Neuropathology of Brain Trauma in Infants and Children

L. J. Dragovic

This chapter focuses on the pathology of brain trauma in infants and children, with particular emphasis on the concepts and controversies of child abuse. Like other deviant expressions of human behavior, child abuse has been with mankind from the beginning. It is a worldwide societal phenomenon representing a broad spectrum of aberrant behavior that results in physical and/or emotional injury of a child. Although the abuse traditionally and commonly takes place in a domestic environment, occurrences in institutional settings have grown more prevalent recently as a result of urban lifestyles and increased family dependence on day care.

EPIDEMIOLOGY

A survey of traumatic brain injury in the United States between 1995 and 2001 revealed 2, 685 average annual deaths for children 0 to 14 years of age. Rates per 100, 000 population were 5.7 for ages 0 to 4 years, 3.1 for 5 to 9 years, and 4.8 for 10 to 14 years. The large majority of head injuries in children are accidental, as shown in Table 16.1, but a substantial proportion are inflicted.¹ Head trauma is the major cause of death in cases of child abuse², ³, ⁴ and is the primary cause of death in more than 80% of cases in some series.⁵

HISTORICAL PERSPECTIVE

Autopsy diagnosis of child abuse was first formulated by Tardieu in the 19th century⁶; the topic was brought into clinical focus by publications of Guthkelch, ⁷ Caffey, ⁸, ⁹ and Kempe¹⁰, ¹¹ during the latter part of the 20th century, with the emergence of the clinical terms “battered child” or “syndrome of child abuse and neglect.” Subsequent attempts to clinically define subsets of the syndrome of child abuse and neglect resulted in the term “whiplash-shaken infant syndrome” in the early 1970s.⁷, ⁸, ⁹ This term encompassed the combination of subdural hemorrhages, retinal hemorrhages, and metaphyseal fractures and allegedly was caused by the violent shaking of the infant without impact to the head. Subsequently, with the realization that physical impact was essential for the craniocerebral injury to take place, the term “shaken impact syndrome” was proposed in the 1980s.¹², ¹³, ¹⁴ These clinical concepts have been ushered into this century under euphemisms for inflicted head injury, such as “abusive head trauma”¹⁵, ¹⁶ and “nonaccidental head injury.” These terms, frequently used in courtrooms, are ambiguous, and their use should be discouraged. The terms “inflicted head injury”¹⁷, ¹⁸ and “inflicted brain injury”⁵ are clearer and are preferable.

CONTINUING DIAGNOSTIC CONTROVERSIES

Nowhere in the realm of medicolegal considerations are there such controversies and confusion as in the assessment and interpretation of central nervous system trauma of infancy and childhood.¹⁹, ²⁰, ²¹, ²² These controversies arise from several sources. First, there is a strong societal intent to prevent and punish child abuse. This intent, however, at times may become misguided and has the potential to blur or bias the conclusions of the forensic pathologist. Second, there is the concept of “shaken baby syndrome.” This syndrome, which includes subdural hemorrhage, retinal hemorrhages, and encephalopathy, ⁸, ⁹, ²³ implies traumatic brain injury without head impact.⁷, ²⁴ Furthermore, it has been postulated that the rotational forces generated by the shaking tear bridging vessels, causing subdural hemorrhages, ⁷ and that encephalopathy is secondary to diffuse axonal injury. All of these concepts pertaining to shaken baby syndrome, widely accepted for three decades, are now under serious scrutiny.¹⁷, ¹⁸, ²¹, ²⁵, ²⁶, ²⁷, ²⁸

