A free‐hand technique for pedicle screw placement in the lower cervical spine

Abstract

Objective:  To describe a free‐hand method for pedicle screw placement in the lower cervical spine with no intraoperative imaging monitors, and to evaluate the safety of this technique.

Methods:  A study of the free‐hand technique of cervical pedicle screw placement was conducted by postoperative radiological review and follow‐up. Thirty‐six patients who had had cervical reconstruction with posterior plate utilizing pedicle screw fixation, and been followed for a minimum of 2 years, were studied. The position of the pedicle screw was evaluated by postoperative oblique radiographs and axial computed tomograms. Clinical outcomes were measured by Odem's criteria.

Results:  A total of 144 screws of diameter 3.5 or 4.0 mm were inserted into the cervical pedicles in 36 patients. Postoperative images showed that 16 (11.1%) of the screws had penetrated the pedicle walls. Among them, 10 (6.9%) screws had penetrated the lateral, 4 (2.8%) the superior and 2 (1.3%) the inferior walls. However, there were no neurological or vascular complications related to the malpositioned screws during a minimum of 2 years follow‐up. In addition, Odem's scores were applied postoperatively in all patients except one with complete neurological deficit.

Conclusion:  Based on 144 screw placements, cervical pedicle screw insertion utilizing a free‐hand technique without intraoperative imaging guidance seems to be safe and reliable. However, solid knowledge of the anatomy of the cervical pedicle and adjacent neurovascular bundles, and careful preoperative review of cervical images, are imperative for successful screw placement in the cervical spine.

Introduction

Cervical pedicle screws can be used effectively in reconstruction of the cervical spine. The superior stability, fixation, and resistance to screw pullout provided by this technique, compared with the lateral mass plating system, has been demonstrated in animal models and in human cadavers ¹ . It has been increasingly used since the concept was first described by Abumi et al. in 1994 ² . This sentence has been reworded for clarity. Has the meaning been retained?

However, there is a significant incidence of vertebral artery or nerve root injuries with this technique. It is therefore important to examine how to place screws into the pedicles of lower cervical vertebrae safely and simply. In the present study, 36 patients whose cervical spines had been fixed with pedicle screws by the freehand technique were reviewed, and the feasibility of the technique was also investigated.

Materials and methods

Patient data

From December 2001 to November 2003, 36 patients were treated with posterior cervical arthrodesis with the AXIS system in our hospital. There were 25 males and 11 females, with an average age of 42.3 years (range, 31–73). Twenty‐four patients were treated for cervical trauma, comprising four cases of C4 fracture, nine of C5‐6 fracture, seven of C6‐7 fracture, three of C4‐5 fracture/dislocation, and one of C3‐4/C6‐7 fracture/dislocation. In addition, there were 11 incomplete and three complete spinal cord injuries. The other 10 patients included four cases of cervical spondylotic myelopathy (CSM), three of cervical spondylotic radiculopathy (CSR), and three of cervical disc herniation, respectively.

A total of 144 screws were successfully inserted in these patients, including 29 patients with four pedicle screws, two with six screws, one with eight screws, and four with two pedicle screws combined with lateral mass screws.

Radiographic evaluation

All patients underwent preoperative anteroposterior, lateral and bilateral oblique radiography and computed tomography (CT) scanning to delineate the osseous anatomy and the courses of the vertebral arteries. This radiographic data enabled careful planning of the entry points and screw trajectories preoperatively.

Surgical technique

Approach and exposure

Under general anesthesia and in the prone position, a standard posterior midline approach was adopted to expose the bilateral laminae and lateral mass. To avoid bleeding, subperiosteal dissection was performed. Reduction was performed for patients with fracture and/or dislocation of vertebrae and/or articular processes. The level was identified objectively according to its anatomical landmarks in all patients.

