RT Journal Article SR Electronic T1 Randomised controlled trial of GM-CSF in critically ill patients with impaired neutrophil phagocytosis JF Thorax JO Thorax FD BMJ Publishing Group Ltd and British Thoracic Society SP 918 OP 925 DO 10.1136/thoraxjnl-2017-211323 VO 73 IS 10 A1 Emma M Pinder A1 Anthony J Rostron A1 Thomas P Hellyer A1 Marie-Helene Ruchaud-Sparagano A1 Jonathan Scott A1 James G Macfarlane A1 Sarah Wiscombe A1 John D Widdrington A1 Alistair I Roy A1 Vanessa C Linnett A1 Simon V Baudouin A1 Stephen E Wright A1 Thomas Chadwick A1 Tony Fouweather A1 Jatinder K Juss A1 Edwin R Chilvers A1 Susan A Bowett A1 Jennie Parker A1 Daniel F McAuley A1 Andrew Conway Morris A1 A John Simpson YR 2018 UL http://thorax.bmj.com/content/73/10/918.abstract AB Background Critically ill patients with impaired neutrophil phagocytosis have significantly increased risk of nosocomial infection. Granulocyte-macrophage colony-stimulating factor (GM-CSF) improves phagocytosis by neutrophils ex vivo. This study tested the hypothesis that GM-CSF improves neutrophil phagocytosis in critically ill patients in whom phagocytosis is known to be impaired.Methods This was a multicentre, phase IIa randomised, placebo-controlled clinical trial. Using a personalised medicine approach, only critically ill patients with impaired neutrophil phagocytosis were included. Patients were randomised 1:1 to subcutaneous GM-CSF (3 μg/kg/day) or placebo, once daily for 4 days. The primary outcome measure was neutrophil phagocytosis 2 days after initiation of GM-CSF. Secondary outcomes included neutrophil phagocytosis over time, neutrophil functions other than phagocytosis, monocyte HLA-DR expression and safety.Results Thirty-eight patients were recruited from five intensive care units (17 randomised to GM-CSF). Mean neutrophil phagocytosis at day 2 was 57.2% (SD 13.2%) in the GM-CSF group and 49.8% (13.4%) in the placebo group, p=0.73. The proportion of patients with neutrophil phagocytosis≥50% at day 2, and monocyte HLA-DR, appeared significantly higher in the GM-CSF group. Neutrophil functions other than phagocytosis did not appear significantly different between the groups. The most common adverse event associated with GM-CSF was fever.Conclusions GM-CSF did not improve mean neutrophil phagocytosis at day 2, but was safe and appeared to increase the proportion of patients with adequate phagocytosis. The study suggests proof of principle for a pharmacological effect on neutrophil function in a subset of critically ill patients.