Introduction

High-flexion knee implants have been developed to accommodate a large range of motion (ROM > 120°) after total knee arthroplasty (TKA). In a recent follow-up study, Han et al. [1] reported a disturbingly high incidence of femoral loosening for high-flexion TKA. The femoral component loosened particularly at the implant-cement interface. Highly flexed knee implants may be more sensitive to femoral loosening as the knee load is high during deep knee flexion [2], which may result in increased tensile and/or shear stresses at the femoral implant fixation.

The objective of this study was to analyse the load-transfer mechanism at the femoral implant-cement interface during deep knee flexion (ROM = 155°). For this purpose, a three-dimensional finite element (FE) knee model was developed including high-flexion TKA components. Zero-thickness cohesive elements were used to model the femoral implant-cement interface. The research questions addressed in this study were whether high-flexion leads to an increased tensile and/or shear stress at the femoral implant-cement interface and whether this would lead to an increased risk of femoral loosening.

Introduction

Within the reconstruction of unicondylar femoral bone defects with morselized bone grafts in revision total knee arthroplasty (TKA), a stem extension appears to be critical to obtain adequate mechanical stability. Whether the stability is still secured by this reconstruction technique in bicondylar defects has not been assessed. Long, rigid stem extensions have been advocated to maximize the stability in revision TKAs. The disadvantage of relatively stiff stem extensions is that bone resorption is promoted due to stress shielding. Therefore, we developed a relatively thin intramedullary stem which allowed for axial sliding movements of the articulating part relative to the intramedullary stem. The hypothesis behind the design is that compressive contact forces are directly transmitted to the distal femoral bone, whereas adequate stability is provided by the sliding intramedullary stem. A prototype was made of this new knee revision design and applied to the reconstruction of uncontained bicondylar femoral bone defects.

Introduction

Hip resurfacing arthroplasty has gained popularity as an alternative for total hip arthroplasty. Usually, cemented fixation is used for the femoral component. However, each type of resurfacing design has its own recommended cementing technique.

In a recent investigation the effect of various cementing techniques on cement mantle properties was studied. This study showed distinct differences in cement mantle volume, filling index and morphology.

In this study, we investigated the effect of these cement mantle variations on the heat generation during polymerization, and its consequences in terms of thermal bone necrosis.

The stability of cemented hip implants relies on the fixation of the cement mantle within the bone cavity. This fixation has been investigated in experiments with cement-bone interface specimens, which have shown that the cement-bone interface is much more compliant than is commonly assumed. Other studies demonstrated that the mechanical response of the interface is dependent on penetration of the cement into the bone. It is, however, unclear how cement penetration exactly affects the stiffness and strength of the cement-bone interface. We therefore used finite element (FE) models of cement-bone specimens to study the effect of cement penetration depth on the micromechanical behavior of the interface.

The FE models were created based on micro computed tomography (micro CT) data of two small cement-bone interface specimens (8x8x4 mm). The specimens had distinct differences with respect to interface morphology. In these models we varied the penetration depth, with six different penetration levels for each model. We then incrementally deformed each model in tension and in shear, until failure of the models. Failure was simulated to occur in the bone and cement when the local ultimate tensile stress was exceeded, by locally reducing the material stiffness to near zero. From the resulting force-displacement curves we established the apparent tensile stiffness and strength for each of the models.

Our results indicated that the strength and stiffness of the cement-bone interface increased with increasing cement penetration depth, both in tension and in shear. However, after reaching a certain penetration depth, both strength and stiffness did not further increase. This depth was dependent on the specific interface morphology. We furthermore found that the strength of the models was higher in shear than in tension. After failure of the models, damage was mainly found in the cement, rather than in the bone.

The FE-based techniques developed for the current study are suitable for exploration of a variety of aspects that may affect the cement-bone interface micromechanics, such as biological changes to the bone and variations of cement material properties.

High-flexion knee replacements have been developed to accommodate a large range of motion (ROM > 120°) after total knee arthroplasty (TKA). Femoral rollback or posterior translation of the femoral condyles during knee flexion is essential to maximise ROM and to avoid bone-implant impingement during deep knee flexion. The posterior cruciate ligament (PCL) has been described as the main contributor to femoral rollback. In posterior-stabilised TKA designs the PCL is substituted by a post-cam mechanism. The main objective of this study was to analyse the mechanical interaction between the PCL and a highflexion cruciate-retaining knee replacement during deep knee flexion. For this purpose, the mechanical performance of the high-flexion cruciate-retaining TKA design was evaluated and compared with two control designs including a highflexion posterior-stabilised design.

Materials & Methods: Prosthetic knee kinematics and kinetics were computed using a three-dimensional dynamic finite element (FE) model of the knee joint. The FE knee model consisted of a distal femur, a proximal tibia and fibula, a quadriceps and patella tendon, a non-resurfaced patella, TKA components and a posterior cruciate ligament in case cruciate-retaining designs were evaluated. Tibio-femoral and patello-femoral contact were defined in the FE knee model and the polyethylene insert was modelled as a non-linear elastic-plastic material. Three different rotating platform TKA systems were analysed in this study: the high-flexion cruciate-retaining PFC Sigma CR150, the high-flexion posterior-stabilised PFC Sigma RP-F and the conventional cruciate-retaining PFC Sigma RP (Depuy, J& J, UK). Both the polyethylene stress characteristics and the tibio-femoral contact locations were evaluated during a squatting movement (ROM = 50° – 150°).

Results: During deep knee flexion (ROM > 120°), the high-flexion cruciate-retaining TKA design showed a lower peak contact stress (74.7 MPa) than the conventional cruciate-retaining design (96.5 MPa). The posterior-stabilized high-flexion TKA design demonstrated the lowest peak contact stress at the condylar contact interface (54.2 MPa), although the post was loaded higher (77.4 MPa). All three TKA designs produced femoral rollback in the normal flexion range (ROM ≤ 120°), whereas the cruciate-retaining designs showed a paradoxical anterior movement of the femoral condyles during high-flexion.

Discussion: PCL retention is a challenging surgical aim and affects the prosthetic knee load and kinematics as shown in this study. In addition, for adequate functioning the PCL should not be too tight or too lax after surgery. Hence, we investigated the effect of PCL laxity on the prosthetic performance and the best-balanced PCL was used in our simulations. Although PCL balancing is not an issue for posterior-stabilized TKA, we found the tibial post to be loaded relatively high for this implant type.

