Knee Dislocations and Fracture-Dislocations

Knee dislocations can occur following a wide variety of mechanisms but occur most frequently following high-energy events

such as motor vehicle collisions or pedestrian versus motor vehicle accidents.¹²⁰,¹²¹
In one large series of 126 patients, 115 (91%) of the dislocations
occurred as a result of a high-energy mechanism. The most common causes
of knee dislocation in that series were motor vehicle collisions (52%),
motorcycle collisions (17%), and motor vehicle versus pedestrian
accidents (16%).¹²¹ The diagnosis of
a knee dislocation in a high-energy multiple-trauma patient is often
difficult, particularly if the knee has spontaneously reduced. There
are frequently other obvious and life-threatening injuries present that
draw the attention of the physicians treating the patient.
Life-threatening injuries to the head, chest, or abdomen are reported
in 27%, associated fractures in 50% to 60%, and multiple fractures in
41% of knee dislocations.¹³⁷
Additionally, ipsilateral skeletal injuries that draw the attention of
the orthopaedic trauma team make examination of the knee very
difficult. In the large series already cited, 56% of the patients had
ipsilateral fractures, with tibial plateau,³⁸ acetabular,¹⁹ and femur¹⁹ fractures occurring most commonly.¹²¹
All of these associated fractures make examination of the knee
extremely difficult. The frequency and severity of associated injuries
are a major reason knee dislocations have often been overlooked in the
past. It is very important to carefully evaluate patients for knee
dislocation in the trauma room, since missing this diagnosis can be
associated with limb threatening vascular injuries.⁹⁰,¹⁴²

Low-energy knee dislocations are less frequent but do
occur. Injuries that occur during athletic competition occasionally
will lead to a knee dislocation. Football and equestrian injuries are
two of the most common athletic events associated with knee
dislocations.¹²¹ Additionally,
low-energy knee dislocations have been noted to occur in obese
patients. These injuries often result from everyday activities such as
stepping down from a curb and are very challenging to treat. They are
associated with a high incidence of neurologic and vascular injuries,
despite the low-energy mechanism.¹¹⁰

The classic teaching is that knee hyperextension is the
most common cause of knee dislocation. Kennedy conducted a study on
cadaver specimens and demonstrated that exaggerated hyperextension will
produce a knee dislocation. In his study on 10 cadaver specimens, the
ACL tore first, followed by the PCL and posterior capsule at 30 degrees
of hyperextension. The popliteal artery tore at approximately 50
degrees of hyperextension.⁵⁶,⁵⁷
A varus or valgus force combined with hyperextension can produce
variability in the pattern of collateral ligament injuries. Knee
dislocations have been reported to have a high incidence of avulsion
injuries to both ligamentous and tendinous insertions⁸¹ (Fig. 54-3). Sisto and Warren¹¹⁴ and Frassica et al.³⁶ have both reported a high incidence of avulsion injuries to the cruciate ligaments (Table 54-1).
Both series involved a higher percentage of low-energy mechanisms of
injury than is seen in most trauma centers. These findings are in
direct contrast to the high frequency of midsubstance tears that are
seen in isolated low-energy (sports) injuries to the ACL. Their
findings are also in direct contrast to the experience of all three
authors of this chapter. The majority of high-energy knee dislocations
do not involve a bony avulsion. However, it is important to recognize
avulsion injuries when they occur, because it will directly impact
treatment options and an avulsion injury frequently simplifies the
treatment plan.

The impact of strain rate on the failure properties of the ACL was investigated by Noyes et al.¹³⁹
They found the primary mode of failure was tibial avulsion at slow
rates and mid-substance ligament failure at faster rates. Kennedy et al.⁵⁷
performed a similar study and found the primary location of failure was
mid-substance at both slow and faster strain rates. While these
laboratory studies termed their varying rates as slow and fast, both
were slow from a clinical perspective. Schenck et al. evaluated the
impact of much higher strain rates (approximately 5400%/sec) on the
PCL. They noted that the high strain rates produced a stripping lesion
of the femoral attachment of the PCL that correlates with the “peel
off” lesion that is frequently encountered following high-energy trauma¹⁰⁶ (Fig. 54-4).
In their cadaveric study, the “peel off” occurred in an avulsion
pattern with very small bony fragments, with the primary injury
occurring through Sharpey’s fibers. When they used a low-velocity
model, they found the PCL tore more consistently in a midsubstance
pattern. They noted that the ACL was more variable in its injury
pattern, with midsubstance tears commonly occurring with both high- and
low-velocity injury rates. The authors hypothesized that the difference
between the ACL and PCL was in part created by the femoral notch. The
ACL is classically torn by the notch transverse to its fibers, whereas
the PCL

is torn by forces parallel to the ligament. This would make the PCL more sensitive to strain rate than the ACL (Fig. 54-5).
It is controversial whether “peel off” lesions can be treated like
avulsions and repaired with good results, rather than reconstructing
them. High-energy trauma exposes the knee to exceptionally high
velocity and strain rates. It is the experience of the lead author that
most ACL tears are midsubstance, while PCL tears can be either
midsubstance or a “peel off” lesion. In the vast majority of cases, we
reconstruct the ligament with either injury pattern.

The history of the injury for patients who have sustained a knee dislocation usually includes a high-energy traumatic injury (Fig. 54-7).
Most patients have no idea of exactly what position their knee was in
during a motor vehicle collision. Occasionally, the history of a
hyperextension injury can be obtained. A history of a dashboard injury
to a flexed knee should increase the suspicion of a knee dislocation.

The physical examination findings associated with a
patient who has sustained a knee dislocation vary widely, ranging from
an irreducible knee dislocation to a spontaneously reduced dislocation
with minimal physical findings. While most patients present with knee
swelling, patients with extensive capsular tears may not demonstrate a
large hematoma following a knee dislocation. The physical findings are
generally obvious in patients who present with the knee dislocated, due
to the gross deformity. It is important to remember that this
represents a minority of the patients who have sustained a knee
dislocation.¹²⁸ Subtle signs of injury such as an abrasion, minimal swelling, or complaints of pain may be the only signs of a knee

injury. An examination of the knee will generally reveal gross ligamentous instability.

