Occiput, C1, and C2 Instrumentation

Published on 26/03/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 2322 times

CHAPTER 298 Occiput, C1, and C2 Instrumentation

The occipitocervical junction (OCJ) represents a complex interface between the cranium and the rostral cervical spine. Composed of the occiput, atlas, and axis, it provides nearly all of the rotation and most of the flexion-extension of the head and neck. This region, with its osseous articulations and ligamentous support structures, must resist forces in eight axes of rotation, including flexion, extension, bilateral lateral bending, bilateral rotation, distraction, and axial loading.¹ The occipitoatlantal joint allows 13 degrees of flexion and extension, 8 degrees of lateral bending, and no rotation, whereas the atlantoaxial complex allows 10 degrees of flexion and extension, no lateral bending, and 94 degrees of rotation.² Therefore, it is imperative that these unique anatomic and kinematic features be taken into consideration when proceeding with occipitocervical fusion. These unique movement capabilities also provide challenges to stabilization compared with the subaxial spine.

Until the past few decades, the mainstay of treatment for cervical instability was external immobilization. However, the nonunion rate and morbidity associated with its use in certain populations precluded its use in some instances.³ In 1910, Mixter and Osgood⁴ described the first surgical treatment of atlantoaxial instability in the literature. They reported securing the posterior arch of the atlas to the spinous process of the axis with a heavy silk thread. Since then, the evolution of the techniques has spanned the gamut from the Gallie type in 1939⁵ to more recent techniques such as the C2 laminar technique.⁶^–⁸

Indications for Fusion—Occipitocervical

Occipitocervical fusion is indicated primarily for instability. Causes of instability include, but are not limited to trauma (including occipitoatlantal dislocation), rheumatoid settling, basilar invagination (primary or secondary), neoplastic (primary or metastatic), congenital abnormalities (e.g., certain cases of Chiari malformation), or iatrogenic instability after surgical decompression. In the setting of trauma, occipitoatlantal dislocation is a devastating injury usually resulting in death or significant neurological morbidity.⁹ If the patient survives the initial injury, occipitocervical fusion is often required. Acquired causes of instability at the craniocervical junction, such as rheumatoid settling or basilar impression, result from pathologic changes of the occipitocervical junction instigated by inflammation of the synovium resulting in bony, cartilaginous, and ligamentous destruction leading to deformation and instability.¹⁰ Such destruction can often cause brainstem compression and/or spinal cord compression, and patients may experience severe neck pain or myelopathy.¹¹ Additional indications for occipitocervical fusion include persistent atlantoaxial instability secondary to failure of previous C1-2 fixation or complicated C1-2 fractures where wiring or screw techniques are considered unsafe or contraindicated in the patient. Also, certain C1 fractures with wide displacement of the lateral mass cannot only destabilize the C1-2 articulation, but also the condyle-C1 articulation, requiring occipital-C2 stabilization.

Indications for Fusion—C1-2

Trauma

Traumatic fractures of C1 or C2 can lead to significant instability of the atlantoaxial complex. Fracture of the dens is the most common traumatic injury encountered requiring stabilization, but fortunately multiple options exist for the management of this fracture.¹² Anderson and D’Alonzo¹³ divided the dens fracture into three groups. Type 1 fractures involve fractures through the upper dens, and although they are generally considered stable, one study suggested that the stability is not absolute.¹⁴ Type 2 fractures occur at the junction of the dens and the body of C2, and type 3 fractures extend into the body of C2. Odontoid screw fixation can address certain C2 fractures; however, it is contraindicated in the following: disruption of the transverse atlantal ligament, associated comminuted fractures of one or both atlantoaxial joints, unstable type 3 fractures, atypical type 2 fracture with oblique comminuted fracture lines, irreducible fractures, associated thoracic kyphosis, and pathologic fractures.¹⁵

Dorsal fusion techniques offer an alternative approach for stabilization of odontoid fractures in complex situations. Such examples include associated fractures to one or both of the atlantoaxial joints or associated Jefferson fractures of the atlas, which places the C1-2 joint at risk for subluxation.¹⁶ Additionally, the patient with extensive, multisystem trauma who requires constant access to the chest and neck, when halo vest immobilization is not recommended, and those patients who have suffered a type 3 dens fracture when surgical fusion could facilitate earlier mobilization can benefit from dorsal fusion.¹⁷

Other indications for dorsal fusion include type 2A fractures. These fractures not only have significant comminution at the base of the dens and behave differently from other type 2 fractures, they exhibit a high rate of nonunion with external immobilization.¹⁸ Posterior displacement of the dens leads to nonunion rates of 70% to 89%,^17,¹⁹ and anterior displacement of more than 6 mm leads to nonunion in 67% of cases compared with 9% in those displaced less than 6 mm.²⁰ Significant displacement in either direction prevents adequate placement of ventral screws through the body and into the dens if the fracture cannot be reduced, therefore requiring dorsal stabilization.

