Although the first generation of total hip arthroplasties
with cement was quite successful in patients older than fifty years, failure
rates were substantially higher in young patients
1,2
. Several studies
3-7
have provided strong evidence that a good cementing technique results
in low rates of loosening and increases the long-term survival of
femoral components in all patients. However, the problem of late
loosening of cemented acetabular components
3,8-10
was not substantially alleviated by contemporary cementing techniques
such as pressurization and centrifugation. Hybrid total hip replacement
consisting of a cementless acetabular component and a cemented femoral
component was developed in an effort to maximize the durability
of fixation and the longevity of the implant
7,11
.
The purpose of this study was to determine the prevalence of aseptic
loosening, polyethylene wear, and osteolysis in a prospective consecutive
series of sixty-four primary hybrid total hip replacements in fifty-five
patients younger than fifty years old.
Two hundred and seventy-five consecutive primary hybrid total
hip replacements were performed in 217 patients by the senior one
of us (Y.-H.K.). Fifty-five of these patients (sixty-four hips)
were younger than fifty years of age. The average duration of follow-up
was 9.4 years (range, eight to ten years). No patient was lost to
follow-up. Forty-three patients were men, and twelve were women.
The average age at the time of the index operation was 43.4 years
(range, twenty-one to fifty years). The average weight of the patients
was 59 kg (range, 45 to 82 kg), and the average height was 163 cm
(range, 143 to 182 cm). The diagnosis was osteonecrosis of the femoral head
in forty-three hips (67%), osteoarthritis in five, osteoarthrosis
secondary to childhood pyogenic arthritis in four, osteoarthrosis
secondary to childhood tuberculous arthritis in three, ankylosing
spondylitis in three, multiple epiphyseal dysplasia in three, developmental
dysplasia of the hip in two, and rheumatoid arthritis in one.
A posterolateral approach was used in all hips. Standing was allowed
on the second day after surgery, and walking with crutches and partial
weight-bearing was begun shortly thereafter. Full weight-bearing
was allowed at six weeks after surgery.
A titanium hemispherical Duraloc series-100 acetabular component
without screw-holes (DePuy, Warsaw, Indiana) was used in all hips.
The acetabular component was press-fit after underreaming of the
acetabulum by 2 mm. A 22-mm (inner diameter) liner made of conventional
415 GUR polyethylene, which was gamma irradiated and ram extruded,
was used in all hips. An Elite plus or Elite femoral component (Ortron
90; DePuy, Leeds, United Kingdom) was cemented into all hips. These
stems are collarless and straight and have a smooth surface (Ra,
0.6 m). A 22-mm zirconia femoral head was used in all hips.
A so-called third-generation cementing technique was used in all
hips. This consisted of porosity reduction of the cement, pressurization
of the cement, and use of a distal centralizer.
Clinical and radiographic follow-up was performed at six weeks;
at three, six, and twelve months; and yearly thereafter. Harris
hip ratings
12
were determined preoperatively and at each follow-up examination.
The radiographic evaluation was done by two independent observers
(H.-K.K. and J.-S.K.). All patients underwent venography on either
the fifth or the sixth postoperative day to screen for deep vein
thrombosis.
The patients subjectively evaluated pain in the thigh with use of
a 10-point visual analog scale.
The preoperative radiographs were assessed for femoral type on
the treated side with use of the isthmus ratio described by Dorr
13
. Anteroposterior, frog-leg lateral, cross-table lateral, and Judet
radiographs made at six weeks were analyzed for the alignment of
the stem and the grade of the cement mantle according to the system
of Barrack et al.
3
. Calcar rounding, resorptive bone remodeling, and cortical hypertrophy
were also assessed.
Subsidence of the femoral component was measured with use of
a perpendicular line drawn from the greater trochanter to the lateral
border of the implant as well as from the proximal-medial portion
of the stem to the lesser trochanter as references. Migration of
the acetabular component was measured vertically between the inferior
margin of the ipsilateral teardrop and horizontally between K�hler's
line and the center of the outer shell of the acetabular cup. A
change of 3 mm compared with the baseline value was considered important.
The center of rotation, limb length, femoral neck length, femoral offset,
and abductor moment arm were measured. Heterotopic ossification
was graded according to the classification system of Brooker et
al.
14
.
Radiographs were examined for evidence of osteolysis of the acetabulum.
The dimensions and locations of osteolytic lesions were recorded
according to the three zones described by DeLee and Charnley
15
.
Radiographs were also examined for evidence of osteolysis around
the femoral stem. Femoral osteolysis was defined as areas of endosteal,
intracortical, or cancellous bone loss that were scalloped or had
the appearance of destruction of bone rather than disuse osteopenia.
