DEFINITION

The term occipitocervical and atlantoaxial instability encompasses a number of varied conditions that compromise the normal function of the O-C1 and C1-C2 joints, resulting in either pain, spinal cord dysfunction, or the threat thereof.

Instability can result from trauma, including fractures of the articular masses and occipital condyles, rupture of the transverse ligament, odontoid fracture, or Jefferson fracture. Nontraumatic causes include inflammatory arthropathy (most commonly rheumatoid arthritis), osteoarthritis, congenital anomalies, rotatory subluxation, tumor, and infection.

Several methods have been described for stabilizing the atlantoaxial complex, as well as the occipitocervical junction, including wiring techniques, transarticular screw fixation, plate and screw fixation, and screw and rod fixation.

We describe our technique for occipitocervical plating; transarticular screw fixation; articular mass screw and rod construct to achieve atlantoaxial arthrodesis with C2 pars, pedicle, and translaminar screw fixation; and C1-C2 wiring techniques.

 

 

ANATOMY

 

The base of the skull is composed of the external occipital protuberance (EOP), the occipital condyles (which articulate with the C1 lateral masses), and foramen magnum. Landmarks noted on posterior dissection are the posterior edge of the foramen magnum, the superior nuchal line, the inferior nuchal line, and the external protuberance (FIG 1).

 

 

 

FIG 1 • Anatomic landmarks and features of the occiput.

 

 

The nuchal lines serve as attachments to the paired neck muscles. The trapezius attaches to the superior line and the rectus capitis attaches to the inferior line

 

The nuchal ligament attaches to the external protuberance.

 

The thickness of the bone in the suboccipital region varies depending on location. In the midline, the internal occipital crest has a mean thickness of 8.3 mm at the level of the inferior nuchal line, increasing to a mean of 13.8 mm at the EOP. The lateral bone is thinner, ranging from a mean of 3.7 mm at the level of the

inferior nuchal line and increasing to a mean of 8.3 mm at the level of the superior nuchal line.1

 

The first cervical vertebra, or the atlas (C1), is unlike any other in that it lacks a vertebral body and spinous process. It consists of an anterior and posterior arch connected by two articular masses, forming a ring that pivots about the odontoid process of C2 (FIG 2A).

 

On each side of the cranial surface of the C1 posterior arch, there is a groove for the vertebral artery, the first cervical nerve, and their associated venous complex (FIG 2B). In a small subset of the population, this groove is covered by an arch of bone, the ponticulus posticus. The resulting foramen is identified as the arcuate

foramen.2

 

The articular masses of C1 give rise to the superior and inferior articular facets, which are broad and articulate, with the occipital condyles superiorly and the axis inferiorly.

 

P.5156

A synovial joint also is located between the posterior aspect of C1 and the odontoid process of the axis.

 

 

 

FIG 2 • A. The atlas consists of an anterior and posterior arch connected by two articular masses. B. Anteroposterior view of first and second cervical vertebrae. Anterior (C) and posterior (D) views of the axis, demonstrating the odontoid process projecting upward from the vertebral body. The pedicle connects the lamina and the vertebral body, projecting superomedially. The pars interarticularis lies between the superior and inferior articular processes. E. The vertebral artery ascends through the foramina transversaria from C6 to C3. It takes a turn laterally through C2 underneath the pars interarticularis. Once it traverses the transverse foramen at C1, it turns medially and lies on the superior surface of the C1 ring. F. After passing medially on the superior surface of the C1 ring, the vertebral artery passes through the foramen magnum and merges with its counterpart to form the basilar artery.

 

 

The axis (C2) has thicker laminae and a larger bifid spinous process than the subaxial cervical vertebra. It is characterized further by an odontoid process that projects upward from the vertebral body. Lateral to the odontoid process, or dens, are the sloping superior articular surfaces, which articulate with the inferior articular facets of C1, forming the atlantoaxial joint. The C2 pedicle can be identified in a zone between the lamina and vertebral body, projecting superomedially (FIG 2C,D).

 

O-C1 articulation: The kidney-shaped lateral masses of the atlas articulate cranially with the kidney-shaped

occipital condyles. The joint allows for 15 to 20 degrees flexion and extension with 5 to 10 degrees of lateral

bending.3 Stability depends on the joint ligaments, the tectorial membrane, and the longitudinal bands of the cruciate ligaments.

 

C1-C2 articulation: The C1-C2 complex is composed of three articulations—two laterally composed of the inferior C1 and superior C2 articular facets, and one anteriorly between the dens and the posterior aspect of the anterior C1 arch.

 

 

The C1-C2 articulation allows for 47 degrees of rotation to either side, which is approximately 50% of the

lateral rotation of the entire cervical spine.4 Panjabi and associates5 showed that in the healthy spine, C1-C2 flexion is 11.5 degrees, extension is 10.9 degrees, lateral bending is 6.7 degrees, and axial rotation to each side is 38.9 degrees.

 

The vertebral artery, which is the first branch of the subclavian artery medial to the anterior scalene muscle, ascends behind the common carotid artery. It then ascends through

 

P.5157

the foramina transversaria from C6 to C1. After traversing through the foramina transversaria at C1, the artery takes a sharp turn medially and posteriorly to course behind the C1 articular mass along the groove in the posterior arch of C1. It then passes through the posterior atlanto-occipital membrane before ascending through the foramen magnum as it merges with its counterpart to form the basilar artery (FIG 2E,F).

