Open Reduction of Traumatic Atlanto-Axial Rotatory Dislocation with Use of the Extreme Lateral Approach. A Report of Two Cases*


Journal of Bone & Joint Surgery - American Volume:
Author Information

†Department of Surgical Neurology, The National Hospital for Neurology and Neurosurgery, Queen Square, London WC1N 3BG, United Kingdom.

Article Outline

Irreducible rotatory injuries at the atlanto-axial level in children may result in torticollis and facial asymmetry4,10,22,25,35. The terminology associated with traumatic rotatory injuries at this level is somewhat confusing, and a clear distinction should be made between rotatory subluxation and rotatory dislocation, as the mechanisms of injury and the optimum management may differ. In this report, we use the term atlanto-axial rotatory dislocation to define a complete and persistent displacement of the adjacent articular surfaces at this level. Ideally, early closed manipulative reduction with use of distraction and derotation with spinal cord monitoring, followed by external bracing, restores both normal anatomical relationships and mobility in many patients. In some patients, however, the maneuver fails, and often the joint fuses spontaneously in malalignment. The exact cause of failure to achieve reduction is unclear.

The purpose of this report is to describe our operative approach and findings in two patients who had atlanto-axial rotatory dislocation. These findings provide new information on the pathomechanics of irreducible atlanto-axial rotatory dislocation that distinguish it from rotatory subluxation. Furthermore, this approach allows for the correction of the deformity. The technique that we describe could potentially be associated with serious problems, such as damage to the neuraxis and vertebral artery, but it is the only method, to our knowledge, that allows restoration of normal anatomical relationships. The operation should not be performed by a surgeon who only occasionally operates on the cervical spine.

Back to Top | Article Outline

Case Reports

CASE 1. A ten-year-old girl had a biopsy of a lymph node in the posterior triangle on the right side of the neck under general anesthesia in November 1990. She awoke with severe neck pain accompanied by dysesthesia over the occiput and neck and had torticollis; she had never had a previous episode of torticollis. Apart from the sensory disturbance, the neurological status was intact. The dysesthesia resolved within forty-eight hours, but the torticollis persisted. Radiographs and computed tomography scans at the time demonstrated atlanto-axial rotatory dislocation and a partial occipito-atlantal dislocation (Fig. 1). Three months later, a manipulation of the cervical spine was done under general anesthesia, followed by application of traction, but there was no improvement. A laminectomy was performed at the first and second cervical levels, again with no clinical or radiographic improvement. The patient was referred to our department for treatment of the deformity. The referring surgeons expressed particular concern about the potential instability of the occipito-atlanto-axial complex after the failed laminectomy and about the possibility of facial asymmetry developing.

With the patient awake, intubation was done with a fiberoptic laryngoscope. The patient then was anesthetized, and traction was applied with skull tongs. Somatosensory evoked potentials were monitored throughout the procedure, and no changes were noted. The patient was placed in the true lateral position (left side down) for the extreme lateral approach to the joint on the right side. The vertebral artery was mobilized, and the posterior arch of the atlas was found lying on the articular surface of the lateral mass of the axis. The dislocated joint surfaces were identified, and a large amount of dense fibrous tissue was seen interposed between the dislocated joints. Rotatory manipulation of the head on the neck did not achieve reduction. The joint surfaces then were denuded of fibrous tissue and articular cartilage. A small osteotome was driven medially across the superior surface of the axis, as far as the base of the odontoid process, to ensure that there was no osseous union with the atlas. This wound then was closed, and the patient was turned into the true lateral position with the left side up, for exploration of the left joint. On this side, because of the rotation, the vertebral artery was lying superficially, so extreme care was needed in the early stages of the dissection. The artery was mobilized from the transverse foramina of the first and second cervical vertebrae. The atlanto-axial joint on this side could not be identified because of osseous union, so a high-speed air-drill was used to disrupt the ankylosis. Additional extension and derotation failed to move the joint, and it became apparent that there was union of the odontoid process, the anterior arch of the first cervical vertebra, and the medial aspect of the left lateral mass of the atlas. Therefore, with use of a three-millimeter-wide osteotome, the base of the odontoid was divided, completely disconnecting the first and second cervical vertebrae anterolaterally. The incision was closed, and the patient was returned to the full lateral position with the right side up. The original wound was reopened, and, with a combination of distraction, derotation, and extension, the dislocation was reduced partially. A small wedge-shaped bone graft was inserted between the lateral masses of the first and second cervical vertebrae. The incision was closed, and the patient was fitted with a graphite halo thoracic jacket. Postoperatively, the neurological status was intact and the position of the head was good. The halo was removed eight weeks postoperatively. At the two-year follow-up examination, the patient had full flexion and extension of the neck, but lateral rotation was limited to 60 degrees in both directions. The neurological status remained intact.

