Brachial amyotrophic diplegia

Patients and methods.

The criteria for inclusion in this study were the presence of progressive motor dysfunction, a clinical examination demonstrating a pattern of weakness limited to the upper extremities bilaterally at least 18 months after onset, and an electrodiagnostic study consistent with a disorder of lower motor neurons. We excluded patients from our analysis if there was any evidence of weakness in neck, bulbar, or respiratory muscles before 18 months. Patients were also excluded if there was bladder or bowel dysfunction, sensory signs or symptoms, onset in the presence of radiating pain, weakness in the distribution of individual motor nerves, or an abnormality on cervical MRIs suggestive of gliosis, a mass lesion, inflammation, or syrinx.

All patients meeting the above criteria continue to be actively followed in our clinics. The electrodiagnostic studies were performed during the first 12 months after the onset of weakness. Results of laboratory data were abstracted from the medical records. Anti-GM1 and anti-asialo GM1 antibodies (GA1) were tested at Athena Diagnostic Laboratories, Worcester, MA.

Case report.

A 53-year-old man presented with a 10-month history of progressive, painless arm and shoulder girdle weakness (figure). Weakness began in the right arm and moved to the proximal left arm 6 months later. Over the next year, the weakness spread throughout the upper limbs in a relatively symmetric pattern. He denied sensory loss and bladder involvement. Medical and family history were unremarkable.

• Download figure
• Open in new tab
• Download powerpoint

Figure. (A) A 59-year-old man with brachial amyotrophic diplegia 7 years after his initial symptom. Note the neurogenic “man-in-the-barrel” phenotype causing the arms to hang flaccidly at his sides. (B) Note the severe neurogenic atrophy throughout the arm and shoulder girdle with normal bulk in the sternocleidomastoid.

On examination there was severe atrophy and weakness of both upper extremities causing the arms to hang limply at his sides. Motor testing demonstrated symmetric 3/5 (Medical Research Council) strength on shoulder abduction, internal and external rotation, elbow flexion and extension, wrist and finger extension, and finger abduction. There was grade 4 strength in the wrist and finger flexors. Facial strength, neck flexors and extensors, and lower extremities were normal. Fasciculations were evident throughout the upper limbs. Sensation was not affected. Muscle stretch reflexes were absent in the upper extremities and normal at the knees and ankles.

Nerve conduction studies (NCS) revealed low-amplitude median and ulnar compound muscle action potentials (CMAPs). There was no conduction block or other evidence of demyelination. Needle EMG demonstrated denervation potentials and large-amplitude polyphasic motor units with reduced recruitment throughout the upper limbs. The sternocleidomastoid and lower-extremity muscles were unremarkable. Cervical MRI was unremarkable. Serum creatine kinase and genetic testing for mutations in the survival motor neuron gene and for X-linked spinobulbar muscular atrophy were negative.

Over the next 2 years weakness in the upper extremities progressed to grade 0 in the deltoid, biceps, triceps, wrist extensors, and interosseus muscles and to 4/5 in the wrist and finger flexors. The pattern of weakness remained stable until the seventh year after onset when the patient first noticed mild leg weakness. Neck weakness developed 6 months later. A recent examination demonstrated symmetric 4/5 proximal and distal lower-extremity weakness and 4/5 neck flexor and extensor weakness. Respiratory and bulbar function was normal, and there was no evidence of corticospinal tract dysfunction.

Results.

Of approximately 400 patients evaluated for motor neuron disorders at the University of Texas Southwestern Medical Center between 1993 and 1997, we identified 8 with the characteristic phenotype and clinical course (2.0%). We also included in the analysis two additional patients evaluated by physicians from Stanford University School of Medicine in 1998. These 10 patients included 6 men and 4 women with a mean age of 58.9 years (range 40 to 78) (table 2). The mean age at onset was 53.3 years (range 35 to 68), and the disease duration ranged from 3 to 11 years. No patient had a family history of neuromuscular disease.

Seven patients reported unilateral weakness at onset that became bilateral over 2 months to 2 years. The onset was bilateral in the remaining three patients. Seven patients first noticed weakness in proximal arm muscles, one in the entire arm, and two in the hand. During follow-up, the distribution of atrophy and reflex loss remained localized to muscles innervated by the C5 and C6 myotomes in one patient and by the C5 through C7 myotomes in one. The remaining seven patients had weakness throughout the entire arm, typically with more proximal than distal involvement. Even as weakness advanced, the physiologic process tended to spare the finger flexors compared with other distal muscle groups. In nine patients the deep tendon reflexes were hypoactive in the upper extremities and normal in the lower extremities. One patient had hyperactive reflexes throughout all four limbs, although muscle tone was not increased and the plantar responses were flexor. Two patients developed mild lower-extremity weakness at 4 and 6 years after onset, and one patient developed lower-extremity and neck weakness during the seventh year after onset. However, no patient has lost the ability to ambulate independently, required assistive devices, or developed bulbar symptoms or signs, or respiratory difficulties. There were no deaths.

Discussion.

We describe 10 patients with a sporadic adult-onset amyotrophy affecting proximal arm and shoulder girdle muscles with variable degrees of distal upper-limb involvement. Respiratory, neck, and bulbar muscles were unaffected. The pattern of weakness resulted in a neurogenic man-in-the-barrel syndrome with the arms hanging flaccidly at the patient’s sides.

