Acetabular Fractures: Extended Iliofemoral Approach

Three main stages characterize the dissection: (a)
elevation of all the gluteal muscles with the tensor fascia lata, (b)
division of the external rotators of the hip, and (c) an extended
capsulotomy along the lip of the acetabulum. The end result is complete
exposure of the outer aspect of the ilium and the whole posterior
column inferiorly to the upper part of the ischial tuberosity.
Furthermore, the approach may be extended to allow a limited exposure
of the internal iliac fossa and the anterior column to the level of the
iliopectineal eminence. This allows simultaneous exposure of both
columns and permits direct visualization of the reduction and fixation
of the anterior and posterior columns (see Fig. 43.4).
The articular surface of the acetabulum along with the femoral head may
also be visualized if this approach is combined with a surgical
dislocation of the hip.

The physician performing an extended iliofemoral
approach requires special training and a familiarity with the complex
anatomy of the pelvis, particularly the many neurovascular structures
that are encountered. Those structures that require identification are
listed below.

General or spinal anesthesia is administered. We prefer
the continuous epidural anesthesia as it provides improved
postoperative pain relief. A Foley catheter is placed in the patient’s
bladder. The patient is supported on a beanbag and placed in the
lateral decubitus position on a radiolucent operating table or fracture
table, depending on the surgeon’s preference.

Vascular access in two separate sites with large-bore
catheters is important for these lengthy procedures in which
significant blood loss is common. The patient’s age and medical
condition often dictate placement of an arterial or central line.

We routinely use an intraoperative cell saver to
minimize transfusion requirements. This permits recycling of about 20%
to 30% of the effective blood loss and is best used when blood loss of
more than 2 L is expected.

The hip is kept extended and the knee flexed throughout
the procedure to minimize sciatic nerve injury. In addition,
intraoperative sciatic-nerve monitoring with spontaneous
electromyography (EMG) and somatosensory evoked potentials (SSEP) is
used in all cases. The entire pelvis, hip, abdomen, and involved
extremity are prepped free, and sterile subdermal electrodes are
inserted. The sensory electrodes are inserted adjacent to the common
peroneal and posterior tibial nerves and the motor adjacent to the
tibialis anterior, peroneus longus, abductor hallucis, and flexor
hallucis brevis. The ground is inserted in the heel.

The incision is in the form of an inverted “J” (Fig. 43.7)
and begins at the posterior–superior iliac spine, extending along the
iliac crest toward the anterior–superior iliac spine (ASIS). From here,
the distal arm of the incision proceeds along the anterolateral aspect
of the thigh for a distance of 15 to 20 cm (Fig. 43.8).
The surgeon has a tendency to make this arm more medial than is
desired. To avoid this, one should visualize a point 2 cm lateral to
the superolateral pole of the patella. With the leg held in neutral
rotation, this location is generally in line with the desired incision.
Furthermore, a gentle posterior curve may be helpful in obese patients.

The fascial periosteal layer at the iliac crest is identified (Fig. 43.9)
and divided sharply along its avascular “white line,” where bleeding
will be minimized. Often it is easiest to start in the area of the
gluteus medius tubercle where landmarks are more obvious and to
progress posteriorly and anteriorly from this point. Posteriorly, the
strong fibrous origins of

the
gluteus maximus should be sharply released from the crista glutei.
Depending on the starting location, the tensor fascia-lata muscle and
the gluteus medius are subperiosteally released in a stepwise fashion
from the outer aspect of the iliac crest (Fig. 43.10).
Using an elevator, the musculature along the external surface of the
iliac wing is released up to the superior border of the greater sciatic
notch and anterosuperior aspect of the hip joint capsule (Fig. 43.11).
During this segment of the exposure, care must be taken to identify the
superior-gluteal neurovascular bundle, which is at risk as it exits
from the notch.

