Cervical Radiculopathy:

Diagnosis and Nonoperative Management

Marc J. Levine, MD, Todd J. Albert, MD, and Michael D. Smith, MD

Abstract

Cervical radiculopathy presents as pain in a dermatomal distribution. This fre- Minnesota, was estimated to be
quently represents compression of an exiting cervical nerve root by either a her- 83.2 cases per 100,000 population,
niated disk or a degenerative cervical spondylotic change. Most patients will with a higher rate for males than
improve with nonoperative treatment, and a small percentage will require fur- females. An age-specific peak
ther diagnostic evaluation and ultimately surgical intervention. An under- (202.9 cases per 100,000) was seen
standing of the normal anatomy and the pathologic changes in cervical radicu- in individuals aged 50 to 54 years.
lopathy will improve the understanding of diagnosis and decision making The radiculopathy was caused by a
regarding nonoperative interventions. An algorithmic approach for decision confirmed disk protrusion in 21.9%
making and a review of nonoperative management are presented. of the patients; spondylosis, with
J Am Acad Orthop Surg 1996;4:305-316 or without disk herniation, was
responsible in 68.4%. Forty-one
percent of the patients had a previ-
ous history of lumbar radiculopa-
Cervical radiculopathy is a clinical physician. Differentiating between thy. The most common presenta-
diagnosis based on a sclerotomal radiculopathy and peripheral tion was C7 monoradiculopathy,
distribution of motor and/or sen- nerve compression is a common followed by C6 monoradiculopa-
sory changes or complaints. Any diagnostic dilemma. Once a pre- thy. During the 4.9-year period of
process that causes impingement liminary diagnosis has been made, surveillance, 31.7% of the patients
of exiting cervical nerve roots appropriate imaging modalities had a recurrence of the condition,
can lead to a radicular disorder. should be utilized to determine the and roughly 25% of the sympto-
Impingement may be brought source of impingement. If any matic cohort underwent surgery.
about by acute pathologic changes question remains as to the cause At last evaluation, 90% of the
or by degenerative changes consis- of sclerotomal changes, nerve- patients who were not treated sur-
tent with cervical spondylosis. Ret- conduction studies may be useful. gically were asymptomatic.
ropulsed disk material, zygo- This review will address these
apophyseal joint hypertrophy, neu- issues as well as the nonoperative
rocentral joint hypertrophy, and management of cervical radicu-
other soft-tissue abnormalities all lopathy. Dr. Levine is in private practice in St.
may cause compression of an exit- Petersburg, Fla. Dr. Albert is Assistant
Professor of Orthopaedic Surgery, Thomas
ing nerve root. Chemical irritation Jefferson University, and Attending Surgeon,
of the nerve root due to neurohu- Epidemiology The Rothman Institute, Philadelphia. Dr.
moral factors has also been Smith is Attending Surgeon, Minnesota Spine
described. In a recent study,1 the records of Center, Minneapolis.
The accurate diagnosis of cervi- over 550 patients seen between
cal radiculopathy begins with a his- 1976 and 1990 with complaints of Reprint requests: Dr. Albert, The Rothman
Institute, 800 Spruce Street, Philadelphia, PA
tory and physical examination. An radiculopathy (average follow-up 19107.
appreciation of cervical spine period of almost 5 years) were
anatomy and the pathophysiology reviewed. The average annual inci- Copyright 1996 by the American Academy of
of radicular pain will greatly dence of cervical radiculopathy in Orthopaedic Surgeons.
enhance the diagnostic skills of the this population in Rochester,

Vol 4, No 6, November/December 1996 305

Cervical Radiculopathy

Anatomy ly by the vertebral artery and the pedicle to enter the neuroforamen.
anterior tubercle. The facet joint The ventral motor nerve root lies
Cervical radiculopathy is largely and the base of the lamina consti- anteroinferiorly in close proximity
secondary to mechanical compres- tute the lateral boundary, and the to the uncovertebral joint. The dor-
sion of exiting nerve roots. An vertebral body is the medial sal sensory nerve root lies near the
appreciation of both the osseous boundary. Because of the relation- superior articular process. As the
anatomy and the neuroanatomy in ship of these structures, osteophyte nerve root enters the neuroforamen,
this region enhances diagnostic formation, cervical instability, disk it is located medially at the level of
and therapeutic acumen. This brief protrusion, or congenital deformi- the tip of the superior articular
overview is restricted to the subaxi- ties may lead to compression and process. It then courses laterally
al cervical spine. subsequent radiculopathy. and inferiorly (Fig. 2). In the distal
Each subaxial cervical spine The neuroanatomy of the cervi- aspect of the neuroforamen, the cer-
motion segment consists of five cal spine is unique to this region vical nerve root (both the anterior
articulations (Fig. 1). Anteriorly, and is unlike that of either the tho- and posterior portions) forms the
the intervertebral disk functions as racic or the lumbar spine. Each cer- dorsal root ganglion. Just distal to
a joint, allowing motion in multiple vical root exits above the pedicle the dorsal root ganglion and out-
planes. Two neurocentral (unco- for which it is named except C8, side the neuroforamina, the anteri-
vertebral) joints lie along the pos- which exits above the T1 pedicle. or and posterior nerve roots join to
terolateral aspect of the vertebral For instance, the C3-4 disk space or form the spinal nerve. The spinal
body and provide articulation foramen transmits the C4 nerve nerve then divides into ventral and
through osseous projections ex- root. dorsal rami. As mentioned previ-
tending to the vertebral body The ventral (anterior) motor ously, the costotransverse lamella
above. These joints of Luschka lie nerve root consists of six to eight transmits the ventral ramus.
between the disk and the nerve- nerve rootlets exiting the spinal In the normal spine of a young
root canal. The facet joints located cord. The dorsal (posterior) senso- person, the cervical nerve root
posteriorly are angled 30 to 50 ry nerve root consists of six to eight occupies approximately one third
degrees to the transverse plane.2 nerve rootlets entering the spinal of the available space in the neuro-
The intervertebral foramina are cord. The two unite to form the cer- foramen. This proportion may
bounded anteriorly by the verte- vical spinal nerve root, which pass- increase with age and degenerative
bral body, the uncinate process, es at an angle of 45 degrees to the changes. In addition, this propor-
and the disk; posteriorly by the coronal plane and inferiorly at 10 tion may increase in an extended
facet joints; and cranially and cau- degrees to the axial plane. The cer- neck because of the relative de-
dally by the pedicles. The subaxial vical nerve root then passes directly crease in foraminal size in this posi-
cervical foramina are approximate- laterally to the corresponding cervi- tion.4 (Henceforth, the terms “cer-
ly 9 to 12 mm in height and 4 to 6 cal disk and over the corresponding vical nerve” and “cervical root”
mm in width. The foramina are
aligned obliquely 45 degrees to the
sagittal plane.3
Projecting laterally from each 1
vertebral body is a rudimentary
2
rib, or costal process, that ends in
the anterior tubercle. An embry- Fig. 1 Superior oblique
view of C4 and C5. 1 =
ologic transverse process extends uncinate process; 2 =
from the lateral masses to the pos- C4
3 superior intervertebral
terior tubercle. These two osseous notch; 3 = foramen trans-
versarium; 4 = posterior
elements fuse laterally to form the 4 tubercle of transverse
true transverse process of the cervi- 7
process; 5 = spinal nerve
cal spine. A groove, known as the foramen; 6 = anterior
tubercle of transverse
costotransverse lamella, transmits process; 7 = C4-5 disk.
the ventral ramus of each exiting C5
spinal nerve and is bounded poste-
riorly by the transverse process and 6 5
the posterior tubercle and anterior-

