Abstract

Background: Intensive versus less stringent glucose control does not confer significant benefits in critical illness. However, insulin may have beneficial effects apart from its role in glucose homeostasis. Methods: Pilot, open-label, non-randomized study of glucose-insulin-potassium (GIK) infusion versus saline placebo in vasopressor dependent septic shock designed to test feasibility, safety, and preliminary efficacy. Inflammatory and endothelial biomarkers were measured before, during, and following treatment. Sidestream dark-field video microscopy and transthorasic echocardiograms were completed prior to and during treatment. Results: 9 patients were enrolled, 4 in the placebo and 5 in the GIK arm. Patients exhibited hyperglycemia at the start of infusion that normalized over time. Inflammatory biomarkers were non-significantly lower (p =0.05-0.25) in the interventional arm following treatment, while some endothelial biomarkers, most notably t-PAI (p = 0.02) were lower in the GIK arm. The microcirculatory measures non-significantly improved in the GIK arm compared to the placebo arm (p = 0.13-0.48). No differences in cardiac function nor vasopressor cessation were noted. Conclusion: GIK infusion is feasible but may cause hyperglycemia. There are potential beneficial effects on inflammation, endothelial activation, and microvasculature independent of macrovascular function. Further investigation of GIK and mechanisms driving these findings may be warranted.

Keywords: Endothelium, GIK, insulin, microcirculation, sepsis.

Graphical Abstract