Objective:To investigate spinal cord morphological changes and risk factors in upper cervical spine surgery using C 2 medial "in-out-in" pedicle screws in patients with atlantoaxial dislocation (AAD) and high-riding vertebral artery (HRVA). Methods:A retrospective analysis was conducted on 41 patients with AAD who underwent C 2 medial "in-out-in" pedicle screw implantation at our hospital between January 2019 and December 2023. The cohort included 12 males and 29 females, with a mean age of 47.6±10.3 years (range: 18-68 years). Among them, 30 patients received unilateral C 2 medial "in-out-in" screws, while 11 patients received bilateral screws. All patients underwent posterior reduction and internal fixation. Measurements included C 2 pedicle height, C 2 pedicle width, C 2 horizontal spinal canal width, screw invasion distance into the spinal canal, and spinal canal invasion rate based on CT findings. MRI evaluations included spinal canal-dura mater distance, dura mater-spinal cord distance, spinal canal-spinal cord distance, and spinal cord cross-sectional area. The change rate of spinal cord cross-sectional area was calculated and categorized into >20% decrease group (decreasing group) and ≤20% decrease group (non-decreasing group). Postoperative outcomes were assessed using the Visual Analogue Scale (VAS) and Japanese Orthopedic Association (JOA) scores. Results:The mean operative time was 165.5±30.1 minutes (range: 120-220 minutes). No spinal cord vascular injuries or severe complications were observed. However, five patients experienced cerebrospinal fluid (CSF) leakage, requiring delayed removal of drainage tubes, but their wounds healed successfully. Follow-ups (range: 6-24 months) showed no loosening of internal fixation, fractures, or bone graft nonunion, with a 100% bone fusion rate at 6 months postoperatively. VAS scores improved significantly from a preoperative median of 5.0 (3.5, 6.0) (range: 1-7) to 2.0 (2.0, 3.0) at 3 months and 1.0 (1.0, 1.0) at 6 months ( Z=36.716, P< 0.001). JOA scores improved from 10.0 (9.0, 12.0) (range: 4-14) preoperatively to 13.0 (11.0, 13.5) at 3 months and 14.0 (12.0, 15.0) at 6 months ( Z=67.093, P<0.001). The height of C 2 pedicle was 5.50±1.78 mm, the width of C 2 pedicle was 2.27±1.23 mm, the width of C 2 horizontal spinal canal was 23.76±4.91 mm, the spinal canal-dura mater distance was 3.08±0.85 mm, dura mater-spinal cord distance was 3.23±0.85 mm, the spinal canal-spinal cord distance was 6.31±1.11 mm, the distance of screw invasion into the spinal canal was 2.80±1.54 mm, the rate of spinal canal invasion was 12.1%±6.8%, the preoperative spinal cord cross-sectional area was 69.81±13.27 mm 2, and the postoperative spinal cord cross-sectional area was 68.81±13.94 mm 2. Based on spinal cord cross-sectional area changes, 32 patients were classified into the non-decreasing group, and 9 patients into the decreasing group. The intraclass correlation coefficient (ICC) values ranged from 0.733 to 0.984, indicating high measurement reliability. There were significant differences in the proportion of C 2 bilateral medial "in-out-in" screws (χ 2=6.903, P=0.009), the incidence of CSF leakage (χ 2=15.391, P<0.001), the distance of screw invasion into the spinal canal ( t=4.990, P<0.001) and the rate of spinal canal invasion ( t=4.045, P<0.001) in the decreasing group versus the non-decreasing group. The JOA scores of the non-decreasing group were significantly higher at 3 and 6 months postoperatively compared to the decreasing group ( P<0.05). No other parameters showed significant differences between the groups. Binary logistic regression analysis identified spinal canal invasion rate [ OR=1.963, 95% CI (1.010, 3.817), P=0.047] as an independent risk factor for spinal cord cross-sectional area reduction. The Jordan index was 0.875, with a spinal canal invasion rate threshold of 14.18%, a sensitivity of 1.000, a specificity of 0.875, and an AUC of 0.983, indicating strong predictive value. Conclusion:The C 2 medial "in-out-in" screw technique provides effective posterior fixation and fusion for AAD patients with HRVA. However, to minimize spinal cord morphological changes and associated risks, the spinal canal invasion rate should be kept below 14.18% when using this technique.

