Occiput

Formerly, fixation to the occiput was generally accomplished by means of stainless steel wire, generally 18 gauge, looped through adjacent burr holes or through the diploic space. Although this provided some degree of stability, several problems were noted, including pull-through, dural laceration with cerebrospinal fluid leakage, and difficult placement after suboccipital craniectomy. As a result, screw-based mechanisms of occipital fixation were developed.1,2

Bicortical placement of occipital screws is preferred. The pullout strength of bicortical screws has been noted to be 50% greater than that of unicortical screws; the pullout strength of unicortical screws is similar to that of 18-gauge wire.3 Screw location is equally important, as screws placed in the midline demonstrate greater insertional torque and greater pullout strength than screws placed more laterally.4 Despite the fact that longer screws may often be placed in the suboccipital region, 3.5-mm titanium screws longer than 7 mm typically break before pulling out,5 suggesting that the use of larger-diameter screws should be considered when possible. Also, though shorter, laterally placed occipital screws may offer less resistance to pullout, and constructs incorporating multiple laterally placed screws may perform equivalently to those with midline screws. There does not appear to be an advantage to the use of constrained (locking) screws in the suboccipital region.6

Promising alternative techniques have been described. Mingsheng et al. described a technique for the placement of screws into the diploic space of the occipital bone.7 Although no biomechanical data were presented, the use of screws with a mean length of nearly 26 mm was found to be possible. Pait et al. described a technique for “inside-outside” screw placement in the suboccipital region, in which a threaded stud is placed through a burr hole with the threads facing outward.8 The pullout strength in synthetic bone was found to be superior to wire cables and bone screws.9

Occiput–C1

In 2001, Grob described a technique for direct posterior occiput–C1 transarticular fixation, describing its use in a patient with occiput–C1 dislocation.10 In the case described, the construct was reinforced with a posterior occiput–C2 plate. Gonzalez et al. described a technique for direct fixation of the occiput–atlas using transarticular screws.11 Though allowing a significantly greater range of motion (ROM) at the occiput–C1 in flexion and extension in pure moment testing than an occipital plate–C2 construct, it was concluded that the technique was a useful alternative, especially in cases where suboccipital bone is deficient, and there is not significant instability at C1–C2.11 Incorporation of posterior occiput–C1 transarticular screws into larger constructs, though theoretically possible, has not been described. Goel and Laheri discussed the use of a C2 pars screw in isolation or in combination with a C1 lateral mass screw for the fixation of the cervical end of the occipitocervical plate/rod.1

Dvorak et al. described a technique for direct anterior occiput–C1 fixation, noting that this construct was as effective as two posterior fixation techniques in lateral bending and axial rotation but less effective when tested in flexion and extension.12 Given that instability in sagittal plane rotation is frequently encountered in pathological conditions affecting the craniovertebral junction, it is interesting to note the reported clinical successes.

Atlas

Traditionally, fixation to the atlas was generally accomplished by means of sublaminar stainless steel wire or, more recently, braided stainless steel or titanium cable. This technique remains useful in cases where the C1 arch is not deficient, fractured, or removed for decompression. It should be kept in mind, however, that the complication rate of sublaminar wire placement in the cervical spine has been estimated to be ~7%.13

Goel and Laheri first described the use of screws in the lateral mass of C1 in 1988, reporting excellent outcomes in 30 patients with atlantoaxial subluxation.1 Some authors have recently introduced polyaxial screws using the basic surgical technique described by Goel.1,14 These techniques obviate some of the deficiencies associated with sublaminar wire placement, particularly in the absence of a viable C1 lamina.

