Trauma

Published on 27/02/2015 by admin

Filed under Pediatrics

Last modified 22/04/2025

Print this page

rate 1 star rate 2 star rate 3 star rate 4 star rate 5 star
Your rating: none, Average: 0 (0 votes)

This article have been viewed 1767 times

Chapter 39

Trauma

Incidence and Etiology

Head trauma and resulting traumatic brain injury (TBI) are the most common causes of morbidity and mortality in children. Head trauma alone, or in combination with injuries to other organs, is responsible for 50% of deaths in children from the ages of 1 to 14 years. Approximately 475,000 cases of TBI occur each year among infants, children, and adolescents younger than 14 years in the United States. Half of the cases occur in children younger than 5 years. Infants and children younger than 4 years and adolescents between 15 and 19 years of age are the two pediatric age groups at highest risk for TBI. Across the age groups, boys are more frequently injured than are girls (2 : 1). Motor vehicle accidents, especially car-versus-pedestrian accidents, are the most common causative events; they are responsible for approximately 30% of TBI cases.

In infants younger than 2 years, the most common causes of head trauma are falls, collision of the head with an object, the infant being dropped, and nonaccidental trauma (NAT). NAT accounts for more than 80% of deaths from head trauma in this group. Skull fractures, subdural hematomas, cerebral edema, and parenchymal contusions are common injuries before the age of 2 years. In children between 3 and 14 years of age, falls, sporting activities, all-terrain vehicle and motor vehicle accidents (when the child is a passenger, pedestrian, or cyclist), and assaults are the most common causes of head trauma. Fractures and parenchymal contusions are the most common injuries in this group. In older adolescents (>14 years), diffuse axonal injury (DAI) related to motor vehicle accidents predominates.

Birth trauma related to cephalopelvic disproportion, large gestational weight, atypical presentation, and the use of forceps or vacuum extraction is discussed in Chapter 23.

Imaging in Pediatric Head Trauma

Computed Tomography

In general, computed tomography (CT) is the initial imaging modality used to evaluate patients with known or suspected head trauma to detect potentially life-threatening conditions that may require immediate surgical intervention. CT is widely available, easy to obtain, and can be performed quickly. It can be used with all manner of life-support and monitoring equipment. The advantages of multidetector CT scanners are a shorter scanning time and the ability to rapidly provide multiplanar and three-dimensional (3D) images. These advances have increased the sensitivity of CT in detecting the sequelae of head trauma, particularly skull base and temporal bone fractures. CT angiography (CTA) is useful in the evaluation of traumatic vascular injury and has the relative advantage of speed and fewer flow-related artifacts compared with magnetic resonance angiography (MRA).

Limitations of CT include low contrast in the immature brain, which decreases its sensitivity in detecting edema in infants, and beam hardening artifact, which may partially obscure small extraaxial collection or subtle cortical contusions adjacent to bone, particularly in the posterior fossa and skull base. Although advances in CT scanners have improved their sensitivity in detecting TBI, CT is known to underestimate the degree and extent of traumatic parenchymal injuries when compared with MR imaging (MRI). Nonetheless, acute imaging findings on CT have been proposed as criteria for grading and predicting outcome in persons with TBI (the Marshall Classification and subsequent modifications by Maas et al.).1

CT has been used liberally in children with scalp contusions. Recently, the exposure of children’s brains to excessive radiation has become a concern after Pearce et al.2 demonstrated that children exposed to radiation to the head from CT scans have an increased risk of the development of brain tumors many years later. For patients with a minor head injury (Glasgow Coma Scale [GCS] score of 13 to 15), clinical guidelines are available regarding indications for obtaining a CT scan of the head, such as the New Orleans Criteria, the Canadian CT Head Rule, and guidelines from the Pediatric Emergency Care Applied Research Network, with high sensitivity for detecting injuries that require neurosurgical intervention.3 The American College of Radiology also publishes Appropriateness Criteria for imaging in the setting of head trauma, including a special section for children younger than 2 years in whom clinical assessment and criteria are less reliable than for older children.4

Adaptive statistical iterative reconstruction is a reconstruction technique that reduces image noise and improves low-contrast detectability and image quality. It provides higher diagnostic performance at a lower dose with no loss of image sharpness, noise, and artifacts.5

Magnetic Resonance Imaging

Because of its more limited availability and longer scanning times, MRI generally is a secondary modality in the evaluation of acute head trauma. MRI usually is performed in the subacute and chronic phases when the findings on CT do not correlate with the patient’s clinical condition, or when patients have unexpected neurologic deterioration or are not responding as expected. MRI offers the advantages of direct multiplanar imaging and greater overall sensitivity and specificity for the detection of TBI. It is particularly useful in the detection of small extraaxial hematomas, nonhemorrhagic intraaxial contusions, edema, brainstem injury, posterior fossa abnormalities, and DAI.6 Contraindications to MRI include incompatible vascular clips, metallic implants, ocular foreign bodies, and most pacemakers.

Specific MRI sequences are invaluable in the evaluation of TBI. Fluid attenuation inversion recovery sequence (FLAIR) uses an inversion recovery pulse with a long inversion time that nulls signal from cerebrospinal fluid (CSF). FLAIR is sensitive for delineation of foci of signal abnormality adjacent to the ventricles and subarachnoid space.

T2* gradient-echo sequence (GRE) uses a pair of bipolar gradient pulses instead of a refocusing 180-degree pulse to enhance magnetic susceptibility caused by magnetic field distortion. Susceptibility-weighted imaging (SWI) is a technique that uses the signal loss from out-of-phase protons with different magnetic susceptibilities. SWI is more sensitive than GRE for detection of blood products such as deoxyhemoglobin, methemoglobin, and hemosiderin.7

Diffusion-weighted imaging (DWI) assesses the diffusion of water protons with use of a bipolar gradient pulse. The normal diffusion of water protons along the gradient reduces the MR signal. Areas of restricted diffusion will retain the MR signal and represent acute cerebral injury in persons with ischemia/hypoxia, trauma, metabolic disorders, and infection. Reduced diffusion is present in lesions with a low nucleus-cytoplasm ratio. DWI is susceptible to the intrinsic T2* signal of the tissue because it is a gradient sequence, known as the T2 shine-through effect. To separate the DWI signal from T2 shine through, the apparent water diffusion coefficient (ADC) is calculated by acquiring images with different gradient duration and amplitude (b-values), thus eliminating the T1 and T2* values.8 Diffusion tensor imaging (DTI) assesses the anisotropy of the brain tissue by evaluating differences in the direction of diffusion of water molecules in normal and abnormal tissues and provides information about the orientation and integrity of the white matter tracts.

