A healthy four-year-old boy restrained in a forward-facing car seat was involved in a high-speed motor vehicle accident in which the car collided with a tree. As reported by paramedics at the scene of the accident, he was conscious and was moving all extremities. Full cervical spine precautions were taken, and he was transported to the emergency room at a level-1 trauma center.
According to the initial emergency room report, the patient was alert, oriented, and moving all four extremities, with chief symptoms of pain in the left arm and chest wall. Initial trauma protocol imaging, including computed tomography (CT) of the cervical and upper thoracic spine, was read as normal. Because of the high-energy mechanism of injury, the young age of the patient, and the on-scene death of the driver, the patient was admitted to the pediatric intensive care unit for twenty-four-hour observation in full cervical spine precautions and wearing an Aspen collar. On subsequent examination three hours after injury, he could no longer move his legs. A neurosurgery consult reported a complete spinal cord injury at the T2 level and a normal upper-extremity examination. There was lower-extremity hyporeflexia, no clonus, and a negative Babinski sign. The bulbocavernosus reflex was not recorded. Magnetic resonance imaging (MRI) was ordered immediately, and it demonstrated a subtle increased T2-weighted signal within the spinal cord from T1-T3. The patient was diagnosed with thoracic SCIWORA. Upper cervical spine damage was noted in the final report by radiology but was not addressed in the chart.
The patient, with the diagnosis of SCIWORA, was transferred to a neurorehabilitation center eight days postinjury wearing an Aspen collar. A spine consultation had been obtained for clearance to participate in an active rehabilitation program. The evaluating surgeon reviewed the imaging and noted asymmetry of the dens position at C1 on CT (Figs. 2-A, 2-B, and 2-C) and bilateral C1-C2 joint dissociation on MRI (Figs. 3-A, 3-B, and 3-C). A diagnosis of unstable C1-C2 distraction injury was made. The patient was treated with C1-C2 posterior spinal fusion and instrumentation (Figs. 4-A and 4-B). Two years postoperatively, the fusion remained solid without neck pain and there was complete paraplegia at the T2 level.
The incidence of injury to the cervical spine in a pediatric trauma patient is approximately 1.5%; motor vehicle accidents are the most common cause2,3. Associated traumatic spinal cord injury tends to occur at, or near, the contiguous level of osteoligamentous disruption as a result of local translational compression of the spinal cord. Alternatively, distraction injuries involving tensile failure of the spinal column without translation represent a unique mechanism for spinal cord injury, allowing local or distant neurologic injury through a mechanism similar to the stretching of taffy from two ends (Fig. 1).
Given the mechanism of injury (a motor vehicle accident with sudden deceleration with the child in a forward-facing car seat anchored at the shoulders), we propose that this patient sustained a distraction-flexion injury without substantial translation. The spine became unhinged at the C1-C2 level, and the remaining distractive inertial energy of the head was dissipated across the length of the spinal cord, with the spinal cord injury occurring at a noncontiguous caudal level. A complete spinal cord injury at the T2 level resulted from C1-C2 osteoligamentous disruption.
The occurrence of spinal cord injury away from the osteoligamentous disruption has been well described in pediatric odontoid fractures caused by distraction injury. In a series of fifty-five patients with odontoid synchondrosis fractures, fifteen resulted in spinal cord injury; eight (53%) had caudal cervicothoracic spinal cord injury away from the osteoligamentous disruption, and only seven (47%) had local spinal cord injury at the C2 level1. The cervicothoracic junction may represent a watershed vascular zone or anchor point for force concentration because of the distribution of exiting nerves and dentate ligaments in this region4. Furthermore, the obstetric literature is replete with analogous injuries involving breech extractions, where pulling the body with the head lodged in the vaginal canal has caused distractive spinal cord injury, most commonly at the cervicothoracic junction5,6.
