The Journal of Arthroplasty Vol. 20 No. 4 Suppl.

2 2005

Preoperative Planning for Revision

Total Knee Arthroplasty
Avoiding Chaos

Kenneth A. Gustke, MD

Abstract: Use of preoperative planning is important in avoiding an unstable

revision total knee arthroplasty. Physical examination should determine the status of
the collateral ligaments so that implants with appropriate constraint are made
available. Radiographic examination should determine if bone loss is present and
whether primary or revision implants will be needed. Preoperative determination of
the joint line position will simplify the surgery and facilitate flexion/extension space
balancing. Three hundred sixty-five revision total knee arthroplasty surgeries were
performed from 1987 to 2003. Of those, 82% were performed with unconstrained
implants. Ten percent of the knees required varus/valgus constraint and 8% knees
used rotating hinge components. Use of preoperative planning techniques resulted
in postoperative stability in 99% of the revision total knee arthroplasties performed.
Key words: revision total knee arthroplasty, preoperative planning, joint line
position, unstable total knee arthroplasty.
n 2005 Elsevier Inc. All rights reserved.

A significant complication seen in unsuccessful
total knee revision surgeries is instability [1].
Instability can occur from inadequate implant
constraint for the ligaments present. An implant
with the required constraint can be determined
preoperatively [2]. Knees with unconstrained or
constrained implants can still be unstable if the
flexion/extension space is not balanced and the
joint line position is incorrect. Because finding
the appropriate joint line during surgery is
From the Division of Orthopedic Surgery, Florida Orthopedic
Institute, University of South Florida College of Medicine, Tampa,
Florida.
Benefits or funds were received in partial or total support of
the research material described in this article from Centerpulse/
Zimmer Orthopedics (Warsaw, Ind).
Reprint requests: Kenneth Gustke, MD, Florida Orthopedic
Institute, 13020 N Telecom Parkway, Temple Terrace, FL 33637.
n 2005 Elsevier Inc. All rights reserved.
0883-5403/05/2004-2012$30.00/0
doi:10.1016/j.arth.2005.03.026
difficult, the joint line position should also be
determined preoperatively. Using these preopera-
tive planning techniques will help in preventing
postrevision instability and simplify the surgery.
Materials and Methods
The author performed 365 revision total knee
arthroplasties from 1987 to 2003. In most cases
(n = 275), the Natural Knee System (Centerpulse/
Zimmer, Warsaw, Ind) was used. Preoperative
planning consisted of a physical examination to
ascertain the status of the collateral ligaments and
therefore the amount of implant constraint re-
quired. Ipsilateral and contralateral radiographs
were used to determine the need for primary or
revision type of implants, the need for bone graft
or spacers, and the proper joint line position.
Primary femoral components without stems were
used in 35% of the cases. Primary femoral

