ORIGINAL ARTICLE

Treatment of basilar invagination with atlantoaxial dislocation:atlantoaxial joint distraction and fixation with transoralatlantoaxial reduction plate (TARP) without odontoidectomy

Hong Xia • QingShui Yin • FuZhi Ai •

XiangYang Ma • JianHua Wang • ZengHui Wu •

Kai Zhang • JingFa Liu • JunJie Xu

Received: 14 December 2013 / Revised: 6 May 2014 / Accepted: 7 May 2014

� Springer-Verlag Berlin Heidelberg 2014

Abstract

Purpose Although direct transoral decompression and

one-stage posterior instrumentation can obtain satisfactory

cord decompression for the treatment of basilar invagination

with atlantoaxial dislocation, surgical injuries run high as

combinative anterior-posterior approaches were necessary.

Furthermore, the complications will rise notably when

involvement of dens and/or clivus in the decompression

necessitates relatively complicated surgical techniques. First

initiated in 2005, transoral atlantoaxial reduction plate

(TARP) works as an internal fixation for the treatment of

basilar invagination with irreducible atlantoaxial disloca-

tion. Therefore, this article aimed to describe several oper-

ative experiences about this approach, which has delivered

successful decompression, fixation and fusion.

Methods 21 consecutive patients with basilar invagina-

tion underwent the TARP operation. The pre- and post-

operative medulla-cervical angles were measured and

compared. The JOA scores of spinal cord function were

calculated pre- and post-operatively. 20 cases (20/21) were

followed up to average 12.5 months.

Results Symptoms of all the 20 cases were relieved in

different degrees. The postoperative imaging showed the

odontoid processes obtained ideal reduction and the internal

fixators were all in good position. The medulla-cervical

angle was correctd from an average (± standard deviation)

128.7� ? 11.9� (n = 20) before surgery to 156.5� ? 8.1�(n = 20) after surgery (P \ 0.01). The average preoperative

and postoperative Japaneses Orthopedic Association scores

were 11.25 (n = 20) and 15.9 (n = 20), respectively, indi-

cating 76 % improvement. Screw-loosening was observed

in one patient due to severe osteoporosis. After a revised

operation with a TARP in another size, the neurological

symptoms showed no obvious improvements. Then the

treatment was terminated.

Conclusions The TARP operation and intra-operative

traction could reduce the odontoid process superiorly

migrating into the foramen magnum, directly ease the

ventral compression of spinal cord, and fix the reduced

atlantoaxial joints through a single transoral approach

without the need of a posterior operation. In this stury, 21

patients were evaluated and 20 did well with TARP oper-

ation. The preliminary clinical result was satisfactory.

Keywords Transoral � Basilar invagination � Atlantoaxial

dislocation � Arthrodesis � Internal fixation

Introduction

Basilar invagination (BI) is a condition characterized by

telescoping of the upper cervical spine into the foramen

magnum [1, 2]. For symptomatic irreducible basilar

invagination with/without atlantoaxial dislocation, transoral

decompression and subsequent posterior fixation have been

the most accepted treatment protocol [3–5]. And odontoi-

dectomy is adopted for this ventral decompression, which

could relieve compression on the brainstem and upper

spinal cord by the dens. Although direct transoral decom-

pression and one-stage posterior instrumentation can obtain

satisfactory cord decompression, it entails a combinative

anterior-posterior approach management, which will

increase surgical injuries. Furthermore, the decompression

H. Xia (&) � Q. Yin � F. Ai � X. Ma � J. Wang � Z. Wu �K. Zhang � J. Liu � J. Xu

Department of Orthopedics, Guangzhou General Hospital of

Guangzhou Military Command (Liuhuaqiao Hospital), No.111

Liuhua Avenue, Guangzhou, People’s Republic of China

e-mail: hong_xia126@126.com

123

Eur Spine J

DOI 10.1007/s00586-014-3378-8

procedure is relatively complicated and more complications

may be induced.

Materials and methods

From September, 2004 to December, 2009, Transoral

Atlantoaxial Reduction Plate (TARP; manufactured by

Weigao Orthopaedic Device Co. Ltd., Weihai City, Peo-

ple’s Republic of China). system, which was originally

designed by Yin [6], was applied in our department on 21

consecutive patients with basilar invagination, of which the

etiology was congenital osseous malformations. The sur-

gical results were inspiring. The surgical technique and the

clinical results were introduced in detail as follows

(Figs. 1, 2).

