Efficacy of Granulocyte Colony-Stimulating Factor and N-Acetylcysteine Therapies in Patients With Severe Alcoholic Hepatitis

Abstract

Patients with alcoholic hepatitis (AH) have high mortality, so new therapies are needed. Administration of granulocyte colony stimulating factor (G-CSF) increases survival times of patients with AH. It is not known whether addition of N-acetyl cysteine (NAC) to G-CSF could further increase survival time. We performed a randomized controlled pilot study to compare the efficacy of standard medical therapy with pentoxifylline to treatment with a combination of G-CSF and standard medical therapy as well as to the combination of NAC, G-CSF, and standard medical therapy in patients with severe AH. We performed an open-label, single-center study of 57 patients with severe AH admitted to a Liver Intensive Care unit in India from October 2014 through March 2017. Patients were randomly assigned to groups that received standard medical therapy (with pentoxifylline) plus G-CSF for 5 days (G-CSF group; n= 18), standard medical therapy plus G-CSF and intravenous NAC for 5 days (combination group; n= 19), or standard medical therapy alone(n= 20). Clinical data and blood samples were collected at baseline; on day 6; and 1, 2, and 3 months after the study began. CD34+ cells were measured in blood samples collected on days 0 and 6. The primary outcome was proportion of patients surviving for 90 days. Secondary outcomes were mobilization of CD34+ cells at day 6, as well as Child Turcotte Pugh, model for end-stage liver disease, and modified discriminant function scores until day90. Significantly higher proportions of patients in the G-CSF group (16/18) and the combination group (13/19) survived for 90 days than in the standard medical therapy group (6/20) (P=.0001 for G-CSF group and P = .037 and combination group). The GGSF and combination groups each had increased numbers of CD34+ cells from baseline until day 6, compared with the standard medical therapy group. The G-CSF group (but not the combination group) had significantly larger median reductions in modified discriminant function scores at study months 1 (reduction of 60.36%), 2 (reduction of 75.36%), and 3 (reduction of 88.73%) vs the standard medical therapy group (P = .02; P = .05; and P = .00, respectively). The G-CSF group had a significantly larger median reduction in model for end-stage liver disease score at 3 months (reduction of 55.77%; P= .01), but not in Child Turcotte Pugh score, compared with the standard medical therapy group. All groups had similar numbers of complications. In a pilot randomized controlled trial, we found administration of G-CSF to improve liver function and increase survival times in patients with severe AH, compared with standard therapy. We found no evidence for benefit of adding NAC to G-CSF. These findings require confirmation in larger trials. ClincialTrials.gov, number: NCT02971306.