MRA and MR venography, which usually are performed without intravenous contrast material using the time-of-flight technique, are useful if a vascular injury is suspected.

MR spectroscopy (MRS) assesses the distribution and quantification of naturally occurring molecules within the central nervous system (see Chapter 25). Various techniques are available. The most commonly used technique is the point-resolved spectroscopy sequence. Both short time to echo (TE) and long TE acquisitions can be obtained. All acquisitions provide sensitive, noninvasive analysis of brain metabolites and cellular biochemical changes. Decreased N-acetyl aspartate (NAA) has been reported to correlate with abnormal neuropsychological function tests in persons with TBI.9

Perfusion MRI of the brain can be assessed either through a dynamic T2/T2* acquisition during a bolus injection of intravenous contrast material or with unenhanced techniques such as arterial spin labeling and blood–oxygen level dependent sequences (see Chapters 27 and 28). Perfusion MRI provides the cerebral blood volume, cerebral blood flow, and mean transit time in targeted areas of the brain.

Magnetization transfer imaging (MTI) relies on the principle that protons bound in macromolecules of tissues exhibit T1 relaxation coupling with protons in the aqueous phase or water. When an off-resonance saturation pulse is applied, it selectively saturates the protons that are bound in macromolecules. These protons subsequently exchange longitudinal magnetization with free water protons, leading to a reduction in the detected signal intensity. The magnetization transfer ratio (MTR) may provide a quantitative index of the structural integrity of tissue. Animal models of rotational acceleration suggest that quantitative MTI offers increased sensitivity to detection of histopathologically proven damage, such as axonal swelling, compared with conventional MRI. Although the precise mechanism for MTR reduction is incompletely understood for mild head trauma, the extent of the abnormality usually increases as the MTR value decreases. Abnormal MTR also has been found in normal-appearing white matter on MRI and is an apparent predictor of poor outcome.10 Functional MRI has shown changes in regional brain activation in patients with TBI and has been used for assessment of clinical outcome of these patients.

Other Imaging Modalities

Magnetoencephalography (MEG) detects the magnetic waves that are created by the electric current along the axons. The advantage of MEG over electroencephalography is that the magnetic waves are less susceptible to distortion caused by the skull than the electric currents (see Chapter 28). This method of imaging has found that brain dysfunction is present in a significantly greater number of patients with minor head trauma and postconcussive syndrome than is shown by MRI or electroencephalography. This method shows excessive abnormal low-frequency magnetic activity, which provides objective evidence of brain injury in these patients that correlates with the degree of symptomatic recovery.

Single-photon emission CT can detect abnormalities in cerebral blood flow; however, alterations in cerebral blood flow are not always associated with traumatic lesions on imaging. A negative initial single-photon emission CT scan after trauma seems to be a strong predictor of a favorable clinical outcome. A worse prognosis is associated with large lesions, multiple lesions, and lesions in the brainstem, temporal lobes, parietal lobes, or basal ganglia.

Positron emission tomography (PET) measures the cerebral metabolism of various substrates (see Chapter 25). Fluorodeoxyglucose is the primary substrate used in the measurement of glucose metabolism, which should correspond to neuronal viability. PET can be used in patients with DAI to determine the extent of damage and prognosis. PET also has been helpful in delineating the extent of reversibility of lesions. The major limitation of PET is the inability to distinguish functional abnormalities from structural damage.

Classification and Mechanisms of Head Injury

Based on different mechanisms of trauma and the associated injuries, the sequelae of head trauma can be classified into direct injury related to impact loading forces and indirect injury related to acceleration/deceleration and rotational forces. Direct injury may be subclassified into penetrating and nonpenetrating closed head injury (CHI). Moreover, brain injury in persons with a CHI may result directly from the impact (coup), or it may happen indirectly (contrecoup). The most common sequelae of direct head injury are scalp hematoma, skull fracture, direct brain contusion caused by a fracture and inward deformation of the skull, brain contusion caused by movement against the rough surface of the skull base, indirect (contrecoup) brain contusion diagonal to the site of impact, stretching and laceration of the brain parenchyma, subarachnoid hemorrhage (SAH) as a result of parenchymal injury, and epidural and subdural hematomas caused by direct vascular injury beneath the site of impact. Nonimpact, acceleration/deceleration injuries are the result of forces of translation (linear), acceleration/deceleration, and rotational and angular acceleration causing shear-strain forces on axons, neurons, and blood vessels. These injuries include contusions, DAI, deep gray matter injury, brainstem injuries, intraparenchymal hematomas as a result of vascular injury, and extraaxial hematomas. The axons are most vulnerable, and the blood vessels are the most resistant to injury from acceleration/deceleration forces (Fig. 39-1).11

Based on the GCS, head injuries are classified as mild (GCS score 13 to 15), moderate (GCS score 9 to 12), and severe (GCS score 3 to 8). Most head trauma results in mild injury; the risk of death from minor head injury in childhood approaches 0%. Moderate head injury in children has a similar rate of occurrence as in adults but a lower mortality rate. Severe head injury occurs less frequently in children compared with adults and has a significantly lower mortality rate. The exception is in infants younger than 2 years, who have a higher mortality rate from severe head injuries. This higher mortality rate is partially attributed to the occurrence of NAT in this age group.12

Skull Fractures

The likelihood of intracranial injury increases significantly when a skull fracture is present in a child. However, the absence of a skull fracture does not preclude intracranial injury and has little prognostic significance in pediatric head trauma. The skull of a younger child is thinner and more pliable than the adult skull. Therefore children have a higher incidence of both skull fractures and traumatic intracranial injury without the presence of a fracture, particularly between the ages of 6 months to 2 years. The parietal and occipital bones are most commonly involved, followed by the frontal and temporal bones. Falls and motor vehicle accidents account for most skull fractures in children.

