Publication

Abstract : Background Sepsis is caused by a dysregulation of immune response to infection that results in very high mortality. Current laboratory tests and clinical criteria are inadequate to diagnose sepsis due to limited sensitivity and specificity. Circulating monocytes are important players in immune homeostasis and their altered HLA-DR expression indicate immune dysregulation. HLA-DR is an MHC Class II cell-surface receptor that can present foreign antigens to helper T cells and mount an inflammatory response. Therefore, we analyzed the variations in HLA-DR expression and the concentration of monocyte subsets for diagnosing post-surgical sepsis. Methods In this double-blinded prospective cohort study, we adopted immunophenotyping and quantification of antigen expression by flowcytometry to detect the changes in circulating monocyte subsets in patients undergoing cardiac surgery. Statistical analysis was performed to identify significant changes and based on the predictive potential of measured variables ROC curve analysis was done. ROC curve permitted the choice of appropriate cut-off values using which a diagnostic protocol was developed. Results We observed that the monocyte subset concentrations in circulation varied differently after surgery. There was a significant downregulation of monocytic HLA-DR on both intermediate (p = 0.0477) and non-classical monocytes (p = 0.0333) at 48 h post-surgery. The monocyte subset analysis clearly showed that the patients with reduced pre-surgical non-classical monocyte count (p = 0.0430) coupled with post-surgical down-regulation of HLA-DR expression on the same subset had a higher incidence of developing sepsis after cardiac surgery. Conclusions Here we are reporting for the first time, the significant influence of non-classical monocytes in inducing dysregulated host response and sepsis after cardiac surgery. Using multiple biomarkers associated with this monocyte subset, we established an algorithm for the diagnosis of sepsis at 48 h post cardiac surgery with 100% sensitivity and 69.23% specificity.