Characterization of sepsis inflammatory endotypes using circulatory proteins in patients with severe infection: a prospective cohort study

Background Sepsis is a heterogeneous syndrome due to a variable range of dysregulated processes in the host immune response. Efforts are made to stratify patients for personalized immune-based treatments and better prognostic prediction. Using gene expression data, different inflammatory profiles have been identified. However, it remains unknown whether these endotypes mirror inflammatory proteome profiling, which would be more feasible to assess in clinical practice. We aim to identify different inflammatory endotypes based on circulating proteins in a cohort of moderately ill patients with severe infection (Sepsis-2 criteria). Methods In this prospective study, 92 inflammatory plasma markers were profiled using a targeted proteome platform and compared between patients with severe infection (Sepsis-2 criteria) and healthy controls. To identify endotypes with different inflammatory profiles, we performed hierarchical clustering of patients based on the differentially expressed proteins, followed by clinical and demographic characterization of the observed endotypes. Results In a cohort of 167 patients with severe infection and 192 healthy individuals, we found 62 differentially expressed proteins. Inflammatory proteins such as TNFSF14, OSM, CCL23, IL-6, and HGF were upregulated, while TRANCE, DNER and SCF were downregulated in patients. Unsupervised clustering identified two different inflammatory profiles. One endotype showed significantly higher inflammatory protein abundance, and patients with this endotype were older and showed lower lymphocyte counts compared to the low inflammatory endotype. Conclusions By identifying endotypes based on inflammatory proteins in moderately ill patients with severe infection, our study suggests that inflammatory proteome profiling can be useful for patient stratification.