A Biomechanical Analysis

COPYRIGHT 2003
BY
THE JOURNAL
OF
BONE
AND JOINT
SURGERY, INCORPORATED
A Biomechanical Analysis
of Polyethylene Liner
Cementation into a Fixed
Metal Acetabular Shell
BY GEOFFREY F. HAFT, MD, ANNELIESE D. HEINER, PHD, LAWRENCE D. DORR, MD,
THOMAS D. BROWN, PHD, AND JOHN J. CALLAGHAN, MD
Investigation performed at the Departments of Orthopaedic Surgery and Biomedical Engineering, University of Iowa, Iowa City, Iowa
Background: A common clinical scenario encountered by an orthopaedic surgeon is a patient with a secure cementless acetabular shell and a failed polyethylene liner. One treatment option is to cement a new liner into the fixed
shell. The purpose of this study was to evaluate technical variables to improve the mechanical strength of such cemented liner constructs.
Methods: The contributions of shell texturing, liner texturing, and cement mantle thickness (between the liner and
the shell) were evaluated by comparing torsional strength (among nine groups of constructs) and lever-out strength
(among eight groups of constructs).
Results: Failure almost always occurred at the cement-liner interface. The two exceptions (failure at the shell-cement
interface) occurred with a polished, untextured shell with no screw-holes. This finding indicates that if a shell has existing texturing (such as holes), further intraoperative scoring of the shell is unnecessary, but some sort of texturing
is necessary to avoid construct failure at the shell-cement interface. Textured liners had significantly ( = 0.05)
greater torsional and lever-out strength than untextured liners. The greatest construct strength occurred when liner
grooves were oriented so as to oppose the applied loading. A 4-mm-thick cement mantle resulted in slightly greater
torsional strength than a 2-mm-thick cement mantle, and a 2-mm-thick cement mantle resulted in considerably
greater lever-out strength than a 4-mm-thick cement mantle, but these differences were not significant.
Conclusions: When cementing a liner into a well-fixed shell, a surgeon should ensure that both the shell and the
liner are textured, as interdigitation of the cement with the shell and the liner is crucial to the mechanical strength of
this construct.
urgeons who perform total joint arthroplasty are frequently faced with a patient who has a cementless acetabular shell that is securely fixed to the pelvis but the
polyethylene liner has failed. Polyethylene wear and osteolysis
associated with a secure cementless acetabular shell are becoming frequent clinical problems, as better designs of the component last into the second decade1 and less optimal designs fail
earlier2-10. In addition to wear, another complication is dislodgment of the modular liner. Numerous case reports over the last
decade have established the ubiquity of this complication in association with a variety of acetabular components11-22.
Several options are available to the surgeon in this situation. One option is to perform a complete acetabular revision,
which may be the preferred treatment in patients who have
poor fixation of the acetabular shell, malpositioning of the
shell, or a very small shell in which a cemented liner would
unduly compromise the polyethylene thickness. However, complete revision of the acetabular component is accompanied by
serious potential complications, including pelvic discontinuity and severe bone loss. Typically, a 6 to 8-mm increase in the
diameter of the replacement acetabular shell is required with
this option, and cutting the original shell into pie-slice pieces
(with the potential for particulate debris generation) is sometimes necessary. Also, although revision acetabular components demonstrate stable interfaces at five to ten years, they
are more commonly associated with radiolucent lines than are
primary components23,24. Another option is simply to replace
the damaged or worn liner with a new version of the original.
However, there are several common situations in which this is
not an option: the locking mechanism of the liner may be
damaged, a replacement polyethylene liner may be unavailable
or of questionable quality (as a result of gamma irradiation in

THE JOUR NAL OF BONE & JOINT SURGER Y JBJS.ORG
VO L U M E 85-A N U M B E R 6 J U N E 2003
A B I O M E C H A N I C A L A N A L Y S I S O F P O L YE T HY L E N E L I N E R
C E M E N T A T I O N I N T O A F I XE D M E T A L A C E T A B U L A R S H E L L
TABLE I Combinations of Shells and Liners Used in Torsion and Lever-out Testing*
Liners
Shells
Cluster hole-unscored
(Summit 54)
Smooth 2-mm
(Summit 50)
Smooth 3-mm
(Duraloc 48)
T, L (2.05 mm)
Cluster hole-scored
(Summit 54)
Smooth 4-mm
(Summit 46)
Vertically
Scored
(Summit 46)
Circumferentially
Grooved
and Nubbed
(Ultima 44)
T, L (4.05 mm)
T, L (4.05 mm)
T, L (3.24 mm)
T (4.05 mm)
T (3.24 mm)
Polished-no hole-unscored
(Duraloc 54)
T, L (3.15 mm)
T, L (3.20 mm)
Polished-no hole-scored
(Duraloc 54)
T, L (3.15 mm)
L (3.20 mm)
*Groups were chosen on the basis of component availability. Five replicate shells and liners were tested in each combination. The components are designated by name and size. Cement mantle thicknesses for shell-liner combinations are also shown (in parentheses). All implants were manufactured by DePuy Orthopaedics (Warsaw, Indiana). T = torsion testing, and L = lever-out testing.
air or a long or unknown shelf life), or a constrained liner may

