REVISION THR SURGERY

REASONS FOR REVISION

These include: -Aseptic loosening

-Loosening from sepsis
-Sepsis
-Recurrent dislocation/subluxation
-Osteolysis with bone loss

EVALUATION OF THE PAINFUL THR

Evaluate the patient in general

The reason for revision
Problems removing the old, what’s there?
Evaluate the expected bone loss
Possibility of infection
Plan and fall-back

Pain in the groin or down the front of the thigh on weightbearing associated with a
painful limp is strongly suggestive of hip pain. Thigh pain is often an indicator of a loose
femoral component. Episodes of subluxation may indicate wear in the acetabulum.

DDx -Low back pain

-Trochanteric bursitis
-Vascular pain

When history is suggestive of hip origin of the pain, then it is very important to
differentiate aseptic loosening from sepsis.

Blood tests

WCC has no value in diagnosis

ESR > 40mm 6 months after surgery is suggestive of infection
Blood cultures of value only in acute infection
CRP > 200, especially with an elevated ESR is highly suggestive of infection. The
sensitivity of these tests ranges from 60% to 96%. The specificity from 85% to 100%.

Aspiration

Most people favor a selective approach. The sensitivity, specificity and accuracy have
been shown to be high. This is recommended in hips < 5 years old, and also in patients
 2

with an elevated ESR. The polymerase chain reaction (PCR) is a new test which can
produce large quantities of specific DNA sequences from relatively small material.

X-rays

Radiolucent lines and endosteal scalloping are seen in both aseptic loosening and in
sepsis. The presence of periosteal new bone is specific to the infected hip joint. This is
usually seen at the junction of the metaphyseal and diaphyseal bone on the medial side.

X-ray indicators of loosening include:

-Distal migration of components is the most significant

-Cement fracture
-The appearance of a pedestal beneath a stem tip with periprosthetic osteolysis
-A 2mm or wider cement-bone interface

Nuclear medicine

WBC labeled Indium scans are thought to be more superior in diagnosing infection, than
the previously used gallium. These are 80-90% sensitive, and 85-100% specific in Dx of
infection.

It takes approximately 1 year after cemented THR for patients to have no significant
uptake of Technetium. With uncemented prostheses, this will take about 2 years.

New work is being done with Indium labeled IgG, which may be very good at diagnosing
low-grade infection. Another new product is Technetium labeled monoclonal antibodies.

THE INFECTED THR

Stages (Fitzgerald)

Stage I -Sepsis within the immediate postoperative period, usually caused by

Staphylococcus Aureus

Stage II -Appearance of infection 6 to 24 months from surgery, usually due to

Staphylococcus Epidermidis

Stage II -Sepsis occurring more than 2 years after surgery. This is presumably due
to seeding from haematogenous spread. (e.g. dental procedure, ulcer).
 3

Microbiology

The more potent organisms are those which produce a glycocalyx. This is a glycoprotein
that facilitates the adhesion of the organism to an implant. This protects them from the
phagocytic uptake and opsonization by antibodies.

Less virulent More virulent

Staph Epidermidis Gm –ve bacilli
Staph Aureus (Meth. Sens) MRSE
Anaerobic Gm =ve cocci MRSA
Streptococci Enterococci

Diagnosis

* In general, with postoperative haemoserous discharge, if the patient is well and blood
indices stable, then this can be ignored. If it continues past 10 days, then suspicion of
infection must be entertained. If this is the case, then the correct procedure is open
drainage and washout of the hip with swabs taken. If a low virulence organism is
cultured, then this irrigation and parental antibiotics is usually all that is needed.

* Acute stage I sepsis is diagnosed as a patient with a fever, increased blood indices and
red wound +/- drainage. Here blood and drainage cultures are important.

*Stage II sepsis is usually diagnosed with all of the above tests. Hip aspiration is very
important in this situation.

* Intraoperatively a frozen section can be performed. The presence of > 5 WBC’s per
high-powered field is thought to be indicative of infection. (Mirra). The most inflamed
tissue and that from the bone-prosthesis interface are obtained.

Treatment

Early stage 1 can be treated with debridement, irrigation, IVI antibiotics and
preservation of THR.

All other stages need to have the prosthesis removed.

One-stage exchange may be considered for low virulence organisms. After debridement,
the new prostheses are replaced using antibiotic impregnated cement.

Two-stage exchange is generally the safest procedure. Here the implant is removed, and
the wound has meticulous debridement with deep cultures taken. It is important to
remove all the cement, thus an osteotomy may be needed. Antibiotic beads (PMMA
cement impregnated with Tobramycin) are often placed into the joint prior to closure.
These are valuably, as the large surface area allows greater leaching of antibiotic. The
 4

patient is then treated with IVI antibiotics, and when the wound has healed, the ESR is
monitored. When this returns to normal, the second stage replacement can be considered,
This has been performed as early as 1 month, although 2 is recommended. This has a
success of up to 93%.

