AJR 2003; 181:1149-1154
© American Roentgen Ray Society
Radicular Pain Avoidance During Needle Placement in Lumbar Diskography
Vibhu Kapoor1,
William E. Rothfus1,
Stephen Z. Grahovac1 and
Richard E. Latchaw1,2
1 Department of Radiology, Division of Neuroradiology, University of Pittsburgh
Medical Center, 200 Lothrop St., Pittsburgh, PA 15213.
2 Present address: Department of Radiology, University of California–Davis
Medical Center, Ste. 3100, 4860 Y St., Sacramento, CA 95817.
Received November 25, 2002;
accepted after revision April 3, 2003.
Address correspondence to V. Kapoor.
Abstract
OBJECTIVE. The objective of our study was to determine whether a
method could be found to reduce iatrogenic radicular pain during needle
placement in lumbar diskography.
MATERIALS AND METHODS. After obtaining permission from the
institutional review board at the University of Pittsburgh Medical Center, we
conducted a study using medical records and existing data that were recorded
for quality control during lumbar diskography. A coaxial technique was being
used for lumbar diskography. We evaluated data for 71 intervertebral disks in
26 patients in which the needle placement was randomly high (superior) or low
(inferior), and the associated pain response during needle placement was
recorded. In an attempt to minimize iatrogenic pain during needle placement,
we identified a potentially "safe window" for needle placement on
MRI of the lumbar spine. On oblique fluoroscopy of the lumbar spine, the safe
window is a triangle formed by the superior articular facet medially, the
superior endplate of the lower vertebra inferiorly, and an imaginary line
joining the tip of the superior articular facet and the superolateral tip of
the vertebral body. This safe window was then used for needle placement in
another 73 intervertebral disks in 27 patients. Pain response to needle
placement was recorded for quality control, and the medical records were
retrospectively compared with the initial 71 intervertebral disks in which
needle placement was random.
RESULTS. In the initial group with random needle placement, lower
extremity radicular pain occurred in 13 (18.3%) of 71 intervertebral disks
with superior needle placement and in 23 (32.4%) of 71 intervertebral disks
with inferior needle placement (total, 50.7%). The pain responses of the
superior and inferior groups were not significantly different (p =
0.27). On MRI, the average distances between the nerve
ganglion–fascicle–rami and the superior articular facets at the
superior disk level were 1.1, 1.4, and 2.5 mm at L3–L4, L4–L5, and
L5–S1, respectively. The average distances between the nerve
ganglion–fascicle–rami and the superior articular facets at the
inferior disk level were 3.0, 3.6, and 6.6 mm at L3–L4, L4–L5, and
L5–S1, respectively. When the safe window was used, only five (6.8%) of
73 patients reported radicular pain. The decrease in radicular pain between
the two groups was significant (p < 0.001).
CONCLUSION. Iatrogenic lower extremity radicular pain is common
during random needle placement at lumbar diskography. High or low needle
placement in the intervertebral disk could not predict whether radicular pain
would be averted. We identified a safe window that can be used for needle
placement during lumbar diskography to minimize iatrogenic lower extremity
radicular pain and thereby improve the reliability of the test.
Introduction
Lumbar diskography has been performed for the past 50 years
[1,
2], and in that time, an
evolution has affected the indications and techniques for the procedure. As
the number of spine procedures for managing lower back pain of discogenic
origin (intradiskal electrothermal therapy, vertebral fusion, and
microdiskectomy) has expanded, so has the need for provocative testing with
diskography. Increasingly, spine surgeons request diskography as a part of the
preoperative workup to identify the disk level that is the source of back
pain, to confirm internal disk abnormalities, and to plan appropriate fusion
surgery. Diskography techniques have gradually evolved with the use of C-arm
fluoroscopy and modification of approaches using dedicated needles
[1,
3–6].
However, the procedure remains a painful one for patients to endure.
Radicular leg pain due to needle placement during lumbar diskography is
very common, presumably because of nerve ganglion–fascicle–rami
irritation during needle placement
[3]. This pain not only
increases patient discomfort but also introduces a variable that leads to
false-positive results [7,
8]. Various techniques for
lumbar diskography have been described, most of which use a 22-gauge needle.
