Halifax interlaminar clamp for posterior cervical fusion: a long-term follow-up review

Transcription

1 J Neurosurg 78: , 1993 Halifax interlaminar clamp for posterior cervical fusion: a long-term follow-up review E. FRANCOIS ALDRICH, M.D., M.MEo., F.C.S., PE'rER B. WEnER, M.D., AND WAYNE N. CROW, M.D. Division of Neurosurgery and Department of Radiology, University of Texas Medical Branch, Galveston, Texas u- Fifty consecutive patients requiring posterior cervical fusion for various pathologies were treated with Halifax interlaminar clamps for internal spinal fixation. Fusion involved the CI-2 level in 17 cases, the C1-3 level in one, and the lower cervical area (C2-7) in 32. No patient was lost to follow-up review, which varied from 6 to 40 months (average 21 months). Fusion failed in five patients, three at the C1-2 level, one at the C1-3 level, and one at the C2-3 level. Screw loosening was the cause of failure in four patients, and in one the arch of C-I fractured. No other complications occurred. Because of the lack of complications, avoidance of the hazards of sublaminar instrumentation, and an excellent fusion rate, this technique is highly recommended for posterior cervical fusion in the lower cervical spine. Atlantoaxial arthrodesis was achieved in only 14 (82%) of 17 patients, however, which might be due to the higher mobility at this multiaxial level. Improved results in this region may be possible by using a new modified interlaminar clamp, by performing adequate bone fusions, and by postoperative external halo immobilization in high-risk patients. K~v WORDS ~ clamp 9 spinal fusion 9 cervical spine stabilization T HE Halifax interlaminar clamp system has become widely accepted instrumentation for posterior cervical fusion because of its ease of use, safety, magnetic resonance (MR) imaging compatibility, and especially because it avoids the hazards of sublaminar wire placement. In our previous experience with 21 patients undergoing posterior cervical fusion with Halifax interlaminar clamps, ~ we reported no complications or mechanical failures occurring after follow-up periods ranging from 1 to 18 months (average 9.2 months). However, with increasing experience and longer follow-up monitoring, it has become clear that the Halifax clamp has an inherent potential for screw loosening in the atlantoaxial region, and the results are therefore not as successful as in the lower cervical area. This report describes our surgical results in 50 consecutively managed patients (including the 21 previously reported cases) treated with Halifax interlaminar clamps for posterior internal fixation in combination with autogenous iliac bone grafting for posterior cervical spine stabilization. Clinical Material and Methods Patient Population This series included 50 consecutively treated patients undergoing posterior cervical spine stabilization proce- dures at the Neurosurgery Service of the University of Texas Medical Branch. This series covers all patients operated on for posterior cervical spine stabilization at this institution between December, 1989, and March, The causes and levels of instability are summarized in Table 1. Presenting symptoms and signs were typical for acute cervical spine disease, and included neck pain, radiculopathy, myelopathy, quadriparesis, quadriplegia, occipital neuralgia, Lhermitte's sign, and Brown-Srquard syndrome. Radiographic evidence of posterior instability was demonstrated preoperatively in all cases but one, in which instability followed bilevel facetectomies in a previously failed anterior cervical fusion. These studies included plain cervical spine x-ray films, cervical spine tomography, myelography with computerized tomography (CT) in flexion and extension, flexion-extension digital video fluoroscopy, and MR imaging. Surgical Technique Awake fiberoptic intubation and general anesthesia were used in all cases. Unless previously immobilized in a halo vest, all patients were placed in skeletal cervical traction using Gardner-Wells tongs. Lateral cervical spine x-ray or video fluoroscopy imaging was obtained throughout the procedure to assure adequate reduction and cervical alignment. A posterior midline approach 702 J. Neurosurg. / Volume 78/May, 1993

