Penetrating Craniocerebral Injury in Pediatric Patients

Abstract

Penetrating head trauma is rare in the pediatric population, and rarer still in the civilian pediatric population. The high rehabilitation potential of children and the higher likelihood of a low-velocity, survivable injury necessitates careful management to minimize morbidity due to secondary injury from ischemia or infection. Management of penetrating injuries includes patient stabilization, appropriate imaging, and if surgery is needed, entry/exit site debridement with dural closure to prevent cerebrospinal fluid leak. Post-operative care includes infection prevention, intracerebral pressure management, and early identification of vasospasm and pseudoaneurysm formation.

1. Introduction

Although rare, penetrating craniocerebral injury is an increasingly recognized cause of emergent neurosurgical admissions in children and adolescents 17 years of age or younger [1]. Outside of war zones, these injuries can occur in the setting of gang-related violence in major metropolitan areas. Penetrating head injury may result from gunshot wounds or stabbings, and less commonly due to non-powder guns (Zip Guns) or transorbital penetration [2, 3]. In this review, we have defined high-velocity injuries as those that result from projectiles/objects traveling at velocities >600 meters per second. These are most commonly the result of rifle injuries, which are more prevalent in military conflicts. Low-velocity injuries are usually the result of handguns in civilian populations. The low-velocity equivalents in military conflicts include both handgun and shrapnel injuries. Additional low-velocity mechanisms of injury in children can include BB and pellet gun injuries, stab injuries, blunt penetrating injury, and trauma resulting from pencils, branches, nails, and other objects.

A relatively small number of case series in varied settings comprise the majority of the pediatric literature on penetrating craniocerebral injury.

Barlow et al. study of gunshot wounds in 108 children (nine cerebral injuries) at Harlem Hospital over 10 years was among the first series to examine the demographic, intentionality, and cause of these types of injuries in children [4]. Most worrisome was that 53% of the guns belonged to the children themselves, and 8% of injuries were inflicted by the police.

Beaver et al. reviewed 132 children under 16 years with fatal firearm injuries in the state of Maryland [5]. The cause of death was a homicide in 46%, accident in 25%, and suicide in 22%. Deaths occurred at home in 75% of cases. The perpetrator was a friend in 21%, family member in 20%, acquaintance in 7%, a bystander in 2%, and self-inflicted in 30%.

In 1993, Levy et al. published their experience with 105 pediatric patients at the University of Southern California/Los Angeles County Medical Center with a diagnosis of a gunshot wound to the brain during an 8-year period [1]. Most of these injuries (72%) were gang-related or secondary to murder-suicide, in contrast to a much higher rate of suicide in the adult population. These findings were contextualized by rising rates of gang-related murders in Los Angeles County and increasing numbers of gang members in that time frame. Similarly, a prior study by Ordog et al. [6] described 34 patients, 10 years of age or younger who were treated for gunshot wounds in Los Angeles between 1980 and 1987, noting that no children in this age range had been treated for gunshot wounds prior to 1980.

A retrospective review of traumatic injuries in children 16 years or younger in the San Francisco bay area from 2000 to 2009 revealed that the incidence of gunshot wounds to the head in children continued to increase over time. This represents another study documenting the vexatious number of these injuries each year in children [7]. These injuries were associated with a 63% mortality rate and were thought to represent mostly intentional injuries.

Bandt et al. reported their experience with 48 patients less than 18 years of age with penetrating intracranial gunshot injuries between 2002 and 2011 in St. Louis, MO [8]. The authors proposed a management paradigm involving more aggressive treatment compared to adults due to the more favorable outcomes experienced by children.

Seventy-one pediatric patients with an intracranial gunshot wound between 1996 and 2013 in Memphis, TN were analyzed by Decuypere et al. [9]. Nearly half of the victims died from their injuries, but over 80% of survivors had a good outcome. As noted in prior studies, the authors reported that variables related to the initial clinical exam and classification of CT findings were found to be the most important predictors of outcome.

Thirty patients under the age of 13 years who suffered craniocerebral gunshot injuries were treated at Red Cross War Memorial Children’s Hospital in South Africa between 1989 and 2001 [10]. Over half of the victims were injured in the crossfire of civilian violence.

