Cervical spondylotic myelopathy (CSM) is one of the most common causes of spinal cord dysfunction in older individuals [
Evidence in the medical literature suggests that the improvement of motor function after surgical decompression in CSM patients may occur via synaptic changes and dendritic sprouting in the cortical and spinal cord neuron pools [
Without medical intervention, the natural recovery process following spinal cord compression is slow and largely depends on the extent of the injury sustained [
An emerging modality used to study functional organization in the human motor cortex is TMS [
With ethics committee (Singapore General Hospital ethics committee) approval, patients presenting with clinical features of CSM of at least 6 months’ duration who were listed for spinal cord decompression surgery were recruited with informed consent obtained. We excluded patients with suspected traumatic spinal injury, or any underlying medical or neurological condition which may confound electrophysiological findings. MRI of the cervical spine was performed in all patients within 1 month before surgery. No physiotherapy sessions were scheduled for these patients after surgery. Every recruited patient underwent TMS and motor function testing 1 month prior to and 4 months after surgery. The operation is usually anterior laminectomy of the cervical spine, or any additional procedure stabilization. We also recruited healthy controls for comparison.
TMS mapping of the left hemisphere was performed using a Medtronic (Medtronic Corporation, USA) figure-of-eight-shaped C-B60 coil with 7 cm internal diameter connected to a Medtronic R8 unit generating a peak magnetic field of 2.2 Tesla. The coil was placed tangentially over the skull with the handle pointing backwards and perpendicular to the direction of the central sulcus at approximately 45 degrees to the midline to evoke an anteromedially directed current in the brain.
The vertex, designated as intersection of the interaural line and the nasion-inion connection, was used as an anatomical landmark for finding the optimal position (hotspot) for eliciting motor-evoked potentials (MEPs) from the right first dorsal interosseous (FDI). This is defined as the position with the lowest stimulation intensity needed to elicit an MEP. At the hotspot, the resting motor threshold (rMT) is determined as the position where the lowest TMS intensity will elicit an MEP at a vertical gain of 50
TMS parameters obtained were the sum amplitude of MEPs (sMEP) of the entire 25-point grid and number of positions (
To better ascertain if corticospinal excitability changes occur at the spinal or supraspinal levels,
Central motor conduction times (CMCT) were also obtained from both upper and lower limbs in all patients before and after surgery. CMCT methodology was in accordance with previously published studies by the same authors [
Apart from clinical history and physical examination, each patient’s motor function was quantitatively assessed using Modified Japanese Orthopaedic Association Score Scale (mJOAS) [
As CSM can result in exclusively upper limb or lower limb complaints as well as mixed upper and lower limb features, we separated patients into two groups. In Group A, all had mixed upper and lower limbs features, but patients in Group B had features exclusive to the lower limbs, in line with the mJOAS described above. None of the patients experienced sphincter disturbances.
Statistical calculations were made using SPSS for Windows software. The Wilcoxon Signed-Rank test was used to compare means and Spearman correlation coefficient was employed to examine the relation between MEP characteristics and functional changes in patients after surgery. A
All 24 patients (16 males, 8 females, mean age ± SD: 58.2 ± 11.5) were right handed as were the 15 healthy age-matched control subjects.
mJOA scores were significantly improved after surgery for all patients (
For all patients, we found that sMEP (
The sMEP (
Separately, for Group A (16 patients), sMEP (
Postoperatively, no significant differences in sMEP for Group A (
For Group A, we found significantly reduced sMEP (
For Group B, there was no significant reduction in sMEP or
We did not find significant differences in CMCT from all 4 limbs and
Table
Summary of experimental results in all patients.
