As mentioned in the previous article (February 2004),
electromyography (EMG) alone is usually not sufficient in the
investigation of neuromuscular disorders and further testing
is warranted.In addition to EMG, measurement
of nerve conduction
velocities (NCV) including
both motor (MNCV) and sensory (SNCV), should be performed. Additional
testing including late responses (F waves, H waves and A waves) and repetitive nerve
stimulation (RNS) will be covered in a future article . Indications for
performing these tests are identical to those given previously
for EMG and can also be beneficial in determining the level
and extent of involvement.
The definition given by the American Association of
Electromyography and Electrodiagnosis (AAEE) for nerve conduction studies (NCS) is as
follows, “recording and analysis of electric waveforms of biologic origin elicited in response to electric or
physiologic stimuli.” In this
article, only electrical stimulation will be covered. Though the techniques may
vary between labs, the principles are the same. Ideally, each lab should work out normal reference values for individual
nerves in each species.
nerve conduction velocity testing (MNCV) involves the
placement of a pair of recording electrodes (either needle
or surface) near a muscle innervated by the nerve of interest
and at least two pairs of stimulating electrodes (usually
needles) along the course of that nerve. A ground electrode is used to minimize interference. The active recording electrode is placed subcutaneously
over the motor point of the muscle with the reference electrode
located several centimeters distally. This placement insures
the optimal configuration of the M wave (a type of compound
muscle action potential, or CMAP) with an initial negative
(upwards) deflection (Figs
1A and 1B). M waves are the result of orthodromic propagation of action potentials down the nerve, acetylcholine
release at the neuromuscular junction and myofiber depolarization. Pairs of stimulating electrodes are placed
at two or three sites along the nerve, ideally with the cathode
and anode straddling the nerve (either deep/superficial or rostral/caudal) and equidistant
from the recording electrodes. Placement can be “fine tuned”
while stimulating by making tiny adjustments in the position
of the active electrode (for recording) or the cathode (for
stimulating) until the desired M wave appears or only a small
amount of current (or voltage) is needed to elicit a response,
respectively. A supramaximal stimulus is applied to record the actual potential of interest. This is calculated by finding the maximal stimulus,
the current value at which M wave amplitude is at its highest
and further increases have no effect, and multiplying it by
1.5 (e.g. if maximal is 1.2 mA, supramaximal would be 1.8 mA).
This process is repeated at the other stimulation sites, though
the recording electrodes must not be moved once their optimal
placement has been established. M waves resulting from stimulation at each site should
look similar, if not, another nerve may be involved (a problem
especially when stimulating at the hip where the peroneal and tibial segments of the sciatic
n. may be difficult to isolate). Latencies are determined for each potential by marking
the point at which the baseline deflects in an upward direction
and amplitudes can be measured by marking the baseline and
negative peak (or some authors use negative to positive peak).
Distances, measured between the tips of the stimulating electrodes,
are entered in to calculate the nerve conduction velocities.
It is this difference in distance over the difference in latency
that determines the MNCV, expressed in meters/second. Multiple
stimulation sites are required to remove the synaptic delay
and muscle depolarization times from the equation. Various
peripheral nerves lend themselves to this technique, such
as the peroneal, tibial, ulnar and radial. Normally M waves are recorded
in the millivolt range, but with
severe disease, it may be necessary to increase the sensitivity
into the microvolt range. In addition, signal averaging may
be indicated. Very
high stimulus intensities may also be required. It is important
to keep the patient warm during these studies, as a drop in
limb temperature will cause a decrease in the conduction velocity
(CV). Technical error must be ruled out before attributing
findings to a disease process.
Fig 1A. Normal feline MNCV following stimulation
of the ulnar n. at the carpus and elbow, recorded from the interosseus m. Note:
the upper value is not a true MNCV as it includes the synaptic
delay and myofiber depolarization time (the terminal velocity).
Fig 1B. Normal canine MNCV following stimulation
of the peroneal n. at the hock, stifle and hip, recorded from the extensor digitalis brevis lateralis m. The upper CV is
the terminal velocity.
When evaluating MNCV
tests, several M wave attributes should be considered. Configuration, duration, amplitude and latency provide important diagnostic information. Dispersion can change a normally biphasic potential into one that
is polyphasic (Fig.
2A). The duration, or time between the M wave’s initial
baseline deflection and the time it takes to return to baseline,
can be prolonged. These changes are indicative of a demyelinating process. A
decrease in M wave amplitude is a non-specific finding in
that it can be seen in neuropathies or myopathies (those involving the muscle used to record the M wave). In addition, the loss of as few as 2 consecutive
internodes of myelin can result in conduction block of a neuron, so diminished amplitude can be seen
with demyelination also. This phenomenon (conduction block) can be seen
in other situations: 1) when metabolic insults occur in a
nerve with preservation of axons (neurapraxia)
or 2) prior to the completion of wallerian degeneration in a recently injured axon (axonotmesis).
