Abstract:

Sepsis-associated encephalopathy (SAE) is the most common and severe neurological complication in patients with sepsis, with an incidence rate as high as 70% in the intensive care unit. SAE not only significantly increases in-hospital mortality but also seriously affects the long-term quality of life of survivors due to the resultant long-term cognitive dysfunction, imposing a heavy burden on society and families. However, the clinical manifestations of SAE are highly heterogeneous and are often obscured by the severe symptoms of the primary disease, and the lack of specific bedside diagnostic tools leads to a clinical recognition rate far lower than the actual incidence. Traditional brain injury biomarkers, such as S100β protein and neuron-specific enolase (NSE), have obvious limitations in clinical application due to insufficient source specificity and interference from hemolysis, among other factors. Meanwhile, neuroimaging examinations, although capable of detecting structural abnormalities, suffer from low positivity rates, cumbersome operation, and difficulty in achieving dynamic monitoring, thus failing to meet the needs for early identification and dynamic assessment of SAE. Therefore, exploring biomarkers that can objectively, sensitively, and dynamically reflect sepsis-induced brain injury has become a critical issue to be addressed in this field. Neurofilament light chain (NfL) is a core component of the neuronal axonal cytoskeleton. When axons are injured by ischemia, inflammation, or metabolic stress, the neurofilament network disintegrates, and NfL is rapidly released into the cerebrospinal fluid and can cross the blood-brain barrier to enter the peripheral blood. In recent years, the maturation of ultra-sensitive detection technologies represented by single-molecule array has made accurate quantification of peripheral blood NfL feasible, laying the methodological foundation for its application in the critical care setting. Existing studies have shown that serum NfL levels are significantly elevated in patients during the acute phase of sepsis and are closely correlated with disease severity, duration of delirium, and poor neurological outcomes, demonstrating promising value for early diagnosis and prognostic stratification. Meanwhile, NfL can comprehensively reflect the cumulative effects of multiple pathological pathways, including neuroinflammation, ischemia-hypoxia, mitochondrial dysfunction, and blood-brain barrier disruption, and is thus regarded as an "integrative biomarker" of brain injury in SAE.This review elaborates on the biological basis and methodological advances in NfL detection, with a focus on summarizing the clinical research evidence regarding its acute-phase dynamic changes in SAE, associations with pathophysiological mechanisms, comparative advantages over traditional brain injury biomarkers, and correlations with imaging alterations. Furthermore, it explores the strategies for early identification and risk stratification based on NfL, as well as its potential as an intervention target, and objectively analyzes the challenges currently facing clinical translation, including lack of standardization, the influence of renal function, and cost-effectiveness, with the aim of providing references for the standardized application of NfL in the clinical management of SAE and for future research directions.

Key words: sepsis, sepsis-associated encephalopathy, neurofilament light chain, biomarker, brain dysfunction, axonal transport