Добавил:
Sekretar
kiopkiopkiop18@yandex.ru
t.me/Prokururor I Вовсе не секретарь, но почту проверяю
Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз:
Предмет:
Файл:Ординатура / Хирургия / Библиотека им академика М.И. Перельмана / Книга_6034_Библиотеки_им_академика_М_И_Перельмана.pdf
CHAPTER 10
Clinical Neurophysiologic and Electrodiagnostic Testing in Disorders of the Lumbar Spine
Jiri Dvorak and Scott Haldeman
Patients with symptoms related to the lumbar spine can be
differentiated into two groups: those with neurologic
deficits and those with more benign pathology causing
pain. This dif ferentiation carries significant clinical importance when considering prognosis as well as the necessity
for nonsurgical or surgical intervention. The patient presenting with neurologic findings suggestive of a spinal
cord or cauda equina lesion may represent a medical or
surgical emergency. Patients with acute or progressive
radiculopathy may respond to nonsurgical care but require
more intense investigation than the patient without radiculopathy and may benefit from surgical decompression.
Patients with chronic neurologic lesions carry a poorer
prognosis than appropriately treated patients with acute
neurologic deficits. For this reason the documentation of
neurologic deficits is one of the primary goals of the
diagnostic process when evaluating patients with disorders of the lumbar spine.
When neurologic deficits are noted on examination of
patients with lumbar radicular pain syndromes due to
disc herniation or stenosis there may be a discrepancy
between clinical and neuroradiologic imaging (magnetic
resonance imaging, computed tomography, myelogram)
findings. Fur ther more, virtually all forms of nondestr uctive pathology noted on an imaging test can exist in the
absence of symptoms. This can make it diff icult to identify the particular nerve root or spinal cord level responsible for the patient’s complaints. In other patients the
clinical examination may be equivocal and there may be
considerable doubt regarding the presence of neurologic
deficits when patients present with vague nonspecific
sensory or motor symptoms in the lower extremities.
Neurologic deficits in the lower extremities, even in
patients with low back pain, may represent disorders that
are not related to the lumbar spine. There are a number of
compression lesions such as peroneal neuropathy and
tarsal tunnel syndrome that can mimic radicular clinical
pictures, especially if the symptoms are diffuse or the
clinical examination is superficial. Surgery to the lumbar
spine in these patients is unlikely to be of any benefit in
reducing such deficits.
A surgeon contemplating surgery often has to answer
two questions. The f irst is the determination of the presence and degree of neurologic loss. The second is the
level of a spinal cord lesion or the nerve root that that
may respond to decompressive surgery. If there is close
concordance between clinical and imaging findings there
is no need to consider further testing. However, in cases
where imaging and clinical findings are not in complete
agreement, the surgeon may require additional testing in
order to make the correct decision on whether to operate
and at what level surgery should be contemplated. It is in
these cases, where there is doubt as to the presence of
neurologic deficits or the level of such deficits, that neurophysiologic and electrophysiologic tests can become
important in the diagnostic process.
There still remain some questions related to the sensitivity, specificity, and positive predictive value of certain
electrophysiologic tests that often raise unreasonable
expectations for these tests. There has also been a trend
toward the indiscriminate ordering of batteries of tests in
patients with sciatica. This has often led to confusing
results that may not be of much help to the treating ph ysician. Despite these shortcomings, the use of electrodiagnostic testing has become routine in most clinical settings
that treat patients with disorders of the lumbar spine. This
chapter will attempt to outline the most common electrophysiologic tests and to describe how they can be of the
113
114 /SECTION I/BASIC SCIENCE
most value to both surgical and nonsurgical treating
physicians.
ELECTRODIA GNOSTIC TESTING
Electrodiagnostic testing can be divided into three dis-
tinct areas of interest.
1. The investigation of a suspected radiculopathy. This
is often the primary goal of electrodiagnostic testing. In
this setting the testing is used to document the presence
of and the level of a radiculopath y as well as to give some
indication of the chronicity of the neurologic loss. The
mainstay of this testing is electromyography (EMG).
Electromyography, however, can be supplemented by the
use of H-reflexes and possibly F-responses in order to
make the testing more meaningful.
2. The investigation of myelopathy. The presence of
symptoms suggestive of a myelopathy may represent a
medical emergency and may have to be confirmed in a
confusing clinical setting. Somatosensory evoked potentials (SEPs) and motor evoked potentials (MEPs) are most
commonly used to in vestigate conduction within the spinal
cord. SEP-techniques can be used to evaluate the sensory
pathways within the spinal cord whereas MEPs allow for
assessment of lesions that affect motor neuron pathways.
Both SEPs and MEPs, however, are impacted by lesions
that affect nerve roots as well as the lumbosacral plexus
and depend on the integrity of peripheral nerves for accurate recording. This can make it diff icult to interpret these
tests if more than one lesion is present in a patient.
3. The differentiation of proximal nerve root lesions
from other peripheral or entrapment neuropathies. The
primary tool for this process is peripheral motor and sensory nerve conduction studies or conventional neurography (electronystagmogram, or ENG). It is often necessary to supplement these studies or at least correlate
findings from ENG studies with F-wave, H-reflex, and
EMG findings. Electromyography of limb and paraspinal
muscles, for example, may allow a distinction to be made
between lesions affecting motor roots and more peripheral nerve elements. Electroneurography, F-wave, and Hreflex studies may be the only manner to distinguish
between proximal root and peripheral nerve disease with
a high degree of conf idence.
