Adjunct to Noninstrumented Fusions

Pseudarthrosis rates are reported at 10% with bone graft alone during fusion procedures for one level, and possibly greater than 30% with more than two or three levels.⁶ Internal fixation has been used extensively to assist with fusion procedures in modern spine surgery.^7,8 Facet fixation specifically has been shown to decrease pseudarthrosis rates over noninstrumented fusion and to have a low incidence of complications.⁹

Dorsal fusions were largely supplanted by dorsolateral fusions in the 1980s owing to a decrease in pseudarthrosis rates. Kornblatt et al. also showed that internal fixation, specifically with facet fixation or pelvic rods, improved the rate of fusion (87% vs. 76% without fixation) and time to radiographic fusion (6.2 months vs. 10.5 months) significantly.⁶ This was also shown by Jacobs et al., using translaminar facet fixation and judging fusion by oblique and flexion-extension films.¹⁰ Both studies used patients who had a pseudarthrosis from a prior procedure. They demonstrated that facet fixation can promote fusion after failed procedures—with the caveat that outcomes deteriorate with each successive surgery in most spinal procedures.¹¹

Adjunct to Anterior or Posterior Lumbar Interbody Fusion

Posterior lumbar fusion may be used for a variety of indications. Because facet fixation does not allow distraction or manipulation of alignment, the majority of uses involve restriction of movement to facilitate fusion, either because a discectomy has been performed or because the patient experiences painful symptoms with motion.

One of the uses that has caused a resurgence in facet screw popularity is as an adjunct to anterior lumbar interbody fusion (ALIF), when there is no need for posterior nerve root decompression. Failure of fusion with ALIF alone has been reported in up to 24% of cases.^12,13 Cadaveric studies have shown that ALIF alone allows more movement during extension with little preload than does the preoperative spine—a risk for graft displacement and poor fusion. Facet fixation with translaminar screws enhances stability, returning motion to the level allowed preoperatively.¹⁴ Kandziora et al. showed the equivalence of ipsilateral facet screws to translaminar screws with regard to range of motion, neutral zone, and elastic zone.¹⁵ However, they also demonstrated improved parameters in all test modes with pedicle screws.

For similar reasons, facet screws are also used as an adjunct to posterior lumbar interbody fusion; this was described as early as 1988, with only 1 complication in 35 patients and fusion apparent in all with the use of postoperative thoracolumbosacral orthosis immobilization.¹⁶ This procedure is useful to enhance fusion acquisition and prevent motion that may lead to graft displacement. With posterior lumbar interbody fusion, no further surgical exposure is necessary to place the screws (Fig. 154-1).

FIGURE 154-1 Model demonstrating orientation of cortical lag screws across inferior facet into superior facet and underlying pedicle.

Segmental Instability

Patients who have become unstable though degeneration or trauma may benefit from a dorsal arthrodesis. Although pedicle screw fixation with rods is increasingly used, facet fixation remains a viable alternative. However, the alignment must already be appropriate because facet fixation does not permit the application of enough leverage to mobilize spinal segments and alter the alignment.

Painful Disc Syndromes

The theory behind posterior lumbar fusion for degenerative disc disease is elimination of motion at the affected segment. Outcomes have been shown to be equivalent to interbody procedures in multiple studies.^17,18 It is thought that the painful portion of the segment is the disc as a whole or the anulus fibrosus and, therefore, dorsal fusion alone may yield poor outcomes unless the disc material is removed, such as with an interbody procedure (see earlier discussion). However, dorsal fusion alone has significant advantages, such as shorter operative time, less risk for complication, lower cost, preservation of the anterior column, and ease of procedure.

