Introduction Severe alcoholic hepatitis (SAH) has a high mortality especially in those who fail to respond to steroid treatment. Early identification of steroid resistant patients may allow rapid implementation of other therapies, which may improve patient outcome. An early change in bilirubin has prognostic value but requires 7 days of steroid treatment first. Our group have previously reported that a simple 48h in vitro method for assessing steroid sensitivity (dexamethasone inhibition of lymphocyte proliferation; DILPA) correlates with 6 month mortality in patients with SAH[1]. We aimed to determine the accuracy of the DILPA in predicting outcome and to compare it to existing models of prognosis in SAH.

Methods Peripheral blood was drawn from consecutive patients with a clinical diagnosis of SAH (Maddrey discriminant function [MDF] > 32). All subjects gave informed consent and ethical approval was obtained from the regional ethics committee. The primary outcome measure was 6 month mortality. Leukocytes, isolated by density gradient centrifugation, were cultured for 48h with the mitogen PHA in the presence or absence of dexamethasone. Tritiated thymidine was added for the final 6h of culture before proliferation was measured on a beta counter. Maximum suppression of lymphocyte proliferation by dexamethasone was calculated (Imax). An Imax of < 60% indicates in vitro steroid resistance as previously determined. The accuracy of the DILPA and existing scoring systems (MDF, MELD, Glasgow and Lille scores) was calculated by area under the receiver operating characteristic (AUROC). Logistic regression of baseline and day 7 parameters was performed to determine independent predictors of outcome.

Results 43 patients were recruited (21% female, median age 45). Patients who survived 6 months had a significantly higher Imax than those who didn’t (79.2% v 41.7%; p = 0.0003). An Imax of > 60% had a 88% sensitivity and 78% specificity in predicting 6 month survival and an AUROC of 0.83. MDF, MELD, Glasgow and Lille scores all had lower AUROC values for predicting 6 months survival than Imax (0.36, 0.4, 0.49 and 0.61 respectively). A stepwise logistic regression demonstrated that only Imax, Day 0 albumin and change in bilirubin at Day 7 of treatment were independent predictors of 6 month survival.

Conclusion The DILPA provides a simple, fast and accurate in vitro method of predicting clinical outcome in patients with SAH. This could be applied in clinical practise to identify patients who will not respond to steroid therapy within 2 days of starting treatment.

Disclosure of Interest None Declared

Reference

1. Di Mambro et al. Hepatology 2011;53:1316–22.