- Email: [email protected]
Axis ring fractures due to simulated head impacts
Clinical Biomechanics 29 (2014) 906–911
Contents lists available at ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Axis ring fractures due to simulated head impacts Paul C. Ivancic ⁎ Biomechanics Research Laboratory, Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, New Haven, CT, USA
a r t i c l e
i n f o
Article history: Received 21 March 2014 Accepted 30 June 2014 Keywords: Axis ring fracture Trauma Cervical spine Biomechanics
a b s t r a c t Background: We investigated mechanisms of axis ring fractures due to simulated head impacts. Methods: Our model consisted of a human upper cervical spine specimen (occiput through C3) mounted to a surrogate torso mass on a sled and carrying a surrogate head. We divided 13 specimens into 3 groups based upon head impact location: upper forehead in the midline, upper lateral side of the forehead, and upper lateral side of the head. Post-impact fluoroscopy and anatomical dissection documented the injuries. Average occurrence times of the peak loads and accelerations were statistically compared (P b 0.05) using ANOVA and Bonferroni pair-wise post-hoc tests. Findings: Of the 13 upper cervical spines tested, 5 specimens sustained axis ring fractures with the most common mechanism being impact to the upper left lateral side of the forehead. The first local force peaks at the impact barrier and neck and all peak head accelerations occurred between 18.0 and 22.8 ms, significantly earlier than the absolute force peaks. The average peak neck loads reached 1761.2 N and the axis ring fractures occurred within 50 ms. Interpretation: We observed asymmetrical fractures of the axis ring including fractures of the superior and inferior facets, laminae, posterior wall of the vertebral body, pars interarticularis, and pedicles. The fracture patterns were related to the morphology of the axis as a transitional vertebra of the upper cervical spine. Understanding the mechanisms of axis ring fractures may help in choosing the optimal reduction technique and stabilization method based upon the specific fracture pattern. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Traumatic fracture of the ring of the axis is the second most common injury of the C2 vertebra; the most common being odontoid fracture (Dvorak et al., 2012). Axis ring fractures occur most often due to blunt head impact in younger individuals during high speed motor vehicle crashes and in older individuals during fall from standing height (Cornish, 1968; Effendi et al., 1981; Levine and Edwards, 1985; Samaha et al., 2000). The classic fracture pattern is bilateral fracture of the pars interarticularis, between the superior and inferior facets of the C2 vertebra, with forward displacement of the anterior fracture fragment. This forward displacement causes widening of the spinal canal and intervertebral foramina at C2 consistent with the low incidence of neurological sequelae associated with the injury observed clinically (Dvorak et al., 2012). White and Panjabi observed that the pars interarticularis of the axis is the most likely site of fracture due to forced hyperextension as it sustains the peak bending moment and is structurally weakened due to its small cross-sectional area and the presence of the foramen transversarium (White and Panjabi, 1990). ⁎ Biomechanics Research Laboratory, Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, 333 Cedar St., P.O. Box 208071, New Haven, CT 06520-8071, USA. E-mail address: [email protected].
http://dx.doi.org/10.1016/j.clinbiomech.2014.06.017 0268-0033/© 2014 Elsevier Ltd. All rights reserved.
Traditionally referred to as Hangman's fracture, the radiographic presentation of axis ring fractures due to modern civilian trauma is often similar to that previously documented due to judicial hanging, though the injury mechanisms differ (Effendi et al., 1981). The mechanism of injury due to judicial hanging with a submental knot has been described as forced hyperextension causing traumatic spondylolisthesis of the axis with death due to distraction of the cord or suffocation (Schneider et al., 1965). The mechanisms of axis ring fractures due to motor vehicle crashes have been described as forced hyperextension combined with axial compression often caused by blunt impact of the head with the vehicle interior and continued momentum of the torso (Effendi et al., 1981). Others hypothesize that flexion–compression or flexion–distraction mechanisms may produce certain axis ring fractures either as primary injury mechanisms or subsequent to forced extension–compression (Levine and Edwards, 1985). Clinical case series have documented fractures of the axis ring primarily due to motor vehicle crashes. In a study of 131 patients, Effendi et al. (1981) observed that the axis ring fractures may occur at any part of the bony canal including the laminae, superior or inferior facets, pars interarticularis, pedicles, or posterior wall of the vertebral body. They found that the fractures were almost always bilateral but rarely symmetrical and did not necessarily occur at the weakest part of the ring. The asymmetrical presentation of the fractures was hypothesized to be due to rotated head posture at the time of head impact.
P.C. Ivancic / Clinical Biomechanics 29 (2014) 906–911
Other case series have observed symmetrical fracture passing transversely through the superior facets immediately posterior to the dens, bordering the foramen transversarium, and through the inferior vertebral notch (Cornish, 1968; Williams, 1975). The prior studies demonstrate that axis ring fracture patterns are highly variable and dependent upon numerous trauma-related and morphological factors. In the prior case series, the clinical researchers have hypothesized that the fracture patterns specific to the axis ring are dependent upon the morphology of the axis as a transitional vertebra of the upper cervical spine (Effendi et al., 1981; Francis et al., 1981; Schneider et al., 1965). These studies describe the traumatic axial load due to head impact transferred inferiorly through the occipital condyles, through the superior and inferior facets of the atlas, and through the superior facets of the axis, and then being split among the body and left and right inferior facets of the axis. Additional loads include hyperextension and posterior shear transferred to the anterior articular surface of the dens through the anterior atlantal arch. Lastly, these loads are combined with those transferred superiorly through the cervical spine due to torso inertia. Most prior biomechanical studies of traumatic spondylolisthesis of the axis have created the injuries manually for evaluation of fixation techniques (Chittiboina et al., 2009; Duggal et al., 2007). Other studies have created realistic injuries in some specimens due to simulated trauma. In their pioneering study of the mechanism of odontoid fracture, Althoff (1979) created bilateral fractures of the pars interarticularis and complete ruptures of the C2/3 disc and anterior and posterior longitudinal ligaments in a single specimen due to impact to the upper lateral side of the head. Others have reported traumatic spondylolisthesis that occurred inconsistently in previous experimental series due to distraction (Dibb et al., 2009) or hyperextension (Nightingale et al., 2002) of upper cervical spine specimens or distractive-extension (Nightingale et al., 1996a, 1997a,b) during vertical drop tests of inverted head–neck specimens. Continued biomechanical research is needed to understand the etiology of axis ring fracture patterns. The goal of this biomechanical study was to investigate axis ring fractures due to simulated head impacts of a model consisting of an upper cervical spine specimen mounted between a surrogate head and surrogate torso mass. We evaluated the previously proposed injury mechanism hypotheses by analyzing the fracture patterns and biomechanical responses of the present specimens.
907
specimen in normal lordotic posture prior to the impact, similar to the in vivo neutral posture (Descarreaux et al., 2003). Cable loads were not measured. No counterbalancing of the head weight was required.
