Between December 1984 and December 1989, 174 total hip arthroplasties
were performed by the senior author (P.F.L.) with use of a single
design of porous-coated acetabular component (Harris-Galante I;
Zimmer, Warsaw, Indiana). Seventy-one arthroplasties were performed
in fifty-six patients who were less than fifty years of age. Five
patients (seven hips) who had a well-functioning prosthesis died
before a minimum duration of follow-up of nine years, and five patients
(six hips) were lost to follow-up. One patient (one hip), in whom
both well-fixed components were revised at another hospital because
of pain, is included in the survivorship analysis only. An additional
patient, who initially had been lost to follow-up but returned because
of a late infection at thirteen years, was excluded. Thus, the present
study included fifty-six hips in forty-four patients who were available
for a clinical and radiographic review after a mean duration of
follow-up of eleven years (range, nine to fourteen years).
The study group included twenty-seven women (thirty-six hips)
and seventeen men (twenty hips). Twelve patients had had a bilateral
total hip arthroplasty; of these, eight had had a two-stage procedure
and four, a one-stage procedure. The mean age of the patients at
the time of the procedure was thirty-seven years (range, twenty-two
to forty-nine years). The mean weight of the patients was 70 kg
(range, 47 to 98 kg). The preoperative diagnosis was osteonecrosis
in eighteen hips (32%), osteoarthritis in fourteen hips (25%), juvenile rheumatoid
arthritis in thirteen hips (23%), adult-onset rheumatoid arthritis
in six hips (11%), posttraumatic arthritis in two hips (4%), and
another diagnosis in three hips (5%).
The Harris-Galante-I porous-coated acetabular component is a
hemispherical shell fabricated of titanium alloy, with a titanium
fiber-metal coating (
Fig. 1
). The component has multiple holes to allow supplemental fixation
with either 4.5 or 5.1-mm titanium-alloy screws. The modular polyethylene
liner (sterilized with gamma irradiation in air) is secured with
three pairs of metal tines that are located at the rim of the prosthesis.
We routinely bent these tines inward before insertion of the liner
to minimize motion between the liner and the shell.
A posterior approach was used in fifty-three hips, and a transtrochanteric
approach was used in three. The acetabulum was prepared with hemispherical
reamers to obtain bleeding cancellous or subchondral bone. The outer
diameter of the acetabular component that was implanted was the
same diameter as the final reamer used. However, the final reamer
was inserted only briefly to the depth of the acetabular rim so
that a so-called press-fit was usually obtained. Additional fixation was
obtained with multiple 4.5 or 5.1-mm titanium-alloy screws. A mean
of four screws (range, three to six screws) were usually directed
into the superior and posterior aspects of the pelvis. A screw was
routinely inserted in the polar hole and was directed toward the
iliopectineal line of the pelvis. Morselized cancellous autograft
was used to fill the empty screw-holes and any acetabular cysts.
A polyethylene liner without an elevated rim was used in all hips.
Fifty-five hips received a Harris-Galante femoral component inserted
without cement, and one received a CDH Precoat femoral component
(Zimmer) inserted with cement. A 28-mm modular chromium-cobalt femoral
head was used in fifty-two hips, a 32-mm head was used in three
hips (all of which received either a 56 or 58-mm acetabular component),
and a 22-mm head was used in one hip. The head size was selected with
the goal of providing a polyethylene thickness of >6 mm. The mean
thickness of the polyethylene liner was 7.8 mm (range, 5.3 to 12.3
mm).
The forty-nine hips that did not have a revision or reoperation were
evaluated with the Harris hip score at each clinical visit. All
hips were evaluated with standardized anteroposterior radiographs
of the pelvis, made with the patient supine and the x-ray beam centered
over the pubis, at six weeks, three months, six months, and yearly
thereafter. The six-week, five-year, and most recent radiographs
were used for analysis in the present study. The angle of abduction
of the acetabular component was measured on the anteroposterior
radiograph and was defined as the angle formed by a line drawn perpendicular
to the edges of the rim of the component and a line drawn between
the bottom of the obturator foramina. Radiolucent lines in the three
zones described by DeLee and Charnley
23
were evaluated on the anteroposterior radiograph. Oblique radiographs
were not made. Pelvic osteolysis was defined as a circular or oval
area of distinct bone loss. Vertical and horizontal migration was
evaluated with the method described by Massin et al.
24
. An acetabular component was considered loose if there was >3 mm
of migration or a change of at least 4° in the angle of
abduction
14
. The thickness of the polyethylene liner was assessed with an electronic
caliper (Ultra-Cal Mark III; Fowler, Newton, Massachusetts) with
use of the Livermore method as modified for uncemented acetabular
components by Latimer and Lachiewicz
14
. The minimum polyethylene thickness was defined as the shortest
distance from the outer edge of the prosthetic femoral head to the
outer edge of the metal acetabular shell. The measured size of the
femoral head was used to correct for magnification. The polyethylene
thickness for the forty-nine unrevised hips was measured at six
weeks, at five years, and at the time of the most recent follow-up
examination. The polyethylene thickness for the seven hips that
were revised or had liner replacement was measured at six weeks,
at five years, and at the time of the most recent follow-up examination before
the revision.
