Electromyography (EMG) accuracy compared to muscle biopsy in childhood

[Guest guest]

J Child Neurol. 2007 Jul;22(7):803-8.

Electromyography (EMG) accuracy compared to muscle biopsy in

childhood.

Rabie M, Jossiphov J, Nevo Y.

Neuropediatric Unit, Hadassah Hebrew University Medical Center,

Mount Scopus, Jerusalem 91240, Israel.

Reports show wide variability of electromyography (EMG) in detecting

pediatric neuromuscular disorders. The study's aim was to determine

EMG/nerve conduction study accuracy compared to muscle biopsy and

final clinical diagnosis, and sensitivity for myopathic motor unit

potential detection in childhood. Of 550 EMG/nerve conduction

studies performed by the same examiner from a pediatric

neuromuscular service, 27 children (ages 6 days to 16 years [10

boys; M:F, 1:1.7]) with muscle biopsies and final clinical diagnoses

were compared retrospectively.

Final clinical diagnoses were congenital myopathies (5 of 27,18%),

nonspecific myopathies (biopsy myopathic, final diagnosis uncertain;

6 of 27, 22%), congenital myasthenic syndrome (3 of 27, 11%),

juvenile myasthenia gravis (1 of 27, 4%), arthrogryposis multiplex

congenita (2 of 27, 7%), hereditary motor and sensory neuropathy (1

of 27, 4%), bilateral peroneal neuropathies (1 of 27, 4%), and

normal (8 of 27, 30%).

There were no muscular dystrophy or spinal muscular atrophy

patients. EMG/nerve conduction studies had a 74% agreement with

final clinical diagnoses and 100% agreement in neurogenic,

neuromuscular junction, and normal categories. Muscle biopsies

concurred with final diagnoses in 87%, and 100% in myopathic and

normal categories. In congenital myasthenic syndrome, muscle

biopsies showed mild variation in fiber size in 2 of 3 children and

were normal in 1 of 3. EMG sensitivity for detecting myopathic motor

unit potentials in myopathies was 4 of 11 (36%), greater over 2

years of age (3 of 4, 75%), compared to infants less than 2 years (1

of 7, 14%), not statistically significant (P = .0879). EMGs false-

negative for myopathy in infants < 2 years of age were frequently

neurogenic (3 of 6, 50%). In congenital myopathies EMG detected

myopathic motor unit potentials in 40%, with false-negative results

neurogenic (20%) or normal (40%). Because our study has no

additional tests for active myopathies, for example Duchenne

muscular dystrophy genetic testing, our sensitivity for myopathies

is lower than if we used a more global view.

In conclusion, EMG detection rate of myopathic motor unit potentials

at a young age was low, improving in children over 2 years of age.

In neurogenic and neuromuscular junction disorders, the EMG has a

very high detection rate. In children with mild to moderate

neurogenic EMG findings and normal nerve conduction, a myopathy

should always be considered.