 

The C1 nerve root, or the suboccipital nerve, exits cranial to the posterior arch of C1 and innervates muscles of the suboccipital triangle. The C2 nerve root, or greater occipital nerve, exits between the posterior arches of C1 and C2, posterior to the superior C1-C2 articulation. It does not exit through a true foramen like the remaining subaxial cervical nerve roots. It traverses inferior to the obliquus capitis inferiorly to ascend through the semispinalis capitis to lie superficial to the rectus capitis. Injury to the greater occipital nerve can lead to dysesthesia of the posterior scalp and can be troublesome to patients.

 

PATHOGENESIS

 

Stability of the O-C1 articulation relies on the ligamentous support and anatomic contour of the occipital condyles on the lateral masses of C1. Occipital condyle fractures may be stable or represent the bony component of an occipitocervical dissociation (OCD). OCD involves disruption of the ligamentous constraints

resulting in an injury with high mortality rate.6

 

Stability of the C1-C2 articulation relies heavily on its ligamentous restraints, including the transverse, alar, and apical ligaments, and the facet capsules. Trauma may disrupt these ligamentous restraints. Also, with the advanced degeneration found in arthritic conditions, these ligamentous structures may become incompetent.

 

Up to 3 mm of anterior translation of C1 on C2, as measured by the anterior atlantodental interval (AADI) on a lateral cervical radiograph, is normal. An atlantodental interval of 3.5 to 5 mm in an adult indicates potential damage to the transverse ligament, whereas an interval greater than 5 mm indicates probable injury to the transverse ligament and accessory ligaments (FIG 3A).

 

In cases of trauma, an atlantodental interval greater than 3.5 mm probably is an indication for further evaluation and most likely requires C1-C2 arthrodesis.

 

In patients with inflammatory arthropathy, including rheumatoid arthritis, a canal diameter identified as posterior atlantodental interval (PADI) smaller than 14 mm is associated with a worse outcome and is an

indication for decompression and fusion.7 The exact AADI measurement is not as relevant in these patients as with trauma patients.

 

Fractures that involve the osseous structures of C1 and C2 also may result in atlantoaxial instability and require arthrodesis (FIG 3B).

 

NATURAL HISTORY

 

In the event of trauma to the occipital condyles, careful evaluation with computed tomography (CT) and possibly magnetic resonance imaging (MRI) is indicated to rule out associated OCD. In the setting of an OCD, patient mortality can be high and operative management is recommended. Any translation or distraction at that level is an indication for intervention. Traction is contraindicated. Immediate immobilization followed by occipitocervical fusion is indicated.

 

In the event of C1-C2 trauma, the potential need for surgery arises in the setting of ligamentous instability, fractures, or a combination of the two. Atlantoaxial instability due to rupture of the transverse ligament represents a threat to the cervical spinal cord with a low likelihood of successful healing. Thus, C1-C2 fusion is indicated.

 

 

 

 

FIG 3 • A. An AADI greater than 5 mm indicates likely injury to the transverse ligament and, in the setting of

trauma, necessitates operative stabilization. B. An avulsion (arrow) of the transverse ligament from the ring of C1 indicates instability and may require arthrodesis of C1-C2.

 

 

Transverse ligament disruption in association with a Jefferson fracture may represent an exception to this rule, in that, successful nonoperative fracture treatment (halo vest) can lead to a “stable” C1-C2 segment on

flexion-extension radiographs.

 

Fractures of the odontoid process may represent a primary indication for C1-C2 fusion if nonoperative means (eg, halo vest immobilization) cannot obtain or maintain an appropriate reduction or if a patient elects surgery to avoid the use of a halo. Displaced odontoid fractures have an increased likelihood of resulting in either non-or malunion in the cases of type II and III fractures, respectively (FIG 4A).

 

Primary atlantoaxial osteoarthritis is quite painful and responds poorly to nonoperative means. C1-C2 fusion affords a high likelihood of symptom relief (FIG 4B).

 

Cervical myelopathy due to either rheumatoid pannus or pseudopannus formation, as seen in older individuals with extensive subaxial spondylosis and spontaneous fusion, is unlikely to improve without fusion of the C1-C2 segment (FIG 4C).

 

C1-C2 instability due to rheumatoid arthritis may be neither symptomatic nor a neurologic threat. Thus, in this case, an ADI exceeding 3.5 mm is not, by itself, an indication for surgery. A PADI of less than 14 mm or the presence of myelopathy is a poor prognostic sign and indicates the need for surgery. Painful C1-C2 rheumatoid involvement in the face of adequate medical therapy also indicates the need for surgery.

 

Basilar invagination or cranial setting of the occipitocervical junction due to rheumatoid arthritis can result in cervicomedullary cord compression. Especially in the presence of myelopathic symptoms, occipitocervical fusion with or without decompression is indicated.

 

 

P.5158

 

 

 

FIG 4 • A. Displaced odontoid fractures (type II) have a higher likelihood of a nonunion and may require a primary C1-C2 fusion. B. Joint space narrowing is a sign of C1-C2 osteoarthritis and responds poorly to nonoperative management. C. Pseudopannus formation behind the dens in patients with rheumatoid arthritis can lead to cervical stenosis and myelopathy. It rarely improves without surgery but will dissolve after C1-C2 fusion. Flexion (D) and extension (E) lateral radiographs demonstrate C1-C2 instability in a patient with rheumatoid arthritis. F. Os odontoideum is another condition associated with instability in which part of the dens is not attached to the axis body.

 

 

Progressive C1-C2 subluxation, especially with cranial settling, also has an unfavorable natural history. C1-C2 fusion in this instance will obviate the need for a future occipitocervical fusion, which has a less favorable influence on the overall condition of the cervical spine (FIG 4D,E).

 

The natural history of asymptomatic C1-C2 instability associated with miscellaneous conditions such as os odontoideum (FIG 4F) and Down syndrome is less clear. When such patients have symptoms, myelopathic signs, or an insufficient PADI, the potential benefits of a C1-C2 fusion probably outweigh the risks of the natural history. The patient's age, lifestyle, and activity level also must be considered in determining the need for surgery.