Attention turns next to the anterior portion of the approach (see Fig. 43.9).
The distal limb of the incision is carried over the fascia covering the
tensor fascia-lata muscle, and the muscle sheath is entered. The
surgeon must stay within the bounds of the sheath, as this will keep
the dissection lateral to the lateral femoral-cutaneous nerve, sparing
the majority of its branches. It is often helpful to open the sheath
from distal to proximal.

Next, the tensor fascia muscle is reflected off its
fascia and retracted laterally and upward to expose the floor of the
sheath and fascia overlying the rectus femoris muscle (see Fig. 43.10).
Small vessels from the superficial circumflex artery are divided and
coagulated close to the bone between the superior and inferior spines.
Distally, the incision must be long enough to expose the inferior
aspect of the muscle belly. This facilitates further release of the
gluteal muscles from the crest.

The fascia overlying the rectus muscle is divided
longitudinally and horizontally, and its reflected head and direct
heads retracted downward and medially to expose a very strong
aponeurosis (the “no name” fascia) over the vastus lateralis muscle
(see Fig. 43.10). When the rectus is
retracted, a constant small vascular pedicle reaching the lateral
border of the muscle always requires coagulation. The aponeurosis can
be divided longitudinally to expose the ascending branches of the
lateral circumflex vessels, which must be isolated and ligated (see Fig. 43.11). Should the upper portion of this exposure be unnecessary, these vessels can occasionally be spared.

Next the thin sheath of the iliopsoas muscle is exposed
and longitudinally incised. This allows the use of an elevator to strip
the fibers of the psoas from the anterior and inferior aspects of the
hip capsule. The exposure of the iliac wing is complete when the
reflected head of the rectus femoris is sharply released from its
insertion.

The gluteus minimus tendon is identified as it inserts
into the anterior edge of the greater trochanter and tagged and
transected, leaving a 3- to 5-mm cuff for repair (Figs. 43.11 and 43.12).
The gluteus minimus muscle also has extensive attachments to the
superior aspect of the hip capsule that may need to be released.
Posteriorly and superiorly, the gluteus medius tendon, measuring 15 to
20 mm in length, is also isolated, tagged, and transected, leaving a 3-
to 5-mm cuff (see Figs. 43.11 and 43.12).
The surgeon must transect and tag these structures sequentially and
carefully for subsequent reattachment. The tensor fascia-lata and
gluteal muscles are held in continuity as a flap and reflected
posteriorly to expose the external rotators and sciatic nerve (see Fig. 43.12).

The tendons of the piriformis muscle, obturator internus
muscle, and the inferior and superior gemelli muscles are tagged and
transected as in the Kocher-Langenbeck approach (Figs. 43.12 and 43.13). The tendinous femoral insertion of the gluteus maximus is identified, tagged,

and released with a cuff for repair (see Fig. 43.13).
It cannot be overemphasized that the quadratus femoris and its blood
supply to the femur via the ascending branch of the medial femoral
circumflex artery must be preserved. The dissection is now complete
(see Fig. 43.1).

The piriformis muscle can be followed toward the greater
sciatic notch and the obturator internus muscle to the lesser sciatic
notch. A Homan or sciatic nerve retractor is then placed into the
lesser notch, allowing complete exposure to the posterior column of the
acetabulum. The surgeon must ensure that the tendon of the obturator
internus maintains its position in the lesser notch between the sciatic
nerve and the retractor. Should additional retraction be required, a
blunt Homan is gently placed into the greater sciatic notch, with the
surgeon aware that no structure is protecting the nerve. The distal
portion of the posterior column can be visualized to the ischial
tuberosity through sharp dissection of the origin of the hamstring
muscles proximally, if necessary.

Although medial exposure of the anterior column is limited by the iliopsoas muscle and the iliopectineal fascia (see Figs. 43.4, 43.9, and 43.13),
further access to the internal iliac fossa and acetabulum is possible.
This access is obtained by subperiosteal dissection beneath the
sartorius and direct head of the rectus or by osteotomy of the superior
and inferior iliac spines, which will, respectively, release these
muscles (see Figs. 43.5 and 43.13).
The insertion of the external oblique muscle onto the crest can also be
subperiosteally released to reveal the inner table of the pelvis, which
is further exposed by the surgeon stripping off the iliacus muscle with
a periosteal elevator. However, extensile exposure of the outer and
inner tables of the iliac wing, especially in the presence of local
fractures, will create a risk of iliac-bone devascularization.

