Abstract
Objective: To investigate specificity and significance of dynamic changes of the cervical dural sac and spinal cord during neck flexion in juvenile muscular atrophy of the distal upper extremity.
Background: The disorder affects young people—predominantly men—and is progressive for several years. One autopsy case showed ischemic necrosis of the cervical anterior horn, suggesting that the disorder is a type of cervical myelopathy. Some authors classify it as monomelic amyotrophy, implying that it is a focal motor neuron disease.
Methods: Neuroradiologic examinations including myelography, CT myelography, and MRI in a fully flexed neck position were performed on 73 patients with this disorder and on 20 disease control subjects.
Results: A distinctive finding in the disorder was forward displacement of the cervical dural sac and compressive flattening of the lower cervical cord during neck flexion. The forward displacement was significantly greater in patients with disease duration less than 10 years than in age-matched control subjects and patients in a late, nonprogressive stage.
Conclusions: Radiologic abnormalities of the lower cervical dural sac and spinal cord support the hypothesis that this disorder is a type of cervical myelopathy.
Juvenile muscular atrophy of the distal upper extremity1-3 has been reported in hundreds of patients, mainly in Japan and other Asian countries. It occurs predominantly in young men and is rarely familial. The onset is insidious and characterized by muscular weakness and muscular atrophy in the hand and forearm; the brachioradialis is spared (oblique amyotrophy). The amyotrophy is unilateral in most patients, asymmetrically bilateral in some, and rarely symmetric. Although there are no fasciculations when the hand is at rest, weak finger extension causes fascicular twitching of the forearm muscles and tremulous movement of the fingers. Subjective and objective sensory disturbances are usually absent, and muscle stretch reflexes are normal. There is no cranial nerve involvement, no pyramidal signs, and no urinary disturbances. Electromyography shows acute and chronic denervation restricted to the upper limbs; peripheral motor conduction is normal. The initial progressive course is followed by a spontaneous arrest within several years after onset. An autopsy performed on a 38-year-old man who died of lung cancer revealed typical amyotrophy in the hands and forearms, predominantly on the left side; he had exhibited these symptoms for 23 years.4 The spinal cord showed flattening in the anteroposterior (AP) direction and ischemic changes of the anterior horn with a loss of large and small neurons and a weak gliosis predominantly on the left, from C5 through T1, maximally at C7 and C8. Another autopsy was performed on a 76-year-old man who had disease onset at age 24, complicated by cervical spondylosis later in life.5 The neuropathologic findings in the anterior horn were similar to those of the first case, but discussion of pathogenesis was difficult because of a concomitant cervical spondylosis.
Pathophysiology of the illness is unknown; our pathologic findings4 prompted us to investigate a possible myelopathic origin. Others consider the illness a monomelic amyotrophy, which implies localized and limited lower motor neuron degeneration.6,7 Routine radiologic examinations, including MRI, may show atrophy of the lower cervical cord,8 but do not shed light on the pathogenesis. We and others have reported dynamic changes of the cervical dural sac and spinal cord during neck flexion in case series of this illness.9-13 We further investigated if those dynamic changes are characteristic of this illness and contribute to the pathogenesis.
Methods.
Subjects.
We studied 73 patients (68 men and 5 women) who were examined at Chiba University Hospital since 1981. Each had the above-mentioned typical clinical features of juvenile muscular atrophy of the distal upper extremity. Age at onset ranged from 11 to 19 (mean, 15.6 ± 1.9) years, and age at radiologic examination ranged from 12 to 52 (mean 22.3 ± 8.1) years. Arm muscle weakness and atrophy were unilateral in 59 patients, asymmetrically bilateral in 13, and symmetrically bilateral in one.
Twenty patients (16 men and 4 women) with other neurologic disorders were examined as control subjects. These subjects underwent neuroradiologic examinations as a diagnostic procedure. Diagnoses included cervical spondylosis (5 patients), spastic spinal paraparesis (4), spinocerebellar degeneration (3), neurosis (2), lumbar spondylosis (1), thoracic outlet syndrome (1), spinal vascular disorder (1), syringomyelia (1), cervical radiculopathy (1), and peripheral neuropathy of the arm (1). Age at radiologic examination ranged from 12 to 56 (mean, 34.9 ± 14.5) years.
