- Email: [email protected]
The effect of motion patterns on edge-loading of metal-on-metal hip resurfacing
Medical Engineering & Physics 33 (2011) 1212–1220
Contents lists available at ScienceDirect
Medical Engineering & Physics journal homepage: www.elsevier.com/locate/medengphy
The effect of motion patterns on edge-loading of metal-on-metal hip resurfacing S.J. Mellon ∗ , Y.-M. Kwon, S. Glyn-Jones, D.W. Murray, H.S. Gill Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, UK
a r t i c l e
i n f o
Article history: Received 10 August 2010 Received in revised form 18 May 2011 Accepted 19 May 2011 Keywords: Biomechanics Finite element analysis Hip resurfacing arthroplasty
a b s t r a c t The occurrence of pseudotumours (soft tissue masses relating to the hip joint) following metal-on-metal hip resurfacing arthroplasty (MoMHRA) has been associated with high serum metal ion levels and consequently higher than normal bearing wear. We investigated the relationship between serum metal ion levels and contact stress on the acetabular component of MoMHRA patients for two functional activities; gait and stair descent. Four subjects with MoMHRA, who had their serum metal ion levels measured, underwent motion analysis followed by CT scanning. Their motion capture data was combined with published hip contact forces and finite element models representing 14% (peak force) and 60% (end of stance) of the gait cycle and 52% (peak force) of stair descent activity were created. The inclination angle of the acetabular component was increased by 10◦ in 1◦ intervals and the contact stresses were determined at each interval for each subject. When the inclination angle was altered in such a way as to cause the hip contact force to pass through the edge of the acetabular component edge-loading occurred. Edge-loading increased the contact stress by at least 50%; the maximum increase was 108%. Patients with low serum metal ion levels showed no increase in contact stress at peak force during gait or stair descent. Patients with high serum metal ion levels exhibited edge-loading with an increase to the inclination angle of their acetabular components. The increase in inclination angle that induced edge-loading for these subjects was less than the intersubject variability in the angle of published hip contact forces. The results of this study suggest that high serum metal ion levels are the result of inclination angle influenced edge-loading but that edge-loading cannot be attributed to inclination angle alone and that an individual’s activity patterns can reduce or even override the influence of a steep acetabular component and prevent edge-loading. © 2011 IPEM. Published by Elsevier Ltd. All rights reserved.
1. Introduction The resurgence in popularity of metal-on-metal hip resurfacing arthroplasty (MoMHRA) is due, in part, to a lower wear rate of these devices which has the potential to reduce the risk of long-term wear induced osteolysis associated with the failure of metal-on-plastic (MoP) bearings. Further advantages of MoMHRA include lower risk of dislocation, a greater degree of bone conservation and restoration of anatomical hip mechanics [1]. MoMHRA is therefore viewed as a treatment option more suited to young and active patients with hip osteoarthritis. Hip resurfacing arthroplasties carried out on patients 55 years or younger accounted for approximately 46% of primary hip arthroplasties performed in the UK on this age group in 2004 [2]. The majority of MoMHRA patients have a longer life expectancy when compared with elderly patients implanted with conventional total hip replacements (THR). The effects of long-term exposure to metal debris from metalon-metal wear are largely unknown. There is, therefore, heightened
∗ Corresponding author. Tel.: +44 (0)1865 227454. E-mail address: [email protected] (S.J. Mellon).
concern over periprosthetic soft tissue masses observed in a small percentage of patients with MoMHRA [3,4]. These masses have been termed pseudotumours due to their morphological similarity to necrotic tumours. Pseudotumours can be extremely destructive, causing one or a number of symptoms such as pain, spontaneous dislocation and nerve palsy [4]. The prevalence of pseudotumours in MoMHRA patients is low but revision, for which the outcome is poor [5], is required in at least 70% of these cases [4]. The cause of pseudotumours in patients with MoMHRA is unknown, but they are associated with high serum and hip aspirate levels of cobalt (Co) and chromium (Cr), the principal elements used in MoMHRA implants [6,7]. The levels of serum Co and Cr ions have been shown to correspond to the amount of linear wear of the femoral component in studies on retrieved MoMHRA components [8–10]. The data would suggest that the occurrence of pseudotumours is related to higher than normal wear of metal-on-metal hip replacement. It was suggested that metal hypersensitivity could be the cause but evidence from lymphocyte proliferation assays does not support this in the majority of cases, however it is probably involved in a small percentage of cases [11]. Edge-loading is a situation in which the femoral component comes into contact with the edge of the acetabular component
1350-4533/$ – see front matter © 2011 IPEM. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.medengphy.2011.05.011
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
1213
Table 1 Subject characteristics including serum metal ion levels. Subject
Gender
Body mass (kg)
Implant type
Femoral component size (mm)
Chromium (g/l)
Cobalt (g/l)
1 2 3 4
Male Female Female Female
85.2 89.2 81.3 56.1
BHR BHR C+ BHR
50 42 42 42
0.9 1.5 13.8 45.2
1.9 1.6 15.2 64.1
and is thought to be a cause of increased wear that leads to high levels of serum Co and Cr ions [8–10,12]. Low metal-on-metal (MoM) wear rates associated with normal loading conditions are attributable to fluid film lubrication [13]. Edge-loading disrupts the fluid-film lubrication between components which can lead to marked localised wear [14]. Steeply inclined acetabular components are thought to be at greater risk of edge-loading. Patients who had acetabular components with inclination angles greater than 55◦ have been shown to have higher serum levels of Co and Cr ions [9]. Furthermore, during revision of steeply inclined acetabular components, large amounts of metallosis have been found [8,9]. In vitro studies using hip simulators have also demonstrated that wear is increased with increased inclination angle [15,16]. Computational modelling has been used to investigate wear in THR. Contact stress has been shown to vary on the polyethylene component, peaking at the edge, indicating an edge-loading mechanism that may result in increased wear [17]. Wear of MoMHRA has not been examined to the same extent using computational methods. Wear of articulating orthopaedic implants is a function of the material properties of the bearing surfaces, the contact stress, the type of lubrication regime and the kinematics of the joint. Motion analysis has been used previously to examine the kinematics of the hip joint during gait in order to obtain information on the wear path in THR patients [18–20]. Furthermore, in an effort to predict wear of total knee replacements (TKR) knee kinematics and measured joint contact forces were combined in patient-specific computational models [21]. Gait and stair activities were shown to provide enough information to predict wear locations and damage depth consistent with retrieval studies. The aim of this study was to examine the relationship between component orientation, functional activity and contact stress in vivo for a group of MoMHRA patients.
These patients underwent further investigations (which had been approved by the local ethics committee) after giving informed consent. All four participants completed gait and stair activities in our motion analysis laboratory followed by computed tomography (CT) scans. In order to investigate the relationship between serum metal ion levels and contact stress, kinematic data from the motion capture were combined with hip contact forces reported by Bergmann et al. (1998) and subject-specific finite element models were created. 2.2. Motion analysis Three-dimensional lower limb motion analysis was conducted in our motion analysis laboratory. The laboratory is equipped with a 12 camera Vicon Nexus MX system (Oxford Metrics Ltd., Oxford, UK) and 2 OR6 AMTI R6-6-1000 force platforms (Advanced Medical Technology Inc., MA, USA). Twenty-three 10 mm diameter spherical retro reflective markers were attached to anatomical landmarks in an established [22] marker configuration, without greater trochanter markers and with extra markers on the medial femoral condyles of the implanted limb, on the tibial tuberosities and the medial malleolii (Fig. 1). Kinematic data were collected with a sampling rate of 100 Hz and force plate data were collected at 1000 Hz. The subjects’ motion was measured during level walking and stair descent. Each trial was repeated six times. The activities recorded showed a high level of repeatability. This was confirmed, for each subject, by examining the variance in kinematic parameters between trials. The subject’s masses and anatomical parameters were recorded.
2. Methods 2.1. Overview As part of an on-going study, a large cohort of MoMHRA patients had their serum metal ion levels measured. In this study, approved by the local ethics committee, there were 158 subjects (201 hips). Twenty-one patients were chosen from this 158 for motion analysis and computed tomography (CT) scans and, based on their metal ion levels, 4 subjects were then chosen for the current study (n = 4). The four subjects had no other joint replacements. Two subjects (Subjects 1 and 2) with low serum metal ion levels and two subjects with high serum metal ions (Subjects 3 and 4) were chosen in order to determine whether any differences in contact stress existed. Subjects 1 and 2 had chromium ion (Cr) levels of 0.9 g/l and 1.5 g/l, and cobalt ion (Co) levels of 1.9 g/l and 1.6 g/l respectively. Subject 3 had Cr and Co ion levels of 13.8 g/l and 15.2 g/l respectively. Subject 4, who had bilateral prostheses, had the highest Cr and Co ion levels at 45.2 g/l and 64.1 g/l respectively. The subjects had been implanted with either the Birmingham Hip Resurfacing (BHR) (Smith and Nephew, Birmingham, UK) or the Conserve Plus (C+) (Wright Medical Technology, Arlington, TN, USA) devices (Table 1). Subject 4 had bilateral prostheses.