Although devascularization of the iliac wing is rare,
Matta warned of its occurrence, especially in associated both-column
fractures. To avoid devascularization of the iliac bone in this case,
he suggested leaving, at a minimum, the direct head of the rectus
femoris and anterior hip capsule attached to the anterior column. Also
of concern with this exposure is the blood supply to the dome of the
acetabulum, which is at risk during dissection of the anterior–inferior
iliac spine.

Displaced acetabular fractures often tear the hip-joint
capsule. If not present, exposure of the acetabular articular surface
can be obtained with a marginal capsulotomy, leaving a cuff of tissue
for repair. Once the hip joint is exposed, distraction with either a
Schanz screw placed into the femoral head or a femoral distractor will
facilitate visualization (Figs. 43.1 and 43.14).
The visualization is important for evaluating the articular reduction,
ruling out any intra-articular hardware, and removing any incarcerated
osteochondral fragments.

Once the exposure of the extended iliofemoral approach
has been completed, the fracture can be reduced according to the
preoperative plan. The soft-tissue flaps must be kept moist with wet
sponges and periodic irrigation throughout the procedure.

Several regions of bone are optimal for screw placement.
These include the iliac crest, the superogluteal ridge, the greater
sciatic buttress (above the sciatic notch and to the anterior–inferior
iliac spine), the anterior column, and the posterior column. Extra-long
screws, ranging from 50 to 120 mm, should be available.

In a transverse fracture, rotational malalignment of the
inferior portion of the acetabulum may be found. In the T-type fracture
patterns, the anterior and posterior fragments may be

separate
and both columns may have become displaced and malrotated. Usually, the
anterior segment has medial displacement of its inferior portion so
that the radius of curvature of the acetabulum is greater than that of
the femoral head. In both transverse and T-type acetabular fractures,
reduction is achieved with a pelvic-reduction clamp attached to 4.5-mm
screws placed proximal and distal to the posterior-column fracture. The
pelvic-reduction clamp initially allows distraction for debriding of
the fracture surfaces, and then facilitates manipulative reduction of
the fracture. A bone spreader in the fracture site can also facilitate
exposure of the fracture or the joint (Fig. 43.15A).
Additional control of rotation is provided by a Schanz screw placed
into the ischium and a pelvic clamp in the greater sciatic notch.

For the reduction of the T-type and more comminuted
variants, the anterior column should be reduced first with respect to
the residual acetabular roof portion of the ilium. The adequacy of
reduction of the posterior column can be visualized by direct
assessment of the articular surface and also with digital palpation
through the greater and lesser sciatic notches.

Before definitive reduction, a gliding hole can be
inserted into the proximal aspect of the posterior column from superior
to inferior (see Fig. 43.15B). This hole helps the surgeon assure that the gliding hole is in the middle of the posterior column.

A gliding hole can also be inserted from the lateral
aspect of the iliac wing into the anterior column distal and medial to
the articular surface. Generally, this requires the insertion of a lag
screw 6 cm proximal to the superior aspect of the articular surface and
2 cm posterior to the gluteal ridge. The lag screw is then angled from
posterosuperior to anteroinferior directly down the superior pubic
ramus to secure the anterior column of the acetabulum. In large
individuals, this can be accomplished with a 4.5-mm cortical screw. In
small individuals, including most women, a 3.5-mm cortical screw is
preferred. Care must be taken to assure that this screw remains
extra-articular and also does not penetrate the anterior aspect of the
superior ramus in the area of the iliopectineal eminence where the
femoral vessels are in close proximity. The use of intraoperative
fluoroscopy for the insertion of this screw is highly recommended.

