- Email: [email protected]
Early predictive value of supine and upright X-ray films of odontoid fractures treated with halo-vest immobilization
The Spine Journal 8 (2008) 612–618
Early predictive value of supine and upright X-ray films of odontoid fractures treated with halo-vest immobilization David H. Kim, MDa,*, Alexander R. Vaccaro, MDb, Jesse Affonso, MDc, Louis Jenis, MDa, Alan S. Hilibrand, MDb, Todd J. Albert, MDb a
Department of Orthopaedic Surgery, Tufts University Medical School, New England Baptist Hospital, 125 Parker Hill Avenue, Boston, MA 02120, USA b Department of Orthopaedic Surgery, Thomas Jefferson University and the Rothman Institute, 925 Chestnut Street, Philadelphia, PA 19107, USA c University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655, USA Received 16 January 2007; accepted 24 March 2007
Abstract
BACKGROUND: Although halo-vest immobilization remains a common form of treatment for type II odontoid fractures, nonunion and C1–2 instability may be the result in up to 20% to 40% of patients. PURPOSE: Supine and upright lateral X-ray films may allow early identification of patients likely to fail halo-vest treatment and earlier surgical treatment with decreased morbidity from prolonged unsuccessful halo-vest immobilization. STUDY DESIGN/SETTING: A prospective cohort study was performed. PATIENT SAMPLE: Twenty patients with type II odontoid fractures. OUTCOME MEASURES: Posttreatment nonunion and C1–2 instability as determined by plain X-ray films and computed tomography scan. METHODS: Both supine and upright lateral X-ray films were obtained immediately after halovest application and at the 2-week, 6-week, and 3-month follow-up. Flexion-extension lateral Xray films were obtained after halo-vest removal. Patients with nonunion or instability underwent computed tomography scan. Upright X-ray films were compared serially to identify loss of reduction. Pairs of supine and upright X-ray films were compared to measure any change in displacement or angulation upon transition from supine to upright position. Nonunion patients were compared with healed patients to determine any difference in fracture behavior based on serial supine and upright X-ray films. RESULTS: Twenty patients with type II odontoid fractures were identified during the study period. Three patients with multiple trauma underwent immediate surgical stabilization. Three elderly patients with nondisplaced fractures were treated in a cervical orthosis. Fourteen patients initiated and completed 3 months of halo-vest immobilization. After halo-vest removal, 4 of 14 patients (29%) showed radiographic nonunion or instability. In all 4 nonunion patients, supine and upright radiographs at 2 weeks revealed change in fracture angulation $5 between the supine and upright positions. In three of these patients standard serial upright lateral X-ray films failed to identify any loss of reduction. In the remaining patient, progressive angulation of 15 was observed. No measurable change in angulation between supine and upright X-ray films was observed in any patient who healed successfully with halo-vest treatment. CONCLUSIONS: Obtaining both supine and upright lateral X-ray films during the follow-up period may identify patients at risk for failure of halo-vest treatment as early as 2 weeks after initiation of treatment. A change in fracture angulation $5 suggests an increased risk of treatment
FDA device/drug status: approved for this indication (halo vest). Nothing of value received from a commercial entity related to this manuscript. 1529-9430/08/$ – see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.spinee.2007.03.009
* Corresponding author. David H. Kim, MD, 125 Parker Hill Ave, Boston, MA 02120. Tel.: (617) 754-6482; fax: (617) 754-5595. E-mail address: [email protected] (D.H. Kim)
D.H. Kim et al. / The Spine Journal 8 (2008) 612–618
613
failure and the potential benefit of early surgical stabilization. Ó 2008 Elsevier Inc. All rights reserved. Keywords:
Odontoid fracture; Halo vest; Evaluation; Nonunion
Introduction Cervical spine odontoid fractures are frequently observed, constituting approximately 10% to 18% of all clinically significant cervical spine injuries in adults [1]. The most commonly used classification system, by Anderson and D’Alonzo [2], defines three basic fracture patterns. Type II fractures occur through the junction of the odontoid process with the C2 vertebral body and represent the most common type of fracture as well as the greatest challenge to the treating clinician. High rates of nonunion and instability have plagued all forms of nonoperative treatment and have resulted in a trend toward early surgical stabilization in many cases. Nevertheless, despite the increasing number of surgical techniques available, halo-vest immobilization remains the most frequently selected form of treatment in many centers. The principal advantages of halo-vest treatment are the potential avoidance of surgery and the perception that it represents a ‘‘noninvasive’’ form of treatment. Unfortunately, halo vests have also been associated with a high rate of complications including pin loosening, pin site infection, intracranial pin penetration with abscess formation, poor cosmetic scarring, and postimmobilization neck stiffness. Halo vests are often poorly tolerated by elderly patients who may suffer skin breakdown or aspiration. Nonunion and cervical spine instability with persistent pain and risk of spinal cord injury remain the major concern after treatment of type II fractures with halo-vest immobilization. Published series have reported nonunion rates of 18% to 41% and have suggested numerous possible risk factors, including fracture displacement, angulation, distraction, comminution, delay in diagnosis, and patient age [1,3–6]. Occasionally, loss of reduction is observed on serial X-ray films after halo-vest application and serves as an early indicator of treatment failure. More typically, failure of halo-vest treatment is not recognized until flexion-extension X-ray films or computed tomographic imaging is obtained at the conclusion of 12 to 15 weeks of immobilization. Anderson et al. [7] have shown that halo vests are relatively less effective at immobilizing the upper cervical spine compared with the mid- and lower cervical spine. Comparing upright and supine lateral X-ray films in patients in halo vests has revealed varying rates of motion that is more pronounced in the upper cervical spine. The study investigators hypothesized that obtaining both upright and supine lateral X-ray films in patients undergoing halo-vest treatment for type II odontoid fractures may allow earlier recognition of fractures at increased risk for treatment failure. If true, early surgical stabilization or conversion to
alternative treatment in these patients may prevent additional morbidity from prolonged halo-vest immobilization.