REVISION FOR ASEPTIC LOOSENING/OSTEOLYSIS

Often the main issue involved with revision for aseptic loosening is that of poor bone
stock. This osteolysis may involve the acetabulum or femur or both. Often this osteolysis
also predisposes the femur to fracture, and both a loose prosthesis and fracture may have
to be treated simultaneously. Thus if there is progressive osteolysis without loosening,
revision surgery is often indicated to prevent fracture from occurring.

REVISION FOR SUBLUXATION/DISLOCATION

The success of this surgery, depends on identifying the cause. As well as this, the
components need to be assessed for loosening and sepsis must be excluded. Often surgery
may involve:
-Replacing the acetabular liner with an elevated rim, or merely rotating it.
-Increasing the neck length
-Removing the acetabulum and improving it’s position
-Removal of impinging bone
-Replacement of both components, with adjustment of offset and version.
-A trochanteric slide if no cause found

SURGICAL APPROACHES

1. Routine anterior and posterior approaches

In general for straight forward revisions the surgeons choice of these is OK. If exposure
of the posterior column of the acetabulum is needed then this approach should be made.
After removing the head, the neck is retracted posteriorly for the anterior approach
(Sciatic N) and it is retracted anteriorly for the posterior approach (Femoral N).

2. Trochanteric osteotomy

This may improve access and help with removal of cement. It can also be turned into a
trochanteric slide to improve stability. The main problem is non-union (20%). An
extended osteotomy with about 12cm of anterolateral femoral cortex can be used, this
helps with component removal, but meticulous reconstruction is needed. This latter
method is now commonly used as it significantly aids in component removal. Must
preserve the vastus lateralis attachment to prevent proximal migration. Look at the pre-op
 5

X-rays, if there is significant trochanteric overhang then # may occur getting the
prosthesis out, or the revision may be placed in varus-Æthus osteotomize the femur.

3. Extensile lateral approach of Mallory and Head

This involves taking the vastus lateralis anteriorly in continuity with the anterior fibers of
gluteus medius and minimus. A sliver of trochanteric bone is taken with the muscle. The
main tendinous portion of medius is left intact.

4. Extensile iliofemoral approach

This gives good access to the anterior column of the acetabulum.

5. Tri-radiate approach

This is reserved for the most difficult revisions, especially if the patient is obese.

FEMORAL RECONSTRUCTION

Aim to restore length, stability and offset.

Removal of the failed component

In order to make this part of revision surgery easier, some surgeons use a smooth-
textured femoral component. This often leaves the cement behind as the prosthesis is
removed. Techniques employed to remove this cement include:
-Ultrasonic tools (e.g. Midas Rex)
-Trochanteric osteotomy, can be an extended osteotomy with part of the
anterolateral cortex.
-Controlled anterior femoral perforations can be used, especially with the distal
plug when using the Midas Rex.
-Flexible lights and fluoroscopes can be used distally
-Large cortical windows can be utilized

With the extended femoral osteotomy for removing a fully porous coated prosthesis, the
gigli saw can be passed around the medial aspect of the stem and directed distally to the
level where the stem becomes cylindrical. Here a metal cutting burr can section the stem,
and the proximal portion is then removed, the distal cylindrical part can be removed with
circular hole drill.
 6

Bone deficiencies

AAOS
1 Segmental (Loss of cortical shell) Grade I-III severity of bone
2 cavitary (Loss down to cortical bone) loss
3 Combined
4 Malalignment Level 1 To the LT
5 Stenosis Level 2 < 10cm below LT
6 Discontinuity Level 3 > 10cm below

OR newer way

Type I =minimal metaphyseal involvement and adequate cancellous bone

Type II =metaphyseal damage, but minimal involvement of the diaphysis

Type IIIA =metaphyseal damage with 4cm of reliable cortex proximal to the isthmus

Type IIIB =metaphyseal damage with 4cm of reliable cortex distal to the isthmus

Type IV =Extensive metaphyseal and diaphyseal damage with ballooned/thin

cortices with widened canal.

General treatment

Modular prosthesis is preferred to adjust the proximal to distal geometry

Use a proximal coated prosthesis where possible
-SROM for proximal fit (I, II)
-Mallory for distal (III)
->16mm shaft then use impaction allograft in the young
->5cm proximal circumferrential loss use allograft

Graft techniques

*Impaction bone grafting has been advocated for treating cavitary defects, and has been
used in type III and IV deficiencies. The morselized graft is impact against a cement
plug, using a tapered oversized femoral component. This cancellous graft however cannot
be relied on to provide stability for an uncemented prosthesis.