306 Journal of the American Academy of Orthopaedic Surgeons

 Marc J. Levine, MD, et al

at some time during their life-

time. 9,10 Lees and Turner 11 re-
viewed the natural history of cervi-
Spinal cord
Dorsal root cal radiculopathy and found that
the condition rarely progressed to a
myelopathic state. In patients
treated nonoperatively, however,
long-term follow-up revealed per-
Dorsal root
Articular ganglion
sistent symptoms in 66% of the
process population. In two other stud-
ies, 12,13 23% of the patients with
persistent neck or radicular pain
Spinal were unable to return to their origi-
nerve nal occupation. It is unclear from
these studies whether there were
specific variations between the
Ventral
root Disk treatment received by the patients
who were able to return to work
Fig. 2 Cross-sectional view of the neural structures of the cervical spine. and the treatment received by
those who were not.

will be used to mean the spinal The vascular response to com- History and Physical
nerve at the cervical level unless pression was studied by Olmarker Examination
otherwise specified.) et al7,8 in a porcine model involving
the cauda equina. These investiga- The presentation of cervical radicu-
tors found that blood flow to some lopathy varies greatly among
Pathophysiology venules stopped with 5 to 10 mm patients. Presenting complaints
Hg of pressure, although the venu- can include pain, paresthesias, and
The mechanical nature in which lar occlusion pressure ranged from motor weakness in different combi-
cervical nerves become compressed 5 to 60 mm Hg. The study also nations and proportions. Classi-
has been well studied both clinical- suggested that intraneural edema cally, most patients complain of
ly and radiographically. However, occurred more readily in nerve significant radicular pain and
the mechanism by which this com- roots than in peripheral nerves. In referred trapezial and periscapular
pression elicits pain is, as yet, poor- addition, edema formation was pain.13 Sensory disturbances asso-
ly understood. Cornefjord et al5 more pronounced in nerves that ciated with nerve compression
used a pig model to study the were rapidly loaded than in nerves often do not follow a strict der-
effects of chronic nerve root and exposed to a slower rate of com- matomal pattern. In fact, in a re-
dorsal ganglia compression. The pression. Similarly, there was a view of the data on more than 840
concentrations of the neuropep- more profound effect on nutritional patients, Henderson et al 14 found
tides substance P and substance status in roots loaded rapidly than that only 55% of patients with a
VIP were measured in compressed in those loaded more slowly. nerve-root compression had pain
roots. The authors found a signifi- in a strictly radicular pattern.
cantly increased concentration of Other studies have shown that
substance P in compressed nerve Natural History motor deficits are present in 60% to
roots after 1 week of compression 70% of patients with radiculopathy
but not after 4 weeks. Numerous Although the natural history of cer- and that roughly 70% have reflex
other chemical mediators of pain vical radiculopathy has not been as changes.15
have been implicated as contribut- well studied as that of lumbar Patients often describe symp-
ing to radicular neck pain. These radiculopathy, it has been estimat- toms that correlate with various
chemical mediators are largely ed that slightly more than half of head positions. Many will find
involved in the inflammatory the adult population will experi- relief with decreased neck motion
response to compression.6 ence neck and radicular symptoms when pain is due to acute cervical-