SEPS
SEPs can be recorded over the scalp adjacent to the
sensory cortex on electric stimulation of the large mixed
motor-sensory nerves, small sensory peripheral nerves,
or the skin over specific dermatomes. Responses can
also be recorded on magnetic stimulation of paraspinal
and peripheral muscles. The recording of these potentials, due to their small amplitude in comparison with
the background electrical noise, requires the computer
averaging of multiple responses. Most laboratories will
record simultaneously the sensory action potential
within the peripheral nerve and, where possible, a
response over the lower lumbar spine. The latter, however can be difficult in older patients and especially in
overweight patients.
The nerves most commonly used for the diagnosis of
spinal cord lesions are the large mixed sensory-motor
posterior tibial and common peroneal nerves of the lower
limbs usually at the level of the ankle. The absolute
latency of scalp response and the difference in latency of
the responses from the two legs can be used as an indication of reduced conduction within the spinal cord. By
recording the peripheral sensory nerve conduction and
measuring the height of the patient, it is possible to calculate the expected normal latency. If a spinal response is
obtained, a central conduction time can be calculated by
subtracting the latency of the spinal response from the
latency of the cortical response.
Attempts to diagnose radiculopathy by stimulating dermatomes and small sensory nerves have led to disappointing results and suggest that the sensitivity, specificity, and reliability of SEPs in radiculopathy is not
sufficient for general use. These tests are also very timeconsuming and technically challenging as the responses
have much smaller amplitude than those obtained from
larger nerves. The use of mixed nerve responses is very
insensitive in the diagnosis of radiculopath y because both
the posterior tibial and peroneal nerves contain fibers
from multiple nerve roots. How e ver, in patients with multilevel radiculopathy or plexopathy marked abnormalities
can be seen on stimulation of these large nerves.
SEPs have, however, been shown to be of value in documenting cauda equina and spinal cord lesions that affect
bowel, bladder, and sexual function. In these patients cortical evoked potentials and bulbocavernosus reflex responses on stimulation of the pudendal nerve may give
some indication whether the bowel, bladder, or sexual dysfunction is due to a lesion within the spinal cord or more
peripherally within the nerve root and pudendal nerve.
MEPS
Barker et al. first introduced the method of painless
magnetoelectric transcranial stimulation of the cerebral
cortex in 1985 (1,2). They applied short magnetic pulses to
the scalp produced by a device designed to stimulate
peripheral nerves, and recorded muscle action potentials
from upper and lower limb muscles. The magnetic field
produced by this instrument passes through scalp and skull
to stimulate the cerebral cortex. Magnetoelectric stimulation can also be used to stimulate deep-lying proximal segments of peripheral nerves and nerve roots (3), thus allowing for evaluation of central and proximal peripheral
pathways. This equipment has also been used to stimulate
paraspinal muscles and record cortical evoked potentials
that can be influenced by muscle spasm.
CHAPTER 10/ CLINICAL NEUROPHYSIOLOGIC AND ELECTRODIAGNOSTIC TESTING / 115
The muscles most commonly used for recording cortically evoked MEPs in the lower extremities are the
quadriceps, tibialis anterior, gastrocnemius, extensor hallucis, and abductor hallucis muscles (4). Surface recording electrodes are placed over the motor end plate. The
segmental innervation of these muscles is used to determine the level of the lesion.
When recording MEPs on stimulation of motor roots at
the lumbar spine, the intensity of the stimulator must be
adjusted so that a potential with a steep negative rise can
be recorded. In this situation the onset latency is not critically dependent on the positioning of the coil or the stimulation strength (3). The site of excitation of the nerve
root is most probably the nerve root as it exits from the
intervertebral foramen (3).
In order to interpret the MEP waveform it is necessary
to obtain an M-wave recording by means of conventional
neurography. The M-wave is the electric potential
recorded from the muscle in response to a supramaximal
stimulus of the peripheral nerve. This provides a measure
of muscle electric response “size” (5) and is used as a reference signal with which transcranial stimulation MEP
amplitude and duration are compared (i.e., MEP amplitude and duration are expressed as ratios of M-wave
amplitude and duration).
F-WAVES
F-waves are long latency responses recorded over distal
muscles on stimulation of motor nerves innervating the
muscle. This is achieved through the stimulation of a Renshaw cell in the anterior horn of the spinal cord through
antidromic stimulation of the motor nerve. Distinct leftright latency differences that exceed normal values or a
reduced number of F-wav es after a given number of supramaximal peripheral stimuli in the presence of normal distal motor conduction can be a sign of a proximal neuronal
lesion in the sciatic plexus or nerve root. F-waves, however, must be interpreted with caution. The F-w ave is often
normal in mild cases of radiculopathy, especially if only
one nerve root is involved.
In conjunction with MEPs, however, F-wave recordings may give information about conduction times in
motor fibers within the proximal segments of spinal
nerves that may be compressed by a disc herniation. Fwave recordings allow for the determination of peripheral
nerve conduction time or peripheral latency (PL), the
time it takes for impulses to travel from the anterior horn
cell to the muscle. This latency includes conduction over
the motor root from the spinal cord through its exit from
the intervertebral foramen to the muscle where it is
recorded. Calculation of PL is especially important in
lumbar spine disorders where motor roots measure 10 to
20 cm (6) and contribute considerably to peripheral
latency. F-wav e recordings can therefore help localize the
site of a lesion (7).
