Soluble ST2 plasma concentrations predict 1-year mortality in acutely dyspneic emergency department patients with pulmonary disease.

Abstract

We evaluated the association between ST2 concentrations and mortality at 1 year in 231 acutely dyspneic patients with pulmonary diseases seen in the emergency department. Blood concentrations of ST2 were ascertained; using 1-year survival as the reference standard, receiver operating characteristic curves with resultant area under the curve (AUC) were measured. Cox proportional hazards models identified independent predictors of 1-year death. Hazard curves compared rates of death as a function of ST2 concentration. Concentrations of ST2 were significantly higher in patients with pulmonary diseases compared with 153 subjects without cardiopulmonary disease (0.23 vs 0.11 ng/mL; P = .01). Among patients with pulmonary diseases, concentrations of ST2 were higher among decedents compared with survivors (1.14 ng/mL vs 0.19 ng/mL; P < .001). ST2 had an AUC of 0.72 as a predictor of death (P < .0001). An ST2 of 0.20 ng/mL had a hazard ratio for death of 6.1 (95% confidence interval, 1.8-21.0; P = .004). Compared with patients with lower ST2 concentrations, mortality rates for patients with an enrollment ST2 of 0.20 ng/mL or more diverged early and rose progressively in 1 year (P < .001). ST2 concentrations are frequently elevated in acute pulmonary diseases and are markedly prognostic for death by 1 year.