Proper placement of pelvic-reduction forceps with
respect to the plane of the fracture and geometry of the osseous
surfaces is crucial for an adequate reduction. A variety of instruments
is available to facilitate reduction: narrow curved osteotomes, bone
hooks, ball spike pushers, King-Tong and Queen-Tong forceps, and the
Farabeuf and pointed reduction clamps (see Fig. 43.15C).

To ensure an anatomic reduction of the acetabulum, the surgeon should work from the periphery toward the acetabulum (Fig. 43.16)
by reducing each fracture fragment sequentially. Once the iliac wing is
stabilized with lag screws, by 3.5-mm laterally applied reconstruction
plates, or both, the posterior column is reduced to the iliac wing as
the surgeon has direct visualization of the acetabular articular
surface.

The posterior-column lag screw and 3.5-mm reconstruction
plate fixation is utilized for transverse and T-type fractures. The
anterior column is then reduced to the intact posterior column. This
reduction can be accomplished with anterior to posterior 4.5-mm lag
screws inserted from the anterior–superior spine into the sciatic
buttress, or anterior-column lag screws from the lateral aspect of the
iliac wing (as described previously), or both. The adequacy of the
reduction is assessed, both by direct visualization of the acetabulum,
with finger palpation of the greater and lesser sciatic notches and
quadrilateral plate, and if necessary, in the internal iliac fossa. The
use of fluoroscopy is essential to assure the adequacy of reduction and
the position of the fixation (Fig. 43.17).

Because intra-articular hardware can lead to rapid
chondrolysis, the surgeon must confirm hardware position before
closure. This confirmation is best achieved radiographically through
use of intraoperative fluoroscopic Judet views (especially the
obturator oblique) and clinically by rotating the hip back and forth
while a finger, to detect any crepitus, is

placed along the quadrilateral surface.

At the completion of osteosynthesis, suction drains are
placed along the external surface of the iliac wing in the vicinity of
the posterior column and vastus lateralis muscle. If the internal iliac
fossa has been exposed, a third drain is placed here. All drains should
exit anteriorly.

The hip capsule is repaired first, followed by
reattachment of the tendinous insertions of the short external rotators
to the greater trochanter through drill holes, and femoral insertions
of the gluteus maximus. Next, the trochanteric insertions of the
gluteus medius and minimus muscles are repaired, through use of five or
six sutures for each tendon, as recommended by Letournel. Finally, the
tensor fascia-lata and gluteal muscles are reattached to their origins
on the iliac crest.

If a medial exposure has been performed, then the
origins of the sartorius and direct head of the rectus femoris muscles
are reattached through drill holes. If osteotomies have been performed,
lag screws should be used.

Finally, the fascia overlying the tensor fascia-lata
muscle is repaired. Then a subcutaneous suction drain is placed and the
skin closed.

Iatrogenic sciatic-nerve injury or worsening of a
preexisting deficit is a significant problem. In our experience,
patients at increased risk include those with preoperative sciatic
nerve compromise and those with fracture patterns that involve the
posterior wall or column. Other authors have identified patients
treated via an ilioinguinal approach to be at increased risk, possibly
related to indirect reduction of the posterior column with the hip
flexed. The peroneal nerve division is most commonly involved. The most
significant factor in reducing the incidence of iatrogenic
sciatic-nerve injury appears to be the experience of the surgical team.

Letournel initially reported an 18.4% incidence of
postoperative, iatrogenic, sciatic-nerve injury using the
Kocher-Langenbeck approach, which he subsequently reduced to 3.3%.
However, he also noted that none of his 114 patients treated with an
extensile approach developed this complication. Matta initially
reported a 9% incidence of iatrogenic nerve palsy, which he reduced to
3.5% after gaining further experience (3.4% of 59 extended iliofemoral
approaches). The incidence of iatrogenic nerve palsy remained elevated,
with an overall incidence rate of 12%, when open reduction and internal
fixation (ORIF) was delayed longer than 3 weeks. The majority of these
injuries involved the sciatic nerve. Alonso et al found postoperative
sciatic-nerve palsy in only 1 of their 21 patients treated with an
extended iliofemoral approach.