Clinical Analysis
Six patients who had revision of the femoral component (including
one patient with pelvic osteolysis) and one patient who had revision
of the acetabular component because of asymptomatic pelvic osteolysis
were excluded from the following detailed clinical analysis, leaving
forty-nine hips in thirty-nine patients available for review.
The overall Harris hip rating was excellent for thirty-five hips (71%),
good for nine (18%), fair for two (4%), and poor for three (6%).
The three poor results were due to juvenile rheumatoid arthritis
with multiple joint involvement, functional deficits secondary to
a head injury, and loosening of the femoral component as a result
of osteonecrosis in one patient each. The patient with the loose
femoral component had refused a revision. The specific results with
regard to pain, limp, use of support, and walking distance are presented
in
Table I
.
Reoperations
Eight additional procedures were performed in seven patients after
the index procedure. Six patients had revision of an uncemented
femoral component because of aseptic loosening (five patients) or
osteolysis (one patient). Two patients, including one of the two
who underwent femoral revision because of osteolysis, had bone-grafting
and liner exchange because of asymptomatic pelvic osteolysis. Both
of these patients had osteonecrosis in association with sickle-cell
disease, and both had severe osteolysis of the ischium and the posterior
acetabular wall. Although both patients were asymptomatic, a reoperation
was performed (at six and one-half and nine years) because of concern
about the size of the lesions. In both patients, the acetabular
component was well fixed and was not revised. In one of these patients,
the well-fixed femoral component was removed so that proximal and
distal femoral osteolytic lesions could be debrided and filled with crushed
cancellous allograft. Both patients had an exchange of the polyethylene
liner, and one had insertion of a smaller femoral head to allow
for the use of a thicker liner. At the time of the most recent follow-up
examination (performed five and seven years after the reoperation),
there was partial healing of the lesions and both patients had an
excellent clinical result.
All six hips that underwent revision of the uncemented femoral
component were found to have a well-fixed acetabular component at
the time of the femoral revision. All of these hips had an incidental
exchange of the polyethylene liner and were therefore excluded from
the radiographic analysis of polyethylene wear after the revision
had been performed. The femoral revision was performed at six and
one-half years in one hip, at seven years in one hip, at seven and
one-half years in two hips, at nine years in one hip, and at ten
years in one hip.
One patient who initially had been lost to follow-up returned after
thirteen years because of a late hematogenous infection and loosening
of the femoral component. At the time of d�bridement, the loose
femoral component was removed but the well-fixed acetabular component
was left intact. There had been no recurrence of infection at one
year after the removal of the femoral component. Reimplantation
was not planned because of multiple medical problems.
At the time of the present report, no acetabular component had
been revised because of aseptic loosening and no revisions were
pending. One patient had had removal of both the acetabular and
the femoral component at another hospital at 6.5 years because of
pain, but the surgeon reported to us that both components were well
fixed.
Radiographic Analysis
Radiographs were analyzed at the six-week, five-year, and most
recent follow-up visits. For hips that had had a reoperation, the
most recent radiograph that was reviewed was the one made immediately
before the revision of the femoral component and exchange of the
polyethylene liner. Fifty-six hips had a complete set of radiographs.
No hip had evidence of a broken screw or liner dissociation, a notable
change in the abduction angle, or horizontal or vertical migration
of the acetabular component at the time of the latest follow-up.
Pelvic osteolysis was seen in thirteen hips (23%). Eleven lesions
were located in the ischial region (
Fig. 2
), and two were located in the superior aspect of the ilium, adjacent
to the acetabular screws. As mentioned previously, two patients
had a reoperation consisting of liner exchange and bone-grafting.
The remaining eleven hips were asymptomatic at the time of the latest
follow up and continued to be followed on an annual basis. A nonprogressive
radiolucent line was seen in one acetabular zone in ten hips (18%)
and in two acetabular zones in six hips (11%). No hip had a radiolucent
line in all three zones.
