Does Contralesional Hand Function After Neonatal Stroke Only Depend on Lesion Characteristics?
Background and Purpose—In children having suffered from neonatal arterial ischemic stroke, the relationship between contralesional hand performance and structural changes in brain areas remote from the infarct site was examined.
Methods—Using voxel-based morphometry, we correlated contralesional gross manual dexterity assessed by the box and block test and whole-brain gray and white-matter volume changes on high-resolution magnetic resonance imaging in 37 7-year-old post–neonatal arterial ischemic stroke children. We also compared the volume of the identified structures with magnetic resonance imaging data of 10 typically developing age-matched children.
Results—Areas showing the highest positive correlation with the box and block test scores were ipsilesional mediodorsal thalamus, contralesional cerebellar lobule VIIa Crus I, and ipsilesional corticospinal tract at the level of superior corona radiata, the posterior limb of the internal capsule, and the cerebral peduncle and the ipsilesional body of corpus callosum. When compared with typically developing age-matched children, post–neonatal arterial ischemic stroke children with severe contralesional hand motor deficit exhibited significant volume reductions in these structures (except the cerebellum), whereas no differences were found with those with good manual dexterity. No negative correlation was found between box and block test scores and brain areas.
Conclusions—Contralesional hand performance after neonatal arterial ischemic stroke is correlated with atrophy in brain areas directly or functionally connected but anatomically remote from the infarct. Our study suggests a role of the cerebellar lobule VIIa Crus I and mediodorsal thalamus in manual dexterity.
Long-term motor deficit after neonatal arterial ischemic stroke (NAIS) is variable and clearly related to the extent of the lesion.1,2 However, most neuroanatomical correlates of motor outcomes after NAIS were derived from lesion category-based studies,2,3 which by definition only analyzed infarct characteristics. As motor performance after brain lesions4,5 cannot be fully explained by infarct characteristics alone, these aforementioned methods may be insufficient to identify reliably all potential brain substrates of motor performance.
The objective of this study was to explore in a voxel-based morphometry analysis which brain structures remote from the site of the NAIS infarct did correlate with contralesional hand motor performance. In addition, by comparing these structures with magnetic resonance imaging (MRI) data of typically developing children, we further attempted to disentangle whether the correlation was mainly because of either atrophy of remote areas in children with poor manual dexterity or to compensatory mechanisms in those children with good manual dexterity or both.
Samples Characteristic and Outcomes
Detailed sample characteristics and outcomes are presented in Table I in the online-only Data Supplement.
We studied 37 children (23 boys; mean age, 87.05 months; SD, 2.29 m) having had unilateral NAIS of the middle cerebral artery territory. The individuals involved in this work were the same as in a previous study2 except for 1 who refused to perform the box and block test (BBT).
Motor performance of the contralesional hand was assessed with the BBT.6 Each individual score was defined as the maximum number of cubes transferred by the contralesional hand from one compartment to the other in 1 minute. The median BBT score was 28 cubes per minute (95% confidence interval, 25–31). Nine children (8 boys, 2 children with right-sided infarct; mean age, 87.78 months; SD, 3.032) exhibited an efficient manual dexterity with a BBT score >75th percentile and constituted the high BBT group. Eight children (5 boys, 3 with right-sided infarct; mean age, 86.75 months; SD, 2.435) showed severe contralesional hand motor deficit with a BBT score <25th percentile (low BBT group).
Typically developing children underwent the same MRI protocol. This control group included 10 age- and sex-matched children with typical development and normal neurological examination (7 boys; mean age, 86.50 months; SD, 2.635).
Written informed consent was obtained from typically developing children/patients/parents, and the study protocol was approved by the University of Angers Ethical Review Committee.
MRI Acquisition and Processing
Parameters of 3D-T1 MRI acquisition and flipping of the right-sided lesions (n=15) to the left hemisphere were reported before.2
Voxel-based morphometry was conducted using standard functionalities of the VBM8 toolbox (http://dbm.neuro.uni-jena.de/VBM8/) implemented in SPM8 (http://www.fil.ion.ucl.ac.uk/spm) with customized Tissue Probability Maps and template created from the study sample using the Template-O-Matic Toolbox (https://irc.cchmc.org/software/tom.php). A modulation of the segmented and normalized gray matter and white matter was undertaken using default parameters (nonlinear effects only) to analyze volume changes corrected for brain size.
A normalized group lesion map was used to exclude the lesion area from the analysis, according to the objective of this study.