H-REFLEX
The H-reflex was first described by Hoffmann in 1918
(8). It is a reflex motor response within a muscle elicited on
electric stimulation of large, low-threshold sensory nerve
fibers within the ner ve from that muscle. The response on
stimulation of this nerve results in excitation of the motor
neuron pool that innervates the muscle (from which the Hwave is recorded) through the same ner ve. It is a monosynaptic reflex response that has a strong correlation with the
tendon jerk but bypasses the muscle spindles.
In adults the H-reflex is recordable in a limited group
of extensor muscles, especially the soleus/gastrocnemius
muscles in the calf innervated by the S1 nerve root. Lo wamplitude voluntary muscle contraction may facilitate the
H-response (9). Stimulation of the tibial nerve at the knee
with slowly increasing intensity from subthreshold to
submaximal levels allows for recording of H-responses
with increasing amplitude from the soleus muscle. Further increase in stimulus intensity elicits M-waves of
increasing size, while the H-reflex diminishes progressively and is eventually replaced by the F-wave on supramaximal stimulus intensity. H-reflexes and F-waves have
similar latencies when stimulus and recording sights are
at the same location. S1 sensory or motor root def icits
reduce H-responses and increase their latency. Right/left
latency differences can be a sensitive indicator of unilateral S1-radiculopathy. Braddom et al. and Aiello et al.
noted a 90% to 100% true-positive rate and 0% true-negative rate in S1 radiculopathies using the H-reflex from
the soleus/gastrocnemius muscles (10,11).
EMG
EMG performed with concentric or monopolar needle
electrodes is the oldest and the most widely used neurophysiologic test for the diagnosis of nerve root compression syndromes (12). It is often used as an extension of
the physical and neurologic examination and the muscles
selected for testing are usually selected based on the clinical findings (Fig. 10-1).
Needle EMG requires the physician to study the muscle under different conditions of muscle contraction. F our
specific forms of electric muscle activity are recorded
and noted for each muscle tested:
1. Insertional activity is evaluated at the time of insertion of the needle into the bulk of the muscle and at each
repositioning of the needle electrode within the muscle. It
is common practice to sample muscle electric activity at
10 to 20 locations within the muscle.
2. Spontaneous activity is studied with the muscle at
rest. At each location within the muscle, the needle electrode is maintained in a stationary and stable positioning
and muscle electric activity is recorded with the muscle at
rest. This allows for the detection of abnormal electric
activity such as fibrillation potentials and positive sharp
116 /SECTION I/BASIC SCIENCE
FIG. 10-1. Electromyogram with concentric needle electrode
from tibialis anterior muscle
waves that are indication of acute denervation of the muscle (Fig. 10-2).
3. Single motor unit action potentials (MUAPs) are
recorded during light voluntary contraction of the muscle
and examined with respect to amplitude, duration, and
number of phases of the electric potentials. An average of
20 MU APs are commonl y e v aluated and can v ary slightly
from muscle to muscle.
4. Motor unit recruitment and the interference pattern
are recorded during a gradual increase of voluntary muscle contraction and during maximal voluntary contraction
to obtain a crude indication of the degree of muscle loss
following denervation (Fig. 10-3).
In normal muscles, MUAPs are only seen during voluntary muscle contraction. The membranes of denervated
muscle fibers become unstable and sensitive to mechanical or chemical irritation. This results in increased insertional activity and spontaneous activity that can be
recorded in the absence of muscle contraction. These
signs of denervation noted on EMG testing become evident at about 14 to 21 days after the nerve lesion. As the
nerves to paraspinal muscles are shorter than those traveling to distal muscles, the spontaneous activity is first
seen in paravertebral muscles followed by proximal and
then distal muscles of the leg. These potentials represent
signs of acute denervation of the muscle.
The analysis of single MUAPs may reveal characteristics that are typical but not specific for lower motor neuron injury that can occur in radiculopathy. The finding of
increased amplitude, increased number of phases, and
increased duration of the motor unit potentials are classically seen only after reinnervation of denervated muscle
fibers as the result of sprouting from adjacent unaffected
fibers. These changes are therefore termed signs of
chronic denervation or reinnervation. Decreased motor
unit recruitment and discharge are crude signs of the
degree of neuronal loss as the result of radiculopathy.
SENSITIVITY AND SPECIFICITY OF
NEUROPHYSIOLOGIC TESTING
There are numerous problems in the interpretation of
published research studies that have looked at the sensitivity and specificity of the different electrodiagnostic
tests. The primary difficulty is the determination of a
gold standard for comparison. The studies that have been
published hav e attempted to correlate the tests with either
clinical examination findings, imaging studies such as
computed tomography (CT), magnetic resonance imaging (MRI) or myelography, or the observation of nerve
root compression noted during surgery. The difficulty in
attempting to correlate electrodiagnostic testing with the
clinical examination is that many clinical findings such
as motor and sensory changes can be equivocal and influenced by pain that the patient ma y be e xperiencing. There
may also be a fair degree of interobserver differences
noted in neurologic clinical findings, especially among
nonneurologists, that can make it difficult to interpret
these results. The dif ficulty in using imaging studies such
as CT, MRI, or myelo graphy as a gold standard is that virtually all findings considered abnormal on these studies,
including some of the most severe lesions that appear to
be causing neuronal compression, can be seen in the
asymptomatic population with a normal examination.
