Background

Periprosthetic osteolysis is the most common long-term complication of a total joint arthroplasty, often resulting in aseptic loosening of the implant. As we aim at developing a safe and minimally invasive implant refixation procedure, thorough characterisation of the properties of the periprosthetic tissue is needed.

Background

Aseptic loosening of prostheses is the most common cause for failure in total joint arthroplasty. Particulate wear debris induces a non-stop inflammatory-like response resulting in the formation of a layer of fibrous periprosthetic tissue at the bone/implant interface. The current treatment is an invasive revision joint replacement surgery. However, this procedure has a high morbidity rate, therefore, a less invasive alternative is necessary. One approach could be to re-establish osseointegration of the joint prosthesis by inducing osteoblast differentiation in the periprosthetic tissue. Therefore, the aim of this study was to investigate the capacity of periprosthetic tissue cells to differentiate into the osteoblast lineage.

Whether it is best to retain the posterior cruciate ligament in the degenerated knee, i.e. using a cruciate-retaining (CR) total knee prosthesis (TKP), or to use a more constraint posterior-stabilized (PS) TKP is of debate. There are limited studies comparing the effect of both methods on implant fixation and clinical outcome, leaving it up to the surgeon to base this decision on anything but conclusive evidence. We assessed the effect of two different philosophies in knee arthroplasty on clinical outcome and tibial component migration measured with radiostereometric analysis (RSA), by directly comparing the CR and PS version of an otherwise similarly designed cemented TKP.

Sixty patients were randomized and received a Triathlon TKP (Stryker, NJ, USA) of either CR (n=30) or PS (n=30) design. RSA measurements (primary outcome) and clinical scores including the Knee Society Score and Knee injury and Osteoarthritis Outcome Score were evaluated at baseline, at three months postoperatively and at one, two, five and seven years. A linear mixed-effects model was used to analyse the repeated measurements.

Both groups showed a similar implant migration pattern, with a maximum total point motion at seven years follow-up of around 0.8 mm of migration (mean difference between groups 95% CI −0.11 to 0.15mm, p=0.842). Two components (one of each group) were considered to have an increased risk of aseptic loosening. Both groups improved equally after surgery on the KSS and KOOS scores and no differences were seen during the seven years of follow-up.

No differences in implant migration nor clinical results were seen seven years after cruciate-retaining compared to posterior-stabilized total knee prostheses.

Summary

In this study we validate that weight-bearing images are needed for accurate polyethylene liner wear measurement in total knee prostheses by measuring the difference in minimum joint space width between weight-bearing and non-weight-bearing RSA views.

The mobile-bearing variant of a single-radius design is assumed to provide more freedom of motion compared to the fixed-bearing variant because the insert does not restrict the natural movements of the femoral component. This would reduce the contact stresses and wear which in turn may have a positive effect on the fixation of the prosthesis to the bone and thereby decreases the risk for loosening. The aim of this prospective randomized study was to evaluate early migration of the tibial component and kinematics of a mobile-bearing and fixed-bearing total knee prosthesis of the same single-radius design.

According to a prospective randomized protocol 20 Triathlon single-radius posterior- stabilized knee prostheses were implanted (9 mobile-bearing and 11 fixed-bearing). Fluoroscopy and roentgen stereophotogrammetric analysis were performed 6 and 12 months post-operatively.

The 1 year post-operative roentgen stereophotogrammetric analysis results showed considerable early migrations in 3 mobile-bearing patients and 1 fixed-bearing patient. The range of knee flexion was the same for the mobile-bearing and fixed- bearing group. The mobile insert was following the femoral component during motion.

This study showed no apparent distinction in early migration and kinematics between mobile-bearing and fixed-bearing single-radius total knee prostheses. Des- pite the mobile insert was following the femoral component during motion, and therefore performed as intended, no kinematic advantages of the mobile-bearing total knee prosthesis were seen. It is concluded that a mobile insert in single-radius total knee prostheses is redundant and will not lead to additional benefits.

