Patient Cohort
Between November 1996 and November 2000, the first 400 surface
arthroplasties with a Conserve Plus prosthesis (Wright Medical Technology,
Arlington, Tennessee) were performed in 355 patients. The study was approved
by the hospital institutional review board. The most common indications for
the procedure included a young age and/or a high level of activity. In some
patients, abnormal proximal femoral morphology was also an indication for the
surface arthroplasty. No patients who met these criteria were excluded because
of severe femoral head cysts or osteopenia. All operations were performed by
the senior author (H.C.A.) who, during that time, also performed sixty-nine
primary conventional total hip arthroplasties. The patients who had the
primary arthroplasties were older (average age, 62.2 years; range, twenty to
ninety-two years) or had important limb-length discrepancies (five had Crowe
class-III developmental dysplasia of the
hip16 and one had a
proximal femoral focal deficiency). One patient had a nonunion of the femoral
neck that did not allow resurfacing, two patients had Gaucher disease, and two
patients had severe juvenile rheumatoid arthritis.
The demographic characteristics of the patients who underwent hip
resurfacing are shown in Table
I, and the etiology of the disease is shown in
Table II. The average age of
the patients who had resurfacing was 48.2 years (range, fifteen to
seventy-seven years). Thirty-two patients had a bilateral arthroplasty at the
same operation, and thirteen patients had sequential bilateral procedures
between 2.5 and thirty-four months after the first side. Fifteen patients had
a subsequent surface arthroplasty of the contralateral hip, but only the first
hip resurfacing, which had the minimum two-year follow-up, is included in this
study.
The Implant
The Conserve Plus acetabular shell is nearly hemispherical (170°). Its
exterior surface has sintered beads ranging from 50 to 150 µm in diameter
for cementless fixation. The one-piece acetabular shell is 5 mm in thickness.
We recommend press-fit insertion by underreaming by 1
mm17.
The femoral component has the same design as the Conserve hemiresurfacing
implant (Wright Medical Technology), which was approved by the Food and Drug
Administration in 1995. This component has a short metaphyseal stem to
facilitate accurate component alignment, and it permits a cement mantle with
an average thickness of 1.25 mm around the femoral head. The femoral component
is greater than a hemisphere (208°), which, in most instances, enables
coverage of all of the reamed bone by the component and maintains the length
of the femoral head and neck (Fig.
1). The surface finish is approximately 0.008 µm (0.3
microinch). The specifications for roundness are substantially more strict for
this design compared with those for the currently available hemisurface
replacement because of the importance of maintaining adequate diametral
clearances for lubrication and for minimizing wear. There are ten component
sizes in 2-mm increments: the acetabular sizes are 46 to 64 mm and the femoral
head sizes are 36 to 54 mm. All components are made of cast F-75
cobalt-chromium molybdenum alloy that is heat treated and solution annealed.
The initial multicenter investigational device exemption trial has been
completed, and the device is classified by the Food and Drug Administration as
investigational. Implantations are currently being performed in ten centers
under "continued access" pending final approval.
Surgical Technique and Hospital Course
A detailed description of the technique and instrumentation has been
published15,18.
Some modifications of the surgical technique were made during the first 100
operations, but the technique has been essentially unchanged since then.
Initially, we used a standard posterior approach but over time changed to a
hockey-stick-shaped incision. We have performed only three trochanteric
osteotomies to improve exposure. After a total capsulectomy is performed, the
acetabulum is reamed to 1-mm under size and is checked carefully for depth and
roundness with gauges. The socket is then impacted into the cavity until it is
fully seated. Radiographs were made during the procedure for the first 100
hips, and, as a result, additional impaction of the acetabular component was
done in three of those hips. After the initial 100 hips, no further
intraoperative radiographs were made to assess the position of the acetabular
component. Currently, we recommend placing the acetabular component at 25°
to 30° of anteversion and 45° of
abduction18.
