Y-TZP Zirconia Run Against Highly Crosslinked UHMWPE Tibial Inserts: Knee Simulator Wear and Phase-Transformation Studies

Riichiro Tsukamoto,1 Paul A. Williams,1 Ian C. Clarke,1 Giuseppe Pezzotti,2 Hiromu Shoji,1 Masao Akagi,3 Kengo Yamamoto4
1

Department Joint Research Center, Peterson Tribology Laboratory, Loma Linda University and Medical Center, Loma Linda, California Ceramic Physics Laboratory, Kyoto Institute of Technology, Kyoto, Japan Department of Orthopaedic Surgery, Kinki University, Osaka, Japan Department of Orthopaedics, Tokyo Medical University, Tokyo, Japan

Received 16 April 2007; revised 4 August 2007; accepted 24 September 2007 Published online 27 December 2007 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.b.30999

Abstract: Background: Zirconia (ZrO2) ceramics combined with highly cross-linked polyethylene appears to be a promising approach to minimize wear in artificial knee joints. The wear performance of yttria-stabilized zirconia (YZr) femoral condyles on 7-Mrad tibial inserts was compared in a knee simulator to CoCr bearing on 3.5-Mrad inserts. Methods: The knee design was the Bi-SurfaceTM type with a 9-year clinical history in Japan (JMM, Japan). A displacement-controlled knee simulator was used with kinematics that included 208 flexion, 658 rotation, and 6 mm anterior/posterior translation. Lubricant was alpha-calf serum, test duration was 10 million cycles (10 Mc), and wear was measured by weight-loss techniques. The wear zones were studied by laser interferometry, scanning electron microscopy, and Raman microprobe spectroscopy. Results: At 10 Mc the wear rates of the CoCr controls averaged 4.5 mm3/Mc. This was within 7% of the prior estimate at 5-Mc duration and comparable to BiSurface wear data from another laboratory. The CoCr condyles increased in roughness (Ra) from <50 nm to average Ra 5 250 nm due to linear scratching. The ceramic condyles remained pristine throughout the wear study (Ra <7nm). With the YZr/7-Mrad combination, the weight change had a positive slope over at 10 Mc, which meant that the actual polyethylene wear was unmeasurable. Microscopic examinations at 10 Mc showed that the zirconia surfaces were intact and there was no detectable change from tetragonal to monoclinic phase. Interpretation: Our laboratory knee wear simulation appeared very supportive of the 9-year YZr/PE clinical results with Bi-Surface total knee replacements in Japan. ' 2007 Wiley Periodicals, Inc. J Biomed
Mater Res Part B: Appl Biomater 86B: 145153, 2008

Keywords:

zirconia; polyethylene; knee prosthesis; wear

INTRODUCTION Total knee replacements (TKR) have shown a 95% survival rate out to 15 years.1,2 However, contemporary high-risk cases include much younger and potentially more active patients with much increased life expectancy. Contemporary TKR enhancements include both highly cross-linked polyethylenes (HXPE) for tibial inserts and zirconia ceramic for femoral condyles37 (Table I).
Correspondence to: I. C. Clarke (e-mail: chalim@llu.edu) Contract grant sponsors: Peterson Family Foundation, Loma Linda University, Japan Medical Materials Corporation, the Western Center for Orthopaedics Research, the Tsukamoto Orthopedic Clinic, the Mine Dental Clinic, and the Department of Orthopaedics, Loma Linda University Medical Center ' 2007 Wiley Periodicals, Inc.

Zirconia was introduced as a high-strength and hightoughness alternative ceramic to alumina.8 Such zirconia designs could be manufactured with virtually identical dimensions to a CoCr condylar design. The two main zirconia ceramics considered for use in medical implants over the past 20 years have been the magnesia and yttria-stabilized types. Of these the yttria-stabilized tetragonal zirconia polycrystal (Y-TZP) type has seen the majority of development and use.912 Zirconia has a crystallographic metastability at body temperature, manifested by a reversible transformation between two of its main polymorphs, the tetragonal and the monoclinic phase (t?m transformation). Signicant strengthening can be induced via this t?m transformation. A volume increase of 34% is produced upon transformation that hinders the progression of
145

146
TABLE I. Summary of TKR Wear Simulation Studies

TSUKAMOTO ET AL.

Study Akagi et al.

