Evaluation of a novel C2 pedicle screw insertion technique: a retrospective comparative clinical study and finite element analysis Authors: Shuai Zhang, Guofeng Zhang, Jingyu Zhang, Yuancheng Zhang, Hua Wang, Jianru Wang, Guowei Han, Hui Liu, Zemin Li Published: 2026 ABSTRACT C2 pedicle screws (C2PS) are extensively used in the treatment of upper cervical spine disorders owing to their superior mechanical properties. However, incorrect screw insertion in the C2 pedicle can lead to catastrophic consequences due to the proximity to critical structures such as the vertebral artery groove and spinal cord. There is currently no universally accepted method for precise C2PS placement. In this study, we evaluated a novel free-hand C2PS insertion technique that uses the superior margin of the C2 lamina and the medial margin of the C2 pedicle as landmarks, thereby avoiding routine exposure of the C2/C3 facet joint capsule. Compared with the traditional Harms technique, the novel method reduced the transverse screw angle and sagittal screw angle, which may facilitate screw insertion in a limited operative field. The overall rate of pedicle breach was not significantly increased. Finite element analysis further showed comparable fixation strength between the two techniques in the modeled normal C2 anatomy. In summary, for anatomically suitable C2 pedicles, the novel C2PS insertion technique provided comparable radiographic accuracy and similar biomechanical stability to the Harms technique, while requiring smaller insertion angles and less lateral exposure. Introduction Certain conditions affecting the upper cervical spine can compromise stability, significantly impacting quality of life and posing life-threatening risks if the spinal cord or vertebral artery is affected. Since Gallie’s seminal publication in 1939, which described the use of C2 sublaminar cables to maintain stability in the atlantoaxial joint, numerous techniques for internal fixation of this joint have been developed1. These include lamina clamps, atlantoaxial transarticular screw technique, screw-plate systems, screw-rod constructs, translaminar screw fixation, and various hybrid approaches2–6. Among these, the technique combining C1 lateral mass screws with C2 pedicle screws (C2PS) has emerged as a gold standard due to its superior mechanical stability, offering robust 3-column fixation7–9. Since Harms et al. introduced polyaxial screws, C2PS internal fixation has become increasingly prevalent in upper cervical spine surgery5. The C2 vertebral body, located at the craniocervical junction, presents unique challenges for screw placement due to its complex morphology and proximity to critical anatomical structures. Accurate screw insertion is especially critical in patients with high-riding vertebral arteries (HRVA) or narrow C2 pedicles (C2P)10–13. To address these challenges, Kepler introduced the “Parsicle” screw fixation technique, which involves the lateral cephalad and medial translation of the C2PS trajectory to minimize the transverse screw angle (TSA) and sagittal screw angle (SSA)14. However, Kepler’s approach did not specify an exact insertion point, provide detailed anatomical parameters for screw placement, or verify the fixation strength of the proposed technique. The medial window technique, proposed by Lee et al., offers an alternative for inserting C2PS in cases involving partially narrow C2P or HRVA. While this technique may facilitate screw placement, it potentially sacrifices spinal epidural space, raising concerns about long-term outcomes that necessitate close monitoring15. Additionally, the C2 cortical bone screw technique introduced by Kim et al. has shown promise, but whether its fixation strength matches that of the C2PS remains to be determined16. The Harms method remains the most widely recognized and commonly used technique for C2PS insertion. However, due to the deep anatomical position of the C2 vertebra and the robust development of its paraspinal muscles, this method presents significant technical challenges. The Harms technique requires TSA and SSA of approximately 25°–30° each, often necessitating extensive retraction of adjacent soft tissues and, in some cases, an extended surgical incision5. Additionally, the technique involves exposing critical anatomical landmarks, including the outer margin of the C2 lateral mass, the C2/C3 joint, and the medial margin of C2P. This extensive exposure can lead to considerable complications, such as unmanageable bleeding from the venous plexus5. Furthermore, excessive lateral exposure increases the risk of vertebral artery injury, while iatrogenic damage to the C2/C3 joint capsule may result in postoperative cervical instability. To address these limitations and facilitate C2PS insertion within a more constrained surgical field, our team developed a novel technique. Using the upper edge of the C2 lamina and the medial margin of the C2P as reference points, and shifting the traditional Harms entry point cephalad and medially, we were able to reduce both the TSA and SSA. These adjustments streamline the insertion process, minimizing tissue disruption and enhancing surgical safety. In 1996, Nabil Ebraheim proposed a similar method for C2PS