Fractures in infants may be linear, depressed, diastatic, compound (stellate or “egg shell”), “ping pong” (buckled), or penetrating. In this age group, fractures are usually the result of a fall from the arms of a caregiver, or they result from an object striking the head. Fractures tend to pass through points of weakness and course toward a suture, synchondrosis, foramen, or canal. As the calvarium becomes more mature, comminuted fractures may occur.

Linear calvarial fractures are more frequent in all age groups, followed by depressed or comminuted calvarial fractures and basilar fractures. Linear fractures usually heal without complication. Complex fractures in which the dura mater is torn may be complicated by herniation of the pia and arachnoid layers, creating a leptomeningeal cyst. The CSF pulsations lead to progressive erosion of skull around the fracture, known as a “growing fracture,” which appears as an angular or linear lytic lesion in the skull with scalloped margins (see Chapter 23). Underlying brain contusion is not uncommon.

Depressed fractures are comminuted skull fractures in which broken bones are displaced inward by at least the thickness of the skull. Depressed skull fractures usually result from blunt force trauma to the head and may result in increased intracranial pressure (ICP), epidural hematoma, and parenchymal contusion or laceration. Depressed fractures usually are treated surgically to prevent complications and also for cosmetic reasons. Fractures displacing the dural sinuses often are treated conservatively because of the potential risk of fatal hemorrhage with intervention (e-Fig. 39-2). Depressed fractures are categorized as compound fractures when a laceration of the overlying scalp is present and as penetrating fractures when an underlying dural tear is present, allowing potential communication between the external environment and the brain. Fractures that communicate with the paranasal sinuses, middle ear, or the mastoid air cells within the intracranial compartment are considered compound fractures. Pneumocephalus indicates the presence of a compound fracture (e-Fig. 39-3). The most serious complications of compound skull fractures are CSF leak and infection. Increased risk factors for infection include visible contamination (hair, skin, fat, or bone), a meningeal tear, and delayed treatment for more than 8 hours after the initial injury. A rare type of compound skull fracture is the elevated skull fracture, which occurs when the fracture is elevated outward above the outer table of the skull.

Diastatic skull fractures involve the cranial sutures. Although diastatic fractures usually involve the lambdoid suture in newborns and infants, they may involve any suture in any age group. The normal suture becomes widened, measuring more than 2 mm. Asymmetric sutures, even if they are less than 2 mm in width, should raise suspicion for a diastatic fracture. An overriding suture is indicative of a diastatic fracture.

Fractures involving the skull base are associated with increased risk of vascular or cranial nerve injury. Basilar fractures most commonly involve the temporal bone, causing bleeding into the middle ear and mastoid air cells. Temporal bone fractures are classified according to their orientation to the long axis of the temporal bone as longitudinal, transverse, or complex. This classification is helpful in determining the risk of complications. Longitudinal fractures represent the majority (70% to 90%) and may be associated with facial nerve injury, incudostapedial dissociation, and pneumocephalus. Transverse temporal bone fractures are less common but have a higher risk of permanent injury to the facial nerve and vestibulocochlear nerve, as well as cochlear disruption, causing facial paralysis, sensorineural hearing loss, and perilymph fistula, respectively. Temporal bone fractures involving the petrous apex or the carotid canal are associated with increased risk of carotid dissection, occlusion, pseudoaneurysm, and/or jugular vein injury.

Studies report a sensitivity of 94% to 99% for the detection of linear or depressed skull fractures with routine radiographis of the calvarium. These radiographs usually are obtained in suspected cases NAT as part of a skeletal survey to document additional areas of injury. Standard CT has a lower overall sensitivity for the detection of linear fractures parallel to the plane of imaging. Images and 3D CT reconstructions often are diagnostic in difficult cases. CT has a relatively high degree of sensitivity for the detection of depressed and basilar skull fractures and fractures of the facial bones, sinuses, and orbits. In children, skull fractures typically heal in 6 to 8 weeks but may remain radiographically apparent for 1 year.

Extraaxial and Intraventricular Hemorrhage

Epidural Hematoma

Epidural hematoma is uncommon in infants and slowly increases in incidence with age, reaching a peak in adulthood. The etiology of epidural hematomas in children is different from that in adults. In younger children, venous epidural hematomas are more common than the arterial epidural hematomas. Tearing of a dural venous sinus, typically the transverse or sigmoid sinus, or an emissary or diploic vein are common causes of venous epidural hematomas. The typical locations of venous epidural hematomas are in the posterior fossa and occipital region from laceration of the transverse or sigmoid sinus, in the middle cranial fossa from injury of the sphenoparietal sinus or middle meningeal veins, and in the parasagittal area from a tear of the superior sagittal sinus. In older children and adolescents, arterial hematomas are more common. The vascular groove for the middle meningeal artery is relatively shallow, with the dura more adherent to the calvarium in children. Trauma to the calvarium causing inward deformity can separate the dura from the inner table and injure the meningeal artery, with hematoma accumulating over the temporal or parietal convexities (Fig. 39-4). The incidence of associated fractures with epidural hematomas is slightly less in children than in adults (83% vs. 93%).

The clinical presentation of epidural hematomas in children typically includes an absence of loss of consciousness at the time of injury, with a lucid interval within the first 24 hours. Several factors are associated with a poor outcome in children with epidural hematomas, including absence of an immediate lucid period, the presence of additional intracranial injuries, and delayed diagnosis and surgical intervention, if indicated.

Epidural collections do not cross suture lines because of periosteal attachments, but they do cross over the falx and tentorium. Posterior fossa epidural hematomas may extend into the supratentorial compartment, whereas subdural hematomas are confined to one compartment. Moreover, epidural collections adjacent to the frontal bones may cross the midline over the superior sagittal sinus, whereas subdural hematomas extend along the falx. These features are much more reliable in differentiating epidural hematomas from subdural hematomas than is a biconvex or lentiform shape, which may not always be present. Venous epidural hematomas frequently are concave, particularly in the posterior fossa. Venous epidural hematomas are usually larger and carry an overall worse prognosis compared with arterial epidural hematomas because of their delayed presentation.

On CT, acute epidural hematomas have increased attenuation with progression to intermediate attenuation in the subacute phase and decreased attenuation, possibly with enhancing membranes, in the chronic phase. Mixed attenuation may represent active bleeding, hypercoagulable states, a dural tear with mixed CSF and blood in the epidural space, or layering serum as a result of clot retraction. Active bleeding may produce a “swirl” effect within the collection. If the mass effect of an epidural hematoma appears disproportionate to the size of the collection, underlying edema related to parenchymal contusion should be suspected.