2 Long-term review of cervical fusion by Halifax clamp TABLE 1 Summary of procedures in 50 patients with posterior cervical spine fi~sion Level of Instability & Etiology No. of Cases CI-2 arthrodesis 17 trauma 10 rheumatoid arthritis 2 Down's syndrome 3 os odontoideum 1 infection 1 C1-3 arthrodesis 1 rheumatoid arthritis 1 C2-7 arthrodesis 32 trauma 30 tumor 1 degeneration 1 was used to gain access to the posterior elements of the unstable and adjacent levels. In the lower cervical area, the upper border of the superior lamina and the lower border of the inferior lamina were freed of ligamentum flavum attachments medially, and small rongeurs and curettes were used to prepare the clamp site to allow adequate clamp positioning. The laminae, spinous processes, and posterior surface of the facets in the region of the fusion were denuded of periosteum, and high-speed drills and curettes were used to roughen the bone until it bled freely. Bilateral clamps were placed in all cases except when the posterior ring was fractured unilaterally. Clamps designed for the lower cervical area were selected, with the correct screw length. It is of the utmost importance to tighten these screws adequately, alternating the tightening between the two clamps, until rigid internal fixation is obtained. Cancellous bone chips harvested from the iliac crest were wedged between the spinous processes and the clamp (below the free edge of the clamp) and placed bilaterally over the prepared fusion sites. Particular consideration was given in cases of atlantoaxial arthrodesis, where specially designed rounded clamps were used to hook around the arch of C-I and the lamina of C-2. Bilateral clamps were used and care was taken to achieve satisfactory positioning, avoiding rotation of the clamps (Fig. 1). During the first onethird of the series, only an onlay lateral bone graft was used, but in the latter part of the series a triangularshaped bone graft was wedged between the arch of C- 1 and the lamina of C-2 below the clamp before the screws were tightened, in addition to a lateral and interspinous onlay bone graft. The correct clamps and screws must be selected to achieve rigid internal fixation. All patients routinely wore a Philadelphia collar for 6 weeks after surgery. A perioperative antibiotic agent (nafcillin or vancomycin) was given: one dose preoperatively, then for 3 days postoperatively. Follow- Up Review No patients were lost to follow-up review. The followup period varied from 6 to 40 months, with an average of 21 months. All patients were followed clinically, and FIG. 1. Lateral cervical spine x-ray film obtained 6 months postoperatively in a 49-year-old patient with traumatic ligamentous instability at C1-2. Bilateral Halifax interlaminar clamps are shown in good position with solid bone fusion between the arch of C-1 and the laminae and spinous process of C-2. routine radiographs were obtained postoperatively at 6 weeks and at intervals thereafter to evaluate vertebral bone element alignment, clamp position, restoration of angulation, and the quality of the bone fusion. Flexionextension studies were obtained in all patients at 6 weeks. In cases where it was difficult or impossible to evaluate the integrity and quality of the bone fusion, additional studies were performed. These included conventional polytomography and axial CT in 15 cases. Magnetic resonance imaging was obtained in 12 cases at variable follow-up periods for evaluation of the spinal cord and adjacent structures, including the neural canal and intervertebral foramina. Results Intraoperative Course Clamps were placed bilaterally in 47 of the 50 patients. Two patients had unilateral laminar fractures, and a clamp was placed contralateral to these fractures. In the remaining patient a unilateral clamp was placed following hemifacetectomies for lateral cervical spine stenosis due to a failed anterior cervical fusion at the same level. Fusions were performed at the C 1-2 level in 17 cases, at the C1-3 level in one, between C-2 and C-7 at a single level in 28 cases, and between C-2 and C-7 at two levels in four cases. Iliac bone grafting was performed on all patients except one, where an anterior cervical fusion had previously been attempted. Rigid internal stabilization was achieved at the time of operation in all cases. This was demonstrated intraoperatively by axial rotation, with flexion-extension force applied by a Kocher clamp. Total reduction with realignment of the vertebral bone elements was achieved in all 32 cases of lower cervical spine fusion (C2-7) regardless of the amount of preoperative subluxation J. Neurosurg. / Volume 78/May,

