Schematic drawing of the pendulum
apparatus impacting the orbit.
"Globe intact" orbit demonstrating
orbital floor and medial wall fractures following soft tissue exenteration
(drop height, 1.1 m). Methylene blue outlines fracture boundaries.
Coronal computed tomographic scan
of orbits demonstrating similar floor and medial wall fractures between the
"globe intact" orbit (A) and the "balloon apparatus" orbit (B) at a drop height
of 0.7 m.
Rhee JS, Kilde J, Yoganadan N, Pintar F. Orbital Blowout FracturesExperimental Evidence for the Pure Hydraulic Theory. Arch Facial Plast Surg. 2002;4(2):98–101. doi:
From the Department of Otolaryngology and Communication Sciences, Medical
College of Wisconsin (Drs Rhee and Kilde); and the Department of Neurosurgery,
Zablocki Veterans Affairs Medical Center (Drs Yoganadan and Pintar), Milwaukee,
Background The mechanism of injury and the underlying biomechanics of orbital blowout
fractures remain controversial. The "hydraulic" theory proposes that a generalized
increased orbital content pressure results in direct compression and fracturing
of the thin orbital bone.
Objective To examine the pure hydraulic mechanism of injury by eliminating the
factor of globe-to-wall contact and its possible contribution to fracture
thresholds and patterns.
Materials and Methods Five fresh human cadaver specimens were used for the study. In each
cadaver head, 1 orbit was prepared to mimic the normal physiologic condition
by increasing the hypotony of the cadaver globe to normal intraocular pressure
(15-20 mm Hg) with intravitreous injection of isotonic sodium chloride solution
(saline). The second orbit served as a "hydraulic control," whereby the globe
and orbital contents were exenterated and replaced by a saline-filled balloon
at physiologic intraocular pressure. A 1-kg pendulum measuring 2.5 cm in diameter
was used to strike the cadaver heads. Drop heights ranged from 0.2 m to 1.1
m (1960 mJ to 10 780 mJ energy). Each head was struck twice, once to
each orbit. Direct visualization, high-speed videography, and computed tomographic
scans were used to determine injury patterns at various heights between the
Results A fracture threshold was found at a drop height of 0.3 m (2940 mJ).
Fracture severity and displacement increased with incremental increases in
drop height (energy). Fracture displacement, with herniation of orbital contents,
was obtained at heights above 0.5 m (4900 mJ). Isolated orbital floor fractures
were obtained at lower heights, with medial wall fractures occurring in conjunction
with floor fractures at higher energies (≥6860 mJ). The globe intact side
and balloon (hydraulic control) side showed nearly identical fracture patterns
and levels of displacement at each drop height.
Conclusions This study provides support for the "hydraulic" theory and evidence
against the role of direct globe-to-wall contact in the pathogenesis of orbital
blowout fractures. In addition, the orbital floor was found to have a lower
threshold for fracture than the medial wall. Preliminary threshold values
for fracture occurrence and soft tissue displacement were obtained.
THE MECHANISM of injury of orbital blowout fractures has long been an
area of debate for otolaryngologists, ophthalmologists, and plastic surgeons.
Sequelae such as diplopia, enophthalmos, hypophthalmos, and sensory disturbances
in the distribution of the infraorbital nerve are well-recognized morbidities
of orbital fractures.1-2 An elucidation
of the underlying mechanisms of orbital fractures is not only of academic
interest, but also of clinical importance in terms of prevention and treatment.
The 2 most accepted theories in the mechanism of orbital floor fractures
fall into 2 categories: increased hydraulic pressure with direct compression
force vs transmitted buckling force via the orbital rim.3-4
Recently, Erling et al5 resurrected an older
theory on the etiology of orbital blowout fractures first espoused by an ophthalmologist
in 1943.6 They proposed that the responsible
mechanism of fracture is a direct globe-to-wall contact; that is, posterior
movement of the globe, in response to an external force, results in a fracture
upon direct contact with an orbital wall.
Of the 3 biomechanical theories, the "buckling mechanism" has been the
most extensively studied.4, 7 The
"hydraulic" theory, as first advocated by Smith and Regan3
in 1957, proposed that a generalized increased orbital content pressure resulted
in direct compression of the orbital floor, thereby causing fracturing of
the thin bone. This classic cadaver study is often quoted as the definitive
article purporting the "hydraulic" theory. Interestingly, this study reports
only a single cadaver specimen and is flawed by questionable methodology and
The role of direct globe-to-wall contact remains unclear, as no experimental
studies have been conducted to validate or disprove this theory. The purpose
of this study was to further examine the "hydraulic" mechanism of orbital
wall fractures, and to elucidate the role of the globe in the pathogenesis
of orbital blowout fractures.
Fresh, unfixed cadaver specimens were used for the study. Each of the
5 cadaver heads was prepared by having 1 orbit as the physiologic control
("globe intact") and the other orbit replaced with a "balloon apparatus,"
following removal of the entire soft tissue orbital contents. This intra-cadaver
setup allowed comparison of the orbits at the same drop heights, thereby eliminating
potential differences between cadaver specimens.
