Pre-infection cerebral cortex structure predicts murine sepsis outcome

Abstract

Sepsis is a life-threatening condition caused by an inappropriate host response to an infection that can lead to multi-organ failure and death. Predictive markers that can inform the trajectory and outcome for a septic patient are necessary to inform treatment and increase the likelihood for patient survival. Here, by leveraging the concept of Lethal Dose 50 (LD50), which is the dose of a pathogen that will kill 50% of a genetically identical host population, we tested the hypothesis that variations in brain structure can be readily used to predict trajectory and outcome in a murine model of sepsis. We found that one week prior to infection, mice that were fated to survive exhibited greater cortical volume, solidity, and thickness compared to those who would succumb to the LD50 challenge, and these metrics were sufficient to train multiple predictive models with 75% to 94% accuracy. Our work reveals a readily measurable non-genetic marker for predicting sepsis prognosis in mice and demonstrates the potential for using pre-infection high-resolution structural brain scans to predict infection outcomes in humans.