The patient demographics, indications for the selection of this component,
surgical protocol and technique, and clinical and radiographic evaluation
techniques were described in the previous
report1. The
standard system of terminology for reporting of results described by Johnston
et al.2 was
used.
Radiographic Evaluation
Twenty-five living patients (twenty-six hips) had had radiographs made at a
minimum of two years postoperatively. The average duration of radiographic
follow-up was 8.2 years (range, 2.0 to 13.2 years) in this group
(Fig. 2). Of the thirty
patients (thirty hips) who had died, fifteen had been followed
radiographically for a minimum of two years (average, 4.7 years; range, 2.0 to
8.8 years). Nine of the patients who died had not been followed clinically for
a minimum of two years. The radiographic criteria for loosening and osteolysis
in this study were the same as those used in the previous
study1,3-9.
Survivorship Analysis
Kaplan-Meier survivorship analysis was performed with four end points: (1)
revision because of recurrent dislocation or mechanical failure of the
constraining mechanism, (2) revision of a component for any reason, (3)
revision because of aseptic loosening of the acetabular component, and (4)
revision because of aseptic loosening of the femoral component.
The average age at the time of surgery was seventy-one years (range,
thirty-one to ninety-two years). At the time of follow-up for the present
study, twenty-five patients (twenty-six hips) were alive, thirty patients
(thirty hips) had died, and no patient had been lost to follow-up. (The
patient who had been lost to follow-up in the previous study was located
through an Internet search.) Hence, clinical follow-up information was
obtained for all fifty-six hips. The living patients were followed clinically
for an average of 10.4 years (range, 7.0 to 13.2 years), and the patients who
died had been followed for an average of 4.6 years (range, 0.1 to 11.1
years).
Redislocation and Failures of the Constraining Mechanism
At the time of the latest follow-up, four (7%) of the fifty-six hips had
had a subsequent dislocation or failure of the component. In one of those
patients, the constrained component had been cemented somewhat proud into a
well-fixed cementless metal-backed acetabular component. Four months after
insertion, the liner dislodged from the cement. A second, smaller, constrained
liner was inserted with cement, and the patient had no additional dislocations
before his death ten months later.
The second patient with a failure was a seventy-seven-year-old man in whom
a constrained acetabular component had been snapped into a previously placed
well-fixed Osteonics shell. The patient had had four dislocations prior to the
revision with the constrained acetabular component. The patient's son reported
that, before the patient's death 11.1 years after the index procedure, the
patient had had a dislocation that had been treated with closed reduction. No
records or radiographs of this episode could be located, and radiographs made
subsequent to the alleged dislocation showed no evidence of component failure
at any of the interfaces. We consider the report of this redislocation to be
questionable but mention it for completeness.
The third patient with a failure was a sixty-six-year-old woman who had had
six previous operations on the hip prior to the revision with the constrained
acetabular component. In one of the previous procedures, an acetabular
component was inserted without cement and a bulk allograft was used to augment
the acetabulum. The patient had sustained six dislocations prior to the
revision with the constrained acetabular liner. Fifty-three months after the
index procedure, the entire acetabular component (the metal backing and the
constrained liner) dislodged from the osseous acetabulum. Half of the metal
backing had been placed against allograft bone. Another constrained acetabular
component was inserted without cement. Forty months later, the patient had had
no additional dislocations, but the component had loosened; it was revised
with use of another constrained component, which was cemented into an
acetabular cage.
The fourth patient with a failure was a thirty-one-year-old woman who had
had nine previous hip operations and five previous hip dislocations. At our
revision, she received a constrained uncemented acetabular component and a
cemented femoral component. Ninety months after the index procedure, the
constrained component failed by separation of the ring that captures the
bipolar head within the polyethylene liner and by dislodgment of the bipolar
head from the larger liner. This was thought to be related in part to
impingement of the femoral neck against the 20° elevated rim of the
polyethylene liner. The patient was treated with a new constrained liner that
was snapped into the well-fixed uncemented acetabular shell. Four months
later, a deep infection developed and was treated with multiple
débridements and ultimately a resection arthroplasty.
Aseptic Loosening of the Acetabular Component Requiring Revision
Two hips (4%) required revision of the acetabular component because of
aseptic loosening at 2.7 and 8.8 years after the index procedure. One other
acetabular component was seen to be loose radiographically, at 10.5 years
after the index procedure. At the time of writing, this patient was awaiting
resection arthroplasty because of acetabular loosening secondary to severe
acetabular insufficiency and a periprosthetic acetabular insufficiency
fracture.