Dorsal fusion also proves to be a valuable alternative in patients with marked thoracic kyphosis and type 2 or shallow type 3 fractures, which often cannot be approached ventrally because the rib cage obscures the correct screw trajectory. Finally, in pathologic fractures from neoplastic disease of the dens, dorsal fusion techniques prove to be a safer option because they prevent instrumentation of the diseased dens.¹²

Ligamentous Injury

Ligaments of the atlantoaxial complex limit flexion of the upper cervical spine. These include the tectorial membrane, cruciform ligament, transverse ligament, and alar ligaments.^21,²² If the atlantodental interval (normal value, 2 to 4 mm) has a value greater than or equal to 5 mm, ligamentous laxity should be suspected. If the interval is more than 10 to 12 mm, complete incompetence of the ligamentous complex is likely.²³ The alar ligaments limit axial rotation of the upper cervical spine and damage to these ligaments increases rotation in the contralateral side by 30%.²⁴ If failure of any component of the atlantoaxial ligament complex is suspected, dorsal surgical fusion is indicated.

Rheumatoid Arthritis

Rheumatoid arthritis (RA) is a systemic disease that can affect the both the atlantoaxial junction and the subaxial spine. RA progression in the cervical spine can cause pain in as many as 49% of patients and symptoms of cord compression or myelopathy in 20% of patients.²⁵ Dorsal fusion has been more successful than ventral approaches for stabilizing the atlantoaxial complex in these patients.²⁶

Congenital

Os odontoideum is the failure of the dens to fuse with the body of the axis. Although it is often confused with type 2 dens fractures, it is considered a stable process. However, both os odontoideum and odontoid agenesis may lead to incompetence of the cruciate ligament, resulting in subsequent atlantoaxial instability.²⁷ The evidence-based indications and treatment options for os odontoideum have previously been reported.²⁸

Preoperative Preparation and Exposure

The patient with occipitocervical instability typically arrives in the operating room with some form of external immobilization, such as a halo orthosis or hard cervical collar. These patients should remain in a flat or semirecumbent position while undergoing awake fiberoptic endotracheal intubation to minimize the risk of neurological injury. This technique allows constant surveillance of the patient’s neurological examination, which in turn allows the anesthesiologist to cease the induction if the examination changes during this critical period. Preparation for intraoperative somatosensory and brainstem auditory evoked potentials for continuous monitoring throughout the operation is done before positioning. Evoked potentials are measured following induction of general anesthesia and compared with those measured after the patient is placed prone. In the supine position, the Mayfield head holder (Integra Life Science, Plainsboro, NJ) is placed on the cranium for fixation or, if a halo brace has been placed preoperatively, a Mayfield head holder adapter for the halo is used. The patient is carefully placed in the prone position and craniocervical alignment is assessed with fluoroscopic guidance before removal of any cervical orthosis. Recent studies suggest that positioning patients with cervical instability may be safer on a rotating bed rather than with conventional logrolling.²⁹

A linear posterior midline incision is made from the inion to the spinous process of the desired caudal level. The skin and subcutaneous tissue are incised down to the dorsal fascia, followed by development of an avascular midline plane between the paraspinal muscles. Soft tissue dissection is continued until the occipital squamosa, foramen magnum, and posterior elements of the first three cervical vertebrae are clearly exposed. The exposure should extend to the lateral margin of the facet joints bilaterally, whereas the occipital surface should be wide enough to support the chosen construct. All soft tissue should be meticulously removed from the bony surfaces of the segments selected, followed by use of a high-speed bur drill to decorticate these same levels to promote fusion before placement of instrumentation.