be needed to combat instability after liner removal and capsular dbridement. In these latter situations, the surgeons best
option may be to leave the fixed shell in place and cement a replacement liner into the shell.
Heck and Murray 25, in 1986, were the first to describe, as
far as we know, the cemented liner technique in a case report
on the revision of a metal-on-metal prosthesis. While there
have been only a few additional case reports on this technique
in the orthopaedic literature12,14,20,26-30, discussions at national
meetings have suggested that many surgeons throughout the
United States have been cementing liners into fixed shells in
recent years. In a retrospective review, seventeen patients with
a cemented acetabular liner were followed for an average of 2.5
years (range, one to 4.7 years)27. There were no radiographic
changes in the bone-shell interface during the follow-up interval, but the first patient in whom this method was used had a
revision because of failure of the cement-liner interface.
Given that surgeons are actively employing this technique, laboratory work is clearly necessary to help to determine the best construct that can be created. Previous studies
of the strength of cemented liner constructs have investigated
a few of the variables involved29-32. The purpose of the present
study was to expand upon these earlier studies and determine
which surgeon-controlled variables would lead to the strongest mechanical construct when a polyethylene liner is cemented into a fixed acetabular shell. The specific objectives of
the study were to determine the contributions of shell texturing, liner texturing, and cement mantle thickness to the overall mechanical strength of the construct.
Materials and Methods
he contributions of the shell-cement and cement-liner interfaces to the cemented liner construct were evaluated
with use of torsional and lever-out tests. Torsional and lever-out
tests were chosen on the basis of previously published studies of
mechanical testing of modular acetabular components33,34 and

cemented all-polyethylene liners32 as well as proposed mechanisms for liner dislodgment in failed modular acetabular components11,14,16,18-20. Push-out testing was rejected as a clinically
unrealistic failure mode, despite its previous use in the testing of liner-locking mechanisms of newly introduced modular components33-35. For both the torsional and lever-out tests,
yield strength and maximum strength were determined.
Experimental groups were designated (Table I) to determine the effects of (1) unscored compared with scored acetabular shells, with and without holes; (2) smooth compared
with textured liners; and (3) cement mantle thickness. Four
different shells and five different liners were used to evaluate
these variables (Fig. 1), with five specimens for each combination studied. The groups were chosen on the basis of component availability. All implants were provided by DePuy
Orthopaedics (Warsaw, Indiana).
The shells included Summit cluster hole size-54 (unmodified) and highly polished Duraloc Enduron size-54 (centralhole-eliminated) components. The shells were scored (by
electron discharge machining) to simulate the intraoperative
scratching that is often done ostensibly to improve cement
interdigitation25,29,31. The scoring consisted of channels centered about the cup apex. The channels had a width of 2 mm
and a depth of 1 mm. The shells were designated as cluster holeunscored, cluster hole-scored, polished-no hole-unscored,
and polished-no hole-scored.
The liners used in the present study included Summit
size-50, Summit size-46, Duraloc Enduron size-48, and Ultima size-44 implants. Equatorial nubs and ridges were removed from all liners except the Ultima liners (which were
unmodified). The Summit liners and the Duraloc liners were
axisymmetric. These liners were designated as smooth and
were further designated by their cement mantle thicknesses
(approximately 2, 3, and 4 mm for the Summit size-50, Duraloc Enduron size-48, and Summit size-46 liners, respectively).

The Ultima liners, which had three circumferential grooves

and a circle of six spacer nubs with a radius of 2 mm, were
designated as circumferentially grooved and nubbed. Summit size-46 liners were vertically scored with a specially ground
cutter on a vertical computer numeric controlled (CNC) milling machine. The scoring consisted of a cross pattern centered
about the cup apex, with 2-mm-wide and 1-mm-deep scores.
The scored Summit liners were designated as vertically scored.
The acetabular shells were potted with dental acrylic
into cylindrical containers (torsional specimens) or rectangular containers (lever-out specimens) to simulate well-fixed
shell backing. Care was taken to keep the potting medium out
of the screw-holes and central impaction hole in the shells
with cluster holes. For the scored shells that were to be evaluated with lever-out testing, the scores were aligned with the
sides of the acrylic pot. For the shells with cluster holes that
were to be evaluated with lever-out testing, the screw-holes
were aligned so that they would be on the tensile side of the
specimen (and aligned with the top of the acrylic pot).
The liners were then cemented into the shells. All specimens were cleaned with soap and water before cementing. The
cement (Surgical Simplex P provided by Howmedica, Rutherford, New Jersey) was mixed by hand, and the liner was cemented five minutes after the cement-mixing was started. For
the purpose of testing reproducibility, the thickness of the cement mantle and the centering of the liner within each shell
were precisely controlled, with use of specially designed hemispherical spacers and an axial-torsional materials testing ma-
chine (model 858; MTS Systems, Eden Prairie, Minnesota).

