PENETRATING CRANIOCEREBRAL INJURIES
20
EXPERIMENTAL MISSILE
WOUNDING OF THE BRAIN
Gunshot wounds to the brain account for
more than 30,000 civilian deaths yearly in
the United States.35 For soldiers engaged in
ground combat, head wounds (largely brain)
have accounted for almost half of all deaths
since World War 11. 15 The author's interest
in the problem of brain wounding developed
after serving as a neurosurgeon in Vietnam.
On returning to the United States and review-
ing war wound literature, it became apparent
that little was known about brain wounding
from missiles in modern biologic terms.*
Amazingly, fewer than 25 laboratory papers
had ever been published in which a brain
wound from a missile had been replicated
in an anesthetized animal and the effects on
the brain studied. Realizing that experimental
research on missile injury to the brain was
an unmet need that would benefit both sol-
diers and civilians, the U.S. Army funded
the author's laboratory.
Experiments cited from the LSU Brain Trauma Labora-
tory were performed by Drs. George Sarna, Dan Torbati,
Joseph Soblosky, and Deepak Awasthi under U.S. Army
Contracts DAMD17-83-C-3145 and DAMD17-86-C-6098.
This peer-reviewed, Department of Defense-funded con-
tract was terminated before completion by Congress,
w hich acted at the behest of animal rights activists. Opin-
ions expressed are those of the author and are not to be
construed as an official Department of the Army position.
*References 7-12, 14, 20- 25, 27, 37, and 39.
From the Department of Neurosurgery, Louisiana State University Medical Center, New Orleans, Louisiana
NEUROSURGERY CLINICS OF NORTH AMERICA
VOLUME 6 •NUMBER 4 •OCTOBER 1995
1042-3680/ 95 $0.00 + .
Michael E. Carey, MD
PHYSICAL CORRELATES OF
MISSILE WOUNDING
Missile wounding can only be understood
in terms of the physical interactions between
the missile and tissue through which it passes.
The primary destructive effect of a missile in-
teracting with tissue is caused by the crushing
action of the missile itself.26• 29• 31• 34• 40 Besides
this direct action, however, a missile moving
in a water or tissue medium generates three
distinct types of pressures within the medium
which it transits. 29• 30• 31
Juxtamissile Pressure
Extremely high pressures (thousands of at-
mospheres) are generated just in front of and
at right angles to a moving missile owing to
flow of the surrounding medium around the
missile. The pressure is confined to the mis-
sile's immediate vicinity.
Longitudinal Strong Shock
Wave Pressure
When a missile strikes an object, a high-
pressure compression front or shock wave is
formed that moves spherically away from the
629
630 CAREY

missile strike-point. A blunt missile may cause
a strong shock wave if it impacts with a strike
velocity as low as 60% of the velocity of sound
in the medium. The velocity of the shock wave
in tissue approximates the velocity of sound in
water (1460 m/ sec) (Fig. 1). The longitudinal,
strong shock pressure wave initially has a steep
front but its shape and duration may be
changed with propagation through a medium,
particularly one that is not homogeneous, such
as tissue. As it travels, therefore, the steep front
of a strong shock wave becomes less steep and,
finally, assumes the characteristics of an ordi-
nary pressure wave (see subsequently). In wa-
ter, the peak pressure of a strong shock wave
may exceed 80 atm, but the entire strong shock
wave event lasts only about 10 µ,sec . In the
absence of a fluid-to-air interface, strong pres-
sure waves are thought not to cause tissue
damage because their brief duration prevents
transfer of sufficient energy to surrounding
tissues to distort and tear them. 34
Pressure Waves from Kinetic Energy
Transfer (Ordinary Pressure Waves)
When a missile passes through tissue, ki-
netic energy is transferred to nearby tissue ele-
ments, which are propeled radially from the
missile track, thereby increasing pressure in
this tissue. A large subatmospheric temporary
cavity forms directly behind the missile as a
consequence of the radial movement of tissue
away from the missile path. When the elastic
limit of the outwardly displaced tissue is
reached, the tissue falls inward from whence
it was displaced. This cycle may be repeated
several times before the deranged tissue comes
to rest around the permanent track created by

4.3 a im
d. p
OP1

OP2
OP3

+ v
11111 111
Time (msecl Pamb
x

Figure 1. Missile (a), black circle travelling through medium (hatched area). The "strong" shock wave (b) moves with
the speed of sound in the medium. This shock wave has an extremely sharp front represented by the vertical line in
graph (d). Its velocity through the medium is indicated by v. The extremely high overpressure ( p) lasts only microseconds.
Owing to kinetic energy transfer, tissue molecules are thrust radially away from the missile's path, creating an overpressure
laterally, OP1, and a near vacuum (white area) directly behind the missile. Once thrust aside, the molecules of the medium
tend to fall back to their original position (first collapse) but then move laterally again to set up another overpressure in
the medium, OP2. This cycle is repeated, OP3, until the kinetic energy is expended. These overpressures are indicated
in oscilloscope trace (c). Overpressures caused by radial displacement of tissue molecules are "ordinary" pressure waves
that last milliseconds and cause tissue damage about the wound track. The authors hypothesize that they also cause
damage at a distance in the brain.
 EXPERIMENTAL MISSILE WOUNDING OF THE BRAIN
the missile. The oscillating, outward and in-
ward rush of tissue creates long-lasting (mil-
liseconds) lower amplitude (20 to 30 atm)
pressure waves, which propagate widely
throughout the medium and may distort and
damage tissues at a distance from the site of
actual missile injury (see Fig. 1).29• 30· 31
ENERGY OF WOUNDING
The kinetic energy (KE) of any missile equals
0.5 X missile mass x (missile velocity) 2• The
amount of tissue damage caused by a mis-
sile may be correlated with the amount of
missile kinetic energy deposited within the
tissue by the missile.34· 40 The missile kinetic
energy of deposit (KEd) = KEen - KEeX! where
KEen = missile kinetic energy of entry and
KEex = missile kinetic energy of exit. If a mis-
sile is retained within the tissue, Eex = 0, and all
missile energy is deposited within the tissue. If
a missile exits, only part of its kinetic energy
is deposited. The same amount of energy may
be deposited by a small-mass missile with a
large velocity or a large-mass missile with
small velocity. Kinetic energy deposited by a
missile is partitioned between that which di-
rectly crushes tissues in its path and that which
displaces tissues adjacent to the missile track. 26
The latter is less destructive than the former
owing to the elastic properties of the displaced
tissue, which may be deformed without being
irrevocably destroyed (L. Sturdivan, Edge-
wood Arsenal, Edgewood, MD, personal com-
munication, 1988).