Materials and methods Study population An application for the following study was submitted and approved by the hospital institutional review board. During the period of January 2003 through December 2004, prospective data were collected on 20 consecutive individuals diagnosed with acute type II odontoid fractures. All patients underwent evaluation and treatment within 48 hours at a level I trauma center. Fourteen patients were neurologically intact and elected to undergo treatment with halo-vest immobilization. Patient age ranged from 18 to 67 years with an average of 43 years. Twelve patients had isolated injuries, whereas two patients were considered polytrauma, both with associated upper-extremity fractures. Motor vehicle accidents (nine patients) and falls (five patients) were the most frequent mechanisms of injury. Six fractures were displaced or angulated and required closed reduction before halo-vest application. During the study period, three patients with severe polytrauma were considered unsuitable for halo-vest immobilization and treated with immediate surgical stabilization. Three elderly and debilitated patients elected to forego treatment with either halovest immobilization or surgery and were treated with hard collar immobilization. Halo-vest treatment After placement of the halo ring, fractures displaced greater than 5 mm or angulated greater than 10 underwent a closed reduction before attachment of the halo vest. Six patients required initial closed reduction, with two patients requiring multiple reduction attempts. No patient during the study period experienced redisplacement once successful closed reduction had been accomplished. After halo-vest application, all patients underwent immediate radiographic imaging consisting of an open-mouth odontoid view and both supine and upright lateral cervical spine X-ray films. Halo pins were retorqued according to manufacturers’ recommendations; and patients and their caretakers were instructed in routine pin-site and halo-vest management before discharge. Radiographic and clinical evaluation All patients underwent routine follow-up evaluations at 2 weeks, 6 weeks, and 3 months after halo-vest application.
614
D.H. Kim et al. / The Spine Journal 8 (2008) 612–618
Patients were interviewed regarding any changes in their pain experience or neurological status. Clinical evaluation included pin-site and halo-vest inspection as well as neurological examination. Radiographic evaluation at each scheduled follow-up visit included an open-mouth odontoid and supine and upright lateral cervical spine views. Before study onset, the decision was made that fracture displacement greater than 5 mm would represent failure of treatment and an indication for surgical stabilization, but this degree of displacement was not observed in any patient. All patients completed 3 months of halo-vest immobilization. After halo-vest removal, all patients underwent lateral flexion and extension cervical spine X-ray films. If X-ray films revealed a minimum of 30 flexion and 30 extension without evidence of measurable displacement and patients reported no significant pain through this range of motion, the fracture was considered healed, and patients were allowed to experience gradual return to normal activities with a short period of soft collar wear and supervised physical therapy on an as-needed basis. If patients were unable to perform flexion and extension through an adequate range of motion or if significant pain was reported during performance of this maneuver, a helical computed tomography (CT) study of the cervical spine was obtained with 1.5mm thin cuts including sagittal and coronal reconstructed images. Radiographic measurements The odontoid process presents an irregular anatomic profile on lateral radiographic images, and no consistent fracture displacement or angulation measurement technique has been universally accepted. The posterior border of the odontoid is typically more linear, but a small degree of rotation can create the appearance of two posterior cortical lines at the level of the centrum. The portion of the odontoid proximal to the atlas ring is often poorly visualized, particularly in elderly individuals with osteopenia and spondylotic degeneration. To determine the most reliable measurement methods for the purposes of this study, fracture displacement and angulation were measured by three different independent physicians using two different measurement techniques. In the anterior method, a visual ‘‘best-fit’’ line was drawn along the anterior border of the odontoid process proximal to the fracture line. A second ‘‘best-fit’’ line was drawn along the anterior border of the axis body. A third line was drawn across the fracture plane defined by the points at which the fracture crossed the anterior and posterior cortices. If these points were poorly visualized on plain X-ray films, the points were extrapolated from the original reconstructed CT images from the time of initial evaluation. Displacement was defined by the distance along the third line from the first line to the second. Angulation was measured as the angle subtended by the first two lines.
In the posterior measurement method, a visual ‘‘best-fit’’ line was drawn along the posterior border of the odontoid process proximal to the fracture line. A second ‘‘best-fit’’ line was drawn along the posterior border of the axis body. The third line was again drawn across the fracture plane as described previously for the anterior method. Displacement and angulation were measured by using the same technique using the new lines. No significant difference was identified between the anterior and posterior measurement methods. Raters expressed the impression that assessment of displacement was easier with the anterior method, whereas assessment of angulation was easier with the posterior method. Therefore, for the purposes of this study, the anterior method was selected to report displacement values, whereas the posterior method was used to report angulation values; 5 was selected arbitrarily as the smallest increment that all raters were confident as representative of a true change in fracture angulation. Statistical analysis Statistical evaluation of the results was performed using the Fisher exact test and the binomial test. Significance was established as a p value #.05.
Results Clinical results During the 2-year study period, 14 patients met study criteria and were included in the study population. All 14 patients successfully completed 3 months of halo-vest immobilization. Five patients developed superficial pin site infections requiring treatment with oral antibiotics. In one patient, persistent unilateral scalp pain in the distribution of the supraorbital nerve led to repositioning of an anterior pin. Seven patients reported varying levels of skin irritation from the vest, but no patient developed clinically significant skin breakdown. There were no episodes of aspiration or pulmonary complications. Serial comparisons of upright X-ray films failed to show progressive displacement in any patient. No patient showed fracture displacement greater than 3 mm on any radiographic view. One patient experienced an incremental progressive posterior angulation of 15 over the course of 3 months without fracture displacement. All patients remained neurologically intact throughout the course of the study. After halo-vest treatment, 4 of 14 patients (29%) developed radiographic nonunion or instability, including the 1 patient with progressive angulation. All four patients reported persistent neck pain. Two patients showed instability on flexion and extension X-ray films, whereas the remaining 2 patients were unable to perform adequate sagittal
D.H. Kim et al. / The Spine Journal 8 (2008) 612–618
range of motion and were diagnosed on the basis of CT imaging. Radiographic results In 13 of 14 study patients, including 3 of 4 patients who later developed nonunions, an evaluation of open-mouth odontoid views and a comparison of supine and upright lateral radiographs immediately after halo-vest application failed to identify measurable displacement or change in angulation. One patient who later developed nonunion showed a 12 change in angulation. For statistical analysis, patients were divided into two groups. Group 1 included the 10 patients who were treated successfully by halo-vest immobilization. Group 2 included the 4 patients who developed a nonunion. Two weeks after halo-vest application, standard upright open-mouth views as well as upright lateral X-ray films did not show measurable fracture displacement or a change in angulation in any patient. A comparison of supine and upright lateral X-ray films at 2 weeks, however, revealed a change in angulation of $5 between the supine and upright positions in all 4 patients in group 1 and none of the patients in group 2 (Table 1). This difference was significant (p5.001); 95% confidence intervals cannot be calculated for this situation. However, the lower level for the 95% odds ratio equals 11.07. The upper level cannot be calculated but is by inference infinity. At 6 weeks follow-up, no measurable fracture displacement was observed in any patient. One patient in group 2 showed an increase in fracture angulation of 9 compared with the 2-week X-ray film on upright lateral imaging. A comparison of supine and upright lateral radiographs showed $5 angulation in 3 of 4 patients in group 2 and 0 of 10 patients in group 1. This difference was not significant (p5.07). At the 3-month follow-up, no measurable displacement was observed in any patient. The group 2 patient with increased fracture angulation at 6 weeks shwed further angulation of 6 on upright lateral imaging. This represented a total change of 15 compared with the 2-week X-ray film. A comparison of supine and upright lateral X-ray films showed $5 angulation in 2 of 4 group 2 patients and 0 of 10 group 1 patients. Again, this difference was not significant (p5.25). Although fracture angulation or displacement between supine and upright radiographic views was the primary
615
variable of interest, secondary analysis was performed comparing other potential risk factors between groups (Table 2). No significant difference was identified in terms of age, smoking status, initial fracture displacement, or angulation or direction of displacement. During the study period, significant fracture comminution was considered an indication for early surgical stabilization. Therefore, there were no patients in the final study population with significant fracture comminution.