 

HISTORY AND PHYSICAL FINDINGS

 

A complete history and physical examination, including a thorough neurologic examination, should be performed when evaluating a patient with occipitocervical and/or C1-C2 pathology. The complaints offered will vary with the presentation (eg, trauma, inflammatory arthritis, developmental, congenital).

 

Patients with a traumatic injury may not only complain of isolated pain but also may present with neurologic

deficits. A low threshold of suspicion should be maintained for patients with blunt trauma to the head or face or with known noncontiguous fractures of the spine.

 

Some patients with rheumatoid arthritis may complain only of axial neck pain, whereas others may present with deteriorating gait and bilateral hand numbness or clumsiness without significant neck pain.

 

Patients with primary atlantoaxial arthritis will complain of severe neck and head pain, most often unilateral, with varying degrees of refusal to rotate their head, especially ipsilaterally toward the pain. Locking or crepitation of the affected joint may be both audible and palpable.

 

Physical examination should include the following:

 

 

Active self-limited rotation of the head, especially toward the side of the pain. Normal rotation is up to 50 degrees of rotation to either side. C1-C2 pathology often causes pain that limits rotation.

 

Palpation of the suboccipital area near the interval between the posterior arches of C1 and C2 may elicit pain. When asked, the patient often can point to the source of their pain.

 

Response to traction versus compression combined with passive C1-C2 rotation. The patient is examined supine with his or her head resting comfortably on a pillow. Passive lateral head rotation is measured with slight manual traction. In cases of C1-C2 arthritis, this maneuver should provide more motion and less pain than similar motion with an axial load. With slight manual traction, head rotation is increased, whereas an axial load may cause pain and result in decreased rotation.

 

 

P.5159

 

In the setting of potential traumatic instability, however, these examination maneuvers are not applicable. The cervical spine must be immobilized until the radiographic and CT findings are known.

 

IMAGING AND OTHER DIAGNOSTIC STUDIES

 

In the setting of blunt trauma, plain radiographs of the cervical spine, specifically the upper cervical spine,

have been shown to be inadequate. McCulloch et al8 reported that plain radiographs have a sensitivity of 52%, a specificity of 98%, a positive predictive value (PPV) of 81%, and a negative predictive value (NPV) of 93%, whereas helical CT had a sensitivity and specificity of 98%, PPV of 81%, and NPV of 93%. They concluded that although helical CT has limited ability to detect pure ligamentous injury, it is superior to plain radiographs when evaluating patients with high-energy trauma for cervical spine fractures.

 

MRI scan is the most effective diagnostic tool for identification of ligamentous and other soft tissue injury.9

 

Patient-controlled flexion-extension plane radiographs may be beneficial in certain situations. However, studies have shown that even extensive damage to the intervertebral structures may inconsistently result in

abnormal findings using this tool.10 In addition, patient effort may play a role in the ability to obtain reliable results.11

 

CT with sagittal and coronal reconstruction is done routinely for diagnostic purposes as well as for preoperative planning.

 

Fractures of either C1 or C2 indicate a significant likelihood of additional cervical spine fractures. As many as 50% of C1 fractures may be associated with other fractures.

 

A vertebral artery angiogram is recommended when there is any question of acute injury or history of injury to the vertebral arteries. A unilateral vertebral artery injury rarely is symptomatic because of sufficient collateral flow through the contralateral vertebral artery as well as the circle of Willis. A patient with a vertebral artery injury who presents with neurologic deficits due to a concomitant spinal cord injury may be especially difficult

to diagnose clinically.

 

We recommend imaging with either an angiogram or an magnetic resonance arthrogram (MRA) in all patients presenting with a significant flexion-distraction injury, fracture that extends into the transverse foramen, or facet dislocation (FIG 5).

 

The method of treatment of vertebral artery injuries associated with cervical trauma remains controversial, especially with asymptomatic injuries. When discovered, symptomatic vertebral artery injuries may be treated with anticoagulation to prevent thromboembolic complications. If a surgical procedure is necessary,

anticoagulation is stopped before and restarted after surgery.12

 

DIFFERENTIAL DIAGNOSIS

 

 

Rheumatoid arthritis: instability, pannus accumulation, cranial settling Degenerative osteoarthritis

 

Trauma: occipital condyle fracture, OCD, articular mass fracture, odontoid fracture, Jefferson fracture, transverse ligament rupture

 

 

Tumor Infection

 

 

 

FIG 5 • Coronal reconstruction of a CT angiogram demonstrating occlusion of flow through the left vertebral artery (right side of the image) in a patient with a C4-C5 facet fracture-subluxation.

 

 

 

Atlantoaxial rotatory subluxation: recurring subluxation, irreducible and fixed subluxation Miscellaneous: Down syndrome, os odontoideum

 

 

NONOPERATIVE MANAGEMENT

 

Stable occipital fractures can be treated with rigid cervical collar immobilization. Stable injuries are typically those that involve the occipital condyle without injury to the tectorial membrane or alar ligaments. Unstable injuries are those that involve occipital avulsion fractures or extensive comminution signaling ligamentous injury. These injuries need halo vest immobilization versus occiput to C2 instrumentation and fusion.

 

In most instances, a hard collar is not adequate for immobilizing an unstable C1-C2 articulation, but one may be considered in an elderly patient who is not a surgical candidate and otherwise cannot tolerate a halo or

Minerva vest (Variteks, Istanbul, Turkey).13

 

For certain fractures, use of a halo vest may be appropriate, and the patient is treated in the orthosis for 3 months. It is a time-tested “nonoperative” option with well-defined success/failure rates.