This is apparent when comparing M waves recorded after stimulation
at different sites. Distal sites may illicit relatively normal responses
that change (often dramatically) upon stimulation at proximal
sites (Fig. 2B). Lastly, latency is necessary to calculate the
conduction velocity. When
evaluating CVs, a patient’s age must be considered. Reference values are slower in young animals and those
that are elderly. Extreme
slowing of MNCV is another indication of loss of myelin. In
many cases results are mixed and it is not always possible
to determine whether myelin changes are primary or secondary
to neuronal disease. Additional
tests (EMG, biopsy) are helpful. Time is also an important consideration. MNCVs, as well as EMG and biopsy findings, can be within
normal limits early in the course of a disease.
Fig 2A. Abnormal canine ulnar MNCV. M waves are dispersed,
low amplitude and CVs are very slow (50% of normal).
These changes are suggestive of demyelination. Diagnosis:
multifocal polyneuropathy (left thoracic limb spared) with
motor and sensory involvement. The dog was previously diagnosed
with malignant melanoma.
Fig 2B. Abnormal canine peroneal MNCV. M wave resulting from distal
stimulation is slightly smaller than expected while those
from the proximal stimulation sites are markedly decreased.
The lack of dispersion and relatively normal CVs (for this
geriatric dog) are consistent with an axonopathy and not demyelination. Diagnosis:
generalized polyneuropathy with motor and sensory involvement of undetermined
Sensory nerve conduction velocity testing (SNCV) is the stimulation
of a sensory branch of a nerve and the recording of the conducted
volley along the course of that nerve. The recording is made directly from the nerve, so a
single site can be used to determine the velocity. As these potentials are only in the microvolt range,
signal averaging is required. Depending on the recording conditions, and neuromuscular
status of the patient, hundreds to thousands of individual
responses may be necessary. Low electrode impedance is needed to minimize interference. Fortunately, the potentials at several sites
can be recorded simultaneously, limited only by the number
of amplifiers available in a given system. For most peripheral
nerves, electrodes previous placed for stimulating the motor
nerve can be used to record the sensory potential, thus optimal
placement (as discussed above) will insure the best response
(Fig. 3A). Stimulating electrodes are placed distally on
either side of a sensory branch of the nerve. SNCVs are routinely recorded from peroneal, ulnar and radial nerves. The tibial nerve can
also be studied but most techniques involve stimulation at
a site where the nerve is mixed, containing both motor and
sensory fibers. Spinal cord dorsum potentials (CD) and somatosensory evoked responses (SEP) can also be recorded
3B). This latter term (SEP) can be used collectively
for both peripheral and central nervous system recordings
following peripheral sensory nerve stimulation. CDs are the result of depolarization of interneurons in the brachial or lumbar plexus and can provide information
regarding sensory input to the spinal cord (especially helpful
in cases with suspected nerve root avulsion).
Fig 3A. Normal canine SEPs following stimulation
of the radial n. at the level of the 4th distal metacarpus, recording
from the elbow, C7 (CD), C1 and off the head.
Fig3B. Normal canine SEPs following stimulation
of the peroneal n. at the level
of the 4th distal metatarsus, recording from the
hock, stifle, hip and L4/5 (CD). Note: although the potentials
appear to be the same size, the sensitivity increases from
top to bottom (see amplitude values at right).
The same analysis applies to both the SNCV and motor MNCVs,
although the sensory nerve action potential (SNAP), is often polyphasic. Temporal dispersion increases with distance
as the result of variable conduction velocities in the different
populations of sensory neurons. A gradual reduction in amplitude is also seen as the
result of phase cancellation between these neurons (Fig.
3B). Excessive dispersion and slow CVs are again
suggestive of demyelination, whereas
a loss of only amplitude suggests either an axonopathy,
if segmental, or a neuronopathy,
if all sites are affected (Figures 4a and 4b).
Fig 4A. Abnormal canine radial SEPs. The SNAP is low amplitude and dispersed,
but the SNCV is normal (indicating sparing of the fastest
fibers). CD is low amplitude. No responses were recorded at
C1 and from the head. Diagnosis: generalized polyneuropathy with primarily sensory involvement (all MNCVs were within normal limits but mild denervation atrophy was found in the quadriceps m. biopsy). This dog also had a C5-C7 myelopathy (cord compression) that could explain the lack
of SEPs rostrally.
Fig 4B. Abnormal canine peroneal SEPs. All three SNAPs are severely reduced in amplitude and SNCVs are slow. Dispersion is difficult to evaluate, as potentials
are barely distinguishable from background noise (despite
averaging 4000 responses). Diagnosis: severe sensory polyradiculoneuropathy. The
acute nature (the study was performed 4 days after the dog
became acutely non-ambulatory) could preclude the detection
of motor involvement (MNCVs, EMG and muscle and nerve biopsies were normal). A toxic
cause was suspected.
Cuddon PA. Electrodiagnosis in Veterinary Neurology: Electromyography, Nerve Conduction Studies,
and Evoked Responses. Loveland, CO.
Dumitru D. Electrodiagnostic Medicine 2nd editon,
Hanley and Belfus, Inc. Philadelphia, PA 2002.
Kimura J. Electrodiagnosis in Diseases of muscle and Nerve:
Principles and practices 3rd edition. Oxford University
Press, New York, New York, 2001.