Patients with a spontaneously reduced knee dislocation
and an ipsilateral fracture are the most challenging diagnostic
dilemma. Frequently, orthopaedic surgeons are concentrating on the
obvious skeletal injuries and ignoring the subtle signs of a knee
injury. A critical but often overlooked diagnostic step is to perform a
good examination under anesthesia after skeletal stabilization. This
will often unmask significant ligamentous knee instability that will
otherwise go undiagnosed. Data we have collected have demonstrated that
26% of tibial plateau fracture patients who sustained their injury as a
result of high-energy trauma have a concomitant bicruciate ligament
injury.¹¹⁸ We have also found a high association of knee dislocations with ipsilateral femur and acetabulum fractures.¹²¹ Accurately diagnosing knee dislocation in multiple trauma patients requires vigilance and a high index of suspicion.

Plain radiographs of the knee often demonstrate subtle
clues of injury in patients with a knee dislocation. Possible findings
include small avulsion fragments, joint space asymmetry between the
medial and lateral joint spaces, or minimal subluxation of the joint.
Stress radiographs under anesthesia in varus or valgus with the knee in
extension can help document collateral ligament disruption.

Surgical decision making can be enhanced by the use of
MRI. The scan does not take the place of physical examination but
provides the surgeon with helpful data. MRI is particularly helpful in
patients with ipsilateral extremity fractures and subtle signs of knee
injury. Skeletal examination is virtually impossible prior to skeletal
stabilization and may be difficult even following stabilization.
Additionally, it is beneficial to identify patients with spontaneously
reduced knee dislocations prior to performing skeletal stabilization of
fractures. The knowledge of a multiple ligament knee injury will alert
the surgeon to the risk of vascular injury and also may lead to
decreased use of a tourniquet during skeletal stabilization procedures
to avoid clot formation adjacent to small intimal tears. To obtain
useful images, it is important to obtain the MRI scan prior to skeletal
stabilization with metallic implants, as metal artifact will often
yield MRI that is virtually useless. The use of MRI on patients with
tibial plateau fractures has been the subject of a number of
manuscripts. Five studies on the subject of MRI findings associated
with tibial plateau fractures have demonstrated a number of significant
findings. These studies report on a total of 197 tibial plateau
fractures, and demonstrate the following injuries: 54 (27%) medial
meniscus tears; 74 (38%) lateral meniscus tears; 63 (32%) ACL tears; 45
(23%) PCL tears; 30 (15%) PMC tears; and 54 (27%) PLC tears.⁹,⁴⁹,⁵⁹,¹¹¹,¹²⁴
Additionally, Yacoubian et al. found that adding MRI to CT and plain
radiographs in 52 patients led to a change in the fracture
classification in 21% and a change in the treatment plan in 23% of
cases.¹⁴¹ Similarly, Holt et al.
reported that MRI led to a change in the fracture classification in 48%
of their tibial plateau fractures, and a change in the treatment plan
in 19% of their patients.⁴⁹
Additionally, they reported a 48% incidence of previously unrecognized
soft tissue injuries in the knee, including two spontaneously reduced
knee dislocations. Stannard et al. have reported on the incidence of
soft tissue injuries in 103 consecutive patients evaluated with MRIs
following tibial plateau fractures.¹¹⁸
Patients in this series were diagnosed with 25 medial meniscus tears,
35 lateral meniscus tears, 45 ACL tears, 41 PCL tears, 16 PMC tears,
and 46 PLC tears. Twenty-six patients were diagnosed with bicruciate
ligament injuries. Published studies have documented between 48% and
90% of patients with high-energy tibial plateau fractures also have
significant associated soft tissue injuries.⁹,¹⁰,¹⁶,⁴⁹,⁵⁹,¹¹¹,¹⁴¹
Due to the high incidence of associated soft tissue injuries, obtaining
MRI appears to be justified following high-energy tibial plateau
fractures.

For many years, knee dislocations were classified according to the classification described by Kennedy.⁵⁷
This system described the position of the tibia with respect to the
femur, and categorized dislocations into five primary types: anterior,
posterior, lateral, medial, and rotatory (Fig. 54-8).
The rotatory dislocations had four subclassifications: anteromedial,
anterolateral, posteromedial, and posterolateral. The posterolateral
rotatory dislocation was described as the most frequently encountered
rotatory dislocation.²,⁶⁸,¹¹⁶ A hallmark of posterolateral dislocations is that they may be irreducible (Table 54-3). This occurs

when the medial femoral condyle buttonholes through the medial capsule
and the medial collateral ligament (MCL) invaginates into the knee
joint, preventing reduction.⁴⁷,⁹²
The sine qua non of this condition is a transverse furrow seen on the
medial side of the knee. Peroneal nerve palsy is frequently associated
as a result of a traction injury of the nerve over the lateral femoral
condyle. Skin necrosis also frequently occurs as a result of pressure
from the medially displaced femur⁴⁷ (Fig. 54-9).

The position classification system is useful for guiding
the surgeon regarding the potential reduction maneuvers needed to
reduce the dislocation. It also may help alert the surgeon to
associated complications or injuries with specific patterns. However,
since more than half of knee dislocations present following spontaneous
reduction, many cannot be classified by this system. Additionally, the
position classification system only suggests which structures may be
torn. It does not specifically guide surgical decision making. For
example, patients can sustain an anterior dislocation and have either a
bicruciate ligament injury or a PCL intact dislocation (Fig. 54-10). The position classification system does not help differentiate between these two patterns.

In contrast, the anatomic classification system is based upon

what structures are torn and is very useful in guiding the surgeon
regarding treatment options. The anatomic classification was initially
described by Schenck and has been modified over the years. It is based
primarily on the findings from physical examination, particularly the
examination under anesthesia.⁶³,¹⁰⁴,¹³⁴
MRI findings can be added to the physical examination, but the primary
tool used to classify these fractures is the physical examination.

The anatomic classification system assigns numbers based
on which ligaments are torn. As a general rule, the higher the number,
the higher is the energy of the dislocation and the greater is the
severity of the injury to the knee. The system considers the knee to be
made up of four ligament groups: ACL, PCL, PMC, and PLC. The following
five major categories are KD I, any knee dislocation with one cruciate
ligament intact; KD II, a bicruciate knee dislocation; KD III, a
bicruciate ligament dislocation with one of the corners torn as well;
KD IV, a bicruciate ligament dislocation with both corners torn; and KD
V, any fracture dislocation. Additional modifiers can be added to
indicate an arterial injury (the letter C) or a neurologic injury (the
letter N). Stannard et al. modified the classification of the fracture
dislocations to identify which structures were torn when there was an
associated fracture¹²¹ (Table 54-4).