3 E. F. Aldrich, P. B. Weber, and W. N. Crow FIG. 2. Lateral cervical spine x-ray film obtained at routine 6-week postoperative evaluation in a 55-year-old woman with rheumatoid arthritis. An asymptomatic C1-2 subluxation was identified with screw loosening and dislocation of one clamp. Anterior sabluxation of C-I with respect to C-2 is present, indicating instability without any evidence of bone fusion. or whether reduction was achieved by traction. In cases of perched or locked facets that did not realign with skeletal traction, open reduction followed by bilateral Halifax interlaminar clamp placement restored alignment. Initial satisfactory intraoperative reduction of the atlanto-dens interval was achieved in all 17 patients undergoing atlantoaxial arthrodesis and in the one patient undergoing C1-3 fusion with bilateral Halifax interlaminar clamp placement. Surgical Failure Surgical failure occurred in five patients in this series. In these cases either the patient became symptomatic or routine diagnostic studies revealed instability at the previously fused level. Atlantoaxial fusion failed in three of 17 cases. Failure occurred in two patients with atlantoaxial instability associated with Down's syndrome. Screw loosening was identified in one of these patients, causing residual atlantoaxial instability, as seen on the routine 6-week follow-up radiographic studies. The second patient became symptomatic (myelopathic) 5 months following the original fusion, and studies revealed a C-1 arch fracture to be the cause of the fusion failure. In both cases alternative fusion methods were employed to achieve atlantoaxial stability. Asymptomatic screw loosening with movement on flexion-extension views was noted in a 68-year-old woman with rheumatoid arthritis at a routine 6-week follow-up evaluation (Fig. 2). At reoperation, the screws were retightened and additional bone grafting was performed. The patient ultimately obtained a solid bone fusion and is currently asymptomatic. One patient with a failed CI-2 wire fusion was referred to our facility for reoperation. At the time of FIG. 3. X-ray films showing preoperative anterior subluxation with angulation of C4-5 (A) and bilateral Halifax interlaminar clamps in position (B) in a 51-year-old man who sustained traumatic C4-5 instability following a motor-vehicle accident. Note the satisfactory realignment of the vertebral bone elements and the restoration of the neural canal to normal. surgery, it was found that the C-2 laminae were eroded and fractured, and clamp placement was not possible. A C1-3 fusion with bone grafting was attempted but failed because of screw loosening 3 weeks following the procedure. An alternative fusion method was used; the patient is currently asymptomatic and doing well. Screw loosening was seen and movement occurred on flexion-extension views in a patient with posterior ligamentous instability at the C2-3 level 4 weeks postoperatively, at which time the Lhermitte's sign reappeared. At reoperation, the screws were tightened again and additional bone grafting was carried out; the patient continued to a solid bone fusion and is currently asymptomatic. There were no clamp failures or other complications in patients with fusion at the C3-7 levels. No other complications have thus far been encountered in this series. Radiographic Results Postoperative vertebral bone element alignment was maintained in all patients with pathology of the lower cervical spine (C2-7) until solid bone fusion was demonstrated. No vertebral body slippage or angulation developed at the level of fusion in any patient, and the neural canal was restored to its original diameter in all cases (Fig. 3). Slippage was noted on routine follow-up x-ray films at 6 weeks in the woman with rheumatoid arthritis mentioned above, resulting in an additional 4- mm atlanto-dens interval malalignment. No movement occurred on flexion-extension studies, and the patient was neurologically intact. She subsequently achieved a solid bone fusion and remains asymptomatic. In addition to the previously mentioned surgical failures, routine follow-up studies of a patient with a type II dens fracture revealed screw loosening. The patient was asymptomatic, and conventional polytomography and CT revealed solid bone fusion with no signs of instability on flexion-extension views. In cases where routine x-ray films did not clearly 704 J. Neurosurg. / Volume 78/May, 1993