The use of intraoperative sciatic-nerve monitoring by
use of SSEP remains controversial. In the studies of Helfet et al,
intraoperative nerve monitoring reduced the incidence of iatrogenic
sciatic nerve injury to 2%. However, more recent studies have
questioned the value of intraoperative SSEPs because the monitoring has
failed to demonstrate a reduction in the rate of iatrogenic nerve
palsies. A high false-positive rate makes unclear the extent to which
intraoperative SSEP changes predict functional outcome. Intraoperative
monitoring of motor pathways with EMG allows for earlier detection of
neurologic compromise and removal of noxious stimuli, and in theory,
decreasing the risk of neurologic sequelae. In one of Helfet’s studies,
the addition of spontaneous EMG to intraoperative SSEP monitoring was
superior to SSEP alone. Because of the significant learning curve that
exists in the treatment of acetabular fractures, most authors agree
that intraoperative monitoring may prove most beneficial among
relatively less-experienced surgeons.

Superior–gluteal vessel injury is difficult to diagnose
and is caused by either the fracture or an iatrogenic insult during
surgery. Letournel reported an incidence of 3.5% in his series. This
potentially lethal occurrence is more likely to occur with severe
displacement of the sciatic notch (e.g., in high-transverse fractures
with marked medial rotation). Acutely, hemodynamic instability with an
arterial injury must be addressed during the initial evaluation and
resuscitation, and it is usually done with arteriography and
embolization. However, once the bleeding has been stopped, concerns may
arise with regard to muscle flap viability.

Because the extended iliofemoral approach completely detaches the gluteal muscles from the iliac wing (see Fig. 43.1),
the superior gluteal vessels are the only blood supply to the flap. If
they are compromised, then complete ischemic necrosis is, in theory,
likely. Mears and Rubash developed their triradiate approach partly in
response to reports of flap necrosis following the extended iliofemoral
approach; however, it has not been established whether the superior
gluteal artery is to blame in these cases. In fact, the incidence of
this complication is relatively low. In over 400 acetabular fractures
addressed with an extended iliofemoral approach by Letournel, Mast,
Martimbeau, and Matta, no one reports abductor flap necrosis. Alonso et
al did not observe this complication when they used either an extended
iliofemoral or a triradiate approach in 59 cases of complex acetabular
fracture.

Furthermore, massive abductor necrosis resulting from a
superior–gluteal artery injury combined with an extended iliofemoral
approach was postulated based on early animal and cadaver studies
alone. Canine studies by Tabor et al showed that although necrosis of
muscle

and
loss of mass occurs after the extended iliofemoral approach in the
presence of gluteal vessel injury, it does not appear to be
functionally significant. In their study, none of the gluteal muscle
flaps sustained complete ischemic necrosis. Thus, some collateral flow
to the abductor muscles must be present and appears to increase in the
presence of superior–gluteal vessel injury.

Bosse et al had recommended that, to assess the
integrity of the superior gluteal artery, a preoperative angiogram be
completed prior to performing an extended iliofemoral approach.
However, a more recent study, based on intraoperative Doppler
examination by Reilly et al, demonstrated only a 2.3% incidence of
absent flow on the superior gluteal artery. No evidence of abductor
muscle ischemia was found in any patient of Reilly et al. This result
does not support the use of routine preoperative angiography in the
management of these injuries.

Letournel reported a 2.3% incidence of in-hospital death
following operative fixation of acetabular fractures; the majority of
the deaths occurred in patients older than 60 years. Although DVT
probably plays a major role, its true incidence after an acetabular
fracture is unknown. However, patients with lower extremity trauma are
particularly a risk. By venography, Kudsk et al demonstrated a 60%
incidence of silent DVT in patients with multiple trauma immobilized 10
days or more. In a prospective study, Geerts et al also demonstrated a
60% incidence of DVT in patients with primary lower-extremity
orthopedic injuries. Letournel reported a 3% incidence of clinically
evident DVT with four fatal and eight minor pulmonary emboli in a
series of 569 patients, most of whom had received anticoagulant
prophylaxis.

