Background
Liver biopsy may be of diagnostic and prognostic value but its role in alcoholic hepatitis (AH) has been controversial.

Aim
To assess the utility of liver biopsy in the assessment of clinically severe AH

Methods
The histological features of alcoholic steatohepatitis (ASH) were recorded and scored in patients enrolled in the Steroids or Pentoxifylline for Alcoholic Hepatitis (STOPAH) trial who underwent liver biopsy. These features were then assessed relative to outcome and established clinical prognostic scores.

Results
The STOPAH trial recruited 1068 patients; biopsies were obtained in 182 (17%). One hundred and sixty‐one biopsies were adequate for histological assessment and 140 (87%) were diagnostic for ASH. Only three biopsies (2%) did not have histological features of alcohol‐related liver injury. In biopsies performed prior to randomisation, ASH was identified in 92.5% of patients meeting clinical trial definitions of severe AH. In biopsies with ASH, taken before or within 48 hours of randomisation, survival differences between Alcoholic Hepatitis Histological Score (AHHS) groups were not significant: comparison of mild / moderate (91%: 21 of 23 patients) with severe (78%: 29 of 37 patients) groups: P = 0.18. The AHHS was not superior to clinical scores of prognosis: area under the curve for 28‐day mortality was 0.728, compared with 0.799 for the Glasgow alcoholic hepatitis score and 0.728 for the MELD score.

Conclusion
Liver histology taken before treatment rarely changes the diagnosis in patients meeting strict criteria for a clinical diagnosis of AH. The AHHS is similar to clinical scores in determining prognosis.

Clinical trial registration

EudraCT reference number: 2009‐013897‐42.

ISRCTN reference number: 88782125.

MREC number: 09/MRE09/59.

UKCRIN ID: 9143.