Vol 4, No 6, November/December 1996 307

Cervical Radiculopathy

disk compression. Patients often

report exacerbation of symptoms C5
C4 C5
with neck hyperextension, particu-
larly when the head is tilted C6
C6
toward the affected extremity. This
position decreases the size of the
neuroforamina. Conversely, pa- C5
tients report relief with a slight T3
C7
amount of neck flexion.
A number of maneuvers can be C5 C7
used during physical examination C7
C6
to support the diagnosis of cervical T2 C6
radiculopathy. As already de- C7 C6
scribed, a combination of neck C6
extension and head tilt toward the
affected limb (the modified Spurling C7
C6
maneuver) may reproduce radicular C7
T1
complaints, as can a Valsalva ma- C7
neuver. The shoulder abduction re- C6
C6
lief sign, as described by Davidson
et al,16 has been shown to be specific C6
C-8
for radiculopathy caused by soft C6
disk herniations.17 C8
A confusing presentation com-
plex may arise in men or women
who complain of breast pain while C7
being evaluated for breast lesions C7
or anginal symptoms. In one
Fig. 3 Radicular patterns of distribution for the C4 through T3 dermatomes.
study,18 18 women had breast pain
that was relieved with therapeutic
measures directed at documented
C6 and C7 radiculopathies. Breast
pain is a much less common pri- C3 radiculopathy, most often Compression of the C5 nerve
mary presenting complaint, but caused by disk disease at C2-3, is root classically produces pain
should be considered when other not common. Presenting com- and/or numbness in an “epaulet”
manifestations of cervical radicu- plaints include headaches and pain pattern that includes the superior
lopathy are absent. Cardiac work- along the posterior aspect of the aspect of the shoulders and the lat-
ups have also been performed for neck that extends to the posterior eral aspect of the upper arm.
this radiculopathic presentation of occipital region and occasionally to Deltoid motor function is often
“cervical angina.” Brodsky19 re- the ear. There are no motor deficits, weakened, as in an intrinsic shoul-
viewed the data on 438 patients which may make differentiation der disorder; the diagnosis of
with such symptoms, 88 of whom from tension headaches difficult. radiculopathy at this site is crystal-
he had treated surgically. Disk protrusions at C3-4 typical- lized by observing the absence of
Full descriptions of the various ly involve the C4 spinal nerve root impingement signs or pain with
nerve-specific radicular symptoms and most commonly present with passive shoulder motion. Patients
are available in many textbooks, neck and trapezial pain. There are may complain of difficulties with
but a brief presentation of typical no motor deficits, and diaphrag- activities of daily living if there is
complaints for specific nerve roots matic involvement has not been involvement of the supraspinatus,
(Fig. 3) will be presented here. well documented. Patients occa- infraspinatus, or elbow flexors.
Radiculopathies above C2 are sionally complain of numbness and Depression of the biceps reflex is
extremely rare. They can cause jaw pain at the base of the neck that an inconsistent finding.
pain and occipital headaches, but extends to the shoulder and scapu- C6 involvement may present as
no motor deficit is seen. lar region. pain or sensory abnormalities

308 Journal of the American Academy of Orthopaedic Surgeons

 Marc J. Levine, MD, et al

extending from the neck to the ly to rule out an intrinsic hand um of degenerative changes that
biceps region, down the lateral problem or compressive neuropa- involve the disk, the two neurocen-
aspect of the forearm to the dorsal thy as the cause of ulnar symptoms. tral joints, and the two facet joints.
surface of the hand, between the Myelopathy must also be ruled out. Uncovertebral osteophytes cause
thumb and index finger, and radiculopathy by compressing the
including the tips of these fingers. nerve root anteriorly. Less com-
The brachioradialis reflex may be Cervical Disk Herniation monly, osteophytes extending from
depressed, and wrist extensor and Degenerative the ventral portion of the superior
weakness is usually present. The Spondylosis articular process can cause compres-
infraspinatus, serratus anterior, tri- sion by neuroforaminal narrowing.
ceps, supinator, and extensor polli- Acute radiculopathies are typically Neuroforaminal narrowing also
cis muscles may also be affected. secondary to disk herniations. More occurs as a result of degenerative
The C7 nerve root is commonly insidious symptoms occur as a result disk disease and the associated
involved. Pain and sensory abnor- of degenerative changes. Acute disk decrease in disk height. 22,23 The
malities extend down the posterior herniation is more common in the term “hard disk” is used to describe
aspect of the arm and the postero- younger population and is referred osteophytes that arise due to degen-
lateral aspect of the forearm and to as a soft disk herniation. Three erative spondylosis and that may
typically involve the middle finger, types of soft disk herniation have compress the spinal cord or nerve
which is rarely affected in C6 disor- been described by Stookey20 and by root. The remainder of this article
ders. Absence of the triceps reflex Rothman and Marvel 21 (Fig. 4). will address radiculopathies caused
is common, and triceps weakness is Intraforaminal herniation is the most by soft and hard disks.
almost always present. The wrist common and is often evidenced by
flexors, wrist pronators, finger radicular symptoms in a derma-
extensors, and latissimus dorsi may tomal distribution. Posterolateral Differential Diagnosis
also be affected. herniation results in predominantly
C8 radiculopathy is least likely motor symptoms, including weak- Clinical evaluation requires the
to be associated with pain. Sensory ness and atrophy. Midline hernia- physician to rule out a number of
changes are usually restricted to tion may result in myelopathy. processes that can mimic cervical
below the wrist; motor involvement Unlike the lumbar spine, where radiculopathy. While the manage-
of the interossei makes differentia- three joints are involved in each ment and etiology of these disor-
tion from ulnar neuropathies and motion segment, in the cervical ders are beyond the scope of this
intrinsic hand disorders difficult. spine each motion segment involves discussion, we will present a brief
The rarity of C8 radiculopathy five articulations. The term “cervical overview and discuss pertinent dif-
should make one search persistent- spondylosis” describes the continu- ferences on clinical examination.

A B C

Fig. 4 Types of soft disk herniation. Intraforaminal (A), posterolateral (B), and midline (C) protrusions are usually associated with dif-
ferent clinical presentations.