MISSILES CAUSING BRAIN WOUNDS
Most wounds in modern combat are caused
by small fragments not bullets. 6 Eighty percent
of all wounds sustained at Dunkirk were
caused by "splinters" 5 mm or less in diame-
ter,13 and missile fragments weighing 110 to
220 mg inflicted most fatal brain wounds in
U.S. soldiers in Vietnam (Carey ME, unpub-
lished data from analysis of casualties sur-
veyed by the Wound Data and Munitions Ef-
fectiveness Team, WDMET, in Vietnam).
A bullet has a streamlined shape that allows
it to maintain a high velocity over many hun-
dreds of yards. Bullet velocity decreases lin-
early with distance. Because of their high veloc-
ities, military bullets possess enormous energies,
631
which, when deposited within the brain, create
destructive ordinary pressure waves that may
widely disrupt the brain. WDMET data pro-
vided an instance of fatal and severe brain
wound by an AK47 round fired from 400 yards.
Owing to their irregular, nonaerodynamic
shape, fragments lose velocities and kinetic en-
ergies exponentially over distance. Lighter
fragments lose velocity and energy with partic-
ular rapidity. Yards from the site of an ord-
nance explosion, the typical shell fragment has
lost much of its initial kinetic energy. Although
such a fragment may penetrate the skull, the
amount of tissue destruction caused directly
by its crushing effect may be small. The ordi-
nary pressure waves that it sets up may not
widely disrupt brain tissue because it does not
deposit much energy within the brain. If its
trajectory does not cross vital brain structures,
such a missile, producing relatively little brain
destruction, may be compatible with life, al-
though it may cause a discrete neurologic injury.
This fundamental difference between bullets
and fragments accounts for the fact that ap-
proximately 90% of all brain wounds operated
on by neurosurgeons in Vietnam were caused
by fragments. Brain wounds from bullets that
the neurosurgeon saw were undoubtedly
caused by tangential wounds in which only
part of the bullet's energy was deposited
within the brain or by spent rounds (as by
ricochet) in which bullet kinetic energy was
greatly decremented and bullet energy depos-
ited within the brain also reduced.
In the past among U.S. civilians, it was not
uncommon to see nonfatal 0.22 caliber bullet
wounds to the brain because the kinetic energy
associated with such bullets is low (in the ki-
netic energy range of some shell fragments).
Currently among U.S. civilians, bullet wounds
to the brain are highly lethal because higher-
velocity (and higher-energy) bullets are being
used. Forty-four magnum bullets have veloci-
ties and energies in the range of military
bullets.
PRIOR EXPERIMENTS ON
BRAIN WOUNDS IN
ANESTHETIZED ANIMALS
During the 19th century, clinicians and phys-
iologists had an intense interest in physiologic,
anatomic, and clinical events associated with
632 CAREY
brain trauma. They recognized that blunt
trauma to the head caused respiratory and cir-
culatory changes in addition to coma. 43 Using
anesthetized dogs, Horsley,32 in 1893, con-
cluded that a bullet wound to the brain could
be fatal by virtue of apnea but the patient might
recover if respiratory support were given. He
hypothesized that the respiratory arrest was
produced by brain stem anemia consequent to
missile-induced increased intracranial pres-
sure. In 1943, Webster and Gurdjian,47 wound-
ing anesthetized mongrel dogs, essentially con-
firmed these observations on apnea. When
animals were wounded through a closed cra-
nial vault, respiratory effects became more pro-
found with increasing missile energy. As had
Horsley, these investigators hypothesized that
the increased intracranial pressure produced
by the bullet affected medullary respiratory
centers. This respiratory effect could be pre-
vented by craniectomy before wounding.
Gerber and Moody,27 in 1972, developed a brain
missile wound model wherein an anesthetized
rhesus monkey was wounded through a 7-mm
diameter trephine by a pellet gun. They con-
cluded that the most important single variable
correlating with death was decreased carotid
blood flow (and, by inference, cerebral blood
flow). Respiratory effects were not prominent.
Clemedson and colleagues19 wounded anes-
thetized dogs with 7.62- or 5.56-mm ammuni-
tion with muzzle velocities ranging from 800
to 1090 m/ sec. Wounding effects resulted from
the projectile itself plus pressure waves, which
caused lacerations, contusions, and intracran-
ial bleeding. All direct cranial hits were fatal.
Lower-velocity missiles produced limited
brain lacerations about the wound track,
whereas higher-velocity missiles caused se-
vere brain lacerations. A frontal wound was
seen to cause medullary hemorrhage, clearly
a distant effect of wounding. A test model ap-
paratus that measured pressures generated by
a missile passing through a water-filled plastic
container showed that peak pressure was a
function of the square of the missile velocity;
projectile weight was of little import.
Djordevic and associates25 used anesthetized
dogs to study the intracranial pressure effects
of brain wounding by 870-mg spheres. Missile
velocities ranged from 122 to 785 m/ sec. In-
creased postwounding intracranial pressure
was attributed to intracranial bleeding and an
increase in cerebrospinal fluid outflow resis-
tance. Vasodilation was not thought to be a
factor in intracranial pressure elevation.
