Secondary Logo

Journal Logo

2018 KNEE SOCIETY PROCEEDINGS

Radiostereometric Analysis Permits In Vivo Measurement of Very Small Levels of Wear in TKA

Teeter, Matthew G. PhD; Wihlidal, Jacob BSc; McCalden, Richard W. MD, MPhil, FRCSC; Yuan, Xunhua PhD; MacDonald, Steven J. MD, FRCSC; Lanting, Brent A. MD, MSc, FRCSC; Naudie, Douglas D. MD, FRCSC

Author Information
Clinical Orthopaedics and Related Research: January 2019 - Volume 477 - Issue 1 - p 80-90
doi: 10.1097/CORR.0000000000000399
  • Free

Abstract

Introduction

Polyethylene wear continues to be a cause for revision of TKA [29, 30]. However, improvements in implant materials and sterilization methods have enhanced wear resistance over historical TKA implants that utilized conventional polyethylene and sterilization in air [18, 26]. For these earlier implants that experienced wear at a high rate, it was possible to measure the magnitude of polyethylene wear in vivo over time using plain radiographs [6, 7]. With this approach, the narrowing of the space between the femoral condyle and tibial baseplate measured with the knee in extension was sufficient to provide a linear wear rate. Today, only the most severely worn implant would have visibly appreciable narrowing on a conventional AP radiograph.

For modern, more wear-resistant TKAs, more advanced methodologies for measuring TKA wear have been developed as research tools utilizing radiostereometric analysis (RSA) and single-plane fluoroscopy [13, 17, 31, 41, 44-47]. To improve the accuracy of linear wear measurements, three-dimensional (3-D) models of the implant components are registered to the radiographs using validated 3-D to two-dimensional (2-D) registration algorithms to more precisely identify the relative distance between the femoral and tibial components [47]. No marker beads are required. These measurements have been found to be more accurate with RSA than with plain radiographs as a result of the calibrated paired orthogonal radiographs obtained in the RSA procedure [46]. The stereo nature of RSA is also advantageous over single-plane images, which lose accuracy in the out-of-plane imaging direction [41]. It has been noted that standing examinations are preferred over supine examinations, because joint laxity in the supine position can change the location of contact [45]. Examinations should be acquired at multiple knee flexion angles to enable the femoral component to drop into different wear pools as the location of contact changes throughout flexion [13, 41, 44]. RSA has a reported accuracy for linear wear measurements in the range of 0.1 to 0.2 mm [31, 47]. Phantom studies have suggested volumetric measurements are likely to be over- or underestimates of wear volume such that volumetric measurements are not as reliable as linear measurements of TKA wear with RSA [41, 44]. Wear measurements can be taken at a single time point or serially in a prospective manner. The advantage of the latter method is that it enables large creep during the early bedding-in period of the first 1 to 2 years of implantation to be separated from the steady-state wear rate (which also includes a smaller amount of creep). With a single time point, both the linear and volumetric measurements are a combination of wear and creep [13].

Implant retrieval studies are the most common method to measure wear, but these failed implants may not truly represent the performance of well-functioning implants [27]. RSA offers the potential to examine measure wear of implants that are still in the patient’s body and contrast this to previous retrieval findings. The Genesis II™ (Smith & Nephew, Memphis, TN, USA) is one example of an implant system that has a long and successful clinical track record [21] and has been widely examined in implant retrieval studies [2, 12, 15, 20, 24, 25, 32, 35, 50]. The Genesis II was designed with strong mediolateral conformity to enable greater contact area, decreasing contact stresses. Substantial conformity in the sagittal plane is limited in the distal radius of the femoral component, leading to a reduction of contact area with increasing flexion. Femoral rollback is encouraged through a dished design. In prior retrieval studies of the Genesis II, linear wear rates of 0.03 to 0.09 mm/year have been reported for the articular surface [12, 35] along with a backside wear rate of 0.008 mm/year [32] and damage to the anterior and posterior aspects of the post in posterior-stabilized designs [25]. Retrieval studies of the Genesis II have also identified that increased surface damage and greater thinning of the medial condyle are associated with increasingly varus leg alignment [20, 24, 50]. Increased linear backside wear has also been associated with thinner polyethylene liners and female versus male patients with the Genesis II [32].

In this study, we implemented an in vivo wear measurement technique to study the Genesis II system to understand the wear performance of well-functioning TKA implants, as opposed to failed retrieved implants, and to evaluate the influence of different factors on TKA wear such as coronal alignment. We sought to determine, at long-term followup, (1) the linear wear rate of the implant using RSA; (2) whether demographic factors were associated with the linear wear rate; and (3) whether limb alignment was associated with the linear wear rate.