Clinical data

21 patients (male 9 and female 12) averagely 38 years old

(ranging from 21 to 64) were diagnosed with basilar

invagination. All the patients manifested neurologic

symptoms in different degrees including limited cervical

motion, neck and shoulder pain, hyper muscular tension of

limbs, positive pathologic reflex, and decreased motor

power and algesia. The patient history spanned from

6 months to 22 years (average 105.8 months). Among

them, three cases had the history of trauma and four cases

had undergone the removal of posterior arch of atlas and

the posterior edge of the foramen magnum through pos-

terior approach but with no obvious curative effects. All

the patients received preoperatively cervical examinations

of three-dimension CT scanning, MRI, and X-ray of AP

Fig. 1 a: TARP internal fixation. b:Atlantoaxial reduction procedure:

(1) closure of the handgrips can impart a loca distraction force

between C1 and C2, and then the plate together with C1 will ascend

through the open slot in the inferior portion of the plate based on the

C2 reduction screw; (2) the nut on the upper arm of the reduction

instrument could be turned, which applies a posterior reduction force

to the upper portion of the plate, displacing C1 posteriorly with

respect to C2

Fig. 2 The sequential reduction

and distraction of the

atlantoaxial joint

Eur Spine J

123

view, lateral view, and opening mouth view. The tips of all

the odontoid processes were over 3 mm higher above the

Chamberlain lines with the medulla-cervical angle from

128.4� to 156.5�. 18 cases were accompanied with atlas

hypoplasia or atlas occipitalization. Three cases were

accompanied with Arnold-Chiari deformation (Fig. 3).

Preoperative preparation

All the patients underwent biaxial skull traction for about

1 week with the weight of 5–10 kgs. Bedside lateral X-ray

was then performed and disclosed that the reduction of

odontoid process was not thorough in all the cases. The

preoperative oral cavity cleansing was conducted with

chlorhexidine collutory for more than 3 days. Dental

cleaning was also performed before the operation.

Surgical techniques

A detailed description of surgical procedures can be found

in the report by Yin et al. [6]; only the main points were

described here.

1. Exposure and release: The patient was placed in

supine position with skull traction. Carry the midline

incision at the posterior pharyngeal wall approximate

4–6 cm. Because the base of the skull is relatively higher

than the normal in cases of basilar invagination, the inci-

sion should be higher than the routine. Sometimes it needs

to be prolonged to the lower part of the clivus. The longus

capitis and collilongus were detached bilaterally to expose

the prevertebral structure and the both lateral atlantoaxial

joints. Contractural scar tissue or hyperplastic osteotylus

between atlas and odontoid vertebra, contractural articular

capsule and scar tissue around lateral mass and articular

cartilage were removed. After sufficient release, atlas

would show some sign of loosening and some reduction by

skull traction in some patients.

2. Reduction and fixation: (1) A suitable TARP plate

was placed on C1 and two screws were anchored in the

bilateral lateral masses of C1 to fix the plate. Then atlas and

the plate formed a complex of reduction. (2) A temporary

reduction screw was immobilized in the vertebrae of axis

with its head 2–3 mm higher than the surface of the plate.

C2 and the temporary fixation screw then formed another

reduction complex (reduction fulcrum). (3) With skull

traction maintained, atlas dislocated downward was braced

upward by popping the upper arm of the reductor that held

cross beam of plate and the inferior arm of the reductor that

held temporary reduction screw (the temporary reduction

screw could pass the runner of the plate). (4) The complex

of atlas and the plate were pushed back by turning the nut

at the superior of the reductor until the atlas was reduced.

(5) It was confirmed that the sagittal diameter of canalis

spinalis recovered and cervical cord was sufficiently

decompressed by C-arm fluoroscopy. (6) After the plate

was fixed and locked with the other two bicortical screws at

C2, the temporary reduction screw was removed. After

above procedures completed, the atlas and axis were fixed

in the condition of reduction. The fixation was strong

enough due to two screws in atlas leaning laterally and two

screws in odontoid vertebra leaning inward.

3. Bone graft: autogenous iliac crest bone was harvested

and filled in the bilateral joint space between C1 and C2

through the plate window.

