|
|
|
        |
NERVE CONDUCTIVE STUDIES
Erik STÅLBERG , Hacer ERDEM
Department of Clinical Neurophysiology, Uppsala University Hospital , Uppsala , SwedenABSTRACT
A review of the principles of motor and sensory conduction studies is presented in this article. These studies are important in the neurophysiological evaluation of patients with nerve muscle disorders. A thorough knowledge of pathophysiology, methodology and analysis of pertinent parameters together define the quality of such an investigation. The review deals with these aspects and contains updated knowledge of the parameters that may be useful in routine work. Examples of the findings in some diseases are also given.
Key-words: motor conduction studies, sensory conduction studies, conduction block, neuropathy, neurography BackgroundThe myelinated nerve axon conducts impulses in a saltatory fashion i.e. depolarisation occurs at the nodes [34]. The currents are prevented from penetrating the membrane between the nodes in the normal nerve due to an isolating myelin sheath. This means that the impulse propagation is much faster than if there was a continuous depolarisation. The conduction velocity is also dependent on the axonal diameter and the properties of the membrane [36]. A normal axon conducts with a speed of 35-60 m/sec [29]. The velocity is reduced if the myelin is defect due to pathological changes, if the ion-channels at the nodal areas are blocked or if the axon diameter is smaller than normal. It is also dependent on temperature.
When the correlation between neurophysiological and morphological parameters was established, the nerve conduction study became an important method used in clinical routines [14;15;18;20;26;31]. Such studies are performed in most EMG laboratories since the 1960’s and have since then become more sophisticated, sensitive and specific.
This summary will give a brief update of the nerve conduction studies, which are performed in clinical routine today.
Factors influencing nerve conduction parameters
The temperature affects the conduction velocity (CV), both locally at the recording site and generally along the nerve.
Locally the amplitude increases as the temperature in the recording site decreases. The amplitude increases by 1.7% per degree Celsius [16].
The temperature also affects the conduction along the nerve segment. The CV decreases as it cools with a factor ranging from 1.2 to 2.4 m/s per degree Celsius. This varies for different nerve [8]. This will reduce the amplitude. These two effects of the temperature on the amplitude neutralize each other. In order to standardize CV and amplitude measurements, it is recommended to keep the skin temperature at above 29° C for the dorsum of the hand and 27° C for the dorsum of the foot [12].
Conduction velocity is age dependent. Full term infants have conduction velocities, which are approximately half of that seen in adults. Conduction velocities rapidly increase from the values recorded in infants to near adult values at around 3-5 years of age. Furthermore pre-term infants have slower values at around 14-28 m/s. In the teens conduction velocities are almost the same as those of adult values [13;21;29].
After the second to fourth decade, conduction velocities start to decrease very slowly. CV decrease by 0.5-1.8 m/s for each decade [3;4]
Longer nerves generally conduct more slowly than shorter nerves [5]. It has been shown that there is a good correlation between CV and estimated axonal length in the peroneal and sural nerves, but not in the motor or sensory fibers of the median nerve. Based on a good correlation between the height of the patient and the length of the nerve, the CV in lower limbs decreases by 2-3 m/s for10 cm increase in height [12] (Fig 1 ).
Nerve impulses propagate faster in the proximal than in the distal nerve segments [14].
Fig1. Correlation between MCV and height in median and peroneal nerves and 95% confidence limits.
GenderIt has been reported that CV is slower in women than that in men [19;32], but the correlation is complex since gender and height are not independent of each other [12]. In our routine, we use the same reference values for women and men.
Reference values
With standardized methods, the technique is sufficiently reproducible to allow the transfer of reference values from one laboratory to the other. A number of techniques and related reference values are given in the literature [12;26].
The principle changes in nerve function are related to demyelination, axonal degeneration and conduction block. There are no absolute dividing lines between these situations; they show some overlap and also dynamic changes from one stage to another due to the interaction between Schwann cells and the axonal condition. In cases of demyelination, the conduction velocity is reduced. In cases of axonal degeneration, there may be normal velocity in the remaining axons, but a weaker muscle response is evoked. In cases of conduction block, no axonal degeneration occurs and therefore a normal response is obtained when stimulating distal to the lesion. When stimulation is performed proximal to the site of abnormality, a reduced number of axon conducts impulses, and a smaller than normal muscle response is obtained (Fig 2 ).
Fig. 2 Schematic explanation of findings in MCS in normal, demyelination, axonal degeneration and conduction block. The amplitude and latency of the evoked CMAP is symbolized by the triangles. Note the various combinations of CMAP at distal and proximal stimulation in different types of pathology.
Motor Conduction Studies
The procedures can be standardised, but still some differences exist between laboratories. The principles used in our laboratory have been summarised elsewhere [12], and is the basis for the description given in this review.
For motor conduction studies (MCS), recording is formed over the belly of the muscle using a surface electrode (Fig ). The active recording electrode is placed over the endplate zone of the muscle in order to record muscle activity at the moment of depolarisation after the nerve impulse has arrived at the endplate. The muscle response, obtained after nerve stimulation is called the compound muscle action potential, CMAP, should have an abrupt negative take-off. If the electrode is away from the endplate, it will show an initial positive phase corresponding to the approaching electrical field of the impulses in the individual muscle fibers. The start of the positive phase corresponds to the start of the depolarisation. However, because of the gradual increase in amplitude from the baseline it may be difficult to determine the exact start. If later part of the CMAP is used for latency measurements, the latency value will be contaminated with some conduction time along muscle fibers and not express only the nerve conduction time. However, when comparing distal and proximal latencies for CV calculations, other points than the take-off can be used, as long as the same parts of the CMAPs are used. It should be mentioned that in some situations, particularly in recording from the abductor digiti minimi (ADM) muscle of the foot in the study of the tibial nerve, underlying muscles with different positions of the endplate zones would contribute to the recording and give rise to positive components. If possible, measurements should be made to the point where the signal leaves the baseline.
