The Adult Hip

Accurate location of landmarks around the hip joint aids
greatly in the proper placement of skin incisions during hip surgery.
Identifiable landmarks include the anterosuperior spine, crest of
ilium, greater trochanter, and shaft of femur. The tip of the greater
trochanter is a reliable marker to the center of the femoral head and
acetabulum (Figure 17-1). Soft tissue
identifiable beneath the skin includes the tensor fascia lata muscle,
which originates from the external lip of the iliac crest below the
anterior iliac spine and inserts into the iliotibial tract, and the
gluteus maximus, which originates from the posterior half of the ilium
behind the posterior gluteal line and inserts three-quarters of its
fibers into the iliotibial tract and one-quarter into the posterior
shaft of the femur below the greater trochanter. These two muscles are
often termed the deltoid of the hip joint, and most surgical approaches
to the hip joint split them in some way.

Gluteus medius and gluteus minimus originate in
different planes from the side of the pelvis, with medius lying
superficial to minimus. What is not well illustrated in anatomy
drawings is that their tendons converge into a single layer that is
inserted into the greater trochanter. The original Hardinge approach
for total hip arthroplasty split these two muscles in one layer, but
separate identification of each and suturing of the split in gluteus
minimus

as
a separate repair will reduce potential dead space, which may decrease
the subsequent risk of postoperative dislocation. Care should be taken
during the lateral approach to protect the superior gluteal nerve that
lies between gluteus medius and minimus. Dissection should not be taken
beyond 5 cm from the tip of the greater trochanter to avoid damage to
this structure. A small bursa often lies under the most anterior fibers
of gluteus medius tendon. This bursa can often be confused with hip
capsule during the direct lateral surgical approach. During this
exposure, a reliable indicator of the outer surface of hip capsule is
the insertion of the anterior fibers of vastus intermedius. The
undersurface of the gluteus minimus may be adherent to the hip capsule
in osteoarthritis and must be dissected free from it prior to a
capsulectomy being performed.

The greater trochanter may have an altered shape due to
underlying disease or previous surgery, and care should be taken to
avoid its fracture. The femoral neck may not be easily identified from
the femoral head if there is extensive osteophyte formation. During
THA, orientation of the femoral neck osteotomy in the correct
anteversion should always be assessed after hip dislocation by
reference to the shaft of the femur with the knee flexed at 90°,
because there is a wide variation in anteversion of the femoral neck.
If the femoral neck osteotomy is made with reference to femoral neck
anteversion, an inaccurate cut may result, and subsequent
malpositioning of the prosthesis may occur.

The acetabulum is made up of three walls and a floor to
form a spherical receptacle for the femoral head. The ischium, which
makes up the posterior column, is the most substantial wall. The ilium
contributes to the superior wall or dome, and a thin contribution from
the pubis completes the anterior column. Medially, the floor of the
acetabulum receives contributions from all three bones. Screws placed
through cementless acetabular cups to aid fixation should be placed
between the 10 and 2 o’clock positions when the acetabulum is viewed
from the side, to avoid damage to the internal iliac vessels
anteriorly, and the sciatic nerve posteriorly. Revision surgery of the
acetabulum often involves bony defects in one or more of the columns,
putting neurovascular structures even more at risk.

A medial defect will often be covered by a
pseudomembrane, which overlies the iliacus muscle. Directly medial to
this lies the obturator nerve and artery, the ureter, and the bladder.
Great care should be taken when preparing the acetabulum to avoid
damage to any of these structures, which may have catastrophic
consequences (Figure 17-2).

This approach is commonly used to access the hip in
cases of suspected septic arthritis. The patient is placed supine on
the operating table. The skin incision runs along the anterior aspect
of the iliac crest, and as the anterior-superior iliac spine is

reached,
extends distally for approximately 10 to 15 cm. The incision extends
through the skin and subcutaneous tissue to the tensor fascia lata
muscle. Care is taken to identify and protect the lateral cutaneous
nerve of the thigh as it penetrates the fascia just below and medial to
the anterior superior iliac spine. The medial edge of tensor fascia
lata is easily seen laterally and the fascia is incised just medial to
this edge. The interval between tensor fascia lata and sartorious is
developed directly down to the tendinous portion of the rectus femoris
muscle.

The reflected head of rectus femoris may need to be
elevated off the edge of the pelvis to allow visualization of the hip
joint capsule. The hip capsule can then be incised over the front of
the femoral neck. This approach allows drainage of the hip in cases of
suspected infection. If a wider exposure is required, then the tensor
fascia lata muscle, and a varying amount of gluteus medius muscle, can
be reflected off the outer aspect of the ilium, and the interval
between tensor fascia lata and sartorious developed further distally.
This distal extension often results in disruption of the lateral
cutaneous nerve of the thigh and extensive dissection proximally may be
followed by the development of heterotopic ossification in the gluteus
medius.

This approach is used for hemiarthroplasty and total hip
arthroplasty. The approach’s major disadvantage is the potential for
damage to the anterior edge of gluteus medius while exposing the
femoral shaft. It should be used with caution in obese patients, as
access to the femur is very restricted. Also, as the approach places
considerable torsional stress on the femur, it should be used with
caution in patients with osteoporosis or other diseases likely to
weaken the femur. This approach does have the advantage of minimal
muscle disruption if performed well and maximum stability of the joint
postoperatively.

The patient may be positioned supine on the operating
table, with the buttock elevated on a bolster, or they may be placed in
the lateral position. The skin incision begins at the iliac crest 5 cm
posterior to the anterior superior iliac spine, runs obliquely down to
the tip of the greater trochanter, and extends posteriorly from the tip
for 3 cm. It then curves anteriorly and distally for a further 10 to 15
cm. The incision is deepened to the fascia lata, and the anterior edge
of gluteus medius is identified beneath the fascia. The fascia is
incised longitudinally just distal to the anterior edge of gluteus
medius, and over the greater trochanter follows the skin incision. This
incision allows the interval between tensor fascia lata muscle and
gluteus medius to be developed. Be aware that the terminal branch of
the superior gluteal nerve crosses this interval in its proximal third
and is susceptible to injury in this area. Clear the fibrofatty tissue
off the front of the hip joint capsule and identify the reflected head
of rectus femoris crossing the capsule proximally. Place a Holman
retractor around the medial side of the capsule, one around the lateral
side, and one under the reflected head of rectus femoris.

The whole of the anterior capsule should now be clearly
visible. The capsule can be opened and the medial and lateral Holman
retractors placed around the neck of the femur within the joint. The
femoral neck can be transected and the head removed. Access to the
acetabulum is excellent, with the femur retracted posteriorly by a
Holman retractor placed immediately behind the acetabulum. Access to
the femoral canal is obtained by adduction and external rotation of the
leg, and a Holman retractor placed deep to the tip of the greater
trochanter. If the lower end of gluteus medius is attached more
distally on the anterior aspect of the greater trochanter, it may need
to be detached to improve access to the proximal femur.

The direct lateral approach allows excellent access to
the acetabulum and proximal femur. It has the potential disadvantage
that it may weaken the abductors with a subsequent limp, and it may not
be extensile enough to allow surgeons complete visualization of the
anterior and posterior acetabular columns during revision hip surgery.
Its major advantage is that dislocation following total hip replacement
is uncommon.

The approach can be performed with the patient supine or
in the lateral decubitus position. If the lateral decubitus position is
used, care should be taken that the pelvis is correctly aligned to
ensure that component positioning is not compromised. The skin incision
begins in the midlateral line approximately 6 cm proximal to the tip of
the greater trochanter and runs distally for 10 cm. The incision is
deepened to the fascia lata, which is incised along the line of the
skin incision. A large self-retaining retractor is placed under the
fascia lata anteriorly and posteriorly, exposing the gluteus medius
proximally and vastus lateralis distally.

The anterior edge of gluteus medius forms a consistent
landmark where it inserts into the most distal aspect of the greater
trochanter. The tip of the greater trochanter is palpated, and
beginning proximally the gluteus medius is split along the line of its
fibers for approximately 3 to 4 cm, toward the anterior superior iliac
spine. This incision is then curved distally over the anterior third of
the greater trochanter, and beyond the trochanter into the vastus
lateralis muscle (Figure 17-4). Avoid splitting
the gluteus medius more than 5 cm proximal to the tip of the greater
trochanter to avoid damage to the superior gluteal nerve. A small
amount of sharp dissection (cutting diathermy) proximal to the greater
trochanter allows the gluteus muscle to be split along the line of its
fibers. Blunt dissection can be used to sweep the fat deep to gluteus
medius superiorly, thus protecting the superior gluteal neurovascular
bundle. The exposed gluteus minimus muscle can then be split along the
same line, giving access to hip capsule. Dissection superiorly up this
line allows placement of a Holman retractor deep to the reflected head
of rectus femoris, thus exposing the capsule up to the anterior
acetabular edge. Hip capsule is most easily identified at the inferior
extent of the greater trochanter, where sharp dissection combined with
external rotation of the leg will eventually expose muscle fibers that
are the capsular insertion of vastus intermedius. The bursa under
gluteus medius is often mistaken for hip capsule and is a much less
reliable indicator.

A complete capsular exposure, extending medially onto
the acetabular margin, proximally under the free edge of the previously
split gluteus minimus tendon (which is often adherent to, and requires
dissection off, the superior hip capsule), and distally under the
vastus intermedius muscle (often a vascular area), allows good
visualization of the femoral neck and head. The remainder of the
exposure is similar to the anterolateral approach. The lateral approach
may be

combined
with trochanteric osteotomy, which allows wide exposure of the
acetabulum and improves access to the proximal femur. Many techniques
of osteotomy have been described, and when using this technique
surgeons should be aware that nonunion of the greater trochanter can be
a particular problem in total hip replacement.

The posterior approach gives excellent exposure of the
proximal femur and acetabulum, is extensile for revision surgery and
involves minimal muscle disruption. However, when used for
hemiarthroplasty or total hip replacement, dislocation is more common.
Careful attention to closure of the deep layers of the incision is
essential to minimize this risk.

