Acta Radiologica

ISSN: 0284-1851 (Print) 1600-0455 (Online) Journal homepage: https://www.tandfonline.com/loi/iard20

Atlantoaxial Subluxation

Yasuyuki Yamashita, M. Takahashi, Y. Sakamoto & R. Kojima

To cite this article: Yasuyuki Yamashita, M. Takahashi, Y. Sakamoto & R. Kojima (1989)
Atlantoaxial Subluxation, Acta Radiologica, 30:2, 135-140

To link to this article: https://doi.org/10.3109/02841858909174652

Published online: 07 Jan 2010.

Submit your article to this journal

Article views: 288

Citing articles: 1 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=iard20

FROM T H E DEPARTMENT OF RADIOLOGY, KUMAMOKI UNIVERSITY SCHOOL O F MEDICINE, KUMAMOTO 860,
JAPAN.
ATLANTOAXIAL SUBLUXATION
Radiography and magnetic resonance imaging
correlated to myelopathy
Y. YAMASHITA,
M. TAKAHASHI,
Y. SAKAMOTO
and R. KOJIMA
Abstract
Twenty-nine patients with atlantoaxial subluxation (18 with
rheumatoid arthritis, 2 due to trauma, 4 with 0s odontoideurn.
and one each with polyarteritis nodosa, rheumatic fever, Klippel-
Feil syndrome, achondroplasia, and cause unknown) were evalu-
ated using a 0.22 tesla resistive MRI unit. Cord compression was
classified into four grades according to the degree on magnetic
resonance imaging. There were 7 patients with no thecal sac
compression (grade 0). 10 with a minimal degree of subarachnoid
space compression without cord compression (grade I ) , 7 with
mild cord compression (grade 2). and 5 with severe cord com-
pression or cord atrophy (grade 3). Although the severity of
myelopathy showed poor correlation with the atlantodental inter-
val on conventional radiography, high correlation was observed
between MR grading and the degree of myelopathy. The high
signal intensity foci were observed in 7 of 12 patients with cord
compression (grades 2 and 3) on T2 weighted images. Other
frequently observed findings in rheumatoid arthritis included soft
tissue masses of low to intermediate signal intensity in the para-
odontoid space, erosions of the odontoid processes, and atlanto-
axial impaction on T1 and T2 weighted images.
Key words: Atlantoaxial subluxation, MRI; myelopathy.
Atlantoaxial subluxation (AAS) secondary to a variety
of etiologic conditions usually has serious prognostic im-
plications (2, 13, 26). Especially patients with AAS pre-
senting with severe persisting pain despite conservative
therapy or with progressive neurologic symptoms need
immediate surgical treatment. However, a poor correla-
tion exists between the degree of subluxation on conven-
tional radiography and the degree of myelopathy (1 1, 21,
22, 25). Magnetic resonance imaging (MRI) has been in-
troduced as a useful diagnostic method for the spine and
spinal cord (4, 5, 9, 15). MRI can directly visualize more
Acta Radiologica 30 (1989) Fasc. 2
-
changes in greater detail, in AAS, than other imaging
techniques.
The present investigation was undertaken to clarify the
clinical usefulness of MRI in AAS.
Material and Methods
Twenty-nine patients ( I 1 men and 18 women, aged
14-79 years, mean 54.2 years) with AAS were examined
with MRI between June 1986 and January 1988 at the
Kumamoto University and its affiliated hospitals. The
clinical diagnoses indicating AAS were rheumatoid arthri-
tis (n=18), polyarteritis nodosa and rheumatic fever (one
case each), 0s odontoideum (n=4), Klippel-Feil syn-
drome, achondroplasia, and cause unknown (one case
each), and trauma (n=2).
The clinical symptoms in these patients were occipital
or cervical pain in 16, decreased range of neck movement
or neck stiffness in 2, and vertigo in 2. Neurologic symp-
toms observed were due to radiculopathy in 4 and cervical
myelopathy in 21 cases. The major symptoms of myelopa-
thy included spasticity, weakness, abnormal reflex, dys-
esthesia, hypesthesia and urinary tract insufficiency. The
duration of myelopathy ranged from I month to 4 years
(average 10.8 months). The clinical symptoms and my-
elopathy were graded according to the method proposed
by PELLICCI et coll. (18) and LAASONEN et coll. (8): I )
Normal-no neurologic symptoms; 2) mild-hyper-re-
flexia, dysesthesia; 3) moderate-mild weakness, hyp-
Accepted for publication 20 September 1988.
135
136 Y. YAMASHITA, M. TAKAHASHI, Y. SAKAMOTO A N D R. KOJIMA