Fig.3 MCS of the median nerve. Recording electrodes are shown over the thenar eminence and stimulating electrodes are indicated (black dots). Plexus stimulation is usually not used for this nerve, since the ulnar nerve is also stimulated. The evoked CMAP is therefore the combined response from all the thenar muscles.
The reference electrode should be placed in such a way that no recordings are taken from the muscle under study. If the reference electrodes are too close to the muscle, e.g. in its tendon, it will contribute with a significant amount of activity. Therefore, more distal positions are preferred, e.g. over the distal interphalangel joint (from the thumb for the median nerve, dig V for the ulnar nerve, toe one for the tibial nerve and toe V for the peroneal nerve).
Usually stimulation is performed at two or more sites along the nerve. It is not adequate to stimulate at only one point and calculate the conduction velocity from the obtained latency between the stimulus and motor response and the distance. This is because the conduction time includes the slower in the conduction in the last segment of the nerve, time in the neuromuscular junction, and possibly some conduction along muscle fibers. Therefore, two stimulation sites well separated from each other along the nerve are used.
In situations of local nerve lesion, short segments should be tested in order to localise the site of the abnormality. This technique has been called “inching” using small distance between stimulation points. We often use distance of 10 mm (“centimetring”). The recording site is kept constant, and the stimulation is performed every 10 mm across the area of suspected lesion. The analysis should focus on sudden jumps with prolonged latency values with more proximal stimulation or abrupt drop in amplitude.
Sometimes somewhat longer segments of the nerve are studied as part of the routine. In ulnar nerve studies, stimulation is often made at the wrist, and above and below the elbow. In studies of the peroneal nerve, stimulation is made at the ankle, and below and above the fibula head. In cases of difference in conduction along the two segments, centimetring is performed.
Generally surface electrodes (felt pad or steel) are used in motor nerve conduction studies. They are fixed on to a plastic bar a fixed distance apart. The electrodes used in our routine (Medtronic 13L36) have an interpole distance of 23 mm.
In a few situations it is preferable to use needle electrodes. A pair of monopolar needle electrodes is used these cases. Alternatively a surface electrode is used as an anode.
It is necessary to use needle electrodes when stimulating a nerve that is located very deeply. Therefore the muscle response can be obtained with less stimulus strength than the one necessary in surface electrodes. In short segment study the use of needle electrode provides more certain localisation than that of surface electrodes.
The output impulse used for MCS is a rectangular wave with a duration of 0.1 or 0.2 ms. Sometimes it may be necessary to increase the stimulus duration to 0.5 or 1 ms in order to get maximal amplitude. In order to ascertain a reliable maximal amplitude of the CMAP, it is advisable to increase the stimulus strength by 10-25% of that which is necessary to obtain maximal amplitude [12]. In some situations a biphasic stimulus pulse is used in order to suppress stimulus artefacts [25].
A number of parameters are of interest in the assessment of different aspects of motor nerve functions (Fig 4).
Fig.4 Measured parameters of the CMAP. Distal (trace 1) and proximal (trace 2) stimulations are shown
The latency is the time between the stimulus and the response. In motor nerve studies, this latency includes the nerve conduction time and also the neuromuscular transmission time.
Distal latency (DLAT) is measured from the distal stimulation point to the first deflection from the baseline.
Proximal latency (PLAT) starts at the proximal stimulation point and ends at the first deflection from the baseline.
The amplitude (AMPL) of the evoked motor response carries important information. It is dependent on the number of axons that conduct impulses from the stimulus point to the muscle, the number of functioning motor endplates and the muscle volume. The amplitude is measured from the baseline to the negative peak.
The area represents a combination of the amplitude and the duration. It therefore reflects the number and synchrony of the muscle fibers activated. A prolongation of the duration can cause a decrease in the amplitude and may be misinterpreted as a conduction block. In this situation there may not be significant difference in the area. Area is the integrated area between the CMAP and the baseline.
The duration (DUR) reflects the synchrony of individual muscle fiber discharges. If there is a significant difference in the conduction velocity among nerve fibers, the duration will be prolonged. This is mainly related to the range of the conduction velocities of the large myelinated fibers. Duration is measured from the onset to the first negative to positive baseline crossing.
Fig.5 CV measurements in two segments in the ulnar nerve; below elbow (traces 1-2) and across elbow (traces 2-3). Note somewhat lower amplitude and longer duration with proximal stimulation.
Calculated parameters ( Fig 5)
The conduction velocity (CV) is calculated by dividing the length of the nerve segment between the two stimulation points by the difference between the proximal and distal latency [12]. In this way the slow distal conduction and any delay in the neuromuscular transmission is eliminated. It is calculated as follows.
CV (m/s) = distance (mm) / LATprox-LATdistal
When motor conduction velocity is calculated in this way it reflects the fastest motor axons.
Since nerve fibers have different conduction velocities, a more proximal stimulation site will give an increased duration of M wave. The change in duration with a proximal stimulation site is called temporal dispersion and is calculated as follows:
DISPERSION = 100 X (DURprox - DURdistal) / DURdistal
In healthy subjects, the maximum dispersion in the ulnar nerve is 10-15% [2;23;27]. In long nerve segments the CV may be lower and the dispersion higher than that seen in short segments [35].
With proximal stimulation, when the duration of the M wave gets longer due to the temporal dispersion, the amplitude and the area of the M wave changes (Fig 5).
DECAY is calculated as shown in these formulas.
