Comprehensive Longitudinal Cytokine Profiling and their Associations with Distinct Immunologic Phenotypes and Outcomes: A Prospective Clinical Study in Venovenous ECMO
Abstract
INTRODUCTION:
Venovenous extracorporeal membrane oxygenation (VV ECMO) is a life‑saving intervention commonly used for severe and refractory respiratory failure. However, patient management and outcomes in this setting remain challenging as well as difficult to predict. Inflammatory reaction, namely by assessing cytokine and chemokine blood levels, may provide prognostic information, but their role in VV ECMO has not been comprehensively characterized.
METHODS:
We prospectively enrolled 35 adult patients receiving VV ECMO for refractory respiratory failure. Blood samples were collected immediately prior to ECMO initiation and longitudinally on Days 1-4, 7, and weekly until decannulation. Plasma concentrations of 48 cytokines and chemokines were measured using a multiplex bead‑based assay, along with several enzymes by enzyme-linked immunosorbent assay. Principal component analysis and hierarchical clustering were used to identify immunologic phenotypes. Associations between biomarker levels, clusters, and clinical outcomes were evaluated using regression models.
RESULTS:
Elevated pre‑ECMO concentrations of TNF‑B and RANTES were significantly associated with reduced survival, while higher IFN-a2, IL-1Ra, IL-12p40, MCP-1, IL-8, MIP-1a, IL-1, and SCF were associated with fewer adverse events (p<0.05). Several cytokines, including IFN‑a2, IP‑10, MCP‑1, MIP‑1a, SCF, and DPP3, were significantly correlated with prolonged ventilator dependence (p<0.05). Cluster analysis revealed distinct immunologic phenotypes among subjects.
CONCLUSIONS:
Inflammatory markers may be associated with clinically meaningful endpoints in VV ECMO, namely, survival, adverse events, and prolonged dependence on mechanical ventilation. To our knowledge, this study represents the most comprehensive cytokine and chemokine profiling in VV ECMO patients to date, using immediate pre‑ECMO sampling and advanced clustering techniques. Larger multicenter studies are warranted to validate these findings and explore their application in guiding individualized VV ECMO management and enhancing patient prognostication.