Abstract
Background: The optimum management of ligamentous injuries of the
elbow is not known. Use of dynamic external fixators has been advocated to
stabilize the joint while maintaining motion, but there are no published data
to corroborate their efficacy. The purpose of this study was to test the
hypothesis that a laterally applied unilateral dynamic external fixator is
capable of stabilizing and restoring normal kinematics to elbows with varying
degrees of soft-tissue injury.
Methods: Six fresh-frozen cadaveric upper extremities, from donors
who were an average of seventy-six years of age at the time of death, were
tested in a custom apparatus with an electromagnetic tracking device to
analyze the kinematic behavior. Testing began with an injury of either the
lateral or the medial collateral ligament, which was followed by a second test
with an injury to the ligament on the contralateral side of the joint. In each
test, the varus-valgus displacement and the forearm rotatory displacement were
measured through the arc of elbow flexion under three loading conditions (hand
weight alone, hand weight plus 3.5 N, and hand weight plus 7 N). After each
test (with each injury), a unilateral external fixator was applied from the
lateral aspect of the elbow, and the same measurements were conducted under
the three loading conditions across the elbow joint.
Results: With varus stress testing, both after injury of the medial
collateral ligament alone and after injury of the lateral collateral ligament
and extensor mass alone, the laterally applied unilateral dynamic external
fixator was capable of maintaining the displacements within the laxity
envelope of an uninjured elbow. With valgus stress testing, after either
lateral or medial ligamentous injury, the fixator was unable to maintain
displacements within the normal laxity envelope when a 7-N load was applied to
the elbow. When both medial and lateral injuries were present, the lateral
fixator maintained varus displacement within normal limits, but valgus
displacement was consistently maintained within normal limits only when no
additional load was applied to the forearm.
Conclusions: A lateral dynamic elbow external fixator is capable of
maintaining varus displacements within normal limits in the presence of medial
and lateral collateral ligament injuries and with a 7-N load added to the
limb. However, valgus displacement is only consistently maintained within
normal limits if no additional displacement force is added to the weight of
the hand and forearm. The maintenance of valgus displacement is more sensitive
to additional load and specifically to the extent of medial soft-tissue
injury.
Clinical Relevance: The use of external fixation of the elbow is
growing in popularity. Yet, there is virtually no information with regard to
the adequacy of various constructs in the context of specific pathological
conditions. We demonstrated that a limited spectrum of soft-tissue injuries
about the elbow can be adequately managed with a laterally applied half-pin
fixator.
With injuries to the elbow, ligamentous and/or osseous integrity can be
compromised. A primary goal of the management of such injuries is to allow
ligamentous and/or osseous healing but prevent joint stiffness. Articulated
external fixators for the elbow have many theoretical advantages over their
static counterparts: most notably, they maintain joint stability in one plane
while allowing joint motion in the flexion plane, with documented clinical
success1,2.
In addition to the complications of joint stiffness and instability, an
important clinical issue is acceptance and compliance by the patient, which
have been lacking with larger frames. The use of a smaller, patient-friendly
frame raises concerns regarding the adequacy of the support afforded because
of its perceived reduced stiffness compared with that of larger frames.
Another important issue concerning smaller, dynamic elbow fixators is whether
they are able to maintain normal elbow kinematics, especially in the face of
substantial, unrepaired soft-tissue injury.
We hypothesized that a small unilateral articulated external fixator
applied laterally would be unable to replicate normal elbow kinematics in the
presence of lateral and/or medial soft-tissue injuries. We also hypothesized
that this configuration would be adequate to stabilize an elbow with lateral
and medial soft-tissue injuries.
Six fresh-frozen cadaveric upper extremities, without gross or radiographic
signs of osseous or soft-tissue abnormality, were studied. There were four
right and two left limbs from three men and three women who were an average of
seventy-six years of age (range, sixty-eight to eighty-five years of age) at
the time of death. All specimens were stored at —20°C from the time
of retrieval and were thawed overnight at room temperature prior to testing.