According to the guidelines planned from the preoperative images, the screw trajectories for C3‐C7 were prepared. The entry point was in the superior and lateral corner of the lateral mass, in the sagittal plane. Trajectories were 10° cephalad for C6 and C7, 10° caudad for C3 and C4, and vertically for C5; in the horizontal plane, they were 40°–47° abducent for C3‐C6, and 30°–40° abducent for C7 ³ . The entry point and screw trajectory at each lateral mass are oriented to the position of the lateral masses, which may alter according to the degree of cervical lordosis or kyphosis. The point and trajectory also are influenced by the patient's individual pathoanatomy.

Initially, cortical bone at the entry point was removed with a high‐speed burr. Then a screw tap of AXIS or an awl produced by ourselves (diameter of its tip is 2 mm) was applied to tap the pedicle from the entry point. The depth was about 2–2.5 mm. Drilling direction was guided by hand feeling of ‘soft area’ in spongy bone. When obvious resistance was encountered, some adjustment was made to the direction in order to locate the ‘soft area’. We found that the abducent angle in the horizontal plane was often smaller than expected.

Usually, a little bleeding, sometimes mixed with some yellow marrow, would be observed after the pedicle hole had been successfully prepared. If hemorrhage was considerable, it might denote bleeding external to the pedicle or injury to the vein in the transverse foramen of the vertebra due to penetration of the lateral wall. Under such circumstances, bone wax was inserted into the hole at once in order to control bleeding.

A blunt probe was used to explore the hole to ensure continuity of the cortical bone. Loss of rigidity along the pedicle wall indicated the cortical wall was not intact. In this case, it was necessary to change the screw trajectory and tap the pedicle again, which usually required radiological guidance.

An alternative method for guiding the screw trajectory was to expose and palpate the medial walls of the pedicles. First, a keyhole in the lamina was created by laminotomy at the correct level and a nerve hook used to palpate the pedicle both superiorly and medially. Next, a 2 mm burr was used to create a starting hole in the previously palpated pedicle. Finally, a power drill with a stop set at 20 mm was used to perform the trajectory ⁴ . A total of thirteen screws were inserted with this technique.

After the screw trajectories had been completed, bone graft was performed on the posterolateral aspect of the spine, following which appropriate AXIS screws and plate were implanted.

The diameter of the screws was usually 3.5 mm or 4.0 mm and the length 24, 26, 28 or 30 mm.

Results

A total of 114 screws were implanted into the lower cervical spine in 36 patients. It was found that 16(11.1%) screws had penetrated the pedicle wall. Among them, 10 (6.9%) screws had pierced the lateral, 4 (2.8%) the superior and 2 (1.3%) the inferior wall. However, there were no neurovascular complications related to these malpositioned screws during a period of 2 years follow‐up. Furthermore, postoperative Odem's scores improved in all patients except one with a complete neurologic deficit.

All patients acquired sound internal fixation and good bone fusion. Eleven patients with incomplete spinal cord injury achieved significant improvement in their nerve function. Four patients with CSM attained significant improvement in sensory and motor functions; while another patient with a ten year history of CSM had a bad result.

Discussion

Pedicle screw fixation offers three‐column stability and has been proven to be the most rigid posterior fixation technique.

There have been many clinical and laboratory studies on cervical pedicle screw fixation ⁵ , ⁶ . Various trajectories have been assessed in terms of their likelihood of causing neurovascular injury.

Abumi et al. ² have suggested that the entry point should be lateral to the center of the facet and close to the posterior margin of the superior articular surface. The angle of insertion of the screw can vary from 25° to 45° medial to the midline in the horizontal plane. In the sagittal plane the angle of insertion should be parallel to the upper endplate of the vertebral body.

Ebraheim et al. ⁷ have studied cadavers and pointed out that the foramen transversarium lines up with the midpoint of the lateral mass. Accordingly, they have suggested that the entry point should be at the superior and lateral corner of the lateral mass, and trajectories be 10° cephalad for C6 and C7, 10° caudad for C3 and C4 and vertically for C5; moreover, in the sagittal plane trajectories should be 40°–47° laterally for C3‐C6, and 30°–40° laterally for C7.

