Introduction Prednisolone has been shown to improve 28 day survival in alcoholic hepatitis (AH). However, use of this therapy is associated with increased infection that could explain the loss of survival benefit by 90 days. A large body of data has also emerged implicating bacterial translocation in the pathogenesis of AH. We therefore sought to elucidate the impact of prednisolone in patients with elevated pre-treatment circulating bacterial DNA (bDNA).
Methods DNA was extracted from the whole blood of patients recruited to the STOPAH study before therapy. In these patients, clinicians confirmed that any prior infection, if present, was controlled. bDNA levels were quantified by quantitative PCR using a Taqman probe that targeted a region within a 380 bp fragment of bacterial 16 S ribosomal DNA. As part of the STOPAH study, patients would then receive therapy with or without prednisolone by random allocation. After unmasking, levels of bacterial DNA were then compared to clinical outcomes such as the development of infection within 7 days, static and dynamic markers of liver function and mortality at 28 and 90 days.
Results bDNA was quantified in 714 whole blood samples. There was no correlation between bDNA levels and baseline characteristics such as age and gender, nor with static markers of liver function such as MELD. Elevated bDNA levels were associated with development of infection within the first 7 days in patients treated with prednisolone (p = 0.001), and remained independently associated with the subsequent development of infection after multivariate analysis (p = 0.02). Importantly, bDNA levels were not associated with the development of infection in patients treated without prednisolone (p = 0.43).
Elevated bDNA levels were associated with death at 90 days (p = 0.05). Mortality in prednisolone treated patients was increased for patients with high bDNA compared to patients with low bDNA (hazard ratio 2; p = 0.02). Accordingly, patients with elevated bDNA were more likely to die if they were randomly allocated prednisolone therapy (OR 2.9, 95% CI 1.2–6.9, p = 0.02). We estimate that if patients with elevated bDNA were instead treated without prednisolone, 90 day mortality in AH could be reduced (odds ratio 0.5, p = 0.03 [one-tailed]).
Conclusion Infection is more likely to develop within 7 days if patients with elevated bDNA are treated with prednisolone. Measuring pre-treatment bDNA could therefore inform decisions to initiate corticosteroid therapy and has the potential to increase survival at 90 days in AH.
Disclosure of Interest None Declared