Brian J. Doherty

Biomechanics of odontoid fracture fixation. Comparison of the one- and two-screw technique.

Direct anterior screw fixation of odontoid fractures has become more prevalent clinically. No biomechanical study, however, has determined whether one or two screws should be used. This study measured the stability of the odontoid process after fracture and internal fixation with one or two screws. Internal fixation of Type II odontoid fractures did not restore the original stability of the intact specimen, direct internal fixation with one or two screws provided 50% of the stability of the unfractured odontoid, and no significant differences between the one- and two-screw technique was found under loading to failure, although the two-screw technique provided increased stiffness in extension loading.

Kinematic evaluation of lumbar fusion techniques.

Study Design Eight human cadaveric lumbosacral spines were biomechanically and kinematically tested in torsion and compression‐flexion. They were retested after simulated posterolateral fusion, anterior lumbar interbody fusion, and circumferential fusion. Objectives To analyze stiffness and motion in the anterior and posterior columns of the index and contiguous spinal motion units of anterior, posterolateral, and circumferential fusions. Summary of Background Data Previous biomechanical studies have not incorporated analysis of motion with six degrees of freedom, consideration of contiguous levels, and comparisons of anterior and posterior column motion. Methods Eight human cadaveric lumbosacral spines were biomechanically tested in compression‐flexion and torsion using an advanced biplanar radiography technique. Each specimen underwent either a simulated posterolateral fusion or anterior fusion followed by a circumferential fusion. Motion and stiffness at the level of the fusion and at contiguous levels were analyzed independently in the anterior and posterior columns of the spine. Results At the level of fusion, the simulated posterolateral and anterior fusions prevented more motion in torsion compared with compression‐flexion. With all specimens, it was shown that circumferential fusions were stiffer than the intact specimen. Our comparison of motion in the anterior and posterior columns found no significant differences within the columns of a single vertebral motion segment. Compared with posterolateral fusions, anterior fusions were found to have the greatest effect in increasing motion at contiguous levels. The effect of circumferential fusions on adjacent level kinematics was not significantly greater than that of anterior fusions. Conclusion There are major biomechanical differences between different fusion techniques. This information should be considered in patients undergoing lumbar spinal fusion.

The viscoelastic responses of the human cervical spine in torsion: Experimental limitations of quasi-linear theory, and a method for reducing these effects

The dynamic torsional viscoelastic responses of the human cadaver cervical spine were measured in vitro. The quasi-linear formulation of time dependent behavior was used to describe and predict the resultant torque as a function of applied angular deflection and time. The performance of the quasi-linear model was good, reaching correlation at the 99% confidence level; however, it tended to underestimate hysteresis energy (mean relative deviation = -19.1%) and observed stiffness. This was in part due to difficulties in establishing the physical constants of the quasi-linear model from finite rate relaxation testing. An extrapolation deconvolution technique to enhance the experimentally derived constants was developed, to reduce the detrimental effects of finite rate testing. The quasi-linear model based on this enhanced derivation showed improved predictive ability and hysteresis energy determination.

A biomechanical study of odontoid fractures and fracture fixation.

The purpose of this study was to measure the stability of the odontoid process after fracture and subsequent screw fixation. To accomplish this, we mechanically reproduced Type II and Type III odontoid fractures on isolated C2 vertebrae by varying the direction of load. These fractures were subsequently stabilized with a single 3.5 mm screw and retested for multidirectional stability and load to failure. Reduced and instrumented specimens were found to have a stiffness equivalent to one half of that of the unfractured odontoid. Load to failure was also slightly less than one half of the original fracture force (average 160 ib). Screw failure was by a cut-out mechanism in all Type III fractures and by bending of the screw in all Type III fractures. Our findings, in conjunction with the existing literature, strongly suggest that Type III fractures result from extension forces, whereas Type II fractures result from lateral or oblique loading forces. Single screw fixation of an odontoid fracture will provide stability equal to approximately one half that of the unfractured bone.