Operative Techniques—Wiring

Wiring Techniques—Occipitocervical

In 1927, Foerester first described reconstruction of the occipitocervical junction with the use of fibular strut grafts.³⁰ Since that time, multiple techniques of fusion have been described, have evolved, and been improved. Simple, stand-alone onlay bone grafting precluded the morbidity from instrumentation, but required prolonged external halo immobilization. Later, however, the addition of wiring to stabilize the segments posteriorly, and securing the bone grafts, enhanced stability in some planes of motion.³¹

Wiring techniques for the occipitocervical junction require the passage of wires between bur holes in the occiput as the rostral fixation point, with spinous process, sublaminar, or facet wires as the caudal fixation points.³²^–³⁹ Although wiring techniques have largely been supplanted by rigid screw fixation, they remain a useful tool in the surgeon’s armamentarium in cases where screw fixation is precluded or undesired.

The occipital bone located near the foramen magnum or the midline nuchal line provides the thickest site of bone for secure fixation. The occipital bone is prepared by enlarging the posterior rim of the foramen magnum and then drilling bur holes into the bone approximately 0.5 cm above the rim of the foramen magnum. The dura is subsequently elevated from the inner table before passing the wire through the bur holes. The wires are then passed from one drill hole to the other. Following preparation of the occipital bone, bone struts or contoured metal rods are fixed to the occiput by twisting the wire tightly and then fixed to the posterior elements of the cervical spine with wire by the following techniques.

Spinous process wiring of C2 and the subaxial spine is achieved by drilling a hole in the base of the spinous processes. The hole is enlarged with a towel clip and then a standard braided 22-gauge stainless steel wire is passed through the hole and around the base of the adjacent spinous process. Fixation between segments is achieved by wiring bone graft struts to the spinous processes.

Sublaminar wiring around C1 and below requires visualization of the underlying dura, which is achieved by making laminotomies and removing the ligamentum flavum. The lamina is also notched where the wire is passed to optimize visualization of the dura. The wire is passed under the lamina by feeding and pulling simultaneously to prevent neurological injury. Then bone struts or rod constructs are wired to the lamina bilaterally.

Facet wiring can be performed if decompression with a laminectomy is indicated and requires the facet joints to be opened and the articular cartilage to be removed. A drill hole is made at a 90-degree angle to the inferior facet and the wire is passed through the hole. Facet wiring alleviates the risk of neurological injury associated with sublaminar wires.

Steinmann Pin Fusion

Sonntag and Dickman ^43,⁴⁴ have described a rod and wire technique using a contoured -inch diameter threaded Steinmann pin. This pin is contoured into a U shape and also bent to accommodate the lordotic curvature of the occipitocervical region. After obtaining the desired curvature and length, the pin is wired to the occiput and cervical lamina or facets. If a suboccipital craniotomy or cervical laminectomy is performed, a plate of cortical iliac crest bone is wired to the central portion of the pin to protect the site of decompression. Fusion rates of 89% were reported in their series using this technique.

C1-2

First reported by Mixter and Osgood,⁴ initial techniques for dorsal atlantoaxial fusion involved variations of wiring together the posterior elements of the axis and atlas. Although these techniques are technically simple and require no special intraoperative equipment, such as fluoroscopy, all of them require rigid postoperative immobilization to obtain successful fusion.³

Because these techniques have largely been supplanted by more recent rigid screw fixation methods, they are only discussed briefly here. Several good review articles describe these techniques in fuller detail.^3,¹²

Gallie Fusion

In 1930, Gallie ⁵ first described the stabilization of a subluxed atlantoaxial complex by using a sublaminar wire placed around the posterior arch of C1 and looped around the spinous process of C2, holding in place a median bone graft notched over the spinous process of C2. Although the procedure is technically simple, it remains the poorest biomechanical construct, therefore requiring supplementation with other techniques. Gallie type of fusion requires an intact posterior arch of C1; therefore, it cannot be used if there is an associated Jefferson fracture or rheumatic involvement.

Brooks-Jenkins Fusion

The Brooks-Jenkins fusion uses doubled 20-gauge wires, which are passed under the laminae of the atlas and axis bilaterally. Two posterolateral autologous iliac crest bone grafts are beveled in to both interlaminar spaces and held in place with the overlying wire.⁴⁰ The Brooks-Jenkins fusion also requires an intact posterior arch of C1.