One of the primary reasons for conducting the electrodiagnostic testing is to document the significance of a suspected compressive lesion. It therefore does not make
sense to use imaging studies as the gold standard. The
problem with surgical observation is that there is considerable subjectivity on the part of the surgeon in documenting the presence of root compression and the sur-
FIG. 10-2. Summar y of different typical spontaneous activity from denervated tibialis anterior muscle
as shown in Figure 10-1 (1, sharp positive wave; 2, fibr illation; 3 and 4, fasciculation).
CHAPTER 10/ CLINICAL NEUROPHYSIOLOGIC AND ELECTRODIAGNOSTIC TESTING / 117
FIG. 10-3. Motor unit recruitment (interference pattern) during voluntary reduced contraction from denervated tibialis
anterior muscle.
geon often does not have full visualization of the root or
explore all potential nerve roots, especially in the newer
minimally invasive procedures.
A number of studies however have attempted to address the issue of sensitivity and specificity of the neurophysiologic assessment of nerve root compression syndromes when compared to clinical, imaging, and surgical
findings (13–15). The results of these studies have varied
greatly from as little as 20% to as much as 90% or better
correlation. The often markedly different results reported
in other studies is due, in part, to the different electrodiagnostic tests used, the number of electrodiagnostic tests
studied, and the basis of documenting the lesion (clinical,
imaging, or surgical). The greater the number of electrodiagnostic tests used and the greater the number of tests
the more sensitive the study is likel y to be, but at the same
time the results are likely to be less specific as each test
has its own unique level of accuracy. The less precise the
imaging or clinical finding the less likely that a correlation will be found.
These studies have been reviewed in detail by a number of authors and we will simply discuss a few of these
papers to illustrate this point. Tullberg et al. (16), for
example, looked at a series of electrodiagnostic tests in
20 patients who had undergone lumbar surgery for CTdocumented disc herniations. They used a wide v ariety of
tests including standard-needle EMG, F-wave responses
on stimulation of the peroneal nerve (L5 root), and tibial
nerve (S1 root) and dermatomal SEPs. They compared
these studies with clinical neurologic findings such as
motor, sensory, or reflex changes, CT scan results, and
surgical observation of root compression. Using multiple
diagnostic tools and multiple points for correlation, it is
not surprising that these authors found poor reliability to
predict results using electrodiagnostic testing. The y noted
that clinically only 4 patients with documented root compression on CT scan had motor loss and only 10 had sensory loss with very little correlation between the different
findings on clinical examination. They noted that 13 of
the 20 patients had abnormal electrodiagnostic test
results but the correlation between tests and between tests
and CT findings was low . As expected EMG was the most
sensitive of these three testing methods for determining
the presence of radiculopathy (45%) but it was less sensitive in determining the level of the disc protrusion
(20%). The sensitivity of the F-wave to document the
presence of a root lesion noted on CT or surgery was
35%, which is in agreement with other investigations
(17–20). However, the results were again unreliable in
predicting the exact level. These authors concluded that
there was no correlation between electrodiagnostic studies and the outcome of surgery.
The lack of correlation between electrodiagnostic studies, clinical findings, and imaging was studied by Haldeman et al. in 100 patients with chronic low back and leg
pain who were undergoing disability evaluations for workrelated injuries (21). The most revealing part of this study
was the lack of correlation between clinical findings and
imaging studies. The conclusion was that, in patients with
chronic persistent back pain, there is a breakdown in the
correlation between the clinical presentation and pathology. This makes it difficult to use patients with chronic
pain complaints as a means of determining the reliability
of any test in documenting disability.
Most studies, however, that have looked at the correlation of a single electrodiagnostic test and a specific clinical or imaging finding have found a correlation of
between 75% and 85% in the documentation of radiculopathy (22,23). The results of Toyokura et al. (24), who
looked at patients with a well-defined lesion rather than
conduct a global study, conflict markedly with the results
of Tullberg et al (16). Tokoyura et al. found that there w as
a significant improvement in F-responses after surgery
that correlated with the improvement of muscle w eakness
after surgery.
There has also been a fairly high correlation between
electrodiagnosis and the evaluation of muscle or motor
function. Carter and Fritz compared EMG findings of
acute denervation in patients with MRI findings of root
compression (25). They compared the f indings on EMG
with the findings on short-time inversion recovery
(STIR) MRI of the muscles affected by the nerve roots.
STIR MRI has been noted to have a strong correlation
with peripheral nerve injury that causes denervation and
associated muscle edema. They noted a 92% cor relation
between denervated muscle on EMG and that was noted
on STIR MRI. Zsu et al. provided more evidence of a
close correlation been different electrodiagnostic tests
and gave some indication how they could be used (26).
They noted that in 227 patients with signs of acute denervation on EMG due to radiculopathy 47% of patients
with L5 radiculopathies had an abnormal peroneal nerve
F-wave. There was an abnormal H-reflex in 73% of cases
with a S1 radiculopathy. The posterior tibial F-response
was less sensitive showing only a 23% abnormal rate in
patients with an S1 radiculopathy. They found no falsepositive results. These authors believe that the use of long
latency responses is primarily to confirm the findings on
EMG and more accurately define the level of the lesion,
but they also believe that these tests should not be performed without EMG because of the large false-negative
results of using F-waves as a freestanding test.
