Neurology India, Vol. 59, No. 3, May-June, 2011, pp. 362-368
Use of spinous processes to determine the laterally angulated trajectory for placement of lateral mass screws: An image analysis
Kuan-Yin Tseng, Chung-Ching Hsia, Yuan-Hao Chen, Shin-I Ma, Chi-Tun Tang
Department of Neurological Surgery, Tri-Service General Hospital, Taipei 114, Taiwan, China
Correspondence Address: Chi-Tun Tang No. 325, Sec. 2, Chenggong Rd., Neihu District, Taipei City, Taiwan China firstname.lastname@example.org
Date of Submission: 02-Aug-2010
Code Number: ni11110
AbstractBackground : Lateral mass screw placement techniques have been broadly described in the literature. Differences in these techniques are related to entry points, lateral angulations and the cephalocaudal axis.
Aim : We evaluated 20 patients who underwent lateral mass screw placement between 2007 and 2009. Computed tomography (CT) scans of the cervical vertebrae were analyzed for each patient.
Material and Methods : We measured the maximal transition from the midpoint of the lateral mass to a proposed intersection point by a line connecting the corresponding spinous process and outermost rim of the transverse foramen at each level. This determined an optimal entry point during the tip of screw tilted on the same level of spinous process.
Results : The results revealed that a screw entry point less than 3 mm medial to the midpoint of the lateral mass could safely avoid violation of the vertebral artery.
Conclusions : The current study uses imaging analysis to demonstrate that spinous processes are an intraoperative landmark to aid surgeons in determining safe lateral mass screw trajectories. The limited-scale case results support our prediction from the image analysis. Depending on intraoperative landmarks, lateral mass screws could be safely and comfortably placed with good clinical outcomes.
Keywords: Cervical spine, fusion, instrumentation, lateral mass screw, spinous process
Operative procedures for posterior cervical spine fixation include: wiring, posterior plate and screw fixation, and pedicle screw constructs.  Lateral mass screw-rod constructs have become widely accepted over the last two decades with fair clinical outcomes. The trajectory of the pedicle screw or the lateral mass screw is of critical importance during posterior cervical fixation as the nerve roots, vertebral arteries, facet joints, and the spinal cord are at risk of potential injury from errant positioning. Jeanneret et al.  have extensively described the techniques of posterior lateral mass screw placement which are being used today.  However, these techniques describe variable starting points and screw directions, and there is considerable intersurgeon and intrasurgeon variability when attempting duplicate screw placements. ,
Recently, a technique using the spinous process to direct the placement of lateral mass screws was reported.  An advantage of this procedure is that the surgeon can use the intraoperative anatomy to direct screw placement. However, it is difficult for inexperienced surgeon to determine the optimal trajectory so as to not violate the neurovascular structure.  In the present study, we used computed tomography (CT) scans and magnetic resonance imaging (MRI) to define a safe entry point for the screw purchase and apply the proposed method in the clinical situations. The limited-scale case results support our prediction from the image analysis. Depending on intraoperative landmarks, lateral mass screws could be safely and comfortably placed with good clinical outcomes.
Material and Methods
We reviewed the hospital records, office visits, and imaging studies of 12 patients who underwent lateral mass screw placement from C3 to C6 over a 24-month period. Six patients were instrumented from the occiput to C3-C5 or C2-C4. Two patients were instrumented from C2 to C5. Two spinal surgeons placed 132 screws in the patients. Indications for posterior cervical fixation included instability due to osseous metastasis, pseudarthrosis, trauma, spondylosis, and myelopathy. Patients with compressive lesions from primarily posterior components were selected to undergo instrumentation along with neural decompression. Posterior instrumentation was also used in patients with prior anterior procedures with resultant myelopathy from progressive kyphotic deformities, as shown in [Table - 1]. Two available instrumentation systems have been introduced to the current market: Summit (DePuy Spine, Raynham, Massachusetts (MA) and Axon (AO Spine, Synthes, North America) systems. Both systems provide well-constructed strength and simple assembly procedures.
