Abstract Body: Aim: This study investigated the relationship between quantitative assessments of gadolinium leakage in the cerebrospinal fluid (CSF) space, as seen on contrast-enhanced magnetic resonance imaging (CE-MRI), and blood-brain barrier (BBB) permeability, as measured directly using a CSF sample, in patients undergoing post-resuscitation care after cardiac arrest.
Methods: This observational study, which used a dataset of prospectively collected cardiac arrest registry, analyzed data from patients who underwent CE-MRI and CSF sampling between January 2019 and December 2023. Quantitative imaging biomarkers representing gadolinium leakage (ΔCon) were derived using advanced image processing techniques. BBB permeability was assessed using the albumin quotient (Qa = CSF albumin/serum albumin). Statistical analyses included generalised estimating equations and generalised linear models to evaluate the relationship between imaging biomarkers and Qa across cardiac arrest severity (i.e. Pittsburgh Cardiac Arrest Category [PCAC] severity groups).
Results: Among the 43 patients (PCAC 2, 3 and 4; 11, 18 and 14 respectively), ΔCon was a statistically significant predictor of Qa in the total cohort (β = 7.102, p = 0.028, pseudo R² = 0.118). Subgroup analysis divided into PCAC groups revealed that ΔCon exhibited the strongest predictive performance in PCAC 3 patients (β = 3.297, p = 0.011, Pseudo R² = 0.351). No significant predictors were identified in the PCAC 2 or 4 groups. No other serum biomarkers reached statistical significance with Qa.
Conclusion: ΔCon, as measured by biotechnological imaging analysis based on CE-MRI, is a feasible and reliable, non-invasive surrogate for evaluating BBB permeability in patients undergoing post-resuscitation care. ΔCon demonstrated robust predictive capacity, particularly in cases of intermediate illness severity (PCAC 3), supporting its potential clinical utility in prognostication and therapeutic decision-making.