In the "globe intact" group, the orbital floor was examined endoscopically
through an anterior maxillary wall antrostomy to ensure that the floor was
intact prior to impact. The orbit was then prepared by injecting the globe
with isotonic sodium chloride (saline). A Schiotz handheld tonometer was used
to measure intraocular pressure. Approximately 4.5 mL of injected saline was
necessary to achieve normal physiologic intraocular pressure (15-20 mm Hg).
The purpose of the "balloon apparatus" was to eliminate the globe as
a possible factor in the pathogenesis of orbital fractures, yet still maintain
the soft tissue content needed for the transmission of hydraulic forces. For
this experimental group, the orbits were prepared by removing the globe and
the orbital soft tissue contents. The orbital floor and medial walls were
then closely inspected to ensure that they were intact and not disrupted by
the exenteration process. The "balloon apparatus" consisted of 2 layered lambskin
condoms with an 8F Kao feeding tube tied into the open end of the condoms.
The feeding tube allowed for the injection of 25 mL of saline, which correlated
to the volume of the orbital contents. The balloon was placed into the orbit
and the eyelid was loosely sutured over the balloon. The free end of the catheter
was attached to a 30-mL syringe that was filled so that the top of the open
column measured 15 cm above the center of the balloon, to approximate physiologic
The cadaver heads were then impacted using a pendulum apparatus. The
pendulum consisted of a 1-kg iron cylinder measuring 2.5 cm in diameter. The
specimen was aligned with the pendulum to ensure that the globe alone would
be impacted (Figure 1). High-speed
videography was used to document that impact occurred directly on the orbital
contents and not the orbital rims. Each cadaver head was struck twice, once
to each orbit.
Following impact, in the "globe intact" group, the orbital contents
were carefully removed leaving the periosteum intact to support the fractured
orbital walls. The fracture patterns were then carefully painted with methylene
blue and digital pictures were taken to document extent of injury (Figure 2). For the "balloon apparatus" group,
the balloon was deflated and removed, and fracture patterns were documented
as in the "globe intact" group. Computed tomographic scans were performed
on 2 of the cadaver heads prior to orbital exenteration to document fracture
patterns before manual manipulation of the orbital contents.
Drop heights were chosen initially to ensure occurrence of fracture,
based upon previous studies.8-9
The drop heights were subsequently decreased until a fracture threshold was
determined. Energy delivered by the pendulum was calculated using the following
equation: U = mgh, where U = energy (millijoules), m = mass (grams), g = gravitational
acceleration (meters per seconds squared), and h = height (meters). Drop heights
ranged from 1.1 m (10 780 mJ) to 0.2 m (1960 mJ).
Analysis of the high-speed videography revealed consistent impact of
the pendulum to the orbital contents without impact to the orbital rim or
other surrounding bones. The orbital rim was noted to be intact in all of
the orbits. In general, fracture severity and bony displacement were greater
with increasing drop heights. Orbital fractures were found with drop heights
of 0.3 m or greater (2940 mJ). No orbital fractures occurred at 0.2 m (1960
mJ) (Table 1).
Bony displacement, with herniation of orbital contents, was obtained
only at heights above 0.5 m (4900 mJ). Fractures of the floor were obtained
at lower heights, with medial wall fractures occurring in conjunction with
floor fractures at higher energies (≥6860 mJ). For each drop height, the
fracture patterns were nearly identical between the "globe intact" and the
"balloon apparatus" groups.
Computed tomographic scans obtained in 2 of the cadaver specimens (drop
heights of 0.5 m and 0.7 m) revealed nearly identical fracture patterns between
the "globe intact" and "balloon apparatus" groups (Figure 3). Fractures patterns revealed by computed tomographic scanning
were found to match the findings on direct orbital wall inspection following
soft tissue exenteration.
The underlying biomechanics in the pathogenesis of orbital blowout fractures
continues to be a controversial subject. The "hydraulic" theory, advocated
by Smith and Regan3 in 1957, proposed that
a generalized increased orbital content pressure resulted in direct compression
of the orbital floor, thereby fracturing the thin orbital bone. In spite of
their conclusions, the "hydraulic" theory was not substantiated by their study,
and no quantitative measures of force needed to create the fractures were
Fujino4 later disputed this theory and
proposed that a direct compression force or buckling force transmitted via
the orbital rim was the causative factor for orbital floor fractures. This
theory of bone conduction or "tsunami" mechanism of injury was first proposed
by LeFort10 and Lagrange8
at the turn of the century. In his series of dried human cadaver experiments
during the mid-1970s, Fujino4 convincingly
demonstrated the occurrence of orbital floor fractures with direct blows to
the orbital rim without fracturing the orbital rim itself. Phalen et al7 later corroborated Fujino's findings by repeating
his experiment on fresh cadaver heads, taking into account the soft tissue
coverage of the orbital rim.