Some authors have suggested that aiming toward the inferior aspect of the disk
or injecting local anesthetic just before annular penetration may decrease
needle pain [3,
4,
9,
10]. However, no comprehensive
studies have focused on the incidence of needle-related pain or on techniques
to avert radicular pain during needle placement. Only one article suggested
that changing the technique of needle placement could alter the incidence of
pain [5].
The purpose of our study was to determine the incidence of radicular pain
during needle placement at diskography and whether a technique that minimized
iatrogenic pain could be identified.
Materials and Methods
Patients
After obtaining permission from the institutional review board at the
University of Pittsburgh Medical Center, we conducted a retrospective study
using medical records and existing data recorded for quality control during
lumbar diskography. During a 5-month period, 53 patients who had undergone
routine lumbar diskography composed our study group. All the patients were
referred from the clinical service for evaluation of back pain that was
considered diskogenic in origin. In phase 1 of the study, the needle was
placed randomly high or low in the intervertebral disk. Pain during needle
placement was recorded on the medical chart. In phase 2, an optimal window was
determined and was used to guide needle placement during phase 3. Again, pain
production during needle placement was recorded for quality control. Informed
consent was obtained from all patients before lumbar diskography. No
interventional procedure other than the standard was performed on any patient
in either group.
Phase 1
Seventy-one disk levels (L2–L3 [n = 5]; L3–L4
[n = 19]; L4–L5 [n = 23]; L5–S1 [n =
24]) were studied in 26 patients, and the occurrence of radicular pain during
needle placement was recorded. The needle was placed high in the disks in 31
patients and low in 40. The technique for needle placement in these groups of
patients was as follows: using a fluoroscopy unit with C-arm capability, we
angled the tube in the craniocaudal plane tangential to the disk being
studied, with a 25–35° lateral to medial angulation. This view
depicted the pedicle, superior articular facet, and lateral margin of the disk
(Fig. 1). We also used a
coaxial technique consisting of an outer 20-gauge 8.89-cm straight needle
(Becton Dickinson, Franklin Lakes, NJ) and an inner 25-gauge, 15-cm spinal
needle (Cook, Bloomington, IN). In most cases, the tip of the 25-gauge needle
was hand-curved to approximately 45° before it was introduced through the
20-gauge needle; in some patients, we did not believe it necessary to curve
the needle.
View larger version (91K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1. —36-year-old man with L4 radiculopathy undergoing L4–L5
diskography. Oblique fluoroscopy image of random needle placement shows X-ray
tube angled 25–35° laterally to medially. This image depicts pedicle
(P), superior articular facet (F), and lateral margin of intervertebral disks.
Needle entry zone is box with vertical sides formed by line joining lateral
margin of disk (x) and line along lateral margins of pedicle (y), and
horizontal sides formed by superior (a) and inferior (b) endplates. Shaded box
is needle entry zone for superior placement; needle (arrowhead)
placement in this patient is inferior.
|
|
The depth of the 20-gauge needle was checked intermittently on lateral
fluoroscopy as the needle was advanced toward the intervertebral disk. After
the needle tip passed the anterior margin of the superior articular facet, the
stylet was removed, and the 25-gauge curved 15-cm needle was advanced through
the 20-gauge needle. The 25-gauge needle was placed randomly in a superior or
inferior site in the disk, and the needle placement was classified as being
high or low as seen on the lateral radiographs obtained after the procedure
(Figs. 2 and
3) by one of two experienced
senior neuroradiologists.
View larger version (111K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3. —40-year-old woman with L4 and L5 radiculopathy. Lateral
L3–L4 diskogram shows superior needle placement. Outer 20-gauge
(arrowhead) and inner 25-gauge (arrows) needles are
distinctly seen.
|
|
The landmark used for needle entry was midway between the superior
articular facet medially and a line drawn along the lateral margin of the
vertebra laterally (Fig. 1).
The needle was introduced into the disk opposite the side of the usual pain.
We documented whether the patient had radicular leg pain as the needle entered
the outer annular fibers.
