To investigate the potential utility of serum uncoupling protein-2 (UCP2) level as a biomarker in septic patients.
Critically ill patients with diagnoses of sepsis – sepsis non-shock group (n = 20) and septic shock group (n = 53), and a control group (n = 15) were enrolled within 24 h of entry into the ICU. Serum levels of UCP2 were measured by enzyme-linked immunosorbent assay (ELISA) at ICU admission for all the groups and at ICU discharge for septic shock group. Clinical parameters and laboratorial tests (APACHE II, SOFA, lactate, etc.) were also collected.
Serum UCP2 concentrations on ICU admission were significantly increased in septic shock group and sepsis non-shock group, compared with control subjects (263.21 ± 29.99 vs 115.96 ± 32.99 vs 60.56 ± 10.05 pg/ml, P < 0.001). Concentrations of UCP2 performed better than other parameters (APACHE II score, SOFA score, PCT, and WBC) in predicting the incidence of sepsis or septic shock on day of ICU admission, as reflected by AUC. On day of ICU admission, the AUC for UCP2 level associated with 28-day mortality was 0.704, higher than the AUC for SOFA and APACHE II score. Patients with higher admission levels of UCP2 (>246.52 pg/mL) had significantly increased 28-day mortality compared with those with lower UCP2 levels (<246.52 pg/mL).
Serum UCP2 levels at admission were markedly increased in patients with sepsis, which is useful for early diagnose and prognostic prediction. UCP2 is a potential biomarker for sepsis, or even a subtype of sepsis.

Author