 

Some patients may require a halo for postoperative immobilization, depending on the fixation quality, the anticipated level of patient compliance with a hard collar, and other unusual circumstances.

 

SURGICAL MANAGEMENT

 

 

Occipitocervical fusion with plate and rod fixation is widely used and is described in the following texts. Several different techniques have been described for successful posterior C1-C2 fixation and fusion.

 

Before Jeanneret and Magerl14 described the transarticular screw technique, posterior fixation was accomplished with Gallie's sublaminar wiring and grafting15 or by the Brooks wiring method.16

 

Some of the newer methods include the C1 articular mass and C2 pedicle screw and rod construct described by Goel and Laheri,17 the use of C1 articular mass and C2 pars screw, and the use of C2 translaminar screw combined with C1 articular mass screw and rod construct.18

 

Biomechanically, the Magerl technique of transarticular fixation provides the best stability compared with traditional

 

P.5160

wiring methods, but it may be technically more demanding than either the Brooks or Gallie method of fixation as well as C1 lateral mass and C2 screw fixation.

 

Malreduction of C1-C2, anomalous position or size of the vertebral arteries, and collapse of the lateral masses of C2 are relative contraindications to the use of the transarticular screw method because of the risk for inadvertent penetration of the vertebral artery.

 

Because of the restrictions on screw trajectory through the C2 pedicle and C1-C2 joint without endangering the vertebral artery, up to 20% of patients cannot have safe placement of bilateral screws using the Magerl

technique.19 Using C1 and C2 independently placed screws, more patients can receive rigid fixation in the setting of anatomic variations.

 

The C2 pars and C2 pedicle are distinct structures. Placement of screws into either the pars or pedicle puts the vertebral artery at risk.

 

C2 translaminar screws involve placement of polyaxial screws into the laminae of C2 in a bilateral, crossing fashion. This screw technique does not put the vertebral artery at risk and does not require fluoroscopic

guidance. It does, however, require intact posterior elements at C2.20

PREOPERATIVE PLANNING

 

All imaging studies should be reviewed with regard to osseous anatomy as well as the course of the vertebral artery. Preoperative CT scanning with reconstructed images in the sagittal plane or in the plane of planned screw trajectory is necessary to view the pertinent anatomy and to avoid injury to the vertebral artery. This should be in addition to standard radiographs to assess overall alignment and MRI to assess neurologic compression.

 

POSITIONING

 

The patient under general anesthesia with an orally placed endotracheal tube is placed in the prone position.

 

The head is rigidly held in place using Mayfield tongs or previously placed halo. The shoulders and arms are tucked at the patient's side. The torso is placed on vertical rolls or gel pads that allow for the abdomen to be free, lowering the intraabdominal pressure.

 

The table is adjusted into a reverse Trendelenburg orientation, with the knees bent and well-padded (FIG 6). The shoulders can be depressed with wide tape if necessary to obtain a true lateral view of the C1-C2 complex.

 

After the patient is positioned, the proper alignment and anatomic reduction of an unstable C1-C2 joint should be confirmed radiographically before proceeding further. Reduction is preferable to avoid vertebral artery injury, neurologic injury, and inadequate bony purchase of the screw. However, mild anterior displacement of C1 on C2 is well tolerated and may facilitate fixing the C1 lateral mass, as long as the PADI remains large enough to accommodate the spinal cord. It is important to confirm that the head is in neutral rotation to avoid an iatrogenic torticollis. Visualization of overlapping of the mandibular rami and the C2-C3 facets can aid in confirmation of a true lateral image.

 

C1-C2 reduction in most trauma conditions can be achieved with longitudinal traction in the awake patient. After successful reduction, a halo vest may be applied to facilitate prone positioning of the patient under anesthesia.

 

The patient's hair should be shaved to above the occipital protuberance especially if occipitocervical fusion is planned.

 

The occiput, posterior neck, and posterior iliac crest should be prepped and draped in standard fashion.

 

 

 

 

FIG 6 • The patient is positioned prone with the head flexed and posteriorly translated to allow the instruments to

achieve the correct C1-C2 trajectory. It is important to confirm reduction of an unstable C1-C2 joint radiographically before proceeding further.

 

TECHNIQUES

• Exposure

   

  A midline posterior skin and subcutaneous incision is made from the occiput to as far distal as necessary. For C1-C2 fusion, the deep subperiosteal dissection is confined from the upper edge of C1 to the inferior margin of the C2 laminae (TECH FIG 1). For occipitocervical fusion, the EOP should be exposed with full-thickness lateral and subperiosteal dissection of the muscular attachments of the trapezius and semispinalis capitis.

   

  When performing Magerl screw fixation, a longer skin incision permits the correct trajectory of the drill guides, which often are tunneled through the posterior cervical extensor muscles.

   

  A shorter skin incision could be used, with the drills, guides, and other instruments passed through percutaneous stab wounds, but we have found the cosmetic results less desirable.

   

  Muscular infiltration with local anesthetic and epinephrine will reduce bleeding.

   

   

   

  TECH FIG 1 • The posterior arch of C1 down to the inferior margin of the lamina of C2 is exposed with meticulous subperiosteal dissection.

   

   

  P.5161

   

  Care should be taken to identify and dissect within the nuchal ligament. This will allow for a relatively avascular dissection to the spinous processes.

   

  The extensor musculature should carefully be dissected off of the C2 spinous process. The C2 spinous process should be preserved in a C1-C2 fusion to allow for suturing of the extensor musculature back to

  bone through drill holes.

   

  The posterior tubercle of the atlas is palpated and musculature attached to the arch is detached with care on both sides. This is done to about 1.5 cm from midline bilaterally to avoid damage to the vertebral artery located in the sulcus of the posterior arch.