insertion17. However, his approach, which relied heavily on cadaveric simulations, maintained the same TSA and SSA as the traditional Harms technique. This failure to modify the screw trajectory led to a higher risk of transverse foramen violation, limiting its adoption in clinical practice. This study aims to evaluate the viability, accuracy, and consistency of this novel C2PS insertion method. Additionally, the study seeks to compare its outcomes to those achieved using the traditional Harms screw insertion technique, with a focus on procedural safety, precision, and fixation reliability. Materials and methods Materials The Institutional Review Board of our institution approved this study ([2023]509). Considering the retrospective nature of this study, the requirement for written informed consent was waived. This study involves methods that comply with the principles outlined in the WMA Declaration of Helsinki and the Belmont Report of the Department of Health and Human Services. This study analyzed preoperative and postoperative cervical computed tomography (CT) (Light Speed VCT, GE Healthcare, Chicago, USA ) images obtained from patients who underwent C2PS internal fixation for upper cervical spine disease at our hospital between January 2018 and June 2024. CT scans taken at 0.625-mm intervals were included, irrespective of the indication or diagnosis for the scan. The inclusion criteria for the analysis were the implantation of C2PS for the treatment of cervical spine disease, and preoperative and postoperative CT imaging of the upper cervical spine region. The exclusion criteria were age < 18 years and the presence of malignancies, infectious diseases, or metabolic bone diseases potentially impacting the C2P, as well as patients for whom pertinent parameters were unattainable. Patients with incomplete data, no postoperative CT, and CT imaging that was unsatisfactory due to factors such as artifacts were also excluded. Surgical technique After administering general anesthesia, the patient was placed in a prone position on a Jackson surgical table. The patient’s head was stabilized by using a Mayfield frame, and the neck was adjusted appropriately into a flexed position (15–30°). Following palpation of the C2 spinous process, a layered incision along the midline of the neck was made through the skin and subcutaneous tissues. This novel insertion technique only necessitates the exposure of the upper half of the C2 laminar and the medial margin of the C2P. The C2PS-insertion point is located 5-mm laterally and 5-mm caudally from the inflection point where the C2P meets the upper edge of the laminar, designated as point N (Fig. 1A). In accordance with the preoperative planning parameters, the C2PSs are then inserted (Fig. 1B). Correspondingly, the Harms technique necessitates a thorough visualization of the posterior arch of the atlas, the C1/C2 lateral mass joint, the outer margin of the C2 lateral mass, and the C2 inferior articular process. The typical entry point (designated as point H) for C2PS insertion is on the line connecting the vertical midpoint of the superior and inferior articular surfaces of the C2 articular process (Fig. 1A). The C2Ps are delineated with a number 4 Penfield. Preoperative CT measurement parameters are used to determine the appropriate screw length, and TSA and SSA are confirmed intraoperatively with C-arm fluoroscopy after screw insertion. In selected cases with high-riding vertebral artery anatomy or an extremely narrow pedicle, we carefully evaluated the anatomy of each C2 pedicle and the vertebral artery on CT and 3-dimensional reconstruction images before surgery. If the anatomical difficulty of either high-riding vertebral artery anatomy or an extremely narrow pedicle, suggested that free-hand insertion would be difficult or unsafe, screw trajectory was individualized on the basis of preoperative simulation and, when available, assisted by robotic or navigation guidance. If safe C2 pedicle screw placement still could not be achieved, or if the anticipated risk of vertebral artery or spinal cord injury remained unacceptably high, C2 pedicle screw placement was abandoned and alternative fixation, including extension to C3 or translaminar fixation, was used instead. Based on the underlying pathology, further reduction or decompression was then performed as required, followed by rod placement, preparation of the fusion bed, drainage tube placement, and layered wound closure. Fig. 1Schematic diagram of the screw entry point in the two C2PS insertion methods. (A) Schematic diagram of the screw entry points of the two methods. (B) Intraoperative scenario for insertion C2PS using the novel method. Black arrow represents the medial edge of the C2P. Blue arrow represents the upper edge of the C2 laminar. Schematic diagram of the screw entry point in the two C2PS insertion methods. (A) Schematic diagram of the screw entry points of the two methods. (B) Intraoperative scenario for insertion C2PS using the novel method. Black arrow represents the medial edge of the C2P. Blue arrow represents the upper edge of the C2 laminar. Measurement of parameters associated with two distinct C2PS insertion techniques The preoperative and postoperative CT imaging data were