Journal of Clinical Neuroscience (2005) 12(1), 62–65

0967-5868/$ - see front matter ª 2004 Published by Elsevier Ltd.

doi:10.1016/j.jocn.2004.03.003

Technical note

Stereotactic atlantoaxial transarticular screw fixation

RW Laherty MBBSMBBS,, RJ Kahler MBBS FRACSMBBS FRACS,, DG Walker PHD FRACSPHD FRACS,, FH Tomlinson PHD MD FRACSPHD MD FRACS

Kenneth G. Jamieson Department of Neurosurgery, Royal Brisbane Hospital, Herston, Qld., Australia

Summary Atlantoaxial stabilisation can be performed using a variety of surgical techniques. Developments in spinal instrumentation and

stereotactic technology have been incorporated into these procedures. We have recently adopted frameless stereotaxy to assist in such

operations. A retrospective study of patients treated by the authors and using frameless stereotaxy from 2001 to 2002 was performed. Each

patient underwent pre-operative fine-cut CT in the position of fixation. Using these images, screw trajectory was planned. Stereotaxis and

fluoroscopy was utilised during fixation. A post-operative CT was performed. There were nine patients. Bilateral screw placement was achieved

in eight. In the remaining case stereotactic planning predicted the single screw fixation. There were no post-operative complications. Post-

operative CT showed screw placement corresponding to the planned trajectory in all 17 screws. Stabilisation was achieved in all. Stereotactic

atlantoaxial screw fixation is an accessible, safe and accurate method for the management of C1-2 instability.

ª 2004 Published by Elsevier Ltd.

Keywords: atlantoaxial, atlas, axis, fixation, spine, stereotactic, stereotaxis, transarticular

INTRODUCTION

Atlantoaxial instability follows a variety of pathologies. It occursmost commonly in patients with rheumatoid arthritis.1;2 It lessfrequently presents as a result of trauma. Less frequent causesinclude neoplastic involvement, metabolic destruction and con-genital subluxation. Instability may also present as a complicationof previous surgical procedures.3;4 Left untreated, this unstablesegment of the axial skeleton has the potential to cause a varietyof symptoms. These range from radicular neck pain and headachesto progressive myelopathy as well as acute cord syndromes andeven sudden death.2;5

Treatment may be non-surgical or surgical. Some fracturesmay be treated non-operatively with manipulation to achievereduction followed by external immobilisation until bony healinghas effected a fusion.

Non-operative methods of immobilisation include halothoracicbracing. Less rigid supports do not provide adequate immobili-sation to maintain reduction. Whilst halothoracic bracing is rela-tively non-invasive, it does require the wearing of a cumbersomeapparatus for a significant period. The most frequent problemsrelate to pin-site infections and pin loosening with loss ofreduction.6 Wearing the apparatus may impact on social interac-tion, physical mobility and personal hygiene. Challenges tomobility increase the risk of deep venous thrombosis.7–9 There arecase reports of major intracranial pathology following falls whilstwearing a halothoracic brace.6

Most lesions will routinely require operative management inorder to achieve all required elements of fusion.10 Surgicalmethods for posterior atlantoaxial stabilisation have evolved overthe latter half of the last century and are now favoured by many,with some authors advocating prophylactic surgery for some pa-tients with degenerative atlantoaxial instability.11

Received 2 September 2003

Accepted 1 March 2004

Correspondence to: R. Laherty, Royal Brisbane Hospital Post Office, P.O. Box

62, Herston, Qld. 4029 Australia. Tel.: +61-7-36367470;

E-mail: richardlaherty@hotmail.com

62

The larger series in the literature are performed by experi-enced spinal surgeons in institutions supporting large popula-tions.1;12;13 Can the demanding technique of transarticular screwplacement described and utilised by spinal experts be modifiedthrough the use of stereotaxis to enable surgeons servicingsmaller populations with less frequent presentations to safelyand effectively treat these patients? The purpose of this study isto review our recent experience using frameless stereotaxy as asurgical adjunct.

MATERIALS AND METHODS

Patients

From July 2001 to June 2002, there were nine patients whoseatlantoaxial instability was treated using stereotactic atlantoaxialtransarticular screw fixation. This was performed using the‘StealthStation’ (Medtronic Sofamor Danek) with the UniversalCannulated Screw System (Medtronic Sofamor Danek).