   

  Dissection lateral to C1 and C2 should be limited to the zygapophyseal joint and not between as there can be significant bleeding from the venous plexus located at that level.

   

  A small-angled curette can be used to detach the soft tissues from the edges of the lamina of C2 and posterior arch of C1 for sublaminar suturing versus wiring at the end of the procedure.

• Occipitocervical Fusion

   

  Posterior dissection is done by exposing to 2 cm above the occipital protuberance and 2 to 3 cm lateral to midline and down to the foramen magnum.

   

  Locate and palpate suboccipital midline keel and paramedian cranium. This will be the location of plate fixation.

   

  Midline screws offer the best bone purchase. Bicortical screws have 50% greater pullout strength than either unicortical screws or wires.21

   

  Size the appropriate occipital plate. Plate size should depend on the rod connection to suboccipital screws. Attempt should be made to minimize medial-lateral rod bending to meet the C1 or C2 screws by appropriately sizing the plate.

   

  Cranial thickness can be measured on preoperative CT; however, typical thickness for midline keel is 8 to 16 mm (thickest at the EOP, and progressively thinner proceeding caudally from the EOP) and 6 mm for paramedian screws.

   

  With the plate held in position, pilot holes are drilled using a handheld power drill with drill guides. Drill guides set to 6 to 8 mm are used for the central keel initially and can be increased up to 14 to 16 mm for maximal purchase at the EOP. Smaller screws are placed distal to the EOP as determined by patient anatomy.

   

  Bicortical screws can be placed with care to maximize bone purchase. Occasionally, cerebrospinal fluid (CSF) or slow venous bleeding can emanate from screw holes. If this occurs, this is typically controlled with placement of the screw.

   

  C1, C2, and other subaxial screws are placed as described in the following section and connected to the occipital plate with a rod and set screw construct.

   

  Decortication to bleeding bone of the arch of C1, posterior occiput, C2, and other fusion levels is completed using a high-speed burr. Bone grafting is performed using local bone and autologous iliac crest or rib, depending on the anatomic circumstances. Corticocancellous harvested graft can be secured between suboccipital bone and arch of C1 or C2 by performing a modified Gallie technique and heavy suture (TECH FIG 2).

   

   

   

  TECH FIG 2 • A lateral radiograph of an occipitocervical fusion.

   

• Magerl Method of C1-C2 Transarticular Screw Fixation14

   

  Sagittal and axial CT images are scrutinized preoperatively. The isthmus of the C2 pars must measure at

  least 4.5 mm in height and width to accommodate a transarticular screw.22 An abnormally large or malpositioned vertebral artery might lead to increased risk of harm to this important structure.

   

  C1-C2 reduction and the ability to obtain a true lateral view of C1-C2 with a fluoroscope are confirmed.

   

  A midline incision is made from the occiput far enough caudally to allow a steep enough angulation of the drill and other instruments.

   

  Posterior C1 and C2 exposure is carried out laterally to visualize the superior and medial surfaces of the C2 pars. Care also should be taken to avoid disturbing the C2-C3 facet capsule.

   

  The starting point for the transarticular screw is at the posterior cortex of the C2 inferior articular process, 2 mm cephalad and 2 to 3 mm lateral to the medial border of the C2-C3 facet joint.

   

  P.5162

   

  The starting point is confirmed with a direct lateral C-arm image and marked with a 2-mm burr to provide a secure starting point for the tip of the drill bit. Caudal-cranial angulation is determined via lateral C-arm fluoroscopic guidance. The sagittal plane orientation is confirmed visually with reference to the superior and medial surfaces of the C2 pars. A Penfield 4 dissector can be placed on the dorsal surface of the C2

  pars to serve as a guide on the lateral fluoroscopy view.

   

  The K-wire is directed superiorly along the C2 pars while aiming toward the anterior arch of C1 as seen on the lateral fluoroscopic images, with slight medial angulation of 0 to 10 degrees (TECH FIG 3A,B). Advance the drill or wire slowly with frequent fluoroscopic visualization.

   

   

   

  TECH FIG 3 • A,B. The guidewire is placed superiorly through the pars, aiming toward the anterior arch of C1 on lateral fluoroscopic images. With the first guidewire in place, a second guidewire is placed on the other side. The K-wire is overdrilled with a drill bit (C,D) and tapped, and the screw is placed on the second side (E) before the same is done on the first side. F. Postoperative radiograph of transarticular screw fixation in a patient who sustained a C1-C2 fracture-dislocation.

   

   

  We recommend leaving the drill bit or K-wire in place on the initial side to transfix the C1-C2 joint, then proceeding to the opposite side. The screw on the second side is inserted before returning to the initial side to remove the drill bit, then tap and insert the second screw. This avoids any problems with loss of reduction (TECH FIG 3C-F).

   

  Bone grafting is performed with autologous iliac crest. After careful decortication of the posterior arches, a modified Gallie technique is employed using either heavy suture or braided titanium cable to secure the graft in place (as described under Gallie Method of Sublaminar Wiring and Grafting).

   

  The extensors at C2 are repaired with drill holes placed through the spinous process.

• Harms/Goel Method of C1 Articular Mass Fixation17

   

  The ponticulus posticus is a common anomaly that can easily be mistaken for a broad posterior arch of the atlas, and the lateral radiograph must be reviewed to check for the presence of an arcuate foramen to avoid injuring the vertebral artery.23

   

  The starting point for the C1 screw is in the middle of the junction of the C1 posterior arch and the midpoint of the posterior inferior part of the C1 lateral mass. The entry point is marked with a 2-mm high-speed burr (TECH FIG 4).

