Chapter 8

Three-Step Technique for Revision

Total Knee Arthroplasty
Kelly G. Vince and Daniel A. Oakes

INTRODUCTION
Revision knee arthroplasty surgery requires that order be restored
to the chaos of failure. Once the failed components, cement,
and useless weak bone have been removed from the knee, a gaping
hole confronts the surgeon. The problems of stability, mobility,
fixation, and the reconstruction of bone defects as well as
restoration of an anatomic joint line all cry out for attention
at once. There are undoubtedly a variety of approaches to the
revision knee surgery. One thing is certain—an organized
approach is essential or the reconstruction is doomed to failure
(Fig. 8.1).
This chapter proposes three steps to the reconstruction of any
knee regardless of the original cause of failure. The surgeon must
(1) reestablish the tibial platform, (2) stabilize the knee in flexion,
and (3) stabilize the knee in extension. These steps have been
described previously1–3 and are based upon the principles of knee
arthroplasty surgery that were developed for the total condylar
knee prosthesis by John Insall, Chit Ranawat, and Peter Walker at
The Hospital for Special Surgery in New York in the early 1970s.4,5
We have applied these concepts to revision knee surgery, expand-
ing them to address the rigors of the failed knee and establishing
an appropriate sequence. Faithful adherence to the proposed
sequence of steps, building one stage upon the other leads to a
successful revision knee arthroplasty (Table 8.1).
Although contemporary instruments have enabled every
surgeon to produce good primary knee arthroplasties, they rely on
bone for reference. This bone simply does not exist in the failed
knee. Consequently, instrument systems have not been reliable for
revision surgery. Missing bone, however, is not the greatest chal-
lenge facing the surgeon. More problematic are the soft tissues.
Working with strong concepts and trial components, the surgeon
will be able to understand the vagaries of lost, plastically deformed,
overly tight, and unreleased ligaments.
 8. THREE-STEP TECHNIQUE FOR REVISION TOTAL KNEE ARTHROPLASTY 105

FIGURE 8.1. Diagram of problems of revision surgery.

This chapter does not deal with the diagnosis of a failed knee
arthroplasty nor with the techniques for the removal of compo-
nents from a failed knee. It must be emphasized, however, that no
revision surgery should be attempted until an accurate mechanical
explanation for the failure has been established. Revision of the
inexplicably painful knee arthroplasty will yield miserable results.

Step 1 Establish Tibial Platform (Figure 8.2)

The tibia is a platform on which to rebuild the knee. The tibial
articular surface is involved with knee function, irrespective of
joint position. Whereas the distal femur bears load only in full
extension and the posterior femur only in flexion, the tibia is con-
stantly part of the articulation. The phrase tibial “platform” is
chosen purposefully. Do not be concerned about the articular
surface at this stage, that will come later.
The proximal tibia will have suffered any amount of insult from
the failed joint. Although good-quality host bone is respected, any
tibial cutting guide can be used to “square up” the surface by
removing obviously weak and dispensable tissue. Defects are iden-
tified at this point, not eliminated. Any bone cut made now must
not sacrifice good bone in an effort to eliminate a bone defect.
In many revision knee arthroplasties, medullary fixation will be
required to enhance fixation. If so, open the medullary canal and
confirm the measurements of endosteal diameter made on pre-
operative radiographs, using hand reamers. Bone should not be
 106

TABLE 8.1. Three-step revision knee arthroplasty

Step Goal Key
1 Establish tibial platform Tibia is common to flexion and extension gaps. It is a foundation to build on.
2 Stabilize knee in flexion Femoral component size and position stabilize the knee in flexion.
2(A) Sizing the femoral component Do not simply fit component to the residual bone.
K.G.VINCE AND D.A. OAKES

2(B) Femoral component rotation The component must not be internally rotated. Feel the residual posterior
condylar bone as a guide. Use posterior lateral augments to correct internal
rotation.
2(C) Joint line In general, a smaller femoral component leads to a higher joint line.
Decision 1 Gap mismatch Flexion gap is so large due to soft tissue failure that the knee cannot be stabilized
in flexion by the size of the femoral component. Need constrained component
or ligament reconstruction.
3 Stabilize knee in extension Seat the femoral component more proximally or distally to create an extension
gap that equals the flexion gap.
Decision 2 Varus-valgus instability The collateral ligaments are incompetent and either a constrained component
or a ligament reconstruction will be required.
 8. THREE-STEP TECHNIQUE FOR REVISION TOTAL KNEE ARTHROPLASTY 107

