Prospective flow cytometry analysis of leucocyte subsets in critically ill patients who develop sepsis: a pilot study

Abstract

AbstractPurpose Sepsis in critically ill patients with injury bears a high morbidity and mortality. Extensive phenotypic monitoring of leucocyte subsets in critically ill patients at ICU admission and during sepsis development is still scarce. The main objective of this study was to identify early changes in leukocyte phenotype which would correlate with later development of sepsis. Methods Patients who were admitted in a tertiary ICU for organ support after severe injury (elective cardiac surgery, trauma, necessity of prolonged ventilation or stroke) were sampled on admission (T1) and 48-72h later (T2) for phenotyping of leukocyte subsets by flow cytometry and cytokines measurements. Those who developed secondary sepsis or septic shock were sampled again on the day of sepsis diagnosis (Tx). Results Ninety-nine patients were included in the final analysis. Nineteen (19.2%) patients developed secondary sepsis or septic shock. They presented significantly higher absolute monocyte counts and CRP at T1 compared to non-septic patients (1030/µl versus 55/µl, p = 0.013 and 5.1mg/ml versus 2.5mg/ml, p = 0.046, respectively). They also presented elevated levels of monocytes with low expression of L-selectin (CD62Lnegmonocytes)(OR[95%CI]: 4.5 (1.4–14.5) p = 0.01) and higher SOFA score (p < 0.0001) at T1 and low mHLA-DR at T2 (OR[95%CI]: 0.003 (0.00-0.17) p = 0.049). Stepwise logistic regression analysis showed that both monocyte markers and high SOFA score (> 8) were independent predictors of nosocomial sepsis occurrence. No other leucocyte count or surface marker nor any cytokine measurement correlated with sepsis occurrence. Conclusion Monocyte counts and change of phenotype are predictive of secondary sepsis in critically ill patients with injury.