The femoral component is aligned with the anatomic main axis of the femoral
neck to avoid notching the neck, especially laterally, and to cover all of the
reamed bone with the femoral
prosthesis18. The
target angle for the femoral component is 140° with the femoral
shaft18,19.
Once the head is cylindrically reamed to size, the dome is removed with an
oscillating saw, a tower alignment guide is applied to ream for the tapered
stem, and the head is chamfered. The bone is meticulously prepared by removing
all soft tissue from cysts, and numerous 0.125-in (0.32-cm) drill-holes are
placed in the dome and chamfered areas. After jet lavage to clean the head, a
femoral suction tip of the same dimensions as the metaphyseal stem is inserted
into the head to suction out the marrow prior to fixation with acrylic
cement.
In this series, all femoral head components were cemented but only a small
number of metaphyseal stems were cemented. Early in the series, the stem was
cemented in twenty hips because of severe osteopenia or large bone defects of
the femoral neck. In thirty-nine of the last forty-four hips in this series,
the stem was routinely cemented to better evaluate the effects of cementation
to improve initial fixation and to evaluate any possible negative consequences
such as stress-shielding. Twelve hips had grafting of the femoral head with
bone paste only and/or with cancellous chips (six hips). Prior to closure, the
hip should be moved through a range of motion and impinging bone should be
removed from the acetabular walls and occasionally from the posterior
trochanteric ridge.
Postoperative Management
All patients were managed with prophylactic antibiotics for two days,
adjusted low-dose warfarin for three weeks, an indomethacin suppository (100
mg) immediately postoperatively, and 25 mg of indomethacin three times a day
for five days. Currently, we give 50 mg of indomethacin preoperatively and 25
mg of indomethacin three times a day postoperatively for four days. Beginning
with the forty-ninth patient in the series, patients undergoing simultaneous
bilateral surgery were given 700 rads of single-dose radiation preoperatively
and indomethacin. Walking was begun on the first postoperative day, with
weight-bearing allowed as tolerated. Crutches were used for four to six weeks,
and a cane was occasionally used for an additional two to three weeks. Sports
were generally permitted at four months
postoperatively18.
Outcome Evaluation
The average duration of follow-up was 3.5 years (range, 2.2 to 6.2 years).
Follow-up visits to measure the range of motion of the hip and to have
radiographs made were scheduled postoperatively at three to four months, at
one year, and then at yearly intervals. All patients were followed
prospectively, and pain, walking, function, and activity were evaluated with
use of the University of California at Los Angeles (UCLA) hip-rating
system20, the Short
Form-12 (SF-12)
questionnaire21,
and the Harris hip-rating
system22. The
Harris hip scores were not collected preoperatively for the first 100 patients
in the study.
Limb-length discrepancy was assessed with use of blocks of different
thickness that were placed under the patient's foot until the pelvis leveled.
This measurement was performed preoperatively and at each follow-up visit
after surgery. Most patients (91%) were examined by the senior author in the
Los Angeles clinic or in one of fifteen special clinics held annually in other
cities in the United States. Online self-evaluation forms were submitted by 8%
of the patients who were clinically evaluated by local orthopaedists who sent
us radiographs for review. These patients were then contacted by telephone to
discuss their progress. Two patients (three hips) died of causes unrelated to
the surgery at twenty-one and twenty-three months postoperatively. Three
patients were lost to follow-up. Fifteen patients (sixteen hips; 4%) failed to
provide radiographs at two years; thus, our radiographic analysis was based on
384 hips. In 350 hips, intraoperative photographs of the femoral head were
made prior to bone-cement fixation and were used to record the quality of the
bone in the femoral head and to measure the size of bone defects. The size of
the defects in the remaining fifty hips was estimated from preoperative
radiographs that were corrected for magnification.
Surface Arthroplasty Risk Index
Beaulé et al. previously developed a risk index for surface
arthroplasty in patients who were forty years of age or
less23. This index
was calculated for all patients, with a maximum of 6 points representing the
highest possible score; the higher the score, the higher the risk of failure.