Preoperative | Postoperative | Significance | |||
---|---|---|---|---|---|
mJOA | 12.7 (2.81) | 13.81 (3.1) |
|
||
sMEP | 1.64 (1.88) | 0.82 (0.89) |
|
||
|
7.86 (3.93) | 5.22 (2.58) |
|
||
|
|||||
Group | A | B | A | B | |
|
|||||
sMEP | 2.03 (1.54) | 0.89 (0.55) | 1.01 (1.05) | 0.48 (0.31) | Group A ( |
|
9.07 (4.00) | 5.60 (2.80) | 5.87 (2.95) | 4.00 (1.41) | Group A ( |
|
|||||
Write | 23.98 (25.49) | 12.11 (5.84) | 20.97 (20.08) | 11.71 (6.10) | Group A ( |
Turn page | 9.56 (9.15) | 5.98 (2.22) | 7.19 (3.53) | 7.89 (4.86) | Group A ( |
Lift small object | 11.43 (7.06) | 9.65 (4.26) | 9.19 (4.13) | 9.54 (6.10) | Group A ( |
Feed | 13.20 (6.74) | 13.06 (7.12) | 12.03 (5.89) | 11.38 (5.74) | Group A ( |
Stack | 5.64 (6.43) | 2.85 (2.17) | 4.00 (6.13) | 2.78 (1.07) | Group A ( |
Lift light can | 4.23 (3.88) | 4.69 (2.87) | 4.21 (3.91) | 4.56 (3.90) | Group A ( |
Lift heavy can | 5.18 (2.34) | 5.14 (3.34) | 4.87 (2.02) | 4.89 (3.33) | Group A ( |
|
|||||
9-hole peg | 68.36 (41.78) | 52.63 (30.00) | 57.39 (28.63) | 59.77 (34.12) | Group A ( |
Tap | 68.63 (12.09) | 70.43 (10.83) | 69.79 (4.57) | 74.79 (8.13) | Group A ( |
Pinch grip strength | 15.55 (7.46) | 20.86 (3.8) | 17.87 (7.33) | 22.64 (4.68) | Group A ( |
|
|||||
CMCT | |||||
R UL | 10.77 (3.22) | 7.89 (2.34) | 9.88 (2.16) | 7.69 (2.87) | Group A ( |
L UL | 11.65 (3.45) | 7.99 (2.77) | 10.45 (2.98) | 7.62 (2.96) | Group A ( |
R LL | 18.34 (3.98) | 19.23 (4.11) | 17.97 (4.06) | 19.86 (4.68) | Group A ( |
L LL | 19.11 (4.07) | 19.25 (4.87) | 18.78 (4.39) | 20.12 (4.61) | Group A ( |
|
|||||
|
1.22 (0.53) | 1.34 (0.45) | 1.21 (0.47) | 1.29 (0.51) | Group A ( |
|
0.82 (0.23) | 0.91 (0.22) | 0.77 (0.36) | 0.86 (0.28) | Group A ( |
Mean values are indicated (standard deviation).
All hand function test results are in seconds.
In healthy controls, mean sMEP was 0.51 (0.28) and
CMCT: central motor conduction time (m/s); R: right; L: left; UL: upper limb; and LL: lower limb.
Figures
sMEP findings graphically. Asterisks denote statistical significance. Preoperative bars are black and postoperative bars are grey. sMEP, sum of MEP amplitudes in mV in vertical axis. Horizontal axis depicts patient and control groups.
Figure
Schematic diagram depicting motor output mapping of a patient in Group A. In the preoperative grid, sMEP is 1.7 mV as sum total of 10 stimulation positions eliciting an MEP (
In the first TMS study of this nature to our knowledge, we sought to provide a vital connection between existing studies using functional imaging and the recovery process after decompression surgery in CSM.