CASE 2. A four-year-old girl was involved in a motor-vehicle accident in which she had been thrown from the vehicle, hitting the right side of the head; there was no record of loss of consciousness or neurological abnormalities. When seen at a local hospital, she had severe neck pain and torticollis but no abnormal neurological signs (Fig. 2-A). Plain radiographs of the cervical spine revealed an atlanto-axial rotatory injury. After several weeks of treatment with analgesics and a collar, there was no change clinically, and repeat radiographs demonstrated persistent atlanto-axial rotatory dislocation. The patient was admitted to another hospital, where several weeks of continuous traction failed to reduce the rotatory dislocation. Six months after the accident, she was admitted to our unit; radiographs made at this time confirmed the persistence of the rotatory dislocation. We elected to treat the patient operatively.

As the right atlanto-axial joint was dislocated, a right-sided lateral approach was used. The second cervical nerve root was identified and, immediately ventral to it, the vertebral artery was noted. The artery then was mobilized out of the transverse foramen with use of a high-speed air-drill and a one-millimeter Kerrison upcut rongeur. The joint between the dislocated lateral masses of the first and second cervical vertebrae was exposed, and abundant tough fibrous tissue was found to be trapped between the two joint surfaces; there also was a frayed segment of the transverse ligament of the atlas with a fragment of bone avulsed from its insertion. The articular cartilage on the joint surfaces was denuded with a high-speed air-drill and then, with a combination of distraction, derotation, and extension, the joint surfaces were completely relocated. The wound was closed, and the child was fitted with a graphite halo thoracic jacket. Postoperative computed tomograms demonstrated relocation of the joint. The halo was removed eight weeks postoperatively, after correction of the deformity had been confirmed on radiographs. Two years later, the patient had no evidence of torticollis (Figs. 2-B, 2-C, and 2-D), had 90 degrees of lateral rotation to the right and 70 degrees to the left, and had no neurological sequelae.

Back to Top | Article Outline

Operative Anatomy and Approach

We prefer to place the patient in the true lateral position; the visual perspective of the surgeon can be altered medially or laterally with use of the lateral tilt facility of the operating table. An important surface landmark is the tip of the transverse process of the atlas, which, in the normal adult, can be palpated approximately one and a half centimeters caudad to and one and a half centimeters anterior to the tip of the mastoid process. To gain access to the cephalad three cervical vertebrae, a ten-centimeter incision is made inferiorly from the mastoid process along the posterior margin of the posterior triangle. The approach utilizes the well defined fat planes that lie deep to the posterior fibers of the sternocleidomastoid muscle and extend caudally to the vertebral artery (Fig. 3). The prevertebral fascia, which covers the floor of the posterior triangle, is divided, and the dissection proceeds through the fat plane that lies posteromedially to the levator scapulae and the splenius cervicis muscles. The posterior fibers of the latter muscle may need to be divided to gain access to the vertical section of the vertebral artery as it passes between the axis and atlas. In this region, the vertical segment of the vertebral artery is crossed horizontally and superficially by the posterior ramus of the second cervical nerve root. Identification of this nerve root is the key to exposure of this area (Fig. 4). The inferior oblique, splenius cervicis, and levator scapulae muscles all have attachments to the tip of the transverse process of the atlas and are detached if the vertebral artery is to be mobilized from the transverse foramen of the first cervical vertebra. The posterior belly of the digastric muscle may have to be retracted superiorly and anteriorly, as it overlies the transverse process of the atlas.

