The Usefulness of Quantitative Analysis of Blood-Brain Barrier Disruption Measured Using Contrast-Enhanced Magnetic Resonance Imaging to Predict Neurological Prognosis in Out-of-Hospital Cardiac Arrest Survivors: A Preliminary Study

Abstract

We aimed to evaluate neurological outcomes associated with blood-brain barrier (BBB) disruption using contrast-enhanced magnetic resonance imaging (CE-MRI) in out-of-hospital cardiac arrest (OHCA) survivors. This retrospective observational study involved OHCA survivors who had undergone CE-MRI for prognostication. Qualitative and quantitative analyses were performed using the presence of BBB disruption (pBD) and the BBB disruption score (sBD) in CE-MRI scans, respectively. For the sBD, 1 point was assigned for each area of BBB disruption, and 6 points were assigned when an absence of intracranial blood flow due to severe brain oedema was confirmed. The primary outcome was poor neurological outcome at 3 months (defined as cerebral performance categories 3-5). We analysed 46 CE-MRI brain scans (27 patients). Of these, 15 (55.6%) patients had poor neurological outcomes. Poor neurological outcome group patients showed a significantly higher proportion of pBD than those in the good neurological outcome group (22 (88%) vs. 6 (28.6%) patients, respectively, p < 0.001) and a higher sBD (5.0 (4.0-5.0) vs. 0.0 (0.0-1.0) patients, p < 0.001). Poor neurological outcome predictions showed that the sBD had a significantly better prognostic performance (area under the curve (AUC) 0.95, 95% confidence interval (CI) 0.84-0.99) than the pBD (AUC 0.80, 95% CI 0.65-0.90). The sBD cut-off value was >1 point (sensitivity, 96.0%; specificity, 81.0%). The sBD is a highly predictive and sensitive marker of 3-month poor neurological outcome in OHCA survivors. Multicentre prospective studies are required to determine the generalisability of these results.

Keywords: heart arrest; magnetic resonance imaging; prognosis.

Figure 1

The blood-brain barrier disruption score using a post-contrast fluid-attenuated inversion recovery image. (A) Leptomeningeal enhancement in the subarachnoid space is not visible; a score of 0. (B) Leptomeningeal enhancement in the right parietal sulcus is visible; a score of 1. (C) Leptomeningeal enhancement in the left sylvian fissure (lateral sulcus separating the frontal lobe from the temporal lobe) is visible; a score of 2. (D) Leptomeningeal enhancement in the right frontal, parietal and temporal sulci is visible; a score of 3. (E-1) Leptomeningeal enhancement in the left temporo-occipital sulci and, (E-2) left fronto-parietal sulci is visible; a score of 4. (F-1) Leptomeningeal enhancement in the left fronto-parietal sulci, (F-2) left temporo-occipital sulci and, (F-3) cerebellar folia is visible; a score of 5. (G-1) Leptomeningeal enhancement is not visible in the subarachnoid space and, (G-2) the intracranial blood flow is absent in the contrast enhanced T1 weighted MR; a score of 6.

Figure 2

Comparisons between the proportion and the score of blood-brain barrier disruption between the good and the poor neurological outcome groups. (A) In the poor neurological outcome group, the proportions of BBB disruption were significantly higher than those in the good neurological outcome group (p < 0.001). (B) In the poor neurological outcome group, the scores of BBB disruption were significantly higher than those in the good neurological outcome group (p < 0.001).

Figure 3

Association of sBD, pBD and Q_A with poor neurological outcomes. (A) Receiver operating characteristic curve for poor neurological outcome on sBD is shown on the panel. (B) Receiver operating characteristic curve for poor neurological outcome on pBD is shown on the panel. (C) Receiver operating characteristic curve for poor neurological outcome on Q_A is shown on the panel.