FIGURE 8.2. The tibial platform is reestablished.

removed from the tibial canal, which has a relatively thin

cortex. A suitable trial rod is selected and attached to a trial tibial
component. Once seated in the medullary canal, the tibial trial
defines the defective bone that will require re-construction. If
the intramedullary rod, fitted into the canal, is not parallel to the
long axis of the tibia, the rod may be too wide for the asymmetric
canal and either a narrower rod or one that is offset may be
required.
There are several complex classification systems for describing
bone defects at revision knee surgery. The simplest approach, in
the course of a demanding surgery, will be the most helpful. Defects
that are contained and have a rim of bone to hold bone graft, can
be filled with particulate graft, be it autograft from the knee,
ground up fresh-frozen bone, or freeze-dried allograft bone chips.
Noncontained defects, as seen when a tibial component has sub-
sided into varus, will most easily be dealt with by modular wedges
or blocks. Combined contained and noncontained defects exist and
respond well to a combined approach—the contained area is filled
with graft and the noncontained area is reconstructed with an
augment on top of the graft (Table 8.2).
Massive defects that offer virtually no host bone on which
to seat any of the component will usually require reconstruc-
tion with structural allograft. These unusual situations should
still be reconstructed following the three steps that are described
here.
108 K.G.VINCE AND D.A. OAKES

TABLE 8.2. A simple approach to bone defects

Bone defect Solution
Contained Particulate bone graft
Noncontained Modular wedge or block
Massive Structural allograft

Tibial defects can be reconstructed at this stage and the tibial

component even cemented into place to save time. This is because
the tibial platform does not affect how we reestablish alignment,
stability, and motion in the knee. We build the knee upon the tibia.
Tension and laxity in flexion and extension are manipulated with
the femoral component. Nonetheless, in the interest of keeping
most of our options open, it is best to leave the trial tibia in place,
noting the type of bone defect if any and how we plan to recon-
struct it when we implant the final components.

Step 2 Stabilize Knee in Flexion (Figure 8.3)

(A) Choose the Size of the Femoral Component That Stabilizes

the Knee in Flexion
Choose the size of the femoral component that stabilizes the knee
in flexion. It is a common and deadly error to measure existing
bone and simply fit the corresponding femoral component to it. In
almost every case, this will lead to the selection of a femoral com-
ponent that is too small and an arthroplasty that is unstable in
flexion or one in which excessive distal femoral bone must be
resected to accommodate an unduly thick articular polyethylene.
Undue resection of distal femur results in an unacceptable proxi-
mal migration of the joint line.
Ignore the residual bone on the distal femur in this step and
visualize the normal bone that was present before any surgery had
been performed. Use the size of the failed component, and lateral
radiographs of the contralateral knee, if unoperated, to estimate
the size of the revision femoral component. The final choice of
revision femoral component size will depend upon the an-
teroposterior dimension that is necessary to stabilize the knee in
flexion. The revision femoral component size will be determined
not by residual bone, so much as by the soft tissues, specifically
the collateral ligaments.
Stability in flexion is determined not only by the size, but also
by the anteroposterior location of the femoral component. Unless
 8. THREE-STEP TECHNIQUE FOR REVISION TOTAL KNEE ARTHROPLASTY 109

the original component was oversized, leaving good posterior

condyles for the revision component, fixation will be compromised
because bone has been lost from the posterior condyles as a result
of the failed knee or the removal of components. That is the
purpose of posterior femoral augments. They exist to fill in bone
defects and consequently to stabilize the knee in flexion by enabling
the surgeon to select an appropriately large femoral component.