Discussion The fundamental principles of treatment for odontoid fractures are widely accepted and include early diagnosis, anatomic reduction, and adequate immobilization to permit healing. Nevertheless, there exists no consensus regarding the ideal form of treatment in most cases. A variety of treatment strategies have been advocated, ranging from initial nonoperative treatment in all patients, to early surgery in selected high-risk populations, to early surgery in all patients [1–3,6,8–16]. Nonoperative treatment recommendations have included halo-vest, cervicothoracic orthosis, and hard-collar immobilization. An evidence-based review of the literature indicates that no standards or guidelines exist for the treatment of type II odontoid fractures, only various treatment options [17–19]. Despite an increasing number of surgical options for stabilization of acute type II odontoid fractures and high reported rates of nonunion after all forms of external immobilization, halo vests remain a frequently selected form of treatment. Several clinical series have reported relatively high complication and failure rates associated with halo-vest treatment. Seybold and Bayley [4] reported a 20% nonunion rate as well as increased complications and problematic postimmobilization stiffness in elderly patients. Stoney et al. [5] reported an 18% nonunion rate and 40% rate of moderate to severe stiffness after halo-vest treatment. Although previous investigators have blamed high nonunion rates on poor vascular supply, a detailed anatomical study has shown a robust anastomotic arterial arcade from both carotid and vertebral systems, which includes both endosteal and ligamentous vessels [20]. Delayed healing or nonunions appear more likely because of the unique histomorphometry of bone at the odontoid base characterized by thin cortical bone and a relatively weak trabecular
Table 1 Change in fracture angulation and displacement between supine and upright tray firms Healed fractures (group 1) Angulation $5 Displacement $2 mm * p!.05.
Nonunion (group 2)
0 wk
2 wk
6 wk
3 mo
0 wk
2 wk
6 wk
3 mo
0/10 0/10
0/10 0/10
0/10 0/10
0/10 0/10
1/4 0/4
4/4* 0/4
3/4 0/4
2/4 0/4
616
D.H. Kim et al. / The Spine Journal 8 (2008) 612–618
Table 2 Nonunion risk factor characteristics of study population
Age (range) Active smokers Fracture initially displaced (requiring reduction) Direction of displacement (anterior/posterior) Fracture initially angulated (requiring reduction) Fracture comminution (%)
Healed fractures (group 1)
Nonunion (group 2)
p value for difference
45 (18–67) 3 of 10 3 of 10
39 (19–65) 3 of 4 3 of 4
Not significant Not significant Not significant
2/1
1/2
Not significant
1 of 10
2 of 4
Not significant
0
0
pattern [21–24]. Cross-sectional analysis has revealed this bone to be mostly cortical in nature [16–19]. Clinical studies have suggested various risk factors for nonunion after type II odontoid fractures, including significant fracture displacement or angulation, posterior displacement, fracture site comminution, patient age, and delay in diagnosis [1,3,4,25–27]. Fracture displacement has been the most commonly identified risk factor for nonunion regardless of treatment modality [1,25,26,28,29]. Consequently, a frequently used strategy for following odontoid fracture patients in halo vests is obtaining serial plain X-ray films in the upright lateral position. Although comparing films performed in the same position facilitates identification of interval fracture displacement, this method may miss persistent instability that occurs when the patient transfers between supine and standing positions. Anderson et al. [7] originally reported the potential for significant fracture-site motion between supine and upright positioning despite halo-vest immobilization in a prospective study of 42 patients with various cervical spine fractures. Five days after injury, the study found an average of 7 angular motion and 1.7-mm translation when moving between the two positions. The greatest amount of motion was observed proximally between the occiput and C1. The rate of nonunion in this study (29%) is in line with previous published reports after halo-vest treatment of odontoid fractures of 18% to 41% [5,6,28]. In this study, when supine and upright X-ray films were compared, an increased rate of fracture angulation was observed at all time points in a majority of patients who eventually developed nonunion (Fig. 1). This angular instability was most pronounced at the 2-week follow-up evaluation when all patients showed angulation $5 . Therefore, this may be the most sensitive time point for identifying patients at increased risk of nonunion using this modality. Of interest, only one of four nonunion patients showed a measurable change in angulation between supine and upright views immediately after halo-vest application. The authors have observed significant cervical muscle spasm in many patients in the acute postinjury period, and it is possible that muscle guarding results in early splinting and pseudostabilization of the upper cervical spine, preventing
immediate detection of fracture instability using supine and upright X-ray films. A trend was also noted toward decreasing amounts of fracture angulation in nonunion patients at the 6-week and 3-month follow-up. In general, neck pain complaints tended to diminish over time, and only one patient with nonunion reported persistent pain throughout the treatment period. The authors speculate that even in the absence of solid osseous healing some degree of soft-tissue healing and fibrous-tissue formation or fracture settling and fragment interdigitation may occur, resulting in cases of pseudostabilization and reduction in neck pain. Of note, no measurable fracture translation was observed in any nonunion patient when comparing supine and upright X-ray films or serial X-ray films over time. This suggests that fracture angulation may be a more sensitive predictor of odontoid fracture instability when following the course of healing with halo-vest treatment. This study suggests that obtaining both supine and upright lateral X-ray films during follow-up can identify odontoid fracture patients at risk for failure of halo-vest treatment as early as 2 weeks after initiation of treatment. This finding may allow earlier surgical treatment of these patients either with anterior screw osteosynthesis or posterior C1–2 instrumented fusion with more rapid recovery and avoidance of unnecessary morbidity from prolonged halo-vest immobilization. The current practice of many specialists is to obtain follow-up anteroposterior and lateral radiographic views in the upright position only. We believe that obtaining both supine and upright lateral X-ray films during follow-up of type II odontoid fractures, particularly at around 2 weeks postinjury, should be considered. A change in angulation can occur between supine and upright position despite no detectable change in displacement or angulation on serial upright X-ray films. Observation of this change in angulation suggests increased risk of failure of halo-vest treatment and may be used to identify patients who might benefit from early surgical stabilization. An obvious limitation of this study is a very small sample size. However, because the positive predictive value of supine and upright X-ray films was so high in terms of predicting nonunion in this prospective series, statistical analysis appropriate for small samples suggests that the associations are significant. Although exact 95% confidence intervals cannot be calculated for the data, the 95% confidence interval for the odds ratio can be determined and suggests that the presence of $5 angular change between supine and upright X-ray films is associated with at least an 11-fold increase in the risk of developing a nonunion. Nevertheless, a larger study would be required to confirm these findings. Another limitation is the determination of fracture healing that was used in this study. We used a standard protocol to assess whether adequate stability had been achieved after the completion of halo-vest treatment. The absence of neck pain and the absence of C1–2 instability or fracture displacement through a minimum of 30 of flexion and 30
D.H. Kim et al. / The Spine Journal 8 (2008) 612–618
617
Fig. 1. (A) Upright and (B) supine lateral cervical spine X-ray films 2 weeks after a halo-vest application in a 37-year-old patient with a type II odontoid fracture. There is a 10 increase in fracture angulation in the supine position. (C) This patient eventually developed a painful unstable nonunion after completion of 3 months of halo-vest treatment and required surgical stabilization.