The hemispherical spacer was first placed in the shell, and the
liner was lowered into the spacer with use of the materials testing machine until a compressive load was detected. The axial
position of the materials testing machine with the liner resting
in the spacer was noted, the liner was raised, and the spacer
was removed. The cement was then poured into the shell, and
the liner was lowered into the cement to the previously noted
level. Excess cement was removed with use of a straight, narrow osteotome, taking care to avoid pulling cement out of the
interface. The liner was kept at this level for fifteen minutes,
and then the specimen (the potting medium, shell, cement,
and liner) was carefully removed from the fixturing of the materials testing machine. The specimens were placed in an incubator at 37C, and the cement was cured for four to five days
before testing.
Torsional Testing
Nine experimental groups of five specimens each, for a total of
forty-five specimens, were tested in torsion (Table I). The liners were rigidly fixed to the actuator of the materials testing
machine with use of twelve small, circumferentially spaced
screws. The potted acetabular shells were fixed to the loadtorque cell of the materials testing machine (Fig. 2-A). An x-y
stage allowed free horizontal motion. An axial load of 70 kg
(690 N) was applied to the liner. The liner was then rotated
about its symmetry axis (Fig. 2-A) at a rate of 1/sec until failure occurred. Torque and rotation angle data were collected
Fig. 1
Acetabular shells (top) and liners (bottom) used for torsional and lever-out testing of cemented liner constructs. The scoring simulated
intraoperative scratching that is done ostensibly to improve cement interdigitation; the channels are 2 mm wide and 1 mm deep. All
implants were manufactured by DePuy Orthopaedics.

geneity and because it provides a conservative estimate of significance. The second method ignores variance heterogeneity
and thus provides a lower limit to the estimate of p. Statistical
power was calculated for = 0.05 for nine experimental
groups of five specimens each and for the realized effect sizes
and a benchmark effect size of one standard deviation37.
Lever-out Testing
Eight experimental groups, with five specimens in each for a total of forty specimens, were evaluated with lever-out testing.
The liners were attached to a 6.35-mm-thick (0.25-in) gripping
ring, with use of twelve small circumferentially spaced screws
(Fig. 3-A). A lever arm (a 3/8"-24 bolt) was screwed into the
center of the bearing surface, and a retaining nut was screwed to
the level of the ring. A low-melting-point (70C) bismuth alloy
(Cerrobend; Cerro Metal Products, Bellefonte, Pennsylvania)
was then poured into the liner. For the smooth 3-mm liners, the
vertically scored liners, and the circumferentially grooved and
nubbed liners, eight holes were drilled into the bearing surface
for interdigitation of the bismuth alloy, to improve purchase.
The specimen was then attached to the load cell of the materials
testing machine. Lever-out torque was applied by means of a
cylindrical platen eccentrically contacting the lever arm. The
platen was lowered at a rate of 1.33 mm/sec (corresponding to a
Fig. 2-A
Figs. 2-A and 2-B Torsional testing of the cemented liner

specimens. Fig. 2-A Cutaway schematic for a smooth 4-mm
liner with a cluster hole-unscored shell. Z indicates the symmetry axis of the liner.
at 0.002-second intervals by the materials testing machine

software (MTS) and analyzed with use of Excel software
(Microsoft, Seattle, Washington). The failure interface was
recorded and photographed.
Yield torque and maximum (ultimate) torque were determined from each torque-angle recording (Fig. 2-B). Yield
torque was defined as either the first abrupt change of slope
for the torque-angle recording or (to account for nonabrupt
slope changes) the intersection of the torque-angle curve with
a 0.01 offset initial slope, whichever was the smaller value.
Each torque measurement was averaged for each shell-liner
combination. We used two methods to examine our data.
First, after testing the assumption of homogeneity of variance
among the nine experimental groups with use of the Brown
and Forsythe method36, one-way analysis of variance was used
to determine whether there was a significant mean difference
( = 0.05) between any of the shell-liner combinations. The
Tukey-Welsch multiple comparison procedure was then used
to determine which shell-liner combinations were significantly different from one another. Second, when the difference between the mean values did not quite reach significance
with the Tukey-Welsch test, we used the common Student t
test. The first method was chosen because of variance hetero-
Fig. 2-B
Representative recording, indicating yield and maximum

torque values.

Fig. 3-A
Figs. 3-A and 3-B Lever-out testing of cemented liner specimens.

Fig. 3-A Cutaway schematic for a circumferentially grooved and
nubbed liner with a cluster hole-unscored shell. P = force.
liner rotation of approximately 1/sec) until failure occurred.