Instability is a major cause for revision surgery in total knee replacement (TKR). With a balanced gap technique, the ligaments are theoretically balanced. However, there is concern that ligament releases needed to align the leg may cause instability. Furthermore, no information is available about the relationship between the amount of varus-valgus laxity directly after implantation and at a later postoperative interval. This prospective clinical study investigated whether ligament releases necessary during total knee replacement (TKR) led to a higher varus-valgus laxity during peroperative examination and after 6 months.

In this prospective cohort study, in 49 patients a primary TKR was implanted using a balanced gap technique. Varus and valgus laxity of the knee was assessed in extension and flexion (70 degrees) per-operative (before and after implant) with a navigation system and post-operative with standardised stress radiographs (both methods 15 Nm stress applied).

Knees were catalogued according to ligament releases performed during surgery: no releases, lateral releases, medial releases with posteromedial condyle (PMC), and medial releases with superficial medial collateral ligament (SMCL). ANOVA was used to test between release groups.

At surgery, before and after implantation of the prosthesis, there was no difference in varus or valgus laxity in extension and flexion between knees that did not need a ligament release (n=22), knees with lateral release (n=5), knees with medial SMCL releases (n=15) and knees with medial PMC releases (n=7). Six months after TKR, varus or valgus laxity in extension and flexion was not significantly different between the release categories.

In conclusion, ligament releases of the SMCL, PMC, and lateral structures performed during a balanced gap technique in TKR do not lead to an increased varus-valgus laxity in extension and flexion at 6 months after surgery. Therefore, routine releases of these structures to achieve neutral leg alignment can safely be performed without causing increased varus-valgus laxity. The results of this study suggest that the reported high incidence of revisions for ligament instability after TKR is not likely to be caused by routine ligament releases when a balanced gap technique is used. Apparently, there is not a ligament instability problem as long as the gaps are properly filled with prosthesis components. We believe that the conclusion of this study would also be valid when bone referenced techniques are applied instead of tensors, as long as the gaps created are balanced.

Bone impaction grafting (BIG) is a surgical technique for the restoration of bone stock loss with impaction of autograft or allograft bone particles (BoP). The goal of a series in-vitro and in-vivo experiments was to assess the suitability of deformable pure Ti (titanium) particles (TiP, FONDEL MEDICAL BV, Rotterdam, The Netherlands) for application as a full bone graft substitute in cemented revision total hip arthroplasty. TiP are highly porous (interconnective porosity before impaction 85 to 90%). In-vitro acetabular reconstructions were made in Sawbones (SAWBONES EUROPE, Malmö, Sweden) to evaluate migration by roentgen stereo photogrammetric analysis and shear force resistance by a lever out experiment. In-vitro femoral TiP reconstructions (SAWBONES, Malmö, Sweden) were used to evaluate micro-particle release and subsidence. Mature Dutch milk goats were used for two in-vivo experiments.

A non-loaded femoral defect model was used to compare osteoconduction of bioceramic coated TiP with BoP and ceramic particles (CeP).

Blood samples were taken for toxicological analysis.

In-vitro: TiP were as deformable as BoP and created an entangled graft layer (porosity after impaction 70 to 75%). Acetabular TiP reconstructions were more stable and resistant to subsidence and shear force than BoP reconstructions (lever-out moment 56 ± 12 Nm respectively 12 ± 4 Nm, p < 0.001). After initial setting, femoral subsidence rates were smaller than seen in femoral bone impaction grafting (0.45 ± 0.04 mm after 300 000 loading cycles). Impaction generated 1.3 mg particles/g TiP (particle Ø 0.7–2 000 μm, tri-modal size distribution). In-vivo: Bioceramic coated (10 −40 μm) TiP showed bone ingrowth rates comparable to BoP and CeP. Reconstructed acetabular defects showed rapid bone ingrowth into the layer of TiP. Serum titanium concentrations slowly increased from 0.60 ± 0.28 parts per billion (ppb) preoperatively to 1.06 ± 0.70 ppb at fifteen weeks postoperatively (p = 0.04).

Mechanical studies showed very good initial mechanical properties of TiP reconstructed defects. The in-vitro study showed micro-particle generation, but in the short-term goat studies, histology showed very few particles and no negative biological effects were found. The in-vivo acetabular study showed very favorable bone ingrowth characteristics into the TiP layer and a much thinner interface with the cement layer compared to similar defects reconstructed with BoP or mixtures of BoP with CeP. Further analysis in a human pilot study should proof that TiP is an attractive and safe alternative for allograft bone in impaction grafting revision arthroplasty.

Balancing the PCL in a PCL-retaining total knee replacement (TKR) is important, but sometimes difficult to execute in an optimal manner. Due to the orientation of the PCL it is conceivable that flexion gap distraction will lead to anterior movement of the tibia relative to the femur. This tibio-femoral repositioning influences the tibio-femoral contact point, which on its turn affects the kinematics of the TKR. So far, the amount of tibiofemoral repositioning during flexion gap distraction is unknown which leads to uncertain kinematic effects after surgery. The goal of this study was to quantitatively describe the parameters of the flexion gap (gap height, anterior tibial translation and femoral rotation) and their relationship while the knee is distracted during implantation of a PCL-retaining TKR with the use of computer navigation. Furthermore, the effect of PCL elevation angle on the flexion gap parameters was determined.

In 50 knees, during a ligament-guided TKR procedure, the flexion gap was distracted with a double-spring tensor with 100 and 200 N after the tibia had been cut. The flexion gap height, anterior tibial translation and femoral rotation were measured intra-operatively using a CT-free navigation system. PCL elevation was calculated based on the femoral and tibial insertion sites as indicated by the surgeon with the pointer of the navigation system.

To identify a relationship between flexion gap height increase and anterior tibial translation, the ratio between anterior translation and gap height increase was determined for each patient between 100 and 200 N.

The mean gap height increased 2.2 mm (SD 0.96) and mean increase in anterior tibial translation was 4.2 mm (SD 1.6). Hence, on average, for each mm increase in gap height, the tibia moved 1.9 mm (SD 0.96) in anterior direction. Knees with a steep PCL showed significantly more AP translation for each mm gap height increase (gap/AP-ratio was 1 : 2.31 (SD 0.63)) compared to knees with a flat PCL (gap/AP-ratio was 1 : 1.73 (SD 0.50)).

The increase in femur (exo)rotation was on average 0.60° (SD 1.4).

