Lack of recovery in monocyte human leukocyte antigen-DR expression is independently associated with the development of sepsis after major trauma

Abstract

Introduction: Major trauma is characterized by an overwhelming pro-inflammatory response and an accompanying anti-inflammatory response that lead to a state of immunosuppression, as observed after septic shock. Diminished monocyte Human Leukocyte Antigen DR (mHLA-DR) is a reliable marker of monocyte dysfunction and immunosuppression. The main objective of this study was to determine the relation between mHLA-DR expression in severe trauma patients and the development of sepsis.

Methods: We conducted a prospective observational study over 23 months in a trauma intensive care unit at a university hospital. Patients with an Injury Severity Score (ISS) over 25 and age over 18 were included. mHLA-DR was assessed by flow cytometry protocol according to standardized protocol. Mann-Whitney U-test for continuous non-parametric variables, independent paired t test for continuous parametric variables and chi-square test for categorical data were used.

Results: mHLA-DR was measured three times a week during the first 14 days. One hundred five consecutive severely injured patients were monitored (ISS 38 ± 17, SAPS II 37 ± 16). Thirty-seven patients (35%) developed sepsis over the 14 days post-trauma. At days 1-2, mHLA-DR was diminished in the whole patient population, with no difference with the development of sepsis. At days 3-4, a highly significant difference appeared between septic and non-septic patients. Non- septic patients showed an increase in mHLA-DR levels, whereas septic patients did not (13,723 ± 7,766 versus 9,271 ± 6,029 antibodies per cell, p = .004). Most importantly, multivariate logistic regression analysis, after adjustment for usual clinical confounders (adjusted OR 5.41, 95% CI 1.42-20.52), revealed that a slope of mHLA-DR expression between days1-2 and days 3-4 below 1.2 remained associated with the development of sepsis.

Conclusions: Major trauma induced an immunosuppression, characterized by a decrease in mHLA-DR expression. Importantly, after multivariate regression logistic analysis, persistent decreased expression was assessed to be in relation with the development of sepsis. This is the first study in trauma patients showing a link between the lack of immune recovery and the development of sepsis on the basis of the standardized protocol. Monitoring immune function by mHLA-DR measurement could be useful to identify trauma patients at a high risk of infection.

Figure 1

Flowchart of inclusion criteria of the study. ISS, Injury Severity Score.

Figure 2

Monocyte human leukocyte antigen-DR (HLA-DR) measurement by flow cytometry. (a) Monocyte identification in whole blood. An ungated leukocyte biparametric representation on the basis of side scatter characteristics (SSC, y-axis) and CD14 expression (FITC-CD14, x-axis) is shown. CD14-expressing population is easily distinguishable as gating region 'CD14+ monocytes'. (b) Gated cells from 'CD14+ monocytes' in (a) are expressed on the basis of HLA-DR expression (monoparametric histogram, PE-HLA-DR). The black histogram depicts isotype control, whereas the gray one represents patient expression (illustrative example). Results are obtained as means of fluorescence intensities (MFI) and then are transformed into number of antibodies per cell (AB/C). FITC, fluorescein isothiocyanate; PE, phycoerythrin.

Figure 3

Time course of monocyte human leukocyte antigen-DR (mHLA-DR) expression in trauma patients. Mean and standard deviation are presented. Results are expressed as numbers of anti-mHLA-DR antibodies bound per cell (AB/C). The independent paired t test was used for comparison between groups. *P < 0.01. (a) mHLA-DR expression in the whole trauma population. (b) mHLA-DR expression in patients with (gray bars) or without (white bars) sepsis.

Figure 4

Receiving operating curve of variation of monocyte human leukocyte antigen-DR expression ratio (days 3 and 4/days 1 and 2) expressed as antibodies per cell for predicting sepsis. Area under curve was 0.80 (P = 0.05, 95% confidence interval 0.69 to 0.88). The best threshold (that is, which maximized sensitivity and sensibility) was 1.2. For a cutoff of 1.2, positive predictive value was 42% and negative predictive value was 87%.