Brown and Crockard and colleagues7, 20- 24 ex-
panded Gerber's model, again using anesthe-
tized rhesus monkeys. The brain wound was
caused by a 310-mg missile fired through a 7-
mm occipital trephine. Missile energies from
0.49 to5.02Joules (J) were created by u sing mis-
sile velocities from 56 to180 m /sec. Wounds of
5 J were highly lethal primarily by their respira-
tory effects, whereas those of 0.49 J produced
few physiologic changes. Wounds of 1.26 J
(missile velocity = 90 m / sec) produced tran-
sient bradycardia, respiratory slowing, and in-
creased tidal volume. Arterial blood gases,
however, were unchanged. In subsequent ex-
periments, the prominent effects of missile
wounding on brain stem function were noted,
but apnea was not notable in these investiga-
tors' standard 1.26 J wound. A highly accurate
prediction of animal survival following a brain
wound of this energy could be made from cere-
bral blood flow (radioactive xenon clearance),
the cerebral metabolic rate of lactate, and mean
arterial blood pressure from 30 to 60 minutes
after brain wounding. Follow-up experiments
indicated that brain wounding in spontane-
ously breathing rhesus monkeys was associ-
ated with decreased cardiac output (CO) and
that this, in turn, might be the underlying cause
of diminished cerebral blood flow associated
with brain wounding. 37 In contrast to the earlier
works of Horsley, Webster, and Curd jian, these
experiments focused on cerebral blood flow
rather than respirations as the critical factor in
a fatal missile wound to the brain.
Maynard and coworkers39 modified the
Gerber-Crockard model wounding anesthe-
tized baboons w ith 4.8-mm, 411-mg steel
sphere through a trephine placed posterior to
the left coronal suture. This resulted in a bi-
frontal wound track. Intracranial pressure was
monitored by bilateral piezoelectric transduc-
ers. In various experiments, missile impact
energies ranged from 7.2 to 29.2 J. With
missile passage, a brief (0.5 msec), energy-
dependent intracranial overpressure devel-
oped that ranged from 68 to 330 p si (470 to
2270 kPa) in the left-sided pressure transducer
and 105 to 570 p si (720 to 3900 kPa) on the right.
Ten of 11 wounded animals failed to main-
tain spontaneous respirations after wounding.
Respiratory complications were common with
missile energies above 7 J, and as higher mis-
sile energies were used postwounding, apnea
 EXPERIMENT AL MISSILE WOUNDING OF THE BRAIN
was the rule, either immediately after wound-
ing or within an hour. All apneic animals were
placed on a respirator. It could not be deter-
mined whether the respiratory changes were
a function of missile energy deposit or whether
they occurred in an all-or-none way above
some threshold energy deposit.
All wounded animals also exhibited hy-
potension, bradycardia, electrocardiagram
changes, and CO reductions. The post-
wounding CO decrements were related to a
so-called scaled energy term (missile impact
energy I animal mass), which provided a linear
relationship between scaled energy and CO
reduction. A simpler analysis made by the au-
thor from Maynard's data also shows a linear
trend between missile impact energy and CO
reduction in nine animals in which missile ve-
locity was measured: impact energy 7 J, CO
reduction 12% (N = 2); missile impact energy
17 J, CO reduction 32% (N = 5); missile impact
energy 28 J, CO reduction 59% (N = 2). Neural
reflexes, both sympathetic and parasympa-
thetic, were thought to underlie the cardiovas-
cular changes because they occurred within
seconds of missile impact. These investigators
noted that both sympathetic and parasympa-
thetic outflows are modulated by hypothala-
mic and brain stem areas, and histologic exam-
ination of the wounded brains revealed that
both the hypothalamus and the brain stem ex-
hibited perivascular ring hemorrhages from
the indirect effects of missile energy.
DEVELOPMENT OF THE LOUISIANA
STATE UNIVERSITY BRAIN
WOUND MODEL
In considering conducting further experi-
ments on brain missile wounds, there appeared
little point in studying massively destructive,
nonsurvivable brain wounds such as are
caused by a high energy of deposit missile such
as a military bullet. Rather the most efficacious
laboratory model should replicate a survivable
brain wound such as would be caused by a
lower energy of deposit missile (i.e., a small
shell fragment or possibly a low-energy bullet,
e.g., a 0.22 caliber). The excellent experiments
of Gerber, Crockard, Brown, Levett, and May-
nard* provided many valuable insights con-
cerning brain wounding, and the use of a pri-
*References 1- 4, 7-12, 14, 19-25, 27, 32, 37, 39, and 47.
633
mate with a human brain configuration (brain
stem more or less at right angles to the midaxial
plane of the cerebral hemispheres) was appeal-
ing. Wounding through a trephine and a large-
mass, slow-moving, heavy projectile did not,
however, accord with the usual fragment
wound of the brain encountered in combat in
which the majority of survivable brain wounds
are caused by small, higher-velocity fragments
that enter through the frontal or parietal bone.
Thus, the criteria for a more realistic and useful
model, in the author's opinion, were (1) wound
through an intact skull; (2) small, high-velocity
missile; (3) frontal entry; and (4) cerebral hemi-
sphere missile trajectory.
The laboratory gun was made by Robert
Carpenter of the Edgewood Arsenal, Edge-
wood, Maryland, who thought that the small-
est feasible m issile that could readily be fired
would be a 2-mm diameter sphere. Such a steel
sphere weighs 31.7 mg. Although most mis-
siles causing brain wounds are not spherical,
spheres are often used in wound simulation
experiments to simplify mathematical analysis
of missile-media interactions. Furthermore, the
author could envision little difference in brain
injury variables that he intended to study
whether the brain injury itself was caused by a
round, smooth missile or an irregularly shaped
one, as would be the case for an actual fragment.
The experimental set-up is depicted in Fig-
ure 2. A solenoid release valve activated by
direct current was placed between the gas stor-
age tank and the gun. This provided missile
velocities closely related to the helium charge
(Fig. 3). Missile accuracy was within 1 to 2 mm
over the short range.
ANIMAL MODEL USED IN THE
LOUISIANA STATE UNIVERSITY
BRAIN WOUNDING EXPERIMENTS
Several considerations determined the
choice of experimental animal: appropriate-
ness relative to physiologic variables to be
studied; mass of the missile relative to the mass
of the brain; use of the "lowest" animal possi-
ble; and cost, not only of the animal, but also
of radioactive isotopes and microspheres re-
quired for the intended studies. This last factor,
in particular, mandated a small animal. Al-
though the rat would have been the least con-
troversial, employing a 31.7-mg missile to in-
jure a 1.0-g brain would have u sed a missile
634 CAREY

Gas
Reservoir
Gun/Barrel B IRespirator I
Release Laser ~
Valve
Head

~~~~~~<-- 0.5 m --7~~

2mm Diameter Velocity Gate
Steel Sphere
Stereotaxic
Frame

I Physiograph I
Figure 2. Experimental set-up. Helium is used as the propellant because of its great coefficient of expansion. Direct
current is used to trigger the solenoid release valve because alternating current gave inconstant missile speed, presumably
because of different valve opening rates . A laser beam allowed very precise aiming. Missile velocity was measured by
the velocity gate from which missile KE was calculated . The stereotaxic frame is about 0.5 meter from the velocity gate.

