VOL. 84-B, NO.

3, APRIL 2002 447

S. Eggli, MD
S. zBrun, MD
R. Ganz, MD
Department of Orthopaedic Surgery, University of Berne, Inselspital, CH-
3010 Berne, Switzerland.
C. Gerber, MD
Department of Orthopaedic Surgery, University of Zrich, Balgrist, CH-
8008 Zrich, Switzerland.
Correspondence should be sent to Dr S. Eggli.
2002 British Editorial Society of Bone and Joint Surgery
0301-620X/02/311344 $2.00
Comparison of polyethylene wear with
femoral heads of 22 mm and 32 mm
A PROSPECTIVE, RANDOMISED STUDY
S. Eggli, S. zBrun, C. Gerber, R. Ganz
From the University of Berne, Switzerland
I
n this prospective, randomised study, we have
compared the wear rate of cemented, acetabular
polyethylene cups articulating with either a 22 mm or
a 32 mm cobalt-chromium head. We evaluated 89
patients who had a total of 484 radiographs. The
mean follow-up period was 71.4 months (SD 29.1). All
the radiographs were digitised and electronically
measured.
The linear wear rate was signicantly higher during
the rst two years and decreased after this period to a
constant value. We suggest that this is partly due to a
run-in process caused by irregularities between
surfaces of the cup and head and an initial plastic
deformation of the polyethylene. The mean volumetric
wear was 120.3 mm
3
/year for the 32 mm head, which
was signicantly higher than the 41.5 mm
3
/year for
the 22 mm heads. The mean linear wear rate was not
signicantly different. We were, however, unable to
nd radiological signs of osteolysis in the patients who
had higher wear rates.
J Bone Joint Surg [Br] 2002;84-B:447-51.
Received 15 May 2000; Accepted after revision 24 May 2001
In 1995 Harris stated that periprosthetic osteolysis is the
main problem in contemporary total hip replacement.
1
Wear particles initiate a foreign-body reaction and are a
major factor in producing osteolysis.
1-4
Recent research has
therefore aimed at reducing the rate of wear of polyethyl-
ene to a minimum.
5-8
Besides other mechanical factors,
several studies have suggested that the size of the femoral
head signicantly inuences wear.
9-12
We have therefore compared in a prospective, random-
ised trial the in vivo polyethylene wear rate of femoral
heads of 22 mm with those of 32 mm and its effect on the
development of osteolysis.
Patients and Methods
Between 1986 and 1987 we performed 126 primary total
hip replacements using the same cemented SLS-86 stem
(Mller straight-stem; implant material, hot-forged
Ti6Al7Nb; head material, CoCrMo; Sulzer Orthopedics,
Baar, Switzerland) combined with a cemented polyethylene
cup by the same manufacturer (ultra-high molecular-weight
polyethylene RCH-1000 Chirulen). The selection of the
size of the femoral head was based on a random assignment
of 22 mm and 32 mm heads determined by a random-
number table. Thirty-seven patients were excluded either
because of previous hip surgery (21), hip infection (1), or
moderate or severe developmental dysplasia of the hip (15),
leaving 89 patients in the study.
There were 37 women and 52 men with a mean age of
66.5 years (SD 8.7) and a mean weight of 64.5 kg (SD 4.3)
for the women and 76.1kg (SD 7.2) for the men. There was
no statistically signicant difference between the groups.
The preoperative diagnosis of the affected hip was primary
coxarthrosis in 73, avascular necrosis of the femoral head
in ten and mild developmental dysplasia in six.
All patients were operated on in the supine position
using a transgluteal approach. In 49 implants the size of the
femoral head was 22 mm and in 40 it was 32 mm. Both
components were cemented with polymethylmethacrylate
cement (PMMA; Palacos, Sulzer Orthopedics). The outer
diameter of the cup ranged between 50 and 58 mm. The
minimal thickness of polyethylene was 10 mm. The mean
thickness of polyethylene for the 32 mm heads was
16.2 mm (SD 2.8) and for the 22 mm heads it was 13 mm
(SD 1.4) which was statistically signicant (p = 0.03). The
stem size, cup size, and leg length were determined before
operation.
