Orbitocranial Penetration of Ball Bearing Gun Trauma: A Prospective Human Cadaveric Study

Abstract

Objective  We completed a prospective human cadaveric study to determine the ability of a ball bearing (BB) pellet to penetrate the orbit and/or surrounding structures.

Methods  A single trained sergeant officer discharged an alloy steel air rifle to eight cadaver orbits from four adult human cadaver heads. Five BB pellets each were aimed at three locations (caruncle, upper eyelid, or lower eyelid) at 10 cm and 1 m, and then less specifically, at the orbital region for 3- and 5-m distances. Computed tomography (CT) of the cadaver heads was performed. Final locations of BB pellets are divided into three categories: intracranial, surrounding orbital structures including the pterygopalatine fossa and infratemporal fossa, and orbit.

Results  Of 40 BB pellets, 37 penetrated soft tissue and were visualized on CT: 19 (51%) rested in the intracranial space, 17 (46%) in surrounding orbital structures, and 1 (3%) within the orbit. The deepest position of a pellet was in the parietal lobe, and most superficial location anterior to the frontal bone. Pellets discharged from 1 m were more likely to rest in the intracranial space compared with those from 10 cm ( p  < 0.001), 3 m ( p  = 0.011), and 5 m ( p  = 0.004). The distance of discharge was associated with final pellet location ( p  = 0.001).

Conclusion  BB guns should be considered dangerous and potentially deadly when aimed at the orbit. Although the thick calvarium can protect the intracranial space from BB penetration, the orbit may be a vulnerable entry point with relatively low resistance, allowing penetration of the intracranial and periorbital spaces.

Introduction

Ball bearing (BB) gun injuries commonly present to the emergency room. Nearly 22,000 children and adolescents each year are assessed in the pediatric emergency setting after nonpowder gun injuries from BB and airsoft guns. ¹ Eighty percent of these injuries occur in children and teenagers, who are particularly vulnerable. ² This may be in part due to the fact that BB guns are marketed as toys and can be easily purchased online, without age verification or requirements for licensure. Moreover, the vast majority of children (98%) presenting with BB gun–related injuries are not wearing any form of eye protection. ³

Ocular trauma is common in such cases, and patients can present with hyphema, traumatic cataract, vitreous hemorrhage, retinal detachment, and/or globe rupture with intraocular foreign body. ⁴ Oftentimes, injury can be severe enough to result in poor visual prognosis including blindness and, in some cases, it is not possible to salvage the globe. ^{2 4} Retrospective studies emphasize the severe ocular injuries that can result from recreational BB guns and caution their use. ^{1 2} When patients present to the emergency room, it is often unclear exactly what type of pellet and gun is used. Imaging can demonstrate intraorbital and/or intracranial foreign bodies that appear round and metallic; however, it is possible these presumed BB pellets are in fact foreign bodies of another kind, such as a buckshot from a higher velocity weapon. The purpose of this prospective study was to determine the capacity and tendency of BB pellets to penetrate the orbital, intracranial, and/or surrounding structures when discharged from an air rifle at close and medium range.

Methods

In this prospective interventional study, a single alloy steel air rifle (Daisy Powerline 880) was used to fire BB pellets at each of eight human cadaver orbits from varying distances, while aiming at various positions. Highly trained sergeant officers affiliated with the University of California, Los Angeles (UCLA) oversaw the project and a single skilled officer from the team fired all BB pellets. This officer had extensive experience in shooting. The experiment was conducted in a safe space approved for research purposes, and eye protection was worn for all staff involved.

Four adult human cadaver heads were obtained and a total of eight human cadaver orbits (two orbits per head) were used. Human cadaver heads were kept refrigerated until several hours before the experiment was conducted. The cadaver heads were secured to a metal stand to ensure their stability during the experiment. One cadaver orbit was found to be enucleated on the day of the experiment. This orbit was not excluded given that our primary outcome was penetration of the orbit, intracranial space, and surrounding structures, not globe trauma. A waiver of consent was obtained from the institutional review board (IRB).

