A girl who was three and a half years old was transported to the hospital emergency department by paramedics after she had presumably drowned in a children's wading pool. Resuscitative efforts were administered en route and in the hospital emergency room but were unsuccessful, and the child was pronounced dead on the same day. The police investigated the death as a suspected accidental drowning, and an autopsy was performed the following day under a coroner's warrant.
At autopsy, external examination revealed minor abrasions and contusions scattered on the face and head, but there was no external evidence of blunt force injury to the torso. On internal examination, there was acute traumatic transection of the abdominal aorta just distal to the level of the origin of the inferior mesenteric artery with massive retroperitoneal hemorrhage. This was associated with a complete fracture-dislocation of the lumbar spine through the L2-L3 intervertebral disc with a completely disrupted anterior longitudinal ligament, hematomas within the paraspinal muscles, and an epidural hematoma around the caudal aspect of the spinal cord (Fig. 1). Histological examination of soft tissue from around the vertebral fracture site showed reparative changes with fibrosis consistent with prior trauma of undetermined age. The vertebrae caudal and cephalad to the fracture site showed periosteal scarring and osteochondral callus formation indicative of a healing fracture. Retroperitoneal soft-tissue fibrosis with remote hemorrhage was also demonstrated. There was no evidence of prior injury at the end plates at the L2-L3 level or of the surrounding ligamentous structures to indicate a previously destabilized spine. Histological examination of specimens from the aorta revealed normal aortic architecture with no evidence of a connective-tissue disorder, no evidence of prior intimal or medial mural injury, and no other inherent aortic disease. Radiographic findings verified the complete fracture-dislocation at the L2-L3 level (Fig. 2). The fracture-dislocation had occurred through the superior end plate of L3 and extended posteriorly with complete circumferential ligamentous disruption through the posterior elements (the anterior longitudinal ligament, posterior longitudinal ligament, facet joints, capsules, and interlaminar space) (Fig. 3).
The remainder of the autopsy examination was noncontributory, as was postmortem toxicological testing. In addition, there was no gross or histological evidence of asphyxiation or drowning. The cause of death was considered to be retroperitoneal hemorrhage due to traumatic rupture of the abdominal aorta associated with a fracture-dislocation of the lumbar spine.
After a review of the above findings, the professional consensus was that the mechanism of injury was a purposeful forced hyperextension of the spine over a fulcrum at an accelerated rate of load application with a possible shear component. The absence of posterior bruising was thought to be related to rapid exsanguination or, less likely, to a constant rate of pressure application. As a result of the autopsy findings, a diagnosis of child abuse was made, charges were filed against the caregiver, and the case was brought to trial. The prosecution was successful in obtaining a conviction for second-degree murder.
Traumatic disruption of the thoracic aorta is a well-described pattern of injury that usually occurs secondary to high-energy mechanisms such as motor-vehicle collisions4,5. Blunt injury to the abdominal aorta is an infrequent occurrence. Of all traumatic injuries of the aorta, only 4% to 8% occur below the diaphragm6-10, and these injuries are exceptionally rare in children11-13. Two case reports of abdominal aortic injuries in children describe a twelve-year-old boy who had an accidental fall and a four-year-old who was allegedly kicked in the abdomen14,15. Neither had a spine fracture, and both were presumed to have sustained the aortic injury as the result of a direct force on the aorta. The mechanism of aortic disruption at the abdominal level in adults and children is not well understood. Several researchers9,16 have hypothesized that laceration or rupture is a result of direct force on the relatively fixed abdominal aorta, with compression of the vessel between the vertebral column and an external nonyielding object. Warrian et al.17 proposed that the aortic disruption occurs as a result of differential deceleration, with the abdominal aorta being fixed by lumbar segmental arteries to the vertebral column18,19. It has also been proposed that the injury is caused by acute dissection (stretching), wherein the forward momentum of the aortic blood column acutely fills the aorta and stretches it cranially while the shift of the abdominal contents into the pelvis stretches the aorta beyond its elastic limit20.
In the case that we presented, the suspected mechanism of injury was forced hyperextension of the child over a fulcrum such as the knee or a banister at an accelerated rate of load application. We acknowledge that the proposed mechanism of injury cannot be proven as the force required to generate such an injury in a child is not well defined. However, the force required to sustain this fatal aortic and spinal injury is clearly of such magnitude that it could not have happened as a result of an accidental fall in or around a child's wading pool.
In the analysis of this case, a flexion-distraction injury was ruled out by virtue of the complete disruption of the anterior longitudinal ligament. The typical hyperflexion spine fracture-dislocation, or Chance fracture, requires a substantial force with a fixed pelvis, such as occurs in passengers wearing a lap seat belt during a motor-vehicle collision. With the typical Chance fracture, the anterior longitudinal ligament is intact and is stripped away from the bone. The combination of aortic injury and a Chance fracture is rare, with the reported cases occurring almost exclusively in adults and with the majority being seat belt injuries. In one large series of 194 aortic injuries, only six associated spine fractures were found, and only one of the six was in a child (an eight-year-old who was crushed by a farm vehicle)4. In addition, we are aware of only one report of a child with an aortic injury combined with an L2-L3 Chance fracture, a ten-year-old boy involved in a motor-vehicle collision when he was belted in the backseat21. To our knowledge, all of the other cases reported in the literature involved adults. In all of the above cases, when the aorta ruptured, the mechanism was a combination of direct and indirect forces. The direct forces involved compression or tethering of the aorta to the spine, and the indirect forces involved a rapid increase in intraluminal pressure and the resultant tensile forces, all of which had to be substantial. It is our view that, if a flexion-distraction injury had been the mechanism of injury in our patient, then the anterior longitudinal ligament would most likely have been intact and would have been only stripped from the anterior surface of the vertebral bodies. We also believe that the aorta would have been able to withstand the flexion-distraction force as it would have followed the path of the anterior longitudinal ligament, despite the autopsy findings of periosteal scarring from previous injuries. In view of the evidence of previous injuries and surrounding fibrosis, the aorta was likely tethered to the anterior longitudinal ligament and/or the bone and, as such, a hyperextension injury caused it to be stretched beyond its tensile capacity over a short segment (Fig. 4).
Hyperextension spinal fracture-dislocations require the application of a substantial force over a fulcrum. In the typical injury, the anterior longitudinal ligament is ruptured with substantial displacement. To our knowledge, only one such case has been reported: in a sixteen-month-old child who had an L3 fracture-dislocation and aortic rupture, presumably due to a kick in the back associated with child abuse3.
In the case that we presented, the injury essentially caused exsanguination into the retroperitoneal space, and death was rapid. It is important to note that the pediatric lumbar spine typically is flexible, compliant, and resilient to injury with day-to-day activities and/or recreational falls. After an extensive review of the literature, we found no existing benchmarks, biomechanical guidelines, or range-of-motion limits with which to gauge or estimate the extent of force required to sustain this type of injury. Nonetheless, the literature shows that sustaining a fracture-dislocation of this severity in a mature spine would require a pedestrian-motor-vehicle collision with an impact velocity of approximately 70 to 100 km/hr15.