Nonimpact Versus Impact

The clinical concept of shaken baby syndrome was promoted through published observations of Guthkelch⁷ and Caffey⁸, ⁹ as a useful operational diagnostic paradigm. The characteristic findings diagnosable by the available imaging modalities of the times included retinal, subdural, and/or subarachnoid hemorrhage. The clinically observed and defined craniocerebral trauma allegedly resulted from the whiplash effect of the infant’s relatively large and heavy head, weak supporting neck musculature, incompletely ossified skull, a relatively large subarachnoid space, and an immature brain subjected to acceleration/deceleration traction forces by a rapid whipping back-and-forth motion. The concept, albeit not supported by an adequate objective postmortem evaluation, has grown into a major misconception among professionals in clinical medicine, with a rather widespread notion of its absolute prevalence as the most important form of brain injury in physically abused infants and small children. The cumulative published clinical and pathologic research over the past several decades¹², ²⁹, ³⁰, ³¹ has identified blunt impact to the head as an undeniable autopsy finding in the great majority of fatal head injuries in infants, and
the concept of shaken baby syndrome was modified into shaken impact syndrome. The proponents of the revised concept offer a process during which an infant is presumably shaken, which results in retinal and subdural hemorrhages, and subsequently, the infant’s head is subjected to impact against some hard surface to account for contusions, fractures, additional hemorrhage, and varied degrees of brain swelling.

TABLE 16.1 Average Annual Numbers, Rates, and Percentages of Traumatic Brain Injury-Related Deaths by Age Groups, United States, 1995-2001¹

	Motor Vehicle Accidents			Falls			Assault			Other/Unknown		Total
Age (years)	No.	Rate^*	%	No.	Rate^*	%	No.	Rate^*	%	No.	%	No.	Rate^*
0-4	499	2.6	45.4	37	0.2	3.4	348	1.8	31.7	215	19.6	1, 099	5.7
5-9	441	2.2	70.2	16	0.1	2.5	64	0.3	10.2	107	17	628	3.1
10-14	541	2.7	56.5	17	0.1	1.8	119	0.6	12.4	280	29.3	957	4.8
^*Average annual rate per 100, 000 population.

In a review of the literature that included 324 autopsies of traumatic brain injury in children, and in which in 54 cases there was an admission of shaking, only 11 did not have objective signs of impact.³² Furthermore, few cases of shaking injuries have been independently documented.³³ Biomechanical studies also cast doubt that shaking by itself can cause significant brain injury. Several biomechanical studies have revealed that the forces generated by shaking are below the threshold necessary to cause subdural hemorrhage and parenchymal brain injury.¹², ³⁴, ³⁵ Because scalp and skull injuries, the mark of a head impact, may be missed in a clinical setting, the forensic pathologist, with the advantage of careful dissection of the scalp and direct observation of the bones of the skull, is in a better position to detect evidence of impact(s) and frequently does so. Drawing from our experience and from the literature, we agree with the opinion of Duhaime et al.¹² that “the shaken baby syndrome, at least in its most severe acute form, is not usually caused by shaking alone. Although shaking may, in fact, be part of the process, it is more likely that such infants suffer blunt impact. The most common scenario may be a child who is shaken and then thrown into a crib or other surface, striking the back of the head.”

Diffuse Axonal Injury as Mechanism for Encephalopathy

The frequently associated finding of shearing type of damage of the white matter in head trauma³⁶, ³⁷ has been incorporated into the acceleration/deceleration mechanism, even though impact has been identified at autopsy. The histopathology of the white matter lesions in traumatic brain injury was championed by Adams et al.³⁸, ³⁹ and Graham et al., ⁴⁰ characterized by widespread abnormal morphology of the involved axons and referred to as diffuse axonal injury. Even though animal models of diffuse axonal injury from Duhaime et al, ¹² Gennarelli, ⁴¹ and Shaver et al.⁴² clearly demonstrated that this type of mechanical damage of the white matter could not be generated in an infant’s brain without an impact, the entrenched “shaken baby” terminology still dominates courtrooms throughout the continent.

Notably, most of the current terminology is misleading. Strictly speaking, diffuse axonal injury is far from being actually diffuse, but is rather multifocal, widespread in some cases, and characterized by predilection for occurrence in certain white matter structures that are subjected to physical force. This results in shearing disruption of the axons and the adjacent small blood vessels that are inseparable from the axons, as described by researchers at the end of the 19th century and subsequently detailed by Holbourn, ⁴³, ⁴⁴ Strich, ³⁶, ³⁷ and Lindenberg and Freytag.⁴⁵ The widely held tenet that diffuse axonal injury is a critical finding in infantile traumatic brain injury and is responsible for its encephalopathy has been challenged by the observations of Geddes et al.¹⁷, ¹⁸ In a study of a large series of cases of presumably inflicted brain injury, these authors found that only a minority had axonal injury, in most instances confined to specific brainstem regions. Moreover, most of the brains displayed signs of hypoxic-ischemic injury. These observations have been supported by neuroimaging studies.⁴⁶

