This article discusses three rapidly growing radiologic approaches to the care of these patients--discography, percutaneous vertebroplasty, and image-guided injection therapy. The focus will be on the clinical indications and efficacy of these procedures with only brief discussions of their technical aspects, which have been detailed in several excellent review articles.1-11
is a third-year Radiology Resident at The University of Pittsburgh
Medical Center. He received both his undergraduate and medical
degrees from The University of Pennsylvania. He plans to pursue a
fellowship in Musculoskeletal Radiology at The University of
Pittsburgh in 2003.
Radiologists are playing an expanding role in the
management of spinal pain through the application of discography,
image-guided injection therapy, and percutaneous vertebroplasty.
Discography demonstrates the source of pain in patients who are
suspected of having discogenic disease, but who have normal or
equivocal findings on MRI. Image-guided injection therapy is a
safe and simple way to deliver steroids to the epidural space,
nerve roots, or facet joints to relieve symptoms from
radiculopathy or degenerative facet disease. Despite widespread
clinical use, however, controversy regarding its efficacy
remains. Finally, vertebroplasty is proving to be an effective
treatment for painful and debilitating vertebral compression
fractures by delivering methylmethacrylate cement percutaneously
into the vertebral body.
Spinal pain is an extremely common cause of morbidity in the
United States and is the second most common reason patients seek
medical attention. It is a challenging and multifactorial disease
that is best managed by a cross-disciplinary approach involving
spine surgeons, radiologists, anesthesiologists, and rehabilitative
medicine specialists, among others. Recent advances in several
image-guided percutaneous therapies have expanded the role of the
procedural radiologist in the treatment of spinal pain.
This article discusses three rapidly growing radiologic
approaches to the care of these patients--discography, percutaneous
vertebroplasty, and image-guided injection therapy. The focus will
be on the clinical indications and efficacy of these procedures
with only brief discussions of their technical aspects, which have
been detailed in several excellent review articles.
Diagnostic imaging has long played a central role in determining
the etiology of back pain, especially with the wide availability of
high-quality magnetic resonance imaging (MRI). In general, MRI has
a high sensitivity for delineating morphologic derangements that
can lead to spinal pain such as degenerative disc disease, facet
arthropathy, annular tears, herniations, and spinal stenoses.
However, MRI is less sensitive for the detection of disc pathology
in the cervical and thoracic spines than in the lumbar spine;
although false negatives do occur even in the lumbar spine.
Furthermore, studies have shown that many normal, asymptomatic
volunteers have disc herniations that are found on MRI.
It is, therefore, crucial that careful clinical correlation be made
when trying to understand the significance of both positive and
negative MRI findings and that correlative studies such as
discography be considered.
For these reasons, discography is playing an expanding role in
localizing discogenic back pain and in identifying occult annular
tears and other internal disc derangements. It is a provocative
test that involves the injection of contrast material into the
suspected pathologic disc and measuring the clinical response in
terms of reproducing the patient's symptoms. Patients may be
referred for discography regardless of what the imaging findings
reveal, especially in cases of suspected cervical and thoracic disc
problems without obvious disc herniations. The clinical goals of
discography include clearly identifying the intervertebral disc as
a source of the patient's back pain and determining which levels
are involved. When performed in a standardized, reproducible
manner, this technique can result in improved outcomes for surgical
therapy--first, by confidently demonstrating that a patient
contemplating spinal surgery does indeed have discogenic pain, and,
second, by identifying the precise disc levels that are pathologic.
Discography also plays a role in identifying candidates who would
benefit from intradiscal electrotherapy (IDET).
Before a discogram is performed, it is important to review the
patient's previous imaging studies to identify the patient's
overall anatomy and likely sites of pathology. Spinal cord and
nerve root compression are also important to identify, as these may
be contraindications to discography.
This is because distension of a disc that is already severely
compressing the spinal cord or nerve roots may worsen the
compression and exacerbate the patient's symptoms. Other
contraindications include active infection anywhere in the body,
uncorrected bleeding disorders, and the inability to lie still for
the duration of the study. Generally, the procedure can be
performed safely with conscious sedation.
