In vitro models of blood-spinal cord barrier in spinal cord injury: A meta-analysis of model design, performance, and physiological relevance.

BACKGROUND: In vitro models of the blood-spinal cord barrier (BSCB) are widely utilized for developing therapeutics against neurological diseases such as spinal cord injury (SCI). However, high variability among existing in vitro BSCB models severely limits their predictive power for preclinical research, highlighting the need for a comprehensive synthesis of current model characteristics and performance. METHODS: We first reviewed in vitro BSCB models under SCI conditions and conducted a comprehensive meta-analysis of mainstream Transwell-based models. The synthesis systematically analyzed critical factors influencing model performance, including cell types, disease modeling strategies (e.g., hypoxic culture, inflammatory stimuli, oxidative stress), and the inclusion of biomaterial matrices. The meta-analysis quantified differences in key BSCB properties, including permeability, transendothelial electrical resistance (TEER), and junctional/inflammatory protein expression, across healthy, diseased, and treatment contexts. RESULTS: The synthesis identified cell composition, disease induction methods, and biomaterial matrix inclusion as core factors determining the ability of in vitro BSCB models to replicate physiological and pathological characteristics. Meta-analysis results revealed significant quantitative differences among three types of BSCB models. Specifically, disease models exhibited consistently higher permeability (measured via FITC-dextran) and lower mean TEER values (97.94 Ω·cm2) compared to healthy models (183.73 Ω·cm2) and treatment groups (146.28 Ω·cm2). Additionally, co-culturing BSCB endothelial cells with glial cells or pericytes was demonstrated to significantly enhance the physiological relevance of the models. CONCLUSIONS: This review consolidates critical insights into in vitro BSCB model design and performance, providing a clear framework for developing more accurate and reliable models. These findings will facilitate the development of effective therapeutic interventions capable of crossing the BSCB, addressing a key bottleneck in mechanistic research and therapeutic development for treating SCI and other neurological diseases.