AMPLDECAY =100 X (AMPLdistal AMPLprox) / AMPLdistal
AREADECAY =100 X (AREAdistal AREAprox) / AREAdistal
In healthy subjects, the mean value of the AMPLDECAY is 4.5-6.2% in the ulnar nerve [27;35], and 5.6-7.7% in the median nerve [11]. The AMPLDECAY is larger in the lower extremities than in the upper extremities. The peroneal nerve has a mean value of 11% and an upper limit of 29% in the AMPLDECAY. On the other hand the AREADECAY is smaller than the AMPLDECAY in peroneal nerves [35].
F-waves (Fig 6 ) When the motor nerve is stimulated, nerve action potentials propagate both in the distal direction to evoke a muscle response, and in the proximal direction as a non-physiological event. Occasionally, the motor neurone depolarisation may evoke a recurrent response by stimulating the first node distal to the neurone. There is only a small chance that the timing of the depolarisation/repolarisation allows this to happen. Normally a recurrent response is evoked in 0.5-5% of the stimulations with some differences between nerves.
Fig.6 Studies of F-waves in normal peroneal (A) and tibial nerves (B). Note the different occurrence of frequency of the F-waves in these two nerves.
The F-waves travel from the stimulation point on the nerve to the neurone and back to the muscle. By subtracting the distal latency, the time taken from the stimulus point to the neurone and back again to the stimulus point can be obtained. This time depends on the conduction distance involved. Instead of measuring the extremity length, we relate the reference values to the height of the patient since arm and leg length are normally correlated to height.
Since each normal nerve contains hundreds of motor axons, it is usual to obtain 5-15 F-waves from 20 stimulations. They differ in latency and shape since they normally represent activity from different motor units. The frequency of occurrence is reduced when there is a conduction block anywhere along the nerve. F-wave measurements thus reflect conduction along the entire nerve and are therefore particularly useful in the study of general polyneuropathy and also when proximal segments are preferentially involved, as in Guillain-Barré syndrome, GBS.
M-satellites (often called A-waves)
M-satellites are responses usually occurring between the CMAP and the F-wave (Fig 7). They have a constant shape and latency and occur in at least 10 per 20 stimulations. In normal conditions they are only seen in the tibial nerve [1]. They are present in various pathological conditions but are unspecific in relation to a given diagnosis. An M-satellite may be generated as an extra discharge in the stimulated axon (intermediate double discharges, IDD) [24], be due to ephaptic transmission between two axons, axon reflex or represent the response from one axon with exceptionally slow conduction velocity. They are for example seen during the fist days of GBS representing IDD.
Fig.7 M-satellites (arrows) seen in the tibial nerve in a patient with normal conduction velocity. The two vertical lines indicate upper normal limit (right) and the estimated shortest F-latency among the obtained responses (left line). The F-wave latency is thus minimally increased. The number of F-waves is normal.
An H-reflex is a monosynaptic reflex that can be elicited by the stimulation of muscle spindle afferents in the limbs. It is possible to evoke H reflexes on most nerves during the first year of life. In adults it can most easily be elicited in the calf muscles and flexor carpi radialis [18]. The H reflex recorded from calf muscles - gastrocnemius, soleus – is mediated via the S1 root. The H reflex recorded from the flexor carpi radialis is mediated C7 root [7].
The H-reflex and F-wave differ in some aspects. The H-reflex contains a sensory and a motor branch. The H-reflex is studied only with a submaximal stimulus and is abolished by supramaximal stimulation. Although consecutive F waves vary in latency and waveform, H reflexes remain constant in response to repetitive stimuli. This is because H reflexes occur from activating the same motor neuron pool. In contrast F-waves represent recurrent discharges from different groups of motor neurons with different conduction characteristics [18].
H-reflexes may be obtained more easily if a long stimulus duration is used i.e. 0.5 or 1 ms. The H reflex habituates and decreases in amplitude with stimulation rates > 0.5 Hz. The voluntary activation of the investigated muscle or Jendrassic’s maneuver will enhance the H-reflex amplitude and shorten the latency [12].
The most common motor neurography parameters are summarised in Table 1.
Table 1.
| Parameter | Significance | Usually measured as | Analysis mode |
| CMAP | |||
| Ampl | # axons, synchronisation | neg. amplitude (mV) | a/m |
| Area | # axons, “ | neg. area (mV *ms) | a |
| Dur | neg. peak duration | (ms) | |
| Ampl decay | cond. block + dispersion | % reduction in ampl | a |
| Dispersion | axonal velocity disp | % increase in dur | a |
| CV | velocity of fastest axons | latency diff. (m/s) | a/m |
| Distal latency | velocity of fastest axons | latency | a/m |
| F-waves | |||
| Latency | cond. of axons along entire nerve | lat (min, mean in ms) | |
| Dispersion | axonal velocity dispersion | min and max lat (ms) | a/m |
| # Of F-waves | # axons and MN excitability | # F-waves 20 stimuli | a/m |
| Amplitude | MUP shape +# F-waves | peak-peak ampl (m V) | a/m not often used |
| M-satellites | |||
| Presence | abnormal excitability or slowly conducting axons | present or not | m |
| H-reflex | |||
| Latency | cond. along reflex arc | H-lat minus M-lat (ms) | a |
| Amplitude | excitability | M ampl / H ampl | a |
Table 1. Note: CMAP = compound muscle action potential; CV= conduction velocity; MN = motor neurone.
All modern EMG systems have programs for neurography. Many have algorithms for automatic measurements. These algorithms vary and reference values may therefore differ somewhat between laboratories.