Force (Fig. 3-A) and displacement data were collected in
0.01-sec intervals by the materials testing machine software
(MTS), were converted to moment and angular data, and
were analyzed with Excel software (Microsoft). The failure
interface was recorded and photographed.
Yield moment and maximum (ultimate) moment were
determined from each moment-angle recording (Fig. 3-B).
Yield moment was defined as the intersection of the momentangle curve with a 0.05 offset initial slope. Each moment measure was averaged for each shell-liner combination. Again, after
testing the assumption of homogeneity of variance among the
eight experimental groups with use of the Brown and Forsythe
method36, one-way analysis of variance was used to determine whether there was a significant mean difference ( =
0.05) between any of the shell-liner combinations. The TukeyWelsch multiple comparison procedure was used first, and then
the Student t test was used to determine which shell-liner combinations were significantly different from one another. Statistical power was calculated for = 0.05 for eight experimental
groups of five specimens each and for the realized effect sizes
and a benchmark effect size of one standard deviation37.
Results
hell texturing, liner texturing, and the cement mantle
thickness each affected both the torsional and the lever-out
strength of the cemented liner constructs (see Appendix). As
an a priori measure, for an effect size of one standard devia-
tion, power was 0.997 for the torsional measures and 0.995 for
the lever-out measures. For the actual one-way analyses of
variance, power was 1.000 for all four measures. Visible failure
occurred only at the cement-liner interface, with the exception
of only two specimens. The two exceptions were both polishedno hole-unscored shells; one involved failure of the shellcement interface with a circumferentially grooved and nubbed
liner, and the other was a combined failure of the cement-liner
and shell-cement interfaces with a smooth 3-mm liner. Because of differences in the design features among the implants,
not all cemented liner constructs could be directly compared;
therefore, only certain subsets are discussed below.
Effect of Shell Texturing
The groups were not compared with respect to the effect of
shell texturing because failure did not occur at the shellcement interface except in the two specimens discussed above.
As long as the shell had some type of texturing, whether existing features (screw-holes) or intraoperative scoring, no failure occurred at the shell-cement interface.
Effect of Liner Texturing
The effect of liner texturing was studied with the cluster holeunscored shells as a constant factor; the smooth 4-mm liners,
vertically scored liners, and circumferentially grooved and
nubbed liners were compared. In torsion, yield torque was
significantly higher for the vertically scored liners ( = 0.05)
(Fig. 4-A). Maximum torque was significantly higher ( =

mean difference of 21.6 N-m for significance with use of the

Tukey-Welsch test; however, when examined with the Student
two-tailed t test, the p value was 0.0008 (Fig. 5-B).
Discussion
he increasing use of the cemented liner technique has generated a compelling need to compare the strength of this
construct with that of modular acetabular components. Meldrum and Hollis31 established that the strength of a cemented
liner construct was similar to that of modular components in
lever-out and push-out testing. The present study evaluated
the mechanical strength of the shell-cement-liner construct in
torsional and lever-out tests, as a function of individual factors
(shell texturing, liner texturing, and cement mantle thickness)
that characterize the construct. The results of the current study
offer guidance to a surgeon who chooses to use a cemented
liner technique.
Shell texturing, in any form, prevented failure at the shellcement interface. As long as the shell had holes or was scored,
visible failure occurred only at the cement-liner interface. This
result suggests that the practice of intraoperative scoring of the
acetabular shell, to improve the strength of the cement-shell interface, is unnecessary provided that the shell has screw-holes or
other existing texturing. The data reported by Dunlop et al.38
and the surgical technique described by LaPorte et al.29 also support the concept that, as long as a shell has screw-holes, additional scoring is not needed. Avoiding scoring of the shell
prevents the creation of metal particulate debris (a likely source
of third-body wear) during the scoring process. However, if a
shell lacks texturing features, a surgeon should score the shell to
avoid failure at the shell-cement interface.
Regarding the effect of cement thickness, the present results suggest that untextured (smooth) liners with a 2-mmthick cement mantle had greater construct strength in the
lever-out test than did the constructs with the 4-mm-thick
cement mantle. We say this because there may be differing
opinions as to the type of statistical test that is appropriate for
these data. The most conservative test, the Tukey-Welsch test,
just missed significance (the test provided no p values), whereas
the less restrictive Student two-tailed t test provided p values
(yield moment, 0.00031; maximum moment, 0.0008) that
were highly significant. Under torsional loading, the differences between the liners with a 4-mm-thick cement mantle
and those with a 2-mm-thick mantle did not appear to be
significant. Bensen et al.30 observed that the mean lever-out
strength for cemented liners with a 4-mm cement mantle was
37 N-m, whereas the liner with a 2-mm cement mantle would
not lever out before the polyethylene yielded about the hole
into which the lever-out rod was inserted (moments as high as
68 N-m were recorded). In torsion, the 2-mm-thick and the
4-mm-thick cement mantles demonstrated no significant differences, although a trend toward better performance was
noted for the thicker (4-mm) cement mantle. A possible explanation may be that interface stresses change minimally as
the cement thickness changes. In a finite element study of
cemented liners under axial loading, Kurtz et al.39 found a
Fig. 3-B
Representative recording, indicating yield and maximum