Purpose of the study: Congruence between the femoral component and the insert has been proposed to decrease wear in total knee arthroplasty (TKA). This congruence should favour unidirectional movement between the components because multidirectional movements carry a risk factor for wear up to 30-fold higher than unidirectional movements. This study explored in vivo displacements between the insert and the femoral component of a prosthesis in order to determine whether they meet the required kinematic criteria.

Material and methods: Twelve patients (7 women, 5 men) aged 45 to 79 years with BMI from 23 to 35 underwent knee surgery for osteoarthritis and were included in this study. The prosthesis was a mobile plateau pros-thesis implanted by the same surgeon using a navigation system. During the procedure, four tantalum beads were implanted in the polyethylene under stereotaxic guidance. The postoperative evaluation was performed at six months with the clinical evaluation (KSS, WOMAC) a 3D fluoroscopic protocol (walking, stairs, get up and go) and a radiostereometric analysis (RSA).

Results: Active flexion under weight bearing was 118 (range 102–125) and the mean KSS 165. The videofluoroscopy combined with RSA showed congruent axial rotation between the femoral component and the insert in the flexion arc 0/60 with a mean difference of 0.38 per degree of flexion (SD 1.85). Beyond 60° flexion, the posterior displacement of the condyle was greater than the insert rotation.

Discussion: Compared with other 3D videofluoroscopic studies, this analysis adds greater accuracy due to the implantation of tantalum beads in the insert, enabling a study of insert displacement in relation to the metal components. This method demonstrates that for the implant studied here, rotation of the insert follows the displacement of the femoral component exactly from 0 to 60° flexion, this is a gliding displacement. Then beyond 60°, a gliding plus rolling movement occurs displacing the femoral component posteriorly.

Conclusion: This in vivo study in patients with a mobile plateau knee prosthesis demonstrates that the insert has a rotation exactly like the femoral component and that complete congruency is maintained between the femoral component and the insert with a pure gliding contact from 0 to 60° flexion. The prerequisite criteria for this type of prosthesis designed to reduce the wear factor are thus confirmed.

Introduction: Approximately 20% of unicompartmental knee replacement (UKR) revisions are related to polyethylene wear. The Phase 1 Oxford UKR was introduced as a design against wear, with a fully congruent mobile bearing. The Phase 2 implant was introduced with new instrumentation (femoral mill) and changes to the bearing shape (lower anterior wall) to reduce the incidence of anterior impingement. We have previously shown that the Oxford UKR has a wear rate of 0.02 mm/year at ten years, in well functioning devices, but that higher wear rates can be seen with impingement or if the congruous articulation is lost. The aim of this study was to determine the 20 year in-vivo wear of the Oxford Phase 1 and Phase 2 UKR, using Roentgen Stereophotogrammetric Analysis (RSA).

Method: We measured the in-vivo wear of 6 Phase 1 (5 patients, mean age 65.24 years) and 7 Phase 2 (4 patients, mean age 63.43) Oxford UKR bearings. Average time since surgery was 22.37 years and 19.46 years for the Phase 1 and Phase 2 implants respectively. Selection criteria included patients who were mobile, with an exercise tolerance greater than 100m as per the American Knee Society Score (AKSS) functional questionnaire. RSA x-rays were taken with the knee in the normal anatomical position on standing and with the knee flexed to 30o. The Oxford knee score (OKS) and AKSS were gained at the RSA examination. Phase 1 and 2 components were reverse engineered by laser scanning, and converted to CAD models. The CAD models of the tibia and femur were pose-estimated in the RSA software (Medis Specials, Leiden, Netherlands). A sphere was fit to the femoral component and the minimum bearing thickness was determined by measuring the shortest perpendicular distance between the sphere and the plane contained on the tibial tray articular surface. The linear wear for each bearing was calculated by subtracting the measured thickness from the corrected nominal bearing thickness. Non-parametric statistics were used to compare the two Phases.

Results: There was no significant difference in age, OKS and AKSS between the two groups. The median wear rate was 0.078 mm/year for Phase 1 and 0.023 mm/year for Phase 2. This difference was statistically significant (p = 0.027).