Motor conduction block
One of the parameters in nerve conduction studies concerns the presence of impulse conduction blocks, CB. A Conduction block is the failure of an action potential to propagate throughout the length of a structurally intact axon. This may be seen at the site of a local nerve entrapment and is typical of autoimmune neuropathies (Guillain-Barré, GBS; chronic inflammatory demyelinating polyneuropathy, CIDP) and in multifocal motor neuropathy with persistent conduction block, MMN. In the hereditary demyelinating polyneuropathies, conduction block is seen only to a slight degree. Other polyneuropathies do not usually show conduction block.
The principal finding is the blocking of an impulse across a local segment without axonal damage. Therefore, stimulation distal to the block gives a normal CMAP whereas a proximal stimulation produces decreased amplitude.
This is normally studied in two ways in the motor nerve; by comparing CMAP at distal and proximal stimulation and by the assessment of the F-wave frequency.
In MCS, there is normally a slightly lower amplitude for proximal stimulation compared with distal and also a slight temporal dispersion. This can be explained simply as the difference in the conduction velocity between individual axons. This gives rise to an incomplete summation of signals and even some degree of so called phase cancellation. The longer the distance, the more pronounced these changes are. These effects are more pronounced in cases of slow conduction velocity due to demyelination, particularly when there is an increased spectrum of velocities among the axons. This means that when a reduction of the CMAP is obtained at proximal stimulation compared to distal response, the differential diagnostic problem is demyelination with general slowing causing abnormal temporal dispersion vs. a pure conduction block. In the first case, the distal-proximal amplitude drop is parallel to the temporal dispersion as expressed by the change in CMAP duration. With pure CB, there is no increase in temporal dispersion.
There have been several attempts to define criteria for conduction block [6;27]. To discriminate between "pure" demyelination and a conduction block, the following criteria have been suggested (Ad Hoc Subcommittee of the American Academy of Neurology AIDS Task Force 1991):
Conduction block is present if there is
>20% AMPLDECAY or AREADECAY and <5% DISPERSION or if there is
>50% AMPLDECAY or AREADECAY, independent of DISPERSION.
It has been [22] demonstrated that both these criteria are equally sensitive in detecting a block.
Our own modifications of these rules are:
>25% (arm) or >40% (leg) AMPLDECAY and <15% DISPERSION or if there is
>50% AMPLDECAY, independent of DISPERSION (in this case there is a combination of CB and demyelination)
The other parameter, which indicates CB, is the frequency of occurrence of F-waves. In cases of proximal CB, there may be no drop in amplitude in the distal part of the nerve, e.g. in early GBS. In this case, the CB can be seen by a reduction in the occurrence of F-waves. The normal frequency of F-waves varies for different motor nerves. When 20 stimuli are given in a normal situation the different nerves normally show the following numbers of F-waves: ulnar 18, median 15, peroneal 10, and tibial 18. As can be seen, it is impossible to detect a slight reduction in the number of F-waves in the ulnar and peroneal nerve. Lack of response is usually taken as an abnormal finding. In the tibial nerve, a better assessment of pathology can be made. It should be noted that in the interpretation of the number of F-waves, the CMAP amplitude needs to be considered. If the CMAP is reduced to half due to axonal degeneration, then the expected number of F-waves is correspondingly reduced.
The number of F-responses should also always be reduced when conventional MCS has shown a conduction block; otherwise a technical error is to suspect in the measurements.
In normal subjects, there could be some anatomical variations in the muscle innervation. Martin-Gruber anastomosis is the most common anomalous innervation of the hand with an incidence of 15-28% [17].
The fibers innervating intrinsic muscles of the hand cross from the median nerve to the ulnar nerve. Sensory fibers are not involved.
Martin-Gruber anastomosis can be divided into three types according to the muscle innervated by the crossing fibers. The most common type is type II, in which the crossing fibers innervate first dorsal interosseus muscle. When the anastomotic fibers end in the ADM and abductor pollicis brevis the MG is classified as type I and type III respectively. Type III is the least common. Martin-Gruber anastomosis can be revealed by stimulating the median and ulnar nerves and recording from the muscles mentioned above. The amplitude of the CMAP wave evoked by the median nerve stimulation at the elbow is found higher than the one evoked by stimulation at wrist level (Fig 8) when recording is performed by muscle with anomalous innervation. The CAMP obtained at the proximal site also has an initial small positive deflection because of the volume of conduction from the deep ulnar innervated muscles. In type I the amplitude of the M wave from the ulnar nerve stimulation shows a reverse discrepancy; lower with elbow stimulation and higher with wrist stimulation.
Fig.8 Median nerve MCS in a case of Martin-Gruber anastomosis. Note the higher amplitude and the initial positive going phase at stimulation at elbow.
When this anastomosis accompanies the carpal tunnel syndrome (CTS), an abnormally fast conduction velocity value, in the forearm segment of the median nerve is found, due to delayed DLAT but a “normal” PLAT.
Another common anatomical variation is the innervation of the extensor digitorum brevis by the accessory peroneal nerve. This muscle is normally supplied by the deep peroneal nerve. In 23 to 28 percent of population, it is innervated by the superficial branch, behind the lateral malleolus. When a low CMAP is obtained distally, stimulation should always be performed behind the malleolus.
Typical findings in pathology
In this table (Table 2), some standard findings are summarised. Note that the clinical presentation is usually a mixture of two classical pathophysiological states (Fig 9).
Fig.9 Schematic summary of the relationship between CV and amplitude parameters in axonal (low amplitude) and demyelinating (low CV) neuropathy.