118 /SECTION I/BASIC SCIENCE
Tullberg et al. (16) found that dermatomal SEP showed
only 15% reliability in documenting the level of a root
lesion, a finding that has been observed in a number
of other studies. The studies of Dermatomal Sensory
Evolved Potentials (DSERs), however, have given conflicting results that have fueled considerable controversy
over the use of these tests. Yazicioglu et al. found that
these tests were misleading in 27% of patients and predicted the presence and level of the lesion in onl y 7.2% of
patients (27). This has led man y authorities, including the
American Academy of Neurology, to issue statements
that DSERs do not add anything significant to the electrodiagnostic evaluation of radiculopathy (19). This has
not, however, eliminated the controversy. Pape et al.
recently reported a strong correlation between SERs and
subsets of patients with sciatica (28). The y found a strong
correlation in patients with sciatica due to facet joint
hypertrophy causing nerve root compression with or
without disc pathology.
Studies on combining motor and sensory evoked
responses that theoretically w ould increase reliability ha ve
also been disappointing. For example, Vohanka and Dvorak (29) correlated the neurophysiologic findings with CT
or MRI findings of the lumbar spine. The quantitative
analysis of motor unit potentials showed 30% sensitivity
in patients with radiculopathy, but without motor deficit.
The MEPs and SEPs combined reached sensitivities of
55%, but the MEPs had 75% false- negative findings.
One of the difficulties in electrodiagnostic studies has
been the lack of lower extremity muscles and easily
accessed peripheral nerves that can be tested for high lumbar disc herniations. Haig et al. (30, 31) has been studying
the sensitivity of needle EMG of paraspinal muscles using
a very precise mapping technique in fairly large samples of
patients with and without low back pain and radiculopath y.
They found that patients who are clinically normal have
few if any EMG abnormalities in the paraspinal musculature despite a high incidence of abnormalities found on CT
and MRI (32) in asymptomatic subjects. How e v er, patients
with radiculopathy as documented on pain drawings had a
high degree of correlation with denervation in needle
EMG of paraspinal muscles. The authors recommend
EMG mapping of paraspinal muscles to rule out false-positive imaging studies. They found that the combination of
paraspinal EMG mapping and lower extremity EMG
showed a very strong correlation with imaging of root
compression with a false-positive rate of only 8% and a
false-negative rate of only 5%. In the small subset of
patients with lack of correlation between imaging studies
and electrodiagnostic testing it is still not possible to indicate the more reliable test for determining the presence of
clinically significant radiculopathy.
Correlation between axial CT imaging and narrow
spinal canal and electrophysiologic conduction studies
has been prospectively evaluated in 132 patients by
Vohanka et al. (33). Neurogenic claudication was initially
declared by 59% of the patients. Twenty-six patients had
one level, 68 had two levels, and 37 had three levels of
central stenosis. No statistically significant relationship
was found among the number of lev els of the stenosis and
the nerve conduction studies. However a significant relationship was found between minimum sagittal spinal
canal diameter and the delay of central conduction time
by transcranial magnetic stimulation. A similar correlation was detected in amplitude of the H-reflex and minimum transversal diameter.
Adamova et al. (34) introduced an exercise treadmill
test in patients with mild lumbar spinal stenosis. It is a
simple examination that can verify walking capacity and
confirm neurogenic claudication described by the patient.
Unfortunately an extensive electrophysiologic testing (Hreflex, F-response, MEP) and the analysis of obtained
data before and after the treadmill test did not show significant changes in comparison with control g roups.
CONCLUSIONS
There are ongoing studies that will hopefully clarify
some of the difficulties clinicians have in interpreting
imaging and electrodiagnostic testing. At this time, however, it is clear that it is not possible to take any one clinical finding, imaging study f inding, or electrodiagnostic
test out of context with other studies or findings. The
most accurate method of determining the presence of a
neurologic lesion is to conduct the electrodiagnostic test
in conjunction with the other tests and clinical findings.
It is increasingly being recommended that the ideal
approach to the study of radiculopathy is the electrodiagnostic consult by a specialist rather than the electrodiagnostic test by a technician. In this situation the electrodiagnostic specialist performs a history and examination of
the patient and reviews all imaging studies. The determination of which electrodiagnostic test should be performed is based on the questions that arise from the
examination of the patient.
If the clinical picture is clear, it may not be necessary
to perform any testing. If there is concern as to whether
an obvious radiculopathy is acute or chronic it may be
sufficient to perform only needle EMG to look for signs
of acute denervation or chronic reinnervation patterns. If
there is concern as to whether a specific lesion on imaging is causing a radiculopathy then it ma y be important to
include H-reflex or peroneal F-wave studies for the documentation of an S1 lesion or F-responses if polyradiculopathy is being considered. Paraspinal EMG mapping
may be most appropriate when an upper lumbar radiculopathy is being considered. If there is confusion as to
whether a neurologic deficit is due to a peripheral metabolic or entrapment neuropathy then it becomes necessary to consider nerve conduction studies. If there is consideration of a myelopathy or cauda equina lesion then
somatosensory evoked responses, bulboca vernosus reflex
CHAPTER 10/ CLINICAL NEUROPHYSIOLOGIC AND ELECTRODIAGNOSTIC TESTING / 119
responses, or cortical motor evoked responses may be
necessary. It may be necessary to consider other neurophysiologic tests such as cystometry, nocturnal penile
tumescence, or specific tests of the autonomic ner vous
system not mentioned in this chapter in order to obtain a
clear picture of the nature of the neurologic lesion causing a patient’s symptoms.
