, Volume 36, Issue 4, pp 630-637,
Open Access This content is freely available online to anyone, anywhere at any time.
Date: 12 Feb 2010

Soluble ST2 plasma concentrations predict mortality in severe sepsis



Patients with sepsis—after surviving the initial hyperinflammatory phase—display features consistent with immunosuppression, including hyporesponsiveness of immunocompetent cells to bacterial agents. Immunosuppression is thought to be facilitated by negative regulators of toll-like receptors, including membrane-bound ST2. We investigated the release of soluble ST2 (sST2), a decoy receptor that inhibits membrane-bound ST2 signaling, during sepsis.


The study population comprised 95 patients with severe sepsis admitted to one of two intensive care units (ICUs) at the day the diagnosis of severe sepsis was made. Blood was obtained daily from admission (day 0) until day 7 and finally at day 14. Twenty-four healthy subjects served as controls. sST2 and cytokines were measured in serum.


Mortality among patients was 34% in the ICU and 45% in the hospital. On admission, sepsis patients had higher sST2 levels [10,989 (7,871–15,342) pg/ml, geometric mean (95% confidence interval, CI)] than controls [55 (20–145) pg/ml, P < 0.0001]. Serum sST2 remained elevated in patients from day 0 to 14 and correlated with disease severity scores (P < 0.001) and cytokine levels on day 0 and during course of disease (P < 0.0001). Nonsurvivors displayed elevated sST2 levels compared with survivors of the intensive care unit (P < 0.0001).


Sepsis results in sustained elevation of serum sST2 levels, which correlates with disease severity and mortality.