Hip simulator and early clinical studies of highly cross-linked ultra-high
molecular weight polyethylene have demonstrated less wear and less femoral
head penetration when compared with conventional ultra-high molecular weight
polyethylene1-6.
However, cross-linking also alters some of the mechanical properties of
ultra-high molecular weight polyethylene, including its ultimate tensile
strength, strain to failure, fracture toughness, and fatigue crack propagation
resistance7-9.
Analyses of early retrieved highly cross-linked components have shown
initiation of surface cracking, which is possibly related to the decrease in
ductility caused by
cross-linking10. A
reduction in ductility, fracture, and fatigue properties is the hallmark of
material embrittlement. Accordingly, newer so-called second-generation highly
cross-linked polyethylenes have been developed in an effort to better retain
the desirable mechanical properties of conventional ultra-high molecular
weight polyethylene as well as the benefits of
cross-linking11.
Wear behavior of conventional and highly cross-linked ultra-high molecular
weight polyethylene acetabular liners can be affected by the geometry of the
femoral head. Geometrical factors include not only roughness associated with
asperities and machining marks but also asphericity of the implant. Roughening
of cobalt-chromium femoral heads occurs in vivo and may increase wear of the
liner12-14.
Root-mean-square roughness values of clinically retrieved cobalt chromium
components heads have been reported to be between 0.15 and 0.20
µm13,14.
Asphericity, or out-of-roundness, of the femoral head influences the bearing
wear of the implant. It is possible to have an extremely smooth, polished
femoral head surface that has relatively large-size scale deviations from the
optimal implant shape. Asphericity can manifest as a region of relative
flatness or as a
protrusion15.
Substantial asphericity of the head has been associated with increased
acetabular
wear16.
This case report focuses on a retrieved cobalt-chromium alloy femoral head
with an apparent protrusion in its main weight-bearing region. We hypothesized
that this protrusion contributed to the observed accelerated damage to the
surface of the highly cross-linked ultra-high molecular weight polyethylene
counter bearing. Our patient was informed that data concerning the case would
be submitted for publication and consented.
A seventy-four-year-old man with a body mass index of 30.7 and history of
diabetes and peripheral vascular disease had a painful nonunion of a femoral
neck fracture nine months after in situ pinning. He was treated with hybrid
total hip arthroplasty. The femoral component was a size-16 cemented VerSys
Heritage (Zimmer, Warsaw, Indiana) high-offset polished stem. The modular
cobalt-chromium head (Zimmer) had a diameter of 32 mm and a neck length of 3.5
mm. The cup was a 62-mm Trilogy (Zimmer) shell that had been subjected to 10
Mrad of electron beam irradiation, and the highly cross-linked Longevity
acetabular liner (Zimmer) had undergone annealing at 150°C (above the melt
temperature of ultra-high molecular weight polyethylene)
(Fig. 1). The acetabular cup
was positioned at 45° of abduction and 19° of anteversion as measured
by the method of Ackland et
al.17. After
surgery, the patient underwent a total hip rehabilitation protocol, including
abductor muscle protection for six weeks with use of crutches, then
progression to weight-bearing as tolerated followed by unrestricted activity.
By three months after surgery, the patient was walking without support and had
no pain.
Seven months after arthroplasty, acute hip pain and a fever developed. The
patient underwent hip aspiration, which revealed group-B streptococcal
infection. An urgent irrigation, débridement, and femoral head and
liner exchange were performed. Intraoperative findings demonstrated purulent
fluid within the hip joint and gross roughening of the acetabular liner
surface. However, no loose cement or osseous or other loose body debris was
identified.
Retrieval Analysis
The retrieved acetabular liner was examined visually and under light
microscopy. It was then quartered, and one quadrant was sputter-coated with
gold-palladium and examined under scanning electron microscopy. The
articulating surface of the cup demonstrated a substantial amount of damage,
characterized by scratching, cracking, or ductile flow. An area close to the
rim of the acetabular shell showed what appeared to be widespread ductile
flow. In the weight-bearing region of the dome, deep crevices in excess of 100
µm in length were prevalent (Figs. 2-A
and 2-B). A second implant quadrant was remelted with use of the
method described by Muratoglu et
al.18,19
and subsequently sputter-coated and examined using scanning electron
microscopy. This remelting technique is used to determine whether cold flow
and permanent deformation are activated as damage modes. The specimen was
equilibrated to room temperature prior to imaging with scanning electron
microscopy. Post-remelting scanning electron microscopic inspection
demonstrated surface cracking that appeared unchanged from pre-melting images,
indicating that the damage was permanent and not recoverable.
The retrieved femoral head was examined with use of three-dimensional laser
micrometry. The laser micrometer has an accuracy of 0.5 µm and a scan rate
of 400 scans/sec (Laser-Linc, Fairborn, Ohio). The equivalent of 360 circular
traces was taken with three-dimensional laser micrometry, and the resulting
scan consisted of more than 600,000 individual coordinate points. Inspection
revealed a relatively large raised area, approximately 10 mm in length and 8
mm in width, suggesting a protrusion on the femoral head
(Fig. 3). Figures E-1 and E-2
in the electronic Appendix show profiles of the actual protrusion, compared
with an ideal spherical surface and with a model surface profile used for the
finite-element analysis (see below).