moment values.
0.05) for the two textured liners (the vertically scored liners
and the circumferentially grooved and nubbed liners), and
these two liners were not significantly different from one another (Fig. 4-A). In the lever-out test, the yield and maximum
moments were significantly higher for the circumferentially
grooved and nubbed liners ( = 0.05) (Fig. 4-B).
Effect of Cement Mantle Thickness
The effect of cement mantle thickness was studied with the
cluster-hole-unscored shell as a constant factor; the smooth
4-mm and smooth 2-mm liners were compared. In torsion,
yield and maximum torque were slightly higher for smooth
4-mm liners, but the difference was not significant (Fig. 5-A).
In the lever-out test, the yield moment for the 4-mm cement
mantle was a mean (and standard deviation) of 6.63 2.36
N-m, whereas the 2-mm cement mantle resisted with a mean
of 22.85 6.41 N-m (Fig. 5-B). The difference between the
means (16.2 N-m) did not meet the requirement of a mean
difference of 17.5 N-m for significance with use of the TukeyWelsch test. However, when the conventional Student twotailed t test was used, the p value was 0.0031. For the maximum
moment in the lever-out test, the 4-mm cement mantle resisted with a mean (and standard deviation) of 23.14 6.32
N-m, whereas the 2-mm cement mantle resisted with a mean
of 42.37 4.40 N-m (Fig. 5-B). The difference between these
means (19.2 N-m) also did not meet the requirement for a

Fig. 4-A
Fig. 4-B
Effect of liner texturing in torsion testing (Fig. 4-A) and lever-out testing (Fig. 4-B). The cluster hole-unscored shell was common to all specimens in this evaluation. Asterisks indicate groups with significantly different average values ( = 0.05).

change of only 9% in cement compressive principal stress as

the cement mantle thickness was changed between 0.5 and 2.0
mm. With no compelling data pointing toward a particular
cement thickness, a surgeon is faced with a choice. One option
is to choose a thicker cement mantle. For example, to ensure a
complete cement mantle of at least 3 mm throughout (and to
prevent the liner from bottoming out), a 4-mm-thick differential (in the radius) between the liner and the shell is probably a good starting point. By erring on the side of caution and
cementing a slightly undersized liner, the surgeon can ensure
adequate liner positioning, containment of the component
within the shell, and at least minimal cement mantle thickness
throughout the construct. (In the present laboratory trials, cement mantle thickness and centering were more meticulously
controlled than is practical intraoperatively.) The other option
is to choose a thinner cement mantle, which means a thicker
liner will be used, resulting in the availability of more polyethylene for wear and for any necessary liner scoring32. However,
a thin cement mantle could lead to bottoming out of the liner
or (in an effort to avoid bottoming out) placement of a proud
liner. It should be noted, though, that if a liner with spacer
nubs was used, bottoming out and malpositioning of the liner
would not be problems.
The most important variable for the surgeon to control
is liner texturing. In almost every cemented liner combination
tested, failure occurred at the cement-liner interface. In torsion, the cemented liner constructs with a vertically scored
liner had the highest average yield torque and maximum
torque; this difference was significant as compared with the
other constructs, with one exception for each measure (both
exceptions having a circumferentially grooved and nubbed
liner) (see Appendix). In lever-out testing, constructs with a
circumferentially grooved and nubbed liner had a significantly
greater yield moment and maximum moment than any other
constructs (see Appendix). Thus, when the groove orientation was directed so as to oppose the applied loading, there
was a much greater resistance to failure. These results suggest
that if a smooth liner is to be used, the surgeon should texture
it with a series of orthogonal grooves prior to cementing it in
place. This conclusion is consistent with the data reported by
Oh32, who tested the effect of grooves on the torsional strength
of a cemented all-polyethylene liner and found that the addition of 1-mm-deep grooves to a liner increased torsional
strength from approximately 28 to 154 N-m. This conclusion
is also supported by Dunlop et al.38, who measured the torsional strength of cemented liner constructs as 2.4 to 14.6
N-m with a smooth liner, 15.6 to 44.4 N-m with a scored liner,
and 40.4 to 78.4 N-m with an all-polyethylene liner.
The U.S. Food and Drug Administration requires testing of the strength of liner capture mechanisms for all commercially available modular acetabular components. The results
of the present study of cemented liner constructs with textured liners compare favorably with those in previously published studies of modular acetabular components. For the
circumferentially grooved and nubbed liner cemented into the
cluster hole-unscored shell in the present study, the average
maximum lever-out moment was 146 N-m. Tradonsky et al.34