Neuroradiologic studies.
All patients and disease control subjects consented to neuroradiologic examinations. Myelograms showing AP and lateral views at the cervical level were obtained with the neck in a neutral position. Thereafter, the lateral views were further studied in fully flexed and extended positions of the neck. On the lateral views, we measured an AP diameter of the dural sac at the superior margin of the C6 vertebral body for each position. Lateral views showing distinct margin of the dural sac were included in statistical analyses. They consisted of 60 sets of myelograms for patients and 17 sets for disease control subjects. AP diameters of the dural sac at a neutral (Dn) and a flexed (Df) neck position were measured for each set. The degree of neck flexion was measured as the angle made by the intersection of lines extending from the posterior margins of the C2 and C7 vertebral bodies.
CT scan about 2 hours after myelography was performed on 65 patients and 16 disease control subjects. Transverse sections of the neck from C3 to C7 midvertebral and intervertebral levels were studied in neutral and fully flexed neck positions. To obtain full flexion of the neck in the CT scanner, the trunk was tilted down rostrally using a pelvic wedge (figure 1). The plane of CT sections was visually determined on a scout view to be perpendicular to the vertebral body axis at each level. The dural AP diameters at a Dn and a Df neck position were measured at the C6 vertebral level. The degree of neck flexion was measured on the scout view in the same manner as on the lateral view of myelogram. The reader of the myelograms and CT myelograms was blind to the diagnosis.
Figure 1. Positioning on the CT table to obtain full flexion of the neck. Maximal elevation of the head against the trunk is possible when the trunk is tilted down rostrally using a pelvic wedge.
MRI of the cervical spine and spinal cord was available in 47 patients and 14 control subjects who were examined after 1986. In addition to sagittal and transverse images in a neutral neck position, those in a fully flexed neck position were studied using the same device as in CT myelography. The MR unit was a superconducting system operated at 0.5 tesla (Toshiba MRT50A). T1-weighted (SE 400/30) and T2-weighted (SE 2000/80) images were obtained with surface coils. Cinematographic MRI was performed on 30 patients by the electrocardiographically referenced field echo method, employing a TR/TE/angle of 55/22/30°, slice thickness of 10 mm, and acquisition matrix of 128 × 128 or 256 × 256 for display.
Statistical methods.
Because juvenile muscular atrophy has a chronological course characterized by onset during the teens, and is progressive for several years and thereafter nonprogressive, statistical analyses of myelographic measurements were made after age matching. Younger groups (less than 30 years old at myelographic examination) included 51 patients with this disorder (mean age, 19.2 ± 3.45 years) and eight disease control subjects (mean age, 19.1 ± 5.72 years). Older groups included nine patients with this disorder (mean age, 38.1 ± 5.90 years) and nine control subjects (mean age, 42.2 ± 6.16 years). The duration after disease onset was less than 10 years for most younger patients and more than 17 years for the older patients. The group means were compared by Student’s t-test. Significance was defined as p < 0.05.
Results.
Myelography.
The AP view of myelogram was normal in all of the patients and control subjects. On the lateral view, in a neutral neck position, the spinal cord was mild to moderately atrophic at around the C6 vertebral level in 39 (65%) patients. Among control subjects, only two showed mild cord atrophy or compression by spondylotic changes. Sixty-four of 73 (88%) patients showed dynamic changes of the position and diameter of the dural sac on neck flexion (figure 2). At the C6 vertebral and neighbor levels, the dural sac shifted anteriorly and its AP diameter was decreased. The epidural space behind the dural sac was radiolucent. These dynamic changes were not observed on neck extension.
Figure 2. Lateral myelogram views in neutral (A) and a fully flexed (B) positions of the neck in an 18-year-old patient. Arrows indicate the posterior margin of the dural sac on neck flexion. The dural sac shifts toward the vertebral body at the C5 to C7 levels on neck flexion. Anteroposterior (AP) diameter of the dural sac is decreased and a radiolucent space appears behind the dural sac. The AP diameters (mm) of the dural sac are measured at the superior margin of the C6 vertebral body.