The patient is placed in the lateral decubitus position,
with the pelvis appropriately positioned for correct orientation of the
components. The incision begins 3 cm distal to the tip of the greater
trochanter and is centered over it. The incision passes posteriorly and
proximally at 45° to the long axis of the femur for 8 to 10 cm. The
incision is deepened to the fascia lata, and beginning distally the
fascia is incised over the trochanter and then proximally over gluteus
maximus along the line of the skin incision. The gluteus maximus muscle
is split along the line of its fibers. A large self-retaining retractor
retracts the gluteus maximus. The leg should then be internally rotated
20° so as to better expose the fatty layer over the external rotators.
This fat is swept posteriorly off the short rotators by blunt
dissection, allowing visualization of the muscles and the sciatic
nerve. The tendon of piriformis is identified and an incision is made
along the proximal edge of piriformis distally, to the posterior margin
of the greater trochanter. It then turns distally along the posterior
margin of the greater trochanter as far as the proximal margin of
quadratus femoris. This incision extends through the capsule and thus
forms an L-shaped flap that is retracted to protect the sciatic nerve (Figure 17-5).
The hip can then be dislocated. Proximal and distal extension of the
exposure is possible to improve access to the acetabulum and femur as
necessary. The relatively high rate of postoperative dislocation (when
compared to other recognized surgical approaches for THA) has been
dramatically improved by repair of the external rotators (Figure 17-6).

Hip dysplasia, labral pathology, pigmented villonodular
synovitis, and loose bodies may present in the young adult. Dysplastic
hips are associated with intermittent postexercise pain, which becomes
more frequent with time. The pain is usually felt in the groin or
proximal thigh, but as with all patients with hip pain it may be felt
in the knee and occasionally even further down the leg. Stiffness is
rarely an early complaint, but as the condition progresses, reduction
in abduction and rotation is common.

The patient may have noted a limp. Labral pathology
causes pain and this may be positional, such that there is pain
associated with extreme flexion and adduction but not with other
positions. The patient may have a feeling of a “catch” in the hip with
certain movements. There is often a history of minor trauma preceding
the onset of pain. Loose bodies cause pain and occasional “catching” in
the hip, although frank locking of the hip is unusual.

At this age, dysplasia of the hip may have developed
into early osteoarthritis. Pain, which may have been a relatively mild
problem in early adulthood, is now a much greater problem. Stiffness
and the development of a leg length discrepancy are not uncommon,
particularly if the acetabular dysplasia is

severe.
Other causes of secondary osteoarthritis, such as gout and pseudogout
may also present at this age. Ankylosing spondylitis and avascular
necrosis of the femoral head commonly present with hip pain in middle
age. In the case of the former there may be a history of back stiffness
but this is not always the case. Patients with avascular necrosis may
have some identifiable risk factors such as alcohol abuse or
corticosteroid ingestion but in many cases no risk factors are evident.
Transient osteoporosis of the femoral head is a recently recognized
condition whose onset is sudden, with rapidly developing pain and
apparently normal radiographs. The diagnosis can only be confirmed on
MRI.

Primary and secondary osteoarthritis are common in the
aging population. Hip, thigh, and knee pain are common and stiffness
influences simple activities such as cutting toenails and putting on
underwear. The history and physical findings are usually diagnostic,
but other diseases such as spinal stenosis and metastatic neoplasm
should always be considered.

Aspects of the history are important and include any
prior treatment of the hip, especially surgery, for this may alter the
already abnormal anatomy. Special aspects of examination include the
Trendelenburg sign, the presence or absence of a fixed flexion
deformity, and the range of internal and external rotation measured
with the patient prone. Signs of labral pathology should be sought with
the impingement test.

Radiographs should include a standard anterior-posterior
(AP) pelvic radiograph, a lateral of the hip, a false profile view to
assess anterior acetabular coverage, and AP views with the hip in
abduction and adduction, and internal and external rotation. From the
radiographs, a number of measurements to determine the type and
severity of the dysplasia can be made. The center-edge angle, the
acetabular index, the teardrop to head distance, and the anterior head
coverage as determined from the false profile view allow accurate
assessment of the pelvic contribution to hip dysplasia. The
abduction-adduction and internal-external rotation views, together with
measurement of the neck shaft angle allow the surgeon to assess the
contribution by the femur to hip dysplasia. The position of the greater
trochanter relative to the center of the femoral head should be
assessed. Occasionally CT scanning of the hip is advisable if there has
been prior surgery as the anatomy of the acetabulum, in particular, may
be hard to define with standard radiographs.

If the patient has pain that cannot be reduced by a 3-
to 6-month program of nonoperative treatment including analgesia,
anti-inflammatory medication, and physiotherapy, then surgery should be
considered. Surgery can take several forms from acetabular redirection
to acetabular supplementation and femoral osteotomy.

There is considerable variation in the approach to
surgery for hip dysplasia. The shape of the femoral head and acetabulum
need to be considered. If the femoral head is nonspherical, a
repositioning osteotomy of the femur or pelvis is generally
contraindicated. If the patient has a fixed deformity of the hip,
however, a repositioning osteotomy of femur may improve hip function
and reduce pain. If the hip joint is congruous, then the site of the
“maximum” dysplasia needs to be identified. In many cases dysplasia is
most marked on one side of the hip and correction of that abnormality
is sufficient to alleviate symptoms.

If there is radiographic evidence on the false profile
view of reduced anterior femoral head coverage, the acetabular index is
at least 15°, and there are no or minimal degenerative changes, then an
acetabular redirecting osteotomy is indicated. Many redirection
osteotomies have been described but ones in contemporary use include
the Ganz, Tonnis, and Ninomiya. The type chosen depends on the
experience and philosophy of the surgeon. The Ganz osteotomy is
performed through a single Smith-Peterson type incision and is a
periacetabular

osteotomy,
which preserves the posterior column of the pelvis allows correction of
deficient anterior coverage, and may allow alteration in the position
of the center of rotation of the hip (Figure 17-7).
The Tonnis osteotomy uses two incisions and divides the pelvis into
two, achieving a similar degree of correction to that of the Ganz. Both
osteotomies have a number of potentially serious complications,
including nerve palsies, nonunion, heterotopic ossification, and
fractures extending into the joint. If these are avoided, however, the
short- to medium-term results are excellent. If there are advanced
degenerative changes then the results tend to be less satisfactory. The
long-term results of periacetabular osteotomies of the Ganz and Tonnis
types have yet to be determined. The long-term (10-23 years) results of
a different type of periacetabular osteotomy, described by Ninomiya,
have been reported as very good, providing that there was little
evidence of osteoarthritis at the time of surgery. If the radiographs
show evidence of advanced degenerative change, then a Chiari
supra-acetabular osteotomy may be considered as a pain relieving
procedure.

Proximal femoral varus-varus plus rotation osteotomy is
a more straightforward procedure, which is indicated if the radiographs
show that proximal femoral dysplasia is the predominant dysplasia.
Another requirement for proximal femoral osteotomy to be considered is
whether the femoral head can be better covered, if the head can be
better centered in the acetabulum, and if the hip is either abducted,
or rotated or both. Careful preoperative planning is necessary to
provide the optimal femoral head coverage. The degree of varus
angulation seldom exceeds 15°. If greater than 15° is planned, then
advancement of the greater trochanter is necessary to avoid a
Trendelenburg gait. Varus femoral osteotomy has the advantage that the
recovery is reasonably short and the risk of serious complications is
low. All varus femoral osteotomies, however, have the potential to
shorten the leg, and this should be explained to the patient
preoperatively. The results in appropriately selected patients are
excellent, although pain over the internal fixation device often
necessitates its removal (Figure 17-8).

The results of osteotomy for the treatment of early
osteoarthritis in other joints strongly suggests that even in ideal
circumstances, osteotomies for the treatment of hip dysplasia can only
be expected to give satisfactory results for 10 to 12 years before pain
becomes a significant problem again.

When preoperatively planning a THR in a patient with a
dysplastic hip, the two components of the hip, the acetabulum and the
femur, should be considered individually and then together. Dealing
first with the acetabulum, it may be present in its normal position, in
a more proximal position, or absent (with or without a false
acetabulum). When it is present in its normal position, it may be
relatively small—in which case the surgeon requires small diameter
acetabular components. Depending on the degree of femoral dysplasia,
the anterior, or less commonly posterior, wall of the acetabulum may be
dysplastic or absent, compromising acetabular fixation. If the surgeon
has any doubts about the geometry of the acetabulum, then preoperative
Judet views or a CT scan should be obtained. If the acetabulum is
present but has migrated proximally, the shape of the ilium causes it
to become progressively more shallow. Also, as previously indicated,
the anterior or posterior walls may be deficient, and the roof may
slope. In these circumstances, a CT scan of the pelvis with appropriate
reconstructions allows the surgeon to plan the surgery.

In cases where a high false acetabulum is present, there
is little or no anterior or posterior columns to aid with cup fixation.
Restoration of the center of the rotation to the anatomical position in
this situation not only improves the biomechanics of the THA but also
improves the amount of host bone available into which an acetabular cup
can be placed (Figure 17-9). Location of the
true acetabular position may be difficult in a high-riding DDH, but can
be achieved in one of three ways: (1) an intact ligamentum teres that
runs from the femoral head down into the

acetabular
fossa; (2) identification of the greater sciatic notch posteriorly and
the anterior column, which runs down and converges at the teardrop; (3)
if neither of these anatomic landmarks is easily identified, an
intraoperative AP radiograph may be performed. Most surgeons prefer
cementless fixation of the acetabular implant with supplemental screw
fixation.

Factors such as leg length, bone deficiency, previous
surgery, and the femoral anatomy need to be taken into consideration.
When the roof is sloping, the femoral head may be used as a graft to
support the acetabular component (Figure 17-10).
When there is a high dislocation, there is at best a false acetabulum,
which generally plays no part in the reconstruction. Many of these hips
do not cause pain until later life, and THR is not indicated. If
surgery is considered desirable, then a CT scan of the true acetabulum
usually reveals a small acetabulum with relatively osteoporotic bone.
This situation requires small diameter acetabular components, and
because of the osteoporosis, great care should be taken to prevent
reaming through the floor.

With regard to the femur, the size and shape of the
femoral canal should be considered, as it is often small or distorted.
It will be necessary to have available a number of small implants,
including micro implants for some patients. Choice of femoral fixation
is still controversial, but surgeons should be aware that cemented
femoral components in the small femora may give an inadequate cement
mantle. In cases of femoral dysplasia there is frequently a marked
increase in the femoral anteversion. This possibility should be
assessed preoperatively with CT scanning. If the anteversion is greater
than 15°, then there are several different strategies available. A
small femoral component may be used in association with a low neck cut
to correct the anteversion. If using uncemented components of a tapered
design, then cutting out a piece of the back of the femoral neck allows
the component to be inserted in the correct degree of anteversion. The
anterior neck may have to be trimmed to prevent impingement. If the
anteversion is severe, then a modular prosthesis that allows the
femoral component to be rotated within a proximal metaphyseal sleeve
may be used. If a femoral shortening needs to be performed, then the
proximal fragment can be rotated back to its correct position.

