W165 - A Multi-modal Longitudinal Model of Sepsis Reveals Immune Dynamics Associated with Mild, Severe and Fatal Disease

Ayelet Alpert – Technion-Israel Institute of Technology; Timothy Cooper – Technion - Israel Institute of Technology; Neta Milman – Technion- Israel Institute of Technology; Danny Eytan – Technion - Israel Institute of Technology; Rambam Health Care Campus; Ashham Mansur – University Medical Center Goettingen; Caspar Mewes – University Medical Center Goettingen; Shai Shen-Orr – Technion - Israel Institute of Technology

Abstract Text: Sepsis remains a significant global health concern, posing complex challenges due to its diverse origins and limited treatment options. Immune responses during sepsis are heterogeneous and may include a mixture of immune paralysis, hypo-responsiveness, or hyper-inflammation. Previous work primarily examined the role of the immune system in sepsis through cross-sectional data or by focusing solely on a singular aspect or data modality, thereby overlooking the complexity of the system. Instead, here, we take a systems-level approach and hypothesize that the timing, sequence, and interaction between multiple players in the immune system are crucial in determining the course and outcome of sepsis. Specifically, we built a multi-modal, longitudinal disease model of sepsis by profiling the peripheral blood of 26 sepsis patients, with varying clinical progressions, across 4-9 time points using mass cytometry (CyTOF), mass-spec proteomics, cytokine arrays, whole-transcriptome sequencing and detailed clinical measurements. Collectively, this rich dataset enabled us to align patients in pseudo-time and construct a shared axis that describes sepsis and recovery. By overlaying features from multiple modalities, we describe the molecular units and biological pathways that participate in sepsis and reveal their dynamics throughout the disease. Finally, by using a clinically heterogeneous cohort, we found specific events, or molecular “misbehaviors” that are associated with worse clinical courses or outcomes. Overall, our work helps to decipher - in detail - the immune system’s failure(s) during sepsis and across different outcomes. This work paves the way towards reliable, actionable, immune monitoring tools and immune-directed interventions for sepsis.