In a first step, for each post-NAIS participant, a voxel-based morphometric correlation analysis was done separately for gray-matter and white-matter maps with the BBT score using multiple regression analysis with lesion volume2 (median, 41.12 mL [95% confidence interval, 10.96–62.98]), sex, hand dominance, head circumference (median, 51.50 cm [95% confidence interval, 50.50–52.00]), and age as nuisance variables. Therefore, a positive correlation found in 1 particular region indicated a higher local volume correlating with better manual dexterity of the contralesional hand, not explained by lesion volume, sex, age, handedness, or brain size. All statistical parametric maps were interpreted with a threshold of P<0.05 corrected for multiple comparisons (false discovery rate) and minimum cluster size of 10 contiguous voxels.
In a second step, volumes of the identified brain structures were compared between the 3 groups of children (high BBT group, low BBT group, and typically developing children) using a Kruskal–Wallis test followed by a post hoc Dunn multiple comparison test, using version 6.01 GraphPad-Prism for Windows (http://www.graphpad.com). The significance level was set at 5%, adjusted for multiple comparisons.
Correlation Between Local Volume and BBT Scores
Gray matter areas showing the highest positive correlation with the BBT scores were the ipsilesional mediodorsal thalamus (MNI [Montreal Neurological Institute] coordinates [−6, −24, 7]; t=6.13) and the contralesional superior posterior lobe of the cerebellum (lobule VIIa Crus I [20, −85, −29]; t=5.29; Figure 1).
White-matter regions whose volume correlated positively with BBT scores were located in the ipsilesional body of corpus callosum ([−17, −16, 31]; t=4.22) and along the ipsilesional corticospinal tract at the level of the superior corona radiata ([−20, −1,4 2]; t=4.64), the posterior limb of internal capsule ([−18, −13, −2]; t=4.69), and the cerebral peduncle ([−12, −16, −17]; t=4.12).
No gray-matter and white-matter regions with negative correlations were found.
Volume Comparison of Identified Brain Structures
Median volumes of the aforementioned clusters were significantly different between the low BBT, high BBT, and control groups (except for the contralesional cerebellum) with significantly lower volumes in the low BBT group compared with both the high BBT and the control groups. No significant differences were found between the high BBT group and the control group (Figure 2). Statistical details are given in Table II in the online-only Data Supplement.
First, the present study confirmed the positive correlation between the volume of ipsilesional cerebral peduncle and hand motor performance previously found in unilateral CP.7 Our results also confirm the relationship between the structural characteristics of the ipsilesional corticospinal tract at different levels (notably the posterior limb of internal capsule) and the upper limb motor outcome after early brain lesions.8
Second, the mediodorsal thalamus is involved in the cerebello-thalamo-cerebral connectivity providing sophisticated interplay between feedforward and feedback mechanisms to generate and control precise, smooth hand movement.9 In addition, the cerebellum is known to contribute to participate in hand function and the lobule VIIa Crus I was activated during complex motor tasks on functional MRI.10 Recently, a significant relationship between Crus I volume and hand motor dexterity was demonstrated.11 The premotor cortex and the primary motor area show connections with the contralateral Crus I via the mediodorsal thalamus.12 Hence, our results would further support the hypothesis of an important combined role for the mediodorsal thalamus and lobule VIIa Crus I in the control of skilled hand movement.10,13
Finally, we show that the correlation with manual dexterity found in brain areas directly or functionally connected but anatomically remote from the infarct is mainly because of atrophy of these structures in impaired children (except for the cerebellum). Atrophy in brain regions distant from the infarct site was also shown in adults after stroke5 and is likely related to Wallerian degeneration, deafferentation, or both.
Together with our previous work,2 the present study provides evidence that motor performance after NAIS is because of not only the characteristic of the infarct itself but also structural loss in ipsilesional remote brain areas being part of the motor network. Our results, thus, suggest a role of the mediodorsal thalamus and cerebellar lobule VIIa Crus I in manual dexterity.
Sources of Funding
This research was supported by the University hospital of Angers (eudract number, 2010-A00976-33), the University hospital of Saint-étienne (eudract number, 2010-A00329-30), and Fondation de l’Avenir (ET0-571).
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.116.013545/-/DC1.
- Received March 23, 2016.
- Revision received April 3, 2016.
- Accepted April 6, 2016.
- © 2016 American Heart Association, Inc.
- Kirton A,
- Deveber G
- Boardman JP,
- Ganesan V,
- Rutherford MA,
- Saunders DE,
- Mercuri E,
- Cowan F
- Gauthier LV,
- Taub E,
- Mark VW,
- Barghi A,
- Uswatte G
- Mathiowetz V,
- Federman S,
- Wiemer D
- Duque J,
- Thonnard JL,
- Vandermeeren Y,
- Sébire G,
- Cosnard G,
- Olivier E
- Schlerf JE,
- Verstynen TD,
- Ivry RB,
- Spencer RM