The quantitative anatomy of the Atlas

Study Design This study is based on direct quantitative caliper measurements of 88 isolated anatomic specimens of the C1 vertebra Objectives The study was undertaken to establish the range and variability of the external dimensions of the atlas and to describe the cortical thicknesses and trabecular distribution of this unlque vertebra. Summary of Background Data Before this study, Francis in 1952 reported the total anterior and posterior diameter of 285 atlas vertebrae. Liu et al reported detailed external dimensions and facet joint surface morphologies on a total of three C1 vertebrae. Methods Measurements were made of overall dimensions, canal diameters, and the dimensions of the anterior and posterior arches of 88 dried human C1 vertebrae. Eight specimens were sectioned in the frontal plane, eight in the sagittal plane, and four in the coronal plane. The anatomy of these sections was documented by radiographic imaging. Cortical thicknesses on the sections were then obtained by direct measurement. Results The canal diameter ranged from 32 mm (SD 2 mm) in the sagittal plane, and 29 mm (SD 2 mm) in the lateral dimension. The mean thickness of the anterior ring was 6 mm (SD 1 mm) and posteriorly was 8 mm (SD 2 mm). Cortical bone was thinnest posteriorly. Conclusions These measurements indicated remarkably constant dimensions for the ring itself, suggesting there may be significant functional restraints on the canal size of this unique vertebra. In contrast, a significant variability was noted in objective measurements of lateral mas height and sagittal plane widths of the entire bone.

The trabecular anatomy of the axis

This article describes the internal anatomy of C2. Although some C2 specimens showed a high density of trabecular bone throughout, a feature noted in all specimens was a void or very hypodense area of bone located immediately beneath the dens. The observed nonuniform distribution of trabecular bone within the axis is considered to have an effect on internal fixation of this bone. The good cancellous bone quality consistently observed in the lateral masses and beneath the facets, as well as near the bony end plates, suggests that these areas may be reasonable sites for the insertion of internal fixation devices. The hypodense area observed in the upper portion of C2 would suggest that fixation devices inserted through this area may obtain relatively poor purchase and may be prone to cut-through failure.

A biomechanical study of Jefferson fractures

Study Design. Fifteen specimens of the first cervical vertebra were tested by the application of pure tensile forces to failure. Seven specimens had intact transverse ligaments, and eight had transection of the transverse ligament before testing. Specimens were tested to failure by the rapid application of laterally directed tensile force to the ring; this force then was exerted through the lateral masses to simulate the mechanism of injury for this fracture as proposed by Jefferson. Objectives. To measure the biomechanical characteristics of the C1 ring, including the fracture patterns created with tensile loading, and to describe the influence of the transverse ligament on the behavior of the ring as it failed under tension. Summary of Background Data. Jefferson fractures have been reproduced in the laboratory by subjecting head and neck preparations to axial load. However, no previous detailed biomechanical studies of the fracture characteristics of the isolated C1 vertebra have been reported. Methods. Specimens were tested to failure by rapid application of laterally directed tensile forces to the ring. Results. Eleven two‐part and three three‐part fractures occurred. The mean tensile strength of the atlas was found to be 2,280 N. The average deformation required to fracture the C1 ring was 1.57 mm. The total energy absorbed by the ring averaged 1.99 N‐m. There was no statistically difference between those specimens with the transverse ligament intact and those without a transverse ligament. Conclusions. The results of this study show that fractures of the C1 ring of greater than two parts can occur with pure tensile loading. The ring will fracture with as little as 1 mm of deformation.

The role of torsion in cervical spine trauma

A dynamic servocontrolled torsion machine has been used to characterize cervical injury due to pure rotation of the head. Resultant force moment, torque, and applied rotation have been measured. Torque applied to the base of the skull resulted in injury to the atlantoaxial joint. No evidence of lower cervical injury was observed by computed tomography, magnetic resonance imaging, in situ fluoroscopy, or visual inspection. Torque applied directly to the lower cervical spine induced ligamentous injury and unilateral facet dislocation; however, the torque to injure the lower cervical spine was significantly greater than the torque to injure the atlantoaxial joint. It was concluded that pure rotation of the head does not mediate lower cervical ligamentous injury because of the comparative weakness of the atlantoaxial joint.

Mathematical modeling of the hybrid III manikin head-neck structure

This paper presents results of an ongoing research program to develop data sets for mathematical models which accurately predict the head-neck kinematics and dynamics of anthropomorphic dummies in crash environments. The mathematical model utilized for this research was the Head-Spine Model (HSM), which was developed at the Armstrong Aeromedical Research Laboratory. In this study, the goal was to develop HSM data sets for the Hybrid III manikin head-neck structure. Modifications of the program source code were made in order to model the asymmetric bending properties of the neck. Data sets were developed, utilizing a 1-element and 4-element neck. The effect of the occipital condyle nodding block joint on head-neck kinematics and dynamics was also investigated and modeled. Tests were conducted to measure the stiffness of this joint in flexion and extension. A data set with a 1-element neck and nodding joint was developed. The Amended Part 572 Head-Neck Pendulum Test, of the Code of Federal Regulations, was simulated to validate these data sets. Small differences were observed between the responses of the 1-element and 4-element necks. The effect of nodding joint on head-neck kinematics and dynamics was negligible. Overall, the model proved to be a reasonable predictor of manikin head-neck kinematics and dynamics.