Patients and Methods

Forty-nine patients who had undergone posterior-stabilized TKA with a Genesis II implant system for varus osteoarthritis with a minimum of 10 years followup (median, 12 years; range, 10-20 years) were selected from our institutional database and studied retrospectively. Between 1997 and 2007, we performed TKAs in 4082 patients. Two thousand eighty-five patients underwent a posterior-stabilized TKA using the Genesis II implant system with a cobalt-chromium femoral component for knee osteoarthritis with a varus deformity; this device was used preferentially by most surgeons at our center over this time period because of their familiarity with the implant and in accordance with hospital contracts. Additionally, it had been demonstrated that the posterior-stabilized device provided higher postoperative ROM than cruciate-retaining devices in the implant system [23]. At 10 years, 190 had died and 54 had undergone revision (of which only one revision was for polyethylene wear). All patients, before surgery, underwent a 3-foot standing radiograph to assess alignment. However, this was not performed routinely after surgery. Therefore, we identified patients who had a Genesis II with at least 10 years of implantation time in one leg and had subsequently underwent a recent TKA of their other leg so that appropriate images were available. We identified 71 patients who fit our criteria, of whom 52 were women and 19 men with a median age at surgery of 67 years (range, 48-83 years) and implantation time of 13 years (range, 10-20 years). Of these, 34 agreed to participate including 21 women and 13 men with a median age at surgery of 66 years (range, 49-77 years) and implantation time of 14 years (range, 10-20 years). In addition, 15 other patients were part of a previous RSA study of implant migration (not wear) [42] and were included here with five women and 10 men with a median age at surgery of 70 years (range, 62-76 years) and implantation time of 11 years (range, 10-12 years). Thus, the final group included 49 patients (Fig. 1). The study was approved by our institutional research ethics board.

Fig. 1
Fig. 1:
The patient selection process for the study is shown. PS = posterior-stabilized; CoCr = cobalt-chromium.

The Genesis II implants that were examined consisted of cobalt-chromium femoral components, titanium tibial components with a polished bearing surface, and tibial inserts machined from polyethylene sheets of compression-molded conventional GUR 1050 (standard posterior-stabilized inserts) or GUR 1020 (high-flexion posterior-stabilized inserts) [21, 22, 25]. The tibial inserts were sterilized with ethylene oxide. The tibial inserts have a symmetric design and are available in four sizes (1/2, 3/4, 5/6, and 7/8) and seven thicknesses (9 mm, 11 mm, 13 mm, 15 mm, 18 mm, 21 mm, and 25 mm). The knees examined in this study included 24 size 3/4 implants (13 of 11 mm, six of 13 mm, three of 15 mm, and two of 18 mm thickness), 15 size 5/6 implants (eight of 11 mm and seven of 13 mm thickness), and 10 size 7/8 implants (seven of 11 mm, two of 13 mm, and one of 18 mm thickness). Six implants utilized high-flexion posterior-stabilized tibial inserts, whereas the remaining implants utilized conventional posterior-stabilized tibial inserts.

Each patient underwent standing, weightbearing stereo RSA examinations at a single time point. Examinations were not collected prospectively and there was no baseline examination, meaning that creep could not be separated from wear. The first 15 patients (who were enrolled as part of the long-term followup study on the Genesis II [42]) were examined with the radiographic sources and cassettes oriented in an AP and lateral direction using a biplane calibration cage (RSA Biomedical, Umea, Sweden). The published results of the earlier study solely examined migration, not wear. Examinations were acquired at 0°, 20°, 40°, 60°, and 80° of flexion. Reference plates were used to link the patient examinations and the calibration cage examination. Examinations for the remaining patients were acquired with the radiographic sources orientated at 40° to each other using a uniplanar calibration cage (RSA Biomedical), which was then inherently registered to the patient examinations. This reduced overlap of the leg segments, enhancing visualization, and enabled examinations to be acquired at 0°, 20°, 40°, 60°, 80°, 100°, and 120° of flexion. The examinations from 0° to 60° are performed with the patient facing the x-ray sources and are similar to performing a squat, whereas the examinations from 80° to 120° are performed with the patient perpendicular to the x-ray sources with the indicated foot raised on a stool and are similar to performing a lunge. The biplane calibration cage utilized small imaging cassettes with a 2364 x 2964 image matrix, whereas the uniplanar calibration cage utilized large imaging cassettes with a 3520 x 4280 image matrix. Both result in images with a 0.1-mm pixel pitch, with no magnification, and have been found to be equivalent in precision [4]. The relative position of the femoral and tibial components was determined on each stereo examination using Model-Based RSA software (RSAcore, Leiden, The Netherlands). The Model-Based RSA algorithm for 2-D/3-D registration of the manufacturer’s CAD models for the femoral and tibial components to each pair of radiographs has been demonstrated to have excellent accuracy with errors of 0.19 mm for translations and 0.52° for rotations [43].