Preoperative planning

Emphasis was placed on pre-operative preparation. Functionalviews were obtained using plain films, computed tomography andfluoroscopy. The position of best radiologic reduction wasachieved and fine-cut (2-mm slices) computed tomography wasperformed.

This fine-cut CT was incorporated into the image guidancesystem and the screw entry point and trajectory was planned.Following the Magerl technique,12–14 the entry site13 was3–4 mm above and 3–4 mm lateral to the inferomedial edge ofthe C2-3 facet joint. The planned trajectory was in an obliquecaudocephalad direction from the posterior aspect of the laminaof the axis. In lateral projection the trajectory was towards theanterior tubercle of the atlas. The virtual screw penetrated thearticular surface in the posteromedial quadrant of the jointsurface and entered the lateral mass of the atlas. The courseof the virtual screw was followed to appreciate its relationshipto the vertebral artery, the spinal cord and the nerve roots.

Stereotactic atlantoaxial transarticular screw fixation 63

Operative technique

Under general anaesthesia, each patient was carefully rolled proneand positioned using fluoroscopic guidance. The patient was heldrigidly in position with either a halothoracic brace secured to theoperating table or a Mayfield three-pin clamp. Once secured, alateral fluoroscopic image was obtained and compared to the‘scanogram’ of the pre-operative fine-cut CT to help confirm theposition of best reduction.

Shoulder and skin retraction was effected when necessary tofacilitate vision.

A midline incision was made from the external occipital pro-tuberance to the mid-cervical region. Bone from the occiput to C3was exposed, with careful dissection to expose all the bony con-tours in the surgical field of the spinous process, the laminae andthe lateral masses of the atlas and axis. The stereotactic referenceframe was fixed to the spinous process of the axis.

The accuracy of the guidance system is improved by selectionof operative landmarks that are recognisable on the stereotacticimage. This is aided by careful dissection to expose and appreciatethe subtleties of the bony contours. Contour anomalies make forbetter reference points than points on a smooth surface. Selectionof landmarks over a greater surface area also increases the accu-racy of the guidance system.

This allows for accurate representation of the screw in itspassage through the axis. Passage through the atlas cannot be soconfident. Optimal reduction both pre- and intra-operatively helpsto ensure that the image of the atlas on the guidance systemcorrelates with its true location.

The guidance system was used to determine where stab inci-sions were sited. A protective sheath was passed through a stabincision along the trajectory of the planned screw to the entrypoint. Guide-wire and screw placement was performed with twooperators. One surgeon was tasked with maintaining the entrypoint and trajectory of the drill guide using the image guidancesoftware. The other was able to perform the drilling with fluoro-scopic guidance.

Autologous bone was harvested from the iliac crest, morcel-lised and placed over the lateral masses after preparing the localcortical surfaces.

Post-operative care

Patients were admitted to the Intensive Care Facility for 24 h.Post-operative computed tomography was performed. A Phila-delphia collar was prescribed for 8–12 weeks. Plain radiographswere obtained at follow-up appointments.

RESULTS

There were nine patients; seven were male. The mean age was 70years (range 31–84 years). The mean follow-up period was 18months (range 13–22 months).

The youngest patient (31 years, male) had congenital sublux-ation. One patient’s instability was secondary to local involvementwith squamous cell carcinoma. He died of this disease fourmonths after fixation. Four male patients had traumatic causes fortheir instability. Three of these had a known history of rheumatoidarthritis. In one of these the rheumatoid process was thought to besignificant as his injury was sustained with minimal force – hestumbled forwards and struck his head on the fridge door. Duringhis hospitalisation he was also found to have non-Hodgkin’s B celllymphoma and died of this 14 months after stabilisation. Theforces associated with the cervical fractures of the other threepatients were enough to contribute to moderate head injury ineach. Three patients (two female and one male) were diagnosed

ª 2004 Elsevier Ltd. All rights reserved.

with rheumatoid associated subluxation. In all, rheumatoidarthritis was present in six patients.

Seventeen screws were placed. In one case, stereotacticplanning predicted single screw fixation, which was achieved.This was due to kyphoscoliosis. Satisfactory screw placementscorresponding to the plans were achieved for all screws. Eachsurgeon assessed this by comparing the screw in the post-operative CT images to the virtual screw in the image guidancesystem.