2.2. Simulated head impact To investigate various fracture patterns to the ring of the axis, we divided the specimens into 3 groups based upon head impact location: upper forehead in the midline (n = 5), upper left lateral side of the forehead (n = 4), and upper left lateral side of the head (n = 4). Each specimen was subjected to a single traumatic impact. A sled apparatus, aligned horizontally relative to the ground, and barrier system were used to simulate head-first impacts. The apparatus consisted of a surrogate torso mass rigidly fixed to a custom sled with the C3 mount of the specimen rigidly fixed at the front of the mass (Fig. 1). The mass,
2. Methods 2.1. Human upper cervical spine model with surrogate head We prepared 13 human osteoligamentous upper cervical spine specimens that were fresh-frozen (occiput to C3 vertebra; 6 males, 7 females; average age: 83.1 years). Apart from typical age related degenerative changes, the specimens did not suffer from any disease that could have affected the osteoligamentous structures. The specimens were rigidly mounted in resin at the occiput and C3 vertebra in normal neutral posture, consistent with the in vivo sagittal upright neutral posture (Descarreaux et al., 2003), and confirmed using fluoroscopy. A Hybrid III surrogate head (4.2 kg mass; 0.0244 kg m2 sagittal plane moment of inertia; Humanetics Innovative Solutions, Plymouth, MI, USA) was rigidly fixed to the occipital mount in anatomical position. The head and spine specimen were stabilized using muscle force replication which simulated the net effects of anterior, lateral, and posterior muscle forces. These simulated muscle forces produced compressive loading to the spine which provided static postural spine stability and passive resistance to motions during the impact. The muscle force replication was symmetric about the midsagittal plane and consisted of six springs with stiffness coefficients of 8.0 N/mm (two anterior, two posterior, and two lateral). The springs originated from the head, connected below the C3 mount, and were preloaded to stabilize the head and upper cervical spine
Fig. 1. High-speed camera photographs of the model immediately prior to impact to the: a) upper forehead in the midline, b) upper left lateral side of the forehead, and c) upper left lateral side of the head. The model consisted of a human upper cervical spine specimen, occiput through C3, mounted to a surrogate torso mass and carrying an anthropometric surrogate head. The head and neck were stabilized using anterior, lateral, and posterior muscle force replication cables. One-component load cells (1) were positioned between the C3 mount and torso mass and at the impact barrier. Accelerometers (2) were rigidly fixed to the head center of mass and torso mass. The global coordinate system (hv) was fixed to the ground and had its positive h-axis oriented horizontally to the right and positive v-axis oriented vertically upward. The head coordinate system (xyz) was fixed to and moved with the head and had its positive x-axis oriented left, positive y-axis oriented superiorly, and positive z-axis oriented anteriorly relative to the specimen.
908
P.C. Ivancic / Clinical Biomechanics 29 (2014) 906–911
18.2 kg, was higher than the 16 kg mass used in prior studies which approximated the fraction of the torso mass that acted on the neck during head-first drop tests (Nightingale et al., 1996b). The orientation of the C3 mount relative to the torso mass determined the head impact location. The C3 mount was: tilted posteriorly by 45° relative to the torso mass to simulate impact to the upper forehead in the midline (Fig. 1a); tilted posteriorly by 45° and axially rotated to the right by 45° to simulate impact to the upper left lateral side of the forehead (Fig. 1b); and tilted posteriorly by 45° and axially rotated to the right by 90° to simulate impact to the upper left lateral side of the head (Fig. 1c). The sled was mounted on low-friction linear bearings, which translated along two precision ground, stainless steel shafts. A highspeed potentiometer was mounted to the sled to determine its horizontal translation. The sled was accelerated using an acceleration generation system consisting of a piston, high-energy compression springs, and computer controlled electromagnetic release. At the time of the electromagnet release, a trigger signal initiated high-speed camera recording at 500 frames/s (MotionPRO, Redlake MSAD, San Diego, CA, USA). The model was abruptly decelerated upon head contact with a padded barrier. The impact apparatus was fully controlled by a personal computer and custom LabVIEW software (LabVIEW 8.5, National Instruments, Austin, TX). Uni-axial load cells, aligned with the direction of sled motion, were used to measure impact barrier loads and neck loads inferior to the C3 mount (40 kN ultimate load, model LCCA-3K, Omega Engineering, Inc., Stamford, CT). Bi-axial accelerometers (50 g capacity; part no. ADXL250JQC, Analog Devices, Norwood, MA, USA) were fixed to the torso mass aligned along the direction of sled motion and to the center of mass of the head aligned in the sagittal and frontal planes. Load cell, accelerometer, and potentiometer data were continuously sampled at 1 kHz using an analog-to-digital converter and the LabVIEW program. 2.3. Injury documentation and data analyses All specimens were examined for axis ring fractures and associated injuries using fluoroscopy and detailed anatomical dissection. Specimen-specific load and acceleration time-history responses during impact were determined for the specimens that sustained ring fractures of the axis using the load cell and accelerometer data. Impact velocity was determined by numerical differentiation of the sled potentiometer data. Horizontal impact forces at the barrier and inferior to the C3 vertebra were expressed in a global coordinate system (hv; Fig. 1a) which was fixed to the ground and had its horizontal axis oriented along the direction of sled translation and its vertical axis oriented upward. Head accelerations were expressed in a head coordinate system (xyz; Fig. 1a) which was fixed to and moved with the head and had its positive x-axis oriented left, positive y-axis oriented superiorly, and positive z-axis oriented anteriorly relative to the specimen. Average peak loads and accelerations and their occurrence times during impact were determined. Accelerometer and velocity data were digitally filtered using a 3rd order, dual pass, Butterworth low-pass filter at a cutoff frequency of 50 Hz. Average occurrence times of the peak loads and accelerations were statistically compared (P b 0.05) using ANOVA and Bonferroni pair-wise post-hoc tests. 3. Results Of the 13 upper cervical spines tested, 5 specimens sustained fractures of the ring of the axis and we focus our further analyses on
Fig. 2. Schematic representations of the fracture patterns and post-trauma fluoroscopic images for each specimen that sustained axis ring fractures. The injuries were due to impact to: a) the upper forehead in the midline in specimen 1; b,c,d) the upper left lateral side of the forehead in specimens 2, 3, and 4; and e) the upper left lateral side of the head in specimen 5.