It is the tailoring of the clinical neurophysiologic test to
the patient and, in particular, the clinical question being
asked that gives these tests their greatest value. The more
qualified the specialist performing the test and the manner
in which the tests are correlated with the clinical f indings
the more reliable the testing can be considered. The indiscriminate use of electrodiagnostic testing by technicians
(even when a ph ysician is acting as a technician) appears to
be the primary reason for the variation in the results in the
different studies. It is hoped that the ev olution of the ordering of electrodiagnostic testing into the requesting of a
consult with a clinician with the capability to examine a
patient, review the imaging studies, and determine the testing approach most likely to answer specif ic questions will
lead to the answers to questions commonly asked by surgeons and other clinicians attempting to determine whether
a patient with low back symptoms has a radiculopathy,
myelopathy, or other neurologic lesion.
REFERENCES
1. Barker AT, Freeston IL, Jalinous R, et al. Magnetic stimulation of the
human brain. J Physiol 1985;369:3P.
2. Barker AT, Jalinous R, Freeston IL. Non-invasive magnetic stimulation
of the human motor cortex. Lancet 1985;1:1106–1107.
3. Britton TC, Meyer BU, Herdmann J, et al. Clinical use of the magnetic
stimulator in the investigation of peripheral conduction time. Muscle
Nerve 1990;13:396–406.
4. Chomiak J , Dv orak J, Antinnes J, et al. Motor evoked potentials: appropriate positioning of recording electrodes for diagnosis of spinal disorders. Eur Spine J 1995;4:180–185.
5. Reiners K, Herdmann J, Freund H-J. Altered mechanisms of muscular
force generation in lower motor neuron disease. Muscle Nerve
1989;12:647–659.
6. Herdmann J , Dv orak J, Rathmer L, et al. Conduction velocities of pyramidal tract fibres and lumbar motor nerve roots: nor mal values. Zent
Neurochir 1991;52:197–199.
7. Dvorak J, Herdmann J, Theiler R, et al. Magnetic stimulation of motor
cortex and motor roots for painless evaluation of central and proximal
peripheral motor pathways. Normal values and clinical application in
disorders of the lumbar spine. Spine 1991;16(8):955 – 960.
8. Hoffmann P. Ueber die Beziehung der Sehnenreflexe zur willkuerlichen
Bewegung und zum Tonus. Zschr Biologie 1918;68(1111): 351–370.
9. Stanley EF. Reflexes evoked in human thenar muscles during voluntary
activity and their conduction pathwa ys. J Neurol Neurosurg Psychiatry
1978;41:1016.
10. Braddom R, Joynson E. Standardization of H reflex and diagnostic use
in S1 radiculopathy. Arch Phys Med Rehabil 1974;55:1661–1666.
11. Aiello I, Serra G, Migliore A. Electrophysiological findings in patients
with lumbar disc prolapse. Clin Neurophysiol 1984. 24(4): p.
3313–3320.
12. Shea P, Woods W, Werden D. Electromyography in diagnosis of nerve
root compression syndrome. Arch Neurol Psychiatry 1950;64:93–104.
13. Knuttsson B. Comparative value of electromyographic, myelographic
and clinical-neurological examinations in diagnosis of lumbar root
compression syndrome. Arch Orthop Scand Suppl 1961;49:1–135.
14. Kimura J. Electrodiagnosis in diseases of nerve and muscle: principles
and practice, 2nd ed. Philadelphia: F.A. Davis, 1989.
15. Wilbourn AJ, Aminoff MJ. The electrodiagnostic examination in
patients with radiculopathies. Muscle Nerve 1998;21:1612–1631.
16. Tullberg T, Svanborg E, Isacsson J, et al. A preoperative and postoperative study of the accuracy and value of electrodiagnosis in patients
with lumbosacral disc herniation. Spine 1993;18(7):837–842.
17. Aminoff M, Goodin D, Parry G. Electrophysiologic evaluation of lumbosacral radiculopathies: electromyo graphy, late response, and somatosensory evoked potentials. Neurology 1985;35:1514–1518.
18. Eisen A, Hoirch M. The electrodiagnostic evaluation of spinal root
lesions. Spine 1983; 8:1(459):98–106.
19. Fisher M, Shivde A, Texeira C, et al. Clinical and electrophysiological
appraisal of the significance of radicular injury in back pain. J Neurol
Neurosurg Psychiatry 1978;41:303–306.
20. Tonzola R, Ackil A, Shahani B, et al. Usefulness of electrophysiological studies in the diagnosis of lumbosacral root disease. Ann Neurol
1981;9:305–308.
21. Haldeman S, Shouka M, Robboy S. Computed tomography, electrodiagnostic and clinical findings in chronic worker’s compensation
patients with back and leg pain. Spine 1988;13:345–350.
22. LaJoie W. Nerve root compression: correlation of electromyographic,
myelographic and surgical findings. Arch Phys Med Rehab 1972;53:
390–392.
23. Lane M, Tamhankar M, Demopoulos J. Discogenic radiculopathy: Use
of electromyography in multidisciplinary management. NY State J
Med 1978;78:32–36.
24. Toyokura M, Ishida A, Murakami K. Follow-up study on F-wave in
patients with lumbosacral radiculopathy. Comparison between before
and after surgery. Electromyogr Clin Neurophysiol 1996;36:207–214.
25. Carter GT, Fritz RC. Electromyography and lower extremity short time
to inversion recovery magnetic resonance imaging f indings in lumbar
radiculopathy. Muscle Nerve 1997;20:1191–1193.