(Hence, they were subjected to a single freeze-thaw cycle.)
Each specimen was transected at the midpart of the humerus and
disarticulated at the radiocarpal joint, without compromising the soft-tissue
structures of the distal radioulnar joint. A weight designed to compensate for
the weight of the hand, equal to 50% of the forearm weight and averaging 3.15
N (range, 1.6 to 5 N), was attached to the radial styloid process. The humeral
origins of the brachialis and triceps were released subperiosteally to the
proximal level of the elbow capsule, and the humeral shaft was embedded in a
Plexiglas tube filled with polymethylmethacrylate dental resin. The embedded
humerus was placed in a fixture that maintained the humerus parallel to the
floor, by tightening the clamps of the elbow test jig. Loosening the fixture
allowed the elbow to be oriented in any of three positions: neutral with
forearm motion in the vertical plane, valgus with forearm motion in the
horizontal plane and the medial epicondyle uppermost, and varus with forearm
motion in the horizontal plane and the lateral epicondyle uppermost. Positions
were maintained by retightening the fixture. The triceps muscle was attached
to a 40-N weight while the biceps and brachialis were concurrently attached to
a motorized pulley. Passive elbow flexion, from a position of full extension
in the plane of the flexion arc, was controlled by the motor at a rate of
30°/sec. In the valgus or varus stress positions, load was applied with a
bag filled with saline solution attached to the distal part of the radius.
Testing Method and Sequence
The testing protocol is shown in Table
I. There were two kinematic assessments of displacement:
varus-valgus and axial rotation. For each assessment, there were two
ligament-alteration groups based on the sequence of release: in Group 1 the
lateral ulnar collateral ligament was released first, followed by the medial
collateral ligament, and in Group 2 the medial collateral ligament was
released first, followed by the lateral collateral ligament. For each
kinematic assessment and ligament state, there were three loading conditions:
hand weight alone, hand weight with a 3.5-N load, or hand weight with a 7-N
load.
The valgus and varus torque loads applied to the forearm were based on the
clinical advice given to patients at our institution. Patients are told that
no more than the weight of an approximately 12-oz (3.5-N) drink should be
handled after surgery requiring stabilization with an external fixator. The
3.5-N standard was doubled to 7 N for an extreme load configuration.
An electromagnetic tracking system (3Space Fastrak; Polhemus, Colchester,
Vermont) was used to record the kinematics of elbow motion and to track the
three-dimensional relationship of the ulna relative to the humerus, with a
sampling rate of 30 Hz. A transmitter source was mounted on the testing table
adjacent and proximal to the humerus. Two receiving sensors were attached, one
to the lateral aspect of the distal part of the humerus and the other to the
medial aspect of the distal part of the ulna. The accuracy of this system has
been documented to be within
0.5°3 and 0.02
mm4. This accuracy
is known to be affected by extraneous ferromagnetic
sources5; hence, any
such interference was minimized by elimination of all ferromagnetic materials
from the testing field. In addition, because of the nature of our proposed
investigation, we conducted a number of pilot studies. These pilot
investigations consisted of studying the kinematic patterns with and without
the presence of the fixator. We observed no effect due to the metal mass of
the fixator with respect to either accuracy or artifactual noise.
Testing began with the intact elbow oriented in the neutral position with
the humerus parallel to the floor and the forearm perpendicular to the floor
when positioned at 90° of flexion. From a position of full flexion, the
elbow was passively moved to full extension, with the muscle loads applied.
The rate of motion was kept constant at 30°/sec. Each testing stage was
performed in triplicate. The elbow was then positioned in the gravity varus
stress position, with the humerus and forearm parallel to the floor, the ulna
inferior, and the radius superior. The testing sequence was repeated with
muscle load and hand weight only, followed by the sequential addition of 3.5 N
and then 7 N. Finally, the elbow was oriented to the gravity valgus stress
position, with the humerus and forearm parallel to the floor, the radius
inferior, and the ulna superior, and the testing sequence was repeated.