Indicated -Young with severe bone loss

-Canal > 16mm (Poor results with distal fixation)
-<4cm of intact diaphysis for fixation

Complications -Dislocation in 3-6%

-Infection in 3-6%
-# or perforation in 5-24%
 7

-Subsidence, massive (>5mm) in 10-20%. Associated with pain.

Occurs at the cement-bone and cement-prosthesis interfaces.
Proximal coated may reduce this by reduced cement-prosthesis
subsidence. The significance of subsidence is unknown, not all
need revision. Best results are 90% survival at 6 years.

*Cortical strut grafts should not be relied on to provide structural support. These are
useful when the proximal bone is deficient, such as type II and III defects. In 99% the
grafts unite at about 8 months, with maturation taking 3-5 years.

*With > 5cm circumferrential loss, a structural allograft can be used. Cannot use a calcar
replacing prosthesis, and tumour prostheses have metal failure and poor adhesion of soft
tissues.

Advantages -Unites well to host bone

-Good soft tissue attachment

Disadvantages -Disease transmission

-Variability of graft quality
-Resorption (Deep freezing decreases the immunogenicity and less
resorption)
-Complete incorporation does not occur, only up to 20%

The long stem component is cemented into the proximal allograft, then the rest of the
stem is press-fit or cemented. The graft host junction can be step cut to give rotational
stability. This is supported by on-lay grafts or circlage wires. (Union of allograft to host
bone is essential for long term outcome). These have shown 70-85% good results at 2
years.

Complications -Non-union in 4-25%

-Infection in 4-15%
-Resorption Not always needing revision, main late concern
-trochanteric non-union.

Cemented femoral stems

Early on significant failures occurred due to loosening at the cement-bone interface. This
may be due to difficulties cementing onto a sclerotic bed. Improvement in results with
cemented femoral prostheses has been seen with the later generation cement techniques.
(Up to 94% success at 10 years). In general it is better to use a standard stem, as long as it
passes at least 2 femoral diameters past the area of cancellous defect. If a perforation is <
than 1/3 of the shaft diameter, then the stem needs to pass at least 2 ½ cortical diameters
below this breach. If this perforation is > 1/3rd of the diameter, then additional on-lay
graft is needed. Once the stem passes beyond the isthmus however, the cement restrictor
becomes ineffective.
 8

Uncemented revision stems

Here it is critical that the implant is stable to allow biological ingrowth. Because of the
concerns about osteolysis, many have been wary to use these, especially the long stem
prostheses. Good results however have been shown at 5 years with fully porous coated
stems. These rely on distal fixation, which is advantageous with proximal loss. Rotational
stability is also very important, thus distal flutes are often used. A minimum of 4cm
interference fit is required for stability, with 6cm preferred. The proximally coated
prostheses can be used in type I and II deficiencies, as these have enough good proximal
bone. These can be used with on-lay and cancellous grafts. It is vital to get proximal
stability, to allow loads to be transferred proximally and avoid shielding. A calcar-
replacing component can be used so that loads are transferred to the LT. Trying to obtain
a good proximal and distal fit with long stem prostheses has been associated with a high
fracture rate. into the host residual femur. Anterior bow is more anatomical, better fit and
less #’s. 96% survival at 5 years. Distal fit is better if the hip is unstable as you can alter
the version.

ACETABULAR RECONSTRUCTION

For acetabular revision you need

-Banked bone
-Variety of cups
-Cages
-reconstruction plates

On the AP X-ray
-Migration of the hip center indicates proximal bone loss
-Ischial osteolysis indicates inferoposterior loss
-Destruction of the teardrop indicates inferoanterior loss
-Destruction of Kohler’s line indicates medial loss

Acetabular bone deficiencies

This is classified by the AAOS as:

Type I Segmental (Involve the supporting cortical shell) = uncontained

Peripheral
-Superior, Anterior or posterior
Central (Medial wall absent)

Type II Cavitary = Contained with supporting bone intact.

Peripheral
-Superior, Anterior or posterior
Central (Medial wall intact)
 9

Type III Combined deficiencies

Type IV Pelvic discontinuity

Type VI Arthrodesis

Paprosky has also classified these according to how well they can support a cementless
component. i.e. how good is the rim, dome, walls and columns.

Type I -Intact walls and columns, >50% supportive bone

Type II -(A-C) Increasing distortion of the supportive rims/columns, but > 50%
host contact.
Type II -Severe distortion of walls/columns, < 50% contact

Bone grafts

Morselized = fragments 5-10mm, better than milled bone.