Discussion: The difference in wear rate is explained by impingement in Phase 1, which was reduced by design changes with the introduction of Phase 2; the Phase 2 is designed to avoid impingement between the femur and the bearing. This study demonstrates that very low wear rates can be maintained with the Phase 2 implant to the end of the second decade after implantation. This is of particular importance when the device is used in younger patients and demonstrates that the Oxford UKR can be a definitive implant for the treatment of isolated compartmental osteoarthritis.

The need for a better understanding of factors that influence surgical outcome has grown as many complications are thought to be avoidable. One approach proven useful in studying surgical procedures is time-action analysis (TAA), a method which objectively determines the efficiency of individual steps. The aim of this study was to assess the surgical process of total knee arthroplasty using TAA, thus enhancing the insight into the procedure, influence of team members and adverse events, eventually leading to process improvement and reduction of error probability.

Methods: In two high output centres and one teaching hospital 37 TKA surgeries were recorded, using 3 different knee systems (NexGen, LCS and Triathlon). The process was analyzed using a fixed taxonomy and the duration, limitations and repetitions were determined using video analysis software. The efficiency of the surgeon was calculated by dividing the time the surgeon spends operating by the time operating plus the time spent talking, thinking or repeating.

Results: Although the two high output centres used different knee systems there was no difference in operating time (47min. (95%CI, 43.2 to 50.1) versus 47min. (42.1 to 51.9)). With an inexperienced nurse the waiting time increased in both hospitals during the femoral osteotomy phase (p= 0.01 and p=0.05). Comparing to a training hospital, the tibial alignment phase showed lower surgical efficiency for both the consultant and 6th year residents (80% vs. 95%, p=0.01). Also the nurse waiting time increased during all phases (18min. vs. 2min., p=0.00). In the teaching hospital more problems (communication, instrument and skill) occurred (mean 19 vs. 5 and 2, p=0.00) and twice as much communication problems existed with residents regardless of nurse experience. Surprisingly the number of problems handling instruments increased inversely with nurse experience (p=0.02) as did the waiting time (27min. vs. 15min.) again being highest in the femoral osteotomy phase (p=0.00).

Conclusion: The similar results in the high output centres show that TKA is a similar and structured process regardless of the knee system, its efficiency mainly dependent on surgical output. The decrease in resident efficiency is caused by less structured use of instruments and miscommunication with the nurse. For nurses the femoral osteotomy phase is most difficult, requiring high attention due to frequent changing of different pins and cutting blocks. Unfamiliarity with instruments (i.e. low volume) results in higher waiting times. Regardless of the knee system the steps of TKA are similar, therefore a consistent surgeon-nurse (OR tech) verbal interaction is advocated especially with a less experienced team. Training should focus knowledge of instruments using uniform names. Extra attention should be paid by those who do not often place knees to communication with the nurse and clear order of the procedure.

Background and purpose: The Souter-Strathclyde total elbow prosthesis is a well-studied replacement therapy for the destructed rheumatoid elbow joint. In the short-term results of a RSA-study we concluded that 8 out of 18 humeral components were at risk for loosening at two years of follow-up. Now we present the long-term results of this study to evaluate these predictions.

Patients and Methods: Twenty-one elbows (18 patients) were included in the RSA-study. At risk for loosening was defined as increase of translation of more than 0.4 mm or increase of rotation of more than one degree during the second postoperative year. The average follow-up was 98 months (range: 12–134 months). RSA-measurements were performed post-operatively, at 3 months, 6 months, 12 months and at yearly intervals thereafter. Prosthetic position and radiolucent lines (RLLs) were examined on conventional radiographs.

Results: Almost all humeral components, including the ones that were defined to be ‘at risk’ for loosening in our short-term study, showed increased and irregular migration in this long-term follow-up study. In contrast to this, the ulnar components were stable. Translations were most prominent in the posterior-anterior direction, the most prominent rotations took place about the transverse axis. No obvious influences of prosthetic alignment on micromotions were found. Four humeral components were clinically loose – three of them could be revised -, but only one of them was defined to be at risk for loosening in the short term study.