The BB pellets used were 0.177-caliber, 4.5-mm round steel BBs (Daisy Ammunition zinc plated). Each pellet weighed 5.1 grains. Of note, a single grain is ∼0.06 g. This description of BB pellet weight was listed on the manufacturer's label, which also states that there are minuscule variations in individual BB weights during production. Before each discharge, 10 pneumatic pumps were used to power the air rifle. A consistent number of pneumatic pumps was used prior to each shot to control for the velocity of the BBs shot, which can vary based on number of pumps. ⁵ The number of pneumatic pumps was recommended by the UCLA sergeant officer as this is a typical level of power preferred by users. This number was kept constant throughout the experiment.

The BB gun was discharged in eight different conditions. A single condition was used for each of the eight cadaveric orbits. Five pellet strikes or trials were completed for each condition/orbit. Two factors were varied in each condition: the distance from eye and location of each attempted pellet penetration. Four distances were tested: 10 cm, 1 m, 3 m, 5 m. For the 10-cm and 1-m conditions, three separate areas were targeted: the caruncle, upper eyelid, and lower eyelid. Due to limits on precision, target shots for the 3-m and 5-m conditions were aimed generally at the ocular adnexa and entry point documented. The 10-cm and 1-m conditions used 15 pellets shot each (5 trials for each of the 3 target locations) using 6 total orbits (orbits 1–6). The remaining 3-m and 5-m conditions had a single target (the orbit) and thus used only 2 orbits (orbits 7 and 8) and 5 trials per condition for a total of 10 shots. In total, there were 40 pellets fired.

All globes were inflated with balanced salt solution intravitreally using a half-inch long 30-gauge needle prior to the experiment. Intraocular pressure was measured by palpation to be roughly physiologic. White chalk was powdered over the eyelids before each discharge to visualize the site of penetration. This was effective in all but one BB pellet penetration. Each shot was observed by two authors from a safe distance of 3 feet and the site of penetration after each shot was examined by chalk disruption due to BB pellet entry. Both observers agreed on all points of penetration.

Computed tomography (CT) scans of the orbit were completed for all eight orbits after completion of the experiment. CT scans from each condition were assessed with the Horos (Horos v3.3.1) imaging suite. The final position of the BB and any ancillary bony defects were noted. Final anatomic location was categorized as orbit, surrounding orbital structures (the maxillary sinus, sphenoid sinus, pterygopalatine fossa, and infratemporal fossa), and intracranial space. Parametric and nonparametric statistical analysis was performed utilizing chi-square tests and Mann–Whitney U test where appropriate. A p -value of <0.05 was considered statistically significant. IRB approval was obtained and the research adhered to the tenets of the Declaration of Helsinki.

Results

Points of penetration were accurately identified in all but one single pellet strike (fifth pellet on orbit 4; Fig. 1B ). The distribution of penetrations for pellets confirmed relatively high shooting accuracy ( Fig. 1A–C ). For the three 10-cm distances, orbits 1, 2 ( Fig. 1A ), and 3 ( Fig. 1B ), five pellets were confirmed to penetrate as per intention in the caruncle, upper lid, and lower eyelid, respectively. Accuracy was similarly high at 1 m, where each of the five pellet strikes for orbits 4 ( Fig. 1B ), 5, and 6 ( Fig. 1C ) again punctured per intention at the caruncle, upper lid, and lower eyelid, respectively. The distribution of pellet penetration was wider when struck from the greater distances of 3 m (orbit 7) and 5 m (orbit 8). These can be noted in Fig. 1D . Of note, one of the eight orbits (orbit 7; Fig. 1D ) was found incidentally to be anophthalmic and had a sphere implant positioned within the orbit.

Fig. 1.

Line drawings of eight orbits and the location of each ball bearing pellet penetration. Each shot is labeled in order of shooting. A total of five pellets were shot for each orbit. ( A ) Orbit 1 (right eye): at 10 cm aiming for the caruncle. Orbit 2 (left eye): at 10 cm aiming for the upper lid. ( B ) Orbit 3 (right eye): at 10 cm aiming for the lower lid. Orbit 4 (left eye): at 1 m aiming for the caruncle. *The exact position of penetration of pellet #5 is unclear. ( C ) Orbit 5 (right eye): at 1 m aiming for the upper lid. Orbit 6 (left eye): at 1 m aiming for the lower lid. ( D ) Orbit 7 (right eye): at 3 m aiming for the orbit. Of note, this orbit was enucleated. Orbit 8 (left eye): at 5 m aiming for the orbit.