P.C. Ivancic / Clinical Biomechanics 29 (2014) 906–911
these specimens. The fractures were due to impacts to: the upper forehead in the midline (specimen 1), the upper left lateral side of the forehead (specimens 2, 3, and 4), and the upper left lateral side of the head (specimen 5). The average age of these 5 specimens was 81.6 years with 3 male and 2 female donors. The average horizontal velocity of the torso mass and C3 vertebra at the time of head impact with the barrier was 3.4 (SD 0.8) m/s with range of 2.7 to 4.7 m/s. On average, the forward displacement of the C3 vertebra in the global coordinate system reached its peak at 67 ms following head impact. The post-trauma fluoroscopic images and schematic drawings of the axis fracture patterns appear in Fig. 2. All 5 specimens sustained multiple asymmetric fractures. The axis fractures were most commonly observed at the superior facet or facets (all specimens), lamina (all except specimen 4), or posterior wall of the C2 vertebral body (all except specimen 2) with less common fracture sites including the pars interarticularis (specimens 2 and 3), inferior facet (specimen 1), and pedicle (specimen 1). No odontoid fractures were observed in any of these specimens. Associated atlas injuries included: a plough fracture variant (specimen 1), right anterior arch and bilateral posterior arch fractures (specimens 2 and 3), and left facet and posterior arch fractures (specimen 4). The specimen-specific time-history responses for forces and accelerations appear in Fig. 3. In all head impact configurations, lag time was observed between the onset times of compression forces at the impact barrier and C3 vertebra (Fig. 3a). Posterior and inferior decelerations of the head were observed immediately following head impact in all impact configurations (Fig. 3b). Right lateral decelerations of the head were initially observed in specimens 2, 3, and 4 due to impact to the upper left lateral side of the forehead and in specimen 5 due to
909
Table 1 Average peak biomechanical response parameters including acceleration (g) and force (N) ordered chronologically among specimens that sustained axis ring fractures. Acceleration peaks include right (Ax), inferior (Ay), and posterior (Az) for the head in the head coordinate system and deceleration of C3 (Ah) in the global coordinate system. Force peaks (Fh) at the impact barrier and neck, first local and absolute, are expressed in the global coordinate system. Event
Time (ms)
Peak
Significant differences
1) Ax, head 2) Az, head 3) Ay, head 4) Fh, impact barrier: first local peak 5) Fh, neck: first local peak 6) Fh, impact barrier: absolute 7) Fh, neck: absolute 8) Ah, C3
18.0 (2.7) 18.6 (3.6) 18.6 (4.0) 19.8 (3.6)
−14.8 (5.2) −17.2 (3.5) −14.7 (4.1) −2038.7 (685.7)
6–8 6–8 6–8 6–8
22.8 (4.0) 38.4 (5.7) 45.0 (9.8) 51.0 (9.1)
−1158.3 (422.1) −2293.7 (411.1) −1761.2 (107.8) −11.1 (1.4)
6–8 1–5,7,8 1–6 1–6
impact to the upper left lateral side of the head. Subsequently, the deceleration peaks of the C3 vertebra occurred (Fig. 3c). The average peak forces and accelerations for the specimens that sustained axis ring fractures are ordered chronologically in Table 1. The first local force peaks at the impact barrier and neck and all peak head accelerations occurred between 18.0 and 22.8 ms, significantly earlier than the absolute force peaks and peak deceleration of the C3 vertebra, between 38.4 and 51.0 ms. Peak head acceleration reached 17.2 g in the posterior direction. The first local force peaks at the neck and impact
Fig. 3. Time-history responses for forces and accelerations for each specimen that sustained axis ring fractures. Time = 0 represents the onset of force at the impact barrier. The loads include: a) impact forces (Fh) at the barrier and C3 vertebra in the global coordinate system. The accelerations include: b) right/left (Ax), superior/inferior (Ay), and anterior/posterior (Az) accelerations of the head in the head coordinate system and c) horizontal deceleration (Ah) of the C3 vertebra in the global coordinate system.
910
P.C. Ivancic / Clinical Biomechanics 29 (2014) 906–911
barrier were 66% and 89% of the absolute peaks, respectively. Peak deceleration of the C3 vertebra reached 11.1 g at 51.0 ms. 4. Discussion Our biomechanical study documented asymmetrical axis ring fractures due to each of the three head impact locations studied: upper forehead in the midline (1 of 4 specimens), upper left lateral side of the forehead (3 of 4 specimens), and upper left lateral side of the head (1 of 4 specimens). Our model consisted of an upper cervical spine specimen mounted between a surrogate head and surrogate torso mass. Odontoid fractures were observed in the remainder of the specimens that sustained impact to the upper forehead in the midline. No other axis injuries were observed with the exception of left superior facet fracture in a single specimen due to impact to the upper left lateral side of the head. Variability existed in the axis ring fractures observed among the specimens (Fig. 2). Our biomechanical results are consistent with prior clinical studies which observed asymmetrical fractures at any part of the axis ring including the laminae, superior or inferior facets, pars interarticularis, pedicles, or posterior wall of the vertebral body (Effendi et al., 1981). These authors provide detailed descriptions and illustrations of the structural anatomy of the axis in relation to specific axis ring fractures. Average peak head accelerations occurred between 18.0 and 18.6 ms, significantly earlier than the peak forces at the impact barrier and neck and peak deceleration of the C3 vertebra (Table 1). The first local peak forces at the impact barrier, 2038.7 N at 19.8 ms, and at the neck, 1158.3 N at 22.8 ms, both occurred significantly earlier than the absolute force peaks: 2293.7 N at the impact barrier at 38.4 ms; 1761.2 N at the neck at 45.0 ms. These data indicate that the axis ring fractures occurred within 50 ms at neck loads at or below 1761.2 N. Our biomechanical results should be considered in the context of the limitations associated with our model and experimental protocol. There exist 3 main limitations associated with the boundary conditions of our model. First, our head–neck–torso model included an occiput through C3 specimen and not the whole cervical spine specimen. Our model enabled torso inertial loads to be transferred directly to the upper cervical spine through the C3 vertebra sufficient to cause axis ring fractures. This boundary condition, which permitted only horizontal translation of C3, was simplified as compared to that which occurs during real-life trauma. Second, our model did not include an active neuromuscular response, however axial preload, postural head and neck stability, and passive resistance to motion were provided by muscle force replication. Third, our results are highly dependent upon the initial posture of the specimen. We investigated 3 head impact locations: upper forehead in the midline, upper left lateral side of the forehead, and upper left lateral side of the head. A finite element model of the cervical spine, validated against experimental trauma data, may be ideally suited to investigate effects of boundary conditions and specimen posture on axis fracture patterns. Further limitations of our model exist. We created fractures of the ring of the axis and used fluoroscopy and anatomical dissections to document the fracture patterns. However, we did not evaluate the ligamentous injuries, which are important clinically when evaluating spinal instability and choosing reduction and fixation methods. The average age of the specimens that sustained ring fractures of the axis, 81.6 years, was older than the younger population at greatest risk of spinal trauma due to high speed motor vehicle crashes. However, the availability of younger cadaveric specimens is limited. We did not measure bone mineral density which has been correlated with vertebral failure force. The present collection rates for load cell, accelerometer, and potentiometer data were below those recommended for impact testing (SAE, 2014). Future studies of cervical spine injury mechanisms may address these limitations. While limited by these factors, our study provides insight into the etiology of fracture patterns of the axis ring based upon head impact location.