On MRI, the signal characteristics of extraaxial collections are more variable than on CT and depend on several factors such as the patient’s hematocrit, local partial pressure of oxygen, pH level, protein concentration, and the strength of the magnetic field. The signal changes over time of intracranial hematomas have been mainly studied for parenchymal hemorrhages and subdural hematomas in adults.

The aging of extraaxial hematomas based on signal characteristics in children has not been validated. The evolution of epidural and subdural hematomas follows a similar pattern (Table 39-1). In contrast to parenchymal hematomas, the hypointense rim on T1- and T2-weighted images representing hemosiderin is rarely seen in extraaxial hematomas because of the absence of a blood-brain barrier.

MRI is useful in distinguishing small epidural from subdural hematomas, in which the displaced dura is visualized as a thin, uniform line of hypointense signal on intermediate and T2-weighted sequences. The inner membrane of a subacute or chronic subdural hematoma may have a similar appearance, but the membrane is thicker and more irregular in contour.

Subdural Hematoma

Subdural hematomas are more common in infants than in older children and adolescents. Typically, subdural hematomas result from shear-strain forces causing stretching and tearing of bridging veins that traverse the inner layer of the dura and the arachnoid membrane. In infants, subdural hematomas are bilateral in 80% to 85% of cases. The most common causes of subdural hematomas are accidental trauma and NAT, shunt placement or overshunting with intracranial hypotension, and blood dyscrasias. Most subdural hematomas in infants younger than 2 years are caused by child abuse. Other findings suggestive of NAT include additional signs of unreported physical trauma, a history of trauma inconsistent with the severity or type of injury, interhemispheric subdural hematomas, subdural hematomas at different stages of blood degradation, and the presence of retinal hemorrhages.

Clinically, infants with subdural hematomas are asymptomatic or may present with vomiting, poor feeding, irritability, a bulging fontanelle, and increasing head circumference. Older children more commonly present with the classic signs of raised ICP, headache, altered level of consciousness, increased systemic blood pressure with decreased heart rate, irregular respiration, asymmetric pupils (anisocoria), and hemiparesis. Large subdural hematomas may manifest with anemia as a result of blood loss.

Subdural hematomas are most commonly located over the frontal, parietal, and temporal lobes. Posterior fossa subdural hematomas, which account for approximately 10% of cases, usually are a result of tearing of the dural sinuses or tentorium. In contrast to epidural hematomas, subdural collections cross sutures but do not cross the falx or tentorium. The normal falx is hyperdense on CT, but it is very thin with a smooth contour. Irregularity or thickening is suggestive of a subdural hematoma.

The CT imaging characteristics of subdural hematomas depend on the age of the hematoma. Hyperacute hematomas representing unclotted blood show attenuation similar to flowing blood in the dural sinuses. After a few hours, an acute subdural hematoma demonstrates increased attenuation related to clotted blood. The hyperattenuating acute blood clot may be mixed with blood from active bleeding, serum from clot retraction, or CSF resulting from a tear in the arachnoid (see Fig. 39-4). Typically within 1 to 3 weeks, the attenuation of a subacute hematoma decreases and becomes isodense to brain parenchyma. After 2 to 3 weeks, the attenuation of a chronic hematoma becomes similar to that of CSF. The hematoma may be difficult to see during the subacute and chronic phases, and associated signs of mass effect will indicate its presence, such as effacement of the adjacent sulci, displacement of the gray-white matter junction, compression of the ventricles, or herniation. If the presence of a hematoma is questionable, intravenous contrast material can be administered to show enhancing inner and outer membranes that are not seen in the acute phase. If the size of the collection increases or if the attenuation values appear greater than expected or are heterogeneous, rebleeding should be suspected. The development of a fluid-fluid level in the chronic phase also is suggestive of rebleeding. The membranes calcify in up to 3% of cases.

MRI is more sensitive than other modalities for the detection of subdural hematomas, particularly when their attenuation is similar to that of the brain parenchyma or of CSF on routine T1 and T2 sequences. Proton-density and FLAIR sequences are the most sensitive for detection of subdural hematomas that are isointense to CSF on T1- and T2-weighted images. The appearance of subdural hematomas on MRI also varies, depending on the phase and organization of the blood products. The evolution of the signal characteristics of subdural hematomas are the same as those previously described for epidural hematomas.

Subdural hygromas are caused by laceration of the arachnoid with accumulation of CSF in the subdural space. Subdural hygromas may occur either alone or in combination with acute hemorrhage after trauma. Subdural hygromas generally manifest 3 to 5 days after injury and are often bilateral (~50% of cases). Hygromas differ from chronic subdural hematomas that are isointense to CSF on T1- and T2-weighted images because they follow the CSF signal in all sequences, including FLAIR and proton-density sequences.

Subarachnoid and Intraventricular Hemorrhage

Traumatic intraventricular hemorrhage is related either to shearing of subependymal veins or decompression of a parenchymal or subarachnoid hematoma into the ventricles. The most common location of injury to the subependymal veins is along the anterior corpus callosum, posterior fornix, or septum pellucidum. Traumatic SAH is usually a result of damage of pia-arachnoid vessels and associated parenchymal injury. It is seen in approximately 18% to 25% of pediatric CHI cases. The amount of blood in the subarachnoid space is typically small and rarely persists for more than 1 week. If a patient has a large SAH without parenchymal damage or extraaxial hematomas, the possibility of an underlying aneurysm or arteriovenous malformation that bled prior to the trauma may be considered. SAH may obstruct the normal CSF resorption at the level of the pacchionian granulations and result in communicating hydrocephalus.

CT is the imaging modality of choice for acute SAH and intraventricular hemorrhage, showing increased attenuation blood within the sulci (Fig. 39-5), layering along the interhemispheric fissure and the tentorium, and in the ventricles. The most common location of traumatic SAH is in the posterior interhemispheric fissure and along the tentorium. Although the normal falx is relatively hyperdense on CT, hyperdense material insinuating into the cerebral sulci is indicative of SAH. Initially, SAH is detected adjacent to the source of bleeding. With time, SAH accumulates in the basal cisterns (particularly the interpeduncular cistern) and Sylvian fissures, over the convexities, and in the occipital horns of the lateral ventricles.