7

Dosage (Mrad) Non 3.5 7 3 3 3 Non Non 3.5 3.5 7 7

Femoral Condyle CoCr CoCr CoCr CoCr Alumina Zr CoCr OxZr CoCr Zr CoCr Zr

Simulator SWM

Wear Rates (mm3/Mc) 11.80 6.40 2.90 3.44 0.67 0.71 21.30 12.40 4.86 1.79 1.59 Weight-gain

Ratio (%) ref 54 25 ref 19 21 ref 58 ref 37 39 N/A

Ueno et al.4

AMTI

Ezzet et al.5 Tsukamoto et al.3

AMTI SWM

Zr, Zirconia; OxZr, Oxidized zirconium; N/A, not available; AMTI, 6 station knee simulator (Boston, Watertown, MA); SWM, 6 station knee simulator (Shore-Western Manufacturing, Monrovia, CA).

any advancing fatigue crack.10 However, similar changes on the implants bearing surfaces may produce grain protrusion and pull-out with signicant impact on surface roughening.11,13 This has the potential to accelerate the wear of mating UHMWPE bearings. For total knee joint replacements (TKR), the yttria-stabilized zirconia (YZr) has been used clinically in since 1989 in Japan.14 The magnesia-stabilized zirconia (Mzr) has also been developed for femoral condyles.15 However, the only ceramic knee with FDA approval to market incorporates a thin zirconia surface overlying metallic femoral condyles16 manufactured from zirconium (OxiniumTM, Smith and Nephew, Memphis, TN). Laboratory studies consistently reported superior wear resistance for zirconia/UHMWPE combinations when compared with controls.1720 Such simulator studies did not comment on any surface roughening after 5 million cycles duration. Additional scanning electron microscopy, X-ray diffraction, and Raman spectroscopic analysis would have been necessary to determine if any phase transformation had occurred in vitro, as has been reported in a small number of zirconia balls retrieved ex vivo.13,2124 One simulator study25 revealed that ZrPE bearings ran at more than double the surface temperature of MPE bearings. Such temperatures likely resulted in considerably increased precipitation rate of serum proteins.26 Therefore, it would appear that the increased thermal insulating properties of the zirconia resulted in higher lubricant temperatures, greater serum degradation and such an artifact led to a reduction in Zr/PE wear.12 However by 1999, one unfortunate manufacturing change by a French vendor brought zirconia ball sales to a sudden halt. This change in sintering processes resulted in a suddenly large number of fractured ProzyrTM balls (St Gobain Desmarquest, France). This initiated a world-wide recall in 1999 that was concluded by 2001.11,2729 The performance of Zr/PE combinations has been studied in knee simulators (Table I). As in the hip joints, all TKR wear studies indicated superior performance for Zr/

PE when compared with controls. However, once again, none appeared to have studied the zirconia bearings by microscopic techniques on conclusion of wear studies. The authors previously reported on a TKR wear study3 of a zirconia knee design that has a 9-year clinical history in Japan.14 This compared wear with combinations of zirconia and CoCr femoral condyles matched with 3.5 and 7 Mrad irradiated tibial inserts. As in all prior Zr/PE wear studies, the zirconia combinations showed wear reductions when compared with controls. Thus up to this point, no laboratory study had looked for microscopic changes in the zirconia femoral heads or femoral condyles. We, therefore, elected to extend our prior knee wear study3 to 10 million cycles (Mc) duration for combinations Zr/7-Mrad and control CoCr/3.5-Mrad combinations to determine if we could (i) better quantify such wear performance and (ii) detect any tetragonal to monoclinic phase transformation in zirconia femoral condyles.

METHODS This study was a 5.510 Mc extension of our previous knee simulator report presented at 5.5 Mc duration3 using the Bi-SurfaceTM knee design (Japan Medical Materials JMM Inc, Osaka, Japan) (Figures 1 and 2). We studied the ZrO2 and CoCr femoral condyles using the posterior-cruciate sacricing conguration.3 The tibial inserts (n 5 6) were made from one lot of ram-extruded PE.3 Gammairradiation sterilization to 3.5-Mrad and 7-Mrad was performed under nitrogen atmosphere. Post-irradiated annealing was conducted at 1308C for 4 h. Shelf aging averaged 24 months from sterilization to initiation of the wear study. Additional tibial inserts were used as soak controls. Knee joints provide an additional level of complexity when compensating for uid absorption. Because of the anteroposterior tracking of the small femoral contact area along the tibial surface, there will be alternating effects of motion, loading, and some frictional heating that cannot be

Journal of Biomedical Materials Research Part B: Applied Biomaterials

Zr WEAR AND PHASE-TRANSFORMATION

147

Figure 1. Posterior view of Bi-SurfaceTM knee with zirconia ceramic femoral condyles showing unique ball-and-socket cam mechanism (Japan Medical Materials, Osaka, Japan). [Color gure can be viewed in the online issue, which is available at www.interscience. wiley.com.]