Following testing of the normal intact elbow, a minimal muscle-splitting
surgical exposure of the medial side and a combined Kocher and common extensor
tendon musculoaponeurotic splitting lateral approach were used. These
exposures were just adequate to visualize the full extent of the medial and
lateral collateral ligaments and the osseous landmarks needed to insert the
axis pin targeting device. The latter relies on two osseous points: the center
of the capitellum laterally and the anteroinferior tip of the medial
epicondyle6. Care
was taken not to compromise the origin of the common extensor tendon from the
lateral epicondyle, the importance of which has been documented in other
studies7.
Radiographs were made at this stage to confirm that the pin was properly
positioned in the humerus. The skin and muscles were reapproximated with
sutures.
The elbow was taken through a test sequence to define the kinematic effect
of the exposure in isolation, without an external fixator. This was followed
by placement of a lateral unilateral articulated external fixation device, the
Dynamic Joint Distractor II (DJD II; Stryker Howmedica, Rutherford, New
Jersey). The placement of the fixator was assisted by the identification of
the medial and lateral isometric points for elbow rotation. The former is the
anteroinferior tip of the medial epicondyle, and the latter is in the center
of a circle best-fitted to the lateral outline of the capitellum. These points
were then used to position the u-shaped targeting clamp, which was secured,
and a lateral stylus was inserted into the center of the capitellum. The clamp
was then removed, and the external fixator was slotted over the stylus. The
position of the humeral pins (4 mm in diameter) and the ulnar pins (3 mm in
diameter) was then determined by the targeting aid. Once the four pins were
inserted, the attaching clamps were tightened.
The next two stages involved the introduction of the pathological
conditions and observation of the effect of the Dynamic Joint Distractor II in
restoring normal kinematics (Fig.
1). Of the six specimens, three first had release of the lateral
collateral ligament and extensor mass (Group 1) and three first had release of
the medial ligament (Group 2). The release involved complete excision of the
ligament from origin to insertion as well as transection of the extensor mass,
to simulate the full extent of ligamentous
deficiency7 because,
with ligament transection alone, the ligament still exerts some effect by
means of its surface attachment to the overlying muscles and their
tendons7. The
simulated-injury sequence was release of the lateral collateral ligament and
extensor mass followed by release of the medial collateral ligament in Group
1, and release of the medial collateral ligament followed by release of the
lateral collateral ligament and extensor mass in Group 2; then all six
specimens were tested with the absence of both collateral ligaments. After
each sequence, the kinematics with and without the dynamic fixator were
assessed.
The final stage of experimentation was the disarticulation of the elbow
joint and digitization of the elbow articular surfaces and the distal part of
the humeral shaft. These data were used to create a coordinate system based on
the osseous anatomy. The collected data consisted of ulnar angulation relative
to the humerus in valgus-varus and internal-external axial rotation as a
function of the degree of elbow flexion.
Statistical Methods
Statistical analysis was performed with three factors taken into account:
(1) two sequences of ligamentous injury (the lateral collateral ligament first
and then the medial collateral ligament, and the medial collateral ligament
first and then the lateral collateral ligament); (2) five joint-integrity
conditions (intact, first injury, first injury plus external fixation, second
injury, and second injury plus external fixation); and (3) three weights (hand
weight, hand weight plus 3.5 N, and hand weight plus 7 N). Four outcome
measures were analyzed (angulation under varus stress, angulation under valgus
stress, axial rotation under varus stress, and axial rotation under valgus
stress). A three-factor analysis-of-variance model was constructed. Because
the sequence of injuries was either first lateral and then medial or first
medial and then lateral, isolated medial and lateral kinematics and combined
medial and lateral kinematic patterns were defined.
The surgical exposure caused no detectable difference in the kinematics
compared with those of the intact elbow. The variation from the control
occurred between 60° and 100° of flexion, so a position of 80° of
flexion was chosen for the detailed analysis to compare the impacts of the
loading modes and the constraint alterations.