   

  The C2 nerve root is retracted in a caudal direction for proper screw placement. If divided, the patient

  may experience troubling neuralgia and numbness postoperatively.

   

  The initial drill hole is made in a straight or slightly convergent trajectory in the sagittal plane and parallel to the plane of the C1 posterior arch in the coronal plane, with the tip of the drill aimed toward the anterior arch of C1 (TECH FIG 5A).

   

  The hole is tapped and measured, and a 3.5-mm polyaxial screw of appropriate length is placed allowing the polyaxial portion of the screw to lie above the arch of C1. An 8-mm unthreaded portion of the C1 screw will also stay above the lateral mass to avoid any irritation to the greater occipital nerve.

   

  P.5163

   

   

   

   

  TECH FIG 4 • C1 lateral mass starting point.

   

   

  Care should be taken when dissecting around the C1-C2 articulation to avoid excessive bleeding from the epidural venous plexus in this area. Hemostasis can be achieved using bipolar electrocautery, powdered Gelfoam with thrombin, and cotton pledgets.

   

  The center of the lateral mass of C1 is the ideal exit point of the C1 lateral mass screw, and the proximity of the internal carotid artery (ICA) places it in danger when placing a bicortical screw. The ICA can vary in location from side to side and may be within 1 mm of the ideal exit point of a bicortical transarticular screw

  or a C1 lateral mass screw.24

   

   

   

  TECH FIG 5 • Postoperative CT scan (A) and lateral radiograph (B) of a patient with a displaced, kyphotic, chronic C2 fracture who underwent C1-C2 posterior fusion using C1 articular mass screws and C2 pars screws.

   

   

  Medial angulation of the screw in the lateral mass of C1 may increase the margin of safety for the ICA, but care should be taken to avoid penetrating the occipitocervical joint by aiming caudally.

• C2 Pedicle/Pars Screw Placement

   

  The starting point of the C2 pedicle is in the midline of the C2-C3 facet joint, 3 to 5 mm cranial to the C2-C3 articulation. The trajectory is 25 degrees of medial convergence and is aimed 25 degrees cephalad while keeping in mind that individual anatomy will vary (TECH FIG 6A).

   

  The starting point of the C2 pars screw is 2 mm superior and lateral to the inferior C2-C3 articulation. It is placed in a craniocaudal trajectory similar to the transarticular screw but does not need to be aimed as much cephalad. It is aimed 20 to 25 degrees medial (TECH FIG 6B).

   

  A no. 4 Penfield dissector is used to feel the medial border of the C2 pars interarticularis, and the superior and medial aspects of the isthmus are palpated during the drilling process.

   

  The drilled hole is then palpated with a blunt ball-tipped probe. The hole is tapped, and a 3.5- or 4.0-mm polyaxial screw is inserted.

   

  C2 pars screw length is typically 16 to 22 mm, depending on the anatomy of the vertebral artery and thickness of the pars. Preoperative CT will aid in estimating length.

   

  The polyaxial screw heads are connected with two rods. If necessary, a reduction of the C1-C2 articulation is performed before fixation with the rods.

   

  The posterior elements of C1 and C2 are decorticated, and a corticocancellous H graft is secured using a modified Gallie technique (TECH FIG 5B).

   

  The extensors at C2 are repaired using drill holes through the spinous process.

   

   

  P.5164

   

   

   

   

  TECH FIG 6 • Screw trajectory and correct identification of the pedicle (A) and pars (B) of the C2 or axis.

   

• C2 Translaminar Screw Placement25

   

  A high-speed burr with a 2-mm tip is used to open a small cortical hole at the junction of the C2 spinous process and lamina starting on the right side. This is done in the cranial half of the C2 lamina.

   

  A hand drill is used to drill the contralateral or left lamina with the drill aligned along the angle of that lamina. This is done to a depth of 25 to 30 mm. Care must be taken to allow the trajectory to be slightly less than the downslope of the lamina so that any cortical breakthrough would occur dorsally and not ventrally toward the spinal canal.

   

  A small ball probe is used to evaluate the drilled hole for any cortical breaches.

   

  Typically, a 4.0- × 30-mm screw is placed with the head of the screw at the junction of the spinous process and lamina on the right. A smaller diameter screw may be needed depending on the height of the

  lamina in order to accommodate two screws.

   

  Using the high-speed burr, a small cortical hole is made at the junction of the spinous process and lamina on the left in the caudal half of the lamina. Using a similar technique as described

   

  P.5165

  earlier, a 4.0- × 30-mm screw is placed in the right lamina with the screw head at the junction of the spinous process and lamina on the left (TECH FIG 7A-C).

   

  C1 lateral mass screws and/or subaxial screws are placed as described.

   

   

   

  TECH FIG 7 • Saw bones model demonstrating coronal (A) and axial (B) appearance of the translaminar screw technique. Axial CT image of translaminar fixation (C).

   

   

  The posterior elements of C1 and C2 are decorticated, and a corticocancellous H graft is secured using a modified Gallie technique (TECH FIG 5B).

   

  If possible, the extensors at C2 are repaired using drill holes through the spinous process.

• Brooks Method of Wire Fixation

   

  Brooks wiring is the most reliable of the traditional wire fixation methods. It does not provide as much stability as other screw options, however, and so must be used in conjunction with significant

  postoperative immobilization, often a halo vest, for optimal likelihood of fusion.16 It also requires passing sublaminar wires at C2, which can be technically demanding.

   

  Midline posterior subperiosteal exposure of C1 and C2 laminae is carried out with careful attention to dissect from midline laterally at C1 to prevent injury to the vertebral artery. The occipital nerves emerge through the interlaminar space between C1 and C2.