Table 2.
| Demyelination | Axonal degeneration | Conduction block | |
| CV and dist latency | | n/ | n |
| Amplitude | n/¯ | ¯ ¯ | ¯ |
| Amplitude decay | n/ | n | |
| Dispersion | | n | |
| F-wave latency | | n/ | n |
| # Of F-waves | n/¯ | ¯ | ¯ |
Sensory Conduction StudiesTable 2. Note: Classical findings in neurography at different types of pathology. = increased; ¯ = decreased; n= normal
The pathophysiological principles regarding sensory nerves are the same as those discussed for motor nerves. Sensory neurography differs in some aspects: The amplitudes are much smaller than those in MCS and since the recordings are carried out on the nerve itself. Furthermore, only one stimulation site is necessary for calculating CV. The stimulation recording can be performed in both the orthodromic or antidromic direction. Some of these things will now be discussed.
The recording may be performed orthodromically or antidromically. The conduction velocity is the same but other parameters are different. In some instances one choice is preferred, but in other cases the choice is based on tradition. Table 3 shows differences between ortho and anti-dromic methods.
Table 3. Differences between orthodromic and antidromic nerve stimulation
| Orthodromic | Antidromic |
| No muscle artifact (+) | Less painful (+) |
| More painful | Larger amplitude (+) |
| Lower amplitude | Muscle artifact in mixed nerves |
Usually surface electrodes are used for recording purpose. The electrodes may be so called ring electrodes, e.g. around the digits. More commonly we use felt pad electrodes with a fixed interelectrode distance between the two recording poles. The electrode is placed along the nerve with the “active” towards the point of stimulation.
If the electrode is placed along the nerve, the two poles record an identical signal, often biphasic in shape with an initial smaller positivity and a larger positive part. In the amplifier the two signals are subtracted, giving rise to a triphasic configuration. The shape, in particular the duration and amplitude of the recorded signal is dependent on the conduction time between the two poles (distance and CV). This means that, the interelectrode distance has to be kept constant within the laboratory.
Reference data must be collected using the same technique that is used in routine studies.
In cases when near nerve needle electrodes are used, all the parameters except CV are different. In this case a special needle electrode is inserted just outside the nerve (the position may sometimes be tested by using the electrode for stimulation. An optimal position is found when a muscle response is obtained with minimal current, often less than 2 mA). The recording may show multiple peaks indicating the difference in conduction velocity among individual axons. This is particularly apparent in cases of pathology. Amplitude parameters are dependent on both the needle position in relation to the nerve, and the distance between the stimulation and recording. It is often not very reproducible in pathology. We use needle electrodes in Morton’s metatarsalgia and meralgia paresthetica.
Stimulation can be performed with a surface electrode, or with a needle electrode. In the first and more common case, the electrode is placed over a sensory nerve, or sometimes over the relevant skin area. The stimulus duration is often 0.1 ms and the frequency around 1 Hz. The stimulus strength is increased, as long the recorded response is increases. With higher stimulus strength, the increase in pain and, depending on the stimulus site, the motor artefacts may significantly disturb the recordings. Here an optimum must be found.
When a needle electrode is used, similar considerations should be made. In our study we used needle electrodes to obtain high selectivity, e.g. in the stimulation of individual digital nerves in cases of Morton’s metatarsalgia.
Parameters and a typical recording from the ulnar sensory nerve are shown in Fig 10.
Fig.10 Sensory nerve action potential. A= parameters, used in all nerve studies. B= Sensory conduction studies in the ulnar nerve. Averaged responses obtained orthodromically at the wrist with stimulation at palm, dig IV and dig V (traces 1-3). Antidromic sensory response obtained in the interdigital space between metacarpals IV and V and stimulating the dorsal ulnar branch. In cases of entrapment at the Guyon´s canal, this response is normal, while the digital responses are abnormal.
The latency is the time from the stimulus to the first positive peak of sensory nerve action potential (SNAP). If there is no clear positive peak in antidromic recording, the latency is measured from the take-off from baseline.
For the normal latency of a nerve there should be approximately two hundred nerve fibers conducting normally and having 10 m V or more of diameter [26].
Some laboratories have a tradition of measuring latency to the first negative peak. Since negative peak latency includes the rise time of SNAP and indicates the temporal dispersion, it is not recommended to use negative peak latency to calculate the CV.
The amplitude of the SNAP should be measured from the first positive peak to the highest negative peak. Some authors measure the amplitude as the maximum peak to peak amplitude or as the amplitude between a line joining the positive peaks as the positive value and the negative peak.
The amplitude reflects the number of nerve fibers having a diameter of 9 m V or more [26].
The area is the integrated area between the signal and baseline over the DUR.
The area represents the combination of amplitude and duration; therefore this reflects the number and synchrony of the activated nerve fibers.
The duration is measured from the first positive peak to the last positive peak. When there is no presence of the initial positive peak, the duration is measured from take-off on the baseline.
The conduction velocity (CV) is calculated by dividing the length of the nerve segment from the stimulus point to the recording point by the positive peak latency. It should be calculated as follows.
CV (m/s) = distance (mm) / LAT
When the sensory conduction velocity is calculated in this way, it reflects the conduction velocity of the fastest sensory fibers [26].
Since physiologic temporal dispersion affects the sensory action potential more than the muscle response, these parameters are not easily used in the routine studies. This is due to the difference in duration of individual unit discharges between nerve and muscle. With short-duration diphasic sensory spikes, a slight latency difference could line up the positive peaks of the fast fibers with the negative peaks of the slow fibers, cancelling both.
Table 4.
| Parameter | Significance | Usually measured as | Analysis mode | Comment |
| Latency | conduction velocity | positive peak (ms) | a/m | |
| CV | conduction velocity | distance/latency (m/s) | a/m | |
| Amplitude | # axons, temporal disp | peak-peak (m V) | a/m | |
| Area | # axons, temporal disp | total area (m V * ms) | a | |
| Duration | dispersion | pos.-pos. peak dur (ms) | a/m | |
| Late components | conduction dispersion | shape | m | in needle rec. |
Table 4 showing parameters usually measured in sensory neurography. Note: for explanation, see Table 1.