MRI is more sensitive than CT in detecting small amounts of SAH, showing increased signal on FLAIR sequences. GRE and SWI sequences are less sensitive than FLAIR, but the demonstration of hypointensity on these sequences confirms the presence of blood. Although extensive or recurrent SAH is rare in children, it may lead to superficial siderosis as a result of the presence of hemosiderin in macrophages along the leptomeninges. Vasospasm resulting from traumatic SAH is very rare in children.

Traumatic Parenchymal Injury

Contusion and Laceration

A brain contusion is a bruise of the parenchyma. The cortex is invariably involved, with variable involvement of the underlying white matter. Brain contusions are twice as common in children as in adults. Brain contusion may occur at the site of impact (coup), diagonally opposite the site of the impact (contrecoup), along the rough calvarial surfaces in the anterior temporal and orbitofrontal regions (acceleration/deceleration), and against the free margins of falx cerebri, tentorium, and foramen magnum (in cases of herniation). Brain edema develops over the first 1 to 2 days and is maximal 3 to 5 days after the initial injury. Approximately 50% of brain contusions are hemorrhagic. Hemorrhage occurs in the Virchow-Robin spaces, perpendicular to the pial surface, and may extend into the subarachnoid space. With increasing severity, the microhemorrhages coalesce into more focal hematomas, which usually develop 2 to 4 days after the injury. The presence of focal parenchymal hematomas has been associated with an adverse outcome. After approximately 1 month, the contused areas evolve into encephalomalacia and associated volume loss.

CT shows hypoattenuation with loss of the gray-white matter differentiation and effacement of the sulci. Hemorrhagic foci appear as areas of increased attenuation involving the cortex (see Fig. 39-5). The attenuation of acute focal parenchymal hematomas increases during the first few days as the clot retracts and then decreases with subsequent proteolysis. Cortical enhancement may be seen for 1 to 2 weeks as a result of proliferation of immature capillaries that lack a blood-brain barrier.

MRI is the imaging modality of choice for parenchymal contusion. In the first hours after the injury when overt edema is not apparent, DWI is the most sensitive sequence for detecting parenchymal contusion, as well as shear injury. After 1 to 2 days, the contused areas appear hyperintense on FLAIR and T2-weighted images related to edema, and the microhemorrhages or focal confluent hemorrhages appear hypointense on T2 GRE or SWI images (e-Fig. 39-6). High signal intensity on T1-weighted images may be seen in the subacute phase, related to the presence of methemoglobin. FLAIR is insensitive to parenchymal contusion in neonates and infants because of the low contrast with the unmyelinated white matter. T2-weighted images are particularly useful in these cases showing loss of gray-matter differentiation and hyperintensity relative to the normal brain.

Lacerations of brain parenchyma are the result of a greater mechanical force that causes tearing of tissue, extending from the cortex into the white matter. Although lacerations are characteristic of penetrating or perforating injury, they also occur in CHI or near a fracture. They are associated with a variable degree of hemorrhage (see Fig. 39-5). The hematoma may rapidly increase in size after the injury, and follow-up imaging in the first 24 to 48 hours is recommended. The most typical locations for parenchymal laceration are the inferior frontal and anterior temporal lobes. Lacerations also may occur in the corpus callosum and brainstem in association with DAI. Lacerations in the pontomedullary junction and cerebral peduncles presumably are related to hyperextension injury. Cerebellar contusions usually are associated with occipital fractures and most frequently involve the inferior cerebellar hemispheres and cerebellar tonsils.

Diffuse Axonal Injury

DAI is one of the most devastating types of TBI and the most common cause of posttraumatic neurologic and cognitive disability and a vegetative state. DAI is more common in children than previously recognized, with infants being particularly susceptible. Different parts of the brain have different consistencies depending on cell morphology, cell concentration, and variable degrees of fixation. DAI results from rapid acceleration/deceleration, resulting in angular and rotational forces that stretch the axons, causing damage to the axonal cytoskeleton, resulting in swelling. Subsequent calcium influx into injured axons is thought to cause further damage.

The degree of impairment of consciousness is usually more severe than with other primary injuries. Patients may present with immediate loss of consciousness and coma. Clinical symptoms often are disproportionate to imaging findings, and CT findings may be normal. Although high-speed motor vehicle accidents are the most common cause of DAI, it also may be caused by any head trauma, including relatively minor trauma. The mechanism of shear-strain injury in infants younger than 12 months is different from the pattern in older children and adults; the incidence of hemorrhagic lesions in infants seems to be lower. The dominant histologic abnormality in infants predominantly is related to hypoxia, which acutely leads to brain swelling and sometimes death. In later stages, extensive encephalomacia and volume loss occur.

All imaging modalities underestimate the extent of DAI. As previously noted, diffuse brain damage may not be evident on CT. Reports have found that CT of acute CHI reveals abnormalities in only 20% to 50% of patients with axonal shear injuries, primarily because of their small size and frequent initial nonhemorrhagic nature. The typical CT findings of DAI include discrete foci of increased or decreased attenuation, which are usually bilateral, less than 1 cm in size, and are oriented in an axis that parallels the tracts. DAI is most commonly located in the subcortical white matter, near the gray-white matter junction of the frontal and temporal lobes, posterior corpus callosum (most commonly the posterior body and splenium), and brainstem (dorsolateral midbrain and upper pons). An important indirect sign of axonal injury involving the corpus callosum is the presence of intraventricular hemorrhage from shearing of the adjacent subependymal veins (Fig. 39-7). With increasing severity, involvement of the internal and external capsules, basal ganglia, thalami, cerebellum, and parietal and occipital lobes occurs. CT also demonstrates diffuse brain edema in infants with severe shear-strain injury. Within 2 to 3 weeks, encephalomalacia and volume loss develop.

MRI is more sensitive than other modalities in detecting shear injuries. In patients in whom DAI is suspected, imaging should be performed at least 3 to 7 days after injury because cellular necrosis and edema are maximal at this time. Fast spin echo imaging is less sensitive to DAI than is conventional spin echo imaging.