Walker et al., Eastlack et al., and Stannard et al. used
the anatomic classification system to classify their injuries and
direct treatment.²⁶,¹²¹,¹³¹,¹³²
They noted that KD III was the most common pattern seen. Some of the
authors also noted that KD IIIL patients fared worse than KD IIIM
patients with regard to arthrofibrosis, instability, disability, and
outcome measures. KD IV injuries have been noted to be higher energy
injuries and to be associated with an increased risk of popliteal
artery injury compared with other dislocations.¹²¹
The anatomic system is very useful because if forces the surgeon to
focus on what structures are torn. It can then be used to direct
surgical treatment. It also allows for an accurate discussion of knee
dislocations between clinicians, as well as allowing accurate
comparison of similar injuries in the literature. We strongly advocate
the anatomic classification system to describe knee dislocations.

Knee injuries are best evaluated by considering the
injury based on the structure(s) torn. Within the diagnosis of a knee
dislocation, multiple structures are frequently injured, with specific
patterns of injury described. As noted earlier in this chapter, the
classification of knee dislocations is most accurately determined by
the ligaments torn (the anatomic system); hence, knowledge of the
structure and function of the knee ligaments is very important in
defining the injury pattern. Furthermore, the associated disruption of
musculotendinous structures (patellar tendon, patellofemoral
dislocation, iliotibial band, biceps femoris, gastrocnemius, and
popliteus) and the importance of associated injuries adds to the
variability of the pathology involved in knee dislocations. Last, the
bony architecture is important, as joint surface fractures can occur
with a knee dislocation and are frequently underreported. Depending
upon the size of fracture fragments, the knee injury can be classified
as a fracture-dislocation when associated with a large condyle fracture
rather than a pure ligamentous injury as is implied with the term knee
dislocation.

Conceptually, defining the knee as a joint composed of
four major ligamentous structures (ACL, PCL, PMC/MCL, and PLC) is very
useful. The ACL and PCL can be torn in tandem (a bicruciate injury),
separately (ACL or PCL intact knee dislocations), or rarely not torn at
all (a cruciate-intact knee dislocation) (Fig. 54-11).
The injuries to the collaterals are frequently complete, with
involvement of at least one of the ligamentous corners. The anatomic
arrangement of the popliteus, the popliteofibular

ligament, and the biceps femoris posterolaterally (Fig. 54-12),
and the semimembranosus/posterior oblique ligament complex
posteromedially allows for injury of those structures in conjunction
with the major ligament injury.

The anterior approach to the knee allows for exploration
of the joint, synovectomy, fracture fixation, and repair of the
extensor mechanism. The patient is placed in the supine position and a
small linen roll is placed under the ipsilateral buttock. A tourniquet
is applied to the proximal thigh. A straight incision beginning
proximal to the superior pole of the patella is carried distally,
directly over the patellar tendon and medial to the tibial tubercle.
The length of incision depends on the pathology being treated. Exposure
of the extensor mechanism involves incision of the arciform layer, the
condensation of layer I anteriorly, just deep to the subcutaneous fat.
Subcutaneous dissection is usually not be made superficial to this
fascial layer as devitalization of the overlying skin can occur,
especially if a concurrent lateral retinacular release is performed.
However, in knee dislocations, there is frequently subcutaneous
stripping from layer one, such that the knee dislocates within the
subcutaneous sleeve. If joint exposure is required, the quadriceps
tendon is incised in its midportion, followed by a medial capsulotomy
carried distally and medially to the tibial tubercle. A cuff of capsule
at least 5 mm wide should be preserved on the medial edge of the
patella for closure. Inadequate longitudinal release of the quadriceps
tendon can result in avulsion of the tendon from the tibial tuberosity
when the patella is displaced laterally. This incision allows exposure
of the anterior knee, distal femur, and proximal tibia. The tibial
tubercle and patellar tendon limit exposure of the lateral knee joint.
Frequently, when open exposure of the intercondylar notch is needed for
cruciate reattachment, an incision from the medial edge of the pole of
the patella can be carried down medially to the anterior tibial plateau
to achieve adequate exposure. Viewing the contents deep within the
notch frequently requires use of a headlamp in such open approaches¹⁰² and is much more difficult than with a purely arthroscopic approach.

The lateral approach is useful for open reduction and
internal fixation of distal femoral condyle fractures (femoral-sided
fracture-dislocations of the knee), tibial plateau fractures, and
reattachment of knee ligaments and lateral tendons. The patient is
positioned supine with a small bolster under the ipsilateral hip. The
incision is placed over the lateral side of the lateral femoral
condyle, anterior to the iliotibial band, and is carried distally over
Gerdy tubercle. The fascia lata is opened parallel to the skin
incision, anterior to the iliotibial band. If proximal extension is
necessary, the vastus lateralis is elevated off the lateral
intermuscular septum and any perforating vessels are ligated.
Retractors are placed subperiosteally over the anterior aspect of the
femur, exposing the entire distal aspect of the lateral femur. The
patellofemoral joint can be exposed by incising the lateral transverse
patellar ligament. Splitting the vastus lateralis muscle through its
belly should be avoided as it can cause excessive bleeding with loss of
control of the perforating vessels. Posterior exposure with this
approach is limited by the fibular collateral ligament and common
peroneal nerve.¹⁰²

The posteromedial approach provides access to the vascular structures of the popliteal fossa, and it is the utilitarian approach

to these vessels. Variations of this approach have been used by Burks and Schaffer¹⁸ (Fig. 54-13) and Berg¹¹ to approach the tibial attachment of the PCL when using the inlay PCL reconstruction technique (Fig. 54-14). Additionally, the cruciates and the PMC can be accessed via this approach. It has also been described by Muscat et al.⁸⁵
for simultaneous repair of vascular injuries and ligaments when
associated with a traumatic knee dislocation. The approaches to the PCL
described separately by Burks and Schaffer and Berg involve a posterior
incision into the flexion crease of the knee combined with a similar
deepplane dissection as described next.