The geometry of the polyethylene tibial inserts was acquired through a validated micro-CT scanning technique [38]. A new, never implanted tibial insert was acquired and scanned for each size of tibial insert that was used in patients in the study. For the size 3/4 tibial inserts, all thicknesses were available, whereas for sizes 5/6 and 7/8, only the 11-mm thicknesses were available. Scans were performed using an isotropic voxel spacing of 50 µm and a beam energy of 90 kVp and 40 mA. Scan volumes were reconstructed at the full isotropic spacing, and the geometry of the entire tibial insert was generated using isosurface rendering from an automatically determined threshold. Reverse engineering the tibial insert geometry has been demonstrated to be more accurate than using manufacturers’ CAD models for wear measurement [37], and the micro-CT technique is accurate to 10 mm3 [38]. Variability between new tibial inserts of this implant system as a result of manufacturing tolerances has been shown to be up to 2% of the total volume in the worst cases [36]. For each patient case, the appropriately sized tibial insert geometry was registered to the appropriately sized tibial baseplate, joining the plane of the baseplate with the plane of the tibial insert backside, centered appropriately within the locking mechanism, using Geomagic Studio® software (Geomagic Inc, Research Triangle Park, NC, USA). This technique has been used previously for fluoroscopic measurements of knee kinematics with a reported accuracy of 0.38 mm [19]. The origin of the tibial baseplate coordinate system was not altered during the addition of the tibial insert to ensure it would maintain alignment with the RSA measurements.

For 10 patients who received size 5/6 or 7/8 tibial inserts, the correct thickness was not available for micro-CT scanning. Instead, the 11-mm thick tibial insert surface of the appropriate size for the patient was translated proximally relative to the tibial baseplate the appropriate number of millimeters that represented the difference in thickness using the Geomagic Studio software. For example, for a size 5/6 tibial insert of 13 mm thickness, the size 5/6 11-mm thick tibial insert was translated proximally 2 mm. This approach was developed after reviewing the manufacturer’s design handbook for the Genesis II to confirm size differences and validated by translating size 3/4 tibial inserts for which the true thickness was available as a comparator. For 50 comparisons from different patient examinations pairing the true thickness to the translated approach (for 13-mm and 18-mm thick tibial inserts), the mean difference in linear penetration was 0.024 mm lower (95% confidence interval, 0.017-0.032 mm lower) using the translation approach. When converted to a wear rate, the error for patients who required this translation is approximately 0.002 mm/year for linear wear rate assuming a 10-year implantation time.

The registration transforms from the Model-Based BRSA software were applied to the appropriate femoral component and tibial component (tibial insert joined to the baseplate) models. Because polyethylene wear of the tibial insert in vivo moves the femoral component closer to the tibial baseplate over time as a result of loss of material, inserting the geometry of a new tibial insert between the registered components results in an intersection between the femoral component and tibial insert geometries, representing the worn region [31, 41]. Linear wear measurements were taken by measuring the greatest depth of the intersection (distance from one geometry surface to the other) between the femoral component and tibial insert geometries on the medial and lateral condyles (Fig. 2) with software previously used to measure contact location and liftoff distance in kinematic studies [40]. Two observers (JW, MGT) completed the measurements of the greatest depth of intersection with an interobserver correlation coefficient of 0.99.

Fig. 2
Fig. 2:
A demonstration of the wear measurement procedure is shown. RSA examinations were performed with the patient’s knee at multiple flexion angles while standing. Models of the femoral and tibial components were registered to the RSA images, and the correctly sized unworn model of the tibial insert was registered to the tibial baseplate. The maximum depth of penetration by the femoral component into the tibial insert was measured on the medial and lateral sides and divided by the length of implantation to determine the wear rate.

Two sets of linear wear measurements were calculated. The maximum depth of penetration on the lateral and medial sides across all measured flexion angles was determined for each patient and divided by implantation time to produce a lateral and a medial maximum linear wear rate. The median depth of penetration on the lateral and medial sides across all measured flexion angles was also determined for each patient and divided by implantation time to produce a lateral and a medial median linear wear rate. The maximum linear wear rate is most representative of what would be measured in a retrieval study but is subject to the greatest amount of error, whereas the median linear wear rate will cancel out over- and underestimates so may be more accurate [13].