Two points were given for femoral head cysts of >1 cm; 2 points, for a
weight of <82 kg; 1 point was given for previous surgery; and 1 point, for
an activity level of =7, according to the UCLA rating system, at one to two
years postoperatively.
Radiographic Analysis
All patients had anteroposterior, modified table-down lateral, and Johnson
cross-table lateral
radiographs24 of
the pelvis made preoperatively and, when possible, during each follow-up
visit. An independent reviewer evaluated all radiographs. A new system was
devised to analyze femoral fixation with use of a scoring system of 0 to 9
points, indicating the presence or absence of radiolucencies in three zones
around the short metaphyseal stem (Fig. 2,
A). Acetabular radiolucencies were identified in three
zones as previously described (Fig. 2,
B)25,26.
Heterotopic ossification was graded according to the Brooker
classification27.
Component Position and Biomechanical Analyses
The preoperative femoral neck-shaft angle was compared with the
postoperative stem-shaft angle on the anteroposterior radiograph. Valgus was
defined as a stem-shaft angle that was greater than the neck-shaft angle by
more than 5°, and varus was defined as a stem-shaft angle that was less
than the neck-shaft angle by >5°. The orientation of the stem in the
sagittal plane was assessed from the Johnson cross-table lateral
radiographs24. A
deviation of =10° of the stem axis from the main axis of the neck was
designated as "anterior" or "posterior." It was
designated as neutral when the deviation was <10°. The hip ratio (the
abductor moment
arm28 divided by
the body moment arm) of the resurfaced hip was compared with that of the
normal hip in patients with Charnley class-A
involvement29 to
assess the accuracy of the biomechanical reconstruction
(Fig. 3).
Statistical Analysis
Two different end points were used for the calculation of the Kaplan-Meier
survivorship curves: (1) the time from surgery to the conversion to a total
hip replacement and (2) the time from surgery to the first appearance of a
stem radiolucency in hips with eventual failure of the femoral component or
when the metaphyseal stem radiolucency score was =7.
The log-rank test was used for comparing Kaplan-Meier survivorship curves.
The Cox multivariate proportionalhazards model was used to determine the
prognostic factors associated with the time to aseptic revision or the
appearance of stem radiolucencies.
Clinical Results
The average duration of follow-up was 3.5 years (range, 2.2 to 6.2 years).
Kaplan-Meier survivorship analysis demonstrated that the overall survival rate
of the component at four years was 94.4% (95% confidence interval, 91% to
98%). The clinical results (the UCLA hip score, Harris hip score, and SF-12
scores) are summarized in Table
III. Only two patients had a UCLA pain score of <7. In one
patient, who had osteonecrosis and lymphoma, the pain was reduced but was not
completely eliminated. The other patient was a candidate for revision because
of femoral component loosening but was lost to follow-up. The average Harris
hip score was 93.5 points (range, 41 to 100 points). Nine hips (one in a
patient with Charnley class-A involvement, one in a patient with Charnley
class-B involvement, and seven in patients with Charnley class-C involvement)
had a Harris hip score of <70 points. An association was found between the
Charnley class and the Harris hip scores. The average Harris hip score was
95.2 points (range, 61 to 100 points) for hips in patients with Charnley
class-A involvement and 93.3 points (range, 66 to 100 points) for hips in
patients with Charnley class-B involvement (p = 0.008). The average Harris hip
score for patients with Charnley class-C involvement was 80.7 points (range,
41 to 100 points), which was inferior to that for both patients with Charnley
class-A and those with class-B involvement (p = 0.001). Postoperatively, the
average scores (and standard deviation) on the SF-12 (50.01 ± 9.69 for
the physical component and 53.10 ± 9.40 for the mental component) did
not differ significantly from those of the general population of the United
States matched for
age29. The range of
motion improved from an average of 85.5° (range, 5° to 150°) of
flexion, 30.5° (range, 0° to 90°) of abduction-adduction measured
in extension, and 18.5° (range, 0° to 85°) in rotation arc
measured in extension to an average of 122.0° (range, 55° to
170°), 69.8° (range, 25° to 130°), and 73.7° (range,
10° to 125°), respectively.
