David Dolinak, MD
R. Ross Reichard, MD
Understanding and properly documenting and interpreting pediatric traumatic brain injury, or TBI, are challenging tasks. Evaluating inflicted traumatic brain injury in infants and young children can be particularly challenging for multiple reasons. First, pediatric TBI may result from varied mechanisms of injury, including a moving head striking a fixed object, an object striking a movable head, or rapid acceleration/deceleration without an identifiable impact site, known as shaking. Second, the history of how the injury occurred may be inaccurate. The caregiver may not recall the event accurately or may provide misleading information intentionally. The lack of a clear sequence of events in most cases of inflicted TBI can make it difficult to correlate autopsy findings with clinical symptoms. Investigative reports and autopsy findings are often integrated at trial to corroborate or refute the statements of witnesses and suspects. Unfortunately, autopsies do not always reveal pathology that is pathognomonic of specific types of injuries. In recent years, β-amyloid precursor protein immunostaining has been used as an adjunct tool to help clarify neuropathological findings.
Suspicious, unexpected, or otherwise unexplained deaths in infants and young children typically fall under medical examiner/ coroner jurisdiction. In young children, the external findings associated with traumatic injury may be subtle or not even visible externally. Thus, even cases of suspected sudden infant death syndrome should be autopsied and are best handled by a forensic pathologist. Autopsy may reveal scalp contusions and even skull fractures (clear evidence of impact injury) that were not appreciated clinically or expected based on external examination of the body. When injury to the scalp, skull, or both is present in conjunction with evidence of intracranial trauma, the cause of death is often attributed to “blunt force head injury” or similar wording. In this type of case, the manner of death is determined by integrating information from investigative reports detailing the known or likely circumstances of how the injury occurred with autopsy findings, review of the child’s medical history, and other case information.
The intracranial pathology in pediatric TBI often differs from that in adults. In both populations, subdural hemorrhage commonly arises from torn bridging veins. In adults, subdural hemorrhage commonly results in mass effect, but in infants and young children, it only infrequently results in mass effect. Therefore, in the appropriate context, in infants and young children, only a small amount of subdural hemorrhage is regarded as a marker of severe TBI, and is not considered to be the primary underlying pathology. Torn bridging veins may become visibly thrombosed when there is extended post-injury survival and may lead to venous infarction of the underlying parenchyma. Children maintained on a ventilator may have poor cerebral circulation, which may result in stasis of blood within cerebral vessels that can mimic thromboses. Careful gross and histological evaluation of the blood vessels and corresponding brain parenchyma typically makes it possible to distinguish the two conditions. Subtle lacerations of the cerebral cortex at the gray-white matter junction are more common in young children than the intraparenchymal hemorrhages and typical cortical contusions seen in adults. Knowledge of the neuropathological features of pediatric TBI is crucial if the findings are to be interpreted properly.
Hypoxic-ischemic brain injury and brain swelling are common findings in cases of inflicted TBI. These findings may be due to an apneic episode, hypotension, dysrhythmia, seizure, or some combination of such deleterious events precipitated by TBI. Also, trauma may cause vasogenic or cellular edema, or both, which contribute to brain swelling. Trauma may also disrupt cerebral vascular autoregulation. If this happens, systemic hypotension may result in cerebral vascular hypotension and hypoxic-ischemic brain injury. These complex pathophysiological changes associated with TBI often result in similar, or overlapping, pathologic findings at autopsy.
Historically, the term “shaken baby syndrome,” or SBS, was coined to describe the mechanism of injury that resulted in subdural, subarachnoid, and retinal hemorrhages in babies that often had metaphyseal fractures. The hypothesis is that shaking causes forces to be transmitted through the torso, which ultimately result in the head snapping back and forth, causing the brain to rotate within the confines of the skull, tearing bridging veins and resulting in subdural and subarachnoid hemorrhage. The rotational injury is also thought to damage axons throughout the brain, resulting in diffuse traumatic axonal injury. This mechanism of injury is also thought to cause retinal and optic nerve hemorrhages. In various jurisdictions the cause of death in these types of cases has been shaken baby syndrome.
Research and experience, however, have called into question the precise mechanism of inflicted TBI in cases that have been described as SBS. Some forensic pathologists consider the current medical literature sufficient evidence to support the concept of SBS, while others disagree. Skeptics of SBS argue that all such cases have head impact(s), whether or not the impact injury can be demonstrated at autopsy. They argue that even a significant head impact may not cause a head contusion or skull fracture, because of, among other reasons, open cranial sutures or the elastic nature of youthful tissue. Skeptics also point out that the hemorrhages described in SBS can result from impact injury alone and therefore are not specifically diagnostic of a shaking episode. In fact, not all of the hemorrhages are necessarily diagnostic of physical injury. Retinal, optic nerve, and subdural and subarachnoid hemorrhage can all be seen in babies who are severely coagulopathic from sepsis or a variety of other underlying disorders.
Supporters of the concept of SBS purport that shaking alone can generate enough force to cause the described hemorrhages and fatal TBI. Proponents of SBS point to the findings of retinal and optic nerve hemorrhages, subdural and subarachnoid hemorrhage, and diffuse traumatic axonal injury as characteristic features of this entity. Additional injuries that may support a diagnosis of severe shaking may include thoracic contusions, posterior rib fractures, and classic metaphyseal fractures of the extremities. The term “shaken impact syndrome” has been used when there is also evidence of head impact. The concept is that the assault consists of a combination of shaking and impact(s) of the head, typically of the head against a fixed object.
Members of the medical profession clearly disagree on the diagnostic implications of particular findings, their relevance in determining the precise mechanism of an injury, and whether an infant was shaken. Even after careful review of all case information, aside from stating there is impact injury of the head, it is usually not possible to state precisely how a particular injury was sustained.
