Scholarly article on topic 'Assessment of Cervical Myelopathy Using Transcranial Magnetic Stimulation and Prediction of Prognosis After Laminoplasty'

Assessment of Cervical Myelopathy Using Transcranial Magnetic Stimulation and Prediction of Prognosis After Laminoplasty Academic research paper on "Clinical medicine"

0
0
Share paper
Academic journal
SPINE
OECD Field of science
Keywords
{""}

Academic research paper on topic "Assessment of Cervical Myelopathy Using Transcranial Magnetic Stimulation and Prediction of Prognosis After Laminoplasty"

SPINE Volume 33, Number 1, pp E15-E20 ©2008, Lippincott Williams & Wilkins, Inc.

Assessment of Cervical Myelopathy Using Transcranial Magnetic Stimulation and Prediction of Prognosis After Laminoplasty

Jun Takahashi, MD,* Hiroki Hirabayashi, MD,* Hiroyuki Hashidate, MD,*

Nobuhiro Ogihara, MD,* Ikuya Yamazaki, MD,t Mikio Kamimura, MD,i Sohei Ebara, MD,§

and Hiroyuki Kato, MD*

Study Design. This study investigated the clinical usefulness of motor-evoked potentials (MEPs) produced by transcranial magnetic stimulation of the brain for cervical myelopathy patients.

Objective. The purpose of this study was to determine the usefulness of MEPs for the assessment of the severity of myelopathy and prediction of the outcome of lamin-oplasty.

Summary of Background Data. Magnetic stimulation has been widely used for examination of the descending excitatory motor pathways in the central nervous system, but little attention has been paid to cervical myelopathy.

Methods. We measured the MEPs of 56 patients who underwent surgery for cervical myelopathy. The MEPs from the abductor pollicis brevis, abductor digiti minimi, tibialis anterior, and abductor hallucis muscle were evoked by transcranial magnetic brain stimulation. The latency from the anterior horn cell of the spinal cord to the hand or foot muscles was also measured, with the F-value [(F + M — 1)/2] calculated. This was followed by estimation of the central motor conduction time (CMCT). Severity of clinical disability was scored on the basis of symptoms according to a modified ADL scale for cervical myelopathy of the Japanese Orthopedic Association (JOA) score.

Results. The average CMCT of the symptomatic side significantly correlated with the preoperative JOA score. The average CMCT of the symptomatic side significantly correlated with the 1-year postoperative JOA score. The average CMCT for patients with poor outcome was significantly longer than that for patients with good outcome. CMCT of 15 milliseconds or more in the upper extremities or that of 22 milliseconds or more in the lower extremities indicated poor prognosis.

Conclusion. In patients with cervical myelopathy, the CMCT significantly correlated with the results of clinical assessment. These findings regarding the duration of CMCT

may be useful parameters in spinal pathology for prediction of the outcome of surgical treatment.

Keywords: transcranial magnetic stimulation, central motor conduction time, cervical myelopathy. Spine 2008; 33:E15-E20

From the ^Department of Orthopaedic Surgery, Shinshu University, School of Medicine, Matsumoto-city, Nagano, Japan; "[Department of Orthopaedic Surgery, Nagano Matsushiro Hospital, Matsushiro-town, Nagano-city, Nagano, Japan; ^Kamimura Clinic, Kotobuki Toyooka, Matsumoto-city, Nagano, Japan; and ^Department of Orthopaedic Surgery, Chigasaki Tokushukai Medical Center, Saiwai-town, Chigasaki-city, Kanagawa, Japan.

Acknowledgment date: January 8, 2007. Revision date: August 5, 2007. Acceptance date: September 18, 2007.

The device(s)/drug(s) is/are FDA-approved or approved by the corresponding national agency for this indication.

No funds were received in support of this work. No benefits in any form have been or will be received from a commercial party related directly or indirectly to the subject of this manuscript. Address correspondence and reprint requests to Jun Takahashi, MD, Department of Orthopaedic Surgery, Shinshu University, School of Medicine, 3-1-1 Asahi, Matsumoto-City, Nagano, 390-8621, Japan; E-mail: jtaka@hsp.md.shinshu-u.ac.jp

With the progressive aging of society, increasing numbers of patients are presenting with cervical myelopathy, and opportunity for diagnosis is increasing. In advanced cases, the symptoms and signs of cervical myelopathy consist of complaints of gait disturbance, clumsiness, and paresthesias of the hands with clinical signs of pyramidal tract and posterior column involvement. However, the clinical presentation can be atypical. For instance, patients may complain of paresthesias or clumsiness, without showing abnormal clinical neurologic signs or it may be difficult to localize the lesion in patients who show pyramidal tract signs.