Limb-Length Discrepancy
Seventy-eight patients had a limb-length discrepancy preoperatively. The
discrepancy was <1 cm in fifty-three patients, 1 to 2 cm in sixteen
patients, 2 to 3 cm in eight patients, and >3 cm in one patient.
Postoperatively, only twenty-five patients had a limb-length discrepancy:
twenty-two had a discrepancy of <1 cm and three had a discrepancy of 1 to 2
cm, but all discrepancies were less than the preoperative measurement.
Radiographic Results
Heterotopic Ossification
One hundred and six hips (36% of the male patients and 12% of the female
patients) had some heterotopic bone. The average UCLA pain score for this
group was 9.4 points (range, 2 to 10 points), which was not significantly
different from that for the rest of the patients. Brooker class-III or IV
heterotopic bone was observed in twenty-eight hips (7%; twenty-six patients)
all of which were in men (10% of the male patients). As a group, these hips
showed a decreased range of motion in the arc of flexion (mean, 109.5°;
range, 55° to 140°) compared with that for the rest of the male
patients (121.9°; range, 85° to 155°) (p = 0.001). All of the
patients had functional arcs of rotation and abduction-adduction. After
implementation of our radiation protocol for patients who had one-stage
bilateral arthroplasty, the overall rate of Brooker class-III or IV
heterotopic ossification was 5.4% (three of fifty-six hips).
Hip Biomechanics
The stem-shaft angle in the first 100 hips (average, 131.1°; range,
110° to 150°) was significantly more valgus than that in the
subsequent hips (average, 137.8°; range, 111° to 153°) (p =
0.001). The stem-shaft angle in the hips that were revised because of femoral
loosening (average, 129°; range, 110° to 148°) was significantly
lower (more varus) than that in the rest of the cohort (average, 136°;
range, 111° to 163°) (p = 0.0255). The stem-shaft angle was negatively
correlated (r = —0.374, p < 0.001) to the abductor moment arm.
In patients with Charnley class-A involvement, the normal hip abductor
mechanics were restored, as denoted by a similar hip ratio on the involved
side (average, 0.584; range, 0.40 to 0.83) and the contralateral, uninvolved
side (average, 0.571; range, 0.40 to 0.78) (p = 0.193).
Acetabular Radiolucencies
Acetabular radiolucencies are reported in
Table IV. Two hundred and
sixty-one hips (68%) had no radiolucencies and 122 hips (32%) had
radiolucencies in one or two zones. Cup protrusion occurred immediately
postoperatively in one hip, and the component was revised to a polyethylene
bearing. There were no progressive radiolucencies in zones I or II.
Metaphyseal Stem Radiolucencies
According to our rating scheme (Fig. 2,
A), sixteen hips (4.2%) that had not been revised had a
femoral metaphyseal stem radiolucency score of =7
(Fig. 4). Only one of these
patients was symptomatic at the time of latest follow-up. The first appearance
of these radiolucencies occurred at an average of twenty-three months (range,
ten to fifty months) postoperatively. The average pain and activity scores of
the group with a radiolucency score of =7 were not significantly different
from those of the rest of the cohort. Several factors correlated with femoral
fixation scores. The main clinical factors associated with the presence of
radiolucencies are summarized in Table
V. In addition, patients were at increased risk of having a
radiolucency if they had large femoral-head cysts (p = 0.029), were female (p
= 0.005), or were of lesser height (p = 0.032). Smaller component size was
significantly (p = 0.005) associated with femoral radiolucencies in male
patients only. None of the fifty-nine cemented metaphyseal stems, even those
with less than optimal bone quality, showed any radiolucency at the time of
the last review. However, this group had had a much shorter duration of
follow-up than the rest of the cohort, and we were not able to detect a
significant relationship (p = 0.206) with use of the log-rank test for
Kaplan-Maier survivorship analysis.
