- Email: [email protected]
The influence of tibial component fixation techniques on resorption of supporting bone stock after total knee replacement
Journal of Biomechanics 44 (2011) 948–954
Contents lists available at ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
The influence of tibial component fixation techniques on resorption of supporting bone stock after total knee replacement Desmond Y.R. Chong a, Ulrich N. Hansen a, Rene van der Venne b, Nico Verdonschot b,c, Andrew A. Amis a,d,n a
Department of Mechanical Engineering, Imperial College London, Exhibition Road, London SW7 2AZ, United Kingdom Department of Orthopaedics, Orthopaedic Research Laboratory, Radboud University, Nijmegen Medical Centre, Nijmegen 6500 HB, The Netherlands c Laboratory for Biomechanical Engineering, Department CTW University of Twente, 7500 AE Enschede, The Netherlands d Department of Musculoskeletal Surgery, Imperial College London, Charing Cross Hospital, London W6 8RF, United Kingdom b
a r t i c l e i n f o
a b s t r a c t
Article history: Accepted 20 November 2010
Periprosthetic bone resorption after tibial prosthesis implantation remains a concern for long-term fixation performance. The fixation techniques may inherently aggravate the ‘‘stress-shielding’’ effect of the implant, leading to weakened bone foundation. In this study, two cemented tibial fixation cases (fully cemented and hybrid cementing with cement applied under the tibial tray leaving the stem uncemented) and three cementless cases relying on bony ingrowth (no, partial and fully ingrown) were modelled using the finite element method with a strain-adaptive remodelling theory incorporated to predict the change in the bone apparent density after prosthesis implantation. When the models were loaded with physiological knee joint loads, the predicted patterns of bone resorption correlated well with reported densitometry results. The modelling results showed that the firm anchorage fixation formed between the prosthesis and the bone for the fully cemented and fully ingrown cases greatly increased the amount of proximal bone resorption. Bone resorption in tibial fixations with a less secure anchorage (hybrid cementing, partial and no ingrowth) occurred at almost half the rate of the changes around the fixations with a firm anchorage. The results suggested that the hybrid cementing fixation or the cementless fixation with partial bony ingrowth (into the porous-coated prosthesis surface) is preferred for preserving proximal tibial bone stock, which should help to maintain post-operative fixation stability. Specifically, the hybrid cementing fixation induced the least amount of bone resorption. & 2010 Elsevier Ltd. All rights reserved.
Keywords: Bone remodelling/resorption Tibial component fixation Total knee replacement Finite element modelling Cemented/cementless fixation
1. Introduction After total knee replacement (TKR), proximal tibial bone resorption due to stress-shielding caused by the stiff implanted prosthesis is a clinical concern: the formation of weakening bone zones and loss of bone–prosthesis support can have detrimental effects on fixation stability and component loosening (Lonner et al., 2001; Davis et al., 2005). Less available bone stock also presents a challenge for revision knee replacement surgery (Backstein et al., 2006; Lotke et al., 2006). Fully cemented—cement applied at both the tibial tray and stem (Abu-Rajab et al., 2006; Lonner et al., 2001), hybrid cementing—cement applied underneath the tibial tray but leaving the stem uncemented (Adalberth et al., 2001; Schai et al., 1998) or cementless—relying on bony ingrowth (Whiteside, 2001) fixation techniques may be chosen. The influence of these different fixation techniques on proximal tibial bone resorption remains unclear due to the paucity of studies in the literature. Increased
n Corresponding author at: Mechanical Engineering Department, Imperial College London, Exhibition Road, London SW7 2AZ, United Kingdom. Tel: + 44 020 7594 7062; fax: + 44 020 7594 5702. E-mail address: [email protected] (A.A. Amis).
0021-9290/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2010.11.026
proximal tibial stress-shielding was observed in-vitro for fully cemented over the hybrid cementing fixation (Jazrawi et al., 2001), but the results were not statistically significant. Bone loss in-vivo at two years in the distal femur after femoral implantation was found to be lower for the cementless than the cemented fixation (Seki et al., 1999). In computational bone remodelling studies of the femoral component fixation, it was found that bone resorption was less extensive when the interfaces between the cement and prosthesis were debonded as compared to when the interfaces were bonded (Barink et al., 2003; van Lenthe et al., 1997). Bone resorption of cemented and cementless tibial fixations has not been investigated in a manner that allows a direct comparison between fixation techniques. In a fixation using cement, the load transfer capability of the stem could be lower if the stem were uncemented or polished (Whiteside, 1993). While sufficient osseointegration increases the fixation strength of the cementless fixation, a greater extent of bone ingrowth into the porous-coated surfaces of the stem could be anticipated to induce greater distal load transfer, thus leading to proximal bone resorption. The objectives of this study were to investigate bone remodelling in the proximal tibia after prosthesis implantation using computational finite element simulations incorporating a strain-adaptive remodelling theory, to
D.Y.R. Chong et al. / Journal of Biomechanics 44 (2011) 948–954
provide a direct comparison between different fixation techniques and to suggest the preferred choice for preserving bone stock and maintaining essential fixation stability. Based on the literature, it was hypothesised that the fully-bonded fixation methods would lead to predictions of more bone resorption than where the fixation stem was not bonded to the surrounding bone.
2. Materials and methods
949
et al., 2007) to mimic daily activities (Table 2). The corresponding antero-posterior shear forces were obtained from Taylor et al. (2004) and shared equally between the medial and lateral compartments. The body weight was taken as a nominal 70 kg. The loads were applied at the tibio-femoral contact points, distributed over areas 5 mm diameter, where the contact points for each physiological load condition varied according to the different angles of knee flexion (Andriacchi et al., 1986). When the knee flexed, the contact point in the lateral compartment moved more posteriorly than in the medial compartment.
2.2. Strain-adaptive bone remodelling simulation
2.1. Finite element models set-up The three-dimensional finite element (FE) tibial fixation models were built based on the CT images of the Visible Human Project female left tibia. The proximal segment of the bone, 150 mm long, was modelled as linear elastic and isotropic, with Poisson’s ratio of 0.3. The heterogeneous properties of the bone were defined by mapping the CT greyscale number to the elastic modulus of each bone finite element based on a linear relationship between the CT number and bone apparent density (r) (Abdul-Kadir et al., 2008), and the elastic modulus assigned by E¼3790r3 (Carter and Hayes, 1977). The generic tibial component was based on commercially available TKR designs, which have common features of the polyethylene bearing resting in a flat metal tray, and a central fixation stem 40 mm long by 12 mm diameter and flanges on both sides (Fig. 1). The proximal tibia was virtually resected at 8 mm below the lateral articulating surface of the tibial plateau. The tray and polyethylene insert sizes were dimensioned to provide a maximum coverage of the resected surface as in clinical practise. Their dimensions and material properties are listed in Table 1. Five different cases were modelled (Fig. 1): (i) fully cemented, (ii) hybrid cementing, (iii) cementless no ingrowth—porous-coated tibial prosthesis assuming the worst scenario of no osseointegration, (iv) cementless partial ingrowth—porous-coated prosthesis assuming a realistic amount of osseointegration, (v) cementless fully ingrown—porous-coated prosthesis assuming the best scenario, that bone has ingrown into the entire prosthesis surface. The interfaces were modelled as bonded, where bone cement was applied. For the cementless cases, the bone–prosthesis and bone–stem interfaces with no bony ingrowth were modelled as unbonded by defining contact between them with Coulomb’s friction. The tibial prosthesis was considered to be porous-coated with beads and a coefficient of friction of 0.6 between the bone and prosthesis was assigned (Hashemi et al., 1996). The regions of bone ingrowth for ‘‘partial ingrowth’’ were determined by analysing the zonal area of the prosthesis with interface micromotion lower than 50 mm, where osseointegration is likely (Chong et al., 2010). Osseointegration was predicted to occur on 33% of the surface area of the tray and stem, in line with Turner et al. (1989) and Sumner et al. (1995). The nodes at these bone–prosthesis interfaces were bonded. For the ‘‘fully ingrown’’ case, all the nodes at the bone–prosthesis interfaces were bonded. In hybrid cementing, a friction coefficient of 0.1 was used at the unbonded interface of the bone to a smooth prosthesis stem (Tissakht et al., 1995; Barker et al., 2005). In all cases, the PE was modelled as bonded to the tibial tray. The entire FE model consisted of approximately 35,000 linear tetrahedral elements, and the convergence tolerance was set at 1%. The MSC.Marc 2005 FE software was used. For the boundary conditions, all nodes of the distal cut surface of the tibia were fully-constrained, and the FE model was loaded with a superposition of three different physiological knee joint load conditions corresponding to 0% and 15% of the walking gait cycle (Shelburne et al., 2006), and 25% of the stair climbing cycle (Zhao
Intact Tibia
Fully Cemented
The change in apparent bone density with time was predicted based on a strainadaptive bone remodelling theory (Huiskes et al., 1987) incorporated into FE simulations. The numerical scheme of the remodelling procedure has been detailed in previous studies of Huiskes et al. (1992) and Weinans et al. (1993), where the strain energy density (SED) was used as a mechanical stimulus to regulate the remodelling process. Two FE models, the pre-operative tibia and the post-operative cases were created (Fig. 1) and loaded under the same conditions. When the SED of the postoperative case (Spost) was within a threshold level of the pre-operative (Spre), no change in the bone density would occur. This threshold level, s, (commonly known as ‘‘dead zone’’) was set at 775% of the physiological SED, with which realistic bone remodelling results for hip (Huiskes et al., 1992; Kerner et al, 1999) and femoral knee (Barink et al., 2003; van Lenthe et al., 1997) implantations were achieved. When Spost increased beyond the dead zone, the bone density would increase, and vice versa. After which, bone density of the post-operative tibia would be updated at the end of each loading. The loading iterations and bone-adaptation process continued until Spost and Spre were matched. The rate of change of the bone density is expressed as: 2
A½Spost ð1 þ sÞSpre , 6 0, 4 dr ¼ A½S ð1sÞS , post pre dt
3 spost Z ð1 þ sÞSpre ð1sÞSpre o Spost o ð1 þ sÞSpre 7 5
ð1Þ
Spost r ð1sÞSpre
0:01 r r r 1:73 g=cm
3
Table 1 Dimensions and material properties of the tibial tray and polyethylene insert modelled. Material
Dimensions
Elastic modulus (GPa)
Poisson’s ratio
Tibial tray (titanium alloy)
Medial/lateral width—74 mm Anterior/posterior depth—46 mm Thickness—4 mm Stem length—40 mm Diameter—12 mm Thickness—8 mm
110 Au et al., (2007)
0.33
1 Au et al. (2005)
0.3
Polyethylene insert
Cementless
Hybrid Cementing
Fig. 1. Finite element models of the pre-operative tibia (a) and different tibial fixation techniques. In the exploded views of the prostheses, the upper component is the polyethylene bearing, which rests on the tray of the titanium alloy tibial component. In the fully cemented case (b), an idealised PMMA cement mantle was interposed between the tibial component and the bone. In the cementless case (c), no cement was applied around the tibial component. In the hybrid cementing case (d), a layer of PMMA was interposed beneath the tray of the tibial component, but not around the fixation stem.
