Professor of Clinical Neurology, Weill Cornell Medical College
Dr. Rubin reports he is a consultant for Merck Sharp & Dohme Corp.
SYNOPSIS: A detailed analysis of electromyography features showed a high correlation with muscle pathology. However, pathologic changes on muscle biopsy may be present even with a totally normal electromyogram.
SOURCE: Sener U, Martinez-Thompson J, Laughlin RS, et al. Needle electromyography and histopathologic correlation in myopathies. Muscle Nerve 2018; Nov. 10. doi: 10.1002/mus.26381. [Epub ahead of print].
Aside from serum creatine kinase and aldolase, the laboratory investigation of suspected myopathy often includes needle electromyography (EMG) and muscle biopsy. Is there a correlation between specific findings on EMG and pathological changes on muscle biopsy that might allow the electromyographer to predict specific pathologic changes on biopsy, based on EMG abnormalities?
In this retrospective study, Sener et al reviewed records of 224 consecutive patients seen for muscle biopsy at Mayo Clinic sites in Jacksonville, Florida, and Rochester, Minnesota. Inclusion criteria required that the EMG be performed prior to the muscle biopsy, and that the muscle biopsied was studied by EMG but was contralateral to the one examined. On review of EMG studies, the authors specifically noted the presence of fibrillation potentials, myotonic discharges, and short duration motor unit potentials. For muscle biopsy, they specifically noted the presence of inflammation, necrosis, fiber splitting, and vacuolar changes. Positive and negative predictive values (PPV and NPV, respectively) were calculated for EMG findings with respect to each pathologic change. PPV was defined as the likelihood that an abnormal finding on EMG would predict a specific pathologic finding on biopsy, and NPV was defined as the likelihood that the absence of an EMG finding would predict the absence of a pathologic finding on biopsy.
Among 224 charts reviewed, six were excluded, as their EMG did not precede their biopsy, leaving 218 patients, 109 each male and female, with a mean age of 54.4 years, for analysis. Of these, 178 (82%) were thought to have EMG findings consistent with myopathy, which was confirmed following biopsy in 143 (80.3%), including 60 with inflammatory myopathy, 18 with muscular dystrophy, eight with congenital myopathy, two with mitochondrial myopathy, 36 with no specific etiology, and 19 miscellaneous. Among the 35 patients with EMG findings suggestive of myopathy but not confirmed on biopsy, diagnoses included hyperCKemia, denervation atrophy, motor neuron disease, neuromuscular junctionopathy, Guillain-Barré syndrome, polyradiculopathy, other neurogenic process, and no final confirmed diagnosis. EMG sensitivity for a confirmed diagnosis of myopathy was 95.3%, with a specificity of 48.5%.
Short duration motor unit potentials were highly sensitive for myopathy but not specific for any particular pathologic change, whereas myotonic discharges were specific, but not sensitive, for pathologic change. Fibrillation potentials were indicative of inflammation, fiber necrosis, splitting, or vacuolar change, with a sensitivity of 65-74% and a specificity of 58-81%, regardless of the presence of short-duration motor unit potentials. Short motor unit potentials and fibrillation potentials had high NPV for inflammation, fiber type splitting, and vacuolar change, whereas PPV was low for any specific EMG abnormality.
Among 24 EMG studies reported as normal, five had myopathy on muscle biopsy, including two cases of carnitine palmityl transferase II carriers and one each of muscular dystrophy, McArdle disease, and dermatomyositis. Another five patients with normal EMG were diagnosed after muscle biopsy with alternative diagnoses, including denervation atrophy, remote polio, compartment syndrome, and central sensitization syndrome. In the final 14 with normal EMG, no alternative diagnosis was made.
In a concomitantly published, almost identical study of 100 patients at Mayo Clinic, Rochester,1 researchers confirmed and extended the above findings. Fibrillation potentials correlated with atrophic and regenerating fibers, fibers reacting for nonspecific esterase, fibers with congophilic inclusions, and increased endomysial connective tissue. Long-duration motor unit potentials correlated with fiber-type grouping, while short-duration motor unit potentials correlated with atrophic, necrotic, and regenerating fibers, and increased endomysial connective tissue. Increased phases of motor unit potentials correlated with atrophic fibers, increased endomysial connective tissue, and fibers reacting for non-specific esterase, whereas increased turns correlated with atrophic and regenerating fibers, increased endomysial connective tissue, and target formations. Early recruitment correlated with regenerating fibers, perimysial inflammation, and increased endomysial connective tissue. Further examination of muscle biopsy and EMG correlation should look into the combination of EMG findings predicting specific pathology.
- Naddaf E, Milone M, Mauermann ML, et al. Muscle biopsy and electromyography correlation. Front Neurol 2018;9:839.