The posteromedial approach can be safely accomplished in
the supine position with the knee flexed and the hip externally rotated
in the figure-four position (Fig. 54-15). The
incision is placed along the posterior aspect of the medial femoral
epicondyle and courses distally along the posterior edge of the
proximal medial tibia. The saphenous vein is identified and should be
protected if possible. The saphenous nerve should be retracted
posteriorly with the skin flap and the saphenous vein to avoid neuroma
formation. For vascular access, the tendons of the pes anserinus are
usually transected 2 cm proximal to their tibial insertion and tagged
for subsequent repair at closure. The proximal portion of the pes is
reflected with the posterior flap, and the distal portion of the pes is
reflected anteriorly. This allows exposure of the medial head of the
gastrocnemius. When using this approach just to repair the PCL, the
medial meniscus, or the PMC, the pes is mobilized but left intact. The
medial head of the gastrocnemius may be detached (partially or
completely) from the medial femur to allow exposure of the popliteal
artery from Hunter canal to its trifurcation. With an approach to the
PCL or the medial knee ligaments, the tendon of the medial head of the
gastrocnemius should be left intact. Exposure requires development of
the plane between the posteromedial knee joint capsule anteriorly and
the medial gastrocnemius posteriorly. All retractors must remain
anterior to the medial head of the gastrocnemius to protect the
popliteal vessels. This approach is complex, and ideally it should be
performed on a cadaver specimen or with an experienced surgeon to
clearly understand the soft tissue planes. It is easy to inadvertently
dissect posterior to the medial gastrocnemius belly or anteriorly into
the joint capsule. Keeping the knee flexed to 70 degrees relaxes the
posterior muscles, simplifying the identification of the plane anterior
to the gastrocnemius. Damage to the saphenous nerve can produce
persistent medial knee pain from a traumatic neuroma and should be
avoided. Furthermore, the incision should be placed posteriorly over
the muscle bellies in the leg, since incisions directly over the tibia
can result in skin necrosis. In the presence of a knee dislocation with
evidence of MCL insufficiency the capsular structures are usually
completely disrupted, allowing exposure of the knee joint itself.
Access to the lateral side of the knee is limited with this approach.
Skin edges should be managed carefully to avoid delayed healing or
wound edge necrosis.

This approach is useful to reconstruct the PLC, to
mobilize the lateral gastrocnemius muscle for use as a muscle pedicle
flap, and to explore and repair the common peroneal nerve. It is
frequently useful in knee dislocations for reattachment or repair of
the cruciate ligaments and especially the PLC. The peroneal nerve is
always isolated prior to deep joint exposure to prevent inadvertent
injury. In dislocations with complete tears of both cruciates, the
ligaments are frequently disrupted circumferentially in a half-circle
starting at the patellar tendon anteriorly, with disruption of the ACL
and the PLC, and with enough force

to
injure the PCL. With such injury, exposure of the tibiofemoral joint
can be facilitated by following the tissue planes created by the
dislocation.¹³³
The patient is placed in the supine position with a tourniquet on the
thigh and a small bump under the ipsilateral buttock. The lateral knee
is most safely exposed with the knee flexed to 90 degrees. This relaxes
the peroneal nerve and allows for better protection of the nerve
throughout the procedure. The skin incision is placed in line with the
fibular head and carried in a straight line proximally, then curved
onto the lateral thigh. With proximal extension, the incision should
curve between the iliotibial band and the biceps femoris tendon. The
incision is carried down to the deep fascia, which is then opened
carefully with scissors. At this point the nerve must be identified (Fig. 54-16).
It can usually be palpated subfascially as it courses from the biceps
femoris through its perineural fat to the fibular neck. This approach
should always include exposure of the peroneal nerve prior to deep
dissection. The peroneal nerve is isolated by palpating its location
and carefully dissecting using fine scissors and forceps without teeth.
Once identified, the nerve is protected with a small vessel loop. It is
important not to clamp any instruments such as a hemostat onto the
vessel loop, as the traction from the instrument can cause an injury to
the peroneal nerve. The nerve can sustain injury with simple exposure,
and the patient should be counseled in this regard preoperatively. Once
the nerve is identified and protected, exposure of the posterolateral
structures is carried out by making two additional fascial incisions.
The first develops the interval between the biceps femoris tendon and
the iliotibial band, giving access to the distal attachments of the
lateral collateral ligament (LCL), the popliteofibular ligament, and
the popliteus tendon. The lateral gastrocnemius is mobilized
posteriorly, exposing the posterolateral capsule. The second fascial
incision splits the iliotibial tract over the lateral epicondyle. This
provides access to the proximal attachments of the LCL and
popliteofibular ligament as well as the femoral attachments of the
posterolateral capsule. In the exposure of combined ligament injuries
of the posterolateral knee, the dissection planes are usually already
formed secondary to the displacement occurring with the knee injury.
Straying posterior to the lateral gastrocnemius places the popliteal
neurovascular structures at risk. The artery is often slightly lateral
of midline at the level of the joint, so great care must be exercised
in this area. This approach provides minimal access to the medial side
of the knee. If the PLC and LCL are intact, the PCL cannot be exposed
from this approach. Furthermore, if the approach is performed for a
complete injury of the PCL and the PLC, an intact ACL will prevent
access medially to the PCL insertion on the tibia. In such a scenario,
a posteromedial incision will be necessary to access the PCL insertion
site¹³³ (Fig. 54-17).

This approach provides access to the popliteal fossa,
the posterior surface of the femoral condyles, the posterior knee
capsule, and the posterior aspect of the distal tibia. It is useful for
resection of tumors from the popliteal fossa or posterior knee, nerve
or blood vessel exploration, reattachment of PCL avulsions from the
tibia, and reconstruction of the PCL.

The patient is placed in a prone position. A tourniquet
is placed on the thigh. The skin incision begins posteromedially over
the semitendinous and semimembranosus tendons, curves transversely
across the knee flexion crease, and then proceeds laterally over the
fibular head and fibula. The sural nerve and short saphenous vein are
identified and the vein is followed proximally to the trunk of the
popliteal vein. The short saphenous vein and sural nerve pierce the
fascia of the posterior knee in the area of the distal flexion crease.
The medial and lateral gastrocnemius muscle heads are identified. The
popliteal nerve, vein, and artery are exposed with ligation of the
genicular branches as required. It is best to approach the vessels from
the medial, proximal end of the fossa and work distally. The vessels
and nerve may be retracted together or, alternatively, the vessels may
be retracted medially and the nerve laterally.

