Posterior cervical spine arthrodesis incorporating C₂ laminar screw fixation in the treatment of cervical spine injury

Abstract

Objective:  To investigate the clinical application and efficacy of an internal fixation technique incorporating C₂ laminar screws for upper cervical spine injury.

Methods:  Using a posterior cervical approach, incorporating C₂ laminar screw fixation and bone grafting were performed on 20 patients with cervical spine injury. There were 12 male and 8 female patients, with a mean age of 45.6 years (range, 32–71 years). All patients were evaluated by X‐ray, computed tomography (CT) and magnetic resonance imaging (MRI).

Results:  The patients were followed up for 11–35 months (mean, 15 months), and bony union was achieved in all patients. There were no spinal cord or vertebral artery injuries during surgery, and only two instances of vein clump injury, in both of which the bleeding was controlled successfully. Postoperative CT scans showed that all the C₂ laminar screws had been placed properly, and were not encroaching on the spinal canal. No spinal instability, evidence of hardware failure or screw loosening was found during the follow‐up period in any patient.

Conclusion:  Crossing C₂ laminar screw internal fixation technique is simple, and is not limited by the position of the vertebral artery in the body of C₂. The laminar screw method avoids arterial injuries and also can be used as a salvage method after previous misinsertion. As all relevant structures are directly visualized during C₂ laminar screw placement, this kind of technique may be applicable to a large number of patients.

Introduction

Instability of the upper cervical spine requiring surgical stabilization can be treated by a variety of techniques. Recently many reports have described C₂ pedicle screw fixation for atlantoaxial stabilization, in conjunction with C₁ lateral mass screws in a rod‐cantilever construct ¹ , ² , ³ . Other reports have described similar C₁ and C₂ screw fixation connected by rigid constructs ⁴ , ⁵ , ⁶ . C₁‐C₂ or C₂‐C₃ rod‐cantilever constructs have been suggested as safer procedures, being applicable to more patients than is transarticular screw fixation ² . C₂ pedicle fixation has also been used by many to incorporate C₂ into subaxial fusion constructs. However, C₂ pedicle screw placement remains technically difficult due to the position of the foramen transversarium of the cervical spine, the varying size of the pedicle of the vertebral arch in different individuals, and the fact that in some cases there is either asymmetry or hypoplasia of the vertebral artery. Therefore, the insertion of these screws carries a high risk of injury to either the adjacent vertebral artery or spinal cord. Vertebral artery injuries can cause serious complications, such as brain infarction in the regions of the cerebellum and posterior cerebral artery, as well as life‐threatening brain stem infarction ⁷ , ⁸ , ⁹ . Some cadaveric studies have also suggested an unacceptably high rate of violation of the foramen transversarium with C₂ pedicle screw placement ¹⁰ .

To overcome the aforementioned shortcomings of C₂ pedicle screw placement, we have developed a technique for inserting a C₂ intralaminar screw which is placed ipsilateral to the aberrant vertebral artery. The pilot hole for the screw starts at the base of the spinous process and its junction with the ipsilateral lamina, and is directed into the contralateral lamina. A polyaxial screw is introduced into the pilot hole, and longitudinal rods are used to connect the C₁ and C₃ screws. Our own limited clinical experience has shown this technique to be effective for promoting atlantoaxial fusion. We herein describe our method for performing atlantoaxial or C₂ to C₃ fixation using intralaminar screws in 20 patients with upper cervical spine injury. The technique was designed, and is performed, with the objective of avoiding damaging to the vertebral artery.