mass 374 times too large relative to the mass
of missiles documented to cause fragment
wounds to the brain in humans. An inexpen-
sive animal with a larger brain was required.
The cat was deemed more appropriate because
of its approximately 25-g brain and because it
has been widely used in neurologic research,
including cerebral edema and cerebral blood
flow experiments that the author and col-
leagues were contemplating.28 • 38• 41 • 44 Although

Velocity (m-s-1)
0 175 250 300 330 370 400
3.0

Ci)
~
:J 2 .0
0
:::?..
>-
...Cl
Q)
c:
w
(,)

:; 1.0
c:
~

Figure 3. Gas reservoir pressure, sphere veloc-

ity, and calculated sphere KE; 5 shots each point.
0 Note the restricted possible wound energies
(dotted lines) because the study focused on a
0 100 200 300 400 500 600 non-fatal missile wound. Skull penetration re-
quired 0.7 J; 67 percent of cats died from apnea
Pressure (psi) at 2.4 J.
 EXPERIMENTAL MISSILE WOUNDING OF THE BRAIN
the rat has little cerebral white matter, the cat,
as does the human, has ample white matter in
which brain edema occurs. Furthermore, in
contrast to dogs, cats' heads have a more uni-
form shape. After fixation in a stereotaxic frame,
one could have a high expectation of producing
a consistently placed brain wound. Because
many animals would have to be used for the
proposed studies, primates, although scien-
tifically appealing, were deemed inappropriate.
Against using the cat were the facts that the
missile-to-brain ratio was still 7 to 15 times too
large, the cat has a bony tentorium not present
in humans, and an "in-line" brain is not the
human brain configuration. Nevertheless, the
author believed that the positive scientific as-
pects of using the cat far outweighed the nega-
tives. Consultants at the Edgewood Arsenal
believed that the missile-brain scaling factor
would not invalidate results.
Pentobarbital anesthesia was selected be-
cause it has been widely used in brain edema
and cerebral blood flow studies. 28• 38• 41 • 44 With
pentobarbital, both brain blood flow and me-
tabolism are simultaneously reduced; flow and
metabolism, thus, remain coupled. (Michenfel-
der J, personal communication, 1989).
ANIMAL PREPARATION
Laboratory personnel deeply sedated the
cats with intraperitoneal pentobarbital and
then infiltrated a groin with 1% local lidocaine
(Xylocaine), performed a cut down, and can-
nulated a femoral vein for further intravenous
pentobarbital as needed. Other required vas-
cular lines were then placed and an endotra-
cheal tube inserted. A 5-cm midline scalp inci-
sion was then made and the skin reflected
laterally. The outer wall of the right frontal
sinus was then removed because preliminary
experiments showed that the small, light-
weight missile merely glanced off of the slop-
ing outer wall of the cat's right frontal sinus
without bony penetration. Removal of the
outer wall allowed the sphere to strike the
nearly vertical, intact posterior bony sinus wall
and achieve skull perforation and brain pene-
tration. To monitor intracranial pressure, a
Camino intracranial pressure monitor was
placed through a small left parietal trephine.
The electroencephalogram was sometimes
monitored via small steel skull screws. For
studies evaluating the effect of missile wound-
635
ing on respirations, the animals breathed on
their own before wounding even though some
prewounding Pa02 values were in the 60 to
80 mm Hg range. After wounding, the cats
were given respiratory assistance by ventilator
as needed. For cerebral blood flow studies in
which control of PaC02 was critical, the ani-
mals were paralyzed with gallamine and ar-
tifically ventilated. After surgical preparation,
the animals were placed in a stereotaxis frame
facing the barrel of the pellet gun. Before
wounding, anesthesia was ensured to be suf-
ficient so the animal would feel no pain.
Because a nonfatal brain wound was de-
sired, missile trajectory was placed in a cere-
bral hemisphere, away from midline and brain
stem structures. The initial trajectory was para-
sagittal, 2 to 3 mm from the falx cerebri. Missile
impact on the bony tentorium almost directly
over the brain stem, however, resulted in
prominent cardiovascular changes in many
animals that would not be analagous to a hu-
man brain wound. To minimize this effect,
missile trajectory was altered to slant 20 de-
grees laterally away from the midline. Serial
coronal sections showed that after penetrating
the right frontal tip, the 2-mm missile track was
usually in cerebral hemisphere white matter
between the outer corner of the lateral ventri-
cal and the cortical gray matter. Because the
author wished to relate all observed physio-
logic effects to missile energy deposit within
the brain, it was a requisite that the missile not
exit the skull (KEex = 0). The author deter-
mined that a threshold missile energy, 0.5 to
0.07 J, was needed to produce frequent skull
perforation through the posterior wall of the
right frontal sinus. Missile energy of 0.9 J en-
sured bone perforation and brain penetration
almost 100% of the time. Missiles w ith energies
as high as 2.4 J rarely exited the skull but
caused severe respiratory problems in a high
percentage of cats (see Fig. 3). As had been
noted by prior investigators, significant intra-
cranial bleeding was not usually a problem. 22• 27
In this respect, too, the experimental paradigm
also mirrored the clinical situation. 18
RESULTS AND DISCUSSION
OF FINDINGS
It quickly became apparent that wounding
through the intact skull produced prominent
respiratory effects even at low missile energies
636 CAREY

even though the missile track was at least
2 cm from the brain stem. 17• 17" To conserve
animals, if a cat became apneic after wound-
ing, it was placed on a respirator, which was
removed approximately every minute to see
whether spontaneous respirations had re-
turned. If respirations had not returned, venti-
latory support was continued until they did
or for 2 hours, when the experiment was termi-
nated. Six minutes of apnea is fatal for a cat,
so in a retrospective analysis relating missile
energy to apnea, the author considered that
cessation of respirations for longer than 6 mi-
nutes to be a fatal apnea. The author's experiments
showed that missile-induced fatal apnea is a direct
function of missile energy deposit. A 0.9 J missile
caused fatal apnea in 14% of cases, a 1.4 Jmis-
sile produced fatal apnea 40% of the time, and
a 2.4 J missile caused fatal apnea 67% of the
time (Fig. 4). Surprisingly, but in accord with
Horsley' s original observations,32 following respira-
tory support of up to 2 hours, many apneic cats
recovered respirations, lived, and were neurologi-
cally intact (Fig. 4). These respiratory effects of
brain wounding accorded with those of Web-
ster and Gurdjian, who also wounded anesthe-
tized animals through an intact skull. 47 It ap-
pears that significant effects of lower energy
of deposit missiles on respirations were
masked in Gerber' s 27 and Crockard' s20• 22 rhesus
monkey and Maynard' s 39 baboon experiments
owing to their use of a skull trephine, which,
as Webster and Gurdjian noted, attenuates
missile effect on respirations. Thus, Crockard
and colleagues noted no significant respiratory
effects with a 1.3 J missile, whereas almost 40%
of the author's animals sustained a fatal apnea
with a 1.4 J wound. Wounding through a tre-
phine required missile energies of 5 to 7 J to
cause a prominent respiratory effect. 20• 22• 39 The
author's data suggest that in a cerebral hemispheral
missile wound not involving midline or brain stem
structures, the crucial physiologic effect determin-
ing life or death is the missile's effect on medullary
respiratory centers. Provision of timely respiratory
support may allow recovery of spontaneous respira-
tions and prevent death . Missile wounding also
may reduce CO, but significant CO reductions

ca
Cl>
c
c.