Radiological assessment based on conventional radiographs showed that a prediction of humeral loosening can be made within four years after surgery.

Interpretation: The RSA results clearly show that almost all humeral components migrate up to several millimeters and several degrees in an irregular manner, but often without clinical consequences.

Introduction: Cementless press-fit cups are the most widely used acetabular implants in total hip arthroplasty today. Their primary and secondary stability is largely determined by the design and choice of surface coating. Porous titanium coatings are used for the majority of cementless acetabular implants. However, an additional hydroxyapatite (HA) layer has been advocated for superior bone ongrowth. We studied the effect of additional HA coating on early micromotion of a porous titanium plasmasprayed cup with a flattened pole. A secondary objective was to compare the extent of micromotion of this well established cementless cup to data of other press-fit cups.

Methods: A total of 44 female patients (45 hips) undergoing total hip arthroplasty for osteoarthritis consented to participate in this prospective, IRB approved study. They were randomized for either a press-fit cup (EP-FIT PLUS^®, Plus Orthopedics AG, Switzerland) with a titanium plasma sprayed surface (Ti-group) or with additional HA coating (HA -group). All cups were used with the same combination of stem, PE liner and ceramic head. Model-based radiostereometry (MBRSA) was used to measure translation and rotation immediately postoperative, at 6 weeks, 3, 6, and 12 months. Statistical analysis of migration was performed utilizing one-sided Mann-Whitney tests and ANOVA.

Results: At one year, mean translation in the HA-group (Ti-group) along the medial-lateral (x), proximal-distal (y) and anterior-posterior (z) axes was −0.01 (0.07), 0.08 (0.09), and 0.03 (−0.06) mm, respectively. Mean rotation around the x-axis (anterior-posterior tilt) was −0.19 (−0.16), the y-axis (anteversion-retroversion) was −0.10 (−0.19), and the zaxis (adduction-abduction) was 0.23 (−0.05) degrees. Our hypothesis that translation and rotation would be different in the two groups was rejected (p< 0.00) for all dimensions except for rotation about the z-axis (p=0.10). The was no evidence for different migration patterns throughout the examined time points. All patients had excellent clinical outcome with a mean Harris Hip Score of 95.4 (HA-group) and 95.3 (Ti-group) (p=0.10). Plain radiographs of the cups showed good osseointegration.

Discussion: With excellent primary stability in both the Ti-group and the HA-group, we conclude, that HA-coating does not significantly increase stability of this flattened pole press-fit cup during the first postoperative year. We were able to show that the early stability of this cup is well comparable to that of the more frequently used hemispherical cups with initial stability being one prerequisite for long-term success. RSA measurements after 2 years will be conducted to confirm the current findings and the cup is also being studied in a long-term observation.

The most widely accepted method to measure kneekinematics is using external movement registration with the aid of skin-mounted markers. However, a large error component appears due to skin movement relative to the underlying bone. The goal of this study is to use fluoroscopy to quantify skin movement artefacts in patients with a total knee prosthesis during a step-up task.

The most widely accepted method to measure knee kinematics is using external movement registration with the aid of skin-mounted markers. However, a large error component appears due to skin movement relative to the underlying bone. The goal of this study is to use fluoroscopy to quantify skin movement artefacts in patients with a total knee prosthesis during a step-up task.

Translational and rotational errors attributed to soft tissue movement were three times larger for the femur than for the tibia about allaxes. The mean of the absolute rotational differences for the femur were2.6, 3.3 and 1.7 degrees about the X, Y and Z axes respectively. Absolute peak differences for individual subjects were 9.1, 12.9 and 10.5 mm along the X, Y and Z axes respectively.

This is the first study examining the 3D relative motion between surface-mounted and bone-anchored markers without the use of cortical pins anchored to the tibia and/or the femur. The results revealed no regular pattern of soft tissue error between subjects indicating the unlikely success of numerical methods for modeling and removing soft tissue motion artifacts when using standard motion capture methods.