Exposure
of the femoral and tibial surfaces is performed through the
neurovascular interval, with care taken to protect these structures.
Distally, the popliteus muscle is found on the posterior aspect of the
tibia and can be retracted to allow exposure of the proximal tibia (Fig. 54-18).
Blunt dissection is safest for access to the proximal tibia. This is
best performed with a finger or blunt scissors to avoid injury to the
neurovascular bundle. Release of the medial head of the gastrocnemius
can be performed to allow better exposure of the proximal tibia.¹³³

The greatest danger is the potential for injury to the
neurovascular structures of the popliteal fossa that are directly
exposed in this approach. Because the patient is in a prone position,
access to the anterior knee is extremely difficult, making PCL
reconstruction impossible without turning the patient supine during a
portion of the procedure (requiring repeat prepping and draping). To
approach the tibial attachment of the PCL, the posteromedial approach
is preferred as it does not require a prone position (and the
associated higher pulmonary and cardiovascular risk with a trauma
patient) followed by reprepping and draping for the supine portion, and
does not require direct exposure of the complicated anatomy of the
popliteal neurovascular structures.

Nonsurgical treatment of the dislocated knee is
important historically and occasionally is an important option today
when circumstances dictate. Historically, knee dislocations were
reduced and treated in a cylinder cast. In the 1960s and 1970s, there
were studies supporting both nonoperative treatment and operative
repair of the torn structures, but there was no clear consensus.^*
More recently, with improved surgical techniques including
arthroscopically assisted ligament reconstruction, and increased
understanding of the ligamentous anatomy and biomechanics around the
knee, there has been evidence supporting surgical reconstruction.³⁰,³¹,³²,³³,¹⁰⁸ These studies demonstrate improved ligamentous stability of the knee as well as improved function postoperatively.

Multiligament knee injuries in specific circumstances
warrant nonoperative treatment. Critically ill patients unable to
tolerate a surgical procedure and patients with grossly contaminated
wounds and/or significant soft tissue injuries around the prospective
surgical site may be candidates for nonoperative management. Also, in
the elderly sedentary person, it may be best to treat the knee
dislocation nonoperatively.¹⁰⁹

The most basic method of nonoperative management is a
long leg or cylinder cast. A long leg knee brace locked in extension
may also be used. The brace may be particularly helpful if there are
significant wounds about the knee that a cast would conceal. The knee
brace is often helpful in the setting of a critically ill patient. The
brace allows easy access and evaluation of the injured extremity. If
the cast or brace does not afford enough stability to maintain the knee
in a reduced position, a knee-spanning external fixator may be
necessary. The external fixator also provides better access to soft
tissue injuries if it can be positioned away from contaminated wounds.
We include spanning external fixation as a definitive form of
nonoperative management even though we understand it is applied in the
operating room. In this situation, it is being used the same way as the
brace or cast, but it is providing better stability and access to the
soft tissues. Regardless of the type of nonsurgical management, it is
critical to obtain high-quality radiographs to establish that the knee
is well reduced with no subluxation. If an adequate reduction has been
obtained, it must be followed by frequent radiographs to verify
continued reduction of the knee. Our protocol for treatment with a cast
or external fixator involves 7 to 8 weeks of immobilization, followed
by removal of the cast or external fixator and manipulation under
anesthesia. Rehabilitation then concentrates primarily on achieving
maximum knee motion.

Improved knee reconstruction techniques have led to a
strong trend toward surgical reconstruction following knee
dislocations. At least nine studies have been published that
specifically compare operative and nonoperative treatment of patients
following knee dislocations.^* All of these studies are
retrospective and feature a wide variety of surgical techniques and
rehabilitation protocols. A total of 172 patients were treated
surgically in these nine studies, compared with 84 patients treated
nonoperatively. All nine authors concluded that surgical treatment is
superior to nonoperative treatment for patients following knee
dislocations. Roman et al.¹⁰⁰ did
note increased stiffness with patients treated surgically, but this
experience was not reported in the other studies. All of the authors
found marked improvements in stability in patients who had surgical
reconstruction of their knee compared with those who had nonoperative
treatment. Richter et al.⁹⁷
published a large series of 77 knee dislocations. They reported that
patients treated surgically had better motion, stability, and return to
work and recreational activities. They also noted improved outcomes
using the Lysholm Knee Scale score (a score of 78 for surgical versus a
score of 65 for nonoperative treatment) and the International Knee
Document Committee (IKDC) score (41% were severely abnormal in the
nonoperative group) following surgical treatment. Wong et al.¹⁴⁰
recently reported a 27% long-term incidence of instability following
reconstruction compared with a 91% incidence of instability following
nonoperative care. Rios et al.⁹⁹ reported improved results using both the Lysholm and Meyers scores in patients following surgical care.

There are very little data in the literature regarding
the ideal time to undertake reconstruction of a knee dislocation. There
are many variables that will always require this decision to be
individualized to each patient. These variables include associated
injuries, whether the dislocation is open or closed, the degree of
contamination if open, and the degree of associated soft tissue injury
if closed. Most surgeons consider surgery performed within the first 3
or 4 weeks to be an acute repair or reconstruction of the knee
dislocation, while anything delayed longer than a month would be a
chronic knee reconstruction. There are advantages to allowing the
patient to recover from the initial effects of the trauma and
performing the reconstructive surgery during the chronic phase. These
advantages include a better nutritional state after recovering from the
trauma, improved local soft tissue condition, and therefore potentially
a decreased risk of infection and wound dehiscence. There are different
advantages to reconstructing a knee dislocation during the acute injury
phase. These advantages include potentially decreased total trauma
recovery time because the patient is recovering from the knee
reconstruction at the same time as he or she is recovering from other
injuries sustained in the trauma, the ability to repair torn structures
if the tissue quality allows, and improved healing of the injury to the
capsule of the knee by taking advantage of the inflammatory reaction
associated with the initial injury.

Two reports have focused on the topic of the ideal time
to initiate the repair or reconstruction of the knee following a
dislocation. Harner et al.⁴⁴
retrospectively evaluated 47 consecutive knee dislocations of patients
who presented to a single level 1 trauma center. Fourteen patients were
eliminated from consideration because of confounding injury variables
such as open dislocations or vascular injuries. The remaining 31 were

divided
into two groups. Nineteen were treated acutely, which the authors
defined as less than 3 weeks following their index injury. The
remaining 12 patients were treated during the chronic period which they
defined as more than 3 weeks following their dislocation. They noted no
significant difference in the final arc of motion obtained by patients
in the two groups, but better final knee stability was seen in the
acute treatment group. The difference was significant when they
evaluated their patients for posterior tibial translation, where 84% of
the patients in the acute group were stable compared with only 50% of
the patients in the chronic group. They also evaluated the patients
using multiple outcome scores. In all cases, patients in the acute
group attained better scores than patients in the chronic group. Using
the Lysholm Knee Scale score, acutely treated patients achieved a mean
score of 91 compared with 80 for the chronic group. On the Meyers
functional rating, 84% of the acute group were rated good or excellent
compared with only 58% of the chronic group. On the Knee Outcome Survey
Activities of Daily Living score, the acute group scored a mean of 91
compared with 84 for the chronic group. Finally, on the Knee Outcome
Survey Sports Activities Scale, patients in the acute group scored a
mean of 89 compared with a mean of 69 for patients in the chronic
group. They concluded that patients with knee dislocations treated
acutely had better functional outcomes than patients treated on a
delayed basis.