With a tensioned PCL the tibia will move anteriorly on average 1.9 mm for every extra mm that the flexion gap is increased. The flexion gap dynamics can be explained in part by the orientation of the PCL: the greater the elevation angle, the more anterior tibial displacement during distraction of the flexion gap. The surgeon must be aware that distraction of the flexion gap influences the tibiofemoral contact point. The tibio-femoral contact point will move posteriorly and stresses in the PCL will rise and produce limited flexion and pain. In case of a conforming insert AP-movement will be limited but high PE stresses may be introduced that can lead to wear. This information may be helpful in selecting the optimal soft tissue balancing procedure and the optimal PE insert thickness in PCL retaining TKR.

Meniscectomy, induces osteoarthritis. Options for repair of a damaged meniscus are an allograft meniscus, an implant made of natural scaffold materials (the collagen meniscus implant; CMI) or an implant made of polymers.

Allograft menisci and the CMI are already clinically used for a considerably number of years. In this educational lecture the focus is on a comparison between the three implant types and the status of a tissue-engineered meniscus.

The allograft meniscus is already used for at least ten years. It is intended for the younger patient with a previous total meniscectomy, with moderate cartilage degeneration and with a good alignment of the knee. The clinical outcome is based on function and pain scores. In this lecture the functional scores, the survival rate and the histology of allograft menisci will be highlighted.

The CMI meniscus implant is intended for a different patient group. To enable implantation of the CMI the rim of the native meniscus should be intact. Patient series that should demonstrate the efficacy of this type of implant are still small and are mainly of the inventors of the implant. In general patients tolerated the implant well. Tissue ingrowth and remodelling into a fibro-cartilaginous tissue was found in animals and patients.

Polymers may be a good alternative for the allograft and CMI implant. Previously they were used to guide vascularized new repair tissue through an ingrowth channel to the avascular lesion. We developed a porous polymer meniscus scaffold with properties to allow tissue infiltration and regeneration of a neomeniscus. It was implanted in dog knees and compared with total meniscectomy. The tissue infiltration and redifferentiation in the scaffold, the stiffness of the scaffold, and the articular cartilage degeneration were evaluated.

Three months after implantation, the implant was completely filled with fibrovascular tissue. After 6 months, the central areas of the implant contained cartilage-like tissue with abundant collagen type II and proteoglycans in their matrix. The foreign-body reaction remained limited to a few giant cells in the implant. The compression modulus of the implant-tissue construct still differed significantly from that of the native meniscus, even at 6 months. Cartilage degeneration was observed both in the meniscectomy group and in the implant group.

The improved properties of these polymer implants resulted in a faster tissue infiltration and in phenotypical differentiation into tissue resembling that of the native meniscus. However, the material characteristics of the implant need to be improved to prevent degeneration of the articular cartilage.

Introduction: Recently, high-flexion knee implants have been developed to provide for a large range of motion (ROM > 120°) after total knee arthroplasty (TKA). High-flexion knee implants are more likely subjected to large knee loads than conventional implants since knee joint forces increase with larger flexion angles. Highly conforming knee replacements are designed to minimise polyethylene peak stresses during (deep) knee flexion.

The Birmingham Knee Replacement (BKR, Jointmedica, UK) is a newly designed knee replacement which combines a high conformity during the complete ROM with the principles of rotating platform and high-flexion TKA. The main objective of this study was to analyze the mechanical performance of the BKR during its full ROM (0°–155°) and investigate whether its high conformity could be maintained during high-flexion. In addition, the BKR polyethylene loading computed in this study was compared with other mobile bearings.

Materials & methods: TKA performance was analyzed using a three-dimensional dynamic finite element (FE) model of the knee joint. The FE knee model consisted of a distal femur, a proximal tibia and fibula, a quadriceps and patella tendon, a non-resurfaced patella and TKA components. Tibio-femoral and patello-femoral contact were defined in the knee model. Three different posterior stabilised rotating platform TKAs were subsequently incorporated: the high-flexion BKR, the high-flexion PFC Sigma RP-F and the standard PFC Sigma RP (Depuy, J& J, USA). The polyethylene insert was modelled as a non-linear elastic-plastic material in each TKA system. Polyethylene loading parameters as well as the tibio-femoral contact point locations were computed during an entire flexion movement (0°–155°).

Results: In the normal flexion range (flexion ≤ 120°) the three knee implants behaved very similar except for the polyethylene loading at the post. At 120° of flexion, the contact stress at the dish was ±45 MPa for all implants whereas the maximal post-cam contact stress came down to 26.7 MPa for the BKR which was half the amount of contact stress experienced by both PFC Sigma implants. During high-flexion (flexion > 120°), the contact stress difference at the post between the BKR and the PFC Sigma RP-F became smaller and came down to 37.9 MPa and 60.7 MPa, respectively. The total amount of plastic deformation at maximal flexion (155°) was smaller for the BKR (577 mm3) in comparison with the Sigma RP-F (2256 mm3). Femoral rollback was negligible for the BKR in the high-flexion range in comparison with the Sigma RP-F (1.9 mm).

Discussion: A comparison between different geometrical models using finite element techniques is jeopardised by differences in element distribution within the various models. These differences may affect calculated parameters such as peak stress values. However, in this study the models were very similar which would indicate that the differences in stress patterns found are due to design differences rather than model artefacts.

The current study therefore indicates that the BKR benefits from its high conformity during the full ROM. Hence, the BKR demonstrated relatively low polyethylene stresses. The quadriceps efficiency during deep knee flexion may be lower in case of the BKR since the femoral rollback was negligible at these flexion angles. Whether this phenomenon is of any clinical relevance is unknown.

We investigated the feasibility of using porous titanium particles (TiP) to reconstruct femoral bone defects in revision hip replacement surgery in stead of using morzelised bone grafts. Questions regarding handling, initial stability and titanium particle release were addressed.

Seven composite femurs (Sawbones) were reamed and filled, stepwise, with 32 grams of large (Ø 3.15 – 4 mm) and 9 grams of smaller (Ø 2.8 – 3.15 mm) pure, 85% porous TiP. Subsequently an Exeter stem was cemented into the graft layer. All reconstructions were loaded axially (0–3000 N) for 300,000 loading cycles at 2 Hz. Subsidence of the stem was measured with radio stereometric analysis (RSA) and possible titanium particle release was measured using the laser diffraction technique.