All patients received DVT prophylaxis with subcutaneousheparin, TED stockings and sequential compression devices.Six patients were mobilised on their first post-operative day.Mobilisation was delayed in the three head-injured malepatients.

There were no post-operative complications.Stabilisation was assessed using flexion–extension plain

radiographs at follow-up and was achieved in all patients.

DISCUSSION

Earlier operative techniques for achieving atlantoaxial stabilityutilised autologous interspinous or interlaminar bone graft heldin position with sublaminar wires.5;15;16 These fusion constructsrequire external immobilisation. The bone was harvested typi-cally from the iliac crest13 and less often from a rib.17 Duringplacement of sublaminar wires, the spinal cord may be in-jured.18 Sublaminar wires have also been associated with suchlate adverse consequences as wire fracture resulting in neuro-logic deficit.19

More recent techniques have attempted to limit sublaminardissection and wire placement. These include interlaminarclamps20;21 and various screw-plate22 and screw-rod systems.23

The use of the iliac crest can threaten local neurovascularelements, can be a source of ongoing pain and has been reportedto contribute to pelvic fractures.24 Artificial bone substances havebeen utilised25 to reduce the incidence of these adverse events.The use of iliac crest remains a valid option and was utilised in ourseries.

Magerl and Seemann14 described the placement of a screwacross the facet joint between the atlas and axis. It was initiallypostulated as an adjunct to the more traditional methods,decreasing the need for post-operative bracing. However, it hassince been shown to offer a more attractive biomechanical profilethan other fusion methods, particularly because of the immediateobliteration of rotational motion26;27 and its faster fusion rate.4;28

The great concern with this technique has been the threat to thevertebral arteries.29 Whilst this is a recognised event, its incidence,significance and pre-disposing factors are difficult to extract fromthe literature.1 Other concerns have related to hardware failure,errant screw placement and hypoglossal nerve damage.1;29 Thesurgery is also technically demanding; prompting some surgeonsto favour other techniques.23

Contraindications include aberrant vertebral artery anatomy,bony dimensions that will not accept a screw and poor bonequality such that screw purchase and regional immobilisation willnot be maintained.1;30 Suitable pre-operative imaging is obviouslyparamount.

A relative contraindication for atlantoaxial screw fixation isincomplete pre-operative reduction.1 Our patient with congenitalsubluxation was thought to have satisfactory, but not perfect,reduction during his pre-operative imaging. This was, however,improved at his post-operative CT. Better reduction may havebeen achieved under anaesthesia. The mechanical advantageof the bi-cortical pull of the screw was also thought to havecontributed.

Journal of Clinical Neuroscience (2005) 12(1), 62–65

64 Laherty et al.

Madawi et al.1 report one of the larger series of posterioratlantoaxial fusions using the Magerl technique. This is one ofthe few papers to address malpositioned screws and their con-sequences, including vertebral artery damage. They advocatedthe transarticular screw technique but highlighted that it wasdemanding and required a high degree of technical precision.

Stereotaxis and image guided surgery have become indis-pensable to many modern intracranial surgeons and are slowlybeing extended to spinal surgery. Welch et al.31 described theuse of image guidance to assist in treatment of complex spinalcases. Four of his 11 cases involved atlantoaxial instability andwere treated with transarticular screws. Welch agreed that theplacement of transarticular screws requires strict technical pre-cision and accuracy and that this is challenged by the limita-tions of surgical exposure in an anatomically complex region.The use of stereotaxis provided a multidimensional appreciationof the anatomic relationships. These relations could be assessedpre-operatively, increasing the efficiency and safety of surgery.

Kawaguchi et al.30 described the use of stereotaxis to aidtransarticular screw placement in two patients. In one, the pre-operative planning identified that only one screw could beplaced. This was due to a medially situated vertebral artery. Inour series there was one patient whose kyphoscoliosis preventedpassage of a screw on one side. The bony anatomy of the atlasand axis indicated that a screw could be accepted. However, theguidance software revealed that the required trajectory would betoo low relative to the ribs. Prior to surgery the operative planwas revised to include wired interlaminar bone graft on thatside.