MRI using SWI and T2* sequences detects DAI in approximately 30% of patients with mild head injury and an initially normal CT scan. DAI is visualized as multiple small, round, ovoid, or elliptical areas of decreased T1 signal, increased T2 and FLAIR signal, and abnormal DWI signal. Hemorrhagic lesions cause susceptibility effect, with hypointensity on T2* and SWI (see Fig. 39-7). Initial nonhemorrhagic lesions may convert to hemorrhagic foci with time, showing typical evolution of signal characteristics of blood, with increasing size as a result of hemorrhage and edema. DWI depicts additional lesions that exhibit facilitated or, most commonly, restricted diffusion. ADC values decrease over a longer period, ~18 days, after injury, which is beyond the usual time frame for cytotoxic edema related to ischemia.13

A reduction in fractional anisotropy has been shown with DTI in early imaging in patients with TBI; this reduction is not apparent with other imaging, including DWI. DTI with 3D fiber tracking has been shown to visualize acute axonal shearing injuries and may have prognostic value for cognitive and neurologic sequelae after TBI.

MTI can show the extent of these lesions to a greater degree. The MTR provides a quantitative index of the magnetization transfer effect and may be viewed as a measure of the structural integrity of tissues. The detection of abnormal MTR in normal-appearing white matter may predict a poor patient outcome. With MTI, the MTR can be derived and can quantitatively measure the structural integrity of tissues. It is more sensitive than T2-weighted images in detecting histologic axonal damage in animal models. Associations have been found between MTI abnormalities and neurologic and cognitive deficits. MRS also can quantify the damage after DAI with decreased NAA, increased choline, and increased glutamate and glutamine levels.

A grading scheme for DAI based on the severity of head trauma and the location of the lesions has been proposed, with mild trauma resulting in more superficial lesions and increasingly severe trauma resulting in involvement of deeper structures. Although DAI is rarely fatal, a greater number of lesions correlates with poorer outcomes. Patients with widespread MRI findings or brainstem injuries usually show no significant neurologic recovery.

Deep Gray Matter Injury

Deep gray matter injury, which usually occurs in severely or fatally injured patients, constitutes less than 5% of all primary intraaxial injuries. This type of injury is found in the thalami, basal ganglia, upper brainstem, and regions around the third ventricle from shear-strain forces disrupting small perforating vessels. The vascular injury can lead to ischemic injury and infarctions with hemorrhagic or nonhemorrhagic lesions. Patients who survive have profound neurologic deficits. CT is usually normal but can show small foci of petechial hemorrhage. FLAIR is the most sensitive MRI sequence for nonhemorrhagic deep gray matter injury, showing foci of hyperintense signal. Other sequences give the expected signal changes of blood products if the lesions are hemorrhagic.

Brainstem Injury

Although primary and secondary brainstem injury is uncommon in the pediatric population, it likely underestimated because CT remains the initial imaging modality of choice in children with head trauma. Clinically, patients with a brainstem injury do not follow the expected course of recovery, warranting further evaluation with MRI.

Primary injuries include brainstem contusion, brainstem shear injury, and the rare pontomedullary separation. Secondary lesions include hypoxic-ischemic injury and Duret hemorrhages. Secondary brainstem injuries cause death in almost 50% of cases of head injury caused by blunt trauma.

Primary brainstem injury can occur from direct forces or, more commonly, indirect forces. This type of injury characteristically involves the dorsal lateral midbrain and superior pons. Less frequent involvement is seen in the lateral midbrain or cerebral peduncle, and infrequent involvement is seen in the periaqueductal region of the midbrain. The lesions involving the periaqueductal region may show hemorrhage and ischemia on MRI related to secondary shear-strain forces on perforating vessels of the brainstem. Indirect forces resulting in brainstem injury can have associated DAI in the corpus callosum and deep white matter of the cerebrum. Although rare, direct forces have been found to result in injury to the upper midbrain from impaction against the free edge of the tentorium. The association of this type of injury with DAI is variable. Pontomedullary separation is a tear of the brainstem at the ventral pontomedullary sulcus as a result of hyperextension forces. This injury is often associated with craniocervical dislocation and is usually fatal.

Secondary brainstem injury can be caused by mechanical forces on the upper midbrain resulting in transtentorial herniation, global injury related to hypoxia-ischemia, or hypoperfusion injury related to hypotension. Transtentorial herniation may be a result of many etiologies, including generalized edema, increased ICP, and intraaxial and extraaxial hematomas. If the cause of transtentorial herniation can be treated, its sequelae have the potential to be reversed. Prolonged compression of the brainstem often results in irreversible injury, typically involving the central brainstem. Brainstem injury related to hypoxia is usually a terminal event.

The tegmentum of the midbrain and the anterior pons are most commonly involved in secondary brainstem injury and can show areas of hemorrhage, infarction, and necrosis. Duret hemorrhages typically are seen in the central pons. They are always seen in association with downward herniation and usually are caused by damage to the medial pontine perforating branches of the basilar artery. The basilar artery is relatively tethered by the circle of Willis, and caudal displacement of the brainstem results in stretching and ultimately laceration of the perforating arteries. The hemorrhage also can result either from vessel wall rupture caused by hypoxic injury or from venous infarction.

Diffuse Brain Swelling and Hypoperfusion Injury

Diffuse brain swelling has been reported to occur in 21% of cases of head trauma in the pediatric age group and is 3.5 times more common in this group, particularly in infants and young children, compared with adults. In head trauma in children, two conditions produce similar but distinguishable imaging patterns of diffuse brain swelling: increased blood volume and axonal shear injury.

After a young child sustains a traumatic injury, the immature vasoregulatory system responds with vasodilation and increased blood flow, resulting in a hyperemic state and subsequent diffuse cerebral swelling or edema. In addition, the release of excitatory amines also causes vasodilation and increased blood volume, and the redistribution of intracranial blood from pial to intraparenchymal vessels seems to contribute further to the hyperemic state. On CT, axonal shear injury in the early stage can have an imaging appearance similar to that of cerebral edema.

The imaging findings in both conditions primarily affect the cerebrum with loss of the gray-white interface, diffuse homogeneously decreased attenuation, and effacement of the basal cisterns and ventricles on CT. The decreased attenuation can cause the circulating blood to appear hyperdense on CT and give a “pseudo-SAH” appearance, and the cerebellum may appear relatively hyperdense in contrast to the edematous cerebral hemispheres (“white cerebellar sign,” Fig. 39-8).