Model-based Roentgen Stereophotogrammetric Analysis (RSA) measures micromotion of an orthopaedic implant with respect to its surrounding bone, without the use of markers on the implant. In previous studies with a total knee prosthesis, Model-based RSA showed to be very accurate. In this study, Model-based RSA is validated in a phantom experiment of a total hip prosthesis.

A metal backed, elliptical shaped EP-FIT PLUS ®cup was used in combination with a SL-PLUS ® hip-stem from PLUS Endoprothetik AG. In vivo conditions were simulated by using sawbones and perspex plates to mimic the bones and soft tissue. Virtual projections of the CAD models of the implant were fitted on the automatically detected contours in nine RSA radiographs and the error inmigration calculation was determined.

The standard deviations of the error in translation for the cup were: 0.03, 0.05, and 0.21 mm. (x, y, z-direction) The standard deviations of the error in orientation were respectively 0.56, 0.48, and 0.18 degrees (n = 10). For the stem, the standard deviations of the error in translation are: 0.09, 0.11, and 0.29 mm and for the orientation: 0.63, 2.03, and 0.24 degrees (n = 0).

The results for the cup are satisfactory, and make Model-based RSA a good alternative for conventional RSA. Especially for this type of metal backed, non hemispherical cup for which no markerless alternative is available. The error in orientation around the y-axis of the stem is of concern. Experiments with models from Reversed Engineering had similar low accuracy. We expect that the cause of these inaccuracies is the rectangular cross sectional shape of this specific hip stem, and we expect better results from experiments with differently shaped stems. The results of this study make very clear that Model-based RSA is avaluable and accurate technique, but phantom studies are always necessary to validate the accuracy for a specific implant shape.

Outcome measures must be valid, reliable and responsive to change criteria. The most common clinical outcome measures are Knee Society Scores, SF-36 quality of life scores, HAQ and DAS scores. However, performance based measures of functioning may not be dependent on patient report or observer judgment.

Examples of objective pre- and post-operative performance outcome measures are surface Electromyography (EMG) of muscles, kinematics and kinetics (gait analysis). For the evaluation of implant survival after joint arthroplasty, Roentgen Stereophotogrammetric Analysis (RSA) is the golden standard to assess micro-motion of the implants.

Surface EMG can be used to asses the stability of joints before and after intervention. Calibrating of raw EMG data is necessary to compare the data between subjects. It was shown that calibration of EMG data by means of isokinetic contractions on a dynamometer during flexion and extension was more reliable and repeatable than using a Maximum Voluntary Contraction in patients after total knee arthroplasty. After total knee arthroplasty RA patients have a lower net knee joint moment and a higher co-contraction than controls, indicating avoidance of net joint load and an active stabilization of the knee joint.

Fluoroscopy can be used to assess the kinematics of joints. In the pre-operative situation the use of CT models of the involved bones can be matched to the assessed fluoroscopic images. In the post-operative situation CAD models of the implants can be used for this purpose. In this way accurate 3D kinematics of joints can be assessed. During a step-up task of RA patients, the rotating platform of a mobile bearing knee showed no- or far less longitudinal rotation than the femur. Therefore, some of the theoretical advantages of this specific rotating platform knee prosthesis can be questioned. Fluoroscopy has also been used to assess soft tissue artifacts that occur in gait analysis i.e. displacements of skin-mounted markers relative to the underlying bone. The large soft tissue artefacts observed (displacements up to 17 mm and 12 degrees) question the usefulness of parameters found with external movement registration.

In order to assess the micromotion of implants after joint arthroplasty a measurement technique with a much higher accuracy than fluoroscopy is needed. RSA uses tantalum markers as landmarks bony structures and as landmarks on the implant. Recently a new RSA technique has been developed that does not rely on the attachment of artificial markers on the implant but uses CAD models of the implant instead. As an example of RSA as outcome measure, results showed that a calciumphosphate coating improves fixation of tibial components in RA patients, thus preventing mechanical loosening and subsequent long-term revision. In another clinical RSA study, it was found that mobile bearing knees are more predictable and forgiving with respect to micromotion compared to posterior stabilized tibial components in RA patients.

The results obtained by the above described performance outcome measures can be valued since the accuracy and precision of the used outcome measures are all published.