Before operations, axial and sagittal slices of CT scans and MRI of each patient were chosen for morphological and mathematical analyses. First, we defined an imaginary line connecting the edges of the corresponding spinous processes and the outermost rim of the transverse foramen from C2 to C6 levels. This provided an intersection point that served as an entry point on the lateral mass edge if screw drilling was directed towards the axis of the line. We measured the distance from the intersection point to the midpoint of the same level of lateral mass. Moreover, the authors found that if the screw entry point located within the point-to-point interval and the screw tail was tilted more laterally, it is free of transverse foraminal penetration. In the zone medial to the imaginary line, there are critical neurovascular structures such as the spinal cord, root sleeve, and vertebral artery. Based on the clinical and anatomical significance of these structures, we divided the lateral mass into a medial dangerous zone and a lateral free-violation zone. The intersection point was defined as the most medial entry point while drilling or taping. The midpoint of lateral masses, which was dissected at the medial and lateral borders, was located lateral to the corresponding intersection point [Figure - 1]. Web-based photo digitizer software (MD-Ruler) was used to measure the distance between the intersection point and the midpoint of the lateral mass from C2 to C6 levels [Table - 2] by the same senior author (S.I. Ma) to diminish inter-observer data bias.
After endotracheal intubation, patients were placed in the prone position. We routinely used the Mayfield skull clamp for head positioning and stabilisation. Head of the patient was placed in the reverse Trendelenburg position because of documented haemodynamic benefits. The surgery began with the standard midline posterior approach to the cervical spine, and the dissection continued down to the underlying fascia with the aid of Bovie electrocautery. A Gelpi retractor was used to reflect the paraspinal muscles to allow full exposure of the borders of the lateral masses. Self-retaining helped in the lateral displacement of the paraspinal musculature. The synovium was denuded from the facet joints at the level of surgery. Intraoperative fluoroscopy was used for level identification. For screw placement, the lateral mass was sectioned equally via imaginary lines drawn in the cephalocaudal and medial to lateral directions. Lateral masses were divided into four quarters and the midpoint was determined. The starting point was medial (approximately 1-3 mm) to the midpoint of the lateral mass. An awl was used to create the starting hole. A 3.5-mm drill bit was set at a depth of 10-14 mm. The tip of the drill was then leaned against the border of the spinous process of the vertebra being instrumented and the lateral angulation was determined [Figure - 2]a-c. Cephalocaudal angulation of the screw placement was oriented parallel to the surface of the superior articular process under C-arm fluoroscopic guidance. Similarly, 3.5×1.4 mm lateral mass screws were placed. Because of the smaller size and ovoid shape of the lateral mass at the C2 level, we shifted the starting pole 1 mm medial and 2-3 mm inferior to the midpoint of the lateral mass. Lateral and cephalocaudal angulations were also determined by the same method. From image analysis, the safety of lateral angulation was dependent on the spinous process of C2 [Figure - 3]. All lateral mass screws were made at bicortical purchases. Lodortically bent rods were fashioned, screw caps were placed, minimal compression was applied, and screws were torqued in accordance with the manufacturers' specifications. Finally, we used intraoperative fluoroscopy to check the placement of the screws. If decompression was warranted, the screws were placed before beginning this aspect of the procedure. To aid in lateral mass arthrodesis, morselised autograft or allograft chips were placed in the lateral gutters and denuded facets. After obtaining adequate haemostasis, the wound was re-approximated in normal anatomic layers. The average operating time was a median of 100 min.
Follow-up was carried out at 24 h postoperatively as well as at 2, 6, and 12 months thereafter. We also compared the Nurick Myelopathy Grade before operations and 24 h after operations. Fusion and well-placed hardware were assessed with plain X-rays and CT scans. Solid arthrodesis was noted if one of the following was demonstrated:  absence of motion involving the instrumented segments on dynamic imaging;  trabecular bone formation between the instrumented levels as seen on CT scans; and  absence of screw construct halo or grossly migrated implants.