950
D.Y.R. Chong et al. / Journal of Biomechanics 44 (2011) 948–954
Table 2 Joint load configurations for bone remodelling simulations. No.
Load case
1 2 3
Load ( body weight)
0% of gait cycle (heel strike) 15% of gait cycle (contra toe-off) 25% of stair climbing cycle a b c
Medial (vertical)
Lateral (vertical)
Antero-posterior
0.9a 2.3a 2c
0.19a 0.46a 1.5c
0.1b 0.08b 1.3b
0 16 32
Shelburne et al. (2006). Taylor et al. (2004). Zhao et al. (2007). Change in Bone Apparent Density with Time
40 Fully Cemented Cementless
30 Change in Apparent Density (%
Knee flexion angle (deg)
Hybrid Cementing Partial Ingrowth
20
Fully Ingrowth
10 0
-10 -20 -30
6
12 36 60 Resected Surface
6
12 36 60 MidStem Time (months)
6
12 36 Stem Tip
60
Fig. 2. Predicted change in the apparent densities of the proximal tibia with time (6, 12, 36 and 60 months after implantation) for different fixation techniques. where A is a constant and the time (t) in months. The above equation can be solved by means of forward Euler’s integration with constant time-steps (Dt) in the FE simulations (Weinans et al., 1993). The density change per time step is calculated from:
Dr ¼ ADt½Spost ð1 7 sÞSpre
ð2Þ
Subsequent to the early studies mentioned above, clinical evidence has suggested that bone suffers ‘memory loss’ and the adaptive process stops after sixty months (Petersen et al., 1995; Saari et al., 2007). Therefore, the bone remodelling following joint replacement was simulated for a period equivalent to sixty months. The remodelling results are presented at three levels in the transverse plane: at the tibial resected surface, at mid-stem and at the stem tip. The apparent density of the bone at each level was computed by averaging the apparent densities of all elements at the particular transverse plane.
fixation (Fig. 3b) occurred by 36 months, while it occurred after 12 months for the fully cemented fixation (Fig. 3a). Greater bone resorption was predicted to occur at the medial condyle, mainly due to the higher joint load acting on the medial side being shielded by the metallic prosthesis. Bone densification was observed at the lateral condyle anterior to the flange for the cementless case, indicating that part of the joint load was transmitted by the tibial tray to the underlying bone. At five years post-operation, a much larger proportion of bone resorption was predicted for the cemented than the cementless fixation at the resected surface and 10 mm underneath. Bone densification was predicted 30 mm below the resected surface and at the stem tip, indicating that the joint load was transferred distally by the cemented stem. For cementless fixation, localised periprosthetic bone densification was observed from the mid-stem level and below. The bone remodelling patterns predicted for the hybrid cementing, partial ingrowth and fully ingrown cases are shown in Fig. 3c–e. Similar to the cementless (no ingrowth) fixation, severe proximal bone resorption began to occur only after 30 months postoperatively for the hybrid and partial ingrowth cases. For the fully ingrown case, severe bone resorption occurred 18 months after tibial implantation. The bone remodelling pattern for the partial ingrowth fixation was similar to the case with no ingrowth, but with a greater amount of proximal bone resorption. With the bone– prosthesis interfaces fully bonded (fully ingrown fixation), the absence of the cement layer led to a prediction of slightly greater bone resorption proximally and at mid-stem, and greater bone densification towards the stem tip when compared to the fully cemented case.
4. Discussion 3. Results The changes in bone apparent density with time after implantation at various levels are illustrated in Fig. 2. At the tibial resected surface, the fully cemented and ingrown cases were predicted to induce the largest amount of proximal bone resorption (26–29% at 60 months), followed by the partial ingrowth case (17%). The cementless (no ingrowth) and hybrid cementing cases experienced the least bone resorption (11%). At the mid-stem level (20 mm below resected surface), bone resorption was observed for the fully cemented, fully ingrown and partial ingrowth cases, though of a lower extent than at the resected surface. Bone densification was predicted for the hybrid case. At the stem tip region, bone densification was predicted for all cases, showing the load transfer capability of the stem. The predicted regions of bone density change at the tibial resected surface and various levels below for all the fixation cases are shown in Fig. 3a–e. Severe bone resorption was defined as where density¼0.01 g/cm3, represented by the light blue region. At the tibial resected surface, severe bone resorption for the cementless
In this study, bone remodelling in the proximal tibia after prosthetic implantation was investigated by finite element simulations, and the influence of fixation techniques (cemented/cementless) were compared. In general, the predictions followed the trend which had been hypothesised. The predictions were compared with published clinical densitometry and radiographic results. At 94 months post-operation, Lonner et al. (2001) found significant bone loss from the proximal tibia for fixation using a fully cemented stemmed tibial prosthesis. The medial tibial plateau experienced a larger amount of bone resorption than the lateral compartment. Consistent with these clinical findings, the bone remodelling simulation predicted severe proximal bone resorption for the fully cemented fixation and greater bone loss at the medial condyle. In other clinical studies, bone hypertrophy was observed at the tip of the stem with hybrid fixation (Bertin et al., 1985) and with uncemented fixation (Whiteside and Pafford, 1989). Bone densification was predicted at the stem tip in the current study. The amount of bone mineral density reduction in the implanted tibia reported by densitometry studies ranged from 10–12% at one to two years post-operatively (Karbowski et al., 1999; Li and Nilsson,
D.Y.R. Chong et al. / Journal of Biomechanics 44 (2011) 948–954
Fully Cemented g/cm
951
Cementless - No Ingrowth
A L Resected surface
M P
10mm below resected surface
20mm below resected surface
30mm below resected surface
Ste m tip
12 months
Time = 0
36 months
60 months
Hybrid Cementing g/cm
Time = 0
Cementless Partial Ingrowth
12 months
36 months
60 months
Cementless Fully ingrown
A L Resected surface
M P
20mm below resected surface
Ste m tip Time = 0
36 months
60 months
Time = 0
36 months
60 months
Time = 0
36 months
60 months
Fig. 3. (a) Bone apparent density plots showing the regions of predicted bone resorption and densification in the transverse plane at the tibial resected surface and various levels below for the fully cemented case. (b) Bone apparent density plots showing the regions of predicted bone resorption and densification in the transverse plane at the tibial resected surface and various levels below for the cementless case. (c). Bone apparent density plots showing the regions of predicted bone resorption and densification in the transverse plane at levels of the tibial resected surface, mid-stem and stem tip for the hybrid cementing case. (d) Bone apparent density plots showing the regions of predicted bone resorption and densification in the transverse plane at levels of the tibial resected surface, mid-stem and stem tip for the cementless partial ingrowth case. (e) Bone apparent density plots showing the regions of predicted bone resorption and densification in the transverse plane at levels of the tibial resected surface, mid-stem and stem tip for cementless fully ingrown case. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)
2000), and up to 23–39% at five to eight years after operation (Levitz et al., 1995; Saari et al., 2007). The current predictions of 11–29% decrease in the proximal bone apparent density at 5 years post-
operation were in reasonable range with the densitometry data, but it becomes difficult to separate the effects of the implant from ageing effects on activity with increasing time.