Introduction Mobile-bearing total ankle arthroplasty has gained more interest in recent years. Clinical results show favourable but varying results, with survival rates between 70% and 90% at 10-year follow-up. Design-specific differences in early migration patterns might explain differences in result and possible modes of failure.

Methods Prospective study of a cementless mobile-bearing total ankle arthroplasty by radiostereometric analysis (RSA). Fifteen total ankle arthroplasties were performed in patients with rheumatoid arthritis. The American Orthopaedic Foot and Ankle Society ankle score and radiostereometric radiographs were evaluated at regular intervals throughout the follow-up period: immediately postoperatively, 6 weeks postoperatively, 3 months, 6 months, and 12 months postoperatively and yearly thereafter.

Results The postoperative clinical results improved. We observed increased migration of the tibial component during the first 3 months, but this stabilized by the 6-months followup. The mean lateral-medial migration was 0.8 mm, distal-proximal migration was 0.9 mm, and posteroanterior migration was −0.5 mm. The latter implicated that the total resultant migration was in anterior and valgus tilting of this tibial component. This resulted in a main mode of migration proximal, anterior and valgus tilting of the tibial component.

Discussion This pilot study showed initial migration of this mobile-bearing ankle prosthesis into upward anterior and valgus tilting. However, migration stabilized at 6 months postoperatively. We think the surgical technique (anterior cortical window for placement) and the method of tibial fixation likely explain this migration.

Introduction: In total knee arthroplasty (TKA) it remains a topic if cementless designs offer long-term stability equivalent to cemented procedures and if the components should be coated with calciumphosphate to enhance fixation. This study compares the three-dimensional migration patters of cemented and uncoated and periapatite (PA) coated tibial trays during a 2-year clinical follow-up study using roentgen stereophotogram-metric analysis (RSA) measurements as a predictor of long-term implant loosening (Ryds definition).

Methods: A double blind randomized prospective study was performed on 101 osteoarhtritic patients receiving 115 Duracon TKA. The tibial tray was either cemented (25), uncoated and uncemented (46) or PA-coated and uncemented (44). The groups were matched for sex, age, BMI and pre-op Insall score. Patients were evaluated at 1 week, 3, 6, 12 and 24 months post-operatively using standard radiographs and Insall scores. At each evaluation RSA measurements determined the translational (medial-lateral (ML), caudal-cranial (CC), anterior-posterior(AP)) and rotational (anterior tilt, external and valgus rotation) displacements of the tibial tray.

Results: Insall scores were not statistically different between the groups. Average component displacement was low for the cemented components in all directions. For the uncemented trays migration was highest in the CC direction (subsidence) and steep during the first 6 weeks. At two years the uncoated trays showed significantly more subsidence (−0.5 0.63 mm, range: −2.1 to 0.5 mm) than the cemented components (0.1 0.17 m, range: −0.2 to 0.4 mm, p< 0.05) and the PA-coated group (−0.1 0.60 mm, range: −2.8 to 0.3 mm, p< 0.05). Average subsidence of the cemented and PA-coated implants was nearly the same but variability was higher for the coated trays (p=0.01). Displacements in all other directions were not significantly different between the groups. Using Ryds definition, a total of 10 tibial trays from the cemented group (40%), 29 trays from the uncoated group (63%) and 11 trays from the PA-coated group (26%) were identified to be at risk for long-term loosening. In seven cases (all cemented) component tilt was critical.

Conclusion: At 2 years no clinical differences were found between cemented, uncoated and PA-coated tibial trays. However, RSA measurements showed significantly different migration patterns and predictions for long-term implant stability. Steep initial subsidence before stabilisation seems an inherent characteristics of uncemented fixation. In contrast, the cement layer below cemented trays can lead to AP tilt. Uncoated uncemented components migrate significantly indicating a high risk of late loosening. PA-coating reduces tray migration and the risk of long-term failure to levels equivalent to cemented fixation.

