| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Harborview Medical Center, University of Washington
School of Medicine, 325 Ninth Ave., Box 359728, Seattle, WA 98104.
2 Present address: Department of Radiology, Austin and Repatriation Medical
Center, Austin Campus, Studley Rd., Heidelberg 3084, Australia.
3 Present address: One Union Square S., #20L, New York, NY 10003.
Received August 9, 2001;
accepted after revision November 8, 2001.
Address correspondence to F. A. Mann.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. We conducted a retrospective review of the findings on conventional radiography, CT, and MR imaging in 95 patients with 107 occipital condyle fractures. We described fracture patterns according to two previously published classification systems. Clinical findings, neurosurgical management, and patient outcome were obtained from the medical records.
RESULTS. Inferomedial avulsions (Anderson and Montesano type III) were the most common type of occipital condyle fracture, constituting 80 (75%) of 107 overall fractures. Unilateral occipital condyle fractures were found in 73 (77%) of 95 patients, and 58 patients were treated nonoperatively; occipitocervical fusion was required in nine patients for complex C1-C2 injuries, and six patients died. Bilateral occipital condyle fractures or occipitoatlantoaxial joint injuries were seen in 22 (23%) of 95 patients. Occipitocervical fusion or halo traction for the craniocervical junction was required in 12 patients, all of whom had CT evidence of bilateral occipitoatlantoaxial joint disruption and six of whom showed normal craniocervical relationships on conventional radiographs. Six patients with nondisplaced fractures were treated nonoperatively, and four patients died. Thirty (32%) of 95 patients showed continued disability, whereas 55 (57.5%) of 95 patients had good outcomes at 1 month. Associated cervical spine injuries were present in 29 (31%) of 95 patients.
CONCLUSION. Given their associated traumatic brain and cervical spine injuries, occipital condyle fractures are markers of high-energy traumas. That conventional radiographs alone may miss up to half of the patients with acute craniocervical instability has not been well established. Avulsion fracture type and fracture displacement are associated with both injury mechanism and the need for surgical stabilization. In this series, most unilateral occipital condyle fractures were treated nonoperatively, whereas bilateral occipitoatlantoaxial joint injuries with findings of instability usually required surgical stabilization.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
In an autopsy study, Saternus [4] was the first to propose a fracture system for the occipital condyle based on of the form of applied strain. The most widely used radiologic classification was described in six patients by Anderson and Montesano [13] (Table 1), who considered fracture morphology, pertinent anatomy, and biomechanics. According to this system, fractures are classified as follows: type I, comminuted impaction fracture due to axial loading (Fig. 2); type II, skull base fracture that extends through the occipital condyle (Fig. 3); and type III, avulsion fracture mediated through tension in the alar ligament (Fig. 4A,4B). Type III injuries may be associated with disruption of the alar ligaments and tectorial membrane and result in craniocervical dissociation [5, 12] (Fig. 5). More recently, Tuli et al. [14] reviewed the published cases of occipital condyle fracture and proposed the following new classification system (Table 2): type 1, nondisplaced occipital condyle fracture (stable); type 2A, displaced occipital condyle fracture with intact ligaments (stable); and type 2B, displaced occipital condyle fracture with radiographic evidence of craniocervical junction instability. Tuli et al. posited that their classification can guide neurosurgical management: type 2B fractures require surgical instrumentation or halo traction, whereas type 2A injuries may be treated with a rigid collar, and type 1 injuries require no specific treatment.
|
|
|
|
|
|
|
The published radiology data are limited to small case series [8, 13,14,15,16,17,18], and little information is available about the range of appearance or clinical significance of these injuries. We therefore aimed to review the imaging features and clinical findings of a large series of patients with occipital condyle fractures to determine the distribution of occipital condyle fractures according to current classifications and to correlate imaging findings with neurosurgical treatment and patient outcome.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All patients underwent cross-table lateral cervical spine radiography and CT of the craniocervical junction. Alignment of the craniocervical junction was evaluated from cross-table lateral radiographs using the basion-dental and basion-axial interval methods described by Harris et al. [19, 20]. Dens to basion distances were considered abnormal if greater than 12 mm. Basion-to-posterior axial line distances were considered abnormal if the basion was greater than 12 mm anterior or greater than 4 mm posterior to the posterior axial line. Standard CT technique comprised 1- to 1.5-mm axial collimation helical or incremental mode scans (CTi HiLight; General Electric Medical Systems, Milwaukee, WI) from the mid clivus to C3. The occipitoatlantal and atlantoaxial articulations were evaluated on coronal and sagittal CT reformations.
