Ipsilateral Corticospinal Tract Excitability Contributes to the Severity of Mirror Movements in Unilateral Cerebral Palsy: A Case Series

Introduction

Rehabilitation interventions in children with unilateral cerebral palsy (UCP) focus on maximizing participation in activities, often involving the upper limb. Many daily living skills require bimanual coordination. This type of coordination primarily develops through early elementary school ages and is associated with maturation of the corticospinal tract (CST). ¹ However, bimanual coordination can be limited due to mirror movements (MM) or involuntary movements of one hand when attempting to isolate movements of the other hand. ² MM are commonly observed in children with UCP and known to have a negative impact on bimanual daily function and rehabilitation outcomes. ²

The persistence of MM throughout childhood and adolescence indicates the continued dysfunction of the corticospinal system and its role in regulating unimanual and bimanual hand control. The CST initially develops with contralateral and ipsilateral projections from each hemisphere to each hand. ³ Typically, the ipsilateral projections weaken over time and the CST maintains primarily contralateral projections from each hemisphere to the opposite hand. ³ However, in children with UCP, CST development may be altered in both hemispheres. In these cases, the contralesional hemisphere maintains projections to both hands (eg, testing the contralesional hemisphere elicits a motor-evoked potential (MEP) from both the hand on the same side as stimulation through ipsilateral projections and the hand opposite stimulation through contralateral projections). Altered CST development, such as maintained ipsilateral projections, is thought to contribute to the presence of MM. ⁴

Clinically, MM can be observed in both hands during functional activities. Prior work examining MM found the etiology differed for the more-affected (ie, weaker) and the less-affected (ie, stronger) hands. The MM of the more-affected hand were associated with the timing of the lesion where children with lesions that occur early in development (ie, periventricular lesions) displayed greater MM in the more-affected hand than children with lesions that occurred after birth (ie, cortico-subcortical lesions and acquired injuries). ⁵ The authors reported no association of MM in the more-affected hand with functional skills as measured by motor tests. In contrast, MM of the less-affected hand were associated with deficits in global hand function (eg, increased MM of the less-affected hand were associated with poor unimanual and bimanual hand skills). ⁵ This prior work incorporated diagnostic findings based on neuroimaging and provides evidence that lesion timing and global hand deficits contribute to MM, further impairing the functional skills of both hands. Investigation into hand deficits utilizing another approach, neurophysiology using brain stimulation testing, further characterized the contralesional hemisphere of children with UCP. Zewdie et al ⁶ reported baseline data that showed a positive association between the responses to brain stimulation testing and hand function, such that greater excitability of both hemispheres was associated with higher performance on the Box and Blocks. Furthermore, increased excitability (ie, decreased inhibition within the contralesional hemisphere), was associated with improved performance on the Assisting Hand Assessment (AHA). Although this study did not measure MM, these results suggest a strong influence of the contralesional hemisphere on functional hand skills in children with CST dysfunction.

Further investigations into MM have focused on the altered CST development of the contralesional hemisphere in children with UCP. Early work by Vandermeeren et al ⁷ mapped the motor representation of each hand in children with UCP. Their findings suggest that when the contralesional hemisphere displayed representation of both hands, MM of the less-affected hand were observed. ⁷ These findings were validated in a recent study in which Marneweck et al ⁸ characterized children where the contralesional hemisphere maintained the representation of both hands (ie, strong ipsilateral projections). Their data suggest that dual representation (ie, both hands represented in the contralesional hemisphere) is associated with not only better hand function but also increased MM in the less-affected hand. ⁸ These data illustrate the complex relationship between the cortical representation of bimanual hand function and MM.

Collectively, these preliminary insights suggest that active contributions of the contralesional hemisphere to movement of both hands lead to MM in some children with UCP. However, the relationship between CST excitability and the presence and severity of MM has not been fully explored. Here, we report a cross-sectional case series of children and young adults with UCP who underwent assessment of MM to characterize the relationships between CST excitability, circuitry, and clinical presentation (AHA, MM, Canadian Occupational Performance Measure [COPM]). We confirm and expand on the current understanding of the relationship between ipsilateral MEPs elicited with single-pulse transcranial magnetic stimulation (TMS) testing and functional skills.