The great benefit offered by the use of stereotaxis is in the pre-operative phase. As others have described, the complex relationalanatomy of the region can be reviewed and appreciated individ-ually. Suitability for screw placement can be assessed and where itis deemed unsuitable, an alternative method of fixation can beplanned well in advance. The software can present the anatomyand the virtual screw such that a simulation of the actual proce-dure can be modelled and practiced. This allows the surgical teamto appreciate the specifics of the individualised procedure aheadof them, affording greater efficiency at surgery. This operativemodel could also be used to help educate the patient as to thesurgical plan and the inherent risks.

These elements translate to improved risk management. Eventhough this is a relatively small series, the ability to plan, prepareand practice in advance of the operation contributes to greatersurgeon comfort and confidence during the procedure.

Performance of the procedure remains much as others havedescribed.12–14 The technique uses fluoroscopy to aid screwplacement. This necessitates a variable radiation dose to the sur-geon. The image guidance system can be used to represent thepassage of the drill or the screw as it passes through the axis. Itmust be remembered that the reference frame is attached to theaxis and therefore the relative position of the atlas may be mis-represented. For this reason fluoroscopy must continue to be used.

It may be possible to decrease radiation exposure by judicioususe of the fluoroscope. We recommend its use during placementof the K wire guide. This does not have to be continuous butscreening should be reasonably frequent, as the flexible wire maybend and skew its trajectory. Fluoroscopy during screw placementmay be even more limited. However care must be taken to avoidadvancing the K wire during subsequent drilling, tapping andscrewing.

The method used for assessing the adequacy of screw place-ment in this series may be questioned. There was no objectivemeasurement or independent review. This reflects the limitationsof a retrospective series. There is, however, only a small margin of

Journal of Clinical Neuroscience (2005) 12(1), 62–65

tolerance whereby a screw could be misplaced and not cause acomplication.

CONCLUSIONS

Stereotactic atlantoaxial transarticular fixation is a modification ofan already proven technique that offers the potential for improvedoutcomes by enabling greater pre-operative assessment andplanning. It also facilitates an individualised surgical model forpre-operative preparation and practice leading to improved sur-gical efficiency and safety.

In our experience the technique is accessible, safe and accurate.

ACKNOWLEDGEMENTS

Medtronic Sofamor Danek – Financial support for travel topresent at Broome Neurosurgical Society of Australasia meeting,2002.

REFERENCES

1. Madawi AA, Casey AT, Solanki GA, Tuite G, Veres R, Crockard HA.Radiological and anatomical evaluation of the atlantoaxial transarticular screwfixation technique. J Neurosurg 1997; 86(6): 961–968.

2. Papadopoulos SM, Dickman CA, Sonntag VK. Atlantoaxial stabilization inrheumatoid arthritis. J Neurosurg 1991; 74(1): 1–7.

3. Dickman CA, Sonntag VK. Surgical management of atlantoaxial nonunions. JNeurosurg 1995; 83(2): 248–253.

4. Grob D. Surgery in the degenerative cervical spine. Spine 1998; 23(24):2674–2683.

5. Dickman CA, Sonntag VK, Papadopoulos SM, Hadley MN. The interspinousmethod of posterior atlantoaxial arthrodesis. J Neurosurg 1991; 74(2): 190–198.

6. Dorfmuller G, Hollerhage HG. Severe intracranial injury from a fall in the haloexternal fixator. J Orthop Trauma 1992; 6(3): 366–369.

7. Cogo A, Bernardi E, Prandoni P, et al. Acquired risk factors for deep-veinthrombosis in symptomatic outpatients. Arch Intern Med 1994; 154(2):164–168.

8. Gebremedhin A, Shamebo M. Deep venous thrombosis in a university teachinghospital, Addis Ababa, Ethiopia. East Afr Med J 1998; 75(7): 432–435.

9. Samama MM. An epidemiologic study of risk factors for deep veinthrombosis in medical outpatients: the Sirius study. Arch Intern Med 2000;160(22): 3415–3420.

10. Ranawat CS, O'Leary P, Pellicci P, Tsairis P, Marchisello P, Dorr L. Cervicalspine fusion in rheumatoid arthritis. J Bone Joint Surg Am 1979; 61(7):1003–1010.