imported to an Advantage Workstation (Version ADW 4.4, GE Healthcare, Chicago, USA) for reconstruction and measurement by using a multi-planar reconstruction and synchronous positioning technology. When the Harms method is employed for screw insertion, the typical entry point for C2PSs is at the midpoint H, which is on the line connecting the vertical midpoint of the superior and inferior articular surfaces of the C2 articular process (Fig. 2A). In the novel screw-insertion technique, the C2PS insertion point (point N) is located 5-mm laterally and 5-mm caudally from the junction where the C2 pedicle intersects with the superior aspect of the C2 laminar (Fig. 2A). To reduce projection-related measurement bias, multiplanar reconstruction (RadiAnt DICOM Viewer, version 2024.1) was used for each patient to align the imaging plane with the pedicle axis. Specifically, the pedicle plane was first localized on the sagittal view and then reconstructed along the pedicle trajectory before axial and sagittal measurements were obtained. After identifying the screw entry point, an axial image passing through the entry point (H or N) and aligned with the pedicle plane was obtained. Line 1 was defined as the line connecting the lateral and medial cortices at the narrowest part of the C2 pedicle on the axial image. Line 2 was defined as the line connecting the screw entry point (H or N) to the midpoint of Line 1 and was regarded as the screw axis for C2PS insertion (Fig. 2B and D). Line 3 represented the midline of the C2 vertebral body, defined as the line connecting the midpoint of the anterior vertebral body margin to the midpoint of the spinous process. The angle between Lines 2 and 3 was defined as the TSA. The theoretical length of the C2PS trajectory was measured from the screw entry point along the screw axis to the anterior cortical margin of the C2 vertebral body. Similarly, after identifying the screw entry point, a sagittal image passing through the entry point and parallel to the medial wall of the C2 pedicle was obtained. Line 4 was established as the line connecting the cranial and caudal cortices at the narrowest point of the C2 pedicle on the sagittal image. Line 5 was defined as the line connecting the screw entry point (H or N) to the midpoint of Line 4. Line 6 was the line parallel to the C2 superior articular surface. The angle between Lines 5 and 6 represented the SSA for C2PS insertion (Fig. 2C and E). The actual screw length implanted during surgery was recorded. The angle measured between the implanted C2PS and Line 3 represented the actual TSA, whereas the angle between the implanted C2PS and Line 6 represented the actual SSA (Fig. 3). Fig. 2Method for measuring parameters related to C2PS insertion. (A) Identify the screw entry points for the insertion of two C2PS on CT 3D reconstruction images of C2. (B) Theoretical screw trajectory length and determination of the TSA for C2PS insertion in the Harms method. (C) Theoretical determination of SSA in the Harms method. (D) Theoretical screw trajectory length and determination TSA for C2PS insertion in the novel method. (E) Theoretical determination of SSA for C2PS insertion in the novel method. N represents the entry point of the C2PS in the novel method, and H represents the entry point of the C2PS in the Harms method. The black dot labeled L signifies the inflection point where the medial edge of the C2P merges with the upper edge of the C2 laminar. Line 1 represents the narrowest part of the pedicle on the axial image. Line 2 represents the axis for C2PS implantation, and the theoretical length of the screw trajectory for C2PS insertion. Line 3 represents the mid-sagittal line of the C2 vertebral body on the axial image. TSA represents transverse screw angle. Line 4 represents the narrowest part of the C2 pedicle in the sagittal plane. Line 5 represents the axis for C2PS insertion. Line 6 is a parallel line to the superior articular surface of the C2 joint process. SSA represents sagittal screw angle. Method for measuring parameters related to C2PS insertion. (A) Identify the screw entry points for the insertion of two C2PS on CT 3D reconstruction images of C2. (B) Theoretical screw trajectory length and determination of the TSA for C2PS insertion in the Harms method. (C) Theoretical determination of SSA in the Harms method. (D) Theoretical screw trajectory length and determination TSA for C2PS insertion in the novel method. (E) Theoretical determination of SSA for C2PS insertion in the novel method. N represents the entry point of the C2PS in the novel method, and H represents the entry point of the C2PS in the Harms method. The black dot labeled L signifies the inflection point where the medial edge of the C2P merges with the upper edge of the C2 laminar. Line 1 represents the narrowest part of the pedicle on the axial image. Line 2 represents the axis for C2PS implantation, and the theoretical length of the screw trajectory for C2PS insertion. Line 3 represents the mid-sagittal line of the C2 vertebral body on the axial image. TSA represents transverse screw angle. Line 4 represents the narrowest part of the C2 pedicle in the sagittal plane. Line 5 represents the axis for C2PS insertion. Line 6 is a parallel line to