Methods

The cohort for this study was a subset of 13 children and young adults with UCP with complete baseline neurophysiologic and behavioral testing from a larger sample who previously participated in a combined behavioral and neuromodulatory intervention (ClinicalTrials.gov NCT02250092).^9,10 All parents, young adults, and children provided consent and assent respectively. The Manual Abilities Classification System (MACS) was used to classify each child’s hand function for daily life. ¹¹ Behavioral testing included the AHA ¹² Woods and Teuber ¹³ test of mirror movements, and COPM. The Woods and Teuber was videotaped and scored by 2 investigators (T.R. and C.Y.C.).

Neurophysiologic responses were determined with single-pulse TMS as described in a prior publication. ¹⁴ Briefly, we recorded bilateral electromyography (EMG) in the first dorsal interosseous muscle of the hand to determine CST circuitry and evaluate the elicited MEPs. First, we identified the motor hotspot, or area of greatest excitability, and the resting motor threshold using up to 85% maximum stimulator output (MSO). Then, we collected 10 single-pulse trials using a TMS intensity of 120% of resting motor threshold, up to 85% MSO. Children were classified as having contralateral, bilateral, or ipsilateral CST circuitry based on the EMG responses (at least 3 of 10 trials) to stimulation of each hemisphere. Contralateral circuitry indicated the presence of an MEP in the hand contralateral to stimulation when testing either hemisphere. Bilateral circuitry indicated bilateral MEPs of similar latency when testing the contralesional hemisphere and a contralateral MEP observed when stimulating the lesioned hemisphere. Ipsilateral circuitry indicated an observed MEP in both hands when stimulating the contralesional hemisphere and no MEP elicited in either hand when stimulating the lesioned hemisphere. MEP amplitudes were analyzed using MATLAB (MathWorks, Natick, MA) with methods previously described. ¹⁴ To examine the relationship between cortical excitability of the contralesional hemisphere to MM scores, we calculated a ratio of MEP amplitudes between the more-affected and less-affected hands elicited during contralesional TMS testing. This ratio was equal to 0 for children with contralateral circuitry and was >0 for children with bilateral or ipsilateral circuitry.

To assess MM, we conducted the Woods and Teuber Assessment. We administered the Woods and Teuber as described in the original paper by positioning the child with their forearms resting on the table. ¹³ The Woods and Teuber involves three motions that are conducted with each hand in the order of (1) tapping the index finger to the thumb, (2) rotation of a fist using pronation and supination, and (3) rapidly touching each finger to the thumb. ¹³ The investigator demonstrated each movement and then asked the child to imitate the movement while being videotaped. The order of movements was the same for each hand and among all participants. Two assessors scored the videotapes using the original criteria of “0 = no clearly imitative movement, 1 = barely discernible repetitive movement, 2 = either slight but unsustained repetitive movement or stronger, but briefer, repetitive movement; 3 = strong and sustained repetitive movement; and 4 = movement equal to that expected for the intended hand.” ¹³ The scores range from 0 to 12 with 0 representing no observed MM. The association between clinical (AHA, MM score, and COPM) and neurophysiologic variables (MEP ratio) was analyzed with nonparametric tests (ie, Spearman’s rank order correlation) given the small sample size and nonnormal distributions in the data, as assessed with Shapiro-Wilk tests for normality.

Results

These 13 children and young adults with UCP (mean age 12 years 2 months) had a wide range in hand skills based on the MACS (I (2), II (9), III (2)) and AHA scores ranging from 34 to 83 AHA 0 to 100 logit-based units (mean 60.5, 95% CI 53.6-67.5 AHA units) (Table 1). We observed MM in the more-affected hand in 62% of our sample (8 children) and the less-affected hand in 92% of our sample (12 children). The CST circuitry of our sample included 5 ipsilateral (ie, no ipsilesional MEP could be elicited), 3 bilateral, and 4 contralateral presentations.

OPEN IN VIEWER

Table 1.