Decrement ratio of the dural AP diameter on neck flexion was calculated with the following equation: (Dn − Df)/Dn. Patients younger than age 30 had a greater decrement ratio (mean, 0.45 ± 0.13) than age-matched control subjects (mean, 0.03 ± 0.07) (p < 0.0001) and the older patients (mean, 0.07 ± 0.12) (p < 0.0001). Older patients did not show a significant decrement of the dural AP diameter compared with the older control subjects (mean, 0.03 ± 0.05). The degree of neck flexion was not different between groups. In patients, the decrement ratio of dural AP diameter was inversely correlated to the duration of illness (figure 3).
Figure 3. Correlation between the decrement ratio ([Dn − Df]/Dn, where Dn is a neutral and Df is a flexed neck position) of the dural AP diameter and disease duration (years) in 60 patients. The decrement ratio is inversely correlated to the disease duration.
CT myelography.
In a neutral neck position, the dural sac had normal location and a normal elliptical shape in all patients (figure 4A). The spinal cord was normal in eight patients (12%) and flattened in the AP direction in 57 (88%). The flattening of the spinal cord was asymmetric in 39 (68%) and symmetric in 18 (32%) patients. These changes were most remarkable at the C6 vertebral level, and extended to one vertebral body length in 11 (19%), to two body lengths in 31 (54%), and beyond two body lengths in 15 (26%) patients.
Figure 4. CT myelography in neutral (A) and a fully flexed (B) positions of the neck, showing the same patient as in figure 2. Transverse sections of the C3, C4, C5, C6, and C7 vertebral levels are shown from top to bottom. The plane of CT sections is perpendicular to the vertebral body axis at each vertebral level in both neutral and flexed neck positions. (A) The spinal cord is slightly flattened anteroposteriorly at the C6 level in a neutral neck position. (B) On neck flexion, the spinal cord and dural sac are displaced forward and remarkably flattened from the C5 to C7 levels.
Of 65 patients who underwent CT myelography, 49 patients were examined in a fully flexed position of the neck. As expected from the results of myelogram, 46 (94%) patients showed forward displacement and tightening of the dural sac (see figure 4B). There was a positive correlation between those of myelography and CT myelography with respect to the degree of neck flexion and the decrement ratio of AP diameter of the dural sac ((Dn − Df)/Dn) (figure 5). Three of four patients over age 30 years had no or small decrement. The spinal cord was flattened and had an eggplant, pear, or boomerang shape. The flattening was asymmetrical in 40 (80%) patients, and the more flattened side corresponded invariably to the more atrophied limb. The flattening located at the C6 vertebral body and its neighboring levels was more extensive than in the neutral position, within two vertebral body lengths in seven (15%), and beyond two body lengths in 39 (85%) patients. Epidural space appeared as a low-density area behind the dural sac. In control subjects, five subjects showed mild spondylotic compression to the dural sac, and three had mild cord atrophy on a neutral neck position; none showed flattening of the spinal cord on neck flexion.
Figure 5. Correlation between measurements on myelography and CT myelography for 46 patients. Both of the degree of neck flexion (A) and the decrement ratio ([Dn − Df]/Dn, where Dn is a neutral and Df is a flexed neck position) of the anteroposterior diameter of the dural sac (B) correlated positively. For the decrement ratio, open circles indicate patients over 30 years of age.
MRI of the spinal cord.
With the neck in a neutral position, the spinal cord was in the normal location in the vertebral canal, and there was no abnormal intrinsic cord signal in all 47 patients who underwent MRI (figure 6, A and D). Twenty-three (49%) patients showed atrophy of the lower cervical cord (see figure 6D). In a fully flexed position of the neck, 41 (87%) patients showed forward displacement and flattening of the lower cervical cord, and a crescent-shaped high signal area behind the cord (see figure 6, B, C, and E). Changes of the intrinsic cord signal on neck flexion could not be judged because of the flattening of the cord. The crescent had high signal on both T1-weighted (see figure 6B) and T2-weighted (see figure 6, C and E) images, and occupied the epidural space behind the displaced dural sac seen in myelography and CT myelography. The center of the crescent located at around the C6 vertebral level in a sagittal section. In 11 patients, there were linear or round signal voids (low signal) in the high signal area (figure 7). Cinematographic MRI showed that the signal voids pulsated synchronously with cardiac beat. Of 47 patients, six showed neither cord flattening nor epidural high intensity signal on neck flexion, but had atrophy of the lower cervical cord. All had a disease duration more than 10 years. All control subjects showed neither cord flattening nor epidural high intensity on neck flexion.