952
D.Y.R. Chong et al. / Journal of Biomechanics 44 (2011) 948–954
Fig. 3. (Continued)
The density–modulus relationship of Carter and Hayes used in this study had previously been correlated with in-vitro tests on cadaveric hip implants (Abdul-Kadir et al., 2008). However, it is known that this relationship varies with anatomical site (Morgan et al., 2003), and further studies, on a subject-specific basis would be valuable. Three different extents of osseointegration – no ingrowth, partial ingrowth and fully ingrown – were modelled for cementless fixations. In retrieval studies, while extensive levels of bone ingrowth (40–90% of the tray area) was found in canine models (Sumner et al., 1994; Turner et al., 1989), only limited bone ingrowth of less than 30% of the tray area was observed clinically (Ranawat et al., 1986; Sumner et al., 1995). In a worst case, 95% of the tibial components retrieved from patients showed less than 5% of bone ingrowth (Cook et al., 1989). The current FE models thus represented the extremes of clinical circumstances. The different fixation techniques were found to influence the bone resorption via a stress-shielding effect, and so alterations of the load transfer mechanism of the prosthesis may be inferred. For example, with bonded interfaces modelled for the fully cemented and fully ingrown cases, the load acting on the tibial tray could be transferred to the bone distally through shear stresses (Askew and
Lewis, 1981) along the bone–stem interfaces. For the uncemented regions and where bone ingrowth did not occur, the unbonded interfaces modelled with contact surfaces and friction were less able to sustain shear stresses. A firm fixation of the stem which transfers more joint load distally would thereby relieve the proximal bone stresses, consequently leading to proximal bone resorption. A greater amount of bone resorption from 0 to 20 mm below the resected surface was thus observed for the fully cemented fixation than the hybrid case. The rate of proximal bone resorption for partial ingrowth cementless fixation was in-between the full and no ingrown cases, as the bone–prosthesis interfaces were partially bonded (33%). Because the stem was uncemented in both hybrid cementing and no-ingrowth fixations, giving comparable load transfer capability of the stem, similar amounts of proximal bone resorption were predicted. The effect of interface bonding has also been investigated in computational studies of knee femoral component fixations (Barink et al., 2003; van Lenthe et al., 1997), where load transfer phenomena similar to those in the present study were observed. Bone resorption was less extensive when the interfaces between the cement and prosthesis were debonded as compared to when the interfaces were bonded. The bonded interface resulted in the joint forces being transferred more
D.Y.R. Chong et al. / Journal of Biomechanics 44 (2011) 948–954
proximally, via the anterior and posterior flanges, thereby leading to bone densification, and severe bone loss distally within the femoral component (van Lenthe et al., 1997). Greater bone densification underneath the stem tip was observed for the uncemented fixation. This is probably due to the direct contact between the metallic stem tip and bone leading to higher bone stresses than with an intermediate layer of cement. The increase in bone density was in agreement with clinical specimens (Bertin et al., 1985). With a complete cement mantle providing a firm interlock between the bone and prosthesis, the load could be distributed more evenly to the entire periprosthetic bone regions rather than concentrated at specific regions underneath the stem tip, leading to current prediction of lower bone hypertrophy at this zone. For the no/partial ingrowth cementless cases, bone densification was also predicted at the lateral condyle of the tibia, indicating high force transmission from the tibial tray to the bone underneath. While the current bone remodelling simulations have shown good correlations with in-vivo studies, some specific limitations were present. The FE modelling assumed that bone was ingrown onto the porous-coated prosthesis immediately after operation for the partial and fully ingrown fixations. In histological analyses, osseointegration was found to occur between 4 weeks (Søballe et al., 1992) and 12 weeks (Bobyn et al., 1981) after implantation. The ingrowth duration was short compared to the simulated bone remodelling period of 60 months, thus the assumption of bonded interfaces between the bone and prosthesis at time zero of bone remodelling should not have a great influence on the end results. It should, however, be noted that the above histological analyses were conducted on canine models and bone fixation might be slower in clinical use. If the central stem was bonded to the bone at a later stage after operation (hence delayed load transfer effect), proximal bone resorption at 5 years for the partial and fully ingrown cases would be less than the currently predicted values. Because this study was based on only one bone model, it has not been validated across the patient population, and so it is not known whether the different implant fixations will have statistically significant effects; however, this study does provide relative comparisons. The bone was modelled as isotropic, but the diaphysis is anisotropic. Although that is not exact, heterogeneous material assignment was found to be adequate for large bone specimens modelling (Baca et al., 2008). The articular cartilage and menisci were not modelled in the pre-operative tibia, so the physiological SED may not have been computed exactly. These soft tissues help to distribute the stresses across the tibial condyles, so the regions of density change predicted at the tibial resected surface could differ. The vertical and shear loadings were obtained from different independent studies, hence the forces prescribed in the FE models were not consistently from the same group of subjects. This could lead to a likely deviation between the joint loads being applied and those in-vivo. In particular, we assumed that the shear force reported by Taylor et al. (2004) was shared equally between the two compartments. Nevertheless, before fully-detailed knee joint loadings can be measured from the same subjects, the current approach of combining loading conditions from different studies remains practical. The study also assumed no interference fit of the tibial component in the hybrid cementing and no-ingrowth fixations. That is, it is assumed that the interference fit will relax or remodel to a negligible level in a timescale short in comparison to the timescale of most of the remodelling response. If a significant interference fit does exist it would increase the shear loading transfer along the bone–stem interface and therefore result in greater proximal bone resorption than predicted in this study. This study assumed a friction coefficient of 0.1 for smooth stems, as in Tissakht et al. (1995), although 0.42 has also been reported
953
(Shirazi-Adl et al., 1993). A sensitivity analysis (Chong, 2009) found that values from 0.1 to 0.8 had small effects ( o10%) on tibial fixation behaviour. A direct comparison between the different tibial fixation techniques revealed that their influences on bone remodelling were diverse. A firm fixation formed by the bonded interfaces between the prosthesis and the bone for the fully cemented and fully ingrown cases induced proximal tibial bone resorption due to load transfer distally. Greater bone resorption would inevitably weaken the bone supporting the tibial prosthesis, leading to potential problems such as aseptic loosening. Although full cementing may provide initial fixation stability, its long-term consequence on bone quality may be an issue. Bone resorption in tibial fixations with a less secure anchorage (hybrid, no ingrowth and partial ingrowth) were predicted to occur at almost half the rate of the changes around the fixations with a firm anchorage. The fully cemented fixation resulted in the most severe proximal bone resorption, while the partial ingrowth cementless fixation led to moderate changes. The least amount of bone resorption followed hybrid cementing. Thus the hybrid cementing technique or the cementless fixation allowing for partial ingrowth may be a preferred choice for preserving tibial bone stock after TKR.
Conflict of interest statement The authors declare that there is no conflict of interest that could inappropriately influence the results of the current research work.
Acknowledgment The studentship for D.Y.R. Chong was sponsored by the National Medical Research Council, Ministry of Health, Singapore. References Abdul-Kadir, M.R., Hansen, U., Klabunde, R., Lucas, D., Amis, A., 2008. Finite element modelling of primary hip stem stability: the effect of interference fit. Journal of Biomechanics 41 (3), 587–594. Abu-Rajab, R.B., Watson, W.S., Walker, B., Roberts, J., Gallacher, S.J., Meek, R.M.D., 2006. Peri-prosthetic bone mineral density after total knee arthroplasty: cemented versus cementless fixation. Journal of Bone and Joint Surgery, British Volume 88, 606–613. ¨ Adalberth, G., Nilsson, K.G., Bystrom, S., Kolstad, K., Milbrink, J., 2001. Allpolyethylene versus metal-backed and stemmed tibial components in cemented total knee arthroplasty. Journal of Bone and Joint Surgery, British 83 (6), 825–831. Andriacchi, T.P., Stanwyck, T.S., Galante, J.O., 1986. Knee biomechanics and total knee replacement. Journal of Arthroplasty 1 (3), 211–219. Askew, M.J., Lewis, J.L., 1981. Analysis of model variables and fixation post length effects on stresses around a prosthesis in the proximal tibia. Journal of Biomechanical Engineering 103, 239–245. Au, A.G., Liggins, A.B., Raso, V.J., Amirfazli, A., 2005. A parametric analysis of fixation post shape in tibial knee prostheses. Medical Engineering and Physics 27, 123–134. Au, A.G., Raso, V.J., Liggins, A.B., Amirfazli, A., 2007. Contribution of loading conditions and material properties to stress shielding near the tibial component of total knee replacement. Journal of Biomechanics 40, 1410–1416. Baca, V., Horak, Z., Mikulenka, P., Dzupa, V., 2008. Comparison of an inhomogeneous orthotropic and isotropic material models used for FE analyses. Medical Engineering and Physics 30, 924–930. Backstein, D., Safir, O., Gross, A., 2006. Management of bone loss—structural grafts in revision total knee arthroplasty. Clinical Orthopaedics and Related Research 446, 104–112. Barink, M., Verdonschot, N., de Waal Malefijt, M., 2003. A different fixation of the femoral component in total knee arthroplasty may lead to preservation of femoral bone stock. Proceedings IMechE Part H—Journal of Engineering in Medicine 217, 325–332. Barker, D.S., Tanner, K.E., Ryd, L., 2005. A circumferentially flanged tibial tray minimizes bone-tray shear micromotion. Proceedings IMechE Part H—Journal of Engineering in Medicine 219, 449–456. Bertin, K.C., Freeman, M.A.R., Samuelson, K.M., Ratcliffe, S.S., Todd, R.C., 1985. Stemmed revision arthroplasty for aseptic loosening of total knee replacement. Journal of Bone and Joint Surgery, British Volume 67 (2), 242–248.
