23

The full migration pattern of tibial components is associated with aseptic loosening: introducing MTPMe-max (MTPM Estimated Maximum)

Bart G Pijls¹, Elise K Laende²

1) Dept Orthopaedics, LUMC, Leiden, The Netherlands

2) QEII Health Sciences Centre, Dalhousie University and Nova Scotia Health Authority, Halifax, Canada.

Introduction: Implant migration evaluation of tibial components with RSA is presently based on early migration (MTPM 6-month or 1-year thresholds) and continuous migration (MTPM 1-2 years). Although these migration indices are valuable for identifying unsafe tibial components, they are limited to specific follow-up moments and do not use the full migration pattern. The purpose of this study was to explore migration indices for tibial components that use the full migration patterns and to evaluate if they are associated with aseptic loosening.

Materials and method: Most migration of tibial components occurs in the first 6 to 12 months followed by stabilization or continuous migration. Such patterns have similarities to the Michaelis-Menten enzyme kinetics model from biochemistry and can be defined for migration as:

MTPM = MTPMe-max * t / (K + t)

where

MTPMe-max = MTPM Estimated Maximum, the estimated maximal MTPM based on the migration pattern

K = is the time (months) at which half of the MTPMe-max value is achieved

t = time in months

We fitted the above equation to migration patterns from a published systematic review on migration of total knee replacements [1] using non-linear regression with least-square differences to determine the MTPMe-max and K values for each implant design. The MTPMe-max and K-value of each implant were plotted against the 5 year revision rate for aseptic loosening [2].

Results: The mean total error of the model was 0.09mm (95%CI 0.04mm to 0.13mm). MTPMe-max up to 1.25mm was considered to be safe based on association with aseptic loosening revision rate while MTPMe-max of more than 1.25mm was unsafe. The K-value could not be used as a predictor for safe versus unsafe implants.

Interpretation and conclusion

MTPMe-max uses the full migration pattern, does not require fixed follow-up moments and could be used for early detection of unsafe tibial components designs.

Keywords: implant safety, migration thresholds, MTPMe-max, TKR

24

Reorienting the Tibial Baseplate Improves the Registration Accuracy of Model-Based Radiostereometric Analysis

Abigail E. Niesen¹, Anna L. Garverick¹, Stephen M. Howell¹, & Maury L. Hull^1,2,3

¹Department of Biomedical Engineering, University of California, Davis,

²Deparment of Mechanical Engineering, University of California, Davis,

³Department of Orthopaedic Surgery, University of California Davis Medical Center

Introduction: The accuracy of model-based radiostereometric analysis (MBRSA) in calculating tibial baseplate migration depends on baseplate shape [1] and orientation relative to the imaging planes. Hence, the primary objectives of this study were to introduce a new method for determining the optimal baseplate orientation to minimize bias error during MBRSA and to demonstrate the clinical usefulness of the method using a knee positioning guide to repeatably orient the baseplate.

Methods: A tibia phantom consisting of a tibia sawbone rigidly fixed to an example baseplate was placed in a reference orientation (i.e. baseplate coordinate system aligned with the laboratory coordinate system) and 23 other orientations with three pairs of radiographs acquired at each orientation. Radiographs were processed in MBRSA software, and the mean maximum total point motion (MTPM), an indicator of bias error during model registration, was plotted as a function of the rotation angles to determine the optimal orientation and a range of acceptable orientations.

Results: The bias error decreased 85% between the reference orientation and the optimal orientation. An acceptable range of orientations was defined by a decrease in bias error more than 50%. Clinical usefulness was demonstrated by repeatedly imaging a knee model placed in a knee positioning guide (simulated clinical positioning) and demonstrating that the mean orientation ± one standard deviation fell within the acceptable range of orientations.

Conclusion: Future researchers can use this method to determine the optimal orientation and a range of acceptable orientations for their specific baseplate to minimize bias error. Use of a knee positioning guide was an effective tool for repeatable patient positioning and should be considered for future RSA studies to maintain consistent positioning during a longitudinal study.