Sonntag’s Modified Gallie Fusion

In this modification technique,³² the sublaminar wires are eliminated under C2, using the spinous process of C2 as a fixation point. A single bicortical bone graft is fit into the interlaminar space by wiring the C1-2 interlaminar space and notched to accommodate the spinous process of C2. This technique provides increased stability without using two levels of sublaminar wires seen in the Brooks-Jenkins technique. To obtain optimal anatomic realignment, patients are kept in a halo vest preoperatively and intraoperatively. The C1-2 interlaminar space is widened with a high-speed bur drill and the spinous process and lamina of C2 are decorticated. The inferior aspect of the spinous process is also notched to seat the wire. Iliac crest graft (4 cm long) is shaped to fit the interlaminar space, placing the concave cortical surface toward the dura. The inferior aspect of the graft is notched to lie over the spinous process of C1 and two strands of no. 24 wire are passed around the posterior arch of C1, over the bone graft, and around the notched spinous process of C2. Finally, wires are tightened to three turns per centimeter. Patients are recommended to stay in a halo for 3 months postoperatively, and then changed to a hard collar for 4 to 6 weeks thereafter. The fusion rate for this technique has been described as 97%.

Rigid Screw Fixation

Previous wiring techniques introduce the risk of neurological injury by passage of the wire adjacent to unprotected dura, and in some patients with incompetent or absent cervical lamina, may prove to be a suboptimal method of stabilization. Therefore, over the past decade, rigid fixation with plate-screw or screw fixation with rod constructs have been studied and popularized. Biomechanical studies ⁴¹^–⁴³ also suggest superiority of screw-based fixation over wiring-based stabilization methods. Increased rigidity of screw-based constructs provide resistance to fatigue and vertical settling, while allowing fusion over fewer segments and decreasing the length of external immobilization after surgery.

Screw Fixation—Occiput

Screw placement for a rigid occipitocervical construct requires specific selection of both cranial and caudal anchors. Occipital screw placement requires both careful measurement of the thickness of the occipital bone and identification of the proximity of the dural sinuses preoperatively. Screws must also be carefully selected to provide optimal purchase, while avoiding violation of the dura, which may cause cerebrospinal fluid (CSF) leaks or cerebellar injury. Anatomic studies ⁴⁴^–⁴⁷ evaluating occipital bone morphology have shown that the occipital protuberance is the area of greatest bone thickness and was consistently located midline on the superior nuchal line. The bone thickness also decreases radially in both lateral and inferior directions from the occipital protuberance.⁴⁷ Ebraheim and others⁴⁴ reported that the maximum thickness of occipital bone ranged from 11.5 to 15.1 mm in males and 9.7 to 12.0 mm in females at the level of the external occipital protuberance (EOP) and that the occipital bone was thicker than 8 mm in an area extending lateral to the external occipital protuberance for up to 23 mm. Therefore, screws up to 8 mm long may be inserted in the region of the superior nuchal line up to 2 cm laterally from the center of the external occipital protuberance, 1 cm from the midline at a level 1 cm inferior to the EOP, and 0.5 cm from the midline at a level 2 cm from the EOP. Haher and associates⁴⁵ examined the pullout strength of unicortical and bicortical screws relative to occipital bone morphology and reported that on average, the pullout strength of bicortical fixation was 50% greater that unicortical screw fixation. However, unicortical screw pullout strength at the occipital protuberance was not significantly different than that of bicortical screws at other anatomic locations. Zipnick and associates⁴⁷ showed that the outer cortex contributed 45% of the total thickness of the occipital bone but the inner cortex only contributed 10%, so unicortical fixation at the occipital protuberance may offer acceptable pullout strength to secure the occiput to prevent the complications of bicortical screw placement.

Inside-Outside Technique

In 1999 Pait and associates ⁴⁸ addressed the complications of outside-inside screw fixation by developing an innovative modification to previous techniques. This technique does not necessitate maximum thickness of the occipital bone. Their inside-outside technique uses lateral mass plating with a cranial bolt that connects the plate to the cervical portion of the construct. The bolt is oriented from the epidural space outward. The occipitocervical plate is held in place to mark the location of the cranial bolt. An elliptic bur hole is then made approximately 1 cm superior to the marked cranial bolt location. The bur hole is connected to the location of the cranial bolt by drilling a trough the same width as the bolt. The bolt (or flathead screw) is then placed in the bur hole with the threads facing outward and slid down the trough to the designated location. The plate is placed over the bolt and loosely held in place with nuts to secure the screws until the desired cervical stabilization is done before completion of the occipitocervical fusion.