Fig. 1. Motion analysis was conducted using an adapted Helen Hayes [22] marker set (pelvis and right leg markers).
1214
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
2.3. Computed tomography (CT) scans Directly following the motion analysis, computed tomography (CT) scans (Siemens Somatom, Siemens Medical Solutions USA, Inc., NY, USA) of each subject’s pelvis and lower limbs were obtained using a metal artefact reduction sequence. In order to ensure registration of internal implant positions to the skin based motion analysis markers, the retro-reflective motion analysis markers were replaced with radio-opaque markers. These multi-modality MM3002 markers (Intermark Medical Innovations, Ltd., Bromley, Kent, UK) consisted of a hydrogel component with a medical grade adhesive. SliceOmatic (Version 4.2 Rev-9b, TomoVision, Virtual Magic Inc., Montreal, Canada) was used to determine the 3D coordinates of multi-modality markers, the anatomical pelvic landmarks, the MoMHRA prosthesis components and the centre of the hip joint within the CT coordinate system. 2.4. Subject specific hip contact force analysis The intersection of the body mass scaled hip contact force vector with the acetabular component was determined for gait and stair descent. This required a combination of the CT and motion capture data, from the four subjects recruited for this study, with reported hip contact forces from telemetered instrumented total hip prostheses [23]. The hip contact force reported for subject ‘KW’ was used in this analysis as his ground reaction force during gait was most similar to that of the four subjects. The hip contact forces (HCF) reported by Bergmann et al. (HCFBERG ) were in a lefthanded femoral coordinate system with the origin at the centre of the femoral head. These were converted to a right-handed femoral coordinate system. The motion capture data for each subject during stair descent and gait were used to determine an identical right-handed femoral coordinate system where the hip joint centre (HJC), determined from the CT scans, was assumed as the centre of the femoral head. Transformation matrices ([TF→P ]i ) required to convert forces from this femoral coordinate system to a pelvic coordinate system were determined for every 2% of the gait cycle (GC). The hip contact forces were then scaled using each subject’s body mass (HCFS ). The scaled hip forces were transformed from the femoral coordinate system to a pelvic coordinate system (HCFSPi ) relative to each subject’s pelvis. An acetabular coordinate system was determined for each subject using points defined from the edge of the acetabular component in the CT images. The hip joint centre was also the origin for the acetabular coordinate system. A normal to the face of the acetabular component passing through the origin was used to determine the anatomical inclination and anteversion of the acetabular component in the coronal and saggital planes. The transformation matrix for pelvic to acetabular coordinates ([TP→A ]i ) was calculated and used to transform the hip contact force (HCFSA ) into the acetabular component reference frame. The location of the intersection of this transformed hip contact force with the inner surface of the acetabular component was then calculated. This vector was used as the input to the subject-specific FE models. In addition, the angle () between the hip contact force vector and the edge of the acetabular component was determined for each patient for stair descent and over the gait cycle (Fig. 2). The angle the hip contact force vector makes in the frontal plane determines the intersection of the force with the acetabular component relative to the acetabular component’s inclination angle. In order to get an indication of the variation in this angle, the hip contact force vectors reported for all four subjects in the study by Bergmann et al. were scaled to each of the four subjects in the present study and analysed for gait and stair descent. The angle this vector made in the frontal plane during 52% stair descent, 14%
Fig. 2. The angle () between the hip contact force vector and an edge vector and the coverage angle (˛).
GC and 60% GC was recorded. The variation in this angle was used as a guide for determining the upper limit of incremental increase in acetabular component inclination modelled during finite element analysis. All analyses were carried out using custom software developed in Matlab (The MathWorks Inc., Natick, MA, USA). 2.5. Finite element analysis Finite element (FE) models of the acetabular and femoral components were constructed, on a subject specific basis. Each subject’s individual acetabular component orientation was modelled. All C+ implants were modelled with an acetabular component with a coverage angle (˛) of 170◦ and a diametrical clearance of 173 m [8]. The coverage angle for the BHR acetabular component was dependent on the size of the implant and varied from 159.1◦ to 166.2◦ (Board & Walter, BHS 2009). The diametrical clearance used for the BHR was 271 m [8]. All acetabular components were modelled with a filleted edge of radius 1.58 mm (based on measurements made on CAD data). Volumes were meshed with 10 node tetrahedral elements. The force vector representing the hip contact force, scaled to each subject’s mass, was tethered to a pilot node on the femoral component defined at the hip joint centre. The pilot node was defined with a rigid surface constraint meaning the contact nodes were constrained to the rigid body motion defined by the pilot node. Contact was detected on Gauss points and friction was set at 0.08 [24]. The gap between the femoral and acetabular components was closed automatically by the solution method. The femoral component was modelled as the target surface and the acetabular component was the contact surface. The target surface was modelled as marginally stiffer than the contact surface to aid convergence. The acetabular components were, therefore, modelled with a Young’s modulus of 200 GPa while the femoral components had an elastic modulus marginally higher at 210 GPa. Each acetabular component was bonded to a substrate, with a Young’s modulus of 17 GPa, representing the bone in the acetabulum [25]. The contact problem was solved using the augmented Lagrange method (ANSYS 11.0, ANSYS Inc., Canonsburg, PA, USA).
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220 Table 2 Orientation of each subject’s acetabular component. Subject
1 2 3 4
1215
Table 3 Mean variation in angle of scaled hip contact force vector in the frontal plane.
Radiographic inclination
Calculated anatomical inclination
Calculated anatomical anteversion
45.6 46.5 60.2 53.7
39.4◦ 46.7◦ 52.5◦ 52.9◦
31.0◦ 20.7◦ 21.8◦ 32.1◦
14% GC
60% GC
52% stair descent
7.7
19.7
11.1
the edge of the acetabular component ranged from 10.2◦ for Subject 1 to 3.4◦ for Subject 3. 3.3. Inter-subject force orientation variability
For each subject, models representing the instances of peak force and edge-loading were constructed. For gait, the peak hip joint force occurred at 14% of the gait cycle (GC). For stair descent, the peak force was at 52% of the activity. During the stance phase of gait, 60% GC represented the point at which the hip contact force consistently came closest to the edge of the acetabular component. Thus for each subject, the orientations of the components at 14% GC, 60% GC and 52% stair descent were separately modelled. For each model, the effect of increasing the inclination of the acetabular component, by up to 10◦ in 1◦ increments, was determined. The resultant contact stress for each angle was recorded. To aid comparison between subjects, the contact stress was normalised by each subject’s body mass and graphed. 2.6. Subject comparison: individual activity pattern versus acetabular orientation Finite element models were created in order to further investigate the relationship between functional activities, acetabular component orientation, contact stress and metal ion levels. In these models the kinematics and scaled HCF for Subject 1 (low metal ion levels) were combined with Subject 3’s (high metal ion levels) acetabular component orientation and size. Models were created for 14% GC and 52% stair descent and for each model, the effect of increasing the inclination of the acetabular component, by up to 10◦ in 1◦ increments, was determined. The normalised contact stress for these models were compared with Subject 3’s data. 3. Results 3.1. Orientation of acetabular component Table 2 shows the inclination as measured from radiographs as well as the calculated anatomical inclination and anteversion. Subject 1 had an acetabular component with an inclination of 39.4◦ and anteversion of 31◦ . Subject 2 had an acetabular component inclination of 46.7◦ and anteversion of 20.7◦ . Subject 3 had an acetabular component inclination of 52.5◦ and an anteversion of 21.8◦ . Subject 4 had an acetabular component inclination of 52.9◦ and an anteversion of 32.1◦ . 3.2. Subject-specific hip contact force analysis At 14% GC the angle between the hip contact force vector and the edge of the acetabular component was greater than 20◦ for Subjects 1 and 2. In contrast, the intersection point was 8.6◦ from the edge for Subject 4 and 11.4◦ from the edge for Subject 3 (Fig. 3). The trend for 52% of stair descent was similar to 14% GC. Subjects 1 and 2 had hip contact force vectors that were more than 25◦ from the edge, and Subjects 3 and 4 had intersection points that were 10.6◦ and 7.2◦ , respectively, from the edge (Fig. 4). Fig. 3 also shows the angle between the hip contact force vector and the edge of the acetabular component at 60% GC. The force vector was closer to the edge here than at any other time during the stance phase of gait (0–60% GC) for all the participants. The angle between the intersection of the hip contact force vector and
