Afifty-five-year-old woman had had recurrent patellar dislocation during
childhood and had been treated with multiple surgical procedures, including
bilateral patellectomy. At fifty years of age, the patient had undergone a
left total knee arthroplasty after the development of bilateral tibiofemoral
arthrosis. One year later, a spontaneous rupture of the quadriceps tendon
occurred and was treated with allograft tendon reconstruction of the extensor
mechanism. Two years after the total knee arthroplasty and one year after the
reconstruction of the extensor mechanism, the reconstruction failed and the
patient underwent a revision total knee arthroplasty with use of a constrained
condylar Insall-Burstein-II implant (Zimmer, Warsaw, Indiana) and a 21-mm
ultra-high molecular weight polyethylene insert that had been sterilized with
inert-gas gamma irradiation. The extensor mechanism was reconstructed with a
whole patellar tendon allograft. Postoperatively, the patient was able to walk
independently with the use of a cane and had no pain. The quadriceps strength
was graded 3 of 5. Three years later, she began to experience episodes of
giving-way and popping of the knee. She described the episodes as a
"total dislocation" of the knee. Initially, she was able to reduce
the knee voluntarily; eventually, however, with progressive worsening of pain
and instability, she had to mechanically reposition the knee before being able
to walk.
The patient was referred to our institution. The physical examination
revealed that the passive range of motion was from 10° of hyperextension
to 100° of flexion, with a 45° extensor lag. The knee was grossly
unstable to both varus and valgus stress. Radiographs demonstrated anterior
displacement of the central metal post in the ultra-high molecular weight
polyethylene insert (Fig. 1).
The patient was unable to walk, so an analysis of gait was not possible. She
also had degenerative spondylolisthesis at L4-L5 with severe spinal stenosis
but no associated lower-extremity weakness.
She was treated with revision total knee arthroplasty and reconstruction of
the extensor mechanism. Intraoperative findings demonstrated dissociation of
the tibial insert. Deformation was present along the anteromedial and
anterolateral portions of the tibial post, which was consistent with anterior
impingement of the femoral component due to hyperextension of the knee
(Fig. 2). The Insall-Burstein
locking mechanism consists of anterior and posterior parallel ridges that were
designed to overhang and lock similar reverse ridges on the ultra-high
molecular weight polyethylene insert. The posterior ridge of the polyethylene
insert was completely absent, a finding that is consistent with abrasive
and/or fatigue wear mechanisms (Fig.
3). The allograft patella was fragmented and attenuated. The knee
was revised to a rotating-hinge prosthesis.
Although the knee had no skin defect that required muscle coverage, a
medial gastrocnemius muscle flap was used to increase the strength of the
extensor mechanism in the manner described by Rhomberg et
al.3. The distal
tendinous portion of the medial aspect of the gastrocnemius was harvested and
was passed through a skin tunnel to the anterior portion of the proximal
aspect of the tibia. The muscular border of the gastrocnemius was sutured to
the tibial periosteum, and the tendinous portion was sutured to the quadriceps
tendon.
Six months after the revision, the range of motion was 0° to 125°,
with a 30° extensor lag. The patient had no knee pain and was walking with
the aid of two canes. During the stance phase of gait, she kept the left knee
fully extended and the left hip flexed. During stance, the center of the trunk
was positioned anterior to the knee. Her ability to walk was limited by back
pain and bilateral weakness of the lower limbs, which were worsening symptoms
of spinal stenosis, and she was scheduled to undergo spinal decompression.
Dissociation of the liner from a metal backing is a well-recognized failure
mechanism of modular acetabular
components4-6.
Torque is applied to the liner during movement at the bearing surface, which
tends to rotate the liner within the metal shell. The forces at the
liner-shell interface can cause tensile stresses at the liner-locking
mechanism that can result in mechanical failure. However, tibial inserts used
in total knee arthroplasty are primarily loaded in compression. Dissociation
of the tibial insert from the baseplate is rarely reported as a complication,
probably because of the relative absence of tensile stresses on the
insert-baseplate locking mechanism.
Hyperextension of the knee may be necessary during gait to provide
stability during the stance phase, a time at which there is quadriceps
insufficiency. Deformation of the anterior portion of the post can result from
hyperextension of the knee. In our patient, hyperextension of the knee
positioned the center of the knee posterior to the center of the upper body
during stance. The patient walked by maintaining the center of mass of the
upper body anterior to the knee. During stance, the knee was extended and the
hip was flexed. When leaning forward, she required two canes to support the
upper body.
My patient had a nonfunctioning extensor mechanism and anterior impingement
of the tibial post. It is likely that anterior impingement during
hyperextension of the knee contributed to posterior lift-off of the insert.
Anterior impingement during hyperextension of the knee would be expected to
result in separation or lift-off of the posterior femoral condyles from the
tibial plateau, rather than posterior lift-off of the tibial insert. However,
with an insert such as the one used in our patient, which is constrained to
varus, valgus, and rotation stresses, posterior lift-off forces could be
transferred to the tibial insert when rotational, varus, or valgus motion
occurs during hyperextension of the knee.
The wear resistance and fatigue strength of ultra-high molecular weight
polyethylene are reduced by gamma irradiation in
air7,8.
However, the implant used in my patient was gamma irradiated in an inert
atmosphere and had been in situ for only three years, so I do not believe that
inferior wear resistance of the insert material was a factor in the
failure.
This case report demonstrates that mechanical failure of the ultra-high
molecular weight polyethylene posterior locking mechanism can result when a
fully constrained posterior stabilized implant is used in a limb with
quadriceps insufficiency.