duplicated by simply loading the polyethylene insert in one xed location (as is done in hip modes). Thus in our study the tibial wear and soak-controls were stored unloaded under deionized water for 60 days prior to testing. During

the wear study the control-soak inserts continued to be unloaded under water and then cleaned and weighed synchronously with the wear inserts.3032 A 6-channel, displacement-controlled, knee simulator was used (Shore-Western Manufacturing, Monrovia, CA) to run the stance-phase kinematics of 208 exion-extension, 6 mm of anteriorposterior translation, and 658 of internal external rotation.7 The Paul load prole33 was used with 2.6 kN peak-load running at 1.8 Hz frequency. The lightly loaded swing phase (40% of gait cycle) was deleted from the kinematics prole. Our test protocols do not generally include the swing-phase of gait. Since the swing phase is very lightly loaded we have excluded it, so that we can better understand the interactions of knee kinetics and kinematics on wear processes during the heavily loaded stance phase.3,7,3436 Our standard wear method differs from that issued by standards organizations such as International Scientic Organization (ISO 14243; parts 1, 2).37,38 Our laboratory routinely omits the swing phase of the knee cycle to focus only on the wear created during the heavily loaded stance phase from extension to 208 exion.3,7,34,35 Like other contemporary knee wear studies, we always run a contemporary control of clinical relevance.4,39 Test durations were to 10 million cycles (10 Mc). The lubricant was 50% alpha calf serum (20 mg/mL protein) with addition of ethylenediamine tetra acetic acid.18 All tibial inserts were cleaned and weighed every 0.5-Mc interval up to 5 Mc duration, then changed at 1-Mc intervals from 5 to 10 Mc duration. Fresh serum was added at each start-up. Wear and uid-absorption trends were analyzed by linear regression techniques and corrections were made for uid-absorption. As an assessment of experimental variance in the sets of 3 implants, we computed total percentage scatter as the (max-avg)/avg and (min-avg)/avg. Phase transformation in zirconia surfaces can be studied by Raman microprobe spectroscopy equipped with a confocal microscope.40,41 Confocal Raman microprobe spectroscopy (irradiation with a blue monochromatic light with wavelength of 488 nm) characterization was carried out on the implants before and after simulation. Raman spectra

Figure 2. Bi-SurfaceTM range of motion and engagement of spherical cam beginning at 808 exion.

Journal of Biomedical Materials Research Part B: Applied Biomaterials

148

TSUKAMOTO ET AL.

morphs, according to the following equation43: Vm 0:5I180 I190 0:5I180 I190 2:2I150 1

Figure 3. (a) Excellent linear trending (weight-gain) evident for 3.5Mrad PE/CoCr combination (N 5 3) from 1 to 10 Mc duration, (b) Excellent linear trending (weight-gain) evident for 7-Mrad PE/zirconia combination (N 5 3) from 1 to 10 Mc duration.

were collected with a triple monochromator spectrometer with wavelength 488 nm (T-64000, ISA, Jovin-Ivon/Horiba Group, Tokyo, Japan) equipped with a charge-coupled detector (high-resolution CCD camera). A confocal conguration for the Raman probe was selected by placing a pinhole aperture with diameter 100 nm in the optical train of the spectrometer and using it to regulate the rejection of outof-focus light.42 All Raman measurements were made with placing the focal plane of the probe on the sample surface. According to this probe conguration, the probe depth below the sample surface was experimentally determined as 6.5 6 0.2 lm, according to defocus probe response function measurements.42 The volume fraction of monoclinic polymorph, Vm, in partially transformed zones can be quantitatively evaluated from the relative intensities of selected Raman bands belonging to the tetragonal (150 cm21 band) and to the monoclinic (180 and 190 cm21 bands) poly-

where I is the integral intensity of a Raman band whose maximum is located at the subscript wavenumber. In this study, each data point represented the average of 625 Raman measurements on the implant surface. The technique was judged accurate to about 0.2% by volume. The precision of the Raman measurement is dictated by the spectroscopic detection limit of the monoclinic doublet located at 180 and 190 cm21. The confocal Raman Spectrometer was used to provide precise mapping of the transformation process. The zirconia femoral condyles were entirely scanned with the confocal microprobe and the Raman spectra belonging to their surface recorded and analyzed.22,4446 The worn and non-worn surfaces of CoCr and zirconia femoral condyles were analyzed after 10 Mc duration by laser interferometry (NewView 600, Zygo, Sunnyvale, CA). The wear patterns were analyzed and compared with the examination by scanning electron microscopy (LV-SEM: Philips, XL30-FEG). Linear regression analysis was performed for each specimen using the SPSS statistical package (version 10: SPSS, Chicago, IL). One-way ANOVA and non-parametric (Kruskal-Wallis) tests were performed with subsequent testing for multiple comparisons (critical p-value 5 0.05). RESULTS The wear trends appeared stable from 1 to 10 Mc duration. The control implants (CoCr/3.5-Mrad) had excellent linear trending with respect to test duration (regression coefcient r [ 0.99), showed good control of experimental variance (\610% with weight-loss averaging 24.26 mg/Mc and corresponding wear rates 4.54 mm3/Mc [Figure 3(a), Table II]. The ZrO2/7-Mrad combination showed equally good linear trending (r [ 0.99) and also excellent control of experimental variance (\65%). However, the net-weight change remained positive throughout the study (average 6 0.54 mg/Mc). This meant there was no quantitative

