The objective of the current study was to quantify and compare the multidirectional flexibility properties of occipital anchor fixation with conventional methods of occipitocervical screw fixation using nondestructive and destructive investigative methods.
Fourteen cadaveric occipitocervical specimens (Oc–T2) were randomized to reconstruction with occipital anchors or an occipital plate and screws. Using a 6-degree-of-freedom spine simulator with moments of ± 2.0 Nm, initial multidirectional flexibility analysis of the intact and reconstructed conditions was performed followed by fatigue loading of 25,000 cycles of flexion-extension (x-axis, ± 2.0 Nm), 15,000 cycles of lateral bending (z-axis, ± 2.0 Nm), and 10,000 cycles of axial rotation (y-axis, ± 2.0 Nm). Fluoroscopic images of the implantation sites were obtained before and after fatigue testing and placed on an x-y coordinate system to quantify positional stability of the anchors and screws used for reconstruction and effect, if any, of the fatigue component. Destructive testing included an anterior flexural load to construct failure. Quantification of implant, occipitocervical, and atlantoaxial junction range of motion is reported as absolute values, and peak flexural failure moment in Newton-meters (Nm).
Absolute value comparisons between the intact condition and 2 reconstruction groups demonstrated significant reductions in segmental flexion-extension, lateral bending, and axial rotation motion at the Oc–C1 and C1–2 junctions (p < 0.05). The average bone mineral density at the midline keel (1.422 g/cm3) was significantly higher compared with the lateral occipital region at 0.671 g/cm3 (p < 0.05). There were no significant differences between the occipital anchor and plate treatments in terms of angular rotation (degrees; p = 0.150) or x-axis displacement (mm; p = 0.572), but there was a statistically significant difference in y-axis displacement (p = 0.031) based on quantitative analysis of the pre- and postfatigue fluoroscopic images (p > 0.05). Under destructive anterior flexural loading, the occipital anchor group failed at 90 ± 31 Nm, and the occipital plate group failed at 79 ± 25 Nm (p > 0.05).
Both reconstructions reduced flexion-extension, lateral bending, and axial rotation at the occipitocervical and atlantoaxial junctions, as expected. Flexural load to failure did not differ significantly between the 2 treatment groups despite occipital anchors using a compression-fit mechanism to provide fixation in less dense bone. These data suggest that an occipital anchor technique serves as a biomechanically viable clinical alternative to occipital plate fixation.