On review of CT imaging, 37 of the 40 (92.5%) BB pellets were identified. The three pellets (3/40; 7.5%) that were not visualized on CT were from orbit 7 ( n  = 2) and orbit 8 ( n  = 1), discharged from 3 and 5 m, respectively. Final positions of the 37 pellets were variable: 19 (51%) were noted in the intracranial space, 17 (46%) were noted in structures surrounding the orbit, and 1 (3%) was noted in the orbit ( Table 1 ).

Table 1. Descriptive data of final location of pellets shot for all eight orbits.

	Intracranial	Surrounding	Orbit	Total
10 cm
Orbit 1 , caruncle	0 (0%)	5 (100%)	0 (0%)	5 (100%)
Orbit 2 , upper lid	3 (60%)	2 (40%)	0 (0%)	5 (100%)
Orbit 3 , lower lid	0 (0%)	5 (100%)	0 (0%)	5 (100%)
	3 (20%)	12 (80%)	0 (0%)	15 (100%)
1 m
Orbit 4, caruncle	4 (80%)	1 (20%)	0 (0%)	5 (100%)
Orbit 5 , upper lid	5 (100%)	0 (0%)	0 (0%)	5 (100%)
Orbit 6 , lower lid	5 (100%)	0 (0%)	0 (0%)	5 (100%)
	14 (93%)	1 (7%)	0 (0%)	15 (100%)
3 m
Orbit 7, a orbit	1 (33%)	1 (33%)	1 (33%)	3 (60%)
5 m
Orbit 8, orbit	1 (25%)	3 (75)	0 (0%)	4 (80%)
Total	19 (51%)	17 (46%)	1 (3%)	37 (92%)

Note: Final positions for 37 of the 40 BB pellet discharges that were confirmed to enter the orbit. The number of pellets and percentage in each category are presented. The three missing pellets on CT were from orbit 7 ( n  = 2; 3 m) and orbit 8 ( n  = 1; 5 m).

^aOrbit 7 was an enucleated eye.

All 15 BB pellets shot from a distance of 10 cm (orbits 1–3) were visualized on CT. Twelve of the 15 (80%) were noted in the surrounding orbital structures and 3 (20%) in the intracranial space. Of note, all five pellets aimed at the lower eyelid from 10 cm came to rest collectively in the pterygopalatine fossa ( Fig. 2C ). Five additional pellets were found in the surrounding structures; one discharged from 1m, one from 3m, and three from 5m. One pellet discharged from 5m was identified in the inferotemporal fossa ( Fig. 2E ).

Fig. 2.

Computed tomography imaging highlighting the variable locations of the ball bearing pellets. ( A ) The pellet that penetrate deepest was seen from orbit 5 (aimed for upper eyelid of the right eye from a distance of 1 m). ( B ) One pellet in orbit 7 (aimed for enucleated right eye from a distance of 3 m) landed within the orbit. Three subcutaneous pellets were seen in orbit 8 (aimed for the left eye from a distance of 5 m) anterior to the frontal bone. These pellets did not penetrate through the orbit. ( C ) All five pellets shot for orbit 3 (aimed for lower lid of the right eye from a distance of 10 cm distance) landed collectively behind pterygopalatine fossa. ( D ) One pellet for orbit 2 (aimed for upper lid of the left eye from a distance of 10 cm) landed in the pterygopalatine fossa. ( E ) One pellet from orbit 7 (aimed for enucleated right eye from a distance of 3 m) landed in the surrounding orbital structures in the lateral wall of the inferotemporal fossa.

The BB pellet with the deepest penetration was seen in the temporoparietal lobe of orbit 5. This pellet was aimed at the upper eyelid from 1 m ( Fig. 2A ). The most superficial pellet location was in the subcutaneous space anterior to the frontal bone ( Fig. 2B , left eye). Three total pellets were found to be lodged in this location, and all could be palpated on physical exam. These three pellets were discharged from a 5-m distance. In addition, only one of all pellet discharges was eventually identified in the orbital space ( Fig. 2B , right eye). This was aimed at the orbit from 3 m in the specimen with an enucleated eye.

All 15 pellets shot from 1 m were identified (orbits 4–6). These were virtually all (14/15; 93%) identified intracranially. None of the pellets discharged from 10 cm or 1 m remained in the orbit.

