P48 Validation of the neutrophil to lymphocyte ratio as a marker of poor outcomes and steroid responsiveness in alcoholic hepatitis: The Gwent experience

Abstract

<h3>Introduction</h3> The STOPAH study showed an improvement in short term mortality with prednisolone but failed to reach statistical significance. Forrest et al subsequently reported that a neutrophil – lymphocyte ratio (NLR) of 5–8 was a predictor of steroid response compared to those with an NLR &lt;5 or &gt;8. We reviewed the validity of the NLR as a marker of severity and steroid response in alcoholic hepatitis patients admitted to the Gwent liver unit. <h3>Method</h3> We undertook a retrospective analysis of consecutive patients admitted with alcoholic hepatitis between January 2014 and March 2019. Patients were identified via coding and discharge notification diagnosis. Only patients diagnosed by a consultant gastroenterologist/hepatologist and with data available on electronic records were included. Baseline NLR and Glasgow alcoholic hepatitis score (GAHS) were recorded with a primary outcome of mortality. Additional observations included acute kidney injury (AKI), readmission within 90 days and day-7 Lille score for those treated with prednisolone. <h3>Results</h3> Sixty-six patients were identified, of which 40 (61.6%) were male, with a median age of 47 (IQR 40–56). There were 12 patients (18.2%) with an NLR of 5–8. Patients with an NLR of 5–8 were more likely to be alive at 28 and 90 days than those with an NLR &lt;5 or &gt;8, however this trend did not continue at 1 year. Patients with an NLR 5–8 had lower rates of AKI (8%) and fewer readmissions within 90 days (33%) compared to those with NLR &lt;5 (22% and 58% respectively) or &gt;8 (52% and 45% respectively). Twenty-seven patients received prednisolone. Patients with an NLR 5–8 were more likely to receive prednisolone than others (66.7%) and tended towards a better response. Patients with NLR 5–8 had more favourable Lille scores with a mean of 0.251 (SD ± 0.207) compared to patients with an NLR &lt; 5 and &gt;8 (Mean 0.459 and 0.406 respectively). Prednisolone was continued past 7 days in 71% of patients with NLR 5–8 compared to those with NLR &lt;5 (44%) or &gt;8 (25%). <h3>Conclusion</h3> Our data supports the previously reported finding that patients with NLR 5–8 are more likely to respond to prednisolone. They had lower mortality up to 90 days and were less likely to have AKI or readmission within 90 days. Furthermore, an NLR &gt;8 was associated with particularly poor 1-year survival and high incidence of AKI. The NLR appears to be another useful method of risk stratification in alcoholic hepatitis.