13
All the hips were assessed radiologically by
preoperative and postoperative anteroposterior (AP) and
lateral radiographs. Further radiological evaluation was
performed at three, six and 12 months after operation. After
this period the patients were reviewed annually. A single
observer made all the measurements.
The amount of wear was evaluated on digitised radio-
graphs using software developed in collaboration with the
Maurice E. Mller Foundation, Berne, Switzerland.
14
The
area of the radiograph containing the acetabular and femo-
ral components was scanned with a resolution of 2000 dots
per inch (dPI) (Fig. 1). The head was then magnied ten-
fold and the contour dened with three points by the
investigator. This information was used to calculate the
centre of the femoral head. Using the known diameter of
the femoral head (accuracy given by the manufacturer, SD
0.001 mm; Sulzer Orthopaedics) the image was calibrated.
The centre of the cup was determined from the centre of a
line crossing the two apices of an ellipse formed by the
metal wire in the border of the cup.
10,11
With this informa-
tion, the software calculated the centres of the femoral head
and the cup and the vertical and horizontal distances of the
two centres in a co-ordinate system dened by a horizontal
line through the teardrop and its perpendicular axis (Fig. 1).
The measurement of each radiograph was repeated four
times to calculate a mean value as well as the standard
deviation and standard error. The direct linear distance
between the two centres was calculated based on Pythago-
ras theorem (d = (x
2
+y
2
) where x is the horizontal
distance and y the vertical distance between the two centres
in the co-ordinate system described above. Volumetric wear
was then calculated by the cylindrical formula
(V = r
2
d) where r is the known diameter of the femoral
head. The accuracy of the described technique of digital
measurement had an SD of 0.01 mm, determined by digitis-
ing in vitro ten radiographs of a 52 mm acetabular cup
together with a 28 mm prosthetic head.
14
The radiological evaluation was based on the criteria
dened by Johnston et al.
15
Osteolysis on the femoral side
was rated according to Gruen, McNeice and Amstutz
16
and
on the acetabular side according to DeLee and
Charnley.
17
Statistical analysis. The mean rates of wear were com-
pared statistically with the size of the head using the
unpaired Student t-test with a signicance level of p t 0.05.
Other parameters tested for a signicance correlation with
the wear rate of the acetabular polyethylene were age,
gender (Students t-test), the size of the cup and the rate of
osteolysis. One-factor ANOVA analysis was used to com-
pare ungrouped numerical data and the level of signicance
was p < 0.05.
Results
The mean clinical and radiological follow-up was 71.4
months (SD 29.1). A total of 484 radiographs was measured
digitally (5.4 follow-ups per patient). Table I shows the
cumulated linear polyethylene wear and the 95% con-
dence intervals (CI) for the 22 and 32 mm heads. Table II
448 S. EGGLI, S. ZBRUN, C. GERBER, R. GANZ
THE JOURNAL OF BONE AND JOINT SURGERY
Table I. Mean cumulative linear polyethylene wear (mm) with 95% CIs for both sizes of
head
22 mm head 32 mm head
Cumulated Cumulated
Follow-up period linear wear 95% CI linear wear 95% CI
3 mths 0.45 0.37 to 0.53 0.53 0.40 to 0.66
6 mths 0.88 0.77 to 0.99 0.92 0.75 to 1.09
12 mths 1.20 1.03 to 1.37 1.31 1.18 to 1.44
2 yrs 1.38 1.27 to 1.49 1.48 1.33 to 1.63
3 yrs 1.49 1.36 to 1.62 1.64 1.49 to 1.79
4 yrs 1.59 1.51 to 1.67 1.80 1.38 to 2.22
5 yrs 1.70 1.43 to 1.97 1.94 1.79 to 2.09
6 yrs 1.80 1.38 to 2.22 2.09 1.84 to 2.34
7 yrs 1.91 1.44 to 2.39 2.22 1.90 to 2.54
8 yrs 2.04 1.51 to 2.57 2.39 2.22 to 2.56
9 yrs 2.16 1.86 to 2.46 2.55 2.19 to 2.91
Fig. 1
Diagram showing the digital measurement of the horizontal and vertical
distances between the centre of the femoral head and the centre of the
cup.
gives the annual linear wear rate and the calculated volu-
metric wear rate for both sizes of head. During the rst 12
months the wear rate was more than ten times higher than
that after this period. After the second year it reached a
constant value of 0.11 (SEM 0.017) mm/year for the 22 mm
head and 0.15 (SEM 0.014) mm/year for the 32 mm head.