The TiP were impacted into a > 3 mm (SD 1.43 mm) thick, highly entangled, graft layer. An average cement mantle of > 2 mm (SD 0.86 mm) was measured and little cement penetration was observed. The average subsidence of only 0.45 mm (SD 0.04 mm) was measured after 300 000 loading cycles. Most titanium particles were found directly after impaction. Most of these particles (87%) were smaller than 10 μm and could therefore be potentially harmful since they can induce osteolysis.

We can conclude that:

A graft layer of impacted TiP can be constructed,

These data warrant further animal tests to assess the biological response to these released impaction particles. Also, animal tests should clarify possible particle release upon loading and its effects.

Recently, high-flexion knee implants have been developed to provide for a large range of motion after total knee arthroplasty. Since knee forces increase with larger flexion angles, it is commonly assumed that high-flex-ion implants are subjected to large loads in the highflexion range (flexion > 120°). However, high-flexion studies often do not consider thigh-calf contact which occurs during high-flexion activities such as squatting and kneeling. We hypothesized that thigh-calf contact is substantial and has a reducing effect on the prosthetic knee loading during deep knee flexion.

The effect of thigh-calf contact on the loading of a knee implant was evaluated using a three-dimensional dynamic finite element knee model. The knee model consisted of a distal femur, a proximal tibia and fibula, a patella, high-flexion components of the PFC Sigma RP-F (Depuy, Warsaw, USA) and a quadriceps and patella tendon. Using this knee model, a squatting movement was simulated including thigh-calf contact characteristics of a typical subject which have been described in an earlier study.

Thigh-calf contact considerably reduced the implant loading during deep knee flexion. At maximal flexion (155°), the compressive knee force decreased from 4.9 to 2.9 times bodyweight. The maximal joint forces shifted from occurring at maximal flexion angle to the flexion angle at which thigh-calf contact initiated (±130°). The maximal polyethylene contact stress at the tibial post decreased from 49.3 to 28.1 MPa at maximal flexion.

This study confirms that thigh-calf contact reduces the knee loading during high-flexion. Both the joint forces and the polyethylene stresses reduced considerably when thigh-calf contact was included.

Bone metastases occur in about 15% of all cancer cases. Pathological fractures that result from these tumours most frequently occur in the femur. It is extremely difficult to determine the fracture risk with the current X-ray methods, even for experienced physicians. The purpose of this study was to assess whether the use of a predictive finite element model could improve the prediction of strength in comparison to an clinical assessment.

Eight human cadaver femora, with and without simulated metastases, were CT-scanned. A solid calibration phantom was included in each scan. From the scans, eight finite element (FE) models were generated using brick elements. The non-linear mechanical properties were based on bone density. After scanning, laboratory experiments were performed. The femora were loaded under compression until failure. During the experiments the failure forces and the course of failure were registered. These experiments were simulated in the FE-models, in which plastic deformation simulated failure of the bones. Six experienced physicians, were asked to rank the femora on strength using X-rays (AP and ML) and additional information on gender and age.

The results showed a strong Pearson’s correlation (r2 = 0.92) between the experimental failure force and predicted failure force. The Spearman’s rank correlations between experiment and predictions ranged between ρ=0.58 and ρ=0.8 for the physicians, whereas it was significantly higher (ρ=0.92) for the FE-model

This study showed that femur specific FE models better predicted femoral failure risk under axial loading than experienced physicians. When the model is further improved by adding, for example, other loading conditions, it can be clinically implemented to predict in vivo fracture risk for patients suffering, for example, bone metastases or osteoporosis.

We present an update of the clinical and radiological results of 62 consecutive acetabular revisions using impacted morsellised cancellous bone grafts and a cemented acetabular component in 58 patients, at a mean follow-up of 22.2 years (20 to 25). The Kaplan-Meier survivorship for the acetabular component with revision for any reason as the endpoint was 75% at 20 years (95% confidence interval (CI) 62 to 88) when 16 hips were at risk. Excluding two revisions for septic loosening at three and six years, the survivorship at 20 years was 79% (95% CI 67 to 93). With further exclusions of one revision of a well-fixed acetabular component after 12 years during a femoral revision and two after 17 years for wear of the acetabular component, the survivorship for aseptic loosening was 87% at 20 years (95% CI 76 to 97). At the final review 14 of the 16 surviving hips had radiographs available. There was one additional case of radiological loosening and four acetabular reconstructions showed progressive radiolucent lines in one or two zones.

Acetabular revision using impacted large morsellised bone chips (0.5 cm to 1 cm in diameter) and a cemented acetabular component remains a reliable technique for reconstruction, even when assessed at more than 20 years after surgery.

Introduction: Finite element (FE) simulation of damage accumulation in the femoral cement mantle is widely used to predict failure of hip prostheses. It is often assumed that the stem-cement interface remains bonded, although debonding is thought to affect cement stress and damage. Rough stems may reduce subsidence, but have been reported to have a detrimental effect on implant survival. Other factors thought to influence cement damage include stem design and orientation and cement thickness. This study investigates the effect of cement mantle thickness and stem malpositioning on cement damage around a smooth, collared implant, and the extent to which this is affected by debonding of the stem-cement interface.

Method: Three FE meshes were built to represent proximal femora with Stanmore Hip prostheses implanted into a thick (2.5 mm) and a thin (1.0 mm) cement mantle, and another thin (1.0 mm) mantle with the implant tilted in varus to achieve a minimal thickness of 0.1 mm laterally. Each model consisted of 4304 eight-noded brick elements with frictional contact at the stem-cement interface. Two analyses were run for each model, in which the stem-cement interface was (a) fully bonded, and (b) fully debonded, with Coulomb frictional contact using a friction coefficient of 0.5. Standardised femur geometry and elastic properties were used. Creep and non-linear damage accumulation in the cement mantle under cyclic loading was modelled using subroutines developed by Stolk et al. (2003). Boundary conditions were applied representing a peak stair-climbing load.

Results: Bonded cases showed extensive cracking around the tip in all cases. Debonded cases had 4–8 times less cracking, which was much more focused at the tip; only the poorly-centralised mantle showed extensive damage elsewhere, in the very thin lateral region. When bonded, the thick mantle had least cracks and the poorly-centralised mantle had most; in the debonded cases, there was no major difference between thick, thin, and poorly-centralised mantles. For each cement mantle geometry, peak maximum principal cement stress was consistently lower in the debonded case than in the bonded case.