The advantages of high viscosity Simplex AF cement (PMMA) compared to low viscosity Simplex P are the low porosity, the high fatigue strength, the lower polymerization time, and the lower maximum polymerization temperature. A prospective, randomized, double-blind clinical study was conducted to assess the in vivo effects of viscosity of bone cement on the micromotion of a polished tapered stem and UHMWP-cup (Exeter, Stryker-Howmedica). Roentgen Stereophotogrammetric Analysis (RSA-CMS, Medis, The Netherlands) was used to measure micromotion.

Twenty patients were included in a Simplex AF group (70 ± 4.3 years), and twenty patients were included in a Simplex P group (71 ± 7.3 years). No significant differences in body mass index and clinical hip scores were observed between the two studied groups.

There was no significant difference of the subsidence of both high and low viscosity cemented Exeter stems. The subsidence was according to the literature and showed that the viscosity of the bone cement did not influence the cement-implant bond of this polished tapered stem design.

The total migration of the cups and the migration along the medial-lateral axis were significantly larger for the Simplex AF cemented cups compared to the Simplex P cemented cups (p=0.037). This can be explained by the higher cement mantle thickness in acetabular Gruen zone 2 (p=0.003) and 3 (p=0.004) of the Simplex AF cemented cups.

We conclude from this study that the viscosity of the bone cement has no effect on the subsidence of polished tapered stems and that a high cement mantle thickness around an UHMWP-cup has a negative effect on fixation.

The aim of this study was on the one hand to compare the fixation of a posterior stabilised prosthesis (PS) and a PCL retaining mobile bearing design (Interax, Howmedica Osteonics, Rutherfort, USA) and on the other hand to measure the mobility of the mobile bearing. All measurements were carried out by means of Roentgen Stereophotogrammetric Analysis (RSA-CMS, Medis, Netherlands).

A prospective,randomised,double-blind study (N=28) was conducted to assess the micromotion of the components. At the one-year follow-up evaluation, the micromotion of the PS-components and the mobile bearing components were not significantly different. The PS-tibial components subsided −0.063 ± 0.177 mm and the mobile bearing knee tibial components subsided 0.067 ± 0.084 mm. The PS tibial component showed a higher variability in the migration results indicating a number of PS with rather large micromotion.

For three patients, the in vivo motion of the mobile bearing with respect to the metal backing was assessed at 30, 60 and 90 degrees of passive flexion. Two of mobile bearings moved posteriorly (2.4 and 2.9 mm) at respectively 60 and 90 degrees of flexion and showed a medial-lateral translation of 0.03 and 3.5 mm. One mobile bearing moved only 0.4 mm posteriorly at 90 degrees of flexion but showed a lateral-medial translation of 3.1 mm.

The broad range of kinematic patterns of mobile bearings during flexion that is observed in fluoroscopic studies is also observed in this study. A possible positive effect of mobile bearing movement may be found in the smaller variability of the micromotion of the mobile-bearing knees compared to the PS knees. The assumption was that shear forces in tibial bones implanted with a mobile bearing prosthesis would be better dissipated from the prosthesis-bone interface resulting in less micromotion. The kinematics of an additional number of mobile bearing knees -already included in the micro-motion study- will have to be assessed in order to determine the relation between mobility and micromotion.

The common factor in all (mechanical) prosthetic failure mechanisms is the induction of osteolysis around the endoprosthesis with subsequent prosthetic migration and finally loosening of the prosthesis. Both initial prosthesis-bone fixation and long-term prosthesis survival depend on the quality of the peri-prosthetic bone mass. The effects of treatment of RA patients with prednison are inhibition of osteoblastic activity and inhibition of calcium resorption from the intestines. The bone mass loss during the first six months of prednison treatment is substantial and will seldom be regained. Bisphosphonates are known to decrease osteoclastic activity and may therefore stop osteolysis at the bone-prosthesis interface.

The aim of the study was to evaluate a possible association of bisphosphonates with reduced migration of total knee prostheses (Interax, Howmedica Osteonics, Rutherfort, USA) in a high-risk group. Roentgen Stereophotogrammetric Analysis (RSA-CMS, Medis, The Netherlands) was used to measure the micromotion.