When examining all pellets that came to rest in the intracranial space, 74% (14/19) of the pellets were discharged from 1 m, 16% (3/19) from 10 cm, and the remaining 10% from both 3 and 5 m. Of the 17 that were found in the surrounding orbital structures, 71% (12/17) were discharged from the shortest distance of 10 cm and 18% (3/17) from the farthest distance of 5 m. The remaining two pellets were discharged from 1 and 3 m, respectively.

BB pellets discharged from 1 m were more likely to penetrate the intracranial space as compared with those fired from a shorter distance of 10 cm ( U  = 30.0; p  < 0.01). Pellets discharged from 1 m were also more likely to be found in the intracranial space when compared with those discharged from 3 m ( U  = 8.5; p  = 0.01) and 5 m ( U  = 9.5; p  = 0.004). No difference was noted in final pellet location when comparing pellets discharged from 10 cm versus 3 m ( U  = 19.5; p  = 0.65). Pellet discharge distance was significantly associated with final pellet location (Pearson's chi-square: 29.8; p  = 0.00).

Discussion

Utilizing an alloy steel air rifle at 10 pumps, BB pellets were found to virtually all enter and pass through the orbit in cadaver specimens. This was noted at multiple distances and through various entry points. Of the identified BBs used in this study, 92.5% were identified as penetrating into the orbit, and all but one traveled through the orbit into surrounding regions such as the paranasal sinuses, pterygopalatine and infratemporal fossae, and the intracranial space.

Given that a large number of BB pellets came to rest in the intracranial space, and only one in the orbit itself, it is possible the orbit acted as a corridor of minimal tissue resistance. Pellets discharged from any distance, ranging from 10 cm to 5 m, were able to penetrate deep in the intracranial space. This was more likely when pellets were discharged from a distance of 1 m, compared with a closer distance of 10 cm. The reason for this is unclear. It is possible that from the shorter distance, the incident angle of the ricochet may have guided to a different resting point. In fact, energy may have been sufficient at this distance to allow for multiple ricochets. Those shot from farther distances of 3 and 5 m were also able to penetrate the intracranial space, although less frequently. This reduced frequency could be a result of the fewer number of total pellets discharged at 3 and 5 m in this study. The closer distance of 10 cm demonstrated a greater propensity for a final position in the pterygopalatine fossa, particularly when aimed at the inferior orbit.

Many of the pellets that landed in the surrounding orbital space congregated near or in the pterygopalatine fossa ( Fig. 2C, D ). In the 10-cm, lower eyelid orbit condition, all five pellets landed in the same location behind the pterygopalatine fossa ( Fig. 2C ). This may suggest a pathway through the inferior orbit to the thicker bone of the skull base posterior to the inferior orbital fissure. The angle of the skull base is suggestive of a ricochet action sending pellets to rest at a position 90 degrees to the angle of penetration. One case report showed gun pellets from an unknown firearm lodged in the pterygopalatine fossa after penetrating through the anterior and posterior wall of the maxillary sinus. ⁶ It appears that while the orbit is a point of vulnerability, the thicker bone of the anterior skull base may serve as a barrier.

In addition, the thick bone of the calvarium provides a robust protection for the intracranial structures, and three BB pellets made contact with the orbital rim superiorly, lodged in a subcutaneous plane extracranially ( Table 1 ; orbit 8; Fig. 2B , left eye). The frontal bone acted as a barrier from the BB pellets. Conversely, pellets discharged from any distance penetrated the intracranial space seemingly with ease, along several access routes. This highlights the vulnerability of the orbit as a corridor to the intracranial space. Although intracranial penetration by BB pellets is an anecdotally seen injury in the emergency room setting, its incidence and the constellation of factors producing this outcome are difficult to examine clinically. Our prospective study found intracranial penetration to be a common sequela for close and medium range BB trauma. Often, the history surrounding BB gun injury in a clinical setting lacks details on where the shot was fired from, from what distance, and what type of rifle was used. Many patients we have seen clinically endorse misfiring their own gun. This study shows that intracranial penetration is possible, even from a distance of 5 m, perhaps as a bystander.

