and 0.608 for the ABCR score, the IBS was 0.135 and 0.128, respectively. Independent predictors of survival in both scores were an increase in Child Pugh ≥2 points and radiological response. In the ABCR score additionally the BCLC stage proved significant (all p < 0.05). Conclusions: The ART and the ABCR score give some indication which patients face a dismal prognosis and may not benefit from further TACE sessions. The ABCR score was of slightly higher predictive value. However, both scores are not sufficient to make clear-cut clinical decisions. Further efforts are necessary to obtain criteria for making valid predictions. THU-092 ALANINE AMINOTRANSFERASE NORMALIZATION DURING ENTECAVIR TREATMENT AND THE HEPATOCELLULAR CARCINOMA RISK IN HEPATITIS B VIRUS-ASSOCIATED CIRRHOSIS S.K. Shin 1 , J.E. Yeon 2 , O.S. Kwon 1 , E.J. Kim 1 , H.N. Lee 1 , S.H. Kang 2 , K.S. Byun 2 , J.H. Kim 3 , S.Y. Kwon 3 , S.J. Suh 4 , H.J. Yim 4 , Y.S. Kim 1 , J.H. Kim 1 . 1 Department of Internal Medicine, Gachon University Gil Medical Center, Incheon; 2 Department of Internal medicine, Korea University Medical College, Guro Hospital; 3 Department of Internal medicine, Konkuk University Medical College, Seoul; 4 Department of Internal medicine, Korea University Medical College, Ansan Hospital, Ansan, South Korea E-mail: kos@gilhospital.com Background and Aims: Sustained abnormal serum alanine aminotransferase (ALT) levels can increase the riskof hepatocellular carcinoma (HCC) in patients with chronic hepatitis B. This study is aimed to confirm the impact of abnormal ALT levels (>30 IU/L) on HCC risk in patients with hepatitis B virus (HBV) associated-cirrhosis during entecavir (ETV) treatment. Methods: A total of 578 treatment-naïve patients with HBV- associated cirrhosis were treated with ETV for more than 1 year. Serum ALT and HBV DNA levels were measured at 3 time points (baseline, 6, and 12 months after ETV treatment) and subjected to risk factor analysis. Results: Mean age was 51 ± 9 years and 366(63.3%) were male. The median follow up period was 43(12–98) months. At baseline, 303 patients (52.4%) were seropositive for HBeAg. The numbers of patients of Child-Pugh class A, B, and C were 417(72.1%), 140 (24.2%), and 21(3.6%), respectively. The number of patients with undetectable HBV DNA (<120 copies/mL) at 6 months or at 12 months were 333(57.6%) and 455(78.7%), respectively. The number of patients with normal ALT at 6 months and at 12 months were 258(44.6%) and 313(54.2%), respectively. Of the 578 patients with HBV-associated cirrhosis, 81 patients (14%) developed HCC during follow-up. Cumulative incidences of HCC at 1, 3, 5, and 7 years were 0.3%, 8.5%, 19.5%, and 30.6%, respectively. Univariate Cox-regression analysis showed that older age, abnormal ALT at 6 months or 12 months, and lower platelet count were significant risk factors of HCC. However, gender, HBeAg positivity, abnormal ALTand HBV DNA levels at baseline, detectable HBV DNA at 6 months or 12 months were not risk factors. Multivariate analysis showed that older age (HR: 1.062, 95% CI: 1.035–1.089, p <0.001), abnormal ALT at 12 months (HR: 1.876, 95% CI: 1.198–2.937, p = 0.006) and lower platelet count (HR: 0.993, 95% CI: 0.987–0.999, p = 0.034) were the risk factors of HCC. Figure shows the cumulative incidences of HCC according to the patterns of ALT change at 6 and 12 months after ETV treatment. Cumulative incidences of HCC were significantly lower in group with sustained normal ALT at 6 months and 12 months than other groups. Conclusions: Abnormal serum ALT levels during ETV treatment are significant risk factor of HCC. Therefore, ALT should be rapidly and sustained normalized to minimize the risk of HCC development in patients with HBV-associated cirrhosis. THU-093 FALSE-NEGATIVE LIVERULTRASOUND LIMITS BENEFITS OF HCC SURVEILLANCE S. Demma 1,2 ,P.O’Donoghue 1 , J. O’Beirne 1 , A. Marshall 1 , T. Meyer 3 , W. Rosenberg 1 , D. Macdonald 1 . 1 Sheila Sherlock Liver Unit and UCL Institute for Liver and Digestive Health, Royal Free Hospital/University College of London, London, United Kingdom; 2 Liver Unit, Department of Experimental and Clinical Medicine, Universityof Catania, Catania, Italy; 3 Department of Oncology, Royal Free Hospital/University College of London, London, United Kingdom E-mail: shirin.demma@gmail.com Background and Aims: The aim of liver ultrasound (US) surveillance in cirrhosis is to detect hepatocellular carcinoma (HCC) at a stage amenable to curative therapy. However, the effectiveness of HCC surveillance is influenced by both timing and the sensitivity of ultrasound. Methods: We conducted a retrospective study of new diagnoses of HCC (n = 481) made at a UK tertiary liver transplant centre (The Royal Free Hospital) between April 2012 and March 2015. Of 481 patients diagnosed with HCC in this period, 108 had previous imaging surveillance for HCC. 97 were identified with no previous surveillance. We assessed the impact of surveillance on tumour diameter at diagnosis, rates of potentially curative intervention and survival. To estimate rates of false-negative US, we retro-extrapolated tumour diameter from diagnosis to the last “normal” surveillance ultrasound using a range of tumour volume doubling times (TVDT). Results: In the surveillance group the mean interval time between imaging was 210 ± 24 days (±95% CI) (target 182 days). Mean tumour size was 2.901 ± 0.33 and 7.922 ± 1.12 cm in the surveillance and non- surveillance group respectively (p < 0.0001). 51% of patients under surveillance received potentially curative treatment (radiofrequency ablation, resection or transplant), compared with 21% of patients presenting for the first time (Fisher’s exact test, p < 0.0001). Only 15% of patients of the surveillance cohort received a liver transplant (vs 1% in the non-surveillance group). Subjects diagnosed under POSTER PRESENTATIONS S340 Journal of Hepatology 2016 vol. 64 | S213–S424