Occipital Plate Fixation

Perhaps most commonly today is placement of occipital screws that are placed either midline with special Y- or T-shaped keel plates or parasagittally in line with cervical instrumentation using eyelet connectors (Fig. 298-1). Lateral fixation has the advantage of allowing the placement of six screws and theoretically gives improved resistance to deformation because of its increased effective moment arm and bilateral purchase. Midline screw fixation allows the placement of only three screws, but uses the thicker occipital keel for longer screw purchase.⁴¹ Anderson and associates⁴¹ compared these two fixation locations in cadaveric specimens. Their small series showed that the location of fixation had only a small effect on the range of motion and was dependent on load direction. Although lateral fixation had significantly less motion in lateral bending, no other significant differences were observed. Surprisingly, the location of occipital fixation may have little influence on biomechanical properties of the fixation, and therefore, the choice of location should be based on the instability pattern and feasibility of plate fixation with the rest of the construct rather than on the specific location of fixating screws on the occiput.

FIGURE 298-1 A, Occipital fixation, keel plate, posterior view. In this illustration, a representative occipital plate is depicted in position. Three midline screws are inserted into the midline keel. Additional points of fixation are possible with the parasagittal screw holes (not shown). The rods can be attached laterally to connect the occipital fixation to the upper cervical spine. The plate needs to be placed carefully below the inion and superior nuchal line to prevent possible injury to the underlying transverse sinus and torcula, and above the foramen magnum to prevent injury to the upper cord. The plate should also be contoured to the bone it slopes away laterally from the midline keel. B, Occipital fixation, keel plate, cutaway lateral view. In this illustration, the depth of the midline occipital screws is appreciated. This view also mimics the midsagittal CT reconstruction of the area, illustrating how the depth of the keel, and the corresponding length of the occipital screw, is measured. Unicortical screws are depicted.

It is our preferred technique to place a midline occipital plate when possible. Screw length is carefully determined based on preoperative imaging. Specifically, a midsagittal reconstructed computed tomography (CT) scan allows determination of the ideal length of the occipital keel screws. When drilling and tapping the screw holes, a stop is strongly recommended to prevent plunging inadvertently into the posterior fossa. Due to the density of the occipital keel, tapping the entire length of the hole is recommended to prevent stripping of the screw.

Complications associated with occipital screws are fortunately rare. Screw pullout is rare, especially if screws are selected for the maximum length possible and at least three screws are placed. If cerebrospinal fluid is encountered, placement of the screw will usually occlude the hole sufficiently and no further intraoperative measures are typically required. If the drill, tap, or screws penetrate the posterior fossa, a cerebellar hematoma is possible. If suspected, postoperative CT should be performed to evaluate this potential fatal complication.

Occipital Condyle Screws

More recently, an interesting technique of occipital fixation was described separately by Uribe and coworkers ⁴⁹ and La Marca and colleagues.⁵⁰ This technique involves screw placement into the occipital condyles, and may be particularly useful when the midline keel is not available, such as in cases of prior suboccipital craniectomy (Fig. 298-2).

FIGURE 298-2 A, Occipital condyle fixation, lateral view. An occipital-cervical construct using occipital condyle screw fixation is shown. The condyle screws are partially threaded to prevent irritation of the vertebral artery coursing below. B, Occipital condyle fixation, AP view. The entry point and trajectory of the condyle screw are illustrated. The entry point is 4 to 5 mm lateral to the medial border of the condyle, with the trajectory 12 to 22 degrees medially. The exact trajectory is determined by preoperative CT imaging and/or intraoperative navigation.

(Courtesy Dr. Juan Uribe)

Establishing the caudal or cervical anchor for the occipitocervical fusion includes the use of C1-2 transarticular screws, C1 lateral mass screws, and C2 pedicle screws. These techniques are used in conjunction with the techniques of occipital screw placement and fixation plates described previously. These challenging C1-2 stabilization techniques are described in further detail in the following section.

Screw Fixation—C1-2

Limitations in stabilization by previous dorsal wiring techniques prompted the development of newer techniques of dorsal atlantoaxial fixations using rigid screw fixation. Although these rigid screw techniques provide significantly higher rates of fusion and less rigid postoperative immobilization, they prove to be more technically challenging due to screw placement near the path of the vertebral arteries, thus requiring the aid of intraoperative fluoroscopy and/or surgical navigation tools and detailed knowledge of the anatomy.³

Transarticular Screw Technique of Magerl

The Magerl transarticular screw technique ⁵¹ proved to be a major advancement in C1-2 fixation. Although it is more technically challenging than dorsal wiring techniques, it offers two main advantages. First, it does not rely on intact posterior elements and, second, it provides immediate stabilization and increasing fusion rates of the construct due to reduction of rotatory movement from the transarticular screw.