In prior case series, clinical researchers have hypothesized that the axis ring fracture patterns are related to the morphology of the axis as a transitional vertebra of the upper cervical spine (Effendi et al., 1981; Francis et al., 1981; Schneider et al., 1965). These prior studies have graphically illustrated the variation in loading throughout the axis that is thought to contribute to ring fractures. Orientation of the facets and inferior endplate of the axis in the sagittal and frontal planes contributes to the bony stress distribution due to traumatic load and the resultant fracture patterns. In the sagittal plane, anterior to the pars interarticularis, compressive load is transferred in part to anterior shear due to the posterior tilt of the superior facets and the anterior tilt of the inferior endplate. Posterior to the pars interarticularis, the compressive load is transferred in part to posterior shear due to the posterior tilt of the inferior facets. This difference in shear load orientation at the pars interarticularis and the offset distance between the superior and inferior facets may predispose the pars to fracture due to traumatic compression load that is accentuated with forced hyperextension. In the frontal plane, the medial-to-lateral orientation of the facets is downward for the superior facets and upward for the inferior facets. The traumatic compressive load is transferred in part to inward radial force at the axis ring due to the frontal plane orientations of the superior and inferior facets. Lamina fractures of the axis were observed in the majority of the present specimens. The fracture pattern of the right lamina of specimen 2 (Fig. 2b) suggested that part of the compressive force was transferred to inward radial force through the inferior facets. The morphology of the upper cervical spine, when considered with the present biomechanical responses and analyses of the high speed movies, may be used to understand the mechanisms of axis ring fractures in our specimens. Previous clinical case series have documented symmetrical fracture passing transversely through the superior facets immediately posterior to the dens, bordering the foramen transversarium, and through the inferior vertebral notch (Cornish, 1968; Williams, 1975). This fracture pattern is consistent with that observed in specimens 1 and 4 (Fig. 2a,d). The associated atlas fractures included a plough fracture variant in specimen 1 and left facet and posterior arch fractures in specimen 4. In specimen 1, the inferior portion of the anterior atlantal arch was displaced anteriorly with the fracture fragment width consistent with that of the dens width. The superior portion of the anterior atlantal arch remained intact. Analyses of the high speed movies indicated that the near-symmetrical fractures passing transversely through the superior facets in both specimens occurred due to continued forward momentum of the torso mass following the rapid deceleration of the head. The fracture patterns suggest that the anterior bony fragment of the axis (dens, body, and anterior portion of superior facets) sheared posteriorly and extended relative to the C2 neural arch prior to flexing forward. Fractures of the pars interarticularis were observed in specimens 2 and 3 (Fig. 2b,c). In specimen 2, the fracture passed transversely through the left pars and directly anterior to the right pars. In specimen 3, a right pars fracture was observed along with left lamina fractures and an anterolaterally directed fracture of the left superior facet. Bilateral fractures of the pars interarticularis are commonly observed in clinical case series due to motor vehicle crashes or traumatic head-first falls (Cornish, 1968; Effendi et al., 1981; Williams, 1975). The pars interarticularis is predisposed to fracture due to its small cross sectional area, the presence of the foramen transversarium, the large bending moment that it sustains between the superior and inferior facets (White and Panjabi, 1990), and the combined anterior and posterior shear forces that it sustains during axial compression due to the sagittal orientation of the superior and inferior facets. Our model produced axis ring fractures due to simulated head impacts consistent with those observed due to real-life motor vehicle collisions and fall from height. Axis ring fractures are generally associated with a low incidence of neurological sequelae due to the widening of the spinal canal and intervertebral foramina. Fundamental studies of
P.C. Ivancic / Clinical Biomechanics 29 (2014) 906–911
the injury mechanisms may help determine the best method of reduction and stabilization of axis ring fractures. Acknowledgments No external support was received for this research. References Althoff, B., 1979. Fracture of the odontoid process. An experimental study. Acta Orthop. Scand. Suppl. 177. Chittiboina, P., Wylen, E., Ogden, A., Mukherjee, D.P., Vannemreddy, P., Nanda, A., 2009. Traumatic spondylolisthesis of the axis: a biomechanical comparison of clinically relevant anterior and posterior fusion techniques. J. Neurosurg. Spine 11, 379–387. Cornish, B.L., 1968. Traumatic spondylolisthesis of the axis. J. Bone Joint Surg. (Br.) 50, 31–43. Descarreaux, M., Blouin, J.S., Teasdale, N., 2003. A non-invasive technique for measurement of cervical vertebral angle: report of a preliminary study. Eur. Spine J. 12, 314–319. Dibb, A.T., Nightingale, R.W., Luck, J.F., Chancey, V.C., Fronheiser, L.E., Myers, B.S., 2009. Tension and combined tension-extension structural response and tolerance properties of the human male ligamentous cervical spine. J. Biomech. Eng. 131, 081008. Duggal, N., Chamberlain, R.H., Perez-Garza, L.E., Espinoza-Larios, A., Sonntag, V.K., Crawford, N.R., 2007. Hangman's fracture: a biomechanical comparison of stabilization techniques. Spine (Phila Pa 1976) 32, 182–187. Dvorak, M.F., Street, J.T., Lenehan, B.J., 2012. Injuries to C1 and C2 (excluding odontoid fractures), chapter 46, In: Benzel, E.C. (Ed.), The Cervical Spine, 5th ed. Effendi, B., Roy, D., Cornish, B., Dussault, R.G., Laurin, C.A., 1981. Fractures of the ring of the axis. A classification based on the analysis of 131 cases. J. Bone Joint Surg. (Br.) 63-B, 319–327.
911
Francis, W.R., Fielding, J.W., Hawkins, R.J., Pepin, J., Hensinger, R., 1981. Traumatic spondylolisthesis of the axis. J. Bone Joint Surg. (Br.) 63-B, 313–318. Levine, A.M., Edwards, C.C., 1985. The management of traumatic spondylolisthesis of the axis. J. Bone Joint Surg. Am. 67, 217–226. Nightingale, R.W., McElhaney, J.H., Richardson, W.J., Best, T.M., Myers, B.S., 1996a. Experimental impact injury to the cervical spine: relating motion of the head and the mechanism of injury. J. Bone Joint Surg. Am. 78, 412–421. Nightingale, R.W., McElhaney, J.H., Richardson, W.J., Myers, B.S., 1996b. Dynamic responses of the head and cervical spine to axial impact loading. J. Biomech. 29, 307–318. Nightingale, R.W., McElhaney, J.H., Camacho, D.L., Kleinberger, M., Winkelstein, B.A., Myers, B.S., 1997a. The dynamic responses of the cervical spine: buckling, end conditions, and tolerance in compressive impacts. Society of Automotive Engineers Paper No. 973344. Nightingale, R.W., Richardson, W.J., Myers, B.S., 1997b. The effects of padded surfaces on the risk for cervical spine injury. Spine (Phila Pa 1976) 22, 2380–2387. Nightingale, R.W., Winkelstein, B.A., Knaub, K.E., Richardson, W.J., Luck, J.F., Myers, B.S., 2002. Comparative strengths and structural properties of the upper and lower cervical spine in flexion and extension. J. Biomech. 35, 725–732. SAE. Safety Test International Standards Committee, 2014. SAE J211/1 201403 surface vehicle recommended practice. Instrumentation for Impact Test — Part 1 — Electronic InstrumentationSociety of Automotive Engineers, Inc., Warrendale, PA. Samaha, C., Lazennec, J.Y., Laporte, C., Saillant, G., 2000. Hangman's fracture: the relationship between asymmetry and instability. J. Bone Joint Surg. (Br.) 82, 1046–1052. Schneider, R.C., Livingston, K.E., Cave, A.J., Hamilton, G., 1965. “Hangman's fracture” of the cervical spine. J. Neurosurg. 22, 141–154. White III, A.A., Panjabi, M.M., 1990. Clinical Biomechanics of the Spine. J.B. Lippincott Co., Philadelphia. Williams, T.G., 1975. Hangman's fracture. J. Bone Joint Surg. (Br.) 57, 82–88.