Materials and methods

From June 2005 to May 2008, 20 cases with cervical vertebral injury underwent incorporating C₂ laminar screw fixation and bone grafting via a posterior cervical approach. There were 12 male and 8 female patients, with a mean age of 45.6 years (range, 32–71 years). Eight patients had type II odontoid process fractures combined with backward dislocation of the atlanto‐axial joint, two had forward dislocation of the atlantoaxial joint, two had nonunion of odontoid process fractures, three had type III odontoid process fractures combined with atlantoaxial joint instability and rupture of the axis transverse ligament, two had type II Hangman fractures combined with instability of C_2,3, 1 with forward dislocation of atlanto‐axial joint combined with rupture of the axis transverse ligament, and two had congenitally loose odontoid processes. All cases had symptoms of atlantoaxial instability. One starting point for fixation of a C₂ laminar screw was at the junction of the spinous process and lamina of C₂, close to the rostral margin of the C₂ lamina, with the drill visually aligned along the angle of the exposed contralateral laminar surface. The trajectory was kept at a slightly smaller angle than the slope of the lamina to ensure that any possible cortical breakthrough would occur dorsally through the laminar surface, rather than ventrally into the spinal canal. Another screw starting point was at the junction of the spinous process and lamina of C₂ on the other side, close to the caudal aspect of the lamina. The Vertex cervical internal system was used for the internal fixation (Vertex; Medtronic Sofamor Danek, Memphis, TN, USA). After successful reduction had been obtained by traction, combined fixation with axis laminar screws and posterior autogenous iliac graft fusion were carried out according to the severity of the defects being treated. All patients were evaluated by X‐ray, CT and MRI.

Surgical techniques

The patients were placed in the prone position with the head and cervical spine maintained in a neutral position by a Mayfield head holder (Ohio Medical Instruments, Cincinnati, OH, USA) Exposure of the posterior upper cervical spine and craniocervical junction was then accomplished in the usual manner. In order to avoid intersection of the trajectories of the bilateral intralaminar C₂ screws, an entry point was selected near the cranial edge for the superior C₂ laminar screw, while the inferior C₂ laminar screw was started near the caudal edge. A unilateral C₂ intralaminar screw was located at the junction of the spinous process and the laminar, midway between the cranial and caudal aspects of the laminar. A hand drill was directed within the contralateral lamina, between the ventral and dorsal cortices (intralaminar), and visually aligned along the angle of the exposed contralateral laminar surface. The trajectory was kept at a slightly smaller angle than that of the slope of the lamina to ensure that any possible cortical breakthrough would occur dorsally through the laminar surface rather than ventrally into the spinal canal. The endpoint was located just caudal to the junction of the pars interarticularis and the lamina. The appropriate screw length was determined by using a routine depth gauge. For fixation of C₂, a high‐speed drill was then used to make a small hole in the cortex at the junction of the spinous process and lamina of C₂. The atlantal pedicle screws and a contralateral C₂ pedicle screw were placed according to the method described by Harms ¹ and Resnick ² with 22 mm (Φ3.5 mm) polyaxial lag screws (Vertex) inserted unilaterally or bilaterally.

Results

All patients were followed up for 11–35 months (mean, 15 months). Favorable bony union was achieved in all patients and all screws remained in a satisfactory position without screw displacement or any perioperative or postoperative complications. No loosening or breakage of screws occurred. Satisfactory results were achieved in all patients. No neurological deterioration occurred. There were no spinal cord or vertebral artery injuries during surgery, and only two cases of vein clump injury, in both of whom the bleeding was controlled successfully. The postoperative CT scans showed that all C₂ laminar screws had been placed properly, and were not encroaching on the spinal canal. There was no spinal instability, evidence of hardware failure or screw loosening during the follow‐up period in any patient.

Typical cases

Two patients with upper cervical spine disorders requiring axis stabilization were treated in our hospital.

Case one

A 52‐year‐old man who was injured in a motor vehicle accident in December 2007 sustained a type II odontoid process fracture and incomplete atlantoaxial backward luxation. He was urgently transferred to our hospital after the injury. Paresthesias in the upper left leg were present after the injury, as well as incomplete paralysis and numbness in the right arm. Radiographic examination revealed a type II odontoid process fracture with incomplete posterior displacement of the odontoid process (Fig. 1A). After three days of skull traction, the atlantoaxial luxation had been reduced. However, instability still existed between C₁ and C₂, and a preoperative CT scan showed that the C₂ pedicles were so small that screws could not be inserted into the C₂ vertebral arch. Thus posterior fusion was performed for C₁ and C₂ fixation with C₁ pedicle screws and C₂ laminar screws (using bilateral‐crossing laminar screws designated as ‘2LAM’ in C₂, Figs. 1B–D). Four weeks after the operation, the patient was able to walk while wearing a brace. By about two months after the operation, the preoperative symptoms and signs had disappeared. At three months after the operation, no instability of the cervical spine was observed and the fixation was stable.