The mean rate of polyethylene wear (and standard deviation) for
the fifty-six hips was 0.15 ± 0.10 mm/yr (range, 0.02
to 0.59 mm/yr). The mean amount of polyethylene wear was 1.47 ±
0.89 mm (range, 0.25 to 3.57 mm). Twenty-two hips had <1 mm
of wear, eighteen had 1 to 2 mm of wear, twelve had >2 to 3 mm of
wear, and four had >3 mm of wear. The wear rate was also analyzed
at three time-intervals: (1) between the initial follow-up and the
five-year follow-up, (2) between the five-year follow-up and the
most recent follow-up, and (3) between the initial follow-up and
the most recent follow-up. No significant differences were noted between
these intervals (p > 0.05). However, logistic regression analysis
showed that wear was positively correlated with the age of the patient
at the time of the arthroplasty (with patients less than thirty-eight
years of age having more wear) and the presence of an excellent
hip score. A repeated-measures analysis with Greenhouse-Geisser
adjustment showed that an excellent hip score (p = 0.004) and a
younger age (p = 0.026) were significantly associated with an increased
rate of polyethylene wear. No association was detected between the wear
rate and the variables of gender (p = 0.99), femoral neck length
(p = 0.66), liner thickness (p = 0.11), abduction angle (p = 0.07),
or weight (p = 0.39).
Kaplan-Meier survivorship analysis was performed for the entire
series of seventy-one hips in patients who were less than fifty
years old, with failure defined as removal of the acetabular metal
shell for any reason. The survival rate at ten years was 98% (95%
confidence interval, 96.9% to 99.9%) (
Fig. 3
).
Since the introduction of total hip arthroplasty in the late 1960s,
the performance of this procedure in younger patients has been controversial.
Caution has been advised in the treatment of young patients because
high rates of failure have been reported
4,25
. The presumed association between younger age and greater activity
levels has raised concerns that total hip arthroplasty may be associated
with a higher rate of mechanical failure and bone destruction in
younger patients than in older patients.
Many investigators have reported high rates of radiographic loosening
and revision of cemented acetabular components in patients with
a variety of diagnoses
1-10
. Barrack et al.
2
, in a study of fifty hips that had been treated with so-called second-generation
cementing techniques, reported that twenty-two hips (44%) had radiographic
signs of loosening of the acetabular cup and eleven hips (22%) had
had revision of the cup after a mean duration of follow-up of twelve
years. Ballard et al.
1
reported that "improved" techniques did not decrease the rate of
loosening of cemented acetabular components in young patients, as
indicated by an acetabular revision rate of 24% (ten of forty-two)
and overall acetabular failure rate of 36% (fifteen of forty-two)
after a minimum duration of follow-up of ten years. Collis
5
reported that the expected rate of survival of cemented acetabular
components in patients younger than fifty years old was only 69%
at fifteen years. Cornell and Ranawat
26
reported that the survival of cemented acetabular components in
active patients who were less than fifty-five years old had decreased
to 75% at thirteen years postoperatively.
Given the high rates of failure that had been reported in association
with a variety of cemented acetabular components
27,28
, it was hoped that uncemented acetabular components would be more
durable in younger patients. However, the rates of loosening and
osteolysis in young patients who have been managed with uncemented
acetabular components have been variable and design-specific. Dowdy
et al.
20
, in a study of forty-one hips that had received an uncemented Mallory-Head
prosthesis, reported that twelve hips (29%) had osteolysis and five
hips (12%) had acetabular loosening after only five years. Piston
et al.
22
, in a study of thirty-five hips that had been treated with several
designs of uncemented acetabular components, reported that two hips
(6%) had had an acetabular revision and six hips (17%) had acetabular
osteolysis after an average duration of follow-up of 7.5 years.
Fye et al.
12
, in a study of sixty-one hips that had been treated with a variety
of uncemented components, reported that one hip (1.6%) had had an
acetabular revision and an additional four hips (6.6%) had radiographic
signs of loosening after a mean duration of follow-up of seven years.
Dunkley et al.
21
, in a study of fifty-five total hip arthroplasties that had been performed
with the Harris-Galante-I component in patients who were fifty years
of age or younger, reported that no component had migrated but that
six liners had been replaced because of polyethylene wear after
a mean duration of follow-up of seven years. The procedures in that
study had been performed by fifteen different surgeons, and forty-five
of the fifty-five hips had received a 32-mm femoral head. Sporer
et al.
17
, in a study of forty-five total hip arthroplasties that had been performed
with the Harris-Galante-I component in patients who were less than
fifty years old, reported that no hip had acetabular loosening after
five to ten years of follow-up. Berger et al.
19
, in a study of sixty-eight total hip arthroplasties that had been performed
with use of the Harris-Galante-I porous-coated acetabular component
in patients who were less than fifty years old, reported that no
hip had acetabular loosening and five hips (7.4%) had osteolysis
after a mean duration of follow-up of 8.8 years. Our results are
similar to those reported by Berger et al., but the mean duration
of follow-up in the present study was more than two years longer.
We believe that the results of the present study demonstrate that
the rate of loosening of the acetabular component in young patients
can be greatly decreased with use of a specific type of uncemented
acetabular component fixed with multiple screws.