A D’Agostino and Pearson test was used to assess the data for normality. The wear data were heterogeneously parametric and nonparametric depending on the flexion angle where it was measured; therefore, we elected to consistently use nonparametric statistics. All demographics are reported as median and range, and all measured values are reported as median and interquartile range (IQR). Correlation analysis was performed using Spearman correlation. Although some factors (such as male sex, height, and tibial insert size) are likely not independent from each other, the patient numbers were too low for rigorous multivariate analysis. The Wilcoxon matched-pairs signed-rank test was used for paired data (such as medial versus lateral side) and the Mann-Whitney test was used for unpaired data (such as wear in male patients versus female patients).

Results

The maximum linear wear rate on the lateral side was 0.047 mm/year (IQR, 0.034-0.066 mm/year) and on the medial side was 0.052 mm/year (IQR, 0.040-0.069 mm/year). This value represents the greatest point of penetration measured for each patient across all flexion angles (0°-120°) and was not different between the medial and lateral sides (median difference = 0.005 mm/year less laterally, p = 0.41). The median linear wear rate across all implants on the lateral side was 0.027 mm/year (IQR, 0.017-0.046 mm/year) and on the medial side was 0.038 mm/year (IQR, 0.022-0.054 mm/year). This value represents an average across all measured flexion angles and was greater medially than laterally (median difference = 0.010 mm/year, p = 0.04). Between flexion angles, there was no difference in measured penetration on the lateral side (Fig. 3A) but there was on the medial side (Fig. 3B). Medial penetration was greater at 20° than 80° (median difference = 0.021 mm/year, p = 0.004), at 40° than 80° (median difference = 0.024 mm/year, p < 0.001), at 40° than 100° (median difference = 0.019 mm/year, p = 0.016), and at 40° than 120° (median difference = 0.019 mm/year, p = 0.015).

Fig. 3 A-B
Fig. 3 A-B:
The linear wear rate (mm/year) measured at each flexion angle on the lateral side (A) and medial side (B) is shown. The solid line is the median and the dashed lines are the IQR.

There were associations (Table 1) between all wear measurements and patient height (Fig. 4B), and between a subset of wear measurements and body mass index (BMI) (Fig. 4D) and tibial insert size (Fig. 4E). There were no associations between any wear measurements and patient age (Fig. 4A), patient weight (Fig. 4C), tibial insert thickness (Fig. 4F), or implantation time (Fig. 4G). All wear measurements were greater in male patients than female patients (Table 2).

Table 1.
Table 1.:
Linear wear rate correlations to demographic variables
Fig. 4 A-G
Fig. 4 A-G:
Plots of median linear wear rates for the medial and lateral sides are given against age (A), height (B), weight (C), BMI (D), tibial insert size (E), and tibial insert thickness (F). A plot of median linear penetration with implantation time is also shown (G).
Table 2.
Table 2.:
Linear wear rates compared between male and female patients

There was an association (Fig. 5) between the median linear wear rate and leg alignment on the medial side (ρ = -0.33, p = 0.02) but not the lateral side (ρ = 0.23, p = 0.12). There was no association between the maximum linear wear rate and leg alignment medially (ρ = -0.27, p = 0.06) or laterally (ρ = 0.25, p = 0.09).

Fig. 5
Fig. 5:
A plot of the median linear wear rates for the medial and lateral sides against leg alignment is shown.

Discussion

Although revision TKA resulting from polyethylene wear is decreasing, long-term wear performance is still a topic of interest to surgeons and device manufacturers. Measurement of wear in vivo with modern implant designs is rarely performed as a result of technical challenges that require the use of RSA over conventional radiographs. In this study, we measured wear in vivo for a single implant design with an excellent record of long-term clinical performance [21], the Genesis II, after a minimum of 10 years of followup. We sought to determine (1) the linear wear rate using RSA; (2) the association between demographic factors and wear rate; and (3) the association between limb alignment and wear rate.