   

  The ligamentum flavum between C1 and the occiput and also between C1 and C2 is sharply divided. A Woodson instrument is used to confirm that there are no dural adhesions in the sublaminar space.

   

  Although Brooks originally described the use of two doubled 20-gauge stainless steel wires passed under each side of the arch of C1 followed by C2 with the aid of a no. 2 Mersilene suture in a cephalad to caudal direction, we routinely use pairs of braided titanium cables rather than stainless steel wire.

   

  After the cables are passed with a loop at the end, two full-thickness rectangular bone grafts measuring approximately 1.25 × 3.5 cm are taken from the iliac crest. The sides of each graft are beveled to fit in the interval between the C1 and C2 laminae and placed on each side.

   

  The bone grafts are then held in place by securing the cables (TECH FIG 8).

   

   

   

  TECH FIG 8 • Brooks wiring technique.

• Gallie Method of Sublaminar Wiring and Grafting

   

  The Gallie method is less stable than the Brooks method and is relatively contraindicated in the presence of any posterior C1-C2 instability.15 Biomechanically, this method provides minimal stabilization in

  rotation with only comparable stabilization in anteroposterior motion in response to motion.26 It also requires significant postoperative immobilization.

   

  Dissection similar to that of the Brooks method is performed.

   

  A sublaminar wire or braided titanium cable is passed under the arch of C1 and looped around the spinous process of C2. We use a suture for this technique when the Gallie graft is employed

   

  P.5166

  in conjunction with Magerl transarticular fixation because the Gallie configuration is relied on for maintenance of graft position, not for mechanical stability (TECH FIG 9).

   

  A corticocancellous bone graft from the iliac crest (TECH FIG 10A) is taken and placed with the

  cancellous side facing down on the posterior elements after the cortical bone has been burred to reveal a nice bleeding cancellous bed (TECH FIG 10B). The small grooves are placed on the superior and inferior edges of the graft to hold the cables in place.

   

   

   

  TECH FIG 9 • Gallie wiring technique.

   

   

   

  TECH FIG 10 • A. The posterior arches of C1 and C2 are decorticated. B. A corticocancellous graft is taken from the iliac crest. This is fashioned into an H shape, and the cancellous side is placed facing down on the decorticated posterior elements of C1-C2. C. A modified Gallie technique is used to secure the graft in place.

   

   

  The cable is tightened, and the graft is secured (TECH FIG 10C).

   

  PEARLS AND PITFALLS

   

   

  Bone grafting ▪ A Gallie H graft is fashioned from the iliac crest and contoured to fit over the posterior arches of C1 and C2, with its cancellous surface applied directly opposing the decorticated surfaces of C1 and C2.

   

  Frameless stereotactic navigation

  • This method registers only one vertebra, and the relation of C1 and C2 obtained on the CT scan may differ from that resulting after positioning on the operating table, resulting in aberrant screw placement and possible injury to the vertebral artery, whereas intraoperative fluoroscopy yields real-time information. Caution should be used in interpreting the information presented on the “virtual” images during surgery.

     

    Injury to the vertebral artery

  • Careful preoperative planning will guide selection of the appropriate procedure to reduce the risk of injury. In the event of injury to a vertebral artery during a Magerl procedure, a short screw may be placed to contain the bleeding. If this occurs while drilling or tapping the first side, it is unwise to attempt a C1-C2 screw on the contralateral side. An alternative fixation technique which does not place the contralateral artery at risk should be employed, such as a Brooks or Gallie procedure.

     

    Venous bleeding in the C1 lateral mass

• Gentle tamponade of the venous sinuses, along with application of hemostatic agents, is recommended. Once the surgical instruments are removed along with the pressure from the retractors, the bleeding is usually controlled with ease. Avoid indiscriminate use of cautery.

   

  Supplemental wire/cable fixation

• Supplemental wiring in conjunction with screw fixation methods at C1 and C2 provides no significant mechanical advantage. However, a suture configuration of a similar nature will hold the graft surfaces in proper apposition to the decorticated host bone, possibly improving the fusion rates.

   

  C1-C2 facet fusion

  • Originally described as a component of the Magerl procedure, direct exposure, decortication, and grafting of the posterior aspect of the C1-C2 facets is not routinely necessary. It may be indicated for revision procedures, patients with incompetent posterior C1 arches, certain fracture patterns, or for high-risk hosts.

 

 

 

POSTOPERATIVE CARE

P.5167

 

Whereas patients undergoing the Brooks or Gallie procedure obtain a maximal fusion rate with postoperative halo vest immobilization, the modern screw fixation methods yield fusion rates in excess of 90% with only cervical collars worn for 6 to 12 weeks.

 

The type of collar used and duration of wear should be in accordance with surgeon judgment about host bone, security of fixation, anticipated patient compliance, and so forth.

 

OUTCOMES

 

Jeanneret and Magerl14 achieved solid fusion in 13 patients stabilized with the transarticular screw technique.

McGuire and Harkey27 showed solid fusion in 8 patients using a transfacet screw technique.

Fielding and associates28 achieved fusion in 45 of 46 patients with fractures using the Gallie technique. Brooks and Jenkins16 used a C1-C2 sublaminar wiring technique to achieve fusion in 14 of 15 patients.

Harms29 reported fusion in all 37 patients with C1 lateral mass and C2 pedicle mini-polyaxial screw and rod construct.

 

 

COMPLICATIONS

Vertebral artery and ICA injuries Infection

Malpositioned screw Nonunion

C2 neuralgia

C1-C2 hyperextension with Brooks or Gallie procedure if the C1 and C2 arches are compressed together

 

 

REFERENCES

1. Roberts DA, Doherty BJ, Heggeness MH. Quantitative anatomy of the occiput and the biomechanics of occipital screw fixation. Spine (Phila Pa 1976) 1998;23(10):1100-1107; discussion 1107-1108.