<a: 0.8
CJ)
c
"O
c ( 12)
::I
0
:;:
-"'
0
a.
0.6

c
E
<O 0.4
/\

-ca
ca
LL

0 0.2

->
:aca (35)
Figure 4. Missile energy and the probability of
causing fatal apnea. Fatal apnea (dotted line)
.c was highly dependent on missile energy of de-
...0
a. o.o~~~~~-+"-'-'-':.t..l..l...~i..U..'-'-'-''-'-"-'-.1..1..1.1..1..1..1.1.~
posit. With timely artificial ventilation , many
cats resumed breathing. Immediate respiratory
o 244 297 396 - mis support lowers the probability of permanent
apnea (solid line) . The likelihood of permanent
0 0.94 1.40 2.48 - Joules apnea (shaded area) could theoretically be de-
creased by measures (i.e ., drugs) that en-
Missile Velocity and Energy hance recovery of respiratory mechanisms.
 EXPERIMENTAL MISSILE WOUNDING OF THE BRAIN
occur at missile energies significantly greater
than those required to cause apnea. Clinically,
unless timely respiratory resuscitation occur-
red, the overriding effect of apnea would make
any possible effects of CO reduction moot.
The author's experiments also suggest that
after bone penetration, a rather narrow limit
of allowable missile energies exists (in the au-
thor's model 0.7 to 2.4 J) before the increased
missile energy deposit gravely affects respira-
tory function and produces death in a high
percentage of animals. By inference, humans
sustaining a missile wound to a cerebral hemi-
sphere also have a relatively narrow window
of missile energies compatible with life before
the indirect effects of the propagated ordinary
pressure waves irreversibly affect medullary
respiratory centers. Pearce and coworkers,42 in
a series of parietal impact wounds, showed
that a 900-mg sphere required about 15 J to
penetrate the parietal bone of a prepared hu-
man skull. A 90 J sphere penetrated the skull
and transited both cerebral hemispheres,
whereas a 250 J sphere severely disrupted the
entire calvarium and produced a large cone-
shaped temporary cavity in the gelatin filling
the skull. This latter wound would clearly be
fatal. Because missiles that have the energy to
transit two cerebral hemispheres are known to
produce severe brain wounds, clinically, one
may estimate the range of survivable hemi-
spheral missile energy deposits in humans to
be approximately 15 to 90 J or less. If the per-
missible range of missile energies in the cat
model (0.7 to 2.4 J) is relevant, the range of
survivable missile energies for humans sus-
taining a cerebral hemisphere wound would
be between 15 and 51 J (maximum KEd =
3.4 x penetration energy). These concepts are
diagramed in Figure 5. Additional effects such
as intracerebral bleeding, which may also con-
tribute to death, are not considered.
In subsequent experiments, it was demon-
strated that, following a brain wound produc-
ing a brief transient apnea, some cats rapidly
regained a normal respiratory rate and sur-
vived, whereas others resumed breathing again
but at a reduced rate.46 These latter cats eventu-
ally developed permanent apnea 10 to 20 mi-
nutes after their injury. CO reductions were not
a prominent feature in these fatally wounded
cats. Mean brain stem regional cerebral blood
flow (rCBF) (microspheres) in cats with respira-
tory depression fell from a control of 33 : :': : 5
mL/100 g/min to 25 : :': : 4 mL/100 g/min. This
637
mild rCBF reduction after wounding resulted
in a brain stem blood flow well above that re-
quired to produce feline ischemia (15 mL/100/
min),33 so, contrary to Horsley' s hypothesis, it
does not appear that brain stem microischemia
is responsible for apnea following a hemisph-
eral missile wound. What biologic factors dis-
tinguish animals that succumb to apnea follow-
ing brain wounding from those that do not
remain to be elucidated so that treatment of im-
paired respirations could be devised.
The author also evaluated brain stem nu-
clear and hypothalamic catecholamines fol-
lowing a 2.4 Jbrain wound, which would have
been expected to produce death from apnea
in greater than half of the animals. 45 Despite
the severe brain stem effects associated with
such a wound, and although epinephrine was
significantly reduced in raphe nuclei and sev-
eral hypothalamic and brain stem nuclei as
well, none of the brain stem or hypothalamic
biogenic amines was totally depleted. The mo-
naminergic systems in these structures ap-
peared to reflect a generalized stress response
pattern seen not only with missile wounding,
but also with elevated intracranial pressure.