Table
MRZ findings in atlantoaxial subluxation according to underlying etiologies
Findings Rheumatoid Other forms Bone ano- 'Ifauma
arthritis of inflamma- maly (n=7) (n=2)
(n=18) tion (n=2)

Atlantoaxial impaction 3 0 2 0
Periodontoid soft tissue 16 2 1 0
Erosion of odontoid process 8 2 0 0
Abnormal bone 0 0 5 0

esthesia; 4) severe-severe weakness, paralysis, severe

muscle atrophy, anesthesia.
The diagnoses of AAS were made from conventional
radiographs in flexion, neutral position and extension.
Conventional tomography was carried out in 18 cases.
The criterion for AAS was a distance of 2.5 mm or greater
between the anterior tubercle of C1 and the anterior as-
pect of the dens or odontoid process on the lateral radio-
graphs or at tomography (6), in which the greatest dis-
tance in three positions (in most cases in flexion) was used
as the maximum atlantodental interval (ADI). Atlantoax-
ial impaction was considered to be present if the tip of the
odontoid process was 8 mm above McGregor's baseline in
men or 9.7 mm above the baseline in women (7). Radio-
graphy was performed within one month of MRI in all
patients.
MR imaging was performed with a 0.22 tesla resistive
MR unit (Toshiba MRT-22A) with a head or surface coil.
Images were obtained with a spin-echo (SE) technique
performed with repetition times (TR) of either 400 or 500
ms and echo-times (TE) of 40 ms to acquire T1 weighted
images; TWTE 1600/40 or 2O00/40 ms for proton density
images; and TR 1600/60 or 2000/80 ms sequences for T2
weighted images. The slice thickness was 5 mm. Four
averagings were used for SE 400-500/40 sequences and
two for SE 1600-2O00/40 and 1600-2O00/60-80 se-
quences. Images were acquired with a 256x256 matrix.
The surface coil was used with a field of view of 6 x
15 cm2. Sagittal MR images were obtained in the maxi-
mum ADI. In the patients who showed atlantoaxial insta-
bility, sagittal images were obtained in flexion, neutral
position and extension (7 patients).
MR grading of cord compression was classified into
four grades (Fig. 1). The subarachnoid space was better
observed on T2 weighted images and the spinal cord was
assessed by T1 weighted images. There were 7 patients
with no thecal sac compression (grade 0), 10 patients with
a minimal degree of subarachnoid space compression
without cord compression (grade l), 7 patients with mild
cord compression (grade 2), and 5 patients with severe
cord compression or cord atrophy (grade 3).
The clinical grading of myelopathy was compared with
AD1 on conventional radiography and the MR grading of
the cord compression. In addition, special attention was
Grade 0 Grade 1 Grade 2 Grade 3
Fig. 1. MRI classification of spinal cord compression. Subarach-
noid space (O), spinal cord (W).
2 5 10 15
AD1 (mm)
Fig. 2. Detectability of AAS by MRI in relation to maximum AD1
measured by conventional radiography. Detectable (01, unde-
tectable (x).
paid to the bony and soft tissue abnormalities around C1
and C2.
Results
The detectability of atlantoaxial subluxation with MRI
in relation to AD1 measured by conventional radiography
is shown in Fig. 2. AAS was recognized as an increase in
the preodontoid space on MRI (Fig. 3). However, 12 of 29
cases of AAS diagnosed by conventional radiography
were not detected because of disappearance of the bone
marrow fat of the odontoid process (Fig. 4).
The MR findings in the bony and soft tissue structures,
for the respective underlying clinical etiologies, are listed
in the Table. Disappearance of the bone marrow fat of the
odontoid process corresponded to bone erosions on con-
ventional radiography. In evaluating erosions of the odon-
toid process, MRI was more sensitive in 6 cases, and
equally sensitive in 6 cases, although changes in the odon-
toid tip were difficult to evaluate because of the possibility
of failure to obtain a true midline section through the
odontoid, which could result in a low signal intensity due
 ATLANTOAXIAL SUBLUXATION 137

Fig. 3 Fig. 4 a Fig. 4 b

Fig. 3. A 54-year-old female with polyarteritis nodosa had devel-
oped myelopathy 12 months previously. The sagittal T1 weighted
SE image (TR/TE = 400/40) in neutral position demonstrates
increased atlantoaxial distance ( H ). A soft tissue mass is
observed in the preodontoid space.
Fig. 4. A 53-year-old female with a 30-year history of rheumatoid
arthritis had, on conventional radiography in neutral position,
AAS (AD1 7 mm) without atlantoaxial instability. a) T1 weighted
SE image ( T W E = 500/40) and b) T2 weighted SE image ( T W E
= 2000/80). The sagittal images in neutral position of both pulse
sequences reveal slight arachnoid space compression at the cervi-
comedullary junction. AAS cannot be detected because of disap-
pearance of the bone marrow fat of the odontoid process.