The prevalence of pelvic osteolysis associated with uncemented
acetabular components
19,22,29-31
has been reported to be higher than that associated with cemented
acetabular components
1-3
in all age-groups. In the present study, osteolysis was mainly localized
to the ischial region and was noted in 23% (thirteen) of the fifty-six
hips after a mean duration of follow-up of eleven years. In the
senior author's previous report, the prevalence of pelvic osteolysis
among patients who were less than fifty years old was only 3% (two
of sixty)
14
. In the entire series of 174 hips that were treated with the Harris-Galante-I
acetabular component between December 1984 and December 1989, thirteen
of the fourteen hips with pelvic osteolysis were in patients who
were less than fifty years old. Obviously, there is serious concern
regarding the increase in the prevalence of pelvic osteolysis between
seven and eleven years of follow-up. Two of these patients had a reoperation,
despite the lack of symptoms, because of the potential for later
catastrophic failure in association with wear and osteolysis. Neither
patient had evidence of loosening of the acetabular component intraoperatively.
Histological examination of tissue obtained from the site of the
osteolytic lesions showed an abundance of polyethylene particles.
Associations have been made between the increased polyethylene
wear and the prevalence of osteolysis in patients with uncemented
acetabular components
31
. The metal shell of the Harris-Galante-I acetabular component has
multiple screw-holes, which some investigators have suggested represent
an avenue for migration of the wear debris
32
. However, osteolysis around the screw-holes was seen in only two
hips in the present study. Schmalzried et al.
33
, in a review of 134 hips, noted that pelvic osteolysis occurred as
frequently or more frequently in association with acetabular components
without screw-holes as it did in association with components with
screw-holes.
Devane et al.
31
noted a significant increase in the wear rates associated with osteolysis
in hips with uncemented acetabular components as compared with hips
with cemented acetabular components. The polyethylene wear rate
associated with cemented acetabular components in younger patients
has been reported to be in the range of 0.09 to 0.12 mm/year
1,3,8
. In previous studies of patients of all ages who were managed with
the Harris-Galante-I acetabular component, the wear rate has been
reported to range from 0.07 to 0.11 mm/year
11,14,18
. In the present study of younger patients who were managed with
the Harris-Galante-I acetabular component, the mean two-dimensional
wear rate was 0.15 mm/year (range, 0.02 to 0.59 mm/yr). In the study
by Latimer and Lachiewicz
14
, the mean wear rate was 0.11 mm/year but the patients had a mean
age of fifty-eight years. Berger et al.
19
also found a higher rate of wear (0.16 mm/yr) in a study of younger
patients who had been treated with the Harris-Galante-I acetabular
component. The increased rate of polyethylene wear is likely related
to the level of activity of these younger patients. Schmalzried
et al.
34
, in a study of thirty-seven hips in patients whose activity was assessed
with a pedometer, suggested that wear is a function of use rather
than duration of follow-up or patient age.
Multivariate analysis showed a significant increase in the rate of
polyethylene wear in patients with an excellent Harris hip score
(p = 0.004) and a younger age (p = 0.026). However, no significant
differences were noted in association with the other variables that
were analyzed, including patient gender and weight, femoral neck
length, abduction angle, and liner thickness. Urquhart et al.
35
reported a significantly increased rate of polyethylene wear in
association with the use of a "long" (extended flange-reinforced)
neck modular femoral component. However, in the present study, there
was no significant association between increased wear and the use
of a femoral component with the same "long" neck length (p = 0.66).
The increased wear rates that were observed in younger patients
and patients with excellent hip scores suggests that such patients
are more active. However, the exact measurement of activity would require
a pedometer
34
or other devices, which were not utilized in the present study.
Despite the prevalence of pelvic osteolysis and the increased rate
of polyethylene wear in this group of patients who were less than
fifty years old, none of the acetabular metal shells were revised
because of aseptic loosening and none of the patients had asymptomatic
loosening or migration of the shell. The survival rate of this uncemented
acetabular component was excellent (98%) after a mean duration of
follow-up of eleven years. We attribute the success of fixation
to the specific porous coating and the number of screws used for
fixation. The senior author now uses an updated version of this component
that permits larger, 6.5-mm screws to be implanted and routinely
uses two or three screws for secure intraoperative fixation.
After a mean duration of follow-up of eleven years, the Harris-Galante-I
porous-coated acetabular component provided excellent fixation in
patients who were less than fifty years old at the time of the operation.
However, the analysis of polyethylene wear suggests that younger,
more active patients are at greater risk for the development of
pelvic osteolysis, a finding that raises serious concerns. However,
whether this asymptomatic osteolysis will lead to increased rates
of symptomatic acetabular loosening of components that have been fixed
with multiple screws is unknown, and continued regular radiographic
follow-up is recommended in this patient population.