Painful Facet Syndrome

In the healthy spine, the nociceptors in the facet joint capsule fire only under supranormal physiologic conditions. However, it has been shown that with inflammation, chemical mediators can sensitize the pain sensors to fire excessively or spontaneously. One study suggested that 15% of patients with low back pain had pain of facet origin.¹⁹ In patients with lumbar facet syndrome, Helbig and Lee demonstrated a positive response to facet block and facet rhizolysis in 50% to 60% of cases.²⁰

The existence of the so-called facet syndrome is still rather controversial. In 1996, Cavanaugh et al. showed in rabbits that algesic chemicals, such as carrageenan or substance P, can sensitize pain receptors or induce persistent firing of these neurons with small movements.⁴

Results of studies examining the therapeutic potential of facet injections have been contradictory, with some showing significant improvement and others showing equivalent results with the injection of normal saline.²¹ Once nonsurgical treatment has been exhausted, facet fixation can eliminate motion at the segment, thereby decreasing or eliminating pain.

Technique

The first description of the translaminar approach was by Magerl in 1984.³ However, the topic of the paper was actually an external fixator and the method of placing facet screws was only vaguely referenced. Diagrams in the article clearly show the path of a translaminar facet screw: entering at the base of the spinous process, traveling the length of the contralateral lamina, crossing the facet joint, and ending at the base of the transverse process (Fig. 154-2). Multiple authors have described percutaneous translaminar screw placement using fluoroscopy.^22,23

FIGURE 154-2 Translaminar screw fixation of facet joints at the thoracolumbar junction and the lumbar spine, where the joints are oriented toward the sagittal plane. To prevent splitting of the lamina, the screws should not be applied as lag screws.

Comparison with Other Internal Fixation Techniques

Translaminar screws are thought to be more stable than the shorter construct associated with ipsilateral placement, but many studies have shown similar outcomes with both types. The ease of learning this technique and the limited time required to become facile make it a relatively convenient adjunct to many procedures. In addition, because both types of facet fixation can be performed percutaneously, they are in accord with the current trend toward minimally invasive spine procedures.

However, translaminar placement does require intact dorsal elements. Many surgeons have reported discectomy, lateral recess decompression, and even partial central decompression with preservation of sufficient lamina to perform the translaminar approach, according to Jacobs et al.¹⁰ An isthmic spondylolisthesis with a pars defect is therefore an obvious contraindication to this procedure.

Because of the longer course of the translaminar screws, more muscle dissection must be performed and greater technical skill is required compared with the ipsilateral approach. The longer screw trajectory also makes canal intrusion more likely than with ipsilateral placement, which limits the path of the screw to the facet joint alone.

Concern also exists regarding the notion that a limited dorsal fusion has the potential for kyphosis over time, as the disc loses height. In addition, although improved pseudarthrosis rates are reported compared with uninstrumented fusion, the rate remains as high as 9% in some reports.¹⁰

Comparison with Pedicle Screws

There are some obvious advantages to the facet fixation technique over the more frequently used pedicle screw fixation technique: lower cost, less dissection, more room for bone graft, lower complication rates, and a lower rate of infection. The attraction of the pedicle screw technique is related to the prevailing notion that facet fixation is less stable. Studies have shown equivalence in biomechanics and stability between the two methods in objective tests, although most such studies have been performed in cadaver models.^23–25 In a prospective study of patients undergoing translaminar facet fixation versus pedicle fixation, Tuli et al. demonstrated a higher nonunion rate, as confirmed by radiographs, in patients undergoing facet fixation: 17.5% versus 2.7%. This finding correlated with symptoms. However, they also found a non–statistically significant difference in end-fusion degeneration, with pedicle screws having the higher rate (13.5% vs. 5%).²⁶

Pedicle screw fixation requires the insertion of more hardware because of the presence of rods connecting the inserted screws, as well as cross bars. This bulk and higher profile may be one of the reasons that hardware removal is more common after pedicle screw than facet screw insertion. In a comparison with pedicle screws in circumferential interbody fusion, Best and Sasso found that less than 5% of the facet screw group had reoperations due to continued pain, whereas 37.5% of the pedicle screw group returned for another procedure.²⁷ As an aside, one of the facet screw patients who was found to have a pseudarthrosis underwent pedicle screw placement without removal of the facet screws.