the superior articular surface of the C2 joint process. SSA represents sagittal screw angle. Fig. 3Schematic diagram of the actual postoperative measurement methods for corresponding parameters in the different C2PS insertion methods. (A-C) The TSA and SSA measurement methods for C2PS insertion as described in the Harms method. (D-F) The TSA and SSA measurement methods for C2PS insertion as described in the novel method. Schematic diagram of the actual postoperative measurement methods for corresponding parameters in the different C2PS insertion methods. (A-C) The TSA and SSA measurement methods for C2PS insertion as described in the Harms method. (D-F) The TSA and SSA measurement methods for C2PS insertion as described in the novel method. To assess measurement reliability, all imaging parameters were independently measured by three observers, and the mean values were used for analysis. One orthopaedic surgeon repeated the measurements after a two-week interval to assess intra-observer consistency. For inter-observer agreement, two orthopaedic surgeons independently evaluated 30 randomly selected pedicle screw placements. Intra- and inter-observer agreements were assessed using Cohen’s kappa coefficient. Evaluation of screw insertion accuracy The accuracy of screw insertion was evaluated on postoperative CT images by using a previously published cortical-breach grading system, using the location and the percentage of the screw diameter over the cortical edge18. In this system 0 = none of the screw diameter; Grade I ≤ 25% of the screw diameter; Grade II 26–50% of the screw diameter; Grade III 51–75% of the diameter; Grade IV 76–100% of the diameter. Each screw insertion was graded independently by 2 of the authors. The grade with the more significant degree of breach in the screw insertion was then used in the accuracy analysis. Finite element analysis (FEA) Establishment of a FE model for the normal C1 to C3 cervical spine Completely normal cervical CT raw data of a 28-year-old man from our hospital’s physical examination center were imported into Mimics 19.0 software (Materialise, Sweden V19.0 version, Belgium) (Figs. 4A–D). Based on the grayscale value regions corresponding to different tissues in CT images, a triangular mesh model of C1–C3 was constructed. Geomagic 2013 software (Geomagic, USA) was then used to construct a Non-Uniform Rational B-Splines (NURB) surface model (Figs. 4E–G). The model was then imported into Hypermesh 2019 software (Altair, USA) for meshing, material assignment, contact definition, and other related operations to generate a standard C1–C3 3-dimensional (3D) FE model (Figs. 4H–J). Fig. 4Establishment of a finite element model for a normal cervical spine C1 to C3. (A-D) The interface for cervical spine 3D reconstruction with Mimics software. (E-G) Geomagic software is used to fit the pattern diagram with the NURB surface model. (H-J) Hypermesh graph. Establishment of a finite element model for a normal cervical spine C1 to C3. (A-D) The interface for cervical spine 3D reconstruction with Mimics software. (E-G) Geomagic software is used to fit the pattern diagram with the NURB surface model. (H-J) Hypermesh graph. Based on the key parameters measured on postoperative CT of the 2 screw insertion methods (average length of C2PS implanted, TSA, and SSA), 2 fixation models were established: one for the Harms method and 1 for the novel method (Fig. 5). The refined mesh components of the C1–C3 segment were then imported into Abaqus 2019 (Dassault, France) FE software for analysis. Fig. 5Schematic diagram of the key parameters used for establishing the finite element analysis model for the two insertion methods. (A, B) Harms method. (C, D) Novel method. Schematic diagram of the key parameters used for establishing the finite element analysis model for the two insertion methods. (A, B) Harms method. (C, D) Novel method. Model parameter The models developed in this study primarily comprised the intervertebral joints, articular cartilage surfaces, and intervertebral discs, including the endplate, annulus fibrosus, and nucleus pulposus, based on the cervical FEA model introduced by Yu et al.19. The average cortical bone thickness was set at 1.5 mm, while the endplate thickness was defined as 0.2 mm. Horizontally, three layers of annular fibrous structures encircled the nucleus pulposus, constituting approximately 43% of the total intervertebral disc area. The gap between articular cartilage surfaces was modeled at 0.5 mm, with a sliding friction coefficient of 0.1 for their contact interfaces. Joint node contact was applied for interactions between the endplate and vertebral body, whereas tied contact was utilized for interactions between the endplate and both the annulus fibrosus and nucleus pulposus. Ligaments were represented using two-node T3D2 elements. The mechanical properties of all relevant biomechanical materials were derived from existing literature, considering the origin and geometric characteristics of each ligament. All models were designed as heterogeneous, continuous, and isotropic. Analysis and computation The cervical spine model was affixed to the lower endplate of the C3 vertebral body, allowing for six degrees of freedom. To simulate