This pattern suggests that a severe brain
wound likely to cause early death does not do
so by totally disrupting the hypothalamic and
brain stem biogenic amine system. Whether
the response of and relative maintenance of the
catecholamine system after brain wounding is
indicative of other chemical mediator systems
in the brain stem is unknown.
At this point, the author hypothesizes, much
as did Horsley and Webster and Gurdjian, that
the fatal effect of a hemispheral missile wound
that does not severely disrupt the brain or di-
rectly damage vital, midline structures is due
to propagated ordinary pressure waves from
missile energy deposit. The propagated ordi-
nary pressure waves most notably affect respi-
ratory nuclei, which lie just under the floor of
the fourth ventricle in close proximity to the
cerebrospinal fluid compartment. Whether
this location makes them particularly suscepti-
ble to mechanical or chemical insults transmit-
ted through or by the cerebrospinal fluid re-
mains to be seen. Possibly the superficial
location of respiratory neurons or the long,
complex respiratory circuits cause the respira-
tory system to be preferentially impaired fol-
lowing even low-energy missile wounds and
other forms of head injury. Cardiovascular
function, although disturbed by head injury,
 1.0 -;---d---/--1 ------------
-
.r:.
ctS
1
I
I
I

-
Q) I I
c : 1b' &.::-.(}' :
0 I ..f;?u ~~ I
> 0.5
:!::::
I I
I ~?
:cctS I
I
I
I
• ·

..0 I I
I I
....0 I I
a. I I
I

0 15(P) 50 100 ? (Joules)

A Missile Energy of Deposit

1.0

-
.r:.
ctS

-
Q)
c
0
~ 0.5
:cctS
..0
0
....
a.

0 15 (P) 50 100 ? (Joules)

B Missile Energy of Deposit

1.0 --------------
I /i1

-
.r:.
ctS
Q)
:~
i/
-
c
0
>
:!:::: 0.5
I
I
I
I
:cctS I
I ?t-+?
..0
I ,-J
....0 C:-t?I
a. Ef9
I

0 15(P) 50 (Joules)

c Missile Energy of Deposit

Figure 5. See legend on opposite page
638
 EXPERIMENTAL MISSILE WOUNDING OF THE BRAIN 639

80

** **
75

...
-Q)

al
;::: 70

?ft.
Figure 6. Percent of brain water in white
matter about the missile track. Tissue so-
dium also rose while tissue potassium
remained constant, characteristic of va-
65
sogenic brain edema. Note that at 48
hours a 2-mm rod injury (a low energy
cutting injury) was not associated with
any increase in water. means :±: S.D.;
** = p < 0.01 ; * = P < 0.05 of control.
White bar = control ; diagonal bar = 0.9 J; 0 6 24 48 72 168
cross-hatched bar= 1.4 J; black bar=
rod. Hours After Brain Lesion

is not usually destroyed. The location of car-
diovascular control nuclei deeper in the core
of the medulla, away from the floor of the
fourth ventricle, may explain why increased
missile energy is required to cause CO de-
creases. If a brain-wounded individual sur-
vives a bout of transient apnea or if, in the
future, respiratory resuscitation becomes pos-
sible, specific treatment may then be required
to prevent concomitant or a later occurring
CO reduction.
Brain edema occurs following traumatic
brain injury, and the author and others 16
sought to measure its significance in brain
wounding from missiles. Brain edema (wet
weight-dry weight method) following missile
wounding appeared confined to the region
about the missile track, w as not severe, and
had the characteristics of vasogenic edema
(VBE). Following missile w ounding (as in cold
injury) VBE peaked in 48 to 72 hours then
receded to be almost gone in a week (Fig. 6).
With the author's sampling techniques, no
difference in the amount of VBE or its time
course between a 0.9 and a 1.4 J missile could
be ascertained. Lack of a distinct graded re-
sponse in the production of VBE in the cerebral
hemisphere probably resulted because missile
energies allowable for a nonfatal wound were
too restricted and the variance in the amount
of VBE produced by a missile of either energy
was too great to allow separation of the VBE
effect between missiles of such close energies.
The lack of a distinct, graded response in this
hemispheral pathobiologic event is in great
counterdistinction to missile effect on medul-

Figure 5. A, Hypothesis of how a missile passing through a nonvital portion of a cerebral hemisphere produces death.
Approximately 15 J are required for skull penetration by a missile. Missile energy deposit close to 15 J causes local
laceration about the missile track only. As missile energy of deposit becomes greater, medullary effects become more
pronounced. Based on feline studies, possibly significant respiratory effects in humans may be seen at 50 J energy
deposit. Fatal brain wounds from medullary dysfunction causing apnea may occur without great brain disruption . The
brain and skull are severely disrupted with massive amounts of missile energy deposit. B, Owing to proximity of the brain
stem itself, one may imagine that less energy of deposit would be required to cause death with a posterior Iossa wound
that did not directly injure the medulla. It may be anticipated that approximately 15 J would be required to perforate the
suboccipital area as has been shown for the parietal bone. C, A missile perforating the thin skull base may do so with little
energy. If the missile tracked though the brain stem it would require very little additional energy to be fatal. R = respiratory;
C = cardiovascular.
640 CAREY

11 ·t··,.·· ·• · · ··•· · · ·· . . ... ..

10

9
. .I
8
. -J

....Q)
0
(.)
7

6
;!
/II
(/) 5

0 1--t--t-+-~~+-~--+~~-+-~~+-~--l
Ao 3 6 9 12 15 rn 21 24 27 30 37 44 51 58 65
Days (Postinjury)
8

l . . ....I . ... . ...l

7
IJ./l I /

l!!
86
(/)

5 ...... . . . . !
/rI
4+--+--1--+--+--+--lf-- ·+--+--+--+-~~t--~-+-~~-+-~--+~~--f
0 3 6 9 12 15 18 21 24 27 30 37 44 51 58 65
B
Days (Postinjury)
Figure 7. A, Before wounding, cats were trained to traverse an elevated beam to achieve a tuna fish reward . Ability was
scored on an 11 point scale before and after wounding and treatment with GM1 ganglioside. Balance beam performance
appeared to return more quickly following treatment. 8 , A normal cat will withdraw its fore limb when a front paw guard
hair touches a surface even though the cat cannot see the surface. This non-visual placing test was evaluated by a 4 point
score before and after wounding and treatment with GM1 ganglioside that seemed to enhance placement performance.