The technical aspects of performing discography vary among
practitioners and depend upon the spinal level being evaluated.
High-resolution C-arm fluoroscopy or CT-fluoroscopy is mandatory
for adequate visualization of bony landmarks during needle
placement and for the detection of annular tears during contrast
injection. Twenty-two- or 25-gauge spinal needles
(Beckton-Dickinson, Franklin Lakes, NJ) of varying lengths are used
depending on patient size. Both coaxial- and single-needle
approaches to the disc have been described and can be used
depending upon operator preference. Post-discography CT is
sometimes performed for better delineation of the size and
locations of annular disruptions but is not routinely used by all
After the patient has been positioned on the table
appropriately, the skin is cleansed with an iodine-based
disinfectant (Betadine, Purdue Pharma, Stamford, CT), followed by
an alcohol rinse. Sterile technique must be carefully applied
during all percutaneous spinal procedures, particularly during
discography, as discitis is a potentially serious, albeit rare,
complication occurring in <0.01% of cases.
For cervical discography, the patient is placed in a supine
position and the needle is inserted from an anterolateral approach.
For lumbar and thoracic discography, the patient is placed in the
prone position and the needle is inserted from a dorsolateral
Once the patient has been prepped and local anesthesia has been
administered, the fluoroscope is aligned in the plane of the target
disc. The needle is advanced toward the inferior margin of the disc
with intermittent fluoroscopy to observe needle position. Care must
be taken in the cervical region to avoid the carotid arteries
branches and jugular veins. When the needle tip reaches the outer
part of the annular disc, slight resistance will be felt, and the
needle is then advanced through the outer annulus into the center
of the disc, where the needle tip position is confirmed by
anteroposterior and lateral fluoroscopy. As the needle tip
approaches the annular margin of the disc, it is not uncommon for
contact to occur with a nerve root, resulting in discomfort or pain
that requires repositioning of the needle. The patient is warned of
this possibility during needle insertion.
Once the needle tip is in the center of the disc, contrast is
injected while the radiologist watches under fluoroscopy.
Typically, 2 to 4 mL of nonionic contrast (Omnipaque 240, Amersham
Health, Princeton, NJ) should be injected to achieve full
distension of the nucleus pulposus, which is crucial to provoking a
Endpoints for injection include detection of an annular tear,
severe pain, or resistance of the disc to accepting further
contrast (figure 1). The patient's description of elicited symptoms
and particularly whether or not his or her clinical symptoms were
reproduced (ie, concordant versus nonconcordant pain) are recorded
carefully, along with the symptom severity and the presence of
annular pathology. When a painful disc is encountered, injection of
preservative-free local anesthetic such as lidocaine into the disc
is useful for temporary pain relief to allow for completion of the
study. This is important for reducing pain that may confuse the
evaluation of adjacent disc levels. After discography, patients are
monitored for 1 to 3 hours prior to discharge. Discharge
instructions should include advising the patient to expect some
pain and discomfort for a few days after the procedure and to
return for medical attention promptly should he/she experience
signs and symptoms of developing discitis, such as worsening back
pain, severe point tenderness, fever, or chills. Progression of
discitis into an epidural abscess or vertebral osteomyelitis are
other thoretical complications.
First developed in the mid-1980s by Deramond and colleagues
in France, percutaneous vertebroplasty has become an increasingly
utilized and effective way of treating painful vertebral
compression fractures from benign and malignant diseases. It was
originally applied successfully to vertebral hemangiomas and
osteolytic metastatases, although its primary application in North
America has become the treatment of osteoporotic compression
Compression fractures can be devastating complications of
osteoporosis leading to severe back pain and complications of
immobility such as progressive bone demineralization,
thromboembolic disease, pneumonia, depression, and markedly
decreased functional status. The incidence of compression fractures
will undoubtedly increase as the U.S. population ages. To date,
conventional therapy for compression fractures has centered upon
conservative measures such as bed rest, nonsteriodal
anti-inflammatory drugs and narcotics for pain control, and bracing
Percutaneous vertebroplasty consists of injecting
polymethylmethacrylate (PMMA), an exothermic semi-liquid cement,
into the vertebral body under radiologic guidance through a
percutaneously inserted cannula. This method of internally casting
the vertebral body results in enhanced vertebral body stability and
loss of micromotion across fracture planes.
