The atlantoaxial articulation consists of two lateral facet joints and one central pivotal joint formed by the odontoid process between the anterior arch and the transverse ligament of the atlas (Fig. 1). Werne (1957) showed that 47° of lateral rotation occurs at the atlantoaxial region.1 Trauma can fracture the odontoid process,2-4 and inflammatory diseases like rheumatoid arthritis and ankylosing spondylitis can destroy the transverse ligament by the ingrowth of synovial pannus,5-" resulting in a dangerous atlantoaxial instability which can lead to high cervical cord compression, obstruction of vertebral arteries,10'16 or even death.10-12 This paper is intended to share our experience in treating 22 patients who needed atlantoaxial arthrodesis, in the hope of minimizing all preventable intraoperative and postoperative complications.
Materials and Methods
Our present study covers a 10 year period, 1969 to 1979. Our data were obtained from patients' medical records, microfilms and roentgenographs. Surgical indications for the 22 atlantoaxial fusions are listed in Table 1 (Fig. 2-6).
We routinely employ a modified Gallie's fusion17 after the atlantoaxial subluxation has been reduced. Usually an unicortical iliac graft is used (Fig. 7), reserving the bicortical iliac graft for patients with severe generalized osteoporosis (Fig. 8).
Fig. 1: The pivot joint formeil between the odontoid process, and anterior arch and transverse ligament of atlas. Sote the two synovial sacs interposed between the odontoid process and the transverse ligament and anterior arch of at /as.
The Unicortical "H" Graft Technique
An incision about 5" in length, extending from the external occipital protuberance to the spinous process of C3, is made along the midline of the posterior aspect of the neck. After stripping the cervical paraspinal muscles from the posterior arch of atlas, and spinous process and laminae of axis, the exposed bone surfaces are roughened with a narrow Wu's ultracurved osteotome. Next, a notch about Va" in depth is made at the superior base of the C2 spinous process with a small rongeur with a narrow beak, and a small hole is also made through the same spinous process slightly above the notch with a sharp rightangled awl. A rectangular unicortical bone graft is removed from the ilium in the usual manner. Using a rongeur with a narrow beak, four notches are cut in the four sides of the iliac graft; the top and bottom notches along the midline and the two lateral notches more proximally based. After passing a size 20 stainless steel wire loop upward under the posterior arch of atlas, the inferior notch of the iliac graft is fitted into the notch on the spinous process of axis, and the upper portion of the same graft is brought into close contact with the posterior arch of atlas. The two ends of the wire are made to engage the two lateral notches of the iliac graft and are passed through the wire loop. Tension is applied to the two ends of the wire simultaneously, and the iliac graft is tightly bound to the posterior arch of atlas by the two wire loops. The two ends of the wire are passed under tension through the hole at the base of C2 spinous process in opposite directions. They are then turned downward and tightly twisted together against the inferior base of C2 spinous process. Cancellous bone chips are packed into the crevices under the iliac graft. The incision is closed in the usual manner.
Fig. 2: Lateral views of lhe cervical spine in flexion and extension, showing ihe greatly increased atlantodental distance inflexion due to rupture of transverse ligament in this rheumatoid patient.
Fig. 3(A): luxt eral views of the cervical spine showing the greatly widened atlantodental distance in extension of this 53-year-old man. who sustained a very rare rupture of his transverse ligament in an automobile accident.
Fig. 3(B): The picture on the right was taken only a few days after atlantoaxial fusion, which had nicely reduced lhe atlantoaxial subluxation shown in (A). The picture on the left was taken six months postoperatively. It shows loss of the original reduction of atlantoaxial subluxation, but the fusion was solid and the patient was compie' ? * asymptomatic.
Fig. 4(A); Lateral cervical spine x-rays showing widening of the atlantoaxial interval in extension in a 47-year-old woman with Down's syndrome. The patient denied any history of trauma to her neck.
Fig. 4(B): The roentgenogram on the right was taken about one year after lhe first atlantoaxial fusion of our patient with Down's syndrome. It shows persistent atlantoaxial subluxation and resorption of ihe hone graft. The roentgenogram on the left shows a second attempt to achieve a solid atlantoaxial fusion with a new iliac graft. The second fusion was successful.
The Bicortical "H" Graft Technique
Due to the thickness of the bicortical graft, a separate size 20 stainless steel wire has to be passed through the base of the spinous process of the axis. Its two ends are twisted together before joining it with the first wire, which has been made to bind the upper portion of the iliac graft to the posterior arch of atlas in a manner identical to the unicortical "H" graft technique.
Skull traction with tongs is used for about a week. Thereafter, depending on how reliable the patient is, either a Minerva cast or a rigid cervical collar is worn for three months, followed by a soft cervical collar for two to three additional months. We routinely use only a soft cervical collar for rheumatoid patients with generalized arthritic involvement of all major joints because they usually do not tolerate Minerva cast or rigid cervical collar well.
Fig. 5: The two roentgenograms on the left show a slightly displaced odontoid fracture, and the roentgenogram on the right was taken soon after atlantoaxial fusion, which improved the alignment of the odontoid fracture.
