged by the
c-statistics (Table 1). The proposed scoringsystem for early HCC
demonstrated the best prognosticdiscrimination of all the evaluated
systems (Table 1).
Table 1. Survival by stage for five existing staging systemsand
a proposed scoring system for early HCC.
5-year survival, by stage/score (%)
AJCC IHPBA JIS CLIP Okuda
Early HCCScoringSystem
T1/Score 0/B 59 59 64 52 45 67T2/Score 1/B 40 52 50 47 53
55Score 2/C No pts. 51 57 50 67 39Score 3 No pts. No pts. 47 100 No
pts. No pts.c-Statistics 0.585 0.518 0.502 0.510 0.508 0.592
Conclusions: Most staging systems perform poorly whenused for
prognostic stratification of pts with early HCC.Although the AJCC
staging provides moderate discrim-inative ability, it only includes
two groupings (T1 and T2)that constitute early HCC. A simple
scoring system forearly HCC provides superior discriminative power
thanexisting staging systems.
30
Liver retransplantation – a 25-year experience
V. SHANMUGAM, C. BHATI, R.MARUTHANAYAGAM, J. BUCKELS, D. F.
MIRZAand S. R. BRAMHALLQueen Elizabeth Hospital, Birmingham, UK
Retransplantation (re-LT) is the only viable option
forirreversible graft failure after primary transplantation.Due to
the growing disparity between the number ofpeople listed for
primary Liver transplantation andorgans available and its inferior
outcomes when com-pared to primary transplantation, re-LT is
certainlycontroversial. The aim of this study was to analyse
theindications and outcomes of re-LT since the beginning ofour
programme.Methods: The clinical and demographic data for
thisretrospective study from 1982 to 2007 was collected fromour
dedicated transplant database. During this period atotal of 2420
adult transplants were performed at ourcentre.Results: A total of
196 re-LT patients underwent 225transplantations (8.09%). The mean
age was 42.99 years(SEM±0.09) with the median age of 44.85 years
(16.4–68.3). The number of retransplants has gone down stea-dily
due to the better techniques and newer immunesuppression. Out of
196 re-LT, 23 had second re-LT(0.95%) and 6 had their fourth re-LT
grafts(0.24%).More then 1/3 in our study (37.6%) requiredregrafting
within the 6 months of primary transplant.The most common
indication was Hepatic arterythrombosis (31.5%), followed by
chronic rejection(22.6%) and recurrent disease (14.2%). The patient
sur-vival and graft survival are as shown in Table 1.Early
retransplants (< 7 days) of the primary trans-plant had a worse
outcome than late retransplants
Abstracts16
� 2009 The AuthorsJournal Compilation � 2009
Hepato-Pancreato-Biliary Association, HPB, 11 (Suppl. 1), 1–104
-
(P = 0.001). Patient survival after retransplantation inpatient
with PNF and Haemorrhagic necrosis had pooroutcome in comparison to
others (P = 0.008). MELDdid not have any impact on patient survival
(P = 0.2)Conclusion: Long terms graft as well as patient survivalin
patients with regraft populations is significantly lowerthen
patients with primary grafts. Patient survival de-pends on the
indication for retransplantation. Retrans-plants within 7 days had
a poor outcome than the latetransplants. Higher MELD did not have
any impact onpatient’s survival. Retransplantation is necessary but
theclinical selection of recipient remains of
paramountimportance.
Table 1. Patient (PS) and graft survival (PS).
First ReLTxPS GS
Second ReLTxPS GS
Third ReLTxPS GS
30 days 83% 82% 64% 59% 67% 60%90 days 73% 72% 59% 59% 24% 40%1
year 66% 65% 45% 45% 24% 20%3 years 61% 59% 40% 36% 0%5 years 57%
54% 40% 32% 0%10 years 47% 43% 25% 24% 0%Total noof patients
196 23 6
31
Initial experience using Hepatitis C positiverenal allografts in
elderly Hepatitis C negativerecipients
T. R. FLOHR, MD, D. KEITH, MD, P. I. LOBO, MD,B. R. SWENSON, MD,
H. BONATTI, MD,T. M. SCHMITT, MD, R. G. SAWYER, MD,T. L. PRUETT, MD
and K. L. BRAYMAN, MD, PHDUniversity of Virginia, Charlottesville,
VA, USA
Introduction: First year mortality for elderly end stagerenal
disease (ESRD) patients is 1.62–2.53 fold higher ascompared to
younger ESRD patients. Renal transplan-tation is an option that
provides improved survival,
however, the lack of suitable organs, longer wait-timesand
worsening co-morbidities are great impedances forthe elderly.
Considering these difficulties, the use ofkidneys from donors with
serology positive for HepatitisC virus (HCV) has been explored. We
report our expe-rience with nine HCV negative elderly patients
whomunderwent renal t