MR imaging of the cervical spine was performed in 37 patients to characterize cord injury or to better assess the patient with inconclusive CT for suspected ligament disruption (1.5-T Signa; General Electric Medical Systems). Standard examination protocol included T1-, T2-, and short tau inversion recovery or fat-suppressed T2-weighted sagittal sequences. Images were assessed for prevertebral and nuchal ligament edema, increased occipitoatlantal and atlantoaxial joint spaces, and presence and location of hemorrhage and cord edema. No attempt was made to assess the integrity of the major extra- or intracranial arteries.
Occipital condyle fractures were categorized according to systems developed by Anderson and Montesano [13] and Tuli et al. [14]. The Anderson and Montesano classification system is based on fracture shape and position and direction of displacement as shown on CT. Using the Tuli classification, we defined fracture displacement to correspond to at least 2 mm of osseous separation. Assessment of craniocervical junction stability was based on basion—dental and basion—axial interval measurements from lateral cervical spine radiographs [19, 20], CT evidence of bilateral occipitoatlantoaxial joint disruption, or presence of ligament injury on cervical spine MR imaging [21].
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
In only one patient was the occipital condyle fracture directly visualized on the lateral cervical spine radiograph. No patient with completely normal findings on lateral radiographs, including normal prevertebral soft-tissue contours, had an unstable occipital condyle fracture. Twenty-eight (29%) of 95 patients had additional cervical spine fractures or ligament or cord injuries. These were localized to the atlantoaxial region in 15 (54%) of 28 patients, the mid or lower cervical region in nine (32%) of 28 patients, and both cervical segments in four (14%) of 28 patients. Atlantoaxial segment fractures included C1 Jefferson's fracture, n = 3; C1 lateral mass, n = 2; C1 anterior or posterior arch, n = 3; C1 superior articular surface, n = 4; odontoid type I, n = 1; odontoid types II and III, n = 5; and C2 pars interarticularis or lateral mass, n = 6. One patient sustained a cervical cord injury without other fractures. Sixty (63%) of 95 patients showed CT evidence of diffuse head injury or focal intracranial hematomas. Twenty-seven patients (28%) had normal findings on head CT scans, and the remaining eight patients had no clinical evidence of head injury.
Injury mechanism was a motor vehicle crash in 56 patients, pedestrian versus vehicle crash in 10, cycling crash in five, fall in 17, assault in five, and other mechanism in two. Ten (10.5%) of 95 patients died in the hospital from multiple injuries without definitive treatment for their occipital condyle fracture, and four of these patients had autopsy or radiographic evidence of complete occipitoatlantal dissociation. Eight patients (8.5%) underwent posterior instrumentation with occipitocervical fusion, and four patients (4%) required halo traction to maintain alignment of the craniocervical junction. A further 10 patients (10.5%) with associated C1-C7 cervical spine injuries were treated by cervical fusion or halo traction. Fifty-four patients (57%) were successfully treated conservatively with immobilization by cervical collar or brace, and six (6.5%) of 95 patients received no specific treatment. Three patients (3%) required surgery to correct displaced impacted skull base fractures.
Thirty (35%) of 85 patients who survived to discharge had a poor outcome at 1 month, requiring continued nursing support, tube feeding, or tracheostomy care; 12 of these patients subsequently gained functional independence with rehabilitation. The remaining 55 (65%) of 85 patients had good outcomes and were independently mobile and self-caring within 1 month. Of 35 patients (91%) without a head injury, 32 showed good recovery and were independent of nursing care by 1 month.
