Separate functional and structural cerebral mechanisms relate to postanoxic coma recovery
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background and objectives
Understanding postanoxic encephalopathy following cardiac arrest is important to provide optimal treatment and predict chances of recovery. Both functional cerebral disturbances (changes in electrophysiological activity or functional connectivity) and structural tissue changes (signs of cytotoxic or vasogenic oedema) have been associated with neurological outcome. We hypothesize that functional and structural changes follow separate pathways towards persistent coma or recovery after a cardiac arrest. We aim to identify common and separate pathways between structural and functional changes using an advanced integration method.
Methods
We performed a prospective multicentre cohort study in comatose cardiac arrest patients in three Dutch hospitals. T1-weighted, diffusion-weighted, and resting-state functional MRI scans were collected on day 2-9 after cardiac arrest. Primary outcome measure was neurological outcome at six months defined by the Cerebral Performance Categories (CPC), dichotomized as CPC 1-2 = good, CPC 3-5 = poor. Maps of grey matter volume, Jacobian deformation, mean diffusivity, skeletonized fractional anisotropy, and eight resting-state networks were used as input for a linked independent component analysis. Group differences in subject loadings on the components between individuals with good and poor outcome were calculated. Relationships between functional and structural components were studied descriptively.
Results
We included 80 patients (31 with poor outcome). Of twenty identified components, six consisted of functional, seven of structural, and five of both structural and functional information. Four components related independently to outcome: two functional and two structural components. None of the components containing both functional and structural information related to outcome. Patients with good outcome showed higher connectivity in multiple resting-state networks (mainly visual, default-mode, and frontoparietal network) and smaller volumes of and higher diffusivity in the cortex and deep grey nuclei. Signs of functional injury were seen in both outcome groups, while only patients with poor outcome showed signs of structural damage. Forty-three patients showed functional without structural, but no showed structural without functional injury.
Discussion
Functional network and structural tissue injury exist independently in postanoxic encephalopathy after cardiac arrest. Our results point towards distinct cerebral mechanisms relating to postanoxic coma outcome, likely including cell swelling, tissue edema, and isolated synaptic failure.
Clinical trial registration
Clinicaltrials.gov NCT03308305