DWI, MRS, and SWI are helpful in differentiating the two conditions. Increased blood flow and interstitial fluid (vasogenic) edema have increased diffusion on DWI (a hypointense DWI signal and a hyperintense ADC signal), whereas acute axonal shear injury with cytotoxic edema shows restricted diffusion (a hyperintense DWI signal and a hypointense ADC signal). DTI also has been found to show decreased anisotropy, providing detection of neuronal injury earlier than conventional DWI. MRS shows decreased NAA levels and increased lactate, choline, and neurotransmitter levels as a result of neuronal (axonal) injury in persons with acute DAI, with metabolites near normal early in the course of cerebral edema. This spectral pattern with axonal injury predicts a poor prognosis. SWI has a high sensitivity for the detection of hemorrhagic lesions and is useful in the evaluation of DAI. Additionally, significant axonal injury leads to evidence of volume loss on follow-up studies, whereas the brain returns to a normal appearance with cerebral edema if significant neuronal damage is not present.

Progressive edema and increased ICP can result in transtentorial, subfalcine, and tonsillar herniation, with subsequent vascular complications, which leads to death in 7% of pediatric cases. Overall, these complications occur less frequently in children than in adults.

Hypoperfusion injury in persons who have sustained a trauma may be due to many etiologies, including severe cerebral edema, hypotension, and shock. This pattern also shows a diffuse loss of gray-white interface; however, relative preservation of the extraaxial spaces and basal cisterns occurs, and ventricles appear relatively normal. Restricted diffusion is present in persons who have a hypoperfusion injury, in contrast to the presence of increased diffusion in persons with vasogenic edema. Distinguishing the three described patterns of injury by imaging has significant clinical implications because the therapy is vastly different for each of the conditions.

The “reversal sign” is a striking CT pattern in children who have had significant anoxic-ischemic injury in which there is decreased attenuation of the cortex diffusely with relative preservation of the attenuation of the thalami, brainstem, and cerebellum. Although children with a diffuse edema pattern have an overall higher mortality rate than do children with the reversal sign, its presence also portends a poor prognosis.

Sequelae of Trauma

Sequelae of trauma may manifest early or as long-term complications. Posttraumatic edema, vascular injuries, and infarction typically manifest acutely. Encephalomalacia and neurologic and cognitive disability occur as the long-term sequelae of trauma. Seizures and hydrocephalus may manifest early or have a delayed onset.

Pediatric traumatic vascular injury is less common in children than in adults. Traumatic arterial dissection may result in infarction (Fig. 39-9). Vasospasm, vascular thrombosis, compression of a vessel from a hematoma, and hypoperfusion also can cause infarction. Intracranial carotid dissection, pseudoaneurysms after skull base fractures, and carotid cavernous fistula are less common sequelae following trauma. Intraoral trauma as a result of a fall with an object in the mouth is the most common cause of extracranial carotid artery injury in children.

Dissection of the vertebral arteries is unusual in children. Dissection occurring at C1-C2 level may be predisposed by the presence a bony bridge over the vessel on the atlas known as the arcuate first neural foramen, which is an anatomic variant.

Venous injury involving the dural venous sinuses is more common as a result of birth trauma and often is accompanied by fractures, subdural hematomas, or epidural hematomas. Thrombosis of the venous sinuses can be related to injury to the epithelial lining, compression as a result of intracranial bleeding or an adjacent fracture, or increased ICP.

CTA can be used for the diagnosis of possible vascular injury in skull base fractures and the diagnosis of possible vascular injury such as laceration, occlusion, dissection, pseudoaneurysm, or arteriovenous fistula.

Pneumocephalus following trauma can be located within any of the extraaxial spaces or within the ventricles and is generally self-limiting, although tension pneumocephalus with mass effect may necessitate urgent surgical evacuation.

CSF leaks can be evaluated with use of CT or radionuclide cisternography or with high-resolution CT alone. High-resolution long T2-weighted MRI also can demonstrate sites of CSF leak.

Infection is an uncommon sequela of trauma and may be seen in penetrating injuries. It can present as meningitis, empyema within the epidural or subdural spaces, cerebritis, or intracranial abscess.

Nonaccidental Trauma

More than five children die each day in the United States as a result of abuse and neglect, with many cases likely unreported. Head injury after NAT is one of the leading causes of mortality and morbidity in infants and children. Radiologic findings are an integral part of the workup and diagnosis of NAT (see Chapter 144). Head injuries in these infants and children include skull fractures, intracranial and retinal hemorrhages, and parenchymal brain injury such as contusions, edema, ischemia, and infarction.

Clinical presentation and findings often may be nonspecific, and children may present with irritability, lethargy, or seizures without external manifestations of injury. Imaging findings therefore may provide the first clue to the diagnosis of NAT. As in other cases of trauma or in the acute clinical presentation of altered mental status, CT is often the initial imaging modality, with MRI used to further define the extent of injury and to look for potential craniocervical injury.

Subdural hematomas are common in NAT because of the most common mechanism of injury. Rotational forces on the head tear the bridging veins in the extraaxial space. Bleeding into the subarachnoid space may coexist and is seen in the sulci and basal cisterns. The epidural hemorrhages seen in infants and young children who have sustained accidental trauma are uncommon in those who have sustained NAT.

Most acute subdural hematomas have high density on CT, but this finding is variable, especially when dilution is present as a result of mixing with CSF from a dural tear. A hematocrit layering effect may be seen. As hematomas age, they gradually become lower in attenuation. Isodense subdural hematomas can be differentiated from prominent subarachnoid spaces by the peripheral location of the veins in the extraaxial space and a more medial location noted when a subdural hematoma is present.

On MRI, the signal intensity of parenchymal injuries varies with time, and when correlated with other clinical findings, it can be helpful to clinicians and child advocacy services. However, the precise aging of parenchymal lesions and extraaxial hematomas should be undertaken with caution. CT hyperdensity is the only reliable indicator of age (Fig. 39-10). Subdural hematomas in infants younger than 2 years and subdural hematomas of different ages are highly suspicious for NAT. Retinal hemorrhages are a cardinal manifestation of NAT (Fig. 39-11). The severity of retinal hemorrhages and their extension to the periphery of the retina correlates with the likelihood of NAT and the severity of associated brain injury.