Kinematic Displacements
Normal
The normal varus displacement in the varus stress position increased
progressively in both Group 1 and Group 2 from hand weight only to hand weight
plus 7 N (Figs. 2 and
3). In the valgus stress
position, the normal valgus displacement also increased progressively in both
groups from hand weight only to hand weight plus 7 N (Figs.
2 and
3). The normal axial rotational
displacements of the ulna with the arm in the varus stress position
progressively increased in pronation in both groups from hand weight only to
hand weight plus 7 N (Figs. 4
and 5). The rotational
displacements with the arm in the valgus stress position progressively
increased in Group 1 but revealed no consistent or progressive change with
increases in the weights in Group 2 (Figs.
4 and
5).
Group 1 (Figs.
2 and
4)
The kinematic testing with the lateral collateral ligament excised and the
extensor mass incised revealed varus angular displacements of —1.0°
± 1.0° with hand weight only to —14.5° ± 6.8°
with hand weight plus 7 N (Fig.
2). When a unilateral frame was applied to the lateral aspect of
the elbow, the displacements changed to 0.1° ± 1.3° with the
weight of the hand to —1.4° ± 1.1° with hand weight plus
7 N.
External rotational displacements of the ulna (pronation) with the arm in
the varus stress position increased from —4.2° ± 0.58°
with hand weight only to —14.3° ± 7.2° with an additional
7-N load. With the fixator in place, the displacements increased to 3.84°
± 2.21° and —3.2° ± 2.0°, respectively.
Displacements in the valgus stress positions are shown in
Figure 4.
Group 2 (Figs.
3 and
5)
With the medial collateral ligament sectioned, there were progressive
increases in valgus angular displacements, from 4.5° ± 1.4° to
9.0° ± 1.7°, with progressive increases in weight
(Fig. 3). When the unilateral
frame was applied to the lateral aspect of the elbow, the valgus displacements
also increased progressively, from 2.6° ± 2.3° to 5.3°
± 1.9°, with progressive increases in weight
(Fig. 3). Internal rotational
displacements of the ulna with valgus stress also increased progressively,
from 1.4° ± 1.07° to 3.1° ± 1.27°, with
increases in weight. With the fixator in place, the internal rotational
displacements decreased to —1.2° ± 3.92° and
—1.7° ± 4.19°, respectively
(Fig. 5).
Medial and Lateral Injury
The second injury to each group created a pool of six specimens with
complete medial and lateral ligamentous deficiency. In the varus stress
position, the varus-valgus displacements with progressively added weights were
—1.17° ± 1.10° with hand weight only to —16.79°
± 7.94° with the addition of 7 N (see Appendix). In the valgus
stress position, the displacements were 4.21° ± 1.89° to
11.01° ± 2.16° (see Appendix). Application of the external
fixator following complete disruption of the medial and lateral ligamentous
structures altered the varus and valgus displacements toward the preinjury
state (Figs. 2 and
3, and Appendix).
With both ligamentous lesions, under the three loading conditions,
rotational displacement of the ulna in the varus position moved progressively
into pronation with the fixator in place, reducing this pronation displacement
back toward the uninjured state (Figs.
4 and
5, and Appendix). Similarly,
rotational displacements with valgus stress were progressively more supinated,
with the effect of the fixator reducing the displacement toward the uninjured
state (Figs. 4 and
5, and Appendix).
Statistical analysis was performed according to the number of injuries,
since there was no difference in the findings between the sequences of injury.
No differences were found when we compared any of the tests with the hand
weight only in the valgus stress position. In the tests of hand weight plus
3.5 N, injury condition 1 (after the first injury only) was significantly
different from the intact state (p = 0.003). Injury condition 2 (after both
injuries) with the fixator was not significantly different from that condition
without the fixator, but injury condition 2 was significantly different from
the intact state and injury condition 1 with the fixator (p = 0.003). With
hand weight plus 7 N, injury condition 1 was significantly different from the
intact state and from injury condition 1 after the fixator was applied (p <
0.001). Injury condition 2 was significantly different from the intact state
and from injury conditions 1 and 2 with the fixator (p < 0.001). In the
varus stress position, there were differences in the hand-weight-only test,
but these were not significant (p = 0.15). With hand weight plus 3.5 or 7 N,
injury conditions 1 and 2 were significantly different from the intact state
and both fixator-stabilized injury states (p < 0.001).