Discussion: Our results show greater, more widely distributed cracking in bonded than debonded cement mantles, in contrast with previous studies involving collarless implants. For a collared stem, calcar contact prevents subsidence, allowing cement stress relaxation. A possible explanation for our result is that debonding enhances the stress relaxation process, reducing and redistributing interfacial and shear stresses; thus reducing damage rates. In contrast, a debonded collarless stem subsides continuously, sustaining high cement stress levels and damage rates. These results may explain the disappointing clinical performance of some rough-surfaced prostheses. Our results suggest that bonding might increase both cement damage and its sensitivity to cement thickness. Similar results for all debonded cement mantles indicate that cement thickness may be less critical than previously thought for smooth, collared prostheses. Bonding should not be assumed in FE studies of smooth stems which clinically are likely to debond; cement damage simulation should be extended to incorporate the debonding process.

Introduction: It is generally accepted that the cement mantle surrounding femoral hip implants should be at least 2–3 mm thick. To achieve that goal, manufactures or surgeons often undersize the stem compared to the broach. However, some implants, such as the Charnley-Kerboul stem, are typically cemented line-to-line i.e. with a broach and stem of the same size. Despite their “minimal” cement mantle, these stems are very successful. This apparent contradiction is known as the “French Paradox”[1]. We used a finite element analysis (FEA) model to investigate the effect of these different cementation philosophies on cement crack propagation and rotational stem stability.

Material and Methods: Based on a CT-scan image of a Charnley-Kerboul plastic stem replica[2], twelve FEA models were created. By decreasing the stem size (4 stems), the average cement mantle thickness increased (1.71–3.77mm). However, the incidence of cement mantle defects (< 1mm) and areas of thin cement (< 2mm) decreased (defects: 34.7–0.0%; thin cement: 40.7–0.0%). The amount of cortical bone support was varied (3 times) between 18.4 and 72.2%. All models were alternately loaded with a cyclic torque load (25.8Nm) and a transversal load (400N) in a ratio of 9:1 for two million cycles. The model predicted fatigue crack formation within the cement and rotational stem stability.

Results: Overall, increasing implant size and increasing the amount of cortical bone support to the cement, improved resistance to accumulated cement damage and rotational stem stability. In both models with undersized stems, more cement cracks and full thickness (FT) cement fractures appeared after less loading cycles than in both models with canal-filling stems. Worst results were obtained with a severely undersized implant surrounded by a thick cement mantle that was poorly supported by cortical bone (first FT crack after < 100 000 cycles, > 220 initiated cracks and 0.6° of implant rotation after 2 million cycles). Best results were obtained with the maximal canal-filling stem surrounded by a thin and deficient cement mantle that was well supported by cortical bone (no FT cracks, < 10 initiated cracks and 0.3° of implant rotation after 2 million cycles).

Conclusion: This study emphasizes the importance of an adequate cementation technique that aims at pressurizing cement up to the cortical bone. This protects the cement mantle against fatigue fracture and stabilises the implant especially if the stem is undersized. From a mechanical point of view, canal-filling stems make sense. They limited the formation of cement cracks and improved rotational stability to the implant. This could explain the excellent results obtained by implants that are cemented line-to-line.

Contemporary knee implants use a variety of methods to control tibiofemoral motions. Posterior stabilized implants have a post and cam to force the femur posterior with flexion. Most posterior cruciate retaining designs rely solely on this ligament and symmetric tibial surfaces to control tibiofemoral translations. However, many studies have demonstrated poor control of tibiofemoral motion in PCL retaining knees. One strategy to augmenting PCL function is to provide a gait-congruent lateral articulation providing definitive stability in extension while allowing lateral condylar translation in deep flexion. It is unknown whether this design strategy, essentially substituting for the ACL, allows the PCL to function more normally.

Fifteen knees in ten patients with a fixed-bearing, PCL retaining, lateral pivot arthroplasty were observed during maximum flexion kneeling and lunging using fluoroscopy. The tibial insert provides a fully conforming lateral articulation from 0°–70° flexion, allowing lateral AP translation at greater flexion. Recruited on the basis of combined KSS scores > 180 points, patients averaged 72 years, 27.5 BMI, and 12 months post-op. Shape matching techniques were used to determine the 3D pose of the implant components.

Skeletal flexion during kneeling averaged 134° (117°–156°) with 11° tibial internal rotation. Medial condylar contact was 3mm posterior, and lateral contact was 11 mm posterior to the tibial AP midpoint. Skeletal flexion during lunging averaged 122° (106°–146°) with 11° tibialinternal rotation. Medial condylar contact was 1mm posterior, and lateral condylar contact was 9mm posterior to the tibial AP midpoint.

Knees with lateral pivot arthroplasty exhibited flexion comparable to the best reported results in North American patients. Tibial rotation was statistically greater than has been reported for symmetric posterior stabilized or PCL retaining implants for the same activities. Posterior translation of the condyles with flexion beyond the range of full articular congruity is consistent with relatively normal PCL function.

Most minimal invasive surgical (MIS) systems use traditional implant systems combined with new instrumentation. In this study we analyzed a THR system that basically implies that all components are implanted through the femoral neck. The cemented femoral component consists of a highly polished tapered design. The acetabular component is made of Alumina and has an outside diameter of 20 mm. The purpose of this study was to investigate the range of motion, the wear characteristics, the fatigue characteristics of the femoral neck and the stability of the femoral component.

The range of motion of the MIS prosthetic system was calculated with a mathematical model that enabled calculation of prosthetic impingement angles. To assess the wear properties, four pairs of Zirconia heads on alumina acetabuli were tested in a hip simulato. To assess the probability of femoral neck fracture, 3 components were tested according to ISO7206. The stability of the femoral components were tested in five fresh cadaver using dynamic loading conditions. After this test, the load was increased until reconstructive failure occurred.

The ROM was in the order of 100 degrees of flexion and at least 30 degrees in other directions. The bearings showed remarkably low wear with a maximum of 0.02 mm3. All three stems survived the ISO-fatigue test. During the dynamic experiments the specimens did not fail, and no macroscopic damage was detected. Migration was only minor and stabilized during testing. The post-testing failure loads varied between 4.1 and 5.5 kN.

The ROM, stem-neck strength and wear properties of the system seem acceptable. The stability of the femoral component was satisfying; but the post-testing strength may be similar to loads that are applied on the hip at a falling accident. We conclude that these results are encouraging and warrant further studies to develop this system.