We only included patients when the characteristic phenotype was present at least 18 months after onset. Eight of the 10 patients meeting these strict criteria had a purely lower motor neuron syndrome with clinical and electrodiagnostic abnormalities fully restricted to the arms. One patient also had a purely lower motor neuron syndrome, but there was subclinical electrodiagnostic evidence of denervation in the lower extremities. Only one patient presented with increased muscle stretch reflexes and showed widespread subclinical neurogenic involvement on EMG, suggesting a diagnosis of ALS. It is uncertain whether these presentations represent variable manifestations of a single disorder or different clinically overlapping conditions.

Bilateral arm weakness is a presenting symptom in 5 to 10% of ALS cases.7-10,16 Therefore, we could not predict the slow clinical course during the early stages. However, BAD became a consideration as the pattern of weakness remained restricted over time. For example, had we set our inclusion criteria at 1 year instead of 18 months, we would have included just one more patient, who also had a lower motor neuron syndrome but developed neck, bulbar, and respiratory weakness before 18 months. Other features that are relatively atypical of ALS may point toward BAD at an even earlier stage. ALS usually presents with weakness that is more prominent in distal muscles,6,17,18 whereas onset of weakness in the shoulders occurs in less than 3% of individuals.19 This can be contrasted with the patients presented here who typically first noticed proximal arm weakness. In addition, 9 of our 10 patients had no pyramidal tract involvement to confirm the diagnosis of ALS.18,20 Finally, our EMGs demonstrated a neurogenic process confined to cervical innervated musculature in 8 of 10 patients.

In 1957, Mulder10 used the term “hanging-arm syndrome” to describe ALS patients whose clinical presentation was similar to what was seen in this series. Hu et al.16 recently called attention to 39 patients from England presenting with severe bilateral proximal arm weakness. These accounted for 10% of their motor neuron disease population and were described by the term “flail arm syndrome.” Some patients had disease durations as short as 6 months or minor weakness in regions outside of the arms at presentation, probably accounting for the higher incidence than was seen here. Defined this way, more than three quarters had upper motor neuron signs, and more than one-half developed bulbar weakness during a follow-up period of 2 years. Patients with the flail arm syndrome had a median survival of 58 months, which represented a trend toward improved survival compared with their overall ALS population. In contrast, after a mean follow-up of 67 months, 7 of the 10 patients reported here failed to develop weakness outside of the upper limbs, and 3 developed only mild weakness in other regions. In addition, by selecting patients with the slowest progression, we have distinguished a subset that almost always lacks pyramidal signs. We think that flail arm syndrome, as defined by the British authors, most likely includes both patients that appear to have classic ALS and a smaller subset with BAD who may account for the overall improved survival.

The term “progressive muscular atrophy” (PMA) refers to the subset of patients with lower motor neuron syndromes that otherwise resemble ALS.21 The first descriptions of this disorder are generally credited to Aran22 and Duchenne23 in the mid-19th century. Two decades later, Charcot and Joffroy24 distinguished two forms of motor neuron disease now considered to be ALS and PMA. More recent estimates from ALS centers suggest that anywhere from 4 to 33% of cases present with a purely lower motor neuron disorder.25-27 This wide range most likely represents a selection bias based on clinical definitions and the interpretation of findings on clinical examination. Although patients with PMA often have clinical courses typical of ALS, with progressive weakness resulting in respiratory failure and death,26,27 natural history studies suggest that, overall, patients who lack pyramidal signs have a greater probability of prolonged survival.21,26-28 BAD may represent a specific phenotype within PMA that further aids in predicting prognosis.

BAD shares some features with monomelic amyotrophy.29,30 Both disorders result primarily from a loss of cervical motor neurons and have rapid progression of weakness in the initial phase followed by slowing or stabilization.29,31,32 Bilateral arm involvement occurs in 30% of patients with monomelic amyotrophy, and, like our patients, a minority will have upper motor neuron signs29 or electrophysiologic evidence of denervation in noninvolved limbs, substantiating a more diffuse process.32 BAD clearly differs from monomelic amyotrophy by the adult onset and the largely symmetric and proximal pattern of weakness. In addition, a number of our patients eventually developed weakness in the legs.

Our patients do not conform to previous reports of hereditary adult-onset proximal muscular atrophies, where weakness in distributions outside of the upper limbs is the rule.33-36 We found two descriptions of adult-onset familial motor neuron disorders presenting with upper-limb paresis,37,38 but patients from both families had rapidly progressive courses with early respiratory failure.

Multifocal motor neuropathy may also cause bilateral upper-extremity weakness.39-41 However, the degree of proximal muscle involvement, absence of antibodies directed against GM1, and lack of demyelinating features weigh against this diagnosis. Pestronk et al.42 described the “proximal greater than distal” lower motor neuron (P-LMN) syndromes, which may have similarities to BAD, although they did not give detailed clinical features. These are slowly progressive disorders most commonly presenting in the upper extremities with late-age onset, male predominance, asymmetry, and absence of conduction block. Thirty percent have selective serum antibody binding to GA1, implicating an immunologic basis, although there is no evidence that the P-LMN syndromes respond to immunomodulating therapy. These antibodies were negative in our patients.

The clinical history generally excluded secondary disorders that potentially cause the neurogenic man-in-the-barrel syndrome, including spinal cord infarction11 and postradiation myelopathy.15 Cervical spondylosis can also rarely present with pure upper-limb weakness, but the diagnosis was ruled out by MRI.13,14 Some patients with juvenile-onset, monomelic amyotrophy develop MRI abnormalities with neck flexion, including abnormal motion of the posterior wall of the dural sac, an increased posterior epidural space, and flattening of the spinal cord.43 These disorders were not present in our patients, but point out that unsuspected mechanical injury or chronic ischemia can selectively injure cervical anterior horn cells.