At 14% GC the mean inter-subject variation for the angle of the hip contact force vector reported by Bergmann et al. (1998) in the frontal plane was 7.7◦ (Table 3). At 60% GC this variation was 19.7◦ . At peak force during stair descent, the mean inter-subject variation was 11.1◦ . 3.4. FE analysis At 14% GC, the contact stresses on the acetabular components of Subjects 1 and 2 remained unchanged for all analysed orientations (Fig. 5). In contrast, the contact stress on the acetabular components of Subject 3 and 4 increased when the inclination was increased. The contact stress on the acetabular component of Subject 3 went from 39.9 MPa in the native orientation to 123.1 MPa when inclination was increased by 10◦ . It only required a 5◦ increase in the acetabular component inclination of Subject 4 to increase the contact stress from 47.2 MPa in the native orientation to 138.9 MPa. For Subject 4, the FE model would not converge for increments in acetabular inclination of 6◦ or more. At 60% GC, all subjects acetabular components demonstrated increases in contact stress as the orientation was altered (Fig. 6). The models representing the acetabular components of Subjects 1 and 2 were found to be edge-loading with the addition of 7◦ and 5◦ to their respective native inclination angles. The contact stress on the acetabular component of Subject 3 was 59.4 MPa at the native inclination angle, which increased to 115.8 MPa with the addition of just 1◦ . The contact stress on the acetabular component of Subject 4 increased by 95% when 2◦ was added to the inclination angle. At peak hip contact force for stair descent (52%) the contact stresses on the acetabular components of Subjects 1 and 2 remained unchanged for all analysed orientations (Fig. 5). Subject 3 showed signs of edge-loading, with an increase of 108% in contact stress, when 9◦ was added to the native inclination angle. The contact stress on the acetabular component of Subject 4 increased from 63.2 MPa at her native orientation to 112.2 MPa when 2◦ was added to the inclination angle (Fig. 7). 3.5. Subject comparison: individual activity versus acetabular orientation At 14% GC the contact stress on the acetabular component of the combined model remained unchanged for all the analysed orientations (Fig. 8). In contrast the contact stress on Subject 3’s acetabular component rose dramatically between 7◦ and 10◦ indicating edgeloading. At 52% of stair descent, the contact stress on the acetabular component of the combined model remained unchanged for every angle analysed (Fig. 9). The contact stress on Subject 3’s acetabular component began rising after 4◦ was added to the inclination angle. 4. Discussion The results in this study suggest that a steeply inclined acetabular component influences the occurrence of edge-loading and this supports the findings of previous studies [26,27]. The overriding view in the orthopaedic community is that recommended values
1216
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
Fig. 3. The hip contact force (% body weight) over the gait cycle and the corresponding angle between the hip contact force vector and the edge of the acetabular component (native orientation) during gait.
for component position can be applied generally to patient cohorts. However, in addition to the importance of component position, we have shown that an individual’s activity patterns have the potential to reduce or override the effects of a steeply inclined acetabular component. The FE analysis indicated that contact stress on the acetabular component increases by at least 50% under edgeloading conditions. Previously, edge-loading has been suggested as the mechanism that causes increased wear of MoMHRAs [8–10] giving rise to elevated levels of serum metal ions [28]. Edge-loading disrupts the fluid film lubrication mechanism in MoMHRA [29] and inadequate lubrication has been shown to be associated with increased wear [30]. Subjects 3 and 4 had high metal ion levels and required the least amount of incremental increase to their acetabular inclination angles to cause edge-loading at 14% GC, 60% GC and stair descent. This result assumes more significance when one considers that Subjects 1 and 2, who had low levels of serum metal ions, do not edge-load at all when up to 10◦ is added to their native inclination angles at 14% of GC and stair descent. In addi-
tion, the inter-subject variability in the angle created by the hip contact force vector in the frontal plane for 14% GC (7.7◦ ), 60% GC (19.7◦ ) and stair descent (11.1◦ ) was less than the increase in inclination angle required to induce edge-loading for Subjects 3 and 4. The blood metal ion levels of Subjects 3 and 4 are consistent with previous studies that have correlated metal ion levels with steep inclination angles. De Haan et al. (2008) showed that serum metal ions in 20% of women who had steep acetabular components (inclination angles of 55◦ or greater) were above 30 g/l, whereas only 2% of men with steep components had serum metal ions at this concentration or above [9]. This could be attributable to a number of different reasons such as women having smaller sized components or inherent differences in the activity patterns of men and women. De Haan et al. (2008) measured the inclination angle using anterior–posterior radiographs. In this study, the only subject who had a steep acetabular inclination according to this method of measurement was Subject 3 (60.2◦ ). Subject 4 was
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
1217
Fig. 4. The hip contact force (% body weight) over the activity cycle and the corresponding angle between the hip contact force vector and the edge of the acetabular component (native orientation) during stair descent.
measured radiographically as having an inclination of 53.7◦ and a calculated anatomical inclination of 62.2◦ . Anatomical inclination differs from radiographic inclination due to the effect of anteversion on measuring inclination from radiographs [31]. It is also worth noting that Subject 4’s anatomical anteversion (32.1◦ ) was outside the recommended zone (20◦ ± 10◦ ). Subjects 1 and 2 were both below 55◦ in their native radiographic and anatomical inclinations. However, adding 10◦ to the inclination angle of the acetabular component of Subjects 1 and 2 put them at inclination angles previously associated with high serum metal ion levels. The contact stress, at these angles, was unaltered at 14% and stair descent suggesting that the individual activity patterns play a role in whether or not edge-loading occurs. The subject comparison, in which Subject 3’s acetabular component orientation was used with Subject 1’s hip contact forces and kinematics, further highlighted the importance of individual activity patterns to the occurrence of edge-loading. The combined model showed no signs of edge-loading for 14% GC or 52% stair descent.
In contrast, the contact stress on Subject 3’s acetabular component increased when 7◦ and 6◦ was added to the inclination angle at 14% GC and 52% stair descent respectively. This suggests that Subject 3’s acetabular component orientation combined with Subject 1’s individual activity patterns decreased the likelihood of edgeloading occurring. This indicates that individual activity patterns have a significant effect on wear and metal ion levels. The largest variability in angles made by the hip contact force vector in the frontal plane was at 60% of GC, which represents the end of the stance phase of gait when the foot is about to leave the floor. At 60% of GC, all the subjects were edge-loading with the addition of a maximum of 8◦ to their native inclination angles. The magnitude of the hip contact force at 60% is less than half the magnitude at peak force, and therefore it can be argued that it is less serious for subjects to edge-load at this point. The magnitude of peak hip contact force during gait and activities such as descending stairs have been shown to be 243% and 316% of body weight respectively [23].
1218
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
Fig. 5. Body mass adjusted contact stress for subjects at 14% of gait cycle.
The angle between the hip contact force vector and the edge of the acetabular component at 14% GC was 11.4◦ for Subject 3 and 8.6◦ for Subject 4. The FE models for Subjects 3 and 4 demonstrated edge-loading with the addition of 10◦ and 3◦ respectively. A similar discrepancy occurred for the same subjects at 52% stair descent. This suggests that for the FE models there was a zone, approximately 5◦ from the edge of the acetabular component, where edge-loading occurred. It is likely that such a zone exists at the edge of the acetabular component in which contact between the
femoral and acetabular components disrupts fluid-film lubrication and increased contact stress occurs; thus edge-loading may effectively occur even if the line of action of the joint reaction force passes through the acetabular component away from the edge. The radial clearance between the acetabular and femoral components permits fluid film lubrication between the components which increases the contact area and reduces wear. Edge-loading is thought to disturb this fluid film between the components resulting in localised metal-on-metal wear. The FE analysis in this study
Fig. 6. Body mass adjusted contact stress for subjects at 60% of gait cycle.
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
1219
Fig. 7. Body mass adjusted contact stress for subjects at 52% of downstairs activity.
did not incorporate a fluid layer between the components and this represents another limitation. It is thought that the addition of a fluid layer between the components, except at the edge of the acetabular component, would increase the differences in contact stress between normal and edge-loading conditions. In the absence of fluid lubrication it is likely that the friction between the components increases during edge-loading. This also represents a limitation in the current study as the inclusion of an accurate representation of this friction would alter the contact stress on the surface of the acetabular component. The subjects chosen for this study all had one MoMHRA with the exception of Subject 4 who had bilateral BHRs. At the time of analysis Subjects 1, 3 and 4 were approximately 2 years post-surgery. Subject 2 was 4 years post-surgery. Subject 4’s exceptionally high metal ion levels are influenced to an unknown extent because she has bilateral resurfacings. Subject 4 has since had both MoMHRAs
revised for psuedotumour and replaced with conventional total hip replacements. A limitation of this study was that the relationship between activity patterns during gait and stair descent, the acetabular orientation and serum metal ion levels were determined using hip contact forces from another study. The extent of the relationship between metal ion levels and hip loading could be further determined if hip contact forces could be calculated/estimated on a subject-specific basis. In conclusion, edge-loading increased contact stress by at least 50%. Patients with low serum metal ion levels showed no increase in contact stress with increasing component inclination at peak force during gait or stair descent. Patients with high serum metal ion levels exhibited edge-loading with a small increase to the inclination angle of their acetabular components that was less than the intersubject variability in the angle of published hip contact forces. This
Fig. 8. Body mass adjusted contact stress for combined and Subjects 1 and 3 at 14% of gait cycle.