It is impossible to determine the effect of a single parameter in clinical or in-vitro knee research. There are also parameters which can not or hardly be determined. These disadvantages can be overcome with a model. The objective of this study was to create a dynamic FE model of a human knee joint after TKA which is applicable to a variety of research question.

The knee model consisted of a femur, tibia and patella, collateral ligaments and a PCL, combined with a CKS cruciate retaining total knee prosthesis. The patella was not resurfaced. An axialload of 150 N and a quadriceps-force of 81N was applied. The model was validated by the model prediction of joint laxities at different flexion-angles and the calculation of the knee kinematics during flexion-extension.

The predicted varus-valgus laxity at different flexion angles was in between 0 and 6.3 degrees. Laxity values decreased towards extension and towards 90 degrees of flexion. The AP test at 20, 30 and 90 degrees of flexion showed a anterior laxity of 3.1, 4.3 and 2 mm, respectively. The posterior laxity was 5.7 mm, but could only be determined at 90 degrees. The model predicted reasonable kinematics, which were identical for two consecutive flexion-extension movements.

The model predictions were well in agreement with reported values, which were measured experimentally. Differences could be well explained by ligament structures which were (still) omitted with in the model. This dynamic model, in which ligaments were actually modelled as bands, combined all major structures within the knee joint. It was well able to predict laxities and kinematics and turned out to be very stable, mathematically. With this model we will be able to address effects of prosthetic and surgical parameters on the stability and kinematics of the knee joint.

Introduction: The purpose of this study was to evaluate the long-term clinical and radiological outcome of instrumented femoral revisions after failed total hip arthroplasties using the impaction bone grafting technique with morsellized bone chips in combination with a cemented polished stem.

Methods: Thirty-three consecutive femoral reconstructions were performed between November 1991 and February 1996 using the X-Change femoral impaction system with fresh frozen morsellized bone grafts and a cemented polished Exeter stem. All patients were prospectively followed. The learning curve with this new technique is included in this report. This technique was used in twenty-four women and nine men; the average age at surgery was sixty-three years (range 33–82). Femoral bone stock defects were classified according to the Endoklinik classification as grade 1 in three hips, grade 2 in fourteen hips, grade 3 in twelve hips and grade 4 in four hips. At a minimal eight years follow-up no patient was lost to follow-up, but eight patients died (at 0.5, 3.5, 3.5, 7.0, 7.0, 7.5 and 9.0 after reconstruction). All were followed until death, none of these deaths was related to the surgery, and none had a re-revision

Results: No femoral reconstruction was re-rerevised at a mean follow-up of 10.4 years (range 8 to 13 years). However, there were three femoral fractures during follow-up (at 3, 6 and 22 months), all at the level of tip of the prostheses. All healed after plating, all femoral implants were left in situ. The average Harris hip score improved from 49 prior to surgery to 85 at review (68–100). The average migration of the stem within the cement mantle was three mm (0–14 mm), most migration was seen in the first year. Radiologically, there were no failures. With an endpoint of femoral revision for any reason, with endpoint aseptic loosening or with endpoint radiological loosening the survival rate using the Kaplan-Meier analysis was hundred per cent in all situations (one-sided 95% C.I. 100–91.3 %)

Conclusions Femoral revision using bone impaction grafting with fresh frozen bone grafts and a cemented polished stem showed an excellent survival at eight to thirteen years follow-up.

The Capital Hip implant was a Charnley-based system which included a flanged and a roundback stem, both of which were available in stainless steel and titanium. The system was withdrawn from the market because of its inferior performance. However, all four of the designs did not produce poor rates of survival. Using a simulated-based, finite-element analysis, we have analysed the Capital Hip system. Our aim was to investigate whether our simulation was able to detect differences which could account for the varying survival between the Capital Hip designs, thereby further validating the simulation.

We created finite-element models of reconstructions with the flanged and roundback Capital Hips. A loading history was applied representing normal walking and stair-climbing, while we monitored the formation of fatigue cracks in the cement.

Corresponding to the clinical findings, our simulation was able to detect the negative effects of the titanium material and the flanged design in the Capital Hip system. Although improvements could be made by including the effect of the roughness of the surface of the stem, our study increased the value of the model as a predictive tool for determining failure of an implant.

Clinical experience of impaction bone grafting for revision knee arthroplasty is limited, with initial stability of the tibial tray emerging as a major concern. The length of the stem and its diameter have been altered to improve stability. Our aim was to investigate the effect of the type of stem, support of the rim and graft impaction on early stability of the tray.

We developed a system for impaction grafting of trays which we used with morsellised bone in artificial tibiae. Trays with short, long thick or long thin stems were implanted, with or without support of the rim. They were cyclically loaded while measuring relative movement.

Long-stemmed trays migrated 4.5 times less than short-stemmed trays, regardless of diameter. Those with support migrated 2.8 times less than those without. The migration of short-stemmed trays correlated inversely with the density of the impacted groups. That of impaction-grafted tibial trays was in the range reported for uncemented primary trays. Movements of short-stemmed trays without cortical support were largest and sensitive to the degree of compaction of the graft. If support of the rim was sufficient or a long stem was used, impacted morsellised bone graft achieved adequate initial stability.

This study presents the clinical and radiological results of 62 consecutive acetabular revisions in 58 patients, at a mean of 16.5 years follow-up (15 to 20). The Kaplan-Meier survivorship for the cup with end-point revisions for any reason, was 79% at 15 years (95% confidence interval (CI); 67 to 91). Excluding two revisions for septic loosening at three and six years, and one revision of a well-fixed cup after 12 years in the course of a femoral revision, the survivorship was 84% at 15 years (95% CI; 73 to 95). At review there were no additional cases of loosening, although seven acetabular reconstructions showed radiolucent lines in one or two zones. Acetabular revision using impacted large morsellised bone chips (0.7 cm to 1.0 cm) and a cemented cup, is a reliable technique of reconstruction, when assessed at more than 15 years.

The increasing success rates of total hip replacements (THR) have led to a younger patient population with an increased probability for revision. The survival of revised components is improved by a good bone quality. This has led to an increased interest in bone preserving THR designs. A novel type of THR was developed of which the femoral component is cemented in the neck. The load carrying area of this prosthesis is reduced in comparison with conventional cemented implants. Whether an adequate stability can be achieved was biomechanically evaluated during simulated normal walking and chair rising. In addition, the failure behaviour was investigated.