Finally, the issue of leg length needs to be addressed.
If the leg is short, do not consider lengthening of over 5 cm. However,
to work within this limit and to achieve the appropriate acetabular and
femoral component positioning, it may be necessary to shorten the femur
through a subtrochanteric osteotomy. This requires special prostheses,
and should not be undertaken without appropriate training. If there is
bilateral high riding DDH, but only one side is symptomatic and
requires THA, consideration has to be given to resulting leg length
inequality and the functional disturbance to gait this might cause. In
this situation, the contralateral hip occasionally requires THA to
maintain function (Figure 17-11).

Often dysplastic hips have had previous surgery and this
always makes hip replacement more difficult. The old incisions may make
access difficult. There may be extensive scarring on the lateral side
of the pelvis and femur, making dissection, access, and restoring leg
length more difficult. The bony anatomy may be quite abnormal, and
preoperative CT scanning is very useful to delineate the features of
the pelvis and proximal femur. The surgical approach adopted for THR in
dysplasia depends on the experience of the surgeon, but as a general
rule the posterior approach is more extensile and allows better
visualization of the acetabulum than the direct lateral. Also dealing
with severe femoral abnormalities may be facilitated through the
posterior approach.

The clinical presentation of AVN depends on the age of
the patient. In the elderly who have sustained a subcapital fracture,
AVN is often heralded by increasing pain and a radiograph that shows
fixation failure, collapse of the femoral head, or both. In younger
patients with idiopathic AVN, pain is the most common presenting
feature. An associated feature is synovitis of the hip, which manifests
clinically with a decrease in the range of movement, and the hip is
irritable (there is pain throughout the range of motion). If the
patient presents late in the evolution of the disease, then the
symptoms and signs will be indistinguishable from advanced
osteoarthritis.

Radiographic investigation begins with plain
anteroposterior and lateral radiographs of the hip. Early in the
disease, radiographs may be normal, but as the disease progresses it
goes through several stages. A number of authors have classified the
stages, the most commonly used classification being that described by
Ficat. Ficat’s classes divide the disease into four stages. Stage 1
refers to a hip that on plain radiographs appears normal. The diagnosis
is made by measuring a raised intraosseous pressure, histologic
examination of biopsy specimens, or an abnormal magnetic resonance
imaging (MRI) examination. Stage 2 avascular necrosis may show patchy
osteoporosis but an otherwise normal radiographic examination.
Clinically, however, the hip is painful, and there may be a reduced
range of motion. Stage 3 shows some change in the contour of the
femoral head, and there may be a subchondral fracture. Stage 4 shows a
loss of joint space and collapse of the femoral head. Radiographs may
be indistinguishable from advanced osteoarthritis of the hip joint.

MRI is the best radiological modality to diagnose AVN
early in the evolution of the disease. Care must be taken, however, not
to misdiagnose transient osteoporosis of the hip as AVN. The former
condition involves the whole of the femoral head and lacks the focal
changes seen in AVN. Transient osteoporosis is a self-limiting
condition with symptoms and MRI findings clearing in 6 to 8 months.

Management of AVN depends on the age of the patient,
size of the lesion, and stage of the disease. In the elderly who
develop AVN following a hip fracture, total hip arthroplasty is the
treatment of choice. In younger patients with idiopathic AVN,
determining the size of the lesion is important as this has a direct
influence on the prognosis. The size is best determined by MRI. Small
lesions (less than 15% of the femoral head) may resolve without
treatment. Large lesions (greater than 50% of the femoral head) tend to
progress to collapse of the head and secondary osteoarthritis, despite
treatment. For intermediate-sized lesions, a number of treatment
options have been proposed. A classification based on the location of
the lesion also has prognostic significance. If the lesion is medial in
the head, then progression is rare; if it is lateral in the head the
prognosis is worst.

There is a paucity of well-conducted trials that show
conclusively that any one treatment is superior. Treatments range from
simple decompression of the femoral head to lower the intraosseous
pressure, to bone grafting of the involved area either directly after
dislocating the hip and elevating the involved segment, or by using a
vascularized fibular graft placed up the center of the femoral neck.
There have been a number of osteotomies described whose purpose is to
remove the affected area of the femoral head from the weight-bearing
axis of the hip. These include varus and valgus osteotomies and more
complex rotational osteotomies. The choice of osteotomy depends on the
location of the lesion and the ability of the proposed procedure to
remove the affected segment from the weight-bearing axis of the hip.
Osteotomies are generally contraindicated for those patients who remain
on corticosteroids or have untreated metabolic bone disease. They are
generally only effective when a small area of the femoral head is
involved.

When considering surgery designed to save the femoral
head and prevent progression of the disease to degenerative arthritis,
surgeons should be aware that, ultimately, conversion to THA is common
following these procedures and the risks and potential benefits of such
surgery should be very carefully weighed in each case. Hemiarthroplasty
or bipolar prostheses should be avoided in AVN of the femoral head, as
histologic assessment of the acetabular cartilage shows abnormalities
in the majority of cases, which may adversely influence the outcome of
such surgery. Idiopathic AVN is frequently bilateral with a reported
incidence of 50 to 80%, and thus MRI should be performed in all cases.

Extensive research continues to be done, but the exact
cause of rheumatoid arthritis remains obscure. Regardless of its cause,
it may be described as an inflammatory process that somehow is
triggered and centers in the joints with articular cartilage. The
inflammation manifests as a stimulus for synovium to hypertrophy and
becomes increasingly hyperplastic

and
hypervascular with increasing cellularity. This hypertrophic synovial
tissue invades and degrades articular cartilage. The actual destruction
of articular cartilage is done in large part by rheumatoid pannus, a
fibrovascular granulation tissue that protrudes from the inflamed
synovium into articular cartilage. It contains fibroblasts, small
vessels, and multiple inflammatory cells that are responsible for the
destruction of articular cartilage and its underlying bone.

Pathology within the hip joint generally involves
varying degrees of articular cartilage loss secondary to the
inflammatory process. This loss usually involves the entire femoral
head, resulting in concentric loss of cartilage and subsequent
concentric joint space narrowing. There may also be varying amounts of
bone loss and even femoral head collapse. Varying degrees of cyst
formation occur, and in about 5% of patients, significant protrusio
acetabuli develops.

A full history and physical examination should be
performed in someone presenting for consideration of surgical treatment
of a rheumatoid hip. Special attention should be taken of what
medications the patient is taking. Prior to any joint replacement, a
review with a rheumatologist to minimize or stop steroids or
antimetabolic drugs should occur, because these medications compromise
fixation of the implant and predispose the patient to an increased risk
of infection. Careful attention to the state of the cervical spine and
jaw for ease of anesthesia, and the state of the upper limbs and
contralateral leg for ease of postoperative mobilization, should be
made.

Radiography of the hip joint shows varying degrees of
osteopenia and loss of joint space. The loss of joint space, in
contrast to that seen in osteoarthritis, is generally concentric, and
varying degrees of protrusio acetabuli may be present (Figure 17-12).
Osteophyte formation is rare, and subchondral sclerosis is not a
prominent feature. Cysts may develop within both the femoral head and
the acetabulum. These changes may culminate in varying degrees of
femoral head collapse, and in severely advanced cases, spontaneous bony
fusion may occur with no range of motion.

Conservative measures are usually the domain of the
rheumatologist and include analgesics, nonsteroidal
anti-inflammatories, steroids, and antimetabolite drugs such as
methotrexate.

Total hip arthroplasty remains the mainstay in the treatment of end-stage rheumatoid arthritis of the

hip. In mature adults with rheumatoid arthritis, care should be taken
during dislocation to prevent fracture in osteoporotic bone, and
reaming of soft acetabular bone. Large cysts should be bone grafted
with reamings from the femoral head or acetabulum.

THA in juvenile rheumatoid arthritis may present an
array of technical complexity because of the small size of the patient
and deficiencies in the femoral and acetabular bone stock. These
problems may require customized or miniature components. Additionally,
although the early result of cemented total hip arthroplasty may be
good, these patients are young, and the total hip replacement must
withstand many years of function.

The indication for THA in young patients who have
hemophiliac arthropathy of the hip should be severe, disabling pain
with activity and at rest that is unresponsive to nonoperative
treatment. Careful preoperative planning to overcome deformities of the
proximal femur, resulting from growth arrests from multiple childhood
bleeds, is often needed. Osteoporosis is common and care should be
taken to avoid fracture of the femur. Full factor VIII replacement is
required not only during the surgery but until the stitches have been
removed. A serious problem affecting the complication rate is a high
prevalence of seropositivity for HIV and the eventual development of
AIDS.

In one multicenter study of 34 THAs performed in 27
patients, at a mean follow-up of 8 years, 3 patients (11%) had
developed a deep infection necessitating prosthesis removal, 7 hips
(21%) had required revision for aseptic loosening, and 9 patients (33%)
had died. Almost all the deaths were related to AIDS.

This relatively high complication rate illustrates the
problems surgeons face when performing THA in young immunocompromised
patients.

The diagnosis of hip pain in young patients has evolved
significantly in the past few years. Referred pain and common hip
pathologies, such as stress fractures of the femoral neck of pelvis
avascular necrosis and early osteoarthritis, can usually be diagnosed
by a combination of careful history and physical examination, followed
up by AP and lateral radiographs and possibly a Technetium-99 isotope
bone scan. If a diagnosis has still not been obtained, an MRI can be
requested.

The MRI diagnosis may include symptomatic acetabular
labral tears, hip capsule laxity and instability, chondral lesions,
osteochondritis dissecans, ligamentum teres injuries, snapping hip
syndrome, iliopsoas bursitis, loose bodies (for example, synovial
chondromatosis), bony impingement, synovial abnormalities, crystalline
hip arthropathy (gout and pseudogout), infection, and posttraumatic
intra-articular debris. Occasionally, MRI arthrography can be a
particularly useful technique for dedicated assessment of hip joint
internal derangements.

Once a diagnosis has been obtained, the correct treatment can be instituted.

Current indications for hip arthroscopy include the
presence of symptomatic acetabular labral tears, hip capsule laxity and
instability, chondral lesions, osteochondritis dissecans, ligamentum
teres injuries, snapping hip syndrome, iliopsoas bursitis, and loose
bodies (for example, synovial chondromatosis). Less common indications
include management of osteonecrosis of the femoral head, bony
impingement, synovial abnormalities, crystalline hip arthropathy

(gout
and pseudogout), infection, and posttraumatic intra-articular debris.
In rare cases, hip arthroscopy can be used to temporize the symptoms of
mild-to-moderate hip osteoarthritis with associated mechanical symptoms.

Technique for hip arthroscopy usually involves
distraction of the joint, usually with a dedicated fracture table, and
image intensification to ensure the portal placement is accurate. The
patient may be positioned supine or in the lateral decubitus position.
Anterior and peritrochanteric portals are most commonly used, giving
good access and working space, to the lateral and superior part of the
hip joint (Figure 17-13).