The pertinent clinical data of our 22 patients are summarized in Table 2.
Atlantoaxial instability can be caused by fracture of odontoid process,2-4 rupture of transverse ligament,5*1"11'"''1 destruction of the two lateral atlantoaxial joints,19-25 absence of odontoid process,26-36 and os odontoidium.18 Several studies have shown that the maximal atlantodental distance in adults is 2 to 4 mm.8'"'37 Ball (1963) demonstrated that section of transverse ligament will produce an atlantoaxial slip of 4 to 5 mm. 1^ However, if the joint capsules and ligaments of the two lateral atlantoaxial joints are also cut, the slip may increase to 10 mm. The patients in our series had an average slip of 9.2 mm, which strongly suggests that both the transverse ligament rupture and destruction of the two lateral atlantoaxial joints are responsible for the slip.
We fully agree with several authors that atlantoaxial arthrodesis is a safe and effective procedure in treating diseases of the atlantoaxial region.2'22'23'38'39 The prevention of severe morbidity and mortality, reversal of neurological deficits, relief of pain, and significant reduction of hospital stay are the great benefits derived from atlantoaxial fusion. When primary bone tumors and metastasis involve the same region, we prefer to fuse from occiput to C3 or C4 because these tumors can cause extensive destruction of the atlas and axis. Stabilization of the unstable atlantoaxial region appeared to be very desirable when our rheumatoid patients subsequently had to undergo major elbow, hand, hip and knee reconstructive operations, because any attempt to intubate a patient with atlantoaxial instability is hazardous and potentially life-threatening.
Twenty-two patients (13 males and 9 females) with an average age of 49.5 years underwent atlantoaxial fusion at Henry Ford Hospital from 1969 to 1979. The indications for their surgery included 15 transverse ligament ruptures, six odontoid fractures and one absence of odontoid process in a patient with Morquio's disease. Their most common preoperative complaint was neck pain, but occipital headaches, neck stiffness, and occasional numbness and weakness of upper or lower extremity could also be present. Examination often revealed neck muscle spasm, limited range of motion of cervical spine, and upper cervical tenderness. Three patients (14%) had demonstrable neurological deficits.
Fig. 6(A): Lateral views of the cervical spine inflexion and extension, showing excessive motion at the atlantoaxial region of this /0-year-old boy with Morquio's disease. Note the absence of odontoid process and the tongue-shaped vertebral bodies, characteristic of Morquio's disease.
Fig. 6(B): Lateral views of cervical spine inflexion and extension of the same patient with Morquio's disease, after atlantoaxial fusion, showing elimination of the excessive motion at the atlantoaxial iunction.
Fig. 7(A): The correct method by which the ?-shaped iliac graft can be locked into the notch at the superior aspect of the spinous process of axis, and simultaneously wired to the posterior arch of atlas.
Fig. 7(B): The iliac graft has been tightly fitted into lhe notch in the axis and securely wired to lhe posterior arch of alias. The two free ends of lhe wire have been passed through the hole in the spinous process of axis and twisted on each cither against the inferior aspect of the same spinous process.
A modified Gallie's fusion using a H-shaped unicortical iliac graft was routinely employed. In addition, a bicortical iliac graft was used for patients with weak and osteoporotic iliac bone. Routine postoperative management included the use of a Minerva cast or a rigid cervical collar for three months, followed by a soft cervical collar for another two to three months. All of our patients with neurological deficits made a significant recovery from their spinal cord compression. Postoperative complications included three superficial wound infections, three cases of pressure sores, one nonunion and one loss of reduction. We feel that atlantoaxial arthrodesis is a safe and effective means in treating various diseases of the atlantoaxial region. When primary or metastatic tumors involve the same regions, we prefer to include the occiput and the third and fourth vertebrae in our fusion because these tumors can cause extensive destruction in the atlantoaxial region.
Fig. H(A): The approximate location from which the bicortical iliac graft is obtained.
Fig. 8(B): This schematic drawing purposely keeps the atlas apart from the axis in order to show the correct way to wire the upper portion of the H-shaped bicortical Iliac graft to the posterior arch of atlas. Note that a second wire has been passed through the spinous process of axis, and its two ends have been twisted on each other.
Fig. 8(C): The completed atlantoaxial fusion with a H-s hoped hicoriieäl iliac graft, which has been intimately wired to the posterior arch of atlas and the spinous process of axis.
1. Werne S-. Studies in spontaneous atlas dislocation. A(\n Orthop Stand ISuppl] 1957; 23.
2. Amyes EW, Anderson FM; Fracture of the odontoid process. AMA Arch Surg 1956; 72:377.
3. Owen -Smith MS: Fractures of the odontoid process of the axis. Proc R Sot Med 1968: 61:41.
4. Blockey HJ, Purser DW: Fractures of the odontoid process of the axis. / Bant joint Surg 1956; 38B:794.
5. Meijers KAE, Voss SFC. Heerhávan. et al: Radiological changes in the cervical spine in ankylosing spondylitis. Ann Rheum Dis 1968; 27:333.
6. Martel WW, Page JW: Cervical vertebral erosions and subluxations in rheumatoid arthritis and ankylosing spondylitis. Arthritis Rheum 1960; 3:546.