Bone mineral density (BMD) was measured in 5 fresh frozen proximal human cadaver femora. The femoral heads were resected and a 20 mm diameter canal was created in the femoral necks. Bone cement was pressurised in this canal and the polished, taper-shaped prosthesis was subsequently introduced centrally. A servohydraulic testing machine was used to apply dynamic loads up to 1.8 kN to the prosthetic head. Radiostereophotogrammetric analysis was used to measure rotations and translations between prosthesis and bone. In addition, the constructions were loaded until failure in a displacement-controlled test.

During the dynamic experiments, the femoral necks did not fail, and no macroscopical damage was detected. The initial stability of the implant did not seem to be sensitive to bone quality. Maximal values were found for normal walking with a mean rotation of about 0.2 degrees and a mean translation of about 120 microns. These motions stabilised during testing. The failure loads in this study varied between 4.1 and 5.5 kN, higher failure loads were associated with higher BMD values. Most specimens showed subtrochanteric spiral fractures.

In conclusion, the stability of the prosthetic device may be adequate under dynamic, physiological loading conditions. The static failure loads were relatively low and require further optimisation of the prosthetic implant.

Malfunctioning of Total Knee Replacements is often related to patella-femoral problems. As the patella groove guides the patella during flexion, the difference between anatomic- and prosthetic groove geometry may be of major influence concerning patella-femoral problems. This study focusses on the orientation or direction of the femoral patella groove, relative to the mechanical axis of the femur. Literature shows a controversy in measured groove orientation: Eckhoff et al. (1996) have measured a lateral groove, and Feinstein et al. (1996) have measured a medial groove, relative to the mechanical axis. Current femoral knee components have a lateral, or neutral directed patella groove. As most TKA surgical techniques subscribe an exorotation of the femoral component during implantation, the prosthetic in vivo situation will show a lateral groove. The objectives were to clarify the described controversy and to determine whether there is a difference in anatomic- and prosthetic groove orientation, which might cause patella-femoral problems.

The patella groove orientation of 100 human femora was measured using a 3-D measurement system. A spherical measurement probe was moved through the groove, starting at the notch and finishing at the cartilage edge, to simulate patella motion. The patella groove angle was defined as the angle between the mechanical axis and the measured groove points, in the frontal plane. A medial patella groove angle of 1.8±2.6° was measured. An implanted situation of a femoral component with neutral groove showed a lateral groove angle of 1.3°. An implanted situation of a femoral component with assymmetrical groove showed a lateral groove angle of 2.6°. The authors measured a medial oriented patella groove. This anatomical groove orientation is in contradiction with current femoral knee component design and surgical practice, because that results in a lateral oriented groove. This difference in anatomic- and prosthetic groove orientation may be a cause of patella-femoral problems.

Bone impaction grafting of the femur is associated with more complications when segmental defects are present. The effect of segmental defect repair on initial stem stability was studied in an in vitro study with fresh frozen goat femora. A standardized medial segmental defect was reconstructed using a cortical strut or a metal mesh. As controls we used intact femora and femora with a non-reconstructed defect. In all four groups impacted bone grafting was performed in combination with a cemented Exeter stem. Each group contained five femora. Reconstructions were dynamically loaded up to 1500N. Migration was measured with Roentgen Stereo-photogrammetric Analysis. All cases with a non-reconstructed segmental defect failed into excessive varus rotation. None of the femora with a reconstructed defect failed. Cortical struts and metal meshes were equally effective in creating a stable stem construction (varus rotation 2.89±2.27 and 2.27±0.57, respectively). Reconstructions with a metal mesh were more reproducible, although the obtained stability was significantly lower (p< 0.01) when compared to impaction grafting in an intact femur (varus rotation 0.58±0.36).

Besides, structural grafts may negatively influence the revascularization of the underlying impacted grafts in contrast to an open wire mesh. So, an in vivo study of 12 goats was done. A standardized medial wall defect was reconstructed with a strut or a mesh in six goats per group. In all femora impaction grafting was performed in combination with a cemented Exeter stem. After six weeks the femora were harvested. A high rate of peri-prosthetic fractures was found (43% and 29% for the strut and mesh groups, respectively). Histological and micro-radiological examination showed different revascularization patterns for both reconstruction techniques. In the strut group revascularized graft was found at the edges of the defect. In the mesh group fibrous tissue and blood vessels penetrated through the mesh and a superficial zone of revascularized grafts was found. Segmental defect reconstruction with a strut reduced the amount of revascularized grafts medially behind the strut (p=0.004). This may interfere with the stability of the stem in the first period after surgery and the incorporation of the impacted grafts on the long-term.

We would recommend segmental defect reconstruction with a mesh. A regime of unloading and long-stem prostheses should be used, irrespective of the reconstruction technique

Aims: The purpose of this study is to validate a novel, quantitative functional test for TKA patients using commonly used subjective questionnaires as standard. Methods: Electrogoniometry was used to get information about the stability of rising from a chair. Thirteen pre Ðoperative TKA patients were measured while rising from a chair at two different heights. Using phaseplots (relationship between joint angle and joint velocity), the relative phase between the hip and knee was calculated. Instability was deþned as the standard deviation over ten trials for every patient and each height. An independent physician administered the WOMAC, KSS and SF-36 questionnaires. The instability factor of rising was correlated (using Spearmanñs Rho) to the outcomes of the questionnaires. Results: The stability of rising correlated moderately high with the questionnaires. Best correlations were calculated for instability and the WOMAC and SF-36 physical functioning (0.45 and 0.56, respectively). The worst correlation was with the KSS pain score (0.05). In general, the correlations for the lowest chair height were slightly higher. Conclusions: As there is no gold standard for functional evaluation of the TKA, a chair rise protocol has been developed. In this study the stability of rising correlates moderately high with the WOMAC, KSS and SF-36 questionnaires. The stability of rising from a lower chair seems to discriminate better than rising from a higher chair. This test will be used as a quantitative functional follow up of TKA patients.

Aims: The effects of bone graft washing and bone graft particle size on initial cup stability in an acetabular model were studied. Methods: From fresh human femoral heads large (8–12 mm) and small (2–4) grafts were produced. Four different groups were tested; large washed, large unwashed, small washed and small unwashed. An AAOS type 3 combined defect was reconstructed in a synthetic acetabular model using a metal mesh and screws. Bone grafts were impacted in a standardized way using a hammer and metal impactors. Cement was pressurized and a PE cup was inserted. Cups were dynamically loaded with 1500 N and 3000 N for 15 minutes using an MTS-device, RSA was used to calculate 3-D cup migrations. Next a lever-out test of the cup was done in a displacement (16¡/minute) controlled test. After testing CT scans of the cups were made to quantify cement layer roughness, as a measurement of cement penetration. Results: Cups reconstructed with large washed bone grafts migrated signiþcantly less during mechanical compression and withstood signiþcantly higher lever-out moments. Washing improved cup stability for both small and large grafts, however only signiþcantly for the larger bone grafts. A signiþcantly higher cement penetration depth was found for all large bone grafts, washing did not inßuence cement penetration depth. Discussion: For acetabular revisions with bone impaction grafting we recommend to use large washed bone grafts.