Figure 6. Cervical MRIs in a neutral (A and D) and a fully flexed (B, C, and E) position of the neck in a 20-year-old patient. The transverse section (D and E) is perpendicular to the vertebral body axis at C6. (A and D) In a neutral neck position, the spinal cord is located normally in the vertebral canal (T1-weighted image). (D) The transverse section in a neutral neck position shows atrophy of the cord on the right. There is no abnormal intrinsic cord signal. (B, C, E) During neck flexion, the spinal cord is displaced forward and flattened from the C5 to T1 vertebral levels. A crescent-shaped high signal area appears behind the spinal cord on both T1-weighted (B) and T2-weighted (C and E) images.
Figure 7. Signal void in the high signal area, reproduced from cinematographic MR imaging, in a 16-year-old patient. (A) No abnormal findings on a neutral neck position. (B) During neck flexion, a high-intensity area with a round signal void appears behind the spinal cord. The signal void pulsated synchronously with the heart beat.
Discussion.
The present neuroradiologic findings suggest a myelopathic mechanism in juvenile muscular atrophy of the distal upper extremity. Mild to moderate atrophy of the lower cervical cord in a neutral neck position was present on MRI in about half of patients. More dramatically, myelography, CT myelography, and MRI reproduced forward displacement of the dural sac and flattening of the lower cervical cord in a fully flexed position of the neck. The cord flattening was asymmetrical, with the more flattened side corresponding to the more atrophied limb. Some Japanese authors reported similar radiologic findings in patients with this disease, and named the phenomenon “over-stretch of the cord,”9,11 “tight dural canal in flexion,”10 or “disproportionate shortening of the dorsal roots”13 based on their hypothetical mechanism of dynamic changes in the spinal canal. Our study, involving a large number of patients, revealed that the dynamic compression of the lower cervical cord on neck flexion is an unequivocal finding confined to a progressive stage of the disease. An absence of forward displacement of the dural sac and cord compression in elderly patients whose disease had arrested suggests that the dynamic compression has a pathogenic significance. However, the exact mechanism of the myelopathy remains unclear. We can only speculate that a possible mechanism of lower cervical anterior horn damage is a chronic circulatory insufficiency or chronic trauma induced by repeated or sustained neck flexion. The anterior horn is the most vulnerable to chronic ischemia, whereas the white matter of the spinal cord is more resistant.14 The epidural high signal behind the displaced lower cervical cord suggests circulatory changes in the spinal canal during neck flexion. Cinematographic MRI showed that the signal void in the epidural space pulsated synchronously with cardiac beat. These findings seem to represent passive dilatation of the epidural venous plexus due to forward displacement of the dural sac, but their pathogenic role is unknown.
In 1983, Gourie-Devi et al.6 introduced the term monomelic amyotrophy to describe 23 patients with single-limb atrophy. Of the 23 patients, 10 had lower-limb involvement and four had proximal or diffuse arm involvement. The other nine patients had distal upper limb involvement similar to the disease discussed here. All the patients shared the common features of insidious onset in the second and third decades of life, male preponderance, and nonprogressive course. As suggested by the authors, monomelic amyotrophy could be a limited and localized lower motor neuron degeneration affecting any spinal cord segments. There is, however, no pathologic evidence to support this hypothesis. In juvenile muscular atrophy of the distal upper extremity, the forward displacement of the lower cervical dural sac and flattening of the cord suggest that a mechanical force working on the lower cervical cord on neck flexion may be operating as a causative or promoting factor.
Acknowledgments
Acknowledgment
The authors thank Masashi Nakajima, MD, for his editorial help.
- Received April 23, 1999.
- Accepted in final form February 10, 2000.
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