Baseline Behavioral Characteristics of Children With UCP Exhibiting Mirror Movements. ^a

ID	Age	Affected Side	MACS	MM-MA Hand	MM-LA Hand	AHA 0-100, Logit-Based Units	COPM-Perf
1	10 y 1 mo	Right	II	6	11	41	1.4
2	8 y 11 mo	Right	II	5	8	64	5.4
3	13 y 5 mo	Right	III	1	2	66	3.8
4	10 y 6 mo	Right	II	0	4	67	2.75
5	8 y 1 mo	Right	II	1	5	66	2.25
6	16 y 10 mo	Right	II	2	4	59	4.75
7	14 y 7 mo	Right	II	3	6	58	2.5
8	15 y 4 mo	Right	I	0	4	57	5.0
9	7 y 4 mo	Left	II	9	12	62	4.0
10	14 y 9 mo	Right	I	0	0	83	3.7
11	19 y 8 mo	Left	II	1	2	75	1.0
12	12 y 4 mo	Left	II	0	1	55	5.0
13	8 y 6 mo	Right	III	0	6	34	3.35

Abbreviations AHA, Assisting Hand Assessment; COPM, Canadian Occupational Performance Measure; LA, less-affected; MA, more-affected; MACS, Manual Abilities Classification System; MM, mirror movements; Perf, performance subtest of the COPM; UCP, unilateral cerebral palsy.

a

Mirror movements were assessed with the Woods and Teuber test of mirror movements. The COPM was child rated.

Clinically, there was no associations between the AHA and MM of the less-affected hand (r = −0.50, P = .08) or with MM of the more-affected hand (r = −0.16, P = .59). Similarly, there were no observed relationships between the self-rating of performance as rated by the COPM and MM of either hand (less-affected: r = −0.11, P = .72; more-affected: r = −0.03, P = .93).

Neurophysiologically, the MEP ratio ranged from 0 to 1.45 (median 0.11) with greater MM observed in children with ratios greater than 0.5 (Table 2). We observed a statistically significant positive relationship between the MEP ratio and the more-affected MM score in our sample of 13 children (r = 0.69, P = .001) (Figure 1). There was no statistically significant relationship between MEP ratio and the less-affected MM score in our sample (r = 0.40, P = .18).

OPEN IN VIEWER

Table 2.

Baseline Imaging and Neurophysiologic Characteristics of Children With UCP Exhibiting Mirror Movements.

ID	Lesioned Hemisphere	Affected Side	Imaging Findings	Krägeloh-Mann Lesion Classification a	CST Circuitry	Lesioned Hemisphere RMT (%MSO)	Contralesional Hemisphere RMT (%MSO)	MEP Ratio
1	Left	Right	Encephalomalacia of whole left MCA territory front-parietal cortex + subcortical white + gray matter+ BG	White and gray matter lesions	Ipsilateral	N/A	55	0.86
2	Left	Right	Moderate to severe porencephalic dilation of left lateral ventricle. Some gliosis and atrophy of the left thalamus + posterior left lentiform nucleus. Mild dilation of right ventricle	White matter lesions	Ipsilateral	N/A	71	0.58
3	Left	Right	Large porencephalic cyst involving the left lateral ventricle with associated thinning of the adjacent cortex and corpus callous. Posterior fossa findings consistent with Dandy-Walker variant	Gray matter lesions	Bilateral	65	57	0.31
4	Left	Right	Left large areas of multicystic encephalomalacia in the left parietal lobe and posterior MCA distribution, thinned corpus callosum	Gray matter lesions	Ipsilateral	Not Est	45	0.09
5	Left	Right	Left MCA distribution, left thalamus, and left temporal lobe	Gray matter lesions	Contralateral	53	43	0
6	Left	Right	Left cerebral encephalomalacia in the left MCA distribution	Gray matter lesions	Ipsilateral	N/A	50	0.04
7	Left	Right	Extensive left MCA infarction	Gray matter lesions	Contralateral	49	43	0.12
8	Left	Right	Injury to the left BG and mild diffuse parenchymal volume loss	Gray matter lesions	Contralateral	50	45	0
9	Right	Left	White matter changes adjacent to the body of the right lateral ventricle	White matter lesions	Ipsilateral	N/A	70	1.45
10	Left	Right	Ex vacuo dilatation of the body of the left lateral ventricle with associated small zone of T2 and FLAIR signal hyperintensity	White matter lesions	Contralateral	44	41	0
11	Right	Left	Cystic encephalomalacia extending from the frontal pole level of the right lateral ventricle through the white matter to involve the cortex at the posterior right frontal lobe level with cephalad extension subsequently almost to the vertex. Mild enlargement of the right lateral ventricle	White and gray matter lesions	Bilateral	44	38	0.20
12	Right	Left	Right thalamus + periventricular white matter + cortical thinning	White and gray matter lesions	Bilateral	75	54	0.36
13	Left	Right	Asymmetric atrophy involving the left frontal lobe with extension to a portion of the left posterior parietal lobe. Ex vacuo dilatation of the left horn Atrophy and encephalomalacia involving the left subinsular, caudate, and lentiform nucleus regions. Cystic encephalomalacia at the left BG extending to the hypothalamic region	Gray matter lesions	Contralateral	77	48	0

