Early co-activation of neutrophil extracellular traps and complement predict mortality from thrombo-inflammatory complications of sepsis: Prospective cohort study

Abstract

Background: Acute systemic disease is often compounded by thrombo-inflammation wherein three innate immune activation and coagulation pathogenesis convergence of interest to promote microvascular thrombosis and organ injury. The role of neutrophilic extracellular traps (NETs) and complement activation in immunothrombosis has been implicated but little is known about the clinicopathologic evidence for cooperative (or synergistic) effects on coagulopathy and outcome. Objective: To determine early NET and complement activation profiles in sepsis, and test of the interaction between them and thromboinflammatory complications including overt DIC and 28-day survival from sepsis. Methods: In this study, a prospective observational cohort study of 100 adults who had sepsis and 50 controls. Blood sampling was done at T0 (0-6 h), serial sampling at 24 h and 72 h of a set at n=80. NET burden was evaluated with the help of biomarkers [citrullinated histone H3 [H3Cit], neutrophil elastase [NE], nucleosomes] based on the enzyme immunoassay (ELI) and summarized in a NET index (z-score composite]. Complement activation was determined by the use of C5a and soluble C5b-9 (sC5b-9) and expressed as a complement index. Coagulopathy was assessed by the use of routine tests of coagulation and overt DIC classification (ISTH score >=5). Multivariable logistic regression and Cox proportional hazards models were used to test for independent associations and an interaction (NET incentivisation of SOA * complement) after adjustment for age and SOFA score. ROC analysis was used to evaluate discriminative performance. Results: At T0, NET and complement biomarkers were found to be significantly up regulated in sepsis compared to controls (H3Cit 420 (310-540) vs 95 (70-120) ng/mL; sC5b-9 590 (450-730) vs 180 (150-220) ng/mL; all p= 0.001). Biomarkers decreased after 72 h but were abnormal (all p<0.001 for trend). NET and complement indices had a high correlation (rho=0.71, p<0.001) and were correlated with SOFA and D-dimer. In adjusted models, with a significant interaction (OR1.68, 95% CI1.12-3.27), NET index (OR2.35, 95% CI1.48-3.71) and complement index (OR2.12, 95% CI1.37-3.27) predicted overt DIC (OR1.68, 95% CI1.12-3.25). Both indices were able to predict 28-day mortality (NET index HR 1.84, 95% CI 1.32-2.56; complement index HR 1.63, 95% CI 1.18-2.26), and the two were significantly associated (HR 1.47, 95% CI 1.05-2.05). The combined dual index model had excellent discrimination for composite thrombo-inflammatory outcomes (AUC time unit model 0.93, 95% CI [88-97]). Conclusions Early co-activation of NETs and complement is related to enhanced thrombo-inflammation in sepsis. A large NET suspension x complement is conducive to support a pathway cooperative effect, besides independent association. Dual pathway biomarker stratification demands external validation and may be used for precision strategy to mitigate sepsis associated coagulopathy.