Use of Time Series Biomarker and Clinical Data to Construct a Time Trajectory Host Response Map

点击翻译按钮获取中文摘要

Principal Investigator/Program Director (Last, first, middle): Reddy, Jr., Bobby Project Summary: Sepsis is an incompletely understood clinical syndrome characterized by a dysregulated host response to infection. In partnership with 8 U.S. hospitals, Prenosis amassed one of the world’s largest datasets and biobanks that combines biomarker and clinical data for patients suspected of infection, housing over 70,000 plasma or serum samples from over 14,000 patients. We also curated a dataset of dense time-series data from each patient’s Electronic Medical Record (EMR), including demographics, vitals, lab results, interventions, outcomes, and many other parameters. To commercialize insights from these data, Prenosis built ImmunixTM, an FHR/HL7 compatible software platform for clinical deployment of diagnostics and clinical decision support tools. Under a previously awarded NIGMS grant (1R44GM139529), Prenosis built a testing pipeline to measure 40 critical protein biomarkers from biobanked samples. To date, we measured these biomarkers on only the initial sample per patient for 6,000 patients and combined these data with EMR clinical parameters to construct 8 endotypes of sepsis. The identification and classification of endotypes—groupings of patients with similar biologic and clinical features—is increasingly becoming recognized as a valuable methodologic approach to assessing patients with sepsis. To complete work for the existing grant, Prenosis will measure the baseline sample for additional patients to total about 10,000 patients to refine and validate the endotypes. In this project, Prenosis proposes to add a critical new dimension to the data by assaying and analyzing longitudinal, time-series biomarker data. We will leverage our pipeline to measure the 40 core biomarkers from 9,000 follow-up samples from ~3,400 patients that we have already collected and stored in the biobank. We will assess the additional value of longitudinal time-series biomarker measurements and clinical data. Initial feasibility data from over 1,000 measured samples demonstrates that longitudinal data provide additional powerful new biologic and prognostic insights. Analytics built upon these data have the potential to improve diagnostic and drug development products for sepsis and COVID. The overall hypothesis of this project is that longitudinal biomarkers will add a valuable biologic and prognostic dimension to endotypes for sepsis; and these longitudinal endotypes will better classify patients who may have a heterogeneous response to sepsis therapies.