Early imaging studies in CSM have focused on morphological changes in operated CSM patients. Fukushima et al. [
The advent of functional imaging, including PET and fMRI, provided new information on brain remodelling by virtue of blood flow changes. In terms of spinal cord lesions, traumatic spinal cord injury (SCI) is known to induce expanded brain activation towards the leg areas, thalamus, and cerebellum as seen in PET studies [
Specifically for CSM, few studies have been published to date addressing fMRI changes before and after decompression. Holly et al. [
Neurobiological evidence certainly exits with regard to the axonal sprouting and contacting of propriospinal neurons in animal experiments after transection of corticospinal projection to the hind limbs [
In summary, TMS mapping of the motor cortex is well recognized to reflect functional plasticity of cortical outputs topographically [
Before decompressive surgery, increased cortical representation of intrinsic hand muscle compared to normal controls is not unexpected and likely reflects an inherent compensatory mechanism in response to cord compromise. The observation is corroborated by functional imaging in spinal cord injury [
In CSM, compression of descending corticospinal tracts results in desynchronization of I-wave volleys evoked with single pulse TMS of the primary motor cortex. The MEPs obtained can be used to calculate the CMCT by subtracting the peripheral conduction time. CMCT is more sensitive measure of corticospinal dysfunction in CSM than somatosensory evoked potentials [
Noteworthy though, we did not find significant CMCT changes before and after surgery in all 4 limbs, despite motor cortex excitability modulation evident with cortical mapping as well as improvement in hand function in relation to MEP changes. In line with these observations, modulation of the ability to facilitate horizontal rather than vertical synaptic connections would be the most likely underlying mechanism at play. As TMS largely stimulates cortical neurons in a transsynaptic fashion [
The lack of
In the light of current knowledge outlined above, it is imperative that our findings can be applied to elucidate modulation of cortical motor control mechanisms in CSM. Based on comparison with healthy controls and within each patient, compensatory expansion of the hand area, in terms of magnitude and spatial representation of cortical excitability postoperatively, is evident. These observations are further strengthened by findings that, for Group A patients, both magnitude and spatial characteristics were larger than controls, whereas for Group B, only spatial characteristic were. This may be related to Group A patients having relatively more upper limb motor deficits compared with Group B, hence, driving enhanced cortical compensatory representation [
We next examined cortical excitability modulation in relation to the functional relevance of these changes. In terms of objective hand function tests, significantly increased grip strength and reduced lifting time for small objects rather than the other tests likely reflected improved direct projections for intrinsic hand musculature. However, significant correlation of changes in magnitude of cortical excitability for both feeding and stacking objects also likely reflects participation of more proximal muscles needed for these tasks which were modulated in terms of horizontal placed connectivity postoperatively. Similarly, spatial changes in terms of number of excitable sites during TMS correlating with writing tasks also reflected functional cortical participation for both intrinsic muscle and wrist action, corroborating the experimental design and TMS both evaluating predominantly motor representation of distal muscles performing more finely skilled tasks.
All these observations, again, were seen exclusively in Group A patients, and all hand function tests were designed to evaluate the upper limb only. It would be interesting to compare our findings with the only fMRI study to date incorporating hand function tests [
It is noteworthy that current knowledge may be limited by several factors. For fMRI, tasks are often limited to motor imagery rather than actual muscle activation due to the presence of movement artefacts. For electrophysiological studies, however, both resting and active tasks can be studied. In an event when both functional imaging and TMS results must be combined, it should thus be noted that findings may not be directly comparable. Overall, published studies are usually small in subject numbers, lacking in standardization of protocols and serialization of data. These deficiencies should be addressed in larger future studies of a similar nature.
In conclusion, we have demonstrated that compensatory expansion of motor cortical representation with a tendency to normalization after surgery occurs largely at cortical rather than spinal level. Cortical plasticity modulation mirrored improvements in relevant tasks requiring utilization of predominantly distal hand muscles. These findings have important implications with regard to the understanding and rehabilitation of patients with lesions involving the cervical spinal cord.
There is no conflict of interests or financial disclosure for all authors.
Andrew Green contributed in manuscript concept, data acquisition, and manuscript preparation. Priscilia W. T. Cheong contributed in data acquisition. Stephanie Fook-Chong helped in data analysis. Rajendra Tiruchelvarayan helped in data acquisition. Chang Ming Guo helped in data acquisition. Wai Mun Yue helped in data acquisition. John Chen helped in data acquisition. Yew Long Lo contributed in manuscript concept, data acquisition, and manuscript preparation.