B
FIGURE 8.3. The femoral component is sized against the one removed. (A)
If the knee was loose in flexion, a large femoral component is selected; or
(B) if the original implant was sized correctly, a comparable revision
femoral component is chosen.
110 K.G.VINCE AND D.A. OAKES

Without them, we would be forced to use components that were

too small.
In the presence of defective bone, due either to defects or soft
quality, enhance fixation with medullary stems. These will influ-
ence the position of the femoral component and accordingly the
stability of the knee in flexion. Stems can create problems. If large
canal filling stems are selected, there will be little latitude for
adjustment of the component position.6 The component may be
positioned in greater varus or valgus, flexion or extension or trans-
lated anteriorly or posteriorly, depending on the morphology of the
femur. The position of smaller stems that are cemented in the canal
(despite the undesirability of methacrylate in the canal) can be
manipulated, anterior and posterior to affect the size of the flexion
gap.
One situation that may arise when trying to determine the size
of the femoral component is a gap mismatch. This important (and
unusual) circumstance must be identified in any revision. Simply
stated, a knee with an irreconcilable gap mismatch has a capacious
flexion gap that, because of soft tissue failure, cannot be balanced
to the likely dimensions of the extension gap with conventional
releases or selection of the correct femoral component size. When
the collateral ligaments, in particular the medial collateral, have
stretched, it seems that we cannot find a femoral component large
enough to stabilize the knee in flexion without an unacceptably
thick polyethylene. The necessary femoral component may be so
large that it no longer fits the medial to lateral dimensions of the
bone. We have a knee that cannot be stabilized in flexion simply
by recreating the anteroposterior dimensions of the femur.
The gap mismatch marks a decision point in the revision
requiring a choice between accepting the laxity in flexion and pro-
tecting the patient with a constrained component or advancing the
collateral ligament on the femur.7 Our preference has been to avoid
linked, constrained devices (hinges) in all cases, and to even recon-
struct ligaments and use a nonlinked constrained device. With this
decision noted, the femoral component size and position estab-
lished and the tibial insert selected, the difficult work of the arthro-
plasty is complete.

(B) Seat the Femoral Component in External Rotation

The femoral component must be correctly rotated in the femoral
canal. Internal rotation leads to patellar maltracking and a host
of extensor mechanism problems. What landmarks exist for
the correct rotation of the revision femur? There are two: the
 8. THREE-STEP TECHNIQUE FOR REVISION TOTAL KNEE ARTHROPLASTY 111

epicondylar axis and the height of residual posterior condylar

bone.
The epicondylar axis, an imaginary line joining the medial and
lateral epicondyles (where the collateral ligaments attach to the
femur) lies in variable amounts of external rotation. It defines the
attachment of the collateral ligaments and accordingly their
functional length. The residual posterior condylar bone, hidden
in the back of the knee, is another reliable guide. Though not
visible, it is palpable. With the knee flexed to 90 degrees, one can
feel the residual bone above the posterior condyles, by running a
finger up onto the posterior femoral cortex (Fig. 8.4). If there is
much more bone left on the medial side as compared to the lateral
side, we know that the failed femoral component had been
implanted in internal rotation. The converse implies external
rotation.
Again, do not be fooled by the residual bone and seat a revision
femoral component in internal rotation. Defective bone should be
reconstructed with augments. Use posterior augments preferen-
tially on the lateral side to correct internal rotation.

(C) Reestablish the Joint Line

We have seen that stability in flexion is determined by the revision
femoral component size and position. To fully stabilize the knee

FIGURE 8.4. Rotational position of the femoral component can be deter-

mined by palpation of the posterior femoral condyles.
112 K.G.VINCE AND D.A. OAKES

requires a polyethylene tibial insert so that the combined height of

the posterior condyles of the femoral component and the tibial
insert match the dimensions of the flexion gap. The point at which
the femoral component meets the tibial polyethylene is the joint
line. Where then does the patella lie? (Fig. 8.5)
The challenge for the surgeon is to match the prosthetic
joint line height as closely as possible to the anatomic joint line.
What is the best remaining landmark for approximating the
anatomic joint line? The location of the inferior pole of the patella,
when the knee is flexed to 90 degrees is an easily identified and reli-
able indicator of desired joint line. Ideally, the joint line should lie
distal to the inferior pole of the patella. In choosing between two
femoral component sizes, both of which stabilize the knee in
flexion, but each of which requires different thicknesses of tibial
polyethylene, select the combination that gives the best patellar
height.