determined the lever-out strength of eight contemporary
modular components; the results ranged from a high of 77.3
N-m (Duraloc; DePuy, Warsaw, Indiana) to a low of 4.9 N-m
(Triloc; DePuy), with a median of 37.5 N-m (Omnifit; Osteonics, Allendale, New Jersey). Bailey et al.33 determined that
the lever-out strength of a Durasul Inter-Op liner (Sulzer Orthopedics, Austin, Texas) was 65.5 N-m.
Davidson et al.40 determined the frictional torque of the
interface of a cobalt-chromium-alloy head articulating against
an ultra-high molecular weight polyethylene shell with use
of a nonrocking biaxial hip-joint simulator. With a 32-mmdiameter head, 5000-N maximum load, and water as a lubricant, the frictional torque measured an average of 0.94 N-m.
This value is lower than even the lowest average maximum
torque measured for the cemented liners (6.1 N-m for a smooth
3-mm liner cemented into a polished-no hole-unscored shell)
and considerably lower than the average maximum torques
measured for the cemented liners that had texturing (51.9 to
65.7 N-m) (see Appendix).
The current study has several limitations. First, the process used to cement the liners into the shells was an idealized
representation of the actual intraoperative procedure. Cement
thicknesses, liner centering, and creation of reproducible scores
in shells and liners were carefully controlled to avoid adding
confounding variables to the model, but they clearly represented a best-case scenario. Second, a larger number of specimens in each shell-liner combination could have provided a
greater statistical power; for instance, the differences in outcome measures between the 2-mm and 4-mm cement mantle
thicknesses might have reached significance with a larger sample size. Third, fatigue failure, which would have been a
potentially informative mode of testing, was not considered in
the current study; liner fatigue damage and subsequent failure could result from numerous instances in which the yield
strength of the liner is exceeded. Fourth, it was not possible to
obtain completely axisymmetric shells, to directly test the hypothesis that at least some shell texturing is necessary; the designs of the shells in this study included an outer liner-locking
mechanism, which provided some stability in torque and
lever-out testing. However, most shells in patients who need
an acetabular revision include these liner-locking mechanisms, making the results in the present study clinically applicable. Finally, the shells and liners tested came from only one
company; the quantitative values might be somewhat different with shells and liners from a different company. However, if similar variables of shell and liner texturing were
studied, the qualitative results and subsequent conclusions
obtained should be similar, regardless of which companys
implants were considered.
Given the importance of liner texturing, the intraoperative time limitations for adequately performing the texturing,
and the potential for weakening a liner by over-texturing it
with a high-speed burr, the authors recommend that manufacturers consider producing polyethylene liners specifically
designed for cementing into a shell. The features of these lin-

Fig. 5-A
Fig. 5-B
Effect of cement mantle thickness in torsion testing (Fig. 5-A) and lever-out testing (Fig. 5-B). The cluster hole-unscored
shell was common to all specimens in this evaluation. No groups had significantly different average values ( = 0.05).