These factors, along with the possible thermal damage that the PMMA
cement does to vertebral body pain receptors, are thought to be the
mechanisms behind the dramatic pain relief this procedure can
provide, although the exact mechanism is not known. Several
published series have shown at least moderate-to-complete pain
relief in approximately 63% to 90% of patients undergoing
vertebroplasty at intermediate-term (approximately 6 months)
Other benefits of vertebroplasty include decreased use of
analgesics and increased mobility.
Large series documenting the long-term effectiveness of
percutaneous vertebroplasty are pending. To maximize effectiveness,
this procedure should be part of a comprehensive medical regimen
directed by an osteoporosis specialist.
Other crucial factors in promoting successful outcomes for
vertebroplasty are appropriate patient selection and work-up.
Vertebral compression fractures clinically present with pain that
is aggravated by flexion of the spine, such as the motion incurred
when moving to a sitting position. It is usually described as a
deep, midline pain that is nonradiating and relieved by lying in
bed. Radicular symptoms should alert one to the possibility that
the pain may be due to other causes, such as disc disease or spinal
or neural foraminal stenoses. A careful physical examination should
be performed, including assessment and documentation of the
patient's baseline neurological status to evaluate for weakness or
radicular symptoms, which may indicate spinal stenosis or nerve
root compromise. On physical examination, every attempt must be
made to correlate the site of the vertebral body collapse with the
patient's symptoms. This can be done by heavily palpating the
patient's back along the midline under fluoroscopy until the
patient's symptoms are reproduced. This type of localization is
critical in selecting candidates who will benefit from the
procedure. When many months or years have passed since the
patient's onset of pain or when there are multiple affected
vertebral bodies with associated degenerative changes, as there
often are, identifying the culprit of the patient's pain can be
difficult. One should keep in mind that in cases of multilevel
collapse, the most morphologically abnormal vertebral body is not
necessarily the cause of the patient's pain, and that pain can be
referred to adjacent levels.
Relative contraindications to vertebroplasty include extreme
collapse of the vertebral body (ie, vertebra plana) and >20%
stenosis of the spinal canal.
While the optimal timing of vertebroplasty is a controversial and
evolving topic, most practitioners advocate some trial of medical
therapy before performing vertebroplasty.
This varies with the severity of pain and the aggressiveness of the
medical therapy. For example, patients who require hospitalization
with heavy narcotic requirements may be considered for
vertebroplasty earlier than others. Improvement of pain on medical
therapy is another relative contraindication to percutaneous
vertebroplasty, and active infection and uncorrected bleeding
diatheses are absolute contraindications. A recent report by
Kaufmann et al
suggests that the likelihood of clinical benefit from
vertebroplasty is not related to the age of the fracture.
Ideally, preprocedural evaluation of vertebroplasty candidates
is performed by MRI, although some rely on plain films (figure 2A)
and bone scintigraphy with SPECT imaging (figure 2B). When
reviewing the MR images (figures 2C and 2D), radiologists should
identify the size and location of the fracture cleft, the presence
of bone edema, the integrity of the posterior cortical margin, and
the presence or absence of spinal canal or neural foraminal
stenoses. T2-weighted or short tau inversion recovery sequences are
particularly useful in identifying areas of bone marrow edema,
which indicate acute fractures to help localize the source of the
The technical aspects of percutaneous vertebroplasty have been
well described in several recent reviews.
The procedure can generally be done under conscious sedation with
the appropriate physiologic monitoring, such as pulse oximetry and
heart rate monitoring by a nurse. General anesthesia is rarely
required except for patients with high levels of anxiety or pain.