26. Zsu Y, Weber R, Li J, et al. F-waves of peroneal and tibial nerve provide unique information in ongoing L5 and S1 radiculopathies. Paper
presented at: Proceedings of the International Society for the Study of
the Lumbar Spine; Cleveland; 2002; No. 158.
27. Yazicioglu K, Ozgul A, Kalyon TA, et al. The diagnostic value of dermatomal somatosensory evoked potentials in lumbosacral disc herniations: a critical approach. Electromyogr Clin Neurophysiol 1999;39:
175–181.
28. Pape E, Eldevik P, Vanvik B. Diagnostic validity of somatosensory
evoked potentials in subgroups of patients with sciatica. Eur Spine J
2002;11:38–46.
29. Vohanka S, Dvorak J. Motor and somatosensory evoked potentials in
lumbar spinal stenosis. Paper presented at 40th Congress of the Czech
and Slovak Neurophysiology. Brno; 1993.
30. Haig AJ, LeBreck DB, Powly SG. Paraspinal mapping. Quantified needle electromyography of paraspinal muscles in persons without low
back pain. Spine 1995;20:715–721.
31. Haig AJ, Vamakawa K, Hudson DM. Paraspinal electromyography in
high lumbar and thoracic lesions. Am J Phys Med Rehabil 2000;79:
336–342.
32. Boden S, McCowin P, Davis D, et al. Abnormal magnetic resonance
scans of the cervical spine in asymptomatic subjects. J Bone Joint Surg
1990;72-A(8):1178–1184.
33. Vohanka S, Adamova B, Dusek L. Correlation between axial CT imaging of the narrow lumbar spinal canal and electrophysiological conduction studies. Eur Spine J 2002;11S:11–12.
34. Adamova B, Vohanka S, Dusek L. The contribution of an exercise
treadmill test to diagnosis in patients with mild lumbar spinal stenosis.
Eur Spine J 2002;11S:54.
CHAPTER 11
Sensorimotor Control of the Lumbar Spine
Sten H. Holm and Aage Indahl
ETIOLOGY OF LOW BACK PAIN
Low back pain has probably been an integral part of most
human lives through the ages. Tattoos on the back of the
“Iceman” recently found in the Swiss Alps ha v e been interpreted as a possible treatment for low back pain. At the
beginning of the 20th century, the sacroiliac joint was
thought of as the main pain generator of the back, and
sacroiliac dysfunction was described (1). With the passing
of time, other structures have come into focus. The term
“facet syndrome” was coined in 1933 by Ghormley,
although the meaning was different from that of today (2).
It is still an undefined entity as is the “sacroiliac syndrome” (3). The major breakthrough came in 1934 when
Mixter and Barr described how herniated nucleus material
from the intervertebral disc, pressing on the nerve root,
was the cause of sciatic pain (4). Although man y details are
now kno wn about what Mixter and Barr described , it is still
not fully understood why disc herniations occur. The
mechanisms behind spinal disorders can either act as single variables or in combination. Derangement in the lumbar intervertebral disc and zygapophysial joints can contribute at the same segmental level or at dif ferent le vels and
be independently painful causing direct and referred pain.
A similar situation can arise when the sacroiliac joint system itself is disturbed, or indirectly affected through
derangement in the lumbar spine or its supporting structures (5–10). The relationship between pain and structural
derangement is still not fully understood.
Low back pain is one of the most common medical
problems of the middle-aged population, and from society’s point of view, it is the most costly musculoskeletal
disease in industrialized countries today (11–13). In the
majority of cases, the origin of the pain remains obscure.
Much of low back pain is thought to arise from damage
to the intervertebral disc or the zygapophysial joints,
either directly through traumatic injuries or disc prolapse,
or indirectly through degenerati v e processes that transmit
unfavorable loading patterns onto other spinal str uctures
(e.g., ligaments, tendons, and supporting musculature) as
well as to the sacroiliac joint (8,14–20).
In the clinical situation, the surgery rate for chronic
back pain is still growing, thus indicating strong beliefs
in pathoanatomic derangements. Stability and instability
are terms that are fundamental in describing the function
of the different back structures. Despite the lack of a
working clinical definition for these terms in a biomechanical system, they are widely used. Instability of one
or more spinal segments is accepted as one cause for low
back pain and the growing number of spinal fusion operations supports this belief. The need for strong trunk
muscles and ergonomic advice to preserve the stability
and integrity of the spine have dominated conservative
treatments (e.g., stabilizing exercises) for many decades.
Even if there is no clear definition of instability, there
seems to be a common understanding that instability is a
situation where a pathologic motion occurs within the
motion segment. Attempts to measure such pathologic
motions using advanced techniques have not been able to
demonstrate instability (i.e., hypermobility) (21). However, this appears not to have had any impact on the clinical belief regarding the existence of instability. Terms
that cannot be defined, processes that cannot be measured, or exercises that have no clear criteria for being a
stabilizing exercise are of little value as scientif ic tools.
For biomechanical systems, it might be more useful to
use the terms such as motion, balance or postural contr ol,
and transfer of loading with regard to the function of the
spine and its motion segments. An increased insight and
understanding of the sensorimotor control system that
takes care of these functions may lead us closer to the
nature of low back pain.