Contents lists available at ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Axis ring fractures due to simulated head impacts Paul C. Ivancic ⁎ Biomechanics Research Laboratory, Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, New Haven, CT, USA
a r t i c l e
i n f o
Article history: Received 21 March 2014 Accepted 30 June 2014 Keywords: Axis ring fracture Trauma Cervical spine Biomechanics
a b s t r a c t Background: We investigated mechanisms of axis ring fractures due to simulated head impacts. Methods: Our model consisted of a human upper cervical spine specimen (occiput through C3) mounted to a surrogate torso mass on a sled and carrying a surrogate head. We divided 13 specimens into 3 groups based upon head impact location: upper forehead in the midline, upper lateral side of the forehead, and upper lateral side of the head. Post-impact fluoroscopy and anatomical dissection documented the injuries. Average occurrence times of the peak loads and accelerations were statistically compared (P b 0.05) using ANOVA and Bonferroni pair-wise post-hoc tests. Findings: Of the 13 upper cervical spines tested, 5 specimens sustained axis ring fractures with the most common mechanism being impact to the upper left lateral side of the forehead. The first local force peaks at the impact barrier and neck and all peak head accelerations occurred between 18.0 and 22.8 ms, significantly earlier than the absolute force peaks. The average peak neck loads reached 1761.2 N and the axis ring fractures occurred within 50 ms. Interpretation: We observed asymmetrical fractures of the axis ring including fractures of the superior and inferior facets, laminae, posterior wall of the vertebral body, pars interarticularis, and pedicles. The fracture patterns were related to the morphology of the axis as a transitional vertebra of the upper cervical spine. Understanding the mechanisms of axis ring fractures may help in choosing the optimal reduction technique and stabilization method based upon the specific fracture pattern. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Traumatic fracture of the ring of the axis is the second most common injury of the C2 vertebra; the most common being odontoid fracture (Dvorak et al., 2012). Axis ring fractures occur most often due to blunt head impact in younger individuals during high speed motor vehicle crashes and in older individuals during fall from standing height (Cornish, 1968; Effendi et al., 1981; Levine and Edwards, 1985; Samaha et al., 2000). The classic fracture pattern is bilateral fracture of the pars interarticularis, between the superior and inferior facets of the C2 vertebra, with forward displacement of the anterior fracture fragment. This forward displacement causes widening of the spinal canal and intervertebral foramina at C2 consistent with the low incidence of neurological sequelae associated with the injury observed clinically (Dvorak et al., 2012). White and Panjabi observed that the pars interarticularis of the axis is the most likely site of fracture due to forced hyperextension as it sustains the peak bending moment and is structurally weakened due to its small cross-sectional area and the presence of the foramen transversarium (White and Panjabi, 1990). ⁎ Biomechanics Research Laboratory, Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, 333 Cedar St., P.O. Box 208071, New Haven, CT 06520-8071, USA. E-mail address: [email protected].
http://dx.doi.org/10.1016/j.clinbiomech.2014.06.017 0268-0033/© 2014 Elsevier Ltd. All rights reserved.
Traditionally referred to as Hangman's fracture, the radiographic presentation of axis ring fractures due to modern civilian trauma is often similar to that previously documented due to judicial hanging, though the injury mechanisms differ (Effendi et al., 1981). The mechanism of injury due to judicial hanging with a submental knot has been described as forced hyperextension causing traumatic spondylolisthesis of the axis with death due to distraction of the cord or suffocation (Schneider et al., 1965). The mechanisms of axis ring fractures due to motor vehicle crashes have been described as forced hyperextension combined with axial compression often caused by blunt impact of the head with the vehicle interior and continued momentum of the torso (Effendi et al., 1981). Others hypothesize that flexion–compression or flexion–distraction mechanisms may produce certain axis ring fractures either as primary injury mechanisms or subsequent to forced extension–compression (Levine and Edwards, 1985). Clinical case series have documented fractures of the axis ring primarily due to motor vehicle crashes. In a study of 131 patients, Effendi et al. (1981) observed that the axis ring fractures may occur at any part of the bony canal including the laminae, superior or inferior facets, pars interarticularis, pedicles, or posterior wall of the vertebral body. They found that the fractures were almost always bilateral but rarely symmetrical and did not necessarily occur at the weakest part of the ring. The asymmetrical presentation of the fractures was hypothesized to be due to rotated head posture at the time of head impact.
P.C. Ivancic / Clinical Biomechanics 29 (2014) 906–911
Other case series have observed symmetrical fracture passing transversely through the superior facets immediately posterior to the dens, bordering the foramen transversarium, and through the inferior vertebral notch (Cornish, 1968; Williams, 1975). The prior studies demonstrate that axis ring fracture patterns are highly variable and dependent upon numerous trauma-related and morphological factors. In the prior case series, the clinical researchers have hypothesized that the fracture patterns specific to the axis ring are dependent upon the morphology of the axis as a transitional vertebra of the upper cervical spine (Effendi et al., 1981; Francis et al., 1981; Schneider et al., 1965). These studies describe the traumatic axial load due to head impact transferred inferiorly through the occipital condyles, through the superior and inferior facets of the atlas, and through the superior facets of the axis, and then being split among the body and left and right inferior facets of the axis. Additional loads include hyperextension and posterior shear transferred to the anterior articular surface of the dens through the anterior atlantal arch. Lastly, these loads are combined with those transferred superiorly through the cervical spine due to torso inertia. Most prior biomechanical studies of traumatic spondylolisthesis of the axis have created the injuries manually for evaluation of fixation techniques (Chittiboina et al., 2009; Duggal et al., 2007). Other studies have created realistic injuries in some specimens due to simulated trauma. In their pioneering study of the mechanism of odontoid fracture, Althoff (1979) created bilateral fractures of the pars interarticularis and complete ruptures of the C2/3 disc and anterior and posterior longitudinal ligaments in a single specimen due to impact to the upper lateral side of the head. Others have reported traumatic spondylolisthesis that occurred inconsistently in previous experimental series due to distraction (Dibb et al., 2009) or hyperextension (Nightingale et al., 2002) of upper cervical spine specimens or distractive-extension (Nightingale et al., 1996a, 1997a,b) during vertical drop tests of inverted head–neck specimens. Continued biomechanical research is needed to understand the etiology of axis ring fracture patterns. The goal of this biomechanical study was to investigate axis ring fractures due to simulated head impacts of a model consisting of an upper cervical spine specimen mounted between a surrogate head and surrogate torso mass. We evaluated the previously proposed injury mechanism hypotheses by analyzing the fracture patterns and biomechanical responses of the present specimens.
907
specimen in normal lordotic posture prior to the impact, similar to the in vivo neutral posture (Descarreaux et al., 2003). Cable loads were not measured. No counterbalancing of the head weight was required.
