Prognostic value of soluble ST2 in an unselected cohort of patients admitted to an intensive care unit - The Linz Intensive Care Unit (LICU) study

Background: Soluble ST2 (sST2) has emerged as a prognostic biomarker in patients with heart disease. We tested the hypothesis that sST2 is an independent predictor of mortality in patients admitted to an intensive care unit (ICU). Methods: We performed measurements of sST2 plasma concentrations in 530 consecutive patients admitted to a medical ICU of a tertiary care hospital during a study period of one year. The patients recruited during the first six months were used for the derivation cohort (n = 274) and the patients recruited during the second six months were used for the validation cohort (n = 256). The endpoint was defined as 90-day all-cause mortality. Results: In the derivation cohort, sST2 was higher among decedents (n 56; median, 146 U/mL) than survivors (n = 218; median 42 U/mL, p<0.001). In multivariate Cox proportional-hazard regression analysis (offering age, sex, BMI, APACHE II score, SAPS II, CRP, IL-6, PCT, creatinine, total cholesterol, albumin, hs-cTnT, BNP and sST2 as independent variables), sST2 was a significant predictor of mortality (risk ratio 1.48,95% CI 1.15-1.90; p = 0.002 per 1 SD increase in log transformed values). In this statistical model, only sST2 and SAPS II contributed independently to mortality prediction. We further observed an additive effect of an sST2 plasma concentration of >84 U/mL and an increased SAPS II for mortality prediction. The findings from the derivation cohort were confirmed in the independent validation cohort. In those patients with a length of stay of >48 h at the ICU (n = 225), sST2 obtained two days after baseline measurement had a better capability than baseline sST2 to predict mortality. Conclusions: In an unselected cohort of patients admitted to the ICU, sST2 was an independent predictor of 90-day all-cause mortality and added prognostic information to the SAPS II. (C) 2011 Elsevier B.V. All rights reserved.