the weight of the head, a 75 N load was applied, with a point on the C1 rotation axis designated as the reference point20–22. This force was distributed uniformly across all nodes of C1 through a distribution coupling mechanism between the nodes on the left and right joint surfaces of C1 and the reference point. Two fixed models were imported into the Abaqus software, and a downward load of 75 N was applied to simulate head weight. While external torques used in other studies vary to account for different dimensions of cervical movement, this study applied a torque of 1.5 N.m to ensure sufficient simulation of physiological cervical activities across all dimensions while minimizing the risk of specimen damage19,20,23. Mechanical analysis was conducted under the following six conditions: forward flexion, backward extension, left and right lateral flexion, and left and right rotation. The relative motion of C1-C2 under these six conditions and the stress-strain distribution within the internal fixation system in various directions of motion was compared between the two screw insertion methods. Statistical analysis Continuous variables were expressed as mean ± standard deviation, while categorical variables were presented as count (percentage). The length of the C2PS, as well as the TSA and SSA, were compared between the two methods using the independent-sample t-test. Screw insertion grade and the presence or absence of breaches were analyzed using a two-way Fisher’s exact test to assess significance, and odds ratios with 95% confidence intervals were calculated. Fisher’s exact test or Pearson’s chi-square test was applied to compare categorical variables. Inter-rater reliability was evaluated using Cohen’s kappa coefficient. All statistical analyses were performed using SPSS version 22.0 (Chicago, IL, USA). Results Patient characteristics In total, 91 patients who underwent C2PS fixation were included in the analysis. Of these, 48 patients (23 females, 25 males; mean age: 51.4 years) had screws inserted using the Harms method. The preoperative diagnoses are detailed in Table 1. A total of 87 C2PS were implanted in this group, with five replaced by C2 translaminar screws due to the presence of HRVA or a narrowed C2P, and four extending the fixation to the C3 vertebra. In the second group, comprising 43 patients (22 females, 21 males; mean age: 53.6 years), fixation was performed using the novel method. A total of 80 C2PS were inserted in this group, with four replaced by C2 translaminar screws for the same reasons—HRVA or a narrowed C2P—and two extending fixation to C3 (Table 1). Among the 91 patients, six experienced postoperative complications. Three patients reported pain in the C2 nerve root innervation area, while three developed superficial incision infections. All six patients recovered fully after standard treatment protocols and were subsequently discharged. Notably, there were no serious complications such as vertebral artery or spinal cord injuries. Table 1Patient characteristics.VariableHarm’s Method(n = 48),No.(%)*Novel Method(n = 43),No.(%)* P Age, mean ± SD (y)51.4 ± 15.453.6 ± 12.40.53Male25(52.1)21(48.8)0.76Female23(47.9)22(51.2)Pathology0.93Basilar invagination2(4.2)1(2.3)Trauma20(41.6)15(34.9)Rheumatoid arthritis2(4.2)1(2.3)Degeneration17(35.3)19(44.2)Congenital2(4.2)3(7.0)Infection1(2.1)2(4.7)Tumor2(4.2)1(2.3)OPLL2(4.2)1(2.3)C2PS placed, No.87(90.6)80(93.0)0.84C2 translaminar screws, No.5(5.2)4(4.7)C3 pedicle screws, No.4(4.2)2(2.3)Complication, No1.0Occipital neuralgia2(4.2)1(2.3)Surgical site of infection1(2.1)2(4.6)OPLL, ossification of the posterior longitudinal ligament; SD, standard deviation.* Values are number (percentage) unless indicated otherwise. Table 1Patient characteristics.VariableHarm’s Method(n = 48),No.(%)*Novel Method(n = 43),No.(%)* P Age, mean ± SD (y)51.4 ± 15.453.6 ± 12.40.53Male25(52.1)21(48.8)0.76Female23(47.9)22(51.2)Pathology0.93Basilar invagination2(4.2)1(2.3)Trauma20(41.6)15(34.9)Rheumatoid arthritis2(4.2)1(2.3)Degeneration17(35.3)19(44.2)Congenital2(4.2)3(7.0)Infection1(2.1)2(4.7)Tumor2(4.2)1(2.3)OPLL2(4.2)1(2.3)C2PS placed, No.87(90.6)80(93.0)0.84C2 translaminar screws, No.5(5.2)4(4.7)C3 pedicle screws, No.4(4.2)2(2.3)Complication, No1.0Occipital neuralgia2(4.2)1(2.3)Surgical site of infection1(2.1)2(4.6)OPLL, ossification of the posterior longitudinal ligament; SD, standard deviation.* Values are number (percentage) unless indicated otherwise. Patient characteristics. OPLL, ossification of the posterior longitudinal ligament; SD, standard deviation. * Values are number (percentage) unless indicated otherwise. Comparative analysis of the two C2PS insertion techniques Preoperative CT measurements showed that the novel C2PS insertion method resulted in shorter theoretical screw path lengths and reduced TSA and SSA than the Harms method. However, no significant differences were observed in pedicle width between the two techniques (Table 2). Postoperative CT measurements confirmed the accuracy of preoperative screw trajectory planning. These results demonstrated that the novel C2PS insertion technique consistently yielded shorter actual screw lengths, TSA, and SSA than the Harms method, mirroring preoperative planning trends (Table 