S172 Poster Session − Friday, April 24 plantation, and identify the predictors of HBV recurrence with emergence of YMDD mutation. Methods: We analysed the outcomes retrospectively in 226 consecutive patients positive for hepatitis B surface antigen, who had undergone liver transplantation from August 2001 to December 2006 at Seoul National University Hospital, a tertiary medical center. Of these, 17 patients were excluded due to the presence of pre-transplant YMDD motif mutant (9/17) and follow-up duration of less than 1 month after transplantation (8/17). All eligible patients received either combination prophylaxis with high dose HBIG and lamivudine (186/209, 89.0%) or HBIG monotherapy (23/209, 11.0%). Results: A total of 209 patients were enrolled. Median follow-up duration was 36.9 months (range, 1−84.4). Twenty-two patients (10.5%) experi- enced post-transplant HBV recurrence. Twelve patients (5.7%) had YMDD motif mutation following transplantation, and were amenable to adefovir add-on therapy. Curves of time-to-viral recurrence and time-to-viral break- through with emergence of YMDD mutation were generated by Kaplan– Meier method, and compared as to clinical variables using log-rank test and Cox regression analysis. Independent predictors of post-transplant HBV recurrence were the recurrence of hepatocellular carcinoma following transplantation (HR 30.09, 95% CI, 9.42−96.10; p < 0.001), insufficient du- ration (<2 months) of HBIG monotherapy (HR 8.53, 95% CI, 1.64−44.40; p = 0.011), lamivudine treatment of more than 1.5 years (HR 3.09, 95% CI, 1.16−8.26; p = 0.024), and high titer (20,000 IU/ml) of HBV DNA before transplantation (HR 2.67, 95% CI, 1.04−6.88; p = 0.042). Independent risk factors of HBV recurrence with emergence of YMDD mutation were high titer of HBV DNA (HR 5.42, 95% CI, 1.54−19.13; p = 0.009) and lamivu- dine treatment of more than 1.5 years (HR 6.28, 95% CI, 1.30−30.30; p = 0.022). Conclusions: HBV recurrence with emergence of YMDD mutation after orthotopic liver transplantation possibly occurs in patients with high viral load before transplantation and prolonged duration of lamivudine treatment. 455 PREDICTORS OF SPONTANEOUS SURVIVAL IN FULMINANT HEPATIC FAILURE PATIENTS MEETING KING’S COLLEGE CRITERIA D.G.N. Craig, K.J. Simpson, P.C. Hayes. Department of Hepatology, Royal Infirmary of Edinburgh, Edinburgh, UK E-mail: darrencraig@doctors.org.uk Introduction: King’s College Criteria (KCC) have high specificity in predicting mortality in fulminant hepatic failure (FHF) but a small pro- portion of patients fulfilling KCC spontaneously survive without requiring orthotopic liver transplantation (OLT). Early identification of this cohort of patients could prevent errors of commissioning when listing for emergency OLT. Aim: To examine factors predicting spontaneous survival amongst patients fulfilling KCC transferred to the Scottish Liver Transplant Unit (SLTU). Methods: Analysis of a prospectively collected database of patients with acute liver injury (ALI) transferred to SLTU between 1992 and 2008. Results: Of 903 patients with ALI transferred to SLTU, 514 (56.9%) had or developed encephalopathy, and thus FHF. 295 (119 (40.3%) male, 176 (59.7%) female) of these patients met KCC. Aetiological factors included acetaminophen overdose (AO)(n = 197, 66.8%), seronegative hep- atitis (n = 45, 15.3%), idiosyncratic drug reactions (n = 22, 7.5%) and other causes (n = 31, 10.5%). 