Geddes⁴⁷ recommends that the term diffuse axonal injury be used only for diffuse traumatic brain injury so that it can be distinguished from other mechanisms of damage, such as widespread (or diffuse) hypoglycemic axonal injury, widespread (or diffuse) vascular injury, or widespread hypoxic axonal injury. Other authors⁴⁸ suggest that physical trauma may not be the principal underpinning of diffuse axonal injury in the first place.

An additional point of confusion is that to many examiners, the observation of axonal injury in a child is synonymous with traumatic origin and, furthermore, with shaken baby syndrome. This notion is incorrect because it does not take into consideration that axons can also be injured by nonmechanical means, specifically by hypoxia and ischemia. Moreover, despite numerous microscopic studies, ³³, ⁴⁹, ⁵⁰, ⁵¹, ⁵² the precise cause of axonal damage demonstrable by b-amyloid precursor protein or neurofilament immunostains⁵³ has not been established and does not discriminate among the different causes of axonal damage.

Origin of Subdural Hemorrhage

The proposal of Guthkelch⁷ that the rotational forces of shaking cause tears of cerebral bridging veins and subdural hemorrhage was based on the work of Ommaya et al.²⁴ on Rhesus monkeys. These animals were subjected to whiplash forces that caused concussion, subdural hemorrhage, and diffuse axonal injury in 18 of 50 animals. The force of the whiplash was such that 11 of the 18 monkeys suffered neck injuries.²⁴ According to Ommaya et al., this force was equivalent to a motor vehicle crash at 30 mph, clearly above the force achieved by manual shaking of an infant.²⁸ More recent studies have questioned the rotational forces as the exclusive cause of subdural hemorrhages and proposed a combination of hypoxia and abnormal hemodynamic situations, such as increases in venous and arterial pressures.¹⁸

Retinal Hemorrhages

Retinal hemorrhages are common (65% to 95%) in cases of inflicted head injury in infants.²⁹, ⁵⁴ These hemorrhages can be
unilateral or bilateral and associated with retinal folds or detachments. The proposed mechanisms of traumatic retinal hemorrhages include tears of retinal vessels by angular deceleration, extravasation of subarachnoid blood, and increased retinal venous pressure.⁵⁵, ⁵⁶ Importantly, retinal hemorrhages are not specific for inflicted head trauma, as they can be present in documented accidental head injuries.⁵⁷

It should also be noted that retinal hemorrhages are commonly seen in vaginally delivered newborns, but resolve within a few weeks.⁵⁸ Retinal hemorrhages can also result from nontraumatic causes, such as sepsis, coagulopathies, hypertension, and subarachnoid hemorrhage.⁵⁹, ⁶⁰, ⁶¹ As concluded by Duhaime et al., ²⁹ “the diagnosis of inflicted head injury cannot rest on the finding of retinal hemorrhage alone, but the finding of severe bilateral retinal hemorrhage with retinal folds or detachments is particularly suggestive of the diagnosis.” (For a detailed discussion of retinal hemorrhages, see Chapter 17).

CONSIDERATIONS OF ANATOMY AND PHYSIOLOGY OF THE DEVELOPING CENTRAL NERVOUS SYSTEM

With opportunity for physical abuse to occur anytime during the lifespan, a consideration must be given to the peculiarities of anatomy and physiology that help define important differences for diagnosis of abuse of newborns, infants, and toddlers, as compared with adults. Pediatricians constantly remind other medical professionals, and society at large, that infants and children are not small adults, even though their growth and development is directed to adulthood. Furthermore, remarkable developmental differences exist among newborns, infants, toddlers, and adults in the central nervous system and its protective skeleton to account for significantly different reactions of the tissues when subjected to mechanical forces.