2). The theoretical screw lengths, pedicle widths, TSA, and SSA values derived from preoperative CT scans exhibited high intra- and inter-observer reliability. The kappa values were as follows: screw length, 0.82 and 0.87; pedicle width, 0.80 and 0.84; TSA, 0.82 and 0.81; and SSA, 0.86 and 0.88, respectively. Similarly, the postoperative measurements of TSA and SSA demonstrated strong intra- and inter-observer consistency, with kappa values of 0.80 and 0.85 for TSA, and 0.88 and 0.84 for SSA, respectively. Table 2Comparison of parameters for the two P2SP methods.Length of the screw trajectoryWidth of the screw trajectoryTSA (°)SSA (°)PreoperativeHarms (n = 182)25.9 ± 2.05.7 ± 1.628.8 ± 2.524.3 ± 3.4New (n = 182)22.0 ± 2.15.6 ± 1.618.6 ± 2.318.3 ± 3.4P0.0000.6440.0000.000PostoperativeHarms (n = 87)23.3 ± 2.03.4 ± 0.325.7 ± 2.021.9 ± 1.7New (n = 80)20.9 ± 2.33.3 ± 0.317.5 ± 1.517.9 ± 1.6P0.0000.3070.0000.000TSA, transverse screw angle; SSA, sagital screw angle. Table 2Comparison of parameters for the two P2SP methods.Length of the screw trajectoryWidth of the screw trajectoryTSA (°)SSA (°)PreoperativeHarms (n = 182)25.9 ± 2.05.7 ± 1.628.8 ± 2.524.3 ± 3.4New (n = 182)22.0 ± 2.15.6 ± 1.618.6 ± 2.318.3 ± 3.4P0.0000.6440.0000.000PostoperativeHarms (n = 87)23.3 ± 2.03.4 ± 0.325.7 ± 2.021.9 ± 1.7New (n = 80)20.9 ± 2.33.3 ± 0.317.5 ± 1.517.9 ± 1.6P0.0000.3070.0000.000TSA, transverse screw angle; SSA, sagital screw angle. Comparison of parameters for the two P2SP methods. TSA, transverse screw angle; SSA, sagital screw angle. Evaluation of C2PS insertion accuracy Postoperative CT scans revealed 15 instances of screw position abnormalities after C2PS insertion using the Harms method, representing a frequency of 17%. These comprised 4 cases of spinal canal invasion, 10 of transverse foramen invasion, and 1 of pedicle upper wall penetration. According to the grading system introduced by Prachya Punyarat, eight were classified as grade I, three as grade II, three as grade III, and one as grade IV (Table 3). With the novel C2PS insertion method, there were 14 instances of screw position abnormalities, representing a frequency of 18%. These included seven cases of spinal canal encroachment, two transverse foramen invasion, and five of pedicle upper wall penetration. Based on the grading system, five cases were classified as grade I, six as grade II, one as grade III, and two as grade IV (Table 3). No significant differences were observed in the frequency or degree of screw position abnormalities between the two insertion methods. The inter-rater reliability of all 167 screws was 93%, with a kappa value of 0.89 between the two physicians. Table 3Accuracy of the two C2PS techniques.Cortical-breach grade(location) (%)Harms’ method (n = 87)New method (n = 80) P Breach0.97No72 (82.8)66 (82.5)Yes15 (17.2)14 (17.5)Grade0.45Ⅰ8 (M2,L5,S1) (9.2)5 (M2,L1,S2) (6.3)Ⅱ3 (M1,L2) (3.4)6 (M2,L1,S3) (7.5)Ⅲ3 (M1,L2) (3.4)1 (M1) (1.3)Ⅳ1 (L1) (1.1)2 (M2) (2.5)Dichotomized grade1.00Acceptable (0-Ⅱ)83 (95.4)77 (96.3)Unacceptable (Ⅲ-Ⅳ)4 (4.6)3 (3.7)M, medial; L, lateral; S, superior. Table 3Accuracy of the two C2PS techniques.Cortical-breach grade(location) (%)Harms’ method (n = 87)New method (n = 80) P Breach0.97No72 (82.8)66 (82.5)Yes15 (17.2)14 (17.5)Grade0.45Ⅰ8 (M2,L5,S1) (9.2)5 (M2,L1,S2) (6.3)Ⅱ3 (M1,L2) (3.4)6 (M2,L1,S3) (7.5)Ⅲ3 (M1,L2) (3.4)1 (M1) (1.3)Ⅳ1 (L1) (1.1)2 (M2) (2.5)Dichotomized grade1.00Acceptable (0-Ⅱ)83 (95.4)77 (96.3)Unacceptable (Ⅲ-Ⅳ)4 (4.6)3 (3.7)M, medial; L, lateral; S, superior. Accuracy of the two C2PS techniques. M, medial; L, lateral; S, superior. Comparison of fixation strength between the Two C2PS insertion techniques FEA revealed that the two fixation methods provided comparable stability in maintaining the relative mobility between C1 and C2 across all tested directions of movement in the modeled normal C2 anatomy (Table 4). The stress-strain diagrams indicated no significant differences in the magnitude of stress experienced by the C2 vertebral body and C2PS or in the regions of highest stress concentration under flexion, extension, left and right lateral flexion, and left and right rotation (Fig. 6). Table 4C1-C2 range of motion of each method under same loading conditions.FlexionExtensionLeft lateral flexionRight lateral flexionLeft rotationRight rotationHarms Method2.13°1.26°0.45°0.26°0.25°0.24°Novel Method2.39°1.34°0.32°0.22°0.24°0.21° Table 4C1-C2 range of motion of each method under same loading conditions.FlexionExtensionLeft lateral flexionRight lateral flexionLeft rotationRight rotationHarms Method2.13°1.26°0.45°0.26°0.25°0.24°Novel Method2.39°1.34°0.32°0.22°0.24°0.21° C1-C2 range of motion of each method under same loading conditions. Fig. 6Stress-strain diagram of the vertebral body and internal fixation system during cervical movement in the two different insertion methods. (A) Diagram showing stress distribution in the C2 vertebral body and C1-C2PS internal fixation during C2PS insertion using the Harms method with cervical flexion-extension movements. (B) Diagram showing stress distribution in the C2 vertebral body and C1-C2PS internal fixation during C2PS implantation using the Novel method with cervical flexion-extension movements. (C) Diagram showing stress