This study has a number of important limitations. First, a single implant design from a single manufacturer was examined, and the results may not extrapolate to other implants. We examined a relatively small number of patients compared with the total number of patients who received this implant at our institution, so those examined may not be representative of the broader population. The number examined also prevented us from using more powerful multivariate analysis methods. Younger patients or those otherwise perceived by the surgeon to be more demanding on their joint might have received oxidized zirconium bearings, which were excluded from analysis, so the examined population may not be the most likely to wear through their implants. Additionally, a patient with a high-wearing implant might not make it to the 10-year time cutoff of the present study, although we have found only a single case of revision of this implant as a result of polyethylene wear in our institutional database. The wear assessment technique measures creep and wear together; therefore, the wear rates reported here are likely overestimations of the steady-state wear rate. Prospectively following the patients would have helped enable the differentiation of bedding-in creep from wear, but this is logistically challenging for long-term followup and some creep would still be included even in the steady-state wear measurements. Wear measurements in many retrieval studies (the most common source of wear performance measurements in the literature) have also not differentiated wear from creep with the exception of certain micro-CT studies [10, 35]. It is impossible for the wear measurement technique used here to separate articular from backside wear, and any actual backside wear would have been counted as articular surface wear. The Genesis II has been reported to have low backside wear and damage as a result of its polished titanium baseplate, although there have been differences found between conventional posterior-stabilized and high-flexion tibial inserts, and a minority of high-flexion inserts were included in the present study [1, 12, 25, 32, 35]. There are potential sources of error in the 2-D/3-D registration steps, although these are known and accepted for RSA [43, 47]. The first 15 patients had RSA examinations with a biplanar rather than uniplanar examination and were not examined past 80°, although most wear was seen in the early flexion angles. There are also potential errors as a result of the manufacturing variability of the tibial inserts and reverse engineering of their surface geometries [36, 38], a problem that is shared with retrieval studies. There was no assessment bias, because the individuals providing care were not involved in study recruitment or data analysis.

The linear wear rates measured in this study can be compared with implant retrieval and wear simulator studies for the Genesis II. A number of retrieval studies have examined the Genesis II using damage scoring techniques and have found it demonstrates good wear resistance [1, 9, 12, 15, 35], but fewer studies have directly measured linear or volumetric wear for the implant. Gascoyne et al. [12] examined 26 pairs of retrieved Genesis II tibial inserts from both cruciate-retaining and posterior-stabilized designs mating with either an oxidized zirconium or cobalt-chromium femoral component. For the latter group, the linear wear rate was approximately 0.09 mm/year for both the lateral and medial sides after a mean implantation time of 23 months. The reasons for revision were primarily infection or instability with no cases revised for wear. Teeter et al. [35] examined 16 retrieved posterior-stabilized Genesis II tibial inserts with a mean implantation time of 32 months. The linear wear rate was 0.049 mm/year for tibial inserts implanted for at least 1 year, and no difference was found between the medial and lateral sides. Infection was the most common reason for revision and no cases were revised for wear. A wear simulator study of cruciate-retaining Genesis II tibial inserts by Teeter et al. [39] found a linear wear rate of 0.06 mm/million cycles on both the lateral and medial sides after the bedding-in period, from 3.2 to 6.1 million cycles. The finding that linear wear depth changes with flexion angle on the medial side would not be possible to determine in a retrieval or simulator study. One explanation could relate to the contact kinematics of the implant design. Ardestani et al. [3] created a computational model of the Genesis II and found that contact area increases from 0° to 50° of flexion, decreases from 50° to 80° of flexion, then increases again from 80° to 100° of flexion in a pattern very similar to the linear wear measurements in the present study. However, errors in the repeatability of the RSA measurements could also contribute to differences across flexion angles. There were no differences between the low to midflexion angles (0°-60°); therefore, a single examination at one of these angles could be representative of wear throughout flexion, which is equivalent to the “maximum” linear wear rate we report. However, it may be more accurate to take linear wear measurements at multiple flexion angles up to 60° and then calculate an average of the measurements, both to find the greatest depth of penetration across wear pools and to minimize error from a single measurement that might under- or overestimate wear. This is the “median” wear rate that we report.

There were associations between wear and patient sex, height, BMI, and implant size. Linear penetration did not correlate with implantation time. Wear and damage have often, but not always, been shown to correlate with implantation time in retrieval studies [12, 24, 27]. The measurement technique used in this study measures creep as well as wear, which might confound the correlation if creep makes up a relatively large proportion of the total volume, as has been demonstrated in a wear simulator study of this implant [39]. In addition, most patients were examined between 10 and 13 years after implantation time, and this relatively narrow range could affect the correlation. Interestingly, greater wear was associated with male patients and taller patients (which were mostly males), a finding that has also been seen for wear of THA [11, 14, 28]. Patients with larger tibial insert sizes (but not thicknesses) also had greater wear rates, although almost all of these patients were also male. Two factors that surgeons often worry are detrimental to implant longevity are younger patient age and higher patient BMI [5, 8, 16, 34]. Age at the time of surgery did not affect wear rates in this study, whereas greater BMI was negatively associated with medial wear and not associated with lateral wear. Greater activity, rather than simply younger age, may result in more wear [27, 28], and we did not track activity of the study patients. Potentially, patients with higher BMI are less active than patients with lower BMI, resulting in the negative association with wear.