A case series from Mashhad, Iran reported 14 penetrating head injuries in children less than 10 years [11]. These injuries were primarily due to low-velocity objects, such as pencils, rods, silverware, and other miscellaneous objects. The authors noted that pediatric civilian low-velocity gunshot injuries were comparatively uncommon in some middle eastern countries and Japan as compared to the United States and Europe. These findings have supported the prior literature which documented the significant number of gunshot-related injuries to the head in the United States as compared to other countries.

A series from Tel Aviv concluded that pediatric craniocerebral gunshot injuries from plastic projectiles resembled those of low-velocity missiles, with similar treatment algorithms and outcomes compared to other penetrating craniocerebral injuries in civilians [12].

Low-velocity penetrating injuries in children are not as common in armed conflict. Given the nature of the high-velocity injuries associated with rifles, most low-velocity injuries are the result of penetrating shrapnel. Low-velocity injuries are those associated with projectiles traveling less than 600 m/sec. This is consistent with the prevalence of handgun injuries in civilian populations. Wani et al. reported on 51 children with penetrating brain injuries who were the victims of armed conflict [13]. Nearly all injuries were from roadside grenade attacks. Victims of grenade attacks had better outcomes than those who suffered bullet injuries to the head. The authors noted that being upright and running away from an audible attack may have predisposed patients to craniocerebral injury, which perhaps could have been prevented by lying down.

A large series of cranial stab wounds was published by Domingo et al. in 1994, including 54 patients less than 14 years. Injuries were due to assault in 58% and accidental in 42%. All patients survived their injury and required surgical debridement [14]. A large series on pediatric penetrating brain injury in Durban, South Africa included more stab wounds (57%) than gunshot wounds (43%) [15]. Interestingly, neurological deficits did not significantly differ by the mechanism of injury when accounting for age and other clinical information.

Irfan et al. reported their experience with four patients in the 2–3 year age range who experienced gunshot wounds to the head, only one of whom did not survive [16]. Penetrating craniocerebral injuries to patients in this age range have the additional nuance of developmental and neurobehavioral impact among survivors [17].

2. Epidemiology

Due to the rarity of penetrating craniocerebral injury in children, it is challenging to identify epidemiological data. This is particularly true given the diversity of data sources reported in the aforementioned case series.

The incidence of pediatric TBI ranges from 12 to 700 per 100,000 population, with most studies reporting between 47 and 280 [18, 19, 20]. While the incidence of penetrating TBI in children remains unknown, it represents a small fraction of those likely between 1% and 7%, and varies significantly by country and setting [15, 21].

An analysis of the National Trauma Data Bank, a curated United States trauma registry, demonstrated that gunshot wounds to the head represented 1.4% of pediatric TBI cases, with increasing incidence over time (from 275 per 100,000 in 2003 to 315 per 100,000 in 2012). The mean age was 14.8 years and 79.2% were male. Most victims were African American (43%) or white (35%). Assault was the most common mechanism (63%), followed by suicide (18.3%) and accident (12.6%). There was a time trend for increasing suicides, decreasing accidental injuries, and a stable assault rate. The location of injury was commonly residential (40.6%) or street (24.9%), and less often a public building (1.9%) or recreational location (0.9%). Mortality was 45.1% overall, and 71.5% in the setting of suicidal intent.

In general, pediatric gunshot wound victims were primarily males with mortality rates ranging from 47.9–65% [1, 6, 8, 9, 10, 15, 22, 23, 24, 25].

Among TBI patients, children in racial minority groups or of low-income status are more likely to be the victim of assault or firearms and are more likely to have poor clinical outcomes [26]. This is likely also true, specifically, for victims of penetrating TBI [21], although data are limited.

3. Management

Initial management of both high- and low-velocity penetrating injury in children involves emergent and aggressive hemodynamic stabilization, correction of coagulopathies, and neurologic assessment. Neurologic imaging after stabilization should include brain computed topography (CT) [27]. CT angiography should be considered to rule out vascular injury. An intracranial pressure monitor is indicated in patients with a low Glasgow coma score (GCS) ≤ 7 or signs of elevated intracranial pressure.

