Functional Connectivity of Hippocampal Circuits and Visual Memory Function in Children and Adolescents With Perinatal Stroke

Abstract

Perinatal stroke is a vascular injury occurring early in life, often resulting in motor deficits (hemiplegic cerebral palsy/HCP). Comorbidities may also include poor neuropsychological outcomes, such as deficits in memory. Previous studies have used resting state functional MRI (fMRI) to demonstrate that functional connectivity (FC) within hippocampal circuits is associated with memory function in typically developing controls (TDC) and in adults after stroke, but this is unexplored in perinatal stroke. Investigating links with visual memory function has the potential to inform prognosis and personalized cognitive rehabilitation strategies. This study aimed to quantify FC within hippocampal circuits of children and adolescents with perinatal stroke and associations with visual memory. We hypothesized that FC would differ between participant groups (AIS, PVI, TDC) and hemispheres (left vs. right stroke), and would correlate with visual memory function. Participants aged 6-19 years with HCP and MRI-confirmed unilateral perinatal stroke (n = 30) arterial ischemic stroke (AIS), n = 38 periventricular venous infarction (PVI) were recruited through the Alberta Perinatal Stroke Project and compared to n = 43 TDC. Resting fMRI volumes (150 volumes, TR/TE = 2000/30 ms, voxels 3.6 mm isotropic, 36 axial slices) were processed to compute FC values between memory-related seeds (including bilateral hippocampi) using a standard pipeline in the CONN toolbox. Seed-to-voxel and seed-to-seed analyses computed FC with each hippocampus. Hemispheric and group differences in FC were examined. A subset of stroke participants (n = 46) completed visual memory testing via CNS Vital Signs (CNSVS), a computerized neurocognitive test battery. Partial correlations assessed associations between FC and visual memory function, factoring out age. We found hemispheric differences in FC within each group. Participants with left perinatal stroke showed greater FC between the hippocampus and lateral prefrontal cortex in the lesioned compared to non-lesioned hemisphere. TDCs had higher hippocampal FC when compared to the lesioned hemisphere of stroke groups. For participants with right hemisphere stroke, associations were observed between hippocampal FC and visual memory function. We describe differences in bilateral hippocampal functional connectivity in children and adolescents with perinatal stroke that are associated with visual memory function. Our findings suggest that developmental plasticity may occur in the non-lesioned hippocampus after perinatal stroke. These findings may inform our understanding of how visual memory function is affected after early unilateral brain injury and facilitate the development of novel therapies for individuals affected by perinatal stroke.

Keywords: functional connectivity; hippocampus; perinatal stroke; visual memory.

FIGURE 1

An illustration of regions of interest (ROIs) from the Harvard‐Oxford Atlas used as seeds for measurements of functional connectivity. Seven ROIs were used in each of the left and right hemispheres (only left shown here). (A) Subcortical and medial ROIs were thalamus, hippocampus, and parahippocampal gyrus. Anterior and posterior divisions of the parahippocampal gyrus were combined into one ROI shown in dark blue. (B) Cortical ROIs were anterior temporal gyrus (combined superior, middle, and inferior gyri, shown in purple), posterior temporal gyrus (combined superior, middle, and inferior gyri shown in medium blue), posterior parietal cortex, and lateral prefrontal cortex.

FIGURE 2

Lesion overlay map illustrating location and overlap of lesions for participants with (A) left hemisphere and (B) right hemisphere arterial stroke lesions. Brighter colors in the color maps represent a greater number of participants with a lesion in that voxel. Shown for comparison are the hippocampus (yellow) and parahippocampal (green) regions of interest. Also shown are the slice numbers in standard space. These images are shown in radiological convention (i.e., left side of image is right hemisphere).

FIGURE 3

(A–C) Group mean seed‐to‐voxel functional connectivity (FC) maps are shown when using the left hippocampus as a seed for the (A) TDC, (B) AIS_Left, and (C) PVI_Left groups. Voxel‐based t‐scores are expressed as heat maps where warm colors (yellow/red) denote positive FC and cool colors (pink/blue) denote negative FC with the left hippocampus seed. AIS—arterial ischemic stroke, PVI—periventricular venous infarction, TDC—typically developing controls, Left—left hemisphere stroke. Images are in neurological convention (right of image is right hemisphere). Threshold: P_FDR < 0.05 (cluster‐size corrected).

FIGURE 4

(A–C) Group mean seed‐to‐voxel functional connectivity (FC) maps are shown when using the right hippocampus as a seed for the (A) TDC, (B) AIS_Right, and (C) PVI_Right groups. Voxel‐based t‐scores are expressed as heat maps where warm colors (yellow/red) denote positive FC and cool colors (pink/blue) denote negative FC with the right hippocampus seed. Threshold: P_FDR < 0.05 (cluster‐size corrected). D. A whole brain group contrast (TDC > AIS_Right) where cool colors illustrate voxels where FC between the right hippocampus and right precuneus was greater in AIS_Right compared to TDC and where warm colors illustrate voxels in the medial frontal lobe where FC was greater in TDC compared to AIS_Right. Threshold: P_FDR < 0.05 (cluster‐mass corrected). AIS—arterial ischemic stroke, PVI—periventricular venous infarction, TDC—typically developing controls. Images are in neurological convention (right of image is right hemisphere).

FIGURE 5

Between hemisphere comparisons of functional connectivity between regions by participant groups. Shown is FC between the hippocampus and (A) lateral prefrontal cortex (B) posterior parietal cortex. Symbols and vertical lines denote mean and standard error. Oblique lines denote FC asymmetry between hemispheres, **p < 0.01, ***p < 0.001. AIS—arterial ischemic stroke, PVI—periventricular venous infarction, TDC—typically developing controls. Note that TDCs were included in all plots for comparison, though they had no lesion damage, and since they were all right‐handed, the left hemisphere could be considered “dominant.”

FIGURE 6

Functional connectivity between left and right hippocampus for (A) Participants with left hemisphere stroke (and controls) and (B) participants with right hemisphere stroke (and controls). Horizontal bars denote statistical significance between groups, *p < 0.05, **p < 0.01.

FIGURE 7

Associations between functional connectivity (FC) and visual memory function in participants with right hemisphere stroke. (A) FC between left hippocampus and left posterior parietal cortex compared with visual memory function (correct hits immediate). (B) FC between right hippocampus and right posterior parietal cortex compared with visual memory function (correct hits delayed). r_s—Spearman's rho, AIS—arterial ischemic stroke, PVI—periventricular venous infarction, shaded areas denote 95% confidence intervals.