This gave a mean volumetric wear rate of 41.5 mm
3
/year
(SEM 4.1) for the 22 mm head and 120.3 mm
3
/year (SEM
10.8) for the 32 mm head (Figs 2 and 3). The difference in
wear rate between the rst and second years was highly
signicant for both sizes of head (Students t-test, 22 mm,
p = 0.007; 32 mm, p = 0.004). Between the second and
third years the difference was signicant for the 22 mm
head but not for the 32 mm head (Students t-test, 22 mm,
p = 0.03; 32 mm, p = 0.2). After the third year there was no
statistically signicant difference in the annual wear rate.
The linear wear rates of the 32 mm head were slightly
higher than those of the 22 mm heads, but because of the
high standard deviation of the measurements the differ-
ences were not signicant (Table II). By contrast, the
calculated volumetric wear rates showed signicantly high-
er values with the 32 mm heads than with the 22 mm heads
(Table II). Neither age (p = 0.095), gender (p = 0.09) nor
the size of the cup (p = 0.122) showed a correlation with
the amount of wear. Osteolysis was present in 17% of the
hips mainly around the proximal stem with no difference
between the 22 mm and the 32 mm heads (chi-squared test,
p = 0.192). No stem showed signs of loosening during this
period. In addition, the mean amount of osteolysis on the
acetabular side was 8% which did not correlate with the
amount of polyethylene wear (chi-squared test, p = 0.32).
Discussion
In the past there has been controversy in the literature with
regard to the measurement of wear.
18,19
Beckenbaugh and
Ilstrup
20
suggested that three-dimensional wear cannot be
calculated on two-dimensional radiographs. Wear is
orientated in the direction of the highest loading, that is, in
the cranial and medial directions in the hip.
21,22
Wear in the
sagittal plane is relatively small and does not signicantly
inuence the wear results.
18,19
Clarke et al
23
found that on
conventional radiographs wear cannot be determined
because the outer contour of the cup is not easily seen. Like
Afi and Jacob
24
we used the metal wire in the rim of the
cup as a landmark to measure the distance between the cup
and the centre of the femoral head (Fig. 1). Ohlin and
Selvick
19
compared the acetabular wear, measured as the
radiological distance between the centre of the femoral
head and that of the ellipse formed by the metal wire in the
cup, with direct measurement in retrieved cups and found it
to differ by less than 5%. In order to decrease the error of
measurement we digitised the radiographs. In this way we
were able to calibrate all the images to obtain the same
enlargement factor, which normally varies between 1.16
and 1.22.
13
A second advantage of the digitisation was that,
compared with the direct measurement on the hard copy of
a radiograph, distances could be enlarged to dene the
measured points more accurately. With a resolution of 200
dPI the discrimination of two measured points was
0.036 mm. In our study, the mean error of measurement
was 7.7%, but because we were not concerned with the
absolute amount of wear but with the relationship between
the wear rates of the two sizes of head this error could be
neglected. With both sizes of head the linear wear rate was
signicantly increased during the rst two years, and
declined to a constant rate subsequently. Dowling et al
25
suggested an initial higher wear rate because of a run-in
caused by irregularities of the surfaces of the head and cup.
Rose et al
26
and Zichner and Willert
27
found in their
investigations in vitro that a major part of the measured
wear rates was initially caused by plastic deformation of
the polyethylene. To our knowledge we are the rst to
demonstrate a reversed exponential wear behaviour of the
polyethylene cup in vivo in relation to time (Figs 2 and 3).
