Serum selenium and glutathione peroxidase-3 activity: biomarkers of systemic inflammation in the critically ill?

To confirm the influence of systemic inflammatory response syndrome (SIRS) on selenium (Se) levels and prospectively evaluate the relationship between serum Se concentration [Se], glutathione peroxidase activity [GPx-3] and injury severity in patients at the time of intensive care unit (ICU) admission. Prospective, observational study. Multidisciplinary University Hospital ICU. A total of 36 ICU patients and 23 healthy volunteer subjects (HVS). Healthy volunteer subjects were designated as controls (Group 1). ICU patients were divided into three groups: without SIRS (Group 2); with SIRS (Group 3); with SIRS and multiple organ dysfunction syndrome (MODS) (Group 4). The latter groups had APACHE II scores > 15. [GPx-3] and [Se] were determined by standard methods within the first 48 h of admission to ICU. Kruskal-Wallis and Mann-Whitney U test were used for analysis of non-parametric continuous variables. The predictive value of [Se] and [GPx-3] for SIRS was calculated using a receiver operating characteristics (ROC) analysis. In SIRS and MODS patients [GPx-3] and [Se] decreased significantly (P = 0.0001 and P = 0.002, respectively). After ICU admission [GPx-3] and [Se] had a predictive value for SIRS ([GPx-3] sensitivity: 90%, specificity: 86.2% (cut-off value: 0.5 U/mL); [Se]: sensitivity 90%, specificity 72.4% (cut-off value: 60 mu g/L). [Se] had predictive value for ICU mortality (P = 0.034). Systemic inflammatory response syndrome and MODS were associated with early decreases in [Se] and [GPx-3]. Low [Se] and [GPx-3] after ICU admission had a predictive value for SIRS, which may aid future selection of patients who could benefit from Se supplementation.