Preoperative planning with plain x-rays and fine-cut CT scans is essential.¹² Traction used to restore anatomic alignment preoperatively may foster a technically simpler procedure. Some surgeons advocate preoperative fiberoptic intubation and perioperative somatosensory evoked potentials.⁵² The patient’s head is fixed in pins or rested on a horseshoe head holder in the prone position. Final positioning is confirmed with real-time fluoroscopy to verify alignment of the atlantoaxial complex.

The spinous process of C1-3 are exposed, followed by the posterior arch of C1, but only 10 to 13 mm on each side of midline to prevent exposure of the vertebral artery. The spinous process and lamina of the C2 are exposed, continuing the dissection out laterally to expose the articular process of C2 and C3, but leaving the C2-3 joint capsule intact.

The C1-2 joint is exposed, followed by removal of the articular cartilage to promote bony fusion. Dissection of the C2 lamina into the joint space laterally allows identification of the neurovascular complex containing the C2 nerve root, which is carefully retracted caudally to prevent inadvertent injury. If C1 is intact, gentle retraction on the posterior arch is achieved by passing a 20-gauge wire around the posterior arch. The use of the wire also facilitates alignment during screw placement.

The entry point for screw placement is 2 to 3 mm lateral to and 2 to 3 mm above the medial aspect of the C2-3 facet. Recognition of the course of the vertebral artery is essential during this step. Using fluoroscopic guidance, a 2.5-mm drill bit is directed sagittally, aiming toward the anterior arch of C1. A percutaneous approach from approximately the level of C7 is required. To prevent possible anterior dislocation of C1 during screw placement, a 3.5-mm tap is used for the entire length of the trajectory, followed by insertion of a 3.5-mm fully threaded screw of the appropriate length, stopping just short of the anterior cortex of C1. After repeating the process on the contralateral side, the laminae of both C1 and C2 are decorticated and autologous iliac crest bone grafts are laid over the exposed surfaces to promote bony fusion.³ Patients are placed in a hard collar for up to 6 weeks postoperatively.

Several clinical and cadaveric studies have shown the reliability in the strength and stability of the transarticular screw construct.¹² Multiple studies have documented fusion rates as high as 95.5% to 96% with the use of this technique^53,⁵⁴ and biomechanical cadaveric studies show the construct to not only be stable in flexion and extension, but also rotation.^16,^52,^55,⁵⁶

Complication rates have been reported as between 2% and 14% for this technique, depending on the series.^16,^52,^55,⁵⁶ The complications include screw malposition and vertebral artery injury. The risk of vertebral artery injury is well documented in the literature⁵⁷ with fatalities reported from bilateral vertebral artery injuries; Madawi and associates⁵⁶ reported intraoperative vertebral artery in 8.2% of their patients. Wright and Lauryssen performed a retrospective study that reported the risk of vertebral artery injury with transarticular screw fixation to be 4.1% per patient or 2.2% per screw inserted.⁵⁷ The main limitation to this technique is that anatomic variation often precludes safe screw placement.

C1-C2 Rod-Cantilever Technique

The Magerl transarticular screw technique remains a technically challenging procedure because a narrow corridor for screw trajectory needs to be maintained to traverse the C2 pedicle and C1-2 joint without injuring the vertebral artery, particularly if there is anatomic variation. These technical limitations prompted several authors to report individualization of C1 and C2 instrumentation by placing C1 lateral mass screws and C2 pedicle screws. Goel and Laheri ⁵⁸ described atlantoaxial fixation where C1 and C2 lateral mass screws were connected using a plate. Later, Harms and Melcher⁵⁹ described C1 lateral mass screws connected to C2 pars screws using rods. This technique circumvents the limitation of anatomic variation for rigid fixation.

The patient is carefully placed in the prone position with use of real-time fluoroscopy to verify alignment and the position of the atlantoaxial complex. The cervical spine is exposed subperiosteally from the occiput to the level of C3-4. The C1-2 articulation is exposed similarly to the technique described in the Magerl technique. After the dorsal root ganglion of C2 is retracted inferiorly, the middle of the junction of the posterior arch of C1 and the midpoint of the inferoposterior part of the C1 lateral mass is exposed. This marks the entry point for the C1 screw, which is marked by using the high-speed bur drill to prevent slippage. The pilot hole is drilled in a straight trajectory in an anteroposterior direction, parallel to the plane of the posterior arch of C1 in the sagittal direction, directing the tip of the drill toward the anterior arch of C1. The hole is tapped, and then a 3.5-mm polyaxial screw of appropriate length is inserted into the lateral mass of C1 bicortically. An 8-mm unthreaded portion of the C1 screw remains above the lateral mass to minimize injury to the greater occipital nerve and to allow the polyaxial portion of the screw to lie above the posterior arch of C1 (Fig. 298-3).