TABLE II. Summary of the Experimental Results of Wear Testing

Femoral Condyle CoCr CoCr CoCr Zirconia Zirconia Zirconia

Dosage (Mrad) 3.5 3.5 3.5 7 7 7

Specimens KL271 KL272 KL273 KL280 KL281 KL282

r-Value 20.995 20.998 20.992 0.990 0.993 0.994

Weight Loss (mg/Mc) 24.75 23.92 24.11 0.52 0.53 0.57

Wear Rate (mm3/Mc) 25.06 24.18 24.38 NA

Avg Wear Rate (mm3/Mc) 24.54

SD 0.43

SD% 9.60%

NA

0.02

4.60%

6SD, standard deviation; NA, not available.

Journal of Biomedical Materials Research Part B: Applied Biomaterials

Zr WEAR AND PHASE-TRANSFORMATION

149

Figure 4. Scanning electron microscopy of the CoCr femoral condyle after 10 Mc; (a) 5 325, (b) 5 31000, and (c) 5 35000.

assessment for ZrO2/7-Mrad wear trends for 10 Mc duration [Figure 3(b), Table II]. Under SEM study the nonworn surfaces of CoCr condyles revealed faint polishing scratches with many carbide formations present. The worn areas of the CoCr femoral condyles examined at very low magnications (253), revealed numerous deep scratches orientated in the antero-posterior direction [Figure 4(a)]. These were up to 1 lm wide and hundreds of microns deep. In contrast, both worn and non-worn areas on the zirconia condyles appeared featureless up to a magnication of 10003 [Figure 5(a)]. The original polishing tracks rst became visible as very ne linear striations [Figure 5(a)]. At mid-level magnications, parallel arrays of surface

scratches were evident and orientated predominantly in the antero-posterior direction [Figure 4(b,c)]. At this level the zirconia surface appeared to have a dimpled appearance. At higher magnications [Figure 5(b)], the dimples corresponded to the diameter of the zirconia grains (0.3 lm). Such zirconia surfaces appeared intact with no evidence of individual grain erosion or larger voids. Laser interferometry comparisons of worn and non-worn CoCr surfaces revealed considerable increase in condylar roughness. Non-worn CoCr bearing surfaces ranged 3050 nm range whereas the worn surfaces ranged 110420 nm (Table III: Ra). This represented a 6-fold increase in CoCr condylar roughness over the 10-Mc duration of the simulator study. In

Figure 5. Scanning electron microscopy of the zirconia femoral condyle after 10 Mc; (a) 5 31000 and (b) 5 320,000.

Journal of Biomedical Materials Research Part B: Applied Biomaterials

150

TSUKAMOTO ET AL.

TABLE III. Surface Roughness of CoCr Femoral Condyle by Laser Interferometry

CoCr Roughness 1 2 3 4 5 Avg SD CV (%)

Ra(nm) Non-Wear 50.56 52.15 32.55 44.60 47.38 45.45 7.78 17.11

Rp (nm) Non-Wear 656.37 405.39 451.57 442.97 773.04 545.87 160.49 29.40

Rz (nm) Non-Wear 621.00 215.37 301.40 437.66 612.26 437.54 181.71 41.53

Ra (nm) Wear 392.20 143.53 174.37 423.96 113.60 249.33 146.51 58.76

Rp (nm) Wear 2645.96 1158.37 1068.55 2074.61 715.84 1532.67 799.35 52.15

Rz (nm) Wear 2457.60 1456.57 1709.78 2415.30 1081.25 1824.10 602.25 33.02

Ra, The average surface roughness; Rp, The highest peak value; Rz, Average of ten-peak height; Non-wear, non-wear zone roughness; Wear, wear zone roughness; Avg, average; SD, standard deviation; CV, coefcient variance.

contrast the zirconia ceramic surfaces averaged \7 nm (Table IV: Ra) for both worn and non-worn regions. The ne polishing scratches seen by SEM were so shallow that they did not register on the oblique laser plots of roughness (Figure 6). Raman spectroscopy in two main wear zones of the zirconia femoral condyles [Figure 7: zones (a) and (b)] conrmed that a tetragonal to monoclinic transformation was not present after 10-Mc duration of wear test. As can be seen from a typical spectrum collected in the main wear zone (Figure 8), only the tetragonal bands (letter t) could be observed. There was no trace of the monoclinic doublet that would have been located at bands 180 and 190 cm21 if monoclinic surface transformation had occurred.