4 Long-term review of cervical fusion by Halifax clamp FIG. 4. Computerized tomography scans in a 39-year-old man who underwent posterior cervical fusion with Halifax interlaminar clamps and iliac bone grafting for a traumatic C5-6 anterior subluxation. Six months postoperatively, solid bone fusion of the iliac graft to the laminae of C-5 (A and B) and the laminae of C-6 (C and D) is demonstrated, indicating solid bone fusion across the fused level. show solid bone fusion, polytomography as well as CT was performed to evaluate the integrity and quality of the bone fusion. All 32 patients with fusion in the lower cervical area (C2-7) demonstrated adequate solid bone fusion (Fig. 4). Following atlantoaxial arthrodesis in two patients, solid bone fusion could not be demonstrated by either method, but no movement occurred on flexion-extension views. In both cases the iliac bone graft had fused to the laminae of C-2, but not to the ring of C-1 (Fig. 5). These two patients have been followed for 13 and 21 months postoperatively, and continue to be asymptomatic. Of the 17 atlantoaxial fusions, 12 were well aligned with solid bone fusion, one was malaligned with solid bone fusion, and two fibrous well-aligned unions occurred; two cases were ultimately fused by alternative techniques. Conventional polytomography and CT showed similar results with regard to evaluation of bone fusion quality and integrity. Discussion Posterior Lower Cervical Fusion Various alternatives are available for posterior cervical fusion, and most of these methods use stainless steel as the internal fixation agent, with or without lilac bone grafting, z6"2z3~ Cervical sublaminar wire fusion can lead to intraoperative and postoperative complications, 14 therefore alternative techniques such as intraspinous or interlaminar wire internal fixation have been developed. Although these techniques avoid sublaminar wire placement, they do not avoid the potential danger of wire fatigue and breakage, or pulling through the bone. Halifax interlaminar clamps provide immediate rigid internal fixation, ~5 avoid the hazards of sublaminar wire, and potentially lessen the chance of metal fatigue. FIG. 5. Computerized tomography scans in a patient following bilateral Halifax interlaminar clamp atlantoaxial arthrodesis with iliac bone grafting. A and B: Incomplete bone fusion of the iliac bone graft to the arch of C-1 is illustrated. C and D: Solid bone fusion of the iliac bone graft to the laminae of C-2 can be seen. Because of MR imaging compatibility, l internal clamp fixation allows noninvasive lifelong follow-up monitoring of the cervical neural elements. Our current results provide further support that interlaminar clamp internal fixation is a very effective way of dealing with posterior cervical instability in the lower cervical area. The ease of the technique, lack of complications, and excellent results lead us to conclude that this might be the method of choice for performing posterior lower cervical fusions. Despite these favorable results, a few unresolved issues persist. Some authors advocate using only a unilateral interlaminar clamp in cases of posterior cervical spine instability.~6'28 In our current series we performed unilateral clamp placement in only three cases, two because of unilateral laminar and facet fractures. All three patients achieved a solid bone fusion without complication, but no meaningful conclusions can be made from these few patients. In general, we believe that, if the trauma was sufficient to cause unilateral facet dislocation, it is likely that anterior and posterior ligamentous injury, disc damage, and possible damage to the contralateral facet occurred. We fail to see the advantage of applying only one clamp when bilateral clamp placement surely adds to the immediate stability. Multilevel clamping was performed in three cases in the lower cervical area where two levels were spanned with the clamps. All three patients continued to a solid bone fusion without complication. Although these patients did well, some caution is warranted for multilevel fusion and care must be taken that anterior displacement of the intermediate segment does not occur. It is especially important to avoid cases in which there is potential posterior ring disruption or pedicle fractures. If more than one level is clamped, it might be effective to use one clamp to span two levels and the contralateral clamp to span only one level. A larger series would be necessary to evaluate multilevel fusion with Halifax interlaminar clamps. The method selected for treatment of cervical spine J. Neurosurg. / Volume 78/May,