Back to Top | Article Outline


The cephalad two cervical vertebrae and the associated ligaments are regarded as part of the craniovertebral complex and have distinctive anatomical and biomechanical properties5,6,13,20,32,34. In children, approximately 70 per cent of all osseous injuries to the cervical spine occur between the first and third cervical vertebrae16. The factors that contribute to these injuries include ligamentous laxity, poorly developed cervical muscles, incompletely developed vertebrae that are wedge-shaped, horizontally oriented facet joints, and a relatively large head14,26.

The biomechanics of the occipito-atlanto-axial area are complex. Approximately 55 per cent of the entire rotation of the cervical spine takes place at the atlanto-axial joint5, axial rotation is coupled with lateral bending to the opposite side23, and the alar and capsular (first and second cervical) ligaments play important roles in limiting both movements under normal conditions. If excessive axial torque (as occurs during trauma, for example) is applied to this joint complex and the resultant angular rotation exceeds 63 degrees3,13,21, then there may be rupture of the alar and capsular ligaments and dislocation of the first and second cervical facet joint13. As a result of the dislocation, the diameter of the spinal canal may be reduced to ten millimeters or less, causing compression of the spinal cord3,21. Nevertheless, injury to the spinal cord in survivors is rare2,14,32. The main long-term problems arise from loss of mobility and tilting of the head or torticollis, which can result in facial asymmetry3,10.

Early recognition of atlanto-axial rotatory dislocation and rotatory subluxation is essential in order for closed reduction to be successful10. An attempt at closed reduction was delayed by months and was not effective in either of our patients. Closed reduction for both atlanto-axial rotatory dislocation and atlanto-axial rotatory subluxation is successful only when the problem has been diagnosed and treated immediately after the traumatic event, and only then will posture and mobility of the head9,17 return to normal.

To our knowledge, no report in the literature has described the pathological anatomy at the atlanto-axial joint in either chronic atlanto-axial rotatory dislocation or rotatory subluxation. A number of hypotheses concerning the mechanism of fixation have been made over the years; they include damage to the synovial membrane3 and rupture of the transverse ligament of the atlas11. In our two patients, abundant fibrous tissue and a segment of the transverse ligament of the atlas were found between the joint surfaces of the lateral masses of the first and second cervical vertebrae, and there was cross union of the atlas and axis; these findings have not been previously described, to our knowledge. In addition, the capsular ligaments on the side of the dislocation were completely disrupted in both patients.

With the exception of the report, by Schmidek et al.27, of an open reduction for atlanto-axial rotatory dislocation, most authors have accepted arthrodesis in the displaced position without inspection of the pathological anatomy of the atlanto-axial joint as treatment for this condition10,24,31. Although this mode of treatment minimizes subsequent neuraxial compression, it does not prevent facial asymmetry and permanent torticollis. The operative approach that we have chosen realigns the vertebrae and restores the head to its natural position11. The extreme lateral approach allows thorough inspection of the atlanto-axial joint and control of the vertebral artery. This approach was used to explore the vertebral artery almost eighty years ago7,19. In 1957, Henry15 gave a concise account of the lateral approach for exposure of the “top segment of the second stage” (the vertical segment between the first and second cervical vertebrae) of the vertebral artery. Extensions and variations of this lateral approach have been used to gain access to the foramen magnum and the cephalad part of the cervical spine8,12,18,28-30,33. To our knowledge, this approach has not been used in the operative treatment of atlanto-axial rotatory dislocation or rotatory subluxation. The structure that is most in jeopardy of damage is the second cervical nerve as it emerges from its dural sheath. If necessary, the nerve can be mobilized and gently retracted or cut, with resultant loss of sensation behind the ear and no noticeable wasting of muscle. Potential difficulties associated with this approach include intraoperative hemorrhage from either the venous plexus on the vertebral artery or the large extradural veins; this risk can be minimized with subperiosteal mobilization of the artery and related veins and by tilting up the head of the operating table. A detailed knowledge of the normal anatomy of the area obviously is essential; this can be obtained by dissection of a cadaver and learning in a workshop environment. Because of the complexity of the operation, we believe that it should be performed by someone who operates regularly on the cephalad part of the cervical spine.