DISCUSSION This appears to be the rst wear study of zirconia bearings run to 10-Mc duration in either hip or knee designs. With the CoCr/3.5-Mrad control set, the wear rate averaged 4.54 mm3/Mc over 10 Mc duration. In the prior study, up to 5 Mc duration, the comparable value was 4.86 mm3/Mc. This 6.6% difference represented a very consistent result for the CoCr/3.5-Mrad controls. The only other laboratory study reporting wear rates with CoCr/3.5-Mrad tibial inserts averaged 3.44 mm3/Mc over 5 Mc duration (Table II). This 32% difference in wear magnitude appeared a reasonable agreement, given knee simulator studies conducted inde-

pendently, under different test conditions but with same design of Bi-Surface total knee replacement (Table I). Unmeasurable wear in HXPE has been a consistent nding in prior work.47,48 The confounding factor in all such studies is the magnitude of the weight-gain due to uid absorption.47,49 Thus the detection of small amounts of polyethylene wear may be confounded by large uidabsorption effects. Thus was not a problem with the CoCr/ 3.5 Mrad control set, which lost weight at the rate of 4.26 mg/Mc (total variation \ 6 12%). In contrast, the YZr/7Mrad combination showed equally stable trending but the weight change was represented by 10.57 mg/Mc (total variation \ 6 6%. This represented a net gain in weight, that is, there was no detection of a net weight loss attributable to wear. The two previous studies of the Bi-Surface with CoCr and 7-Mrad inserts did produce measurable wear (Table I: 1.6 and 2.9 mm3/Mc). Even extending our Zr/7-Mrad knee study from 5 to 10 million cycles failed to identify such wear. Thus, compared with the CoCr data, the YZr/7Mrad combination appeared superior both at 5 and 10 Mc durations. Therefore, there continued to be a benecial effect with highly polyethylene crosslinked inserts combined with zirconia femoral condyles but it was not quantiable with current methods. This appears to be the rst examination of surface roughness evident in ceramic and CoCr femoral condyles following a simulator wear study. The difference in roughness values was dramatic. The zirconia roughness was

TABLE IV. Surface Roughness of Zirconia Femoral Condyle by Laser Interferometry

Zirconia Roughness 1 2 3 4 5 Avg Std CV (%)

Ra (nm) Non-Wear 6.16 6.52 6.15 6.01 6.35 6.24 0.20 3.18

Rp (nm) Non-Wear 33.42 39.46 32.62 35.59 37.96 35.81 2.91 8.13

Rz (nm) Non-Wear 65.09 65.10 57.73 61.61 68.45 63.60 4.07 6.41

Ra (nm) Wear 4.98 6.76 4.99 5.27 6.39 5.68 0.84 14.75

Rp (NM) Wear 24.88 30.35 31.90 25.02 45.71 31.57 8.50 26.93

Rz (nm) Wear 69.91 46.29 50.83 49.34 54.99 54.27 9.29 17.11

Ra, The average surface roughness; Rp, The highest peak value; Rz, Average of 10-peak height; Non-wear, non-wear zone roughness; Wear, wear zone roughness; Avg, average; SD, standard deviation; CV, coefcient variance.

Journal of Biomedical Materials Research Part B: Applied Biomaterials

Zr WEAR AND PHASE-TRANSFORMATION

151

Figure 8. Typical Raman spectrum showing only tetragonal phase after 10 Mc duration. Figure 6. Comparison of the non-worn and worn surfaces of CoCr and zirconia femoral condyles analyzed after 10 Mc by laser interferometry.

unchanged at 10 Mc duration (Ra \ 8 nm) whereas the CoCr wear zone had increased from generally an average Ra \ 50 nm to average Ra 5 250 nm (range 100400 nm) by end of study. This 5-fold increased roughness was a result of large, linear, and parallel scratches that developed in the antero-posterior sliding direction. What was interesting was that these scratched surfaces had no discernable effect on the overall wear rates of the 3.5-Mrad polyethylene inserts from 1 to 10 Mc duration. The regression coefcients averaged [0.99, which represented excellent linear trending for the control set. It was possible that these scratches developed early in the study but we had not quantied any early changes in roughness (from 0 to 1 Mc).