Simulated structural is bone from another region fit to the acetabulum
Anatomical structural = donor acetabular graft.

Structural grafts fail due to resorption or fragmentation, thus cannot rely on these for long
term support. These merely unite to host bone, do not remodel. It is better though to use a
structural graft beneath a cage to share the load, as morselized graft can only do this
when it remodels, by this time the cage may have failed.

For complete acetabular grafts, cut the host from AIIS to notch, form a tongue in groove
for the graft, hold with screws. Use a cemented cup.

Treatment principles

*Best option if possible is to use an uncemented prosthesis. Cemented acetabulum

revision prosthesis have been shown to have a failure rate of up to 50% at 8 years. This
contrasts with the survival of 87% of uncemented cups after 11 years.
 10

*Need at least 50% supportive host bone to use an uncemented cup

*If < 50% then a cage is needed with a cemented cup

Minor cavitary defects, >50% contact

-Larger components are usually required to enable rim fit at revision.

-Usually need support with screw fixation.
-Contained small cavitary defects can be filled with morselized graft (Up to 25%)
-The cup may be medialized to enable good superolateral coverage, with breaching of the
medial wall not associated with a poor outcome. (Dorr).
-Excellent results at 5 and 10 years (87-90%)

Minor segmental/combined defects, > 50% contact

-Small segmental defects can be ignored if they are peripheral, use a larger cup.

-If combined, but still > 50% contact, then can use a simulated allograft. Here the femoral
head graft can be shaped and fixed (Plate or screws) to the pelvic rim- this is usually
performed for superior defects. It is shaped like a number 7 and reamed, then an
uncemented prosthesis used. At these areas ingrowth will occur, whereas the allograft
will unite onto the cup, but no ingrowth will occur. Initial results at 4 years showing good
results, but at 10 years, 47% were loose.

Larger defects, <50% contact

*Large cavitary defects, with < 50% host contact can be treated with a roof reinforcing
ring, morselized graft and cemented cup.

*Moderate combined defects can be treated with a spanning acetabular cage, supporting
morselized or femoral head graft, and an uncemented cup.

*Can used an oblong cup for superior deficiencies to increase cortical contact to > 50%
and maintain the hip center. Hemispherical cups can be placed high with good results, but
need a trochanteric advancement and a long neck calcar replacing femur to maintain leg
length. Oblong components can be custom made off CT or used off the shelf. Problem
with fit may occur. Must lateralize the cup.

*Massive defects can be treated with entire acetabulum allografts. These have a 60%
failure rate at 16 years, but may further revision easier.

*Use of bipolar prostheses with acetabular bone grafts have had poor results.
 11

Reconstruction rings

When acetabular defects are severe (Type II and IV), supporting metal rings are a means
of optimizing the remaining acetabular bone. If less then 50% host contact is possible
with a cup, then an antiprotusio cage is indicated. These will bridge the deficiency, and
allow graft to be packed behind them. These stabilize the graft, and prevent motion
between the allograft and the acetabular component. The component used, is usually an
all polyethylene cup which is cemented inside the reconstruction cage.

Types of cages used include;

*Containing-convert an uncontained defect (Segmental or combined) into

contained defect. E.g acetabular rim, medial mess cages.

*Lateralizing – Transfer weightbearing forces to the roof and rim of the

acetabulum whilst compressing the underlying graft. I.e these rest on the rim. E.g.
Muller. These are primarily indicated for the treatment of protrusio. For large
contained defects, the inferior ring must be supported by host bone, otherwise a
spanning cage is needed.

*Bridging- Span major defects. E.g Ganz, Birch-Schneider

The Muller ring is fixed to the ilium alone. This is the least bulky and is used in mainly
central cavitary and segmental defects. Failure occurs in 50% if used inappropriately

The Ganz ring has a hook which curves around the teardrop and enters the obturator
foramen. This hook only aids in positioning, with the ring screwed into the ilium alone.

The Burch-Schneider antiprotusio cage is much larger, and can span large defects and
obtain stability in relation to both the ilium and ischium. It can be used for large
segmental and combined defects, as well as be useful in treating pelvic discontinuity.
(With a posterior plate). Here the ischial flange is driven into a slot cut in the
ischium.(Better than ischial screws). 24% failure at 2-11 years, half of these were from
sepsis.

Must be stable to host bone. Better also if they have some intrinsic stability. Plate pelvic
dissociation before inserting the ring.

If unable to hold the cage to the host, then bulk allograft may be needed.

Salvage options for failed cage

May need a more substantial cage

Structural allograft
Girdlestones