Abbreviations: BG, basal ganglia; CST, corticospinal tract; FLAIR, fluid-attenuated inversion recovery; LA, less-affected; MA, more-affected; MCA, middle cerebral artery; UCP, unilateral cerebral palsy; Not Est, not established due to testing fatigue; %MSO, percentage of maximum stimulator output.

a

Krägeloh-Mann lesion classification as described in Krägeloh-Mann et al (2004). ¹⁵

OPEN IN VIEWER

Figure 1.

(A) Scatterplot showing relationship between MM score and MEP ratio. MEP ratios greater than 1 indicates that ipsilateral MEP responses were larger on average than contralateral MEP responses during contralesional TMS testing. (B) Scatterplot showing relationship between MM score and MEP ratio by lesion location according to the Krägeloh-Mann lesion classification (2004). CST, corticospinal tract; GM, gray matter; MEP, motor-evoked potential; MM, mirror movements; WM, white matter.

Discussion

In our cross-sectional sample of children, we observed that those with larger amplitude MEPs elicited from the more-affected hand relative to MEPs elicited from less-affected hand experienced greater severity in MM. Therefore, greater relative strength in excitability of ipsilateral CST projections relates to greater intensity of MM. This evidence supports other investigations by demonstrating the presence of more-affected hand motor representations in the contralesional hemisphere and their link to the presence of MM. ⁸

Similar to others, we found a higher presence of MM of the less-affected hand as compared to the more-affected hand. ¹⁶ Others described a positive relationship between the presence of MM and a negative impact on hand function as measured by COPM and AHA.^2,5,17 Our small sample size may have limited our ability to detect similar relationships.

A recent study reported improvements in bimanual goal-directed tasks following bimanual training but no reduction in MM. ¹⁸ This research supports that behavioral interventions alone may not be sufficient to address MM, particularly in children with ipsilateral CST projections. Examining CST excitability combined with behavioral performance of functional skills may contribute to identification of therapeutic targets for neuromodulatory interventions (eg, noninvasive brain stimulation) and lead to effective, individualized treatment approaches that induce long-lasting neuroplastic changes. For example, bimanual motor training may be more effective if combined with neuromodulation targeting the excitability of ipsilateral projections originating from the contralesional hemisphere. However, such targeted and individualized therapy approaches have yet to be directly tested.

Our results are limited by the study design and sample size. Our study design included children and young adults at varying stages of maturation given our wide age range (ie, 7-19 years of age) with diverse lesions. Furthermore, this subset of data is reported at baseline only so we cannot state the relationship of MM and the potential for change following intervention. Our small sample prevents further statistical analyses such as the relationships between MM and CST circuitry, age, lesion location, or lesion type.

We chose to evaluate MM using the Woods and Teuber assessment as this method could be easily implemented into clinical observations during rehabilitation evaluations. EMG-based methodology could also be implemented to standardize assessment of MM, although limitations exist with integrating EMG into the clinical environment. One study confirmed that clinical assessments of MM are strongly correlated with EMG-based methodology. ¹⁸ EMG-based methodologies could provide visual feedback to the child on the level of MM and categorization of MM as suggested by others. ¹⁹ Providing this additional description in reporting research findings would allow for a rich description of the clinical presentation of the studied population. Clinically, the presence of MM is described as present or absent with a subjective clarifying statement designating mild, moderate, or severe. Categorizing children with MM could assist with determining what aspects of motor-based interventions are most effective for children with MM (e.g. focus on object stabilization and bimanual hand function). Future studies with larger samples comparing the measurement methodology for MM would provide psychometric support to guide clinical practice assessment guidelines.

Conclusions

Our results advance the understanding of the neurophysiologic basis for MM by combining TMS cortical excitability assessments with feasible clinical testing. MM continue to be difficult to evaluate objectively in the clinic and the impact of MM on optimal treatment selection is not understood. Our data indicate that the CST from the contralesional hemisphere maintains ipsilateral projections to the more-affected hand in some children with observable MM. Neuromodulatory interventions that target the contralesional hemisphere, paired with motor training, have the potential to influence contralesional excitability and improve bimanual motor function.