Occipital condyle fractures can be classified into three main types according to the Anderson and Montesano scheme.² These fractures are seen in 1% to 3% of cases of blunt trauma to the craniocervical region. Type 1 fractures usually result from axial loading injuries and are comminuted. Type II fractures are linear fractures that originate in the squama of the occipital bone and extend into the condyle. Type III fractures are avulsion fractures of the condyles; these fractures are most prone to instability and atlanto-occipital dislocation.

C1 Fractures and Transverse Ligament Injuries

Fractures of the atlas are usually defined in relation to the lateral mass and extent of arch involvement.³ They can involve any parts of the ring in isolation or in combination, ranging from single unilateral fractures to burst-type fractures involving all four aspects, which is known as a Jefferson fracture. Since isolated atlas fractures without ligamentous injury are stable and heal with simple immobilization, the clinical importance of fractures of the atlas is to understand the possible involvement of the transverse ligament, the vertebral artery, and other associated spinal fractures. The most commonly cited radiographic criteria indicating unstable disruption of the transverse ligament include the Rule of Spence⁴ (lateral displacement of C1 lateral masses over C2 greater than 6.9 mm) and the atlantodental interval being greater than 3 mm. However, when feasible, this author prefers MRI evaluation of all atlas fractures to assess for concomitant ligamentous injury. Transverse ligament disruption, as with other cases of atlantoaxial instability, is an indication for surgical fixation.

Nontraumatic disruption of the atlantoaxial ligaments can also lead to gross atlantoaxial instability. C1-2 rotatory subluxation is a rare condition usually seen after inflammatory and/or infectious conditions of the pharynx and tonsils in the pediatric population. In the elderly, rheumatoid arthritis (discussed later) can lead to atlantoaxial instability requiring surgical stabilization.

C2 Fractures

Odontoid process fractures affect the elderly far more often than younger people and are, unfortunately, relatively common. The most common classification scheme for fractures of C2, the Anderson and D’Alonzo scheme,⁵ relies on the location of the fracture line within the odontoid process or body of C2. In this scheme, type I fractures involve the tip of the dens, type II fractures run through the junction of the dens and the body of C2, and type III fractures course through the vertebral body of C2.

Type I fractures are an avulsion of the alar ligament and are usually stable. Cervical collar immobilization for symptomatic management is usually sufficient.

Type II fractures (Figure 28-1) are the most common type of dens fracture and are more often subject to nonunion, especially in patients older than 50 years of age when displacement is greater than 5 mm. When choosing treatment strategies for type II odontoid fractures, the surgeon must consider the integrity of the transverse ligament, age and orientation of the fracture, displacement and/or angulation of the fractured process, and patient-specific factors such as medical comorbidities, and body habitus. For example, certain body habitus features, such as a barrel chest, can make anterior odontoid screw placement impossible.

FIGURE 28-1 Type II dens fracture.

Type III fractures extend into the C2 vertebral body. This fracture type can be mechanically unstable but usually heals well with immobilization. As such, treatment usually entails cervical immobilization in either a rigid cervical orthosis or a halovest for 12 weeks, and the majority of patients heal by bony union.

Fractures of the C2 pedicles (also known as traumatic spondylolisthesis or hangman’s fractures) are often classified based on the mechanism of injury,^6,⁷ where flexion (type III) and flexion-distraction (type IIa) are often unstable and require surgical fixation, especially type IIa injuries with greater than 4 mm distraction and/or greater than 11 degrees of angulation. Other fractures of the axis can include isolated fractures of the C2 vertebral body or fractures of the C2 spinous process or lamina, which are usually stable and can achieve good union with nonoperative immobilization.

Craniocervical Manifestations of Rheumatoid Arthritis

Between 10% and 85% of patients with rheumatoid arthritis (RA) have neck pain and 10% to 60% have neurological deficits.⁸ Spinal column manifestations of RA are most often seen at the craniocervical junction. This is usually a late finding in the disease course; therefore, a significant proportion of RA patients with craniocervical abnormalities are elderly.

RA of the upper cervical spine, similar to RA in peripheral joints, is an inflammatory condition that results in degenerative synovitis, ligament laxity, pannus formation, and bony erosion. These pathological changes can lead to atlantoaxial subluxation and are present in up to 86% of patients with RA. ⁸RA can also lead to degeneration of the occipital condyle-C1 joints, leading to cranial settling. Degeneration of the C1-2 and O-C1 joints can also lead to vertical migration of the odontoid process into the foramen magnum (basilar invagination), resulting in myelopathy from odontoid compression of the lower brainstem. Myelopathy can also be caused by pannus formation around the dens and consequent narrowing of the spinal canal.