Severe myelopathy often shows poor prognosis even after surgical treatment. In diagnosis of myelopathy, severe spinal cord compression is often accompanied with mild symptoms, and mild compression with severe neurologic symptoms, when spinal compression is evaluated by magnetic resonance imaging (MRI) or myelogram. If complication with other neurologic disorder (spinal cord disorder other than of cervical spine, such as cauda equina disorder, diabetic peripheral neuropathy, or cerebral infarction, etc.) exists, it results in presentation of an atypical clinical symptom. In the case of patients with cervical sprain who present increased tendon reflexes and with myelopathy-like symptoms, surgical indication is difficult to determine. Electrodiagnosis, a functional diagnosis method, is feasible in such cases,1'2 and spinal cord evoked potential measurement is used routinely in Japan, however, it is slightly invasive. On the other hand, noninvasive and painless stimulation of the deep nervous system including cerebral cortex and peripheral nerve could be achieved by transcranial magnetic stimulation, which would enable implementation of multiple tests, before and after surgery, and could provide objective evaluation of the motor function which is necessary in ADL evaluation.3-5

In this report, we will introduce the basics and cautions regarding the transcranial magnetic stimulation method and investigate its usefulness and problems in assessment of cervical myelopathy on the basis of our

experience in using the method in severity assessment of myelopathy surgical indication determination and prediction of postoperative prognosis.

■ Materials and Methods

Subjects of this study included 56 patients (42 males and 14 females) who underwent surgery with diagnosis of cervical myelopathy and had central motor conduction time (CMCT) measured before surgery between June 1997 and April 2001 (3-year and 11-month period). Patients were aged 46 to 86 years (mean 64.5 ± 10.3) at the time of surgery, and the preoperative clinical disability scores according to a modified ADL scale for cervical myelopathy of the Japanese Orthopedic Association (JOA) (JOA score: 0-17; 17 represents full function) (Table 1) were 5 to 15 points (mean 10.6 ± 2.8). The study group consisted of patients with cervical spondylotic myelopathy (39 patients), cervical posterior longitudinal ligament ossification (14 patients), cervical disc herniation (2 patients), and cervical yellow ligament calcification (1 patient), who were operated by French door laminoplasty (52 patients) or open door lamin-oplasty (4 patients). JOA score measurement (before and 1 year after surgery) and a 10-second grip and release test were performed for clinical disability assessment, and the Hirabayashi method was used for recovery rate evaluation.

For statistical analysis, Student t test and Spearman rank correlation test were performed using StatView (SAS Institute, Cary, NC), with significance of P < 0.05.

Mechanism of Transcranial Magnetic Stimulation

Principle. Faraday's principle of electromagnetic induction is applied for the method. An alternating magnetic field is generated by short-term discharges of current (time constant: <0.3 milliseconds) from a capacitor to a coil to induce eddy currents in the body tissue. These eddy currents flow in reverse direction to the current in the coil. From this property and information about the arrangement and direction of deep nervous system to

Table 1. ADL Scale for Cervical Spondylotic Myelopathy

Motor dysfunction of upper extremity

0—Unable to feed oneself

1—Unable to handle knife and fork, able to eat with spoon.

2—Handle knife and fork with considerable difficulty

3—Handle knife and fork with slight difficulty

4—None

Motor dysfunction of lower extremity

0—Unable to walk

1—Able to walk on a flat floor with walking aid

2—Able to walk up and/or down stair with handrail

3—Lack of stability and smooth reciprocation, but able to walk unaided

4—No dysfunction Sensory deficit

A. Upper extremity

0—Severe sensory loss

1—Mild sensory loss

2—None

B. Lower extremity (same as A)

C. Trunk (same as A) Sphincter dysfunction

0—Unable to void

1—Marked difficulty in micturition

2—Difficulty in micturition

3—None

Modified from the scale proposed by the Japanese Orthopedic Association*.

*Adapted from Yonenobu et al.18

be stimulated, optimal stimulation sites and directions can be determined.

