Bihemispheric developmental alterations in basal ganglia volumes following unilateral perinatal stroke

Abstract

Introduction: Perinatal stroke affects millions of children and results in lifelong disability. Two forms prevail: arterial ischemic stroke (AIS), and periventricular venous infarction (PVI). With such focal damage early in life, neural structures may reorganize during development to determine clinical function, particularly in the contralesional hemisphere. Such processes are increasingly understood in the motor system, however, the role of the basal ganglia, a group of subcortical nuclei that are critical to movement, behaviour, and learning, remain relatively unexplored. Perinatal strokes that directly damage the basal ganglia have been associated with worse motor outcomes, but how developmental plasticity affects bilateral basal ganglia structure is unknown. We hypothesized that children with perinatal stroke have alterations in bilateral basal ganglia volumes, the degree of which correlates with clinical motor function.

Methods: Children with AIS or PVI, and controls, aged 6-19 years, were recruited from a population-based cohort. MRIs were acquired on a 3 T GE MR750w scanner. High-resolution T1-weighted images (166 slices, 1 mm isotropic voxels) underwent manual segmentations of bilateral caudate and putamen. Extracted volumes were corrected for total intracranial volume. A structure volume ratio quantified hemispheric asymmetry of caudate and putamen (non-dominant/dominant hemisphere structure volume) with ratios closer to 1 reflecting a greater degree of symmetry between structures. Participants were additionally dichotomized by volume ratios into two groups, those with values above the group mean (0.8) and those below. Motor function was assessed using the Assisting Hand Assessment (AHA) and the Box and Blocks test in affected (BBTA) and unaffected (BBTU) hands. Group differences in volumes were explored using Kruskal-Wallis tests, and interhemispheric differences using Wilcoxon. Partial Spearman correlations explored associations between volumes and motor function (factoring out age, and whole-brain white matter volume, a proxy for lesion extent).

Results: In the dominant (non-lesioned) hemisphere, volumes were larger in AIS compared to PVI for both the caudate (p < 0.05) and putamen (p < 0.01) but comparable between stroke groups and controls. Non-dominant (lesioned) hemisphere volumes were larger for controls than AIS for the putamen (p < 0.05), and for the caudate in PVI (p = 0.001). Interhemispheric differences showed greater dominant hemisphere volumes for the putamen in controls (p < 0.01), for both the caudate (p < 0.01) and putamen (p < 0.001) in AIS, and for the caudate (p = 0.01) in PVI. Motor scores did not differ between AIS and PVI thus groups were combined to increase statistical power. Better motor scores were associated with larger non-dominant putamen volumes (BBTA: r = 0.40, p = 0.011), and larger putamen volume ratios (BBTA: r = 0.52, p < 0.001, AHA: r = 0.43, p < 0.01). For those with relatively symmetrical putamen volume ratios (ratio > group mean of 0.8), age was positively correlated with BBTA (r = 0.54, p < 0.01) and BBTU (r = 0.69, p < 0.001). For those with more asymmetrical putamen volume ratios, associations with motor function and age were not seen (BBTA: r = 0.21, p = 0.40, BBTU: r = 0.37, p = 0.13).

Conclusion: Specific perinatal stroke lesions affect different elements of basal ganglia development. PVI primarily affected the caudate, while AIS primarily affected the putamen. Putamen volumes in the lesioned hemisphere are associated with clinical motor function. The basal ganglia should be included in evolving models of developmental plasticity after perinatal stroke.

Keywords: Basal ganglia; Cerebral palsy; Motor function; Pediatric; Perinatal Stroke; Volumetrics.

Fig. 1

3D volume generation from ROI series. A) The caudate was traced slice by slice on the sagittal view, creating a series of ROI slices. B) This series was then used by Horos to compute a 3D volume. C & D) The putamen was segmented similarly but used axial slices for maximal visibility. Structures in each hemisphere were segmented and volumes recorded separately.

Fig. 2

Example of a severely damaged putamen segmentation (left side of the image). A) No apparent evidence of putamen tissue. B) Putamen tissue is successfully segmented. While not evident, navigating through surrounding slices reinforce the selection of this tissue as part of the putamen. C) Putamen tissue and shape clearly visible, although not intact. Dominant hemisphere structure is carefully kept separate from globus pallidus. D) White matter tracts inform the segmentation of both structures in this slice, and all subsequent slices.

Fig. 3

A) Distal middle cerebral artery occlusion: possible damage to primary motor cortex, but basal ganglia structures are spared. B) Occlusion closer to middle cerebral artery: basal ganglia are partially damaged. C) Proximal middle cerebral artery occlusion: M1 and basal ganglia are severely damaged.

Fig. 4

Group differences in relative structure volume for the dominant hemisphere caudate (A) and putamen (B), as well as the non-dominant hemisphere caudate (C) and putamen (D). PVI - periventricular venous infarction, AIS - arterial ischemic stroke, TDC - typically developing control, *p < 0.05, **p < 0.01.

Fig. 5

Box-and-blocks test scores with corresponding relative structure volume corrected for RWM and Age. BBTU scores are plotted against relative caudate and putamen volumes in the dominant hemisphere. BBTA scores are plotted against relative caudate and putamen volumes in the non-dominant hemisphere. RWM - Relative White Matter, BBTA - box-and-blocks test of the affected hand, BBTU - box-and-blocks test of the unaffected hand, *p < 0.05.

Fig. 6

Motor assessments and inter-hemispheric structure volume ratio (non-dominant structure/ dominant structure) corrected for RWM and age. RWM - Relative White Matter, AHA - Assisting Hand Assessment, BBTA - box-and-blocks test of the affected hand, **p < 0.01, ***p < 0.001.

Fig. 7

Motor assessment score plotted against age in subgroups with relatively intact caudate, damaged caudate, intact putamen, or damaged putamen. Groups were defined by structure volume ratio (non-dominant hemisphere structure/dominant hemisphere structure) such that a ratio greater than the mean (0.8) reflected a relatively intact structure, while a ratio less than the mean indicated damage to the structure in the non-dominant hemisphere. BBTA - box-and-blocks test of the affected hand, BBTU - box-and-blocks test of the unaffected hand, *p < 0.05, **p < 0.01, ***p < 0.001.