4. Incision closure: plate was covered with paravertebral

muscle. Muscular layer and membrana mucosa of posterior

pharyngeal wall were sutured respectively.

Postoperative management

Nasal trachea cannula were maintained for at least 2 days

with nasogastric nutrition fed for 2 days after the operation.

Oral cavity ultrasonic nebulization was performed three

times a day for one week. Prophylactic antibiotics was

administered for 5–7 days and mannitol and methylpred-

nisolone twice a day for 3 days. Neck collar protection was

kept for 2–3 months. If obvious osteoporosis existed, halo

vest should be placed after operation and maintained for

3 months.

Statistical analysis

Means and standard deviation were obtained. Preoperative

and postoperative medulla-cervical angle measurements

were compared using the student t test with statistical

significance set at a P value of \ 0.01.

Results

20 patients were followed up for average 12.25 months

(ranging from 4 to 36 months) whose symptoms of were

relieved to different degrees. The postoperative imaging

visualized ideal reduction of the odontoid processes with

their upper edges 3 mm lower than the Chamberlain lines

and the internal fixators all in good position. The medulla-

cervical angle was correctd from an average(± standard

deviation) 128.7� ? 11.9� (n = 20) before surgery to

156.5� ? 8.1� (n = 20) after surgery (P \ 0.01). The

average preoperative and postoperative Japaneses Ortho-

pedic Association scores were 11.25 (n = 20) and 15.9

(n = 20), respectively, indicating 76 % improvement

(Table 1, 2). Screw-loosening was observed in one patient

due to severe osteoporosis. After a revised operation with a

TARP in another size, the neurological symptoms failed to

improve obviously. Then the treatment was terminated.

Eur Spine J

123

Fig. 3 Compared with

preoperative X-ray, CT and

MRI (Left) postoperative results

(Right) showed that the upward

and backward migrated

odontoid had been relocated

ideally, and the ventral

compression on the spinal cord

had been decompressed

thoroughly

Eur Spine J

123

Bony fusion was obtained in all patient at the last time of

follow-up.

Discussion

A complex occipitocervical deformity resulting from

abnormal skull development, Basilar invagination with

atlantoaxial dislocation can cause cervicomedullary com-

pression and inflict progressive neurologic deficit on

patients. Its treatment protocol requires not only to relieve

the compression, but also to provide stabilization after

decompression [7].

Since predominantly the anterior brain stem and upper

cervical cord were pressed in patients with basilar invagi-

nation, the odontoid process, and the lower clivus if nec-

essary can be resected to achieve ventral decompression

whose procedure can be operated through a transoral,

transnasal [5] or a standard anterior cervical approach [8].

However, as the dens should be gradually ground with a

high-speed burring tool and the alar ligament and apical

ligament resected during its procedure, Odontoidectomy is

of high risk through any aforementioned approach and

could incur spinal cord injury and dural tear, the most

frequent complications [9, 10].

Nevertheless, odontoidectomy was spared in all our

cases because enough distraction between the atlantoaxial

joint was realized instead. Goel who secured two titanium

spacers within the bilateral atlantoaxial joints through the

posterior approach, observed that the distraction and

impaction of the titanium spacer within the joint cavity

increased the height of the lateral masses, reduced the

basilar invagination and restored the craniovertebral

alignments [4, 11]. In his report in 2005, a significant,

though incomplete, reduction of basilar invagination was

garnered in both cases and the clinical symptoms of the

patients improved dramatically. Francesco Cacciola [12]

reported a case with basilar invagination treated surgically

through the posterior approach for atlantoaxial distraction

and also proved atlantoaxial distraction as a good way for

decompression. These correspond to our research findings.