618
D.H. Kim et al. / The Spine Journal 8 (2008) 612–618
of extension on lateral X-ray films were considered satisfactory criteria to determine that the injury had healed. If patients reported pain or could not perform adequate flexion-extension range of motion, a CT scan was obtained to assess the degree of osseous healing. Although this is a common protocol used by many spine surgeons, there may be a number of patients who appear stable based on flexion-extension X-ray films but who would not show solid osseous healing if a CT scan had been obtained. These patients may be considered to have a ‘‘stable fibrous nonunion,’’ and, although there is no consensus regarding the best treatment in this setting, observation without specific intervention appears a reasonable option. The findings of this study, therefore, are specific for predicting a painful or unstable nonunion as opposed to a strict definition of osseous healing. References [1] Clark CR, White AA 3rd. Fractures of the dens. A multicenter study. J Bone Joint Surg Am 1985;67:1340–8. [2] Anderson LD, D’Alonzo RT. Fractures of the odontoid process of the axis. J Bone Joint Surg Am 1974;56:1663–74. [3] Hadley MN, Browner CM, Liu SS, Sonntag VK. New subtype of acute odontoid fractures (type IIA). Neurosurgery 1988;22:67–71. [4] Seybold EA, Bayley JC. Functional outcome of surgically and conservatively managed dens fractures. Spine 1998;23:1837–45; discussion 1845–6. [5] Stoney J, O’Brien J, Wilde P. Treatment of type-two odontoid fractures in halothoracic vests. J Bone Joint Surg Br 1998;80:452–5. [6] Ekong CE, Schwartz ML, Tator CH, Rowed DW, Edmonds VE. Odontoid fracture: management with early mobilization using the halo device. Neurosurgery 1981;9:631–7. [7] Anderson PA, Budorick TE, Easton KB, Henley MB, Salciccioli GG. Failure of halo vest to prevent in vivo motion in patients with injured cervical spines. Spine 1991;16(10 suppl):S501–5. [8] Crockard HA, Heilman AE, Stevens JM. Progressive myelopathy secondary to odontoid fractures: clinical, radiological, and surgical features. J Neurosurg 1993;78:579–86. [9] Graziano G, Jaggers C, Lee M, Lynch W. A comparative study of fixation techniques for type II fractures of the odontoid process. Spine 1993;18:2383–7. [10] Apfelbaum RI, Lonser RR, Veres R, Casey A. Direct anterior screw fixation for recent and remote odontoid fractures. J Neurosurg 2000;93(2 suppl):227–36. [11] Bohler J. Anterior stabilization for acute fractures and non-unions of the dens. J Bone Joint Surg Am 1982;64:18–27.
[12] Borne GM, Bedou GL, Pinaudeau M, Cristino G, Hussein A. Odontoid process fracture osteosynthesis with a direct screw fixation technique in nine consecutive cases. J Neurosurg 1988;68: 223–6. [13] Esses SI, Bednar DA. Screw fixation of odontoid fractures and nonunions. Spine 1991;16(10 suppl):S483–5. [14] Etter C, Coscia M, Jaberg H, Aebi M. Direct anterior fixation of dens fractures with a cannulated screw system. Spine 1991;16(3 suppl): S25–32. [15] Geisler FH, Cheng C, Poka A, Brumback RJ. Anterior screw fixation of posteriorly displaced type II odontoid fractures. Neurosurgery 1989;25:30–7; discussion 37–8. [16] Jeanneret B, Magerl F. Primary posterior fusion C1/2 in odontoid fractures: indications, technique, and results of transarticular screw fixation. J Spinal Disord 1992;5:464–75. [17] Julien TD, Frankel B, Traynelis VC, Ryken TC. Evidence-based analysis of odontoid fracture management. Neurosurg Focus 2000;8:e1. [18] Maak TG, Grauer JN. The contemporary treatment of odontoid injuries. Spine 2006;31(11 suppl):S53–60; discussion S61. [19] Isolated fractures of the axis in adults. Neurosurgery 2002;50(3 suppl):S125–39. [20] Schiff DC, Parke WW. The arterial supply of the odontoid process. J Bone Joint Surg Am 1973;55:1450–6. [21] Heller JG, Alson MD, Schaffler MB, Garfin SR. Quantitative internal dens morphology. Spine 1992;17:861–6. [22] Schaffler MB, Alson MD, Heller JG, Garfin SR. Morphology of the dens. A quantitative study. Spine 1992;17:738–43. [23] Amling M, Hahn M, Wening VJ, Grote HJ, Delling G. The microarchitecture of the axis as the predisposing factor for fracture of the base of the odontoid process. A histomorphometric analysis of twenty-two autopsy specimens. J Bone Joint Surg Am 1994;76: 1840–6. [24] Amling M, Posl M, Wening VJ, Ritzel H, Hahn M, Delling G. Structural heterogeneity within the axis: the main cause in the etiology of dens fractures. A histomorphometric analysis of 37 normal and osteoporotic autopsy cases. J Neurosurg 1995;83:330–5. [25] Schatzker J, Rorabeck CH, Waddell JP. Fractures of the dens (odontoid process). An analysis of thirty-seven cases. J Bone Joint Surg Br 1971;53:392–405. [26] Dunn ME, Seljeskog EL. Experience in the management of odontoid process injuries: an analysis of 128 cases. Neurosurgery 1986;18: 306–10. [27] Lennarson PJ, Mostafavi H, Traynelis VC, Walters BC. Management of type II dens fractures: a case-control study. Spine 2000;25:1234–7. [28] Hadley MN, Browner C, Sonntag VK. Axis fractures: a comprehensive review of management and treatment in 107 cases. Neurosurgery 1985;17:281–90. [29] Apuzzo ML, Heiden JS, Weiss MH, Ackerson TT, Harvey JP, Kurze T. Acute fractures of the odontoid process. An analysis of 45 cases. J Neurosurg 1978;48:85–91.