In cases with clear evidence of impact of the head, such as scalp contusions and skull fractures, attributing the cause of death to blunt force injury is not difficult. In a small number of cases there are no contusions or fractures, but there is subdural and/or subarachnoid hemorrhage, brain swelling, and optic nerve sheath hemorrhage—findings that although characteristic of trauma (accidental or inflicted), may also be seen, albeit rarely, in natural disease processes. It is these types of cases that are most controversial. A pathological marker that definitively distinguishes traumatic injury from nontraumatic injury would help markedly in determining the cause of death.
The advent of β-amyloid precursor protein, or β-APP, immunostaining, a sensitive marker of axonal injury, has furthered our appreciation of axonal injury but fails to provide all of the answers. β-APP is present in many cell types, and immunostaining may detect its expression in different cells and structures, including glia, dorsal root ganglion cells, and leptomeninges. β-APP is a much more sensitive means of detecting injured axons than H&E or silver stains, and it highlights axons earlier in their injury process (in as little as two hours or so). β-APP is transported by rapid anterograde transport, and when normal axonal flow is disrupted, it accumulates at the site of axonal injury and thus becomes detectable on immunostaining. However, β-APP accumulates anywhere there is axonal injury, regardless of the etiology, be it from, for example, trauma, infarction, multiple sclerosis, or metabolic abnormalities (hypoglycemia). The sensitivity of β-APP makes it possible to detect axonal injury before changes are identifiable by routine H&E staining and thus may demonstrate “zig-zag” areas of early infarction (vascular axonal injury). β-APP immunostain use markedly increases the sensitivity for detecting axonal injury, but also makes interpreting the findings more complex.
Children generally react to TBI differently than adults, and infants appear to react to TBI differently than older children. Children older than one year of age who have sustained severe TBI tend to have diffuse traumatic axonal injury (of the corpus callosum, cerebral hemispheres, and brain stem). In contrast, children less than one year of age with severe TBI tend to have traumatic axonal injury of the brain stem limited to the corticospinal tracts, along with supratentorial vascular axonal injury. Differentiating traumatic from vascular axonal injury and its distribution has helped in our understanding the pathophysiology of inflicted TBI.
When diffuse traumatic axonal injury or brain stem traumatic axonal injury, or both, can be demonstrated, it can provide not only useful information about the cause of death but also cautious insight into the clinical symptoms the child may have experienced. These findings can then be evaluated in light of the provided history. That is, comments such as “the child became limp,” “his eyes rolled back in the head,” or “the child was breathing funny” may be supported or refuted based on the degree and distribution of traumatic axonal injury. Each piece of information can then be woven together to understand the case to the best of our ability.
Even when there is clear autopsy evidence of TBI, β-APP immunostains may not detect any axonal injury. In some cases of fatal TBI there is no detectable diffuse traumatic axonal injury, but the injury may prove fatal because of brain swelling or global hypoxic-ischemic injury. In this situation, only vascular axonal injury may be detected. In other cases, there may be diffuse traumatic axonal injury, but the survival interval may be insufficient to allow for the detection of β-APP within injured axons. Production and transport of β-APP is an active process; thus, if there is no or only limited cerebral perfusion, even with more than two hours of post-injury survival time, axonal injury may be present but not detectable by β-APP immunostain.
β-APP immunostains have also been used to evaluate the spinal cord and optic nerves. Cervical spinal nerve roots may have axonal swellings/bulbs highlighted by β-APP immunostains. We have detected traumatic spinal nerve root axonal injury at all levels of the cord. However, our experience has been that rare cases with marked brain swelling, in the absence of TBI (for example, near-drowning deaths), may have cervical nerve root axonal injury. Thus, we cautiously interpret axonal injury of the cervical nerve roots. Our experience with β-APP immunostaining of optic nerves indicates that it can detect traumatic axonal injury; however, the sensitivity is low, involving only about one-third of cases. Our knowledge of β-APP immunostain results on the spinal cord and optic nerves is still evolving.
As with any injury, with survival time, there is ongoing tissue reaction, and thus axonal swellings/bulbs are only transiently visible histologically. After approximately one-month survival time, most axonal swellings/bulbs are no longer visible. At this point, all that may be seen histologically are reactive astrocytes, macrophages, or both located within the white matter in the distribution of diffuse traumatic axonal injury. Those who survive head injury for months to years and are severely disabled or in a persistent vegetative state often have gross sequelae of diffuse traumatic axonal injury. These findings may include a thin, possibly torn corpus callosum, focal cavitary lesions within the corpus callosum or dorsal rostral quadrants of the brain stem, and marked loss of white matter with concomitant hydrocephalus ex-vacuo.
Autopsies of abused infants and young children who die of inflicted TBI often are complex and involve both traumatic and hypoxic-ischemic brain injury. The forensic pathologist must carefully integrate investigative information, medical records, radiologic findings, and gross and microscopic findings; interpret β-APP immunostains; and evaluate metabolic, vitreous electrolyte, and toxicology studies to understand the entirety of the case. β-APP immunostains are now a useful diagnostic tool in the neuropathological evaluation of inflicted TBI. β-APP helps determine whether there is traumatic, vascular, or metabolic axonal injury, or all three. As is true for the use of any immunostain, a correct interpretation requires knowledge of the antibody’s strengths and weaknesses. The β-APP immunostain results are but one part of the complete case investigation and evaluation.
Dr. Dolinak, a member of the CAP Neuropathology Committee, is a forensic pathologist with the Cuyahoga County Coroner’s Office, Cleveland. Dr. Reichard is with the Office of the Medical Investigator, University of New Mexico, Albuquerque. Their full discussion of β-amyloid precursor protein immunohistochemistry will be published in the May issue of Archives of Pathology & Laboratory Medicine.