When there was enough distraction between the atlanto-

axial joint, the tip of the dens would recede taking the

Chamberlain’s line as reference, thus completing the

decompression. Moreover, to achieve distraction between

the atlantoaxial joints, an essential conduit is through pre-

and intra-operative skull traction as traction in cases with

basilar invagination and atlantoaxial subluxation results in

a significant improvement in the craniovertebral align-

ments [13]. Another way is to use the specially designed

Table 1 Clinical manifestations and neurological outcomes

No Sex Age Neck pain Bilateral Babinski Hypoesthesia Muscle weakness

Preoperative Postoperative Preoperative Postoperative Preoperative Postoperative Preoperative Postoperative

1 M 29 Yes No – – Yes No 4.5 5

2 F 64 Yes Yes ? ? Yes Yes 2 3

3 F 26 Yes No ? ? Yes No 4 5

4 F 32 Yes No ? – Yes No 4.5 5

5 M 21 No No – – No No 4.5 5

6 F 55 Yes No ? ? Yes Yes 3 4.5

7 M 28 Yes No ? – Yes No 4 5

8 M 31 Yes No ? – Yes No 4.5 5

9 F 51 Yes No ? ? Yes Yes 4 5

10 F 33 Yes No ? – Yes No 3 5

11 M 39 Yes No ? ? Yes Yes 3 4

12 M 34 Yes No ? ? Yes No 4.5 5

13 F 35 Yes No ? – Yes No 4.5 5

14 F 25 Yes No – – Yes No 5 5

15 F 37 Yes Yes ? – Yes No 5 5

16 M 54 Yes No ? ? Yes Yes 3 4.5

17 F 42 Yes No ? – Yes Yes 3 4

18 F 43 Yes No ? ? Yes No 4 5

19 M 31 Yes No ? – No No 5 5

20 F 43 Yes No ? – Yes No 4.5 5

21 M 45 Yes No ? ? Yes No 3 4

Eur Spine J

123

reduction which has been described as technique of trans-

oral atlantoaxial reduction plate (TARP) surgery for the

treatment of atlantoaxial dislocation from the year 2005 [6,

14–17]. By this technique, the atlantoaxial joint was dis-

tracted widely with the specially designed reductor, and the

distraction induced was well maintained by the TARP

fixation in this one-stage surgery. The anteriorly dislocated

atlas could also be reduced with the help of this reductor.

The distraction could be maintained via the tricortical bone

grafts harvested from the iliac crest and inserted into the

bilateral atlantoaxial joint spaces. Resection of local scar

and articular capsule is also very important for distraction.

One of the advantages of this TARP surgery is that it

could achieve decompression and fixation in a single

approach. The screw for the axis fixation hereof was ped-

icle axis screw. Usually, the entrance point of the anterior

axis pedicle screw was 5 mm under the superior facet of

the axis and 7 mm to the midline of the axis, at lateral and

inferior angles respectively of 25� and 20� However, some

patients are clinically not candidates for anterior axis

pedicle screw fixation, particularly the subset of patients

with basilar invagination due to the deep invagination of

C2. Axis trans-facet screw fixation could therefore be used

instead and would effectively prevent vertebral artery

injury. The entrance point of this axis screw is 5 mm under

the superior facet of the axis and 7 mm to the midline of

the axis, managed at 25� lateral and parallel to the facet.

There were three cases accompanied with Arnold-Chiari

deformation in this series. The treatment of chiari malfor-

mation (CM) includes a standard suboccipital craniectomy,

with cervical laminoplasty with or without duraplasty.

Although this posterior surgery could relief patients’

symptoms, patients with CM and basilar invagination (BI)

with a syrinx may not require posterior decompression if

cerebrospinal fluid (CSF) flow at the level of the foramen

magnum is restored via anterior decompression [18].

During our clinical practices, if a patient was found to have

BI with CM, anterior transoral decompression and fusion

was the first priority. According to the improvement of the

patient’s clinical manifestations and/or syringomyelia, we

could decide whether a posterior surgery is needed. In this

series, the clinical symptoms and the syrinx of the three

cases with CM had improved dramatically. Anterior

decompression via distraction of the atlantoaxial joint was

performed without the need for posterior decompression of

the CM.

Table 2 Clinical outcomes and radiographic measurements

No Sex Age JOA Score Neurologic

improvement (%)

Complications Cervicomedullary angle (CMA)

