studies

HEPATOLOGY, VOLUME 62, NUMBER 1 (SUPPL) AASLD ABSTRACTS 819A
had a positive test and 14 were negative. The 14 drinking subjects
who were PEth-negative reported low levels of use. This
study confirms that, despite advice to abstain, more than one
quarter of alcohol-addicted LT recipients return to alcohol use.
PEth identifies alcohol use in LT recipients who deny drinking.
Liver transplant centers may want to include PEth as part of
routine post-LT care in order to detect alcohol use in recipients
who deny drinking and thereby provide them with early interventions,
designed to restore sobriety.
Disclosures:
Michael R. Lucey - Advisory Committees or Review Panels: Galectin; Grant/
Research Support: Vertex, Abbvie, Gilead, Salix
The following authors have nothing to disclose: Michael F. Fleming, Jenny Vue
1232
A New Score based on explant pathology allows an
individualized prediction of HCC-recurrence after liver
transplantation
Charlotte E. Costentin 1 , Giuliana Amaddeo 1 , Christophe Duvoux 1 ,
Julien Calderaro 2 , Alexis Laurent 3 , Ariane Mallat 1 , Françoise
Roudot-Thoraval 4 ; 1 Hepatology, Hôpital Henri Mondor, Creteil,
France; 2 pathology, Henri Mondor Hospital, Creteil, France;
3 Surgery, Henri Mondor Hospital, Creteil, France; 4 Public Health,
Henri Mondor Hospital, Creteil, France
Aim: After liver transplantation (LT) for hepatocellular carcinoma
(HCC), a standardized prediction of recurrence would
valuable in order to define patients who might benefit from post
LT adjuvant therapy or early changes in immunosuppressive
regimens. We recently compared the accuracy of 4 explantbased
models for prediction of recurrence (EASL 2015). In
this study, the “Up to seven model”, using 2 variables (size
of largest nodule and number of nodules) appeared to be the
best tool to identify patients with a high risk of recurrence at
5 years post LT. The aim of this study was to design a new
explant-based model including vascular invasion and tumor differentiation
as predictors of recurrence and to test its accuracy
against the “Up to seven” model. Methods: Following uni- and
multi-variate analysis of pathological predictors of recurrence
in a series of 372 liver explants of patients transplanted for
HCC between 2003 and 2005 and followed prospectively for
5 years, we built 2 predictive models. New score 1 included
all independent pathological predictors (number of nodules,
size of the largest, tumor differentiation, micro/macrovascular
invasion, tumor burden (uni or bilobar)) and ranged from 0 to
9 points. New score 2 included all above mentioned variables
except tumor differentiation (a variable that is not available
when tumor is totally inactivated after bridging therapy) and
ranged from 0 to 11 points. The accuracy of these 2 scores
and the “Up to seven” model without vascular invasion to predict
5 year-HCC recurrence was assessed by comparison of
AUCs of ROC curves which were subsequently compared using
the Hanley&McNeil method. Results: The “Up to seven model”
identified two distinct risk groups for 5 year-recurrence of HCC
(10.8±2.2% if score ≤7 ; 54.5±4.5% si score >7). New scores
1 and 2 identified 4 groups of patients with different levels
of risks of recurrence ranging from 10 to >70% at 5 years.
AUCs were 0.7915 [0.73394 - 0.84907] for the “Up to seven
model, 0.7881 [0.73135 - 0.8448] for New score 1 and
0.7889 [0.7352 - 0.8427] for New score 2 (p=0.743). Tumor
differentiation did not improve accuracy to predict 5 year-HCC
recurrence (New score 1 vs New score 2, p=0.49). Using
New score 2, a nomogram was built, giving the probability of
recurrence at 1, 3 and 5 years post transplantation, according
to the score value (fig) Conclusion: A new score using number
of nodules, size of the largest, vascular invasion and tumor
burden on explant has similar accuracy compared to the “Up
to seven” model to predict 5 year-HCC recurrence but allows a
more individualized prediction of recurrence based on patients’
profile of explant findings.
Disclosures:
Christophe Duvoux - Advisory Committees or Review Panels: Novartis, Roche,
Novartis, Roche, Novartis, Roche, Novartis, Roche; Speaking and Teaching:
Astellas, Astellas, Astellas, Astellas
Françoise Roudot-Thoraval - Advisory Committees or Review Panels: Roche; Consulting:
LFB biomedicaments; Speaking and Teaching: gilead, Janssen, BMS,
Roche
The following authors have nothing to disclose: Charlotte E. Costentin, Giuliana
Amaddeo, Julien Calderaro, Alexis Laurent, Ariane Mallat
1233
Comparative analysis of renal dysfunction in recipients
with NASH compared to controls with alcoholic liver
disease after liver transplantation
Cheri Ogwo 1 , Satheesh Nair 2,1 , Jason Vanatta 2,1 , Vinaya Rao 2,1 ,
Chibuzor Iwelu 1 , James Eason 2,1 , Sanjaya K. Satapathy 2,1 ; 1 Surgery,
University of Tennessee Health Sciences Center, Memphis,
TN; 2 Surgery, Methodist University Hospital, Memphis, TN
BACKGROUND: Association between NAFLD/NASH and
chronic kidney disease (CKD) has been well recognized.
Release of pro-inflammatory mediators from steatotic liver or
through the contribution of metabolic syndrome (Obesity, type
2 DM, dyslipidemia, and hypertension) has been postulated
as the etiology. AIM: We aim to assess if the predisposition
for renal dysfunction persists after liver transplantation (LT) in
recipients with NASH compared to controls with alcoholic liver
disease (ALD). METHODS: Charts of 806 adult LT recipients
(6/2006- 12/2012) reviewed; 108 recipients with NASH,
and 122 with ALD were identified. After exclusion of SLK transplants
(9/11), re-transplants (3/5), death ≤ 90 days (3/7); 92
NASH recipients were compared with 99 recipients with ALD.
eGFR before LT, at 3 months, 1 yr, and 3 yrs. were compared.
RESULTS: NASH patients are more likely to be older (57.7 yrs.
vs 55 yrs., p=0.01), obese (32±5.4 vs. 28.4±5.6, p=0.001),
and diabetic (42.4% vs. 38.4%, p=0.003), and have lower
MELD score at listing (20.63±7.32 vs. 22.08±6.12, p=0.04).
Both groups received our standard steroid-free protocol with
rabbit ATG induction with low-dose Tacrolimus. eGFR pre-LT in
NASH and ALD group were comparable (76±45 vs. 78±33,
p=0.19). A significant decline in eGFR (Figure 1) was noted
in NASH recipients at 3 mos. (61±30 vs. 69±36, p=0.15), 1
yr. (60±20 vs. 70±25, p=0.03), and 3 yrs. (51±17 vs.68±18,
p=0.001). Based on eGFR groups (Gr 1: eGFR < 45, Gr II:
eGFR ≥45-60, eGFR >60), a distinct pattern of either inferior
recovery of renal function or decline in renal function was noted
in NASH recipients post LT (Figure 1). Declining eGFR in the
NASH group was more likely in diabetics (46±16 vs. 67±23,
p=0.02), but not in non-diabetics (57±17 vs. 68±17, p=0.17)
at 3 yrs. of follow up. CONCLUSION: LT recipients with NASH
phenotype have an increased propensity for renal dysfunction
compared to ALD particularly in diabetics. Early conversion to
calcineurin inhibitor sparing agents in LT recipients with NASH
and diabetes should be considered as a strategy to preserve
renal function.