Mortality and morbidity in children with NAT often are due to parenchymal injury. Hypoxic ischemic injury and brain edema as a result of the trauma are more common than are primary parenchymal contusions and DAI. Hypoxic ischemic injury is more often diffuse, and although the exact mechanism is unclear, it is most likely related to hypoperfusion.

Imaging findings show a diffuse loss of gray-white matter differentiation with sparing of the basal ganglia. Severe insults show more extensive brain edema, and the basal ganglia and the posterior fossa structures may be involved. Hypoxia resulting from strangulation involves the territory of the anterior circulation as a result of compression of the carotid vessels with relative sparing of the vertebral circulation. MRI and DWI are useful to delineate the full extent of ischemic injury, and MRS may demonstrate neuronal loss by demonstrating a decrease in the neuronal marker NAA and an increase in lactate, indicating anaerobic metabolism in the affected areas of the brain (see Fig. 39-11).

A multidisciplinary approach should be used in the diagnosis of NAT when radiologic findings are suspicious or inconsistent with the provided clinical history. Certain metabolic disorders such as glutaric aciduria type 1, an inherited autosomal-recessive metabolic disorder that can present with subdural hematomas, should be considered in the differential diagnoses. Congenital or acquired coagulopathies can give rise to intracranial hemorrhage or vascular thrombosis. Birth injury can be associated with intracranial hemorrhage in infants, which generally resolves within a month. Severe forms of meningoencephalitis can present with extraaxial collections and parenchymal hemorrhage.

Vigilance, familiarity with the radiologic findings, communication with the referring physician, relevant clinical history, and recommendation for further imaging, if needed, all aid in the diagnosis of NAT.

Prognosis

Children are more likely than adults to recover from focal brain injury and have a higher likelihood of survival after severe injury. However, children seem to be more vulnerable to long-term cognitive and behavioral dysfunction after diffuse brain injury. Some degree of learning disability is found in 50% of survivors of head trauma, with severe motor, sensory, cognitive, and behavioral deficits present in many patients. Overall, children younger than 6 years have the worst prognosis, likely because of the increased risk and prevalence of shear injury in the relatively immature brain. Ultimately, functional outcome depends on how many neurons are preserved after injury. The location and extent of injury and the ability of existing neurons to reorganize their connections to recover function are critical.

Summary

Head trauma is a frequent cause of morbidity and mortality in the pediatric population. A wide spectrum of traumatic head injuries occurs in children, many of which are unique to the pediatric population. The type of injury primarily depends on the mechanism of injury, the force sustained, and the age of the patient. Newer imaging modalities are useful in the evaluation of these patients and have the potential to predict recovery and outcome.

Key Points

Head trauma and resulting TBI are the most common causes of morbidity and mortality in children.

In general, CT is the initial imaging modality used to evaluate patients with known or suspected head trauma.

The risk of death from minor head injury in childhood approaches 0%.

The likelihood of intracranial injury significantly increases when a skull fracture is present in a child. The absence of a skull fracture, however, does not preclude intracranial injury and has little prognostic significance in pediatric head trauma.

In younger children, venous epidural hematomas are more common than arterial epidural hematomas.

A brain contusion is a bruise of the parenchyma. The cortex is invariably involved, with variable involvement of the underlying white matter.

DAI is one of the most devastating types of TBI and is the most common cause of posttraumatic neurologic and cognitive disability.

Diffuse brain swelling and hypoperfusion injury is more common in children and can be differentiated using DWI, MRS, and SWI techniques.

Children are more likely than adults to recover from focal brain injury and have a higher likelihood of survival after severe injury. However, children seem to be more vulnerable to long-term cognitive and behavioral dysfunction after sustaining a diffuse brain injury.

References

1. Maas, AI, et al. Prediction of outcome in traumatic brain injury with computed tomographic characteristics: a comparison between the computed tomographic classification and combinations of computed tomographic predictors. Neurosurgery. 2005;57(6):1173–1182.

2. Pearce, MS, Salotti, JA, Little, MP, et al. Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study. Lancet. 2012;380(9840):499–505.

3. Stiell, IG, et al. The Canadian CT Head Rule for patients with minor head injury. Lancet. 2001;357(9266):1391–1396.

4. Davis, PC, et al. Head trauma. American College of Radiology. ACR Appropriateness Criteria. Radiology. 2000;215(suppl):507–524.

5. Vorona, GA, et al. The use of adaptive statistical iterative reconstruction in pediatric head CT: a feasibility study. AJNR Am J Neuroradiol. 2012. [May 24 [Epub ahead of print]].

6. Lee, H, et al. Focal lesions in acute mild traumatic brain injury and neurocognitive outcome: CT versus 3T MRI. J Neurotrauma. 2008;25(9):1049–1056.

7. Sehgal, V, et al. Clinical applications of neuroimaging with susceptibility-weighted imaging. J Magn Reson Imaging. 2005;22(4):439–450.

8. Provenzale, JM. Imaging of traumatic brain injury: a review of the recent medical literature. AJR Am J Roentgenol. 2010;194(1):16–19.

9. Tollard, E, et al. Experience of diffusion tensor imaging and 1H spectroscopy for outcome prediction in severe traumatic brain injury: preliminary results. Crit Care Med. 2009;37(4):1448–1455.

10. Bagley, LJ, et al. Magnetization transfer imaging of traumatic brain injury. J Magn Reson Imaging. 2000;11(1):1–8.

11. Zhu, GW, Wang, F, Liu, WG. Classification and prediction of outcome in traumatic brain injury based on computed tomographic imaging. J Int Med Res. 2009;37(4):983–995.

12. Maas, AI, Stocchetti, N, Bullock, R. Moderate and severe traumatic brain injury in adults. Lancet Neurol. 2008;7(8):728–741.

13. Galloway, NR, et al. Diffusion-weighted imaging improves outcome prediction in pediatric traumatic brain injury. J Neurotrauma. 2008;25(10):1153–1162.

Related posts:

Acquired Abnormalities (Stone Disease and Infection) Prenatal Imaging Alignment Disorders Spinal Tumor and Tumorlike Conditions Embryology, Anatomy, and Normal Findings Abdominal Trauma