Recently, a computer navigation system has been used for pedicle screw placement ⁸ . A neuronavigation system together with intraoperative fluoroscopy can minimize the risk of complications, especially where there are cervical deformities. Neuronavigation can provide simultaneous axial, coronal and sagittal views, which can not be provided by fluoroscopy. However, the navigation system is expensive and very complicated to manipulate; besides, it is very time consuming.

The freehand technique has been proved to be safe and simple. We have applied the technique successfully in 36 posterior cervical arthrodesis patients, with minimal complications (Fig. 1).

Figure 1.

C7 burst fracture in a 52‐year‐old male managed with pedicle screws. (a) Preoperative magnetic resonance imaging (MRI). (b) Postoperative radiographs. (b1) Anteroposterior view. (b2 and 3) Oblique views. (c) Postoperative CT scan showing the screw centered in the pedicle.

Anatomic considerations

To accurately and safely place a transpedicular screw in the cervical spine, currently available anatomic data must be analyzed ⁹ .

The angle formed by the pedicle and central sagittal plane of a vertebra in horizontal plane is called the abduction angle. The average abduction angle from C2 to C7 is 33.59°, 42.43°, 45.6°, 42.37°, 37.89° and 34.43°, respectively ¹⁰ . The pedicles of C3 and C4 incline to cranial about 10° in sagittal plane, and those of C6 and C7 incline to caudal about 10°. The pedicle of C5 is just parallel to the horizontal line in the sagittal plane ¹¹ . The width of the pedicle is between 5 and 6 mm, the height is about 7 to 8 mm, and the smallest diameter is 4 mm. The distance from the cortex of the lateral mass to the middle of the cortex of the vertebra is about 30 mm in the horizontal plane. The medial cortex of the pedicle is the thickest, followed by the inferior, superior and lateral cortices (Fig. 2c). So it is recommended that the pedicle be tapped along its medial cortex.

Figure 2.

C4‐5 fracture and dislocation in a 40‐year‐old male with some neurological symptoms. (a) Preoperative radiographic images. (a1) Lateral view. (a2) CT scan. (a3) MRI. (b) Postoperative radiographs. (b1) Lateral view. (b2) Anteroposterior view. (c) Postoperative CT scan. The arrow indicates where a screw has encroached into the vertebral artery foramen by 40%–60% of its diameter without vascular complications.

Problems and countermeasures

Previous studies have reported some complications with cervical pedicle screws ¹² , ¹³ , ¹⁴ , ¹⁵ . In our study, ten screws loosened after placement. This may be related to pedicle wall breakage, malposition of the screw, or too large a hole for the screw due to repeated tapping. Under such circumstances, the entrance point and trajectory should be carefully adjusted, and a screw of greater diameter inserted. The laminoforaminotomy technique can also be of some help in adjusting the entrance point and trajectory.

Bleeding occurred in 11 screw holes, at a rate of 150–450 ml per 10 min. This was usually addressed by immediate insertion of bone wax or a screw (Fig. 3).

Figure 3.

C5 burst fracture in a 32‐year‐old male. (a) Preoperative lateral radiograph. (b) Postoperative radiographs, a screw seems to be in the pedicle. (b1) Lateral view. (b2 and 3) Oblique views. (c) Postoperative CT scan. The arrow indicates where a screw has encroached into the vertebral artery foramen by 100% of its diameter with effusive bleeding, which was successfully stopped.

Error in screw trajectory: If a freehand transpedicular insertion fails, an alternative funnel technique is necessary, which usually requires exposure of the medial wall of the pedicle and implantation of the screw under direct vision ¹⁶ .

In the present study, no progression of nerve symptoms, or injury of vessels or viscera due to errors in screw placement, occurred in the 36 patients. One patient complained of deterioration in muscle strength and disturbance of sensation in the upper limbs. However, this might have been related to reactive edema after decompression of the cervical canal, or a very long history prior to surgery. Luckily enough, the symptoms were significantly alleviated by rehabilitation.

Pedicle screw fixation can provided the greatest stability to the cervical spine. Because we did not use fluoroscopy or image guidance, it was paramount to analyze each patient's cervical anatomy individually. Further studies with long‐term follow up are required.