The values of linear wear rate are in agreement with other
449 COMPARISON OF POLYETHYLENE WEAR WITH FEMORAL HEADS OF 22 MM AND 32 MM
VOL. 84-B, NO. 3, APRIL 2002
Table II. The mean annual linear wear rate (mm/year) and the mean calculated volumetric wear (mm
3
/year) with 95% CI for both sizes of
head
22 mm head 32 mm head
Volumetric Volumetric
Follow-up Linear wear 95% CI wear 95% CI Linear wear 95% CI wear 95% CI
3 mths 1.8 1.50 to 2.10 669.6 556.5 to 782.7 2.12 1.59 to 2.65 1668.4 1249.8 to 2087.1*
6 mths 1.72 1.50 to 1.94 639.8 556.9 to 722.8 1.56 1.27 to 1.85 1227.7 999.5 to 1456.0*
12 mths 0.64 0.56 to 0.72 119.0 101.9 to 136.2 0.78 0.70 to 0.86 613.9 551.7 to 676.0*
2 yrs 0.18 0.17 to 0.20 67.0 61.4 to 72.5 0.17 0.15 to 0.19 133.8 120.4 to 147.2*
3 yrs 0.11 0.10 to 0.12 40.9 37.2 to 44.6 0.16 0.15 to 0.18 125.9 114.1 to 137.7*
4 yrs 0.10 0.10 to 0.11 37.2 35.3 to 39.1 0.16 0.12 to 0.20 125.9 96.8 to 155.0*
5 yrs 0.11 0.90 to 0.13 40.9 34.647.2 0.14 0.13 to 0.15 110.2 101.5 to 118.8*
6 yrs 0.10 0.08 to 0.12 37.2 28.6 to 45.8 0.15 0.130.17 118.1 103.9 to 132.2*
7 yrs 0.11 0.08 to 0.14 40.9 30.9 to 51.0 0.13 0.11 to 0.15 102.3 87.4 to 117.3*
8 yrs 0.13 0.10 to 0.16 48.4 35.7 to 61.0 0.17 0.16 to 0.18 133.8 124.4 to 143.2*
9 yrs 0.12 0.10 to 0.14 44.6 38.3 to 51.0 0.16 0.14 to 0.18 125.9 107.8 to 144.0*
* p < 0.05
measurements in vitro and in vivo, which range between
0.11 and 0.15 for the 22 mm head and 0.1 and 0.214 for the
32 mm head.
10,11,28
The differences are caused by varia-
tions in the techniques of measurement, the types of pros-
thesis used and the quality of the polyethylene, and do not
allow comparison of the results. Nevertheless, the ndings
are in agreement in that 22 mm heads have less linear wear
than 32 mm heads. Statistically, we could not nd a sig-
nicant difference in the linear wear between the two sizes
of head because of the high deviation of the measured
values, but a comparison of the volumetric wear showed
that it was approximately three times higher with the
32 mm head than with the 22 mm head, which was sig-
nicant. Although the different wear rates during the rst
two years cannot be compared because most of the wear
may represent plastic deformation of the polyethylene, the
values after this time are consistent and clearly indicate that
the 32 mm head produces signicantly more wear. We
agree with Kesteris et al
10
that these measurements may not
be taken as absolute values but are valid when comparing
the wear rate between the two sizes of head. Possible
causes for the increased wear rate in 32 mm heads are an
increased sliding distance,
10-12,14
and in 22 mm heads bet-
ter stress distribution
10,12-14,29
and increased polyethylene
thickness.
11,14
We now know that the primary reason for aseptic loosen-
ing is osteolysis caused by a foreign-body reaction initiated
by polyethylene wear.
