Elevated Levels Of Circulating Bacterial DNA In Alcoholic Hepatitis Identify Patients Who Will Not Respond To Corticosteroid

TY - JOUR

T1 - Elevated Levels Of Circulating Bacterial DNA In Alcoholic Hepatitis Identify Patients Who Will Not Respond To Corticosteroid

AU - Vergis, Nikhil

AU - Atkinson, Steven

AU - Knapp, Susanne

AU - Patel, Vishal

AU - Trial Group STOPAH

AU - Antoniades, Charalambos

AU - Forrest, Ewan

AU - Thursz, Mark

PY - 2015

Y1 - 2015

N2 - Background and Aims: Prednisolone has been shown to improve 28-day survival in alcoholic hepatitis (AH). However, use of this therapy is associated with increased infections that could explain the limited benefit of corticosteroid beyond 28 days. Interestingly, a large body of data suggests that bacterial translocation plays an important role in the pathogenesis of AH. The present study sought to elucidate the impact of prednisolone in patients with elevated circulating bacterial DNA (bDNA). Methods: DNA was extracted, pre-treatment (Day0), from the whole blood of 156 patients recruited to the multi-centre Steroids or Pentoxyfilline for Alcoholic Hepatitis (STOPAH) study. Patients were matched for MELD, age, gender, therapy and the presence of infection. Subsequently, real-time PCR was performed using a Taqman probe targeting a region within a 380bp fragment of bacterial 16S rDNA for bDNA quantification. Levels of bacterial DNA were compared to Maddrey’s discriminant function (mDF), SIRS score, Lille score, the development of an infectious serious adverse event within the first 7 days of therapy, and mortality.Results: bDNA levels correlated with mDF although this did not achieve statistical significance [r=0.146, p=0.07]. However, there was strong correlation between bDNA levels and Lille model [r=0.29, p=0.0016]. The Day0 level of circulating bDNA in Lille non-responders (>0.45) was higher than in Lille responders (<0.45) [21 vs 14pg/ml; p=0.02]. Importantly, this association was only seen in patients treated with prednisolone [prednisolone treated: r=0.4, p=0.0046 vs no prednisolone: r=0.18, p=ns]. Day0 bDNA level >18.4pg/ml in prednisolone treated patients was strongly associated with Lille non-response [odds ratio 6, p<0.0001].There was no association between Day0 bDNA level and the development of infection (22 vs 19 pg/ml, p=ns); SIRS criteria (r=0.02, p=ns), 28-day or 90-day mortality (18 vs 21pg/ml, p=ns and 17 vs 21pg/ml, p=ns).Conclusions: Pre-treatment circulating bDNA levels may be used to predict poor outcome in AH. Further, patients with elevated circulating bDNA are unlikely to respond to prednisolone therapy and so could be spared the infectious complications associated with this agent. Further work should validate these findings in a larger cohort of AH patients.

AB - Background and Aims: Prednisolone has been shown to improve 28-day survival in alcoholic hepatitis (AH). However, use of this therapy is associated with increased infections that could explain the limited benefit of corticosteroid beyond 28 days. Interestingly, a large body of data suggests that bacterial translocation plays an important role in the pathogenesis of AH. The present study sought to elucidate the impact of prednisolone in patients with elevated circulating bacterial DNA (bDNA). Methods: DNA was extracted, pre-treatment (Day0), from the whole blood of 156 patients recruited to the multi-centre Steroids or Pentoxyfilline for Alcoholic Hepatitis (STOPAH) study. Patients were matched for MELD, age, gender, therapy and the presence of infection. Subsequently, real-time PCR was performed using a Taqman probe targeting a region within a 380bp fragment of bacterial 16S rDNA for bDNA quantification. Levels of bacterial DNA were compared to Maddrey’s discriminant function (mDF), SIRS score, Lille score, the development of an infectious serious adverse event within the first 7 days of therapy, and mortality.Results: bDNA levels correlated with mDF although this did not achieve statistical significance [r=0.146, p=0.07]. However, there was strong correlation between bDNA levels and Lille model [r=0.29, p=0.0016]. The Day0 level of circulating bDNA in Lille non-responders (>0.45) was higher than in Lille responders (<0.45) [21 vs 14pg/ml; p=0.02]. Importantly, this association was only seen in patients treated with prednisolone [prednisolone treated: r=0.4, p=0.0046 vs no prednisolone: r=0.18, p=ns]. Day0 bDNA level >18.4pg/ml in prednisolone treated patients was strongly associated with Lille non-response [odds ratio 6, p<0.0001].There was no association between Day0 bDNA level and the development of infection (22 vs 19 pg/ml, p=ns); SIRS criteria (r=0.02, p=ns), 28-day or 90-day mortality (18 vs 21pg/ml, p=ns and 17 vs 21pg/ml, p=ns).Conclusions: Pre-treatment circulating bDNA levels may be used to predict poor outcome in AH. Further, patients with elevated circulating bDNA are unlikely to respond to prednisolone therapy and so could be spared the infectious complications associated with this agent. Further work should validate these findings in a larger cohort of AH patients.

KW - Bacterial DNA

KW - Alcoholic Hepatitis

KW - Corticosteroid

KW - Pentoxyfilline

KW - STOPAH

M3 - Meeting abstract

SN - 0168-8278

JO - Journal of Hepatology

JF - Journal of Hepatology

T2 - 50th The International Liver Congress - European Association for the Study of the Liver

Y2 - 22 April 2015 through 26 April 2015

ER -