We defined bilateral injury to the occipitoatlantoaxial joint complex as the presence of either bilateral occipital condyle fractures or unilateral occipital condyle fracture with contralateral widening of the occipitoatlantal (>2 mm) or atlantoaxial (>3 mm) joints determined from sagittal and coronal CT reformations (Figs. 4A,4B, 5, and 7A,7B). On the basis of these criteria, unilateral occipital condyle fractures were found in 73 patients, 58 (79%) of whom were successfully treated with a cervical brace or collar. Bilateral occipitoatlantal injury was seen in 22 (23%) of 95 patients, four of whom died and 12 (55%) of whom were treated by occipitocervical fusion or halo traction for the cervicocranium. However, only six (50%) of these 12 patients showed definite criteria on conventional radiographics for occipitoatlantal dissociation on lateral cervical spine radiographs [19, 20]. Three patients with crescent occipital condyle fractures extending to the basiocciput had clinically unstable injuries despite normal basion—dental and basion—axial interval measurements (Fig. 8A,8B,8C,8D).
|
|
|
|
|
|
MR imaging showed 16 (38%) of 42 occipital condyle fractures in 37 patients who were involved in this investigation. Prevertebral or nuchal ligament edema and hemorrhage were visible in 27 (73%) of 37 patients. Extradural or subdural hemorrhage was present at the foramen magnum in 11 (30%) of 37 patients, and cord edema or hemorrhage was present in nine (24%) of 37 patients. Injuries to the alar ligaments and tectorial membrane could not be reliably diagnosed using our standard imaging protocols.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Occipital condyle fracture is thus a rare injury, and the radiology literature is limited to small case series [8, 13,14,15,16,17,18]. In 1988, Anderson and Montesano [13] described their classification system in six patients who had occipital condyle fractures. Four of these patients had avulsion (type III) fractures. The largest clinical series to date presented 15 patients with occipital condyle fractures, eight of whom had extension of skull base fractures through the occipital condyle (type II) and five of whom had avulsion injuries (type III) [18]. Another study of 55 consecutive patients with high-risk head injuries who underwent screening craniocervical junction CT found 11 occipital condyle fractures in nine patients [8].
We believe that the relatively large number of occipital condyle fractures in our study is explained by the interstate referral practice and the rigorous institutional policy for imaging the craniocervical junction. During the early 1990s, the initial screening evaluation of the craniocervical junction in this patient group was performed on coronal tomography supplemented by thin-section CT. Subsequently, the use of routine craniocervical junction CT in patients requiring head CT became standard practice. In 1997, we extended our institutional CT screening strategy for high-risk patients to include the entire cervical and upper thoracic spine.
Using the Anderson and Montesano classification system [13], we found type III avulsion occipital condyle fractures to be the most common (80/107, 75%) and type I impaction occipital condyle fractures to be the most unusual. However, the distinction between a noncomminuted type I and a nondisplaced type III fracture may be difficult. Applying the Tuli et al. classification system [14] to our patients, we found that type 1 nondisplaced occipital condyle fractures are the most frequent, with displaced unstable type 2B fractures seen in 18 (19%) of 95 patients. Assessment of craniocervical junction stability in the severely injured patient with multiple fractures is problematic. Definitive imaging evaluation by dynamic radiography or MR imaging must often be deferred while spinal immobilization is maintained and other life-threatening conditions are addressed. Bloom et al. [8] have suggested that alar ligament thickening, stretching, or discontinuity is a useful, direct CT sign of disruption. However, we cannot confidently diagnose the integrity of the alar ligaments using these criteria because of the frequent presence of minor positional asymmetry.
We believe that subdividing Anderson and Montesano type III fractures [13] into two groups—stable and unstable—could effectively combine the principles of both classification systems (Table 3). We suggest that the CT finding of bilateral occipitoatlantoaxial joint complex injury (defined as either bilateral occipital condyle fractures or unilateral occipital condyle fracture with contralateral widening of the occipitoatlantal [>2 mm] or atlantoaxial [>3 mm] joint) be used as a marker for instability (Figs. 4A,4B, 5, and 7A,7B). This association between occipital condyle fractures and craniocervical instability underscores the importance of considering the occipital condyle fracture as a potential component of a more extensive injury to the occipitoatlantoaxial joint complex.
|
Craniocervical junction CT is readily performed in critically injured patients and can provide a basis for occipital condyle fracture characterization and treatment planning. The neurosurgical management of these rare injuries is tailored to individual patient circumstances and is not purely based on radiologic criteria. Additional important considerations include the presence of associated cervical spine injuries and the neurologic prognosis.
