Abstract Rationale Glycemic dysregulation is common in sepsis and contributes to higher mortality. The receptor for advanced glycation endproducts (RAGE) pathway responds to inflammatory and metabolic stress, yet the RAGE relationship with metabolic derangements in sepsis is unknown. We aimed to evaluate associations between RAGE pathway markers including soluble RAGE (sRAGE), advanced glycation endproducts (AGE), AGER mRNA expression, and early measures of glucose and lactate dysregulation. We hypothesized that increased RAGE pathway activation would be associated with greater disturbances in glucose and lactate homeostasis, and the effect of glucose dysregulation on mortality would be partially mediated by plasma sRAGE concentrations. Methods We conducted a prospective cohort study of critically ill subjects with sepsis (n = 674). We collected day 0 plasma and PAXgene whole blood. AGER mRNA expression was measured via Affymetrix Human Genome array (n = 161). We quantified plasma concentrations of sRAGE and AGE using enzyme-linked immunosorbent assays. We measured peaks and changes in glucose and lactate measures (Δ) within 24 hours of admission. We ascertained mortality at 30 days. Multivariable linear regression was used to test associations between log-transformed RAGE pathway markers and glycemic measures adjusting for age, sex, race, and interaction by diabetes mellitus. We applied causal mediation analysis (CMA) using generalized structural equation modeling to determine if plasma sRAGE concentrations partially mediate the association between glucose and mortality. Results Plasma sRAGE concentrations were positively associated with peak glucose (β = 13.8, 95%CI4.9-22.6; P0.01) and peak lactate levels (β = 0.8, 95%CI0.5-1.2; P0.01). Plasma AGE concentrations similarly were associated with peak glucose and peak lactate levels; plasma sRAGE and AGE levels were both associated with Δ-lactate measures while plasma sRAGE alone associated with Δ-glucose measures (Table). AGER expression was also associated with glucose levels (β = 44.2, 95%CI3.67-84.75; P=0.03). In the CMA, a 50 mg/dL increase in glucose was associated with a significant indirect effect on mortality via plasma sRAGE (OR = 1.05, 95%CI1.05-1.47; P0.01). The total effect of glucose on mortality was significant (OR = 1.10, 95%CI1.00-1.22; P=0.03), indicating partial mediation. The proportion of the total effect mediated by sRAGE was 25% (95%CI5%-43%; P0.01). Conclusion We demonstrate that RAGE pathway activation is associated with early glucose and lactate dysregulation in sepsis and partially mediates risk of sepsis-associated mortality. These findings suggest that RAGE pathway markers may reflect metabolic and inflammatory stress in sepsis and may identify high-risk metabolic phenotypes. Future work is warranted to understand whether RAGE pathway targeting can mitigate the impact of metabolic dysregulation on sepsis-associated mortality. This abstract is funded by: HL149851 (T.K.J.), HL161196 (N.J.M)
Storm et al. (Fri,) studied this question.