There was an association found between the median wear rate calculation on the medial side and alignment with greater wear with increasingly varus leg alignment. In the present study, all cases had a target of neutral mechanical alignment. Alignment targets outside of mechanically neutral are currently being examined as a potential method to enhance patient satisfaction [48, 51]. Multiple retrieval studies have found greater polyethylene damage and volumetric wear for tibial inserts retrieved from patients with varus-aligned limbs [20, 24, 33, 49]. Our findings are therefore in agreement with these retrieval studies.

In the population examined in this study, there was good wear resistance in patients with well-performing implants at long-term followup. If this patient group is representative of the broader patient population receiving TKA, this suggests that targeting further improvements in polyethylene wear may not be necessary. However, the most demanding and wear-inducing patients may have been excluded from the examined population as a result of the study design. Greater wear rates were associated with male patients, taller patients, patients with lower BMIs, and patients with larger tibial inserts. There was also an association between greater medial wear and more varus leg alignment. Applying the in vivo wear measurement techniques described here to study other modern total knee implant systems at long-term followup would provide complementary information to retrieval studies and valuable data for surgeons and device manufacturers.

Acknowledgments

We thank Nicole Dearing, Mehul Garach, and Codie Primeau for their assistance with the study.

References

1. Abdel MP, Gesell MW, Hoedt CW, Meyers KN, Wright TM, Haas SB. Polished trays reduce backside wear independent of post location in posterior-stabilized TKAs. Clin Orthop Relat Res. 2014;472:2477–2482.
2. Anderson FL, Koch CN, Elpers ME, Wright TM, Haas SB, Heyse TJ. Oxidised zirconium versus cobalt alloy bearing surfaces in total knee arthroplasty: 3D laser scanning of retrieved polyethylene inserts. Bone Joint J. 2017;99:793–798.
3. Ardestani MM, Moazen M, Maniei E, Jin Z. Posterior stabilized versus cruciate retaining total knee arthroplasty designs: conformity affects the performance reliability of the design over the patient population. Med Eng Phys. 2015;37:350–360.
4. Broberg JS, Yuan X, Teeter MG. Investigation of imaging magnification in radiostereometric analysis. Proc Inst Mech Eng H. 2017;231:92–95.
5. Chen JY, Lo NN, Chong HC, Bin Abd Razak HR, Pang HN, Tay DK, Chia SL, Yeo SJ. The influence of body mass index on functional outcome and quality of life after total knee arthroplasty. Bone Joint J. 2016;98:780–785.
6. Collier M, Engh C, Hatten K, Ginn S, Sheils T, Engh G. Radiographic assessment of the thickness lost from polyethylene tibial inserts that had been sterilized differently. J Bone Joint Surg Am. 2008;90:1543–1552.
7. Collier M, Jewett B, Engh C Jr. Clinical assessment of tibial polyethylene thickness: comparison of radiographic measurements with as-implanted and as-retrieved thicknesses. J Arthroplasty. 2003;18:860–866.
8. D'Apuzzo MR, Novicoff WM, Browne JA. The John Insall Award: Morbid obesity independently impacts complications, mortality, and resource use after TKA. Clin Orthop Relat Res. 2015;473:57–63.
9. Dolan MM, Kelly NH, Nguyen JT, Wright TM, Haas SB. Implant design influences tibial post wear damage in posterior-stabilized knees. Clin Orthop Relat Res. 2011;469:160–167.
10. Engh CA Jr, Zimmerman RL, Hopper RH Jr, Engh GA. Can microcomputed tomography measure retrieved polyethylene wear? Comparing fixed-bearing and rotating-platform knees. Clin Orthop Relat Res. 2013;471:86–93.
11. Gallo J, Havranek V, Zapletalova J. Risk factors for accelerated polyethylene wear and osteolysis in ABG I total hip arthroplasty. Int Orthop. 2010;34:19–26.
12. Gascoyne TC, Teeter MG, Guenther LE, Burnell CD, Bohm ER, Naudie DR. In vivo wear performance of cobalt-chromium versus oxidized zirconium femoral total knee replacements. J Arthroplasty. 2016;31:137–141.