Keywords: total knee replacement; tibial baseplate migration; biplanar radiographs; model registration; maximum total point motion

25

Femoral prosthesis design and patient positioning during RSA examination do influence accuracy and precision of MBRSA-EGS

Zhuang Kang¹, Jing Xu¹, Raimund Forst¹, Frank Seehaus¹

¹Department of Orthopaedic Surgery, University of Erlangen-Nuremberg, Erlangen, Germany

Introduction: Model-based RSA using elementary geometrical shape (MBRSA-EGS) has been validated using a straight-tapered stem design [1, 2]. Aim of this study was to investigate if MBRSA-EGS could be influenced by the femoral stem design and leg alignment within RSA set up (e.g. bending contraction of the leg).

Material and method: Seven different femoral stem component designs were investigated: n=2 had straight stem with medial-/lateral-located taper; n=3 had curved stem with round, spiky or semi-round tip; n=2 had long straight or curved stem. A femoral stem-bone-model (FSBM) and a phantom model was used to mimic two patient leg alignment scenarios: (I) a leg flexion of 0, 5, 10 deg around medial-lateral axis; (II) internal/external leg rotation of -10, 0, 10 deg around superior-inferior axis. The protocol was repeated n=10 for each position scenarios.

Results: The stem design affects migration measurement errors: For translation along anterior-posterior axis (T_a.p.) (mean±SD from -0.020±0.053 to -0.044±0.183 mm) and rotation around superior-inferior axis (R_s.i.) (mean±SD from 0.062±0.255 to -0.323±0.785 deg). The leg alignment scenario (I) mean±SD from -0.009±0.079 to -0.016±0.118 mm (T_a.p.) and 0.068±0.412 to 0.025±0.690 deg (R_s.i.) and scenario (II) from 0.020±0.111 to -0.040±0.163 mm (T_a.p.) and -0.044±0.498 to 0.203±0.780 deg (R_s.i.) influences migration measurement errors as well.

Interpretation and conclusion: This result indicated that stem design and leg positioning affect MBRSA-EGS accuracy and precision. For investigated femoral stem components with long stem, lateral-located taper and spiky tip, less precision was observable. Using MBRSA-EGS for migration detection has to be therefore carefully considered. In general, model-based RSA using CAD/RE models or marker-based RSA should be applied preferentially.

Keywords: Hip, RSA, Stem design, Patient positioning

26

RSA in revision TKA: difficulties and recommendations

Maartje Belt¹, Malou te Molder¹, Petra Heesterbeek¹
1. Research Department, Sint Maartenskliniek, Nijmegen, the Netherlands

Introduction: While the number of revision total knee arthroplasties (rTKA) continues to increase, RSA studies with rTKAs are not as common as in primary TKAs. Although of clinical importance, the characteristics of these implants and patients present some more challenges in clinical practice than already known from primary TKA. The aim was to quantify the number and reasons for unanticipated drop-outs, and difficulties regarding RSA analyses in studies with rTKA prostheses.

Material and methods: We analyzed all RSA images of patients included in one of the three ongoing rTKA RSA studies, with a follow-up of 2 and 10 years. We descriptively reported the number of and reason for exclusion, loss to follow-up, and other difficulties regarding the RSA analyses that presented during the studies.

Results: 68 patients signed informed consent, of whom 7 patients were eventually not included as study participant for the following reasons: no markers (N=1), not enough markers (N=2), or per-operative decision to implant another type of prosthesis (N=4). During follow-up, 12 patients were loss-to-follow for the following reasons: they passed away (N=6), removal of study implant (N=1), markers removed by extensor mechanism reconstruction (N=1), other health issues or withdrawal of consent (N=4). We encountered occluded markers in 24 patients. This could be solved with a marker model in 18/24, the others needed to be excluded from the analyses (6/24). One patient was excluded due to a high condition number.

Interpretation and conclusion: 26/68 potential study participants were either not included in the study, or dropped out of the study unanticipated. We recommend thorough planning of the study in advance, and to account generously (~40%) for loss-to-follow or other problems in the sample size. Also, it might be helpful to add a successful clinical postoperative RSA scene as condition for inclusion.