5 E. F. Aldrich, P. B. Weber, and W. N. Crow instability must counteract the forces that produced the instability to ensure optimum results. Most instances of instability treated by posterior cervical fusion result from flexion forces, and this flexion force should therefore be counteracted by the fixation device. The design of the Halifax clamp negates the flexion-extension forces seen in this area, and is therefore well suited for posterior cervical fusion. Because of the clamp design, great reduction forces can be applied by slowly alternating tightening of the screws, thus leading to reduction and realignment of the subluxed vertebral bone elements at the unstable level, thereby restoring the vertebral canal to normal. Atlantoaxial Arthrodesis The atlantoaxial joint is a multiaxial joint with great mobility, accounting for one-half of the cervical rotation; 29 it therefore differs greatly from the lower cervical spine, where flexion-extension provides most of the movement. The atlantoaxial region is further complicated by a variety of pathological processes affecting this level, and therefore upper cervical fusion results are less than perfect. A variety of surgical options have been described to attain atlantoaxial arthrodesis. Wire internal fixation combined with autogenous bone grafts is now the most frequently used treatment, ~.-,5 with the methods described by Gallie ~2 and Brooks and Jenkins s being the most common. Many variations of this technique have also been described ~8'2~ as well as alternative methods, including screw fixation and instrumentation with rods. 4A~ The frequency of osseous union associated with a Brooks or Gallie technique without external halo fixation varies from 60% to 85%. 3~ ~.24.ze In a recent report, Dickman, etal., 9 reported a 97% union rate in 36 patients undergoing atlantoaxial arthrodesis for a variety of causes using a modified wire placement and iliac graft technique. Chan, etal., 7 reported fusion in 10 (91%) of 11 rheumatoid arthritic patients treated with a Gallie technique. In these two series the patients were immobilized postoperatively in a halo vest for 3 months. Cybulski, etal., 8 reported no failures in their series of eight patients undergoing atlantoaxial arthrodesis with Halifax interlaminar clamps in combination with methyl methacrylate and iliac bone grafting. We previously reported our experience with eight cases in which no fusion failures or mechanical difficulties were encountered. ~ However, with more experience and a longer follow-up period, it has become evident that fusion failures do occur at this level, mostly because of screw loosening and in patients at risk for nonunion. In our current series we obtained a fusion rate of 82% (14 of 17 cases). It is well known that instability resulting from trauma yields better results than that caused by rheumatoid arthritis, congenital abnormalities, or tumors. This was also the case in our series, where all the traumatic C 1-2 instabilities fused, but fusion failed in two of the three patients with Down's syndrome and in the one patient with rheumatoid arthritis. In two of these three cases, screw loosening was the cause of the fusion failure. Seex and Johnston 27 reported their ex- perience with interlaminar clamps for posterior cervical fusion; failure occurred in two of their four C1-2 fusions and in one case of attempted CI-3 fusion. Our only C 1-3 fusion failed because of screw loosening, and we recommend that alternative fusion methods be utilized should a CI-3 fusion be necessary. Recently Moskovich and Crockard 2~ reported their experience with Halifax interlaminar clamps for atlantoaxial anhrodesis for various causes and obtained an 80% fusion rate (20 of 25 cases). Screw loosening occurring between 3 and 6 weeks postoperatively led to fusion failure in four of our 50 cases. It would therefore seem, at least in our series, that longer periods of postoperative immobilization would not have prevented this complication. The only other failure occurred 5 months postoperatively when the arch of C-1 fractured in a patient with Down's syndrome. We believe that a good fusion technique is a prerequisite. It is important to use the specially designed clamp with a rounded, deeper curvature to hook around the arch of C-1 and the lamina of C-2. Use of the flatter clamp designed for the lower cervical area should be avoided, as this will lead to clamp dislocation. The clamp site should be adequately prepared so that clamp rotation is avoided. The correct screw length should be used, and the technique can only be successful if rigid internal stabilization is achieved at the time of surgery. Any movement at the level of internal fixation might lead to potential loosening of the screw, and solid bone fusion may not occur. Despite following a rigid surgical