Cervical Discography
Cervical Discography
Cervical Discography
Cervical Discography:
Clinical Implications From 12 Years of Experience
Stephen A. Grubb, MD, and Carol K. Kelly, PhD
1382
Results
Of the 173 discograms performed, 24 (18 from the saline
group and 6 from the contrast group) were negative (i.e.,
no pain responses were provoked). Of the discs injected
with contrast, one study was radiographically normal in
appearance; one showed evidence of early degenerative
changes and provoked nonconcordant (unfamiliar) pain
at four levels; three studies showed early degenerative
Table 1. Number of Discs Injected and Concordant and Nonconcordant Pain Responses at Each Cervical Disc Level
Level
Number of
Discs Injected
Concordant
Pain Responses
Concordant
Responses (%)
Nonconcordant
Responses
Nonconcordant
Responses (%)
C2C3
C3C4
C4C5
C5C6
C6C7
C7T1
156
168
164
152
146
21
48
94
98
91
69
4
30
55
58
55
45
17
14
17
11
14
13
0
9
10
7
8
8
0
C23
C34
C45
C56
C67
C7T1
Number of Cases
156
168
164
152
146
21
3
7
7
13
6
1
Total 37
2
1
6
5
5
4
4
4
Total 31
6
1
1
1
1
12
4
1
2
9
Total 38
9
3
1
1
10
1
Total 25
Total 17
Total 1
Diagnosis
Myelopathy
2
5
Postlaminectomy kyphosis
Instability
1 at C1C2
2 Traumatic
2 Degenerative
DDD
HNP
17
1
1
Hangmans fracture
C3C4
C4C5
C5C6
C6C7
C7T1
vestigated one to four discs per patient, selecting the levels based on the purported pain distribution and/or suspicious findings from imaging studies. Whitecloud and
Seago32 suggested performing discography on three levels, usually C4 C5, C5-C6, and C6 C7. If it had been
assumed that C4 C5 was asymptomatic and therefore
functioning as a control level, this strategy would have
missed inspecting the C3C4 intervertebral level, which
in the current series was found to be almost as frequently
abnormal as C6 C7. Again, on the basis of the high
percentage of patients with multilevel disc disease in this
study, the authors believe that discography is inadequate
if not performed on all accessible cervical discs.
The proportion of cervical discs identified as symptomatic likewise varies among studies. In the current series, 50% of the discs capable of injection produced concordant pain. Parfenchuck and Janssen24 obtained the
slightly higher incidence of 57%. Other studies found
similar rates.1,2 Connor and Darden8 examined 31 patients with discography, reporting that 26 of these (84%)
experienced concordant pain. Although they identified
the C5C6 and C6 C7 disc spaces as those most commonly symptomatic, Connor and Darden did not include
information on how many levels were injected per patient. Shinomiya et al28 compared the proportion of positive pain responses elicited from symptomatic patients
with those of asymptomatic volunteers and found that
65% of the discs injected in the symptomatic group reproduced familiar pain versus 50% of the asymptomatic
group. No information was given regarding nonconcordant pain responses in the symptomatic group. For a
thorough understanding of discographic results, information on concordant and nonconcordant pain responses is essential.
Schellhas et al26 also looked at discographic outcomes
in an asymptomatic group. They found that in 17 of 20
discs found to be normal on MRI, discography identified
painless, anular tears. This result supports the observation that injection of a nonionic contrast material will
not of itself produce pain in asymptomatic patients, even
in the presence of anular tears. Furthermore, the use of
contrast is essential to visualize both painful and nonpainful lesions of the anulus.
This investigation aimed to examine a complete series
of cervical discographic studies, and to determine
whether reproducible patterns of characteristic pain
could be associated with each symptomatic cervical vertebral level identified during discography. As Cloward
first described,7 there is a consistent and predictable area
of pain provoked at each cervical spine segment during
discography. Similar pain distributions have been reported by other authors, most recently by Dwyer et al,9
although they were investigating pain patterns produced
by the stimulation of zygapophysial joints. For the patients in the current study, the pain distributions were
not pathognomonic. There was a great deal of overlap
between the pain produced at each of the cervical spine
segments. Clinically, however, patients with a positive
Unilateral Pain
(%)
Bilateral Pain
(%)
54
66
61
54
52
50
46
34
39
46
48
50
cause half of the patients will have concordant pain produced by stimulation of both the disc and the facet joints,
or because an asymptomatic disc will be associated with
a painful ipsisegmental zygapophysial joint.