37
38 The Journal of Arthroplasty Vol. 20 No. 4 Suppl. 2 June 2005
components with stems were used in 7% of the
cases. Revision-thickness femoral components
with stems were used in 58% of the cases.
Thirty-six (10%) knees required varus/valgus
constraint and 30 (8%) knees used rotating hinge
components. Most cases (82%) used uncon-
strained implants.
Results
Five (1%) revision knees had instability requir-
ing further surgery. Four cases had persistent
posterior cruciate ligament instability. In 2 cases,
the original revision was performed using stan-
dard congruent liners. They became stable after
re-revision with ultracongruent (deep dished)
tibial liners, which were not available at the time
of the initial revision. Two patients had persistent
posterior instability caused by hamstring spasticity
secondary to previous cerebral vascular accidents.
One was revised to a rotating hinge and the
other underwent an arthrodesis. The other case
had lateral collateral ligament insufficiency in
flexion only that was not recognized preopera-
tively. It was revised to a varus/valgus con-
strained implant.
Discussion
Revision total knee arthroplasty is a challenging
experience not only for surgeons but also for
surgical technicians, the operating room nurses,
and the implant representatives. Planning for a
revision knee arthroplasty is complicated because
there are many different implant options available
from which to choose. There are primary and
revision implants, unconstrained posterior cruciate
retaining and substituting implants, implants with
varus/valgus constraint, and rotating hinges. A goal
in revision surgery should be to use the least
constrained implants that will achieve stability.
Varus/valgus constrained and hinged implants
produce higher implant cement bone stresses that
may lead to higher radiolucencies and earlier
loosening [3]. Even if constrained implants are
used, knee stability cannot be guaranteed. If the
flexion/extension spaces are not balanced, a varus/
valgus constrained knee or rotating hinge can jump
the post and dislocate. However, choice of implants
with inadequate constraint in the absence of
collateral ligaments will result in an unstable knee.
The last thing surgeons want is to be in surgery and
not have what they need. However, if I ask my
implant representative to bbring everything just in
case,Q I will receive 23 pans of instruments and
24 bins with 347 boxes of implants and have chaos
for everybody involved. The advantage of preoper-
ative planning is that it will help surgeons deter-
mine what implants will be needed ahead of time
to simplify the surgery and avoid chaos.
One could theoretically use primary total knee
components in cases with minimal bone loss and
good bone quality. In most instances, a primary
component will not be stable with missing or soft
bone without the added stability of a stem. A
potential problem with the use of a primary
femoral component is that it may raise the joint
line [4]. With a proximally positioned joint line, if
the flexion space is appropriately tensioned, insta-
bility in extension will be present. Or the opposite,
if the extension space is appropriately tensioned,
the flexion space will be too tight and the knee will
have poor flexion. In most revision situations,
there is bone loss from the distal and posterior
aspect of the femur from component removal,
movement of a loose implant, or osteolysis. To
restore the joint line with distal bone loss, a
revision femoral component with either a fixed
distal buildup or a primary-thickness component
with spacers is required.
At knee revision surgery, scarring makes it
difficult to tell whether the posterior cruciate
ligament is intact or not. If a large angular deformity
needs to be corrected, the posterior cruciate liga-
ment usually needs to be resected to balance the
knee. If unconstrained implants are appropriate,
then posterior cruciate ligament–substituting
implants must be available for revisions. A deep-
dished ultracongruent liner was preferred because
its geometry provides posterior restraint through
the entire range of motion. A classic post-and-cam
posterior stabilizer was used in cases requiring a
major posterior capsular release because of the
static stabilization that this design provides.
It is most important preoperatively to determine
the status of the collateral ligaments with physical
examination and radiographs so that implants with
the appropriate constraint are made available. The
knee is examined with varus/valgus stress applied
not only in extension but also in flexion to
determine if the collateral ligaments are intact.
Next, the radiographs are reviewed for axial
malalignment, femoral component malrotation, or
joint line malposition. These can also be causes of
varus/valgus instability. If these are absent in the
presence of varus or valgus instability, then collat-
eral ligament insufficiency is present and additional
constraint is required. Radiographs will also show
the presence of bone loss. This will indicate the

