What is the role of EMG and NCS in the workup of Charcot-Marie-Tooth (CMT) disease?

Updated: Feb 19, 2019
  • Author: Francisco de Assis Aquino Gondim, MD, PhD, MSc, FAAN; Chief Editor: Nicholas Lorenzo, MD, MHA, CPE  more...
  • Print


Compared with acquired neuropathies, CMT1 is typically associated with diffuse and uniform conduction slowing. Nerve conduction is stable and secure, so conduction block or dispersion is rare, contrasting with acute or chronic inflammatory demyelinating polyradiculoneuropathies. Little difference exists between proximal and distal findings on nerve conduction studies (NCS). F-wave responses are usually prolonged and EMG shows evidence of denervation. Brainstem auditory evoked potentials can demonstrate delay in wave 1.

Median motor nerve conduction velocities are below 38 m/s in CMT1 and above in CMT2, although some studies have proposed a cut-off of 42 m/s. Nerve conduction studies in CMT2 typically reveal mild slowing, with median nerve velocities above 38 m/s, reduced CMAP amplitudes to less than 4 mV, and reduced sural nerve sensory action potential (SNAP) amplitudes to less than 10 mV. Sural nerve sensory responses can be absent. Phrenic CMAP also shows reduced amplitudes. EMG reveals signs of chronic denervation.

Although the separation of neuronal and nonneuronal forms is an important etiologic and pathogenic distinction, even in CMT1 the clinical deficits appear to correlate better with progressive axonal degeneration than with slowed nerve conduction velocities (see below). This is not surprising because demyelination disturbs axonal structure and transport.

The distinction between demyelinating and nondemyelinating CMT is not always clear. Relatively normal nerve conduction velocities have been reported in younger members of a family with an MPZ mutation, whereas older relatives had severely slowed nerve conduction velocities. Conduction values are symmetric in CMT1, and few differences exist between proximal and distal nerve segments. Nerves often are refractory to stimulation or require higher amplitude and prolonged stimulation. Sensory nerve conduction velocities in all forms of CMT1 are reduced and often unrecordable. Sensory loss correlates with median sensory nerve conduction velocities and CMAP amplitudes.

In one large study of people with CMT1A, mean motor nerve conduction velocities in the median and peroneal nerves were 20 m/s (5-34) and 17 m/s (10-20). Responses in the legs are often absent because of complete denervation of small foot muscles. In other studies, nerve conduction velocities have ranged from 10-42 m/s, again illustrating that the distinction of CMT1 and CMT2 cannot rest on NCS findings alone. In general, nerve conduction velocities are slower in CMT1A because of PMP22 point mutations versus duplication. In a study of 42 patients with CMT1A, weakness correlated with axonal loss as measured by CMAP amplitudes but not with nerve conduction velocities. This suggested that disability results from loss or damage to large-caliber motor and sensory axons.

In CMT1B, a similar nerve conduction velocity range of 4-59 m/s has been documented, although typical nerve conduction velocities are 5-15 m/s. In a study of a single pedigree, nerve conduction velocities were significantly slower than in CMT1A patients, whereas in a comparison of 119 CMT1A and 10 CMT1B patients, no differences were found.

In CMT1D due to EGR2 mutations, the nerve conduction velocity range is 16-41 m/s.

In CMTX1 due to CX32 mutations, NCS findings are more variable and axonal features are more common. Asymmetry may be prominent, and conduction block and dispersion have been observed. The nerve conduction velocity range is 25-43 m/s in men and 31-50 m/s in women. Not surprisingly, because of the CNS expression of CX32, subclinical CNS involvement has been documented in some patients who have abnormal visual, motor, and brainstem auditory evoked potentials. MRI abnormalities have been reported as well.

In children, NCS findings are normal at birth, except for children with CHN and DSS. As the PNS matures, abnormal nerve conduction velocities develop and then are stable for life. Changes are fully manifest at age 2-4 years, even in asymptomatic patients.

In HNPP, background polyneuropathy independent of superimposed entrapment neuropathy, which becomes more prevalent with age, is typically present. The variability within families may be considerable. In a recent series of 99 patients with a PMP22 deletion, a multifocal polyneuropathy with diffusely increased distal motor latencies (DML) was typical, with more normal motor conduction velocities, diffuse reduction of sensory nerve action potentials, and multiple instances of focal slowing at anatomic entrapment sites. These features, including focal slowing, were also observed in several patients with a clinical CMT rather than an HNPP phenotype; this indicates that NCS findings suggest PMP22 deletion, even when the clinical features do not suggest this. Neurophysiologic findings were similar in oligosymptomatic and asymptomatic patients and became characteristic as early as the second decade of life.

EMG findings are normal in proximal muscles but may show distal changes with increased duration and amplitude motor unit potentials. Signs of active denervation such as increased insertional activity and fibrillation potentials are not prominent in muscles unaffected by weakness. Diffusely slow sensory nerve conduction velocities independent of nerve entrapment were found in another study, consistent with a background dysmyelinative polyneuropathy. Slowed motor conduction was less common in HNPP, although DML were frequently prolonged, indicating that a distal motor polyneuropathy is present, similar to that observed in immunoglobulin M (IgM) monoclonal gammopathy against myelin-associated glycoprotein or sulfated glucuronyl paragloboside.

HNPP may be characterized electrically by a profile of slow nerve conduction velocities in most sensory nerves, relatively less frequent and more minor motor slowing, prolonged DML, and F-wave latencies. Other authors proposed diagnostic criteria for the disorder based on bilaterally delayed median DML, slowed median sensory nerve conduction velocities at the wrist, and prolonged DML or motor conduction slowing in the peroneal nerves. Bilaterally normal median DML and sensory nerve conduction velocities at the wrist appear to exclude HNPP.

In HNA, evidence for a generalized neuropathy is absent.

Did this answer your question?
Additional feedback? (Optional)
Thank you for your feedback!