Management of craniocervical abnormalities in patients with rheumatoid arthritis depends on the severity of clinical symptoms and the extent of craniocervical instability. Similar to craniocervical and atlantoaxial instability induced by traumatic events, measurement of the Powers ratio and the atlantodental interval can be used to assess occipitoatlantal and atlantoaxial instability, respectively. C2 vertical subluxation can be assessed by a number of radiographic lines (Chamberlain’s, McRae’s and McGregor’s lines). Frank craniocervical instability requires surgical stabilization.

Conservative Management of Occipitocervical Injuries in the Aging Spine

Once evidence of occipitocervical injuries is discovered in the aging spine, the treating clinician has to decide whether to pursue surgical or nonsurgical management of these conditions. The initial step in the management of all craniocervical region injuries is to determine whether the injury is stable or unstable. Some instances of craniocervical abnormalities, like occipitoatlantal dislocation, result in evidence of clear instability and are therefore surgical emergencies. However, stable injuries such as type I odontoid fractures can be managed with cervical orthoses while bony union is achieved.

Conservative management of upper cervical injuries is usually achieved with rigid immobilization of the cervical spine either with rigid cervical collars, such as the Philadelphia collar or Miami J collar, or with halo vests. Cervical collars provide good sagittal motion restriction in the upper cervical and subaxial spine. However, they are easy to remove and thus have variable rates of user adherence.

Halo vests provide good upper cervical and subaxial sagittal motion restriction. They also provide superior axial plane motion restriction compared to cervical collars. In addition, halo vests are secured to the skull and cannot be easily removed by users. Halo vests are associated with a higher morbidity and mortality rate, especially for elderly patients. For these reasons, halo vest use in the aging population, while sometimes unavoidable, should be approached with caution.

Surgical Approaches and Techniques

Ventral vs. Dorsal Approaches

Isolated odontoid fractures, any large C2 pannus with ventral compression of the spinal cord, and bony tumors of C1 or C2 (especially those located in the midline and anterior to the spinal cord) can be approached ventrally. Ventral approaches to the high cervical spine can be accomplished through the neck, the posterior pharyngeal wall, the maxilla, or the mandible. The ventral retropharyngeal approach gains access to the ventral aspects of C1 and C2. With this approach, care must be taken to preserve the cervical branches of the facial nerve and the hypoglossal nerve as these structures traverse the neck. The transoral approach utilizes an incision in the dorsal pharyngeal wall and provides excellent access to the ventral midline cervical spine. It can be used for odontoid resection and for resection of clival lesions. Transoral approaches may be associated with relatively higher complication rates, including CSF leak, wound dehiscence, retropharyngeal abscess, and lingual edema. Extended maxillary approaches (described by Crockard and colleagues) and mandibular approaches can be utilized in cases where wider surgical corridors are required.

Dorsal approaches to the occipitocervical region are used far more frequently than ventral approaches. The patient is positioned in the standard prone position and common surgical principles such as dissection along the avascular midline plane and subperiosteal dissection are employed.

Occipitocervical Fusion

The instrumented technique for achieving rigid fixation across the occipitocervical junction was popularized by Ransford and colleagues in 1986. They described the use of a contoured steel loop and sublaminar wiring to establish a fairly rigid fixation across the OC junction. Although the use of sublaminar wires increases the risk of injury to neural structures when compared to uninstrumented, onlay fusion procedures, the vast improvement in fusion rates after sublaminar wiring popularized its use. However, in spite of the improved level of fixation after sublaminar wiring, patients still required the use of halo vests before complete solid fusion could be established. The desire for fixation techniques that obviate the need for halo vests led to the techniques being used today. The most common surgical treatment for OC instability today involves the use of a contoured occipital plate that is connected by a rod to cervical screws (Figure 28-2).

FIGURE 28-2 Occipital-cervical fixation.