It is also not clear how condylar roughness effects would accelerate polyethylene wear in vivo.50 Very little has been reported for CoCr roughness effects in TKR designs, either in vitro or in vivo. What our study does show was that the CoCr condylar surfaces were susceptible to 3rd-body abrasion in a presumed pristine laboratory test. The most likely source for 3rd-body abrasive particles would be either the carbides released from the CoCr bearing surfaces as a result of the wear process or polymethylmethacrylate particles released by cement degradation under the femoral condyles. This appears to be the rst examination of Y-TZP zirconia bearings for phase transformation in a simulator wear model.12,23,44,45 At 10 million cycles duration, microscopic examination by a highly spatially-resolved, confocal Raman probe did not detect any transformation on the surfaces of the zirconia condyles. In addition the SEM study showed a pristine surface with rounded sub-micron zirconia grains visible. There was no surface dimpling or grain eruption that might have signied that the tetragonal to monoclinic phase transformation had initiated.12,51,52 The roughness proles were also reassuring in that the various surface parameters showed no change over 10-Mc duration. On the other hand, a prior review of the literature showed that no hip simulator study had observed such transformation effects in vitro and yet such changes clearly had occurred in the various retrieval studies of Y-TZP zirconia femoral heads.12,51 Therefore, it is not known how predictive our knee simulator data will be for in vivo performance of the zirconia femoral condyles. Our laboratory has not had the opportunity to analyze any retrieved zirconia knee joints. However, since such Y-TZP zirconia knee implants have been used clinically for about a decade14 there may also be an expectation that the transformation risks are reduced in TKR designs. The use of alumina-doped zirconia11 may lead to a superior stability as compared to the non-doped zirconia used previously.12 It is also possible that the hydrothermal wear effects in zirconia knee implants may be totally different from that present between the zirconia femoral heads and polyethylene acetabular liners.
The authors thank Donaldson Arthritis Research Foundation (Colton, CA) for the use of the laser interferometry equipment.

Figure 7. Raman spectroscopic sites (a, b) studied on zirconia femoral condyles. [Color gure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Journal of Biomedical Materials Research Part B: Applied Biomaterials

152

TSUKAMOTO ET AL.

REFERENCES
1. Ranawat CS, Flynn WF, Saddler S, Hansraj KK, Maynard MJ. Long-term results of the total condylar knee arthroplasty. Clin Orthop Rel Res 1993;286:94102. 2. Ritter MA, Berend ME, Meding JB, Keating EM, Faris PM, Crites BM. Long-term followup of anatomic graduated components posterior cruciate-retaining total knee replacement. Clin Orthop Rel Res 2001;388:5157. 3. Tsukamoto R, Chen S, Asano T, Ogino M, Shoji H, Nakamura T, Clarke IC; Improved wear performance with crosslinked and zirconia implants in knee simulation. Acta Orthop Scand 2006;77:505511. 4. Ueno M, Ikeuchi K, Nakamura T, Akagi M. Comparison of the wear properties of polyethylene plate in total knee prostheses (TKP) using different femoral component materials. Key Eng Mater 2003;15:801804. 5. Ezzet KA, Hermida JC, Colwell CW, DLima DD. Oxidized zirconium femoral components reduce polyethylene wear in a knee wear simulator. Clin Orthop Relat Res 2004;428:120 124. 6. White SE, Whiteside LA, McCarthy DS, Anthony M, Poggie RA. Simulated knee wear with cobalt chromium and oxidized zirconium knee femoral components. Clin Orthop Relat Res 1994;309:176184. 7. Akagi M, Asano T, Clarke IC, Niiyama N, Kyomoto M, Nakamura T, Hamanichi C. Wear and toughness if crosslinked polyethylene for total knee replacements: A study using a simulator and small-punch testing. J Orthop Res 2006;24:20212027. 8. Cales B. Zirconia as a sliding material. Clin Orthop Relat Res 2000;379:94112. 9. Drouin JM, Cales B, Chevalier J, Fantozzi G. Fatigue behavior of Zirconia hip joint heads: Experimental results and nite element analysis. J Biomed Mater Res 1997;34:149155. 10. Piconi C, Maccauro G. Review zirconia as a ceramic biomaterial. Biomaterials 1999;20:125. 11. Clarke IC, Green DD, Williams P, Pezzotti G, Sedal L, Donaldson T, Gustafson GA. Current status of zirconia used in total hip implants. J Bone Joint Surg (Am) 2003;85A:7384. 12. Clarke IC. Metastable nature of zirconia femoral heads from a 20-year perspective of clinical and simulator wear studies. Sem Artroplast 2006;17:165178. 13. Haraguchi K, Sugano N, Nishii T, Miki H, Oka K, Yoshikawa H. Phase transformation of a zirconia ceramic head after total hip arthroplasty. J Bone Joint Surg (Br) 2001;83B:9961000. 14. Akagi M, Nkamura T, Matsusue Y, Ueo T, Nishijyo K, Ohnishi E. The Bisurface total knee replacement: A unique design for exion. J Bone Joint Surg 2000;82A:16261632. 15. Roy ME, Whiteside LA, Katerberg BJ, Steiger JA. Phase transformation, roughness, and microhardness of articially aged yttria- and magnesia-stabilized zirconia femoral heads. J Biomed Mater Res A, 2007. Submitted for publication. 16. Laskin RS. An oxidized Zr ceramic surfaced femoral component for total knee arthroplasty. Clin Orthop Relat Res 2003;416:191196. 17. McKellop H, Lu B, Benya P. Friction, lubrication and wear of cobalt-chromium alumina and zirconia hip prostheses compared on a joint simulator. Paper presented at 38th Annual Meeting of the Orthopaedic Research Society, Washington, DC, 1992. P402. 18. Derbyshire B, Fisher J, Dowson D, Hardaker C, Brummitt K. Comparative study of the wear of PE with zirconia ceramic and stainless steel femoral heads in articial hip joint. Med Eng Phys 1994;16:229236. 19. Saikko V. Wear of polyethylene acetabular cups against zirconia femoral heads studied with a hip joint simulator. Wear 1994;176:207212.