Using a combination of perioperative mechanical
prophylaxis and postoperative anticoagulation prophylaxis, venous
thrombosis and PE rates of less than 3% and 1% respectively, have been
achieved. Improved detection of venous thromboembolism through use of
magnetic resonance venography has also led to a lower incidence of PE
because it has inspired aggressive treatment of asymptomatic DVTs in
the pelvis and proximal thigh. However, other data suggest that
magnetic resonance venography has a high false-positive rate for the
detection of thrombi in the pelvic veins, and its usefulness as a
screening tool is still debated. Borer et al recently reviewed 973
patients with pelvis or acetabular fractures and found that the overall
rate of PE was 1.7%, and the overall rate of fatal PE was 0.31%.
Routine preoperative screening for DVT had no effect on the incidence
of PE in this study.

The incidence of infection has been reported to be as
high as 19% but probably lies between 4% and 5%. Matta noted a 5%
incidence of postoperative wound infection in 262 patients. Of 59
patients who received extended iliofemoral approaches, 5 (8.5%)
developed deep infection. Mayo found a 4% overall infection rate, which
was 19% in 26 patients who underwent an extended iliofemoral approach.
Letournel reported 24 postoperative infections in 569 patients (4.2%)
with nine superficial, ten early deep, and five delayed or late
infections. Furthermore, he observed skin necrosis in 1.8% (10.2% of
extended iliofemoral approaches) and hematomas in 6.7% of cases. To
minimize wound problems, he advocated the use of prophylactic
antibiotics, multiple suction drains in all recesses to prevent
hematoma formation, surgical evacuation of hematomas, and if present,
debridement of the Morel-Lavalle lesion over the greater trochanter.
Other factors such as morbid obesity and burns must also be taken into
consideration as they may render the patient more susceptible to
infection.

The most common complication following the operative
fixation of acetabular fractures through the extended iliofemoral
approach is HO (Fig. 43.18), with an incidence ranging from 18% to 90%. However, functional limitation in patients with HO occurs in only 5% to 10%

of cases. Nevertheless, heterotopic bone formation is more common and
severe with the extended iliofemoral approach because of the external
surface of the iliac wing is stripped. Letournel reported its
occurrence in 46% of his extended iliofemoral approaches performed
within 4 months of injury; a 21% incidence was found in 635 other
approaches. Prior to his use of prophylaxis, these rates were 69% and
24% respectively. Matta noted a significant loss of motion in 20% and
Letournel observed severe HO (Brooker III and IV) in 35% of patients
treated with this approach within 3 weeks of injury. Both indomethacin
and low-dose radiation therapy (single or multiple fractions) have been
shown to decrease the incidence and severity of HO in patients with
acetabular fractures. However, concerns remain about the cost and the
long-term effects of radiotherapy, particularly in the younger trauma
population.

Despite prophylaxis with indomethacin, Alonso et al and
by Johnson et al have reported rates of HO ranging from 86% to 88% in
patients treated with an extended iliofemoral approach. Of these
patients, Brooker class III or IV ossification was present in 14% and
13%, respectively. In the Johnson et al study, the majority of patients
in the treated group had Brooker class 0-II ossification, with the
untreated group having mostly Brooker III and IV ossification. A more
recent study by Moed et al showed a 50% incidence of HO after extensile
exposures in patients treated with indomethacin. Only one patient in
the treated group had severe (Brooker III-IV) ossification.
Indomethacin clearly does not eliminate the occurrence of HO, but it
significantly decreases its severity.

The incidence of avascular necrosis (AVN) after
operative treatment of acetabular fractures has generally ranged from
3% to 9%, with the majority of cases identified between 3 and 18 months
after surgery. However, an increased incidence of AVN of the femoral
head is found in cases presenting after 3 weeks and those associated
with a posterior fracture/dislocation. In all probability, the fate of
the femoral head is determined at the time of the injury.