Step 3 Seat Femoral Component to Stabilize Knee in Extension

(Figure 8.6)
This part is easy. The femoral component must be seated on the
distal femur so that there is neither recurvatum nor a flexion con-
tracture. If the trial components result in recurvatum, the femoral
component may be seated more distally by using distal femoral
augments. This will be the case in the majority of revisions in which

FIGURE 8.5. The point at which the femoral component meets the tibial
articular surface is the joint line. The patella height is then noted.
 8. THREE-STEP TECHNIQUE FOR REVISION TOTAL KNEE ARTHROPLASTY 113

FIGURE 8.6. With the provisional components in place the knee is brought
to full extension.

bone is missing as a result of failed primary. Selecting a thicker

polyethylene tibial insert instead of a distal femoral augment will
unbalance the stability that had been achieved in Step 2, in which
the knee was stabilized in flexion.
Rarely, the surgeon may resect additional distal femur to sta-
bilize the knee in extension. This may occur for the knee that had
failed with a fixed flexion contracture, especially if the joint line
had been lowered during the primary arthroplasty. When distal
femoral resection is contemplated, check that it is not for the
purpose of accommodating an inordinately thick tibial insert that
is going to result in proximal joint line migration. This could be a
gap mismatch.

DECISION POINT: SOFT TISSUE BALANCE IMPOSSIBLE?

In trying to stabilize the knee in extension it may become appar-
ent that one or both of the collateral ligaments is deficient. A failed
medial collateral, producing valgus instability, is the most dis-
abling. When the medial collateral has suffered true plastic failure,
it will not be possible to stabilize the knee by releasing the lateral
side. Despite extensive lateral collateral releases, the medial liga-
ment remains lax. All further releases simply lengthen the lateral
side, and increasingly thick articular polyethylene creates a flexion
contracture because the posterior structures are intact.8 This is a
decision point. The revision cannot be left without a functional
114 K.G.VINCE AND D.A. OAKES

FIGURE 8.7. The revision arthroplasty should be stable in (A) flexion and
(B) extension.
 8. THREE-STEP TECHNIQUE FOR REVISION TOTAL KNEE ARTHROPLASTY 115

MCL and either a constrained implant or a ligament reconstruc-

tion (or both) will be required.

CONCLUSION
Having (1) reestablished the tibial platform, (2) stabilized the knee
in flexion, and (3) stabilized the knee in extension, the revision
arthroplasty is effectively complete. The trial components can be
removed and the bone prepared for implantation of the permanent
components. The three steps lend themselves to whatever implant
is planned for the revision. Although posterior-stabilized implants
generally provide a higher degree of stability for the revision, these
steps can lead to a sound reconstruction when cruciate-retaining
implants are selected. As has been indicated by the “decision
points,” circumstances arise when the pathology of the deformity
dictates the best choice of implant.
The three steps to revision knee arthroplasty presented here
provide the surgeon with an orderly approach based on sound sur-
gical principles. Meticulous preoperative planning and adherence
to the steps should allow the knee surgeon to overcome the daunt-
ing challenge of the revision knee arthroplasty (Fig. 8.7).

References
1. Vince KG. Revision Knee Arthroplasty Instructional Course Lectures of
the AAOS. Mosby; 1992.
2. Vince K. Revision knee arthroplasty. In: Chapman, ed. Operative Ortho-
pedics. Philadelphia: JB Lippincott; 1993:1981–2010.
3. Vince K. Planning revision total knee arthroplasty. Seminars in Arthro-
plasty; 1996.
4. Insall JN. Total knee replacement. In: Insall JN, ed. Surgery of the Knee.
New York: Churchill Livingstone; 1984:587–696.
5. Vince KG, Insall JN. The total condylar knee arthroplasty. In: Laskin, ed.
Total Knee Arthroplasty. New York: Springer Verlag; 1991.
6. Vince KG, Long W. Revision knee arthroplasty. The limits of press fit
medullary fixation. Clin Orthop. August 1995; (317):172–177.
7. Vince K, Berkowitz R, Spitzer A. Ligament Reconstructions in Difficult
Primary and Revision TKR. Accepted for presentation at the Annual
Meeting of the Knee Society. San Francisco, California, February 1997.
8. Vince KG. Limb length discrepancy after revision total knee arthroplasty.
Techniques in Orthop. 1988; 3:35–43.