Phase 2
The second part of the study consisted of retrospectively evaluating the
lumbar MRIs of 10 patients with normal L3–L4, L4–L5, and
L5–S1 intervertebral disks. The distances between the nerve
ganglion–fascicle–rami and the superior articular facets at the
superior and inferior intervertebral disk levels were measured (Fig.
4A,
4B,
4C). All the MRIs were
evaluated on a PACS (picture archiving and communication system) workstation
(Agfa, Ridgefield Park, NJ). The shortest distance that could be measured with
this system was 0.1 mm. Using these measurements, we identified a safe window
(Fig. 4A,
4B,
4C) through which a needle
might pass without coming into contact with a nerve ganglion, fascicle, or
rami. The safe window as seen on MRI was at the inferior disk level (Fig.
4A,
4B,
4C) between the exiting nerve
root and the base of the adjacent superior articular facet.
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A. —23-year-old man with tingling and paresthesia of right foot
who underwent MRI of lumbar spine. Axial unenhanced T1-weighted image (TR/TE,
500/25) obtained at superior portion of L5–S1 intervertebral disk level
shows that distance between right superior articular facet
(arrowhead) and adjacent exiting nerve (arrow) is
approximately 1.5 mm.
|
|
View larger version (129K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B. —23-year-old man with tingling and paresthesia of right foot
who underwent MRI of lumbar spine. Unenhanced axial T1-weighted image (500/25)
obtained at inferior portion of L5–S1 intervertebral disk level shows
distance between right superior articular facet (arrowhead) and
adjacent exiting nerve (arrow) to be approximately 7 mm.
|
|
View larger version (140K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C. —23-year-old man with tingling and paresthesia of right foot
who underwent MRI of lumbar spine. Unenhanced axial T1-weighted image (500/25)
obtained at inferior L5–S1 intervertebral disk level shows benefit of
using curved inner 25-gauge needle as opposed to straight needle. Curved
needle (a) pierces posterolateral aspect of intervertebral disk at 90°
with final position of its tip in posterior third of nucleus pulposus, as
opposed to straight needle (b) that pierces disk tangentially with final
position of its tip in annulus of disk. Alternatively, straight inner needle
that is tangential to disk margin may get deflected (c) and pass lateral
relative to intervertebral disk. Curved arrow marks outer 20-gauge needle, and
straight arrow marks exiting nerve.
|
|
Phase 3
Using the information from MRI, we altered the technique with the aim of
decreasing patient pain related to needle placement. We reviewed the medical
records of 27 patients, representing 73 disk levels (L2–L3 [n =
4]; L3–L4 [n = 24]; L4–L5 [n = 25]; L5–S1
[n = 20]), for radicular pain during needle placement using the safe
window as determined on MRI. On oblique fluoroscopy of the lumbar spine, the
safe window is a triangle (Fig.
5A,
5B,
5C,
5D) formed by the superior
articular facet medially, the superior endplate of the lower vertebra
inferiorly, and an imaginary line joining the tip of the superior articular
facet and the superolateral tip of the vertebral body. The technique of
diskography was similar to our initial technique except that the target for
the needle placement was the safe window (Figs.
5A,
5B,
5C,
5D and
6). Using this safe window, we
documented radicular leg pain production as the 25-gauge curved spinal needle
entered the outer annular fibers of the intervertebral disk at each level.
View larger version (94K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A. —40-year-old man with L5 radiculopathy undergoing lumbar
diskography with needle placement through safe window. Oblique radiograph of
lumbar spine obtained during L4–L5 diskography, shows safe window
(triangle) to be formed by superior articular facet medially,
superior endplate of lower vertebra inferiorly, and imaginary line joining tip
of superior articular facet and superolateral tip of vertebral body.
|
|
View larger version (101K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B. —40-year-old man with L5 radiculopathy undergoing lumbar
diskography with needle placement through safe window. Oblique radiograph of
lumbar spine obtained during L4–L5 diskography, without shaded triangle,
shows outer 20-gauge straight needle (straight arrow) in safe window.
