Low-energy ligamentous injuries at the Lisfranc joint frequently occur in
younger and athletic patients and can lead to substantial long-term
disability1-4.
There are two patterns of low-energy Lisfranc injuries: transverse and
longitudinal. Both patterns are thought to involve the interosseous Lisfranc
ligament between the first cuneiform and the base of the second
metatarsal5,6.
For the purpose of operative decision-making, these injuries are classified
clinically as either stable or unstable. Unstable injuries are generally
thought to be optimally treated with open reduction and internal fixation. The
threshold for a diagnosis of an unstable injury is often stated to be =2 mm
of diastasis at the interval between the first cuneiform and the base of the
second metatarsal as seen on plain
radiographs2,4,7-9.
However, no single diagnostic method has been found to consistently confirm
the diagnosis of an unstable injury. The most accepted images of diagnosing
unstable injuries of the Lisfranc joint are either weight-bearing
radiographs1-4
or manual stress radiographs, which can be made with use of an
anesthetic1,10.
Computed tomography has also been
suggested11, but in
the absence of a clear diastasis the presence of periarticular bone fragments
provides only indirect evidence of ligament injury and cannot be used to
diagnose instability. Therefore, findings on a computed tomography scan are
often thought to be unreliable for the same reason that non-weight-bearing
radiographs are thought to be unreliable—i.e., because there is no
stress on the injured joints.
Controversy remains regarding which test is most sensitive for diagnosing
an unstable
injury1-4,10,11.
To our knowledge, no previous investigators have evaluated this question.
Furthermore, we have found no specific guidelines regarding how much weight is
needed to make a weight-bearing radiograph effective for diagnosis or
regarding which specific manual stress radiographs are best for diagnosing
these two types of Lisfranc injuries. Also, while both the longitudinal and
the transverse forms of the low-energy Lisfranc injury are thought to involve
disruption of the interosseous first cuneiform-second metatarsal ligament, it
has not been determined if additional ligamentous injury must be involved to
produce the unstable forms of these injuries. Further clarification of the
actual pathological characteristics of these injuries could provide the
surgeon with a deeper understanding of the nature of instability and may lead
to more accurate and reliable reduction and fixation methods.
Ligamentous stability of the Lisfranc articulation is provided by the
Lisfranc ligament (the interosseous first cuneiform-second metatarsal
ligament), the interosseous ligament between the first and second cuneiforms,
and the plantar cuneometatarsal ligaments. These include the first
cuneiform-first metatarsal ligament as well as the plantar ligament between
the first cuneiform and the second and third
metatarsals12-14,
which originates from the inferolateral surface of the first cuneiform and
divides into two bands. The weaker, deep band attaches to the base of the
second metatarsal, and the stronger, superficial band attaches broadly to the
base of the third metatarsal. There is no plantar ligament between the second
cuneiform and the second metatarsal. Figure
1 depicts the critical stabilizing ligaments of this joint
complex.
Preliminary work with cadavera led to our hypothesis that sectioning of the
Lisfranc ligament alone would not be sufficient to reliably create either a
transverse or a longitudinal pattern of instability. We also observed that
weight-bearing loads did not lead to substantial instability. This finding was
in agreement with the observation in another study, in which the
tarsometatarsal ligaments were sectioned but no displacement was observed with
weight-bearing3.
One purpose of this study was to define which ligaments must be disrupted
to produce each of the patterns of an unstable Lisfranc injury. The other
purpose was to compare the effectiveness of weight-bearing radiographs with
that of injury-pattern-specific stress radiographs for detecting each pattern
of instability. We hypothesized that an abduction stress radiograph would
demonstrate instability between the first cuneiform and the second metatarsal
and between the second cuneiform and the second metatarsal in the transverse
pattern and that a radiograph made with an adduction stress across the midfoot
would demonstrate instability between the first cuneiform and the second
metatarsal and between the first and second cuneiforms in the longitudinal
pattern.