The cervical spines of patients were evaluated at the C2-C6 level (126 lateral masses) by CT scan. The imaginary line connecting the corresponding spinous process and outermost rim of the transverse foramen intersected at the same level of the lateral mass on the critical point. This imaginary line divided the lateral mass into the lateral zone: free-violation zone and medial zone: dangerous zone surrounding important neurovascular structures, such as the spinal cord, exiting nerve root, and vertebral artery. The results of the distance between the critical point and midpoint of the lateral mass at the C2-C6 level are shown in [Table - 2]. The mean values of the distance are expressed as means±standard deviations. Mean values of C3 and C4 were 3.98±0.58 mm and 4.78±0.8 mm, respectively. Mean values of C5 and C6 were 4.43±0.68 mm and 3.89±0.53 mm, respectively. Therefore, for C3, C4, C5, and C6 lateral masses, the proposed entry points at the most medial 3 mm to midpoint of the lateral mass were located in the free-violation zone. By means of the determined points and with the screw leaning against the corresponding spinous process, we checked the tract of the lateral mass screw to see whether violation of important neurovascular structures would occur. However, mean of the distance between the critical point and midpoint of the lateral mass for C2 was 3.42±0.49 mm. Therefore, the entry point of the C2 lateral mass screw was modified to a corresponding point 1 mm medial to the midpoint. Lateral and cephalocaudal angulations were determined by the same method. On clinical inspection, lateral angulation was also considered safe even though the tip of the screw tilt was on the C3 spinous process.
All patients underwent posterior instrumentation involving lateral mass screws. Included in this review were 20 patients (10 female, 10 male), with age range 26-70 years (mean, 51 years). Patients were followed up for an average of 14 months. Postoperative complications, malpositioned screws, and cerebrospinal fluid leaks were not found. The most common complication was superficial infection in three patients. This was determined intraoperatively if gross evidence of purulence in the subfascial layers or growth of bacteria in culture media occurred. There were no neurovascular injuries or mortality directly related to the procedure. Furthermore, the postoperative Nurick Myelopathy Grade  was not worse than the preoperative Grade for all patients [Table - 1]. One patient (Case 2) had the screw pull-out of C4 during the surgical procedure. It was postulated that the screw was too short (12 mm) to engage the superior and lateral portions of the ventral cortex of the superior articular process. Fortunately, we revised the initial 3.5×12 mm screw to a 3.5×16 mm screw in the same entry point and achieved bicortical purchase. The lateral and cephalocaudal angulations were based on the same method. Postoperatively, this patient had no screw lucency at a 7-month follow-up [Figure - 4]. For patients available for follow-up longer than 6 months (n=19), radiographic evidence of fusion was demonstrated on dynamic plain roentgenograms and/or a 64-slice helical CT scan.
The indications for instrumentation of the posterior cervical spine include instability secondary to pseudarthrosis, metastasis, trauma, and spondyloarthropathy.  Screw techniques other than the pedicle screw have evolved from various wiring techniques to transarticular, transfacet, translaminar, pars, and lateral mass screws. , Although the lateral mass screw construct is proved to be a safe procedure, hazardous injury of critical structures can occur, which can be catastrophic. Traditional posterior-plate screw techniques can be associated with potential problems, including blunt or penetrating injury to the vertebral artery, nerve root, facet joints, and spinal cord. , Currently, there is no standardised intraoperative method to warrant the safe trajectory angle of insertion of a lateral mass screw, as proposed by the techniques described by An, Anderson, Margel, or Roy-Camille. Moreover, there is considerable individual variation among surgeons in determining the intraoperative trajectory angle for screw placement. , In the past literature, Chin et al. used a cadaveric analysis to determine that while intrumenting C3, C4, C5, and C6 levels by using the tips of spinous process caudal to the level of screw placement, the lateral angle was accurately determined.  The lateral and cephalad trajectory angles at each spinous process relative to the starting hole were compared with 30° and 15° respectively. It decreased intersurgeon and intrasurgeon variability when attempting duplicate screw placements. However, some debates the cephalocaudal trajectories of C3 and C4 were overestimated.  In addition, because the flexion, extension, or distraction injury of the subaxial spine would change the lateral trajectories, the usage of spinous process caudal to the level of screw purchasing will not be appropriate in patients with distraction or extension injury of subaxial spine. According to these disadvantages, Stevens et al. presented a new technique for the placement of lateral mass screws from C3 to C7. The midpoint of lateral mass is used as the entry point. The lateral trajectory is parallel to the corresponding spinous process.  