Keywords: revision TKA, RSA, occluded markers, drop-outs, clinical practice

27

Propagation of Registration Error into Maximum Total Point Motion to Analyze Tibial Baseplate Stability at Six Months Using Marker-Based and Model-Based RSA

Abigail E. Niesen¹ & Maury L. Hull^1,2,3 ¹Department of Biomedical Engineering, University of California, Davis, ²Deparment of Mechanical Engineering, University of California, Davis, ³Department of Orthopaedic Surgery, University of California Davis Medical Center.

Introduction: Maximum total point motion (MTPM), the largest movement of a point on the baseplate relative to the tibia, is the most commonly used metric when developing stability limits which seek to predict aseptic loosening of tibial baseplates after total knee arthroplasty. However, the propagation of registration error into MTPM for stable baseplates (i.e. baseplates with MTPM < 0.5 mm) manifested as bias (i.e. systematic error) and precision (i.e. random error) has not been quantified and compared to the 6-month stability limit for marker-based RSA and model-based RSA [1], which are subject to different magnitudes of registration error [2].

Methods: A simulation applied specified true displacements, true rotations, and registration errors in six degrees of freedom to an example tibial baseplate. The registration errors were selected from a publication which computed these errors from double examinations using both marker-based and model-based RSA [2]. The bias in MTPM, precision in MTPM, and relative bias in MTPM for baseplates with true MTPM = 0.5 mm (i.e. bias in MTPM/0.5 mm × 100%) were computed for combinations of true displacement and true rotation for both marker-based and model-based RSA using 10,000 simulations.

Results: Results revealed that the bias in MTPM for stable baseplates with model-based RSA is about three times that for marker-based RSA (12%-27% vs. 5%-8%, respectively), and that the precision in MTPM for stable baseplates with model-based RSA is double that for marker-based RSA (0.11-0.16 mm vs. 0.05-0.07mm).

Conclusion: The 6-month stability limit of 0.5 mm, which was derived from MTPM using marker-based RSA, may need to be adjusted to incorporate the increase in bias and imprecision in MTPM when applying this stability limit to model-based RSA.

Keywords: total knee replacement; tibial baseplate migration; model-based radiostereometric analysis; stability limit; registration error

28

Propagation of Registration Errors into the Change in Maximum Total Point Motion to Analyze Tibial Baseplate Stability at Two Years Using Marker-Based and Model Based RSA

Abigail E. Niesen¹ & Maury L. Hull^1,2,3

¹Department of Biomedical Engineering, University of California, Davis,

²Deparment of Mechanical Engineering, University of California, Davis,

³Department of Orthopaedic Surgery, University of California Davis Medical Center

Introduction: The change in maximum total point motion (ΔMTPM) is used to predict long-term risk of tibial baseplate loosening by comparing ΔMTPM between 1 and 2 years to the continuous migration stability limit of 0.2 mm [1]. However, effects of registration error on ΔMTPM have not been quantified for marker-based and model-based radiostereometric analysis (RSA) and compared to this stability limit for stable baseplates.

Methods: A simulation applied specified true displacements, true rotations, and registration errors in six degrees of freedom to an example baseplate. The registration errors were selected from a publication which computed these errors from double examinations using both marker-based and model-based RSA [2]. The simulation assumed that true displacements and true rotations were identical at 1 and 2 years, representative of a stable baseplate; however, apparent MTPM at 1 and 2 years differed due to registration errors. To compute ΔMTPM between 1 and 2 years, MTPM (i.e. the largest movement of a point on the baseplate) at 1-year relative to the reference exam was subtracted from MTPM between the reference exam and 2-years. The bias and precision in ΔMTPM and the proportions of baseplates which fell above the continuous migration stability limit were determined for both marker-based and model-based RSA using 10,000 simulations.

Results: No bias error occurred in ΔMTPM; however, the precision of ΔMTPM was twice as large for model-based than marker-based RSA, resulting in about 25% of stable baseplates falling above 0.2 mm for model-based RSA.

Conclusion: Researchers should be aware that the false positive rate is 25% when applying the 0.2 mm stability limit to assess baseplate stability using model-based RSA. To reduce the false positive rate, the stability limit would need to be increased for model-based RSA. Keywords: continuous migration; tibial baseplate migration; radiostereometric analysis; error analysis