Once the patient has been prepared and draped in the prone position
and local anesthesia has been administered to the skin and
periosteum, the vertebral body is accessed from a transpedicular
approach with an 11-gauge bone biopsy needle through a small skin
incision (figures 2E and 2F). The tip of the needle is advanced to
the anterior third of the vertebral body near the midline, which is
confirmed by biplane fluoroscopy at our institution. A vertebrogram
can be performed prior to PMMA injection to exclude the possibility
that the needle communicates with major venous structures, although
not all authors perform this step routinely. The PMMA powder is
then prepared with an organic solvent and a sterile opacifying
barium agent (approximately 30% wt/vol) (Bryan Corporation, Woburn,
MA) and instilled into the vertebral body under careful biplane
fluoroscopic visualization. The injection is terminated when there
has been filling of the fracture cleft, extension to the posterior
vertebral margin, or evidence of extraosseous cement leakage. It is
important to note that the percentage of vertebral body filling has
no relationship to pain relief; therefore, overaggressive filling
of the vertebral body, which can increase the chance of leak or
extravasation, should be avoided.
At the end of the procedure, the patient is monitored for 1 to 2
hours, at which time he or she is encouraged to ambulate to
evaluate for pain improvement. The patient is discharged home the
same day with instructions for limited progressive activity and
reduction of narcotic use as tolerated to judge response to
Complications from vertebroplasty are relatively low, occurring
in 1% to 10% of patients.
They include the following: failure of pain relief or worsening of
pain; spinal cord or nerve root compression; rib fractures from
lying in the prone position; paravertebral hematomas; swallowing
difficulty from esophageal compression; and pulmonary embolism from
venous cement extravasation.
Complications are associated with poor visualization of cement flow
during injection, poor needle positioning, inadequate aseptic
technique, and inability of the patient to cooperate. They may
occur somewhat more frequently in patients with neoplastic disease
due to the higher incidence of cortical destruction. The value of
careful attention to cement flow with strict lateral fluoroscopy,
as well as meticulous needle positioning to prevent paraspinal
cement leaks and venous extravasation, cannot be overemphasized.
The small associated risk of spinal cord or nerve root compression
from cement leaks requires the availability of a spine surgeon in
case emergent decompressive surgery should become necessary.
Epidural steroid injections, nerve root injections, and
Although it has been in use for several decades and many
clinical trials have been performed, the therapeutic efficacy of
steroid injection therapy for spinal pain remains controversial.
While the basic pathophysiologic mechanisms of spinal pain and
sciatica are incompletely understood, Nygaard et al
and Lee et al
have suggested that disc herniations are associated with an
abnormal inflammatory response mediated by phospholipases and other
compounds, resulting in nerve root irritation. These irritant
effects are thought to be countered by the anti-inflammatory
effects of epidural steroids.
Nelemans and colleagues
recently published a comprehensive review of 21 randomized,
controlled trials of injection therapies, and found that the
overall methodologic quality of these studies was poor; only three
of the trials were performed in a well-designed, placebo-controlled
manner. These three trials showed a trend toward short-term and
long-term pain relief, but this trend was not statistically
significant in the pooled data. Despite the numerous clinical
trials published, the authors concluded that there remains a
paucity of high-quality data either to support the use of steroid
injection therapy or to recommend abandoning its use, and that
further clinical trials need to be undertaken.
Notwithstanding the controversies regarding its efficacy,
clinicians continue to refer patients to radiologists for these
procedures. Many of the trials evaluating steroid injection
therapies, especially the earlier ones, have been criticized for
their lack of needle position documentation. In clinical practice,
injection therapies are often performed blindly by nonradiologists
without imaging guidance. However, Stitz and Sommer
have shown that blind needle placement results in inappropriate
positioning of the needle tip approximately 25% of the time, thus
supporting the use of imaging guidance. As for their current
clinical role, injection therapies, such as nerve root blocks and
facet joint injections, may have diagnostic and therapeutic value
in identifying or confirming the source of spinal pain when surgery
is being considered.