Proper pre-clinical testing of cemented THA implants may help to prevent bad implants from entering the market. Within the frame of a multinational EU-program, a finite element (FE) simulation was developed, for FE-based pre-clinical testing of cemented THA stems against the damage accumulation failure scenario. The simulation allows monitoring of cement crack formation and implant migration in cemented THA reconstructions. The current study is concerned with the clinical validation of the test. The damage accumulation failure scenario was simulated for four cemented hip stems, with well-known survival rates. The question was: Can the FE simulation rank the stems according to their clinical survival rates?

Four stems were analysed: the Lubinus SPII, the Exeter, the Charnley and the Mueller Curved. The Swedish hip register [1] reports survival rates of 4, 5, 8 and 13%, respectively, at 10 years after surgery. Four FE models were created, representing cemented THA reconstructions with the four stems in composite femurs. The stem-cement interfaces were unbonded (m = 0.25). A loading history was applied to the models, representing 20 million cycles of alternating walking and stair climbing. Using a 3D continuum damage mechanics approach, the damage accumulation and creep processes in the cement, and subsequent prosthetic migration were monitored.

The Mueller C. produced a considerably higher number of cement cracks than the other three stems. Cracks were formed around the entire stem. The cracked zones often extended over the thickness of the mantle. The Charnley performed better, with a lower number of cracks. Proximo-distal damage pathways were formed, although at a much lower rate than around the Mueller C. The Exeter performed better. Full thickness crack zones were produced only in the proximo-medial region. The Lubinus performed best, with the lowest number of cement cracks. No full thickness cracks were formed. Concerning migration, the Exeter migrated more than the other stems. From the collared implants, the Lubinus SPII showed the lowest migration values.

When considering the number of cement cracks produced in the simulation, the ranking of the stems would be, from superior to inferior: Lubinus SPII, Exeter, Charnley, Mueller Curved. This ranking corresponds to a ranking based on clinical survival rates. The stems behaved according to their design concepts, with the highest migration values for the Exeter stem. In conclusion, the FE simulations produced a clinically valid ranking of four cemented THA implants. This corroborates the use of the FE simulation for pre-clinical testing purposes.

The properties of impacted morsellised bone graft (MBG) in revision total knee arthroplasty (TKA) were studied in 12 horses. The left hind metatarsophalangeal joint was replaced by a human TKA. The horses were then randomly divided into graft and control groups. In the graft group, a unicondylar, lateral uncontained defect was created in the third metatarsal bone and reconstructed using autologous MBG before cementing the TKA. In the control group, a cemented TKA was implanted without the bone resection and grafting procedure. After four to eight months, the animals were killed and a biomechanical loading test was performed with a cyclic load equivalent to the horse’s body-weight to study mechanical stability. After removal of the prosthesis, the distal third metatarsal bone was studied radiologically, histologically and by quantitative and micro CT.

Biomechanical testing showed that the differences in deformation between the graft and the control condyles were not significant for either elastic or time-dependent deformations. The differences in bone mineral density (BMD) between the graft and the control condyles were not significant. The BMD of the MBG was significantly lower than that in the other regions in the same limb. Micro CT showed a significant difference in the degree of anisotropy between the graft and host bone, even although the structure of the area of the MBG had trabecular orientation in the direction of the axial load. Histological analysis revealed that all the grafts were revascularised and completely incorporated into a new trabecular structure with few or no remnants of graft. Our study provides a basis for the clinical application of this technique with MBG in revision TKA.

The use of impacted, morsellised bone grafts has become popular in revision total hip arthroplasty (THA). The initial stability of the reconstruction and the effectiveness of any subsequent process of revitalisation and incorporation will depend on the mechanical integrity of the graft. Our aim in this study was to document the time-dependent mechanical properties of the morsellised graft. This information is useful in clinical application of the graft, in studies of migration of the implant and in the design of the joint.

We used 16 specimens of impacted, morsellised cancellous bone from the sternum of goats to assess the mechanical properties by confined compression creep tests. Consideration of the graft material as a porous, permeable solid, filled with fluid, allowed determination of the compressive modulus of the matrix, and its permeability to fluid flow.

In all specimens the compression tests showed large, irreversible deformations, caused by flow-independent creep behaviour as a result of rolling and sliding of the bone chips. The mean permeability was 8.82 *10⁻¹² m⁴/Ns (SD 43%), and the compressive modulus was 38.7 MPa (SD 34%). No correlation was found between the apparent density and the permeability or between the apparent density and the compressive modulus. The irreversible deformations in the graft could be captured by a creep law, for which the parameters were quantified.

We conclude that in clinical use the graft is bound to be subject to permanent deformation after operation. The permeability of the material is relatively high compared with, for example, human cartilage. The confined compression modulus is relatively low compared with cancellous bone of the same apparent density. Designs of prostheses used in revision surgery must accommodate the viscoelastic and permanent deformations in the graft without causing loosening at the interface.

It has been suggested that the endurance of cemented femoral reconstructions in total hip arthroplasty is affected by the creep of acrylic cement, but it is not known to what extent cement creeps under loading conditions in vivo, or how this affects load transfer. We have simulated the long-term creep properties of acrylic cement in finite-element models of femoral stem constructs and analysed their effects. We investigated whether subsidence rates measured in vivo could be explained by creep of acrylic cement, and if polished, unbonded, stems accommodated creep better than bonded stems.

Our findings showed that polished prostheses subsided only about 50 μm as a result of cement creep. The long-term prosthetic subsidence rates caused by creep of acrylic cement are therefore very small and do not explain the excessive migration rates which have sometimes been reported. Cement creep did, however, relax cement stresses and create a more favourable stress distribution at the interfaces. These trends were found around both the bonded and unbonded stems. Our results did not confirm that polished, unbonded, stems accommodated creep better than bonded stems in terms of cement and interface stress patterns.