Among 59 porcine eye specimens submitted to paintball impact in the 1- to 13-J range, 10 (17%) disengaged completely from the firmly attached manometric probe and surrounding orbital gelatin mount (
Table 1). The orientation of the optic nerve stalk within these mounts approximated the anatomic norm closely, since each globe was secured to a cannula passing through the Perspex orbital apex, and the anterior globe margin was centralized and appropriately positioned flush with the orbital rim. In each instance of globe expulsion, the paintball penetrated the orbit adjacent to the globe, producing rotation and eventual globe repulsion, dramatically evident on high-speed film images (
Fig. 2). Because of this unique behavior, these eyes were excluded from the analysis of the effects of blunt direct ocular trauma published previously.
8 These expulsive events occurred on a regular basis, on multiple days throughout the test series, despite a uniform and highly standardized method for preparing the ballistic mounts. Clinically, traumatic transection of the optic nerve is rarely accompanied by complete globe evulsion,
2 since the suspensory ligaments, extraocular muscles, periorbita, lids, orbicularis muscle, and associated skin tend to retain the severely distracted globe. Nevertheless, within the confines of the orbit, substantial rotation and anteroposterior displacement of the globe can occur with minimal restraint from the extraocular musculature.
9 Thus, although the expulsion of relatively unrestrained, gelatin-mounted porcine eyes was not deemed to be representative of actual clinical pathology, the dramatic forces involved could not be reasonably overlooked.
To better understand the nature of this less common but experimentally frequent phenomenon, supercomputer modeling was performed with CTH, programmed with 10-mm off-center, oblique impact and with 20-mm tangential impact, with the paintball speed set at 300 m/s. Globe-expulsive events were observed when the paintball impact was sufficiently offset from the globe center and when pyramidal bony orbital constraints were in place, but not when these were removed. Twenty-millimeter tangentially offset (grazing) impact sheared the nerve flush with the globe via a strain rate effect within 260 μs, with minimal posterior displacement and just 5° of globe rotation (
Fig. 3; top series). Ten-millimeter offset midperipheral impact produced compressive globe distortion and posterior displacement, followed by rebound and tractional nerve avulsion 10 mm behind the lamina, after 700 μs and 20° of globe rotation (
Fig. 3; bottom series).
Impact thresholds for various clinically relevant forms of intraocular injury have been detailed previously.
8 In 10 eyes that underwent total avulsion from the orbit, the paintball had entered the orbit adjacent to the globe, forcefully propelling the globe anteriorly once the loculated paint mass proceeded posteriorly beyond the equator of the globe. The speed and force of this “ricochet repulsion” was dramatically evident on high-speed film images.
Supercomputer modeling experiments produced comparable results when orbital constraints were in place, but not when these were removed. Using CTH, The 300-ft/s, 10- and 20-mm offset impacts were run to simulate the ballistic experiments wherein the paintball ejected the eye (
Fig. 4 showing a 2-D plane cut through the center of the eye and acrylic holder to allow for viewing into the interior of the simulated orbit). The predictions of the numerical models appear to be borne out in the statistical analysis of the associations of impact location and globe displacement in the empirical porcine ballistic studies.
Table 1 shows the ocular distortion and intraorbital globe displacement and rebound data for the 59 eyes submitted to paintball trauma. The energy and histopathologic data for these impact studies have been presented previously.
8 Posterior intraorbital displacement (measured from the optic nerve head) was 4.76 ± 0.76 mm (SEM) for the 26 eyes sustaining direct central impact, three times higher than that observed among the ten ejected globes (1.57 ± 0.59 mm;
P = 0.016). Of some interest was the difference in posterior displacement of the ejected globes documented as having sustained an offset “hit” as opposed to a “miss” (4.85 ±0.75 vs. 0.75 ± 0.23;
P < 0.0001), indicating the likelihood of different mechanisms of nerve disengagement such as that suggested by the numerical modeling exercises. The maximum change in axial length of the globes during impact was correspondingly different, with those sustaining central impact shortening by 6.4 ± 0.38 mm, while ejected offset–miss globes shortened by only 0.9 ± 0.7 mm (
P < 0.0001). Further comparisons of this kind may be readily performed incorporating the energy and pathologic data published previously, reaffirming that eyes likely to sustain the greatest rotational or rebound traction trauma to the optic nerve were often those demonstrating the least intraocular pathology.