13. Gill H, Waite J, Short A, Kellett C, Price A, Murray D. In vivo measurement of volumetric wear of a total knee replacement. Knee. 2006;13:312–317.
14. Goyal P, Howard JL, Yuan X, Teeter MG, Lanting BA. Effect of acetabular position on polyethylene liner wear measured using simultaneous biplanar acquisition. J Arthroplasty. 2017;32:1670–1674.
15. Heyse TJ, Chen DX, Kelly N, Boettner F, Wright TM, Haas SB. Matched-pair total knee arthroplasty retrieval analysis: oxidized zirconium vs. CoCrMo. Knee. 2011;18:448–452.
16. Keeney JA, Eunice S, Pashos G, Wright RW, Clohisy JC. What is the evidence for total knee arthroplasty in young patients?: a systematic review of the literature. Clin Orthop Relat Res. 2011;469:574–583.
17. Kellett CF, Short A, Price A, Gill HS, Murray DW. In vivo measurement of total knee replacement wear. Knee. 2004;11:183–187.
18. Kurtz SM, Muratoglu OK, Evans M, Edidin AA. Advances in the processing, sterilization, and crosslinking of ultra-high molecular weight polyethylene for total joint arthroplasty. Biomaterials. 1999;20:1659–1688.
19. Lam-Tin-Cheung K, Yuan X, Nikolov HN, Lanting BA, Naudie DD, Teeter MG. Marker-based technique for visualizing radiolucent implant components in radiographic imaging. J Orthop Res. 2017;35:2017–2022.
20. Li Z, Esposito CI, Koch CN, Lee YY, Padgett DE, Wright TM. Polyethylene damage increases with varus implant alignment in posterior-stabilized and constrained condylar knee arthroplasty. Clin Orthop Relat Res. 2017;475:2981–2991.
21. McCalden RW, Hart GP, MacDonald SJ, Naudie DD, Howard JH, Bourne RB. Clinical results and survivorship of the GENESIS II total knee arthroplasty at a minimum of 15 years. J Arthroplasty. 2017;32:2161–2166.
22. McCalden RW, MacDonald SJ, Bourne RB, Marr JT. A randomized controlled trial comparing ‘high-flex' vs ‘standard' posterior cruciate substituting polyethylene tibial inserts in total knee arthroplasty. J Arthroplasty. 2009;24:33–38.
23. McCalden RW, MacDonald SJ, Charron KD, Bourne RB, Naudie DD. The role of polyethylene design on postoperative TKA flexion: an analysis of 1534 cases. Clin Orthop Relat Res. 2010;468:108–114.
24. Pang HN, Jamieson P, Teeter MG, McCalden RW, Naudie DD, MacDonald SJ. Retrieval analysis of posterior stabilized polyethylene tibial inserts and its clinical relevance. J Arthroplasty. 2014;29:365–368.
25. Paterson NR, Teeter MG, MacDonald SJ, McCalden RW, Naudie DD. The 2012 Mark Coventry award: a retrieval analysis of high flexion versus posterior-stabilized tibial inserts. Clin Orthop Relat Res. 2013;471:56–63.
26. Pitta M, Esposito CI, Li Z, Lee YY, Wright TM, Padgett DE. Failure after modern total knee arthroplasty: a prospective study of 18,065 knees. J Arthroplasty. 2018;33:407–414.
27. Pourzal R, Knowlton CB, Hall DJ, Laurent MP, Urban RM, Wimmer MA. How does wear rate compare in well-functioning total hip and knee replacements? A postmortem polyethylene liner study. Clin Orthop Relat Res. 2016;474:1867–1875.
28. Schmalzried TP, Shepherd EF, Dorey FJ, Jackson WO, dela Rosa M, Fa'vae F, McKellop HA, McClung CD, Martell J, Moreland JR, Amstutz HC. The John Charnley Award. Wear is a function of use, not time. Clin Orthop Relat Res. 2000;381:36–46.
29. Schroer WC, Berend KR, Lombardi AV, Barnes CL, Bolognesi MP, Berend ME, Ritter MA, Nunley RM. Why are total knees failing today? Etiology of total knee revision in 2010 and 2011. J Arthroplasty. 2013;28:116–119.
30. Sharkey PF, Lichstein PM, Shen C, Tokarski AT, Parvizi J. Why are total knee arthroplasties failing today–has anything changed after 10 years? J Arthroplasty. 2014;29:1774–1778.
31. Short A, Gill HS, Marks B, Waite JC, Kellett CF, Price AJ, O'Connor JJ, Murray DW. A novel method for in vivo knee prosthesis wear measurement. J Biomech. 2005;38:315–322.
32. Sisko ZW, Teeter MG, Lanting BA, Howard JL, McCalden RW, Naudie DD, MacDonald SJ, Vasarhelyi EM. Current total knee designs: does baseplate roughness or locking mechanism design affect polyethylene backside wear? Clin Orthop Relat Res. 2017;475:2970–2980.