Vol 4, No 6, November/December 1996 309

Cervical Radiculopathy

Myelopathy and weakness of the flexor pollicis longus, the thenar musculature,
Cervical myelopathy is a dis- longus, pronator quadratus, and and the median nerve flexors to the
tinct entity from both a diagnostic flexor digitorum profundus of the index and long fingers. Compres-
and a treatment perspective. index finger, but no sensory sion of the ulnar nerve in Guyon’s
Patients with cervical myelopathy deficits. These findings may be canal typically affects the superfi-
may present with complaints of confused with those of C8 radicu- cial and deep branches, leading to a
bowel and bladder changes, sexu- lopathy except for the absence of a sensory deficit along the volar por-
al dysfunction, gait disturbance, sensory deficit. From a motor tion of the ulnar one and a half dig-
and difficulty with fine-motor standpoint, true C8 radiculopathies its. Dorsal sensation in these digits
function of the hand (writing, but- are characterized by weakness in remains normal in this syndrome
toning, handling change). Clinical all ulnar nerve–innervated mus- because the nerves to this region do
examination is remarkable for cles. not pass through the canal. Motor
upper motor neuron findings The classic carpal tunnel syn- deficits are consistent with involve-
including, but not limited to, drome also mimics a C6-C7 radicu- ment of the muscles of the deep
crossed and inverted radial reflex- lopathy from a sensory standpoint. motor branch of the ulnar nerve.
es, clonus, a Babinski sign, and a However, the triceps and wrist A true C8 or T1 radiculopathy is
finger-release sign (Hoffman sign). extensor muscles are not weak- characterized by sensory distur-
While peripheral radiculopathic ened, as they are innervated above bances on both the volar and the
symptoms may accompany mye- the carpal tunnel. The thenar dorsal surface and causes motor
lopathy, the presence of upper motor weakness associated with deficits of median nerve–innervat-
motor neuron changes requires a carpal tunnel syndrome may sug- ed muscles, such as the T1-depen-
different treatment protocol. It gest T1 radiculopathy except that dent thenar muscles. Remem-
must be remembered that a small other T1 nerve–innervated muscles, bering that C8 and T1 radicu-
number of patients with myelopa- including the hypothenar and lopathies are remarkably rare and
thy may present with only hand ulnar nerve–innervated dorsal using electrodiagnosis judiciously
dysfunction as an initial complaint. interosseous muscles, are normal. should ensure the appropriate
Diagnosis of entrapment of a diagnosis of a peripheral compres-
Entrapment Syndromes palmar cutaneous nerve is based sion syndrome.
A number of entrapment syn- on the absence of motor deficits The radial nerve is commonly
dromes can mimic cervical radicu- despite sensory changes in the C6 compressed at the elbow by a num-
lopathy. Diabetes, smoking, alco- distribution. Electrodiagnosis ber of structures. Usually, only the
hol consumption, rheumatoid (measurement of nerve-conduction motor branch (posterior inter-
arthritis, and hypothyroidism are velocities) is useful in identifying a osseous nerve) is involved, affect-
generally considered risk factors peripheral neuropathy. ing the extensor digitorum commu-
for peripheral nerve entrapment.24 Diagnostically, peripheral ulnar nis, extensor carpi ulnaris, abduc-
Peripheral median neuropathies neuropathies can also be difficult to tor pollicis longus, and extensor
may be confused with cervical differentiate from true radicu- pollicis longus. This is also consis-
radiculopathies. The pronator syn- lopathies. Cubital tunnel syndrome tent with C7 radiculopathy, but
drome may mimic a C6-C7 radicu- typically presents with weakness of unlike that condition, there is no
lopathy with sensory involvement ulnar nerve–innervated muscles sensory change and no involve-
of the radial three and a half fin- distal to the elbow, with corre- ment of the triceps or wrist flexor
gers and median nerve–innervated sponding sensory changes, and musculature.
muscles. The muscles affected in may be confused with a C8 or T1
the pronator syndrome include the radiculopathy. The flexor carpi Thoracic Outlet Syndrome
pronator teres and the flexor carpi ulnaris, flexor digitorum profundus Symptoms of thoracic outlet
radialis; the radial nerve–innervated to the ring and little fingers, in- syndrome often involve the contri-
muscles of the C6 and C7 derma- terosseous, and hypothenar mus- butions of the lower cervical roots
tomes (wrist extensors and triceps) cles are affected in both this entrap- to the brachial plexus and present
are spared. ment syndrome and C8 or T1 as changes in median and ulnar
Entrapment of the anterior radiculopathy. Muscles that are not distribution. Either vascular or
interosseous nerve (a motor branch affected in cubital tunnel syndrome neurogenic causation is possible.
of the median nerve) may present but are affected in C8 or T1 radicu- On physical examination, the pres-
with pain in the proximal forearm lopathy are the flexor pollicis ence of vascular bruits, asymmetric

310 Journal of the American Academy of Orthopaedic Surgeons

 Marc J. Levine, MD, et al

pulses, or thenar muscle wasting Evaluation An MR imaging study should,

greater than interosseous muscle at a minimum, include both T1-
wasting is more consistent with An algorithmic approach is justi- and T2-weighted sequences. The
thoracic outlet syndrome than fied for the logical workup and T1 sagittal examination provides
radiculopathy. Radiographs show- treatment of cervical radiculopathy an excellent survey of the cervical
ing cervical ribs also implicate tho- (Fig. 5). spine, which is another valuable
racic outlet syndrome rather than asset of this modality. The axial T1
C8 or T1 radiculopathy. Imaging Modalities images provide insight into both
Other disease processes that can The role of plain radiography is intraspinal and extraspinal dis-
mimic cervical radiculopathy in- somewhat limited in the evaluation orders, as well as the intrathecal
clude reflex sympathetic dystro- of the nerve roots. It remains an nerve root anatomy. On T1 images,
phy, herpes zoster, brachial neuri- important initial study to rule out a hypointense signal is common for
tis, and rotator cuff and shoulder instability or pathologic changes in herniated degenerative disks, calci-
girdle injuries.21 the bone. Oblique views of the cer- fied ligaments, and bone spurs,
vical spine can show narrowing of making differentiation of these
Other Causes of Compression the neuroforamina secondary to structures more difficult. The T2-
While most nontraumatic cervi- degenerative changes. Cervical in- weighted sequence or variants
cal radiculopathies are caused by stability may be visualized with thereof may provide a “myelo-
acute disk disease and degenera- dynamic flexion and extension graphic” view of the cervical spine.
tive changes, there are other, less films. Unlike MR imaging of the lumbar
common causes of compression. A Plain radiography may be used spine, imaging of the cervical spine
thorough history and physical as an initial study for evaluating may be less accurate because of
examination and the judicious use neck pain associated with radicu- problems with motion artifact.
of imaging modalities can expedite lopathy as long as the sensitivity Other sequences (proton-density
effective diagnosis and treatment. and specificity of this test are and fat-suppressed fast spin-echo
Both intraspinal and extraspinal understood. Friedenberg and images) may add further informa-
tumors can cause radicular com- Miller 25 showed that by the fifth tion.
plaints by direct compression or decade 25% of their asymptomatic Numerous studies have com-
secondary to structural collapse of patients had evidence of degenera- pared the accuracy of myelogra-
bone elements. In general, most tive changes; by the seventh phy, CT (with and without con-
malignant tumors will cause decade, this number rose to 75%. trast-material enhancement), and
myelopathic symptoms bilaterally. When 92 asymptomatic patients MR imaging. 26,27 Modic et al 28
Unilateral radicular changes may were compared with a group of prospectively compared the accura-
be seen with osteochondromas matched symptomatic patients, the cy of MR imaging, myelography,
extending from the posterior ele- only radiographic difference was a and CT-myelography for evalua-
ments. Schwannomas that arise higher rate of degenerative changes tion of cervical radiculopathy.
from the nerve sheath also cause at the C5-6 and C6-7 spaces in the Magnetic resonance imaging was
unilateral radiculopathies, which symptomatic population. Plain as sensitive as CT-myelography for
often progress to myelopathy. radiographs should be obtained identifying a diseased segment but
These lesions are more commonly only after conservative manage- was less accurate for identifying the
intradural and may be exacerbated ment for 4 to 6 weeks has failed. exact disease process. The authors
by the Valsalva maneuver. Extra- Magnetic resonance (MR) imag- concluded that an MR imaging
spinal radiculopathies may be ing has had a significant impact on study accompanied by nonen-
caused by direct extension of thy- the radiographic evaluation of cer- hanced CT provides excellent visu-
roid, esophageal, and pharyngeal vical radiculopathy. Soft-tissue alization of the cervical spine. In a
tumors. Pancoast tumors have visualization is unsurpassed by prospective study of 100 patients,29
eroded through the pedicles of C7 that possible with any other modal- MR imaging was shown to be as
and T1, causing C8 radiculo- ity. The physician should recog- accurate as postmyelography CT in
pathies. Cervical radiculopathies nize that visualization of foraminal the evaluation of cervical radicu-
have also been caused by soft-tis- stenosis may not be appreciated as lopathy. In another prospective
sue compression secondary to sar- well on this study as on computed study, Neuhold et al30 correlated
coidosis and arteriovenous malfor- tomography (CT) performed with MR imaging findings with intraop-
mations. or without a contrast medium. erative pathologic findings and