Fig. 9. Body mass adjusted contact stress for combined and Subjects 1 and 3 at 52% of downstairs activity.
1220
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
suggests that serum metal ion levels are influenced by acetabular orientation and individual activity patterns. A future study will utilise the techniques in this study to examine absolute cup position relative to edge-loading depending on acetabular component design. Future work will also look at the influence of acetabular anteversion on contact stress during functional activities and will incorporate subject specific hip reaction forces calculated using a musculoskeletal model. Acknowledgement The authors wish to thank Dr. David Simpson for assistance with the FE modelling. Conflict of interest The work described in this paper contains no conflict of interest. References [1] Quesada MJ, Marker DR, Mont MA. Metal-on-metal hip resurfacing: advantages and disadvantages. Journal of Arthroplasty 2008;23(7 (Suppl. 1)):69–73. [2] National Joint Registry for England and Wales; 2nd annual report; 2005. http://www.njrcentre.org.uk/NjrCentre/Portals/0/Documents/England/ Reports/NJR AR 2.pdf. [3] Hart AJ, Sabah S, Henckel J, Lewis A, Cobb J, Sampson B, et al. The painful metalon-metal hip resurfacing. Journal of Bone and Joint Surgery – British Volume 2009;91B(6):738–44. [4] Pandit H, Glyn-Jones S, McLardy-Smith P, Gundle R, Whitwell D, Gibbons CLM, et al. Pseudotumours associated with metal-on-metal hip resurfacings. Journal of Bone and Joint Surgery – British Volume 2008;90-B(7):847–51. [5] Grammatopolous G, Pandit H, Kwon YM, Gundle R, McLardy-Smith P, Beard DJ, et al. Hip resurfacings revised for inflammatory pseudotumour have a poor outcome. Journal of Bone and Joint Surgery – British Volume 2009;91B(8):1019–24. [6] Kwon YM, Ostlere S, McLardy-Smith P, Gundle R, Whitwell D, Gibbons CLM, et al. Metal ion levels in pseudotumours associated with metal-on-metal hip resurfacings. In: The 55th orthopaedic research society annual meeting. Las Vegas, USA; 2009. [7] Langton DJ, Sprowson AP, Joyce TJ, Reed M, Carluke I, Partington P, et al. Blood metal ion concentrations after hip resurfacing arthroplasty. A comparative study of articular surface replacement and Birmingham hip resurfacing arthroplasties. Journal of Bone and Joint Surgery – British Volume 2009;91B(10):1287–95. [8] Campbell P, Beaule PE, Ebramzadeh E, LeDuff M, De Smet K, Lu Z, et al. The John Charnley Award: a study of implant failure in metal-on-metal surface arthroplasties. Clinical Orthopaedics and Related Research 2006;453:35–46. [9] De Haan R, Pattyn C, Gill HS, Murray DW, Campbell PA, De Smet K. Correlation between inclination of the acetabular component and metal ion levels in metal-on-metal hip resurfacing replacement. Journal of Bone and Joint Surgery – British Volume 2008;90B(10):1291–7. [10] de Smet K, de Haan R, Calistri A, Campbell PA, Ebramzadeh E, Pattyn C, et al. Metal ion measurement as a diagnostic tool to identify problems with metalon-metal hip resurfacing. Journal of Bone and Joint Surgery – American Volume 2008;90A:202–8. [11] Young-Min K, Peter T, Burkhard S, Hemant P, Adrian T, David B, et al. Lymphocyte proliferation responses in patients with pseudotumors following metal-on-metal hip resurfacing arthroplasty. Journal of Orthopaedic Research 2009.
[12] Morlock MM, Bishop N, Zustin J, Hahn M, Ruther W, Amling M. Modes of implant failure after hip resurfacing: morphological and wear analysis of 267 retrieval specimens. Journal of Bone and Joint Surgery – American Volume 2008;90A:89–95. [13] Chan FWP, Bobyn JDP, Medley JBP, Krygier JJCET, Tanzer MMD. Wear and lubrication of metal-on-metal hip implants. Clinical Orthopaedics and Related Research 1999;369:10–24. [14] Liu F, Jin ZM, Roberts P, Grigoris P. Importance of head diameter, clearance, and cup wall thickness in elastohydrodynamic lubrication analysis of metal-onmetal hip resurfacing prostheses. Proceedings of the Institution of Mechanical Engineers Part H – Journal of Engineering in Medicine 2006;220(H6): 695–704. [15] Leslie I, Williams S, Isaac G, Ingham E, Fisher J. High cup angle and microseparation increase the wear of hip surface replacements. Clinical Orthopaedics and Related Research 2009;467(9):2259–65. [16] Williams S, Leslie I, Isaac G, Jin ZM, Ingham E, Fisher J. Tribology and wear of metal-on-metal hip prostheses: influence of cup angle and head position. Journal of Bone and Joint Surgery – American Volume 2008;90A:111–7. [17] Barbour PSM, Barton DC, Fisher J. The influence of contact stress on the wear of UHMWPE for total replacement hip prostheses. Wear 1995;181–183(1): 250–7. [18] Bennett D, Humphreys L, O’Brien S, Kelly C, Orr J, Beverland DE. The influence of wear paths produced by hip replacement patients during normal walking on wear rates. Journal of Orthopaedic Research 2008;26(9):1210–7. [19] Bennett D, Humphreys L, O’Brien S, Kelly C, Orr JF, Beverland DE. Gait kinematics of age-stratified hip replacement patients – a large scale, long-term follow-up study. Gait & Posture 2008;28(2): 194–200. [20] Bennett D, Humphreys L, O’Brien S, Kelly C, Orr JF, Beverland DE. Wear paths produced by individual hip-replacement patients – a large-scale, long-term follow-up study. Journal of Biomechanics 2008;41(11):2474–82. [21] Fregly BJ, Sawyer WG, Harman MK, Banks SA. Computational wear prediction of a total knee replacement from in vivo kinematics. Journal of Biomechanics 2005;38(2):305–14. [22] Kadaba MP, Ramakrishnan HK, Wootten ME. Measurement of lower extremity kinematics during level walking. Journal of Orthopaedic Research 1990;8(3):383–92. [23] Bergmann G, Deuretzbacher G, Heller M, Graichen F, Rohlmann A, Strauss J, et al. Hip contact forces and gait patterns from routine activities. Journal of Biomechanics 2001;34(7):859–71. [24] Vassiliou K, Elfick APD, Scholes SC, Unsworth A. The effect of ‘running-in’ on the tribology and surface morphology of metal-on-metal Birmingham hip resurfacing device in simulator studies. Proceedings of the Institution of Mechanical Engineers Part H – Journal of Engineering in Medicine 2006;220(H2): 269–77. [25] Dalstra M, Huiskes R, Vanerning L. Development and validation of a 3dimensional finite-element model of the pelvic bone. Journal of Biomechanical Engineering – Transactions of the Asme 1995;117(3):272–8. [26] Brodner W, Grubl A, Jankovsky R, Meisinger V, Lehr S, Gottsauner-Wolf F. Cup inclination and serum concentration of cobalt and chromium after metal-onmetal total hip arthroplasty. Journal of Arthroplasty 2004;19(8):66–70. [27] De Haan R, Campbell PA, Su EP, De Smet KA. Revision of metal-onmetal resurfacing arthroplasty of the hip – the influence of malpositioning of the components. Journal of Bone and Joint Surgery – British Volume 2008;90B(9):1158–63. [28] Zahiri CA, Schmalzried TP, Szuszczewicz ES, Amstutz HC. Assessing activity in joint replacement patients. Journal of Arthroplasty 1998;13(8):890–5. [29] Barrett ARW, Davies BL, Gomes M, Harris SI, Henckel J, Jakopec M, et al. Computer-assisted hip resurfacing surgery using the Acrobot (R) Navigation System. Proceedings of the Institution of Mechanical Engineers Part H – Journal of Engineering in Medicine 2007;221(H7):773–85. [30] Udofia IJ, Jin ZM. Elastohydrodynamic lubrication analysis of metal-on-metal hip-resurfacing prostheses. Journal of Biomechanics 2003;36(4):537–44. [31] Murray DW. The definition and measurement of acetabular orientation. Journal of Bone and Joint Surgery – British Volume 1993;75(2):228–32.