Recently, Ganz from Bern, Switzerland, has developed a
new technique for surgical dislocation of the hip, based on detailed
anatomical studies of the blood supply. It combines aspects of
approaches that have been reported previously and consists of an
anterior dislocation through a posterior approach with a “trochanteric
flip” osteotomy. The external rotator muscles are not divided and the
medial femoral circumflex artery is protected by the intact obturator
externus. He reported his experience using this approach in 213 hips
over a period of 7 years. Perfusion of the femoral head was verified
intraoperatively and none subsequently developed avascular necrosis. It
allows the treatment of a variety of conditions, which may not respond
well to other methods, including arthroscopy. The most common
conditions treated with this technique include impingement of the
femoral neck on the bony acetabulum, and debridement of osteophytes and
osteochondral lesions in patients with early osteoarthritis of the hip
who are not suitable candidates for THA.

In order to understand how some modern concepts
surrounding total hip arthroplasty have evolved to their current form,
it is important to review the history of hip arthroplasty. This history
can be broken down into three distinct eras based around the
innovations of the man considered to be the father of total hip
arthroplasty, Sir John Charnley.

Charnley’s contributions to THA between 1954 and 1974
represent this century’s most significant developments in orthopaedic
surgery.

The aging population with a higher incidence of hip
osteoarthritis, combined with a decreasing patient acceptance of
disability, is leading to increasing strain on a the health care
budget, accounting for up to 1 to 2.5% of a county’s gross national
product. In the last 15 years, surgeons have been asked to evaluate the
success of a THA not only for prosthesis longevity, but also cost
effectiveness. Health economists can now create models for comparing
the cost effectiveness of a THA with coronary bypass and renal
transplantation. These models measure cost-benefit analysis using a
Health Related Quality of Life Index, and for patients in their fifth
and sixth decade of life, THA comes out significantly ahead of these
other more complex procedures. Governments use these economic
indicators and other outcome measures to allocate health budgets more
appropriately.

One of the most important variables in cost-benefit
analysis (CBA) of THA is the age of the patient. Despite an increase of
cost over benefit with increasing age, THA has been shown to be
beneficial in the octogenarian population. The benefit of THA can now
be accurately measured, and none would argue that it is a huge success,
leading politicians and health providers to look more closely at cost
containment. With the advent of new design, materials and surface
finishes over the last 20 years, nowhere is this more applicable than
implant costs. The original Charnley cemented stem and cemented cup
costs approximately 30% of a cementless porous-coated femoral and
acetabular component. In young active total hip recipients, a new
prosthetic design, which offered a 90% improvement in survivorship over
15 years and a 15% reduction in the cost of revision surgery, could be
sold at a price of 2 to 2½ times that of conventional cemented
components such as the Charnley prosthesis and still remain cost
effective. Using more likely estimates of the improved performance of
new technology, however, the upper limit of cost effectiveness is an
increase of 1½ to 1. Only a very small increase in the cost of a
prosthesis could ever be justified for older patients of either sex.

In many institutions cost concerns have led to the
concept of implant matching, where cementless porous-coated prostheses
are reserved for younger more active patients, and cemented THA
reserved for elderly more sedentary patients. Another area that is
being evaluated for cost savings is length of hospitalization. The time
a patient spends in hospital contributes the largest proportion to the
overall cost of THA. Day of surgery admission, and improved surgical
technique, rehabilitation, and home services all allow patient
discharge on day 3 or 4.

Perhaps the most extreme expression of improved length
of stay is the newly emerging miniincision surgery (MIS) techniques.
Although promising from a surgical perspective, length of
hospitalization is probably more influenced by age, patient
co-morbidities, patient expectations, and home circumstances. The
potential benefits of MIS have yet to be proven and careful analysis of
the long-term results will be needed.

Specific scoring systems have been widely used to assess
the clinical results following primary THA. Most studies attempt to
measure the outcomes of technique and procedure, and do not assess the
effect on general function and the satisfaction of the patient.
Measurement of outcome in THA is usually achieved using two different
systems, one that measures health-related quality of life (HRQOL), and
the second group that contain joint-specific tools. Only validated
outcome measures should be used for assessment.

The clinical results of the so-called first-generation
porous-coated cementless acetabular components with a minimum 10-year
follow-up have now been reported by a number of centers, independent of
their developers. From a clinical and radiologic perspective, reliable
fixation to the acetabular socket is reported in the range 95 to 99% of
cases. During the development of titanium fiber mesh, plasma spray, or
sintered cobalt-chrome beads as a surface finish

applied
to the outer diameter of metal shells, animal models suggested that
only 20 to 40% of the roughened surface achieved histologic bony
ingrowth (Figure 17-19).
Time has proven that this figure seems to give satisfactory stability
for prolonged loading, and the results do not appear to deteriorate
over time.

The noningrown areas of the acetabular cup have,
however, produced a new problem. Access of particulate debris,
especially polyethylene from the bearing surface to these noningrown
areas, gives rise to large areas of bone resorption that we now term osteolysis.
Although this phenomena of bone resorption was seen with cemented
acetabular cups and femoral stems, it was present at a much lower rate,
size of the lesions were much less, and perhaps most importantly, their
increase in size over time was much less than is seen with cementless
metal-backed acetabular components.

This increase in osteolysis rate is thought to be due to
design alterations in the porous-coated acetabular components.
Increased thickness of the metal shell to increase its stiffness, and
increase in femoral head size to 32 mm to improve range of hip movement
and joint stability led to a large decrease in thickness in the
polyethylene insert. In many cases, thickness of the insert fell below
the critical level of 6 mm, which is now known to cause a dramatic
increase in polyethylene wear at the bearing surface. Two other factors
played a part in this early problem: introduction of a secondary
bearing surface between the polyethylene liner and the inner surface of
the metal shell, and poorly designed locking mechanisms.

In many early designs, little attention was given to
congruity between the inner surface of the metal shell and the outer
surface of the polyethylene liner, and in some cases, there was up to 2
mm of gap between the two surfaces. Subsequent loading of the liner by
the femoral head transmitted the force through the polyethylene in a
nonuniform manner, leading to stress concentration, liner deformation,
and high wear not only at the primary bearing surface but also on the
outer surface of the liner, so-called back-side wear. An additional
problem with this lack of conformity also occurred. Any fluid in the
gap between the shell and liner is hydraulically forced out by the
bottoming out of the liner during weight bearing. This fluid, laden
with polyethylene particles, follows the path of least resistance
through screw holes in the shell, and so gains access to the
prosthesis-bone interface and leads to osteolysis (Figure 17-20).
Fluid can also be forced, under this hydraulic pressure, into other
areas of the prosthesis bone interface of both the femoral and
acetabular component where bone ingrowth has not occurred, again
causing osteolysis. This process has led to the concept of the
“effective joint space,” where surfaces of the prosthesis that are not
bony ingrown can potentially be bathed in joint fluid rich in debris
particles that cause osteolysis.

Locking mechanisms between the metal shell and
polyethylene liner involve the use of peripheral mechanisms around the
opening of the shell that capture the polyethylene liner. Incorporation
of this locking mechanism often involved additional sacrifice of
polyethylene thickness in an area that often was subjected to the high
stress and subsequent polyethylene wear. With time, these locking
mechanisms often lost their rigid grip on the liner, allowing movement
that further aggravated back-side polyethylene wear.

Despite these shortcomings, a prospective report was conducted of 100 hips with a porous-coated

anatomic (PCA) THA (Howmedica, Rutherford, NJ) with a mean age of 58
years. At 15-year follow-up, 17% (17 hips) of the entire cohort and 23%
(15 hips) of the living cohort (64 hips in 55 patients) had undergone
revision due to loosening of the acetabular component or osteolysis.
These results compare favorably with a study of cemented acetabular
components with a 25-year follow-up reported by the same author.
Patients still alive 25 years after surgery had a mean age of 56 years
at the time of operation. These patients received a cemented Charnley
THA with first generation cementing techniques. There was a 15%
revision rate for acetabular loosening, and a further 36% that
demonstrated definite or probable radiographic loosening.

The component that has been studied most extensively is
the Harris-Galante I (Zimmer, Inc, Warsaw, IN). This component
comprises a titanium fiber mesh coating that was usually implanted with
line-to-line reaming (last acetabular reamer used is exactly the same
diameter as prosthesis to be implanted) and screw fixation. A 9- to
14-year follow-up study of 71 hips in 56 patients less than 50 years
old at the time of surgery, revealed no revision or radiological
loosening of the components, but a 23% incidence of pelvic osteolysis.

Contemporary acetabular cup design trends now minimize
the number of screw holes through the acetabular metal shell to prevent
fluid gaining access to the prosthesis bone interface. The central
hole, present to allow the surgeon to ensure the cup is fully seated in
the acetabular bed after cup insertion, can now be blocked with a plug.
The locking mechanism has been redesigned to ensure the polyethylene
shell is solidly fixed to the metal shell, congruity between the inner
surface of the metal shell and outer surface of the polyethylene liner
has improved, and that the inner surface of the shell is smooth to
reduce abrasive wear against the liner. The minimum thickness of the
polyethylene (even with incorporation of the locking mechanism) is 8
mm. This has necessitated a return to smaller femoral head size when
smaller cups are used. The use of 32 mm metal femoral heads
articulating against conventional polyethylene has disappeared.

There are two principles of cementless acetabular
fixation. The first involves implantation of the component using a
technique that ensures initial

stability
of the metal shell against the acetabular bone. Supplemental fixation,
such as screws through the cup, spikes, lugs, or fins in the periphery,
may be used to achieve stability. Another method is to under-ream the
acetabular bone bed by 1 to 2 mm, and then use the roughness of the
outer surface of the metal shell to achieve a “scratch fit.” Other
initial fixation modes include an expansion cup, where the cup diameter
is reduced with a special instrument, the cup is implanted, and then
allowed to return to its initial diameter; and screw in designs, where
the cup has a thread incorporated into its outer surface, and is
screwed into the acetabular bone.

Achieving initial stability allows the second principle
to occur, that of bony ingrowth onto and into the metal shell.
Different surface finishes of the cup, including titanium fiber mesh,
cobalt-chromium beads, titanium plasma spray, tantalinium fiber metal,
and grit-blasting of the surface, are currently used by manufacturers
to obtain bony ingrowth. Hydroxyapatite may be used as supplemental
fixation and to improve initial stability, but unless used in
conjunction with a porous ingrowth surface, it does not appear to
maintain long-term fixation on its own.