Vol 4, No 6, November/December 1996 311

Cervical Radiculopathy

Acute radiculopathy

Nonprogressive Progressive deficit

nondisabling deficit or disabling weakness
of antigravity muscles

Nonoperative management
(collar, traction, NSAIDs,
heat) for 10-14 days Cervical spine series
(including flexion-extension),
MR imaging

Physical therapy
for 2-3 weeks

Inconclusive Positive
Reevaluation

CT-myelography
Decreased Unclear No change Progressive
symptoms diagnosis deficit

Continue EMG Cervical spine Negative Positive Nonpathologic

nonoperative series (including or (correlative disease process
management flexion-extension) inconclusive at specific
root level)

Negative Positive Neurologic consultation Surgery Rheumatologic

(vertebral destruction (MR imaging of brain, or further
and instability) EMG, possibly neurologic
spinal tap) workup
Continue nonopera-
tive management
Immediate evaluation
for 4 weeks
(MR imaging, tumor
workup, medical
evaluation)
Reevaluation

No improvement

MR imaging

Positive study with Negative Positive study

correlation with showing evidence
specific root findings of pathologic
(especially motor process
weakness)
Consider
rheumatologic
and/or further Immediate evaluation
Consider surgical neurologic (MR imaging, tumor
management, CT- workup workup, medical
myelography if nec- evaluation)
essary for specificity

Fig. 5 Algorithm for temporal sequence of diagnosis and nonoperative management of acute cervical radiculopathy. CT = computed
tomography; EMG = electromyography; MR = magnetic resonance; NSAIDs = nonsteroidal anti-inflammatory drugs.