7. Crellin RQ, Mac Cabe JJ. Hamilton EBD: Severe subluxation of the cervical spine in rheumatoid arthritis. / Bane joint Surg 1971; 52B:244-251.
8. Sharp J, Purser DW: Spontaneous atlantoaxial dislocation in ankylosing spondylitis and rheumatoid arthritis. Ann Rheum Dis 1961; 20:47-77.
9. Mathews JA: Atlantoaxial subluxation in rheumatoid arthritis. Ann Rheum Dis 1969; 28:260-266.
10. Webb FWS, Hickman JA, Brew D St J: Death from vertebral artery thrombosis in rheumatoid arthritis. Br Med } 1968; 2:537.
11. Martel WW. Abell MR: Fatal atlantoaxial luxation in rheumatoid arthritis. Arthritis Rheum 1963; 6:224.
12. Davis FW, Markley HE: Rheumatoid arthritis with death from medullary compression. Ann Intern Med 1951; 35:451.
13. Meijers KAE, Van Beusekom GT, Luyendijk W, et al: Dislocation of the cervical spine with cord compression in rheumatoid arthritis. / Bone joint Surg 1974; 56B:668-680.
14. Pratt-Thomas HR, Berger KE: Cerebellar and spinal injuries after chiropractic manipulation. JAMA 1947; 133:600.
15. Rowland LP, Shapiro JH, Jacobson HC: Neurological syndromes associated with congenital absence of the odontoid process. AMA Arch Neurol 1958; 80:286-291.
16. Ford FR: Syncope, vertigo and disturbances of vision resulting from intermittent obstruction of the vertebral arteries due to defect in the odontoid process and excessive mobility of the second cervical vertebra. Bulletin of the Johns Hopkins Hospital 1952; 91:168.
17. Wu KK: Surgical techniques for arthrodesis of two to four adjacent spinal vertebrae throughout the entire spinal column. Henry Ford Hosp Med j 1975; 23:39-50.
18. Wollin DG: The os odontoideum. / Bone joint Surg 1963; 45A:1459.
19. Ball J: The articular pathology of rheumatoid arthritis. International Study Center for Rheumatic Disease, Amsterdam U.S.R.A.), Symposium, 25, 1963.
20. Bland IH: Rheumatoid arthritis of the cervical spine. / Rheum 1974; 1:319-341.
21. Conlon PW; lsadale IC, Ròse BS: Rheumatoid arthritis of the cervical spine. An analysis of 333 cases. Ann Rheum Dis 1966; 25:120126.
22. Cregan ICF: Internal fixation of the unstable rheumatoid cervical spine. Ann Rheum Dis 1966; 25:242-252.
23. Forsyth HF, Alexander E Jr, Davis C Jr, et al: The advantages of early spine fusion in the treatment of fracture-dislocation of the cervical spine. / Berne inini Surg 1959; 41A:17.
24. Sharp J, Purser DW, Lawrence )S: Rheumatoid arthritis of the cervical spine in adult. Ami Rheum Dis 1958; 17:303-313.
25. Robinson RA: The diagnosis and treatment of acute and degenerative lesions of the cervical spine. Report on proceedings of the Boston Orthopedic Club. / Bone joint Surg 1957; 39A:983.
26. Fullenlove FM: Congenital absence of the odontoid process. Report of a case. Radiologi/ 1954; 63:72-73.
27. Giannestras NJ, Mayf ield FH, Provenicio FP, et al: Congenital absence of the odontoid process. / Bone joint Surg 1<>64; 46A:838.
28. Cellman EL: Congenital absence of the odontoid process of the axis. Report of a case. / Boue joint Surg 1<»59; 41A:345-348.
29. Ivie 1 McK: Congenital absence of odontoid process. Radiology 1046; 46:268-269.
30. Miyakawa G: Congenital absence of odontoid process. A case report. / Bone joint Surg 1952; 34A:671-677.
31. Raskin N: Congenital absence of the odontoid process. Dis Nerv System 1953; 14:236-238.
32. Roberts SM: Congenital absence of the odontoid process resulting in dislocation of the atlas on the axis, j Bone joint Surg 1933; 15:988-989.
33. Scannell RC: Congenital absence of the odontoid process. A case report. / Bone joint Surg 1945; 27:714-715.
34. Schultz EH Jr, Levy RW, Russo PE: A genesis of the odontoid process. Radiology 1956; 67:102-105.
35. Stiefel DM: Congenital absence of the odontoid process. Report of a case. / Bone inini Surg 1950; 32A:946-947.
36. Weiler HG: Congenital absence of the odontoid process of the axis with atlantoaxial dislocation. / Bone Joint Surg 1942; 24:161165.
37. Hinck VC, Hopkins CE: Measurement of the atlantodental interval in the adult. Am j Roe tit 1960; 84:945-951.
38. McGraw RW, Rusch RM: Atlantoaxial arthrodesis. / Bone joint Surg 1973; 55B-.482-489.
39. Simmons EH, Fielding JW: Atlantoaxial arthrodesis. / Bow joint Surg 1967; 49 A: 1022.