Altered Resting-State Connectivity within Executive Networks after Aneurysmal Subarachnoid Hemorrhage

Abstract

Aneurysmal subarachnoid hemorrhage (aSAH) is associated with significant mortality rates, and most survivors experience significant cognitive deficits across multiple domains, including executive function. It is critical to determine the neural basis for executive deficits in aSAH, in order to better understand and improve patient outcomes. This study is the first examination of resting-state functional Magnetic Resonance Imaging in a group of aSAH patients, used to characterize changes in functional connectivity of the frontoparietal network. We scanned 14 aSAH patients and 14 healthy controls, and divided patients into "impaired" and "unimpaired" groups based on a composite executive function score. Impaired patients exhibited significantly lower quality of life and neuropsychological impairment relative to controls, across multiple domains. Seed-based functional connectivity analysis demonstrated that unimpaired patients were not significantly different from controls, but impaired patients had increased frontoparietal connectivity. Patients evidenced increased frontoparietal connectivity as a function of decreased executive function and decreased mood (i.e. quality of life). In addition, T1 morphometric analysis demonstrated that these changes are not attributable to local cortical atrophy among aSAH patients. These results establish significant, reliable changes in the endogenous brain dynamics of aSAH patients, that are related to cognitive and mood outcomes.

Fig 1. Median test scores for patient and control groups, on domains of (a) the Neuropsychological Impairment Scale (NIS), (b) the Stroke-Specific Quality of Life Scale (SS-QoL), and (c) different measures of the Trail-Making test and Digit Span.

The error bars indicate the range within each group. A ‘*’ indicates significant difference between groups after False Discovery Rate correction.

Fig 2. Median test scores for impaired patients and control groups on sub-scales of the Neuropsychological Impairment Scale (NIS) and Stroke-Specific Quality of Life Scale (SS-QoL), for which impaired patients scored significantly worse than controls after False Discovery Rate correction; error bars indicate the range for each group.

Fig 3. Median test scores for unimpaired patient, impaired patient and control groups, shown for cognitive tests where the impaired patients scored significantly worse than both the unimpaired patients and the healthy controls, after False Discovery Rate correction.

Error bars represent the range for each group.

Fig 4. Brain regions showing significant differences in seed-based functional connectivity maps between control and patient populations.

Results are shown for seeds in left and right BA46 (dorsolateral prefrontal cortex). Maps are based on pairwise comparisons between controls, aSAH patients, and patient subgroups (“impaired” and “unimpaired”, based on CEFS scores), using voxel-wise 2-sample t-tests (p < 0.05), with cluster-size correction (α = 0.05 significance level).

Fig 5. Brain regions showing significant covariation with cognitive and self-report measures, within the aSAH patient group.

Results are shown for seeds in Left and Right BA46 (dorsolateral prefrontal cortex), examining the Composite Executive Function Scores (CEFS), and the “Mood” subscale of the Quality of Life (SS-QoL) scale. All significant regions exhibit negative correlation between connectivity and behavioural measures (p < 0.05), with cluster-size correction (α = 0.05 significance level).

Fig 6. Brain regions showing significant volumetric change, when comparing (left) Patients to Controls, and (right) Impaired to Unimpaired patients.

Significant regions were identified via Bootstrapped estimates of the mean difference between groups, with multiple comparison correction at False Discovery Rate 0.05.