Early predictive value of supine and upright X-ray films of odontoid fractures treated with halo-vest immobilization David H. Kim, MDa,*, Alexander R. Vaccaro, MDb, Jesse Affonso, MDc, Louis Jenis, MDa, Alan S. Hilibrand, MDb, Todd J. Albert, MDb a
Department of Orthopaedic Surgery, Tufts University Medical School, New England Baptist Hospital, 125 Parker Hill Avenue, Boston, MA 02120, USA b Department of Orthopaedic Surgery, Thomas Jefferson University and the Rothman Institute, 925 Chestnut Street, Philadelphia, PA 19107, USA c University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655, USA Received 16 January 2007; accepted 24 March 2007
Abstract
BACKGROUND: Although halo-vest immobilization remains a common form of treatment for type II odontoid fractures, nonunion and C1–2 instability may be the result in up to 20% to 40% of patients. PURPOSE: Supine and upright lateral X-ray films may allow early identification of patients likely to fail halo-vest treatment and earlier surgical treatment with decreased morbidity from prolonged unsuccessful halo-vest immobilization. STUDY DESIGN/SETTING: A prospective cohort study was performed. PATIENT SAMPLE: Twenty patients with type II odontoid fractures. OUTCOME MEASURES: Posttreatment nonunion and C1–2 instability as determined by plain X-ray films and computed tomography scan. METHODS: Both supine and upright lateral X-ray films were obtained immediately after halovest application and at the 2-week, 6-week, and 3-month follow-up. Flexion-extension lateral Xray films were obtained after halo-vest removal. Patients with nonunion or instability underwent computed tomography scan. Upright X-ray films were compared serially to identify loss of reduction. Pairs of supine and upright X-ray films were compared to measure any change in displacement or angulation upon transition from supine to upright position. Nonunion patients were compared with healed patients to determine any difference in fracture behavior based on serial supine and upright X-ray films. RESULTS: Twenty patients with type II odontoid fractures were identified during the study period. Three patients with multiple trauma underwent immediate surgical stabilization. Three elderly patients with nondisplaced fractures were treated in a cervical orthosis. Fourteen patients initiated and completed 3 months of halo-vest immobilization. After halo-vest removal, 4 of 14 patients (29%) showed radiographic nonunion or instability. In all 4 nonunion patients, supine and upright radiographs at 2 weeks revealed change in fracture angulation $5 between the supine and upright positions. In three of these patients standard serial upright lateral X-ray films failed to identify any loss of reduction. In the remaining patient, progressive angulation of 15 was observed. No measurable change in angulation between supine and upright X-ray films was observed in any patient who healed successfully with halo-vest treatment. CONCLUSIONS: Obtaining both supine and upright lateral X-ray films during the follow-up period may identify patients at risk for failure of halo-vest treatment as early as 2 weeks after initiation of treatment. A change in fracture angulation $5 suggests an increased risk of treatment
FDA device/drug status: approved for this indication (halo vest). Nothing of value received from a commercial entity related to this manuscript. 1529-9430/08/$ – see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.spinee.2007.03.009
* Corresponding author. David H. Kim, MD, 125 Parker Hill Ave, Boston, MA 02120. Tel.: (617) 754-6482; fax: (617) 754-5595. E-mail address: [email protected] (D.H. Kim)
D.H. Kim et al. / The Spine Journal 8 (2008) 612–618
613
failure and the potential benefit of early surgical stabilization. Ó 2008 Elsevier Inc. All rights reserved. Keywords:
Odontoid fracture; Halo vest; Evaluation; Nonunion
Introduction Cervical spine odontoid fractures are frequently observed, constituting approximately 10% to 18% of all clinically significant cervical spine injuries in adults [1]. The most commonly used classification system, by Anderson and D’Alonzo [2], defines three basic fracture patterns. Type II fractures occur through the junction of the odontoid process with the C2 vertebral body and represent the most common type of fracture as well as the greatest challenge to the treating clinician. High rates of nonunion and instability have plagued all forms of nonoperative treatment and have resulted in a trend toward early surgical stabilization in many cases. Nevertheless, despite the increasing number of surgical techniques available, halo-vest immobilization remains the most frequently selected form of treatment in many centers. The principal advantages of halo-vest treatment are the potential avoidance of surgery and the perception that it represents a ‘‘noninvasive’’ form of treatment. Unfortunately, halo vests have also been associated with a high rate of complications including pin loosening, pin site infection, intracranial pin penetration with abscess formation, poor cosmetic scarring, and postimmobilization neck stiffness. Halo vests are often poorly tolerated by elderly patients who may suffer skin breakdown or aspiration. Nonunion and cervical spine instability with persistent pain and risk of spinal cord injury remain the major concern after treatment of type II fractures with halo-vest immobilization. Published series have reported nonunion rates of 18% to 41% and have suggested numerous possible risk factors, including fracture displacement, angulation, distraction, comminution, delay in diagnosis, and patient age [1,3–6]. Occasionally, loss of reduction is observed on serial X-ray films after halo-vest application and serves as an early indicator of treatment failure. More typically, failure of halo-vest treatment is not recognized until flexion-extension X-ray films or computed tomographic imaging is obtained at the conclusion of 12 to 15 weeks of immobilization. Anderson et al. [7] have shown that halo vests are relatively less effective at immobilizing the upper cervical spine compared with the mid- and lower cervical spine. Comparing upright and supine lateral X-ray films in patients in halo vests has revealed varying rates of motion that is more pronounced in the upper cervical spine. The study investigators hypothesized that obtaining both upright and supine lateral X-ray films in patients undergoing halo-vest treatment for type II odontoid fractures may allow earlier recognition of fractures at increased risk for treatment failure. If true, early surgical stabilization or conversion to
alternative treatment in these patients may prevent additional morbidity from prolonged halo-vest immobilization.
Materials and methods Study population An application for the following study was submitted and approved by the hospital institutional review board. During the period of January 2003 through December 2004, prospective data were collected on 20 consecutive individuals diagnosed with acute type II odontoid fractures. All patients underwent evaluation and treatment within 48 hours at a level I trauma center. Fourteen patients were neurologically intact and elected to undergo treatment with halo-vest immobilization. Patient age ranged from 18 to 67 years with an average of 43 years. Twelve patients had isolated injuries, whereas two patients were considered polytrauma, both with associated upper-extremity fractures. Motor vehicle accidents (nine patients) and falls (five patients) were the most frequent mechanisms of injury. Six fractures were displaced or angulated and required closed reduction before halo-vest application. During the study period, three patients with severe polytrauma were considered unsuitable for halo-vest immobilization and treated with immediate surgical stabilization. Three elderly and debilitated patients elected to forego treatment with either halovest immobilization or surgery and were treated with hard collar immobilization. Halo-vest treatment After placement of the halo ring, fractures displaced greater than 5 mm or angulated greater than 10 underwent a closed reduction before attachment of the halo vest. Six patients required initial closed reduction, with two patients requiring multiple reduction attempts. No patient during the study period experienced redisplacement once successful closed reduction had been accomplished. After halo-vest application, all patients underwent immediate radiographic imaging consisting of an open-mouth odontoid view and both supine and upright lateral cervical spine X-ray films. Halo pins were retorqued according to manufacturers’ recommendations; and patients and their caretakers were instructed in routine pin-site and halo-vest management before discharge. Radiographic and clinical evaluation All patients underwent routine follow-up evaluations at 2 weeks, 6 weeks, and 3 months after halo-vest application.