20. Goldsmith AAJ, Dowson D. A multi-station hip joint simulator study of he performance of 22 mm diameter zirconiaultra-high molecular weight polyethylene total replacement hip joints. Proc Instn Mech Engr 1999;213:7790. 21. Hernigou P, Bahrami T. Zirconia and alumina ceramics in comparison with stainless-steel heads. J Bone Joint Surg (Br) 2003;85:504509. 22. Walter WL, Skyrme AD, Richard S, Chia M, Green DD, Walter WK, Zicat B. Polyethylene wear rates with zirconia and cobalt chrome heads, Paper Presented at 51st Annual Meeting of the Orthopaedic Research Society, San Francisco, CA, 2004. P1194. 23. Green DD, Pezzotti G, Sakakura S, Ries M, Clarke IC. Two and 10 years retrievals of zirconia femoral heads: XRD, SEM and Raman spectroscopy studies. Bioceramics: Materials and Applications IV. In: Sundar V, Rusin RP, Rutiser CA, Westerville, OH, The American Ceramic Society; 2003. pp 133 140. 24. Santos EM, Vohra S, Catledge A, McClenny MD, Lemons J, Moore KD. Examination of surface and material properties of explanted zirconia femoral heads. J Arthroplasty 2004;(19 Suppl 2):3034. 25. Lu Z, McKellop H. Frictional heating of bearing materials tested in a hip joint wear simulator. Proc Inst Mech Eng H 1997;211:101108. 26. Liao YS, Benya PD, McKellop HA. Effect of protein lubrication on the wear properties of materials for prosthetic joints. J Biomed Mater Res 1999;48:465473. 27. Prozyr St. Gobain. On February 4, 2002, summarized their report of 162 fractures of ProzyrTM zirconia femoral heads made in tunnel furnace from January 1998 to September 1999. (19981992). Available at www.prozyr.com. curite Sanitaire des 28. AFSSPS (2001): Agence Franc aise de Se (France) distributed Advisory and Warning Produits de Sante letter regarding their ongoing investigation into high fracturerates associated with ProzyrTM product made by St Gobain Advancees Desmarques (SGAD) using tunnel-furnace sintering process (July 2001). 29. Food and Drug Adminstration (FDA): Recall of zirconia ceramic femoral heads for hip implants. 2001. Available at http://www.fda.gov/cdrh/recalls/zirconiahip.html. 30. Clarke IC. Friction and wear studies: Validity of wear-screening protocols. Eng Med 1981;10:115122. 31. Clarke IC, McGuire PJ, Hull DB, Okuda R, Sarmiento A. Quantitative Laboratory Wear Studies Of Hip Prostheses. Paper Presented at 27th Annual Meeting of the Orthopaedic Research Society, Las Vegas, NV, 1981. P185. 32. Lu Z, McKellop H, Liao P, Benya P. Potential thermal artifacts in hip joint wear simulators. J Biomed Mater Res 1999;48:458464. 33. Paul JP. Forces transmitted by joints in the human body. Proc Instn Mech Engr 1966;181 (Part 37):815. 34. Kawanabe K, Clarke IC, Tamura J, Akagi M, Good VD, Williams PA, Yamamoto K. Effects of A-P translation and rotation on the wear of UHMWPE in a total knee joint simulator. J Biomed Mater Res 2001;54:400406. 35. Tamura J, Clarke IC, Kawanabe K, Akagi M, Good VD, Williams PA, Masaoka T, Schroeder D, Oonishi H. Micro-wear patterns on UHMWPE tibial inserts in total knee joint simulation. J Biomed Mater Res 2003;61:218225. 36. Manaka M, Clarke IC, Yamamoto K, Shishido T, Gustafson A, Imakiire A. Stripe wear rates in Alumina THRcomparison of microseparation simulator study with retrieved implants. J Biomed Mater Res B 2004;69:149157. 37. ISO 142431: 200203: Implants for surgeryWear of total knee joint prosthesesPart 1: Loading and displacement parameters for wear-testing machines with load control and corresponding environmental conditions for tests.