Of note, the proximity of pedicle screws to the spinal cord causes significant imaging artifact that can make evaluation and diagnosis of complications difficult. The distance of facet screws from the canal reduces this artifact.

Technique

Ipsilateral insertion of facet screws has the advantage of not requiring an intact lamina and spinous process, as does the translaminar approach. In 2009, Su et al. analyzed 80 cadaveric facet joints from 37 spines in an anatomic study of appropriate screw placement.²⁸ The study was designed to determine the radiographic characteristics during percutaneous screw placement that would be expected to traverse the center of the facet joint without complications. They stated that the L2-3 segment could not be instrumented in this fashion because of the vertical orientation of the facet joint. For segments L3 through S1, they observed an appropriate orientation in the mediolateral and craniocaudal planes at each level. Rostrocaudally, the screw trajectory should begin over the caudal end plate of the rostral level. Mediolaterally, the view should show the screw’s starting point to be located over the medial border of the pedicle.

The angle of the screw should be approximately 15 degrees in the axial plane and approximately 30 degrees in the sagittal plane—varying slightly depending on the levels being fused (Fig. 154-3). Another technique orients the screw such that it is parallel to the caudal edge of the lamina.²⁹ The tip of the screw should end up caudally and laterally in the caudal pedicle on the anteroposterior view, and in the caudal half of the pedicle on the lateral view. Su et al. also state that a 35-degree oblique angle on fluoroscopy is optimal for viewing final placement.²⁸

FIGURE 154-3 A, Caudal angulation (ϕ) of the pin in the sagittal plane relative to a line drawn parallel to the intervertebral disc of the instrumented level. Note that the pin terminates in the inferior half of the pedicle at the pedicle-vertebral body junction. B, Lateral angulation (β2) of the pin in the axial plane relative to the midline spinous process.

Comparison with Translaminar Screws

The initial impetus for developing ipsilateral screws was to eliminate the long path of translaminar screws and the requirement for intact dorsal elements (spinous process and laminae). Multiple studies have shown equivalence between ipsilaterally placed screws and translaminar screws.¹⁵

The ipsilateral method of fixation requires a smaller exposure and can easily be added to a dorsal decompression with little or no additional dissection. Even when not placed percutaneously, ipsilateral facet screws require less retraction than translaminar placement.

There are significant limitations with ipsilateral placement, however, predominantly with regard to anatomy and biomechanics. First, certain levels do not lend themselves to the application (L2-3 and above) because of the orientation of the facets and the difficulty associated with obtaining the correct angle of insertion. The vertical and sagittal facet orientation dictates that the screw and driver be positioned in the space occupied by the spinous process, which could be accomplished only with dorsal element removal. This eliminates the percutaneous placement option and removes one of the advantages of ipsilateral placement over the translaminar approach. Therefore, translaminar fixation would be more appropriate at the upper lumbar levels.

Neither type of facet fixation permits the distraction and manipulation of spinal alignment, as can be accomplished with pedicle screws. This makes them less useful in the setting of traumatic fracture or malalignment from malignant or infectious etiologies, when correction of an angulation is required.

Comparison with Pedicle Screws

Although facet screws have always been considered inferior to traditional pedicle screw fixation in rigidity and stability, data suggest that ipsilateral facet screw placement may be equivalent biomechanically. In cadaveric models, Ferrara et al. conducted short- and long-term cyclical testing of facet screws and pedicle screws with flexion-compression and torsional loading. Although the two systems were equal in most parameters assessed, they found facet screws provided 30% more stiffness during flexion in short-term cycling. Agarwala et al. performed another cadaver model study showing that the Boucher method of facet screw placement demonstrated a trend toward reduced motion compared to pedicle screws in flexion extension, but not in lateral bending or axial rotation.³⁰ This trend was only shown in primary fusion procedures without ALIF, and did not meet statistical significance. Although performed in vitro, these studies provide support for the theory that facet fixation may be equivalent in stability to pedicle screw fixation and able to maintain its mechanical integrity over long-term cycling.²⁹