In order to understand sensorimotor control, it is necessary to have knowledge of the different structures that
are involved. All clinical entities for the treatment of low
back pain are unfortunately lacking a physiologic backup
and verification from adequate experimentally controlled
trials. Injuries and structural degeneration cannot be
120
CHAPTER 11/SENSORIMOTOR CONTROL OF THE LUMBAR SPINE / 121
FIG. 11-1. Neuromuscular network connecting the central nervous system to peripheral structures
(shown in boxes): inter vertebral disc, zygapophysial joint, skin, and spinal muscles. (Reprinted with
permission from Holm S, Indahl A, Solomonow M. Sensorimotor control of the spine. J EMG & Kinesiol
2002;12:219-34.)
properly studied in the human situation because of the
ongoing aging process. There remains an absolute need
for interdisciplinary studies and experimental models in
order to evaluate the neuromuscular interaction and the
muscular control in the spine (22,23).
The peripheral part of the intervertebral disc, as well as
the zygapophysial joint capsules, are richly innervated by
different nerves that serve the function of pain and
mechanical reception (20,24–26) (Fig. 11-1). Both injury
as well as noxious stimulation of the spinal structures
have been shown to cause spasm of the lumbar muscles
and hamstrings (10,27), and may induce perturbations in
the proprioceptive function (14). Such observations indicate possible mechanisms for long-term activation of the
musculature as an important factor in low back problems.
LUMBAR AND SACRAL STRUCTURES AND
INNERVATION PATTERNS
This section contains a brief summary of important
structures and mechanisms involved in control and move-
ments of the back. Also discussed is how lesions in the
avascular supporting structures, depending on location,
size, and degree of inflammation, can cause perturbations
to the proprioceptive function of the different receptors
and result in increased or prolonged muscle activation
that may cause pain. Irritation of low threshold nerve
endings in the sacroiliac joint, intervertebral disc, or the
zygapophysial joint tissue may trigger a reflex activation
of the gluteal and paraspinal muscles that may become
painful over time. To come closer to a solution to many
low back problems, a better understanding of muscle
function and their interactions with the passive structures
through the neural structures is needed.
The Lumbar Intervertebral Disc
The intervertebral disc is a deformable connective
structure that allows mobility and transfer loads between
the vertebrae. A normal intervertebral disc consists of a
gel-like central nucleus pulposus designed to sustain
pressure. Surrounding the nucleus pulposus is a special
122 /SECTION I/BASIC SCIENCE
arrangement of layers of collagen fibers, known as the
annulus fibrosus, which can resist movements in all
directions due to their alternating oblique orientation. A
cartilaginous end plate joins the vertebral body and the
intervertebral disc, and provides a nutritional pathway to
the disc, which is an avascular structure (28).
In the superficial layers of the disc, nerves form simple
free endings in the fetal stage, which increase in number as
the fetus matures. During the postnatal period, various
types of receptors develop, and in adult material, five types
of nerve terminations can be found. The complexity of the
receptors on the surface of the annulus increases with age.
Within a given disc, the receptors are not uniformly distributed (Fig. 11-2).After postnatal development, there is a
relative decrease in the number of receptors in the anterior
region. In adults the greatest number of endings are found
in the lateral regions of the disc, a smaller number occur in
the posterior region, and the least number occur in the
anterior region. The source of the nerve endings in the lumbar disc is the lumbar sinuvertebral nerves and branches of
the lumbar ventral rami and the gray rami communicantes.
Each lumbar sinuvertebral nerve supplies the disc at its
level of entry into the vertebral canal and the disc above.
The posterolateral corner of each lumbar disc receives
branches from the lumbar ventral rami that originate just
outside the intervertebral foramina. This region of the disc
receives a branch from the gray ramus communicantes
before its connection with the ventral ramus. Branches of
the gray rami communicantes innervate discs at various
levels. Even though the lumbar intervertebral discs are
innervated by branches of the sympathetic nervous system,
it does not necessarily mean that afferent fibers from these
structures return to the nervous system through the sympathetic trunk. It has been suggested that somatic afferent
fibers from the discs simply use the course of the rami
communicantes to return to the ventral rami (20).
The presence of nerve endings in the lumbar intervertebral disc raises the question as to their function. Malinsky
(29) proposed a proprioceptive function based on its morphology; however Kumar and Davis (30) did not f ind any
evidence to support this theory. Two studies have demonstrated that mechanoreceptors are present in the outer annulus fibrosus of the intervertebral disc (20,31). Furthermore,
it has been shown that in the rat the dorsal portions of intervertebral discs from L1-L2 to L4-L5 are multisegmentally
innervated by the T11-L5 dorsal root ganglions (32).
For acute pain to occur, nerves must be involved. The
disc itself may be an intrinsic source of pain originating
from mechanical or chemical disturbances (33) (Fig. 11-
3). As previously described, it has been established that
the intervertebral disc receives innervation and that there
are pain potentials in the outer part of the annulus f ibrosus. It is possible that discogenic pain from a single level
may involve more than one recurrent branch of the spinal
nerves. Free nerve ending associated with blood vessels
in the disc may be considered as having a vasomotor or
vasosensory function, but because the annulus fibrosus
contains so few blood vessels, this is less likely to be the
function for the majority of the nerve f ibers in the disc.
Although there is no absolute explicit evidence that disc
pain can be ascribed to a particular type of nerve ending
in the disc, there is abundant evidence suggesting that the
disc can be painful (33,34).
The Zygapophysial Joints
Together with the intervertebral disc, the lumbar
zygapophysial joints, or more commonly called facet
FIG. 11-2. Schematic of the lumbar intervertebral disc showing nonuniform innervation in the peripheral part.
FIG. 11-3. Schematic of a lumbar inter vertebral disc showing lesions that may or may not cause pain.