Although acetabular cup modularity (separate acetabular
shell and polyethylene liner assembled at the time of surgery) was
introduced to reduce surgical inventory, increase surgical options, and
allow future liner exchanges, a disadvantage is that it introduces a
further interface from which wear debris can be generated. This has
lead surgeons back toward using a single piece acetabular cup, where
the bony ingrowth surface and polyethylene come as one piece.
Disadvantages of this design, however, are that visualization of the
medial acetabular margin during implantation is not possible (ensuring
the cup is fully seated), and also that few options exist for
additional fixation with screws through the metal shell if initial
stability is not achieved with a scratch fit.

Reports of increased rates of polyethylene wear with use
of cementless acetabular components is a major concern. Surgeon
selection of a porous ingrowth acetabular cup should balance wear
issues with concerns of initial stability (i.e., unusual acetabular
anatomy requiring additional fixation) and bony ingrowth (previously
damaged bone from irradiation or fracture seems to reduce the amount of
bony ingrowth) for each patient.

Three types of ceramic, alumina, zirconia, and more
recently oxinium, have been used as femoral head bearing surfaces
against conventional polyethylene. Alumina oxide has also been used in
a ceramic-on-ceramic (hard-on-hard) bearing surface. Early hip
simulator studies have suggested a five- to tenfold reduction in
polyethylene wear rate compared to metal femoral heads, but more recent
simulator studies are variable with some showing a similar rate to
metal on polyethylene, depending on variables such as temperature of
the polyethylene, type of lubrication, and introduction of third body
debris. In clinical studies of polyethylene wear rates made from
clinical radiographs, there are conflicting results, with some showing
that ceramic femoral heads confer an advantage and an equal number
showing no difference.

The claimed material advantage of ceramic femoral heads
over their metal counterparts is their self-polishing ability, making
them resistant to scratches. Generation of particulate debris is a
timedependent process, meaning that the advantage of

ceramics
only becomes evident after a given time period (claimed to be 5 to 7
years). This phenomena might explain the variability of the clinical
studies, where only those with a follow-up greater than 5 to 7 years
show a difference, but this has yet to be proven.

Two major issues with ceramic femoral heads include the
potential for fracture and the cost. With improved manufacturing
techniques, the risk of ceramic femoral head fracture was lowered to
approximately 1 in 50,000 implants. In 1999, a defect in the
manufacturing process of the supplier of 80% of the world’s zirconia
femoral heads led to a significant number of zirconia femoral head
fractures, which prompted a worldwide recall of all zirconia femoral
heads. Use of zirconia femoral heads, because of this defect and recall
has been limited.

When ceramic femoral heads were first released, they
were approximately three times as costly as metal femoral heads, and
although these costs have declined, the clinical studies to date make
it difficult for surgeons to justify their use on the basis of improved
prosthesis longevity.

Alumina oxide ceramic has been used in Europe for many
years as a hard-on-hard bearing surface. Wear rates with this
articulation have been reported as 10 times less than the lowest
conventional polyethylene wear rates. An identified drawback of
ceramic-on-ceramic is the limited head and neck sizes available. There
is also concern about the state of the bearing surface and the rim of
the acetabular component if a dislocation occurs. The elimination of
polyethylene does not eliminate the problem of wear debris and
osteolysis, as at least one alumina-on-alumina hip of early design had
a high incidence of wear and osteolysis.

This bearing couple has been around as long as
metal-on-polyethylene THA, but results from early designs were poor
because of poor manufacturing tolerances between the acetabular cup and
femoral head, leading to binding of the surface and eventual loosening
of the prosthesis. This result was compounded by the early prostheses
not being fixed to the bone, but even the introduction of acrylic
cement could not improve the results sufficiently to compete with the
results of metal-on-polyethylene THA. Improved manufacturing techniques
and research into lubrication methods at the interface have led to a
resurgence of interest in metal-on-metal THA. There is no doubt that
there is a dramatic reduction in wear rate to negligible levels, using
both radiographic wear measurement techniques and retrieval specimens,
but there remains major concern regarding the production of cobalt and
chromium metallic debris, and its elimination from the body.

A recent randomized clinical trial comparing one type of
metal-on-metal with metal-on-polyethylene had to be abandoned when very
high levels of erythrocyte and urine cobalt-chromium were detected.
Metallic wear particles have also been isolated in lymph nodes, livers
and spleens of patients with THA, raising concerns about the long-term
effects these materials might produce.

There have been many attempts to modify the wear
characteristics of polyethylene by altering its material properties.
Highly crystalline polyethylene (Hylamer) and Poly-Two (carbon
fiber-reinforced polyethylene) were developed to improve the wear
performance of conventional polyethylene, but clinical results have
been disappointing and their use has been discontinued. The most
promising current area for improving the wear performance of
polyethylene is by increasing the amount of cross-linking that occurs
between the individual polyethylene molecules.

Several studies have demonstrated the favorable
long-term results of this cross-linked polyethylene using cemented
nonmetal backed all-polyethylene cups. A report on 62 ultra high
molecular components gamma irradiated in air with 100 megarads of
ionizing radiation and a control group of 10 nonirradiated polyethylene
cups implanted between 1971 and 1978. Twenty-eight hips (45%) were
available for follow-up at a mean 17.3 years after operation. At 2
years, the cross-linked group had a linear wear rate of 0.15 mm/year
and the control group a linear wear rate of 0.39 mm/year. This high
initial wear rate was attributed to a highly oxidized surface layer.
After 2 years, the rates decreased to a steady state of 0.06 mm/year
for the cross-liked group and 0.29 mm/year for the nonirradiated cups.
These figures suggest a 79%

reduction in linear wear when using cross-linked polyethylene.

In 1999, a report on 14 chemically cross-linked
all-polyethylene acetabular cups articulating with a 22 mm diameter
zirconia femoral head. There were four non-cross-linked
all-polyethylene cemented acetabular cups with a 22 mm diameter
zirconia femoral head used as a control group. Follow-up ranged between
10 and 11.3 years. Rate of linear wear in the cross-linked polyethylene
group was 0.04 mm/year, while rate of wear in the control group was
0.16 mm/year—a 75% reduction in wear rate using cross-linked
polyethylene. Historical studies of THJR polyethylene wear using the
same implants, similar follow-up, and same wear measurement technique
gave a wear rate of 0.07 mm/year, demonstrating a 45% reduction in wear
when cross-linked polyethylene is used.

Results of these two important studies should be
interpreted with caution, however, because the method of cross-linking
was radically different from that currently used by manufacturers,
resulting in a material so brittle that a locking mechanism for use in
modular acetabular components could not be machined. Careful attention
to heating and cooling temperatures as well as cooling times by
manufacturers, has reduced this problem, and the material stiffness is
now much closer to traditional polyethylene.

The cross-linking process involves the bonding together
of carbon atoms in adjacent molecular chains, achieved by replacing
carbon-hydrogen bonds with carbon-carbon bonds. An unwanted side effect
of this process is the generation of free hydrogen atoms, charged
particles known as free radicals. Free radicals are known to cause
surface oxidation, which in turn increases polyethylene wear, and so
their reduction is critical during the manufacturing process.

Each of the THJR implant manufacturing companies uses
its own proprietary process for cross-linked polyethylene manufacture.
This process involves four steps, and each surgeon who uses a
particular brand of cross-linked polyethylene should be aware of the
steps used in its manufacture. There are choices available at each step
in the manufacture of cross-linked polyethylene.

Each choice has advantages and disadvantages, and may
result in differences in the wear performance and material properties
of cross-linked polyethylene from the different manufacturers. To date,
all wear testing of cross-linked polyethylene has been performed using
multidirectional wear simulators. Crossing path wear tracts are used in
an attempt to reproduce in vivo THJR conditions. Depending on the
degree of cross-linking (determined mainly by the dosage of radiation
of strength of chemical cross-linking), a 70 to 90% reduction in
polyethylene wear using a hip simulator can be achieved.

Since 1999, the Food and Drug Administration has
approved a number of highly cross-linked polyethylene components for
clinical use. The 2-year results of many clinical studies comparing
conventional and highly cross-linked polyethylene are expected in the
near future, but it will be some years before the true benefits of
cross-linked polyethylene (if any!) are known.

The use of cementless fixation for the femoral component
is a popular approach. Most early failures of cementless components
occurred on the acetabular side associated with accelerated
polyethylene wear and osteolysis. Porous-coated femoral stems have been
relatively successful in achieving constant fixation with a minimum of
unwanted problems. Assuming good initial fixation and a porous-surface
that encouraged bony ingrowth, the major problems have been with thigh
pain and stress shielding. Clinical results with the first generation
cementless stems are now out to 15 years, and assuming a good initial
femoral stem design is used with a good acetabular component,
porous-coated stems do not appear to show the deteriorating results
with time that have been reported with cemented components.

Hydroxyapatite is the most popular osteoinductive
calcium phosphate coating used in total hip arthroplasty. It was
originally used as the sole coating on the surface of a prosthesis to
obtain initial stability. Although early clinical results were good,
over time it appears to resorb leading to loosening and failure
necessitating revision. In one study comparing the same cup design with
either hydroxyapatite coating or porous coating, at a minimum follow-up
of 5 years, the aseptic loosening rate was 11% (21 in 188) for the
hydroxyapatite component versus 2% (2 in 109) for the porous-coated
component.

There is also concern that hydroxyapatite could increase
polyethylene wear and increase the incidence of osteolysis. This is not
shown, however, in one study of 314 hips (274 patients) in which a

tapered
hydroxyapatite-coated titanium femoral stem (ABG, Howmedica,
Rutherford, NJ) was assessed at a minimum 10-year follow-up. No stems
showed aseptic loosening and there were no cases of distal endosteal
femoral osteolysis. In another study of the same prosthesis, a mean
6-year follow-up (4-10 years) of 97 patients showed that although there
were very good clinical results and 100% survivorship of the femoral
stem, there was an alarmingly high rate of polyethylene wear, 0.24
mm/year (range, 0.05-0.76 mm/year).

The place of hydroxyapatite as a femoral stem coating has not been established in the literature at this time.

Surface hip replacement consists of resurfacing the
acetabulum with a thin layer of bearing surface, and replacement of
only the femoral head (not neck) with a metal ball. Historical failure
of surface replacement has been due to the production of wear debris
with subsequent bone resorption, loosening, and failure. Recently, to
avoid these problems a surface replacement using a metal-on-metal
bearing, allowing thin components and femoral design and
instrumentation to avoid varus alignment, has been designed. McMinn,
the designer of this prosthesis, reported on his experience with 235
joints over a 5-year period. There were no femoral neck fractures and
no dislocations. Over that time, problems with premature failure of
component fixation necessitated changes in fixation technique from
press-fit, to cemented, to the current system that uses a peripherally
expanded hydroxyapatite coated acetabular cup and a cemented metal
head. Potential advantages of this system over conventional THA include
conservation of bone for later revision surgery if necessary,
improvement in range of motion, and reduction in dislocation rate. The
major concern is early femoral neck fracture, which seems to occur more
commonly in females where osteoporosis might be a factor.