Journal of Biomedical Materials Research Part B: Applied Biomaterials

Zr WEAR AND PHASE-TRANSFORMATION

153

38. ISO 142432: 200010: Implants for surgeryWear of total knee joint prosthesesPart 2: Methods of measurement. 39. Muratoglu OK, Bragdon CR, Jasty M, OConnor D, Knoch V, Harris WH. Knee simulator testing of conventional and cross-linked polyethylene tibial inserts. J Arthroplasty 2004; 19:887897. 40. Pezzotti G, Porporati AA. Phase transformation analysis of phase-transformation and stress patterns in zirconia hip joints. J Biomed Opt 2004;9:372384. 41. Pezzotti G. Raman piezo-spectroscopic analysis of natural and synthetic biomaterials. Anal Bioanal Chem 2005;381:577590. 42. Wan K, Zhu W, Pezzotti G Determination of in-depth probe response function using spectral perturbation methods. J Appl Phys 98:113101113107. 43. Katagiri G, Ishida H, Ishitani A, Masaki T. Direct determination by Raman microprobe of the transformation zone size in Y2O3 containing tetragonal ZrO2 polycrystals. Advances in Ceramics, Vol. 24: Science and Technology of Zirconia III. Westerville, OH: The American Ceramic Society; 1988, pp 537544. 44. Clarke IC, Pezzotti G, Sakakura S, Sugano N. Phase transformation and residual stress in retrieved zirconia ball implant. Key Eng Mater 2003;240242:777780. 45. Pezzotti G, Clarke IC, Jobe C, Donaldson T, Yamamoto K, Tateiwa T. Kumakura T, Tsukamoto R, Ikeda J. Confocal Raman spectroscopic analysis of ceramic hip joints. Key Eng Mater 2006;309311:12111214.

46. Brown S, Clarke IC. Triggering the phase transformation of zirconia balls in vivo and in vitro. In: Proceedings of Trans Soc Biomat, Pittsburgh, PA, April 2629, 2006. p 103. 47. Bragdon CR, Jasty M, Muratoglu OK, Harris WH. Thirdbody wear testing of a highly cross-linked acetabular liner. J Arthroplasty 2005;20:379385. 48. Muratoglu OK, Bragdon CR, Daniel BS, Jasty M, Harris WH. A novel method of cross-linking ultra-high-molecular-weightpolyethylene to improve wear, reduce oxidation, and retain mechanical properties. J Arthroplasty 2001;16:149160. 49. Clarke IC, Starkebaum W, Hosseinian A, McGuire P, Okuda R, Salovey R, Young R. Fluid-sorption phenomena in sterilized polyethylene acetabular prostheses. Biomaterials 1985;6: 184188. 50. Wang A, Polineni VK, Essner A, Sun DC, Stark C, Dumbleton JH. Effect radiation dosage on the wear of stabilized UHMWPE evaluated by hip and knee joint simulators. Paper Presented at Annual Meeting Society for Biomaterials, New Orleans, LA, 1997. p 394. 51. Green DD, Williams PA, Donaldson TK, Clarke IC. BioloxForte vs Biolox-Delta under micro separation test mode in the USA. Paper Presented at 51st Annual Meeting of the Orthopaedic Research Society, San Francisco, CA, 2004. P0239. 52. Brown SS, Green DD, Pezzotti G, Donaldson T, Clarke IC. Possible triggers for phase transformation in zirconia hip balls. J Biomed Mater Res B, 2007; DOI:10.1002/jbmb. 30964.

Journal of Biomedical Materials Research Part B: Applied Biomaterials