954
D.Y.R. Chong et al. / Journal of Biomechanics 44 (2011) 948–954
Bobyn, J.D., Pilliar, R.M., Cameron, H.U., Weatherly, G.C., 1981. Osteogenic phenomena across endosteal bone-implant spaces with porous surfaced intramedullary implants. Acta Orthopaedica Scandinavica 52, 145–153. Carter, D.R., Hayes, W.C., 1977. The compressive behavior of bone as a two-phase porous structure. Journal of Bone and Joint Surgery 59-A (7), 954–963. Chong, D.Y.R., 2009. Biomechanical analysis of fixation and bone remodelling of total knee replacement, PhD Thesis, Imperial College London, UK. Chong, D.Y.R., Hansen, U.N., Amis, A.A., 2010. Analysis of bone–prosthesis interface micromotion for cementless tibial prosthesis fixation and the influence of loading conditions. Journal of Biomechanics 43 (6), 1074–1080. Cook, S.D., Barrack, R.L., Thomas, K.A., Haddad, R.J., 1989. Quantitative histologic analysis of tissue growth into porous total knee components. Journal of Arthroplasty 4 (Supplement 1), S33–S43. Davis, A.M., Damani, M., White, L.M., Wunder, J.S., Griffin, A.M., Bell, R.S., 2005. Periprosthetic bone remodeling around a prosthesis for distal femoral tumors—longitudinal follow-up. Journal of Arthroplasty 20, 219–224. Hashemi, A., Shirazi-Adl, A., Dammak, M., 1996. Bidirectional friction study of cancellous bone-porous coated metal interface. Journal of Biomedical Materials Research (Applied Biomaterials) 33, 257–267. Huiskes, R., Weinans, H., Grootenboer, H.J., Dalstra, M., Fudala, B., Slooff, T.J., 1987. Adaptive bone-remodeling theory applied to prosthetic-design analysis. Journal of Biomechanics 20 (11/12), 1135–1150. Huiskes, R., Weinans, H., van Rietbergen, B., 1992. The relationship between stress shielding and bone resorption around total hip stems and the effects of flexible materials. Clinical Orthopaedics and Related Research 274, 124–134. Jazrawi, L.M., Bai, B., Kummer, F.J., Hiebert, R., Stuchin, S.A., 2001. The effect of stem modularity and mode of fixation on tibial component stability in revision total knee arthroplasty. Journal of Arthroplasty 16 (6), 759–767. Karbowski, A., Schwitalle, M., Eckardt, A., Heine, J., 1999. Periprosthetic bone remodelling after total knee arthroplasty: early assessment by dual energy X-ray absorptiometry. Archives of Orthopaedic and Trauma Surgery 119 (8), 324–326. Kerner, J., Huiskes, R., van Lenthe, G.H., Weinans, H., van Rietbergen, B., Engh, C.A., Amis, A.A., 1999. Correlation between pre-operative periprosthetic bone density and post-operative bone loss in THA can be explained by strain-adaptive remodelling. Journal of Biomechanics 32, 695–703. Levitz, C.L., Lotke, P.A., Karp, J.S., 1995. Long-term changes in bone mineral density following total knee replacement. Clinical Orthopaedics and Related Research 321, 68–72. Li, M.G., Nilsson, K.G., 2000. Changes in bone mineral density at the proximal tibia after total knee arthroplasty: a 2-year follow-up of 28 knees using dual energy X-ray absorptiometry. Journal of Orthopaedic Research 18 (1), 40–47. Lonner, J.H., Klotz, M., Levitz, C., Lotke, P.A., 2001. Changes in bone density after cemented total knee arthroplasty—influence of stem design. Journal of Arthroplasty 16 (1), 107–111. Lotke, P.A., Carolan, G.F., Puri, N., 2006. Impaction grafting for bone defects in revision total knee arthroplasty. Clinical Orthopaedics and Related Research 446, 99–103. Morgan, E.F., Bayraktar, H.H., Keaveny, T.M., 2003. Trabecular bone modulus–density relationships depend on anatomic site. Journal of Biomechanics 36, 897–904. Petersen, M.M., Olsen, C., Lauritzen, J.B., Lund, B., 1995. Changes in bone mineral density of the distal femur following uncemented total knee arthroplasty. Journal of Arthroplasty 10 (1), 7–11.
Ranawat, C.S., Johanson, N.A., Rimnac, C.M., Wright, T.M., Schwartz, R.E., 1986. Retrieval analysis of porous-coated components for total knee arthroplasty. Clinical Orthopaedics and Related Research 209, 244–248. ¨ Saari, T., Uvehammer, J., Carlsson, L., Regne´r, L., Karrholm, J., 2007. Joint area constraint had no influence on bone loss in proximal tibia 5 years after total knee replacement. Journal of Orthopaedic Research 25, 798–803. Schai, P.A., Thornhill, T.S., Scott, R.D., 1998. Total knee arthroplasty with the PFC system—results at a minimum of ten years and survivorship analysis. Journal of Bone and Joint Surgery, British 80 (5), 850–858. Seki, T., Omori, G., Koga, Y., Suzuki, Y., Ishii, Y., Takahashi, H.E., 1999. Is bone density in the distal femur affected by use of cement and by femoral component design in total knee arthroplasty? Journal of Orthopaedic Science 4, 180–186. Shelburne, K.B., Torry, M.R., Pandy, M.G., 2006. Contributions of muscles, ligaments, and the ground-reaction force to tibiofemoral joint loading during normal gait. Journal of Orthopaedic Research 24, 1983–1990. Shirazi-Adl, A., Dammak, M., Paiement, G., 1993. Experimental determination of friction characteristics at the trabecular bone/porous-coated metal interface in cementless implants. Journal of Biomedical Materials Research 27, 167–175. ¨ Søballe, K., Hansen, E.S., Rasmussen, H.B., Jørgensen, P.H., Bunger, C., 1992. Tissue ingrowth into titanium and hydroxyapatite-coated implants during stable and unstable mechanical conditions. Journal of Orthopaedic Research 10, 285–299. Sumner, D.R., Kienapfel, H., Jacobs, J.J., Urban, R.M., Turner, T.M., Galante, J.O., 1995. Bone ingrowth and wear debris in well-fixed cementless porous-coated tibial components removed from patients. Journal of Arthroplasty 10 (2), 157–167. Sumner, D.R., Turner, T.M., Dawson, D., Rosenberg, A.G., Urban, R.M., Galante, J.O., 1994. Effect of pegs and screws on bone ingrowth in cementless total knee arthroplasty. Clinical Orthopaedics and Related Research 309, 150–155. Taylor, W.R., Heller, M.O., Bergmann, G., Duda, G.N., 2004. Tibio-femoral loading during human gait and stair climbing. Journal of Orthopaedic Research 22, 625–632. Tissakht, M., Eskandari, H., Ahmed, A.M., 1995. Micromotion analysis of the fixation of total knee tibial component. Computers and Structures 56 (2/3), 365–375. Turner, T.M., Urban, R.M., Sumner, D.R., Skipor, A.K., Galante, J.O., 1989. Bone ingrowth into the tibial component of a canine total condylar knee replacement prosthesis. Journal of Orthopaedic Research 7 (6), 893–901. van Lenthe, G.H., de Waal Malefijt, M.C., Huiskes, R., 1997. Stress shielding after total knee replacement may cause bone resorption in the distal femur. Journal of Bone and Joint Surgery, British 79 (1), 117–122. Weinans, H., Huiskes, R., van Rietbergen, B., Sumner, D.R., Turner, T.M., Galante, J.O., 1993. Adaptive bone remodeling around bonded noncemented total hip arthroplasty: a comparison between animal experiments and computer simulation. Journal of Orthopaedic Research 11, 500–513. Whiteside, L.A., 1993. Cementless revision total knee arthroplasty. Clinical Orthopaedics and Related Research 286, 160–167. Whiteside, L.A., 2001. Long-term followup of the bone-ingrowth ortholoc knee system without a metal-backed patella. Clinical Orthopaedics and Related Research 388, 77–84. Whiteside, L.A., Pafford, J., 1989. Load transfer characteristics of a noncemented total knee arthroplasty. Clinical Orthopaedics and Related Research 239, 168–177. Zhao, D., Banks, S.A., D’Lima, D.D., Colwell Jr., C.W., Fregly, B.J., 2007. In vivo medial and lateral tibial loads during dynamic and high flexion activities. Journal of Orthopaedic Research 25 (5), 593–602.