33. Srivastava A, Lee GY, Steklov N, Colwell CW Jr, Ezzet KA, D'Lima DD. Effect of tibial component varus on wear in total knee arthroplasty. Knee. 2012;19:560–563.
34. Stambough JB, Clohisy JC, Barrack RL, Nunley RM, Keeney JA. Increased risk of failure following revision total knee replacement in patients aged 55 years and younger. Bone Joint J. 2014;96:1657–1662.
35. Teeter MG, Brandt JM, Naudie DD, Bohm ER, McCalden RW, Holdsworth DW. Measurements of surface and subsurface damage in retrieved polyethylene tibial inserts of a contemporary design. J Long Term Eff Med Implants. 2012;22:21–31.
36. Teeter MG, Milner JS, MacDonald SJ, Naudie DD. Manufacturing lot affects polyethylene tibial insert volume, thickness, and surface geometry. Proc Inst Mech Eng H. 2013;227:884–889.
37. Teeter MG, Naudie DD, Bourne RB, Holdsworth DW. How do CAD models compare with reverse engineered manufactured components for use in wear analysis? Clin Orthop Relat Res. 2012;470:1847–1854.
38. Teeter MG, Naudie DD, McErlain DD, Brandt JM, Yuan X, Macdonald SJ, Holdsworth DW. In vitro quantification of wear in tibial inserts using microcomputed tomography. Clin Orthop Relat Res. 2011;469:107–112.
39. Teeter MG, Parikh A, Taylor M, Sprague J, Naudie DD. Wear and creep behavior of total knee implants undergoing wear testing. J Arthroplasty. 2015;30:130–134.
40. Teeter MG, Perry KI, Yuan X, Howard JL, Lanting BA. Contact kinematic differences between gap balanced vs measured resection techniques for single radius posterior-stabilized total knee arthroplasty. J Arthroplasty. 2017;32:1834–1838.
41. Teeter MG, Seslija P, Milner JS, Nikolov HN, Yuan X, Naudie DD, Holdsworth DW. Quantification of in vivo implant wear in total knee replacement from dynamic single plane radiography. Phys Med Biol. 2013;58:2751–2767.
42. Teeter MG, Thoren J, Yuan X, McCalden RW, MacDonald SJ, Lanting BA, Naudie DD. Migration of a cemented fixed-bearing, polished titanium tibial baseplate (Genesis II) at ten years: a radiostereometric analysis. Bone Joint J. 2016;98:616–621.
43. Valstar ER, de Jong FW, Vrooman HA, Rozing PM, Reiber JH. Model-based roentgen stereophotogrammetry of orthopaedic implants. J Biomech. 2001;34:715–722.
44. van Ijsseldijk EA, Lebel B, Stoel BC, Valstar ER, Gouzy S, Vielpeau C, Kaptein BL. Validation of the in vivo volumetric wear measurement for total knee prostheses in model-based RSA. J Biomech. 2013;46:1387–1391.
45. van Ijsseldijk EA, Valstar ER, Stoel BC, de Ridder R, Nelissen RG, Kaptein BL. Measuring polyethylene wear in total knee arthroplasty by RSA: differences between weight-bearing and non-weight-bearing positioning. J Orthop Res. 2014;32:613–617.
46. van Ijsseldijk EA, Valstar ER, Stoel BC, Nelissen RG, Kaptein BL. A model-based approach to measure the minimum joint space width of total knee replacements in standard radiographs. J Biomech. 2012;45:2171–2175.
47. van Ijsseldijk EA, Valstar ER, Stoel BC, Nelissen RG, Reiber JH, Kaptein BL. The robustness and accuracy of in vivo linear wear measurements for knee prostheses based on model-based RSA. J Biomech. 2011;44:2724–2727.
48. Vandekerckhove PJ, Lanting B, Bellemans J, Victor J, MacDonald S. The current role of coronal plane alignment in total knee arthroplasty in a preoperative varus aligned population: an evidence based review. Acta Orthop Belg. 2016;82:129–142.
49. Vandekerckhove PJ, Teeter MG, Naudie DD, Howard JL, MacDonald SJ, Lanting BA. The impact of wear and lift-off on coronal plane alignment in TKA and implications to future constrained revision: a retrieval study. J Arthroplasty. 2015;30:2017–2020.
50. Vandekerckhove PTK, Teeter MG, Naudie DDR, Howard JL, MacDonald SJ, Lanting BA. The impact of coronal plane alignment on polyethylene wear and damage in total knee arthroplasty: a retrieval study. J Arthroplasty. 2017;32:2012–2016.
51. Young SW, Walker ML, Bayan A, Briant-Evans T, Pavlou P, Farrington B. The Chitranjan S. Ranawat Award: No difference in 2-year functional outcomes using kinematic versus mechanical alignment in TKA: a randomized controlled clinical trial. Clin Orthop Relat Res. 2017;475:9–20.
© 2018 by the Association of Bone and Joint Surgeons