New MRI technology could predict outcome after cardiac arrest

Brain damage with severe cognitive and functional impairment are common after cardiac arrest, but far from universal. Even after thorough testing, most comatose patients receive an indeterminate neurologic prognosis post-arrest.

Physicians faced with this uncertainty have a difficult task in counseling family members trying to make decisions to maximize their loved one's chances for recovery -- without that recovery leading to a disabled state a patient would never have wanted.

Unfortunately, bedside neurologic exams, electroencephalogram (EEG), and conventional brain imaging (CT/MRI) all are poor at predicting neurologic status after cardiac arrest.

A new study suggests that diffusion tensor imaging (DTI) -- advanced processing of MRI images using specialized software -- might reliably identify permanent brain injury and improve prediction of poor outcomes, thus helping families make decisions.

European investigators retrospectively analyzed MRIs of 150 cardiac arrest patients (derivation cohort) who had been comatose for at least 7 days, of whom 22% had a favorable neurologic outcome (able to function independently). Applying their algorithm to 50 comatose cardiac arrest patients in a validation cohort, severe abnormalities in DTI predicted a poor outcome with 90% sensitivity and 100% specificity--far better than any other known method.

DTI measured whole-brain fractional anisotropy, which in this application represented white-matter tract damage.

The longer a patient is comatose after cardiac arrest, generally speaking, the worse her prognosis. However, a minority of patients with prolonged coma will recover fair or good neurologic function. Early termination of life support for such patients who were destined to improve creates the "self-fulfilling prophecy bias" that corrupts most studies about outcomes after cardiac arrest.

The use of this MRI modality shows promise as a potentially valuable tool to more accurately predict poor outcomes after cardiac arrest, and ease the agony of indecision families experience after a loved one's anoxic brain injury. As it would not require hospitals to purchase additional capital equipment, hire new physicians, etc., if further validated, it could be widely deployed relatively easily.

Source: Lancet Neurology