Square = control ; circle= SM-1.

lary respiratory function, in which a distinct,
graded physiologic effect occurred with mis-
siles of even close energies, 0.9 and 1.4 J.
Coagulopathy has been well described fol-
lowing head injury. 36 The author and others
showed that 2 hours following missile wound-
ing, total platelets are significantly decreased,
but by 6 hours, they had returned to normal.
Fibrinogen levels began decreasing 4 hours
after injury and were still depressed at 6 hours.5
 EXPERIMENTAL MISSILE WOUNDING OF THE BRAIN
DRUG STUDIES
The experimental missile wound model has
been used to test agents that might improve
physiologic or overall brain function after in-
jury. Using Gerber's and Crockard's missile
wound model, it was demonstrated that man-
nitol given shortly after wounding markedly
improved cerebral perfusion pressure, cere-
bral blood flow, and CMR02 consumption out
of proportion to any observed intracranial
pressure reduction, possibly through direct
effects on blood flow and metabolism.7· 8• 9
Dimethyl sulfoxide also improved post-
wounding mean arterial blood pressure, cere-
bral perfusion pressure, cerebral blood flow,
and CMR0 2 through many possible but not
clearly understood mechanisms. 10• 11
In the author's laboratory, Soblosky and
Sarna developed a detailed cat behavioral par-
adigm that allowed accurate behavioral assess-
ment of wounded cats after they had awak-
ened from anesthesia. Although their behavior
was depressed for a few days following anes-
thesia and injury, no cat ever appeared in pain.
The neuroprotective effect of GM 1 ganglioside,
given intraperitoneal at 20 mg/kg 10 minutes
after wounding and continued daily for 10
days, was evaluated. These experiments were
terminated by animal rights activists acting
through Congress before statistical signifi-
cance could be achieved. The missile wound-
GM1 ganglioside data are presented, however,
to show that it is possible to use the feline
missile wound model to test the effects of neu-
roprotective drugs and because GM 1 ganglio-
side is suggestively neuroprotective after mis-
sile wound injury, both accelerating and
enhancing recovery of neural function (Fig. 7).
SUMMARY
If a missile penetrates a cerebral hemisphere
and does not severely disrupt the brain or tran-
sit a vital brain structure, it is hypothesized
that the indirect effect of ordinary pressure
waves set up by the interaction of missile and
tissue and that impinge on brain stem respira-
tory nuclei determines life or death. The likeli-
hood of fatal apnea is a direct function of mis-
sile energy of deposit within the brain. With
brain wounding, a reduction in CO may also
occur, but missile energy required to produce
a significant CO decrease is in excess of that
required to produce respiratory problems. Un-
641
less the individual managed to survive a pe-
riod of apnea or respiratory resuscitation oc-
curred, the effects of apnea would overshadow
any CO decreases. Although transmitted ordi-
nary pressure waves might interfere with the
reticular activating system within the brain
stem and produce persistent coma, specific
long-lasting neurologic defects from a missile
wound usually result from direct missile dam-
age to the cerebral cortex or cortical projections.
In designing treatments for missile wounds
of the brain, two distinct entities must be kept
in mind: the brain stem and the cerebral cortex.
To decrease the immediate mortality from
brain wounding, prompt treatment has to be
devised to aid dysfunctional respiratory nuclei
and possibly cardiac control nuclei. To de-
crease long-term neurologic morbidity, drug
therapy has to be instituted to help injured
cerebral cortical neurons for days to weeks
after wounding. Totally different strategies
and drugs may be needed to treat the brain
stem as opposed to the cerebral cortex.
References
1. Aleksandrov LN, Dyskin EA, Ozeretskovskii LB, et al:
Mechanisms of gunshot wound of the cranium and brain
(experimental study). Vestn Khir (USSR) 104:81, 1970
2. Allen IV, Crockard HA, Maynard RL, et al: Pathological
Changes Following Experimental Medium Velocity
Penetrating Head Injury. New York, Raven Press, 1982
3. Allen IV, Kirk J, Maynard RL, et al: An ultrastructural
study of experimental high velocity penetrating head
injury. Acta Neuropathol (Berl) 59:277, 1983
4. Allen IV, Scott R, Tanner JA: Experimental high-
velocity missile head injury. Injury 14:183, 1982
5. Awasthi D, Rock WA, Carey ME, et al: Coagulation
changes after an experimental missile wound to the
brain in the cat. Surg Neural 36:441, 1991
6. Beyer JC, Arisma JK, Johnson OW: Enemy ordnance
material. In Coats JB Jr (ed): Wound Ballistics. Wash-
ington, DC, Office of the Surgeon General, 1962
7. Brown FD, Johns LM, Crockard HA, et al: Response
to mannitol following experimental cerebral missile
injury. in Popp AJ, Bourke RS, Nelson LR, et al (eds):
Neural Trauma. New York, Raven Press, 1979, p 281
8. Brown FD, Johns L, Jafar JJ, et al: Detailed monitoring
of the effects of mannitol following experimental head
injury. J Neurosurg 50:423, 1979
9. Brown FD, Johns LM, Jafar JJ, et al: Systemic and
cerebral hemodynamic response to mannitol after ce-
rebral missile injury. Surg Forum 29:525, 1978
10. Brown FD, Johns L, Mullan S: Dimethylsulfoxide ther-
apy following penetrating brain injury. Ann NY Acad
Sci 41:245, 1983
11. Brown FD, Johns LM, Mullan S: Dimethylsulfoxide in
experimental brain injury, with comparison to manni-
tol. J Neurosurg 53:58, 1980
 642 CAREY
12. Brown FD, Johns LM, Mullan JF: DMSO treatment in
an experimental missile injury. Surg Forum 30:444, 1979
13. Burns BD, Zuckerman S: The wounding power of
small bomb and shell fragments. British Ministry of
Supply Advisory Council on Scientific Research and
Technical Development, RC 350, 1942
14. Calica AB, Crockard HA, Mullan JF: Predicting injury
to the brain-the death equation. Surg Forum 27:
470, 1976
15. Carey ME: Learning from traditional combat mortality
and morbidity data used in the evaluation of combat
medical care. Milit Med 152:6, 1987
16. Carey ME, Sarna GS, FarrellJB: Brain edema following
an experimental missile wound to the brain. J Neuro-
trauma 7:13, 1990
17. Carey ME, Sarna GS, Farrell JB, et al: Experimental
missile wound to the brain. J Neurosurg 71:754, 1989
17a. Carey ME: Exp erimental missile wounding. In Nara-
yan RK, Wilberger J, PolvashockJ: Neurotrauma. New
York, McGraw Hill, 1995
18. Carey ME, Young HF, Mathis JL: The neurosurgical
treatment of craniocerebral missile wounds in Viet-
nam. Surg Gynecol Obstet 135:386, 1972
19. Clemedson CJ, Falconer B, Frankenberg L, et al: Head
injury caused by small-caliber, high velocity bullets.