Important lessons are to be learned from our experience. As with other injuries or fractures of a joint, failure of early closed reduction probably indicates that either soft tissue or a fragment of bone is interposed between the joint surfaces1. Irreducible rotatory injuries at the first and second cervical joint should not be treated differently from other dislocations. If early reduction is impossible, prolonged traction is not the answer. If computed tomography or magnetic resonance imaging suggests that there is tissue in the joint, this is more evidence that open reduction is probably necessary. We suggest that, if closed manipulation fails, early open exploration of the joint allows the best chance of reduction and maintenance of normal posture and mobility. When radiographic investigations indicate the likelihood of cross union of the atlas and axis and the decision is made to attempt to correct the deformity, posterior arthrodesis alone is not effective. Even with extensive bilateral mobilization of the atlanto-axial joints, relocation may be extremely difficult. The use of an osteotome to divide the base of the odontoid process or to open the atlanto-axial joints can potentially cause cross union of the atlas and axis or fusion of the atlanto-axial joints. Although we accept this as a potential problem, our justification of the procedure is that the head is placed in the correct anatomical alignment, which prevents the development of facial asymmetry.

*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

Investigation performed at the Department of Surgical Neurology, The National Hospital for Neurology and Neurosurgery, London

1. Apley, A. G., and Solomon, L.: Apley's System of Orthopaedics and Fractures. Ed. 6, p. 338. London, Butterworth Scientific, 1982.
2. Corner, E. M.: Rotatory dislocations of the atlas. Ann. Surg., 45: 9-26, 1907.
3. Coutts, M. B.: Atlanto-epistropheal subluxations. Arch. Surg., 29: 297-311, 1934.
4. Coventry, M. B., and |and |Harris, L. E.: Congenital muscular torticollis in infancy. Some observations regarding treatment. J. Bone and Joint Surg., 41-A: 815-822, July 1959.
5. Crisco, J. J., III; Panjabi, M. M.; and |and |Dvorak, J.: A model of the alar ligaments of the upper cervical spine in axial rotation. J. Biomech., 24: 607-614, 1991.
6. Crisco, J. J., III; Oda, T.; Panjabi, M. M.; Bueff, H. U.; Dvorak, J.; and |and |Grob, D.: Transections of the C1-C2 joint capsular ligaments in the cadaveric spine. Spine, 16(10S): 474-S479, 1991.
7. Drüner, L.: Über die unterbindung der Arteria vertebralis. Zentralbl. Chir., 44: 670-673, 1917.
8. du Toit, G.: Lateral atlanto-axial arthrodesis. A screw fixation technique. South African J. Surg., 14: 9-12, 1976.
9. El-Khoury, G. Y.; Clark, C. R.; and |and |Gravett, A. W.: Acute traumatic rotatory atlanto-axial dislocation in children. A report of three cases. J. Bone and Joint Surg., 66-A: 774-777, June 1984.
10. Fielding, J. W., and |and |Hawkins, R. J.: Atlanto-axial rotatory fixation. (Fixed rotatory subluxation of the atlanto-axial joint.). J. Bone and Joint Surg., 59-A: 37-44, Jan. 1977.
11. Fiorani-Gallotta, G., and |and |Luzzatti, G.: Sublussazione laterale e sublussazione rotatorie dell'atlante. Arch. Ortop., 70: 467-484, 1957.
12. George, B.; Dematons, C.; and |and |Cophignon, J.: Lateral approach to the anterior portion of the foramen magnum. Application to surgical removal of 14 benign tumors: technical note. Surg. Neurol., 29: 484-490, 1988.
13. Goel, V. K.; Winterbottom, J. M.; Schulte, K. R.; Chang, H.; Gilbertson, L. G.; Pudgil, A. G.; and |and |Gwon, J. K.: Ligamentous laxity across C0-C1-C2 complex. Axial torque-rotation characteristics until failure. Spine, 15: 990-996, 1990.