In painless and noninvasive stimulation of the deep nerve system, magnetic stimulation has advantages compared with electric stimulation, because magnetic stimulation reaches deeper tissue with smaller attenuation.

Stimulation Device. A MAGSTIM 200 monophasic transcranial stimulator (Magstim Co., Wales, UK) with a round coil (outside diameter, 110 mm) was used in this study.

Evoked Potential Recording System and Electrodes. A general evoked potential recording system or electromyograph with 4 channels or more is preferable. Recording electrodes should be placed on the left and right sides individually and on surfaces of upper and lower extremities (negative, belly muscle; positive, tendon).

Testing Positions of Patients. Patients are tested while awake and sitting on a chair or bed. Paralyzed patients are tested in the supine position on a bed.

Stimulation and Recording Method and Evaluation

After selecting the most adequate stimulation and recording methods, motor-evoked potentials (MEPs) should be evoked in extremity muscles by magnetic stimulation of the head, neck, and lower back. In the head stimulation, anterior horn cells in motor cortex are stimulated both directly and indirectly, and in the neck and lower back stimulation, the nerve root is stimulated at intervertebral foramen. From measured difference in latency time between head and neck, and head and lower back, (simple) CMCTs that represent latency time from brain to nerve root of neck and lower back are determined. F-waves and M-waves are then recorded, and F-values are calculated to determine CMCT (F), which represents the conduction time from brain to spinal anterior horn. In this study, no response or a latency time of 1 standard deviation (SD) or higher than that of the mean value in healthy volunteers was judged abnormal.

Measurement Method and Caution

The coil (double core type) should be placed on the parietal region and stimulations of 20% to 30% above threshold (maximum 100%) should be applied to the motor area at least 2 times, to ensure obtainment of reproducible MEP values. In the general procedure, records from 4 channels are obtained from both right and left sides at abductor pollicis brevis (APB), abductor digiti minimi (ADM), tibialis anterior (TA), and abductor hallucis (AH). To determine simple CMCT, stimulation is applied first to the head, followed by stimulation of cervical spine (C7 spinous process) and lumbar spine (middle point between L5 and S1 spinous process) to record latency times of APB, ADM, TA, and AH in the right and left sides of the body. Differences between left and right in these latency times represent simple CMCTs, which are conduction times from the mo-

Table 2. Preoperative JOA Score Versus Symptomatic Side CMCT Data

Correlation Coefficient P

JOA score

vs. CMCT (APB) -0.48 0.0006

vs. (ADM) -0.40 0.0051

vs. (TA) -0.34 0.0213

vs. (AH) -0.45 0.0025

Table 3. Preoperative the Upper and Lower Limb Motor Function Versus Symptomatic Side CMCT

Correlation Coefficient P

Upper limb motor function score

vs. CMCT (APB) —0.58 <0.0001

vs. (ADM) — 0.43 0.0026

Lower limb motor function score

vs. CMCT (TA) —0.55 0.0001

vs. (AH) —0.55 0.0001

tor area to the spinal nerve root. To determine CMCT (F), F-waves and M-waves are recorded by electric stimulation of median nerve (carpal region), ulnar nerve (carpal region), common peroneal nerve (fibra head), and tibial nerve (tarsal region), and F-values can be calculated from Kimura's formula6 ((F + M — 1)/2). The difference between F-value and latency time from head is then calculated as CMCT (F) = (latency time from head) — (F + M —1/2), which is the conduction time from the motor area of the brain to the spinal anterior horn. Following these procedures, approximate severity of motor pathway disorder and difference between the right and left sides can be estimated, and conduction speed from spinal cord and cauda equina can also be obtained.

Safety

Safety studies of this procedure have been conducted from multiple aspects and no adverse reaction has been reported in human or animal studies. The maximum current evoked by this magnetic stimulation system reported by Barker et al5 was approximately 50 ^c/pulse, which is only 1/2000 to 1/20,000 of the current volume used in electroconvulsive therapy (100 mc to 1C), and the maximum current evoked was approximately 0.25 A, smaller than the current evoked by the electric stimulation system of Merton et al.7 Heat generation in tissue by this method is estimated at 2 mW or less, which is less than 1/300 of the international standard. The maximum magnetic field strength used is similar to that of static magnetic fields used in MRI, and the action time is very short. Various studies regarding relation to atrial fibrillation also concluded that there is no safety-related concern.8

As mentioned above, this method was demonstrated to be highly safe, but careful and strict selection of subjects is essential in clinical implementation. Patients with history of epilepsy, electrical implants such as cardiac pacemakers, or affixed magnetic bodies such as intracranial surgery clips should be ruled out. In cases such as cervical myelopathy with mechanical pressure in spinal canal, the patient should wear a neck collar to prepare for body movement.