Contents lists available at ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
The influence of tibial component fixation techniques on resorption of supporting bone stock after total knee replacement Desmond Y.R. Chong a, Ulrich N. Hansen a, Rene van der Venne b, Nico Verdonschot b,c, Andrew A. Amis a,d,n a
Department of Mechanical Engineering, Imperial College London, Exhibition Road, London SW7 2AZ, United Kingdom Department of Orthopaedics, Orthopaedic Research Laboratory, Radboud University, Nijmegen Medical Centre, Nijmegen 6500 HB, The Netherlands c Laboratory for Biomechanical Engineering, Department CTW University of Twente, 7500 AE Enschede, The Netherlands d Department of Musculoskeletal Surgery, Imperial College London, Charing Cross Hospital, London W6 8RF, United Kingdom b
a r t i c l e i n f o
a b s t r a c t
Article history: Accepted 20 November 2010
Periprosthetic bone resorption after tibial prosthesis implantation remains a concern for long-term fixation performance. The fixation techniques may inherently aggravate the ‘‘stress-shielding’’ effect of the implant, leading to weakened bone foundation. In this study, two cemented tibial fixation cases (fully cemented and hybrid cementing with cement applied under the tibial tray leaving the stem uncemented) and three cementless cases relying on bony ingrowth (no, partial and fully ingrown) were modelled using the finite element method with a strain-adaptive remodelling theory incorporated to predict the change in the bone apparent density after prosthesis implantation. When the models were loaded with physiological knee joint loads, the predicted patterns of bone resorption correlated well with reported densitometry results. The modelling results showed that the firm anchorage fixation formed between the prosthesis and the bone for the fully cemented and fully ingrown cases greatly increased the amount of proximal bone resorption. Bone resorption in tibial fixations with a less secure anchorage (hybrid cementing, partial and no ingrowth) occurred at almost half the rate of the changes around the fixations with a firm anchorage. The results suggested that the hybrid cementing fixation or the cementless fixation with partial bony ingrowth (into the porous-coated prosthesis surface) is preferred for preserving proximal tibial bone stock, which should help to maintain post-operative fixation stability. Specifically, the hybrid cementing fixation induced the least amount of bone resorption. & 2010 Elsevier Ltd. All rights reserved.
Keywords: Bone remodelling/resorption Tibial component fixation Total knee replacement Finite element modelling Cemented/cementless fixation
1. Introduction After total knee replacement (TKR), proximal tibial bone resorption due to stress-shielding caused by the stiff implanted prosthesis is a clinical concern: the formation of weakening bone zones and loss of bone–prosthesis support can have detrimental effects on fixation stability and component loosening (Lonner et al., 2001; Davis et al., 2005). Less available bone stock also presents a challenge for revision knee replacement surgery (Backstein et al., 2006; Lotke et al., 2006). Fully cemented—cement applied at both the tibial tray and stem (Abu-Rajab et al., 2006; Lonner et al., 2001), hybrid cementing—cement applied underneath the tibial tray but leaving the stem uncemented (Adalberth et al., 2001; Schai et al., 1998) or cementless—relying on bony ingrowth (Whiteside, 2001) fixation techniques may be chosen. The influence of these different fixation techniques on proximal tibial bone resorption remains unclear due to the paucity of studies in the literature. Increased
n Corresponding author at: Mechanical Engineering Department, Imperial College London, Exhibition Road, London SW7 2AZ, United Kingdom. Tel: + 44 020 7594 7062; fax: + 44 020 7594 5702. E-mail address: [email protected] (A.A. Amis).
0021-9290/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2010.11.026
proximal tibial stress-shielding was observed in-vitro for fully cemented over the hybrid cementing fixation (Jazrawi et al., 2001), but the results were not statistically significant. Bone loss in-vivo at two years in the distal femur after femoral implantation was found to be lower for the cementless than the cemented fixation (Seki et al., 1999). In computational bone remodelling studies of the femoral component fixation, it was found that bone resorption was less extensive when the interfaces between the cement and prosthesis were debonded as compared to when the interfaces were bonded (Barink et al., 2003; van Lenthe et al., 1997). Bone resorption of cemented and cementless tibial fixations has not been investigated in a manner that allows a direct comparison between fixation techniques. In a fixation using cement, the load transfer capability of the stem could be lower if the stem were uncemented or polished (Whiteside, 1993). While sufficient osseointegration increases the fixation strength of the cementless fixation, a greater extent of bone ingrowth into the porous-coated surfaces of the stem could be anticipated to induce greater distal load transfer, thus leading to proximal bone resorption. The objectives of this study were to investigate bone remodelling in the proximal tibia after prosthesis implantation using computational finite element simulations incorporating a strain-adaptive remodelling theory, to
D.Y.R. Chong et al. / Journal of Biomechanics 44 (2011) 948–954
provide a direct comparison between different fixation techniques and to suggest the preferred choice for preserving bone stock and maintaining essential fixation stability. Based on the literature, it was hypothesised that the fully-bonded fixation methods would lead to predictions of more bone resorption than where the fixation stem was not bonded to the surrounding bone.
2. Materials and methods
949
et al., 2007) to mimic daily activities (Table 2). The corresponding antero-posterior shear forces were obtained from Taylor et al. (2004) and shared equally between the medial and lateral compartments. The body weight was taken as a nominal 70 kg. The loads were applied at the tibio-femoral contact points, distributed over areas 5 mm diameter, where the contact points for each physiological load condition varied according to the different angles of knee flexion (Andriacchi et al., 1986). When the knee flexed, the contact point in the lateral compartment moved more posteriorly than in the medial compartment.
2.2. Strain-adaptive bone remodelling simulation
2.1. Finite element models set-up The three-dimensional finite element (FE) tibial fixation models were built based on the CT images of the Visible Human Project female left tibia. The proximal segment of the bone, 150 mm long, was modelled as linear elastic and isotropic, with Poisson’s ratio of 0.3. The heterogeneous properties of the bone were defined by mapping the CT greyscale number to the elastic modulus of each bone finite element based on a linear relationship between the CT number and bone apparent density (r) (Abdul-Kadir et al., 2008), and the elastic modulus assigned by E¼3790r3 (Carter and Hayes, 1977). The generic tibial component was based on commercially available TKR designs, which have common features of the polyethylene bearing resting in a flat metal tray, and a central fixation stem 40 mm long by 12 mm diameter and flanges on both sides (Fig. 1). The proximal tibia was virtually resected at 8 mm below the lateral articulating surface of the tibial plateau. The tray and polyethylene insert sizes were dimensioned to provide a maximum coverage of the resected surface as in clinical practise. Their dimensions and material properties are listed in Table 1. Five different cases were modelled (Fig. 1): (i) fully cemented, (ii) hybrid cementing, (iii) cementless no ingrowth—porous-coated tibial prosthesis assuming the worst scenario of no osseointegration, (iv) cementless partial ingrowth—porous-coated prosthesis assuming a realistic amount of osseointegration, (v) cementless fully ingrown—porous-coated prosthesis assuming the best scenario, that bone has ingrown into the entire prosthesis surface. The interfaces were modelled as bonded, where bone cement was applied. For the cementless cases, the bone–prosthesis and bone–stem interfaces with no bony ingrowth were modelled as unbonded by defining contact between them with Coulomb’s friction. The tibial prosthesis was considered to be porous-coated with beads and a coefficient of friction of 0.6 between the bone and prosthesis was assigned (Hashemi et al., 1996). The regions of bone ingrowth for ‘‘partial ingrowth’’ were determined by analysing the zonal area of the prosthesis with interface micromotion lower than 50 mm, where osseointegration is likely (Chong et al., 2010). Osseointegration was predicted to occur on 33% of the surface area of the tray and stem, in line with Turner et al. (1989) and Sumner et al. (1995). The nodes at these bone–prosthesis interfaces were bonded. For the ‘‘fully ingrown’’ case, all the nodes at the bone–prosthesis interfaces were bonded. In hybrid cementing, a friction coefficient of 0.1 was used at the unbonded interface of the bone to a smooth prosthesis stem (Tissakht et al., 1995; Barker et al., 2005). In all cases, the PE was modelled as bonded to the tibial tray. The entire FE model consisted of approximately 35,000 linear tetrahedral elements, and the convergence tolerance was set at 1%. The MSC.Marc 2005 FE software was used. For the boundary conditions, all nodes of the distal cut surface of the tibia were fully-constrained, and the FE model was loaded with a superposition of three different physiological knee joint load conditions corresponding to 0% and 15% of the walking gait cycle (Shelburne et al., 2006), and 25% of the stair climbing cycle (Zhao
Intact Tibia
Fully Cemented
The change in apparent bone density with time was predicted based on a strainadaptive bone remodelling theory (Huiskes et al., 1987) incorporated into FE simulations. The numerical scheme of the remodelling procedure has been detailed in previous studies of Huiskes et al. (1992) and Weinans et al. (1993), where the strain energy density (SED) was used as a mechanical stimulus to regulate the remodelling process. Two FE models, the pre-operative tibia and the post-operative cases were created (Fig. 1) and loaded under the same conditions. When the SED of the postoperative case (Spost) was within a threshold level of the pre-operative (Spre), no change in the bone density would occur. This threshold level, s, (commonly known as ‘‘dead zone’’) was set at 775% of the physiological SED, with which realistic bone remodelling results for hip (Huiskes et al., 1992; Kerner et al, 1999) and femoral knee (Barink et al., 2003; van Lenthe et al., 1997) implantations were achieved. When Spost increased beyond the dead zone, the bone density would increase, and vice versa. After which, bone density of the post-operative tibia would be updated at the end of each loading. The loading iterations and bone-adaptation process continued until Spost and Spre were matched. The rate of change of the bone density is expressed as: 2
A½Spost ð1 þ sÞSpre , 6 0, 4 dr ¼ A½S ð1sÞS , post pre dt
3 spost Z ð1 þ sÞSpre ð1sÞSpre o Spost o ð1 þ sÞSpre 7 5
ð1Þ
Spost r ð1sÞSpre
0:01 r r r 1:73 g=cm
3
Table 1 Dimensions and material properties of the tibial tray and polyethylene insert modelled. Material
Dimensions
Elastic modulus (GPa)
Poisson’s ratio
Tibial tray (titanium alloy)
Medial/lateral width—74 mm Anterior/posterior depth—46 mm Thickness—4 mm Stem length—40 mm Diameter—12 mm Thickness—8 mm
110 Au et al., (2007)
0.33
1 Au et al. (2005)
0.3
Polyethylene insert
Cementless
Hybrid Cementing
Fig. 1. Finite element models of the pre-operative tibia (a) and different tibial fixation techniques. In the exploded views of the prostheses, the upper component is the polyethylene bearing, which rests on the tray of the titanium alloy tibial component. In the fully cemented case (b), an idealised PMMA cement mantle was interposed between the tibial component and the bone. In the cementless case (c), no cement was applied around the tibial component. In the hybrid cementing case (d), a layer of PMMA was interposed beneath the tray of the tibial component, but not around the fixation stem.