Also, a radiolucent zone that was linear but >2 mm wide was deemed
to be osteolysis. The dimensions and locations of osteolytic lesions
were recorded according to the zones of Gruen et al.
16
. The length and width of the osteolytic lesions were measured.
The cementless acetabular components were considered loose when
there was a change in position or a continuous radiolucent line
wider than 2 mm on both anteroposterior and lateral radiographs.
Only definite loosening of the stem was considered in this study.
Definite loosening is defined as subsidence of the stem or the mantle
of cement, bending or breakage of the stem, or a crack in the mantle
of cement. Debonding of the cement-stem interface was evidenced
by a radiolucent line of any width at this interface.
Linear wear of the polyethylene liner was measured by a software
program (AutoCAD R13; Autodesk, Sausalito, California). Head penetration
into the polyethylene liner was determined at annual intervals from
anteroposterior pelvic radiographs with use of a computer system
that allowed us to calculate the change in position of the head
center relative to the cup center. The amount of penetration on
the six-week postoperative radiographs was assigned the "zero position," and
wear was calculated as subsequent penetration from this initial
point. Volumetric wear was calculated with the equation: V = pr
2
W, where V = volumetric wear, r = the radius of the femoral head,
and W = measured linear wear. Anteversion and the abduction angle
of the acetabular component were measured, and an attempt was made
to correlate these values with wear of the polyethylene liner.
Statistical analysis was performed with the use of chi-square tests
with Yates correction, Student two-tailed t tests, and analysis
of variance
17
. The Kaplan-Meier curve method
18
was used in the analysis of the failure-free rate. Also, Greenwood's
formula
19
was used to calculate the confidence interval of the failure-free
rate at a certain time-point.
The mean preoperative Harris hip score of 44 points (range, 5 to
66 points) improved to 95 points (range, 76 to 100 points) at the
final follow-up evaluation, at which time eight hips (13%) were
slightly painful after prolonged walking and the remaining fifty-six
hips (88%) were not painful.
Before the operation, four patients (7%) used no support for walking,
two (4%) used a cane full time, twenty-one (38%) used a crutch,
and twenty-eight (51%) used two crutches. At the final follow-up
examination, no patient used any support. Before the arthroplasty,
every patient had a limp: five (9%) had a mild limp; twenty-six
(47%), a moderate limp; and twenty-four (44%), a severe limp. At
the latest follow-up evaluation, thirty-nine patients (71%) had
no detectable limp; nine (16%), a slight limp; and seven (13%),
a moderate limp.
Two patients had transient thigh pain. No patient had thigh pain
at one year after the operation.
Sixty-two (97%) of the sixty-four femoral stems were in neutral
position, and two were in varus on postoperative radiographs.
According to the system of Barrack et al.
3
, the mantle of cement around the femur, as seen on the early postoperative
radiographs, was classified as grade A in fifty hips (78%), grade
B in six (9%), and grade C1 in eight (13%). The Dorr ratio ranged
from 0.26 to 0.48. Fifty-nine hips (92%) were Dorr type A, and five
hips (8%) were type B. No stem demonstrated subsidence or aseptic
loosening (
Figs. 1-A
and
1-B
). The average lateral opening of the acetabular components was
46.4° (range, 35° to 55°). The average anteversion
of the acetabular components was 22.5° (range, 15°
to 25°). The values for the center of rotation, limb length,
femoral neck length, femoral offset, and abductor moment arm were
similar between the untreated and treated hips or between the two
treated hips. No cup demonstrated migration or aseptic loosening.
A review of the serial radiographs revealed that fifty-eight femora
(91%) had cortical thinning or cancellization of the cortex, limited
to the calcar femorale (zone 7). Twenty femora (31%) had distal
cortical hypertrophy (in zones 3 and 4). Seven hips (11%) had an
incomplete radiolucent line of <1 mm at the interface between
the bone and cement in zones 1 and 7. There were no radiolucent
lines in the other five zones. The remaining fifty-seven hips (89%)
had no radiolucent lines. In two hips, an early postoperative gap
between the acetabulum and the acetabular component was filled in
by bone.
Six hips (9%) had mottled radiolucencies in the calcar femorale
that were presumed to be due to osteolysis. These lesions were all <1
cm in diameter and had not progressed on the serial radiographs.
No distal osteolysis was seen in any hip, and no osteolysis was
seen around the acetabular component.