The temporary relief of pain that steroid injections provide may
also allow more aggressive and effective physical rehabilitative
Image-guided steroid injections can be performed from a
transforaminal approach for selective nerve root blocks, from a
midline interlaminar approach for epidural injections, and directly
into the facet joint for facet injections. Recently, Silbergleit et
presented a well-written review of the technical details of these
three types of injection therapy. Selective nerve root injections
are used for patients with radicular symptoms, often due to acute
disc pathology that is resistant to conservative therapy or is due
to degenerative neural foraminal stenosis.
Fluoroscopy, CT, or CT-fluoroscopy can be used for guidance
during selective nerve root blocks depending on operator
preference, although CT is believed by many to be more accurate, as
it images the nerve root target directly.
The needle is usually inserted from an oblique dorsolateral
approach (figures 3 and 4), and as the tip approaches the target
nerve root, the patient's radicular symptoms may be reproduced. The
nerve root sleeve is injected with a standard mixture of a
long-acting corticosteroid such as betamethasone (Schering
Corporation, Kenilworth, NJ) or methylprednisolone (Pharmacia &
Upjohn, Kalamazoo, MI) and a local anesthetic (0.5% bupivocaine).
Facet joint injections can also be accomplished under either CT or
fluoroscopic guidance with the same injection mixture as that used
for nerve root blocks. Fluoroscopic guidance is adequate for
epidural steroid injections, which are usually performed with an
interlaminar approach, but can also be performed with a sacral
hiatal or transforaminal approach.
Needle tip position in the epidural space is suggested by sudden
loss of resistance during needle advancement and is confirmed by
epidurography with 2 to 3 mL of nonionic iodinated contrast
material (Omnipaque 240, Amersham Health, Princeton, NJ) (figure
5). A standard dose of corticosteroid (eg, 1 to 2 mL of
methylprednisolone acetate [40 mg/mL]) and a local anesthetic (0.5%
bupivocaine) is injected. Complications from epidural steroid
injections are extremely uncommon and are usually minor, including
small epidural hematomas and vasovagal responses.
Infectious complications and inadvertent intrathecal administration
of medication are theoretical risks as well.
Percutaneous spinal interventions will continue to evolve
rapidly in the near future. The application of percutaneous
vertebroplasty may expand to include the delivery of bioresorbable
bone-strengthening polymers, antibiotic-impregnated materials,
osteogenic growth factors, and even antitumoral agents. Ongoing
research will undoubtedly result in the development of cements that
have higher tensile strengths and are easier to use. Cement
delivery systems will also become easier to use and more widely
available. Beyond vertebroplasty, newer techniques (such as
kyphoplasty, which involves treating compression fractures by
inserting a balloon inflation device into the vertebra to restore
vertebral body height), are currently under evaluation.
Intradiscal electrotherapy (IDET) and radiofrequency neurotomy are
two other developing image-guided treatments for discogenic and
facet-related back pain.
Magnetic resonance guidance systems and combined CT-fluoroscopic
devices may bring additional improvements in safety, speed, and
efficacy of injection procedures. Before these promising techniques
achieve wider clinical use, a clearer definition of long-term
outcomes, complications, and effect upon patients' quality of life
must be determined by large randomized prospective clinical
Percutaneous spinal interventions have become integral
components in the multidisciplinary approach to the management of
spinal pain, which is a serious and costly healthcare problem
debilitating millions of Americans each year. Procedural
radiologists are ideally suited to perform these exciting
procedures safely and accurately due to their skill in image-guided
intervention and intimate knowledge of anatomy. Competition from
spine physicians and anesthesiologists to perform these procedures
will likely continue as future applications broaden this rapidly
growing field and as clinical indications and efficacy are defined
The author would like to thank Dr. William E. Rothfus, Dr.
Stephen Grahovac, and Dr.Vibhu Kapoor of The University of
Pittsburgh Medical Center, for reviewing this manuscript and
providing the clinical images.