312 Journal of the American Academy of Orthopaedic Surgeons

 Marc J. Levine, MD, et al

found MR imaging to be a viable including flexion and extension tude proportional to muscle atro-
alternative to postmyelography CT. views, should be the initial diag- phy. Significant alterations may be
They suggested the use of MR nostic imaging modality used to seen in polyradiculopathies with
imaging as the initial diagnostic evaluate the possibility of cervical multiple muscle involvement.
modality when cervical disk dis- radiculopathy and to rule out insta- Dramatic changes are more com-
ease is suspected. bility or pathologic changes in monly seen in lumbar stenosis but
A cervical myelogram outlines osseous structures. If the clinical may occur in cases of severe cervi-
the spinal cord and exiting nerve findings support a diagnosis of cer- cal spondylosis.
roots with radiopaque dye. The vical radiculopathy due to nontrau- Nerve-conduction velocity and
study is performed by introducing matic soft disk or hard disk latency changes are not typically
dye into one of two areas. The changes, MR imaging should be found in cervical radiculopathies
water-soluble agent may be injected performed. If surgical treatment unless there is extreme demyelina-
via the C1-2 interval, allowing the becomes necessary and the MR tion of axons. Because the lesion is
dye pool to gravitate caudally. This images are nonspecific (especially proximal to the region tested, the
expeditious means of visualization in a case of multilevel spondylosis), usefulness of peripherally oriented
may be complicated by the inherent myelography and postmyelogra- studies is limited.32
risks of introducing a needle at this phy CT can be pursued for greater Another alternative for electrical-
interspace. If dye is introduced into specificity. ly evaluating cervical radicu-
the lumbar region, the patient must lopathies is cervical root stimulation
be placed in a position that forces Electrodiagnosis (CRS). With this technique, cervical
the dye pool cranially. Although Nerve compression may lead to roots are stimulated by placing
needle placement is less critical in motor, sensory, and/or autonomic monopolar needles in the para-
this region, a delay in visualization, changes. In cases of polyradicu- spinal muscles, and compound-
dilution of the dye load for cervical lopathy or difficult clinical diag- muscle action potentials are record-
imaging, and the risk of dye pass- noses, the use of modalities that ed in the biceps, triceps, and abduc-
ing through the foramen magnum utilize electrical stimulation may be tor digiti minimi muscles.
detract from the attractiveness of necessary to help differentiate Electromyography has histori-
this protocol. As with all imaging radiculopathy from peripheral cally been the modality of choice
modalities, clinical correlation is compression syndromes. for differentiating cervical radicu-
imperative in formulating a treat- The usefulness of electromyog- lopathies from more peripheral dis-
ment plan. Hitselberger and raphy (EMG) and nerve-conduc- turbances. Electromyographic
Witten31 reported a 21% incidence tion velocity studies is dependent changes represent a continuum
of cervical filling defects in 300 on their ability to detect motor that begins with a decrease in
asymptomatic patients who under- changes occurring as a result of motor-unit potentials and progress-
went myelography for evaluation nerve compression. In radiculopa- es to fibrillation potentials of multi-
of acoustic tumors. thy, abnormalities in sensory-nerve ple muscles. Many of the changes
The accuracy of cervical myelog- action potentials (SNAPs) are seen with chronic radiculopathies
raphy alone has been estimated to uncommon. Typically, compres- are not unique to radiculopathy
range from 67% to 92%.26 For this sion that leads to cervical radicu- and require careful interpretation.
reason, cervical myelography is lopathy occurs proximal to the dor- Electromyography has been shown
often accompanied by CT. Axial sal root (sensory) ganglion. Unless to correlate better with clinical
images of dye-enhanced neural ele- the dorsal root ganglion at the dis- symptoms than does plain radiog-
ments offer excellent visualization talmost aspect of the neuroforamen raphy. In a retrospective review of
of nerves in relation to surrounding is involved, the SNAPs will remain 108 patients,33 the disk height and
osseous structures. The combina- normal. In the case of a compres- neuroforaminal size were of little
tion of myelography and postmye- sive brachial plexopathy, the use in predicting clinical findings,
lography CT provides important SNAPs are routinely abnormal uni- in contrast to EMG.
details required for preoperative laterally because the encroachment The literature currently favors
planning after failure of nonopera- is distal to the sensory ganglion. CRS over EMG for accurate differ-
tive management. Bilateral SNAP changes are sugges- entiation of cervical radiculopathy.
In summary, we believe that tive of peripheral polyneuropathy. Berger et al34 compared CRS with
after the initial observation period Compound-muscle action po- conventional EMG and evaluation
of 4 to 6 weeks, plain radiography, tentials show a decrease in ampli- of nerve conduction and late

Vol 4, No 6, November/December 1996 313

Cervical Radiculopathy

responses in 34 patients. Of the 18 this generalization is a progressing radicular symptoms. With this pil-
patients with clinical evidence of neurologic deficit or a deficit that low arrangement, the head is
radiculopathy, 11 had abnormal disables the patient (severe deltoid flexed slightly, and the shoulders
EMG studies, and all 18 had abnor- or wrist extensor weakness). Com- are internally rotated. The neck
mal responses to CRS. Of the 16 pression secondary to trauma, in- flexion serves to enlarge restricted
patients with symptoms but no fection, tumor, or other pathologic neuroforamina, and the internal
signs of radiculopathy, 5 had changes in the soft tissues is rotation of the shoulders decreases
abnormal EMG studies, and 9 had excluded from this discussion. The the stretching of the cervical
abnormal responses to CRS. efficacy of treatment modalities is nerves. This position may be
In a more recent study, 35 the often related to the pathophysio- specifically suited for patients with
authors compared the CRS, EMG, logic processes that cause the pain. radicular symptoms.41 There are
and nerve-conduction velocity The acuity of the presenting symp- commercially available cervical pil-
findings in 32 patients with both toms is a factor in the selection of lows that simulate the inverted-V
clinical signs and symptoms sug- appropriate treatment protocols. pillow arrangement.
gestive of cervical radiculopathy. Acute neck pain due to cervical Home traction devices that
Conventional EMG was positive in radiculopathy can be treated initial- attach to door frames have provid-
a little over half of the patients, ly with a short course of cervical ed relief for some patients with
while CRS was abnormal in more immobilization in a soft collar. radicular symptoms. Traction
than 75%. The CRS study was pos- Patients with acute pain typically forces of 8 to 12 lb are generally
itive in 25 patients. Thirteen of the present to their physician within 2 applied for 15- to 20-minute peri-
25 patients subsequently under- weeks after the onset of symp- ods. Theoretically, traction forces
went surgery, which documented toms.36 Immobilization serves to relieve pressure from compressed
intraoperative findings consistent decrease the acute inflammatory nerves. Some believe that traction
with radiculopathy. Only 10 of the response and helps to decrease may increase blood flow and
13 had a positive EMG study. pain. Prolonged immobilization decrease ischemia while flushing
Overall, the role of electrodiag- should be avoided, however, out inflammatory by-products.
nostic testing is to assist in difficult because the cervical musculature The angle of traction application
diagnostic situations and to rule atrophies rapidly. The duration of has been studied by Colachis and
out peripheral neuropathies, but immobilization should not exceed Strohm.42 The maximum interver-
not to be an additional test for con- 10 days to 2 weeks and should be tebral distance was achieved with
firmation of a clear monoradicu- followed by gradual weaning. traction forces applied at an angle
lopathy, which is a clinical diagno- During the weaning period, the of approximately 24 degrees of
sis. Whether the physician chooses paraspinal muscles can be strength- flexion. The application of traction
EMG or CRS, the temporal se- ened with isometric exercises.37 should not be initiated until muscle
quence remains constant. These A review of the literature dis- spasms have been alleviated. This
studies should follow plain radiog- closed that the efficacy of soft col- technique is usually contraindicat-
raphy and a period of conservative lars in treating cervical radiculopa- ed in patients with myelopathic or
management and should precede thy is still unclear. While some long-tract signs.41 Care should be
more complex imaging modalities. have reported benefits from their taken that traction does not hyper-
In practice, we rarely order electro- use, others have found cervical col- extend the neck, thereby compress-
diagnostic testing in a workup lars of little value.38-40 While many ing the foramina. A survey of the
other than for patients with unusu- patients are comfortable with the literature reveals conflicting
al presentations or diabetes or to higher part of the collar anteriorly, reports regarding the benefits of
rule out a peripheral compression the extension this necessitates cervical traction.38,43
syndrome. aggravates the condition in other The pharmaceutical manage-
patients, who prefer to reverse the ment of cervical radiculopathy can
collar to encourage neck flexion. be divided into three categories of
Nonoperative Management A very short course of bed rest medication. Narcotic analgesics
can also serve as a form of cervical can be used in the acute setting but
Initial nonoperative management is immobilization and has the benefit should be used cautiously because
appropriate in almost all cases of of eliminating axial forces caused of the addictive and depressive
cervical radiculopathy caused by by gravity. The inverted-V pillow side effects. Drugs directed at
soft or hard disks. The exception to arrangement can further relieve muscle spasm also may serve a role