614
D.H. Kim et al. / The Spine Journal 8 (2008) 612–618
Patients were interviewed regarding any changes in their pain experience or neurological status. Clinical evaluation included pin-site and halo-vest inspection as well as neurological examination. Radiographic evaluation at each scheduled follow-up visit included an open-mouth odontoid and supine and upright lateral cervical spine views. Before study onset, the decision was made that fracture displacement greater than 5 mm would represent failure of treatment and an indication for surgical stabilization, but this degree of displacement was not observed in any patient. All patients completed 3 months of halo-vest immobilization. After halo-vest removal, all patients underwent lateral flexion and extension cervical spine X-ray films. If X-ray films revealed a minimum of 30 flexion and 30 extension without evidence of measurable displacement and patients reported no significant pain through this range of motion, the fracture was considered healed, and patients were allowed to experience gradual return to normal activities with a short period of soft collar wear and supervised physical therapy on an as-needed basis. If patients were unable to perform flexion and extension through an adequate range of motion or if significant pain was reported during performance of this maneuver, a helical computed tomography (CT) study of the cervical spine was obtained with 1.5mm thin cuts including sagittal and coronal reconstructed images. Radiographic measurements The odontoid process presents an irregular anatomic profile on lateral radiographic images, and no consistent fracture displacement or angulation measurement technique has been universally accepted. The posterior border of the odontoid is typically more linear, but a small degree of rotation can create the appearance of two posterior cortical lines at the level of the centrum. The portion of the odontoid proximal to the atlas ring is often poorly visualized, particularly in elderly individuals with osteopenia and spondylotic degeneration. To determine the most reliable measurement methods for the purposes of this study, fracture displacement and angulation were measured by three different independent physicians using two different measurement techniques. In the anterior method, a visual ‘‘best-fit’’ line was drawn along the anterior border of the odontoid process proximal to the fracture line. A second ‘‘best-fit’’ line was drawn along the anterior border of the axis body. A third line was drawn across the fracture plane defined by the points at which the fracture crossed the anterior and posterior cortices. If these points were poorly visualized on plain X-ray films, the points were extrapolated from the original reconstructed CT images from the time of initial evaluation. Displacement was defined by the distance along the third line from the first line to the second. Angulation was measured as the angle subtended by the first two lines.
In the posterior measurement method, a visual ‘‘best-fit’’ line was drawn along the posterior border of the odontoid process proximal to the fracture line. A second ‘‘best-fit’’ line was drawn along the posterior border of the axis body. The third line was again drawn across the fracture plane as described previously for the anterior method. Displacement and angulation were measured by using the same technique using the new lines. No significant difference was identified between the anterior and posterior measurement methods. Raters expressed the impression that assessment of displacement was easier with the anterior method, whereas assessment of angulation was easier with the posterior method. Therefore, for the purposes of this study, the anterior method was selected to report displacement values, whereas the posterior method was used to report angulation values; 5 was selected arbitrarily as the smallest increment that all raters were confident as representative of a true change in fracture angulation. Statistical analysis Statistical evaluation of the results was performed using the Fisher exact test and the binomial test. Significance was established as a p value #.05.
Results Clinical results During the 2-year study period, 14 patients met study criteria and were included in the study population. All 14 patients successfully completed 3 months of halo-vest immobilization. Five patients developed superficial pin site infections requiring treatment with oral antibiotics. In one patient, persistent unilateral scalp pain in the distribution of the supraorbital nerve led to repositioning of an anterior pin. Seven patients reported varying levels of skin irritation from the vest, but no patient developed clinically significant skin breakdown. There were no episodes of aspiration or pulmonary complications. Serial comparisons of upright X-ray films failed to show progressive displacement in any patient. No patient showed fracture displacement greater than 3 mm on any radiographic view. One patient experienced an incremental progressive posterior angulation of 15 over the course of 3 months without fracture displacement. All patients remained neurologically intact throughout the course of the study. After halo-vest treatment, 4 of 14 patients (29%) developed radiographic nonunion or instability, including the 1 patient with progressive angulation. All four patients reported persistent neck pain. Two patients showed instability on flexion and extension X-ray films, whereas the remaining 2 patients were unable to perform adequate sagittal
D.H. Kim et al. / The Spine Journal 8 (2008) 612–618
range of motion and were diagnosed on the basis of CT imaging. Radiographic results In 13 of 14 study patients, including 3 of 4 patients who later developed nonunions, an evaluation of open-mouth odontoid views and a comparison of supine and upright lateral radiographs immediately after halo-vest application failed to identify measurable displacement or change in angulation. One patient who later developed nonunion showed a 12 change in angulation. For statistical analysis, patients were divided into two groups. Group 1 included the 10 patients who were treated successfully by halo-vest immobilization. Group 2 included the 4 patients who developed a nonunion. Two weeks after halo-vest application, standard upright open-mouth views as well as upright lateral X-ray films did not show measurable fracture displacement or a change in angulation in any patient. A comparison of supine and upright lateral X-ray films at 2 weeks, however, revealed a change in angulation of $5 between the supine and upright positions in all 4 patients in group 1 and none of the patients in group 2 (Table 1). This difference was significant (p5.001); 95% confidence intervals cannot be calculated for this situation. However, the lower level for the 95% odds ratio equals 11.07. The upper level cannot be calculated but is by inference infinity. At 6 weeks follow-up, no measurable fracture displacement was observed in any patient. One patient in group 2 showed an increase in fracture angulation of 9 compared with the 2-week X-ray film on upright lateral imaging. A comparison of supine and upright lateral radiographs showed $5 angulation in 3 of 4 patients in group 2 and 0 of 10 patients in group 1. This difference was not significant (p5.07). At the 3-month follow-up, no measurable displacement was observed in any patient. The group 2 patient with increased fracture angulation at 6 weeks shwed further angulation of 6 on upright lateral imaging. This represented a total change of 15 compared with the 2-week X-ray film. A comparison of supine and upright lateral X-ray films showed $5 angulation in 2 of 4 group 2 patients and 0 of 10 group 1 patients. Again, this difference was not significant (p5.25). Although fracture angulation or displacement between supine and upright radiographic views was the primary
615
variable of interest, secondary analysis was performed comparing other potential risk factors between groups (Table 2). No significant difference was identified in terms of age, smoking status, initial fracture displacement, or angulation or direction of displacement. During the study period, significant fracture comminution was considered an indication for early surgical stabilization. Therefore, there were no patients in the final study population with significant fracture comminution.