2,4,30-32
Several long-term studies
have shown that there is a close correlation between the
amount of wear debris and osteolysis. Sochart
33
found that
the 25-year survivorship exceeded 90% for arthroplasties
with a wear rate of less than 0.1 mm per year, but that the
20-year survivorship of acetabular components with a wear
rate greater than 0.2 mm per year was below 30%. Also
Maloney et al
34
showed that there was a signicant correla-
tion between the amount of wear and loosening of either
the acetabular or femoral components. Murray
29
showed in
his survivorship analysis that a prosthesis with a 32 mm
head had a shorter survival than that with a 22 mm articula-
tion. In our study we could not nd a signicant correlation
between the wear rate and the amount of osteolysis, but we
consider that the mean follow-up time of 71.4 months was
too short to allow conclusions to be made.
Our findings have shown that 32 mm heads produce sig-
nificantly higher polyethylene acetabular wear in vivo than
22 mm heads. The measured wear distance shows a reversed
exponential course but we cannot comment on how much of
this distance is produced by real wear and how much by
plastic deformation. We could not find a significant correla-
tion between the wear rate and the amount of osteolysis after
a mean follow-up period of 71 months.
No benets in any form have been received or will be received from a
commercial party related directly or indirectly to the subject of this
article.
References
1. Harris WH. The problem of osteolysis. Clin Orthop 1995; 311:
46-53.
2. Glant TT, Jacobs JJ, Molnar G, et al. Bone resorption activity of
particulate-stimulated macrophages. J Bone Miner Res
1993;8:1071-9.
3. Willert H-G, Ludwig J, Semlitsch M. Reaction of bone to methacry-
late after hip arthroplasty: a long-term gross, light microscopic, and
scanning electron microscopic study. J Bone Joint Surg [Am]
1974;56-A:1368-82.
4. Murray DW, Rushton N. Macrophages stimulate bone resorption
when they phagocytose particles. J Bone Joint Surg [Br]
1990;72-B:988-92.
5. Fruh HJ, Willmann G. Tribological investigations of the wear couple
alumina-CFRP for total hip replacement. Biomaterials
1998;19:1145-50.
6. Devane PA, Horne JG, Martin K, Coldham G, Krause B. Three-
dimensional polyethylene wear of a press-t titanium prosthesis:
factors inuencing generation of polyethylene debris. J Arthroplasty
1997;12:256-66.
7. Besong AA, Tipper JL, Ingham E, et al. Quantitative comparison of
wear debris from UHMWPE that has and has not been sterilised by
gamma irradiation. J Bone Joint Surg [Br] 1998;80-B:340-4.
450 S. EGGLI, S. ZBRUN, C. GERBER, R. GANZ
THE JOURNAL OF BONE AND JOINT SURGERY
Fig. 2
Comparison of the linear wear rates (mm/year) of the 22 mm and the
32 mm heads.
Fig. 3
Comparison of the volumetric wear rates (mm
3
/year) of the 22 mm and the
32 mm heads.
8. Cates HE, Faris PM, Keating EM, Ritter MA. Polyethylene wear in
cemented metal-backed acetabular cups. J Bone Joint Surg [Br]
1993;75-B:249-53.
9. Jasty M, Goetz DD, Bragdon CR, et al. Wear of polyethylene
acetabular components in total hip arthroplasty: an analysis of 128
components retrieved at autopsy or revision operations. J Bone Joint
Surg [Am] 1997;79-A:349-58.
10. Kesteris U, Ilchmann T, Wingstrand H,

Onnerf alt R. Polyethylene

wear in Scanhip
R
arthroplasty with a 22 or 32 mm head: 62 matched
patients followed for 7-9 years. Acta Orthop Scand 1996;67:125-7.
11. Livermore J, Ilstrup D, Morrey B. Effect of femoral head size on
wear of the polyethylene acetabular component. J Bone Joint Surg
[Am] 1990;72-A:518-28.
12. Saikko VA, Paavolainen PO, Sl atis P. Wear of the polyethylene
acetabular cup: metallic and ceramic heads compared in a hip sim-
ulator. Acta Orthop Scand 1993;64:391-402.
13. Eggli S, Pisan M, M uller ME. The value of preoperative planning for
total hip arthroplasty. J Bone Joint Surg [Br] 1998;80-B:382-90.
14. zBrun. Pfannenabrieb bei M uller Geradschftsprothesen mit 22 mm
und 32 mm K opfen. Thesis, University of Berne/Switzerland, 1999.
