REAL-TIME EVALUATION OF BRAIN INJURY SEVERITY IN SEPTIC MICE USING A NOVEL SCORING SYSTEM

Objectives : The difficulty in translating findings from basic research on sepsis associated encephalopathy (SAE) into clinical practice may be attributed to the suboptimal assessment methods currently in use. The objective was to develop an assessment index to dynamically evaluate brain injury severity in the septic acute phase, making the experimental data more representatives of clinical SAE patients. Methods : We independently developed the Sepsis-Associated Brain Injury Score (SABIS) based on Quick Sequential Organ Failure Assessment (qSOFA), Full Outline of Unresponsiveness, and Confusion Assessment Method of ICU methodologies. Under blind conditions, we validated SABIS's effectiveness and accuracy by assessing its correlation with brain tissue pathology and its ability to predict mortality, and compared the variance of SABIS and classical scoring system from different evaluators in the same batch of models to verify its standardization and generalizability. We used condition-matched male and female mice to establish cecal ligation and puncture, feces intraperitoneal injection, and endotoxemia models, monitoring SABIS changes to investigate gender and modeling method effects. Results : At the same time point and detection region, the Spearman correlation analysis between SABIS and three brain injury pathological indicators showed positive results. SABIS predicts short-term and long-term mortality as well as the classical modified murine sepsis score, and its operator-derived heterogeneity index is significantly lower. Evaluating SABIS can reveal the impact of gender and modeling method on sepsis-related brain injury characteristics. Conclusions : Our novel SABIS can robustly and accurately assess brain injury severity in various sepsis animal models. The scoring system demonstrates good generalizability and high consistency with pathological indicators, enhancing the translational potential of sepsis research.