Fig. 5 a Fig. 5 b Fig. 6

Fig. 5. A 53-year-old female with a 34-year history of rheumatoid
arthritis. She developed paresthesia of both extremities and gait
disturbance 2 months prior to the present MRI examination.
Conventional radiography showed AAS with AD1 of 15 mm in
flexion and neutral position. a) TI weighted SE image (TR/TE =
500/40) and b) T2 weighted SE image (TR/TE = 2000/80) in
neutral positions demonstrate a low signal intensity mass in the
preodontoid space and the spinal cord is constricted at the cervi-
comedullaryjunction. T2 weighted image reveals increased signal
intensity foci at that site.
Fig. 6. A 52-year-old female with a 20-year history of rheumatoid
arthritis had developed occipital pain and paresthesia of both
upper extremities 16 months previously. T1 weighted SE image in
neutral position shows atlantoaxial impaction associated with
AAS (AD1 6 mm on conventional radiography). The medulla is
compressed by the tip of the odontoid process, while the cervical
cord is compressed by the posterior arch of C1.
138 Y. YAMASHITA, M. TAKAHASHI, Y. SAKAMOTO A N D R. KOJIMA

-
0
0
._ normal 0 0 0 0 ma3
-
._
V
5 10 15
AD1 (mm)
Fig. 8. Degree of myelopathy in relation to maximum AD1 meas-
ured by conventional radiography. Atlantoaxial impaction pres-
ent (0)and not present (0).

). severe 0 02
5
a
-
u

.
moderate 0 rm a
c

01
- COJ)
mild 0 m 0
W
Fig. 7. A 39-year-old male with a 4-year history of severe myelo-
pathy. Conventional radiography showed AAS (AD1 5 mm) due
to 0s odontoideum and the spinal canal diameter at CI was small.
TI weighted SE image (TR/TE 400/40) in flexion reveals marked
Grade 0 Grade 1 Grade 2 Grade 3
atrophy of the spinal cord. High signal intensity due to 0s odon-
toideum is also observed (+).
MR Grading of Myelopathy

Fig. 9. Degree of myelopathy in relation to MRI grading of cord

compression. Foci of increased signal intensity (0).

to volume averaging of the low signal through the odon-

toid cortex.
Soft tissue masses of low to intermediate signal intensi-
ty on T1 and T2 weighted images were observed in the
periodontoid space in 15 patients with rheumatoid arthri-
tis or other inflammatory lesions (Fig. 5). Three patients
with rheumatoid arthritis showed a soft tissue mass of low
signal intensity on T1 and of high signal intensity on T2
weighted images. An intermediate signal intensity mass
was also observed in a patient with a lesion of unknown
etiology.
Atlantoaxial impaction associated with AAS was ob-
served in 3 cases of rheumatoid arthritis, and in 1 case
each of 0s odontoideum and Klippel-Feil syndrome. It
was also demonstrated on conventional radiography, but
MRI allowed more detail and clear demonstration of the
odontoid process (Fig. 6). The 0s odontoideum and other
bone anomalies were clearly visible on MRI (Fig. 7).
The degree of myelopathy.did not necessarily correlate
with the distance of subluxation on conventional radiogra-
phy, but MR grading showed a good correlation with the
degree of myelopathy (Figs 8,9). The etiology of AAS had
no relation to the degree of myelopathy. The cord was
extremely atrophic in patients with long-standing myelo-
pathy (Fig. 7). Seven of 12 patients (58.3%) with cord
compression (grades 2 and 3) showed foci of increased
signal intensity in the spinal cord on T2 weighted images
(Fig. 5 ) . This high signal intensity was often observed in
patients with severe myelopathy. The duration of myelo-
pathy, with this finding, ranged from 1 month to 2 years.
Discussion
Atlantoaxial subluxation may occur as a result of var-
ious conditions such as a trauma, congenital maldevelop-
ment of the odontoid process, and destructive diseases
including rheumatoid arthritis, infections or neoplasms (2,
13, 26); of these, rheumatoid arthritis was the most fre-
quent in our series.
Approximately 30 to 40 per cent of all patients with
chronic rheumatoid arthritis have radiologic evidence of
cervical spine subluxation (3, 10, 12, 16). Several types of
C 1 X 2 subluxation can be distinguished radiologically:
anterior, posterior, lateral, rotatory subluxation of C1 on
C2, and atlantoaxial impaction. Anterior subluxation
(AAS) is the most common form of these abnormalities.
Several synovial bursae and their associated ligaments lie
around the odontoid. Chronic inflammation of these bur-
sae and ligaments leads to erosion of the bones, with
subsequent loss of stability. Abnormal synovial prolifera-
tion and pannus formation are observed (22). The soft tis-
sue masses around the odontoid process, of low to inter-
mediate signal intensity on T1 and T2 weighted images,
 ATLANTOAXIAL SUBLUXATION 139