15. Johnston RC, Fitzgerald RH Jr, Harris WH, et al. Clinical and
radiographic evaluation of total hip replacement: a standard system of
terminology for reporting results. J Bone Joint Surg [Am]
1990;72-A:161-8.
16. Gruen TA, McNeice GM, Amstutz HC. Modes of failure of
cemented stem-type femoral components: a radiographic analysis of
loosening. Clin Orthop 1979;141:17-27.
17. DeLee JG, Charnley J. Radiological demarcation of cemented sock-
ets in total hip replacement. Clin Orthop 1976;121:20-32.
18. Bankston AB, Ritter MA, Keating EM, Faris PM. Measurement of
polyethylene thickness in total hip arthroplasty. J Arthroplasty
1994;9:533-8.
19. Ohlin A, Selvik G. Socket wear assessment: a comparison of three
different radiographic methods. J Arthroplasty 1993;8:427-31.
20. Beckenbaugh RD, Ilstrup DM. Total hip arthroplasty: a review of
three hundred and thirty-three cases with long follow-up. J Bone Joint
Surg [Am] 1978;60-A:306-13.
21. Wroblewski BM. Direction and rate of socket wear in Charnley low-
friction arthroplasty. J Bone Joint Surg [Br] 1985;67-B:757-61.
22. Wroblewski BM. Wear and loosening of the socket in the Charnley
low-friction arthroplasty. Orthop Clin North Am 1988;19:627-30.
23. Clarke IC, Black K, Rennie C, Amstutz HC. Can wear in total hip
arthroplasties be assessed from radiographs? Clin Orthop
1976;121:126-42.
24. Afi KF, Jacob HA. Verschleissmessung bei H ufttotalendoprothesend
mit Poly athylenpfanne (RC-1000) und hartverchromtem Protasul-10
Kopf. Z Orthop Ihre Grenzgeb 1981;119:157-63.
25. Dowling JM, Atkinson JR, Dowson D, Charnley J. The character-
istics of acetabular cups worn in the human body. J Bone Joint Surg
[Br] 1978;60-B:375-82.
26. Rose RM, Nusbaum HJ, Schneider H, et al. On the true wear rate of
ultra high-molecular-weight polyethylene in the total hip prosthesis. J
Bone Joint Surg [Am] 1980;62-A:537-49.
27. Zichner LP, Willert H-G. Comparison of alumina-polyethylene and
metal-polyethylene in clinical trials. Clin Orthop 1992;282:86-94.
28. Kabo JM GJ, Loren G, Amstutz HC. In vivo wear of polyethylene
acetabular components. J Bone Joint Surg [Br] 1993;75-B:254-8.
29. Murray BF. Joint replacement arthroplasty. Vol. 1. New York/NY:
Churchill Livingstone, 1992.
30. Anthony PP, Gie GA, Howie CR, Ling RSM. Localised endos-
teal bone lysis in relation to the femoral components of cemented
total hip arthroplasties. J Bone Joint Surg [Br] 1990;72-B:971-
9.
31. Schmalzried TP, Jasty M, Harris WH. Periprosthetic bone loss in
total hip arthroplasty: polyethylene wear debris and the concept of
the effective joint space. J Bone Joint Surg [Am] 1992;74-A:849-
63.
32. Allen M, Brett F, Millett P, Rushton N. The effects of particulate
polyethylene at a weight-bearing bone-implant interface: a study in
rats. J Bone Joint Surg [Br] 1996;78-B:32-7.
33. Sochart DH. Relationship of acetabular wear to osteolysis and
loosening in total hip arthroplasty. Clin Orthop 1999;363:135-50.
34. Maloney WJ, Galante JO, Anderson M, et al. Fixation, polyethylene
wear, and pelvic osteolysis in primary total hip replacement. Clin
Orthop 1999;369:157-64.
451 COMPARISON OF POLYETHYLENE WEAR WITH FEMORAL HEADS OF 22 MM AND 32 MM
VOL. 84-B, NO. 3, APRIL 2002