Contents lists available at ScienceDirect
Medical Engineering & Physics journal homepage: www.elsevier.com/locate/medengphy
The effect of motion patterns on edge-loading of metal-on-metal hip resurfacing S.J. Mellon ∗ , Y.-M. Kwon, S. Glyn-Jones, D.W. Murray, H.S. Gill Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, UK
a r t i c l e
i n f o
Article history: Received 10 August 2010 Received in revised form 18 May 2011 Accepted 19 May 2011 Keywords: Biomechanics Finite element analysis Hip resurfacing arthroplasty
a b s t r a c t The occurrence of pseudotumours (soft tissue masses relating to the hip joint) following metal-on-metal hip resurfacing arthroplasty (MoMHRA) has been associated with high serum metal ion levels and consequently higher than normal bearing wear. We investigated the relationship between serum metal ion levels and contact stress on the acetabular component of MoMHRA patients for two functional activities; gait and stair descent. Four subjects with MoMHRA, who had their serum metal ion levels measured, underwent motion analysis followed by CT scanning. Their motion capture data was combined with published hip contact forces and finite element models representing 14% (peak force) and 60% (end of stance) of the gait cycle and 52% (peak force) of stair descent activity were created. The inclination angle of the acetabular component was increased by 10◦ in 1◦ intervals and the contact stresses were determined at each interval for each subject. When the inclination angle was altered in such a way as to cause the hip contact force to pass through the edge of the acetabular component edge-loading occurred. Edge-loading increased the contact stress by at least 50%; the maximum increase was 108%. Patients with low serum metal ion levels showed no increase in contact stress at peak force during gait or stair descent. Patients with high serum metal ion levels exhibited edge-loading with an increase to the inclination angle of their acetabular components. The increase in inclination angle that induced edge-loading for these subjects was less than the intersubject variability in the angle of published hip contact forces. The results of this study suggest that high serum metal ion levels are the result of inclination angle influenced edge-loading but that edge-loading cannot be attributed to inclination angle alone and that an individual’s activity patterns can reduce or even override the influence of a steep acetabular component and prevent edge-loading. © 2011 IPEM. Published by Elsevier Ltd. All rights reserved.
1. Introduction The resurgence in popularity of metal-on-metal hip resurfacing arthroplasty (MoMHRA) is due, in part, to a lower wear rate of these devices which has the potential to reduce the risk of long-term wear induced osteolysis associated with the failure of metal-on-plastic (MoP) bearings. Further advantages of MoMHRA include lower risk of dislocation, a greater degree of bone conservation and restoration of anatomical hip mechanics [1]. MoMHRA is therefore viewed as a treatment option more suited to young and active patients with hip osteoarthritis. Hip resurfacing arthroplasties carried out on patients 55 years or younger accounted for approximately 46% of primary hip arthroplasties performed in the UK on this age group in 2004 [2]. The majority of MoMHRA patients have a longer life expectancy when compared with elderly patients implanted with conventional total hip replacements (THR). The effects of long-term exposure to metal debris from metalon-metal wear are largely unknown. There is, therefore, heightened
∗ Corresponding author. Tel.: +44 (0)1865 227454. E-mail address: [email protected] (S.J. Mellon).
concern over periprosthetic soft tissue masses observed in a small percentage of patients with MoMHRA [3,4]. These masses have been termed pseudotumours due to their morphological similarity to necrotic tumours. Pseudotumours can be extremely destructive, causing one or a number of symptoms such as pain, spontaneous dislocation and nerve palsy [4]. The prevalence of pseudotumours in MoMHRA patients is low but revision, for which the outcome is poor [5], is required in at least 70% of these cases [4]. The cause of pseudotumours in patients with MoMHRA is unknown, but they are associated with high serum and hip aspirate levels of cobalt (Co) and chromium (Cr), the principal elements used in MoMHRA implants [6,7]. The levels of serum Co and Cr ions have been shown to correspond to the amount of linear wear of the femoral component in studies on retrieved MoMHRA components [8–10]. The data would suggest that the occurrence of pseudotumours is related to higher than normal wear of metal-on-metal hip replacement. It was suggested that metal hypersensitivity could be the cause but evidence from lymphocyte proliferation assays does not support this in the majority of cases, however it is probably involved in a small percentage of cases [11]. Edge-loading is a situation in which the femoral component comes into contact with the edge of the acetabular component
1350-4533/$ – see front matter © 2011 IPEM. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.medengphy.2011.05.011
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
1213
Table 1 Subject characteristics including serum metal ion levels. Subject
Gender
Body mass (kg)
Implant type
Femoral component size (mm)
Chromium (g/l)
Cobalt (g/l)
1 2 3 4
Male Female Female Female
85.2 89.2 81.3 56.1
BHR BHR C+ BHR
50 42 42 42
0.9 1.5 13.8 45.2
1.9 1.6 15.2 64.1
and is thought to be a cause of increased wear that leads to high levels of serum Co and Cr ions [8–10,12]. Low metal-on-metal (MoM) wear rates associated with normal loading conditions are attributable to fluid film lubrication [13]. Edge-loading disrupts the fluid-film lubrication between components which can lead to marked localised wear [14]. Steeply inclined acetabular components are thought to be at greater risk of edge-loading. Patients who had acetabular components with inclination angles greater than 55◦ have been shown to have higher serum levels of Co and Cr ions [9]. Furthermore, during revision of steeply inclined acetabular components, large amounts of metallosis have been found [8,9]. In vitro studies using hip simulators have also demonstrated that wear is increased with increased inclination angle [15,16]. Computational modelling has been used to investigate wear in THR. Contact stress has been shown to vary on the polyethylene component, peaking at the edge, indicating an edge-loading mechanism that may result in increased wear [17]. Wear of MoMHRA has not been examined to the same extent using computational methods. Wear of articulating orthopaedic implants is a function of the material properties of the bearing surfaces, the contact stress, the type of lubrication regime and the kinematics of the joint. Motion analysis has been used previously to examine the kinematics of the hip joint during gait in order to obtain information on the wear path in THR patients [18–20]. Furthermore, in an effort to predict wear of total knee replacements (TKR) knee kinematics and measured joint contact forces were combined in patient-specific computational models [21]. Gait and stair activities were shown to provide enough information to predict wear locations and damage depth consistent with retrieval studies. The aim of this study was to examine the relationship between component orientation, functional activity and contact stress in vivo for a group of MoMHRA patients.
These patients underwent further investigations (which had been approved by the local ethics committee) after giving informed consent. All four participants completed gait and stair activities in our motion analysis laboratory followed by computed tomography (CT) scans. In order to investigate the relationship between serum metal ion levels and contact stress, kinematic data from the motion capture were combined with hip contact forces reported by Bergmann et al. (1998) and subject-specific finite element models were created. 2.2. Motion analysis Three-dimensional lower limb motion analysis was conducted in our motion analysis laboratory. The laboratory is equipped with a 12 camera Vicon Nexus MX system (Oxford Metrics Ltd., Oxford, UK) and 2 OR6 AMTI R6-6-1000 force platforms (Advanced Medical Technology Inc., MA, USA). Twenty-three 10 mm diameter spherical retro reflective markers were attached to anatomical landmarks in an established [22] marker configuration, without greater trochanter markers and with extra markers on the medial femoral condyles of the implanted limb, on the tibial tuberosities and the medial malleolii (Fig. 1). Kinematic data were collected with a sampling rate of 100 Hz and force plate data were collected at 1000 Hz. The subjects’ motion was measured during level walking and stair descent. Each trial was repeated six times. The activities recorded showed a high level of repeatability. This was confirmed, for each subject, by examining the variance in kinematic parameters between trials. The subject’s masses and anatomical parameters were recorded.
2. Methods 2.1. Overview As part of an on-going study, a large cohort of MoMHRA patients had their serum metal ion levels measured. In this study, approved by the local ethics committee, there were 158 subjects (201 hips). Twenty-one patients were chosen from this 158 for motion analysis and computed tomography (CT) scans and, based on their metal ion levels, 4 subjects were then chosen for the current study (n = 4). The four subjects had no other joint replacements. Two subjects (Subjects 1 and 2) with low serum metal ion levels and two subjects with high serum metal ions (Subjects 3 and 4) were chosen in order to determine whether any differences in contact stress existed. Subjects 1 and 2 had chromium ion (Cr) levels of 0.9 g/l and 1.5 g/l, and cobalt ion (Co) levels of 1.9 g/l and 1.6 g/l respectively. Subject 3 had Cr and Co ion levels of 13.8 g/l and 15.2 g/l respectively. Subject 4, who had bilateral prostheses, had the highest Cr and Co ion levels at 45.2 g/l and 64.1 g/l respectively. The subjects had been implanted with either the Birmingham Hip Resurfacing (BHR) (Smith and Nephew, Birmingham, UK) or the Conserve Plus (C+) (Wright Medical Technology, Arlington, TN, USA) devices (Table 1). Subject 4 had bilateral prostheses.