11. Grob D. Atlantoaxial immobilization in rheumatoid arthritis: a prophylacticprocedure. Eur Spine J 2000; 9(5): 404–409.

12. Grob D, Jeanneret B, Aebi M et al. Atlantoaxial fusion with transarticular screwfixation. J Bone Joint Surg (Br) 1991; 73: 972–976.

13. Haid Jr RW. C1-C2 transarticular screw fixation: technical aspects.Neurosurgery 2001; 49(1): 71–74.

14. Magerl F, Seemann P. Stable posterior fusion of the atlas and axis bytransarticular screw fixation. In: Kehr P, Weidner A (eds) Cervical Spine I.Springer, New York 1987; 322–327.

15. Brooks A, Jenkins E. Atlantoaxial arthrodesis by the wedge compressionmethod. J Bone Joint Surg (Am) 1978; 60: 279–284.

16. Gallie W. Fractures and dislocations of the cervical spine. Am J Surg 1939; 46:495–499.

17. Brockmeyer DL. A bone and cable girth-hitch technique for atlantoaxial fusionin pediatric patients. Technical note. J Neurosurg 2002; 97(3 Suppl): 400–402.

18. Fraser AB, Sen C, Casden AM, Catalano PJ, Post KD. Cervical transduralintramedullary migration of a sublaminar wire. A complication of cervicalfixation. Spine 1994; 19(4): 456–459.

19. Blacklock JB. Fracture of a sublaminar stainless steel cable in the uppercervical spine with neurological injury. Case report. J Neurosurg 1994; 81(6):932–933.

20. Holness RO, Huestis WS, Howes WJ, Langille RA. Posterior stabilization withan interlaminar clamp in cervical injuries: technical note and review of the longterm experience with the method. Neurosurgery 1984; 14(3): 318–322.

21. Moskovich R, Crockard HA. Atlantoaxial arthrodesis using interlaminarclamps. An improved technique. Spine 1992; 17(3): 261–267.

22. Goel A, Desai KI, Muzumdar DP. Atlantoaxial fixation using plate and screwmethod: a report of 160 treated patients. Neurosurgery 2002; 51(6): 1351–1356.

23. Harms J, Melcher RP. Posterior C1-C2 fusion with polyaxial screw and rodfixation. Spine 2001; 26(22): 2467–2471.

ª 2004 Elsevier Ltd. All rights reserved.

Stereotactic atlantoaxial transarticular screw fixation 65

24. Hu RW, Bohlman HH. Fracture at the iliac bone graft harvest site after fusionof the spine. Clin Orthop 1994; 309: 208–213.

25. Bryan WJ, Inglis AE, Sculco TP, Ranawat CS. Methylmethacrylatestabilization for enhancement of posterior cervical arthrodesis in rheumatoidarthritis. J Bone Joint Surg Am 1982; 64(7): 1045–1050.

26. Grob D, Crisco 3rd JJ, Panjabi MM, Wang P, Dvorak J. Biomechanicalevaluation of four different posterior atlantoaxial fixation techniques. Spine1992; 17(5): 480–490.

27. Richter M, Schmidt R, Claes L, Puhl W, Wilke HJ. Posterior atlantoaxialfixation: biomechanical in vitro comparison of six different techniques. Spine2002; 27(16): 1724–1732.

ª 2004 Elsevier Ltd. All rights reserved.

28. Coyne T, Fehlings M, Wallace M, Bernstein M, Tator C. C1-2 posteriorcervical fusion: Long term evaluation of results and efficacy. Neurosurgery1995; 37: 688–692.

29. Haid Jr RW, Subach BR, McLaughlin MR, Rodts Jr GE, Wahlig Jr JB. C1-C2transarticular screw fixation for atlantoaxial instability: a 6-year experience.Neurosurgery 2001; 49(1): 65–68.

30. Kawaguchi Y, Ishihara H, Ohmori K, Kanamori M, Kimura T. Computer-assisted Magerl's transarticular screw fixation for atlantoaxial subluxation. JOrthop Sci 2002; 7(1): 131–136.

31. Welch WC, Subach BR, Pollack IF, Jacobs GB. Frameless stereotactic guidancefor surgery of the upper cervical spine. Neurosurgery 1997; 40(5): 958–963.

Journal of Clinical Neuroscience (2005) 12(1), 62–65