Figure 1.

A 52‐year‐old man with a type II odontoid process fracture and incomplete posterior atlantoaxial luxation. (A) Preoperative CT scan with sagittal and coronal reconstruction of the cervical spine: the arrows show the overhang of the atlantal lateral mass on the axis. (B) Postoperative X‐ray film. (C) Postoperative CT: Because the size of the C₂ pedicle was small, we used bilateral‐crossing intralaminar screws in the axis and pedicle screws in the atlas. The arrows demonstrate reduced C₁‐C₂ articulation. (D) Postoperative CT showing good screw placement of C₁ pedicle screws and C₂ laminar screws, bone grafting was performed with bone harvested from the iliac crest.

Case two

A 41‐year‐old man with a Hangman's fracture complained of intolerable pain and dysfunction in the neck. In June 2007, the patient had been sitting in the front passenger seat of a car when it fell into a river three meters below the road. Radiographic examination revealed a Hangman's fracture with an unstable C₂‐C₃ teardrop fracture (Fig. 2A). However, the patient had no neurologic symptoms and signs. Because of the gradually increasing instability between C₂ and C₃, C₂‐C₃ fixation was performed with the laminar screw system. When a guide wire was inserted under the guidance of C‐arm, there was massive bleeding from the hole. The bleeding was controlled by plugging the hole with bone wax and compressing it. Since re‐insertion of the guide wire or a screw could result in further vascular injury, the laminar screw system was used for C₂‐C₃ fixation (application of an intralaminar screw and a single pedicle screw in C₂ described in the data as ‘1PED1LAM’) and two pedicle screws were applied in C₃ (Figs. 2B–D). The day after the operation, the patient was encouraged to walk while wearing a soft neck brace. At six months after the operation, neck pain had improved and the fixation was stable. The range of neck movement was not restricted. This method not only rebuilt the stability of C₂‐C₃ but also preserved the rotation function between atlas and axis.

Figure 2.

A 41‐year‐old man with Hangman's fracture. (A) Lateral plain radiograph and CT scan revealing Hangman's fracture (left black arrow) with an unstable C₂‐C₃ teardrop fracture (right black arrow). (B) During C₂‐C₃ fixation, massive bleeding was observed at the insertion site of a pedicle screw on the right side of C₂, so an intralaminar screw and a single pedicle screw were used in C₂ and two pedicle screws in C₃. (C) Postoperative radiograph showing that the screw‐rod is properly constructed. (D) CT scan showing good screw placement of the C₂ pedicle screw, and C₂ laminar screws can be seen on the left side.

Discussion

Historically the standard fixation for atlantoaxial segment instability has been posterior wiring using the Gallie or Brooks techniques ¹¹ , ¹² , ¹³ . However, such techniques generally require postoperative halo immobilization and are associated with relatively high failure rates. Also, wiring techniques cannot be applied when the posterior C₁ arch has been disrupted. Transarticular C₁‐C₂ screws using the Jeanneret‐Magerl technique provide very rigid fixation across the atlantoaxial joint ¹⁴ , but they are technically demanding due to the danger of vertebral artery injury. Cadaveric studies have shown an incidence of up to 20% of anomalies that preclude safe transarticular screw placement ¹⁵ , and clinical experience suggests that the incidence of vertebral artery injuries during screw placement is 4% ¹⁶ . Since Abumi et al. clinically applied cervical pedicle screws in 1994 ¹⁷ , such screws have been broadly used instead of posterior wiring for unstable cervical spines and spinal deformity, posterior wiring having been reported to be associated with a high frequency of pseudoarthrosis and nonunion (about 18%) ¹⁸ , ¹⁹ . Biomechanically, cervical pedicle screws are the most stable screws ¹⁵ . However, because the cervical pedicle is small and located adjacent to the vertebral artery laterally, the spinal cord medially, and the nerve roots vertically, their application carries a high risk of neurovascular complications during screw insertion ²⁰ . Lateral mass screws have also been reported to involve a risk of neurovascular injuries ²¹ , ²² . The laminar screw method is thus thought to be useful since arterial injuries can be avoided and it can also be used as a salvage modality after previous misinsertion.