Arrowhead marks superior articular facet, and curved arrow marks superior
endplate of lower vertebra.
|
|
View larger version (99K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5C. —40-year-old man with L5 radiculopathy undergoing lumbar
diskography with needle placement through safe window. Oblique radiograph of
lumbar spine during L4–L5 diskography shows paths of outer 20-gauge
8.89-cm straight needle (arrow) and inner 25-gauge 15-cm hand-curved
spinal needle (arrowhead) through safe window.
|
|
View larger version (109K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5D. —40-year-old man with L5 radiculopathy undergoing lumbar
diskography with needle placement through safe window. Anteroposterior
radiograph of lumbar spine obtained during L4–L5 diskography shows final
position of tips of outer 20-gauge straight needle (arrow) and inner
25-gauge curved spinal needle (arrowhead).
|
|
Statistical Analysis
To analyze the data, we used a repeated measures logistic regression model
(PROC GENMOD component, SAS Institute, Cary, NC) to assess pain production
from random placement of needle, superior versus inferior placement in the
disk, and pain production from random needle placement versus placement
through the safe window. This statistical method takes into consideration the
correlations when more than one procedure is performed on a patient.
Results
Phase 1
Lower extremity radicular pain during random needle placement occurred at
36 (50.7%) of 71 disk levels during lumbar diskography in the first part of
the study. It occurred in 13 (18.3%) of 71 disks with superior needle
placement, and in 23 (32.4%) of 71 disks with inferior needle placement. This
difference was not statistically significant (p = 0.27).
Phase 2
Review of axial MRIs (Fig.
4A,
4B,
4C) showed the average
distances between nerve ganglion–fascicle–rami and the superior
articular facets at the superior disk level to be 1.1, 1.4, and 2.5 mm, with a
range of 0–2.9, 0–3.8, and 0.8–5.2 mm at L3–L4,
L4–L5, and L5–S1, respectively. The average distance between the
nerve ganglion–fascicle–rami and the superior articular facets at
the inferior disk level was 3.0, 3.6, and 6.6 mm, with a range of
1.2–4.1, 1.9–5.0, and 3.2–10.2 mm at L3–L4,
L4–L5, and L5–S1, respectively. The distance between the exiting
neural structures and the identifiable landmark on fluoroscopy (i.e., the
superior articular facet) was longest at the inferior disk level.
We identified a safe window (Fig.
4A,
4B,
4C) that is represented on
oblique fluoroscopy by a triangle formed by the superior articular facet
medially, the superior endplate of the lower vertebra inferiorly, and an
imaginary line joining the tip of the superior articular facet and the
superolateral tip of the vertebral body (Fig.
4A,
4B,
4C).
Phase 3
With the needle targeted through the safe window (Figs.
5A,
5B,
5C,
5D and
6), lower extremity radicular
pain occurred in only five (6.8%) of 73 intervertebral disks levels
(L2–L3 [n = 0]; L3–L4 [n = 3]; L4–L5
[n = 1]; L5–S1 [n = 1]). The decrease in radicular
pain in this safe window group relative to the random group was significant
(p < 0.001).
Discussion
Lumbar diskography, or nucleography, has been performed for approximately
50 years [1,
2]. It has been used to
diagnose herniated nucleus pulposus; to evaluate fissuring of the posterior
annulus fibrosus, part of preoperative planning for localization of
symptomatic diskogenic disease; and to differentiate recurrent disk herniation
from scar tissue [3,
11–13].
Increasingly, diskography has been used to evaluate internal disruptions of
the disk to identify patients who could benefit from conservative therapy
versus more aggressive therapy such as intradiskal electrothermy or
disk-fusion surgery.
Diskography is an inherently painful procedure, and the pain caused by
needle placement may result in false-positive or confusing results
[7,
8,
14]. In certain patients with
abnormal psychosocial profiles (as determined by psychometric testing), it may
result in back pain that persists for up to a year after the procedure
[15]. The goal of this project
was to determine the approach that would have the least possibility of causing
radicular pain during needle placement so as to reliably minimize the
possibility of a false-positive result and improve the reproducibility of
diskography. We noted iatrogenic pain with random needle placement in
approximately 50% of our patients, and we have described a safe window for
needle placement during diskography to minimize this complication.