ers would resemble the backside designs of cemented allpolyethylene acetabular components. The liners should have
vertical grooves to increase torsional strength at the cementliner interface and circumferential grooves to increase the
lever-out strength. The liners should also incorporate spacers to prevent bottoming out or malpositioning of the liner
in the shell; these spacers can also contribute to the strength
of the cement-liner interface. Trial liners, which the surgeon
can use to ensure a proper fit in a variety of shell types, are
also crucial. Along these lines, manufacturers should also
consider making a constrained liner with these characteristics, as the cemented liner technique is gaining popularity in
patients with recurrent dislocations26,27. Given the increasing
prevalence of patients who will require cemented acetabular liners in the coming years, the availability of a line of
prefabricated textured liners would be useful in revision hip
surgery.
Appendix
Results of cemented liner constructs tested in torsion
and lever-out are available with the electronic versions of
this article, on our web site at www.jbjs.org (go to the article
citation and click on Supplementary Material) and on our
quarterly CD-ROM (call our subscription department, at 781449-9780, to order the CD-ROM).
Geoffrey F. Haft, MD
Thomas D. Brown, PhD
Department of Biomedical Engineering, University of Iowa, 2181
Westlawn Building, Iowa City, IA 52242. E-mail address for G.F.
Haft: geoff-haft@uiowa.edu. E-mail address for T.D. Brown:
tom-brown@uiowa.edu
Anneliese D. Heiner, PhD
John J. Callaghan, MD
Department of Orthopaedic Surgery, University of Iowa, 200
Hawkins Drive, Iowa City, IA 52242. E-mail address for J.J.
Callaghan: john-callaghan@uiowa.edu. E-mail address for A.D.
Heiner: anneliese-heiner@uiowa.edu
Lawrence D. Dorr, MD
The Dorr Arthritis Institute, Centinela Hospital Medical Center, 501
East Hardy Street, Suite 300, Inglewood, CA 90301. E-mail address:
centinela.appts@tenethealth.com.
In support of their research or preparation of this manuscript, one or
more of the authors received grants or outside funding from DePuy, Warsaw, Indiana. In addition, one or more of the authors received payments
or other benefits or a commitment or agreement to provide such benefits
from commercial entities (DePuy and Howmedica, Rutherford, New Jersey). Also, a commercial entity (DePuy) paid or directed, or agreed to pay
or direct, benefits to a research fund, foundation, educational institution,
or other charitable or nonprofit organization with which the authors are
affiliated or associated.
References
1. Schulte KR, Callaghan JJ, Kelley SS, Johnston RC. The outcome of Charnley
total hip arthroplasty with cement after a minimum twenty-year follow-up. The
results of one surgeon. J Bone Joint Surg Am. 1993;75:961-75.
13. Brien WW, Salvati EA, Wright TM, Nelson CL, Hungerford DS, Gilliam DL.
Dissociation of acetabular components after total hip arthroplasty. Report
of four cases. J Bone Joint Surg Am. 1990;72:1548-50.
2. Bankston AB, Cates H, Ritter MA, Keating EM, Faris PM. Polyethylene wear
in total hip arthroplasty. Clin Orthop. 1995;317:7-13.
14. Gonzalez della Valle A, Ruzo PS, Li S, Pellicci P, Sculco TP, Salvati EA. Dislodgment of polyethylene liners in first and second-generation Harris-Galante
acetabular components. A report of eighteen cases. J Bone Joint Surg Am.
2001;83:553-9.
3. Callaghan JJ, Kim YS, Brown TD, Pedersen DR, Johnston RC. Concerns and
improvements with cementless metal-backed acetabular components. Clin
Orthop. 1995;311:76-84.
4. Clohisy JC, Harris WH. The Harris-Galante porous-coated acetabular component with screw fixation. An average ten-year follow-up study. J Bone Joint
Surg Am. 1999;81:66-73.
5. Engh CA Jr, Culpepper WJ 2nd, Engh CA. Long-term results of use of the anatomic medullary locking prosthesis in total hip arthroplasty. J Bone Joint Surg
Am. 1997;79:177-84.
6. Hernandez JR, Keating EM, Faris PM, Meding JB, Ritter MA. Polyethylene
wear in uncemented acetabular components. J Bone Joint Surg Br. 1994;
76:263-6.
7. Latimer HA, Lachiewicz PF. Porous-coated acetabular components with
screw fixation. Five to ten-year results. J Bone Joint Surg Am. 1996;
78:975-81.
8. Maloney WJ, Peters P, Engh CA, Chandler H. Severe osteolysis of the pelvis
in association with acetabular replacement without cement. J Bone Joint Surg
Am. 1993;75:1627-35.
9. Nashed RS, Becker DA, Gustilo RB. Are cementless acetabular components
the cause of excess wear and osteolysis in total hip arthroplasty? Clin
Orthop. 1995;317:19-28.
10. Sychterz CJ, Moon KH, Hashimoto Y, Terefenko KM, Engh CA Jr, Bauer TW.
Wear of polyethylene cups in total hip arthroplasty. A study of specimens
retrieved post mortem. J Bone Joint Surg Am. 1996;78:1193-200.
11. Barrack RL, Burke DW, Cook SD, Skinner HB, Harris WH. Complications
related to modularity of total hip components. J Bone Joint Surg Br. 1993;
75:688-92.
12. Beaver RJ, Schemitsch EH, Gross AE. Disassembly of a one-piece metalbacked acetabular component. A case report. J Bone Joint Surg Br. 1991;
73:908-10.
15. Kauschke T, Zilch H. [An unusual early complication in cementless replacement of the hip joint. Case report]. Unfallchirurgie. 1994;20:329-32. German.
16. Kitziger KJ, DeLee JC, Evans JA. Disassembly of a modular acetabular component of a total hip-replacement arthroplasty. A case report. J Bone Joint
Surg Am. 1990;72:621-3.
17. Louwerse RT, Heyligers IC. Late failure of the polyethylene liner fixation in an
uncemented total hip arthroplasty. J Arthroplasty. 1999;14:391-6.
18. OBrien RF, Chess D. Late disassembly of a modular acetabular component.
A case report. J Arthroplasty. 1992;7 Suppl:453-5.
19. Retpen JB, Solgaard S. Late disassembly of modular acetabular components.
A report of two cases. Acta Orthop Scand. 1993;64:193-5.
20. Ries MD, Collis DK, Lynch F. Separation of the polyethylene liner from
acetabular cup metal backing. A report of three cases. Clin Orthop.
1992;282:164-9.
21. Star MJ, Colwell CW Jr, Donaldson WF 3rd, Walker RH. Dissociation of modular hip arthroplasty components after dislocation. A report of three cases at
differing dissociation levels. Clin Orthop. 1992;278:111-5.
22. Castro FP Jr, Chimento G, Munn BG, Levy RS, Timon S, Barrack RL. An analysis of Food and Drug Administration medical device reports relating to total
joint components. J Arthroplasty. 1997;12:765-71.
23. Silverton CD, Rosenberg AG, Sheinkop MB, Kull LR, Galante JO. Revision of
the acetabular component without cement after total hip arthroplasty. A
follow-up note regarding results at seven to eleven years. J Bone Joint Surg
Am. 1996;78:1366-70.
24. Weber KL, Callaghan JJ, Goetz DD, Johnston RC. Revision of a failed cemented total hip prosthesis with insertion of an acetabular component without cement and a femoral component with cement. A five to eight-year
follow-up study. J Bone Joint Surg Am. 1996;78:982-94.