2.2. Simulated head impact To investigate various fracture patterns to the ring of the axis, we divided the specimens into 3 groups based upon head impact location: upper forehead in the midline (n = 5), upper left lateral side of the forehead (n = 4), and upper left lateral side of the head (n = 4). Each specimen was subjected to a single traumatic impact. A sled apparatus, aligned horizontally relative to the ground, and barrier system were used to simulate head-first impacts. The apparatus consisted of a surrogate torso mass rigidly fixed to a custom sled with the C3 mount of the specimen rigidly fixed at the front of the mass (Fig. 1). The mass,
2. Methods 2.1. Human upper cervical spine model with surrogate head We prepared 13 human osteoligamentous upper cervical spine specimens that were fresh-frozen (occiput to C3 vertebra; 6 males, 7 females; average age: 83.1 years). Apart from typical age related degenerative changes, the specimens did not suffer from any disease that could have affected the osteoligamentous structures. The specimens were rigidly mounted in resin at the occiput and C3 vertebra in normal neutral posture, consistent with the in vivo sagittal upright neutral posture (Descarreaux et al., 2003), and confirmed using fluoroscopy. A Hybrid III surrogate head (4.2 kg mass; 0.0244 kg m2 sagittal plane moment of inertia; Humanetics Innovative Solutions, Plymouth, MI, USA) was rigidly fixed to the occipital mount in anatomical position. The head and spine specimen were stabilized using muscle force replication which simulated the net effects of anterior, lateral, and posterior muscle forces. These simulated muscle forces produced compressive loading to the spine which provided static postural spine stability and passive resistance to motions during the impact. The muscle force replication was symmetric about the midsagittal plane and consisted of six springs with stiffness coefficients of 8.0 N/mm (two anterior, two posterior, and two lateral). The springs originated from the head, connected below the C3 mount, and were preloaded to stabilize the head and upper cervical spine
Fig. 1. High-speed camera photographs of the model immediately prior to impact to the: a) upper forehead in the midline, b) upper left lateral side of the forehead, and c) upper left lateral side of the head. The model consisted of a human upper cervical spine specimen, occiput through C3, mounted to a surrogate torso mass and carrying an anthropometric surrogate head. The head and neck were stabilized using anterior, lateral, and posterior muscle force replication cables. One-component load cells (1) were positioned between the C3 mount and torso mass and at the impact barrier. Accelerometers (2) were rigidly fixed to the head center of mass and torso mass. The global coordinate system (hv) was fixed to the ground and had its positive h-axis oriented horizontally to the right and positive v-axis oriented vertically upward. The head coordinate system (xyz) was fixed to and moved with the head and had its positive x-axis oriented left, positive y-axis oriented superiorly, and positive z-axis oriented anteriorly relative to the specimen.
908
P.C. Ivancic / Clinical Biomechanics 29 (2014) 906–911
18.2 kg, was higher than the 16 kg mass used in prior studies which approximated the fraction of the torso mass that acted on the neck during head-first drop tests (Nightingale et al., 1996b). The orientation of the C3 mount relative to the torso mass determined the head impact location. The C3 mount was: tilted posteriorly by 45° relative to the torso mass to simulate impact to the upper forehead in the midline (Fig. 1a); tilted posteriorly by 45° and axially rotated to the right by 45° to simulate impact to the upper left lateral side of the forehead (Fig. 1b); and tilted posteriorly by 45° and axially rotated to the right by 90° to simulate impact to the upper left lateral side of the head (Fig. 1c). The sled was mounted on low-friction linear bearings, which translated along two precision ground, stainless steel shafts. A highspeed potentiometer was mounted to the sled to determine its horizontal translation. The sled was accelerated using an acceleration generation system consisting of a piston, high-energy compression springs, and computer controlled electromagnetic release. At the time of the electromagnet release, a trigger signal initiated high-speed camera recording at 500 frames/s (MotionPRO, Redlake MSAD, San Diego, CA, USA). The model was abruptly decelerated upon head contact with a padded barrier. The impact apparatus was fully controlled by a personal computer and custom LabVIEW software (LabVIEW 8.5, National Instruments, Austin, TX). Uni-axial load cells, aligned with the direction of sled motion, were used to measure impact barrier loads and neck loads inferior to the C3 mount (40 kN ultimate load, model LCCA-3K, Omega Engineering, Inc., Stamford, CT). Bi-axial accelerometers (50 g capacity; part no. ADXL250JQC, Analog Devices, Norwood, MA, USA) were fixed to the torso mass aligned along the direction of sled motion and to the center of mass of the head aligned in the sagittal and frontal planes. Load cell, accelerometer, and potentiometer data were continuously sampled at 1 kHz using an analog-to-digital converter and the LabVIEW program. 2.3. Injury documentation and data analyses All specimens were examined for axis ring fractures and associated injuries using fluoroscopy and detailed anatomical dissection. Specimen-specific load and acceleration time-history responses during impact were determined for the specimens that sustained ring fractures of the axis using the load cell and accelerometer data. Impact velocity was determined by numerical differentiation of the sled potentiometer data. Horizontal impact forces at the barrier and inferior to the C3 vertebra were expressed in a global coordinate system (hv; Fig. 1a) which was fixed to the ground and had its horizontal axis oriented along the direction of sled translation and its vertical axis oriented upward. Head accelerations were expressed in a head coordinate system (xyz; Fig. 1a) which was fixed to and moved with the head and had its positive x-axis oriented left, positive y-axis oriented superiorly, and positive z-axis oriented anteriorly relative to the specimen. Average peak loads and accelerations and their occurrence times during impact were determined. Accelerometer and velocity data were digitally filtered using a 3rd order, dual pass, Butterworth low-pass filter at a cutoff frequency of 50 Hz. Average occurrence times of the peak loads and accelerations were statistically compared (P b 0.05) using ANOVA and Bonferroni pair-wise post-hoc tests. 3. Results Of the 13 upper cervical spines tested, 5 specimens sustained fractures of the ring of the axis and we focus our further analyses on
Fig. 2. Schematic representations of the fracture patterns and post-trauma fluoroscopic images for each specimen that sustained axis ring fractures. The injuries were due to impact to: a) the upper forehead in the midline in specimen 1; b,c,d) the upper left lateral side of the forehead in specimens 2, 3, and 4; and e) the upper left lateral side of the head in specimen 5.