Although there has been a recent resurgence of interest
in performing THA through very small incisions by both the public and
implant companies, there are several reports in the literature of using
small incisions 15 years or more ago. Interest in this as a technique
for performing THA has probably followed from a resurgence of interest
in performing unicompartment total knee replacement approximately 5 to
7 years ago.

There are two types of minimally invasive THA: the
single-incision technique and the two-incision technique. Almost all
THA done in this manner are press-fit using porous-coated femoral and
acetabular components because of difficulty cementing through a small
incision. The single-incision technique can be performed as a limited
anterior approach as described by Hardinge, where a very limited
splitting of gluteus medius occurs through a 5 cm incision, or a
posterior approach through a 5 cm incision based over the femoral neck
posteriorly.

The two-incision technique employs an incision 1 cm
greater than the femoral head diameter, based over the femoral neck
anteriorly. Through this, the hip is dislocated anteriorly and a
femoral neck osteotomy performed. Acetabular preparation is performed
with the aid of an image intensifier, which is also used to ensure
correct positioning of a press-fit acetabular cup. A separate 4 cm
incision is made over the tip of the greater trochanter, and femoral
canal preparation and stem insertion are again aided by an image
intensifier.

Potential advantages of performing less tissue
dissection include quicker postoperative mobilization and a shorter
hospital stay. Studies that have attempted to demonstrate an
improvement in these outcome measures have found that patient
comorbidities and domestic circumstances have a greater impact on these
than the length of the surgical incision. Long-term results of
prosthesis survival using current minimally invasive surgical
techniques are not available.

Death associated with THA may be broken into several
categories: intraoperative, in-hospital but postoperative, 30-day
rates, and 90-day rates. Intraoperative rates in one study from the
Mayo Clinic database on 38,488 THAs in 29,431 patients between 1969 and
1997 showed 23 deaths during surgery. Predisposing factors included
elderly patients with preexisting cardiovascular conditions with acute
hip fractures or pathological fractures (17 of

23).
The strongest predictor of sudden death was use of cement for fixation
in all 23 cases. The authors recommend avoidance of cement if possible
in predisposed patients with acute fracture or malignancy.

The postoperative in-hospital mortality after primary
THA is in the range 0.1 to 0.8%, 30-day mortality between 0.15 to
1.42%, and 90-day mortality between 0.2 to 0.74%. Ninety-day mortality
after revision surgery is approximately the same as for primary THA
(0.6 to 0.9%).

Nerve palsy following a THA is an uncommon (prevalence
1%) but concerning outcome for both the surgeon and the patient.
Informed consent should include this complication as a possible
outcome. Predisposing factors include female gender, revision surgery,
lengthening of the limb, and bleeding tendencies. The most common nerve
injured is the sciatic (78%), followed by the femoral (13.2%) and the
obturator (1.6%).

The sciatic nerve is composed of independent tibial
(medial) and common peroneal (lateral) divisions that usually (90%) are
united as a single nerve down to the lower border of the thigh.
Although grossly united, the funicular bundles of the two are separate
and there is no exchange of bundles between them. In 10% of cases the
two divisions exit the pelvis as distinct nerves. In these situations
the tibial nerve always passes below the piriformis muscle while the
common peroneal component can pass above or through piriformis. The
common peroneal nerve is more prone to mechanical injury than the
tibial nerve because it has larger and more tightly packed funiculi
with less protective connective tissue, and because of its more lateral
position and tethering in the sciatic notch and head of fibula.

Nerve palsy can occur after THA from one of three
reasons: direct physical damage from dissection, poor retractor
placement, or entrapment during cerclage femoral wiring from excessive
traction or from compression usually by postoperative hematoma
associated with anticoagulation. During surgery attention should always
be given to sharp retractor placement over the anterior of posterior
columns. Formal dissection of the sciatic nerve is not usually
considered necessary, but its position should always be kept in mind
during surgery. If cerclage wiring of the femur is required, care
should be taken to split the posterior linea apsera as close to the
femoral shaft as possible. The wire should always be passed from
posterior and lateral to anterior and medial, and the leg should be
placed in the anatomic, not dislocated position during passing of the
wire.

If a nerve palsy is noted in the recovery room, an
attempt should be made to identify its origin. After checking that it
is not anesthesia related (patient received a partial spinal or
epidural anesthetic), the operative procedure should be reexamined. If
there is any chance of the nerve being trapped by a cerclage wire, or
the limb has been lengthened more than 3 cm, the patient should be
returned immediately to the operating room for either removal of the
wire or shortening of the leg (femoral head exchange if possible).
Hematoma compressing the nerve may give clinical signs of ongoing
bleeding and buttock pain. Discontinuation of anticoagulation may be
considered, but there is no absolute indication for reoperation to
evacuate a hematoma. Good outcomes have been reported with conservative
management.

During preoperative planning for THA, if there is the
possibility of the leg being lengthened more than 2 to 3 cm (such as
commonly occurs with reconstruction of a congenital hip dislocation), a
wake-up test can be discussed with the patient. This involves telling
the patient he or she will be asked to “wriggle their toes” on both
feet before extubation. The anesthetist must ensure no motor block or
muscle relaxant, and the ability to increase the patient’s level of
consciousness to allow understanding of commands. If the patient is
able to dorsiflex his or her toes on the contralateral foot but not the
toes on the operated leg, the patient is assumed to have a sciatic
nerve palsy, and consideration should be given to reoperation.

If the nerve palsy is treated nonoperatively, there is a
41% chance of complete recovery, a 44% chance of partial recovery, and
15% of patients will have a poor outcome. Prognosis is good if limited
motor function is present immediately after surgery or if some motor
function returns in the first 2 weeks postoperatively. Femoral nerve
injury generally has a better outcome than sciatic nerve injury.

One of the commonest findings noted by the patient after
primary THA is a leg-length inequality (approximately 25%). Most
studies of large groups of people show that approximately 30% of the
population have a discrepancy between 1 and 2 centimeters,

however,
and often the patients’ complaint of a longer leg after operation is
not borne out with clinical or radiologic measurement. This functional
leg-length difference is caused by contractures around the hip that
create a pelvic obliquity. Commonly, this is caused by tightness of the
gluteus medius muscle in a hip that has been short and has had a
decreased offset and then is corrected to the normal hip length and
offset. Over time, the feeling of a “long leg” will become less
obvious, but it is important to warn the patient of this preoperatively
so they are not worried that something was done incorrectly at the time
of surgery.

A true lengthening of the leg at the time of surgery can
be avoided by accurate preoperative planning, anatomic component
geometry, and intra-operative assessment (including use of outrigger
jigs, center of head to lesser trochanter distance, and relationship of
knees and heels before and after reconstruction as well as
intraoperative radiographs if there is still uncertainty. Correct
restoration of offset and leg length in a diseased hip is always
balanced against risk of dislocation by making the soft tissues too
loose, but both goals can always be achieved with careful surgical
technique.

A recent large study of 58,521 elective primary THA
(excluding acute fractures) in the United States reported a 3.9%
dislocation rate during the first 6 months after surgery, and for
12,956 revision THAs, a 14.4% dislocation rate. These figures are much
higher than are generally quoted for single surgeon series of THA,
which are approximately 0.5 to 2%. Factors that influence rate of
dislocation include patient factors (including cerebral dysfunction and
alcoholism), surgical approach (traditional posterior approach has a
fourfold greater risk of dislocation compared to Hardinge approach),
surgical factors (correct positioning of acetabular and femoral
component, trimming of marginal osteophytes from the acetabulum), and
implant factors (head size, head-neck ratio, elevated rim acetabular
liner, choice of neck length, choice of offset).

Recent techniques designed to repair the posterior
capsule and external rotators to the hip have reduced dislocations with
the posterior approach to 0.5%. A large study from the Mayo Clinic
demonstrated no benefit in dislocation rates with a larger head, but a
definite benefit with use of an elevated rim liner (2.2% incidence if
elevated rim used, 3.85% if no rim). Laboratory data suggests decreased
impingement between the femoral neck and acetabular rim with a 28mm
femoral head compared to a 22 mm femoral head, but no additional
benefit is gained by using a 32 mm femoral head for this purpose.

An important recent study has shown a large increase in
dislocation rate when the acetabular cup diameter was 32 mm or greater
than the femoral head diameter. The authors postulated that site of
attachment of pseudocapsule is moved further from the femoral head,
potentially allowing greater laxity, and they recommended use of a 22
mm diameter femoral head for acetabular cups less than 50 mm diameter,
and a 32 mm femoral head for acetabular cups greater than 62 mm outer
diameter.

Treatment of a dislocation depends on the suspected
cause and the number of previous dislocations. Most first-time
dislocations where there is no obvious cause such as osteophyte
impingement or component malposition are treated by closed reduction
and gradual mobilization (Figure 17-21). There
is no evidence that prolonged immobilization or physiotherapy after
closed reduction decreases the likelihood of recurrence. If an obvious
cause is identified, then operative treatment is indicated. The late
dislocation (occurring at least 2 to 5 years after surgery) raises the
possibility of wear of the acetabular component, which may require
revision.

Multiple dislocations are of great concern to the
patient because they cannot be sure of the circumstances when they will
occur. Often these patients request surgery. If the anteroposterior and
lateral radiograph fail to show acetabular malposition that can be
corrected by revision surgery, a decision must be made as to the best
form of treatment. Most recurrent dislocations are posterior, and in
the majority of these cases the original THA was performed through a
posterior approach. There is little guidance in the literature as to
which surgical approach should be performed in this situation, but the
authors prefer a Hardinge type approach through nonscarred tissue.

The least complicated option is to increase tissue
tension by increasing the length (and offset) of the femoral neck and
performing a careful soft tissue repair. A further option is
augmentation of the cup with a custom-made wedge of polyethylene (often
a liner cut into thirds), held to the existing cup with screws, placed
in the position where the dislocation was occurring. Trochanteric
advancements have not been reported as useful in the literature.
Complete revision of both components is a drastic option if both are
well-fixed and there is no obvious component malposition.