In the first three years after the operation, linear wear was probably
related to the so-called bedding-in process. Subsequently, the annual
wear rate was reduced substantially. The average linear wear (and
standard deviation) was 0.96 ± 0.066, with an average
annual rate of 0.096 ± 0.013 mm (
Table I
). The average volumetric wear was 364.7 ± 25.2 mm
3
, with an average annual rate of 43.4 ± 3.5 mm
3
. There was a significant relationship between wear of the polyethylene
liner and patient age (younger than forty years old) (p = 0.034),
male gender (p = 0.028), and the abduction angle of the acetabular
component (p = 0.038). With the numbers available, there was no
significant relationship between wear and diagnosis (p = 0.17),
patient weight (p = 0.13), hip score (p = 0.1), range of motion
(p = 0.24), or amount of anteversion (p = 0.36).
No femoral or acetabular component was revised because of aseptic
loosening. One hip (2%) had a revision because of late infection.
The Kaplan-Meier survivorship analysis with failure defined as revision
of either the femoral component or the acetabular component, or
both, revealed a 98% chance of survival (95% confidence interval,
0.93 to 1.0) at 9.4 years with follow-up of all patients. The probability
of survival with failure defined as aseptic loosening of either
the femoral component or the acetabular component, or both, was
100% (95% confidence interval, 0.93 to 1.0) at 9.4 years.
There were two nondisplaced intraoperative longitudinal fractures
of the proximal-medial portion of the femur. These fractures healed
completely after treatment with a cerclage wire, and the prostheses
were considered solidly fixed at the final follow-up examination.
One hip had a late staphylococcal infection at one year after the
operation. It was revised with a cementless femoral stem with use
of an extended osteotomy of the greater trochanter. One hip dislocated
at two weeks after the operation. It was treated with closed manipulation, and
the patient had no additional episodes of dislocation. Two hips
had grade-II heterotopic ossification, and one hip had grade-III
heterotopic ossification.
We used no prophylaxis against thromboembolism. There were seven
cases (11%) of venographically documented silent deep venous thrombi
in the calf on the side of the operation. None of these thrombi
was treated, and there were no cases of pulmonary embolism.
The predominant diagnosis in this series was osteonecrosis of the
femoral head (forty-three of the sixty-four hips). Young patients
and patients with osteonecrosis of the femoral head have been reported
to have a higher rate of failure following total hip arthroplasty
1,3,20,21
.
No femoral component in this series was revised because of aseptic
loosening. We believe that these gratifying clinical results are
related to five factors: (1) improved cementing technique, (2) better
stem design, (3) tight fixation of the femoral component in the
strong trabecular bone in these young patients, (4) the utilization
of a 22-mm femoral head with a thicker polyethylene liner, (5) a
relatively light patient weight (mean, 59 kg), and short patient
stature (mean, 163 cm).
We attribute the good results of the cemented stems to the use of
an improved cementing technique (porosity reduction, pressurization
of the cement, and use of a distal centralizer) and a better stem
design (undercut flange for even stress distribution in the cement
mantle and a vaquasheen surface finish [Ra, 0.6 m] for optimal implant-cement
interlocking). Evidence of the improvement in the cementing technique
is that there was no hip with a grade-C2 or D cement mantle.
Clohisy and Harris
22
reported that none of the cementless acetabular components in their
study migrated or was revised because of aseptic loosening. We also
noted no aseptic loosening of an acetabular component in our study.
Several studies have shown the performance of a ceramic head
to be superior to that of a metal head
23-26
. The opposite finding has been reported by other authors
27-29
, including Sychterz et al.
30
and Kim et al.
31,32
, who found that increased wear was associated with a young age.
The overall rate of wear with the zirconia femoral heads used in
our study was high (mean, 0.096 mm/yr). Sequential annual wear measurements
showed that the high wear rate in the first three years after the
operation gradually decreased in the subsequent years. The initial
high wear rate was probably related to the bedding-in process. The
reasons for the high overall wear rate remain uncertain. Third-body
debris and the level of activity of the young patients may have
been factors.
The rate of femoral osteolysis in this study (six of sixty-four hips;
9%) was about the same as the rates reported by Mulroy et al. (nine
of 102 hips; 9%)
5
and Zicat et al. (six of fifty-one hips, 12%)
33
. The prevalence of acetabular osteolysis in our study (0%) was
lower than the 4% (three of seventy-four hips) reported by Zicat
et al. The low prevalence of osteolysis around the acetabular and
femoral components in our study suggests that the spread of particulate
debris
34
is limited by solid osseous integration between the cup and strong
acetabular bone as well as by the satisfactory cementing technique
in the femur. Also, it is possible that the low prevalence of osteolysis
may have been a result of our relatively short duration of follow-up.