149 (50.5%) of KCC positive cases were deemed suitable transplant candidates and 115 (77.2%) underwent emergency OLT. There were no significant differences in outcome according to aetiology (table 1). KCC +ve pts Died (no OLT) Survived (no OLT) Died with OLT Survived without OLT AO, n (%) 130 (66) 15 (7.6) 15 (7.6) 37 (18.8) Non-AO, n (%) 32 (32.7) 3 (3.1) 13 (13.3) 50 (57.5) AO vs. non-AO c 2 = 0.10, p = 0.75 c 2 = 1.04, p = 0.31 Within the AO cohort (n = 197), spontaneous survivors were significantly less likely to be hypotensive at admission (c 2 =29.5, p < 0.001) or subse- quently require inotropic support (c 2 =21.5, p < 0.001) and had significantly lower admission creatinine (p = 0.003) and higher pH (p = 0.005, Wilcoxon rank sum tests) compared with non-survivors. Within the non-AO co- hort (n = 98), admission hypotension (p = 0.006), inotrope requirement (p = 0.018, both Fishers exact test) and serum lactate (p = 0.022, Wilcoxon rank sum test) predicted a poor outcome. There were no significant differ- ences between spontaneous survivors and non-survivors in either cohort with respect to: age; gender; admission encephalopathy grade; requirement for renal replacement therapy; presenting blood urea, prothrombin time or liver function tests. Conclusion: A small number of FHF patients fulfilling KCC survive with conservative management alone. Overall clinical condition, in particular the presence of hypotension, should be considered when planning emer- gency OLT for FHF. 456 HCC PERSISTENCE OR RECURRENCE AFTER BRIDGING THERAPY HELPS PREDICT TRANSPLANT LIST DROPOUT AND GENERATE A MORE EQUITABLE EXCEPTION POLICY M. De Giorgio 1 , S. Vezzoli 1 , E. Cohen 2 , E. Armellini 1 , M.G. Luc` a 1 , G. Verga 1 , D. Pinelli 1 , R. Nani 1 , M. Colledan 1 , S. Fagiuoli 1 , M. Strazzabosco 2,3 . 1 Center for Liver Research (CeLiveR), Ospedali Riuniti of Bergamo, Bergamo, Italy; 2 Internal Medicine, Section of Digestive Diseases and Liver Center, Yale University, New Haven, USA; 3 Clinical Medicine, University of Milano Bicocca, Milano, Italy E-mail: madegio@hotmail.it Introduction: Earlier diagnosis and more effective treatments are im- proving the survival of HCC patients. On the other hand, priority for liver transplantation among HCC patients misses uniformity in Countries outside UNOS area and remain a critical issue. Aim of the study was to generate an updated prognostic model for HCC applicable to organ allocation. Material and Methods: we analysed disease progression (death or progression of HCC beyond T2 stage) in a cohort of 177 consecutive patients that presented with HCC inside the Milan criteria. Patients were treated according to the BCLC algorithm and prospectively followed by imaging every 3 months; 76 patients (43%) were listed for transplantation according to AISF/CNT guidelines. Transplanted patients were censored at the time of transplantation. Results: Median follow up was 16 months. HCC progressed beyond T2 stage in 75 cases; 8 patients died while in T2 stage and 68 were transplanted. To verify the ability of the UNOS policy to predict HCC progression rate (PR), we compared the expected PR according to the UNOS policy with the observed PR for the respective population at risk. The current policy resulted in a large overestimation of PR in T2 patients. Two sensitivity analyses were conducted to identify risk factors (RF) for progression using the log-rank method: in the first one, transplanted patients were considered as disease progression; in the second one, only non-transplanted patients were analysed. RF significant in any analysis was considered for multivariable analysis. RF with the strongest relationship to progression included tumor persistence after local therapy or recurrence inside Milan criteria (HR 2.52−3.16; p < 0.0001). Using strongest RF for progression and the patient’s natural MELD, an adjusted model applicable to organ allocation was generated, which decreased the discrepancy between expected and observed PR. Conclusions: Tumor persistence/recurrence inside Milan criteria after treatment strongly predicted the risk of tumor progression beyond Milan criteria. The current MELD exception largely overestimates progression in T2 patients treated according to BCLC. An adjusted model that begins with the patient’s natural MELD, and stratifies progressive risk according to the response to local treatment, better captures the PR of T2 HCC.