Modern EMG equipment has programs to measure these parameters. Often averaging is necessary in order to obtain a good signal to noise ratio.
In this table (Table 5), some standard findings are summarised. Note that the clinical presentation usually is a mixture of two of the classical pathophysiological states.
| Demyelination | axonal degeneration | ||
| CV and dist latency | | n/ | n |
| Amplitude | n/ | ¯ ¯ | ¯ |
| Amplitude decay | n/ | n | |
| Duration | | n | |
Table 5. Classical findings in neurography at different types of pathology. = increased; ¯ = decreased; n= normal
Autonomic nerve testingIn this review we do not include a description of the various tests of autonomic function but refer to articles [10;30;33]. It should however be noted that this is important in the general neurophysiological investigation of a patient with neuropathy. Some of the most commonly used methods are indicated in Table 6.
Table 6. Autonomic tests
| Autonomic testing includes Heart rate variation | |
| (At deep breathing, Valsalva, tilt test) | parasympathetic (sympathetic) |
| SSR | sympathetic |
| Pletysmography | sympathetic |
Neurographic findings in some examples of neuropathies
Polyneuropathy may be expressed as axonal or demyelinating, sensory, motor or sensory-motor. The type of electrophysiological picture depends on the underlying cause. In a typical case of sensory-motor demyelinating polyneuropathy (diabetes mellitus) the following may be found; Abnormalities are more pronounced in the legs than in the arms. Sensory findings are often more abnormal than motor. Note that this may not always reflect a more severe sensory involvement, but may be due to the fact that sensory nerves are usually investigated in a more distal segment than the motor nerves.
The CV is reduced relatively more than the amplitude drops.
F-waves in the motor nerves are delayed but not necessarily reduced in frequency. The F-waves include conduction along the entire nerve and is therefore sensitive parameters in early polyneuropathy (Fig 11)
Sensory amplitudes are sensitive to temporal dispersion and may show distinct general reduction.
No conduction block is seen.
Fig.11 F-waves in polyneuropathy compared to normal. A= normal condition, B= case of diabetic polyneuropathy. Note the prolonged latency in polyneuropathy. CV of right peroneal nerve was 36 m/sec There is also one M-satellite.
The hallmark is a proximal slowing with conduction block (Fig 12A), more in motor than in sensory nerves. The motor CV in the forearm or lower leg may be normal while the F latency is increased. The number of F waves is reduced due to proximal conduction block. Often M-satellites are seen (Fig 12B). Sensory CV may be normal in early the phase. Autonomic tests are often abnormal.
Fig.12 Guillain-Barré syndrome. Note the distal conduction block (A), loss of F waves (asterisk) and presence of M-satellites (arrows) (B).
As in GBS, the pathological hallmark of the CIDP is inflammatory demyelination of root and peripheral nerves.
Motor and sensory conduction velocities are usually less than 80% of the lower limits of normal. The conduction velocity may be normal when the early stage of the disease and when the process remains confined to proximal nerves or segments. If proximal nerve segments have a greater degree of reduction of conduction velocity than distal segments, a variable degree of nerve conduction block is found. The decrease in amplitudes (due to dispersion and loss of units) is also characteristic features of the muscle and nerve action potentials in this disorder.
Sensory nerve action potentials can not usually be elicited (from ulnar, median and sural nerves) using conventional methods [9].
Electrophysiological studies reveal increased central conduction time as well slowing of peripheral conduction velocity [18].
Hereditary Motor Sensory Neuropathies
The hypertrophic (type I) and neuronal (type II) varieties of Charcot-Marie-Tooth disease are the most common forms of hereditary motor sensory neuropathy (HMSN). These usually show an autosomal dominant inheritance. The hypertrophic variety shows enlargement of the peripheral nerves, segmental demyelination and remyelination with onion-bulb formation, and axonal atrophy. A very slow nerve conduction velocity is a hallmark of the HMSN type I. Prolonged terminal latencies in the early stages indicates prominent distal slowing. Despite slowing the degree of temporal dispersion is limited, which indicates homogeneity of the pathologic process. Such uniformity helps in differentiating this entity from acute inflammatory polyneuropathy.
In the neuronal variety of HMSN, patients have neither hypertrophic nerves nor prominent segmental demyelination. Electrophysiological studies reveal mild slowing of the nerve conduction velocities, and reduction in amplitude of the sensory and muscle action potentials [18].
Mononeuropathy; lesion, entrapment (CTS)
In cases of lesions and entrapment, e.g. across the elbow, the study of short segments (inching, centimetring) may be very important. In cases of local demyelination, a local slowing is seen. The site of a conduction block is revealed by short segment studies (Figs 13, 14 ). In some instances needle stimulation may be helpful to localise the lesion or entrapment with precision. It should be noted that both proximal and distal stimulation is important to give the complete picture. Comparison between sides regarding CMAP amplitude and latency may also provide important information.
EMG helps to assess the dynamics of the process (ongoing denervation, late or recent reinnervation)
Fig.13 Short segment studies. Centrimetring across the elbow over the ulnar nerve and recording from the abductor digiti minimi muscle. Note the gap in latency (demyelination) and the amplitude drop at more proximal stimulation (conduction block). A= raster mode, B= superimposed mode.
Fig.14 Sensory centimetering across the transcarpal ligament from wrist (top) to midpalm (bottom). Note the abrupt change in latency at a position corresponding to the distal border of the ligament. The sensory conduction studies show normal condition.