950
D.Y.R. Chong et al. / Journal of Biomechanics 44 (2011) 948–954
Table 2 Joint load configurations for bone remodelling simulations. No.
Load case
1 2 3
Load ( body weight)
0% of gait cycle (heel strike) 15% of gait cycle (contra toe-off) 25% of stair climbing cycle a b c
Medial (vertical)
Lateral (vertical)
Antero-posterior
0.9a 2.3a 2c
0.19a 0.46a 1.5c
0.1b 0.08b 1.3b
0 16 32
Shelburne et al. (2006). Taylor et al. (2004). Zhao et al. (2007). Change in Bone Apparent Density with Time
40 Fully Cemented Cementless
30 Change in Apparent Density (%
Knee flexion angle (deg)
Hybrid Cementing Partial Ingrowth
20
Fully Ingrowth
10 0
-10 -20 -30
6
12 36 60 Resected Surface
6
12 36 60 MidStem Time (months)
6
12 36 Stem Tip
60
Fig. 2. Predicted change in the apparent densities of the proximal tibia with time (6, 12, 36 and 60 months after implantation) for different fixation techniques. where A is a constant and the time (t) in months. The above equation can be solved by means of forward Euler’s integration with constant time-steps (Dt) in the FE simulations (Weinans et al., 1993). The density change per time step is calculated from:
Dr ¼ ADt½Spost ð1 7 sÞSpre
ð2Þ
Subsequent to the early studies mentioned above, clinical evidence has suggested that bone suffers ‘memory loss’ and the adaptive process stops after sixty months (Petersen et al., 1995; Saari et al., 2007). Therefore, the bone remodelling following joint replacement was simulated for a period equivalent to sixty months. The remodelling results are presented at three levels in the transverse plane: at the tibial resected surface, at mid-stem and at the stem tip. The apparent density of the bone at each level was computed by averaging the apparent densities of all elements at the particular transverse plane.
fixation (Fig. 3b) occurred by 36 months, while it occurred after 12 months for the fully cemented fixation (Fig. 3a). Greater bone resorption was predicted to occur at the medial condyle, mainly due to the higher joint load acting on the medial side being shielded by the metallic prosthesis. Bone densification was observed at the lateral condyle anterior to the flange for the cementless case, indicating that part of the joint load was transmitted by the tibial tray to the underlying bone. At five years post-operation, a much larger proportion of bone resorption was predicted for the cemented than the cementless fixation at the resected surface and 10 mm underneath. Bone densification was predicted 30 mm below the resected surface and at the stem tip, indicating that the joint load was transferred distally by the cemented stem. For cementless fixation, localised periprosthetic bone densification was observed from the mid-stem level and below. The bone remodelling patterns predicted for the hybrid cementing, partial ingrowth and fully ingrown cases are shown in Fig. 3c–e. Similar to the cementless (no ingrowth) fixation, severe proximal bone resorption began to occur only after 30 months postoperatively for the hybrid and partial ingrowth cases. For the fully ingrown case, severe bone resorption occurred 18 months after tibial implantation. The bone remodelling pattern for the partial ingrowth fixation was similar to the case with no ingrowth, but with a greater amount of proximal bone resorption. With the bone– prosthesis interfaces fully bonded (fully ingrown fixation), the absence of the cement layer led to a prediction of slightly greater bone resorption proximally and at mid-stem, and greater bone densification towards the stem tip when compared to the fully cemented case.
4. Discussion 3. Results The changes in bone apparent density with time after implantation at various levels are illustrated in Fig. 2. At the tibial resected surface, the fully cemented and ingrown cases were predicted to induce the largest amount of proximal bone resorption (26–29% at 60 months), followed by the partial ingrowth case (17%). The cementless (no ingrowth) and hybrid cementing cases experienced the least bone resorption (11%). At the mid-stem level (20 mm below resected surface), bone resorption was observed for the fully cemented, fully ingrown and partial ingrowth cases, though of a lower extent than at the resected surface. Bone densification was predicted for the hybrid case. At the stem tip region, bone densification was predicted for all cases, showing the load transfer capability of the stem. The predicted regions of bone density change at the tibial resected surface and various levels below for all the fixation cases are shown in Fig. 3a–e. Severe bone resorption was defined as where density¼0.01 g/cm3, represented by the light blue region. At the tibial resected surface, severe bone resorption for the cementless
In this study, bone remodelling in the proximal tibia after prosthetic implantation was investigated by finite element simulations, and the influence of fixation techniques (cemented/cementless) were compared. In general, the predictions followed the trend which had been hypothesised. The predictions were compared with published clinical densitometry and radiographic results. At 94 months post-operation, Lonner et al. (2001) found significant bone loss from the proximal tibia for fixation using a fully cemented stemmed tibial prosthesis. The medial tibial plateau experienced a larger amount of bone resorption than the lateral compartment. Consistent with these clinical findings, the bone remodelling simulation predicted severe proximal bone resorption for the fully cemented fixation and greater bone loss at the medial condyle. In other clinical studies, bone hypertrophy was observed at the tip of the stem with hybrid fixation (Bertin et al., 1985) and with uncemented fixation (Whiteside and Pafford, 1989). Bone densification was predicted at the stem tip in the current study. The amount of bone mineral density reduction in the implanted tibia reported by densitometry studies ranged from 10–12% at one to two years post-operatively (Karbowski et al., 1999; Li and Nilsson,
D.Y.R. Chong et al. / Journal of Biomechanics 44 (2011) 948–954
Fully Cemented g/cm
951
Cementless - No Ingrowth
A L Resected surface
M P
10mm below resected surface
20mm below resected surface
30mm below resected surface
Ste m tip
12 months
Time = 0
36 months
60 months
Hybrid Cementing g/cm
Time = 0
Cementless Partial Ingrowth
12 months
36 months
60 months
Cementless Fully ingrown
A L Resected surface
M P
20mm below resected surface
Ste m tip Time = 0
36 months
60 months
Time = 0
36 months
60 months
Time = 0
36 months
60 months
Fig. 3. (a) Bone apparent density plots showing the regions of predicted bone resorption and densification in the transverse plane at the tibial resected surface and various levels below for the fully cemented case. (b) Bone apparent density plots showing the regions of predicted bone resorption and densification in the transverse plane at the tibial resected surface and various levels below for the cementless case. (c). Bone apparent density plots showing the regions of predicted bone resorption and densification in the transverse plane at levels of the tibial resected surface, mid-stem and stem tip for the hybrid cementing case. (d) Bone apparent density plots showing the regions of predicted bone resorption and densification in the transverse plane at levels of the tibial resected surface, mid-stem and stem tip for the cementless partial ingrowth case. (e) Bone apparent density plots showing the regions of predicted bone resorption and densification in the transverse plane at levels of the tibial resected surface, mid-stem and stem tip for cementless fully ingrown case. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)
2000), and up to 23–39% at five to eight years after operation (Levitz et al., 1995; Saari et al., 2007). The current predictions of 11–29% decrease in the proximal bone apparent density at 5 years post-
operation were in reasonable range with the densitometry data, but it becomes difficult to separate the effects of the implant from ageing effects on activity with increasing time.
