Elevated Plasma Angiopoietin-2 Levels Are Associated With Fluid Overload, Organ Dysfunction, and Mortality in Human Septic Shock

OBJECTIVES:Angiopoietins modulate endothelial permeability via endothelial cell junctions. Angiopoietin-2 blocks the angiopoietin-1/Tie-2 interaction that stabilizes these junctions, and elevated plasma angiopoietin-2 levels are associated with vascular leakage. We hypothesized that plasma angiopoietin-1 and angiopoietin-2 levels are associated with indirect markers of increased vascular permeability, organ dysfunction, mortality, and plasma proinflammatory cytokine levels in human septic shock. DESIGN:Multicenter observational cohort study derived from a randomized controlled trial (Vasopressin and Septic Shock Trial of vasopressin versus norepinephrine in septic shock). SETTING:ICUs of hospitals in Canada, Australia, and the United States. PATIENTS:Three hundred forty-one patients in the randomized, controlled Vasopressin and Septic Shock Trial trial of vasopressin versus norepinephrine in septic shock. INTERVENTIONS:None. MEASUREMENT AND MAIN RESULTS:We measured plasma levels of angiopoietin-1 and angiopoietin-2 at study baseline and determined their association with percent fluid overload and acute organ dysfunction and generated a receiver operating characteristic curve for plasma angiopoietin-2 levels versus acute kidney injury. We also determined the association of angiopoietin-1 and angiopoietin-2 levels with hemodynamics, mortality, and plasma cytokine levels. Plasma angiopoietin-2 levels were directly associated with percent fluid overload at baseline (rs = 0.18; p = 0.0008) and at 6 hours (rs = 0.13; p = 0.023), but not at 24 hours (rs = 0.041; p = 0.46). Plasma angiopoietin-2 levels were associated with the development of hepatic (p < 0.0001) and coagulation (p < 0.0001) dysfunction and acute kidney injury (p < 0.0001). Receiver operating characteristic curve had an area under the curve of 0.73 for acute kidney injury. angiopoietin-2 levels were also inversely associated with days alive (r = –0.24; p = 0.010) and positively associated with increased 7-day (log-rank trend chi-square = 5.9; p = 0.015) and 28-day (log-rank chi square = 4.9; p = 0.027) mortality. A threshold of angiopoietin-2 levels above the first quartile (> 5,807 pg/mL) was observed to be associated with increased mortality risk, which aligns with prior studies. Plasma angiopoietin-2 levels were positively associated with plasma cytokine levels, including tumor necrosis factor-α and interleukin-6 at baseline (rs = 0.39; p < 0.0001 and rs = 0.51; p < 0.0001) and at 24 hours (rs =