Value of Biomarkers of Brain Injury for Predicting Psycho-cognitive Sequelae Secondary to Sepsis: a Prospective Multicenter Study

Sepsis is a major cause of admission to intensive care units and is responsible for approximately 11 million deaths worldwide each year. It may be complicated by an acute brain dysfunction known as sepsis-associated encephalopathy (SAE), which affects about 50% of patients and typically manifests as delirium or coma. The diagnosis of SAE is primarily clinical, with EEG and brain MRI providing supportive information. Risk factors for SAE are mainly related to patient characteristics, including advanced age, chronic kidney disease, and pre-existing neurodegenerative or cognitive disorders. The pathophysiology of SAE involves three major mechanisms: neuroinflammation, endothelial dysfunction with blood-brain barrier disruption, and mitochondrial dysfunction leading to neuronal injury. These mechanisms likely explain both acute neurological symptoms and long-term psycho-cognitive sequelae. Following sepsis, 30-60% of patients develop psychiatric disorders and cognitive impairments comparable to those observed after moderate traumatic brain injury or early Alzheimer's disease. These sequelae are part of post-intensive care syndrome (PICS), supporting the need for structured post-ICU follow-up. However, the optimal target population and organization of such follow-up remain unclear, as some studies have reported reduced quality of life in patients receiving long-term follow-up. Identifying early biomarkers predictive of psycho-cognitive outcomes is therefore crucial, as current data on neuronal and glial biomarkers remain limited. Such biomarkers could improve prognostication, guide cognitive rehabilitation and psychological support, and contribute to the development of future neuroprotective strategies.

Primary objective:

To evaluate the value of biomarkers of brain injury for predicting cognitive impairment (memory impairment assessed by a MoCA score \< 26/30) at 3 months.

Secondary objectives:

* To evaluate the value of brain injury biomarkers for predicting delirium in the ICU, including hypoactive, hyperactive, and mixed phenotypes * To assess the association between brain injury biomarkers and the duration of delirium in the ICU * To evaluate the value of brain injury biomarkers for predicting psychological sequelae (anxiety, post-traumatic stress disorder, and depression) at 3 months * To evaluate the value of brain injury biomarkers for predicting functional neurological outcomes using the Glasgow Outcome Scale-Extended at ICU discharge and at 3 months * To assess the association between brain injury biomarkers and Post-Intensive Care Syndrome (PICS) * To assess the association between brain injury biomarkers and EEG abnormalities in the ICU * To assess the association between brain injury biomarkers and MRI abnormalities at 3 months

• Coma- and/or delirium-free days [Time frame: Day 10 after inclusion]
• Duration of delirium in the intensive care unit (ICU) [Time frame: 3 months]
• Type of ICU delirium: hypoactive, hyperactive, or mixed [Time frame: 3 months]
• Psychiatric disorders (anxiety, depression : ( 0 =min; 21 =max) , or PTSD) [Time frame: 3 months]
• Glasgow Outcome Scale-Extended (GOSE) [Time frame: 3 months]
• Delirium duration [Time frame: 3 months]
• Delirium phenotype [Time frame: 3 months]
• Psychiatric outcomes [Time frame: 3 months]
• Functional outcome [Time frame: 3 months]
• PICS : Post-Intensive Care Syndrome (PICS) [Time frame: 3 months]
• Neurological, psychiatric, or rehabilitation follow-up proposed to the patient [Time frame: 3 months]