Patients undergoing occipitocervical fusion are usually placed in a Mayfield clamp and secured in a prone position, taking care to avoid excessive motion at the craniocervical junction during positioning. Since fixation of the occiput to the cervical spine eliminates the natural range of motion at the OC-C1 joint, care must be taken to maintain the spine in a neutral position in order to prevent patients from assuming a permanent flexed or extended position after surgery. An incision is usually made from the external occipital protuberance down to C3 or C4 and a subperiosteal muscular dissection is performed at all levels where screws are to be placed. The occipital bone is thickest in the midline and thins out laterally, so the length of the occipital screws must be chosen carefully and in accordance with the shape of the bone. Depending on the integrity of the bony structures in the atlas and axis, lateral mass screws can be placed at C1 and translaminar or pedicle/pars screws may be used at C2. Transarticular C1-C2 screws are also an option. The cervical spine screws are then secured via a rod to the occipital plate.

Odontoid Screw

When feasible, an excellent option for treatment of type 2 odontoid fractures is direct fixation of the fracture with an anterior odontoid screw (Figure 28-3). Preoperative considerations include intact transverse ligament, fracture line orientation, and acuity of injury (given concern for nonunion with sclerotic fracture edges). Depending on displacement of the fractured odontoid process, reduction can be first achieved with external immobilization prior to, or at the time of, surgery.

FIGURE 28-3 Anterior odontoid screw with C4-6 lateral mass fixation.

Practical preoperative considerations include patient anatomy and operative positioning to allow proper screw trajectory. Limiting factors can include barrel chest, short craniocaudal neck dimension, or rigid cervical spine preventing extension to achieve necessary trajectory. When discussing the operative plans and obtaining patient consent, possible plans for aborting screw placement and proceeding with C1-2 posterior fusion can be helpful.

Operative planning, positioning, and set-up are critical for appropriate odontoid screw placement. Patients are positioned and C-arm biplanar fluoroscopy is utilized to achieve adequate working views in the AP and lateral planes and optimal fracture reduction prior to incision.

Skin incision is planned based on necessary screw trajectory and cosmesis, often centered around C5, and neck dissection should proceed with standard attention to developing a safe corridor between the carotid sheath and trachea/esophagus to access the anterior cervical spine. Placement of the screw is performed over a K-wire under fluoroscopic guidance and an appropriate entry point is chosen at the anterior-inferior body of C2, depending upon the planned screw trajectory. Optimal placement can be facilitated by drilling a recess into the body of C3 and removing a piece of the C2-3 disc to allow for the screw trajectory and entry point at C2. A single lag screw is utilized for fracture fixation and reduction, and an attempt should be made to achieve bicortical purchase through the odontoid fragment to maximize biomechanical stability of the construct. Great care is taken at the time of K-wire and screw placement to avoid injury to the vertebral-basilar complex and cervical cord/brainstem dorsal to the fracture fragment. Advantages of odontoid fracture fixation with an odontoid screw include direct fracture reduction/stabilization, preservation of some C1-2 motion, decreased time of immobilization, and decreased morbidity associated with halo placement or posterior surgical approach.

C1-2 Harms

Multiple options exist for posterior C1-2 fixation. In cases of fractures involving both the atlas and odontoid, consideration must be given to the stability of the atlantal arch in immobilizing the C1-2 complex, and, when necessary, fixation can be extended to the occiput. Otherwise, posterior C1-2 fixation techniques are useful in cases of atlantoaxial instability including type II odontoid fractures, degenerative disease of the C1-2 complex, osteoinvasive malignancy of the C1-2 complex, and nonunion of odontoid fracture.

C1 lateral mass-C2 pars/pedicle screw fixation, known as the Harms construct ⁹, is an effective posterior fusion construct for appropriately selected patients, and, unlike an odontoid screw, it can be utilized in patients with a disrupted transverse ligament. Advantages include direct visualization of fusion surfaces, flexibility in timing of surgery (can be utilized in acute and chronic treatment of instability), and, as a polyaxial screw and rod construct, it can easily be extended to the occiput or subaxial spine, if necessary.

Operative planning is critical to safe and effective treatment and should include a preoperative cervical spine CT scan to delineate the bony anatomy; when necessary, this can be supplemented with vascular imaging to define vertebral artery anatomy. Patients are positioned prone with the head immobilized in a halo or Mayfield pins that are secured to the table. The neck is maintained in a neutral position, and C-arm fluoroscopy or other navigation tools are utilized.