A study on the operative treatment of 116 recurrently
dislocating hips using one of the above operations, showed only a 61%
success rate. An increasingly popular alternative is revision of the
acetabular cup with a constrained liner. A constrained liner is one
that “captures” the femoral head and prevents it from dissociating from
the acetabulum. A 5-year follow-up of 56 patients who were treated with
a constrained acetabular component showed a 97% success rate at
preventing further dislocation. The potential disadvantage of decreased
range of movement and component loosening (because the internal surface
of the polyethylene is greater than a hemisphere causing impingement
and greater stress transmission to the prosthesis-bone interface) has
not been seen in medium-term follow-up studies of this component. It
remains the most effective solution to date for this distressing and
very difficult problem.

Those patients at risk of heterotopic ossification (HO),
the laying down of bone in tissues, should be identified preoperatively
and some form of prophylaxis instituted. Risk factors include
ankylosing spondylitis, previous acetabular or pelvic fracture, or
previous problems with heterotopic ossification during THA on the
contralateral hip.

Prophylaxis for this problem can be given with
pharmacological means or radiation. Aspirin has been shown to be not
effective in preventing HO, and most surgeons would now recommend
indomethacin, which is the most studied nonsteroidal anti-inflammatory
drug used for this purpose. One study showed that in 123 male patients
undergoing primary THA, given a 10-day course of 25 mg Indomethacin
three times daily, there was a 7.6% incidence of mild HO and no cases
of Brooker grade III or IV compared to the normal quoted prevalence of
3 to 10% of patients who experience functional impairment in the form
of diminished range of motion due to HO.

A randomized prospective study directly comparing 3 ×
3.3 Gy of radiation with 150 mg of diclofenac for 3 weeks showed that
radiation was slightly more effective than nonsteroidal
anti-inflammatory tablets in reducing the incidence of HO, but that
both were very effective. The dose of radiation that is effective at
preventing HO has steadily decreased over the last 20 years, and now
0.7 Gy is shown to be effective.

If HO does occur after THA and it is symptomatic, with
decreased range of motion after maturity (as determined by isotopic
bone scan usually 6 months after surgery), the HO may be excised and
the patient treated with postoperative radiation. Often a preoperative
CT scan is helpful to document exactly where the HO is located, and an
appropriate surgical approach can be used. With this regime an average
increase of 45° flexion and 25° of abduction can be anticipated (Figure 17-22).

Treatment of an acute periprosthetic infection depends
on the elapsed time since surgery. Acute infections occurring within 4
to 6 weeks of surgery, where the components are judged to be stable and
there is no reaction in the surrounding bone, can be treated with
thorough debridement of infected material, pulsatile lavage, and
prolonged IV antibiotics, yielding a very high prosthesis retention
rate (>95% depending on the infecting organism). The majority of
organisms in early infection are staphylococcus, and the antibiotics of
choice here are flucloxacillin combined with rifampicin.

In patients who sustain an infected THA more than 6
weeks after surgery, there is a greater chance of infected material
having gained access to the prosthesis patient interface. This does not
always equate with prosthetic loosening, but almost always necessitates
removal of the prosthesis. Removal of the prosthetic components can be
performed as a one- or two-stage revision procedure. The two-stage
exchange remains the standard for treatment, with the first stage
consisting of removal of all components and foreign material such as
cement, cement restrictors and wires, meticulous extensive debridement,
and accurate identification of the infecting organism. Use of an
articulating spacer containing antibiotic to maintain tissue tension is
becoming more popular, but even if not used, some form of antibiotic
impregnated cement should be placed in the acetabulum and upper femoral
canal.

Postoperative management consists of IV antibiotics for
2 to 4 weeks, followed by 2 to 4 weeks of oral antibiotics. Debate
exists about whether the patient should undergo a period without
antibiotic cover before reimplantation, and the value of hip aspiration
to ensure no organisms are present prior to reimplantation. The
patient’s erythrocyte

sedimentation
rate and C-reactive protein should be close to normal before
reimplantation, and this usually occurs 6 to 10 weeks following
prosthesis removal. Reimplantation is usually performed with a cemented
femoral component to allow further direct antibiotic delivery to the
bone. Acetabular revision may be performed with a cemented or
cementless component. Success rates for this procedure with a
nonresistant organism exceed 95%.

Recent studies have supported selective use of a
one-stage exchange, where the organism is considered of low virulence
and responsive to antibiotic therapy, an adequate debridement can be
achieved, and the patient’s medical condition contraindicates further
major surgery. Success rates in excess of 80% can be achieved when this
approach is used in carefully selected cases.

The incidence of fractures around the femoral component
of a THA, both intraoperative and postoperative, has been increasing
over the last 20 years. Intraoperative fractures are generally caused
by the need to achieve good initial stability of porous-coated
components. They are usually stable, involving a split in the femoral
cortex, and can be treated by keeping the patient from weight bearing
in the early post operative period. If an intraoperative fracture is
identified as unstable, cerclage wires can be passed around the femur.
Postoperative fractures are occurring due to an increase in the number
of THAs, younger patients, increasing survival of the prosthesis in
patients who are simultaneously loosing bone mass, and fractures
through osteolytic lesions.

The Vancouver classification (A, B, C) for these
fractures is based on the level of the fracture and stability of the
femoral component. Type A_G is a fracture of the greater trochanter. Type A_L
is a fracture of the lesser trochanter. Type A fractures are treated
conservatively if stable and operatively if unstable. Type B fractures
occur around the lower stem or tip of the prosthesis. Type B1 includes
fractures where the femoral component is solidly fixed. They are
treated with plating or strut allografts. Type B₂ includes fractures where the femoral component

is loose and are treated with revision of the loose femoral component
to a long-stemmed prostheses, usually cementless, to bypass the
fracture, and fixation of the fracture with wires or plates (Figure 17-23). Type B₃ fractures occur when there is severe underlying bone stock loss, and the femoral component is almost always loose (Figure 17-24). They are treated in a similar fashion to Type B₂
fractures, often with additional allograft. Type C fractures occur
below the tip of the femoral stem and are treated independently of the
prosthesis.

Nonunion of periprosthetic fractures is usually treated
by revision to a long-stemmed porous-coated component, but treatment is
difficult with a high rate of complications and a poor functional
outcome.

Fractures of the acetabulum are less common than of the
femur, and usually occur during insertion of a porous-coated component
into an acetabulum that has been under-reamed by 1 to 2 mm to gain good
initial stability. Fractures occur more frequently in women with
osteopenic bone. If an intraoperative acetabular fracture is
recognized, assessment of stability of both the fracture and the
component is required. An unstable component requires removal. If the
medial wall has been breached, over-reaming and a larger component with
rim fit can achieve a good result. If a column fracture has occurred,
it will need to be exposed and plated, followed by implantation of an
acetabular cage or a larger diameter cup. If a fracture occurs but the
cup appears stable, the situation can be accepted, and the fracture
will usually heal, but the component has a higher chance of not
achieving bony ingrowth, and thus requiring revision at a later date.

Improved fixation has led to emergence of other modes of
failure in THA. Wear is the removal of material, with the generation of
wear particles that occurs as the result of relative motion between two
opposing surfaces under load. The mechanical consequences (mechanical
thinning) of polyethylene wear have been recognized since the early
1970s when the first wear studies on THA suggested a limited life span
of the polyethylene bearing. This concept was partly responsible for
the development of modular acetabular components, which allowed
exchange of the plastic liner without disturbance of the bone
prosthesis interface (Figure 17-25).

Since the 1980s, it has been recognized that the
clinical consequences of wear—the release of excessive wear particles
into the biological environment—is the predominant cause of late
failure of THA, and this still remains the case today. When particles
within a certain size range are phagocytized in sufficient amounts by
macrophages, they enter a state of activated metabolism, with the
release of cytokines that result in periprosthetic bone

loss leading to eventual loosening and necessitating component revision (Figure 17-26).
The concept of the effective joint space includes all areas of the
joint that can be accessed by synovial fluid laden with wear particles.
This includes all areas of the bone and prosthesis (cementless
implants) and bone and cement (cemented implants) interfaces, which
often extend far beyond the areas of capsular attachment associated
with a nonreplaced hip. Synovial fluid can excite this response,
leading to areas of osteolysis (focal areas of progressive
periprosthetic bone loss) in locations removed from the primary
articular surface (Figure 17-27).

Wear can be classified into two areas: wear mechanisms
(adhesion, abrasion, and fatigue) and wear modes (conditions under
which the prosthesis was functioning when the wear occurred). Adhesive
wear occurs when two surfaces are bonded together under load, and
material is pulled away from one or more surfaces when movement occurs
between them. Abrasive wear is a mechanical process wherein asperities
on the harder surface removes material form the softer surface. Fatigue
occurs when local stresses exceed the strength of the material, causing
it to fail, usually after a particular number of cycles. Bearing
surfaces are designated primary (intended to articulate) or secondary
(not intended to move against any other material, whether this is the
patient’s bone or another component of the THA prosthesis).

Mode 1 wear occurs as a result of motion (which is
intended to occur) between two primary bearing surfaces. Mode 2 wear
occurs when a primary bearing surface moves against a secondary surface
(dislocation). Mode 3 wear takes place when a third substance gains
access between two primary bearing surfaces (often referred to as third
body wear). Mode 4 wear occurs when two secondary bearing surfaces rub
together, such as impingement of the femoral neck on the rim of the
acetabular cup, motion of the prosthesis in the bone, or motion between
modular connections of the femoral stem (morse taper fretting) or
acetabular cup (back-side wear).

Surgeons choosing a particular primary THA system for
implantation should evaluate all of these areas for that system
carefully, because the type and incidence of wear modes for that
prosthesis vary widely, a well as variation in when they occur over the
service life of that prosthesis. In most primary THA, the greatest
contribution to wear is from mode 1 wear—polyethylene particles
generated from the primary bearing surface (Figure 17-28). Two types of assessment for this are available: simulator studies and measurement from clinical radiographs.

Modern simulator studies of polyethylene wear now
incorporate the critical fact that, in the hip as opposed to the knee,
the path that a point on the femoral head makes within the polyethylene
is multidirectional, and there are often crossing paths. Although this
allows comparison of two bearing surfaces within the same simulator,
many surgeons are skeptical that simulator wear studies can truly
reproduce the in vivo situation under which a THA functions. Variables
still to be fully understood include lubrication medium, presence of
third body particles, temperature of the primary bearing surface, and
loading (there is some recent radiologic evidence that the primary
bearing surfaces of a THA dissociate by up to 1 mm during a normal gait
cycle). Wear simulator results vary greatly but tend to show
polyethylene wearing at a rate of 0.06-0.10 mm/million cycles.
Historically one million cycles were taken to represent one year’s
wear, but a recent pedometer study shows variation between 0.4 to 4
million cycles/year among study patients.