A more recent cadaver study by Waterhouse et al9
appeared to validate both the "hydraulic" and "buckling mechanism" theories.
The study showed that each mechanism could independently produce orbital floor
fractures via different biomechanics. In addition, they found that the fracture
patterns differed between orbits struck on the orbital rim vs directly on
the globe. The "hydraulic" mechanism produced larger fractures with involvement
of the floor and medial wall, where herniation of orbital contents was frequent.
The "buckling mechanism" produced smaller fractures involving the mid-medial
floor, without significant orbital content herniation. However, the role of
the globe in the pathogenesis of these fractures was not investigated.
Erling et al5 recently resurrected an
older theory on the etiology of orbital fractures that was first described
by Pfeiffer6 in 1943. They proposed that the
responsible mechanism of fracture is direct globe-to-wall contact; that is,
posterior movement of the globe, in response to an external force, results
in a fracture upon direct contact with an orbital wall. In their analysis
of computed tomographic scans of clinical cases of orbital blowout fractures,
they found that the size of the orbital wall displacement often exactly fit
the size of the globe. However, no corroborative experiments were conducted,
and no such studies exist to support this theory, to date.
This study examined the role of the globe in the pathogenesis of orbital
fractures by devising a mechanism to replace the orbital contents of 1 orbit
with a "globeless" apparatus, and thereby effectively remove the globe as
a factor in the fracture equation. Our findings demonstrate nearly identical
fracture patterns between the "globe intact" and "balloon apparatus" groups
at every drop height. Though these findings do not definitively rule out the
possibility of the globe contributing to orbital fractures, this similar pattern
of injury between the orbits suggests that direct globe-to-wall contact is
an unlikely factor in the pathogenesis of orbital fractures. Furthermore,
this study provides additional evidence for the "hydraulic" mechanism in its
Our data suggest that the threshold of energy needed to produce blowout
fractures appears to be in the range of 1960 to 2940 mJ. Our study is the
first to attempt to quantify the threshold using the fresh cadaver model for
the "hydraulic" mechanism.9, 11-12
Our threshold value is similar to that obtained by Green et al,11
the only in vivo study, human or nonhuman primate, to quantify the threshold
for orbital fractures. In their study with Macaca fascicularis monkeys, they found that fractures of the orbital floor were consistently
produced at and above 2080 mJ. Interestingly, in Phalen and colleagues'7 human cadaver study of the "buckling mechanism," a
similar threshold at 2500 mJ was found.
This threshold energy needed to cause orbital fractures is relatively
low compared with the kinetic energies of missiles generated during sporting
events. Racquet sports, in particular, can generate projectiles with enormous
energies; a tennis ball (0.57 kg) hit at 100 mph will generate 11 400
mJ, while a squash ball (0.24 kg) hit at 130 mph will generate 8100 mJ. Even
the human fist, once set into motion, is capable of causing orbital fractures
with energy measurements ranging from 900 to 3700 mJ.11
Not surprisingly, the severity of the fracture, in terms of size, number
of walls involved, and degree of orbital content herniation, correlated with
increasing drop heights. Though larger number of drops need to be performed
at each height before any definitive conclusions can be drawn, it was interesting
to note that higher energies were needed to cause fractures of the medial
wall. In addition, no medial wall fractures occurred without concomitant floor
fractures. Though we are cautious in extrapolating our experimental findings
to the clinical setting, our study suggests that in the absence of orbital
rim or facial skeleton trauma (pure hydraulic mechanism), the orbital floor
is more prone to fracture than the medial wall, and that isolated medial orbital
wall fractures cannot occur without trauma to the surrounding bony framework
(eg, orbital rim and nasal bone).
In the clinical realm, orbital fractures may occur by way of the "buckling"
or "hydraulic" mechanism, with a combination of the 2 mechanisms being the
most likely scenario. A potential future study examining fracture patterns
and threshold values with combined blows to the orbital rim (medial, inferior,
superior, and lateral) and to the orbital soft tissue would allow for investigation
into the additive effects of both mechanisms of injury.
The shortcomings of this study include the small number of specimens
and the cadaver model itself. Larger numbers of drops are needed at each height
to apply statistical analyses and thereby provide more powerful data. In addition,
the age of the cadavers is well above the average of the population that usually
sustains these injuries. However, previous studies have demonstrated that
age is not thought to have a significant effect on facial bone strength.13
In conclusion, this study provides validation of the pure "hydraulic"
theory and evidence against the role of direct globe-to-wall contact in the
pathogenesis of orbital blowout fractures. In addition, the orbital floor
was found to have a lower threshold for fracture than the medial wall, and
preliminary threshold energy values for fracture occurrence and bony displacement
Accepted for publication July 26, 2001.
Corresponding author and reprints: John S. Rhee, MD, Department of
Otolaryngology and Communication Sciences, MCW Clinic at Froedtert West, Milwaukee,
WI 53226 (e-mail: firstname.lastname@example.org).