Surface Arthroplasty Risk Index
The average surface arthroplasty risk index (and standard deviation) for
the whole group was 2.73 ± 1.55 (range, 0 to 6). When conversion to
total hip replacement was used as the end point, a surface arthroplasty risk
index of >3 points was significantly associated (p = 0.004) with an earlier
time to revision (Fig. 5-A).
The rate of survival of the implants at four years was 97% (95% confidence
interval, 94% to 100%) for patients with a low risk of revision (surface
arthroplasty risk index of =3 points) compared with 88.8% (95% confidence
interval, 80.0% to 97.6%) for patients with a high risk (surface arthroplasty
risk index of >3 points). The average surface arthroplasty risk index was
3.77 ± 1.45 points (range, 0 to 6 points) for the group with
metaphyseal stem radiolucencies compared with 2.66 ± 1.53 points
(range, 0 to 6 points) for the remaining hips (p = 0.0004). For patients with
a surface arthroplasty risk index of >3 points, the chance that a femoral
radiolucency would develop was 4.2 times greater than that for patients with a
risk index of =3 points. Figure
5-B shows the survivorship curves at four years with the first
appearance of a radiolucency around the metaphyseal stem used as the end
point. No femoral stem radiolucency was seen at four years in 73.8% (95%
confidence interval, 60.2% to 87.3%) of the high-risk group and 94.5% (95%
confidence interval, 90.2% to 98.8%) of the low-risk group (p = 0.001).
Femoral Cyst and Bone Defect Assessments
The bone defects of the femoral heads were categorized according to the
size of the largest defect and whether they were single or multiple. The
various categories are summarized in Table
VI, which illustrates the association between the size of the
defects and the clinical or radiographic results.
Conversion to Total Hip Replacement
Twelve hips (3%) were converted to total hip replacement. Seven revisions
were done because of loosening of the femoral component and three, because of
a fracture of the femoral neck. Two fractures of the femoral neck occurred
within the first six weeks postoperatively, and the third occurred at twenty
months. One hip was revised to a total hip replacement because of recurrent
subluxations. In one patient with rheumatoid arthritis, a late hematogenous
infection developed and the surface replacement was converted to a total hip
replacement (a direct exchange) at thirty-six months. The patient was free of
infection at eighteen months after the revision.
The demographic characteristics of the seven patients who had loosening of
the femoral component were not different from those of the overall group.
However, the component stem-shaft angle was 128.3° (range, 110° to
148°), which was significantly different from that of the rest of the
patients (136.2°; range, 111° to 163°) (p = 0.0255). The time to
first observation of the metaphyseal stem radiolucency was twenty months
(range, 12.5 to thirty-six months), and the time to the first symptoms was
twenty-seven months (range, sixteen to fifty-one months). The time to the
revision was thirty-five months (range, twenty-three to sixty-one months). All
but one of the failed femoral components were revised to a total hip
replacement with use of a unipolar head that was size-matched to the inside
diameter of the existing well-fixed socket.
In five of the patients who had a revision to a total hip replacement
because of loosening of the femoral component, large areas of cystic
degeneration of the head and remaining osseous defects were seen even after
bone preparation that diminished the surface area available for fixation. In
three hips, the components were "proud," as suggested by a thick
mantle of cement in the dome area, and the surrounding bone was sclerotic.
Four patients had a high activity level (an activity score of 8 and 9 in two
each).
In addition to the twelve hips that had a conversion to a total hip
replacement, one hip in a patient who had had a bilateral surface arthroplasty
had an isolated revision of the acetabular component on the right side after
the shell acutely protruded into the medial acetabular wall on the second
postoperative day when the patient had been walking with a four-point gait. An
infection subsequently developed in the hip, and an isolated direct socket
exchange was performed with use of a cup inserted with bone cement containing
antibiotics.
