Magnetic Resonance Imaging As a Outcome Predictor in Comatose Survivors after Cardiac Arrest

Chun, Young Ho; Park, Kyu Nam; Choi, Seung Pil; Kim, Young Min; Lim, Yun Soo; Kim, Se Kyung

Abstract

PURPOSE: The prognosis of comatose survivors after cardiac arrest is predicted by various methods such as neuralagic examination, magnetic resonance spectroscopy (MRS), biochemical markers, and somatosensory-evoked potentials (SSEP). The objective of this study was to investigate differences in magnetic resonance imaging (MRI) findings between good (CPC 1-2) and poor (CPC 3-5) outcome for comatose survivors after cardiac arrest.
METHODS
We retrospectively studied the cases of 19 comatose survivors after cardiac arrest who underwent MRI between March 1, 1997 and August 31, 1999. All MRI readings were performed by a neuro radiologist blindly and best CPC during 6 months after ROSC was used as final outcome.
RESULTS
Sixteen of the 19 patients had abnormal findings on the T2-weighted images: hippocampal infarct in 13 (68%), basal ganglia/thalamic infarct in 9 (47%), cortical infarct in 7 (37%), cortical watershed infarct in 6 (32%), and cerebellar infarct in 5 (26%). Among 6 patients with good outcome (CPC 1-2), 4 patients (67%) had hippocampal infarct, 1 (17%) basal ganglia/thalamic infarct, 1 (17%) white matter infarct, and 1 (17%) cerebellar infarct. Among the 13 patients with poor outcomes (CPC 3-5), 9 patients (69%) had hippocampal infarct, 8 (62%) basal ganglia/thalamic infarct, 7 (54%) cortical infart, 6 (46%) cortical watershed infarct, and 4 (31%) cerebellar infarct. Cortical and cortical watershed infarcts were only detected in patients with poor outcomes (p<0.05).
CONCLUSION
On the MRI findings, comatose survivors after cardiac arrest were susceptible to hippocampus, basal ganglia/thalamus, cortex, cortical watershed, and cerebellar infarts. Cortical and cortical watershed infarcts may be useful as a prognostic tool for comatose survivor after cardiac arrest.