14. Hadley, M. N.; Zabramski, J. M.; Browner, C. M.; Rekate, H.; and |and |Sonntag, V. K.: Pediatric spinal trauma. Review of 122 cases of spinal cord and vertebral column injuries. J. Neurosurg., 68: 18-24, 1988.
15. Henry, A. K.: Extensile Exposure, pp. 58-66. Baltimore, Williams and Wilkins, 1957.
16. Hudgins, P. A., and |and |Hudgins, R. J.: Radiology of cervical spine trauma. Clin. Neurosurg., 37: 571-595, 1991.
17. Johnson, D. P., and |and |Fergusson, C. M.: Early diagnosis of atlanto-axial rotatory fixation. J. Bone and Joint Surg., 68-B(5): 698-701, 1986.
18. Kratimenos, G. P., and |and |Crockard, H. A.: The far lateral approach for ventrally placed foramen magnum and upper cervical spine tumours. British J. Neurosurg., 7: 129-140, 1993.
19. Küttner, H.: Die Verletzungen und traumatischen Aneurysmen der Vertebralgefässe am Halse und ihre operative Behandlung. Beitr. klin. Chir., 108: 1-60, 1917.
20. Maiman, D. J., and |and |Yoganandan, N.: Biomechanics of cervical spine trauma. Clin. Neurosurg., 37: 543-570, 1991.
21. Mazzara, J. T., and |and |Fielding, J. W.: Effect of C1-C2 rotation on canal size. Clin. Orthop., 237: 115-119, 1988.
22. Ono, K.; Yonenobu, K.; Fuji, T.; and |and |Okada, K.: Atlantoaxial rotatory fixation. Radiographic study of its mechanism. Spine, 10: 602-608, 1985.
23. Panjabi, M.; Dvorak, J.; Crisco, J. J., III; Oda, T.; Hilibrand, A.; and |and |Grob, D.: Flexion, extension, and lateral bending of the upper cervical spine in response to alar ligament transections. J. Spine Disord., 4: 157-167, 1991.
24. Richaud, J.: New technique for posterior intra-articular arthrodesis using a single block graft in cases of atlantoaxial instability. Case report of unstable atlantoaxial rotatory luxation in a 10-year-old girl. Childs Nerv. Syst., 6: 412-415, 1990.
25. Rinaldi, I.; Mullins, W. J., Jr.; Delaney, W. F.; Fitzer, P. M.; and |and |Tornberg, D. N.: Computerized tomographic demonstration of rotational atlanto-axial fixation. Case report. J. Neurosurg., 50: 115-119, 1979.
26. Ruge, J. R.; Sinson, G. P.; McLone, D. G.; and |and |Cerullo, L. J.: Pediatric spinal injury: the very young. J. Neurosurg., 68: 25-30, 1988.
27. Schmidek, H. H.; Smith, D. A.; Sofferman, R. A.; and |and |Gomes, F. B.: Transoral unilateral facetectomy in the management of unilateral anterior rotatory atlantoaxial fracture/dislocation: a case report. Neurosurgery, 18: 645-652, 1986.
28. Sen, C. N., and |and |Sekhar, L. N.: An extreme lateral approach to intradural lesions of the cervical spine and foramen magnum. Neurosurgery, 27: 197-204, 1990.
29. Simmons, E. H., and |and |du Toit, G., Jr.: Lateral atlantoaxial arthrodesis. Orthop. Clin. North America, 9: 1101-1114, 1978.
30. Spetzler, R. F., and |and |Grahm, T. W.: The far-lateral approach to the inferior clivus and the upper cervical region. Technical note. BNI Quart., 6: 35-38, 1990.
31. Turoczy, L.; Kenez, J.; and |and |Veres, R.: Fixed traumatic rotatory atlantic dislocation through interposed bony fragments. Neuro-Orthop., 11: 125-131, 1991.
32. VanGilder, J. C.; Menezes, A. H.; and Dolan, K. D.: The Craniovertebral Junction and Its Abnormalities, pp. 195-215. Mount Kisco, New York, Futura, 1987.
33. Verbiest, H.: A lateral approach to the cervical spine: technique and indications. J. Neurosurg., 28: 191-203, 1968.
34. White, A. A., III, and |and |Panjabi, M. M.: The clinical biomechanics of the occipitoatlantoaxial complex. Orthop. Clin. North America, 9: 867-878, 1978.
35. Wortzman, G., and |and |Dewar, F. P.: Rotatory fixation of the atlantoaxial joint: rotational atlantoaxial subluxation. Radiology, 90: 479-487, 1968.
Copyright 1996 by The Journal of Bone and Joint Surgery, Incorporated