Other Complications
Four hips required a reoperation, including a cup exchange because of
component mismatch, removal of heterotopic bone from two hips in one patient,
and wire removal in a hip with trochanteric bursitis. The prevalence of
dislocation was 0.75% (three of 400). Two hips had an early postoperative
dislocation, and one had recurrent dislocation. The latter hip was in a
patient with hip dysplasia who had undergone multiple osteotomies and had
abnormal femoral geometry (coxa valga of 160°). This hip was eventually
converted to a total hip replacement, as mentioned in the previous section.
The other two dislocations required closed reduction, but the hips were stable
at the time of the most recent follow-up.
Femoral Component Size and Interface Bone-Cement Area
The distribution of the femoral components according to size in the male
and female patients is shown in the Appendix. The contact areas between bone
and cement, calculated from the inside dimensions of the component by adding
the cylindrically reamed area, chamfered reamed area, resected dome area, and
approximated area of the drilled holes, also appear in the Appendix.
The results of total hip arthroplasty have been excellent in older
age-groups; however, for patients who are forty years of age or younger, the
failure rates have ranged from 21% to 28% at five
years1,3,30,31.
The concept of surface arthroplasty for the treatment of advanced arthritis of
the hip in young and active patients has many attractive features because of
its ability to preserve femoral bone.
Although our early clinical results are promising, thirteen hips needed a
revision, suggesting that all patients may not be suitable candidates for this
procedure. The appearance of femoral stem radiolucencies is of concern,
although only one patient was symptomatic. At the present time, the major
factors related to femoral stem radiolucency are female gender and the
presence of large cysts in the femoral head. These reflect the importance of
femoral fixation, a factor that we consider crucial to long-term durability.
The most important factor that determines good fixation appears to be the area
available for cement fixation. Individuals with smaller reamed femoral heads
and small component sizes (which includes most female and lighter male
patients) had more loosening and radiolucencies than did those with larger
head sizes. These observations are consistent with previous findings with the
Tharies (total hip articular replacement with internal eccentric shells)
surface replacement, which demonstrated a lower rate of survival for the
smaller femoral components (39%) compared with the larger sizes (59%) at
eleven years of
follow-up32. In
recognition of the higher risk for this group, we now recommend insertion of
the femoral stem with cement and use of additional drill-holes in the prepared
femoral head to increase the fixation area, especially for women and in the
presence of extensive cysts. The area available for fixation in a femoral
surface arthroplasty with cement is related to the quality of the bone, and it
is apparent that more cysts and/or larger cyst size and a smaller femoral head
adversely affect the area available for fixation.
Careful patient selection is critical to the success of any given
procedure. By acknowledging that certain types of patients do better with this
procedure, at least as it is currently performed, than others do, we can
minimize early failures. There are two main modes of femoral failure following
surface arthroplasty: neck fracture and aseptic loosening. The occurrence of
early femoral neck fracture in this series was very low (0.75%), probably
reflecting the experience of the surgeon and the surgical technique. It is
important to avoid or at least to minimize notching the neck and to cover all
of the reamed bone with the component. If the component is not fully seated,
the uncovered reamed area behaves as a circumferential stress-riser. Special
care must be taken, when reaming cylindrically at the recommended angle of
140°, to stop the reaming before the reamer touches the lateral cortex. A
large osteophyte, which usually forms anteriorly on the femoral neck, should
be removed only if there is a substantial impingement of the range of motion
at 90° of flexion and internal rotation.
The hips that had aseptic loosening of the femoral component and required
revision appear to have had several risk factors, which are enumerated in
Table VI. Four of the hips were
among the first 100 that were resurfaced before additional fixation holes were
added to the chamfered area and the femoral suction tip technique was
developed. Our current femoral component has a short metaphyseal stem that
serves as an "antenna," and we believe that this provides an early
indication of the quality of fixation at the femoral bone-cement-implant
interfaces. We believe that a radiolucency score of =7 on our rating system
is substantial and that the radiolucency is likely to progress to component
loosening over time.