No differences in axial ulnar rotation could be identified in the varus
stress position with hand weight only, but with valgus stress there was
significantly less rotational displacement with fixator-stabilized injury
condition 2 than with the intact state, injury condition 1, or injury
condition 2 (p = 0.01). In the valgus stress position, all injury patterns
stabilized with a fixator showed significantly less rotational displacement
than the unstabilized injuries (p = 0.004 for hand weight plus 3.5 N, and p =
0.003 for hand weight plus 7 N) but showed no difference when compared with
the intact state. In the varus stress position, with hand weight plus 3.5 N,
injury conditions 1 and 2 were significantly more displaced than the intact or
the fixator-stabilized conditions (p = 0.002). With hand weight plus 7 N, the
second injury group was significantly more displaced than the intact state and
both stabilized-injury conditions (p = 0.001).
Our data show that our first hypothesis—that an external fixator
could not allow replication of normal kinematics after soft-tissue
injury—was incorrect. In fact, the articulated fixator applied laterally
as a half-frame introduces only several degrees of alteration of normal elbow
kinematics. The second hypothesis—that this external fixator, as applied
in our study, can stabilize an elbow with selected lateral and medial ligament
injuries—was confirmed.
Dynamic hinged external fixators for the elbow were first described in
1975, by Volkov and Oganesian, for the restoration of joint motion in cases of
acute and chronic elbow
stiffness8. Few
publications have provided laboratory data regarding the stiffness and optimal
placement of an articulated hinged fixator about the
elbow9-13.
We are also not aware of any reports that have addressed the question posed by
the present study.
During activities of daily living, the vast majority of functions generate
a sustained varus stress across the elbow, with only occasional short-lived
valgus stresses. Following trauma with subsequent internal and/or external
stabilization, the goal of rehabilitation is to minimally stress the elbow
until osseous and soft-tissue healing has occurred. Maintenance of active
motion is sought to avoid stiffness. Because of concern about the proximity of
the ulnar nerve, most surgeons avoid placing an external fixator medially.
Therefore, we assessed the effectiveness of the smaller, less rigid, lateral
unilateral Dynamic Joint Distractor-II fixator for stabilizing the elbow after
lateral and medial soft-tissue injuries. These simulated soft-tissue injuries
were chosen because they most commonly accompany elbow
dislocations14. Our
data support the notion that the fixator adequately protects the lateral
soft-tissue injury from a varus force and also effectively resists moderate
valgus force. The fixator resists both varus and valgus displacement even with
increasing displacement loads. In our experience, the medial 1° to 2°
of displacement still present after the fixator has been applied is less than
a clinician or patient can observe with daily function.
An ideal clinical function of an articulated external fixator is provision
of sufficient rigidity to compensate for the effect of soft-tissue injuries by
decreasing the force on healing articular fractures while replicating the
kinematics of the elbow joint. On the basis of our data, a clinician may apply
this articulated fixator using half-pins on the lateral side of the elbow and
reliably restore varus, valgus, and rotational stability to normal.
Furthermore, one should be able to reproducibly apply the device referable to
the axis of rotation. Because of the lack of comparative literature, it is not
yet possible to fully understand the relevance of absolute axis accuracy and
fixator rigidity as a function of kinematic replication. Anatomical variations
of the lateral and medial collateral ligament complexes are reported to be
present in as many as 50% of the normal
population15. These
poorly understood variations support the concept that a less rigid fixator
might be able to more adequately compensate for those variables than would a
more rigid device. While the fixator was unable to precisely replicate normal
kinematics, our data demonstrated the ability of the device to accommodate
moderate varus forces in the presence of complete soft-tissue injury.
The cadaveric nature and small number of test specimens were limitations of
this study. Also, the elbows were from elderly subjects and hence may not
represent the younger patient population in whom these injuries commonly
occur.