The intralaminar screw technique, the newest technique described in the literature, has many advantages ²³ . First, it seems that it is the safest technique in regard to vertebral artery injury. Also, there is no need for an acute angle for placement of screws. C₂ intralaminar screws can be placed under direct vision without image guidance. The only drawback to the technique is its requirement for intact and adequately sized laminae. Biomechanically, this technique is comparable to posterior transarticular screw fixation and the C₁ lateral mass‐C₂ pedicle screw system with additional cable fixation in flexion‐extension. Although it is not very rigid in resisting lateral bending, the movement is still less than that of the intact spine. Lapsiwala's study demonstrated that additional cable fixation to the posterior transarticular screw improves resistance to flexion and extension, and this finding is consistent with previous research ²⁴ , ²⁵ . It has also been reported that intralaminar screw fixation possesses biomechanically equivalent strength to the transarticular technique ²⁶ , thus it is suggested that intralaminar screw fixation is an excellent method. The present authors recently compared the C₁ lateral mass‐C₂ pedicle screw technique and the C₁ lateral mass‐C₂ intralaminar screw technique, and found that those two techniques provide equivalent stability to the C₁‐C₂ complex.

The intralaminar screw is useful in patients in whom pedicle and/or transarticular screws cannot be placed. It is a safe technique and has been biomechanically validated ²⁶ . One major advantage of this technique is that there is no risk of vertebral artery injury because the screws remain in the posterior elements of the C₂ vertebrae. In addition, an intra‐operative navigation system is not necessary for this technique since the screws can be inserted into the laminae under direct vision. Also, intra‐operative navigation systems are not available in all facilities, and it has also been reported that 100% accuracy cannot be guaranteed even with computer‐assisted systems ²⁷ .

One potential drawback of this technique is that ventral penetration of a laminar screw into the spinal canal cannot easily be observed ²³ . Matsubara et al. believe that patients with unilateral occlusion or asymmetry of the vertebral artery are good candidates for this new technique ²⁸ . Even when patients have normal pedicle anatomy, the risk of screw malpositioning should not be overlooked ²⁹ . A literature review revealed that the incidence of asymmetry or hypoplasia of the vertebral artery is around 15%, and that damage to the dominant artery results in brainstem ischemia ³⁰ . The fatality rate for brainstem ischemia is reported to be 75% to 86% ³¹ . A strong fixation is meaningless if a lethal complication occurs. Laine et al. have stated that ‘to make placement of pedicle screws safer is obviously quite important in the current medicolegal climate, and it is important because of our dedication to patient safety’ ³² . We wholeheartedly agree with this statement. We therefore measured the laminar diameters preoperatively and evaluated the vertebral arteries with CT and MRA to ensure a good outcome.

To avoid vertebral artery injuries, we select intralaminar screws rather than pedicle or lateral mass screws for the same side as asymmetry or hypoplasia of the vertebral artery, and on the opposite side where there is a unilaterally occluded vertebral artery. This kind of technique is applied when screws cannot be inserted due to a small C₂ pedicle. Furthermore, in cases with screw misinsertion, we use intralaminar screws instead of again inserting screws into the vertebral arch or lateral mass.

We now list some key points:

• 1The intralaminar screw is useful in patients in whom pedicle and/or transarticular screws cannot be placed.
• 2The intralaminar screw method can avoid vascular injuries, and the technique can be used as a salvage treatment after misinsertion.
• 3Intralaminar screws, rather than pedicle or lateral mass screws, should be selected for the same side as asymmetry or hypoplasia of the vertebral artery, and on the opposite side where there is a unilaterally occluded vertebral artery.
• 4Intralaminar screw fixation possesses strength which is biomechanically equivalent to that of the transarticular technique.