FIGURE 298-3 A, C1 lateral mass fixation, posterior view. The screw is illustrated in place on the place, and the entry point of the C1 lateral mass screw is as depicted on the right. In the final position, the screw head sits dorsal to the C2 root, and the screw threads rostral to the C2 root. The entry point may be obscured by an overhanging posterior arch, and some of the arch may need to be carefully removed with the high-speed drill, taking care to avoid the adjacent horizontal portion of the vertebral artery. The initial entry point can be made with the high-speed drill, with careful preservation of the C2 root traversing caudally, the vertebral artery coursing laterally, and the thecal sac medially. The head of the screw sits above the level of the posterior arch, such that a portion of the screw is exposed from the screw head down to the entry point of the lateral mass. Because the exposed screw is immediately adjacent to the traversing C2 root, a partially threaded lag screw is useful to prevent irritation of the root. B, C1 lateral mass fixation, axial view. The trajectory of the C1 lateral mass screw is illustrated. Although this is confirmed by reviewing the preoperative axial CT imaging, the trajectory is typically 10 to 15 degrees lateral to medial. Again noted is a portion of exposed screw from the entry point in the lateral mass to the screw head at the level of the posterior arch. Partially threaded lag screws are used to prevent possible irritation of the traversing C2 root.

A no. 4 Penfield is used to delineate the medial border of the pars interarticularis of C2 and then the entry point for the C2 screw is made with the high-speed bur drill in the cranial and medial quadrant of the isthmus surface of C2 (Fig. 298-4). A pilot hole is made with a 2-mm bit, just perforating the opposite cortex. The direction of the bit is approximately 20 to 30 degrees in a lateral to medial and cephalad trajectory. The integrity of the pilot hole is confirmed with the ball probe. The hole is tapped, and then a 3.5-mm polyaxial screw of the appropriate length is inserted bicortically. Contoured rods are placed in the screw heads and secured in position. The exposed bone of C1 and C2 is decorticated and autologous iliac crest cancellous bone is placed over the decorticated bone. The patient is placed in a soft cervical collar for 2 to 3 weeks, postoperatively.

FIGURE 298-4 A, C2 pedicle screw fixation—posterior view. The entry point is depicted on the right, with the final position of the screw head on the left. The entry point is 5 mm lateral and 7 mm inferior to a point determined by the intersection of a horizontal line over the C2 lamina and a vertical line along the lateral aspect of the spinal canal.⁶⁰ B, C2 pedicle screw fixation, axial view. The trajectory of the C2 pedicle screw is approximately 20 to 40 degrees in the transverse plane. This can be verified by review of preoperative axial CT imaging and intraoperatively by palpating the medial border of the C2 pedicle. C, C2 pedicle screw fixation, lateral view. The trajectory of the C2 pedicle screw is 20 degrees in the sagittal plane.

Harms and Melcher ⁵⁹ reported satisfactory screw placement and reduction in all 37 patients in their series. They reported no neural or vascular damage related to the technique. This technique is more widely applicable and simpler than the transarticular technique, yet some patients with narrow C2 pars or medially located foramen transversarium still pose a threat to safe screw placement.^60,⁶¹

Bilateral, Crossing C2 Laminar Screws

Wright ⁶^–⁸ and Leonard⁶ described a new technique for rigid screw fixation of the axis and incorporation into atlantoaxial fixation or subaxial cervical constructs with insertion of polyaxial screws into the laminae of C2 in a bilateral, crossing fashion (Fig. 298-5). These screws are connected to C1 lateral mass screws similarly to the C1-2 rod-cantilever technique. This technique provides the benefit of allowing safe rigid fixation of C2 without placing the vertebral arteries at risk of injury. Although C1 screw placement still requires aid in visualization with fluoroscopy, the C2 screws can also be placed without fluoroscopic guidance or surgical navigation. This technique requires intact posterior elements of C2, unlike the previous techniques described.³