technique, there is still an inherent potential for screw loosening; a new Halifax interlaminar clamp is currently available that is designed to lessen or prevent this complication. A major difference among series describing atlantoaxial arthrodesis is the wide variation in the type and regimens of postoperative immobilization. It has been shown that a halo vest is the most effective way of immobilizing the upper cervical spine ~7 and that collars and braces are inadequate to control rotation. In our patients as well as those in other published Halifax interlaminar clamp series, the only postoperative immobilization used was a soft cervical collar or a Philadelphia collar for 6 weeks. When halo external fixation was not used, the routine wire autogenous bone fusion techniques achieved fusion rates ranging from 60% to 85%, with an average below 80%. It seems at this stage that superior results (90% to 97%) can be obtained by wire fusion with autogenous bone grafting combined with external halo immobilization for 3 months. We believe that higher atlantoaxial fusion rates can be achieved in the future using the redesigned Halifax interlaminar clamp for the C1-2 area to prevent screw loosening, with meticulous attention to the bone fusion, and with external halo immobilization for a 3-month period in the nontraumatic group of patients at risk for nonunion. Bone Graft Originally, Tucker 28 advocated the use of interlaminar clamps without bone grafting and reported fusion 706 J. Neurosurg. / Volume 78 /May, 1993

6 Long-term review of cervical fusion by Halifax clamp across the facet joints in most patients. Holness, et al., j6 reported radiographically documented anterior and/or posterior bone fusion in all patients followed longer than 4 years when only interlaminar clamps were used. Although it is difficult to argue with these authors' excellent results in their large series, not using a bone graft with internal fixation in the cervical spine remains controversial. The internal fixation devices provide only temporary fixation and cannot be depended upon for long-term stability. The success of the fusion, therefore, ultimately depends on the quality of the bone fusion. For that reason, it is our policy to perform bilateral onlay iliac bone grafts in all patients with fusion in the lower cervical spine, as we believe that this greatly enhances the chance of a solid bone fusion occurring and therefore providing long-term stability. Using this technique, we achieved good-quality solid bone fusion across the unstable level in all cases. Because of the anatomical configuration of the posterior elements of C-I and C-2, particular attention should be paid to the bone graft in this area. Most bone fusion failures in our series occurred because of inadequate fusion of the bone graft to the posterior ring of C-1. Onlay bone grafts in this area have a tendency to form a pseudoarthrosis and therefore in the latter twothirds of our series we adopted a technique in which a triangular-shaped piece of bone was wedged between the clamp and the posterior elements of C-I and C-2. In addition to this bilateral triangular bone graft, onlay bone was grafted lateral as well as medial to the clamps. Adequate decortication of the posterior elements of C-1 and C-2 is mandatory before clamp and bone graft placement to ensure fusion. Bone grafts should not be placed between the arches of C-1 and C-2 before the screws are tightened, as slippage would leave a gap resulting in clamp loosening. Using this technique, we obtained an 82% rate of solid bone fusion (14 of 17 patients). In recent reports, stable osseous fusions were reported in 76% :3 and 83% 9 of cases where atlantoaxial fusion was attempted using an interspinous method of posterior atlantoaxial arthrodesis. Radiographic Evaluation of Bone Fusion Flexion-extension studies, necessary to evaluate the stability of the fusion, are routinely carried out at 6 weeks postoperatively. Although the integrity of the bone fusion across the facets and lamina can be evaluated adequately on a routine x-ray film, this may sometimes be difficult. In these cases, conventional polytomography or axial CT yields excellent information regarding integrity of the fusion. In our series, both methods yielded the same information but, because of the ease of technique and interpretation, we prefer to obtain CT bone scans when in doubt. It is especially important when evaluating patients with atlantoaxial arthrodesis, where pseudoarthrosis is a more common occurrence. References 1. Aldrich