Landmarks for the placement of C1 lateral mass screws have been studied in detail.15,16 Gupta and Goel performed a cadaveric anatomical study of the C1 lateral mass and C2 nerve root, noting that the mean distance from the midline to the midpoint of the C1 lateral mass was 17.6 mm, and the mean width of the posterior arch at the point of the vertebral artery overpass was 4.7 mm.16 With an entry point above the C2 nerve root (or the space available after the sectioning of the C2 root), at the junction of the C1 posterior arch and the midpoint of the posteroinferior part of the C1 lateral mass, screws can be placed with a medial angulation of 9.8 to 21.6° and a superior angulation of 17.8 to 28.8°, depending on the individual patient’s anatomy.1 Using this technique, a screw thread length of 19.1 to 25.9 mm can be accommodated within the lateral mass.17 Resnick et al. examined computed tomography (CT) scans in 50 consecutive patients without known pathology of the craniocervical junction and found that placement of screws into the lateral mass of C1 was possible in every instance.18 Although C2 nerve root irritation as a result of C1 lateral mass placement inferior to the C1 arch is a potential complication, the incidence has been reported to be 12%, and the symptoms are temporary.19 C1 lateral mass screws, when placed using a standard entry point and trajectory, exhibited a pullout force of 1818.16 N, comparable to that of C2 pedicle screws.17 Incorporation of C1 lateral mass screws into a longer occipitocervical instrumentation construct, though possible in many instances, may be cumbersome secondary to crowding of screw heads at the occipitocervical rod bend ( Fig. 7.1 ).

A variant technique for placement of C1 lateral screws has been described by Tan et al., using an entry point 18 to 20 mm lateral to the midline and 2 mm superior to the inferior border of the arch.20 The authors determined that a screw with a mean length of 29.65 mm could be placed in 92% of 50 anatomical specimens. The primary limiting factor was the thickness of the posterior arch in the region of the vertebral artery groove.

Axis

The use of sublaminar wires or cables remains a viable option for fixation at C2, particularly in the setting of diminutive or malformed posterior elements, so long as the C2 lamina remains structurally intact. However, the development of several screw-based fixation options has greatly decreased the general applicability of this technique.

Screws placed into the pars interarticularis of C2 have gained wide acceptance as caudal anchors for occipitocervical instrumentation constructs.1 Following a modification of the technique used in the subaxial cervical spine, screws are placed using an entry point at the junction of the inferior border of the lamina and the midpoint of the inferior facet process of C2, with a trajectory 15° medial and parallel to the superior surface of the lateral mass (usually ~35° superior) in the sagittal plane.1,21 Bicortical screw purchase is not used. In case of a deep vertebral artery groove, screw length may be shortened to avoid injury. A minor, though potentially useful variant of this technique employs a less medially directed trajectory, similar to that of a C1–C2 transarticular screw (see the later description), through the lateral mass of C2.1 This technique allows the later conversion to a C1–C2 transarticular screw, if necessary.

Goel and Laheri described placement of screws into the pedicle of C2 ( Fig. 7.2 ).1 A great deal of confusion exists in the literature regarding the nomenclature C2 “pedicle.” For the purposes of this discussion, a C2 pedicle screw is defined as one that passes through the C2 pedicle along its axis and terminates within the body of C2 ( Fig. 7.3 ). Usually with the assistance of intraoperative neuronavigation techniques, the pedicle of C2 may be cannulated using a 39° medial trajectory parallel to the inferior end plate in the sagittal plane.22 Radiographic analyses have shown that 91% of C2 pedicles could accept a screw at least 4 mm in diameter.18,23 C2 pedicle screws have been shown to have significantly higher resistance to pullout than C2 pars interarticularis screws22 and to provide caudal fixation equivalent to C1–C2 transarticular screws in finite element modeling.24 C2 pars interarticularis and pedicle screws are readily incorporated into longer occipitocervical constructs.

Goel and Kulkarni placed the screw into the spinolaminar junction, incorporating the thickness of the spinous process.25 Wright described the technique for placement of intralaminar C2 screws.26 Using an entry point at the base of the C2 lamina, crossing screws may be placed into the cancellous bone between the cortices ( Fig. 7.4 ). Gorek et al. demonstrated equivalent ROM reduction in C1–C2 constructs anchored by intralaminar C2 screws and C2 pedicle screws.27 When employed as a salvage technique, C2 lamina screws exhibit a pullout strength at least equivalent, if not superior, to screws placed into the pars interarticularis of C2.28 Though incorporation into longer constructs generally requires three-plane rod bending or special off set connectors, C2 lamina screws may be particularly useful in situations where the C2 lateral mass or pedicle is attenuated. The placement of the screw at the spinolaminar junction as described by Goel and Kulkarni25 appears to provide more stability to the construct than the placement of the screw in the lamina.