Discussion The fundamental principles of treatment for odontoid fractures are widely accepted and include early diagnosis, anatomic reduction, and adequate immobilization to permit healing. Nevertheless, there exists no consensus regarding the ideal form of treatment in most cases. A variety of treatment strategies have been advocated, ranging from initial nonoperative treatment in all patients, to early surgery in selected high-risk populations, to early surgery in all patients [1–3,6,8–16]. Nonoperative treatment recommendations have included halo-vest, cervicothoracic orthosis, and hard-collar immobilization. An evidence-based review of the literature indicates that no standards or guidelines exist for the treatment of type II odontoid fractures, only various treatment options [17–19]. Despite an increasing number of surgical options for stabilization of acute type II odontoid fractures and high reported rates of nonunion after all forms of external immobilization, halo vests remain a frequently selected form of treatment. Several clinical series have reported relatively high complication and failure rates associated with halo-vest treatment. Seybold and Bayley [4] reported a 20% nonunion rate as well as increased complications and problematic postimmobilization stiffness in elderly patients. Stoney et al. [5] reported an 18% nonunion rate and 40% rate of moderate to severe stiffness after halo-vest treatment. Although previous investigators have blamed high nonunion rates on poor vascular supply, a detailed anatomical study has shown a robust anastomotic arterial arcade from both carotid and vertebral systems, which includes both endosteal and ligamentous vessels [20]. Delayed healing or nonunions appear more likely because of the unique histomorphometry of bone at the odontoid base characterized by thin cortical bone and a relatively weak trabecular
Table 1 Change in fracture angulation and displacement between supine and upright tray firms Healed fractures (group 1) Angulation $5 Displacement $2 mm * p!.05.
Nonunion (group 2)
0 wk
2 wk
6 wk
3 mo
0 wk
2 wk
6 wk
3 mo
0/10 0/10
0/10 0/10
0/10 0/10
0/10 0/10
1/4 0/4
4/4* 0/4
3/4 0/4
2/4 0/4
616
D.H. Kim et al. / The Spine Journal 8 (2008) 612–618
Table 2 Nonunion risk factor characteristics of study population
Age (range) Active smokers Fracture initially displaced (requiring reduction) Direction of displacement (anterior/posterior) Fracture initially angulated (requiring reduction) Fracture comminution (%)
Healed fractures (group 1)
Nonunion (group 2)
p value for difference
45 (18–67) 3 of 10 3 of 10
39 (19–65) 3 of 4 3 of 4
Not significant Not significant Not significant
2/1
1/2
Not significant
1 of 10
2 of 4
Not significant
0
0
pattern [21–24]. Cross-sectional analysis has revealed this bone to be mostly cortical in nature [16–19]. Clinical studies have suggested various risk factors for nonunion after type II odontoid fractures, including significant fracture displacement or angulation, posterior displacement, fracture site comminution, patient age, and delay in diagnosis [1,3,4,25–27]. Fracture displacement has been the most commonly identified risk factor for nonunion regardless of treatment modality [1,25,26,28,29]. Consequently, a frequently used strategy for following odontoid fracture patients in halo vests is obtaining serial plain X-ray films in the upright lateral position. Although comparing films performed in the same position facilitates identification of interval fracture displacement, this method may miss persistent instability that occurs when the patient transfers between supine and standing positions. Anderson et al. [7] originally reported the potential for significant fracture-site motion between supine and upright positioning despite halo-vest immobilization in a prospective study of 42 patients with various cervical spine fractures. Five days after injury, the study found an average of 7 angular motion and 1.7-mm translation when moving between the two positions. The greatest amount of motion was observed proximally between the occiput and C1. The rate of nonunion in this study (29%) is in line with previous published reports after halo-vest treatment of odontoid fractures of 18% to 41% [5,6,28]. In this study, when supine and upright X-ray films were compared, an increased rate of fracture angulation was observed at all time points in a majority of patients who eventually developed nonunion (Fig. 1). This angular instability was most pronounced at the 2-week follow-up evaluation when all patients showed angulation $5 . Therefore, this may be the most sensitive time point for identifying patients at increased risk of nonunion using this modality. Of interest, only one of four nonunion patients showed a measurable change in angulation between supine and upright views immediately after halo-vest application. The authors have observed significant cervical muscle spasm in many patients in the acute postinjury period, and it is possible that muscle guarding results in early splinting and pseudostabilization of the upper cervical spine, preventing
immediate detection of fracture instability using supine and upright X-ray films. A trend was also noted toward decreasing amounts of fracture angulation in nonunion patients at the 6-week and 3-month follow-up. In general, neck pain complaints tended to diminish over time, and only one patient with nonunion reported persistent pain throughout the treatment period. The authors speculate that even in the absence of solid osseous healing some degree of soft-tissue healing and fibrous-tissue formation or fracture settling and fragment interdigitation may occur, resulting in cases of pseudostabilization and reduction in neck pain. Of note, no measurable fracture translation was observed in any nonunion patient when comparing supine and upright X-ray films or serial X-ray films over time. This suggests that fracture angulation may be a more sensitive predictor of odontoid fracture instability when following the course of healing with halo-vest treatment. This study suggests that obtaining both supine and upright lateral X-ray films during follow-up can identify odontoid fracture patients at risk for failure of halo-vest treatment as early as 2 weeks after initiation of treatment. This finding may allow earlier surgical treatment of these patients either with anterior screw osteosynthesis or posterior C1–2 instrumented fusion with more rapid recovery and avoidance of unnecessary morbidity from prolonged halo-vest immobilization. The current practice of many specialists is to obtain follow-up anteroposterior and lateral radiographic views in the upright position only. We believe that obtaining both supine and upright lateral X-ray films during follow-up of type II odontoid fractures, particularly at around 2 weeks postinjury, should be considered. A change in angulation can occur between supine and upright position despite no detectable change in displacement or angulation on serial upright X-ray films. Observation of this change in angulation suggests increased risk of failure of halo-vest treatment and may be used to identify patients who might benefit from early surgical stabilization. An obvious limitation of this study is a very small sample size. However, because the positive predictive value of supine and upright X-ray films was so high in terms of predicting nonunion in this prospective series, statistical analysis appropriate for small samples suggests that the associations are significant. Although exact 95% confidence intervals cannot be calculated for the data, the 95% confidence interval for the odds ratio can be determined and suggests that the presence of $5 angular change between supine and upright X-ray films is associated with at least an 11-fold increase in the risk of developing a nonunion. Nevertheless, a larger study would be required to confirm these findings. Another limitation is the determination of fracture healing that was used in this study. We used a standard protocol to assess whether adequate stability had been achieved after the completion of halo-vest treatment. The absence of neck pain and the absence of C1–2 instability or fracture displacement through a minimum of 30 of flexion and 30
D.H. Kim et al. / The Spine Journal 8 (2008) 612–618
617
Fig. 1. (A) Upright and (B) supine lateral cervical spine X-ray films 2 weeks after a halo-vest application in a 37-year-old patient with a type II odontoid fracture. There is a 10 increase in fracture angulation in the supine position. (C) This patient eventually developed a painful unstable nonunion after completion of 3 months of halo-vest treatment and required surgical stabilization.