25. Heck DA, Murray DG. In vivo construction of a metal-backed, high-molecularweight polyethylene cup during McKee-Farrar revision total joint arthroplasty.
A case report. J Arthroplasty. 1986;1:203-6.
26. Goetz DD, Capello WN, Callaghan JJ, Brown TD, Johnston RC. Salvage of a
recurrently dislocating total hip prosthesis with use of a constrained acetabular component. A retrospective analysis of fifty-six cases. J Bone Joint Surg
Am. 1998;80:502-9.
27. Haft GF, Heiner AD, Callaghan JJ, Brown TD. Stability of acetabular liners cemented into cementless shells. Presented as a poster exhibit at the Annual
Meeting of the American Academy of Orthopaedic Surgeons, San Francisco,
California, February 28-March 4, 2001.
28. Implex. Continuum hip system: Hedrocel acetabular cup: replacement insert
technique. Product literature. Allendale, NJ: Implex; 1999.
tigue performance. Trans Am Soc Biomech. 2000;24:165-6.

34. Tradonsky S, Postak PD, Froimson AI, Greenwald AS. A comparison of the
disassociation strength of modular acetabular components. Clin Orthop.
1993;296:154-60.
35. American Society for Testing and Materials, editors. Standard test method
for determining the axial disassembly force of a modular acetabular device.
In: Annual book of ASTM standards. Volume 13, Medical devices; emergency
medical services. Philadelphia: American Society for Testing and Materials;
2003. E1820-97.
36. Conover WJ, Johnson ME, Johnson MM. Comparative study of tests for homogeneity of variances, with applications to the outer continental shelf bidding data. Technometrics. 1981;23:351-61.
29. LaPorte DM, Mont MA, Pierre-Jacques H, Peyton RS, Hungerford DS. Technique for acetabular liner revision in a nonmodular metal-backed component.
J Arthroplasty. 1998;13:348-50.
37. Borenstein MT, Rothstein H, Cohen J, Shoenfeld D, Berlin J, Lakatos E.

Power and precision, version 2. A computer program for statistical power
analysis and confidence intervals. Mahwah, NJ: Lawrence Erlbaum Associates; 2001.
30. Bensen CV, Del Schutte H Jr, Weaver KD. Mechanical stability of polyethylene liners cemented into acetabular shells. Crit Rev Biomed Eng. 2000;
28 (1-2):7-10.
38. Dunlop D, Oxland T, Gordon J, Garbuz D, Masri B, Duncan C. The biomechanics of cementing a new liner into a well-fixed metal acetabular shell. Trans
Orthop Res Soc. 2002;27:973.
31. Meldrum RD, Hollis JM. The strength of a cement acetabular locking mechanism. J Arthroplasty. 2001;16:748-52.
39. Kurtz S, Hozak W, Ray R, Edidin A. Biomechanical analysis of cementing an

acetabular liner in a metal shell for revision total hip arthroplasty. Trans
Orthop Res Soc. 2001;26:1078.
32. Oh I. A comprehensive analysis of the factors affecting acetabular cup fixation and design in total hip replacement arthroplasty: a series of experimental and clinical studies. In: The hip proceedings of the eleventh open scientific
meeting of The Hip Society. St. Louis: CV Mosby; 1983. p 129-77.
33. Bailey A, Radefeld R, Krevolin J. Liner stability in acetabular component fa-
40. Davidson JA, Schwartz G, Lynch G, Gir S. Wear, creep, and frictional heating
of femoral implant articulating surfaces and the effect on long-term performancePart II, Friction, heating, and torque. J Biomed Mater Res. 1988;22
(A1 Suppl):69-91.

A Biomechanical Analysis

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Biomechanical Analysis

Uploaded by

Copyright:

Available Formats

COPYRIGHT 2003

air or a long or unknown shelf life), or a constrained liner may

mechanical testing of modular acetabular components33,34 and

The Ultima liners, which had three circumferential grooves

chine (model 858; MTS Systems, Eden Prairie, Minnesota).

Figs. 2-A and 2-B Torsional testing of the cemented liner

at 0.002-second intervals by the materials testing machine

Representative recording, indicating yield and maximum

Figs. 3-A and 3-B Lever-out testing of cemented liner specimens.

liner rotation of approximately 1/sec) until failure occurred.

mean difference of 21.6 N-m for significance with use of the

Representative recording, indicating yield and maximum

change of only 9% in cement compressive principal stress as

maximum lever-out moment was 146 N-m. Tradonsky et al.34

tigue performance. Trans Am Soc Biomech. 2000;24:165-6.

37. Borenstein MT, Rothstein H, Cohen J, Shoenfeld D, Berlin J, Lakatos E.

39. Kurtz S, Hozak W, Ray R, Edidin A. Biomechanical analysis of cementing an

You might also like