If use of iliac crest autograft is planned, positioning, prepping, and draping the patient should be modified accordingly. Incision and dissection is carried through the midline ligamentum nuchae to expose the caudal edge of the occipital bone and the cephalad edge of the C3 lamina, and subperiosteal lateral dissection is extended to the C1-2 joint and the lateral aspect of C2 (while preserving the C2-3 facet capsule). Great care must be taken to avoid injuring the vertebral artery, including limiting lateral dissection to the medial one third of the cephalad atlantal arch and, when present, recognizing the ponticulus posticus identified on preoperative CT scan.

Screw placement requires adequate exposure of the lateral mass of C1, which requires identification, and often retraction of, the C2 (greater occipital) nerve, along with meticulous hemostasis, as there is often significant bleeding from a venous plexus. The middle of the C1 lateral mass at the junction with posterior atlantal arch is a reliable entry point for the C1 lateral mass screw. The screw is inserted with a slight medial trajectory as the medial wall of C1 is palpated to ensure maintenance of its integrity. Lateral fluoroscopy (or other navigation tool) should be utilized and the tip of the screw should be aimed at the anteriormost part of the anterior arch.

The superolateral quadrant of the C2 lateral mass is the approximate entry point for a C2 pedicle screw. The screw is placed with a medial and cephalad trajectory (about 30 degrees in each plane). A C2 pars screw is an alternative to the pedicle screw; it is very similar but has a more inferior and medial entry point and thus has a steeper cephalad trajectory and less medial trajectory. It is essential that preoperative CT scans be studied carefully, as there is a high variability in the position and course of the vertebral arteries in this area.

Placement of rods is performed in a standard fashion. Use of autograft and/or allograft is done at the preference of the surgeon, and careful attention is directed to decortication and preparation of the fusion surfaces including the C1-2 articulating surfaces.

C1-2 Transarticular Screws

An alternative method for posterior atlantoaxial fixation is a C1-2 transarticular screw construct (Figure 28-4), where an appropriatelysized lag screw traverses the pars interarticularis of C2, the atlantoaxial joint, and the lateral mass of C1. The indications for its use and its biomechanical stability are similar to C1-2 posterior fixation screw-rod constructs.

FIGURE 28-4 C1-2 transarticular screws with supplemental translaminar wiring.

Preoperative planning is similar to that for C1-2 posterior screw-rod fixation techniques, with an emphasis on the importance of vertebral artery anatomy. The patient should be positioned in Mayfield or halo pins rigidly fixed to the operating room table. C-arm fluoroscope should be positioned for AP and lateral imaging, and sterile prep and drape should include the caudal extension of the sterile field to the upper thoracic spine for possible percutaneous placement of the transarticular screws. Careful analysis of fluoroscopic visualization of atlantoaxial spine and ability to achieve appropriate alignment of C1-2 for screw placement should be performed prior to incision. Sublaminar wiring can augment the transarticular screw fixation construct, and may be employed at the surgeon’s discretion.

The appropriate trajectory of the transarticular screw requires a steep cephalad angle that, depending upon individual patient anatomy, may not be technically feasible in the wound utilized for dissection of the atlantoaxial spine. Therefore, use of a percutaneous tunneling device through a separate stab incision over the lower cervical or upper thoracic spine may be necessary to achieve the optimal angle, under fluoroscopic guidance.

Beginning with a K-wire under fluoroscopic guidance, the entry point for the transarticular screw is approximately 3 mm lateral to the medial edge, and 3 mm superior to the inferior edge of the C2 inferior articular process. The trajectory proceeds in a steep cephalic and slightly medial angle across the pars of C2. After traversing the pars, the screw can be visualized in the surgical field prior to entering the lateral mass of C1, where attention should be directed toward retracting/protecting the C2 nerve and ganglion. For optimal placement, the tip of the screw should engage the cortex of the anterior-superior lateral mass of C1. Use of a 4- to 5- cm lag screw (size can be planned based on preoperative CT) can achieve firm bony purchase and tight compression of the C1-2 joint for optimal fusion. Attention should be directed toward decortication of fusion surfaces, often including placement of a tricortical strut graft between lamina of C1 and process of C2 to augment fusion.

C2 Laminar Screws

In 2004, Wright and Leonard reported a case series of C2 fixation using crossing laminar screws at C2 (Figures 28-5 and 28-6). Since then, the C2 laminar screw has emerged as a viable alternative to C2 pedicle/pars screws and C1/2 transarticular screw techniques. The growth of this technique can be attributed to ease of placement, lower incidence of vertebral artery injury, and a similar biomechanical profile when compared to C2 pedicle/pars or C1-2 transarticular screw placement.