Liow et al.⁶² also
evaluated the time to treatment for patients following knee
dislocations. Their retrospective study evaluated 21 patients with 22
knee dislocations. They had two very distinct groups. Eight patients
had their knee dislocation treated within 2 weeks of the index injury,
while the remaining 14 knees were not treated until at least 6 months
following injury. Again, there was no significant difference between
the two groups in total arc of motion attained. They did note that the
only patient in their series who developed arthrofibrosis was in the
acute treatment group. The patients in the acute treatment group had a
mean Lysholm Knee Scale score of 87 compared with a score of 75 for the
chronic treatment group. The Tegner Activity Scale levels were similar,
with a mean score of 5.0 for the acute group and 4.4 for the chronic
group.

The initial data appear to favor acute repair or
reconstruction following knee dislocations. Clearly, additional data
are needed before we can strongly recommend acute treatment.
Additionally, the surgeon will still need to individualize care to each
patient considering the associated injuries.

Twenty-two studies have been published that document range of motion following knee dislocations.^* Analysis of these studies reveals some outcome trends. When comparing studies published prior to 1994³,³⁶,⁷⁶,¹⁰⁰,¹¹⁰,¹¹⁴ with those published in the last 15 years,^†
there has been a trend toward an improving arc of motion over time. The
mean arc of motion in the pre-1994 studies was 106 degrees, compared
with 123 degrees in the more contemporary studies. The reasons for the
improvement are most likely multifactorial, including improved
reconstruction techniques and the use of aggressive rehabilitation
protocols. The use of continuous passive motion machines was common in
many studies.^‡ Despite these impressive final ranges of
motion, arthrofibrosis has remained a major problem for patients
following knee dislocations. The remarkable improvements seen in total
arc of motion have frequently required secondary surgical treatment in
the form of either manipulation under anesthesia or arthroscopic lysis
of adhesions. Reports of the need for surgical intervention to treat
arthrofibrosis have varied from a low of 5%⁶² to a high of 71%⁸⁸
following acute dislocations. The mean reported incidence of
arthrofibrosis requiring surgical intervention was 29% of the patients
from 14 different studies.^§ The incidence of arthrofibrosis
is improving with time. A similar analysis of the literature completed
5 years ago revealed an incidence of arthrofibrosis of 38%¹²⁰
compared with a rate of 29% in the current analysis. In the final
analysis, motion has improved remarkably over the years for patients
following knee dislocations. However, while the incidence of
arthrofibrosis is improving, it remains a major problem. Great
vigilance by the treating surgeon regarding motion is required in the
postoperative period. Nearly one third of patients will require
surgical intervention to treat arthrofibrosis and maximize the final
arc of motion.

Most surgeons believe that pain and loss of motion
rather than instability are the primary problems for patients treated
surgically following knee dislocation. While published outcome studies
support that pain and loss of motion are major problems, the data
demonstrate that recurrent instability following reconstruction remains
a major problem as well. One difficulty in assessing the literature is
the wide range of methods used for reporting knee stability in the
literature. For the purposes of this chapter, instability has been
defined as grade 2 or 3 on a scale 0 to 3, KT 1000 or 2000 Ligament
Arthrometer (MEDmetric Corporation, San Diego, CA) data that are
greater than a 3-mm side-to-side difference for either anterior or
posterior translation or greater than 5 mm for total anteroposterior
translation, or groups C or D on the IKDC score.

Six outcome studies have reported stability data following nonoperative treatment of patients with knee dislocations.³,³⁶,⁷⁶,⁹⁷,¹⁰⁰,¹⁴⁰
In five of the six studies, the authors reported 100% of the patients
had residual instability in at least one direction following treatment.
In the remaining study,¹⁴⁰ the authors reported 91% of their patients had residual instability as a final result using the criteria defined above.

Twenty-one outcome studies have reported stability
outcome data following surgical repair or reconstruction of knee
dislocations.^* All of the studies that report on all the
ligament groups torn in the dislocation list at least an 18%
instability rate. The range is from 18% to 100%, with a mean of 42% of
patients reported to have at least one ligament with chronic
instability following repair or reconstruction. In many cases, more
than one ligament in an individual patient meets the criteria for
chronic instability. Anterior and/or posterior instability is reported
more frequently than instability of the posteromedial

or
PLCs. It is interesting that while motion results have improved
slightly since our analysis 5 years ago, chronic knee instability has
gotten slightly worse.

Pain is the most frequent problem for patients following
knee dislocation, and it occurs to some degree in virtually every
patient. It can vary from occasional pain that has minimal impact on
daily activities to severe and disabling pain. There are a wide variety
of potential causes of knee pain following dislocation, including:
articular cartilage injury, chronic instability, arthrofibrosis, and
posttraumatic arthritis. Severe pain may cause a patient with a stable
knee and excellent motion to have a poor functional outcome. It is very
common for patients with arthrofibrosis to experience significant pain
at their limits of motion.

The incidence of pain varies from less than 25% of patients in studies by Yeh et al.¹⁴³ and Marinek et al.⁷² to 46% to 68% in four other series.³,⁷⁰,⁹⁷,¹¹⁴
Two studies found that chronic pain was less prevalent in patients who
had their dislocation reconstructed acutely compared with those
patients who had reconstruction of a chronic dislocation.³,⁸⁸
Two possible causes of increased pain in patients treated chronically
are simply the pain associated with chronic instability and subsequent
osteoarthritis.⁴ Additional outcome
studies that specifically address the issue of chronic pain are needed.
Patients who sustain knee dislocations must be prepared for the
possibility of significant long-term pain.