The techniques for lumbar diskography have been described and modified
since the time it was first performed in 1952
[1]. We have used a
modification of the coaxial technique described by Sachs et al.
[6] and Johnson et al.
[16] in well over 2000 disks
(having various degrees of degeneration, 20–30% of the patients in our
practice being obese) with excellent results. A particular advantage to this
coaxial technique is the use of a small curved needle. There is less chance of
irritating the segmental nerve root with a 25-gauge needle placed into the
disk. Placement of a curved needle perpendicular to the annular fibers averts
a pure annular injection, and because the curved needle is steerable, fine
modifications of needle direction can be achieved (both in and outside the
disk). Although precurved coaxial diskography needle sets are commercially
available and work well through the safe window, we prefer to hand-curve our
25-gauge needles so that we can modify the angle of curvature. The angle of
the hand-curved needle can be adjusted as determined by the final position of
the base needle. We have used a variety of needle lengths, up to 20 cm in
obese patients. We emphasize that the safe window technique is not restricted
to curved needles but can also be used for oblique placement of straight
needles [3,
4].
Kostelic et al. [17,
18] have described the course
of the nerve roots, ganglion, fascicles, and the rami on cross-sectional
imaging and anatomic sections in great detail. As the dorsal and ventral nerve
roots exit the neural foramen, they divide into a variable number (7–15)
of fascicles distal to the dorsal root ganglion at the upper intervertebral
disk level [17]. These
fascicles are approximately 1 mm posterior and superior in relation to the
intervertebral disk and are in closer relationship to the disk than are the
nerve roots [18]. The
fascicles then regroup to form the dorsal and ventral rami at the level of the
superior endplate [17]. The
dorsal ramus gives off a median branch to the facet joint
(Fig. 6B), the irritation of
which may also be a cause of back pain during needle placement.
Other authors have described techniques to avoid contact with the
extraforaminal nerve during diskography
[3,
4] or percutaneous diskectomy
[19], recommending placement
at the inferior aspect of the disk, or contacting, then sliding off, the
superior facet. We describe a similar method, but with a well-defined,
reproducible target. On the basis of our observations, we believe there is a
safe window through which the needle is least likely to produce radicular
pain. On an oblique fluoroscopic image of the lumbar spine, the safe window is
a triangle (Fig. 5A,
5B,
5C,
5D) formed by the superior
articular facet medially, the superior endplate of the lower vertebra
inferiorly, and an imaginary line joining the tip of the superior articular
facet and the superolateral tip of the vertebral body. The longest side of
this triangle represents the course of the nerve at it exits the foramen. We
found the distance between the nerves and the superior articular facets to be
short, on the order of millimeters, especially at the superior disk level and
in particular at the L3–L4 level. We also confirmed that the nerve
follows an oblique path anteroinferiorly as it exits the neural foramen and
that at the L5–S1 intervertebral disk level, it follows a less vertical
path than at the L3–L4 disk level.
These observations were based on MRI measurements obtained in subjects with
normal disk height (Fig. 4A,
4B,
4C). With disk degeneration,
there is a decrease in disk height with concomitant shrinkage of the safe
window. Avoiding radicular pain in this setting becomes more challenging. For
the safe window, we chose the articular facet as one of the reference points
because it is easily identifiable on C-arm fluoroscopy, which is used for all
our diskograms.
We emphasize the need for using the smallest diameter needle possible to
enter the intervertebral disk. The outer diameter of a commonly used 22-gauge
spinal needle is 0.72 mm, and that of a 25-gauge spinal needle is 0.51 mm. The
clearance between the superior articular facet and the nerve is only a few
millimeters, an average of 3.0, 3.6, and 6.6 mm at the inferior disk level at
L3–L4, L4–L5, and L5–S1, respectively, in healthy patients.
At L5–S1, it is the maximum; however, in our experience, the approach to
this level is technically the most challenging. Not only is angulation
steeper, but degenerative disk disease frequently causes severe disk-space
narrowing. Any diminution in disk height decreases the size of the safe
window, and thus the margin of error between the needle and the nerve. The use
of a 25-gauge needle may make the difference in avoiding radicular leg pain.
