Bactericidal activity response of blood neutrophils from critically ill patients to in vitro granulocyte colony-stimulating factor stimulation

K Yang, É Azoulay, L Attalah, JR Zahar… - Intensive care …, 2003 - Springer
K Yang, É Azoulay, L Attalah, JR Zahar, A Van de Louw, C Cerf, CJ Soussy, P Duvaldestin…
Intensive care medicine, 2003Springer
Objective Neutrophil function impairment is common in nonneutropenic critically ill patients.
Whether granulocyte colony-stimulating factor (G-CSF) may be useful for preventing
nosocomial infection in these patients is debated. The response of blood neutrophils from
critically ill patients to G-CSF was investigated in vitro. Design and setting Prospective study,
laboratory investigation in two intensive care units. Patients 52 critically ill patients without
immunosuppression. Measurements Neutrophils obtained from 52 patients on the 5th day of …
Objective
Neutrophil function impairment is common in nonneutropenic critically ill patients. Whether granulocyte colony-stimulating factor (G-CSF) may be useful for preventing nosocomial infection in these patients is debated. The response of blood neutrophils from critically ill patients to G-CSF was investigated in vitro.
Design and setting
Prospective study, laboratory investigation in two intensive care units.
Patients
52 critically ill patients without immunosuppression.
Measurements
Neutrophils obtained from 52 patients on the 5th day of their intensive care unit stay were incubated with and without G-CSF (1, 10, 100 ng/ml). Reactive oxygen species (ROS) release and bactericidal activity against Staphylococcus aureus and Pseudomonas aeruginosa were evaluated. Plasma cytokines (interleukin 10, tumor necrosis factor α, and G-CSF) were measured.
Results
Median values (25th–75th percentiles) indicated no stimulatory effect of G-CSF on neutrophil bactericidal activity against either organism: S. aureus, 100% (95–109) of the unstimulated condition with 1 ng/ml G-CSF, and P. aeruginosa , 102% (98–109) with 1 ng/ml G-CSF. However, wide interindividual variability was found, ranging from marked inhibition to marked stimulation. Similar variability was found for ROS release. No correlations were found between ROS release and bactericidal activities against either bacterial strain. Inhibition of neutrophil bactericidal activity by G-CSF was associated with significantly higher plasma interleukin 10 concentrations. Plasma G-CSF levels were significantly higher in patients whose neutrophil bactericidal activity was unresponsive to G-CSF, suggesting G-CSF receptor downregulation.
Conclusions
The effect of G-CSF on in vitro neutrophil bactericidal activity varied widely, depending on endogenous levels of G-CSF and was not predictable based on severity scores.
Springer