Finite-Element Analysis
We performed a finite-element analysis to generate an estimate of the
effect of a bearing-surface protrusion on the contact stresses in an
ultra-high molecular weight polyethylene counter bearing. All finite-element
analysis simulations were performed using ABAQUS (v.6.5; HKS, Pawtucket, Rhode
Island). The model geometry for both the femoral head and the ultra-high
molecular weight polyethylene acetabular cup was generated with use of Patran
(2001; MSC Software, Santa Ana, California). The model was discretized into a
two-dimensional axisymmetric mesh consisting of approximately 3000 four-node
bilinear elements for the body and surface-to-surface contact elements for the
contact surface between the femoral head and the ultra-high molecular weight
polyethylene liner. To capture contact mechanical predictions on a fine
resolution, the mesh was graded, with the smallest elements near the
anticipated contact region. We modeled a head with a diameter of 32 mm, an
elastic modulus of 200 GPa, and a Poisson ratio of 0.28. The inner diameter of
the modeled liner was 32 mm, and the cup was 10-mm thick. The liner was given
a Poisson ratio of 0.48, and the inelastic deformation response was piecewise
linear fit to true stress-strain behavior as determined from the data given in
the work of Kurtz et
al.20. The radial
clearance between the two model surfaces was 100 µm. The geometry of the
model protrusion consisted of a cone (8 mm in width and 9 µm in height)
that was superimposed onto an ideal spherical head at the centerline (see
Appendix). The femoral head was loaded to 700 N axially, which approximates
the full body weight of the patient (see Appendix).
The results of the simulation for the ideal sphere (and for the non-ideal
geometry incorporating the cone-on-sphere protrusion) are shown in the
Appendix and are summarized in Table
I. The maximum von Mises effective stress in the presence of the
protrusion was 60% higher than in the ideal case (see Appendix). The maximum
contact pressure was 98% higher than in the ideal case. The point of maximum
effective stress was substantially closer to the surface in the perturbed
case, indicating a possible worsening of surface and subsurface damage and
therefore wear phenomena (Figs.
4-A and
4-B).
The sphericity of the femoral head may have a substantial effect on
ultra-high molecular weight polyethylene
wear16. Ito et al.
found considerable variability in sphericity and surface profiles when
comparing femoral heads from different
manufacturers21.
They reported that while the mean deviation (± standard deviation) from
an ideal sphere for ball bearings measured only 0.3 ± 0.0 µm, the
mean deviation ranged from to 0.4 ± 0.1 µm to 12.1 ± 4.6
µm for femoral heads from various manufacturers. Of interest, femoral heads
manufactured in 1995 were found to have inferior sphericity compared with
those made in 1999 or
200021.
The protrusion studied in this work is important to consider for several
reasons. First, it resulted in a dramatic local increase in the maximum stress
and incident surface pressure due to contact. Further, it represents a type of
geometric manufacturing flaw that might be entirely missed in a manufacturing
quality-control process. For instance, Ito et al. reported sphericity by
measuring only two orthogonal circular traces (sagittal and
transverse)21.
However, to obtain a more accurate quantification of asphericity, a small
number of traces must be made of the roundness of a sphere to determine its
surface geometry15.
Nevertheless, the space between these traces will always allow sufficiently
small protrusions to escape detection. For example, if only two orthogonal
traces are made, a protrusion smaller than a quadrant of the head might not be
identified. Also, roughness measurements are typically taken over a small
region of the surface to identify microscopic asperities and are not a
suitable method of detecting the millimeter-size protrusion described in this
report.
Our finite-element analysis simulation predicted a substantial increase in
the maximum magnitude of the effective stress resulting from contact-loading
with a perfectly smooth femoral head that has a protrusion. The volume of
material affected by this predicted increase in stress is near the surface,
while the corresponding background stress away from the surface is the same as
that predicted for a spherical head without a protrusion. These results are
not surprising since contact stresses generally increase and the depth of
maximum stress decreases as contacts become smaller or more concentrated. In
our study, we documented a raised contact region superimposed on a larger
spherical one, resulting in a region of high stress near the surface and a
surrounding field of ordinary stress.
The value of the maximum effective stress in our simulation, 3.66 MPa, is
not by itself sufficient to cause yielding in cross-linked ultra-high
molecular weight polyethylene, which has a yield strength of 20 to 25
MPa7. Rather, when
ultra-high molecular weight polyethylene surfaces are damaged by the femoral
head in total hip arthroplasty, it is typically the sharp microscopic
asperities that comprise femoral head roughness and that cause severe
near-surface stresses during articulating contact. Roughness-related
micro-asperities are driven into the counter-bearing surface by the incident
contact pressure, and the resulting localized stresses can exceed the yield
strength and cause plastic deformation, surface damage, and other traditional
wear phenomena22.
Therefore, increased contact pressure causes more severe local stresses at the
microscale and magnifies surface damage and wear phenomena associated with
micro-asperities. Thus, we infer that a protrusion, due to its associated
increased contact pressure, can substantially increase the surface damage
effects in the region contacted by the protrusion during weight-bearing joint
motion. Further, the concomitant loss of ductility, strength, and fracture
resistance that accompanies cross-linking of ultra-high molecular weight
polyethylene renders this polymer substantially more susceptible to apparently
brittle damage mechanisms (e.g., cracking). Therefore, a highly cross-linked
ultra-high molecular weight polyethylene surface that undergoes exacerbated
contact stresses (e.g., those predicted by our finite-element analysis) may be
prone to accelerated surface damage, such as the extensive cracking and
delamination damage observed in this case.
Figures describing the finite-element analysis 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). ?