Fracture of the occipital condyle: the forgotten part of the neck
- Correspondence to: Mr Kelly, Staff Grade (e-mail: )
- Accepted 8 December 1999
A case of occipital condylar fracture in a multiply injured and unconscious motorcyclist is reported. This injury was clinically unsuspected but found on the lowest cuts of head computed tomography. It is shown that this site is often inadequately imaged when scanning the head and neck in victims of trauma. The Anderson and Montesano classification of occipital condylar fracture is described. It is noted that types 1 and 2 are stable injuries but type 3 is potentially unstable. A retrospective analysis of 30 head computed tomography scans in trauma cases revealed that in only 16 were the occipital condyles adequately imaged. It is emphasised that vigilance is required to detect fractures of the occipital condyle and that it should be standard practice to include this area when performing computed tomography of the head in trauma victims.
A 16 year old male road traffic accident victim was found lying beside his motorcycle with the helmet some distance away. The exact speed and circumstances of the incident were unclear. Paramedic roadside opinion was that the victim had suffered a tumbling fall with consequent potential for twisting head injury. In the prehospital phase a semirigid cervical collar was applied and fluid resuscitation given.
On arrival at hospital the patient was haemodynamically stable with a Glasgow Coma Scale score of 9 and other than bleeding from an occipital scalp wound no external injury was seen. He was electively intubated and ventilated. Cervical spine control was maintained at all times. Fluid resuscitation was continued and initial radiographs revealed bilateral pulmonary contusions. Cervical spine plain films were normal.
Immediate computed tomography of the head chest and abdomen was performed and a frontal lobe haematoma with mild cerebral oedema was seen. The lowermost head cuts revealed a displaced fracture of the right occipital condyle (fig 1). Ventilation was continued and ongoing care given by intensivists and the neurosurgical team. Intracranial pressure was monitored and remained normal. Halo traction was applied before his extubation on day four. He required three months rigid fixation in a halo vest. Follow up computed tomography showed healing at the fracture site without evidence of subluxation. Residual cognitive impairment at two weeks necessitated referral to the head injury rehabilitation unit and at the time of writing cognitive function and speech are now normal and a residual swallowing problem is improving constantly.
Occipital condylar fracture was originally described by Charles Bell in 1817.1 This was a postmortem diagnosis of a hospital patient who sustained a fall at the time of discharge! His demise was attributed to medullary compression by the condylar fragment. Before the wide availability of computed tomography diagnosis in life was rare but by 1996 61 cases had been reported in the literature. In addition other occipital condylar fractures found at postmortem examination have been described.
Often patients with this diagnosis have loss of consciousness because of associated head injury and indeed most early neurological deficits in patients with occipital condylar fracture are attributable to such related injury. In those patients who are conscious on presentation neck pain is usual and indeed may be the only symptom although neurological deficit including hemiparesis (attributable to contusion of ipsilateral adjacent spinal cord) is recognised. Lower cranial nerve deficits (that is, IX–XII) are a “classic” sign. However they occur in only a third of cases and within this group a third of such deficits present late.2
Cervical spine radiographs are usually normal2 although prevertebral soft tissue swelling and asymmetry of the odontoid peg as seen on the anteroposterior view are recognised signs. In addition computed tomography of the head often neglects fully to image the craniocervical junction and the recognition of this injury therefore presents a diagnostic challenge. Because neurological deficit can be late in onset it is however necessary to make the diagnosis at the time of presentation. It is therefore important to retain the same index of suspicion for skull base injury as is widespread for the cervicodorsal junction.
It is best to consider the paired occipital condyles and the first two cervical vertebrae as a functional unit. The stability of the atlantooccipital junction within this unit is maintained by several structures.3 Firstly, the tectorial membrane is a continuation of the posterior longitudinal ligament and attaches to the foramen magnum. This limits flexion and extension. Secondly, the paired alar ligaments run directly from the odontoid peg to each occipital condyle; their function is to check lateral flexion and rotation. It is these above structures that primarily stabilise this part of the neck. Finally, the anterior and posterior atlanto-occipital membranes interconnect the margins of the foramen magnum and the ring of the atlas while in addition the apical ligament extends from the odontoid peg to the anterior margin of the foramen magnum.
Anderson and Montesano described a commonly used classification system based on injury mechanism and computed tomography fracture morphology.4 Class I fractures are impacted condylar fractures consequent upon axial loading injuries. Type II fractures are extensions of a basilar skull fracture. Both these types are functionally stable consequent upon integrity of the tectorial membrane and contralateral alar ligament. Type III fractures, sustained during extreme rotation and/or lateral bending are avulsion fractures of the condyle by the alar ligament. This type of fracture is potentially unstable because of loading of the tectorial membrane and contralateral alar ligament. Tuli et al2 recommend an alternative classification. They propose that instability is determined more by ligamentous injury as detected by computed tomography alignment criteria and magnetic resonance imaging findings than on fracture morphology.
Optimal treatments are not fully agreed upon but most sources propose semirigid collar provision for stable injury and rigid immobilisation of varying design for the potentially unstable injury.
To monitor the adequacy of our computed tomography head scans in imaging the craniocervical junction we retrospectively assessed a series of 30 computed tomography scans of head injured patients examined before the case under discussion. The hard copies were examined for inclusion of the occipital condyles. In only 16 of 30 patients was there satisfactory demonstration of these structures. It is already our policy to print “bone window” images in addition to standard “brain windows” in acutely head injured patients and we now ensure that computed tomography in patients with head injury starts at the C1 ring.
In conclusion, vigilance is required to diagnose occipital condylar fracture and in particular we advise that all unconscious trauma victims undergoing computed tomography head imaging need careful examination of the craniocervical base. This should include 3 or 5 mm axial sections to cover the occipital condyles, with inspection of the resulting images on “bone windows” in addition to the conventional “brain” settings.
Andrew Kelly initiated the case report, searched the literature and wrote the report. Richard Parrish conceived the idea of a retrospective analysis of our CT scans, performed this review and contributed materially to the discussion.
Andrew Kelly is the guarantor for this paper.
Thanks to Sallie Waring for medical illustration.