Knee dislocation is a severe injury that is frequently
associated with multiple injuries. The vast majority of patients are
unable to return to work immediately following treatment for a knee
dislocation. Eight outcome studies address the issue of patient
employment following knee dislocations.^* Most patients are
able to resume gainful employment, with many being able to resume
high-intensity occupations following contemporary reconstruction
techniques and extended rehabilitation. The combined results from the
eight studies show that 93% of patients were able to resume some type
of work following knee dislocation, with 31% having to accept a job
that was either light duty or less physically demanding than their
previous job. Again, there was a higher rate of return to employment
for patients who had treatment in the acute period compared with those
who had chronic reconstructions. Richter et al.⁹⁷
reported that only 58% of their patients with a repaired chronic
dislocation returned to work compared with 85% of those undergoing an
acute repair. It is clear from these data that it is possible for
patients to resume work, including high physical demand fields,
following reconstruction of knee dislocations.

Functional outcome scores are being used increasingly to
report on patient results following knee dislocations. One problem is
that there are a variety of scores used in the literature. The most
commonly used scores are the Meyers,⁷⁷ Lysholm, and IKDC scores. Five authors reported their results using Meyers’ criteria.³⁶,⁴⁴,⁷⁷,⁹⁹,¹⁰¹
There were excellent results in 29%, good in 53%, fair in 11%, and poor
in 7% of those patients treated surgically. When evaluating
nonoperative treatment with the Meyers score, there were no excellent
results and 6% good, 16% fair, and 78% poor results. Clearly, the
outcomes reported using the Meyers score favor surgical treatment of
knee dislocations.

Thirteen studies reported results using the Lysholm Knee Scale score.^*
A score of 90 to 100 is considered excellent, 80 to 89 is good, 70 to
79 is fair, and below 70 is a poor result. The mean score for all of
these studies for patients treated with surgical reconstruction was 84,
with a range from 75 to 91. Three studies reported Lysholm scores on
patients treated nonoperatively.⁸²,⁹⁷,⁹⁹
The mean Lysholm score for patients in these three studies was 64.
Clearly, surgical treatment yields better results than nonoperative
treatment using the Lysholm score. Two studies compared Lysholm scores
for patients treated surgically in the acute time period with those
treated during the chronic period.⁴⁴,⁶²
The mean Lysholm score for patients treated acutely was 89, while the
score for those treated chronically was 78. Again, acute reconstruction
appears to have better results as measured by the Lysholm score in
these two studies.

Seven outcome studies have reported their results using the IKDC scoring system.⁴⁴,⁶²,⁷⁰,⁷²,⁹⁷,¹⁰¹,¹²⁹
This system uses the following designations: Group A, normal knee;
Group B, nearly normal knee; Group C, abnormal knee; and Group D,
severely abnormal knee. Combining the seven studies, we get the
following results: Group A, 4%; Group B, 37%; Group C, 41%; and Group
D, 17%. The results, using this scoring system, are still
disappointing, with nearly 60% of patients falling into the abnormal or
severely abnormal categories. Richter et al.⁹⁷ reported on IKDC results in nonoperatively treated patients and found that 94% were either abnormal or severely abnormal knees.

Assessing patient outcomes regarding return to athletic
activities is difficult because of the variability of the definition of
return to activities among studies. Many patients with knee dislocation
are not involved in competitive athletics. The reported level of return
of patients to some recreational activity following knee dislocation
varies from 0%³ to 97%¹¹⁹
with a mean of 65%. Richter et al. reported that 56% of their patients
treated surgically returned to athletic activity compared with only 17%
of their patients treated conservatively. They reported that 49%
returned to their prior level of play, 40% decreased by one level of
play, and 10% decreased by two levels.⁹⁷
Seven studies reporting on patients treated surgically noted that 76%
of patients returned to some form of athletic or recreational
activities. Only 39% of patients were able to return to their preinjury
level of activity.^**

In summary, the number of published manuscripts that
report outcome scores continues to increase. Despite the limitations of
small series and retrospective data already mentioned, combining the
data can yield valuable information. Surgical treatment using
contemporary techniques clearly yields results that are superior to
nonoperative treatment. There is also a small amount of data that
suggest that treatment in the acute time period may yield results that
are superior to the results obtained for patients treated on a delayed
basis in the chronic time period.

Motion,
pain, and instability are all major problems for patients following
knee dislocations. Motion is improving slowly over the years, possibly
as a result of aggressive rehabilitation protocols. However, chronic
instability remains a problem for up to 42% of patients following
reconstruction. This number has risen slightly over the past 5 years.
Most patients are able to achieve adequate stability to perform most
activities of daily living, but stability to resume athletic activities
and other high stress activities is much more elusive. The outcomes
regarding patient employment are very encouraging and suggest patients
can return to productive lives following knee dislocation. Finally,
outcome scores using the Meyer and Lysholm scores are frequently good
to excellent. The outcomes using the IKDC score are far less favorable.

The concept of a double-bundle PCL reconstruction is an
attempt to recreate the functional anatomy of the PCL. The larger
bundle is the anterolateral bundle, which is tensioned with the knee
place in 70 degrees of flexion. The smaller posteromedial bundle is
tightened with the knee in relative extension (approximately 30 degrees
of flexion). The combination of these two bundles may improve the
stability of the PCL through the entire arc of motion. A recently
published study of 30 PCL reconstructions combining the inlay technique
and double-bundle technique reported excellent stability 2 to 3 years
following reconstruction.⁵¹ Apsingi et al.⁹
evaluated nine cadaver knees, comparing single- to double-bundle PCL
reconstruction in both posterior and posterolateral laxity. They found
double-bundle reconstruction to prevent posterior laxity better than
single-bundle reconstruction at full extension, the difference being
1.6 mm; there was no difference at all other flexion angles.

Berg initially reported the concept of creating a trough
on the back of the tibia and inlaying a graft at the anatomic insertion
of the PCL, rather than drilling a transtibial tunnel. He described the
problem of the graft having to traverse around the “killer turn” when
the transtibial tunnel is used and that this might lead to graft
failure.¹¹,²⁰
Markolf confirmed in a biomechanical study that traversing the “killer
turn” caused one third of cadaver grafts to fail and the remainder to
have significantly more thinning and elongation than inlay grafts.⁹³,¹²⁵ Seon and Song¹⁰⁸ and MacGillivray et al.⁶⁷
both retrospectively compared the inlay to the transtibial technique
with a minimum 2-year follow-up. The former study cited good clinical
outcomes in both groups, while the latter demonstrated that neither
technique fully recreates anterior-posterior stability with a
single-bundle femoral attachment site.