distribution in the C2 vertebral body and C1-C2PS internal fixation during C2PS insertion using the Harms method with cervical lateral bending movements. (D) Diagram showing stress distribution in the C2 vertebral body and C1-C2PS internal fixation during C2PS insertion using the Novel method with cervical lateral bending movements. (E) Diagram showing stress distribution in the C2 vertebral body and C1-C2PS internal fixation during C2PS insertion using the Harms method with cervical rotation movements. (F) Diagram showing stress distribution in the C2 vertebral body and C1-C2PS internal fixation during C2PS insertion using the Novel method with cervical rotation movements. Max represents the point of maximum stress concentration. Stress-strain diagram of the vertebral body and internal fixation system during cervical movement in the two different insertion methods. (A) Diagram showing stress distribution in the C2 vertebral body and C1-C2PS internal fixation during C2PS insertion using the Harms method with cervical flexion-extension movements. (B) Diagram showing stress distribution in the C2 vertebral body and C1-C2PS internal fixation during C2PS implantation using the Novel method with cervical flexion-extension movements. (C) Diagram showing stress distribution in the C2 vertebral body and C1-C2PS internal fixation during C2PS insertion using the Harms method with cervical lateral bending movements. (D) Diagram showing stress distribution in the C2 vertebral body and C1-C2PS internal fixation during C2PS insertion using the Novel method with cervical lateral bending movements. (E) Diagram showing stress distribution in the C2 vertebral body and C1-C2PS internal fixation during C2PS insertion using the Harms method with cervical rotation movements. (F) Diagram showing stress distribution in the C2 vertebral body and C1-C2PS internal fixation during C2PS insertion using the Novel method with cervical rotation movements. Max represents the point of maximum stress concentration. Discussion C2PS techniques are commonly used to treat upper cervical spine conditions by ensuring spinal stability and restoring alignment through 3-column fixation7–9,24–26. However, the C2 vertebral body is situated at the transition between the head and spine, and the significant anatomical variations, along with the complex surrounding structures, make screw placement challenging, often requiring a steep learning curve18,27,28. Hence, some researchers have proposed using lower-risk pars screws instead. However, biomechanical studies have demonstrated that C2 pars screws cannot effectively maintain the 3-column stability of the spine8. Similarly, replacing C2PS with C2 translaminar screws does not provide sufficient fixation strength, as their effectiveness depends on the integrity of the anatomical structures such as the lamina, rendering them unsuitable for many patients29–32. To enhance the reliability and safety of internal fixation techniques, some researchers have suggested creating an intraoperative screw insertion model based on preoperative imaging data to improve the precision of C2PS insertion, facilitating a more targeted approach33. However, due to variations in intraoperative positioning and interference from surrounding soft tissues, it is difficult to maintain perfect alignment with a preoperative model. Additionally, while intraoperative CT or navigation technology has been proposed to guide screw placement, comparative studies have shown that the accuracy does not significantly differ from free-hand insertion28,34,35. Since Goel first introduced the C2PS technique in 1994, and Harms et al. introduced polyaxial screws, the most widely accepted method for C2PS insertion is to locate the entry point midway between the superior and inferior articular processes4,5. The screw is then placed parallel to the upper articular surface of C2, under direct visualization of the medial margin of the C2P. While the Harms method provides detailed anatomical localization, it necessitates extensive exposure of the outer edge of the C2 lateral mass, C1/C2 and C2/C3 articular processes, and the medial margin of the C2P. This extensive exposure often leads to unpredictable bleeding due to the proximity of critical structures, and may also cause damage to the C2 nerve root or vertebral artery, increasing the risk of postoperative degeneration of the C2/C3 lateral mass joint. In contrast, when inserting the C2PS using the N point proposed by our team as the entry point, only the upper edge of the C2 lamina and the medial margin of the C2P need to be exposed. Relative to the insertion point (H) in the traditional Harms technique, the N point is positioned more cephalad and medially. Preoperative and postoperative CT scans demonstrated that using the N point significantly reduces the TSA and SSA, making the insertion process more straightforward for the surgeon. Additionally, this method minimizes the need for excessive exposure or retraction of surrounding soft tissues, thereby reducing surgical trauma. The primary challenges of C2PS insertion include the presence of HRVA and a narrow pedicle. Previous studies have reported a frequency of HRVA at the C2 level of 14–18%, and a frequency of the narrowest pedicle