Persistently Elevated Soluble Triggering Receptor Expressed on Myeloid Cells 1 and Decreased Monocyte Human Leucocyte Antigen DR Expression Are Associated With Nosocomial Infections in Septic Shock Patients

Abstract

Sepsis-acquired immunosuppression may play a major role in patients' prognosis through increased risk of secondary infections. Triggering receptor expressed on myeloid cells 1 (TREM-1) is an innate immune receptor involved in cellular activation. Its soluble form (sTREM-1) has been described as a robust marker of mortality in sepsis. The objective of this study was to evaluate its association with the occurrence of nosocomial infections alone or in combination with human leucocyte antigen-DR on monocytes (mHLA-DR).

Design: Observational study.

Setting: University Hospital in France.

Patients: One hundred sixteen adult septic shock patients as a post hoc study from the IMMUNOSEPSIS cohort (NCT04067674).

Interventions: None.

Measurements and main results: Plasma sTREM-1 and monocyte HLA-DR were measured at day 1 or 2 (D1/D2), D3/D4, and D6/D8 after admission. Associations with nosocomial infection were evaluated through multivariable analyses. At D6/D8, both markers were combined, and association with increased risk of nosocomial infection was evaluated in the subgroup of patients with most deregulated markers in a multivariable analysis with death as a competing risk. Significantly decreased mHLA-DR at D6/D8 and increased sTREM-1 concentrations were measured at all time points in nonsurvivors compared with survivors. Decreased mHLA-DR at D6/D8 was significantly associated with increased risk of secondary infections after adjustment for clinical parameters with a subdistribution hazard ratio of 3.61 (95% CI, 1.39-9.34; p = 0.008). At D6/D8, patients with persistently high sTREM-1 and decreased mHLA-DR presented with a significantly increased risk of infection (60%) compared with other patients (15.7%). This association remained significant in the multivariable model (subdistribution hazard ratio [95% CI], 4.65 [1.98-10.9]; p < 0.001).

Conclusions: In addition to its prognostic interest on mortality, sTREM-1, when combined with mHLA-DR, may help to better identify immunosuppressed patients at risk of nosocomial infections.

Keywords: human leucocyte antigen-DR on monocytes; immunosuppression; mortality; nosocomial infection; septic shock; soluble triggering receptor expressed on myeloid cells 1.

Figure 1.

Soluble form of triggering receptor expressed on myeloid cells 1 (sTREM-1) concentration and human leucocyte antigen-DR on monocytes (mHLA-DR) expression in survivors and nonsurvivors septic shock patients. A, sTREM-1 concentrations were measured by enzyme-linked immunosorbent assay in survivors (white boxes, n = 87) and nonsurvivors septic patients (gray boxes, n = 29) at D1/D2, D3/D4, and D6/D8. Results were expressed as pg/mL. B, HLA-DR expression on monocytes was measured in survivors and nonsurvivors septic patients and expressed as number of anti-HLA-DR antibodies bound per cell (AB/C). Results are presented as Tukey box plots. Mann-Whitney tests were used for comparisons between groups at each time point. Only p values below 0.05 are shown.

Figure 2.

Soluble form of triggering receptor expressed on myeloid cells 1 (sTREM-1) concentration and human leucocyte antigen-DR on monocytes (mHLA-DR) expression in septic shock patients with or without nosocomial infection. A, sTREM-1 concentrations were measured by enzyme-linked immunosorbent assay in patients with secondary infection (gray boxes, n = 25) and in patients with no secondary infections (white boxes, n = 85) at D1/D2, D3/D4, and D6/D8. Results were expressed as pg/mL. B, HLA-DR expression on monocytes was measured in septic patients with or without secondary infections and expressed as number of anti-HLA-DR antibodies bound per monocyte (AB/C). Results are presented as Tukey box plots. Mann-Whitney tests were used for comparisons between groups at each time point. Only p values below 0.05 are shown.

Figure 3.

Cumulative incidence curves of nosocomial infection occurrence in septic shock patients stratified based on human leucocyte antigen-DR on monocytes (mHLA-DR) and/or soluble form of triggering receptor expressed on myeloid cells 1 (sTREM-1) values at D6/D8. A, Septic shock patients were stratified based on median mHLA-DR value measured at D6/D8 (i.e., 6668 antibodies bound per monocyte [AB/C]). The occurrence of nosocomial infections was monitored in both groups during 28 d after ICU admission. Red full line represents the group of patients with mHLA-DR at D6/D8 below median value. Black dotted line represents the group of patients with mHLA-DR value above median. Cumulative incidence curves were assessed using a Fine and Gray competing risk model. B, Septic shock patients were stratified based on median mHLA-DR value (i.e., 6668 AB/C) and median sTREM-1 value (i.e., 392 pg/mL) measured at D6/D8. The occurrence of nosocomial infections during 28 d after ICU admission was monitored in patients with both mHLA-DR below median value and sTREM-1 above median value at D6/D8 (red full line) and in the rest of the cohort (black dotted line). Cumulative incidence curves were assessed using a Fine and Gray competing risk model.