PTU-034 Histological remission in autoimmune hepatitis: can it be predicted?

Introduction A recent report1 suggested that initial treatment of Autoimmune Hepatitis (AIH) with high-dose prednisolone results in more rapid normalisation of serum ALT and (by implication), improved outcome. However, we have reported2 that persisting histological activity in AIH (Ishak necroinflammatory score (NIS) >3), despite serum ALT normalisation, is associated with fibrosis progression and increased mortality. Here, we aimed to identify, retrospectively, disease or treatment related factors associated with attaining histological remission. Methods We studied 111 patients with AIH (93 female, median (range) age at diagnosis 57 (14–77) years treated with prednisolone plus azathioprine/mycophenolate, who had attained normalisation of serum ALT and had available follow-up biopsy. Results Starting daily doses of prednisolone and of azathioprine were 30 (10–60) and 50 (25–150) mg respectively. Time taken to achieve normal serum ALT after starting treatment was 8 (0–144) weeks and showed no correlation with any of: (a) gender, age, serum ALT, or globulin at presentation (b) NIS or fibrosis grade at diagnostic biopsy, (c) starting dose of prednisolone or azathioprine. Biopsy was repeated 26 (12–90) months after starting treatment. 72 patients (65%) attained histological remission. Comparing these with the 37 patients not in histological remission, there were no significant differences in age, gender, presenting serum ALT, or in NIS or fibrosis stage at diagnostic biopsy. Neither were there differences in either (a) starting or cumulative dose of prednisolone or azathioprine (either absolute or corrected for body weight) or (b) time to achieve normal serum ALT. Only serum globulin at presentation was lower in those who achieved histological remission (42 vs 49 g/l, p=0.01). On multivariate analysis, NIS at follow-up biopsy was independently associated with serum ALT at time of biopsy (p≤0.001) but with none of above baseline, treatment or early response -related variables. Conclusion We could not identify baseline or treatment -related variables associated with (and thus, potentially predictive of) histological remission. Specifically, we could not establish that histological remission was related to either time to serum ALT normalisation or to starting or cumulative dose of prednisolone. Competing interests None declared. References 1. Schramm C, et al. Hepatology 2010;52:2247–82. 2. Hoeroldt BS, et al. Gut 2009;58:A18.