952
D.Y.R. Chong et al. / Journal of Biomechanics 44 (2011) 948–954
Fig. 3. (Continued)
The density–modulus relationship of Carter and Hayes used in this study had previously been correlated with in-vitro tests on cadaveric hip implants (Abdul-Kadir et al., 2008). However, it is known that this relationship varies with anatomical site (Morgan et al., 2003), and further studies, on a subject-specific basis would be valuable. Three different extents of osseointegration – no ingrowth, partial ingrowth and fully ingrown – were modelled for cementless fixations. In retrieval studies, while extensive levels of bone ingrowth (40–90% of the tray area) was found in canine models (Sumner et al., 1994; Turner et al., 1989), only limited bone ingrowth of less than 30% of the tray area was observed clinically (Ranawat et al., 1986; Sumner et al., 1995). In a worst case, 95% of the tibial components retrieved from patients showed less than 5% of bone ingrowth (Cook et al., 1989). The current FE models thus represented the extremes of clinical circumstances. The different fixation techniques were found to influence the bone resorption via a stress-shielding effect, and so alterations of the load transfer mechanism of the prosthesis may be inferred. For example, with bonded interfaces modelled for the fully cemented and fully ingrown cases, the load acting on the tibial tray could be transferred to the bone distally through shear stresses (Askew and
Lewis, 1981) along the bone–stem interfaces. For the uncemented regions and where bone ingrowth did not occur, the unbonded interfaces modelled with contact surfaces and friction were less able to sustain shear stresses. A firm fixation of the stem which transfers more joint load distally would thereby relieve the proximal bone stresses, consequently leading to proximal bone resorption. A greater amount of bone resorption from 0 to 20 mm below the resected surface was thus observed for the fully cemented fixation than the hybrid case. The rate of proximal bone resorption for partial ingrowth cementless fixation was in-between the full and no ingrown cases, as the bone–prosthesis interfaces were partially bonded (33%). Because the stem was uncemented in both hybrid cementing and no-ingrowth fixations, giving comparable load transfer capability of the stem, similar amounts of proximal bone resorption were predicted. The effect of interface bonding has also been investigated in computational studies of knee femoral component fixations (Barink et al., 2003; van Lenthe et al., 1997), where load transfer phenomena similar to those in the present study were observed. Bone resorption was less extensive when the interfaces between the cement and prosthesis were debonded as compared to when the interfaces were bonded. The bonded interface resulted in the joint forces being transferred more
D.Y.R. Chong et al. / Journal of Biomechanics 44 (2011) 948–954
proximally, via the anterior and posterior flanges, thereby leading to bone densification, and severe bone loss distally within the femoral component (van Lenthe et al., 1997). Greater bone densification underneath the stem tip was observed for the uncemented fixation. This is probably due to the direct contact between the metallic stem tip and bone leading to higher bone stresses than with an intermediate layer of cement. The increase in bone density was in agreement with clinical specimens (Bertin et al., 1985). With a complete cement mantle providing a firm interlock between the bone and prosthesis, the load could be distributed more evenly to the entire periprosthetic bone regions rather than concentrated at specific regions underneath the stem tip, leading to current prediction of lower bone hypertrophy at this zone. For the no/partial ingrowth cementless cases, bone densification was also predicted at the lateral condyle of the tibia, indicating high force transmission from the tibial tray to the bone underneath. While the current bone remodelling simulations have shown good correlations with in-vivo studies, some specific limitations were present. The FE modelling assumed that bone was ingrown onto the porous-coated prosthesis immediately after operation for the partial and fully ingrown fixations. In histological analyses, osseointegration was found to occur between 4 weeks (Søballe et al., 1992) and 12 weeks (Bobyn et al., 1981) after implantation. The ingrowth duration was short compared to the simulated bone remodelling period of 60 months, thus the assumption of bonded interfaces between the bone and prosthesis at time zero of bone remodelling should not have a great influence on the end results. It should, however, be noted that the above histological analyses were conducted on canine models and bone fixation might be slower in clinical use. If the central stem was bonded to the bone at a later stage after operation (hence delayed load transfer effect), proximal bone resorption at 5 years for the partial and fully ingrown cases would be less than the currently predicted values. Because this study was based on only one bone model, it has not been validated across the patient population, and so it is not known whether the different implant fixations will have statistically significant effects; however, this study does provide relative comparisons. The bone was modelled as isotropic, but the diaphysis is anisotropic. Although that is not exact, heterogeneous material assignment was found to be adequate for large bone specimens modelling (Baca et al., 2008). The articular cartilage and menisci were not modelled in the pre-operative tibia, so the physiological SED may not have been computed exactly. These soft tissues help to distribute the stresses across the tibial condyles, so the regions of density change predicted at the tibial resected surface could differ. The vertical and shear loadings were obtained from different independent studies, hence the forces prescribed in the FE models were not consistently from the same group of subjects. This could lead to a likely deviation between the joint loads being applied and those in-vivo. In particular, we assumed that the shear force reported by Taylor et al. (2004) was shared equally between the two compartments. Nevertheless, before fully-detailed knee joint loadings can be measured from the same subjects, the current approach of combining loading conditions from different studies remains practical. The study also assumed no interference fit of the tibial component in the hybrid cementing and no-ingrowth fixations. That is, it is assumed that the interference fit will relax or remodel to a negligible level in a timescale short in comparison to the timescale of most of the remodelling response. If a significant interference fit does exist it would increase the shear loading transfer along the bone–stem interface and therefore result in greater proximal bone resorption than predicted in this study. This study assumed a friction coefficient of 0.1 for smooth stems, as in Tissakht et al. (1995), although 0.42 has also been reported
953
(Shirazi-Adl et al., 1993). A sensitivity analysis (Chong, 2009) found that values from 0.1 to 0.8 had small effects ( o10%) on tibial fixation behaviour. A direct comparison between the different tibial fixation techniques revealed that their influences on bone remodelling were diverse. A firm fixation formed by the bonded interfaces between the prosthesis and the bone for the fully cemented and fully ingrown cases induced proximal tibial bone resorption due to load transfer distally. Greater bone resorption would inevitably weaken the bone supporting the tibial prosthesis, leading to potential problems such as aseptic loosening. Although full cementing may provide initial fixation stability, its long-term consequence on bone quality may be an issue. Bone resorption in tibial fixations with a less secure anchorage (hybrid, no ingrowth and partial ingrowth) were predicted to occur at almost half the rate of the changes around the fixations with a firm anchorage. The fully cemented fixation resulted in the most severe proximal bone resorption, while the partial ingrowth cementless fixation led to moderate changes. The least amount of bone resorption followed hybrid cementing. Thus the hybrid cementing technique or the cementless fixation allowing for partial ingrowth may be a preferred choice for preserving tibial bone stock after TKR.
Conflict of interest statement The authors declare that there is no conflict of interest that could inappropriately influence the results of the current research work.
Acknowledgment The studentship for D.Y.R. Chong was sponsored by the National Medical Research Council, Ministry of Health, Singapore. References Abdul-Kadir, M.R., Hansen, U., Klabunde, R., Lucas, D., Amis, A., 2008. Finite element modelling of primary hip stem stability: the effect of interference fit. Journal of Biomechanics 41 (3), 587–594. Abu-Rajab, R.B., Watson, W.S., Walker, B., Roberts, J., Gallacher, S.J., Meek, R.M.D., 2006. Peri-prosthetic bone mineral density after total knee arthroplasty: cemented versus cementless fixation. Journal of Bone and Joint Surgery, British Volume 88, 606–613. ¨ Adalberth, G., Nilsson, K.G., Bystrom, S., Kolstad, K., Milbrink, J., 2001. Allpolyethylene versus metal-backed and stemmed tibial components in cemented total knee arthroplasty. Journal of Bone and Joint Surgery, British 83 (6), 825–831. Andriacchi, T.P., Stanwyck, T.S., Galante, J.O., 1986. Knee biomechanics and total knee replacement. Journal of Arthroplasty 1 (3), 211–219. Askew, M.J., Lewis, J.L., 1981. Analysis of model variables and fixation post length effects on stresses around a prosthesis in the proximal tibia. Journal of Biomechanical Engineering 103, 239–245. Au, A.G., Liggins, A.B., Raso, V.J., Amirfazli, A., 2005. A parametric analysis of fixation post shape in tibial knee prostheses. Medical Engineering and Physics 27, 123–134. Au, A.G., Raso, V.J., Liggins, A.B., Amirfazli, A., 2007. Contribution of loading conditions and material properties to stress shielding near the tibial component of total knee replacement. Journal of Biomechanics 40, 1410–1416. Baca, V., Horak, Z., Mikulenka, P., Dzupa, V., 2008. Comparison of an inhomogeneous orthotropic and isotropic material models used for FE analyses. Medical Engineering and Physics 30, 924–930. Backstein, D., Safir, O., Gross, A., 2006. Management of bone loss—structural grafts in revision total knee arthroplasty. Clinical Orthopaedics and Related Research 446, 104–112. Barink, M., Verdonschot, N., de Waal Malefijt, M., 2003. A different fixation of the femoral component in total knee arthroplasty may lead to preservation of femoral bone stock. Proceedings IMechE Part H—Journal of Engineering in Medicine 217, 325–332. Barker, D.S., Tanner, K.E., Ryd, L., 2005. A circumferentially flanged tibial tray minimizes bone-tray shear micromotion. Proceedings IMechE Part H—Journal of Engineering in Medicine 219, 449–456. Bertin, K.C., Freeman, M.A.R., Samuelson, K.M., Ratcliffe, S.S., Todd, R.C., 1985. Stemmed revision arthroplasty for aseptic loosening of total knee replacement. Journal of Bone and Joint Surgery, British Volume 67 (2), 242–248.