EF, Crow WN, Weber PB, et al: Use of MR imaging-compatible Halifax interlaminar clamps for pos- terior cervical fusion. J Neurosurg 74: , Alexander E Jr: Posterior fusions of the cervical spine. Clin Neurosurg 28: , Anderson LD: Fractures of the odontoid process of the axis, in The Cervical Spine Research Society (eds): The Cervical Spine. Philadelphia: JB Lippincott, 1983, pp Barbour JR: Screw fixation in fracture of the odontoid process. South Austral Clin 5:20-24, Brooks AL, Jenkins EB: Atlanto-axial arthrodesis by the wedge compression method. J Bone Joint Surg (Am) 60: , Cahill DW, Bellegarrigue R, Ducker TB: Bilateral facet to spinous process fusion: a new technique for posterior spinal fusion after trauma. Neurosurgery 13:1-4, Chan DPK, Ngian KS, Cohen L: Posterior upper cervical fusion in rheumatoid arthritis. Spine 17: , Cybulski GR, Stone JL, Crowell RM, et ai: Use of Halifax interlaminar clamps for posterior C 1-2 arthrodesis. Neurosurgery 22: , Dickman CA, Sonntag VKH, Papadopoulos SM, et al: The interspinous method of posterior atlantoaxial arthrodesis. J Neurosurg 74: , DuToit G Jr: Lateral atlanto-axial arthrodesis. A screw fixation technique. S Aft J Surg 14:9-12, 1976 I I. Fielding JW: Current concepts review. The status of arthrodesis of the cervical spine. J Bone Joint Surg (Am) 70: , Gallie WE: Fractures and dislocations of the cervical spine. Am J Surg 46: , Geisler FH, Cheng C, Poka A, et ah Anterior screw fixation of posteriorly displaced type II odontoid fractures. Neurosurgery 25:30-38, Geremia GK, Kim KS, Cerullo L, et ah Complications of sublaminar wiring. Surg Neurol 23: , Grob D, Crisco JJ III, Panjabi MM, et al: Biomechanical evaluation of four different posterior atlantoaxial fixation techniques. Spine 17: , Holness RO, Huestis WS, Howes W J, et at: Posterior stabilization with an interlaminar clamp in cervical injuries: technical note and review of the long term experience with the method. Nenrosurgery 14: , Johnson RM, Hart DL, Simmons EF, et al: Cervic,~d orthroses. A study comparing their effectiveness in restricting cervical motion in normal subjects. J Bone Joint Sorg (Am) 59: , Lipson S J, Hammerschlag SB: Atlantoaxial arthrodesis in the presence of posterior spondyloschisis (bifid arch) of the atlas. A report of three cases and an evaluation of alternative wiring techniques by computerized tomography. Spine 9:65-69, McCarron RF, Robertson WW: Brooks fusion for atlantoaxial instability in rheumatoid arthritis. South Med J 81: , Mitsui H: A new operation for atlanto-axial arthrodesis. J Bone Joint Surg (Br) 66: , I. Moskovich R, Crockard HA: Atlantoaxial arthrodesis using interlaminar clamps. An improved technique. Spine 17: , Murphy M J, Southwick WO: Posterior approaches and fusions, in The Cervical Spine Research Society (eds): The Cervical Spine. Philadelphia: JB Lippincott, 1983, pp Papadopoulos SM, Dickman CA, Sonntag VKH: Atlantoaxial stabilization in rheumatoid arthritis. J Neurosurg 74:1-7, Paradis GR, Janes JM: Posttraumatic atlantoaxial instability: the fate of the odontoid process fracture in 46 cases. J. Neurosurg. / Volume 78/May,

7 E. F. Aldrich, P. B. Weber, and W. N. Crow J Trauma 13: , Pierce DS, Barr JS Jr: Fractures and dislocations at the base of the skull and upper cervical spine, in The Cervical Spine Research Society (eds): The Cervical Spine. Philadelphia: JB Lippincott, 1983, pp Schatzker J, Rorabeck CH, Waddell JP: Fractures of the dens (odontoid process). An analysis of thirty-seven cases. J Bone Joint Surg (Br) 53: , Seex K, Johnston RA: lntedaminar clamp for posterior fusions. J Neurosurg 75:495, 1991 (Letter) 28. Tucker HH: Technical report: method of fixation of subluxed or dislocated cervical spine below CI-C2. Can J Neuroi Sci 2: , White AA III, Panjabi MM: Clinical Biomechanies of the Spine. Philadelphia: JB Lippincott, 1978, p Whitehill R, Schmidt R: The posterior interspinous fusion in the treatment of quadriplegia. Spine 8: , Yashon D: Surgical management of trauma to the spine, in Schmidek HH, Sweet WH (eds): Operative Neurosurgical Techniques. Indications, Methods and Results, ed 2. Orlando, Ha: Grune & Stratton, 1988, Vol 1, pp Manuscript received June 26,! 992. Accepted in final form October 2, This paper was presented in part at the Annual Meeting of the American Association of Neurological Surgeons, San Francisco, California, April 11-16, Address reprint requests to: E. Francois Aldrich, M.D., Division of Neurosurgery, E-I 7, University of Texas Medical Branch, Galveston, Texas J. Neurosurg~ / Volume 78/May, 1993