Finding Appropriate Subtypes in a Trial of Balanced versus nOrmaL Saline FlUid in Sepsis

PROJECT SUMMARY/ABSTRACT Sepsis, defined as a dysregulated host response to infection, is responsible for 75,000 childhood hospitalizations annually in the United States and 50% of inpatient pediatric deaths worldwide. Despite multiple trials, there are no targeted therapies for adult or pediatric sepsis, and supportive care with timely antibiotics and fluid resuscitation remains the mainstay of therapy. Fluid therapy is a cornerstone of sepsis resuscitation, although the choice of fluid remains controversial. Recent trials in adults showed lower rates of major adverse kidney events within 30 days (MAKE30) with balanced crystalloid, such as lactated Ringer’s (LR), relative to normal saline (NS). MAKE30 is a composite endpoint incorporating persistent kidney dysfunction, initiation of dialysis, or death. Balanced solutions like LR have a composition more similar to normal human serum, whereas NS has been associated with worse kidney function, albeit via unclear mechanisms. While much of pediatric sepsis management is extrapolated from adults, children have a distinct epidemiology and outcome profile, making application of adult data problematic. Thus, to specifically assess whether LR or NS resuscitation improves MAKE30 in children, the PRagMatic Pediatric Trial of Balanced versus nOrmaL Saline FlUid in Sepsis (PRoMPT BOLUS) was undertaken, with planned enrollment of 8,800 children. However, heterogeneity has contributed to negative trials in sepsis, as therapies effective in some patients are ineffective in others. To mitigate this heterogeneity, biomarkers have been proposed for both risk stratification and identification of sub-phenotypes with shared pathophysiology. Our study, entitled Finding Appropriate Subtypes in a Trial of Balanced versus nOrmaL Saline FlUid in Sepsis (FAST BOLUS), will assess for heterogeneity of treatment effect of LR versus NS in the PRoMPT BOLUS trial. Leveraging our group’s extensive experience with biomarker-defined risk stratification and sub-phenotyping, we will measure plasma biomarkers in 800 children collected at randomization and assess the for differential effects of fluid assignment on MAKE30 across risk strata using two separate biomarker-based mortality prediction models (Aim 1). Additionally, we will test for differential effects of fluid assignment on MAKE30 after stratifying subjects according to one of two biomarker-defined inflammatory sub-phenotypes (Aim 2). Lastly, we will leverage this biobank to investigate potential mechanisms between fluid choice and MAKE30 by measuring markers of endothelial cell (angiopoietin-2) and endothelial glycocalyx (syndecan-1) damage (Aim 3). These Aims will assess the utility of established biomarker-based strategies for prognostic (Aim 1) and predictive enrichment (Aim 2) strategies by leveraging an ongoing pediatric sepsis trial, which is a necessary analysis of randomized trials conducted in heterogeneous critical illness syndromes. Successful completion of the Aims will provide evidence for personalized fluid resuscitation in pediatric sepsis and establish proof of concept data for incorporation of biomarker-based methods to reduce heterogeneity in future trials.