314 Journal of the American Academy of Orthopaedic Surgeons

 Marc J. Levine, MD, et al

in the acute setting. Spasms occur radicular disorder.45 Recommend- will continue to treat patients on
as a result of increased muscle ten- ed injections include lidocaine and the basis of available retrospective
sion at insertion sites, which leads methylprednisolone acetate or tri- information and anecdotal experi-
to avascularity and buildup of amcinolone diacetate. The anti- ence.
anaerobic metabolic by-products. 41 inflammatory effects of cervical
Antispasmodic agents disrupt the epidural injections can be repeated
cycles of repetitive spasm. with multiple injections; however, Summary
Nonsteroidal anti-inflammatory the risks of needle placement
drugs decrease pain brought about should be weighed. Diagnosis of cervical radiculopa-
by the inflammatory process. Appropriate physical therapy thy requires the physician to
These agents prevent the formation protocols require a coordinated appreciate not only the cervical
of various substances in the cyclo- effort between the physician and anatomy but also the numerous
oxygenase pathway, which have the therapist. Patients with initial disease processes that can mimic
been implicated as contributing to acute radicular symptoms may cervical radiculopathy. Initial
the pain response in cervical benefit from immobilization, fol- plain radiographs are useful as a
radiculopathy. Systemic monitor- lowed by heat and cold therapy. preliminary study to check for
ing, particularly of the liver and the Electrical stimulation applied by instability and gross structural
gastrointestinal system, is impor- the therapist can help break spasm change. Conservative manage-
tant with any long-term usage. cycles. During the weaning period ment protocols should be started
Oral corticosteroids are generally from use of a cervical collar, iso- almost immediately in an effort to
not recommended for cervical metric neck-strengthening proto- control the inflammatory process.
radiculopathies because of the asso- cols are introduced. Stretching Electrodiagnostic tests can be used
ciated risk factors. Olmarker et al44 exercises can also be instituted at in particularly confounding cases
used a porcine model to explore the this time. If the patient is free of but should not be considered part
effects of intravenous administra- pain after 6 weeks, more aggressive of the routine workup. Imaging
tion of methylprednisolone on nu- exercise regimens can be intro- modalities are useful in confirming
cleus pulposus–induced nerve root duced to build up the paraspinal clinical diagnoses and identifying
injury. Their results suggested that muscles and protect the neck from the cause of compression when ini-
high doses of this agent may reduce further attacks. tial nonoperative protocols fail.
nerve root damage secondary to It should be noted that none of More aggressive therapy should be
compression by disk material if the above-mentioned nonoperative tailored to the patient’s symptoms
administered in the first 24 to 48 approaches to managing cervical and the chronicity of pain. When
hours. radiculopathy has been subjected nonoperative management is inef-
Epidural administration of corti- to prospective randomized efficacy fectual or disabling weakness, pro-
costeroids has been shown to be trials to gauge efficacy, nor have gressing radiculopathy, or myelop-
most beneficial in patients with they been compared with observa- athy is present, consideration of
both signs and symptoms of a tion alone. Until this occurs, we surgical intervention is warranted.

References
1. Radhakrishnan K, Litchy WJ, O’Fallon 4. Rauschning W: Anatomy and pathol- 7. Olmarker K: Spinal nerve root com-
WM, et al: Epidemiology of cervical ogy of the cervical spine, in Frymoyer pression: Nutrition and function of the
radiculopathy: A population-based JW, Ducker TB, Hadler NM, et al (eds): porcine cauda equina compressed in
study from Rochester, Minnesota, 1976 The Adult Spine: Principles and Practice. vivo. Acta Orthop Scand Suppl 1991;
through 1990. Brain 1994;117(pt 2): New York: Raven Press, 1991, vol 2, 242:1-27.
325-335. pp 907-928. 8. Olmarker K, Rydevik B, Holm S:
2. Panjabi MM, Oxland T, Takata K, et al: 5. Cornefjord M, Olmarker K, Farley DB, Edema formation in spinal nerve
Articular facets of the human spine: et al: Neuropeptide changes in com- roots induced by experimental, grad-
Quantitative three-dimensional anato- pressed spinal nerve roots. Spine ed compression: An experimental
my. Spine 1993;18:1298-1310. 1995;20:670-673. study on the pig cauda equina with
3. Czervionke LF, Daniels DL, Ho PS, 6. Chabot MC, Montgomery DM: The special reference to differences in
et al: Cervical neural foramina: pathophysiology of axial and radicular effects between rapid and slow onset
Correlative anatomic and MR imaging neck pain. Semin Spine Surgery of compression. Spine 1989;14:569-
study. Radiology 1988;169:753-759. 1995;7:2-8. 573.