Objective:To investigate spinal cord morphological changes and risk factors in upper cervical spine surgery using C 2 medial "in-out-in" pedicle screws in patients with atlantoaxial dislocation (AAD) and high-riding vertebral artery (HRVA). Methods:A retrospective analysis was conducted on 41 patients with AAD who underwent C 2 medial "in-out-in" pedicle screw implantation at our hospital between January 2019 and December 2023. The cohort included 12 males and 29 females, with a mean age of 47.6±10.3 years (range: 18-68 years). Among them, 30 patients received unilateral C 2 medial "in-out-in" screws, while 11 patients received bilateral screws. All patients underwent posterior reduction and internal fixation. Measurements included C 2 pedicle height, C 2 pedicle width, C 2 horizontal spinal canal width, screw invasion distance into the spinal canal, and spinal canal invasion rate based on CT findings. MRI evaluations included spinal canal-dura mater distance, dura mater-spinal cord distance, spinal canal-spinal cord distance, and spinal cord cross-sectional area. The change rate of spinal cord cross-sectional area was calculated and categorized into >20% decrease group (decreasing group) and ≤20% decrease group (non-decreasing group). Postoperative outcomes were assessed using the Visual Analogue Scale (VAS) and Japanese Orthopedic Association (JOA) scores. Results:The mean operative time was 165.5±30.1 minutes (range: 120-220 minutes). No spinal cord vascular injuries or severe complications were observed. However, five patients experienced cerebrospinal fluid (CSF) leakage, requiring delayed removal of drainage tubes, but their wounds healed successfully. Follow-ups (range: 6-24 months) showed no loosening of internal fixation, fractures, or bone graft nonunion, with a 100% bone fusion rate at 6 months postoperatively. VAS scores improved significantly from a preoperative median of 5.0 (3.5, 6.0) (range: 1-7) to 2.0 (2.0, 3.0) at 3 months and 1.0 (1.0, 1.0) at 6 months ( Z=36.716, P< 0.001). JOA scores improved from 10.0 (9.0, 12.0) (range: 4-14) preoperatively to 13.0 (11.0, 13.5) at 3 months and 14.0 (12.0, 15.0) at 6 months ( Z=67.093, P<0.001). The height of C 2 pedicle was 5.50±1.78 mm, the width of C 2 pedicle was 2.27±1.23 mm, the width of C 2 horizontal spinal canal was 23.76±4.91 mm, the spinal canal-dura mater distance was 3.08±0.85 mm, dura mater-spinal cord distance was 3.23±0.85 mm, the spinal canal-spinal cord distance was 6.31±1.11 mm, the distance of screw invasion into the spinal canal was 2.80±1.54 mm, the rate of spinal canal invasion was 12.1%±6.8%, the preoperative spinal cord cross-sectional area was 69.81±13.27 mm 2, and the postoperative spinal cord cross-sectional area was 68.81±13.94 mm 2. Based on spinal cord cross-sectional area changes, 32 patients were classified into the non-decreasing group, and 9 patients into the decreasing group. The intraclass correlation coefficient (ICC) values ranged from 0.733 to 0.984, indicating high measurement reliability. There were significant differences in the proportion of C 2 bilateral medial "in-out-in" screws (χ 2=6.903, P=0.009), the incidence of CSF leakage (χ 2=15.391, P<0.001), the distance of screw invasion into the spinal canal ( t=4.990, P<0.001) and the rate of spinal canal invasion ( t=4.045, P<0.001) in the decreasing group versus the non-decreasing group. The JOA scores of the non-decreasing group were significantly higher at 3 and 6 months postoperatively compared to the decreasing group ( P<0.05). No other parameters showed significant differences between the groups. Binary logistic regression analysis identified spinal canal invasion rate [ OR=1.963, 95% CI (1.010, 3.817), P=0.047] as an independent risk factor for spinal cord cross-sectional area reduction. The Jordan index was 0.875, with a spinal canal invasion rate threshold of 14.18%, a sensitivity of 1.000, a specificity of 0.875, and an AUC of 0.983, indicating strong predictive value. Conclusion:The C 2 medial "in-out-in" screw technique provides effective posterior fixation and fusion for AAD patients with HRVA. However, to minimize spinal cord morphological changes and associated risks, the spinal canal invasion rate should be kept below 14.18% when using this technique.