Ten fresh-frozen lower-extremity specimens from five cadavers were divided
into two groups of five specimens (one from each cadaver) to study the two
forms of low-energy Lisfranc injuries. The cadavers were of three female and
two male donors, whose average age was sixty-six years (range, forty-six to
ninety-one years) at the time of death. Specimens were kept frozen at
—40°C and were thawed twenty-four hours prior to use. Each specimen
was cut through the proximal part of the tibia and dissected according to a
specific protocol to remove all of the overlying dorsal soft tissues at the
first and second tarsometatarsal and first intercuneiform joints, while
preserving the interosseous and plantar ligaments. The plantar fascia and long
plantar ligaments were undisturbed, thereby preserving the natural tie-rod
mechanism of the arch. A drill was used to make small holes in the dorsal
cortex of the second metatarsal, first cuneiform, and second cuneiform, and
threaded metallic markers were placed for later radiographic measurements. The
markers were placed at the proximal-medial aspect of the base of the second
metatarsal, the medial-distal border of the second cuneiform, and the
lateral-distal border of the first cuneiform
(Fig. 2). In half the
specimens, which would later be included in the longitudinal group as
described below, a second, medial marker was placed in the first cuneiform to
ensure that the measured distances were orthogonal to the first-second
cuneiform joint. No specimen was noted to have preexisting pathological
changes either on direct visual inspection of the joints studied or on
non-weight-bearing radiographs made prior to ligament sectioning.
The specimens were mounted on a custom testing apparatus with use of a
threaded rod cemented into the tibia such that the total height of the
specimen and rod was 56 cm. The apparatus was designed to allow rotation in
the sagittal and axial planes. For weight-bearing radiographs (described
below), iron barbell flat-plate weights were added to the top of the apparatus
to directly transfer the load to the specimen. The apparatus could be set to
allow low-friction pistoning against a plastic x-ray-cassette cover. These
conditions allowed for natural pronation and internal rotation of the tibia
with weight-bearing. The limb was loaded with the axis of the weight slightly
(7 cm) anterior to the ankle joint, simulating the normal position of the body
center of gravity. For the stress radiographs, the specimens could be
suspended just above the x-ray-cassette cover without limiting rotation in any
of the three planes. Abduction and adduction stress tests were performed by
one evaluator (J.F.). Both the weight-bearing and the stress radiographs were
anteroposterior radiographs made with 15° of cephalad tilt of a portable
x-ray machine 36 in (91.4 cm) above the x-ray cassette at 60 kV for 0.04
second.
After preparation and mounting of the specimen, baseline radiographs with
and without 222.4 N (50 lb) of simulated body weight were made of all
specimens. During preliminary trials of the specimens, we noticed, with direct
visualization of the involved joints, that only a relatively small amount of
force was needed to demonstrate the maximum tarsometatarsal displacement. We
chose 222.4 N for the weightbearing load to dynamically induce observable
movement at the midfoot joints. In preliminary trials of four specimens
(unrecorded data), each from a different cadaver, the addition of more weight
did not cause further displacement. To our knowledge, there is no published
standard for weight-bearing on an injured lower extremity; however, we thought
that most patients with an acute painful injury could not endure more than
this much weight.
Radiographs were made of all mounted specimens prior to any ligament
sectioning. The interosseous first cuneiform-second metatarsal ligament was
then sectioned in all ten specimens, and radiographs were made with
weight-bearing and with abduction and adduction stress. Five of the specimens
then became the transverse injury group, and the plantar ligament between the
first cuneiform and the second and third metatarsals was sectioned at the
plantar aspect of the second cuneiform-second metatarsal joint. The
weight-bearing and abduction stress radiographs were repeated. The remaining
five specimens became the longitudinal group, and the interosseous ligament
between the first and second cuneiforms was sectioned. Weight-bearing and
adduction stress radiographs were then made.
The injury-specific stress tests were developed during our preliminary
testing. The abduction stress test, for detecting an unstable transverse
injury, was performed with the ankle in plantar flexion. The second metatarsal
head was grasped and abducted while counter pressure was provided by the other
hand grasping the calcaneus and applying thumb pressure over the
calcaneocuboid area. The adduction stress test, for detecting an unstable
longitudinal injury, was also performed with the ankle in plantar flexion. The
first metatarsal head was grasped, and the first ray was adducted and
pronated. Counter pressure was provided by the other hand grasping the
calcaneus and applying thumb pressure medially over the navicular with the
hindfoot slightly inverted to relax the abductor hallucis.
The radiographs were digitized (Vidar Systems, Herndon, Virginia), and the
marker distances were measured digitally by a musculoskeletal radiologist who
was blinded to the study sequences. Each measurement was made three times with
the digital software, and the average of the three was recorded. There was
never more than a difference of 1 pixel (0.0353 mm) among the measurements
because the digitized radiographs were analyzed on an invisible grid (SIENET
MagicView 1000 VE42; Siemens Medical Solutions, Malvern, Pennsylvania), which
removed any subjective variables of interpretation. Absolute distances were
recorded. Changes in joint diastasis were determined by subtracting
measurements on each radiograph from those on the initial weight-bearing
radiograph of the specimens with no ligaments sectioned. A threshold of 2 mm
was used to define clinical instability, and each radiograph was then judged
to be either diagnostic or nondiagnostic for instability. Descriptive
statistics including means and standard deviations were qualitatively analyzed
because of the small number of specimens. Throughout the testing, qualitative
visual examinations of the distances between joints were also performed.