diameter being less than 4 mm in 10–32% of cases14,36,37. These anatomical variations contribute to an accuracy rate of only 70–80% for C2PS implantation, whether performed manually or with navigation assistance38–41. Incorrect placement of C2PS can encroach on the transverse foramen, potentially damaging the vertebral artery, or penetrating the medial margin of C2P, risking spinal cord injury, both of which may lead to catastrophic outcomes. The frequency of abnormal screw placement using the novel method and the Harms method was comparable (approximately 17%). Additionally, no significant differences were found in the location or degree of C2PS breach between the two techniques. Although medial and superior wall breaches appeared numerically more frequent with the novel technique, this pattern should be interpreted cautiously because the difference was not statistically significant in the present cohort. The more cephalad and medial entry point used in the novel technique may account for this trend, whereas the reduced rate of transverse foramen encroachment may reflect the medially shifted trajectory. Fortunately, due to the ample space within the C2 spinal canal, screws that inadvertently entered the spinal canal did not result in severe complications, such as spinal cord or nerve injury. Conversely, the reduced rate of transverse foramen encroachment may reflect a different breach pattern associated with the modified trajectory. However, this finding should not be extrapolated to patients with HRVA or other anatomically challenging pedicles, because such an inference is not directly supported by the present data. Accordingly, our data suggests a different distribution of breach patterns rather than a proven reduction in clinically important risk. Larger studies are needed to determine whether the medial and transverse breach profile of the novel trajectory differ meaningfully from that of the Harms technique and to clarify its clinical relevance. The diameter and length of C2PS are important determinants of fixation strength. With the novel technique, the TSA and SSA for C2PS implantation were smaller than those used in the Harms method, resulting in a shorter actual screw length. Therefore, it was necessary to determine whether the shorter trajectory would compromise fixation strength. Finite element analysis demonstrated that both methods maintained comparable stability in terms of relative C1-C2 motion and showed no significant differences in the stress distribution within the fixation construct during different cervical motions. However, these biomechanical conclusions should be restricted to the modeled condition because the finite element model was based on a single normal C2 anatomy and did not incorporate high-riding vertebral artery anatomy or markedly narrowed pedicles. Accordingly, the FEA findings support the feasibility of the novel trajectory under the tested anatomical conditions, but they should not be interpreted as evidence of superiority or universal applicability. Limitations Nevertheless, several limitations to this study should be considered. First, the study was retrospective and conducted at a single center; therefore, future multicenter, prospective studies are needed to confirm the effectiveness of the novel C2PS insertion technique. Second, due to the lack of standardized long-term follow-up, it is not possible to determine whether the novel method influences long-term outcomes, such as spinal fusion rates. Third, in selected cases with high-riding vertebral artery anatomy or markedly narrowed pedicles, the novel C2PS technique may not be feasible, and alternative fixation should be performed. Thus, the novel technique should be cautiously generalized to cases with anatomical difficulties. Fourth, the finite element model was constructed from a single normal C2 anatomical configuration and did not include high-riding vertebral artery anatomy or markedly narrowed pedicles. Therefore, the biomechanical results should be interpreted only within the modeled condition. Finally, the relatively small sample size may have limited our ability to detect differences in specific breach patterns, particularly medial breach, between the two techniques. Larger sample size studies should be conducted to clarify whether there is clinical significance. Conclusion In conclusion, this study highlights that using the inflection point where the upper edge of the C2 lamina meets the medial margin of the C2P as a reference for determining the entry point in C2PS insertion avoids the need for extensive exposure of structures such as the outer edge of the C2 lateral mass joint and the C2/C3 joint. This approach results in smaller TSA and SSA for C2PS insertion, making the procedure more straightforward and less invasive. In anatomically suitable C2 pedicles, the novel insertion method achieved radiographic accuracy comparable to that of the Harms technique and demonstrated similar fixation strength in the modeled normal anatomy, without a significant increase in the overall rate of abnormal screw positioning. These findings support the feasibility of this modified free-hand technique, while further validation is required in patients with challenging pedicle anatomy.