A new concept in the reconstruction of knee dislocations
involved the use of a hinged external fixator (Compass Knee Hinge,
Smith & Nephew, Memphis, TN) and early motion following surgical
reconstruction of the knee. A preliminary report indicated that less
than 10% of patients using the hinge had any ligament instability
following early motion rehabilitation, compared with approximately 40%
of patients treated with a conventional brace.¹²²
The data are very preliminary but create possible treatment options for
patients with marked knee instability. Indications for the use of the
hinged external fixator combined with early motion following
reconstruction include KD III, KD IV, and KD V knee dislocation
patterns.

The conventional wisdom has been that the PLC should be
repaired if surgery is accomplished within 3 or 4 weeks of injury and
the tissue quality appears good. However, there were no published
series supporting that view. Stannard et al.¹¹⁵
recently presented a series that compared repair using suture anchors
to bone with a modified two-tailed reconstruction technique that
recreates the popliteus tendon, the popliteofibular ligament, and
lateral collateral ligament. Both groups underwent early motion
rehabilitation. The authors found a significant improvement in
stability of the PLC with reconstruction compared with repair. These
findings may not occur if early motion rehabilitation is not used or if
alternative repair techniques are used.

For proper management of fracture-dislocations of the
knee, attention must be given in particular to the normal anatomy of
the tibial and femoral condyles. The femoral condyles are asymmetric in
size and shape. The medial femoral condyle is approximately 1.7 cm
longer than the lateral condyle in its outer circumference. This
asymmetry in length produces axial rotation of the tibia on the femur
during flexion and extension.¹⁹ With
regard to width, the lateral condyle is slightly wider than the medial
when measured at the center of the intercondylar notch. In the sagittal
axis of the knee, the lateral femoral condyle is longer or more
anterior than the medial condyle. In the coronal or anterior-posterior
plane, the medial femoral condyle projects farther distally than does
the lateral condyle.⁵,⁴¹
When the femur is viewed along its anatomic axis, this appearance
becomes obvious. However, in normal weight-bearing alignment, the
condyles appear to be equal in length. The parallel condylar surfaces
are created by the mechanical axis configuration of the lower extremity.

The tibial plateau joint surface is complex. The normal
tibial articulation involves the menisci to provide congruity with the
distal femoral condyles, and in reality the menisci should be
considered tibial extensions. The menisci function to create conformity
between the flat tibial and curved femoral surfaces. The medial condyle
of the tibia is nearly flat and has a larger surface area than the
lateral condyle. The lateral condylar surface is slightly concave. Both
condyles have a 10-degree posterior inclination to the tibial shaft in
the saggital plane. Bordering the notch are the tibial spines (or
tubercles), both medial and lateral, which function to stabilize the
condyles from side-to-side motion. The interspinous area is void of
hyaline cartilage and is the insertion site for the meniscal horns and
the cruciate ligaments.

The cruciates insert on the intertubercular sulcus and not on the spines themselves.⁴,³⁹,⁴¹,⁷⁸,⁸⁴
Due to the asymmetry of the tibial condyles and the relative covering
of each respective meniscus, the medial tibial plateau does not
tolerate joint surface incongruity as well as does the lateral tibial
plateau. With the more extensive covering of the lateral meniscus, up
to 10 mm of lateral plateau displacement may be tolerated clinically,
depending on the exact configuration of the fracture. Although anatomic
reduction is considered the goal in fracture management, the clinician
should remember the differences specifically between the medial and
lateral tibial plateaus.

The anatomy of the popliteal neurovascular structures
explains their susceptibility to injury. The popliteal artery and vein
course from the fibrous tunnel of the adductor hiatus through the
popliteal space (giving off the five geniculates) and exit through the
fibrous arch of the soleus muscle. Being securely fixed proximally at
the adductor hiatus and distally at the soleus arch, the popliteal
artery and vein can be torn or stretched with the exaggerated
tibiofemoral displacement or hyperextension that produces the
dislocation.

Surgical treatment of knee fracture-dislocations
frequently involves the use of open approaches. Although the lateral
compression type of injuries with an associated contralateral ligament
injury can be treated using arthroscopic-assisted approaches, the
clinician is usually faced with a large condylar fragment, gross
capsular disruption, and the need for open reduction of the fracture
fragment. Furthermore, with acute injuries to both the condyle and
ligament, the risk of arthroscopic fluid extravasation is a concern, as
it can result in a compartment syndrome or vascular insufficiency.
Certainly delayed ligament reconstruction is best performed with
arthroscopic approaches; in contrast, early surgery involving fracture
fixation and frequently ligament repair or reconstruction requires an
open approach. Selection of the specific approach depends upon the
fracture configuration and the ligaments involved; treatment may
require the use of combined incisions. In such scenarios, the
importance of maintaining an adequate soft tissue bridge between
incisions must be kept in mind. The posteromedial, posterolateral, and
anterior approaches are commonly used in the treatment of these
fracture-dislocations.

Moore type I fractures (Fig. 54-35)
can be treated through either a direct posterior or posteromedial
approach. Galla and Lobenhoffer described a direct posteromedial
approach that does not cross the popliteal crease and does not divide
the tendon of the medial head of the gastrocnemius.³⁸ This report was introduced into the English language by Fakler et al.²⁸
With the patient prone and a thigh tourniquet in use, a 6- to 8-cm skin
incision is made along the medial border of the medial head of the
gastrocnemius, its proximal extent being at the joint line. The
popliteal crease is not crossed. The small saphenous vein is identified
between the medial and lateral heads of the gastrocnemius. Medial to
the vein is the medial sural cutaneous nerve; the nerve is not
retracted, as dissection is carried medial to the medial head of the
gastrocnemius. The medial head is then retracted laterally, carefully
placing the retractor to avoid the major neurovascular structures. The
semimembranosus is also dissected and retracted medially; its insertion
on the posteromedial tibial plateau is not disrupted. The superior
border of the popliteus is elevated off the posterior tibia, exposing
the fracture fragment. If further distal or medial exposure is
required, the soleus origin or the semimembranosus insertion can be
elevated, respectively. Reduction of the posterior fracture fragment
can be accomplished by placing the knee in either flexion, to relax the
posterior capsule, or hyperextension.¹³,²⁸