FIGURE 298-5 A, C2 translaminar screw fixation—posterior view. The intralaminar portion of the screws is shown in a transparent fashion for clarity. The entry point is at the junction of the spinous process and lamina. Care must be taken to leave room for both screws by placing one on an entry point toward the rostral aspect of the lamina and the other toward the caudal aspect. If the first entry point is made midway, it may be difficult to place the second screw. B, C2 translaminar screw fixation, axial view. The trajectory and length of each screw are depicted. This illustration approximates the axial CT image necessary to determine screw length, with the length measured from the entry point to a termination short of the C2-3 facet. The downslope of the lamina from the spinous process to facet is also depicted. This slope needs to be visualized intraoperatively and the trajectory of the screw matched accordingly. The screw trajectory should be slightly less than the visualized downslope so that any possible violation occurs dorsally rather than ventrally into the canal.

Patients are placed in the prone position with the head and cervical spine maintained in a neutral position by use of the Mayfield head holder (Integra Life Science). Exposure of the posterior upper cervical spine and craniocervical junction is performed to identify the posterior arch of C1 to the lateral aspect to visualize the bilateral lateral masses. The spinous process, lamina, and medial lateral masses of C2 are then exposed. C1 lateral mass screws are placed using the Harms technique described in the previous section.

The high-speed bur drill is used to open a small cortical window at the junction of the C2 spinous process and lamina on the right, close to the rostral margin of C2 lamina. Using a hand drill, the contralateral (left) lamina is carefully drilled to a depth (typically 28 to 30 mm, but this distance is ascertained from the preoperative axial CT images), with the drill visually aligned along the angle of the exposed contralateral laminar surface. The trajectory is kept slightly less than the downslope of the lamina to ensure that any possible cortical breakthrough would occur dorsally through the laminar surface rather than ventrally into the spinal canal. A small ball probe is used to palpate the length of the hole to verify that no cortical breakthroughs into the spinal canal have occurred. A 3.5- or 4.0-mm diameter polyaxial screw is carefully inserted along the same trajectory. The length of the screw is typically 28 to 30 mm, but is again determined from the preoperative axial CT scan. In the final position, the screw head remains at the junction of the spinous process and lamina of the right, with the length of the screw within the left lamina.

A small cortical window is made at the junction of the spinous process and lamina of C2 on the left, close to the caudal aspect of the lamina. Using the same technique as described earlier, a screw is placed into the right lamina, with the screw head remaining on the left side of the spinous process. After screw placement, all exposed laminar surfaces are decorticated with the high-speed drill, while taking care not to unroof the intralaminar screws at C2. The C1 lateral mass screws are connected to the crossing, bilateral C2 laminar screws with posterior rods. Autologous iliac crest bone grafts are wedged under the rods between the lamina of C1 and the spinous process and lamina of C2, as well as into the C1-2 joint. Patients are treated with a cervical collar for 6 weeks, postoperatively.

It is important to palpate the drilled hole before screw insertion to minimize the chance of the screw violating the canal. Intraoperative and postoperative radiographs are not predictive of proper screw placement.⁶² A variation of Wright’s technique as reported by Jea and associates⁶³ involves placement of a small cortical window at the lateral aspect of the lamina, near the laminar-facet junction, to visualize the distal tip of the screw to verify that the screw has not violated the spinal canal.

In the initial series of 10 patients studied by Wright, no intraoperative or immediate postoperative complications were encountered. All C2 screws were placed satisfactorily without any technical problems during screw insertion. There were no neurological or vascular complications noted. All patients demonstrated stability on flexion-extension radiographs obtained at 6 weeks. The first two patients treated with this technique had thin-cut CT scans with sagittal and coronal reconstructions obtained at 6 months after surgery. Both patients exhibited solid arthrodesis with bridging bone from the posterior arch of C1 to the lamina of C2.³²

A larger series with 20 patients with a follow-up period of at least 1 year reported 100% fusion rates and no complications.⁸ Other authors have subsequently published their experience with C2 laminar fixation⁶⁴^–⁶⁷ with excellent results. Although in some cases the screws were found to violate the spinal canal, no neurological or vascular injury occurred.

Initial biomechanical testing suggests atlantoaxial stabilization with C2 translaminar screws to be equivalent to transarticular C1-2 fixation and C1-2 rod-cantilever techniques. Gorek and associates ⁶⁸ compared C2 translaminar screw fixation with C1-2 transarticular fixation and reported equivalent stability and Lapsiwala and associates⁶⁹ concluded that C2 translaminar screws provide adequate stiffness compared with anterior and posterior transarticular screws, and C2 pedicle screws.