An experimental study. Z Rechtsmed (West Germany)
73:103, 1973
20. Crockard HA, Brown FD, Calica AB, et al: Physiologi-
cal consequences of experimental cerebral missile in-
jury and use of data analysis to predict survival.
J Neurosurg 46:784, 1977
21. Crockard HA, Brown FD, Calica AB, et al: ICP CVR
and cerebral metabolism follo wing experimental mis-
sile injury. In Beks JWF, Bosch DA, Brock M (eds):
Intracranial Pressure III. New York, Springer-Verlag,
1976, p 73
22. Crockard HA, Brown FD, Johns LM, et al: An experi-
mental cerebral missile injury model in primates.
J Neurosurg 776, 1977
23. Crockard HA, Brown FD, Trimble J, et al: Evoked
potentials, cerebral blood flow and metabolism follow-
ing cerebral m issile trauma in monkeys. Surg Neural
7:281, 1977
24. Crockard HA, Johns L, Levett J, et al: "Brainstem"
effects of experimental cerebral trauma. In Popp AJ,
Bourke RS, Nelson LR, et al (eds): Neural Trauma.
New York, Raven Press, 1979, p 19
25. Djordevic M, Lofgren J, Steiner H, et al: Intracranial
pressure effects of missiles. In Beks JWF, Bosch DA,
Brock M (eds): Intracranial Pressure III. New York,
Springer-Verlag, 1976, p 79
26. Fackler ML: What's Wrong with the Wound Ballistics
Literature and Why. San Francisco, Letterman Army
Institute of Research, 1987
27. Gerber AM, Moody RA: Craniocerebral missile injury
in the monkey: An experimental physiological model.
J Neurosurg 36:43, 1972
28. Ginsberg MD, Budd WW, Welsh FA: Diffuse cerebral
ischemia in the cat: I. Local blood flow during severe
ischemia and recirculation. Ann Neural 3:482, 1978
29. Harvey EN, Korr IM, Oster G, et al: Secondary damage
in wounding due to pressure changes accompanying
the passage of high velocity missiles. Surgery
21 :218, 1947
30. Harvey EN, McMillen JH: An experimental study of
shock waves resulting from the impact of high velocity
missiles on animal tissues. J Exp Med 85:321, 1947
31. Harvey EN, McMillen JH, Butler EH, et al: Mechanism
of wounding. In Coats JB Jr (ed): Wound Ballistics.
Washington, DC, Office of the Surgeon General,
143-235, 1962
32. Horsley V: The destructive effects of small projectiles.
Nature 50:104, 1894
33. Hossman KA, Hossman V, Tayagi S: Microsphere
analysis of local cerebral and extracerebral blood flow
after complete ischemia of the cat brain for one hour.
J Neural 218:275, 1978
34. Janzon B: High Energy Missile Trauma. Goteborg, Mi-
nab / Gotab, 1983
35. Kaufman HH: Civilian gunshot wounds to the head.
Neurosurgery 32:962, 1993
36. Kaufman HH, Moake JL, Olson JD, et al: Delayed and
recurrent intra cranial hematomas related to dissemin-
ated intravascular clotting and fibrinolysis in head
injury. Neurosurgery 7:445, 1980
37. Levett JM, Johns LM, Replogle RL, et al: Cardiovascu-
lar effects of experimental cerebral missile injury in
primates. Surg Neural 13:59, 1980
38. Long DM, Maxwell RE, Choi KS, et al: Multiple thera-
peutic approaches in the treatment of brain edema
induced by a standard cold lesion. In Reulen HJ, Schur-
mann K (eds): Steroids and Brain Edema. New York,
Springer Verlag, 1972, p 87
39. Maynard RL, Cooper GJ, Evans VA, et al: Pathophysio-
logical effects of exp erimental medium velocity pene-
trating head injury. CDE Technical Note No: 567, CDE,
Parton Down, Salisbury Wilts, England 1983
40. Owen-Smith MS: High Velocity Missile Wounds. Lon-
don, Edward Arnold, 1981
41. Pappius HM, McCann WP: Effects of steroids on cere-
bral edema in cats. Arch Neural 20:207, 1969
42. Pearce BP, Stauber MC, Watkins FP: Physical effects
of the penetration of head simulants by steel spheres.
NATO, Defence Research Study Group II, 1983
43. Polis A: Recherches experimentales sur la commotion
cerebral. Rev Chir 274:645, 1884
44. Shalit MN, Cotev S: Interrelationship between blood
pressure and regional cerebral blood flow in experi-
mental intracranial hypertension. J Neurosurg 40:
594, 1974
45. Soblosky JS, Rogers NL, Adams JA, et al: Central and
peripheral biogenic amine effects of brain missile
wounding (BMW) and increased intracranial pressure.
J Neurosurg 76:119, 1992
46. Torbati D, Jacks AF, Carey ME, et al: Cerebral cardio-
vascular and respiratory variables after an experimen-
tal brain missile wound. J Neurotrauma 9(suppl
1):5143, 1922
47. Webster JE, Gurdjian ES: Acute physiological effects
of gunshot and other penetrating wounds of the brain.
J Neurophysiol 6:255, 1943
Address reprint requests to
Michael E. Carey, MD
Department of Neurosurgery
Louisiana State University Medical Center
1542 Tulane A venue
New Orleans, LA 70112