are due to the synovial proliferation or pannus-containing inability of MRI to image the cortical bone in the odontoid
fibrous tissue or hemosiderin, according to their stages (1, process.
19). These masses were also observed in other inflamma-
Request for reprints: Dr Yasuyuki Yamashita, Department of
tory processes, such as polyarteritis nodosa and rheumat- Radiology, Kumamoto University School of Medicine, 1-1-1
ic fever, in our series. Only one patient with an anomaly Honjo, Kumamoto 860,Japan.
showed this finding. SZE et coll. (23) reported three cases
with a mass in chronic AAS as a result of mechanical
REFERENCES
irritation due to degenerative disease and congenital dys-
I . AISENA. M., MARTEL W., ELLISJ. H. and MCCUNEW. J.:
plasia of the dens.
Cervical spine involvement in rheumatoid arthritis. MR
Conventional radiographs of the neck readily demon- imaging. Radiology 165 (1987), 159.
strate AAS. Several authors emphasize poor correlation 2. BREENBERG A. D.: Atlantoaxial dislocations. Brain 91 (l968),
between radiographically demonstrated AAS and clinical 655.
signs of myelopathy (11, 21, 22, 25). Because of the wide 3. CABOTA. and BECKER A.: The cervical spine in rheumatoid
arthritis. Clin. Orthop. 131 (1978), 130.
spinal canal in the upper cervical region, severe AAS can 4. DEE G. J., BELLOJ. A. and HILALS. K.: High field, thin
exist without any neurologic symptoms, but patients with section nuclear magnetic resonance imaging of the cervical
such abnormalities are clearly susceptible to sudden death spine. Cardiovasc. Intervent. Radiol. 8 (1986). 283.
due to medullary compression (14,20). On the other hand, 5. HAN J. S., KAUFMANB., EL YOUSEFS. J. et coll.: NMR
imaging of the spine. Amer. J. Neuroradiol. 4 (1983). 1151.
patients with a congenital narrow spinal canal at C 1 X 2 6. HINCKV. C. and HOPKINS C. E.: Measurement of the atlanto-
may have serious neurologic defects in spite of mild AAS. dental interval in the adult. Amer. J. Roentgenol. 84 (1960).
Our examination revealed that the cord was compressed 945.
not only by the bony structures, but also by a radiolucent 7. _ _ and SAVARA B. S.: Diagnostic criteria of basilar im-
soft tissue mass, in patients with rheumatoid arthritis or pression. Radiology 76 (1961), 572.
8. LAASONEN E. M., KANKAANPAEAE U., PAUKKU P., SANDELIN
other inflammatory lesions. In addition, patients with J., SERVOA. and SLATISP.: Computed tomographic myelo-
long-standing myelopathy showed severe cord atrophy. graphy (CTM) in atlanto-axial rheumatoid arthritis. Neurora-
These conditions may also be responsible for discrepancy diology 27 (1985), 119.
between the degree of myelopathy and the AD1 on con- 9. LEEB. C. P., DECKM. D. F., KNEELAND J. B. and CAHILL P.
T.: MR imaging of the craniocervical junction. Amer. J.
ventional radiography. Neuroradiol. 6 (1985), 209.
Although the detection of AAS by MRI is less reliable 10. MARTEL W.: Pathogenesis of cervical discovertebral destruc-
than with conventional radiography, MRI can provide tion in rheumatoid arthritis. Arthritis Rheum. 20 (l977), 1217.
useful information regarding the spinal cord and the sur- 1 1 . MATHEWS J. A.: Atlanto-axial subluxation in rheumatoid ar-
rounding structures. There was good correlation between thritis. A 5-year follow-up study. Ann. Rheum. Dis. 33
(1974), 526.
the MR grading of the cord compression and the degree of 12. MEIKLEJ. A. and WILKINSON M.: Rheumatoid involvement of
myelopathy. CT myelography can demonstrate cord com- the cervical spine. Radiologic assessment. Ann. Rheum. Dis.