The rate of surgical intervention ranges from 51 to 100% for injuries due to gunshot wounds or explosives [9, 10, 22, 23, 24], 70–94% in pellet gun injuries [28, 29], 60% in dog bites [30], and up to100% in stab wounds [14]. Lower rates of intervention among high-velocity injuries likely reflect nonsurvivors or those with inoperable injuries.

Goals of surgery depend on the extent of the injury [31]. Small head wounds without significant intracranial pathology may simply require local wound care and closure, whereas devitalized scalp or compromised bone and dura may require more extensive debridement. Intracranial injuries with hematomas and/or mass effect may require debridement of necrotic brain tissue and accessible bone fragments, or hematoma evacuation. Debridement of the missile tract and the aggressive pursuit of bone/metallic fragments is not recommended when there is no significant mass effect (Figure 1). As noted in adult populations a primary goal of surgical intervention in children is dural closure to avoid CSF leaks. Antiepileptic and antimicrobial recommendations are discussed in the following section.

Figure 1.

A–D. Entry wound and exit wound should be debrided in the operating room with removal of superficial foreign material and repair of any dural injury.

4. Complications

4.1 Structural

Direct impact from the penetrating object can cause a variety of damage to tissue. Skull fracture and cerebrospinal fluid leak due to dural laceration are common. Children’s skulls do not become fully ossified until two years of age; therefore, they are more susceptible to skull fracture after non-missile or low-velocity penetrating trauma. The most common entry locations are the thin-walled orbit and the squamous temporal bone [32, 33, 34]. Bihemispheric injury and ventricular transgression are associated with worse outcomes (Figure 2).

Figure 2.

A–B. Bihemispheric injury is associated with a worse prognosis.

Low-velocity objects are, however, less likely to cause contra-coup injuries, thermal or blast injuries than are high-velocity or missile objects [34]. Depending on the course of the object through brain tissue, there may be a hematoma or cerebral contusion causing a mass effect.

4.2 Neurologic

The significance of the trajectory of the object/projectile is that it has been found to be a powerful determinant of outcome (both morbidity and mortality) as previously noted. The trajectory of the object impacts the likelihood of a transient or permanent neurologic deficit, which most commonly are weakness and cranial neuropathy. A review of 223 patients with a near-even mix of high and low-velocity injury found that the mean age of children with the neurologic deficit was higher (11.72 years) than those who were neurologically intact (8.96 years) [15].

The routine use of antiepileptics is controversial. Trauma literature supports the use of antiepileptics for the prevention of early (<7 days) post-traumatic epilepsy but not late post-traumatic epilepsy [34, 35]. Low GCS is independently associated with developing post-traumatic seizures in pediatric patients [36]. Left-sided injury is also reported to be associated with a higher likelihood of seizures [15].

Newer antiepileptics, such as levetiracetam, have not been as rigorously studied for use in this patient population despite their popularity. Additionally, there is a disparity in pediatric literature compared to adult literature regarding the routine use of antiepileptics after traumatic brain injury. It seems to be routine practice in moderate to severe penetrating head trauma to administer prophylactic antiepileptics due to the low-risk profile, though this remains at the discretion of the treating neurosurgeon and there is not enough literature support to define management standards.

Blunt traumatic brain injury has been associated with a high rate of new diagnoses of ADHD, oppositional defiant disorder, mood disorders, PTSD, OCD, bipolar disorder, and antisocial or aggressive behaviors. Although perhaps of less importance in the acute period, penetrating head trauma, particularly to the frontal lobe, may cause devastating neuropsychiatric sequelae in children. Neuropsychiatric evaluation and referral to child psychiatry should always be considered [37].

4.3 Infectious

The routine use of broad-spectrum empiric antibiotics after penetrating head trauma is common but not universally agreed upon. Complications of grossly contaminated wounds include cerebritis, intracerebral abscess, ventriculitis, and meningitis. The most common pathogens are Staphylococcus species and gram-negative bacteria, though anaerobic species have been reported in both pediatric and adult cases with high mortality [34, 38]. Empiric antibiotic therapy should cover for these most common pathogens. There is considerable variability in the literature as to the exact type of antibiotics and the length of therapy. Early surgical intervention and debridement <12 hours from presentation decrease the risk of subsequent infection, with watertight dural closure being one of the most significant variables related to minimizing infection [34, 39].