The smaller the diameter of the inner curved needle, the less the chance of
causing lower extremity radicular pain during needle placement through the
safe window, especially in patients with significant disk degeneration.
In summary, lumbar diskography is an inherently painful procedure, and
false-positive results caused by needle placement are a problem. Even when a
25-gauge needle is used, a large proportion of patients experience pain during
random placement of the needle into the disk. However, we found that when the
safe window is used, patient discomfort and the reliability of the procedure
can be markedly improved.
Acknowledgments
We thank Eric Jablonowski and Mike Purvis for their help with the
illustrations and Howard E. Rockette and Wei Li for their help with the
statistical analysis in the paper.
References
- Erlacher PR. Nucleography. J Bone Joint Surg
Am 1952;34:204
–210
- Feinberg SB. The place of diskography in radiology as based on 2320
cases. AJR1964; 92:1275
–1281
- Schellhas KP. Diskography. Neuroimaging Clin N
Am 2000;10:579
–596[Medline]
- Gibby WA. Interventional techniques and diagnostic procedures of
the spine. In: Zimmerman RA, Gibby WA, Carmody RF, eds.
Neuroimaging: clinical and physical principles. New
York: Springer 2000:1440
–1445
- Kumar N, Agorastides ID. The curved needle technique for accessing
the L5/S1 disk space. Br J Radiol2000; 73:655
–657[Abstract]
- Sachs BL, Spivey MA, Vanharanta H, et al. Techniques for lumbar
diskography and computed tomography/diskography in clinical practice.
Orthop Rev1999; 19:775
–778
- Carragee E, Tanner C, Khurana S, et al. The rates of false-positive
lumbar diskography in select patients without low back complaints.
Spine 2000;25:1373
–1380[Medline]
- Carragee E, Tanner CM, Yang B, Brito JL, Truong T. False-positive
findings on lumbar diskography: reliability of subjective concordance
assessment during provocative disc injection. Spine1999; 24:2542
–2547[Medline]
- Tehranzadeh J. diskography 2000. Radiol Clin North
Am 1998;36:463
–495[Medline]
- McCulloch JA, Waddell G. Lateral lumbar diskography. Br
J Radiol 1978;51:498
–502[Abstract/Free Full Text]
- Colhoun E, McCall IW, Williams L, Cassar Pullicino VN. Provocation
diskography as a guide to planning operations on the spine. J Bone
Joint Surg Br 1988;70:267
–271
- Hodge JC, Ghelman B, Schneider R, et al. Recurrent disk versus scar
in the postoperative patient: the role of computed tomography
(CT)/diskography, and CT/myelography. J Spinal Disord1994; 7:470
–477[Medline]
- Jackson RP, Cain JE, Jacobs RR, Cooper BR, McManus GE. The
neuroradiographic diagnosis of lumbar herniated nucleus pulposus. II. A
comparison of computed tomography (CT), myelography, CT-myelography, and
magnetic resonance imaging. Spine1989; 14:1362
–1367[Medline]
- Walsh TR, Weinstein JN, Stratt KF, Lehmann TR, Aprill C, Sayre H.
Lumbar diskography in normal subjects: a controlled, prospective study.
J Bone Joint Surg Am1990; 72:1081
–1088[Abstract/Free Full Text]
- Carragee EJ, Chen Y, Tanner CM, Hayward C, Rossi M, Hagle C. Can
diskography cause longterm back symptoms in previously asymptomatic subjects?
Spine 2000;25:1803
–1808[Medline]
- Johnson JC, Deeb ZL. Coaxial technique for lumbar puncture in the
morbidly obese patient. Radiology1991; 179:874[Abstract/Free Full Text]
- Kostelic J, Haughton VM, Sether L. Proximal lumbar spinal nerves in
axial MR imaging, CT, and anatomic sections. Radiology1992; 183:239
–241[Abstract/Free Full Text]
- Kostelic J, Haughton VM, Sether L. Lumbar spinal nerves in the
neural foramen: MR appearance. Radiology1991; 178:837
–839[Abstract/Free Full Text]
- Onik GM. Percutaneous diskectomy in the treatment of herniated
lumbar disks. Neuroimaging Clin N Am2000; 10:597
–606[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