ALSIn motor neuron disease, the motor and sensory nerve conduction velocities are normal in preserved axons. A slowing may be seen when severe atrophy has developed and the CMAP is distinctly reduced. The change in the CV is considered to be due to the random loss of axons with no preferential early involvement of large axons.
The sensory nerve conduction velocities and amplitudes are normal in MND. This is an important finding which helps to differentiate ALS from axonal neuropathies.
Since pure motor nerve neuropathies do occur and can produce the same nerve conduction findings as MND, the ultimate diagnosis will depend on the EMG findings and clinical presentation.
MMN is a disorder that also shows conduction block [23;28], called Multifocal motor neuropathy (motor neuropathy with persistent multifocal conduction block), MMN. This disorder may be clinically similar to motor neuron disease. Therefore, in each patient with suspected motor neuron disease, AMPL DECAY must be studied in several motor nerves. (Fig 15). Not only the usual segments in the forearm should be tested, but also the nerves should be stimulated at proximal sites in the axilla and the supraclavicular fossa of the ulnar nerve. Sensory nerves do not show conduction block. In many respects multifocal motor neuropathy is similar to CIDP [28].
Fig.15 Conduction block in median motor nerve in a case of MMN (A). The sensory conduction studies show normal condition (B).
Although a neurophysiological diagnosis is basically based on EMG findings in radiculopathies, nerve conduction studies should be carried out in order to differentiate them from peripheral neuropathies. The results from sensory nerve conduction studies will be normal in opposition to what is usually found in most neuropathies including plexopathies. MCS can show some abnormalities parallel to the loss of the motor fibers.
The diagnosis is similar to that of radiculopathies in which diagnosis is based the pathological EMG findings limited to the denervated muscles. The presence of distal sensory potential serves as a criterion for differentiating preganglionic roots avulsion from plexopathy [18].
The clinical symptoms of sensory and motor dysfunction are related to types of pathophysiology. Some are explained by the peripheral lesion, others are due to the effects of central processing.
The functional disturbances may be divided into positive and negative effects according to the following table 7.
Table 7.
SYMPTOMS
| Sensory nerve fibers: | |
| 1. Local excitability | paraesthesia |
| 2. Neurapraxia | nerve impulse blocking with loss of sensibility |
| 3. Demyelination | changed sensation depending on type of nerve involvement |
| 4. Axonal degeneration | loss of sensation |
|
Motor nerve fibers: |
|
| 5. Local hyperexcitability | muscle cramps, twitches. Normal force. |
| 6. Neurapraxia | reduction in force, no atrophy |
| 7. Demyelination | no symptoms |
| 8. Axonal degeneration | force reduction, atrophy |
| Autonomic nerve fibers: | |
| 9. Neurapraxia or axonal degeneration | disturbed autonomic function, i.e. sweat function, gastrointestinal disturbance |
In all neurophysiological studies of nerve pathology, it is of importance to choose the optimal methods in order to extract pertinent information.
The following table is a short summary of the relationship between information obtainable from EMG, sensory/motor conduction studies and force measurements.
Table 8.
FUNCTION EFFECT
| Force | Neurography | EMG | |
| Sensory nerve fibers: | |||
| 1. Local hyperexcitability | - | - | - |
| 2. Neurapraxia | - | +(spec) | |
| 3. Demyelination | - | + (sep from 4) | - |
| 4. Axonal degeneration | - | + (sep from 3) | - |
|
Motor nerve fibers: |
|||
| 5. Local hyperexcitability | - | - | + |
| 6. Neurapraxia | + | +(spec) | + |
| 7. Demyelination | - | + | - |
| 8. Axonal degeneration | + | + | +(spec) |
|
Autonomic nerve fibers: |
|||
|
9. Neurapraxia or axonal degeneration |
- | + | - |
Neurography has become an essential part of the investigation the peripheral nervous system. The methods can be highly standardised, the analysis gives quantitative data and the various methods allow different types of nerve pathology to be tested.
Neurophysiological investigations are performed as a result of the suspicion of a special disease in the peripheral nerve, of a special subjective symptom or of findings during a physical examination. Some indications are listed below based on these three classifications.
Polyneuropathy
Mononeuropathy
Autoimmune required neuropathy
Critical illness
ALS
MMN
Weakness
Numbness
Pain
Tingling
Abnormal sensations
Autonomic dysfunction
SignsWeakness
Disturbed sensibility
Atrophy
Weak reflexes
Acknowledgements
The work was supported by the Swedish Medical Research Council (ES, Grant 135).
1. Bischoff C, Stålberg E, Falck B, Puksa L. Significance of A-waves recorded in routine motor nerve conduction studies. Clin Neurophys 1996; 101: 528-533.
2. Brown WF, Feasby TE. Conduction block and denervation in Guillain-Barré polyneuropathy. Brain 1984; 107: 244.
3. Buchthal F, Rosenfalck. Sensory conduction velocity along the limbs. Brain Res 1966; 3: 34-42.
4. Buchthal F, Rosenfalck A. Evoked action potentials and conduction velocity in human sensory nerves. Brain Res 1966; 3: 1-122.
5. Campbell WW, Ward L, Swift T. Nerve conduction velocity varies inversely with height. Muscle Nerve 1981; 3: 436-437.
6. Cornblath DR, Sumner AJ, Daube JR, Gilliatt RW, Brown WF, Parry GJ, Albers JW, Miller RG, Petajan JH. Conduction block in clinical practice. Muscle Nerve 1991; 14: 869-871.
7. Deschuytere J, Roselle N, De Keyser C. Monosynaptic reflexes in superficial forearm flexors in man and their clinical significance. Journal of Neurology, Neurosurgery, and Psychiatry 1976; 39: 555-565.