■ Results

Right/left (normal control) preoperative CMCT values for APB, ADM, TA, and AH were 12.3 ± 4.4 millisec-

Table 4. Preoperative 10-Second Grip and Release Test Versus Upper Limb CMCT

Correlation Coefficient P

10-second grip and release test vs. CMCT (APB) —0.24 0.0247

vs. (ADM) —0.23 0.0317

Table 5. Symptomatic Side CMCT Data Versus 1-Year Postoperative JOA Score

Correlation Coefficient P

CMCT (APB) is. JOA score —0.40 0.0097

(ADM) is. —0.42 0.0071

(TA) is. —0.32 0.0405

(AH) is. —0.40 0.0087

onds/12.1 ± 4.2 milliseconds (9.4 ± 1.0 milliseconds), 11.5 ± 4.0 milliseconds/12.2 ± 3.1 milliseconds (6.2 ± 1.4 milliseconds), 18.3 ± 5.4 milliseconds/17.8 ± 5.7 milliseconds (14.3 ± 2.3 milliseconds), and 18.5 ± 5.3 milliseconds/18.9 ± 4.2 milliseconds (14.9 ± 2.0 milliseconds), respectively. These values showed prolongation compared with those of normal subjects reported by Kameyama et al9 or Iizuka.10

As for the relationship between the preoperative JOA score and symptomatic side CMCT data, the preopera-tive JOA score and CMCT (TA) showed slight correlation, whereas the preoperative JOA score and CMCT (APB), CMCT (ADM), and CMCT (AH) were significantly related (Table 2). Comparing the upper or lower limb motor function score and symptomatic side CMCT data, the preoperative upper limb motor function score showed significant correlation with CMCT (APB) and CMCT (ADM), whereas data for preoperative lower limb motor function and CMCT (TA) and CMCT (AH) were also significantly related (Table 3). In a comparison of the 10-second grip and release test and upper limb CMCT data, significant relationship existed between the results from the grip and release test and CMCT (APB) and CMCT (ADM), though the correlation coefficients were small (Table 4). Between the symptomatic side CMCT data and 1-year postoperative JOA score, the JOA score and CMCT (APB), CMCT (ADM), and CMCT (AH) showed significant correlation, i.e., subjects with prolonged preoperative CMCT showed lower 1-year postoperative JOA scores (Table 5). As for the correlation between symptomatic side CMCT and upper and lower limb motor function score, subjects with prolonged CMCTs (except for CMCT (AH)) showed lower 1-year postoperative upper and lower limb motor function scores (Table 6).

Evaluating the subjects by JOA score (Hirabayashi method), we classified subjects with recovery rates 60% or more as good outcome and 40% or less as poor out-

Table 6. Symptomatic Side CMCT Data Versus 1-Year Postoperative Upper and Lower Limb Motor Function Score

Correlation Coefficient P

CMCT (APB) vs. upper limb motor function score —0.34 0.0248

(ADM) vs. —0.34 0.0273

(TA) vs. lower limb motor function score —0.50 0.0015

(AH) vs. —0.27 NS

Figure 1. Comparison of preoperative upper limb CMCT between good and poor prognosis groups. Significant prolongation of CMCT (APB) amd CMCT (ADM) were observed in poor prognosis group compared to good prognosis group.

CMCT(APB)

CMCT(ADM)

p<0.05

p<0.05

0 2 4 6 8 10 12 14 16 ms

Means ± S.D.

Good outcome patients : Recovery rate >60% n=17 Poor outcome patients : Recovery rate <40% n=16

come. Preoperative CMCT (APB) values in good and poor outcome groups were 10.8 ± 3.6 milliseconds (mean ± SD) and 13.7 ± 2.9 milliseconds, respectively, and CMCT (ADM) values were 11.5 ± 3.5 milliseconds and 14.9 ± 1.2 milliseconds, respectively, showing significant prolongation in the poor outcome group (Figure 1). In lower limb CMCT, CMCT (TA) levels in good and poor outcome groups were 16.7 ± 4.8 milliseconds and 21.4 ± 5.3 milliseconds, respectively and CMCT (AH) values were 18.3 ± 4.9 milliseconds and 21.9 ± 3.5 milliseconds, respectively, showing significant prolongation in the poor outcome group (Figure 2).

