Abstract
Introduction
Surgical fixation of the axis (C2) includes screw positioning through bony corridors close to vital structures. However, human bones exhibit large variations in size and shape across individuals and populations. A 3D statistical model of C2 using Quantitative Computed Tomography (QCT) was created to analyze the anatomical variations and to understand their implications for screw positioning.
Material and Methods
106 standard clinical QCTs from 46 adult female and 60 male European patients, aged 52.0 years ± 19.9 years, with intact C2s have been included in this study. The mean image resolution was 0.5mm × 0.5mm in the axial plane and 0.6mm in craniocaudally direction. After anonymization 3D statistical modeling of C2 was performed. This included the computation of averaged 3D surface and volumetric bone mineral density (vBMD) models and principal component analysis (PCA). Transpedicular and odontoid screw templates were virtually implanted and their corridors were analyzed.
Results
PCA revealed a highly variable anatomy of C2 in which size was the predominant variation in the 1st principal component (PC), whereas shape changes were primarily described by the remaining PCs. The largest shape variation was observed at the spinous and transverse processes and the transverse foramina. Comparison of the averaged 3D surface models of C2 for men and women separately revealed mainly a difference in size: The model for males was ~7% larger in the axial plane (anterior-posterior and left-right directions) and ~9% in craniocaudal direction. 3.5mm odontoid screw templates could be virtually implanted in all C2. The average corridor length was 39.2mm ± 2.8mm and the median screw length 38mm. The average corridor length difference between the genders was ~8%. The corridor exhibited the lowest average vBMD value (199.0 mgCaHA/ml) underneath the basis of the odontoid process and a maximum value (911.2 mgCaHA/ml) at its upper part near the cortical shell. Virtual implantation of a 3.5mm C2 pedicle screw with 1mm safety zone was not possible in 31.1% due to interference with the vertebral artery. In 26.4% a 3.5mm screw and in 42.5% a 4.5mm screw could be positioned. The median screw length was 32mm. The average pedicle corridor length was 28.7mm ± 1.9mm with an average convergence of 21.0° ± 3.6° to the sagittal plane and an average ascent of 18.7° ± 3.5° in cranial direction. The lowest average vBMD value for the corridor was located in the trabecular bone of the vertebral body (314.8 mgCaHA/ml) whereas the maximum value (689.2 mgCaHA/ml) was measured near the anterior cortical shell of the vertebral body. The average entry point was ~5mm lateral from the medial border of the inferior articular process and ~10mm cranial from its caudal border.
Conclusion
We established a 3D statistical model of C2 using QCTs. It revealed largely variable surfaces, bone quality and corridor dimensions and allowed the efficient 3D assessment of parameters relevant for screw positioning. There are anatomical conditions that allowed/not allowed for screw positioning. Surgical decision making must rely on both, a thorough anatomical understanding and the given individual situation.
Acknowledgments
This work was supported by AOSpine TK of the TK System of the AO Foundation, Davos, Switzerland. D. Gehweiler received a research fellowship grant from the AO Research Institute Davos, Davos, Switzerland.