To improve patient selection, Beaulé et
al.23 reviewed the
early clinical results after metal-on-metal surface arthroplasty in patients
who were forty years of age or younger. This group is considered a target
population for the procedure and has the most to gain from a more conservative
prosthetic solution. In that series, the surface arthroplasty risk index was
significantly higher (4.7 points, p < 0.001) for the thirteen hips with a
complication (those converted to total hip replacement or those with femoral
lucencies) compared with 2.6 points for the group of eighty-one hips that did
not have a complication. In patients with an index of >3 points, the
relative risk of early problems was twelve times greater than that in patients
with an index of =3 points. In the present series, patients with an index
of >3 points had a relative risk of early problems that was 4.2 times
greater than that in patients with an index of =3 points. The surface
arthroplasty risk index appears to be a useful tool in determining which
patients can safely benefit from metal-on-metal surface arthroplasty.
Ultimately, the control of postoperative factors is up to the patients, but
it is the duty of the surgeon to properly inform them of the potential risks.
At the present time, we do not have definitive contraindications to the
surgery, but patients who have compromised bone stock, particularly large
femoral head cysts and small femoral heads, warrant special scrutiny as
candidates for the procedure. Patients also should be informed that high
activity levels (especially impact sports), although they have not been found
to be significant at this time, are likely to shorten the life of the implant
in a manner similar to the way in which those activities affect a total hip
replacement.
The prevalence of Brooker class-III and IV heterotopic ossification of 10%
in the male patients seemed high; however, the prevalence of 7% in the overall
group was comparable with the prevalence of 5.8% reported by Dorn et al., in a
study in which 40% of the patients were male and the authors advocated the use
of indomethacin for four days as prophylaxis against heterotopic
ossification33. We
anticipated some increase in the frequency of heterotopic ossification because
of the additional stretching of the muscles required for surface arthroplasty
to facilitate work around the head and neck and to obtain access to the
acetabulum. After the first of thirty-two patients (sixty-four hips) managed
with simultaneous bilateral arthroplasty had formation of Brooker class-III or
IV heterotopic bone, we changed our protocol when operating on both hips at
the same operation and provided radiation at a single preoperative dose of 700
rads. This protocol proved to be effective, as Brooker class-III or IV
heterotopic ossification developed in only 5.4% (three) of the fifty-six hips
(twenty-eight patients) treated with radiation. Prophylaxis with indomethacin
remains our recommendation for patients who have a unilateral procedure.
Despite the lack of clear evidence linking metal-on-metal total hip
arthroplasty with long-term problems, there are lingering concerns over the
local and possible systemic effects of metal wear products, including ions.
Recent reports of unusual lymphocytic aggregates in the tissues around failed
metal-on-metal stem-type total hip
replacements34,35
prompted us to investigate the tissues around the failed components in our
study, and we found that such features were present in approximately one-third
of the hips36. At
the present time, there does not appear to be an association between these
lymphocytic aggregates and clinical outcome, but this is the subject of
ongoing research. The levels of metal ions in the blood and urine of
approximately fifty patients with the Conserve Plus implants are being
monitored over time, and the results to date are comparable with those from
conventional total hip replacements with metal-on-metal
bearings37.
Finally, at the present time, there is no defined cause-and-effect
relationship between the metal-on-metal implants that were first implanted in
the 1960s and
cancer38.
Comparisons of the rate of survival of the implants in the present series
and that of other implants should take into account the high activity levels
(average score, 7.7) of these patients. For 54% of the patients, the score for
activity level on the UCLA rating system was >7. Continued close follow-up
is needed to better define the results and indications for this procedure.
A table showing the area of bone-cement interface for the various component
sizes and a graph showing the distribution of the component sizes by gender in
this series are available with the electronic versions of this article, on our
web site at
(go to the article citation and click on "Supplementary Material")
and on our quarterly CD-ROM (call our subscription department, at
781-449-9780, to order the CD-ROM).