A latency time of 1 SD or higher than that of the mean value in healthy volunteers9,10 was used for the evaluation of the CMCT prolongation. CMCT values for APB, ADM, TA, and AH on the symptomatic side showed prolongation in 42 cases (75%), 50 cases (89.3%), 42 cases (75%), and 37 cases (66.1%), respectively, among the 56 patients. Of the 56 patients, 28 (50%) showed

increased CMCT in APB, ADM, TA, and AH. Of the 56 patients, 52 (94.6%) patients showed intramedullary high signal intensity in MRI whereas the 3 (5.4%) patients without intramedullary high signal intensity did not show increased CMCT in APB, ADM, TA, and AH on the symptomatic side.

■ Discussion

Since the introduction of transcranial magnetic stimulation of the human motor cortex, the painless and nonin-vasive assessment of central motor conduction has become a useful diagnostic method. There are different reasons for delayed CMCT: slowed conduction in demy-elinated corticospinal fibers, conduction along other oli-gosynaptic pathways, or reduction of size and synchrony of corticospinal volleys reaching the anterior horn cells.11

In function tests of myelopathy, electrophysical methods including needle electromyogram and F wave and W

CMCT(TA)

CMCT(AH)

p<0.05

Good outcome patients Poor outcome patients

p<0.05

20 25 Means ± S.D.

Recovery rate >60% n=17 Recovery rate <40% n=16

Figure 2. Comparison of preoperative lower limb CMCT between good and poor prognosis groups. Significant prolongation of CMCT (TA) and CMCT (AH) were observed in poor prognosis group compared to good prognosis groups.

wave measurements are routinely employed. However, these are indirect measurements and direct evaluation of spinal conduction function is not possible with them. Another spinal conduction test, the somatosensory-evoked potential, is mainly focused on sensory pathway and does not reflect motor pathway function. The recent development of transcranial magnetic stimulation has enabled evaluation of the motor nerve system's conduction pathway,12 and this method is applied to objective evaluation of myelopathy.

In diagnosis of cervical myelopathy, conventional diagnostic methods such as neurologic findings, MRI, and myelogram are usually performed, but conclusive diagnosis is sometimes difficult because many symptoms tend to be differentiated from the existing disease. In this regard, transcranial magnetic stimulation could be an effective diagnostic tool. Transcranial magnetic stimulation also has potential to enable early diagnosis of myelopathy, by detecting changes in CMCT and in MEP wave forms that reflect conduction times of the pyramidal tract, because early pathologic changes of cervical spondylotic myelopathy includes demyelinating of the pyramidal tract.13,14 In accordance with the result of a study by Iizuka et al15 and Kameyama et al,9 which demonstrated significant prolongation of upper limb CMCT in subjects with cervical myelopathy compared with control and correlation with severity of symptom, this study showed correlation between upper/lower limb CMCT and JOA score, presenting the possibility of quantitative evaluation of the myelopathy by transcranial magnetic stimulation. Furthermore, this test method has advantages in the evaluation of patients who have developed severe joint deterioration due to rheumatoid arthritis or who have difficulty in communication and, therefore, for whom neurologic detection is difficult.

In 1987, Ono et al14 reported specific hand deformity, loss of function, and marked lateral spine demyelination in pathology of cervical myelopathy, and defined my-elopathy of hand by presentation of finger escape sign and slow grasp/release in a 10-second grip and release test. In this study, the results of 10 second tests correlated with upper limb CMCT and demonstrated the usefulness of these variables in quantitative evaluation of cervical myelopathy.

In a comparison between CMCT data and surgical outcome, Kameyama16 reported poor prognosis in patients with prolonged preoperative CMCT and with enhancement of intensity in spinal compressed region in T2 contrast MRI image, because of irreversible changes in spinal cord. Okada17 also reported a significant relationship between CMCT and postoperative JOA score or recovery rate determined by the Hirabayashi method, demonstrated by regression analysis of preoperative variables in cervical myelopathy to detect the relationship. This report also showed significant prolongation in preoperative upper and lower limb CMCT in poor prognosis patients compared with those with good prognosis, and preoperative CMCT in upper and lower limbs

longer than 13 milliseconds and 21 milliseconds, respectively, predicted poor prognosis.