954
D.Y.R. Chong et al. / Journal of Biomechanics 44 (2011) 948–954
Bobyn, J.D., Pilliar, R.M., Cameron, H.U., Weatherly, G.C., 1981. Osteogenic phenomena across endosteal bone-implant spaces with porous surfaced intramedullary implants. Acta Orthopaedica Scandinavica 52, 145–153. Carter, D.R., Hayes, W.C., 1977. The compressive behavior of bone as a two-phase porous structure. Journal of Bone and Joint Surgery 59-A (7), 954–963. Chong, D.Y.R., 2009. Biomechanical analysis of fixation and bone remodelling of total knee replacement, PhD Thesis, Imperial College London, UK. Chong, D.Y.R., Hansen, U.N., Amis, A.A., 2010. Analysis of bone–prosthesis interface micromotion for cementless tibial prosthesis fixation and the influence of loading conditions. Journal of Biomechanics 43 (6), 1074–1080. Cook, S.D., Barrack, R.L., Thomas, K.A., Haddad, R.J., 1989. Quantitative histologic analysis of tissue growth into porous total knee components. Journal of Arthroplasty 4 (Supplement 1), S33–S43. Davis, A.M., Damani, M., White, L.M., Wunder, J.S., Griffin, A.M., Bell, R.S., 2005. Periprosthetic bone remodeling around a prosthesis for distal femoral tumors—longitudinal follow-up. Journal of Arthroplasty 20, 219–224. Hashemi, A., Shirazi-Adl, A., Dammak, M., 1996. Bidirectional friction study of cancellous bone-porous coated metal interface. Journal of Biomedical Materials Research (Applied Biomaterials) 33, 257–267. Huiskes, R., Weinans, H., Grootenboer, H.J., Dalstra, M., Fudala, B., Slooff, T.J., 1987. Adaptive bone-remodeling theory applied to prosthetic-design analysis. Journal of Biomechanics 20 (11/12), 1135–1150. Huiskes, R., Weinans, H., van Rietbergen, B., 1992. The relationship between stress shielding and bone resorption around total hip stems and the effects of flexible materials. Clinical Orthopaedics and Related Research 274, 124–134. Jazrawi, L.M., Bai, B., Kummer, F.J., Hiebert, R., Stuchin, S.A., 2001. The effect of stem modularity and mode of fixation on tibial component stability in revision total knee arthroplasty. Journal of Arthroplasty 16 (6), 759–767. Karbowski, A., Schwitalle, M., Eckardt, A., Heine, J., 1999. Periprosthetic bone remodelling after total knee arthroplasty: early assessment by dual energy X-ray absorptiometry. Archives of Orthopaedic and Trauma Surgery 119 (8), 324–326. Kerner, J., Huiskes, R., van Lenthe, G.H., Weinans, H., van Rietbergen, B., Engh, C.A., Amis, A.A., 1999. Correlation between pre-operative periprosthetic bone density and post-operative bone loss in THA can be explained by strain-adaptive remodelling. Journal of Biomechanics 32, 695–703. Levitz, C.L., Lotke, P.A., Karp, J.S., 1995. Long-term changes in bone mineral density following total knee replacement. Clinical Orthopaedics and Related Research 321, 68–72. Li, M.G., Nilsson, K.G., 2000. Changes in bone mineral density at the proximal tibia after total knee arthroplasty: a 2-year follow-up of 28 knees using dual energy X-ray absorptiometry. Journal of Orthopaedic Research 18 (1), 40–47. Lonner, J.H., Klotz, M., Levitz, C., Lotke, P.A., 2001. Changes in bone density after cemented total knee arthroplasty—influence of stem design. Journal of Arthroplasty 16 (1), 107–111. Lotke, P.A., Carolan, G.F., Puri, N., 2006. Impaction grafting for bone defects in revision total knee arthroplasty. Clinical Orthopaedics and Related Research 446, 99–103. Morgan, E.F., Bayraktar, H.H., Keaveny, T.M., 2003. Trabecular bone modulus–density relationships depend on anatomic site. Journal of Biomechanics 36, 897–904. Petersen, M.M., Olsen, C., Lauritzen, J.B., Lund, B., 1995. Changes in bone mineral density of the distal femur following uncemented total knee arthroplasty. Journal of Arthroplasty 10 (1), 7–11.
Ranawat, C.S., Johanson, N.A., Rimnac, C.M., Wright, T.M., Schwartz, R.E., 1986. Retrieval analysis of porous-coated components for total knee arthroplasty. Clinical Orthopaedics and Related Research 209, 244–248. ¨ Saari, T., Uvehammer, J., Carlsson, L., Regne´r, L., Karrholm, J., 2007. Joint area constraint had no influence on bone loss in proximal tibia 5 years after total knee replacement. Journal of Orthopaedic Research 25, 798–803. Schai, P.A., Thornhill, T.S., Scott, R.D., 1998. Total knee arthroplasty with the PFC system—results at a minimum of ten years and survivorship analysis. Journal of Bone and Joint Surgery, British 80 (5), 850–858. Seki, T., Omori, G., Koga, Y., Suzuki, Y., Ishii, Y., Takahashi, H.E., 1999. Is bone density in the distal femur affected by use of cement and by femoral component design in total knee arthroplasty? Journal of Orthopaedic Science 4, 180–186. Shelburne, K.B., Torry, M.R., Pandy, M.G., 2006. Contributions of muscles, ligaments, and the ground-reaction force to tibiofemoral joint loading during normal gait. Journal of Orthopaedic Research 24, 1983–1990. Shirazi-Adl, A., Dammak, M., Paiement, G., 1993. Experimental determination of friction characteristics at the trabecular bone/porous-coated metal interface in cementless implants. Journal of Biomedical Materials Research 27, 167–175. ¨ Søballe, K., Hansen, E.S., Rasmussen, H.B., Jørgensen, P.H., Bunger, C., 1992. Tissue ingrowth into titanium and hydroxyapatite-coated implants during stable and unstable mechanical conditions. Journal of Orthopaedic Research 10, 285–299. Sumner, D.R., Kienapfel, H., Jacobs, J.J., Urban, R.M., Turner, T.M., Galante, J.O., 1995. Bone ingrowth and wear debris in well-fixed cementless porous-coated tibial components removed from patients. Journal of Arthroplasty 10 (2), 157–167. Sumner, D.R., Turner, T.M., Dawson, D., Rosenberg, A.G., Urban, R.M., Galante, J.O., 1994. Effect of pegs and screws on bone ingrowth in cementless total knee arthroplasty. Clinical Orthopaedics and Related Research 309, 150–155. Taylor, W.R., Heller, M.O., Bergmann, G., Duda, G.N., 2004. Tibio-femoral loading during human gait and stair climbing. Journal of Orthopaedic Research 22, 625–632. Tissakht, M., Eskandari, H., Ahmed, A.M., 1995. Micromotion analysis of the fixation of total knee tibial component. Computers and Structures 56 (2/3), 365–375. Turner, T.M., Urban, R.M., Sumner, D.R., Skipor, A.K., Galante, J.O., 1989. Bone ingrowth into the tibial component of a canine total condylar knee replacement prosthesis. Journal of Orthopaedic Research 7 (6), 893–901. van Lenthe, G.H., de Waal Malefijt, M.C., Huiskes, R., 1997. Stress shielding after total knee replacement may cause bone resorption in the distal femur. Journal of Bone and Joint Surgery, British 79 (1), 117–122. Weinans, H., Huiskes, R., van Rietbergen, B., Sumner, D.R., Turner, T.M., Galante, J.O., 1993. Adaptive bone remodeling around bonded noncemented total hip arthroplasty: a comparison between animal experiments and computer simulation. Journal of Orthopaedic Research 11, 500–513. Whiteside, L.A., 1993. Cementless revision total knee arthroplasty. Clinical Orthopaedics and Related Research 286, 160–167. Whiteside, L.A., 2001. Long-term followup of the bone-ingrowth ortholoc knee system without a metal-backed patella. Clinical Orthopaedics and Related Research 388, 77–84. Whiteside, L.A., Pafford, J., 1989. Load transfer characteristics of a noncemented total knee arthroplasty. Clinical Orthopaedics and Related Research 239, 168–177. Zhao, D., Banks, S.A., D’Lima, D.D., Colwell Jr., C.W., Fregly, B.J., 2007. In vivo medial and lateral tibial loads during dynamic and high flexion activities. Journal of Orthopaedic Research 25 (5), 593–602.