Fig. 1. Motion analysis was conducted using an adapted Helen Hayes [22] marker set (pelvis and right leg markers).
1214
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
2.3. Computed tomography (CT) scans Directly following the motion analysis, computed tomography (CT) scans (Siemens Somatom, Siemens Medical Solutions USA, Inc., NY, USA) of each subject’s pelvis and lower limbs were obtained using a metal artefact reduction sequence. In order to ensure registration of internal implant positions to the skin based motion analysis markers, the retro-reflective motion analysis markers were replaced with radio-opaque markers. These multi-modality MM3002 markers (Intermark Medical Innovations, Ltd., Bromley, Kent, UK) consisted of a hydrogel component with a medical grade adhesive. SliceOmatic (Version 4.2 Rev-9b, TomoVision, Virtual Magic Inc., Montreal, Canada) was used to determine the 3D coordinates of multi-modality markers, the anatomical pelvic landmarks, the MoMHRA prosthesis components and the centre of the hip joint within the CT coordinate system. 2.4. Subject specific hip contact force analysis The intersection of the body mass scaled hip contact force vector with the acetabular component was determined for gait and stair descent. This required a combination of the CT and motion capture data, from the four subjects recruited for this study, with reported hip contact forces from telemetered instrumented total hip prostheses [23]. The hip contact force reported for subject ‘KW’ was used in this analysis as his ground reaction force during gait was most similar to that of the four subjects. The hip contact forces (HCF) reported by Bergmann et al. (HCFBERG ) were in a lefthanded femoral coordinate system with the origin at the centre of the femoral head. These were converted to a right-handed femoral coordinate system. The motion capture data for each subject during stair descent and gait were used to determine an identical right-handed femoral coordinate system where the hip joint centre (HJC), determined from the CT scans, was assumed as the centre of the femoral head. Transformation matrices ([TF→P ]i ) required to convert forces from this femoral coordinate system to a pelvic coordinate system were determined for every 2% of the gait cycle (GC). The hip contact forces were then scaled using each subject’s body mass (HCFS ). The scaled hip forces were transformed from the femoral coordinate system to a pelvic coordinate system (HCFSPi ) relative to each subject’s pelvis. An acetabular coordinate system was determined for each subject using points defined from the edge of the acetabular component in the CT images. The hip joint centre was also the origin for the acetabular coordinate system. A normal to the face of the acetabular component passing through the origin was used to determine the anatomical inclination and anteversion of the acetabular component in the coronal and saggital planes. The transformation matrix for pelvic to acetabular coordinates ([TP→A ]i ) was calculated and used to transform the hip contact force (HCFSA ) into the acetabular component reference frame. The location of the intersection of this transformed hip contact force with the inner surface of the acetabular component was then calculated. This vector was used as the input to the subject-specific FE models. In addition, the angle () between the hip contact force vector and the edge of the acetabular component was determined for each patient for stair descent and over the gait cycle (Fig. 2). The angle the hip contact force vector makes in the frontal plane determines the intersection of the force with the acetabular component relative to the acetabular component’s inclination angle. In order to get an indication of the variation in this angle, the hip contact force vectors reported for all four subjects in the study by Bergmann et al. were scaled to each of the four subjects in the present study and analysed for gait and stair descent. The angle this vector made in the frontal plane during 52% stair descent, 14%
Fig. 2. The angle () between the hip contact force vector and an edge vector and the coverage angle (˛).
GC and 60% GC was recorded. The variation in this angle was used as a guide for determining the upper limit of incremental increase in acetabular component inclination modelled during finite element analysis. All analyses were carried out using custom software developed in Matlab (The MathWorks Inc., Natick, MA, USA). 2.5. Finite element analysis Finite element (FE) models of the acetabular and femoral components were constructed, on a subject specific basis. Each subject’s individual acetabular component orientation was modelled. All C+ implants were modelled with an acetabular component with a coverage angle (˛) of 170◦ and a diametrical clearance of 173 m [8]. The coverage angle for the BHR acetabular component was dependent on the size of the implant and varied from 159.1◦ to 166.2◦ (Board & Walter, BHS 2009). The diametrical clearance used for the BHR was 271 m [8]. All acetabular components were modelled with a filleted edge of radius 1.58 mm (based on measurements made on CAD data). Volumes were meshed with 10 node tetrahedral elements. The force vector representing the hip contact force, scaled to each subject’s mass, was tethered to a pilot node on the femoral component defined at the hip joint centre. The pilot node was defined with a rigid surface constraint meaning the contact nodes were constrained to the rigid body motion defined by the pilot node. Contact was detected on Gauss points and friction was set at 0.08 [24]. The gap between the femoral and acetabular components was closed automatically by the solution method. The femoral component was modelled as the target surface and the acetabular component was the contact surface. The target surface was modelled as marginally stiffer than the contact surface to aid convergence. The acetabular components were, therefore, modelled with a Young’s modulus of 200 GPa while the femoral components had an elastic modulus marginally higher at 210 GPa. Each acetabular component was bonded to a substrate, with a Young’s modulus of 17 GPa, representing the bone in the acetabulum [25]. The contact problem was solved using the augmented Lagrange method (ANSYS 11.0, ANSYS Inc., Canonsburg, PA, USA).
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220 Table 2 Orientation of each subject’s acetabular component. Subject
1 2 3 4
1215
Table 3 Mean variation in angle of scaled hip contact force vector in the frontal plane.
Radiographic inclination
Calculated anatomical inclination
Calculated anatomical anteversion
45.6 46.5 60.2 53.7
39.4◦ 46.7◦ 52.5◦ 52.9◦
31.0◦ 20.7◦ 21.8◦ 32.1◦
14% GC
60% GC
52% stair descent
7.7
19.7
11.1
the edge of the acetabular component ranged from 10.2◦ for Subject 1 to 3.4◦ for Subject 3. 3.3. Inter-subject force orientation variability
For each subject, models representing the instances of peak force and edge-loading were constructed. For gait, the peak hip joint force occurred at 14% of the gait cycle (GC). For stair descent, the peak force was at 52% of the activity. During the stance phase of gait, 60% GC represented the point at which the hip contact force consistently came closest to the edge of the acetabular component. Thus for each subject, the orientations of the components at 14% GC, 60% GC and 52% stair descent were separately modelled. For each model, the effect of increasing the inclination of the acetabular component, by up to 10◦ in 1◦ increments, was determined. The resultant contact stress for each angle was recorded. To aid comparison between subjects, the contact stress was normalised by each subject’s body mass and graphed. 2.6. Subject comparison: individual activity pattern versus acetabular orientation Finite element models were created in order to further investigate the relationship between functional activities, acetabular component orientation, contact stress and metal ion levels. In these models the kinematics and scaled HCF for Subject 1 (low metal ion levels) were combined with Subject 3’s (high metal ion levels) acetabular component orientation and size. Models were created for 14% GC and 52% stair descent and for each model, the effect of increasing the inclination of the acetabular component, by up to 10◦ in 1◦ increments, was determined. The normalised contact stress for these models were compared with Subject 3’s data. 3. Results 3.1. Orientation of acetabular component Table 2 shows the inclination as measured from radiographs as well as the calculated anatomical inclination and anteversion. Subject 1 had an acetabular component with an inclination of 39.4◦ and anteversion of 31◦ . Subject 2 had an acetabular component inclination of 46.7◦ and anteversion of 20.7◦ . Subject 3 had an acetabular component inclination of 52.5◦ and an anteversion of 21.8◦ . Subject 4 had an acetabular component inclination of 52.9◦ and an anteversion of 32.1◦ . 3.2. Subject-specific hip contact force analysis At 14% GC the angle between the hip contact force vector and the edge of the acetabular component was greater than 20◦ for Subjects 1 and 2. In contrast, the intersection point was 8.6◦ from the edge for Subject 4 and 11.4◦ from the edge for Subject 3 (Fig. 3). The trend for 52% of stair descent was similar to 14% GC. Subjects 1 and 2 had hip contact force vectors that were more than 25◦ from the edge, and Subjects 3 and 4 had intersection points that were 10.6◦ and 7.2◦ , respectively, from the edge (Fig. 4). Fig. 3 also shows the angle between the hip contact force vector and the edge of the acetabular component at 60% GC. The force vector was closer to the edge here than at any other time during the stance phase of gait (0–60% GC) for all the participants. The angle between the intersection of the hip contact force vector and
At 14% GC the mean inter-subject variation for the angle of the hip contact force vector reported by Bergmann et al. (1998) in the frontal plane was 7.7◦ (Table 3). At 60% GC this variation was 19.7◦ . At peak force during stair descent, the mean inter-subject variation was 11.1◦ . 3.4. FE analysis At 14% GC, the contact stresses on the acetabular components of Subjects 1 and 2 remained unchanged for all analysed orientations (Fig. 5). In contrast, the contact stress on the acetabular components of Subject 3 and 4 increased when the inclination was increased. The contact stress on the acetabular component of Subject 3 went from 39.9 MPa in the native orientation to 123.1 MPa when inclination was increased by 10◦ . It only required a 5◦ increase in the acetabular component inclination of Subject 4 to increase the contact stress from 47.2 MPa in the native orientation to 138.9 MPa. For Subject 4, the FE model would not converge for increments in acetabular inclination of 6◦ or more. At 60% GC, all subjects acetabular components demonstrated increases in contact stress as the orientation was altered (Fig. 6). The models representing the acetabular components of Subjects 1 and 2 were found to be edge-loading with the addition of 7◦ and 5◦ to their respective native inclination angles. The contact stress on the acetabular component of Subject 3 was 59.4 MPa at the native inclination angle, which increased to 115.8 MPa with the addition of just 1◦ . The contact stress on the acetabular component of Subject 4 increased by 95% when 2◦ was added to the inclination angle. At peak hip contact force for stair descent (52%) the contact stresses on the acetabular components of Subjects 1 and 2 remained unchanged for all analysed orientations (Fig. 5). Subject 3 showed signs of edge-loading, with an increase of 108% in contact stress, when 9◦ was added to the native inclination angle. The contact stress on the acetabular component of Subject 4 increased from 63.2 MPa at her native orientation to 112.2 MPa when 2◦ was added to the inclination angle (Fig. 7). 3.5. Subject comparison: individual activity versus acetabular orientation At 14% GC the contact stress on the acetabular component of the combined model remained unchanged for all the analysed orientations (Fig. 8). In contrast the contact stress on Subject 3’s acetabular component rose dramatically between 7◦ and 10◦ indicating edgeloading. At 52% of stair descent, the contact stress on the acetabular component of the combined model remained unchanged for every angle analysed (Fig. 9). The contact stress on Subject 3’s acetabular component began rising after 4◦ was added to the inclination angle. 4. Discussion The results in this study suggest that a steeply inclined acetabular component influences the occurrence of edge-loading and this supports the findings of previous studies [26,27]. The overriding view in the orthopaedic community is that recommended values
1216
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
Fig. 3. The hip contact force (% body weight) over the gait cycle and the corresponding angle between the hip contact force vector and the edge of the acetabular component (native orientation) during gait.