After the interosseous first cuneiform-second metatarsal (Lisfranc)
ligament was sectioned, the mean change (and standard deviation) in the first
cuneiform-second metatarsal distance seen on weight-bearing radiographs was
0.7 ± 1.2 mm compared with the distance measured on the weight-bearing
radiographs of the intact specimens. The weight-bearing radiographs were
diagnostic for one of the ten specimens on the basis of the displacement at
this location. The mean displacement at this level was 1.6 ± 1.7 mm on
the abduction stress radiographs, which were diagnostic for two of the five
specimens, and —0.4 ± 0.8 mm on the adduction stress radiographs,
which were not diagnostic for any of the five specimens.
In the transverse injury group (sectioning of the plantar ligament between
the first cuneiform and the second and third metatarsals at the second
cuneiform-second metatarsal joint), the mean displacement at the first
cuneiform-second metatarsal joint was 1.1 ± 1.0 mm on the
weight-bearing radiographs, which were diagnostic for one of the five
specimens on the basis of the displacement at that location. The mean
displacement at the second cuneiform-second metatarsal joint was —0.3
± 0.2 mm of apparent joint-space narrowing, and the weight-bearing
radiographs were not diagnostic for any of the five specimens on the basis of
the displacement at that location. Weight-bearing caused a visibly decreased
distance across the tarsometatarsal joints and dorsal translation of the
second metatarsal base on the second cuneiform. The mean displacement at the
first cuneiform-second metatarsal joint was 4.6 ± 1.4 mm on the
abduction stress radiographs, which were diagnostic for all five of the
specimens on the basis of the displacement at that location. The mean
displacement at the second cuneiform-second metatarsal joint was 4.7 ±
2.0 mm, and again the abduction stress radiographs were diagnostic for all
five of the specimens on the basis of the displacement at that location
(Fig. 3). The increased
displacement between the first cuneiform and the second metatarsal and between
the second cuneiform and the second metatarsal on an abduction radiograph
(compared with that seen on a weight-bearing radiograph) in a transverse-group
specimen is shown in Figures 4-A and
4-B.
In the longitudinal injury group (sectioning of the interosseous ligament
between the first and second cuneiforms), the mean displacement at the first
cuneiform-second metatarsal joint was 2.2 ± 3.2 mm on weight-bearing
radiographs, which were diagnostic for one of the five specimens on the basis
of the displacement at that location. The mean displacement between the first
and second cuneiforms was 1.5 ± 2.6 mm, and the weight-bearing
radiographs were diagnostic for one of five specimens on the basis of the
displacement at that location. Weight-bearing caused a visibly decreased
distance across the tarsometatarsal joints and dorsal translation of the first
cuneiform on both the second metatarsal and the second cuneiform. The mean
displacement at the first cuneiform-second metatarsal joint was 0.5 ±
1.5 mm on the adduction stress radiographs, which were diagnostic for one of
five specimens on the basis of the displacement at that location. Despite a
small amount of displacement measured on the radiographs, all five specimens
displayed a visibly wide diastasis at the first cuneiform-second metatarsal
joint (Fig. 5). The mean amount
of displacement between the first and second cuneiforms was 3.3 ± 1.3
mm on the adduction stress radiographs, which were diagnostic for four of five
specimens on the basis of the displacement at that location
(Fig. 6). The increased
displacement between the first and second cuneiforms with adduction stress,
compared with the amount of displacement with weight-bearing, in this
longitudinal injury group was observed radiographically
(Figs. 7-A and 7-B).
The findings of this cadaver study suggest that disruption of both the
interosseous first cuneiform-second metatarsal ligament and the plantar
ligament between the first cuneiform and the second and third metatarsals is
necessary to produce the transverse Lisfranc injury pattern and that
disruption of both the interosseous first cuneiform-second metatarsal ligament
and the interosseous ligament between the first and second cuneiforms is
necessary to produce the longitudinal injury pattern. We found that sectioning
of the Lisfranc ligament alone was insufficient to produce evidence of
instability (as defined by a displacement of at least 2 mm) at the first
cuneiform-second metatarsal joint on weight-bearing radiographs.