This technique for placement of lateral mass screws yielded adequate fixation without any appreciable neurovascular complications. However, this trajectory dictated by aligning the screw with spinous process would still be variation among surgeons in determining the intraoperative trajectories for screw placement. In the current study, we continued using the method of Chin et al. The technique is slightly different in that, we attempted to lean the tip of the drill against the border of the spinous process of the vertebra being instrumented. It facilitated the corresponding spinous process to act as an exact reproducible visual guide for lateral trajectory of lateral mass screws. We also used CT scans or MRI of the cervical spine to determine entry points not exceeding 3 mm medial to the midpoint of the lateral mass to facilitate screw placement dependent on the corresponding spinous process. The findings of our study differ from previous reports in that lateral angulation is determined more confidently by visualising the spinous process of the corresponding level being instrumented. The anatomic study of An et al. also showed that the nerve-root exit point was at the anterolateral portion of the superior facet.  Therefore, the less lateral or more cephalic the drill trajectory, the more likely the exiting nerve root will be impinged. The positions of Margel demonstrated that the nerve and artery were safe from violation at levels C3-C6, with lateral angulation of 25°C and even 30°C.  Consequently, in our series, the lateral trajectory was dictated by the screw tilt against the spinous process, and the screw placement was more towards the safe zone. The well-described triggered electromyographic response is also used to aid screw positioning away from neural breach.  We did not encounter nerve breach after triggering a response when checking the position. The results show that our method is safe and that it does not violate the nerve structure.
For the C2 lateral mass, the distance between the intersection point and the midpoint differed by less than 3 mm in some patients. This phenomenon could be explained because the C2 lateral mass tends to be smaller and more ovoid than other levels. An entry point 1 mm medial to the midpoint of the lateral mass was sufficiently far to prevent injuring critical neurovascular structures during placement. Because of the larger bifida of the C2 spinous process, the lateral angulation of C2 determined by the above-mentioned method can be modified in a similar way to the standard Anderson technique, which proved to be the safest way with good bone purchase.  The tip of the C3 spinous process was also used to start the trajectory axis. Some reviews report that the anti-torque strength between different screws is similar. Because C7 is a transitional level, the lateral mass is shallow and elongated in the sagittal view and screw placement is seldom satisfactory. In most individuals, the vertebral artery emerges at the C6-C7 level and curves into the sixth transverse foramen, or sometimes the fifth or seventh transverse foramina. Screw placement when applying standard techniques is consistently hazardous to the nerve root and vertebral artery.  For patients in whom C7 vertebra fixation is involved and lateral mass screws cannot be successfully placed, the alternative option of a transpedicular route would be considered.
In the same protocol using a preoperative image survey, reconstructed CT scans and MRI can be used to visualise the path of the bilateral extracranial vertebral artery, especially at the atlantoaxial complex where the normal horizontal vascular loop or aberrant vertebral artery ectasia may occur. These variant findings preclude the catastrophic complication during placement or maybe bypass to adjacent fixation. Moreover, the usage of spinous processes as a landmark for safe angulation of the lateral mass screws will not be appropriate in patients with spinal deformities or for patients who have previously undergone laminectomy with no spinal process in situ. Hence, these patients may be approached through Margel or Roy-Camille methods. The aim of this study might stress on bringing out an alternative to the standard skills.
Although the vertebral artery courses through each corresponding transverse foramen, two precautions must be taken to protect the vascular structure:  careful review of the vertebral artery by CT or MRI because the density or signal maybe too sluggish to be visualised; and  surgeons must look for abnormal trauma-related pseudoaneurysms or true obstruction of the vertebral artery. MRI is better for mural assessment of the vessel than CT. Postoperative images prove that our proposed method is an alternative to achieving safer cervical arthrodesis without violation of critical neurovascular structures. This image-guided practice offers spinal surgeons a fast, reliable, and reproducible option.
Copyright 2011 - Neurology India
The following images related to this document are available:
Photo images[ni11110f1.jpg] [ni11110t1.jpg] [ni11110f3.jpg] [ni11110f4.jpg] [ni11110f3 (2).jpg] [ni11110t2.jpg] [ni11110f2.jpg]