Several factors contribute to the increased vulnerability of the brain of infants and children to trauma. Throughout childhood, head size is a dominant physical feature of the child’s body, and the proportional size of head to body remains relatively large for the first 5 to 7 years of life. This feature and the relative weakness of neck muscles result in a greater vulnerability of the brain to traumatic forces. The high water content of the infant and toddler brain and limited myelination of the structures deemed to become the white matter add to the heightened vulnerability to mechanical forces. The pliability of the young skull bones when a stationary head is subjected to blunt force results in greater likelihood of direct trauma to the cortex immediately underneath.

Whereas the developing brain is snuggled in the developing skull, the bones of which form by membranous ossification, the growing dura mater is the essential inner lining of the skull and adheres tightly to it. The dura covers the surface of the brain, but does not adhere to it, and allows for limited sliding of its inner surface over the arachnoid membrane. The cranial and spinal dura mater are continuous with each other at the foramen magnum. Also, the cranial dura is continuous with the optic nerves’ sheaths. Note that the subdural space is practically nonexistent, a virtual space, with no fluid present under normal anatomical circumstances; it is referred to as the subdural space only in situations of abnormal fluid accumulation. From both sides of the brain’s parasagittal surfaces along the falx, eight to 12 small-caliber bridging veins (approximately 0.8 mm) drain into and regulate the pressure of the superior sagittal sinus. This is the major venous blood-collecting structure of the brain, and its walls are formed by foldings of the dura mater. The bridging veins entering the dural sinus are thin vessels, with delicate walls with remarkable elasticity that allow for stretching three to four times their actual length (maximum 3 mm) without rupture or tearing. This remarkable physical feature enables rapid change of the direction of the fast-moving head, without detrimental structural or functional consequences. These vessels open into the sinus almost at a right angle in a developing brain, although the veins located posteriorly become oriented obliquely forward against the current in the sinus because of the posterior growth of the cerebral hemispheres. Nonetheless, impact of the fast-moving head, face (chin or cheeks), back of the head, or buttocks against any type of unyielding surface (hard or padded) or excessive squeezing or pulling up of the skull during birth may result in disruption of several of these delicate vessels. The unilateral or bilateral tear of these veins results in bleeding into the otherwise virtual space beneath the dura. (Further detailed discussion of biomechanical principles of head trauma is available in Chapter 5.) These accumulations are often less dramatic in infants and small children than in adults and do not result in space-occupying effect; however, the detrimental effect of the force on the young brain results in functional and structural damage, with serious consequences and likely eventual deadly outcome.

Normal brain function depends on adequate blood flow, metabolic activity, and a balance of brain fluids. The metabolic rate of particular central nervous system tissue structures is significantly different in a developing brain compared with an adult brain. In adults, the gray matter has approximately four times more active metabolism than the white matter. In infancy and early childhood, brain growth occurs largely as a result of development of the white matter, which is the result of intense myelination of the axons. The intense buildup of myelin increases the metabolic activity in the white matter, which also has a higher water content. Hence, the reaction to blunt injury of a developing brain is commonly observed as diffuse swelling, with collapse of the ventricular system and paucity of identifiable cortical damage. The only tell-tale sign of blunt trauma to the cortex is the presence of the localized subarachnoid hemorrhage in cases of very short survival before death.

Normal blood flow through the central nervous system of the adult averages 50 to 55 mL/100 g of tissue per minute, which requires approximately 750 mL/min, or 15% of the total resting cardiac output. The situation is different in an infant or small child, in whom a much larger percentage of the cardiac output (which is overall much smaller compared with that of an adult, owing to the smaller size of the growing heart) is earmarked to supply the central nervous system.

Total cessation of blood flow to the brain causes unconsciousness within 5 to 10 seconds. Cerebral blood flow is influenced by hypoxia, hypercapnia, changes in blood pH, and hypoglycemia, as well as increased intracranial pressure and brain swelling resulting from any acute trauma to the brain (see Chapters 7 and 8). Brain tissue reacts to acute disturbance of physical, chemical, or biologic nature by swelling and to chronic disturbances by atrophy, gliosis, and compensatory accumulation of cerebrospinal fluid.

Only gold members can continue reading. Log In or Register to continue