Abstract 410: Early Imaging Incompletely Predicts Subsequent Cerebral Physiological Derangements After Cardiac Arrest

Abstract

Introduction: Decreased grey-white matter ratio (GWR) on early head CT scans correlates with poor outcome after cardiac arrest. However, it is unknown if early imaging features fully predict of subsequent derangments in cerebral physiology that cause secondary brain injury. Hypothesis: GWR on early head CT correlates with average intracranial pressure (ICP), brain tissue oxygen (PbtO2), and pressure reactivity (a proxy for cerebral autoregulation) after cardiac arrest. Methods: A total of 30 post-arrest patients were identified over a 3-year period who underwent head CT within 1 day of arrest as well as intracranial monitoring (IM) during hospitalization. 5 patients were excluded either due to unavailable IM recordings (n=1) or motion degraded head CTs (n=4). ICP, P bt O 2 , and mean arterial pressure (MAP) were recorded continuously using a bedside recording unit (CNS Monitor, Moberg Research Inc). Pressure reactivity index (PRx), was calculated as the Pearson correlation coefficient between ICP and MAP over 5 minute intervals, updated every minute. GWR was measured for the basal ganglia from Hounsfield unit measurements made from initial head CT. Relationships between average ICP, PRx, P bt O 2 and GWR was calculated using linear regression. Results: Average GWR for the cohort was 1.2154. Average ICP was weakly correlated with initial GWR (R 2 =0.257, p=0.01). A trend towards higher PRx values was seen with lower initial GWR; however, this did not reach statistical significance (R 2 =0.15, p=0.07). There was no relation between average P bt O 2 and GWR (R 2 =0.02, p=0.56). Conclusions: Lower initial GWR is associated with higher average ICP, and may also be associated with disorderd cerebral autoregulation. However, no relationship could be identified between initial imaging findings and average brain oxygen levels. These data imply that physiological stressors such as brain hypoxia are not predicted by initial imaging findings and suggests a role for invasive neuromonitoring to identify and treat processes that promote secondary brain injury.