P.C. Ivancic / Clinical Biomechanics 29 (2014) 906–911
these specimens. The fractures were due to impacts to: the upper forehead in the midline (specimen 1), the upper left lateral side of the forehead (specimens 2, 3, and 4), and the upper left lateral side of the head (specimen 5). The average age of these 5 specimens was 81.6 years with 3 male and 2 female donors. The average horizontal velocity of the torso mass and C3 vertebra at the time of head impact with the barrier was 3.4 (SD 0.8) m/s with range of 2.7 to 4.7 m/s. On average, the forward displacement of the C3 vertebra in the global coordinate system reached its peak at 67 ms following head impact. The post-trauma fluoroscopic images and schematic drawings of the axis fracture patterns appear in Fig. 2. All 5 specimens sustained multiple asymmetric fractures. The axis fractures were most commonly observed at the superior facet or facets (all specimens), lamina (all except specimen 4), or posterior wall of the C2 vertebral body (all except specimen 2) with less common fracture sites including the pars interarticularis (specimens 2 and 3), inferior facet (specimen 1), and pedicle (specimen 1). No odontoid fractures were observed in any of these specimens. Associated atlas injuries included: a plough fracture variant (specimen 1), right anterior arch and bilateral posterior arch fractures (specimens 2 and 3), and left facet and posterior arch fractures (specimen 4). The specimen-specific time-history responses for forces and accelerations appear in Fig. 3. In all head impact configurations, lag time was observed between the onset times of compression forces at the impact barrier and C3 vertebra (Fig. 3a). Posterior and inferior decelerations of the head were observed immediately following head impact in all impact configurations (Fig. 3b). Right lateral decelerations of the head were initially observed in specimens 2, 3, and 4 due to impact to the upper left lateral side of the forehead and in specimen 5 due to
909
Table 1 Average peak biomechanical response parameters including acceleration (g) and force (N) ordered chronologically among specimens that sustained axis ring fractures. Acceleration peaks include right (Ax), inferior (Ay), and posterior (Az) for the head in the head coordinate system and deceleration of C3 (Ah) in the global coordinate system. Force peaks (Fh) at the impact barrier and neck, first local and absolute, are expressed in the global coordinate system. Event
Time (ms)
Peak
Significant differences
1) Ax, head 2) Az, head 3) Ay, head 4) Fh, impact barrier: first local peak 5) Fh, neck: first local peak 6) Fh, impact barrier: absolute 7) Fh, neck: absolute 8) Ah, C3
18.0 (2.7) 18.6 (3.6) 18.6 (4.0) 19.8 (3.6)
−14.8 (5.2) −17.2 (3.5) −14.7 (4.1) −2038.7 (685.7)
6–8 6–8 6–8 6–8
22.8 (4.0) 38.4 (5.7) 45.0 (9.8) 51.0 (9.1)
−1158.3 (422.1) −2293.7 (411.1) −1761.2 (107.8) −11.1 (1.4)
6–8 1–5,7,8 1–6 1–6
impact to the upper left lateral side of the head. Subsequently, the deceleration peaks of the C3 vertebra occurred (Fig. 3c). The average peak forces and accelerations for the specimens that sustained axis ring fractures are ordered chronologically in Table 1. The first local force peaks at the impact barrier and neck and all peak head accelerations occurred between 18.0 and 22.8 ms, significantly earlier than the absolute force peaks and peak deceleration of the C3 vertebra, between 38.4 and 51.0 ms. Peak head acceleration reached 17.2 g in the posterior direction. The first local force peaks at the neck and impact
Fig. 3. Time-history responses for forces and accelerations for each specimen that sustained axis ring fractures. Time = 0 represents the onset of force at the impact barrier. The loads include: a) impact forces (Fh) at the barrier and C3 vertebra in the global coordinate system. The accelerations include: b) right/left (Ax), superior/inferior (Ay), and anterior/posterior (Az) accelerations of the head in the head coordinate system and c) horizontal deceleration (Ah) of the C3 vertebra in the global coordinate system.
910
P.C. Ivancic / Clinical Biomechanics 29 (2014) 906–911
barrier were 66% and 89% of the absolute peaks, respectively. Peak deceleration of the C3 vertebra reached 11.1 g at 51.0 ms. 4. Discussion Our biomechanical study documented asymmetrical axis ring fractures due to each of the three head impact locations studied: upper forehead in the midline (1 of 4 specimens), upper left lateral side of the forehead (3 of 4 specimens), and upper left lateral side of the head (1 of 4 specimens). Our model consisted of an upper cervical spine specimen mounted between a surrogate head and surrogate torso mass. Odontoid fractures were observed in the remainder of the specimens that sustained impact to the upper forehead in the midline. No other axis injuries were observed with the exception of left superior facet fracture in a single specimen due to impact to the upper left lateral side of the head. Variability existed in the axis ring fractures observed among the specimens (Fig. 2). Our biomechanical results are consistent with prior clinical studies which observed asymmetrical fractures at any part of the axis ring including the laminae, superior or inferior facets, pars interarticularis, pedicles, or posterior wall of the vertebral body (Effendi et al., 1981). These authors provide detailed descriptions and illustrations of the structural anatomy of the axis in relation to specific axis ring fractures. Average peak head accelerations occurred between 18.0 and 18.6 ms, significantly earlier than the peak forces at the impact barrier and neck and peak deceleration of the C3 vertebra (Table 1). The first local peak forces at the impact barrier, 2038.7 N at 19.8 ms, and at the neck, 1158.3 N at 22.8 ms, both occurred significantly earlier than the absolute force peaks: 2293.7 N at the impact barrier at 38.4 ms; 1761.2 N at the neck at 45.0 ms. These data indicate that the axis ring fractures occurred within 50 ms at neck loads at or below 1761.2 N. Our biomechanical results should be considered in the context of the limitations associated with our model and experimental protocol. There exist 3 main limitations associated with the boundary conditions of our model. First, our head–neck–torso model included an occiput through C3 specimen and not the whole cervical spine specimen. Our model enabled torso inertial loads to be transferred directly to the upper cervical spine through the C3 vertebra sufficient to cause axis ring fractures. This boundary condition, which permitted only horizontal translation of C3, was simplified as compared to that which occurs during real-life trauma. Second, our model did not include an active neuromuscular response, however axial preload, postural head and neck stability, and passive resistance to motion were provided by muscle force replication. Third, our results are highly dependent upon the initial posture of the specimen. We investigated 3 head impact locations: upper forehead in the midline, upper left lateral side of the forehead, and upper left lateral side of the head. A finite element model of the cervical spine, validated against experimental trauma data, may be ideally suited to investigate effects of boundary conditions and specimen posture on axis fracture patterns. Further limitations of our model exist. We created fractures of the ring of the axis and used fluoroscopy and anatomical dissections to document the fracture patterns. However, we did not evaluate the ligamentous injuries, which are important clinically when evaluating spinal instability and choosing reduction and fixation methods. The average age of the specimens that sustained ring fractures of the axis, 81.6 years, was older than the younger population at greatest risk of spinal trauma due to high speed motor vehicle crashes. However, the availability of younger cadaveric specimens is limited. We did not measure bone mineral density which has been correlated with vertebral failure force. The present collection rates for load cell, accelerometer, and potentiometer data were below those recommended for impact testing (SAE, 2014). Future studies of cervical spine injury mechanisms may address these limitations. While limited by these factors, our study provides insight into the etiology of fracture patterns of the axis ring based upon head impact location.
