DIscovering SubPhenotypes and ENdotypes in pediatric SEpsis-induced MODS: the DISPENSE Study

Project Summary/Abstract Sepsis is a “syndromic” definition reflecting patients with great heterogeneity in age, comorbidities, pathophysiology, and inciting triggers. This heterogeneity likely contributes to a myriad of negative randomized controlled trials (RCTs). The identification of higher-risk subgroups (e.g., subphenotypes) within these heterogeneous disease states is a critical gap in knowledge prohibiting development of tailored treatments targeting subgroups with similar biology (e.g., endotypes). Despite years of research, aside from timely antibiotic administration, pharmacologic treatment has proven ineffective in decreasing mortality and morbidity in children with sepsis associated multi-organ dysfunction syndrome (MODS) likely due, in part, to heterogeneity in these patients. There is limited to no understanding of the progression of pathophysiologic changes across the time course of pediatric severe sepsis, nor how those changes may be impacted by tailored treatments. Clinical and/or plasma biomarker-based variables have been successfully used to identify subphenotypes, while gene expression techniques have aided in identifying more mechanistically based endotypes. The PRECISE GRACE2 (NCT05266001) and TRIPS (NCT05267821) trials offer an unprecedented opportunity to identify subphenotypes in children with severe sepsis-induced MODS using latent class analysis (LCA) and to gain insight into mechanistic differences between both LCA-defined subphenotypes and PRECISE-defined subphenotypes over time and in response to treatment. Specific Aims are: 1. To test the hypotheses that a) in children with sepsis-induced MODS, de novo LCA will identify subphenotypes, and that b) clinical outcomes and c) response to treatment will vary between these subphenotypes. 2. To test the hypotheses that genome-wide expression profiling will identify a) de novo transcriptomic subphenotypes that will differ in outcome and response to treatment and that b) gene expression will identify treatment-triggered molecular changes in either the transcriptomic, and/or LCA-, and/or PRECISE-defined subphenotypes. Consistent with NICHD Pediatric Trauma and Critical Illness Branch research priorities, completion of this work will confirm whether subphenotypes in children with sepsis-induced MODS exist and reflect underlying mechanistic differences, thus allowing stratification into groups with differing risk (prognostic enrichment) and/or potential responsiveness to targeted therapies (predictive enrichment) allowing identification of potential treatable traits and new precision-based diagnostic and therapeutic targets in these highest-risk children.