Retrospectively a group of nine RA patients treated with prednison (non-bisphosphonates group) and a group of fourteen RA patients (bisphosphonates group) treated with prednison in combination with bisphosphonates (Etidronate) were included from a prospective randomized study of 82 patients (Nelissen et al., 1998).

At the two-year follow-up evaluation, functional scores and knee scores did not differ significantly among the two groups. At the two-year follow-up evaluation, the non-bisphosphonates group subsided −0.47 ± 0.8 mm, and the bisphosphonates components subsided 0.07 ± 2.9 mm. In the analysis of variance with repeated measurements, with correction for follow-up time, sedimentation rate, and prosthesis fixation type, the bisphosphonates group migrated 1.20 mm less in the total migration (95% c.i.: 1.07–1.30 mm) compared to the non-bisphosphonates group.

In this study, bisphosphonates medication in addition to corticosteroid medication was associated with reduced migration of knee prostheses compared to corticosteroid medication alone.

Early micromotion of joint prostheses with respect to the bone can be assessed very accurately by a method called Roentgen Stereophotogrammetric Analysis (RSA); a method that uses two simultaneous X-ray exposures of the joint and has an accuracy of 0.1 mm for translations and 0.3 degree for rotations [1]. In order to reach this accuracy, metallic markers are inserted into the bone and attached to the surface of the prosthesis. These markers can then be identified automatically in the two radiographs [2]. Since the adjustments to the prosthesis are difficult, time-consuming and expensive, RSA has only been applied in a limited number of clinical trials.

In a previous study we have developed a Model-based RSA algorithm, which does not require the attachment of markers to the prosthesis [3]. This algorithm is based on minimisation of the non-overlapping area (NOA) between the automatically detected contour of the prosthesis from the roentgen image, with the virtually projected contour of a three-dimensional model of the prosthesis.

Because the accuracy of this NOA algorithm was not as high as the accuracy of the currently used Marker-based RSA, we have studied alternative algorithms for Model-based RSA. From a simulation study in which we used models of the Interax Total Knee Prosthesis (Stryker-Howmedica) and the G2 Hip Prosthesis (Johnson & John-son), we found that the results of the NOA algorithm can be improved substantially. The newly developed Model-based RSA algorithm is based on minimisation of the mean distance between the points of the actual contour and the virtually projected contour. The simulation study shows that the new algorithm is superior to the NOA-algorithm in situations where part of the contour is occluded, as well as in situations where the contour is distorted by noise. With the new algorithm, the residual position error can be reduced to 0.1 mm. and also the residual orientation error can be reduced to 0.3 degree, making Model-based RSA a future alternative to Marker-based RSA.

To measure micromotion of an orthopaedic implant with respect to its surrounding bone, Roentgen Stereo-photogrammetric Analysis (RSA) was developed. A disadvantage of conventional RSA is that it requires the implant to be marked with tantalum beads. This disadvantage can potentially be resolved with model-based RSA, whereby a 3D model of the implant is used for matching with the actual images and the assessment of position and rotation of the implant. In this study, an improved model-based RSA algorithm is presented and validated in phantom experiments. This algorithm is capable to process projection contours that contain drop-outs. To investigate the influence of the accuracy of the implant models that were used for model-based RSA, we studied both Computer Aided Design (CAD) models as well as models obtained by means of Reversed Engineering (RE) of the actual implant.

The results demonstrate that the RE-models provide more accurate results than the CAD models. If these RE models are derived from the very same implant, it is possible to achieve a maximum standard deviation of the error in the migration calculation of 0.06 mm for translations in x- and y-direction and 0.14 mm for the out of plane z-direction, respectively. For rotations about the y-axis, the standard deviation was about 0.1 degree and for rotations about the x- and z-axis 0.05 degree. For the femur component, it was also possible to reach these accurate results for non-scanned components. The results show that the new algorithm is an improvement with respect to a study we presented earlier [1].

Studies with clinical RSA-radiographs must prove that these results can also be reached in a clinical setting, making model-based RSA a possible alternative for marker-based RSA.