Introduction Standard treatment of AIH is corticosteroid-based. However there is no consensus regarding optimal dose or duration of steroid therapy. Budesonide had similar short-term efficacy to Prednisolone in a large randomised trial; but data are lacking on its longer term efficacy. We conducted an audit on management of AIH in 28 UK centres of varying size and report outcomes in relation to type, dose and duration of steroid treatment in this cohort.

Methods We aimed to collect all prevalent cases since 2000 and all incident cases since 2007 by searching electronic patient letters, histology databases and hospital coding. Validation was by 1999 IAIHG diagnostic criteria. Information, on diagnosis, severity, treatment and outcome was entered into a web-based system.

Results 1267 patients (80% female age (mean (range) 55 years (8–86)) were followed up for 4 (0–14) years. Initial treatment was Prednisolone in 1091 patients, Methyl-Prednisolone/Hydrocortisone in 20 patients (pooled with Prednisolone as outcomes similar) and Budesonide in 61 patients. 984 patients also received a steroid sparing agent (962 Thiopurines). The remaining 95 patients did not receive steroids (12 received Thiopurine monotherapy). Of these 95, 10 required liver transplantation, 3 of these >1 year after diagnosis. Compared to the 1172 patients receiving steroids, the 95 who did not were older (58 vs 50 years p < 0.001) but had similar prevalence of cardio-respiratory comorbidity, cirrhosis and decompensation at presentation. They had a higher all-cause and liver-related death/transplant rate (p < 0.001), which persisted in multivariate analysis, and also (all-cause only) when patients with decompensation at presentation (n = 282) were excluded. All-cause and liver-related death/transplant rate were also associated with shorter duration of steroid therapy (p = 0.001 and 0.008) and with its discontinuation after initial normalisation of transaminases (p = 0.007 and 0.016), but not associated with initial dose of Prednisolone.

Of the 61 patients initially receiving Budesonide, treatment was changed to Prednisolone in 13 after 8 (0.2–29) months. Compared to those initially receiving Prednisolone, those receiving Budesonide had (a) similar age and gender distribution (b) lower peak serum ALT (c) surprisingly, a similar frequency of cirrhosis (26% vs 25%) at presentation (d) similar serum ALT values and percentages achieving normal values after 6 (64% vs 68%) and 12 months (78% vs 75%) treatment (e) similar all-cause and liver-related death/transplant rate.

Conclusion In this Multicentre UK Audit (a) 7% of patients with AIH did not receive steroids and had higher mortality than those receiving steroids (b) 5% of patients received Budesonide (one quarter with cirrhosis) and had similar mortality to those receiving Prednisolone.

Disclosure of Interest None Declared