Pediatric series of penetrating head trauma report a higher rate of infection than adults, up to 40–50% [40]. Risk factors for infection include the foreign body being a porous material, such as wood, fragmentation of the object, cerebrospinal fluid leak, nasal or mastoid sinus involvement, and gross contamination of the entry or exit site (Figure 3).

Figure 3.

Dural laceration and the cerebrospinal fluid leak should be addressed urgently in the operating room to reduce the likelihood of infection.

Low-velocity penetrating injuries, which are more common in young patients, are more likely to be grossly contaminated with skin, hair, and bone along the projectile tract. They are also more likely to involve porous material that can fragment, further increasing the risk of infection.

While most studies recommend immediate initiation of broad-spectrum antibiotics on arrival to the emergency department, a large retrospective review of adult trauma patients in Cleveland, OH described a very low incidence of infection despite most patients only receiving one dose of Ancef [39].

4.4 Vascular

Morbidity after penetrating head trauma can be significantly impacted by both immediate and delayed vascular pathology. Immediate injury to the vasculature or direct tissue injury can lead to significant blood loss, space-occupying hematoma, or cerebral ischemia, all of which may cause neurologic deficit or secondary injury due to cerebral edema and elevated intracranial pressure [41].

Traumatic intracranial pseudoaneurysms are a relatively common sequelae of both blunt and penetrating head trauma; 20% of traumatic aneurysms are related to penetrating head trauma. The most common vessels involved are the MCA, followed by the ACA and the ICA [Alvis]. The rupture risk of traumatic pseudoaneurysms resultant from penetrating injury is not well quantified, and they may grow or shrink with time. However, the presence of subarachnoid hemorrhage after penetrating head trauma has been significantly associated with mortality (Figure 4). At 48 hours post-injury, 17% of survivors and 68% of nonsurvivors had SAH on imaging [42]. Vascular imaging should be strongly considered for any penetrating trauma in which the object tract traverses the Sylvian fissure or any major subarachnoid space or is adjacent to vascular structures. CTA is rapid and convenient, however, may be obscured by an artifact of metallic objects and, therefore, formal angiography might be necessary. Aneurysms can form in the days following trauma or may form in a delayed fashion. Surgical or interventional treatment is recommended due to the high risk of rupture [43, 44].

Figure 4.

Subarachnoid hemorrhage is a significant predictor of mortality.

Vasospasm after penetrating pediatric head trauma may be an underrecognized phenomenon contributing to preventable “secondary” brain injury by reducing cerebral perfusion. A prospective study of pediatric patients with blunt traumatic brain injury demonstrated a moderately high prevalence of vasospasm on transcranial doppler (TCD), which was correlated with the severity of the injury. This study further identified post-resuscitation GCS < 8, mechanism of injury (motor vehicle accident), and fever at admission as significant predictors of subsequent development of vasospasm [45]. In adults with penetrating trauma, the incidence of vasospasm is as high as 40% and has a strong association with the presence of subarachnoid hemorrhage [46]. Vasospasm occurs in 21% of blunt moderate to severe TBI patients, and 33.5% of severe TBI patients studied with TCD. Peak onset is 2–4 days post-injury and resolves in 2–3 days [45]. Vasospasm following low-velocity penetrating trauma has also been described. Given that these injuries are less likely to be lethal and have the potential for a good outcome, prevention of secondary injury is paramount [47].

The anticipation and presumptive treatment to avoid concomitant infection are essential. Associated cerebritis, abscess, or sepsis can be additionally associated with stroke. This is an important consideration in an immunocompromised patient. Additionally, stroke-associated pneumonia (SAP) can increase morbidity and mortality following these injuries.

5. Conclusions

Pediatric penetrating head trauma in civilian populations is rare. Children in these settings are more likely to be struck by low-velocity or non-missile objects than by firearms, which confer a higher likelihood of survival. Therefore, surgical debridement or decompression, closure of dural violations to prevent infection, and diligent medical management to prevent secondary injury are critical to maximizing recovery in this resilient patient population.