Transcranial magnetic stimulation showed potential in quantitative diagnosis and qualitative diagnosis in chronic compressive cervical myelopathy. This method is comparable to the magnetic stimulation system reported by Barker et al.12 This evaluation method is also useful in the process of informed consent by estimating postoperative recovery rate.

■ Conclusion

In patients with cervical myelopathy, the CMCT significantly correlated with the results of clinical assessment. These findings regarding the duration of CMCT may be useful parameters in spinal pathology for prediction of the outcome of surgical treatment.

■ Key Points

• The average CMCT of the symptomatic side significantly correlated with the preoperative JOA score, i.e., patients with lower JOA scores had longer CMCTs.

• The average CMCT of APB and ADM muscles significantly correlated with the grip and release test results, and average CMCT of the symptomatic side with the 1-year postoperative JOA score.

• The average CMCT for patients with poor outcome (recovery rate <40%, n = 16) was significantly longer than that for patients with good outcome (recovery rate >60%, n = 17).

• CMCT of 15 milliseconds or more in the upper extremities or that of 22 milliseconds or more in the lower extremities indicated poor prognosis.

References

1. Cracco RQ. Spinal evoked response: peripheral nerve stimulation in man. Electroenceph Clin Neurophysiol 1973;35:379-86.

2. Hoffmann P. Demonstration eines Hemmungsreflexes im menschlichen Reckenmark. Zeitschrift fur Biologie 1919;70:515-25.

3. Eisen AA, Shtybel W. AAEM minimonograph #35: clinical experience with transcranial magnetic stimulation. Muscle Nerve 1990;13:995-1011.

4. Barker AT, Freeston IL, Jalinous R, et al. Magnetic stimulation of human brain. J Physiol (Lond) 1985;369:3-6.

5. Barker AT, Freeston IL, Jalinous R, et al. Magnetic stimulation of the human brain and peripheral nervous system: an introduction and the results of an initial clinical evaluation. Neurosurgery 1987;20:100-9.

6. Kimura J. Electrodiagnosis in Disease of Nerve and Muscle: Principle and Practice. FA Davis, Philadelphia 1983:353-77.

7. Merton PA, Morton HB. Stimulation of the central cortex in the intact human subject. Nature 1980;285:227.

8. McRobbie DM, Foster MA. Cardiac response to pulsed magnetic fields with regard to safety in NMR imaging. Phys Med Biol 1985;30:695-702.

9. Kameyama O, Shibano K, Kawakita H, et al. Transcranial magnetic stimulation of the motor cortex in cervical spondylosis and spinal canal stenosis. Spine 1995;20:1004-10.

10. Iizuka M. Diagnosis in magnetic stimulation: motor pathway evaluation by magnetic stimulation method in cervical myelopathy. MB Orthop 1997;10:37-49.

11. Felix D, Wiesendanger M. Pyramidal and non-pyramidal motor cortical effects on distal forelimb muscles of monkeys. Exp Brain Res 1990;40:332-9.

12. Baker AT, Jalinous R, Freeston IL, et al. Non-invasive magnetic stimulation of the human motor cortex. Lancet 1985;1:1106-7.

13. Ogino H, Tada K, Okada K, et al. Canal diameter, anteroposterior com-

pression ratio and spondylotic myelopathy of the cervical spine. Spine 1983;8:1-15.

14. Ono K, Ebara S, Fuji T, et al. Myelopathy hand. New clinical signs of cervical cord damage. J Bone Joint Surg Br 1987;69:215-9.

15. Iizuka M, Tanaka H, Sugioka H, et al. Motor pathway evaluation in cervical myelopathy using percutaneous magnetic stimulation. Spine Spinal Cord 1989;2:611-6.

16. Kameyama O. Prediction of post-operative recovery rate in cervical myelopathy: is prognosis prediction possible? Electrodiagnosis of Spine [Sekizui denki shindangaku] 1995;17:81-4.

17. Okada F. Spinal motor function evaluation in cervical myelopathy using transcranial magnetic stimulation. Cent Jpn J Orthop Trauma 2000;43:1-5.

18. Yonenobu K, Okada K, Fuji T, et al. Causes of neurologic deterioration following surgical treatment of cervical myelopathy. Spine 1985;11:818-23.