Defining Circulating Micro-RNA Biomarkers for the Early Diagnosis and Prognosis of Sepsis

Prospective observational cohort of patients along the entire sepsis continuum in National University Hospital Emergency Department (ED) will be enrolled along with non-infective controls. Whole blood samples will be separated immediately into plasma and serum for storage at the Tissue Repository. Those who are subsequently admitted to the general ward, high dependency (HD) or intensive care unit (ICU) will have their 2nd and 3rd samples obtained at 24-48 hours and 48-72 hours respectively. The 2nd and 3rd samples will be used as prognostic markers (Objective #2). Patients who are discharged directly from the ED will be tracked for any clinical recurrence of their disease within 28 days to ensure the diagnostic accuracy of the first sample of biomarkers that are extracted.

Samples will be batch-processed via the circulating miRNA profiling workflow comprising of the pre-analytics, analytics, and data processing and analysis phases. This qPCR-based miRNA profiling has much smaller feature to sample ratio thus allowing a global sequential forward selection to optimally combine a host of features with complementary prediction strengths to form the biomarker panel which has the least number of components and the maximized classification power (ROC AUC). Panel robustness measurement using computational generated noises and validation with a larger set of blinded samples will be performed.

The final objective will be assessed by measuring the concentrations of secreted cytokines, chemokines and reactive oxygen species after synthesis and transfection the newly-derived miRNA panel into monocytes and monocytic cell lines.