Preoperative Postoperative Preoperative Postoperative

1 M 29 16 17 100 No 138 153

2 F 64 6 – – Screw-loosening 122 –

3 F 26 13 17 100 No 129 163

4 F 32 16 17 100 No 107 146

5 M 21 15 16 50 No 121 163

6 F 55 7 14 70 No 125 150

7 M 28 11 16 83.3 No 121 152

8 M 31 12 15 60 No 152 168

9 F 51 15 17 100 No 138 166

10 F 33 8 13 55.6 No 148 165

11 M 39 7 13 60 No 144 162

12 M 34 15 17 100 No 127 147

13 F 35 15 17 100 No 133 159

14 F 25 16 17 100 No 123 152

15 F 37 16 17 100 No 121 151

16 M 54 9 16 87.5 No 119 145

17 M 42 10 15 71.4 No 124 150

18 F 43 14 17 100 No 141 170

19 M 31 15 17 100 No 118 157

20 F 43 16 17 100 No 135 166

21 F 45 9 13 50 No 110 145

Neurologic improvement [25] = ([Postoperative JOA] - [Preoperative JOA])/(17 - [Preoperative JOA])

JOA Japanese Orthopedic Association

Eur Spine J

123

In recent years, minimally invasive endoscopic odontoi-

dectomy through transnasal approach has been carried out by

some researchers [9, 19]. Inspiring as the clinical results are,

absence of rigid internal fixation and fusion might not be able

to prevent further cranial settling after an anterior decom-

pression procedure [20–22]. Meanwhile, some authors

introduced the aggressive posterior reduction surgery for the

treatment of basilar invagination without the need for ante-

rior decompression [1, 11, 23, 24]. However, the value of the

anterior transoral approach should not be underestimated.

For such occipitocervical deformities as defects of posterior

arch of C1, congenital small pedicles and occipital dysplasia,

posterior operations for stabilization are identified extremely

difficult to perform. For these patients, the TARP operation

could provide surgeons with an option to perform a direct

anterior fixation so as to achieve stabilization.

Conclusion

The TARP operation and intra-operative traction could

reduce the odontoid process superiorly migrating into the

foramen magnum, directly mitigate the ventral compres-

sion of spinal cord, and fix the reduced atlantoaxial joints

through a single transoral approach without the need of a

posterior operation. In this stury, 21 patients were evalu-

ated and 20 did well with TARP operation. The pre-

liminary clinical result was satisfactory.

Acknowledgments We thank Associate Professor Jin Xu from

Medical English Center at Southern Medical University, Guangzhou,

China for her proofreading of this paper for language. Grant provided

by key project of the Twelfth Five-year Project of Military Medicine

(BWS11C065) and Medical Scientific Research Foundation of

Guangdong Province, China (B2012289).

Conflict of interest None.

References

1. Chandra PS, Kumar A, Chauhan A, Ansari A, Mishra NK,

Sharma BS (2013) Distraction, compression, and extension

reduction of basilar invagination and atlantoaxial dislocation: a

novel pilot technique. Neurosurgery 72(6):1040–1053

2. Klimo P Jr, Rao G, Brockmeyer D (2007) Congenital anomalies

of the cervical spine. Neurosurg Clin N Am 18(3):463–478

3. Hansen MA, da Cruz MJ, Owler BK (2008) Endoscopic trans-

nasal decompression for management of basilar invagination in

osteogenesis imperfecta. J Neurosurg Spine 9(4):354–357

4. Goel A, Pareikh S, Sharma P (2005) Atlantoaxial joint distraction

for treatment of basilar invagination secondary to rheumatoid

arthritis. Neurology India 53(2):238–240

5. Dasenbrock HH, Clarke MJ, Bydon A, Sciubba DM, Witham TF,

Gokaslan ZL, Wolinsky JP (2012) Endoscopic image-guided

transcervical odontoidectomy: outcomes of 15 patients with

basilar invagination. Neurosurgery 70(2):351–359

6. Yin Q, Ai F, Zhang K, Chang Y, Xia H, Wu Z, Quan R, Mai X,

Liu J (2005) Irreducible anterior atlantoaxial dislocation: one-

stage treatment with a transoral atlantoaxial reduction plate fix-

ation and fusion. report of 5 cases and review of the literature.

Spine (Phila Pa 1976) 30(13):e375–e381

7. Nishikawa M, Ohata K, Baba M, Terakawa Y, Hara M (2004)

Chiari I malformation associated with ventral compression and

instability: one-stage posterior decompression and fusion with a

new instrumentation technique. Neurosurgery 54(6):1430–1434

8. Wolinsky JP, Sciubba DM, Suk I, Gokaslan ZL (2007) Endo-

scopic image-guided odontoidectomy for decompression of bas-

ilar invagination via a standard anterior cervical approach.