for component position can be applied generally to patient cohorts. However, in addition to the importance of component position, we have shown that an individual’s activity patterns have the potential to reduce or override the effects of a steeply inclined acetabular component. The FE analysis indicated that contact stress on the acetabular component increases by at least 50% under edgeloading conditions. Previously, edge-loading has been suggested as the mechanism that causes increased wear of MoMHRAs [8–10] giving rise to elevated levels of serum metal ions [28]. Edge-loading disrupts the fluid film lubrication mechanism in MoMHRA [29] and inadequate lubrication has been shown to be associated with increased wear [30]. Subjects 3 and 4 had high metal ion levels and required the least amount of incremental increase to their acetabular inclination angles to cause edge-loading at 14% GC, 60% GC and stair descent. This result assumes more significance when one considers that Subjects 1 and 2, who had low levels of serum metal ions, do not edge-load at all when up to 10◦ is added to their native inclination angles at 14% of GC and stair descent. In addi-
tion, the inter-subject variability in the angle created by the hip contact force vector in the frontal plane for 14% GC (7.7◦ ), 60% GC (19.7◦ ) and stair descent (11.1◦ ) was less than the increase in inclination angle required to induce edge-loading for Subjects 3 and 4. The blood metal ion levels of Subjects 3 and 4 are consistent with previous studies that have correlated metal ion levels with steep inclination angles. De Haan et al. (2008) showed that serum metal ions in 20% of women who had steep acetabular components (inclination angles of 55◦ or greater) were above 30 g/l, whereas only 2% of men with steep components had serum metal ions at this concentration or above [9]. This could be attributable to a number of different reasons such as women having smaller sized components or inherent differences in the activity patterns of men and women. De Haan et al. (2008) measured the inclination angle using anterior–posterior radiographs. In this study, the only subject who had a steep acetabular inclination according to this method of measurement was Subject 3 (60.2◦ ). Subject 4 was
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
1217
Fig. 4. The hip contact force (% body weight) over the activity cycle and the corresponding angle between the hip contact force vector and the edge of the acetabular component (native orientation) during stair descent.
measured radiographically as having an inclination of 53.7◦ and a calculated anatomical inclination of 62.2◦ . Anatomical inclination differs from radiographic inclination due to the effect of anteversion on measuring inclination from radiographs [31]. It is also worth noting that Subject 4’s anatomical anteversion (32.1◦ ) was outside the recommended zone (20◦ ± 10◦ ). Subjects 1 and 2 were both below 55◦ in their native radiographic and anatomical inclinations. However, adding 10◦ to the inclination angle of the acetabular component of Subjects 1 and 2 put them at inclination angles previously associated with high serum metal ion levels. The contact stress, at these angles, was unaltered at 14% and stair descent suggesting that the individual activity patterns play a role in whether or not edge-loading occurs. The subject comparison, in which Subject 3’s acetabular component orientation was used with Subject 1’s hip contact forces and kinematics, further highlighted the importance of individual activity patterns to the occurrence of edge-loading. The combined model showed no signs of edge-loading for 14% GC or 52% stair descent.
In contrast, the contact stress on Subject 3’s acetabular component increased when 7◦ and 6◦ was added to the inclination angle at 14% GC and 52% stair descent respectively. This suggests that Subject 3’s acetabular component orientation combined with Subject 1’s individual activity patterns decreased the likelihood of edgeloading occurring. This indicates that individual activity patterns have a significant effect on wear and metal ion levels. The largest variability in angles made by the hip contact force vector in the frontal plane was at 60% of GC, which represents the end of the stance phase of gait when the foot is about to leave the floor. At 60% of GC, all the subjects were edge-loading with the addition of a maximum of 8◦ to their native inclination angles. The magnitude of the hip contact force at 60% is less than half the magnitude at peak force, and therefore it can be argued that it is less serious for subjects to edge-load at this point. The magnitude of peak hip contact force during gait and activities such as descending stairs have been shown to be 243% and 316% of body weight respectively [23].
1218
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
Fig. 5. Body mass adjusted contact stress for subjects at 14% of gait cycle.
The angle between the hip contact force vector and the edge of the acetabular component at 14% GC was 11.4◦ for Subject 3 and 8.6◦ for Subject 4. The FE models for Subjects 3 and 4 demonstrated edge-loading with the addition of 10◦ and 3◦ respectively. A similar discrepancy occurred for the same subjects at 52% stair descent. This suggests that for the FE models there was a zone, approximately 5◦ from the edge of the acetabular component, where edge-loading occurred. It is likely that such a zone exists at the edge of the acetabular component in which contact between the
femoral and acetabular components disrupts fluid-film lubrication and increased contact stress occurs; thus edge-loading may effectively occur even if the line of action of the joint reaction force passes through the acetabular component away from the edge. The radial clearance between the acetabular and femoral components permits fluid film lubrication between the components which increases the contact area and reduces wear. Edge-loading is thought to disturb this fluid film between the components resulting in localised metal-on-metal wear. The FE analysis in this study
Fig. 6. Body mass adjusted contact stress for subjects at 60% of gait cycle.
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
1219
Fig. 7. Body mass adjusted contact stress for subjects at 52% of downstairs activity.
did not incorporate a fluid layer between the components and this represents another limitation. It is thought that the addition of a fluid layer between the components, except at the edge of the acetabular component, would increase the differences in contact stress between normal and edge-loading conditions. In the absence of fluid lubrication it is likely that the friction between the components increases during edge-loading. This also represents a limitation in the current study as the inclusion of an accurate representation of this friction would alter the contact stress on the surface of the acetabular component. The subjects chosen for this study all had one MoMHRA with the exception of Subject 4 who had bilateral BHRs. At the time of analysis Subjects 1, 3 and 4 were approximately 2 years post-surgery. Subject 2 was 4 years post-surgery. Subject 4’s exceptionally high metal ion levels are influenced to an unknown extent because she has bilateral resurfacings. Subject 4 has since had both MoMHRAs
revised for psuedotumour and replaced with conventional total hip replacements. A limitation of this study was that the relationship between activity patterns during gait and stair descent, the acetabular orientation and serum metal ion levels were determined using hip contact forces from another study. The extent of the relationship between metal ion levels and hip loading could be further determined if hip contact forces could be calculated/estimated on a subject-specific basis. In conclusion, edge-loading increased contact stress by at least 50%. Patients with low serum metal ion levels showed no increase in contact stress with increasing component inclination at peak force during gait or stair descent. Patients with high serum metal ion levels exhibited edge-loading with a small increase to the inclination angle of their acetabular components that was less than the intersubject variability in the angle of published hip contact forces. This
Fig. 8. Body mass adjusted contact stress for combined and Subjects 1 and 3 at 14% of gait cycle.