Injury-specific abduction stress radiographs were more useful than
weight-bearing radiographs for the detection of the transverse Lisfranc injury
pattern. Abduction stress radiographs were diagnostic for all of the
specimens, whereas weight-bearing stress radiographs were diagnostic for only
one of five specimens. In the specimens with the longitudinal pattern of
injury, which was also not detected with weight-bearing radiographs, adduction
stress produced visibly wide displacement both between the first cuneiform and
second metatarsal and between the first and second cuneiforms. However, stress
radiographs demonstrated instability only between the first and second
cuneiforms.
The first cuneiform-second metatarsal displacement was not well detected on
adduction stress radiographs because the direction of the x-ray beam was too
oblique with respect to the joints of interest and obscured the degree of
actual diastasis observed visually. This observation is in agreement with the
findings reported by Coss et al., who demonstrated variability in radiographic
measurements with variations in foot pronation-supination or external and
internal rotation of the x-ray
beam10. The failure
to detect a first cuneiform-second metatarsal diastasis may also have been due
to dorsal translation of the first cuneiform, which could have made the
three-dimensional diastasis more vertically oriented. The use of fluoroscopy,
as recommended by other authors, would eliminate this
problem1,15.
The diagnosis of low-energy ligamentous Lisfranc injuries with use of
non-weight-bearing radiographs is notably difficult. Weight-bearing
radiographs are commonly used to detect these subtle
injuries1-4.
Abduction stress radiographs have been advocated by
some1,10,
but they have been described as being made with application of an abduction
force against the medial border of the forefoot. We found that the amount of
transverse instability could be emphasized on abduction stress radiographs
only by abducting the lateral four metatarsals.
In the absence of the interosseous and plantar ligaments between the first
and second rays, weight-bearing actually resulted in a visible decrease in the
diastasis at the tarsometatarsal joints in the unstable specimens. This was
apparently due to the tie-rod effect of the plantar fascia, which leads to the
truss-like behavior of the midfoot joints as described by
Hicks16. As
observed visually during testing, the apparent joint closure with
weight-bearing coincided with dorsal translation of the second metatarsal base
on the second cuneiform in the transverse instability group and with dorsal
translation of the first cuneiform on the second metatarsal and second
cuneiform in the longitudinal instability group. This abnormal movement at
these joints may be a causative factor in the generation of the degenerative
changes seen after these injuries.
There were limitations to our study. Our testing model allowed measurement
of dorsal joint translation in only two dimensions. Diastasis from rotation
about an axis plantar to the points measured is best represented in two
dimensions as dorsal surface translation. Three-dimensional analysis of the
tarsometatarsal diastasis associated with these injuries would be the research
goal of a follow-up study. However, we believe that our radiographic data are
meaningful because our methods are comparable with the current clinical
practice for making weight-bearing and stress
radiographs1-4,10.
Another limitation is that the results are based on a small number of
specimens in each study group. Also, the use of these particular abduction and
adduction stress radiographs has not been validated in clinical studies.
Clinically, the presence of intact skin overlying the dorsum of the foot may
diminish the effectiveness of these tests in demonstrating instability. We
removed the dorsal ligaments from the joints that were studied in order to
place the metallic markers. While this resulted in a non-physiologic
condition, the dorsal ligaments have not been shown to be important
stabilizers of the
midfoot15. Finally,
we chose 222.4 N of weight for simulated weight-bearing on the basis of our
observations that more weight did not cause further displacement; however,
whether a patient could actually bear this much weight after an acute injury
is not known.
In conclusion, sectioning of the interosseous first cuneiform-second
metatarsal (Lisfranc) ligament alone was insufficient to produce
radiographically identifiable instability of the Lisfranc joint with
weight-bearing or with manual stress. In this cadaver model, both the
interosseous first cuneiform-second metatarsal ligament and the plantar
ligament between the first cuneiform and the second and third metatarsals had
to be disrupted to produce the transverse injury pattern. Both the
interosseous first cuneiform-second metatarsal ligament and the interosseous
ligament between the first and second cuneiforms had to be disrupted to
produce the longitudinal injury pattern. On the basis of the results of this
study, we recommend further evaluation of the effectiveness of injury-specific
stress radiographs made with fluoroscopy for patients with suspected
instability of the Lisfranc joint to determine both the presence of
instability and its pattern. ?
Note: The authors thank Charles Roehm, Dennis Kayner, Lisa Doro,
MSE, Loretta Popp, MA, and Bryson Lesniak, MD, for their assistance with this
study.