Measurement from clinical radiographs assumes that a
change in displacement of the femoral head between two radiographs
taken at different times represents polyethylene wear. Calculation of
volumetric polyethylene wear is then made on the basis of femoral head
diameter and direction of displacement. Measurements are based on
locating the center of the femoral head, and either measuring the
shortest distance to the edge of the acetabular cup, or locating the
center of the cup and measuring the difference between head and cup
center. Radiostereometric analysis (RSA), originally developed for
measurement of component migration, can be used if beads are
prospectively implanted in the pelvis at the time of surgery, the
acetabular cup has beads implanted, and a dedicated RSA setup for
taking radiographs is available. It can only be used prospectively, and
expense usually constrains numbers of subjects that can be recruited.

Retrospective measurements from clinical radiographs can
be made by placing radiographs on a tablet and using a stylus to
digitize points (manual method) or by taking points directly from the
digital image of an radiograph (usually automated). Recent studies have
shown good correlation between radiologic measurements and retrieved
prostheses. Hips with linear wear rates of greater than 0.2 mm/year
have been associated with an increased prevalence of osteolysis, while
hips with linear wear less than 0.1 mm/year have infrequently been
associated with osteolysis.

The historical mode of failure of cemented femoral
acetabular and femoral components was by aseptic loosening associated
with mild or moderate osteolysis, and moderate or severe component
migration. With the advent of modularity, hybrid THA, multiple
revisions, porous-coated femoral and acetabular components, and
alternate problems (fractured ceramic femoral heads), different modes
of failure require different solutions to be applied. The primary
philosophy is to achieve a solidly fixed, stable component with minimal
bone loss (or bone restoration where necessary). The minimum surgery,
carrying with it the minimum risk of complications to achieve this aim,
is often the best solution.

If a well-designed component with a good clinical track
record is found to be solidly fixed at the time of revision surgery,
and there is minimal evidence of wear of its primary bearing surface
(or it can be replaced, as with a modular femoral head or modular
acetabular cup), then consider retaining the component at the time of
revision. This decision must also take into account ease of exposure
(exposure of an acetabular defect around a solidly fixed femoral stem),
leg length and tissue tension issues (increased risk of postoperative
dislocation if neck length options are limited), and head size (a
monoblock femoral stem with a femoral head diameter that cannot be
matched to options available for the acetabular revision).

Revision for progressive periprosthetic osteolytic
defects in the presence of well-fixed components often involves
grafting of the defect with allograft, removal of screws from the cup,
and downsizing the diameter of the primary bearing surface from 32 to
28 or less millimeters (to increase polyethylene thickness) by
replacing the polyethylene liner (Figure 17-29). Increasing the neck length,

to rectify poor tissue tension caused by the exposure (which can lead to an increased risk of dislocation), is often necessary.

Bony defects of the femoral shaft can be assessed by
good quality anteroposterior and lateral radiographs of the hip, as
well as proximal femur beyond the tip of the femoral prosthesis.
Defects are always more extensive than they appear on radiograph.
Acetabular defects often require additional imaging, such as Judet
views to assess the anterior and posterior columns, and occasionally CT
if a pelvic dissociation (loss of bony continuity through both anterior
and posterior columns) is suspected. If there is intrapelvic migration
of the acetabular component, pelvic angiography to document the
proximity of the iliac vessels is often required.

Choice of acetabular revision components is often
dictated by the experience, training and philosophy of the operating
surgeon, and tempered by the extent of any bony defect (or other major
structural consideration, such as pelvic dissociation) present. The
most commonly used grading system for acetabular bone loss emphasizes
the presence or absence of the acetabular rim, deficiencies of the
acetabular dome and walls, and integrity of the anterior and posterior
columns. Type I defects, with intact rims, walls, and columns, may be
dealt with using cementless components with or without supplemental
screw fixation, or cemented components using impaction grafting
techniques. Twelve-year results of the Harris-Galante titanium fiber
mesh socket used for revision of type I defects approach the 95% good
or excellent results they enjoy in primary THA.

Type II defects have a distorted rim and wall, but an
intact anterior and posterior column, and may be treated with a
cementless acetabular component with supplementary screw fixation.
Defects are filled with compacted allograft bone chips. If the major
acetabular deficiency is superior, with minimal anterior or posterior
column bone loss, a slightly higher hip center may be accepted to
prevent over-reaming and sacrifice of good bone from the anterior and
posterior columns.

Type III defects have a missing rim, the walls are
severely compromised, and the columns offer no support. Two choices are
present here—a very large (“jumbo”) acetabular component (outer
diameter greater than 66 mm) or an antiprotrusio cage. The guarded
long-term prognosis of structural acetabular allografts (failure rates
up to 50% at 10 years have been reported), and the difficulty and
expense of supporting a dedicated structural allograft bone bank, have
limited their use outside of dedicated hospital units. For most
patients (elderly with a multiply revised hip and component migration)
with a type III defect, the component of choice is either a jumbo or
bilobed cup. Jumbo cups are a good option when there is a large central
defect. The acetabular margin is reamed down to create a
horseshoe-shaped lateral acetabular margin, and an oversized acetabular
cup is then placed on this margin and held with multiple screws. The
central defect can be allografted. Alternatively, if an oval defect is
present and the surgeon wishes to restore the hip center to normal, a
bilobed cup may be used.

In a younger patient with a type III defect, where bone
stock preservation and restoration must be maximized to allow future
procedures during that patient’s lifetime, compacted allograft bone
chips can be packed into defects (including structural allograft
supplementation to prevent bone chip migration). If there is likely to
be less than 50% contact between host bone and the acetabular
prosthesis, an antiprotrusio cage should be used. This is attached to
the ilium ischium and pubis, overcoming problems with column or medial
wall deficiency. Bypassing of the acetabulum allows load sharing to
occur with underlying allograft, encouraging its incorporation for
restoration of bone stock. A polyethylene acetabular socket is then
cemented into the cage. Care should be taken with component
orientation, because bony landmarks that can usually be identified will
no longer be available (Figure 17-30).

Pelvic discontinuity cannot always be predicted
preoperatively, but where there are major column defects and acetabular
component migration, it should be suspected, and appropriate component
inventory for intraoperative treatment of discontinuity should be
available. Three choices are available:

Selection of an appropriate revision femoral component
should take into account any difficulties with potential femoral
fractures, cortical windows, or extended femoral osteotomies that might
be encountered during extraction of the existing implant. Well-fixed
cemented and cementless stems may be difficult to remove, and vigorous
extraction attempts that might lead to bony fracture should be avoided (Figure 17-32).
Cemented implants should be initially removed from the cement mantle. A
decision can then be made as to whether the cement can be removed from
the canal proximally, or whether a distal cortical window will be
required. An extended femoral osteotomy may be performed if there is
varus femoral remodeling. Most well-fixed cementless femoral implants
have to be removed using thin power instruments to divide the areas of
bone where ingrowth has occurred into the porous-coated surfaces of the
prosthesis.

Cemented femoral revision relies on interdigitation
between the cement and the bone for a mechanically durable interface.
The loss of cancellous bone seen at revision surgery compromises this
interdigitation. Although the 10-year results of cemented femoral
revisions have an 80 to 85% prosthesis survival, improved cementless
prosthesis design has decreased the role of cemented femoral revision.

Occasionally, it will be necessary to change a
well-fixed cemented femoral component due to unusual head size, or leg
length issues. If the cement mantle is radiologically intact, another
component may be cemented into the intact cement mantle without its
removal, with good medium-term (5 to 7 year) results.

Cementless femoral revision has become increasingly
popular as it provides a stable initial construct and has good
long-term results. Proximal porous-coated implants provide the
theoretical advantage of biologic fixation with a minimum of stress
shielding (loss of bone from the proximal femur when it is not loaded).
This benefit is outweighed by the

disadvantage
of obtaining initial stability and long-term fixation in the damaged
proximal femur, and their use in the revision setting is limited.

The most popular method for femoral revision is
bypassing the damaged proximal bone using diaphyseal fixation. Most
femoral stems designed for diaphyseal fixation are either extensively
porous-coated devices with a cylindrical distal shape, or grit-blasted
fluted devices with a conical distal shape. Titanium stems have a lower
modulus of elasticity to potentially reduce stress shielding, but
cannot be fully porous coated because notch sensitivity reduces their
strength, increasing the risk of stem fracture. The stress shielding
associated with diaphyseal fixation has not yet become a major clinical
problem, but concern remains about extraction of an extensively
porous-coated device.

There is little literature suggesting the best surgical
approach to employ in revision surgery. The posterior approach makes an
extensile exposure of the femoral shaft easier, but, as in primary THA,
suffers from a higher dislocation rate. If an anterior approach has
been used previously, there is no contraindication to using a posterior
approach for revision, and the converse also applies.

A trochanteric osteotomy may be required if extensive
exposure of an anterior or posterior column deficiency is required,
such as when an antiprotrusio cage is used. A recent advance in
revision surgery has been the extended greater trochanteric osteotomy,
where in the greater trochanter and the lateral third of the
circumference of the femur are raised as a flap, with their soft
tissues still attached, for approximately 10 cm down the femoral shaft (Figure 17-33).
As well as good acetabular exposure, it has the advantage of
facilitating removal of well-fixed cemented and cementless implants in
a bone sparring manner. Indications include:

The vast majority of femoral and acetabular revisions
currently use cementless fixation of both the acetabular and femoral
components. A technique popularized

by
the Exeter group has used an alternative for many years with good
medium-term clinical results—impaction grafting with cement. This
method seeks to restore cancellous structure to the femoral canal using
densely packing cancellous bone chips. These bone chips are packed
tightly enough to provide axial and rotational stability to a femoral
stem when it is cemented into them. Technical difficulty has limited
its use, however, with many centers reporting a high incidence of
femoral fracture and component subsidence. Ten-year results from the
center that initiated the technique have recently been published, but
until long-term results show a definite improvement over cementless
femoral revision with regards to restoration of bone stock, its use
will remain limited to a few dedicated revision centers.

These constructs are required if there is severe
segmental proximal femoral bone loss that is inadequate to provide
initial stability for conventional revision femoral stems. Allograft
prosthetic composites, where the prosthesis is cemented into the
allograft and has a porous-coated surface for bony ingrowth of host
bone where available, allows better soft tissue attachment than does an
all-metal tumor prosthesis. A tumor prosthesis is reserved for the
older low-demand patients who need a faster operation and earlier
weight-bearing mobilization.

Allograft prosthetic composite results in the medium
term have been good, but there remains a high rate of dislocation and
infection. Healing of the allograft-host bone junction usually requires
3 to 6 months of protected weight bearing. Segmental allografts remain
largely avascular, with a zone of revascularization only extending for
1 to 2 cm into the allograft from the junction with host bone. Greater
trochanteric union to an allograft with a stable fibrous or bony
interface occurs in approximately 70% of cases.