Fig. 1. Correct femoral component size is determined
from the lateral radiograph.
need for spacers or bone graft. If there is a lateral
insufficiency, one needs to determine if it is caused
by lateral collateral ligament deficiency or by
iliotibial band deficiency. If there is lateral laxity
in flexion, there is lateral collateral insufficiency or
femoral component malrotation. If a lateral collat-
eral ligament is absent, the additional constraint of
a varus/valgus prosthesis is necessary and the
lateral collateral ligament is reconstructed with a
Fig. 2. The level of the joint line is measured up to a
common point such as the medial epicondylar ridge.
Preoperative Planning ! Kenneth A. Gustke 39
Fig. 3. Joint line measurements are transferred to the
radiograph of the knee being revised.
5-mm Dacron suture. If there is just an iliotibial
band deficiency, one option is to use unconstrained
implants and either brace the knee for 3 months or
preferably use a varus/valgus constrained implant.
A medial collateral ligament deficiency is more
difficult to manage. If a varus/valgus constrained
implant is used when there is no medial collateral
ligament present, the knee is going to eventually
fail from central stem failure, implant loosening, or
dislocation as the remaining medial soft tissues
stretch out. For elderly patients with a medial
collateral ligament totally absent, use of a rotating
hinge prosthesis is preferred. In young patients, the
medial collateral ligament is reconstructed with an
Fig. 4. Correct-sized femoral template is placed at the
proper joint line on the knee to be revised. The position
of the new distal femoral resection is determined
relative to the distal end of the femoral component to
be removed.
40 The Journal of Arthroplasty Vol. 20 No. 4 Suppl. 2 June 2005
Achilles tendon allograft and a varus/valgus im-
plant is used to provide an internal splint. If neither
of the collateral ligaments is present, there is no
choice other than using a rotating hinge.
A knee with unconstrained implants and intact
collateral ligaments can still be unstable if the
flexion/extension space is not balanced because of
an inappropriate joint line position. Joint line
malposition is a common error made in revision
surgery. It leads to instability in flexion or exten-
sion. It is difficult to determine the correct joint line
intraoperatively because the distal and posterior
femoral articular surfaces are not available to be
used as landmarks. Preoperative radiographic tem-
plating and measurements from anatomical points
on the contralateral radiograph will show where
the correct joint line position is relative to the
existing femoral component.
The correct joint line position is determined by
the femoral component position and size. The
correct femoral component size is determined from
the contralateral lateral knee radiograph (Fig. 1).
The level of the joint line is measured up to a
common point such as the medial epicondylar ridge
(Fig. 2). The joint line measurements are trans-
ferred to the radiograph of the knee being revised
(Fig. 3). The correct-sized femoral template is
placed at the proper joint line on the radiograph
of the knee to be revised (Fig. 4). The position of
the new distal femoral resection is determined
relative to the distal end of the femoral component
to be removed (Fig. 4). At surgery, the level of the
new distal femoral resection is marked on the
femur using this measurement from the distal end
of the existing femoral component. This will place
the new femoral component at the anatomical joint
line. With a properly positioned joint line, flexion/
extension space balance will be present in most
cases, reducing the number of times the distal
femur and the chamfer cuts are made to balance
the flexion and extension spaces.
Conclusion
Instability is a major issue in the failure of
revision total knee arthroplasty. Preoperative plan-
ning will help prevent this complication and allow
use of the least constrained implants. One should
not automatically use constrained implants just
because the procedure is a revision knee arthro-
plasty. In my series, they were required only 18% of
the time. Preoperative physical examination is most
important to determine the status of the collateral
ligaments. Varus/valgus constrained implants can
be used if lateral collateral ligament insufficiency is
present and in cases of medial collateral ligament
insufficiency as long as the medial collateral liga-
ment is reconstructed. If the medial collateral
ligament is absent and will not be reconstructed,
then use of a rotating hinge device is appropriate.
Preoperative radiographs are used to verify that
the cause of instability is from collateral ligament
insufficiency rather than from component mala-
lignment. The radiographs will show whether
primary components can be used or whether
revision components with stems, spacers, or bone
graft are required. Radiographic templating will
determine the appropriate femoral component size
and joint line position. Knowing this ahead of time
will simplify the surgery.
Taking time to do preoperative planning will
minimize having an excessive number of instru-
ments and implants at surgery that will only add to
its complexity. It should also shorten the operative
time. Excellent preoperative planning should result
in knee stability for patients with the least con-
straint, in a simpler surgery for the representative
and the operating room team, and most impor-
tantly in a simpler surgery for surgeons.
References
1. Peters CL, Hennessey R, Barden RM, et al. Revision
total knee arthroplasty with a cemented posterior-
stabilized or constrained condylar prosthesis; a mini-
mum 3-year and average 5-year follow-up study.
J Arthroplasty 1997;12:896.
2. Gustke K. Preoperative planning in revision total knee
replacement. Orthopedics 2002;25:975.
3. Hartford JM, Goodman SB, Schurman DJ, et al.
Complex primary and revision total knee arthroplasty
using the condylar constrained prosthesis. J Arthro-
plasty 1998;13:380.
4. Gustke KA. Planning and techniques for uncon-
strained revision total knee arthroplasty. In: Vince
KG, editor. Knee surgery. Baltimore, MD: Williams &
Wilkins; 1994. p. 1556.