In prior case series, clinical researchers have hypothesized that the axis ring fracture patterns are related to the morphology of the axis as a transitional vertebra of the upper cervical spine (Effendi et al., 1981; Francis et al., 1981; Schneider et al., 1965). These prior studies have graphically illustrated the variation in loading throughout the axis that is thought to contribute to ring fractures. Orientation of the facets and inferior endplate of the axis in the sagittal and frontal planes contributes to the bony stress distribution due to traumatic load and the resultant fracture patterns. In the sagittal plane, anterior to the pars interarticularis, compressive load is transferred in part to anterior shear due to the posterior tilt of the superior facets and the anterior tilt of the inferior endplate. Posterior to the pars interarticularis, the compressive load is transferred in part to posterior shear due to the posterior tilt of the inferior facets. This difference in shear load orientation at the pars interarticularis and the offset distance between the superior and inferior facets may predispose the pars to fracture due to traumatic compression load that is accentuated with forced hyperextension. In the frontal plane, the medial-to-lateral orientation of the facets is downward for the superior facets and upward for the inferior facets. The traumatic compressive load is transferred in part to inward radial force at the axis ring due to the frontal plane orientations of the superior and inferior facets. Lamina fractures of the axis were observed in the majority of the present specimens. The fracture pattern of the right lamina of specimen 2 (Fig. 2b) suggested that part of the compressive force was transferred to inward radial force through the inferior facets. The morphology of the upper cervical spine, when considered with the present biomechanical responses and analyses of the high speed movies, may be used to understand the mechanisms of axis ring fractures in our specimens. Previous clinical case series have documented symmetrical fracture passing transversely through the superior facets immediately posterior to the dens, bordering the foramen transversarium, and through the inferior vertebral notch (Cornish, 1968; Williams, 1975). This fracture pattern is consistent with that observed in specimens 1 and 4 (Fig. 2a,d). The associated atlas fractures included a plough fracture variant in specimen 1 and left facet and posterior arch fractures in specimen 4. In specimen 1, the inferior portion of the anterior atlantal arch was displaced anteriorly with the fracture fragment width consistent with that of the dens width. The superior portion of the anterior atlantal arch remained intact. Analyses of the high speed movies indicated that the near-symmetrical fractures passing transversely through the superior facets in both specimens occurred due to continued forward momentum of the torso mass following the rapid deceleration of the head. The fracture patterns suggest that the anterior bony fragment of the axis (dens, body, and anterior portion of superior facets) sheared posteriorly and extended relative to the C2 neural arch prior to flexing forward. Fractures of the pars interarticularis were observed in specimens 2 and 3 (Fig. 2b,c). In specimen 2, the fracture passed transversely through the left pars and directly anterior to the right pars. In specimen 3, a right pars fracture was observed along with left lamina fractures and an anterolaterally directed fracture of the left superior facet. Bilateral fractures of the pars interarticularis are commonly observed in clinical case series due to motor vehicle crashes or traumatic head-first falls (Cornish, 1968; Effendi et al., 1981; Williams, 1975). The pars interarticularis is predisposed to fracture due to its small cross sectional area, the presence of the foramen transversarium, the large bending moment that it sustains between the superior and inferior facets (White and Panjabi, 1990), and the combined anterior and posterior shear forces that it sustains during axial compression due to the sagittal orientation of the superior and inferior facets. Our model produced axis ring fractures due to simulated head impacts consistent with those observed due to real-life motor vehicle collisions and fall from height. Axis ring fractures are generally associated with a low incidence of neurological sequelae due to the widening of the spinal canal and intervertebral foramina. Fundamental studies of
P.C. Ivancic / Clinical Biomechanics 29 (2014) 906–911
the injury mechanisms may help determine the best method of reduction and stabilization of axis ring fractures. Acknowledgments No external support was received for this research. References Althoff, B., 1979. Fracture of the odontoid process. An experimental study. Acta Orthop. Scand. Suppl. 177. Chittiboina, P., Wylen, E., Ogden, A., Mukherjee, D.P., Vannemreddy, P., Nanda, A., 2009. Traumatic spondylolisthesis of the axis: a biomechanical comparison of clinically relevant anterior and posterior fusion techniques. J. Neurosurg. Spine 11, 379–387. Cornish, B.L., 1968. Traumatic spondylolisthesis of the axis. J. Bone Joint Surg. (Br.) 50, 31–43. Descarreaux, M., Blouin, J.S., Teasdale, N., 2003. A non-invasive technique for measurement of cervical vertebral angle: report of a preliminary study. Eur. Spine J. 12, 314–319. Dibb, A.T., Nightingale, R.W., Luck, J.F., Chancey, V.C., Fronheiser, L.E., Myers, B.S., 2009. Tension and combined tension-extension structural response and tolerance properties of the human male ligamentous cervical spine. J. Biomech. Eng. 131, 081008. Duggal, N., Chamberlain, R.H., Perez-Garza, L.E., Espinoza-Larios, A., Sonntag, V.K., Crawford, N.R., 2007. Hangman's fracture: a biomechanical comparison of stabilization techniques. Spine (Phila Pa 1976) 32, 182–187. Dvorak, M.F., Street, J.T., Lenehan, B.J., 2012. Injuries to C1 and C2 (excluding odontoid fractures), chapter 46, In: Benzel, E.C. (Ed.), The Cervical Spine, 5th ed. Effendi, B., Roy, D., Cornish, B., Dussault, R.G., Laurin, C.A., 1981. Fractures of the ring of the axis. A classification based on the analysis of 131 cases. J. Bone Joint Surg. (Br.) 63-B, 319–327.
911
Francis, W.R., Fielding, J.W., Hawkins, R.J., Pepin, J., Hensinger, R., 1981. Traumatic spondylolisthesis of the axis. J. Bone Joint Surg. (Br.) 63-B, 313–318. Levine, A.M., Edwards, C.C., 1985. The management of traumatic spondylolisthesis of the axis. J. Bone Joint Surg. Am. 67, 217–226. Nightingale, R.W., McElhaney, J.H., Richardson, W.J., Best, T.M., Myers, B.S., 1996a. Experimental impact injury to the cervical spine: relating motion of the head and the mechanism of injury. J. Bone Joint Surg. Am. 78, 412–421. Nightingale, R.W., McElhaney, J.H., Richardson, W.J., Myers, B.S., 1996b. Dynamic responses of the head and cervical spine to axial impact loading. J. Biomech. 29, 307–318. Nightingale, R.W., McElhaney, J.H., Camacho, D.L., Kleinberger, M., Winkelstein, B.A., Myers, B.S., 1997a. The dynamic responses of the cervical spine: buckling, end conditions, and tolerance in compressive impacts. Society of Automotive Engineers Paper No. 973344. Nightingale, R.W., Richardson, W.J., Myers, B.S., 1997b. The effects of padded surfaces on the risk for cervical spine injury. Spine (Phila Pa 1976) 22, 2380–2387. Nightingale, R.W., Winkelstein, B.A., Knaub, K.E., Richardson, W.J., Luck, J.F., Myers, B.S., 2002. Comparative strengths and structural properties of the upper and lower cervical spine in flexion and extension. J. Biomech. 35, 725–732. SAE. Safety Test International Standards Committee, 2014. SAE J211/1 201403 surface vehicle recommended practice. Instrumentation for Impact Test — Part 1 — Electronic InstrumentationSociety of Automotive Engineers, Inc., Warrendale, PA. Samaha, C., Lazennec, J.Y., Laporte, C., Saillant, G., 2000. Hangman's fracture: the relationship between asymmetry and instability. J. Bone Joint Surg. (Br.) 82, 1046–1052. Schneider, R.C., Livingston, K.E., Cave, A.J., Hamilton, G., 1965. “Hangman's fracture” of the cervical spine. J. Neurosurg. 22, 141–154. White III, A.A., Panjabi, M.M., 1990. Clinical Biomechanics of the Spine. J.B. Lippincott Co., Philadelphia. Williams, T.G., 1975. Hangman's fracture. J. Bone Joint Surg. (Br.) 57, 82–88.