Magnetic resonance imaging findings associated with cardiac arrest.

Abstract

The frequency and prognostic significance of neuroradiological findings after cardiac arrest are unknown. Using healthy volunteers as control subjects, we studied the magnetic resonance imaging (MRI) findings associated with cardiac arrest, adjusted for confounding factors. The presence of cerebral infarcts, leukoaraiosis, atrophy, and edema on ultra-low-field MRI was assessed in 88 community volunteers and 52 cardiac arrest survivors enrolled in a placebo-controlled, randomized, double-blind trial of nimodipine in out-of-hospital ventricular fibrillation. Cardiac arrest was an independent risk factor for the presence of infarcts in a logistic regression model adjusted for age, sex, and history of myocardial infarction, stroke, coronary heart disease, cardiac failure, and hypertension (odds ratio, 3.6; 95% confidence interval, 1.3 to 9.9; P = .01). Leukoaraiosis was associated with increasing age but not with cardiac arrest. Adjusted for age, the delay of advanced life support had an inverse correlation with the degree of atrophy in placebo-treated patients (r = -.62, P < .0001) but not in patients treated with nimodipine (r = -.10, P = .43). Lack of age-related atrophy, possibly implicating the presence of brain edema, predicted poor outcome after cardiac arrest (odds ratio, 4.6; 95% confidence interval, 1.4 to 15.8; P = .01). Cardiac arrest was associated with deep cerebral infarcts but not with leukoaraiosis. MRI findings did not predict the functional outcome at 1 year. Nimodipine treatment had no significant effect on the MRI findings, but delayed resuscitation was associated with probable brain edema only in placebo-treated patients.