pression in rheumatoid AAS (8, 17), but MRI is more 30 (1971), 154.
sensitive and non-invasive for evaluating the spinal cord. 13. MESCHENI.: Cervical 'spine. In: Roentgen sign in clinical
practice, p. 654. W. B. Saunders Company, Philadelphia,
It has been reported that MRI in cervical flexion and London 1966.
extension is useful for evaluating spinal compression in 14. MIKULOWSKI P., WOLLHEIM F. A,, ROTMILP. and OLSENI.:
AAS. However, as it is very painful or sometimes danger- Sudden death in rheumatoid arthritis with atlanto-axial dislo-
ous for patients with severe myelopathy to hold the same cation. Acta Med. Scand. 198 (1971), 445.
15. MODICM. T., WEINSTEIN M. A., PAVLICEK W. et coll.: Nucle-
position for a long time this was performed routinely only
ar magnetic resonance imaging of the spine. Radiology 148
in the maximum AD1 position, except in patients who (1983), 757.
showed atlantoaxial instability. 16. NAKANO K. K.: Neurologic complications of rheumatoid ar-
The foci of increased signal intensity in the spinal cord thritis. Orthop. Clin. Amer. 6 (1979, 861.
were frequently observed in patients with cord compres- 17. OSBORNE D., ' R ~ O LP.,
O DUBOISP., DRAYER B. and HEINZE.:
Assessment of craniocervical junction and atlantoaxial rela-
sion with varying etiology (24). The incidence of foci of
tion using metrizamide-enhanced CT in flexion and exten-
high signal intensity is high in AAS. This is probably sion. Amer. J. Neuroradiol. 4 (1983), 843.
because the spinal cord is more severely constricted in 18. PELLICCI P. M., RANAWAT C. S., TSAIRISP. and BRYAN W. J.:
this condition. Although its pathogenesis has not been A prospective study of the progression of rheumatoid arthri-
clarified, myelomalacia with edema, demyelination or tis of the cervical spine. J. Bone Jt Surg. 16 (1981). 342.
19. REYNOLDS H., CARTERS. W., MURTAGH F. R., SILBIGER M.
gliosis of the spinal cord have been suggested as possible and RECHTINE G. R.: Cervical rheumatoid arthritis. Value of
important factors (17). flexion and extension views in imaging. Radiology 164 (1987),
In conclusion, MRI is useful for the preoperative and 215.
follow-up examinations of AAS, because it directly visu- 20. SMITHH. P., CHALLAV. R. and ALEXANDER E.: Odontoid
compression of the brainstem in a patient with rheumatoid
alizes the spinal cord, the subarachnoid space, and sur- arthritis. J. Neurosurg. 53 (1980), 841.
rounding bony and soft tissue structures. However, the 21. - BENNR. T. and SHARP J.: Natural history of rheumatoid
degree of subluxation may be underestimated because of cervical luxations. Ann. Rheum. Dis. 31 (1972), 431.
140 Y. YAMASHITA, M. TAKAHASHI, Y. SAKAMOTO A N D R. KOJIMA
22. STEVENS J. C., CARTLIDGE N. E. F., SAUNDERS M., APPLEBY
A., HALLM. and SHAWD. A,: Atlanto-axial subluxation and
cervical myelopathy in rheumatoid arthritis. Q. J. Med. 40
(1971), 391.
23. SZE G., BRANT-ZAWADZKI M. N., WILSON C. R., NORMAN D.
and NMON T. H.: Pseudotumor of the craniovertebral junc-
tion associated with chronic subluxation. MR imaging stud-
ies. Radiology 161 (1986), 391.
24. TAKAHASHI M., SAKAMOTO Y.,MIYAWAKI M. and BUSSAKA H.:
Increased MR signal intensity secondary to chronic cervical
cord compression. Neuroradiology 29 (1987), 550.
25. WEISSMAN B. N. W., ALIABADI P., WEINFELD M. S., THOMAS
W. H. and SOSMAN J. L.: Prognostic features of atlantoaxial
subluxation in rheumatoid arthritis patients. Radiology 144
(1982), 745.
26. WORTZMAN G. and DEWARF. P.: Rotary fixation of the atlan-
toaxial joint. Rotational atlantoaxial subluxation. Radi-
ology 90 (1%8), 479.