Introduction Severe alcoholic hepatitis (SAH) is associated with high short-term mortality. In addition, it is associated with a high risk of infection, at presentation (baseline) and after commencing treatment (incident). Current tools for the prediction of survival and development of infection are imperfect. Elevated serum procalcitonin has been described in systemic inflammation and, in SAH, as able to discriminate between bacterial infection and a sterile systemic inflammatory response. The aim of this study was to determine whether serum procalcitonin was able to predict the incident infection and mortality in SAH.
Method Cases with SAH were recruited prospectively through the Steroids or Pentoxifylline for Alcoholic Hepatitis trial. Procalcitonin was measured by ELISA in serum samples taken at baseline (n=708). Median procalcitonin levels were compared between groups of interest by the Mann-Whitney U test, correlations were tested using Spearman’s rank and association with outcomes was tested by logistic regression.
Results Serum procalcitonin levels were generally elevated (median 0.20 ng/ml, IQR 0.08–0.52 ng/ml; normal <0.05 ng/ml) and did not differ between groups defined by baseline infection (p=0.45). Procalcitonin correlated with serum creatinine (rho 0.15, p<0.0001) at baseline but not bilirubin, white cell count, albumin or INR. Serum procalcitonin did not correlate with circulating bacterial DNA levels (p=0.90). A weak correlation with baseline MELD (rho 0.08, p=0.05) was observed but not with either DF or GAHS (p=0.1 and p=0.08, respectively). Procalcitonin levels did not predict incident infection or Lille response. However, procalcitonin levels did predict both 28 day (OR 1.10, 95% CI 1.01–1.21, p=0.03) and 90 day mortality (OR 1.20, 95% CI 1.07–1.36, p<0.01). The association between procalcitonin and mortality, at either time point, was independent of age, neutrophil count, INR, serum bilirubin, creatinine and albumin.
Conclusion In the context of SAH, serum procalcitonin is associated with baseline renal function and is an independent predictor of mortality. In this cohort, serum procalcitonin was not associated with baseline infection nor did it predict incident infection. This may indicate that in SAH procalcitonin levels are driven by inflammation rather than infection. However, the requirement that baseline infection was clinically controlled prior to randomisation and sample collection may explain the lack of association with procalcitonin levels.
Disclosure of Interest None Declared
- Severe alcoholic hepatitis