123 STRUCTURE OF THE HEPATITIS B VIRUS RNase H, A TARGET FOR NEW ANTIVIRAL DRUG DEVELOPMENT, UNRAVELED BY ULTRA-DEEP PYROSEQUENCING AND MOLECULAR MODELING

Abstract

a single tertiary center. The date and cause of death were recorded and checked using the national death registry. Results: Baseline characteristics of 361 included patients were: age: 60.0±9.8 yrs, male: 71.7%, Child Pugh score: 6.1±1.8 (A: 69.7%, B: 23.3%, C: 7.0%), Meld score: 11.3±4.5, follow-up: 3.1±1.4 yrs, death rate: 28.8% (liver related death in 62.5%). The 144 patients with non-selective BB were compared to the 190 controls without BB; 27 patients with selective BB were excluded. At baseline, there were higher frequencies of esophageal (p < 0.001) or gastric varices (p = 0.019) and alcohol withdrawal (p = 0.001) in the BB group. Child–Pugh (p =0.56) and Meld (p =0.25) scores did not differ. The independent predictors of overall survival were Meld (p < 0.001) and age (p =0.028) by Cox model. The independent predictors of liver related survival included Meld (p < 0.001) and gastric varices (p = 0.026), and a significant interaction between Meld and BB (p =0.035). Therefore, we assessed the BB effect according to baseline Meld score: in the lowest tertile, BB group had an improved overall survival as compared to controls (logrank: p =0.071); a beneficial BB effect was also observed in the second tertile until 4 years of follow-up; in contrast, in the highest tertile, patients treated with BB had decreased survival beyond 2 years (p = 0.095). Liver survival showed the same trends in the first (p = 0.140) and second (p =0.306) tertiles; in the highest tertile, the poorer survival in BB group was more pronounced than for overall survival (p = 0.019). Conclusion: The BB effect on survival differed according to liver disease severity in alcoholic cirrhosis: in good condition patients, BB tended to improve survival during the first years whereas there was a significantly decreased liver survival in the most severe patients.