Only one pellet remained within the orbit ( Fig. 2B , right eye). This was seen when aimed from a distance of 3 m. Interestingly, this was the one orbit that was enucleated. It is possible that the scarred tissues in the orbital space postenucleation acted as a net for this pellet. Perhaps, the pellet struck the implant and reduced velocity in this manner. It is also important to note that though few pellets came to rest in the orbit, it is likely that in passing through the orbit, significant damage to the periocular structures may have occurred. For obvious reasons, it was not possible to assess the physiology of injury with this experimental design. The purpose of this study was not to determine the degree of globe injury as this has been previously studied, but rather evaluate the orbital and intracranial trauma from BB gun injury.

In our study, 3 of the 40 pellets were not found on CT imaging. These were aimed from farther distances of 3 and 5 m. The two observers used a chalk film to cover the external orbital space between each shot pellet and marked each location of penetration. At the time of the experiment, there was only one site of penetration not visualized by the two observers. Both observers identified all sites of penetration for the 3- and 5-m shooting conditions, so we presume that these shots passed completely through facial soft tissues. While we can hypothesize about the trajectory of each pellet, the exact pathway of each BB is somewhat open to interpretation.

The variance in the BB pellets themselves and firing of the BB gun were also limiting factors. The BB pellets had a large variability in weight listed by the manufacturer; however, this mirrors real-world settings and improves the generalizability of our results. BB guns are also innately less accurate than other firing devices such as a high-power rifle. This introduces variability of BB delivery that is difficult to standardize. Although an experienced shooter fired all shots and aimed for varying anatomical regions, variability in the point of BB entry was inevitable. Our analyses showed that the intended target (caruncle, upper, and lower lid) did not impact the final location of the BB pellet. Thus, while there is some inevitable degree in inaccuracy with the shooting methodology used in this study, it is unlikely that this impacted the final conclusions regarding frequency of intracranial penetration. Additionally, we did not quantify intraocular pressure of each globe prior to BB pellet firing. Given the extent of intracranial injury and entry that was observed, change in intraocular pressure likely has a limited effect on BB penetration.

Furthermore, our study used older human adult cadavers, which may not generalize to children and adolescents who are at highest risk of orbital BB injuries. The rate of change in globe dimensions appears to stabilize after 5 years of age, so our results may be less applicable to this subset of the population. ⁷ While our study population does not directly represent the cohort of younger individuals who are most often affected by BB injuries, it is unlikely that there would be significant biomechanical differences in the tissues or relative anatomic positions of foramina between children older than 5 years and adults. Thus, generalizable conclusions can still be drawn from our study. The variance of soft tissue structures and orbital parameters including the degree of brow projection may also affect the pathway of a BB pellet shot. However, given the ease of intracranial penetration, these factors may not be as significant.

We identified a single experimental study in the literature that also prospectively assessed BB pellet injuries to the orbit of pig eyes. ⁵ The study concluded that the pump-type BB gun (Daisy Powerline Model 880—the same model gun used in this study) produced velocities over 600 feet per second, which can be fatal. ⁵ This type of air rifle is one of the most common BB guns purchased in the United States, although airsoft guns and paintball guns are increasingly popular among children. Ocular injuries from airsoft guns include corneal abrasion and hyphema. ⁸ Paintball pellets can also cause ruptured globe injury. ⁹ Eye protection to prevent or significantly reduce ocular injury from BB gun trauma cannot be overly emphasized. The American Society for Testing and Materials has created a “Standard Specification for Eye Protective Devices for Paintball Sports” updated in 2021 to enforce this. ¹⁰ Effective eyewear should provide both frontal and side protection to prevent injury. ¹¹ In addition to protecting the globe, eye protection can theoretically prevent the BB from transorbital access to the intracranial space. Future studies could assess the ability of various safety glasses to stop BB pellets and measure the protective efficacy on orbital and intracranial penetration.

This prospective study used human adult cadaveric orbits to demonstrate that BB pellet injuries can be severe, resulting in penetration of the periorbital and intracranial spaces. BB guns can readily fire pellets that end up in the intracranial space from both close and farther distances up to 5 meters. Additional studies quantifying the efficacy of various safety goggle designs and materials for preventing ocular, orbital, and intracranial injuries may be of use.

Funding Statement

Funding This work is supported by an Unrestricted Grant from Research to Prevent Blindness, Inc., to the Department of Ophthalmology at UCLA.