Occipital condyle fractures typically occur in association with multiple injuries sustained from high-energy blunt trauma, although a range of presentations, clinical significance, and outcomes were observed in this study. The presence and severity of head injury were the main determinants of outcome. Isolated unilateral occipital condyle fractures without head injury were usually, but not universally, stable with good recovery. All patients shown to have bilateral occipital condyle fractures or occipitoatlantal or atlantoaxial joint space widening required operative stabilization (occipitocervical fusion or halo traction). This finding supports the hypothesis that disruption of all cranio-cervical ligaments is required for complete occipitoatlantal dissociation [10, 11], and that the intact contralateral alar ligament and tectorial membrane continue to stabilize a unilateral condyle avulsion fracture. However, recent experimental work suggests that even partial resection of one occipital condyle may result in altered biomechanics [24].
The anterior extension of an occipital condyle avulsion fracture around the foramen magnum is rarely reported in the radiology literature [22, 23] (Fig. 6A,6B), although ring and partial ring fractures of the skull base are described in autopsy studies [25]. We observed extension to the basiocciput in 21 (26%) of 80 type III occipital condyle fractures and in such cases found the alignment of the craniocervical junction difficult to assess from conventional lateral cervical spine radiographs (Fig. 8A,8B,8C,8D). In such cases, the basion—dental and basion—axial intervals may be normal in the presence of craniocervical dissociation [19, 20].
The large number (29/95, 31%) of patients with additional cervical spine injuries is not surprising given the number of motor vehicle crashes and frequency of head injuries [9]. The 15 cases of associated C1 or C2 fracture can be considered part of the spectrum of the occipital condyle fracture. In four patients, the presence of a fracture of either the anterior or posterior margin of the superior articular surface of the lateral mass of C1 led us to postulate that rotation was an important contributing injury mechanism.
Emergent lateral cervical spine radiography has a limited role in the detection of occipital condyle fractures. Although no unstable occipital condyle fracture was present when craniocervical alignment and prevertebral soft tissues were completely normal, endotracheal intubation and pooling of pharyngeal secretions rendered most lateral cervical spine radiographs indeterminate.
Cervical spine MR imaging was invaluable for the detection of spinal cord impingement and intrinsic cord injury and for confirming occipitoatlantal dissociation [21, 26, 27]. However, our standard imaging protocols for the cervical spine were unable to directly and reliably show either the existence of an occipital condyle fracture or the presence of alar ligament or tectorial membrane injury. Dedicated three-dimensional gradient-echo sequences with small fields of view are said to show these ligaments in volunteer subjects [28], but the efficacy of such techniques in the critically ill trauma patient remains uncertain.
We have arbitrarily defined upper normal limits for the width of the occipitoatlantal (2 mm) and atlantoaxial (3 mm) articulations in adult patients, as determined from sagittal and coronal CT reformations. Accurate determination of these distances is limited by partial volume averaging, poor z-axis resolution, and errors of small distance measurement. The upper normal limits are therefore intended as an approximate guide for the diagnosis of a disrupted joint and to aid the recognition of bilateral occipitoatlantoaxial joint complex injuries.
On the basis of our series of patients with occipital condyle fractures, which is one of the largest studies to date, we suggest a clinically useful modification of the commonly used fracture classifications based on clinical stability. Patients with occipital condyle fractures may have a wide variety of clinical outcomes, and recognition of those with unstable injury patterns should afford more timely and appropriate therapy. Because additional cervical spine injuries that may influence neurologic outcome are commonly associated with occipital condyle fractures (31% in our series), thorough imaging evaluation of the entire cervical spine must be rigorously pursued in all patients in whom an occipital condyle fracture is found.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J. C. Bertozzi, C. A. Rojas, and C. R. Martinez Evaluation of the Pediatric Craniocervical Junction on MDCT Am. J. Roentgenol., January 1, 2009; 192(1): 26 - 31. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Alcelik, K. S. Manik, P. S. Sian, and S. E. Khoshneviszadeh Occipital condylar fractures: REVIEW OF THE LITERATURE AND CASE REPORT J Bone Joint Surg Br, May 1, 2006; 88-B(5): 665 - 669. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Schuster, K. Waxman, B. Sanchez, S. Becerra, R. Chung, S. Conner, and T. Jones Magnetic Resonance Imaging Is Not Needed to Clear Cervical Spines in Blunt Trauma Patients With Normal Computed Tomographic Results and No Motor Deficits Arch Surg, August 1, 2005; 140(8): 762 - 766. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Neeman, A. I. Bloom, E. S. Lustrin, and S. P. Karakas Occipital Condyle Fractures in the Pediatric Population * Drs Lustrin and Karakas respond: RadioGraphics, November 1, 2003; 23(6): 1699 - 1701. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