J Neurosurg Spine 6(2):184–191

9. Yu Y, Wang X, Zhang X, Hu F, Gu Y, Xie T, Jiang X, Jiang C

(2013) Endoscopic transnasal odontoidectomy to treat basilar

invagination with congenital osseous malformations. Eur Spine J

22(5):1127–1136

10. Wu JC, Huang WC, Cheng H, Liang ML, Ho CY, Wong TT, Shih

YH, Yen YS (2008) Endoscopic transnasal transclival odontoi-

dectomy: a new approach to decompression: technical case

report. Neurosurgery 63(ons suppl 1):94–96

11. Goel A, Shah A (2008) Atlantoaxial joint distraction as a treat-

ment for basilar invagination: a report of an enperience with 11

cases. Neurol India 56(2):144–150

12. Cacciola F, Patel V, Boszczyk B (2013) Novel use of bone

cement to aid atlanto-axial distraction in the treatment of basilar

invagination: a case report and technical note. Clin Neurol

Neurosurg 115(6):787–789

13. Menezes AH, VanGilder JC, Clark CR, el-Khoury G (1985)

Odontoid upward migration in rheumatoid arthritis: an analysis of

45 patients with ‘‘cranial settling’’. J Neurosurg 63(4):500–509

14. Ai F, Yin Q, Wang Z, Xia H, Chang Y, Wu Z, Liu J (2006)

Applied anatomy of transoral atlanto-axial reduction plate inter-

nal fixation. Spine (Phila Pa 1976) 31(2):128–132

15. Xu J, Yin Q, Xia H, Wu Z, Ma X, Zhang K, Wang Z, Yang J, Ai

F, Wang J, Liu J, Mai X (2013) New clinical classification system

for atlantoaxial dislocation. Orthopedics 36(1):e95–e100

16. Ai FZ, Yin QS, Xu DC, Xia H, Wu ZH, Mai XH (2011) Transoral

atlantoaxial reduction plate internal fixation with transoral tran-

spedicular or articular mass screw of C2 for the treatment of

irreducible atlantoaxial dislocation: two case reports. Spine (Phila

Pa 1976) 36(8):E556–E562

17. Zhang K, Xu J, Wang Q, Wang G, Wu Z, Xia H, Yin QS (2012)

Treatment of dens fractures with posterior atlantoaxial disloca-

tion with transoral atlantoaxial reduction plate surgery: case

report and introduction of a novel treatment option. Spine (Phila

Pa 1976) 37(7):e451–e455

18. Dickman CA, Kalani MYS (2012) Resolution of cervical syrin-

gomyelia after transoral odontoidectomy and occipitocervical

fusion in a patient with basilar invagination and Type I Chiari

malformation. J Clin Neurosci 19:1726–1728

19. Kassam AB, Snyderman C, Gardner P, Carrau R, Spiro R (2005)

The expanded endonasal approach: a fully endoscopic transnasal

approach and resection of the odontoid process: technical case

report. Neurosurgery 57(1 Suppl):E213

20. Dickman CA, Crawford NR, Brantley AG, Sonntag VK (1995)

Biomechanical effects of transoral odontoidectomy. Neurosur-

gery 36(6):1146–1152

21. Naderi S, Pamir MN (2001) Further cranial settling of the upper

cervical spine following odontoidectomy. report of two cases.

J Neurosurg 95(2 Suppl):246–249

22. Di Lorenzo N (1992) Craniocervical junction malformation

treated by transoral approach. A survey of 25 cases with emphasis

on postoperative instability and outcome. Acta Neurochir (Wien)

118(3–4):112–116

Eur Spine J

123

23. Goel A (2004) Treatment of basilar invagination by atlantoaxial

joint distraction and direct lateral mass fixation. J Neurosurg

Spine 1(3):281–286

24. Jian FZ, Chen Z, Wrede KH, Samii M, Ling F (2010) Direct

posterior reduction and fixation for the treatment of basilar

invagination with atlantoaxial dislocation. Neurosurgery

66(4):678–687

25. Hirabayashi K, Watanabe K, Wakano K, Suzuki N, Satomi K,

Ishii Y (1983) Expansive open-door laminoplasty for cervical

spinal stenotic myelopathy. Spine (Phila Pa 1976) 8(7):693–699

Eur Spine J

123