Although the cord injury pattern of this case follows that of a distraction-flexion odontoid fracture, the phenotype more closely approximates atlantooccipital dissociation, occurring instead at the atlantoaxial segment. The typical mechanism of atlantooccipital dissociation is a motor vehicle accident with sudden deceleration with a child restrained in a forward-facing car seat, with mortality rates of approximately 50%3,7,8. The larger head in a young child is carried forward, and the cervical spine undergoes longitudinal tension along its axis since the thorax is anchored by the shoulder harness. Early diagnosis may be difficult because of spontaneous reduction, with up to 60% of these injuries missed on initial presentation9,10. This occurs since the resting position of any injury represents only the minimal residual displacement, not the maximal transient displacement that occurs during the injury. Coexistent severe neurologic deficit should alert the provider of a more unstable osteoligamentous pattern despite the residual displacement. In atlantooccipital dissociation, late sudden death or neurologic deterioration of patients without initial neurologic injury may also occur, and it is critical to recognize and stabilize an unstable segment11,12. Ligamentous injury, in particular, has less healing potential than osseous injury, and can remain unstable both acutely and chronically9,11,12,13,14,15. Our patient with C1-C2 dissociation seems a variant of the atlantooccipital dissociation pattern, and failure to stabilize the segment could have led to sudden death or C2-ventilator-dependent tetraplegia from a secondary translational injury after a minor fall or trauma. Thus, a higher level osteoligamentous injury with a lower level spinal cord injury would be more important than the opposite, since a secondary lower level translational injury would not cause additional neurologic deterioration in a patient with a higher level complete spinal cord injury.
In our patient, the knowledge of SCIWORA likely contributed to the misdiagnosis. Pang and Wilberger originally defined SCIWORA as the presence of myelopathy after trauma without evidence of fracture or instability on radiograph or tomography16. With widespread MRI, there are few injuries without any radiographic findings, and some authors have suggested the term spinal cord injury without neuroradiographic abnormality (SCIWONA) in the modern era17. The question is not whether there are any true traumatic cases of SCIWONA, but whether many distraction injuries are incorrectly labeled as such. If it is not considered that distractive osteoligamentous injury may bring about spinal cord injury at a distant noncontiguous level, then any seemingly unexplainable spinal cord injury can be considered as SCIWONA. Even in their original description, Pang and Wilberger reported a young child who sustained a lap belt distraction-flexion injury with an L2 Chance fracture and concomitant T6 SCIWORA16. In their patient, we believe it likely that the spinal column became unhinged at L2, without substantial local translation, allowing distraction along the length of the spinal cord, leading to spinal cord injury at T6. The pattern of distraction injuries was similar in both of these cases, except the spinal cord injury in our patient occurred caudal to the osteoligamentous injury. Again, this difference is important since the risk of secondary spinal cord injury at a more cephalad level because of subsequent translational injury at the unstable segment is more critical.
Although we do not dispute SCIWORA or SCIWONA in full, these should remain diagnoses of exclusion. Multiple authors cite Leventhal’s work from 1960 that the pediatric spinal column can stretch 2 in (5.1 cm) and the spinal cord can stretch 0.25 in (0.6 cm) before disruption as evidence for the mechanism behind SCIWORA and SCIWONA6. The reproduction of these exact numbers (with use of English notation) throughout the modern literature corroborates this origin. Leventhal, however, merely reported a small case series of six neonates with cord injury after breech extraction, with one death. An autopsy was performed, and only a single sentence was devoted to the claim concerning spinal cord and spinal column elasticity. We doubt this disparate ratio. Given the number of falls children experience on a daily basis, one would expect a much higher incidence of spinal cord injury, which remains rare in young children. In addition, any effects of maternal relaxing hormones on neonatal ligamentous structures would quickly wane after birth.
In conclusion, it is important to meticulously review the osteoligamentous and spinal cord injury pattern throughout the entire pediatric spine after trauma since injury can occur at noncontiguous levels. Any diagnosis of SCIWORA or SCIWONA should be one of exclusion. It is particularly important if the osteoligamentous injury is above the spinal cord injury because the risk of a secondary osteoligamentous translation can lead to a higher level of neurologic injury. This case report stresses the importance of recognition of multilevel injury patterns in the pediatric spine and the prevention of additional neurologic deficits, incorrect treatment, or complications.