Nevertheless, the purpose of discography is to confirm
the identity of painful, abnormal intervertebral disc
spaces when surgery is being considered as a treatment
option for nonradicular discogenic pain. Discography
confirms the diagnosis of discogenic pain. In the experience of the current authors, most patients have pain resulting from multiple disc spaces, and for these patients,
surgical options are severely limited. In patients for
whom surgical intervention is being considered, however, it is imperative to confirm the morphologic integrity of the disc spaces that will be transitional after surgery. Cervical discography, using a contrast medium as
the provocation agent, shows that morphologic abnormalities exist in many cervical disc spaces, even among
asymptomatic volunteers.26 The current authors believe
that this information is essential for avoiding surgical
failure.
Data from this series corroborate observations found
elsewhere, namely that degenerative discs at C6 C7 may
produce anterior chest pains in addition to pain in the
more expected lower neck and shoulder girdle region.26
The discrepancy between patient inclusion of chest pain
on a pain diagram and mention of chest pain in either
clinic notes or during discography is curious and perhaps
evidence of a bias in the pain reports that clinicians ex-
Yes
No
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Chest Pain
Reproduced as
Part of
Concordant Pain
Yes
No
Chest Pain
Included in
Patient Pain
Drawing
Yes
Y
Y
Y
Y
N
N
N
N
N
N
N
N
N
N
No
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
References
1. Aprill C, Bogduk N. The prevalence of cervical zygapophyseal joint pain: A
first approximation. Spine 1992;17:744 7.
2. Aprill C, Dwyer A, Bogduk N. Cervical zygapophyseal joint pain patterns II:
A clinical evaluation. Spine 1990;15:458 61.
3. Bogduk N, Aprill C. On the nature of neck pain, discography and cervical
zygapophyseal joint blocks. Pain 1993;54:21317.
4. Bogduk N, Marsland A. The cervical zygapophyseal joints as a source of
neck pain. Spine 1988;13:610 17.
5. Brodsky AE, Binder WF. Lumbar discography: Its value in diagnosis and
treatment of lumbar disc lesions. Spine 1979;4:110 20. 6. Butler D, Trafimow
JH, Andersson GBJ, McNeill TW, Huckman MS. Discs degenerate before facets.
Spine 1990;15:11113.
7. Cloward RB. Cervical diskography: Technique, indications, and use in diagnosis of ruptured cervical disks. Am J Roentgenol Radium Ther Nuclear Med
1958;79:56374.
8. Connor PM, Darden B VIII. Cervical discography complications and clinical
efficacy. Spine 1993;18:2035 8.
9. Dwyer A, Aprill C, Bogduk N. Cervical zygapophyseal joint pain patterns: I.
A study in normal volunteers. Spine 1990;15:4537.
10. Fraser RD, Osti OL, Vernon-Roberts B. Discitis after discography. J Bone
Joint Surg [Br] 1987;69:26 35.
11. Gore DR, Sepic SB, Gardner GM. Roentgenographic findings of the cervical
spine in asymptomatic people. Spine 1986;11:521 4.
Point of View
Nikolai Bogduk, MD, PhD, DSc
Newcastle Bone and Joint Institute
University of Newcastle
Royal Newcastle Hospital
Newcastle, NSW, Australia
Conviction is not a substitute for truth. Surgeons may be
convinced that they know about neck pain. They may be
convinced that it most commonly arises from the C5C6
disc. They might even test their beliefs by ordering discography at C5C6, and perhaps at either or both of the adjacent levels. When this test proves positive, it appears to
confirm their diagnosis, and repeated confirmations of this
sort reinforce their conviction. But they can be wrong.
When surgeons sample only at their preferred level,
they introduce a sampling bias. They may well prove to
themselves that the C5C6 disc does hurt as suspected,
but when they sample only that disc or its immediate
neighbors, they do not show that other discs are not also
painful. It could be that discogenic neck pain is not unifocal. In which case, testing only the conventional or
preferred level creates illusions of successful diagnosis.
This is what Grubb and Kelly demonstrated in their articles in this issue. By testing all available levels in their
patients, they found that single-level discogenic pain is