Vol 4, No 6, November/December 1996 315

Cervical Radiculopathy

9. Dillin W, Booth R, Cuckler J, et al: 23. Connell MD, Wiesel SW: Natural his- et al: Cervical root stimulation in the
Cervical radiculopathy: A review. tory and pathogenesis of cervical disk diagnosis of radiculopathy. Neurology
Spine 1986;11:988-991. disease. Orthop Clin North Am 1987;37:329-332.
10. Hult L: The Munkfors Investigation. 1992;23:369-380. 35. Tsai CP, Huang CI, Wang V, et al:
Acta Orthop Scand (Suppl) 1954;16:1-76. 24. Bracker MD, Ralph LP: The numb Evaluation of cervical radiculopathy
11. Lees F, Turner JWA: Natural history arm and hand. Am Fam Physician 1995; by cervical root stimulation.
and prognosis of cervical spondylosis. 51:103-116. Electromyogr Clin Neurophysiol
BMJ 1963;2:1607-1610. 25. Friedenberg ZB, Miller WT: Degen- 1994;34:363-366.
12. DePalma AF, Subin DK: Study of the erative disc disease of the cervical 36. Roberts WA, Garfin SR, White AA III:
cervical syndrome. Clin Orthop 1965; spine: A comparative study of Degenerative disorders: An algorithm
38:135-142. asymptomatic and symptomatic for the diagnosis of neck pain, in
13. Garvey TA, Eismont FJ: Diagnosis patients. J Bone Joint Surg Am 1963; The Cervical Spine Research Society
and treatment of cervical radiculopa- 45:1171-1178. Editorial Committee (eds): The Cervi-
thy and myelopathy. Orthop Rev 1991; 26. Bell GR, Ross JS: Diagnosis of nerve cal Spine, 2nd ed. Philadelphia: JB
20:595-603. root compression: Myelography, com- Lippincott, 1989, pp 611-616.
14. Henderson CM, Hennessy RG, Shuey puted tomography, and MRI. Orthop 37. Shelokov AP: Evaluation, diagnosis,
HM Jr, et al: Posterior-lateral fora- Clin North Am 1992;23:405-419. and initial treatment of cervical disc
minotomy as an exclusive operative 27. Larsson EM, Holtas S, Cronqvist S, et disease, in Regan JJ (ed): Cervical Spine
technique for cervical radiculopathy: al: Comparison of myelography, CT Disease. Philadelphia: Hanley &
A review of 846 consecutively operat- myelography and magnetic resonance Belfus, 1991, pp 167-176.
ed cases. Neurosurgery 1983;13: imaging in cervical spondylosis and 38. British Association of Physical
504-512. disk herniation: Pre- and postopera- Medicine: Pain in the neck and arm: A
15. Lunsford LD, Bissonette DJ, Jannetta tive findings. Acta Radiol 1989;30: multicentre trial of the effects of phys-
PJ, et al: Anterior surgery for cervical 233-239. iotherapy. BMJ 1966;1:253-258.
disc disease: Part 1. Treatment of later- 28. Modic MT, Masaryk TJ, Mulopulos 39. Huston GJ: Everyday aids and appli-
al cervical disc herniation in 253 cases. GP, et al: Cervical radiculopathy: ances: Collars and corsets. BMJ 1988;
J Neurosurg 1980;53:1-11. Prospective evaluation with surface 296:276.
16. Davidson RI, Dunn EJ, Metzmaker JN: coil MR imaging, CT with metriza- 40. Naylor JR, Mulley GP: Surgical col-
The shoulder abduction test in the mide, and metrizamide myelography. lars: A survey of their prescription and
diagnosis of radicular pain in cervical Radiology 1986;161:753-759. use. Br J Rheumatol 1991;30:282-284.
extradural compressive monoradicu- 29. Van de Kelft E, van Vyve M: Diag- 41. Murphy MJ, Lieponis JV: Nonoper-
lopathies. Spine 1981;6:441-446. nostic imaging algorithm for cervical ative treatment of cervical spine pain,
17. Beatty RM, Fowler FD, Hanson EJ Jr: soft disc herniation. J Neurol Neuro- in The Cervical Spine Research Society
The abducted arm as a sign of rup- surg Psychiatry 1994;57:724-728. Editorial Committee (eds): The Cer-
tured cervical disc. Neurosurgery 1987; 30. Neuhold A, Stiskal M, Platzer C, et al: vical Spine, 2nd ed. Philadelphia: JB
21:731-732. Combined use of spin-echo and Lippincott, 1989, pp 670-677.
18. LaBan MM, Meerschaert JR, Taylor RS: gradient-echo MR-imaging in cervical 42. Colachis SC Jr, Strohm BR: A study of
Breast pain: A symptom of cervical disk disease: Comparison with mye- tractive forces and angle of pull on
radiculopathy. Arch Phys Med Rehabil lography and intraoperative findings. vertebral interspaces in the cervical
1979;60:315-317. Neuroradiology 1991;33:422-426. spine. Arch Phys Med Rehabil 1965;46:
19. Brodsky AE: Cervical angina: A cor- 31. Hitselberger WE, Witten RM: Abnor- 820-830.
relative study with emphasis on the mal myelograms in asymptomatic 43. Tan JC, Nordin M: Role of physical
use of coronary angiography. Spine patients. J Neurosurg 1968;28:204-206. therapy in the treatment of cervical
1985;10:699-709. 32. Gnatz SM, Simpson JM: Non-surgical disk disease. Orthop Clin North Am
20. Stookey B: Compression of spinal evaluation, treatment and rehabilita- 1992;23:435-449.
cord and nerve roots by herniation of tion of cervical spine disorders, in An 44. Olmarker K, Byrod G, Cornefjord M,
the nucleus pulposus in the cervical HS, Simpson JM (eds): Surgery of the et al: Effects of methylprednisolone on
region. Arch Surg 1940;40:417-432. Cervical Spine. London: Martin nucleus pulposus-induced nerve root
21. Rothman RH, Marvel JP Jr: The acute Dunitz, 1994, pp 147-165. injury. Spine 1994;19:1803-1808.
cervical disk. Clin Orthop 1975;109: 33. Hong CZ, Lee S, Lum P: Cervical 45. Ferrante FM, Wilson SP, Iacobo C, et
59-68. radiculopathy: Clinical, radiographic al: Clinical classification as a predictor
22. Heller JG: The syndromes of degener- and EMG findings. Orthop Rev 1986; of therapeutic outcome after cervical
ative cervical disease. Orthop Clin 15:433-439. epidural steroid injection. Spine
North Am 1992;23:381-394. 34. Berger AR, Busis NA, Logigian EL, 1993;18:730-736.

316 Journal of the American Academy of Orthopaedic Surgeons