8. Dioszeghy P, Stålberg E. Changes in motor and sensory nerve conduction parameters with temperature in normal and diseased nerve. Clin Neurophysiol 1992; 85: 229-235.
9. Dyck PJ, Prineas J, Pollard J. Chronic Inflammatory Demyelinating Polyradiculoneuropathy. In: Peripheral Neuropathy. Dyck PJ, Thomas PK, eds. Philadelphia: W.B. Saunders Company; 1993. pp 1498-1523.
10. Ewing DJ. Practical bedside investigation of diabetic autonomic failure. In: A textbook of clinical disorders of the autonomic nervous system. Bannister R, ed. Oxford: Oxford University Press; 1983. pp 371-405.
11. Falck B, Andreassen S, Groth T, Lang H, Melander M, Nurmi A, Puusa A, Rosenfalck A, Stålberg E, Suojanen M. The Development of a Multicenter Database for Reference Values in Clinical Neurophysiology - Principles and Examples. Comput Programs Biomed 1992; 34: 145-162.
12. Falck B, Stålberg E. Motor Nerve Conduction Studies: Measurement principles and interpretation of findings. Journal of Clinical Neurophysiology 1995; 12: 254-279.
13. Gamstorp I, Shelburne SA. Peripheral sensory conduction of ulnar and median nerves of normal infants, children, and adolescents. Acta Paediatr Scand 1965; 54: 309-313.
14. Gilliatt RW, Thomas PK. Changes in nerve conduction with ulnar lesions at the elbow. Journal of Neurology, Neurosurgery, and Psychiatry 1960; 23: 312-320.
15. Hodes R, Larrabee MG, German W. The human electromyogram in response to nerve stimulation and the conduction velocity of motor axons. Arch Neurol Psychiat 1948; 60: 240-365.
16. Hopf HC, Maurer K. Temperature dependence of electrical and mechanical responses of the adductor pollicis muscle in humans. Muscle Nerve 1990; 13: 259-262.
17. Kimura J. Electrophysiological study of anomalous innervation of intrinsic muscles. Arch Neurol 1976; 33: 842-844.
18. Kimura J. Electrodiagnosis in Diseases of Nerve and Muscle: Principles and practice. Philadelphia: F.A. Davis; 1989.
19. LaFratta CW, Smith OH. A study of the relationship of motor conduction velocity in the adult to handedness and sex. Arch Phys Med Rehabil 1964; 54: 475-477.
20. Lambert EH. Diagnostic value of electrical stimulation of motor nerves In. 1962. pp 9-16.
21. Lang AH, Puusa A, Hynninen P, Kuusela V, Jäntti V., Sillanpää M. Evolution of nerve conduction velocity in later childhood and adolescence. Muscle Nerve 1985; 8: 38-43.
22. Lange DJ, Trojaborg W, Latov N, Hays AP, Younger DS, Uncini A, Blake DM, Hirano M, Burns SM, Lovelace RE, Rowland LP. Multifocal motor neuropathy with conduction block: is it a distinct clinical entity? Neurol 1992; 42: 487-505.
23. Lewis RA, Sumner AJ. Multifocal demyelinating neuropathy with persistent conduction block. Neurol 1982; 32: 958-964.
24. Magistris MR, Roth G. Motor axon reflex and indirect double discharge: ephaptic transmission? A reappraisal. Clin Neurophys 1992; 85: 124-130.
25. Nilsson J, Ravits J, Hallett M. Stimulus artifact compensation using biphasic stimulation. Muscle Nerve 1988; 11: 597-602.
26. Oh SJ. Clinical Electromyography. Nerve Conduction studies. University Park Press, Baltimore 1984.
27. Olney RK, Budingen HJ, Miller RG. The effects of temporal dispersion on compound muscle action potential area in human peripheral nerve. Muscle Nerve 1987; 10: 728-733.
28. Parry GJ. Motor neuropathy with multifocal conduction block. In: Peripheral neuropathy . Dyck PJ, Thomas PK, Griffin JW, Poduslo JF, eds. Philadelphia: WB Saunders Company; 1993. pp 1518-1524.
29. Rosenfalck P, Rosenfalck A. Electromyography-sensory and motor conduction. Findings in normal subjects. Lab of Clin Neurophysiology, Copenhagen 1975; 1-49.
30. Shahani BT, Halperin JJ, Boulu P, Cohen J. Sympatetic skin response - a method of assessing unmyelinated axon dysfunction in perpheral neuropathies. Journal of Neurology, Neurosurgery, and Psychiatry 1984; 47: 536-542.
31. Simpson JA. Fact and fallacy in measurement of conduction velocity in motor nerves. Journal of Neurology, Neurosurgery, and Psychiatry 1964; 27: 381-385.
32. Stetson DS, Albers JW, Silverstein BA, Wolfe RA. Effects of age, sex and antropomethric factors on nerve conduction measures. Muscle Nerve 1992; 15: 1095-1104.
33. Stålberg E, Nogues MA. Automatic analysis of heart rate variation, Part I: Method and reference values in healthy controls. Muscle Nerve 1989; 12: 993-1000.
34. Tasaki I, Takeuchi T. Weitere studien Uber den Aktionsstrom der markhaltigen Nervenfaser und uberdie elektrosaltatorische Ubertragung des Nervenimpulses. Pflug Arch Ges Physiol 1942; 245: 764-782.
35. Taylor PK. CMAP dispersion, amplitude decay and area decay in a normal population. Muscle Nerve 1993; 16: 1181-1187.
36. Waxman SG. Conduction in myelinated, unmyelinated, and demyelinated fibers. Arch Neurol 1977; 34: 585-589.