Fig. 9. Body mass adjusted contact stress for combined and Subjects 1 and 3 at 52% of downstairs activity.
1220
S.J. Mellon et al. / Medical Engineering & Physics 33 (2011) 1212–1220
suggests that serum metal ion levels are influenced by acetabular orientation and individual activity patterns. A future study will utilise the techniques in this study to examine absolute cup position relative to edge-loading depending on acetabular component design. Future work will also look at the influence of acetabular anteversion on contact stress during functional activities and will incorporate subject specific hip reaction forces calculated using a musculoskeletal model. Acknowledgement The authors wish to thank Dr. David Simpson for assistance with the FE modelling. Conflict of interest The work described in this paper contains no conflict of interest. References [1] Quesada MJ, Marker DR, Mont MA. Metal-on-metal hip resurfacing: advantages and disadvantages. Journal of Arthroplasty 2008;23(7 (Suppl. 1)):69–73. [2] National Joint Registry for England and Wales; 2nd annual report; 2005. http://www.njrcentre.org.uk/NjrCentre/Portals/0/Documents/England/ Reports/NJR AR 2.pdf. [3] Hart AJ, Sabah S, Henckel J, Lewis A, Cobb J, Sampson B, et al. The painful metalon-metal hip resurfacing. Journal of Bone and Joint Surgery – British Volume 2009;91B(6):738–44. [4] Pandit H, Glyn-Jones S, McLardy-Smith P, Gundle R, Whitwell D, Gibbons CLM, et al. Pseudotumours associated with metal-on-metal hip resurfacings. Journal of Bone and Joint Surgery – British Volume 2008;90-B(7):847–51. [5] Grammatopolous G, Pandit H, Kwon YM, Gundle R, McLardy-Smith P, Beard DJ, et al. Hip resurfacings revised for inflammatory pseudotumour have a poor outcome. Journal of Bone and Joint Surgery – British Volume 2009;91B(8):1019–24. [6] Kwon YM, Ostlere S, McLardy-Smith P, Gundle R, Whitwell D, Gibbons CLM, et al. Metal ion levels in pseudotumours associated with metal-on-metal hip resurfacings. In: The 55th orthopaedic research society annual meeting. Las Vegas, USA; 2009. [7] Langton DJ, Sprowson AP, Joyce TJ, Reed M, Carluke I, Partington P, et al. Blood metal ion concentrations after hip resurfacing arthroplasty. A comparative study of articular surface replacement and Birmingham hip resurfacing arthroplasties. Journal of Bone and Joint Surgery – British Volume 2009;91B(10):1287–95. [8] Campbell P, Beaule PE, Ebramzadeh E, LeDuff M, De Smet K, Lu Z, et al. The John Charnley Award: a study of implant failure in metal-on-metal surface arthroplasties. Clinical Orthopaedics and Related Research 2006;453:35–46. [9] De Haan R, Pattyn C, Gill HS, Murray DW, Campbell PA, De Smet K. Correlation between inclination of the acetabular component and metal ion levels in metal-on-metal hip resurfacing replacement. Journal of Bone and Joint Surgery – British Volume 2008;90B(10):1291–7. [10] de Smet K, de Haan R, Calistri A, Campbell PA, Ebramzadeh E, Pattyn C, et al. Metal ion measurement as a diagnostic tool to identify problems with metalon-metal hip resurfacing. Journal of Bone and Joint Surgery – American Volume 2008;90A:202–8. [11] Young-Min K, Peter T, Burkhard S, Hemant P, Adrian T, David B, et al. Lymphocyte proliferation responses in patients with pseudotumors following metal-on-metal hip resurfacing arthroplasty. Journal of Orthopaedic Research 2009.
[12] Morlock MM, Bishop N, Zustin J, Hahn M, Ruther W, Amling M. Modes of implant failure after hip resurfacing: morphological and wear analysis of 267 retrieval specimens. Journal of Bone and Joint Surgery – American Volume 2008;90A:89–95. [13] Chan FWP, Bobyn JDP, Medley JBP, Krygier JJCET, Tanzer MMD. Wear and lubrication of metal-on-metal hip implants. Clinical Orthopaedics and Related Research 1999;369:10–24. [14] Liu F, Jin ZM, Roberts P, Grigoris P. Importance of head diameter, clearance, and cup wall thickness in elastohydrodynamic lubrication analysis of metal-onmetal hip resurfacing prostheses. Proceedings of the Institution of Mechanical Engineers Part H – Journal of Engineering in Medicine 2006;220(H6): 695–704. [15] Leslie I, Williams S, Isaac G, Ingham E, Fisher J. High cup angle and microseparation increase the wear of hip surface replacements. Clinical Orthopaedics and Related Research 2009;467(9):2259–65. [16] Williams S, Leslie I, Isaac G, Jin ZM, Ingham E, Fisher J. Tribology and wear of metal-on-metal hip prostheses: influence of cup angle and head position. Journal of Bone and Joint Surgery – American Volume 2008;90A:111–7. [17] Barbour PSM, Barton DC, Fisher J. The influence of contact stress on the wear of UHMWPE for total replacement hip prostheses. Wear 1995;181–183(1): 250–7. [18] Bennett D, Humphreys L, O’Brien S, Kelly C, Orr J, Beverland DE. The influence of wear paths produced by hip replacement patients during normal walking on wear rates. Journal of Orthopaedic Research 2008;26(9):1210–7. [19] Bennett D, Humphreys L, O’Brien S, Kelly C, Orr JF, Beverland DE. Gait kinematics of age-stratified hip replacement patients – a large scale, long-term follow-up study. Gait & Posture 2008;28(2): 194–200. [20] Bennett D, Humphreys L, O’Brien S, Kelly C, Orr JF, Beverland DE. Wear paths produced by individual hip-replacement patients – a large-scale, long-term follow-up study. Journal of Biomechanics 2008;41(11):2474–82. [21] Fregly BJ, Sawyer WG, Harman MK, Banks SA. Computational wear prediction of a total knee replacement from in vivo kinematics. Journal of Biomechanics 2005;38(2):305–14. [22] Kadaba MP, Ramakrishnan HK, Wootten ME. Measurement of lower extremity kinematics during level walking. Journal of Orthopaedic Research 1990;8(3):383–92. [23] Bergmann G, Deuretzbacher G, Heller M, Graichen F, Rohlmann A, Strauss J, et al. Hip contact forces and gait patterns from routine activities. Journal of Biomechanics 2001;34(7):859–71. [24] Vassiliou K, Elfick APD, Scholes SC, Unsworth A. The effect of ‘running-in’ on the tribology and surface morphology of metal-on-metal Birmingham hip resurfacing device in simulator studies. Proceedings of the Institution of Mechanical Engineers Part H – Journal of Engineering in Medicine 2006;220(H2): 269–77. [25] Dalstra M, Huiskes R, Vanerning L. Development and validation of a 3dimensional finite-element model of the pelvic bone. Journal of Biomechanical Engineering – Transactions of the Asme 1995;117(3):272–8. [26] Brodner W, Grubl A, Jankovsky R, Meisinger V, Lehr S, Gottsauner-Wolf F. Cup inclination and serum concentration of cobalt and chromium after metal-onmetal total hip arthroplasty. Journal of Arthroplasty 2004;19(8):66–70. [27] De Haan R, Campbell PA, Su EP, De Smet KA. Revision of metal-onmetal resurfacing arthroplasty of the hip – the influence of malpositioning of the components. Journal of Bone and Joint Surgery – British Volume 2008;90B(9):1158–63. [28] Zahiri CA, Schmalzried TP, Szuszczewicz ES, Amstutz HC. Assessing activity in joint replacement patients. Journal of Arthroplasty 1998;13(8):890–5. [29] Barrett ARW, Davies BL, Gomes M, Harris SI, Henckel J, Jakopec M, et al. Computer-assisted hip resurfacing surgery using the Acrobot (R) Navigation System. Proceedings of the Institution of Mechanical Engineers Part H – Journal of Engineering in Medicine 2007;221(H7):773–85. [30] Udofia IJ, Jin ZM. Elastohydrodynamic lubrication analysis of metal-on-metal hip-resurfacing prostheses. Journal of Biomechanics 2003;36(4):537–44. [31] Murray DW. The definition and measurement of acetabular orientation. Journal of Bone and Joint Surgery – British Volume 1993;75(2):228–32.