618
D.H. Kim et al. / The Spine Journal 8 (2008) 612–618
of extension on lateral X-ray films were considered satisfactory criteria to determine that the injury had healed. If patients reported pain or could not perform adequate flexion-extension range of motion, a CT scan was obtained to assess the degree of osseous healing. Although this is a common protocol used by many spine surgeons, there may be a number of patients who appear stable based on flexion-extension X-ray films but who would not show solid osseous healing if a CT scan had been obtained. These patients may be considered to have a ‘‘stable fibrous nonunion,’’ and, although there is no consensus regarding the best treatment in this setting, observation without specific intervention appears a reasonable option. The findings of this study, therefore, are specific for predicting a painful or unstable nonunion as opposed to a strict definition of osseous healing. References [1] Clark CR, White AA 3rd. Fractures of the dens. A multicenter study. J Bone Joint Surg Am 1985;67:1340–8. [2] Anderson LD, D’Alonzo RT. Fractures of the odontoid process of the axis. J Bone Joint Surg Am 1974;56:1663–74. [3] Hadley MN, Browner CM, Liu SS, Sonntag VK. New subtype of acute odontoid fractures (type IIA). Neurosurgery 1988;22:67–71. [4] Seybold EA, Bayley JC. Functional outcome of surgically and conservatively managed dens fractures. Spine 1998;23:1837–45; discussion 1845–6. [5] Stoney J, O’Brien J, Wilde P. Treatment of type-two odontoid fractures in halothoracic vests. J Bone Joint Surg Br 1998;80:452–5. [6] Ekong CE, Schwartz ML, Tator CH, Rowed DW, Edmonds VE. Odontoid fracture: management with early mobilization using the halo device. Neurosurgery 1981;9:631–7. [7] Anderson PA, Budorick TE, Easton KB, Henley MB, Salciccioli GG. Failure of halo vest to prevent in vivo motion in patients with injured cervical spines. Spine 1991;16(10 suppl):S501–5. [8] Crockard HA, Heilman AE, Stevens JM. Progressive myelopathy secondary to odontoid fractures: clinical, radiological, and surgical features. J Neurosurg 1993;78:579–86. [9] Graziano G, Jaggers C, Lee M, Lynch W. A comparative study of fixation techniques for type II fractures of the odontoid process. Spine 1993;18:2383–7. [10] Apfelbaum RI, Lonser RR, Veres R, Casey A. Direct anterior screw fixation for recent and remote odontoid fractures. J Neurosurg 2000;93(2 suppl):227–36. [11] Bohler J. Anterior stabilization for acute fractures and non-unions of the dens. J Bone Joint Surg Am 1982;64:18–27.
[12] Borne GM, Bedou GL, Pinaudeau M, Cristino G, Hussein A. Odontoid process fracture osteosynthesis with a direct screw fixation technique in nine consecutive cases. J Neurosurg 1988;68: 223–6. [13] Esses SI, Bednar DA. Screw fixation of odontoid fractures and nonunions. Spine 1991;16(10 suppl):S483–5. [14] Etter C, Coscia M, Jaberg H, Aebi M. Direct anterior fixation of dens fractures with a cannulated screw system. Spine 1991;16(3 suppl): S25–32. [15] Geisler FH, Cheng C, Poka A, Brumback RJ. Anterior screw fixation of posteriorly displaced type II odontoid fractures. Neurosurgery 1989;25:30–7; discussion 37–8. [16] Jeanneret B, Magerl F. Primary posterior fusion C1/2 in odontoid fractures: indications, technique, and results of transarticular screw fixation. J Spinal Disord 1992;5:464–75. [17] Julien TD, Frankel B, Traynelis VC, Ryken TC. Evidence-based analysis of odontoid fracture management. Neurosurg Focus 2000;8:e1. [18] Maak TG, Grauer JN. The contemporary treatment of odontoid injuries. Spine 2006;31(11 suppl):S53–60; discussion S61. [19] Isolated fractures of the axis in adults. Neurosurgery 2002;50(3 suppl):S125–39. [20] Schiff DC, Parke WW. The arterial supply of the odontoid process. J Bone Joint Surg Am 1973;55:1450–6. [21] Heller JG, Alson MD, Schaffler MB, Garfin SR. Quantitative internal dens morphology. Spine 1992;17:861–6. [22] Schaffler MB, Alson MD, Heller JG, Garfin SR. Morphology of the dens. A quantitative study. Spine 1992;17:738–43. [23] Amling M, Hahn M, Wening VJ, Grote HJ, Delling G. The microarchitecture of the axis as the predisposing factor for fracture of the base of the odontoid process. A histomorphometric analysis of twenty-two autopsy specimens. J Bone Joint Surg Am 1994;76: 1840–6. [24] Amling M, Posl M, Wening VJ, Ritzel H, Hahn M, Delling G. Structural heterogeneity within the axis: the main cause in the etiology of dens fractures. A histomorphometric analysis of 37 normal and osteoporotic autopsy cases. J Neurosurg 1995;83:330–5. [25] Schatzker J, Rorabeck CH, Waddell JP. Fractures of the dens (odontoid process). An analysis of thirty-seven cases. J Bone Joint Surg Br 1971;53:392–405. [26] Dunn ME, Seljeskog EL. Experience in the management of odontoid process injuries: an analysis of 128 cases. Neurosurgery 1986;18: 306–10. [27] Lennarson PJ, Mostafavi H, Traynelis VC, Walters BC. Management of type II dens fractures: a case-control study. Spine 2000;25:1234–7. [28] Hadley MN, Browner C, Sonntag VK. Axis fractures: a comprehensive review of management and treatment in 107 cases. Neurosurgery 1985;17:281–90. [29] Apuzzo ML, Heiden JS, Weiss MH, Ackerson TT, Harvey JP, Kurze T. Acute fractures of the odontoid process. An analysis of 45 cases. J Neurosurg 1978;48:85–91.