Aseptic Loosening of the Femoral Component Requiring Revision
Three hips (5%) required revision of the femoral component because of
aseptic loosening at an average of 3.8 years (range, 1.1 to 7.4 years) after
the index procedure. No other femoral component was seen to be loose
radiographically.
Osteolysis of the Pelvis and Femur
Two hips demonstrated acetabular osteolysis. One was revised with another
constrained liner and shell at 6.5 years following the index procedure. The
original implant was bone-ingrown at the time of the revision and had been
radiographically stable five years following the index procedure. The second
case of acetabular osteolysis was in the patient described above who was
awaiting a resection arthroplasty at the time of writing. Osteolysis was not
the primary mechanism of failure in this patient. No patient had substantial
femoral osteolysis.
Other Reoperations
In the previous study, seven hips underwent a reoperation during the
follow-up period for reasons other than redislocation or instability. In the
subsequent five years, an additional seven hips (for a total of fourteen, or
25%) required a reoperation for reasons other than redislocation or
instability. Of these fourteen hips, five required a reoperation because of
infection (two were treated with débridement only, and three required
removal of all implants), three required revision of the femoral component
because of aseptic loosening, two required revision of the acetabular
component because of aseptic loosening, two required internal fixation of a
periprosthetic femoral fracture distal to the stem, one required repair of an
intraoperative laceration of the femoral nerve, and one required prophylactic
revision of the acetabular component because of severe acetabular
osteolysis.
Clinical Results
Of the twenty-six hips in the twenty-five living patients evaluated at an
average of 10.4 years, fourteen (54%) were not painful and two (8%) were
severely painful. Twelve patients (twelve hips) had no limp, and five patients
(six hips) were unable to walk.
Survivorship Analysis
Survivorship curves derived with the
Kaplan-Meier10
method are presented in Figures
3-A,
3-B,
3-C, and
3-D. The survival rate at
twelve years was 90% with revision because of recurrent dislocation or
mechanical failure of the constraining mechanism as the end point, 68% with
revision of a component for any reason as the end point, 92% with revision
because of aseptic loosening of the acetabular component as the end point, and
92% with revision because of aseptic loosening of the femoral component as the
end point.
Revision with this tripolar type of constrained acetabular component for
treatment of recurrent dislocations prevented additional dislocation or
component failure in all but four hips. Two of the failures occurred in the
first five years, and the other two occurred between five years and the time
of the longer (average, ten-year) follow-up. The 4% and 5% rates of revision
because of aseptic loosening of the acetabular and femoral components,
respectively, are lower than those reported for revisions done with
cement11,12
and similar or only slightly higher than those reported for revisions done
without
cement13-16.
No additional femoral components and one additional acetabular component were
loose according to radiographic criteria. Pelvic osteolysis occurred in two
hips (4%); one of those hips was revised, and revision was pending for the
other at the time of writing. The rate of osteolysis in this series is only
slightly higher than that reported in other series of cementless acetabular
revisions14,16.
The overall rate of implant revision or removal of 21% (twelve hips) is high.
However, when one considers that these components were utilized in the worst
cases of hip dislocation following total hip replacement in two large complex
hip practices, these results are better than those in large series in which
hip dislocations were treated with multiple
modalities17-23.
We continue to use and recommend the use of this device in patients with
hip dislocation following total hip replacement when other modalities do not
or probably will not provide adequate stability (such as in patients with
abductor deficiency, excessive femoral bone loss requiring a tumor prosthesis
or femoral allografts, multiple previous operations on the hip, or dementia)
and in low-demand patients. The occasional implant-related failures of this
device as well as this extensive experience have led us to make five important
recommendations: (1) if the liner is cemented into a secure shell it should
not be left proud, and the backside of the liner should be scored to achieve
better cement fixation; (2) the liner should not be inserted into a grossly
malpositioned shell, and the elevated rim of the liner should not be
positioned where impingement might occur; (3) a constrained shell and liner
should not be used when <50% of the shell is against host bone, unless
additional fixation such as a cage or cementless fixation with multiple screw
augmentation into the host bone is used; (4) as many screws as possible should
be used to optimally secure the cementless acetabular shell; and (5) all
patients should use a brace for a minimum of six weeks postoperatively.
?