    

    

2. Stubbs D. The arcuate foramen. Variability in distribution related to race and sex. Spine 1992;17(12):1502-1504.

    

    

3. Wang JC, Mummaneni PV, Haid RW Jr. Fixation options in the occipitocervical junction. In: Mummaneni PV, Lenke LG, Haid RW Jr, eds. Spinal Deformity: A Guide to Surgical Planning and Management. St. Louis: Quality Medical Publishing, 2008:223-2400.

    

    

4. White A, Panjabi M. The clinical biomechanics of the occipitoatlantoaxial complex. Orthop Clin North Am 1978;9(4):867-878.

    

    

5. Panjabi M, Dvorak J, Duranceau J, et al. Three-dimensional movements of the upper cervical spine. Spine 1988;13(7):726-730.

    

    

6. Ben-Galim PJ, Sibai TA, Hipp JA, et al. Internal decapitation: survival after head to neck dissociation injuries. Spine (Phila Pa 1976) 2008;33(16):1744-1749.

    

    

7. Boden SD, Dodge LD, Bohlman HH, et al. Rheumatoid arthritis of the cervical spine: a long-term analysis with predictors of paralysis and recovery. J Bone Joint Surg Am 1993;75A:1282-1297.

    

    

8. McCulloch PT, France J, Jones DL, et al. Helical computed tomography alone compared with plain radiographs with adjunct computed tomography to evaluate the cervical spine after high-energy trauma. J Bone Joint Surg Am 2005;87(11):2388-2394.

    

    

9. Muchow RD, Resnick DK, Abdel MP, et al. Magnetic resonance imaging (MRI) in the clearance of the cervical spine in blunt trauma: a meta-analysis. J Trauma 2008;64(1):179-189.

    

    

10. Brown T, Reitman CA, Nguyen L, et al. Intervertebral motion after incremental damage to the posterior structures of the cervical spine. Spine 2005;30(17):E503-E508.

     

     

11. Hwang H, Hipp JA, Ben-Galim P, et al. Threshold cervical range-ofmotion necessary to detect abnormal intervertebral motion in cervical spine radiographs. Spine 2008;33(8):E261-E267.

     

     

12. Cothren CC, Moore EE, Ray CE Jr, et al. Cervical spine fracture patterns mandating screening to rule out blunt cerebrovascular injury. Surgery 2007;141(1):76-82.

     

     

13. Chaudhary A, Drew B, Orr RD, et al. Management of Type II odontoid fractures in the geriatric population: outcome of treatment in a rigid cervical orthosis. J Spinal Disord Tech 2010;23(5): 317-320.

     

     

14. Jeanneret B, Magerl F. Primary posterior fusion of C1/2 in odontoid fractures: indications, techniques, and results of transarticular screw fixation. J Spinal Disord 1992;5(4):464-475.

     

     

15. Gallie W. Fractures and dislocations of the cervical spine. Am J Surg 1939;46:495-499.

     

     

16. Brooks AL, Jenkins EW. Atlanto-axial arthrodesis by the wedge compression method. J Bone Joint Surg Am 1978;60(3):279-284.

     

     

17. Goel A, Laheri V. Plate and screw fixation for atlanto-axial subluxation. Acta Neurochir (Wien) 1994;129(1-2):47-53.

     

     

18. Gorek J, Acaroglu E, Berven S, et al. Constructs incorporating intralaminar C2 screws provide rigid stability for atlantoaxial fixation. Spine 2005;30(13):1513-1518.

     

     

19. Madawi AA, Casey AT, Solanki GA, et al. Radiological and anatomic evaluation of the atlantoaxial transarticular screw fixation technique. J Neurosurg 1997;86(6):961-968.

     

     

20. Wright NM. Posterior C2 fixation using bilateral, crossing C2 laminar screws: case series and technical note. J Spinal Disord Tech 2004;17(2):158-162.

     

     

    P.5168

     

21. Haher TR, Yeung AW, Caruso SA, et al. Occipital screw pullout strength: a biomechanical investigation of occipital morphology. Spine 1999;24(1):5-9.

     

     

22. Mandel IM, Kambach BJ, Petersilge CA, et al. Morphologic considerations of C2 isthmus dimensions for

    the placement of transarticular screws. Spine 2000;25(12):1542-1547.

     

     

23. Young JP, Young PH, Ackermann MJ, et al. The ponticulus posticus: implications for screw insertion into the first cervical lateral mass. J Bone Joint Surg Am 2005;87(11):2495-2498.

     

     

24. Currier BL, Todd LT, Maus TP, et al. Anatomic relationship of the internal carotid artery to the c1 vertebra: a case report of the cervical reconstruction for chordoma and pilot study to assess the risk of screw fixation of the atlas. Spine 2003;28(22):E461-E467.

     

     

25. Menendez JA, Wright NM. Techniques of posterior C1-C2 stabilization. Neurosurgery 2007;60(1 suppl 1):S103-S111.

     

     

26. Grob D, Crisco JJ III, Panjabi MM, et al. Biomechanical evaluation of four different posterior atlantoaxial fixation techniques. Spine 1992;17(5):480-490.

     

     

27. McGuire RA, Harkey HL. Modification of technique and results of atlantoaxial transfacet stabilization. Orthopaedics 1995;18(10): 1029-1032.

     

     

28. Fielding JW, Hawkins RJ, Ratsan SA. Spine fusion for atlanto-axial instability. J Bone Joint Surg Am 1976;58(3):400-407.

     

     

29. Harms J, Melcher RP. Posterior C1-2 fusion with polyaxial screw and rod fixation. Spine 2001;26(22):2467-2471.