Tables showing data from the test under the various conditions are
available with the electronic versions of this article, on our web site at
(go to
the article citation and click on "Supplementary Material") and on
our quarterly CD-ROM (call our subscription department, at 781-449-9780, to
order the CD-ROM). ?
Ring D, Jupiter JB. Compass hinge
fixator for acute and chronic instability of the elbow. Oper Orthop
Traumatol. 2005;17:
143-57.17143
2005
[PubMed][CrossRef]
Stavlas P, Gliatis J, Polyzois V,
Polyzois D. Unilateral hinged external fixator of the elbow in complex elbow
injuries. Injury. 2004;35:
1158-66.351158
2004
[PubMed][CrossRef]
Morrey BF, Tanaka S, An KN. Valgus
stability of the elbow. A definition of primary and secondary constraints.
Clin Orthop Relat Res.
1991;265:
187-95.265187
1991
[PubMed]
Luo ZP, Niebur GL, An KN. Determination
of the proximity tolerance for measurement of surface contact areas using a
magnetic tracking device. J Biomech.
1996;29:
367-72.29367
1996
[PubMed][CrossRef]
Milne AD, Chess DG, Johnson JA, King GJ.
Accuracy of an electromagnetic tracking device: a study of the optimal range
and metal interference. J Biomech.
1996;29:
791-3.29791
1996
[PubMed][CrossRef]
Cheng SL, Morrey BF. Treatment of the
mobile, painful arthritic elbow by distraction interposition arthroplasty.
J Bone Joint Surg Br.
2000;82:
233-8.82233
2000
[PubMed][CrossRef]
Dunning CE, Zarzour ZD, Patterson SD,
Johnson JA, King GJ. Ligamentous stabilizers against posterolateral rotatory
instability of the elbow. J Bone Joint Surg Am.
2001;83:
1823-8.831823
2001
[PubMed]
Volkov MV, Oganesian OV. Restoration of
function in the knee and elbow with a hinge-distractor apparatus. J
Bone Joint Surg Am. 1975;57:
591-600.57591
1975
Bottlang M, O'Rourke MR, Madey SM,
Steyers CM, Marsh JL, Brown TD. Radiographic determinants of the elbow
rotation axis: experimental identification and quantitative validation.
J Orthop Res. 2000;18:
821-8.18821
2000
[PubMed][CrossRef]
Sekiya H, Neale PG, O'Driscoll SW, An
KN, Morrey BF. An in vitro biomechanical study of a hinged external fixator
applied to an unstable elbow. J Shoulder Elbow Surg.
2005;14:
429-32.14429
2005
[PubMed][CrossRef]
Madey SM, Bottlang M, Steyers CM, Marsh
JL, Brown TD. Hinged external fixation of the elbow: optimal axis alignment to
minimize motion resistance. J Orthop Trauma.
2000;14:
41-7.1441
2000
[PubMed][CrossRef]
Fukuda Y, Takai S, Yoshino N, Murase K,
Tsutsumi S, Ikeuchi K, Hirasawa Y. Impact load transmission of the knee
joint-influence of leg alignment and the role of meniscus and articular
cartilage. Clin Biomech (Bristol, Avon).
2000;15:
516-21.15516
2000
[PubMed][CrossRef]
Radin EL, Swann DA, Paul IL, McGrath PJ.
Factors influencing articular cartilage wear in vitro. Arthritis
Rheum. 1982;25:
974-80.25974
1982
[CrossRef]
Lill H, Korner J, Rose T, Hepp P,
Verheyden P, Josten C. Fracture-dislocations of the elbow joint—strategy
for treatment and results. Arch Orthop Trauma Surg.
2001;121:
31-7.12131
2001
[PubMed][CrossRef]
Beckett KS, McConnell P, Lagopoulos M,
Newman RJ. Variations in the normal anatomy of the collateral ligaments of the
human elbow joint. J Anat.
2000;197Pt 3:
507-11.197507
2000
[PubMed][CrossRef]